# A grammar of Pichi

Kofi Yakpo

Studies in Diversity Linguistics 23

#### Studies in Diversity Linguistics

#### Editor: Martin Haspelmath

#### In this series:


# A grammar of Pichi

Kofi Yakpo

Yakpo, Kofi. 2019. *A grammar of Pichi* (Studies in Diversity Linguistics 23). Berlin: Language Science Press.

This title can be downloaded at: http://langsci-press.org/catalog/book/85 © 2019, Kofi Yakpo Published under the Creative Commons Attribution 4.0 Licence (CC BY 4.0): http://creativecommons.org/licenses/by/4.0/ Indexed in EBSCO ISBN: 978-3-96110-133-7 (Digital) 978-3-96110-134-4 (Hardcover) 978-3-96110-135-1 (Softcover) ISSN: 2363-5568 DOI:10.5281/zenodo.2546450 Source code available from www.github.com/langsci/85 Collaborative reading: paperhive.org/documents/remote?type=langsci&id=85

Cover and concept of design: Ulrike Harbort

Typesetting: Redmer Kronemeijer, Kofi Yakpo, Sebastian Nordhoff, Hing-Yuet Fung Proofreading: Amir Ghorbanpour, Bev Erasmus, Christian Döhler, Eitan Grossman, Felix Anker, Felix Kopecky, Hing-Yuet Fung, Ivica Jeđud, Jeffrey Pheiff, Jeroen van de Weijer, Ludger Paschen, Mykel Brinkerhoff, Sebastian Nordhoff, Vadim Kimmelman, Valentin Vydrin, Yvonne Treis Fonts: Linux Libertine, Libertinus Math, Arimo, DejaVu Sans Mono Typesetting software: XƎLATEX

Language Science Press Unter den Linden 6 10099 Berlin, Germany langsci-press.org

Storage and cataloguing done by FU Berlin

To Yèni and Fodé








## **Acknowledgments**

I wish to express my gratitude to the people in Equatorial Guinea whose support is the foundation of this work. I am deeply indebted to Françoise Tatchouop for her hospitality and generosity during my stays in Equatorial Guinea. I am most grateful to my language teachers, the late Natalia "Abuela" Toichoa Borico, María Rosa "Mami Rose" Buesule Toichoa, Rudolfo "Djunais" Beaka Chale, Ildefonso "Boye" Ntutumu, Sonia "Lage" Belobe Toichoa, Sandra Eyang Ncoha Belobe, Miguel Ángel Ñat Buesule, Fermín Beaka Chale, Juan Antonio Tonca Toichoa, Agustín Gaspar Nguema Eñeso, Rosalía Ndjoku, Maura Toichoa Lopete, María Fernanda Aboki Sami, Hilda Méndez Sami, Eduardo Mejía, Ursus Megua Kofi, Miguel Fernández Ndonjo, as well as María, Beatriz, Lindo, and Charlie of Ela Nguema. I also wish to thank Trinidad Morgades Besari, the late Samuel Ebuka, the late Edward Jones, and Ruth Jones for sharing their wealth of knowledge on the history and the peoples of Bioko with me.

A special note of thanks goes to Danae Perez for copy editing and the correction of Spanish examples, as well as to Hing-Yuet Fung for a final round of editing and problem fixing in LaTeX. I am grateful to Sebastian Nordhoff and Redmer Kronemeijer for assistance with the conversion to LaTeX and typesetting, and to the members of the Language Science Press community for proofreading. I also wish to thank three anonymous reviewers for their detailed comments and suggestions.

I am obliged to Pieter Muysken, whose support secured funding from the Centre for Language Studies at Radboud University Nijmegen for field research between 2003 and 2007. I equally wish to acknowledge the support of grant no. 27613316 of the Research Grants Council of the Government of Hong Kong, which provided me with funding for complementary research in 2016.

> Kofi Yakpo Hong Kong, December 2018

## **Symbols and abbreviations**




### **1.1 The language and its speakers**

Pichi is an Afro-Caribbean English-lexifier Creole language (Ethnologue code "fpe") spoken on the island of Bioko, Equatorial Guinea (cf. Map 1). Pichi is the most widely spoken language of the country's capital Malabo next to Spanish, and it serves as a primary language to a large proportion of the capital's inhabitants. Pichi is also used as a primary language in a number of villages and towns along the Coast of Bioko, amongst them Sampaca, Fiston, Basupú, Barrio Las Palmas, and Luba (Morgades Besari, p.c.), and it is spoken as a lingua franca throughout Bioko (cf. Map 2 below). The language is also used by a sizeable community of people originating from Bioko in Bata, the largest town on the continental part of the country. In the literature, Pichi is known under the names "Fernando Po Creole English" (Simons & Fennig 2017), "Fernando Po Krio" (Berry 1970), "Fernandino Creole English" (Holm 1988), "Pidgin (English)" (Morgades Besari, p.c.) "Broken English" (Zarco 1938), and "Pichinglis" (Lipski 1992). While older speakers sometimes refer to the language as "Krio" or "Pidgin", most present-day speakers refer to it as "Pichinglis", "Pichin" with a nasalised final vowel, or "Pichi" *tout court*.

Pichi descends from 19th century Krio, which first arrived in Bioko, the former Fernando Po, with African settlers from Freetown, Sierra Leone, in 1827 (Fyfe 1962: 165). Krio, in turn, emerged as the principal language of the urban population of Freetown, Sierra Leone, from the late 18th century onwards (Huber 1999). Modern Krio and Pichi are therefore both descendants of Early Krio. Linguistic and historical evidence suggests that the diffusion of Krio along the west coast of Africa in the 19th century also contributed significantly to the formation of Nigerian Pidgin, Cameroon Pidgin, and Ghanaian Pidgin English (Huber 1999).

No linguistic census data exist in Equatorial Guinea, but probably up to 70 per cent of the population of Bioko island, hence well above 100,000 speakers, regularly use Pichi at various levels of nativisation and in various multilingual and multilectal constellations in and outside their homes (Yakpo 2013: 194). Next to Pichi, at least fourteen languages are spoken by the peoples of Equatorial Guinea (Hammarström et al. 2017). Fang has the largest number of speakers, but its use is largely limited to the continental part of the country (also referred to as "Río Muni"). Bube is probably the second most widely spoken African language of the country, but its use is, in turn, limited to Bioko. There is an established pattern of language shift to Pichi and Spanish in Malabo and other larger agglomerations of Bioko, and there are indications that Bube is under increasing pressure from these two languages. Equatorial Guinea also harbours the Portugueselexifier creole Fa d'Ambô, spoken by the people of the island of Annobón (cf. Map 1). Fa

d'Ambô shares historical and linguistic ties with the other Portuguese-lexifier creoles of the Gulf of Guinea, namely Lungwa Santome and Lunga Ngola (Angolar) in São Tomé, and Lung'Ie in Príncipe (Post 2013).

Mutual intelligibility between Pichi, Krio, Cameroon Pidgin, Nigerian Pidgin, and Ghanaian Pidgin English is relatively high. However, an impediment to fluid communication between speakers of Pichi and its African sister languages is the divergent path of development of Pichi since 1857. In that year, Spain began to actively enforce colonial rule in Equatorial Guinea. From then onwards, Pichi was cut off from the direct influence of English. Pichi has therefore escaped the phonological, grammatical, and lexical convergence with English that has been documented for English-lexifier creoles spoken alongside English (see e.g. Sala & Ngefac 2006 for Cameroon Pidgin). At the same time, Pichi has been in intense contact with Spanish for over a century and has undergone substantial lexical and some structural influence from the colonial language of Equatorial Guinea .

Equatorial Guinea has three *de jure* official languages, namely Spanish, French, and Portuguese. From the primary to the tertiary levels, instruction is given alone in Spanish, which is therefore the only *de facto* official language of the country. There is no legally or politically defined role for education in African languages (Yakpo 2011; 2016). However, the national education bill currently in vigour (Ley Núm. 5/2.007 2007) offers the optional use of indigenous languages in education (Olo Fernandes 2012). The socio-linguistic status of Pichi is particularly unfavourable among the natively spoken languages of Equatorial Guinea. During colonial rule, Pichi was considered an impoverished, debased form of English by Spanish colonial administrators and missionaries (see Zarco 1938: 5–7 for a pungent exposition of this view). Pichi, like the other creole languages of the Atlantic Basin, still has to struggle with this difficult legacy. In spite of its great importance as a community language and as a national and regional lingua franca, Pichi enjoys no official recognition nor support, is conspicuously absent from public discourse and the official media, and until today, has no place in the educational policy of Equatorial Guinea (Yakpo 2016).

The lingering colonialist perspective on Pichi and its sister languages in West Africa and across the Atlantic stands in stark contrast to the fact that these languages epitomise the achievements of African and African-descended peoples who, in resisting and adapting to the ignominious system of European slavery and colonialism, carved out in Africa and the Americas one of the largest, and today most vibrant cultural and linguistic zones of the world.

### **1.2 Contact with Spanish**

Spanish has left a deep imprint on the lexicon and grammar of Pichi. Codemixing is an integral part of the linguistic system of Pichi (Yakpo 2009b, Yakpo 2018). The pervasive influence of Spanish on Pichi is for one part the consequence of language policy. Since colonial rule and the independence of Equatorial Guinea in 1968, Spanish has remained the sole medium of instruction at all levels of the educational system (Lipski 1992: 35–

Map 1 Continental and insular Equatorial Guinea (in bold)

Map 2 Towns with Pichi-speaking communities in Bioko (in bold)

36). There is a widespread competence in different registers of Spanish by Pichi speakers in Malabo and Equatorial Guinea as a whole (Lipski 1985; Castillo-Rodríguez & Morgenthaler García 2016). In Malabo, the acquisition of Spanish begins in early childhood, even for many working-class Equatoguineans with little or no school education.

Another factor favouring codemixing is the positive attitude towards multilingualism in a highly polyglot society, against the background of a tenacious vitality of Pichi as a symbol of social identity. Presumably, Pichi-Spanish codemixing has for a long time served as a badge of identity for the population of Bioko in the course of a long history of immigration by speakers of other varieties of West African English-lexicon Creoles. Today, the language also plays an important role for the self-identification of those who grew up on the island in the face of an accelerated pace of internal migration by Equatoguineans from the mainland. *Bɔ́n na yá, gró na yá* 'born here, grown up here' is the mark which distinguishes Pichi-speaking islanders, irrespective of their ethnic background, from the late arrivals of mainland origin who speak little or no Pichi.

Equally, the burgeoning oil economy of Equatorial Guinea has led to increased urbanisation, extending multi-ethnic social networks and the spread of Pichi as a native language. In such a socio-economic environment and amidst a high general competence in the official language Spanish, codemixing between Pichi and Spanish, rather than being exceptional, is consciously and confidently articulated in daily life (cf. chapter 11 for a detailed description of codemixing). Pichi is also in contact with other African languages spoken in the region, amongst them Fang and Bube, as well as Nigerian and Cameroonian Pidgin (Yakpo 2013 discusses influences on Pichi from these languages).

### **1.3 Variation**

The variation recorded in Pichi appears to be determined by a mixture of the factors age, language background, and social class. Phonological variation is particularly conspicuous. Some of the variation in Pichi may be captured by an albeit oversimplified division of speakers into two groups. Group 1 principally consists of the Fernandinos, the old commercial and social elite of Bioko (Lynn 1984) that inhabits the historical centre of Malabo and has used Pichi as a home language since the 19th century. Group 1 also comprises people of diverse ethno-linguistic backgrounds who grew up in Malabo in the ambit of Fernandino culture. The lexicon, grammar, and phonology of Group 1 reflects an earlier chronolect of Pichi, which is also closer to (early) Krio.

Group 2 is larger and culturally more diverse by incorporating "nuevos criollos" (Morgades Besari, p.c.) who have been accultured more recently into the Pichi-speaking urban culture of Malabo. It encompasses a large number of speakers with a Bube cultural background who have shifted to Pichi as a primary language (Bolekia Boleká 2007), and it includes large numbers of speakers with varying degrees of nativisation. Group 1 is shrinking at the expense of Group 2 through rapid urbanisation, immigration, and language shift. The terms "Mesopidgin" and "Acropidgin" employed by Morgades Besari (2011) capture some of the socio-linguistic differences between Group 1 and Group 2. The distinction between Group 1 and 2 is also reflected in apparent-time differences,

#### 1.3 Variation

where older speakers (principally those who came of age in the colonial era and the first decade of independence) tend to use the Group 1 lect, and the young majority population of Malabo and Bioko tends to use the Group 2 lect.

In this work, I privilege the description of the language of Group 2 in the wish to represent how Pichi is spoken by the young and multi-ethnic majority in the homes and streets of Malabo today. I nevertheless account for variation by employing alternate forms where they exist (e.g. *nɔ́bà~nɛ́a* 'neg.prf', *tínap~tánap* 'stand (up)'), and some of them may reflect differences between Groups 1 and 2. In the following, I present a few generalisations of the variation present in my corpus.

For Group 2 speakers, there is no phonemic contrast between the alveolar fricative [s] and the postalveolar fricative [ʃ] (1–2), and this is systematically applied to all words where Group 1 speakers use [ʃ] (3). Group 2 speakers also insert a palatal glide [j] between [s] and a following mid vowel where Group 1 uses [ʃ] alone (4–5):


Group 2 speakers tend to neutralise the phonemic distinction between close-mid and open-mid vowels (6–7):


Group 2 speakers also tend to nasalise [i]-final words with an H.L tonal configuration (8) and to prenasalise [j]-initial words as in (9). This may lead to the formation of homophones like (10) and (11) for Group 2 speakers:


There is also some variation in the use and acceptance of certain grammatical structures. For example, Group 2 speakers seem to prefer the negative perfect marker *nɛ́a* over *nɔ́ba*. Equally, a serial verb construction (SVC) featuring the verb *sté* 'be long time' is not

readily accepted as grammatical by many Group 1 speakers (cf. §11.2.5) and may therefore be a more recent development. Conversely, other types of SVCs are more common with Group 1 than with Group 2. Amongst them are SVCs involving the verb *ték* 'take' (cf. §11.2.3) and motion-direction SVCs involving the verbs *gó* 'go' and *kán* 'come' (cf. §11.2.1). *Ték*-serialisation is very common in modern Krio and all other African English-lexifier creoles. Group 2 speakers instead tend to employ a combination of a verb and a prepositional phrase in these contexts. A final area characterised by variation is the extent of Pichi-Spanish language contact. For example, the names of weekdays and numerals are almost exclusively expressed in Spanish by Group 2 speakers. Group 1 speakers have access to both English- and Spanish-derived lexicon. They may employ *lunes* 'Monday' in a codemixed sentence, but are equally likely to use *mɔ́nde* 'Monday'. Further, Englishderived numbers above five are rarely used by Group 2 speakers (cf. §13.3.1). In contrast, Group 1 speakers master a wider range of the Pichi numeral system. However, even with this group, Pichi numbers above ten are seldom heard.

### **1.4 Affiliation**

Pichi belongs to the grouping of languages referred to in the literature by various appelations, among them "English-based Afro-American" (Alleyne 1980), "Atlantic Anglophone Creoles" (Hancock 1986; 1987) "Atlantic English-based Creoles" (e.g. Muysken & Smith 1990), "Atlantic English Creoles" (e.g. Baker 1999). In this work and others, I employ the term "Afro-Caribbean English-lexifier Creoles" (abbreviated AECs) (Faraclas 2004) as a label that includes information about the speaker population ("Afro-", i.e. people of African ancestry) and the two world regions where the languages are mainly spoken ("Afro-Caribbean", i.e. Africa and the Caribbean). The use of "lexifier" underscores the dynamic character of the English input to the lexicon, which varies in size and nature between the different languages.

All Afro-Caribbean English-lexifier Creoles are transmitted and learned in various ways within the family and serve as means of communication and identification to linguistic communities. I therefore dispense with the term "pidgin" with its socio-structural connations and use "creole" alone. When referring to the linguistic grouping, "Creole" is written with an initial capital letter. The generic term is written "creole" in lower case.

With well over 100 million speakers, the Afro-Caribbean English-lexifier Creoles and Pidgin-Creoles (henceforth AECs) spoken in Africa and the Americas together constitute one of the largest lectal continua of the Western hemisphere in speaker numbers and geographical extent (Yakpo 2016: 22–23). Besides Pichi, the African sub-grouping of the AECs contains Krio (Sierra Leone), Aku (Gambia), Ghanaian Pidgin English, Nigerian Pidgin, and Cameroonian Pidgin (Huber & Görlach 1996; Huber 1999; Baker & Huber 2001). There are also historical connections and cross-influences with varieties of Liberian English (Singler 1997). Even if many details are still unclear, the evidence that there is a degree of common ancestry between the African and Caribbean AECs is compelling (e.g. Hancock 1986; 1987; Smith 1987; 2015). There are also indications of a historical relation of the AECs with African American English(es) (Dillard 1973; Rickford 1999; Winford 2017).

#### 1.4 Affiliation

Within the African AECs, Pichi is most directly related to the Krio language of Sierra Leone. A comparison of the two languages yields systematic lexical and structural correspondences. But it also reveals some differences. To begin with, both languages share a large percentage of non-basic vocabulary, as shown in (12a), with the same tonal configurations. However, the Yoruba (b), Mende (c), and Temne (d) component of the Pichi lexicon appears to be much smaller than that of Krio and is limited to a few words in the corpus (data from Fyle & Jones 1980):


The two languages also share a number of lexical items common to numerous African and American English-lexicon Creoles. These were first compiled by (Smith 1987; 2001b; 2015) and termed "Ingredient X, Y, and Z". In (13), I list all the relevant words contained in the Pichi corpus. They comprise "Ingredient X" words of African origin (a), "Ingredient Y" words of Portuguese origin (b), "Ingredient Z" words of English origin (c), as well as a few function words of diverse origin (d):


Some of the differences in vocabulary between the two languages owe to the same phonological characteristics that differentiate the members of Group 1 (Pichi) and Group 2 (Krio) in the preceding section. Hence, most speakers of Pichi make no phonemic distinction between alveolar and postalveolar fricatives (14a); the phonemic distinction between close-mid and open-mid vowels is neutralised by most speakers (b).

In addition, the distinction between velar and labial nasal consonants tends to collapse in word-final position (c); phonological processes create preferred CV sequences (d), voiced obstruents are normally devoiced in word-final position (e), while other words have different coda consonants (f). In general terms, present-day Pichi as spoken by the majority of its speakers exhibits a tendency towards the reduction of phonemic contrasts when compared to Krio.


Other differences in vocabulary, phonology, and grammar stem from the divergent sociopolitical development that Equatorial Guinea and Sierra Leone have gone through in the last hundred years. In Sierra Leone, British colonisation and the retention of political, economic, and linguistic ties with Britain after independence have reinforced the relationship between Krio and English. In Equatorial Guinea, the direct link with English

1.4 Affiliation

was severed in 1858 when Spanish assumed the role of the dominant language. Equally, the influence of Krio on Pichi had petered out by the first decades of the 20th century as Spanish colonialism gradually put a stranglehold on relations between Fernando Po and Sierra Leone.

The role of the respective superstrates English (for Krio) and Spanish (for Pichi) can be read from the impact of these two languages on institutional and administrative terminology (15a), the numeral system above ten (b), and other lexical items (c). The use of a larger number of English-derived lexical items in Krio corresponds with a stronger presence of Spanish-derived lexicon in Pichi:


There is a high degree of correspondence between the forms of Pichi and Krio function words and the categories they express. For example, the forms and functions of the TMA markers in (16) are largely coterminous:


However, the distribution of the markers in (16) is not always identical in the two languages. For example, the Krio data reveals more combinatorial possibilities of the habitual marker *kin* 'hab' with other TMA markers (cf. Smith 2001a), while the Pichi imperfective marker *de* 'ipfv' seems to have a broader range of functions than the Krio cognate form. Moreover, Krio has at least two auxiliary constructions which are not attested in my data. The verb *blánt* is only employed as a lexical verb with the meaning 'reside' in Pichi. In Krio, the element *blant* is a preverbal TMA element that expresses habitual aspect. Further evidence for grammaticalisation is that *blant* is L-toned in this function. Consider the following example (Krio sentences are marked Krio):

(17) Krio *Olú* name *blant* hab *gó* go *London* place *fɔ* prep *Krísmɛs.* Christmas 'Olu always goes to London for Christmas.' (Yillah & Corcoran 2007: 181)

Further, Krio employs the locative-existential copula *dé* 'be.loc' together with the preposition *pan* 'on' in an, albeit lectally restricted, auxiliary construction to express progressive aspect (18). The construction is rejected by Pichi speakers (19):

(18) Krio *Olú* name *dé* be.loc *pan* on *kám.* come 'Olu is coming (right now).' (Yillah & Corcoran 2007: 179)

(19) \* *A* 1sg.sbj *dé* be.loc *pan* on *chɔ́p.* eat Intended: 'I'm eating.' [ye07je 025]

Conversely, there is no data to suggest the existence in Krio of the Pichi egressive aspect construction involving the auxiliary verb *kɔmɔ́t* 'go/come out' (20) or, obviously, the continuative aspect construction featuring the Spanish-derived verb *sigue* 'continue' (21). Equally, an adverbial SVC involving the V1 *sté* 'stay, be a long time' appears to be unique to Pichi (22):


The literature on Krio also indicates a wider range and a more pervasive use of SVCs than attested for Pichi. For instance, Krio has a resultative SVC featuring dynamic verbs in the V2 position (23) and a give-type SVC in order to mark a recpient or beneficiary (24). Both types of construction are not attested in Pichi:

(23) Krio

*Di* def *húman* woman *kúk* cook *rɛ́s* rice *sɛ́l.* sell 'The woman cooked rice and sold it.' (Finney 2004: 72) 1.5 Previous research on Pichi

(24) Krio *I* 3sg.sbj *báy* buy *klós* clothing *gí* give *in* 3sg.poss *pikín.* child 'He bought some clothes for his child.' (Finney 2004: 72)

In contrast, resultative state of affairs similar to (23) above may only feature stative property items as secondary verbs. Such constructions in Pichi are best seen to involve secondary predication (25):

(25) *Dɛn* 3pl *dɔ́n* pfv *bíl* build *di* def *hós* road *strɔ́n.* be.strong 'The house is solidly built.' [ra07ve 069]

At the same time, Pichi only employs a less integrated type of multiverb construction, namely clause chaining, in order to express a sentence like (24), involving a dynamic V2. Note that unlike the Krio sentences above, the Pichi example in (26) exhibits resumptive subject marking, i.e. the subject is repeated with the second verb in the series:

(26) *Yu* 2sg *ték* take *di* def *mɔní* money *yu* 2sg *gí* give *mí.* 1sg.indp

'You took the money (and) gave it to me.' [ro05de 033]

Numerous questions, however, remain open with regard to the extent of differences between the two languages. A considerable obstacle to comparative research is the lack of fresh data on Krio since the 1980s.

### **1.5 Previous research on Pichi**

Yakpo 2009a (in English) and 2010 (in Spanish) are the first in-depth descriptions of the phonology and grammar of Pichi. Zarco (1938) is a language guide with a vocabulary list and a short grammar section. Trinidad Morgades Besari, former Vice-Chancellor of the National University of Equatorial Guinea and a well-known philologist of the country, has written about the use of Spanish and Pichi in Equatorial Guinea (Morgades Besari 2005; 2011). Morgades Besari's unpublished work encompasses wordlists, a collection of stories and proverbs and proposals for an orthography of Pichi (see Yakpo 2011 for a discussion of the orthography). CEIBA Ediciones (Barcelona) has published a series of works dealing with the precolonial and colonial history and the political economy of Fernando Po, as well as the pivotal role of the Fernandinos in the making of present-day Bioko (e.g. Martín del Molino 1993; Cantús 2006).

### **1.6 Standardisation and orthography**

No commonly accepted standard orthography is in use for Pichi. The transcription used in this work follows the Krio orthography employed in the seminal Krio-English Dictionary (Fyle & Jones 1980) and subsequent revisions (Coomber 1992), which, in turn, relies

on the IPA-based Africa Alphabet (International African Institute 1930) and the African Reference Alphabet (UNESCO 1981). In the Krio/Pichi orthography, the grapheme <ɛ> renders the open-mid front vowel [ɛ], and <ɔ> renders the open-mid back vowel [ɔ]. Other vowel graphemes approximately also represent the corresponding IPA sounds. Pichi consonant phonemes and graphemes are presented in Table 2.1. In codemixed sentences, Spanish material is rendered using the standard Spanish orthography.

Tone is marked on all Pichi words throughout this book. H-toned syllables bear an acute accent, e.g. *wét* [wét] 'wait', and L-toned syllables remain unmarked, e.g. *wet* [wèt] 'with'. Tonal notation applies to the morpheme (i.e. the root), not the phonological word. In multimorphemic words, each morpheme therefore receives separate tone marks, e.g. *ús=tɛ́n* {*ús* 'q', *tɛ́n* 'time'} 'when', *fáyn-wán* {*fáyn* 'nice', *-wán* 'adv'} 'nicely'. Acute accents over Spanish words are orthographic, and hence not tone marks.

### **1.7 Methods and data**

This grammatical description of Pichi is based on the analysis of a corpus of 46,060 words of dialogues, narratives, procedural texts, and elicitations. The data was collected during three stays of four weeks each in Malabo between 2003 and 2007 as part of the research for my PhD thesis (Yakpo 2009a). Recordings were conducted in the quarters of Ela Nguema, Nyumbili, and the historical centre of Malabo. Recordings were done with a digital mini disc recorder and transcribed and analysed using the SIL Toolbox 1.5 programme. The analysis of tone was done from connected speech and words spoken in isolation using the Praat 5.0 software (Boersma & Weenink 2008). Much of my approach is guided by linguistic typology and the descriptive apparatus developed in research on African languages. I try to describe as much variation as feasible. I largely avoid comparative or etymological observations with respect to English and African languages and try to look at Pichi "from the inside". This grammar has also been published in Spanish (Yakpo 2010) in an abridged version for use in Equatorial Guinea by researchers and university students, teachers, and educationists.

In Equatorial Guinea, I worked with altogether seventeen language consultants. All speakers have been using Pichi continuously since childhood onwards. Without exception, they are inhabitants of Malabo since birth or infancy. Most of them use Pichi more often than any other language, and most speakers view Pichi as the language they know best. Additionally, all speakers also know at least one of the following other languages in varying degrees of proficiency: Fang, Bube, Fa d'Ambô, Kombe, Lungwa Santome, Nigerian Pidgin, Twi, Spanish, French, English, and German. There is a bias in the data towards speakers with a Bube ethno-linguistic background, reflective of the circumstance that the majority of people who use Pichi as their primary language are from a Bube background. The numerical dominance by these "nuevos criollos" over the "old" Creole community of Fernandino descent (Morgades Besari, p.c.) represents a significant shift in the social dynamics of the language which is reflected in my choice of speakers.

A few words are in order on aspects of my linguistic background and communicative approach during the research leading to this book. During my first stay in Malabo, I

#### 1.7 Methods and data

used Ghanaian Pidgin English and Spanish as working languages. During subsequent visits, when I felt confident enough to use Pichi without impeding fluid communication, I conducted my research exclusively in Pichi. My acquisition of Pichi and integration into social networks in Malabo was greatly facilitated by fluency in Ghanaian Pidgin English, competence in, and exposure to other Afro-Caribbean English-lexifier Creoles and West African languages, and a cultural and communicative *savoir faire* acquired during a childhood spent in Ghana. Fluency in French and Portuguese were also important resources in navigating the plurilingual landscape of Malabo and Bioko at various junctures during my research.

Table 1.1 lists relevant information on language consultants. Speakers are sorted alphabetically along the "code" column. The symbol "N.N." in the last row of the "speaker" column stands for incidental data collected from strangers in the streets, markets, and other public places in Malabo. Not included in the list is my own speaker code (ko). My participation in recorded conversations was kept to a minimum, but due to the nature of the method, it was more extensive during elicitations. Utterances of mine are, however, nowhere included in the analyses and interpretation of data. The symbols for gender are (F)emale and (M)ale. Age is provided in brackets of 10+, 20+, 30+, etc. The column "languages" specifies self-identified language knowledge. The symbol (h) in the "languages" column indicates home languages used for interaction within the (extended) family. Languages are listed in alphabetical order but home languages come first. Basic information on social class can be deduced from the "activity" column. The column "residence" indicates the neighbourhood of Malabo in which the respective speakers are domiciled. Detailed information on the corpus is provided in Table 1.2 further below.


Table 1.1 Language consultants


Table 1.1 Language consultants

Table 1.2 provides information on the corpus. The list is sorted alphabetically according to the "text code" column, which lists the name of the text (e.g. 03ab). Text names

were given according to mnemonic principles. An "e" at the end of text code indicates that the text consists of elicited data (e.g. 05ae). The "type" column indicates the text genre, "contents" provides a short description of the text. The column entitled "word count" provides an indication of the relative length of texts. An asterisk (\*) after the "text code" indicates that the corresponding text is contained (in part or in full length) in the text section of this book.


#### Table 1.2 Corpus


Table 1.2 Corpus

The corpus presented in Table 1.2 consists of altogether thirty-four texts of different genres totalling 46,060 words. Based on the figures of the "word count" column, narratives constitute approximately 37 per cent of the total corpus (the word count of texts with two genres has been divided by two). This genre encompasses life stories and family histories, illness and near-death accounts, supernatural encounters and other emotionally charged experiences, as well as travel and life abroad. Conversations amount to 25 per cent of the corpus. The topics range from house-building to gender relations, from jesting and joking to metalinguistic discussions during elicitation. In many of the conversations recorded, in particular those involving peer-to-peer communication, form is just as important as content. These conversations "for their own sake" are characterised by emphatic, expressive, and figurative language.

Procedural texts form some 7 per cent of the corpus. They describe various types of routines, for example the preparation of dishes, car maintenance and repair, medical treatment and sorcery, habits and ways of doing things. Elicitation makes up about 33 per cent of the corpus. I employed oral (Spanish to Pichi and monolingual Pichi-based) elicitation to obtain data chiefly on grammatical relations, the classification of situations (i.e. dynamic vs. non-dynamic verbs vs. adjectives), complementation, relativisation, and derivation. I made use of visual, video-based elicitation to uncover the expression of spatial relations including caused positions, the expression of certain complex events ("staged events"), and reciprocity. The video clips of the Language and Cognition Group of the Max-Planck Institute for Psycholinguistics in Nijmegen provided the basis for these elicitiations. Most elicitations were conducted in groups of two or three speakers. 1.8 Presentation of the data

This produced valuable data on variation and encouraged vivid metalinguistic discussions during the exercise.

### **1.8 Presentation of the data**

In examples, the free translation is followed by a text code in squared brackets. Whenever an example features elicited data, the second letter of the text code is an "e", e.g. [dj07ae 137] and [ra07ve 069]. Common parentheses in the free translation line contain supplementary and disambiguating translation material. Squared brackets provide contextual or other relevant meta-information. Punctuation in the Pichi examples follows intonation: A full stop indicates an utterance-final boundary tone, a comma continuative intonation. A slash denotes a speech interruption and hence an incomplete sentence. Spanish words are rendered in the Spanish orthography. I do not provide category labels for Spanish grammatical morphemes where they occur, since this would have complicated interlinear glossing and given Spanish material undue prominence.

A final note is in order on the notion of frequency employed throughout this work. When an exact percentage is not given, certain expressions may indicate the relative frequency or importance of a phenomenon. The expressions in the left column of Table 1.3 correspond approximately to the percentages given in the right column (Michaelis et al. 2013).



## **2 Segmental phonology**

The phonological system of Pichi features a phoneme inventory of twenty-two consonants and seven vowels. There is a good deal of free and allophonic variation in the use of these phonemes. Phonological processes include nasalisation, the use of clitics and the appearance of a linking /r/ during cliticisation, as well as the reduction of consonant clusters by deletion and insertion. In general, however, Pichi speakers tend to fully articulate consonants and vowels. The majority of Pichi words consist of one or two syllables. There are no phonemic long vowels but words may feature clusters of up to three consonants. The segmental system of Pichi interacts in various ways with the suprasegmental system (cf. chapter 3).

### **2.1 Consonants**

The maximal inventory of twenty-two consonant phonemes in Pichi is presented in IPA symbols in Table 2.1. Orthographic symbols are in parentheses, where these differ from IPA symbols. Details on the status and distribution of these phonemes are discussed in sections §2.2 and §2.6.2.1.


Table 2.1 Consonant and approximant phonemes

The (near-)mininal pairs in Table 2.2 establish the phonemic status of the segments contained in Table 2.1.

### **2.2 Consonant allophony and alternation**

#### /**b**/ and /**v**/:

The voiced labio-dental plosive /v/ is a phoneme in its own right in a small number of words, where it does not alternate with /b/, e.g. *greví* [grèví] 'gravy' and *gív=an* [gívàn] 'give him/her/it'. In a second group of words, /v/ is in free variation with /b/, e.g. *vájin*

#### 2 Segmental phonology

Table 2.2 Consonant phoneme minimal pairs


[bádʒìn~vádʒìn] 'virgin', *ívin* [íbìn~ívìn] 'evening', *óva* [óbà~óvà] 'over, be excessive', *sɛven* [sɛ́bèn~sɛ́vèn] 'seven', and *ríva* [ríbà~rívà] 'river'. Free variation is also encountered in the Spanish-derived lexicon of most speakers, as in *abuela* [abwɛla~aßwela~avwɛla] 'grandmother'.

In a third group of words, we only find /b/, which therefore does not alternate with /v/. Hence, we find *fíba* [fíbà] 'resemble', *líba* [líbà] 'liver', *súb* [súb] 'shove', *híb* [híb] 'throw', *bába* [bábà] 'cut hair', and *dɛ́bul* [dɛ́bùl] 'devil'. The orthographic representation chosen for words of the second group, in which we find free alternation between [b] and [v], is <v>. Alternating words are given with both variants in the Pichi-English vocabulary section.

#### /**tʃ**/ and /**dʒ**/:

The voiceless postalveolar affricate tends to be unstable with many speakers and optionally alternates with the voiceless palatal plosive [c] and sometimes with the voiceless postalveolar fricative [ʃ], particularly in word-final position. Hence we find *tɔ́ch* [tɔ́tʃ~tɔ́c~tɔ́ʃ] 'touch'. A small number of speakers, all of which belong to Group 1 (cf. §1.3) exhibit allophonic variation between /tʃ/ and /dʒ/ in some words, with the latter allophone appearing in word-final position before the clitic *=an* '3sg.obj', i.e. *jɔ́ch=an* [dʒɔ́dʒàn] 'judge him/her/it'.

The vast majority of speakers, however, and Group 1 speakers in particular, use wordfinal /tʃ/ in every environment including ones which are not prone to devoicing, i.e. *chénch=an* [tʃéntʃàn] 'change him/her/it'. I have accounted for the fact that most speakers exhibit no such variation by opting for <ch> in the orthography even though word-final /tʃ/ may be an allophone of /dʒ/ for a minority of speakers in words like *jɔ́ch* 'judge' (but not in others, e.g. *kéch* 'catch').

#### 2.2 Consonant allophony and alternation

/**s**/:

The voiced alveolar fricative [z] is attested as a free variant of the voiceless alveolar fricative between two vowels in word-medial position, e.g. *ísi* [ízì~ísì] 'be easy' and *lési* [lézì~lésì] 'be lazy'. I take [z] to be a non-phonemic variant of /s/ in these words.

Furthermore, most Group 1 speakers (cf. §1.3) apply an opposition between /s/ and /ʃ/ (rendered by the grapheme <sh>), which produces minimal pairs like *só* [só] 'sew' and *shó* [ʃó] 'show'. For Group 2 speakers, this opposition is, however, neutralised in favour of /s/, and they employ the voiceless alveolar fricative [s] in any position in which Group 1 speakers may use the voiceless postalveolar fricative [ʃ]. Group 2 speakers therefore produce homonyms like *só* [só] 'sew' and *só* [só] 'show'.

Additionally, Group 2 speakers usually insert a palatal glide /j/ between /s/ and either of the mid vowels /e/ and /ɔ/ where Group 1 speakers only employ /ʃ/. This inter-group variation applies to the following words in the data: *kwɛ́sɔn* [kwɛ́sjɔ̀n~kwɛ́sʃɔ̀n] 'question', *nésɔn* [nésjɔ̀n~néʃɔ̀n] 'nation(ality)', *séb* [sjéb~ʃéb] 'share', *sék* [sjék~ʃék] 'shake', *sém* [sjém~ʃém] 'shame', *sɔ́t* [sjɔ́t~ʃɔ́t] 'be short, shirt', *sén* [sjén~sén] 'same', and *sɔ́p* [ʃɔ́p] 'shop'. Although the insertion of /j/ is optional, it is very common with the words listed. The insertion of /j/ is, however, not generalised to two other words in the corpus featuring a sequence of the phonemes /sé/. Hence, we find *sé* [sé] 'quot' and *fɔséka* [fɔ̀sékà] 'due to'.

The orthography does not represent the segment /j/ in words to which insertion applies. The words that exhibit this alternation are listed in the preceding paragraph and are additionally identified in the Pichi-English vocabulary.

/**n**/ and /**m**/:

The realisation of the alveolar nasal /n/ and the bilabial nasal /m/ is conditioned by a number of factors, which are covered in §2.5.2.

#### /**nj**/ and /**ɲ**/:

A prothetic /n/ is optional (and present in at least half of the occurrences recorded) in a specific group of words with an underlying word-initial /j/. The relevant words are *yandá* [jàndá~njàndá] 'yonder', *yún* [jún~njún] 'be young' and *yús* [jús~njús] 'use'. In this group of words, I therefore analyse the combination of these segments as a cluster consisting of the alveolar nasal /n/ and the palatal approximant /j/.

In a second, equally small group of words, I posit the phoneme /ɲ/, compare the minimal pair *nyú* [ɲú] 'be new' vs. *yú* [jú] '2sg.indp'. The other words that do not alternate in my data and therefore appear to feature a word-initial /ɲ/ rather than the cluster /nj/ are *nyangá* [ɲàŋgá] 'put on airs', *nyankwé* [ɲànkwé] '(the) nyankwé (dance)', *nyɔ́ní* [ɲɔ́ní] 'ant', and *nyús* [ɲús] 'news'. The phoneme /ɲ/ is also found in a word-medial, syllable onset position in two words in the corpus, namely in the place name *Panyá* [pàɲá] 'Spain' and in the ideophone *ményéményé* [méɲéméɲé] 'whine, nag in a childlike fashion'.

A third group of words with a word-initial /j/ does not usually exhibit nasal prothesis at all, e.g. *yɛ́s* [jɛ́s] 'yes', *yét* [jét] 'yet', *yɛ́stadé* [jɛ́stàdé] 'yesterday', and *yáy* [jáj] 'eye'. In the orthography, I only render an initial /n/ with the second group of words, i.e. words that feature the phoneme /ɲ/. Words with an optional prothetic /n/ are listed above and given with their alternate forms in the Pichi-English vocabulary.

#### 2 Segmental phonology

#### /**j**/:

This voiced palatal approximant is a phoneme in its own right in words like *yú* [jú] '2sg.indp', *yá* [já] 'here', *yɛ́s* [jɛ́s] 'yes' and *yét* [jét] 'yet'. Besides that, some words with a word-initial /j/ optionally appear with a prothetic /n/ (cf. on /n/ below). The segment /j/ is also optionally inserted between /s/ and one of the mid-vowels /e/ and /ɔ/ in another group of words (cf. on /ʃ/ below).

Further, /j/ is optionally inserted between either of the velar consonants /g/ and /k/ and the front vowels /a/ and /ɛ/. However, this process only applies to a few relevant words of English origin with which it occurs in the majority of instances. The corpus contains the following words to which this applies: *gádin* [gádìn~gjádìn], *gál* [gál~gjál] 'girl', *gɛ́l* [gɛ́l~gjɛ́l] 'girl', *káp* [káp~kjáp] 'cap', *kápinta* [kápìntà~kjápìntà] 'carpenter', and *kɛ́r* [kɛ́r~kjɛ́r] 'carry'. In contrast, a /j/ is not normally inserted in other words of English origin like *gɛ́t* [gɛ́t] 'get', *kán* [kán~kám] 'come', and *káyn* [kájn] 'kind', as well as a group of words of non-English origin with an L.H pitch pattern, amongst them *garí* [gàrí] 'garí', *kaká* [kàká] 'defecate', *kasára* [kàsárà] 'cassava', and *kandá* [kàndá] 'skin'.

The orthography does not render the epenthetic /j/ in words that feature it. All relevant words are listed above and are identified in the Pichi-English vocabulary section.

/**r**/:

The symbol /r/ varies in pronounciation between that of a voiced uvular fricative [ʁ] and a velar fricative [ɣ]. Some speakers use an alveolar tap [ɾ] instead of these two segments, and I have also occasionally heard an uvular trill [ʀ]. We therefore find variants like the following: *máred* [máʁèd~máɣèd~máɾèd] 'marry', *dríng* [dʁíng~dɣíng~dɾíng] 'drink', *kɛ́r* [kɛ́ʁ~kɛ́ɣ~kɛ́ɾ] 'carry', and *rɛ́s* [ʁɛ́s~ɣɛ́s~ɾɛ́s] 'rice'. The orthography represents this segment as <r> and as [r] for phonemic and phonetic transcriptions.

/**h**/:

This voiced glottal fricative is phonemic in a small group of words which is delineated by minimal pairs like *hól* [hól] 'hole, hold' vs. *ól* [ól] 'be old'. The group contains words like *hát* [hát] 'hurt, heart', *hála* [hálà] 'shout', *hós* [hós] 'house', and *héd* [héd] 'head'. The group also includes two words with a word-medial /h/, namely *bihɛ́n* [bìhɛ́n] 'behind' and *wahála* [wàhálà] 'trouble'.

With a second and larger group, /h/ may be inserted at the beginning of the vowelinitial word. Such a prothetic /h/, although optional, occurs more often than not with most words in this group. Hence we find variants like *ánsa* [ánsà~hánsà] 'respond', *áks* [áks~háks] 'ask', *ópin* [ópìn~hópìn] 'open', and *évi* [évì~ébì~hévì~hébì] 'be heavy'. In some instances, it is however impossible to determine whether a word-initial /h/ is prothetic or part of the segmental structure of a word, because the data contains no recorded instance without an initial /h/. Some of the words to which this applies are *húman* 'woman', *hɛ́lp* 'help', *hébul* 'be able', *hía* 'year', *hásis* 'ashes', and *hós* 'house'. I have chosen to render these words with an initial <h>.

A third group of vowel-initial words is not attested with a prothetic /h/, e.g. *óva* [óvà] 'be excessive, over'; *ónli* [ónlì] 'only', *áfta* [áftà] 'then', and *éch* [étʃ] 'age'. In the orthography, the segment /h/ is only represented with words that always appear with a wordor syllable-initial /h/.

2.3 Vowels

/**gb**/ and /**kp**/:

These two voiced and voiceless labiovelar plosives are marginally phonemic and only occur in a handful of ideophones, e.g. *nák gbin* 'hit ideo' = 'hit hard and unexpectedly', *sút kpu* 'shoot ideo' = 'shoot followed by the sound of a dull impact on the body'.

### **2.3 Vowels**

Pichi has seven vowel phonemes, which are presented in IPA symbols in Table 2.3. The IPA symbols for vowels are also used as orthographic symbols in this work. Vowel length is not distinctive. Vowel allophony and alternation are discussed in §2.4.

The following (near-)minimal pairs establish the phonemic status of the segments contained in Table 2.3:



#### 2 Segmental phonology

### **2.4 Vowel allophony and alternation**

Pichi shows some lexically determined vowel alternation. Hence we find alternate forms like *kɛ́r~kɛ́ri~kári* 'carry, take', *lɛ́k~láyk* '(to) like', *gɛ́l~gál* 'girl', *unu~una* '2pl', *wɔ́nt~wánt* 'want'. Other than that, there is some variation in the use of mid-vowels, with a tendency towards the reduction of phonemic contrasts. Furthermore, Pichi has vowel-vowel combinations, as well as sequences consisting of an approximant and a vowel. There are no phonemic long vowels in Pichi. The properties of sequences of non-identical vowels are covered in §2.6.2.2.

/**e**/ and /**ɛ**/:

Minimal pairs such as *yɛ́s* [jɛ́s] 'yes' vs. *yés* [jés] 'ear' establish the phonemic status of the unrounded close-mid front vowel /e/ and the unrounded open-mid front vowel /ɛ/. However, many speakers collapse the phonemic contrast between /e/ and /ɛ/ by raising /ɛ/ towards /e/. The opposite direction is far less common. Hence, variants like the following ones are attested: *lɛ́k* [lɛ́k~lék] 'like', *chɛ́k* [tʃɛ́k~tʃék] 'check', *kɛ́r* [kɛ́r~kér] 'carry', and *nɛ́k* [nɛ́k~nék] 'neck'. The use of either variant of a content word also often conditions the vowel quality of preceding or following function words (cf. §2.5.3).

/**o**/ and /**ɔ**/:

The phonemic status of the rounded close-mid back vowel /o/ and the rounded openmid back vowel /ɔ/ is evident in minimal pairs like *kól* [kól] 'be cold' vs. *kɔ́l* [kɔ́l] 'call' and *fɔ* [fɔ̀] 'prep' vs. *fó* [fó] 'four'. Nonetheless, many speakers also neutralise this phonemic contrast by raising /ɔ/ towards /o/. With content words, this neutralisation is less common than the /e~ɛ/ alternation. However, it is almost generalised with Group 1 speakers (cf. §1.3) in words with grammatical functions, such as the associative preposition *fɔ* [fɔ̀~fò] 'prep', the comparative adverb *mɔ́* [mɔ́~mó] 'more', the negator *nó* [nó~nɔ́] 'neg', the coordinator *ɔ* [ɔ̀~ò] 'or', the TMA marker *nɔ́ba* [nɔ́bà~nóbà] 'neg.prf'. The negative focus marker *cum* negative identity copula *nóto* 'neg.foc' is however routinely pronounced [nótò].

### **2.5 Phonological processes**

Phonological processes include lenition and fortition, nasalisation, vowel assimilation, deletion and insertion, as well as cliticisation.

### **2.5.1 Lenition and fortition**

Lenition, the weakening of segments, may affect stops in intervocalic position as in *bigín* [bìɣín] 'begin'. Strengthening, or fortition, affects voiced obstruents, which are generally devoiced in word-final position. Devoicing therefore produces the following word-final variant of segments. The details regarding lenition and fortition outside of these specific contexts require further investigation:

2.5 Phonological processes


### **2.5.2 Nasals and nasal place assimilation**

A number of processes involve nasals and nasalisation. These apply in diverse ways to different groups of words. We have seen that /n/ prothesis or prenasalisation is optional with a group of words featuring an initial /j/ (cf. §2.2). Secondly, the following group of verbs with a word-final /i/ and an H.L pitch configuration is optionally (and very frequently) subjected to word-final nasalisation (realised as /n/ or nasalisation of the final /i/): *grídi* [grídì~grídìn] 'be greedy', *hángri* [hángrì~hángrìn] 'be hungry', *hɔ́nti* [hɔ́ntì~hɔ́ntìn] 'hunt', *hɔ́ri* [hɔ́rì~hɔ́rìn] 'hurry', *ísi* [ísì~ísìn] 'be easy', *lési* [lésì~lésìn] 'be lazy', *lɔ́ki* [lɔ́kì~lɔ́kìn] 'be lucky', *sɔ́ri* [sɔ́rì~sɔ́rìn] 'be sorry', *wɔ́ri* [wɔ́rì~wɔ́rìn] 'worry', and *tɔ́sti* [tɔ́stì~tɔ́stìn] 'be thirsty'. This group of words may be contrasted with a second group that also features a word-final /i/, but exclusively occurs with a word-final nasal. In this latter group, we find words such as *físin* [físìn] '(to) fish', *ívin* [ívìn] 'evening', *mɔ́nin* [mɔ́nìn] 'morning', and *pikín* [pìkín] 'child'.

A third group of words features a word-final /i/, but is not attested with a final /n/. This group includes words with an L.H pitch configuration, such as *rɛdí* [rɛ̀dí] 'be ready', *greví* [grèví] 'gravy', and *dɔtí* [dɔ̀tí] 'be dirty'. It also contains monosyllabic words like *mí* [mí] '1sg.indp', *sí* [sí] 'see', and *grí* [grí] 'agree'.

A fourth group involves function words that are subjected to nasal place assimilation. The relevant words are the personal pronouns *=an* '3sg.obj', *dɛn* '3pl', and *dɛ́n* '3pl.indp', the preposition *frɔn* 'from', the locative noun *bɔtɔ́n* 'under(side)', the TMA marker and verb *kán* 'pfv, come', the determiner *sɔn* 'some, a', and the pronominal *sén* 'same'. In these words, the final nasal is conditioned by the place of articulation of the following segment:


2 Segmental phonology

(6) *Pút=an* put=3sg.obj *dé!* there [pútà**n d**é] 'Put it there!'

Anticipatory nasalisation of a vowel preceding the nasal consonant of these function words is also commonplace (7). The word-final nasal of these words may be deleted altogether, in which case a nasal trace is left behind with the preceding vowel (8):


Before a pause, hence when there is no assimilatory pressure from following segments, the word-final nasal in these function words may either be realised as [n] or [m], as in (9) and (10), respectively. The analysis of a subcorpus revealed that two thirds of prepausal instances of the word-final nasal were realised as [n], with the remaining third being realised as [m]. Instances of prepausal *kán* necessarily involve the content word 'come' rather than the homonymous preverbal aspect marker *kán* 'pfv'. The Pichi equivalent of the content word 'come' is more often pronounced as [kám] than [kán] (11):


The orthographic choice of <n> for the word-final nasal with these grammatical words reflects these tendencies. Nevertheless, the content word 'come' is also written as *kán* in order to preserve the orthographic unity of the etymologically related aspect marker and content word.

### **2.5.3 Vowel assimilation**

Pichi features a tongue root vowel harmony targeting mid-vowels. The distinction between the [+high] vowel /e/ and the [-high] vowel /ɛ/, and between [+high] /o/ and

2.5 Phonological processes

[-high] /ɔ/ is collapsed in stem vowels. Enclitics and adjoining function words harmonise with the stem. Hence we find *dɛn de kéch dɛ́n* → [den de kéch dén] 'they [ipfv] catch them'. Compare (12) and (13). Note that in (12), the speaker also collapses the phonemic contrast between /e/ and /ɛ/ in *mék* /mék/ 'make' (cf. §2.4):


These harmonic processes are reflective of a general tendency of function words to be phonologically assimilated to adjoining words.

### **2.5.4 Insertion and deletion**

We have seen that the insertion of consonants affects various types of words (cf. §2.5.2 and the entries /h/, /s/, /j/, and /n/ in §2.6.2.1). Deletion is less frequent. In general, vowels and consonants of content words tend to be fully articulated (except cf. 16–17). Nevertheless, high-frequency (function) words tend to be phonologically reduced or fused with adjoining words to a greater degree than other words. One function word, the TMA marker *nɛ́a* 'neg.prf', is not pronounced as the fuller variant [nɛ́và~nɛ́bà] in natural speech in the corpus. The virtually complete sound change of this TMA marker is reflected in the orthographic choice of *nɛ́a* (15).

This contrasts with the pronunciation of the functionally equivalent word *nɔ́ba* [nɔ́bà ~nɔ́à] 'neg.prf' which occurs equally often in the reduced and full variants. Note that segment deletion may have repercussions for the use of tone (cf. §3.2.2):

(15) *Dɛn* 3pl *nɛ́a* neg.prf *rích* arrive *dé.* there [dɛ̀n nɛ́à rích dé]

'They haven't arrived there yet.'

Pichi speakers exhibit a systematic tendency to break up onset consonant clusters in which the first segment is the fricative /s/ and the second a liquid or nasal. Both insertion and deletion are employed to achieve this end. The biconsonantal clusters /sl/, /sn/, and /sm/ are very often broken up by insertion of the vowels /i/ or /u/. Thus we have *slíp* [slíp~sìlíp] 'lie down', *smɔ́l* [smɔ́l~sìmɔ́l~sùmɔ́l] 'be small', and *snék* [snék~sìnék] 'snake'. Biconsonantal sequences of /sk/ and /sp/ are not reduced – hence *skín* [skín] 'body' and *spún* [spún] 'spoon'.

#### 2 Segmental phonology

Optional reduction can be observed with onset clusters involving a sequence of the fricative /s/, a stop, and a fricative or approximant, namely the biconsonantal cluster /st/ and the triconsonantal clusters /str/, /skr/, and /skw/. The possibility of reduction is, however, lexically restricted to specific words in the corpus. Therefore \*[tímà] is, for example, rejected for *stíma* [stímà] 'ship'. The pronunciation of the initial /s/ is optional in the following words, with either variant being equally common: *skrách* [skrátʃ~krátʃ] 'scratch', *skwís* [skwís~kwís] 'squeeze', *stík* [stík~tík] 'tree', *stón* [stón~tón] 'stone', *strít* [strít~trít] 'street', and *strɔ́n* [strɔ́n~trɔ́n] 'be strong'. Next to the words listed above, four additional words occur with an initial /s/ only once in the corpus, namely *tínap* [stínàp~tínàp] and its variant *tánap* [stánàp~tánàp] 'stand (up)', *pínch* [spíntʃ~píntʃ] 'pinch', and *trímbul* [strímbùl~trímbùl] 'tremble'. I therefore assume that these alternants are the result of spontaneous back-formation. Words to which optional /s/ deletion applies are given with their alternate forms in the Pichi-English vocabulary list.

The tendency to avoid clustering also frequently leads to the insertion of an epenthetic vowel into coda consonant clusters featuring liquid-stop sequences. Hence, with the three possible coda clusters /lp/, /lt/, and /lk/ (cf. Table 2.8), insertion produces free variants like *hɛ́lp* [hɛ́lp~hɛ́lɛ̀p] 'help', *bɛ́lt* [bɛ́lt~bɛ́lɛ̀t], and *milk* [mílk~mílìk] 'milk'. In addition, Pichi speakers manifest a marked tendency to avoid the clustering of consonants across word boundaries. This leads to the deletion of word-final consonants as in (16) and (17) below.


The deletion of word-final consonants and the reduction of word-initial clusters is indicative of a general tendency towards CV syllable structures where this is possible. Other processes in which insertion is relevant are covered in §2.2, §2.6.3 and §3.3. The latter section also treats the insertion of a linking /r/.

### **2.6 Phonotactics**

The distribution of some consonants and vowels has already been touched upon in §2.2 and §2.4. The following sections provide details on the ordering principles of Pichi phonemes. Pichi also exhibits an instance of tone-conditioned suppletive allomorphy, a phenomenon relating to suprasegmental phonotactics covered after the basics of the tone system have been described (cf. §3.2.5).

2.6 Phonotactics

### **2.6.1 The word**

The vast majority of Pichi words are mono- and bisyllabic. In addition, most words carry a single H tone over their only, penultimate, or final syllable (cf. §3.1.3). The presence of a single H tone per word and knowledge of the possible tonal configurations therefore provides a means of metrically delineating the prosodic word in very much the same way as the position of stress does in intonation-only languages.

### **2.6.2 The syllable**

The syllable template in Pichi is (C)(C)(C)(V)V(C)(C). A vowel constitutes the syllable nucleus. There are a few single-vowel roots, all of which are function words, e.g. *a* '1sg.sbj', *e* '3sg.sbj', or *ó* 'sp'. There are no phonemic long vowels in Pichi, adjacent vowels are invariably heterosyllabic.

Pichi has many words with initial biconsonantal clusters. Some word-initial clusters consisting of three consonants also exist. But both bi- and triconsonantal word-initial onsets tend to be broken up by deletion and insertion (cf. §2.5.4). Word-final consonant clusters contain up to two segments and involve nasals, liquids and approximants as the penultimate segment, or the fricative /s/ as the final segment of the coda. In connected speech, a word-final consonant, whether as the final consonant of a clustered coda or the only consonant of a coda, is often deleted.

#### **2.6.2.1 Distribution of consonants**

Table 2.5 presents the distribution of the twenty-two Pichi consonants in syllables (syllableinitial in the onset and syllable-final in the coda) and words (initial, medial, and final). The following abbreviations apply: IO = word-initial onset; MO = word-medial onset; MC = word-medial coda; FC = word-final coda.


Table 2.5 Distribution of consonant phonemes

Table 2.5 allows the conclusion that all twenty-two consonant phonemes save /ŋ/ occur as word-initial onsets. All consonants except /ŋ/, /kp/, and /gb/ occur as word-medial

#### 2 Segmental phonology

onsets as well. The latter two phonemes are only attested as word-initial onsets in ideophones. Eleven consonants appear in word-medial codas out of which two consonants appear as word-medial onsets in only two words each, namely /ɲ/ (*Panyá* 'Spain, Spanish' and *ményéményé* 'whine, nag in a childlike fashion') and /h/ (*bihɛ́n* 'behind' and *wahála* 'trouble'). Sixteen consonants occur in word-final codas. Examples of the distribution of consonants follow in Table 2.6.

Only roots are taken into account in Table 2.6, not phonological words. In compounds, all consonants that may appear in word-final position in roots may additionally do so in word-medial coda position at the morpheme boundary. Compare the opaque compound *big-dé* 'big.cpd-day' = 'festivity', the reduplicative compound *tɔch-tɔ́ch* 'touch repeatedly', and the lexicalised reduplication and ideophone *gbogbogbo* 'in haste'.

More than one consonant may appear in syllable onsets and codas. Table 2.7 lists the possible permutations of consonant clusters in syllable onsets, and Table 2.8 lists consonant combinations in the coda. Table 2.7 shows that up to three consonants may cluster in onsets. Clusters of three consonants may be broken up by deletion and insertion (cf. §2.5.4). The sequences /gj/, /kj/, and /sj/ may be said to arise through phonological processes alone (cf. also §2.2). The sequences /gj/ and /kj/ surface through optional /j/ epenthesis in words like *gál* [gál~gjál] 'girl' and *kɛ́r* [kɛ́r~kjɛ́r] 'carry', while the sequence /sj/ appears in variants like *sɔ́p* [sɔ́p~sjɔ́p] 'shop' (cf. also §2.2).

Coda clusters are limited to maximally two consonants. Coda clusters always involve nasals or continuants, and liquid-stop sequences may also be broken up by epenthetic vowels (e.g. *hɛ́lp* [hɛ́lɛ̀p] 'help'). Possible cluster permutations in the coda are listed in Table 2.8.

#### **2.6.2.2 Distribution of vowels and approximants**

All Pichi vowels may occur in word-initial position. In general, however, vowels only appear in word-initial position in a small number of words. The majority of Pichi words, and content words in particular, either have a consonant, an approximant or a prothetic /h/, sometimes a prothetic /y/ or /w/, in the onset of their initial syllable.

Most words that do have an initial vowel are function words: personal pronouns (e.g. *a* '1sg.sbj', *e* '3sg.sbj', *una* '2pl', and *ín* '3sg.indp), question words (e.g. *údat* 'who' and all words featuring the clitic question particle *ús=* 'q'), clause linkers (e.g. *adɔnkɛ́* 'even if', *ɛf* 'if', and *áfta* 'then'), locative nouns (e.g. *ínsay* 'inside' and *ɔntɔ́p* '(on)top'), quantifiers (e.g. *ɔ́da* 'other', *ɛ́ni* 'every'), and interjections (e.g. *ékié* 'good gracious', *áy* 'expression of pain'). Some content words also feature a word-initial vowel (e.g. *aráta* 'rat', *éch* 'age(-grade)', *ívin* 'evening', and *ɛ́nta* 'enter'). In contrast, vowels in word-final position are very common and we find them throughout all word classes (e.g. *mí* '1sg.sbj', *butú* 'stoop over', *sóté* 'until', *nó* 'know', *bɛlɛ́*'belly', *fɔ* 'prep', and *sísta* 'sister'). There are certain restrictions on sequences of vowels. Not only are there no phonemic strings of two identical vowels (i.e. long vowels) in Pichi; vowel-vowel sequences are heterosyllabic. In such cases of vowel hiatus, the immediately adjacent nuclei bear polar tones, e.g. *bi.ó* [L.H] 'behold', *klí.a* [H.L] 'clear' vs. *\*fɔ=an* [L.L] 'for him/her'). This tonotactic restriction triggers a tone-conditioned suppletive allomorphy of two forms instantiating 3sg ob-


Examples of consonant distribution

Table

2.6

2.6 Phonotactics

#### 2 Segmental phonology


Table 2.7 Onset consonant clusters

2.6 Phonotactics


Table 2.8 Coda consonant clusters

ject case, a typologically interesting phenomenon not attested in other Afro-Caribbean English-lexifier Creoles (cf. §3.2.5). There are also only certain types of admissable vowel combinations, provided in Table 2.9.

Table 2.9 Vowel sequences


i ìó íɛ̀ íà

Sequences involving an approximant and a vowel are presented in Table 2.10. Pichi features both falling and rising sequences. In the former, the vowel comes first (e.g. /ɔj/), while in rising sequences, the vowel follows the approximant (e.g. [wi]). The logically possible sequences \*[ji] and \*[ɔw] are not attested in the corpus.

#### 2 Segmental phonology

Table 2.10 Sequences involving an approximant and a vowel


A comparison of Table 2.9 and Table 2.10 shows that opening sequences are realised as vowel-vowel sequences, while closing sequences are realised as vowel-approximant strings. The circumstances surrounding cliticisation speak to the validity of differentiating between vowel-vowel and vowel-approximant sequences. Due to a restriction imposed by tonal phonotactics, *=an* may not encliticise to a vowel-terminal host if the final vowel of the host carries a low tone (cf. §3.2.5). Monosyllabic verbs featuring an approximant as the final segment may, however, take the object pronoun *=an*. Compare the verb *báy* 'buy' in (18):

(18) *Yu* 2sg *wánt* want *báy=an* buy=3sg.obj *na* loc *puerto* port (*…*)

'(If) you want to buy it at the port (…)'

If the word-final approximant /j/ in *báy* [báj] 'buy' were an [i], i.e. a vowel, and a tonebearing segment in its own right, it should be low-toned in accordance with Pichi tonal phonotactics (since it is preceded by a high-toned vowel [á]). A low-toned final vowel would, in turn, block the encliticisation of *=an* as it does with other verbs with a final low tone. This is, however, not the case, since the sequence [áj] is monomorphemic and bears a single high tone. There is thus no restriction on the encliticisation of *=an*. The same principle applies to other verbs with a final approximant, e.g. *aláw=an* 'allow=3sg.obj' = 'allow her/him'.

The distribution of approximants in the syllable may be read from the tables given in §2.6.2.1. Some observations are in order here on variation in strings of approximants and vowels. The verb *drɛ́b* 'drive' features the variants [drɛ́b~drájb]. However this free alternation is not encountered with other words to which it could potentially apply. Hence on the one hand, we find *bɛ́t* [bɛ́t] and *fɛ́t* [fɛ́t] 'fight'. On the other hand, words like *bráyt* [brájt] 'be bright', *táyt* [tájt] 'be tight', and *wáyp* [wájp] 'wipe' do not have less complex variants with a monosegmental [ɛ] instead of the bisegmental [aj].

The series [ɔj] is found in two groups of words. The first group consists of only two words in the corpus. A second group of words exhibits a free alternation between the strings [ɔj] and [wɛ] with a preference for the latter sequence. A third group of words invariably features [wɛ] and is not attested with the [ɔj] variant:

2.6 Phonotactics

(19) Group 1 *bɔ́y* [bɔ́j] 'boy' *ɔ́yl* [ɔ́jl] 'oil' Group 2 *spwɛ́l* [spwɛ́l~spɔ́jl] 'spoil, spend' *bwɛ́l* [bwɛ́l~bɔ́jl] 'boil' *jwɛ́n* [dʒwɛ́n~dʒɔ́jn] 'join' Group 3 *swɛ́la* [swɛ́là] 'swallow' *kwɛ́nch* [kwɛ́ntʃ] 'die off' *kwɛ́sɔn* [kwɛ́sjɔ̀n] 'question' *wɛ́l* [wɛ́l] 'be well'

Note that group 1 contrasts with group 2 in that [ɔj] in group 1 is either word-final (i.e. *bɔ́y*) or word-initial and the nucleus of a syllable without an onset (i.e. *ɔ́yl*). In turn, words in group 3 are either bisyllabic (i.e. *kwɛ́sɔn* and *swɛ́la*) and feature a consonant cluster in the coda (i.e. *kwɛ́nch*) or begin with the alternating feature (i.e. *wɛ́l*). Hence the characteristic environment for the [wɛ́~ɔ́j] alternation is a monosyllabic word with a heavy syllable, a single consonant in the coda, and an onset featuring a stop (or a stop component like the affricate [dʒ]).

### **2.6.3 Cliticisation**

Pichi has at least two clitics which participate in forming phonological words. The proclitic question particle *ús=* 'q' attaches to mostly generic nouns in order to form basic question words. The enclitic object pronoun *=an* '3sg.obj' attaches to verbs, prepositions, locative nouns, and in double-object constructions to other object pronouns (i.e. "the hosts").

Cliticisation in Pichi is characterised by segmental reduction, the loss of morphosyntactic independence, and inseparability from the host. Two elements can be considered full clitics by these criteria: The object pronoun *=an* '3sg.obj' and the question particle *ús=* 'q'. Other elements are clitic-like to a lesser degree: Dependent person pronouns may be said to be enclitic to the following element of the predicate, the pluraliser *dɛn* 'pl' to the preceding noun.

The question element *ús=* 'q' is proclitic to generic nouns in question words. These question words form single prosodic words, and the proclitic is phonologically adapted to the host; hence *ús=tín* [útín] 'what' and *ús=káyn* [úkájn] 'which'.

The object pronoun *=an* '3sg.obj' is enclitic to the preceding verb, preposition, or locative noun with which it forms a single phonological word. The pronoun *=an* '3sg.obj' may also encliticise to a preceding H-toned object pronoun in double-object constructions (cf. §9.3.4). The pronoun undergoes a higher than usual degree of segmental reduction, hence we find the variants [=àn~ã̀~à]. Under certain conditions, the enclisis of *=an* triggers a tone-conditioned suppletive allomorphy, a (tonal) phonotactic phenomenon described in §3.2.5.

## **3 Suprasegmental phonology**

Pichi is a tone language. In previous work, I posited that Pichi has a mixed prosodic system in which individual words are either specified for pitch accent or tone (Yakpo 2009a,c), similar to systems claimed for other European-lexifier creoles of the Atlantic basin (e.g. Rountree 1972; Alleyne 1980; Devonish 1989; 2002; Good 2004; Rivera-Castillo 1998; Rivera Castillo & Faraclas 2006). In subsequent work on Pichi and comparative work on the prosodic systems of other Afro-European contact varieties (e.g. Bordal Steien & Yakpo 2018), I found no evidence that the Pichi lexicon is stratified and that "tonal" and "pitch-accented" words differ with respect to their pitch-related properties or the tonal processes described in this chapter. I therefore treat the prosodic system of Pichi as a tonal system *tout court*. In the following, the term "tone class" designates the various fixed pitch patterns that Pichi words fall into.

The pitch analyses in this chapter were done from connected speech and from words pronounced in isolation using the Praat 5.0 software. The analyses are presented in figures containing a pitch trace and a syllabic segmentation of the utterance. The transcription employed for rendering syllabic segments is orthographic. Nonetheless, phonetic tones are marked on each syllable in the figures for easier recognition.

The approximate pitch values of each syllable are given in Hertz (Hz) on the vertical axis. The horizontal axis provides the time elapsed (1.0 = 1 second). In the examples in this chapter, the second line contains a phonetic tonal notation of the Pichi utterance above. When a tonal process is described, the relevant Pichi sentence is sometimes repeated after the arrow (→). The second line of the Pichi utterance following the arrow then provides phonetic tone, i.e. the actual pronunciation of the sentence after the tonal process under discussion has taken place. For clarity of presentation, text codes have been omitted with examples in this chapter.

### **3.1 Characteristics of tone**

Pichi has two distinctive tonemes, namely a High (H) and a Low (L) tone. The language employs lexical and morphological tone, and there is an unevenly distributed number of tone classes. Boundary tones at the right edge of utterances fulfil the pragmatic and grammatical functions of intonation (cf. §3.4).

The tone-bearing unit in Pichi is the syllable. Vowels and sonorants serve as tonebearing segments. Evidence comes from the interaction of lexical tones and boundary tones over utterance-final syllables. In utterance-final position, a boundary tone will associate with the final tone-bearing unit of the utterance. The sonorants /n/, /m/, /l/, and /r/ may bear phonetic tone in Pichi. Hence, an utterance-final /n/, for example, may carry a boundary tone.

#### 3 Suprasegmental phonology

Consider the citation form of *tɛ́n* 'time' in Figure 3.1. Here the declarative L% (L boundary tone), which follows the lexical H tone over /ɛ́/, is spread out over the vowel and the final /n/. Sonorants like /n/ do not, however, bear lexical tone by themselves. Rather, they always bear the tone of the left-adjacent, i.e. preceding, vowel. In contrast, with non-sonorant final segments, tone is only borne by the preceding vowel. The final obstruent in *tɔ́k* 'talk' in Figure 3.2 cannot bear tone, so the utterance-final declarative L% is borne by the vowel alone.

Figure 3.1 Citation form of *tɛ́n*

Figure 3.2 Citation form of *tɔ́k*

(2) *Tɔ́k.* hl% 'Talk.' 3.1 Characteristics of tone

When the utterance-final word is a light (vowel-final) monosyllable, the vowel may be lengthened, sometimes up to two beats. I assume that the lengthening of light monosyllables is caused by the metric preference of Pichi for footed tonal domains within the word boundary. Heavy monosyllables with a final non-tone-bearing segment like *tɔ́k* 'talk' block the creation of footed domains in utterance-final position. But light syllables leave room for this option. The vowels of the light monosyllables in Figure 3.3 and Figure 3.4 have been lengthened in order to accommodate the HL contour consisting of the lexical H tone of the monosyllable and the declarative L% boundary tone.

Figure 3.3 Citation form of *só*

Figure 3.4 Citation form of *dé*

3 Suprasegmental phonology

(4) *Dé.* h**l%** 'There.'

### **3.1.1 Distinctive tones**

Pichi contrasts two level tones, a high tone (H) and a low tone (L). H tone is the more active tone in tonal processes: H rather than L participates in tone spreading and is more active in pitch register expansion. Contour tones do not constitute tonemes in their own right. Instead, they result from the succession of a lexical tone and a polar floating tone over a single tone-bearing unit (cf. §3.2.2).

Figure 3.5 and Figure 3.6 present the pitch trace and segmentation of the two words *hasis* /H.L/ 'ashes' and *dɔtí* /L.H/ 'be dirty' said in isolation. The two words *hasis* and *dɔtí*

Figure 3.6 L.H pattern

#### 3.1 Characteristics of tone

represent the tone patterns of the two most frequent tone classes of Pichi (cf. Table 3.1). The mean pitch on the L-toned syllable of *dɔtí* is 109.17 Hz, that of the H-toned syllable 129.27 Hz. Hence, the difference in pitch between the H- and L-level tones amounts to 20.1 Hz. With *hásis*, the mean pitch of the H tone is 108.59 Hz, while the mean L tone stands at 99.72 Hz. The difference in mean pitch between H and L therefore stands at 8.87 Hz. This difference is just about half of that between L and H in *dɔtí*.

Table 3.1 Pitch values


The relatively small difference in mean pitch between the syllables of *hásis* arises due to the fact that the H tone over the first syllable is carried over into the first half of the following L-toned syllable. In contrast, the L tone of the first syllable of *dɔtí* shows no signs of rightward spreading.

Words may bear a single or more H or L tones. Compare the pitch traces of the utterance-final tonal words *nyɔ́ní* 'ant' and *Bata* 'Fang' in the collocations *lɛk nyɔ́ní* 'like ants' and *tɔ́k Bata* 'speak Fang' in Figure 3.7 and Figure 3.8.

Equatoguinean Spanish has been analysed as a tone language, in which the lexical stress characteristic of Spanish has been converted to lexical tone due to contact with

Figure 3.7 H.H pattern

#### 3 Suprasegmental phonology

Figure 3.8 L.L pattern

the tone languages of Equatorial Guinea (Lipski 2015; Bordal Steien & Yakpo 2018). Words codeswitched or borrowed from Equatoguinean Spanish are therefore specified for lexical tone just like Pichi words.

Figure 3.9 Pitch over Spanish *abril*

Figure 3.9 and Figure 3.10 feature the utterance-final Spanish words *abril* 'April' and *nigeriano* 'Nigerian', the latter in the collocation *na nigeriano* 'foc Nigerian' *=* 'He is a Nigerian'. The pitch configurations over these two words conforms to those of Pichi words with a word-final (Figure 3.9) and a penultimate (Figure 3.10) H tone, respectively.

#### 3.1 Characteristics of tone

Figure 3.10 Pitch over Spanish *nigeriano*

### **3.1.2 Lexical and morphological tone**

A small number of monosyllabic roots are distinguished from each other by pitch alone. The list in (5) contains most words in the corpus to which this applies. In conformity with a general pattern, (more) functional words are L-toned, while the corresponding content words are H-toned.


However, there are also numerous homophones, which can neither be distinguished segmentally, nor by their pitch properties. The following list contains most homophones in the corpus.



#### 3 Suprasegmental phonology


Morphological tone is employed in the personal pronoun paradigm in order to distinguish morphologically different forms of the same lexeme from one another (e.g. *mi* '1sg.poss' – *mí* '1sg.indp', *dɛn* '3pl' – *dɛ́n* '3pl.indp'). Pichi also features a morphological tonal process (cf. §3.2.4). In addition, there are three items that derive from a common etymon and are distinguished by pitch alone: *de* 'ipfv' – *dé* 'be.loc', *di* 'def' – *dí* 'this', *go* 'pot' – *gó* 'go'). All low-toned monosyllabic roots are words with more or less grammatical functions, such as personal pronouns (e.g. a '1sg.sbj'), determiners (e.g. *di* 'def'), TMA markers (e.g. *bin* 'pst', *kin* 'hab'), clause linkers (e.g. *ɛf* 'if'), or prepositions (e.g. *pan* 'on'). Low-toned function words, except dependent personal pronouns, are listed in (7).


There are, however, limits to this pattern of functional differenciation by tone. The monosyllabic roots *dɔ́n* 'down, done, prf', *kán* 'come, pfv', *mék* 'make, sbjv', *sé* 'say, qot', and *wán* 'one, a' also have a more grammatical meaning besides their lexical one. Yet, their different functions are covered by segmentally and suprasegmentally identical forms.

Pichi also exhibits one morphological tonal process. In compounds and morphological reduplication, the H tones over all non-final components are deleted and replaced by an L tone (cf. §3.2.4).

3.1 Characteristics of tone

### **3.1.3 Tone classes**

About 95 per cent of roots contained in my lexical data-base carry a single H tone over their only, penultimate, or final syllable. Other syllables in these words are L-toned. The remaining 5 per cent of roots feature diverse tone patterns with more than one H, or no H tone. Many (e.g. *nyɔ́ní* 'ant' < Mende *yɔ́ní* 'red ant') but not all (e.g. *ápás* 'after' < English 'half-past') of these words originate from African languages or are monosyllabic function words with an L tone over their only syllable (cf. 7), while words with a single H tone are mostly English-derived. This circumstance speaks to the fact that stress-totone conversion took place in the formation of the proto-language of Pichi, as in many other Afro-European creole and non-creole contact languages (e.g. Berry 1970; Criper 1971; Criper-Friedman 1990; Alleyne 1980; Gussenhoven & Udofot 2010; Bordal Steien 2015).

Table 3.2 contains a listing of the tone classes of the simplex roots contained in the lexical data base of the corpus (cf. Faraclas 1996; Good 2004, for pitch classes in Nigerian Pidgin and Saramaccan). A few examples are provided for each tone class. Not included in this table are ideophones, which feature numerous idiosyncratic tonal patterns and often involve lexicalised reduplication and triplication (cf. §4.5.3 and §12.1 for a detailed treatment).

Members of the monosyllabic L-toned tone class only contribute a total of nineteen roots and 2.5 per cent of the total in terms of individual entries and are hence listed as belonging to a minor tone class. The members of this class are, however, mostly function words that constitute the backbone of the grammatical system of Pichi: the personal pronouns *a* '1sg.sbj', *e* '3sg.sbj', *=an* '3sg.obj'; the TMA markers *de* 'ipfv', *go* 'pot', *bin* 'pst'; the preposition *fɔ* 'prep' and the homonymous forms *na* 'loc' and *na* 'foc' outrank any other root of the language in a frequency count. This makes this tone class perceptually as salient as the H and H.L tone classes. In contrast, the members of the other minor tone classes are each composed of relatively few lexical words, which together make up 6 per cent of roots in the corpus.

Table 3.2 points to additional characteristics of the corpus. With 54.1 per cent, about half the roots are H-toned monosyllables. Another 25.2 per cent are polysyllabic roots with an H tone over the penultimate syllable (of which a mere 1.8 per cent have more than two syllables). Together, these two groups constitute an overwhelming majority of 79.3 per cent of all roots. An additional 15.3 per cent bear an H tone over the final syllable. Most roots in the corpus, namely 94.6 per cent, therefore carry an H tone over the only syllable, the penultimate syllable, or the final syllable.

It should also be mentioned that many of the Spanish items that find their way into codemixed Pichi sentences bear a penultimate H tone in accordance with their original Spanish penultimate syllable stress. This holds in particular for the invariant 3sg present insertion form of the Spanish verb (cf. §13.2.2). Spanish-origin items therefore align with the majority tone classes of Pichi.

#### 3 Suprasegmental phonology


#### Table 3.2 Distribution of tone classes over types

### **3.2 Tonal processes**

Pitch changes conditioned by various factors may take place within a tonal domain. A tonal domain may be confined to the word, cut across a word boundary in specific phonosyntactic phrases, and involve a whole clause or sentence. The tonal processes attested in the data are described in §3.2.1 to §3.2.4. A summary of these processes is given in Table 3.3.

### **3.2.1 Tone spreading**

H tones may spread to right-adjacent L-toned syllables within the word boundary. The H tone over the first syllable of *prɔ́mis* 'promise' in Figure 3.11 spreads to the second syllable.

(8) *Yu* l 2sg *bin* l pst *prɔ́mis* h.l promise *mí* h 1sg.indp *mɔní.* l.h money *→ Yu* l *bin* l *prɔ́mís* **h.h** *mí* h *mɔní.* l.h 'You promised me money.'


Table 3.3 Tonal processes

Figure 3.11 H tone spreading

#### 3 Suprasegmental phonology

An environment that is particularly conducive to rightward tone spreading is when the L-toned syllable of a bisyllabic word with an H.L. pattern is hemmed in by the preceding H tone and the H tone of a following object. In Figure 3.12, the L-toned syllable of *fínis* 'finish' is raised in pitch approximately to the level of the following object *skúl* 'school'. The pitch trace in Figure 3.13 exemplifies the same process with *vɔ́mit* 'vomit' and the following object *chɔ́p* 'food.

Figure 3.12 H tone spreading

Figure 3.13 H tone spreading

(9) *Wé* h sub *a* l 1sg.sbj *go* l pot *fínis* h.l finish *skúl,* h school (*…*) *→ Wé* h *a* l *go* l *fínís* **h.h** *skúl,* **h** (*…*) 'When I finish school, (…)'

3.2 Tonal processes

(10) *E* l 3sg.sbj *de* l ipfv *vɔ́mit* h.l vomit *chɔ́p.* h food *→ E* l *de* l *vɔ́mít* **h.h** *chɔ́p.* **h** 'He is vomiting (the) food.'

A second phono-syntactic environment that favours rightward H tone spreading is a modifier-noun phrase. The L-toned syllable of a bisyllabic property item in prenominal position and with an H.L pattern may be raised to H if it is immediately followed by a noun with an initial (or only) H tone. An example for this process is provided in (16) further below. In the NP, the L-toned syllable of the modifier *fúlis* 'foolish' is raised to an H tone because it is followed by the H-toned noun *mán* 'man'.

### **3.2.2 Floating**

Pichi makes extensive use of floating boundary tones for the purpose of intonation. Aside from that, a lexical tone may be set afloat when two adjoining vowels merge or one of two adjoining vowels is deleted. Tone floating is particularly likely to occur in the contact zone between an H-toned high-frequency function word and a following L-toned vowel. In Figure 3.14, the final consonant /k/ of *mék* 'sbjv' is deleted. This creates a vowel hiatus, which in turn leads to the deletion of the first, higher /e/ of *mék* in favour of the second, lower vowel /à/. The rising-falling contour over *mâ (mék=à)* is clearly visible.

In Figure 3.15, the final segment of *háw* 'how' is deleted and the lexical H tone is set afloat. The vowel merger between /a/ and the following low-toned dependent personal pronoun *e* creates an HL contour tone.

Figure 3.14 Vowel deletion sets tone afloat

#### 3 Suprasegmental phonology

Figure 3.15 Vowel merger sets tone afloat


### **3.2.3 Downdrift and downstep**

Downdrift and downstep contribute to a general downward cline of pitch in utterances. An utterance normally begins with a high pitch onset and declines progressively with every lexical tone. Downdrift (indicated by ↓H) causes an H to be lowered by a preceding L tone as in Figure 3.16.The overall effect of downdrift is visible by the roughly equivalent pitch over the initial L-toned personal pronoun *a* '1sg.sbj' and the final H-toned noun *hós* 'house'.

(13) *A* l 1sg.sbj *mítɔp=an* h.h=l meet=3sg *yɛ́stadé* ↓**h**.l.↓**h** yesterday *na* l loc *in* l 3sg.poss *hós.* ↓**h** house 'I met him yesterday in his house.'

The second phenomenon involving declination is downstep (indicated by –H). In a series of adjacent H tones, each tone may be lowered successively in relation to the preceding one. Downstep is exemplified by the two successive homophones in Figure 3.17 and the iteration in Figure 3.18. We also find downdrift in both examples.

Figure 3.16 Downdrift

Figure 3.17 Downstep

Figure 3.18 Downstep

3 Suprasegmental phonology


### **3.2.4 Deletion**

Tone deletion occurs in two contexts. In compounds (including reduplications), the lexical H tone over the first component is deleted (also see Yakpo 2012a). The syllable whose tone has been deleted becomes L-toned. The second component retains its original tone pattern. Tone deletion therefore forms an intrinsic part of a derivational process in Pichi (cf. §4.3). The second context in which tone deletion occurs is when a boundary tone overrides the utterance-final lexical tone of a word (cf. §3.4.4).

Figure 3.19 presents the pitch trace of an NP headed by the noun *mán* 'man'. The noun is modified prenominally by the verb *fúlis* '(be) foolish', which has an H.L tone pattern. The pitch of the utterance-final H tone over *mán* stands at roughly the same level (albeit slightly downstepped and falling due to declarative intonation) as that of the preceding H tones over the first and second syllables of *fúlis*. Note that the second, lexically L-toned syllable of *fúlis* bears a phonetic H tone due to tonal plateauing (cf. §3.2.1).

Figure 3.19 Simplex noun

#### 3.2 Tonal processes

Figure 3.20 Compound noun


In contrast, the pitch trace in Figure 3.20 above exemplifies tone deletion. The head noun *mán* 'man' is also modified by a verb with an H.L pattern, namely *máred* 'marry, be married'. However, *máred* and *mán* form a single phonological word, the compound noun *mared-mán* 'married man'. The H tone over the first syllable of *máred* has been deleted in the process and replaced by L (the downward cline over the first syllable is caused by a pitch reset at the beginning of the utterance). At the same time, *mán*, the final component of the compound, retains its H tone (which falls slightly due to its utterancefinal position).

Reduplicated verbs show the same suprasegmental characteristics as compound nouns. The pitch trace of the reduplicated (and sentence-medial) monosyllabic *rɔ́n* 'run' in Figure 3.21 shows an L.H pitch configuration over the two identical components. This parallels the pitch trace over the compound *wach-mán* 'watchman' above. Reduplication therefore involves the same derivational process as compounding. The lexical H-tone over the first component is deleted and replaced by an L tone.

#### 3 Suprasegmental phonology

Figure 3.21 Monosyllabic reduplicated verb

Figure 3.22 Bisyllabic reduplicated verb

	- '(…) this running around (…)'
	- '(…) now, (it was) constant shouting.'

3.2 Tonal processes

### **3.2.5 Pitch range expansion**

In Pichi, certain phonetic features may increase the prominence of a (series of) syllable(s). Segments may be lengthened or may be pronounced with increased volume, they may be pronounced with a breathy or creaky voice, and the speech rate may be slowed down or accelerated for stylistic effect. But there is no stress in Pichi in the sense of an automatic, metrically conditioned culmination of phonetic features as in intonation-only languages. Nor does Pichi make use of intonational melodies spanning the entire (or parts of the) utterance for the realisation of pragmatic functions, since these would override the lexical tone of individual words. Instead, pitch range expansion, and an extra-high tone in particular, are exploited to signal focus and emphasis. Focused or emphasised constituents may bear a higher than usual pitch, an extra-high tone on their H-toned syllable(s). The extra-high tone may spread rightwards onto following L-toned syllables until the word boundary is reached (cf. §3.2.1).

Figure 3.23 features the clefted verb *drɔ́ngo* 'be dead drunk'. In the pitch trace, the emphatic character of the predicate cleft construction is evident in two ways. The Htoned syllable of *drɔ́ngo* bears an extra-high tone, and the segment /r/ is lengthened for emphasis. The utterance in Figure 3.23 shades off into a chuckle from the fifth syllable onwards, which produces a wavering pitch trace:

Figure 3.23 Predicate cleft and extra-high tone for emphasis

(20) *Na* l foc *[drrrɔ́ngò]* **+h**.l be.dead.drunk *yu* l 2sg *dɔ́n* h prf *drɔ́ngo.* h.l be.dead.drunk 'You're absolutely dead drunk.'

Elements that fulfil central functions in pragmatically marked contexts are particularly common with extra-high tone, e.g. question elements like *háw* 'how', *wétin* 'what', *údat* 'who', *ús=tín* 'what', the negator *nó*, modifications of degree via repetition like *bíg bíg*

#### 3 Suprasegmental phonology

'very big', and the degree adverb *bád* 'bad, extremely'. Both components of the repetition *bíg bíg* 'be very big' in Figure 3.24 carry an extra-high tone. There is no sign of downstep within the reduplicated sequence:

Figure 3.24 Extra-high tone


'They built him a huge house.'

Entire clauses or sentences may also be placed under focus by (a series of) extra-high tones, which thereby (cumulatively) fulfil(s) the same function as emphatic intonation covered in §3.4.2 further below. There are two principal means of emphasising sentences, which are often used together. The last H tone of the utterance may be raised to an extrahigh pitch as in Figure 3.25. Here the H tone of the utterance-final word *mán* 'man' has been raised to an extra-high level. The sentence nonetheless bears declarative intonation. The word *mán* still exhibits the utterance-final fall characteristic of declarative intonation (cf. §3.4.1) but at a significantly higher pitch level than in a non-emphatic context (22).

(22) *Yu* l 2sg *húman* h.l woman *de* l ipfv *mék* h make *jɔmba* l.l lover *wet* l with *mi* l 1sg.poss *mán.* **+h**l% man. 'Your wife makes love with my husband.'

Secondly, the use of an utterance-final extra-high tone is often accompanied by "pitch range expansion" (Yip 2002: 276). Alternatively, pitch range expansion may be accompanied by the use of the emphatic boundary tone instead of the utterance-final extra-high tone (cf. §3.4.2). During pitch range expansion, the pitch range between H and L tones

#### 3.2 Tonal processes

Figure 3.25 Utterance-final extra-high tone for emphasis

is widened throughout the entire utterance by pronouncing H tones with a higher-thanusual pitch and, optionally, L tones with a lower-than-usual pitch. This creates a strongly undulating pitch contour over the entire utterance.

Figure 3.26 graphically depicts the dramatic rises and falls that may characterise pitch range expansion. The female speaker begins with an L-toned *na* at 190 Hz, rises to 490 Hz with H-toned *só*, then falls to an all-time low with *dɛn* at 145 Hz, until the pitch range gradually evens out towards the end of the utterance (23).

Figure 3.26 Pitch range expansion for emphasis

3 Suprasegmental phonology

(23) *Na* l foc *só* **+h** like.that *dɛn* **+l** 3pl *de* **+l** ipfv *tɔ́k=an.* **+h**=lh% talk=3sg.obj 'That's how they say it.'

### **3.3 Tone-conditioned suppletive allomorphy**

Pichi features a tone-conditioned suppletive allomorphy (TCSA) of the two pronominal variants *=an* '3sg.obj' and *ín* '3sg.indp', which may both instantiate (direct and indirect) object case (cf. §5.4.1 for an overview of the inflection of personal pronouns). Suppletive allomorphy is conditioned by a tonotactic prohibition of immediately adjoining or "string-adjacent" (Suzuki 1998) identical tones (cf. also §2.6.2.2). Suppletive allomorphy therefore relies on the conditioning environment of vowel hiatus. Further, there are no phonemic long vowels in Pichi. String-adjacent vowels within the same lexical word are always heterosyllabic, and in addition, invariably carry polar tones (cf. §2.6.2.2). TCSA can therefore only be triggered when the enclisis of *=an* '3sg.obj' creates a phonological word. A head with an L-toned vowel-final syllable may therefore not take the vowelinitial L-toned clitic object pronoun *=an*. Instead, the independent (emphatic) personal pronoun *ín* '3sg.indp' is recruited as a suppletive allomorph. Allomorph distribution according to the phonological class of the host is summarised in Table 3.4.


There is no tonotactic restriction on the enclisis of *=an* with consonant-final hosts like *máred* 'marry', since the condition of tonal string-adjacency is not met (24).

(24) *E* 3sg.sbj *go* pot *máred=an.* marry=3sg.obj 'S/he'll marry him/her.'

There are no restrictions on the enclisis of *=an* with vowel-final hosts with a word-final H-tone like *trowé* 'throw, pour away', since the vowel sequence across the morpheme boundary bears a polar [H.L] tone:

(25) *A* 1sg.sbj *fít* can *ték* take *di* def *wɔtá* water *a* 1sg.sbj *trowé=an.* throw=3sg.obj 'I can take the water (and) pour it away.'

#### 3.3 Tone-conditioned suppletive allomorphy

If the word-final vowel of the host is L-toned, as with *fíba* 'resemble', the pitch configuration after enclisis of *=an* across the clitic boundary would be [L.L]. This is an illicit pitch configuration over string-adjacent vowels in Pichi phonological words and triggers the use of suppletive *ín* '3sg.indp'. Compare the following two examples:


The class of words that features the allomorph *ín* as an object pronoun also includes verbs of Spanish origin. Spanish verbs are always inserted into Pichi clauses in the Spanish 3sg present tense form, irrespective of their tense-aspect (cf. §13.2.2). Examples follow with the verbs *fírma* 'sign' (< Span. *firmar*) from the Spanish 1st conjugation class, and *sube* 'go/bring up' (< Span. *subir*) from the 3rd conjugation class:


Pichi has a second mechanism next to tone-conditioned suppletive allomorphy to ensure that the requirement of a string-adjacent polar [H.L] tone is not breached. A buffer consonant /r/ can be inserted at the clitic boundary. Epenthesis forestalls the crossmorphemic vowel hiatus and makes the use of the allomorph *ín* unnecessary:

(30) *Yu* 2sg *fíba[r]=an* resemble=3sg.obj *bɔkú.* a.lot 'You resemble him a lot.'

Once the epenthetic segment is present, there is no phonotactic difference with a word in which the final consonant forms an integral part of the root like *máred* 'marry' in (24). Another example featuring epenthesis follows, involving the general associative preposition *fɔ* 'prep'. In (31), we find /r/ epenthesis, in (32), suppletive allomorphy:

(31) *E* 3sg.sbj *tót=an* carry=3sg.obj *fɔ[r]=an.* prep=3sg.obj 'He carried it for her.'

3 Suprasegmental phonology

(32) *Dán* that *tín* thing *dé* be.loc *fɔ* prep *ín.* 3sg.indp 'That thing is hers.'

Three aspects are noteworthy with respect to /r/ epenthesis in Pichi. Firstly, /r/ insertion is exceedingly rare in natural discourse. In the Pichi corpus, there are less than a dozen instances of /r/ epenthesis in natural discourse, involving a mere handful of lexemes, among them *kɔ́ba[r]=an* 'cover it', *klía[r]=an* 'clear it', *fía[r]=an* 'fear him/her', *fíba[r]=an* 'resemble him/her, *drɔ́ngo[r]=an* 'get him/her drunk', and *fɔ[r]=an* 'for him/her'. By contrast, the corpus contains hundreds of syntagmas involving the suppletive allomorph *ín.* I could therefore only uncover the distribution of the epenthetic /r/ and its role in TCSA by means of elicitation. Secondly, elicitation revealed that the availability of /r/ epenthesis is subject to considerable idiolectal variation. For some speakers, the use of epenthesis with many verbs is not acceptable, i.e. \**fála[r]=an* 'follow him/her', for others it is. All speakers, however, accepted TCSA with all verbs and prepositions, whether belonging to the native Pichi or the non-native Spanish lexical layer.

The third aspect of interest is that /r/ epenthesis is ungrammatical with Spanish derived verbs, cf. (33). Epenthesis is limited to the native layer of the Pichi vocabulary, thus excluding inserted Spanish verbs from the application of /r/ epenthesis, and limiting them to TCSA alone, hence the ungrammaticality of the following example.

(33) \* *Yu* 2sg *gɛ́t* get *fɔ* prep *fírma[r]=an.* sign=3sg.obj Intended: 'You have to sign it.'

Pichi words with a word-final L-toned /ì/, e.g. *wɔ́ri* 'worry', merit some attention in the context of epenthesis. Such words exhibit the conditioning feature but neither trigger /r/ epenthesis nor TCSA, compare the ungrammatical sentences (34) and (35). Other verbs in this group are *sɔ́ri* 'feel sorry', *grídi* 'be greedy', *hángri* 'be hungry', *lési* 'be lazy', and *tɔ́sti* 'be thirsty'.


Instead, a word-final nasal /n/ appears at the clitic boundary, thus avoiding the LL vowel hiatus that should trigger suppletive allomorphy, as in (36):

(36) *Di* def *tín* thing *sɔ́rin=an* make.sorry=3sg.obj *bɔkú.* much 'This made her feel very sorry.'

3.4 Intonation

Outside of the clitic environment, the wordfinal /ì/ in these words may, but need not be pronounced as a nasalised vowel, as shown in the phonetic transcription in (37):

(37) *A* 1sg.sbj *sɔ́ri* feel.sorry *[sɔ́rĩ] sé ̀* qot *e* 3sg.sbj *kíl* kill *di* def *dɔ́g.* dog 'I felt sorry that she killed the dog.'

The word-final /n/ in examples like (37) is therefore not epenthetic. It is morphologically affiliated to the verbal root and is realised in the clitic environment. The word-final /n/ in verbs like *sɔ́ri* (group 1) has been constructed by analogy with words like *físin* '(to) fish', *hɔ́ntin* '(to) hunt', *mɔ́nin* 'morning', *ívin* 'evening', and *plantí* 'plantain' (group 2). The construction of a word-final /n/ in group 1 words probably occurred in response to the ban on string-adjacent identical tones in the context of cliticisation.

### **3.4 Intonation**

The functions of intonation are realised by sentence-final particles and utterance-final boundary tones. Pichi boundary tones are floating tones, which are inserted at the right edge of an utterance. These boundary tones serve pragmatic functions by differentiating sentence types, such as declaratives from questions. They also fulfil grammatical functions by linking clauses.

Four boundary tones and contours, represented by <%> (Pierrehumbert 1980), were identified in the corpus. Their functions with declaratives and questions are summarised in Table 3.5 (cf. Hirst & Di Cristo 1998: 18–20).


Table 3.5 Utterance type and boundary tones

A boundary (contour) tone (henceforth only "boundary tone") associates with the last syllable of an utterance. A boundary tone (BT) may either form a contour by itself (e.g. question intonation) or arise if the lexical tone (LT) of the utterance-final syllable is polar to the following BT. Otherwise, a BT produces a fall or a level tone over the utterancefinal syllable.

Table 3.6 shows how LTs and BTs interact. The leftmost column contains the wordfinal LT over the last syllable of the utterance. The top row contains the relevant BT. The

#### 3 Suprasegmental phonology

boxes in the table contain the (contour) tones over the utterance-final syllable that result from the interaction of LT and BT. These tones represent the phonetic output, the way the tone is actually pronounced. Some of these output tones are level tones, others are contour tones of varying complexity.

LT/BT Example Declarative L% Emphatic LH% Cont./Emph. ∅% Question LH% L *dɛn* '3pl' *Píchi* 'Pichi' L H.L L (fall) LH L (level) LH H *gó* 'go' *pikín* 'child' H L.H HL HLH H LH H *bɔbí* 'breast' L.H H HLH H LH

Table 3.6 Interaction of lexical tones and boundary tones

LTs are not overridden by BTs save in one instance. In yes-no questions, the utterancefinal LT is deleted and replaced by the question boundary contour tone. This is why the rightmost column in Table 3.6 features the same LH% boundary tone in the utterancefinal position with all tone classes.

### **3.4.1 Declarative intonation**

Non-emphatic declaratives feature an L%, which is also found on the right edge of the citation form of words. The declarative L% causes an utterance-final fall to the bottom of the pitch register. Compare the word-final L-toned syllable of *kɔ́ntri* 'country' in Figure 3.27.

(38) *A* l 1sg.sbj *bin* l pst *wánt* h want *kɔmɔ́t* l.h go.out *na* l loc *dís* h this *kɔ́ntri.* h.l**l%** country 'I wanted to leave this country.'

In contrast, polysyllabic vowel-final words with a final lexical H tone do not usually feature an utterance-final fall in non-emphatic declaratives. They retain their word-final H tone. Compare *bɔbí* 'breast' in Figure 3.28.

(39) *A* l 1sg.sbj *de* l ipfv *gí=an* h=l give *bɔbí.* l.**h** =3sg.obj 'I'm breast-feeding her.'

Figure 3.27 Declarative L% over H.L word

Figure 3.28 Unpronounced declarative L% over L.H word

Content questions feature the same boundary tone as declaratives. Compare the utterancefinal fall over the monosyllable in Figure 3.29.

(40) *Wétin* h.l what *mék* h make *dán* h that *wán?* **hl%** one 'What causes this?'

#### 3 Suprasegmental phonology

Figure 3.29 L% with content question

### **3.4.2 Emphatic intonation**

Emphatic intonation expresses meanings like extra-emphasis, insistence, impatience or reproach. There are two ways of signalling emphasis at the sentence level in Pichi. One way involves the use of the emphatic LH% boundary tone. A second way involves the use of pitch range expansion (cf. §3.2.5).

The emphatic LH% is an additive contour tone. It succeeds the lexical tone of the utterance-final syllable, which may therefore count up to three beats in length. Additionally, the last lexical H before the LH% boundary contour tone is often pronounced with an extra-high tone due to emphasis. This peculiar combination of an extra-high lexical tone and a contour boundary tone creates a highly perceptible utterance-final tonal melody.

Phonemically, an utterance-final L to which the emphatic LH% boundary tone associates bears an LHH sequence of tones. Phonetically, the utterance-final syllable is realised as an LH contour. Figure 3.30 depicts the utterance-final rise over the L-toned monosyllable *=an* '3sg.obj'.

```
(41) E
      l
      3sg.sbj
               de
               l
               ipfv
                    lé=an.
                    h=lh%
                    lay=3sg.obj
      'She is laying it (on the table).'
```
When the emphatic boundary tone links with an utterance-final H-toned syllable the resulting contour features an initial rise, an intermediate fall, and a final rise. The utterance-final, extensively lengthened syllable thus bears an HLH contour. Compare the utterance-final H-toned monosyllables *ín* '3sg.indp' and *gó* 'go' in Figure 3.31 and Figure 3.32.

Figure 3.30 Emphatic LH% over L-final word

Figure 3.31 Emphatic LH% over H-final word

Figure 3.32 Emphatic LH% over H-final word

3 Suprasegmental phonology

(42) *Na* l foc *ín.* **hlh%** 3sg.indp 'That's it [you should know that].'

(43) *A* l 1sg.sbj *go* l pot *gó.* **hlh%** go 'I'll go [you don't need to remind me to].'

An utterance-final, H-toned syllable of a polysyllabic word also bears this contour. Compare *bɔbí* 'breast' and *chukchúk* 'thorn' in Figure 3.33 and Figure 3.34.The two words were pronounced with emphatic intonation during vocabulary elicitation because the speaker expected me to be familiar with them.

Figure 3.33 H% over vowel-final L.H word

Figure 3.34 H% over obstruent-final L.H word

3.4 Intonation

(44) *Bɔbí.* l.**hlh%** breast 'Breast [that's self-evident!].'

(45) *Na* l foc *chukchúk.* l.**hlh%** thorn 'It's a thorn [that's self-evident].'

The LH% boundary contour tone is a loan from (Equatoguinean and, ultimately, European) Spanish together with the meanings associated with it. The LH% contour boundary tone is also employed for list intonation (cf. §3.4.3). Figure 3.35 presents the pitch trace of an utterance in Equatoguinean Spanish.

Compare the contour over the utterance-final L-toned syllable with that borne by the utterance-final L-toned syllable in Figure 3.30. Also compare the emphatic contour over the phonologically independent *sí* 'yes' with that of the high-toned *ín* '3sg.indp' in Figure 3.31.

Figure 3.35 Emphatic intonation in European Spanish

(46) *Sí* h**lh%** yes *vengo.* h.**lh%** I.come 'Yes [you should know that!], I'll come.'

### **3.4.3 List intonation**

The additive LH% boundary tone employed for emphatic intonation is also used for list intonation. As in emphatic declaratives, LH% associates with the final syllable and cre-

#### 3 Suprasegmental phonology

ates an LH contour over an utterance-final L-toned syllable and an HLH contour over an utterance-final H-toned syllable. The same intonation contour is once more found in Equatoguinean (and European) Spanish with a similar range of meanings.

The following three pitch traces form part of a list. Take note of the LH contour over the L-toned dependent pronoun *dɛn* '3pl' before the short pause, as well as the LH contour borne by the L-toned final syllable of *manicura* 'manicure' in Figure 3.36 and *chía* 'chair' in Figure 3.37. Compare this with the declarative L% over *dé* 'there', the closing sentence of the list in Figure 3.38.

	- '(…) so I was getting my money there.'

Figure 3.36 List intonation

Figure 3.38 Declarative L% over final item in list

### **3.4.4 Continuative intonation**

The absence of a boundary tone, usually before a prosodic break (a brief but audible pause), signals continuative intonation. With continuative intonation, the lexical tone of the relevant syllable simply maintains its pitch and is therefore pronounced with the same pitch as it would in utterance-medial position. Continuative intonation functions as a floor-holding device, a juncture marker on the right edge of utterances in order to prepare the ground for following material. Continuative intonation therefore plays an important role in signalling topic and focus next to the particles employed for this purpose (cf. §7.4).

In Figure 3.39, the topical NP *mi láyf* 'my life' is set off from the rest of the utterance by a pause. The monosyllable *láyf* 'life' bears continuative intonation. Compare this to the

#### 3 Suprasegmental phonology

utterance-final monosyllable *bád* 'bad', which features declarative intonation, signalled by L%. The symbol [p] indicates a pause. The pitch trace of the pronoun *e* '3sg.sbj' is slighty distorted due to creaky voice.

Figure 3.39 Continuative intonation with topicalisation


Continuative intonation is also employed as a juncture marker between linked clauses. Here, it may occur alone as a prosodic clause linker between juxtaposed clauses, or in conjunction with an overt clause linker. Figure 3.40 and Figure 3.41 are two clauses linked in a sequential, temporal relation. The adverbial time clause is introduced by *di dé wé* '(the day) when' in Figure 3.40. In the example, continuative intonation is found over the rightmost L-toned monosyllable *=an* '3sg.obj'. The absence of the utterance-final L% of declarative intonation halts the fall of the lexical L tone to the bottom of the pitch register.

(51) *Di* l def *dé* h day *wé* h sub *yu* l 2sg *go* l pot *wánt* h want *plánt=an,* h=**l**∅**%** plant=3sg.obj (*…*) 'The day you would want to go plant it (…)'

The second clause in sequence features a lexical H over the utterance-final syllable. Here, continuative intonation produces no effect other than the maintenance of the lexical H tone. Compare *dɔtalɔ́* 'daughter-in-law' and *sɔnilɔ́* 'son-in-law' in Figure 3.41.

Figure 3.40 Continuative intonation with clause linkage

Figure 3.41 Continuative intonation over non-final clause


Continuative intonation is also used as a stylistic device in 'unfinished' utterances, such as the one in Figure 3.42. The final syllable retains its H tone or may even rise slightly towards the end. This emphatic variant of declarative intonation is employed for dramatic effect. Compare the utterance-final, H-toned monosyllable *dé* 'there'.

#### 3 Suprasegmental phonology

Figure 3.42 Continuative intonation for stylistic effect

(53) *Sɔn* L some *hángri* h.l be.hungry *kéch* h catch *mí* h 1sg.indp *dé.* h∅**%** there

'I became really hungry there [you wouldn't believe how much].'

### **3.4.5 Question intonation**

Yes-no questions are formed with an LH% contour boundary tone. Contrary to emphatic intonation, question intonation is substitutive: The lexical tone over the utterance-final syllable is replaced by the question LH%. In this way, the utterance-final syllable of a yesno question invariably bears an LH contour, irrespective of its original tone. Compare the pitch contour over the L-toned second syllable of *Píchi* 'Pichi' in Figure 3.43.

(54) *Yu* l 2sg *go* l pot *lán* h teach *mí* h 1sg.indp *Píchi?* h.**lh%** Pichi 'Will you teach me Pichi?'

The H tone of the LH% contour may vary in pitch. While non-emphatic questions exhibit a gentle final rise and may therefore be similar in pitch to continuative intonation, more emphatic questions yield steeper rises. The more dramatic the rise, the more the question may additionally convey emphatic nuances like counter-expectation or insistence. I assume that in instances where the rise is particularly steep, the H tone component of the LH% boundary contour tone is raised to extra-high, thus rendering L+H%. Such an extra-steep rise is particularly common in rhetorical questions, optionally over the L-toned utterance-final question tag *nɔ́* as in the following example.

Figure 3.43 Non-emphatic yes-no question

Figure 3.44 Emphatic yes-no question

(55) *Una* l.l *gɛ́fɔ* h.l *píl* h *dɛ́n* l *nɔ́?* **l+h%**

> 2pl have.to peel 3pl.indp intj

'You [pl] have to peel them, right [you should know that]?'

The utterance-final syllable in the question above exhibits a particularly steep rise. At the same time, emphasis is additionally expressed through pitch range expansion. The contrast between H and L tones is widened across the entire utterance as can be seen by the deep troughs in the pitch trace.

## **4 Morphology**

Pichi nouns and verbs constitute two major word classes. Adjectives, prepositions, and adverbs constitute minor word classes with a few members each. Pichi word formation strategies are predominantly analytic. Besides that, the use of one (adverb-deriving) affix and morphological tone play a role in Pichi derivation and inflection.

### **4.1 Word classes**

Pichi word classes are differentiated by their syntactic functions (e.g. a noun may head an NP), distribution within the sentence (e.g. a preposition may not be preceded by an article), the morphosyntactic categories that may be specified for them (e.g. verbs may be specified for tense, aspect, and modality), their derivational potential (e.g. personal pronouns and prepositions are not normally reduplicated, and adverbs do not function as nouns), as well as semantic criteria (dynamic states-of-affairs and property concepts are generally expressed as verbs).

The major underived word classes, with the most members and the potential to occur in the largest range of environments, are nouns and verbs. The noun-verb distinction in Pichi is quite strong: although verbs may function as nouns in specific (e.g. in emphatic) contexts, the reverse is not usually the case. The verb-adjective distinction is weak. There are only a handful of adjectives, which are indistinguishable from verbs in most environments. The minor word classes consist of adjectives, prepositions, adverbs, as well as various sentence elements that contribute to the meaning of the sentence.

### **4.1.1 Nominals**

Nouns appear as one of up to three core participants of a verb, i.e. as subjects or up to two objects. Nouns also occur as objects of prepositions, and they may function as adverbials. They may be modified by other elements of the noun phrase (e.g. *di* 'def', *dá(n)* 'that', *sɔn* 'some, a' or *dɛn* 'pl'), including other nouns in associative constructions and compounds. The vast majority of nouns bears a single H tone and belongs to one of the major tone classes (cf. Table 3.2). Underived nouns typically denote time-stable object concepts. Nouns also belong to an open class which may be extended by compounding, conversion, and borrowing from Spanish.

Personal pronouns, pronominals, and compound question words are subsets of nominals that exhibit a more restricted distribution. Personal pronouns are found in the same syntactic positions as noun phrases but do not cooccur with preposed modifiers. The latter usually also holds true for the pronominals *nátin* 'nothing', *sɛ́f* 'self', and *yón* 'own'.

#### 4 Morphology

The pronominals *káyn* 'kind' and *wán* 'one' have a wider distribution but are also characterised by specific syntactic preferences. Locative nouns form a further subclass of nominals characterised by distributional specificities. Locative nouns are not often preceded by modifiers or determiners, and their distribution overlaps with that of prepositions.

### **4.1.2 Verbs and adjectives**

Verbs occupy the centre of the predicate. The predicate is best seen to include a number of functional elements that form a tightly-knit unit with the verb in order to constitute clauses: TMA markers, preverbal adverbs, the negator, dependent personal pronouns, as well as the clitic 3sg object pronoun. Verbs are usually preceded by a subject noun, pronoun, or both. Verbs may optionally be followed by objects. They are typically monoor bisyllabic and usually belong to one of the three major tone classes.

There are numerous subclasses of verbs which can be defined along formal and semantic lines: Aspectual and modal verbs, transfer and communication verbs, stative, inchoative-stative, and dynamic verbs, labile verbs, and copula verbs. Other than reduplication, Pichi only has marginally productive means of verb derivation through compounding. There are numerous other strategies for the creation of new verbal meanings, e.g. light verb constructions, involving *gɛ́t* 'get, have', *mék* 'make', or *gí* 'give', as well as systematic borrowing from Spanish.

There is just a handful of adjectives in Pichi. A small set of property items alternates between uses as inchoative-stative verbs and as adjectives (cf. Table 7.11 in §7.6.5). The overwhelming majority of property concepts are lexicalised as inchoative-stative verbs in Pichi. The following "semantic types" (Dixon 2004: 3) are expressed through inchoativestative verbs: Dimension (e.g. *bíg* 'be big', *smɔ́l* 'be small', and *lɔ́n* 'be long'), age (e.g. *ól* 'be old' and *yún* 'be young'), value (e.g. *bád* 'be bad', *fáyn* 'be good', and *trú* 'be true'), and colour (e.g. *blák* 'be black', *wáyt* 'be white', *rɛ́d* 'be red', and *yɛ́lo* 'be yellow'). Most physical properties are also lexicalised as inchoative-stative verbs (e.g. *hád* 'be hard', *sáf* 'be soft', *sók* 'be wet', *évi* 'be heavy', *hɔ́t* 'be hot', *swít* 'be tasty').

Human propensities are divided between inchoative-stative (e.g. *gudhát* 'be goodhearted', *wíkɛd* 'be wicked', *badhát* 'be mean', *klɛ́va* 'be clever') and dynamic verbs (e.g. *gládin* 'be glad', *jɛ́lɔs* 'be envious') according to whether they denote intrinsic or transient properties. Resultatives are exclusively expressed through the stative readings of labile change-of-state verbs (e.g. *brók* 'break, be broken', *chɛ́r* 'tear, be torn', *lɔ́s* 'lose, be lost' and *wɛ́r* 'be dressed'). Semantic types like position or location are expressed through other means, such as copula clauses featuring the locative-existential copula *dé* (cf. e.g. 281–282) in combination with adverbials, or through locative verbs like *lé* 'lie' and *tínap* 'stand (up)' (cf. §8.1.3).

### **4.1.3 Other word classes**

Most prepositions must be followed by an object, although some may be stranded, that is, they may occur in the clause-final position. Prepositional phrases are found in the clauseinitial or -final position. A majority of prepositions is monosyllabic, a few are bisyllabic.

4.2 Inflection

Pichi exhibits a division of labour between prepositions, locative nouns, locative adverbs, and locative verbs in order to express spatial relations. The language has a small number of underived adverbs, amongst them a group of four preverbal adverbs.

Each of the following groups of modifiers may also be said to constitute minor word classes unto themselves, because they occupy distinct syntactic positions in the noun phrase or predicate: the article, demonstratives, quantifiers, prenominal attributive modifiers, numerals, the pluraliser, emphasis markers, topicalisers, TMA markers, aspectual and modal verbs, the general negator, interjections, and ideophones. Certain elements modify sentences in their entirety with respect to pragmatic status (e.g. question words, tags, focus particles, interjections) or link sentences with each other (e.g. clause linkers and conjunctions). These sentential elements may also each be considered a separate word class due to their functions and syntactic behaviour.

### **4.2 Inflection**

Most grammatical functions are realised analytically by independent words without the morphological modification of heads or dependents. Participant-marking is taken care of by prepositions and locative nouns, serial verb constructions, and word order, and nominal modification by juxtaposition of adjectives and other modifiers. Number-marking is achieved by post-nominal modification.

The verbal category of number is signalled by personal pronouns and reduplication. Complementisers, preverbal TMA markers, serial verb constructions, and adverbs participate in expressing the grammatical categories of tense, modality, and aspect. Comparison is expressed by adverbs of degree, ideophones, verbs, phrasal expressions, suprasegmental modification, serial verb constructions, and prepositions. There are, however, exceedingly rare cases of number marking on *gɛ́l/gál* 'girl' and *bɔ́y* 'boy' by an apparently marginal plural affix {-s}, hence *gɛ́l-s*, *bɔ́y-s*.

A description of the only inflectional morphological processes follows. The expression of the grammatical relations of subject, object, and possessive case may be seen to involve the use of (tonal) suprafixation, summed up in Table 4.1 (cf. §5.4.1 for the full pronominal paradigm and examples).

Table 4.1 Suprafixation with personal pronouns


Tone-conditioned suppletive allomorphy also fulfils inflectional functions in Pichi, even if it involves outright substitution rather than morphological modification (cf. §3.3). It has been suggested that the cognate form of the Pichi imperfective marker *de* 'ipfv' be analysed as an inflectional verbal prefix in Jamaican Creole (Farquharson 2007: 30).

#### 4 Morphology

In Pichi too, the use of resumptive imperfective marking with the preverbal aspectual adverbs *jís/jɔ́s* 'just' and *stíl* 'still' suggests a tighter-than-usual syntagmatic relation between the imperfective aspect marker and the verb it modifies:

(1) *Náw* now *dɛn* 3pl *de* ipfv *jís* just *de* ipfv *kán.* come 'Now, they're just coming.' [ye07je 179]

### **4.3 Derivation**

Pichi makes use of morphological processes for the purpose of derivation. One is a tonal process which derives compounds, including reduplications. The other is adverbderiving suffixation. Compounding and reduplication are two highly productive derivational processes in Pichi.

### **4.3.1 Affixation**

Table 4.2 summarises the derivational processes found in Pichi. This section covers formal aspects of compounding and reduplication, which both receive a more detailed functional treatment in §4.4 and §4.5.1, respectively. Adverb-deriving suffixation is covered in this section in both its formal and functional aspects.


Table 4.2 Derivational processes

Compounding and reduplication both make use of the same tonal derivation. Reduplication is therefore best seen as a form of (self-)compounding in Pichi. In the process, the H tone over the initial component(s) is deleted and replaced by an L tone. The final component retains its original tone configuration. The resulting compound word then features a single H tone like most Pichi words. Pichi compounds are therefore rightheaded; the L-toned initial component(s) function as modifier(s) to the final component, which is the head.

#### 4.3 Derivation

Nouns, verbs, adjectives, and adverbs participate in compounds. The resulting structures may function as nouns or verbs. Personal pronouns may also participate as modifiers in compound personal pronouns (cf. §5.4.2). Compounding is fairly productive (cf. §4.4 for details). Compare the compound in (2) featuring the modifier noun *kɔ́ntri* 'country, home town' and the modified noun *chɔ́p* 'food'. While *kɔ́ntri* loses the H tone over its first syllable, the head noun *chɔ́p* retains its original H tone:

(2) *Na* foc *in* 3sg.poss *kɔntri-chɔ́p.* country.cpd-food 'That's his local food.' [au07ec 007]

Compounding through tone deletion also characterises the reduplication of dynamic verbs in order to derive verbal number (3). This kind of derivation is fully productive for all dynamic verbs. Equally, it can be observed with a small number of lexicalised reduplications involving other word classes (cf. §4.5.3):

(3) *Kán* pfv *tót* carry *bɛlɛ́,* belly *bigín* begin *de* ipfv *hala-hála* red.cpd-shout *náw,* now *hala-hála.* red.cpd-shout '(Then she) became pregnant, (and) began lamenting and lamenting.' [ab03ay 118]

Adverbs are derived from verbs and adjectives by means of the suffix -*wán* 'adv', etymologically related to the numeral *wán* 'one'. Amongst its numerous other uses (cf. §5.3.1), the cardinal numeral *wán* 'one' serves as a pronominal substitute for nouns in NPs featuring attributively used property items (i.e. *di blák wán* 'the black one'; *di bíg wán* 'the big one'). When such NPs appear in an adverbial slot in the clause, the resulting structure functions as a manner adverb.

The semantic link between the function of -*wán* 'adv' as an adverbialising suffix and the meaning of *wán* in other contexts is opaque. This warrants the analysis of -*wán* 'adv' as a suffix rather than seeing it as the second component of a compound word. The derivation of adverbs is a derivational process distinct from compounding and does not involve the tone deletion that accompanies the latter kind of word formation. In the following examples, the property items *fáyn* '(be) fine' (4) and *smɔ́l* '(be) small' (5) and the affix *-wán* retain their lexically assigned H tone. The resulting adverbs are bisyllabic, bimorphemic words with an H-H (downstepped H) tone configuration:


The derivation of manner adverbials through the suffixation of -*wán* is not particularly productive. In the corpus, it is unanimously accepted with a limited number of

#### 4 Morphology

monosyllabic property items denoting physical properties, such as *smɔ́l* 'be small', *kól* 'be cold', *hɔ́t* 'be hot', *fáyn* 'be fine'. In contrast, the formation of adverbials with many other property items was rejected by informants, amongst them *dɔtí* 'be dirty', *bád* 'be bad', *bɛlfúl* 'be satiated', *nékɛd* 'be naked', *táya* 'be tired', *lét* 'be late', *frɛ́s* 'be fresh', *rɛ́p* 'be ripe', and *sáful* 'be slow, diligent'.

The generic noun *tɛ́n* occurs in a small number of more or less lexicalised expressions functioning as sentence and time adverbs. All of the expressions contained in the corpus are listed in (6). Like derived adverbs featuring the suffix -*wán* 'adv', these bisyllabic expressions are not compounds, since there is no tonal derivation.

The meanings of these expressions are semantically distinct from the meanings of their components in varying degrees. The degree of semantic opaqueness of each collocation is reflected in the orthographic choice of writing them as single or separate words. A good indicator of the degree of semantic unity of the collocations in (6) is their behaviour during repetition for emphasis (cf. 55 further below). Even in the lexicalised expressions (e.g. *bádtɛn* 'unfortunately) each morpheme nevertheless retains its original pitch, as shown by tone marking. This renders complex words with a sequence of two H tones (the second H undergoes downstep).


The largely unpredictable meanings of the adverbs in (6) are reason enough to consider them as lexicalised phrasal expressions, rather than analysing *tɛ́n* as a productive adverbialising suffix.

### **4.3.2 Conversion**

Some word classes are characterised by multifunctionality. They may undergo conversion and appear in a syntactic position reserved for another class without morphological derivation. Table 4.3 provides an overview of productive conversion. Some processes are unidirectional, others bidirectional. Arrows indicate the direction of conversion. The productivity of conversion varies with word class and is often subject to lexical idiosyncracies.

Verbs may be employed in the syntactic position of nouns. This process of conversion is very productive. The meanings of such nominalised verbs vary in accordance with their lexical aspect. A dynamic verb used as a noun denotes the nominalised activity,

4.3 Derivation


Table 4.3 Conversion

while an (inchoative-)stative verb used in such a way denotes the corresponding nominalised state.

In (7), the dynamic verb *hála* 'shout' is used as a dynamic noun or "action nominal", (Comrie & Thompson 1985). In (8), the inchoative-stative verb *gúd* 'be good' is employed as a stative noun or "state nominal" (Comrie & Thompson 1985). The use of nominalised verbs as cognate objects is common for emphasis (cf. §9.3.3). Cognate objects behave no differently from other nominalised verbs:


'Then through that goodness that I treated you with, that's where my husband would think that my anger has finished.' [ro05rr 003]

A verb can also appear in the nominal position together with its object, although this is rarely heard in natural speech:

(9) *Na* foc *di* def *wás* wash *klós,* clothing *na* foc *di* def *tín* thing *mék* make *yu* 2sg *táya.* be.tired 'It's the washing of clothing, that's why you're tired.' [dj05be 039]

In contrast, very few nouns are attested in the syntactic position of verbs. The noun *bɛlɛ́* 'belly' (10) may be used as a verb with the meaning 'impregnate' (11). Other nounverb pairs in the corpus that may be employed in a similar way are *kaká* 'defecate, faeces', *pipí* 'urinate, urine', *rút* 'root, uproot', *latrín* 'toilet, go to toilet'. These rare cases are not listed in Table 4.3 because they are lexicalised, and there is hence no productive nounverb conversion.

#### 4 Morphology


Other word classes are also characterised by multifunctionality. Members of the small adjective class of Pichi may be used as inchoative-stative verbs without a change in form (cf. §7.6.5). Property items, whether adjectives or verbs, may be employed as attributive adjectives (i.e. prenominal modifiers, cf. §5.2.1), and nouns may modify other nouns in associative constructions without an overt process of derivation (cf. §4.4.2). Further, labile verbs may be used in their respective lexical aspect classes without any formal change (cf. §9.2.3). Such multifunctionality with respect to lexical aspect class is very productive. It is lexically restricted to the class of labile verbs, which constitutes a large verb class in Pichi. Aside from that, members of the small class of adverbs are not usually employed as nouns or verbs.

### **4.4 Compounding**

Pichi makes extensive use of compounding in order to derive nouns, verbs, and personal pronouns. Compound words are formed by combining two, sometimes more lexical items. Most types of compounding are covered in §4.4. Reduplication, which also involves compounding, is covered separately in section §4.5.1. Aspects of the morphophonology of compounding are covered in §3.2.4.

### **4.4.1 General characteristics**

Compounding forms part of a continuum of possessive constructions or relations of modification between constituents (cf. also §5.2.3). I only refer to those possessive constructions as "compounds" which form single phonological words via the tonal derivation described in §3.2.4. I nevertheless use the term "compounding" as a generic term to designate the formative processes that derive compounds associative constructions and *fɔ*-constructions. Compounds relate in interesting ways to associative constructions and *fɔ*-prepositional phrase constructions. The two latter types of possessive constructions are formed by syntactic concatenation alone. In the following, I refer to the individual lexical items occurring in these three types of possessive constructions as "components". Table 4.4 provides an overview of relevant characteristics of the three types of compounding:

Phonological and semantic factors determine the choice between compounding and the use of associative constructions for word formation. Speakers may opt to use a compound when the relevant concepts are commonly associated with each other, and the

4.4 Compounding


Table 4.4 Characteristics of compounding

entire structure is conventionalised or lexicalised. In contemporary Pichi, there is no formal difference between compounds that may have been carried over from English (e.g. *pan-kék* 'pancake', *ren-sísin* 'rain(y) season') and language-internal formations (e.g. *kɔntri-chɔ́p* 'local food'). The meanings of both groups may be more compositional or more idiosyncratic, and both undergo the same tonal derivation characteristic of compounding:


Some semantically opaque compounds also exist, in which one component has no independent meaning (13a) or where one component is obsolete (b). It is noteworthy that the initial components of the first two compounds below exhibit a regular soundmeaning relation with the verbs *spót* 'be stylish' and *lúk* 'look', respectively, although there is no nominalising suffix *\*-in* in Pichi. However, there is one verb-noun pair in the corpus, in which the noun (*bɛ́rin* 'burial') is the action nominal to a verb (*bɛ́r* 'bury'). The compound in (c) is therefore transparent and fully segmentable. Opaque and exocentric compounds are written without a hyphen in this work and their components are separated by a dot where relevant:


Other collocations are also partially opaque but exhibit the prosodic characteristics of either associative constructions or compounds. These are structures that have inherited varying degrees of semantic opacity and lexicalisation from English, cf. (14–16). In the compounds in (14), both components before and after the dot retain their original pitch

#### 4 Morphology

configurations. In collocations involving the generic noun *dé* 'day' as a modified noun, the "modifier" has no meaning of its own:


The structure of two sets of kinship terms is also of interest. The root *gran-* 'grand-' is segmentable and has a discernible meaning. However, the root is never found independently of the word it modifies. It only appears in compounds (15a), which can, in turn, be preceded by the prenominal modifier *grét* 'great' (b):


The second set of kinship-denoting compounds contains the segmentable root *lɔ́*'law' as the final component. The composite meanings of these compounds are idiosyncratic. Additionally, some of the structures are fully segmentable, with the first component constituting an independent word (16a). Further, we find variants of group (a) compounds with slightly altered initial components (b). With these, the etymology is clear, but the altered initial component never occurs on its own. A final group contains an opaque initial element, which is a fossilised English morpheme that does not exist (any longer) in contemporary Pichi (c):


In Spanish compounds and neologisms involving Spanish components (e.g. *busca-blanco* 'female sex worker specialised to white men'), the initial component(s) is/are always low-toned, while the final component bears H tone on the penultimate or only syllable (17). This also holds for reduplicative compounds involving Spanish-derived dynamic verbs. The H tone is therefore found on the syllable that is stressed in standard Spanish. However, when these Spanish-derived compounds are employed in Pichi clauses, the H

#### 4.4 Compounding

tone over the final component may not be shifted to other components of the compound for focus or emphasis (as the placement of stress may be in Spanish). This speaks for an analyisis of these collocations as Pichi-style compounds featuring the tonal derivation that other compounds have:


Although in many cases conventionalisation is a good indicator for the use of compounding, phonology may override semantics. Compounds are shunned in favour of associative constructions where the first component belongs to the L.H tone class featuring a word-final H tone. We have seen that this tone class remains unaffected by other tonal and intonational processes as well (cf. e.g. §3.4.1). Hence the concepts in (18), although conventionalised, are expressed as associative constructions, syntactic phrases consisting of prosodically independent components:


The tonal derivation characteristic of compounding also distinguishes lexicalised compound verbs (19a) from verb-object phrases (b) (cf. also §4.4.3):


### **4.4.2 Compound nouns**

Compound nouns function as nouns in a clause. Their final component is always a noun, while their initial component(s) may be a noun, verb, or an adverb. Compound nouns are the most common type of compound in the corpus. They instantiate a relation of

#### 4 Morphology

modification, with the first component serving as the modifier and the second as the modified element.

In a large number of collocations in the corpus, the modified noun is one of the generic nouns listed in (20), which serve other important functions in the language as well (cf. Faraclas 1996: 252):


The tendencies of nominal compounding are summarised in Table 4.5. The column "modifier/modified" in Table 4.5 lists the types of modification relations attested in the data. I have added the third relevant possessive construction, the "*fɔ*-construction" for comparison. The columns headed by "compound", "associative construction", and "*fɔ*-construction" contain a cross (×) if the structure is employed to express the corresponding relation in the leftmost column. A blank space indicates that the structure is not employed for this purpose.

Compounds, associative constructions, and *fɔ*-prepositional constructions form part of a continuum of "possessive" constructions. In this continuum, associative constructions may express the widest range of modification relations, including most relations that may also be expressed as compounds and *fɔ*-prepositional constructions (cf. also §5.2.3). Table 4.5 shows that compound nouns are only used to express "kind of/entity" relations – the "activity/agent" relation being a subtype of the "kind of/entity" relation in which the first component is a dynamic verb and the second a human-denoting noun.

In turn, associative constructions represent the conventional means of expressing a measurement relation (referred to as "measure/entity" in Table 4.5), a "group/member of" relation featuring the modified noun *pikín* 'child', and a "gender of/creature" relation featuring the gender nouns *mamá* 'mother' and *papá* 'father', *mán* 'man' and *húman* 'woman', or *bɔ́y* 'boy' and *gál* 'girl' in the modifier position.

4.4 Compounding


Table 4.5 Tendencies of nominal compounding

Secondly, associative constructions are the default option for expressing "kind of/ entity" relations when these are not expressed as compounds. One criterion that determines the use of an associative construction as a default option is the nature of the modifier noun. Modifier nouns with an L.H pitch configuration and/or with more than two syllables are less likely to undergo the tone deletion that derives compound nouns. A second, subsidiary criterion is the lack of conventionalisation or lexicalisation of the collocation. In all other cases, "kind of/entity" relations, including "activity/agent" relations are usually expressed through compounds. Nevertheless, allowance must be made for numerous lexicalised exceptions to these tendencies.

In "kind of/entity" compounds, the first component modifies the second as to certain qualities. These compounds encompass bicomponental food items and dishes (21a) and body parts (b), as well as other concepts commonly associated with each other (c). Note that *kaka-rás* 'arse' in (b) is a lexicalised compound and an exception to the tendency for collocations featuring an L.H modifier noun to be realised as associative constructions (the other most common exception being *bɛlɛ́*'belly' when used in the modifier position of a compound, cf. 27). Compounds are also employed to form highly conventionalised quantifier compounds which express ordinal numerals (d) as well as dual and *ɔ́l* 'all' extensions of the pronominal system (e).

In sum, the use of "kind of/entity" compounds therefore reflects the degree of conventionalisation of the linkage between the participating nouns and in that a certain degree of inalienability:


#### 4 Morphology


Certain "kind of/entity" relations follow in (22) that are expressed through associative constructions rather than compounds. Group (a) features collocations, in which the modifier noun belongs to the L.H tone class. Here we also find some highly conventionalised collocations (b). The words in (c) contain associative constructions that involve trisyllabic modifier nouns from different tone classes. Other concepts are not sufficiently conventionalised or lexicalised to appear in compounds even if they present no formal obstacles (d). Also note the "kind of/entity" relations listed in (22):


Other "kind of/entity" relations are also expressed through associative constructions, although they do not present any phonotactic or semantic obstacles either. For example, the generic noun *tɛ́n* 'time' is only recorded as a modified noun in the associative constructions listed in (23), even though these structures are lexicalised and occur very frequently. Note, however, that other, lexicalised collocations involving *tɛ́n* are not expressed as compounds either (cf. 6 above):


Compounds involving *sáy* 'side, place' are equally scarce. This noun is only attested as a modified noun in three compounds in the corpus, all of which have partially idiosyncratic meanings (24a). Other equally conventionalised collocations involving *sáy* are expressed through associative constructions (b) or via *fɔ*-prepositional constructions (c):

4.4 Compounding


"Group/member of" structures feature the human-denoting noun *pikín* 'child' in the modified position. The conventional way of expressing this relation is through the associative construction. The modified noun *pikín* may acquire quite an idiosyncratic meaning in the collocations listed under (25b). In these associative constructions, *pikín* 'child' denotes a typical member of the group specified by the modifier noun rather than a kind of child (cf. Heine et al. 1991: 91–97). For example, the construction *Guinea pikín* is best translated as 'person of Equatoguinean stock, typically Equatoguinean person':


"Gender of/creature" structures in which the modifier noun specifies the gender of a modified noun are also expressed as associative constructions. Compare the following collocations involving nouns with diverse pitch configurations:


The human-denoting nouns *mán* 'man, person', *húman* 'woman', *pípul* 'people', and *pɔ́sin* 'person' usually appear as modified nouns in compounds only (27). The list also contains two compounds featuring *bɛlɛ́* 'belly' as a modifier noun. *Bɛlɛ́* and *kaká* 'faeces' are the only attested nouns with an L.H pattern that are subjected to the tonal derivation characteristic of compounding. In the two compounds, the H tone over *bɛlɛ́* has been deleted:

#### 4 Morphology


The noun *mán* 'man' is encountered in "activity/agent" compounds in which the first component is a dynamic verb with *mán* instantiating the agent or "doer". Such compounds are a subtype of the "kind of/entity" type of compound and serve to form agentive nouns as in the examples provided in (28):


Certain compounds involving *mán* 'man' are neutral in their gender reference (29a) and equivalent to the far less common *pɔ́sin* 'person' (b) in "activity/agent" compounds. However, *mán* is also employed with the meaning 'person' in other contexts (e.g. *na mán* 'foc man' = 'that's a human being'). Hence the gender-neutral use of *mán* is not necessarily an indication of the generalisation of its function. In fact, *húman* 'woman' always occurs as the "doer" when a female reference is desired (c) (cf. also *mákit-húman* 'market woman' in 27 above). The generic noun *mán* 'man' therefore falls short of functioning as an agentive suffix, in spite of its general, gender-neutral meaning in some contexts:


### **4.4.3 Compound verbs**

Three types of compounds may function as verbs in a clause: verb-verb reduplications, adverb-verb degree compounds, and verb-noun property compounds. The latter two are treated in this section; reduplication is extensively covered in section §4.5.1.

4.4 Compounding

A verb may appear as the head of a compound featuring the multifunctional word *óva* 'over, be excessive, too much' as the first component. The resulting compound verb expresses an excessive degree of the situation denoted by the verb. It is therefore normally formed with verbs denoting properties, such as *dráy* 'be dry, lean' (30), or verbs whose meaning contains an implicit gradation, such as *dríng* 'drink (alcohol)' (31).

Such compounding is therefore an integral part of the Pichi system of comparison and emphasis (cf. §6.9.1). Other degree compounds found in the data are *ova-stáwt* 'be too corpulent', *ova-hɔ́t* 'overheat, be too hot', *ova-klín* 'clean excessively, be excessively clean', and *ova-fáyn* 'be excessively beautiful':


Many speakers do not accept degree compounds formed with verbs that are not property items. The alternative to the ungrammatical example (32) is provided in (33):

(32) \* *A* 1sg.sbj *dɔ́n* prf *ova-blánt* over.cpd-reside *na* loc *Panyá.* Spain Intended: 'I have lived in Spain for too long.' [au07ec 052]

(33) *A* 1sg.sbj *dɔ́n* prf *tú* too *mɔ́ch* much *sté* stay *na* loc *Panyá.* Spain 'I have stayed in Spain for too long.' [au07ec 053]

Equally, degree compounding is not accepted with a degree verb like *bɔkú* 'be much' (34). Instead, *óva* may be employed as a degree verb on its own (35):

(34) \* *Di* def *chɔ́p* food *ova-bɔkú.* over.cpd-much Intended: 'The food is too much.' [au07ec 041]

(35) *Di* def *chɔ́p* food *óva.* be.over 'The food is too much.' [au07ec 042]

Property compounds are lexicalised compounds consisting of a property item and noun. Many of these compounds denote human propensities and emotions and involve a body part as the second component. The resulting structures are idiosyncratic and unpredictable in their meanings. Most property compounds are therefore exocentric. Consider *bad-hát* 'bad.cpd-heart' = 'be mean' in (36):

4 Morphology

(36) *Dɛn* 3pl *nó* neg *lɛ́k* like *pɔ́sin,* person *dɛn* 3pl *tú* too *bad-hát.* bad.cpd-heart

'They don't like people, they're too mean.' [ma03hm 012]

Other compounds of this type are *trɔn-yés* 'strong.cpd-ear' = 'be disobedient', *trɔn-héd* 'strong.cpd-head' = 'be stubborn', *gud-hát* 'good.cpd-heart' = 'be good hearted', *brok-hát* 'break.cpd-heart' = 'be broken-hearted', and *opin-yáy* 'open.cpd-eye' = 'be enlightened, cultivated' (cf. 19 above).

There are also some semantically transparent endocentric compounds in the corpus involving dynamic verbs that nevertheless denote properties. Compare the nominalised compound verb *chɔp-mɔní* 'eat.cpd-money' = 'expensive' in (37):

(37) *Dán* that *sáy,* side *na* foc *chɔp-mɔní.* eat.cpd-money 'That place, it's expensive.' [ro07fn 203]

### **4.5 Iteration**

This section describes structures that involve the full iteration of a word. There are two distinct types of iteration in Pichi. Reduplication involves a morphological operation in addition to iteration, namely the tonal derivation also used in compounding (cf. §3.2.4). Repetition involves iteration alone, and is therefore limited to syntactic concatenation. Reduplication is only employed with dynamic verbs and expresses various meanings associated with verbal number. Repetition is attested with a wider range of word classes than reduplication and produces distributive, emphatic, and intensifying meanings (Yakpo 2012a).

A limited number of Pichi words consist of identical components that cannot be separated and used on their own. Such unsegmentable, lexicalised iterations are found in various word classes, including ideophones. In spite of the formal differences between them, reduplication and repetition are characterised by a functional overlap. Both types of iteration are associated with quantification. Table 4.6 summarises relevant features of the two types of iteration in Pichi.

### **4.5.1 Reduplication**

As a productive derivational process, reduplication is only attested with dynamic verbs. However, the pattern is also found in a few lexicalised iterations involving nouns (cf. §4.5.3). Reduplication involves a complex morphological process consisting of the two distinct and simultaneous processes of iteration and tonal derivation. In the process, the verb is reduplicated, and the high tone over the first, reduplicated component is deleted and replaced by an L tone.

Therefore, this kind of reduplication is formally no different from compounding, except that the first component is a copy of the root; hence it involves "self-compounding"

4.5 Iteration


Table 4.6 Types of iteration

(Downing 2001: 6) (cf. §3.2.4 for a detailed treatment of the pitch-related aspects of reduplication). The application of the morphological process of tone deletion to the first component of the reduplicated verb suggests that Pichi reduplications, like compounds, are right-headed (cf. Odden 1996: 117).

Reduplication modifies the meaning of the verb root. The reduplicated verb may therefore appear in any syntactic position that a non-reduplicated verb may be found in. In (38), a reduplicated *wáka* 'walk' appears as a V2 in an SVC. Sentence (39) features a reduplicated *rɔ́n* 'run' as a nominalised verb preceded by the demonstrative *dí* 'this':


In the same vein, reduplication may be applied to a complement verb irrespective of its reduced finiteness:

(40) *Kán* pfv *tót* carry *bɛlɛ́,* belly *bigín* begin *de* ipfv *hala-hála,* red.cpd-shout *náw* now *hala-hála.* red.cpd-shout '(Then she) became pregnant, (and) began lamenting and lamenting.' [ab03ay 118]

Reduplication expresses verbal number. The range of meanings associated with verbal reduplication spans the semantically close notions of iterative aspect, dispersive, distributive, low intensity, and casualness. A befitting cover term for these functions therefore is "temporal and/or spatial disaggregation". Reduplication also often co-occurs with several nominal participants. Pichi reduplication is "event-internal" (Cusic 1981: 238); it

#### 4 Morphology

denotes the reiteration of a single event on a single occasion, consisting of repeated internal phases. Therefore reduplication does not express habitual aspect and is only found with dynamic verbs (cf. §6.3.6 for details on the expression of iterative aspect).

The iterative notion expressed by reduplication harmonises with the meanings expressed by imperfective aspect. There is a much stronger tendency for reduplicated predicates to co-occur with the imperfective aspect marker *de* 'ipfv' than with any other TMA marker. The presence of the imperfective marker and the reduplicated verb *rɔ́b* 'rub' in (41). Since the unmarked reduplicated verb acquires a factative reading (hence past and perfective) by default, the presence of *de* 'ipfv' provides an imperfective sense to the clause:

(41) *Na* foc *ús=káyn* q=kind *tín* thing *mék* make *yu* 2sg *de* ipfv *rɔb-rɔ́b* red.cpd-rub *yu* 2sg *sɛ́f* self *nía* near *mí* 1sg.indp *bifó* before *mi* 1sg.poss *fámbul?* family

'Why are you constantly rubbing yourself up to me [getting all cosy with me] in front of my family?' [ge07fn 129]

Further, iterative reduplication is also attested with the potential mood marker *go* 'pot', as in the following example, and the habitual marker *kin* (cf. 45):

(42) *A* 1sg.sbj *nó* neg *wánt* want *nó* neg *nátín* nothing *wé* sub *go* pot *tayt-táyt* red.cpd-tighten *mi* 1sg.poss *skín.* body 'I don't want anything [clothes] that would be too tight for me (in various places).' [ra07fn 045]

Further, the interaction of verbal and nominal plurality often characterises the use of iterative aspect. The presence of plural referents generally induces a sense of iterativedistributive action of the situation denoted by the verb. For example, the light verb construction in (43) features the reduplicated nominalised verb *jwɛ́n* 'join'. The presence of the plural subject *mí wet Rubi* 'me and Rubi', which is picked up by the resumptive pronoun *wi* '1pl', induces a cumulative meaning of the reduplicated and deverbal noun *jwɛ́n* 'join':

(43) *Mí* 1sg.indp *wet* with *Rubi* name *wi* 1pl *mék* make *jwɛn-jwɛ́n,* red.cpd-join *wi* 1pl *báy* buy *pía,* avocado *wi* 1pl *báy* buy *sadín,* sardine *wi* 1pl *báy* buy *tomates,* tomatoes *wi* 1pl *desayuna.* breakfast

'Me and Rubi, we joined up, we bought avocados, we bought sardines, we bought tomatoes, we had breakfast.' [ye03cd 152]

In turn, the presence of the plural object *nɔ́mba dɛn* 'numbers' in the following sentence renders an iterative and distributive reading of the reduplicated verb *chénch* 'change'.

4.5 Iteration

(44) *Wétin* what *yu* 2sg *de* ipfv *chench-chénch* red.cpd-change *nɔ́mba* number *dɛn* pl *só?* like.that

'Why do you constantly change (telephone) numbers like that?' [ye03cd 131]

The iterative-distributive sense of the reduplicated verb is particularly evident in a reciprocal construction like (45). We have seen that a single form, the pronominal *sɛ́f* 'self, emp' is employed as both the reflexive and reciprocal anaphor. Hence there is room for ambiguity between the reflexive and reciprocal senses when a clause features a plural subject. One disambiguating feature amongst others is the presence of a reduplicated verb. There is no formal feature contained in (45) that would categorically force a reciprocal interpretation on the clause. But the use of reduplication, the presence of plural referents, and the meaning of the verb *cháp* 'chop' and its instrument object *kɔ́tlas* 'cutlass' collude to induce a reciprocal rather than a reflexive meaning of the clause:

(45) *Dɛn kin de chap-cháp dɛn sɛ́f kɔ́tlas ó.*

3pl hab ipfv red.cpd-chop 3pl self cutlass sp

'(Mind you) they have the habit of chopping each other up with cutlasses [referring to political violence in northern Nigeria].' [ye07fn 239]

Conversely, where there are no plural subjects or objects, the iterative meaning of the reduplicated verb shades off into the nuances of low intensity or casualness of the action denoted by the verb. Once again, it is the cumulative meaning of the various elements of the clause that tilts the balance towards this particular reading.

In (46), the intransitive use of the reduplicated verb *tɔ́n* 'turn', in concert with the singular subject *e* '3sg.sbj', favours the related readings of low intensity or casualness. Further examples for these nuances are the reduplication of *rɔ́b* 'rub' in (41) above, and of *táyt* 'tighten' in (48) below. All these examples may also be seen to involve a nuance of lack of control by the subject:

(46) *E* 3sg.sbj *sé* qot *e* 3sg.sbj *wánt* want *kán* come *tɔn-tɔ́n* red.cpd-turn *fɔ* prep *Guinea.* Equatorial.Guinea 'He said he wanted to come move around a little in Equatorial Guinea.' [ed03sb 190]

The distribution of verbal reduplication in my corpus also suggests that it principally occurs in contexts of low transitivity, even if reduplication does not categorically function as a detransitivising device. Hence, preceding examples featuring reduplication for one part involve verbs characterised by a low transitivity, such as locomotion verbs (*wáka* 'wáka', *rɔ́n* 'run') and other verbs denoting body movement (*tɔ́n* 'turn, move around', *rɔ́b* 'rub (oneself)', as well as verbs of sound emission (*hála* 'shout', *kráy* 'cry') in intransitive clauses.

Further, where reduplicated verbs (irrespective of their semantic class) do appear in transitive clauses, these clauses involve less prototypical transitivity, such as reflexive and reciprocal constructions, lexicalised verb-noun collocations (*chénch nɔ́mba* 'change

#### 4 Morphology

one's telephone number') or verbs followed by quantifier phrases like *ɔ́l sáy* 'all place' = 'everywhere'. The latter type of phrase is functionally equivalent to an adverbial indefinite and is therefore not a prototypical undergoer object either:

(47) *Dɛn* 3pl *de* ipfv *lɔk-lɔ́k* red.cpd-lock *ɔ́l* all *sáy.* side 'They're constantly closing every place.' [pa07fn 467]

Additionally, where reduplicated verbs with a higher transitivity occur, they are far more frequent in intransitive clauses. In the following sentence, the reduplicated Spanishorigin verb *pica* 'snip, cut up' appears without a patient object:

(48) *A* 1sg.sbj *bigín* begin *de* ipfv *pica-píca,* red.cpd-cut.up *wi* 1pl *fráy* fry *patata,* potato *wi* 1pl *fráy* fry *plantí.* plantain 'I began to (casually) snip (the trimmings), we fried potatoes, we fried plantain.' [ye03cd.172]

### **4.5.2 Repetition**

Repetition in Pichi is a syntactic operation during which an item is duplicated or triplicated (more repetitions are not attested in the data). Although a pause or boundary tone is not normally inserted between the repeated elements, repetition does not involve the tonal process that characterises compounding and reduplication. Hence every repeated constituent retains its lexically determined tone pattern. Repetition involves syntactic concatenation. Normally, there is no pause or boundary tone between the repeated elements. Hence, the morphological operation characteristic of compounding and reduplication is not employed with this kind of iteration. Repetition is attested with a wider range of word classes than reduplication. My data features repetition of nouns, verbs, attributively used property items, adverbs, and ideophones.

Repetition produces a range of emphatic, intensifying nuances. The core meaning of repetition is augmentative, hence an iconic "more of the same". However, the expression of plural number does not lie within the functional range of repetition. In the following three examples, we witness the use of intensifying repetition for emphasis with the temporal adverb *náw* 'now' (49), the locative noun *dɔ́n* 'down' (50), the common noun *fámbul* 'family', and the attributively used property item *bɔkú* '(be) much' (51):


4.5 Iteration

(51) *Fɔ* prep *mi* 1sg.poss *fámbul* family *fámbul* rep *fámbul* rep *a* 1sg.sbj *nó* neg *sabí* know *bɔkú* much *bɔkú* rep *pɔ́sin* person *dɛn.* pl 'Within my immediate family I don't know that many people.' [fr03wt 031]

The repetition of numerals renders a distributive sense. Clauses in which numerals are used with a distributive sense very often also feature plural nominal participants. In this example, the repetition *tú tú* 'two rep' functions as a depictive adjunct and is oriented towards the plural object pronoun *dɛ́n* '3pl.indp':

(52) *Yu* 2sg *fít* can *kɛ́r* carry *dɛ́n* 3pl.indp *tú* two *tú.* rep 'You can carry them in pairs.' [bo07fn 231]

Numerals of Spanish origin may be repeated for distributive meaning in the same way as Pichi numerals. Sentence (53) features the threefold repetition of the Spanish numeral *quinientos* 'five hundred'. It is worthy of note that repeating the numeral more than twice merely extends the distributive sense to additional participants rather than providing an additional emphatic nuance as with the repetition of members of other word classes:

(53) *Quinientos* five.hundred *quinientos* rep *quinientos.* rep 'Five hundred each.' [hi03cb 058]

The preceding examples have shown that various syntactic categories may be subjected to repetition. Nevertheless, the by far most commonly repeated categories are property items functioning as prenominal attributive modifiers like *bɔkú* in (51) above, distributive numerals used as depictive modifiers like *tú* 'two' in (52) above, and time expressions like *náw* 'now' in (49) above. This distribution points towards the fact that repetition is strongly associated with gradable, quantity- and quality-denoting lexical items, as well as with distribution.

The quantificational essence of repetition also transpires when it is applied to time expressions. The corpus contains numerous instances of repeated time expressions with an emphatic, quantificational meaning. The repetition of a temporal adverb like *náw* 'now' (49) above or a temporal noun like *mɔ́nin* 'morning' in the following sentence renders an intensive meaning 'early in the morning, at dawn':

(54) *Áfta* then *a* 1sg.sbj *de* ipfv *mít=an* meet=3sg.obj *nía* near *di* def *klós* clothing *dɛn* pl *di* def *mɔ́nin* morning *mɔ́nin* rep *tɛ́n.* time 'Then I ran into her by the clothes at dawn.' [ru03wt 037]

Other time expressions that allow some form of gradation are also frequently repeated in this way. For example the property item *lɔ́n* '(be) long' in the collocation *lɔ́n tɛ́n* 'long time ago' is very often repeated in order to indicate a larger degree of time-depth:

#### 4 Morphology

(55) *E* 3sg.sbj *bin* ipfv *dɔ́n* prf *pás* pass *lɔ́n* long *tɛ́n,* time *nóto* neg.foc *lɔ́n* long *lɔ́n* rep *tɛ́n.* time 'It happened long ago, not very long ago.' [ma03sh 001]

The repetition of time expressions involving the generic noun *tɛ́n* 'time' depends in form on the degree of semantic independence of the components of the collocation. When the collocation is endocentric, only the modifier element is reduplicated. In the following sentence, only *wán* 'one' is therefore repeated rather than the entire expression *wán tɛ́n* 'once'. The same holds for *lɔ́n tɛ́n* 'long ago' in the preceding example:

(56) *Na* foc *wán* one *wán* rep *tɛ́n* time *dásɔl.* only 'It's just once in a while.' [fr03ft 053]

In contrast, once the two words *wán* and *tɛ́n* are employed as part of the lexicalised expression *wántɛn* 'at once', the entire collocation is repeated:

(57) *Na* foc *wán* one *mán* man *wé* sub *de* ipfv *abraza* embrace *tú* two *húman* woman *wántɛn* at.once *wántɛn* rep *só.* like.that 'That's a man embracing two women at once.' [dj07re 038]

Further, the repetition of periods of the day other than *mɔ́nin (tɛ́n)* 'morning (time)' is not encountered in the data. Expressions like *ívin tɛ́n* 'evening' or *sán tɛ́n* 'noon' do not appear to lend themselves to some concept of quantification or gradation. This is possibly so because the corresponding period is of no cultural relevance, while 'at dawn' in (54) above is, since this is when people usually get up. Hence, for example, there is no instance of ?*sán sán tɛ́n* with the intended reading 'exactly at noon'.

We are therefore once more dealing with a degree of lexical specialisation here. Such lexicalisation is also attested with other common repetitions. For example, the two dimension concepts *bíg* '(be) big' and *smɔ́l* '(be) small' are two of the most commonly encountered repeated property items in the corpus. Compare the following two examples:


In the rarer cases where verbs that function as predicates rather than prenominal modifiers are repeated, these are usually not property items. Property items are most commonly repeated when they precede a head noun as attributive modifiers; there is

4.5 Iteration

not a single instance of a repeated property item functioning as a predicate, e.g. ?*e bíg bíg* 'it is very big'.

The meanings of repeated verbs are closely tied to their semantic structure. Hence, a verb like *kɔ́t* 'cut' may imply a series of cyclic repetitions, particularly in the context of cooking as in (60). The resulting meaning of the repetition is very close to that of iterative reduplication in an example like (48) above. Note that this verb is repeated together with its clitic object pronoun *=an* '3sg.obj':

(60) *Di* def *dé* day *yu* 2sg *bwɛ́l* boil *jakató* bitter.tomato *yu* 2sg *kɔ́t=an* cut=3sg.obj *kɔ́t=an* rep *kɔ́t=an* rep *yu* 2sg *báy* buy *wán* one *sardina* sardine.

'The day you boil bitter tomato, you cut it up into small bits (and) you buy a sardine.' [ro05rt 063]

A similar case can be made for the repetition of the locomotion verb *júmp* 'jump'. This verb also naturally lends itself to a cyclical movement. In (61), reduplication and the simultaneous use of repetition of the reduplicated sequence build up to an emphatic iterative sense with a cyclical meaning:

(61) *Sɔntɛ́n* perhaps *e* 3sg.sbj *bin* pst *de* ipfv *jump-júmp* red.cpd-jump *jump-júmp,* rep *pero* but *e* 3sg.sbj *strét* be.straight *náw.* now 'Let's assume she was constantly jumping around but she's upright now.' [ye07je 111]

Two words in the corpus allow partial iteration. With the two inchoative-stative verbs and property items *wɔwɔ́* '(be) ugly, messed up' and *lílí* '(be) little, tiny', one syllable rather than the entire word may be iterated. Both words share the characteristic that they already constitute lexicalised iterations or at least appear so by their their segmental structure. Sentence (62) exemplifies the partial iteration of *lílí* '(be) little'. A simplex word \**lí* does not exist in Pichi. Since there is no sign of tone deletion over the first component of the iteration, I analyse *lílí-lí* as an instance of partial repetition rather than reduplication:

(62) *Pero* but *como* since *di* def *harina* flour *tú* too *lílí-lí,* little-rep *kɔ́n* corn *tú* too *smɔ́l* be.small *náw,* now *a* 1sg.sbj *mezcla* mix *ín* 3sg.indp *ɔ́l.* all

'But since the flour is too little, the corn is too little now, I mixed all of it [in making the porridge].' [dj03do 044]

Now compare the fully (63) and partially iterated (64) alternatives for *wɔwɔ́*'(be) ugly, messed up'. In both examples, the property item *wɔwɔ́* is employed as a prenominal modifier. Note that a monosyllabic root *\*wɔ* does not exist in Pichi:

#### 4 Morphology


The tonal characteristics of the partial iteration of *wɔwɔ́* in (64) above are of interest. In the example, the original lexical H tone over the first syllable of the *wɔ-wɔwɔ́* before the ligature has been replaced by an L tone. The presence of tone deletion points to the operation of partial reduplication rather than repetition. This contrasts with the iteration of other, attributively used property items in a similar way. In (58) and (59) above, *bíg* and *smɔ́l* undergo repetition, not reduplication. Although this example stands alone, it may be indicative of an area of transition between reduplication and repetition not only in meaning but also in form.

There is often no sharp distinction in meaning between the repetition of single words and the iteration of larger chunks of a sentence. This is particularly so if the repeated elements are not separated from each other by a pause or declarative intonation (hence an utterance-final fall) as in the sentence below. The iteration of the NP *in estómago* 'her stomach' in (65) conveys a repetitive and emphatic meaning in very much the same way as the verb-object phrase *kɔ́t=an* 'cut=3sg.obj' in (60):

#### (65) *Nɔ́,* intj *in* 3sg.poss *estómago* stomach *in estómago* rep *in estómago.* rep

'[She would repeatedly say] No, (it's) her stomach, her stomach, her stomach [rather than a pregnancy].' [ab03ay 122]

### **4.5.3 Lexicalised iteration**

A limited number of Pichi words consist of identical components that cannot be separated and used on their own. Such unsegmentable, lexicalised iterations are found in various word classes. An example follows featuring the ideophonic noun *wuruwúrú* 'confusion'. The (lexicalised) iteration of ideophones is covered in section §12.1.

(66) *Dɛn* 3pl *de* ipfv *mék* make *wuruwúrú.* confusion 'They're causing confusion.' [be07fn 147]

The pitch structure of lexicalised iteration is characterised by diversity. Some words feature a pitch configuration suggestive of reduplication, others feature a configuration that points towards repetition. The former group comprises cases of lexicalised iterations (67a) with no attested simplex form but whose etymology can be established. It also encompasses words with identical components, of which the origin of the simplex form is

#### 4.5 Iteration

difficult or impossible to establish – these words are probably reflexes of English or Portuguese lexicalised iterations (b). The group also contains words which have a deducible, but idiosyncratic semantic relation with a simplex form (c). With all these words, we find an L tone over the first component of the word, while the second component bears an H tone. Hence this is the pitch configuration that we have already seen with iterative, verbal reduplication in section §4.5.1. The only difference is that (67) also includes nouns:


## **5 The nominal system**

Nouns are modified grammatically and pragmatically by means of pre- and postnominal elements. Common nouns are not inflected for number, case or gender in Pichi. In the personal pronoun paradigm, number and case are, however, morphologically marked. Generally, a noun phrase (henceforth NP) headed by a common noun has the structure given in Figure 5.1, which provides a (constructed) complex NP for exemplification.


'As for all these my two last nice stories here of today that (…)'

Figure 5.1 Structure of the noun phrase

The possibilities for modifying nouns with determiners (def and dem) and quantifiers (qnt) depend on their lexical class. Pichi nouns fall into three lexical classes: count nouns (e.g. *hós* 'house') including collective nouns (e.g. *pípul* 'people'), mass nouns (e.g. *watá* 'water'), and proper nouns (e.g. place names, such as *Panyá* 'Spain', as well as personal names like *Tokobé).*

The slot def/dem indicates that the definite article *di* (def) and the proximal and distal demonstratives *dí* and *dán* (dem) do not cooccur. Possessive pronouns (pron) precede the head and may co-occur with demonstratives but not with the definite article. NP constituents in other slots featuring a single function label in Figure 5.1 may coocur.

There are two quantifier slots. The quantifiers *ɔ́l* 'all' and *dásɔl* 'only' (qnt) can be floated and may occur either in a pre- or post-head position (hence the postnominal *ɔ́l* in brackets). The possessor in compounds, associative constructions, and dislocated possessive constructions is best seen to fill the modifier (mod) slot. Several modifiers can therefore co-occur (e.g. *bíg blák kichin-písis* 'big black kitchen rag'). The possessor in a *fɔ*-prepositional construction follows the head, but its exact position in the postnominal slot may depend on pragmatic factors, e.g. either before or after *sɛ́f* or *náw* depending on the scope of foc or top. Relative clauses (relc) invariably follow the head noun.

### **5.1 Determination**

This section covers the distribution and functions of the definite article, indefinite determiners, demonstratives, and number marking. Quantifiers are treated separately in section §5.3.

#### 5 The nominal system

### **5.1.1 Definiteness and specificity**

Definiteness and specificity of nouns are marked by the prenominal definite article *di* 'def' and the indefinite determiners *wán* 'one, a' and *sɔn* 'some, a'. In addition, bare nouns without a preceding determiner are marked for definiteness and specificity by default. Some relevant characteristics of definiteness marking are presented in Table 5.1. The use of bare nouns is covered in more detail in §5.1.4.

Table 5.1 Characteristics of definiteness marking


The definite article *di* signals definiteness of a noun phrase. It is neutral as to number and can be used with count, mass, and proper nouns alike. *Di* may precede NPs headed by full nouns (cf. 5 below), the numeral *wán* 'one' in its function as a pronominal (1), or any element functioning as a noun, such as the deverbal noun *dú* in (2):


Proper nouns, such as the place name *Camerún* 'Cameroon' and personal names, do not usually co-occur with the article (3), but may appear with it if required (4):


#### 5.1 Determination

The definite article *di* is employed in contexts in which a noun is specific, identifiable, and familiar to discourse participants either through its presence in the immediate physical surrounding (e.g. *maíz* 'maize') (5), or through situational inference (e.g. *mɔ́nin mɔ́nin tɛ́n* 'early in the morning') (6):


The associative use of the article is exemplified in (7). The referent *leche* 'milk' has been established earlier on in discourse. The Spanish noun *animal* 'animal' is therefore definite by association with the antecedent *leche*:

(7) *Es* it.is *que,* that *e* 3sg.sbj *fáyn* be.fine *wé* sub *yu* 2sg *nó* neg *sabí* know *sé* qot *e* 3sg.sbj *kɔmɔ́t* come.out *fɔ* prep *di* def *animal.* animal

'It's that it is fine when you don't know that it [the milk] has just come out of the animal.' [ed03sp 105]

The anaphoric use of the article can be seen in the following examples. The referent *mán* 'man' is introduced in (8a) by the speaker abbreviated as (hi) (cf. Table 1.1 in §1.7) and taken up as a definite NP by speaker (bo) in (b). Note the presence of the Nigerian Pidgin form *haws* 'house' instead of Pichi *hós* in (b):

(8) a. *Dɛn* 3pl *kin* hab *fíɛ* fear *dɛn* 3pl *mán* man *dán* that *káyn* kind *stáyl.* style

'They (usually) fear their husbands and the like.' [hi03cb 131]

b. *Yu* 2sg *de* ipfv *fíɛ* fear *di* def *mán* man *mék* sbjv *e* 3sg.sbj *nó* neg *bít* beat *yú* 2sg.indp *ɔ* or *mék* sbjv *e* 3sg.sbj *nó* neg *drɛ́b* drive *yú* 2sg.indp *fɔ* prep *haws* house *ó.* sp

'You fear your man lest he should beat you or drive you out of the house.' [hi03cb 132]

Cataphoric use of the article – where the identity of the definite noun is established in following discourse – can be seen in the relative construction in (9):

#### 5 The nominal system

(9) *Yu* 2sg *nó* neg *fít,* can *porque* because *yu* 2sg *mamá* mother *nó* neg *go* pot *hébul* be.capable *pé* pay *ɔ́l* all *di* def *wók* work *wé* sub *dán* that *mán* man *dɔ́n* prf *dú* do *fɔ* prep *yú* 2sg.indp

'You can't because your mother wouldn't be able to pay all that work that the man has done for you.' [ab03ay 021]

Singular count nouns are marked for indefiniteness with the cardinal numeral *wán* 'one' (10), or with the quantifier *sɔn* 'some, a' (cf. 11 below). The numeral *wán* is not a fully grammaticalised indefinite article. In many contexts,*wán* retains its lexical meaning of 'one'. *Wán* also has pronominal functions and can itself be preceded by the demonstratives *dí* and *dán* and the definite article *di* (e.g. 1).

(10) *A* 1sg.sbj *gɛ́t* get *wán* one *bíg* big *sísta* sister *wé* sub *na* foc *mulata.* African-European.f 'I have a/one big sister who is African-European.' [fr03ft 022]

When used with count nouns, *wán* usually signals a higher degree of specificity than *sɔn*. However, there is no categorical distinction between specific and non-specific deixis in Pichi. This can be seen in the following two sentences. Here the noun *fébɔ* 'favour' appears with *sɔn* 'some, a' in (11) and *wán* 'one, a' in a specific and emphatic setting in (12):


Given that Pichi does not mark number on nouns morphologically, *wán*, rather than *sɔn*, is used to express that singular number is a significant feature of the referent as in *wán motó* 'one car' (13). Here an interpretation of *wán* as a numeral would appear awkward, since the speaker does not have more than one car in mind:

(13) *Yu* 2sg *sabí* know *sé* qot *wán* one *motó* car *fɔ* prep *wán* one *mún* month *na* foc *cincuenta* fifty *dólar,* dollar *ɛf* if *yu* 2sg *hól* hold *wán* one *motó fɔ wán mún.*

car prep one month

'You know that a car for one month is fifty dollars, if you keep a car for only one month.' [ed03sp 076]

5.1 Determination

*Wán* rather than *sɔn* is also common in emphatic contexts. The data does not contain a single sentence in which a noun is preceded by *sɔn* in an equative clause of the type in (14), in which the identified entity is highly specific. The numeral *wán* may also signal additional emphasis when it precedes a noun under cleft focus in a presentative construction, as in (15) (cf. also §7.4.4):

(14) *Na* foc *wán* one *ɔnkúl* uncle *directo,* direct *fɔ* prep *mi* 1sg.poss *mamá* mother *in* 3sg.poss *papá* father *in* 3sg.poss *fámbul* family *pát.*

part

'(He) is a direct uncle on my mother's father's family's side.' [fr03ft 051]

(15) *E* 3sg.sbj *dé* be.loc *complicado,* complicated *na* foc *wán* one *tín* thing *dat.* that 'It's complicated, it's one (kind of a) thing.' [ye07de 017]

Contrary to what one would expect of a cardinal numeral that signals singular number, *wán* can also modify a noun containing a numeral above one (16). Such usage of *wán* is often found in conjunction with Spanish numerals and head nouns and is likely to be a case of structural borrowing from Spanish. In Spanish, the plural indefinite article (*unos*/*unas*) fulfills an identical function (cf. also §13.3.1):

(16) *Áfta* then *wi* 1pl *kán* real *mít* meet *layk* like *wán* one *seis* six *años* year.pl *después.* afterwards 'Then we met again some six years later.' [fr03ft 191]

With plural count nouns, indefiniteness is signalled through the presence of *sɔn* alone (17) or the absence of a definiteness expression altogether (cf. §5.1.4). Mass nouns may only be modified by *sɔn* for indefiniteness, or they occur devoid of any determiner (18):


'Blood was coming out of her nose, it was coming out of her mouth.' [ab03ay 125]

Furthermore, *wán*, but not *sɔn*, may occur with NPs that are within the scope of negation, even if only with an emphatic meaning (19). In the absence of emphasis, NPs do not usually appear with a marker of indefiniteness in negative clauses (20) (cf. §7.2.2 for details):

#### 5 The nominal system


Both *wán* and *sɔn* can function as pronominals and refer anaphorically to a preceding indefinite NP. While *wán* is limited to anaphoric reference of a singular count noun, *sɔn* may be used to refer to preceding singular or plural count and mass nouns.

In both (21) and (22), *wán* and *sɔn* refer to a preceding NP *televisión* 'TV set'. When referring to a plural noun, *sɔn* may optionally be followed by the pluraliser *dɛn* 'pl' (23):


*Sɔn* and *wán* may also be used with a partitive reading when followed by a definite possessed noun. Once more the nominal referent preceded by *sɔn* tends to receive a less specific reading than the one featuring *wán*. The same meaning may alternatively be expressed if *sɔn* or *wán* are followed by a definite *fɔ*-prepositional phrase (cf. e.g. 83):


Finally, only the quantifier and indefinite, non-specific determiner sɔn appears in NPs which function as nominal and adverbial indefinite pronouns and involve generic nouns like *tín* 'thing', *pɔ́sin* 'person', *tɛ́n* 'time', *sáy* 'side', *plés* 'place', *áwa* 'hour, time', and *stáyl* 'style'. Compare the following two examples (cf. §5.4.3 for a complete listing):

5.1 Determination


'One day somebody came that, he was saying that you had given a blow somewhere [you had fathered a child somewhere].' [fr03wt 185]

### **5.1.2 Demonstratives**

Pichi has a two-term demonstrative system that serves to express the notions of proximity and distance with the speaker as the deictic centre. The demonstratives *dí/dís* 'this' and *dá/dán/dat* 'that' and sometimes *dɛn* 'those' express the spatial, temporal, and discourse functions of proximal and distal reference respectively. Table 5.2 gives an overview of the forms and functions of Pichi demonstratives.


Table 5.2 Demonstratives

It is unclear whether *dí* and *dá* are distinct realisations or phonological variants with a deleted final consonant of the forms *dís* and *dán/dát*. The differentiation between *dán* and *dat* suggests that the "short" and the "long" forms may be distinct developments from their respective English etymons (< 'this/that'). Likewise, the use of either form as attributive demonstratives could not be correlated to any (socio-)linguistic conditioning factor.

In contrast, it is very likely that *di* 'def' is a reflex of English *the*, while *dí* 'this' is a reflex of the English proximal demonstrative *this*. The evidence is prosodic. Pichi *di* 'def' was lexicalised as L-toned because English *the* is usually unstressed, while *dí* 'this' received a lexical H because *this* is usually stressed in English.

Demonstratives may be used attributively as prenominal modifiers. The forms *dí* and *dís* are equivalent in function, although *dí* is more common as a proximal demonstrative (28):

#### 5 The nominal system

(28) *Djunais* name *tɔ́k* talk *sé,* qot *nɔ́* intj *Rubi* name *dí* this *gɛ́l* girl *lɛ́k* like *yú.* 2sg.indp 'Djunais said, really Rubi, this girl likes you.' [ru03wt 021]

The two forms *dá* and *dán* serve as distal attributive demonstratives (29). The form *dán* is used in the majority of cases, irrespective of the word-initial onset of the following noun. NPs featuring an attributively used demonstrative are pluralised in the usual way by means of the postposed pluraliser *dɛn* (29):

(29) *Ɔ́l* all *dán* that *pikín* child *dɛn* pl *na* foc *dán* that *mán* man *in* 3sg.poss *yón.* own 'All those children are that man's.' [hi03cb 190]

*Dí* and *dís* (30), as well as *dá* and *dán* (cf. e.g. 38) may combine with the numeral and pronominal *wán* 'one', in order to form singular (30) and plural (31) demonstrative pronominals:


The forms *dís* and *dat* may be employed as independent pronominals on their own, although this use is marginal compared to that involving the pronominal *wán*:


*Dís* and *dát*, but never *dí* and *dá/dán*, also occur in sentence-final position in a presentative construction of the type presented in (34) and (35), where the demonstratives are anaphoric to an antecedent focused NP (cf. §7.4.4):

(34) *Sé* qot *na* foc *ín* 3sg.indp *dís,* this *na* foc *yu* 2sg *húman* woman *dís,* this *yu* 2sg *wánt* want *ɔ* or *yu* 2sg *nó* neg *wánt,* want *na* foc *in* 3sg.poss *dís.* this '(She said) this is her, this is your wife, you like it or not, this is her.' [ed03sp 009]

5.1 Determination

(35) *Na* foc *in* 3sg.poss *vida* life *dát.* that 'That's his (kind of) life.' [he07fn 228]

Demonstrative adjectives do not co-occur with the definite article. They may, however, precede proper nouns (36) and possessive pronouns (37):

(36) *Lúk=an,* look=3sg.obj *di* def *dé* day *wé* sub *dís* this *Paquita* name *in* 3sg.poss *papá* father *bin* pst *kán* come *ték=an,* take=3sg.obj *e* 3sg.sbj *pé* pay *avioneta.* small.aircraft 'Look at this, the day that Paquita's father came to take her, he hired a small

(37) *Cuñado,* brother-in-law *mí* 1sg.indp *gɛ́fɔ* have.to *fɛ́n* look.for *dán* that *mi* 1sg.poss *prima* cousin.f *ó,* sp *Cristina.* name 'Brother(-in-law), I [emp] really have to look for that my (female) cousin, Cristina.' [ge07ga 046]

Demonstratives are often reinforced through the deictic locative adverbs *yá* 'here', *dé* 'there', and sometimes *yandá* 'yonder, over there' (38):

(38) *Ɛhɛ́,* intj *wán* one *glás* glass *watá* water *aparte,* separate *yu* 2sg *pút=an* put=3sg.obj *ínsay,* inside *dán* that *wán* one *dé,* there *yu* 2sg *fít* can *ték* take *medio* half *fɔ* prep *dán* that *sén* same *glas* glass (*…*)

'Exactly, one glass of water separately, you put it inside, as for that one, you can take half in that very glass (…)' [dj03do 054]

The idiom *dís-tín* 'this-thing' may substitute for an inanimate noun. Example (39) shows that this expression has been lexicalised to an extent which allows the occurrence of the demonstrative *dán* 'that' with its full referential meaning:

(39) *A* 1sg.sbj *ték* take *tú* two *peso* peso *a* 1sg.sbj *báy* buy *dán* that *dís-tín* this-thing (*…*) 'I took two pesos (and) I bought this whatsit (…)' [ed03sp 083]

The 3pl dependent personal pronoun and pluraliser *dɛn* occasionally occurs in the determiner position at the very left of the NP. In this position, *dɛn* simultaneously functions as a plural definite article and a demonstrative with a largely discourse deictic function. Prenominal *dɛn* usually also has emphatic force. This use of *dɛn* however is marginal in the corpus. Note the additional presence of *dɛn* as a pluraliser after the noun *fronteras* 'borders' (40):

aircraft.' [ab03ay 140]

#### 5 The nominal system

(40) *Wet* with *ɔ́l* all *dɛn* those *fronteras* borders *dɛn* pl *wé* sub *dɛn* 3pl *de* ipfv *chénch.* change 'With all those borders that are changing.' [fr03ft 102]

In their function as markers of spatial deixis, the proximal and distal demonstratives serve to locate referents in physical space with the speaker as the deictic centre (41):

(41) *Wi* 1pl *de* ipfv *gó* go *dɔ́n,* down *wi* 1pl *de* ipfv *gó* go *lɛ́f=an* leave=3sg.obj *di* def *sáy* side *dán* that *motó* car *dé.* be.loc 'We're going down, we are going to leave it where that car is.' [ma03ni 043]

The demonstrative pronouns also serve to express discourse-pragmatic deixis. I reiterate example (38) above in (42) below in context. In the excerpt, speaker (dj) explains how to cook corn porridge. The interjection *ɛhɛ́*'exactly' confirms the interruptive question posed in (42a). The topical *dán wán dé* in (b) is therefore anaphoric to the process explained just beforehand in the same sentence.

The anaphoric function of the distal demonstrative pronoun is frequently made use of in order to refer to preceding NPs, phrases, and entire sentences. *Dán sén glás* 'that very glass' represents in (38) an additional means of referent tracking via the use of the focus and emphasis marker *sén* 'same, very':

(42) a. *Wán* one *glás* glass *watá?* water

'One glass of water?' [fr03do 053]

b. *Ɛhɛ́,* exactly *wán* one *glas* glass *watá* water *aparte,* separate *yu* 2sg *pút=an* put=3sg.obj *ínsay,* inside *dán* that *wán* one *dé* there *yu* 2sg *fít* can *ték* take *medio* half *fɔ* prep *dán* that *sén* same *glas* glass (*…*)

'Exactly, one glass of water separately, you put it inside, that one [that method], you can take half in that very glass (…)' [dj03do 054]

### **5.1.3 Number**

Pichi marks plural number via the postposed pluraliser *dɛn* which is identical to the 3pl dependent pronoun. The pluraliser is clitic-like in one respect: It may not be separated from the noun it refers to by any constituent. Typically, the pluraliser occurs with count nouns (43), but it may also follow collective nouns like *pípul* 'people' (44):


5.1 Determination

The pluraliser is also encountered with mass nouns denoting liquids such as *watá* 'water' (45) or *leche* 'milk' in (46):


NPs featuring a cardinal numeral can also optionally be marked for plural number (47), although in the majority of instances, speakers prefer not to use the pluraliser together with a numeral (48):


Furthermore, the pluraliser may co-occur with quantifiers that indicate plurality of the referent such as *ɔ́l* 'all' (49), and *bɔkú* 'many, much' (50), although the absence of plural marking is equally common (51):


The pluraliser is also consistently made use of with inserted Spanish nouns marked with the Spanish plural morpheme {-s} (52). The same is true of the few instances in the corpus, in which the nouns *bɔ́y* 'boy' and *gál* 'girl' are marked for plural with the marginal Pichi plural morpheme {-s} as in (53):

5 The nominal system


Personal names may be pluralised in order to form an associative plural (54). The resulting meaning is 'X and those associated with her/him habitually or at the time of reference':

(54) *A* 1sg.sbj *dɔ́n* prf *explica* explain *Boyé* name *dɛn,* pl *sé* qot *na* foc *só* like.that *mi* 1sg.indp *de* ipfv *mɛ́mba,* remember *ɔ́l* all *tín.* thing

'I have explained to Boyé and the others that this is how I remember everything.' [ru03wt 045]

Plural number need not be marked on the head noun of a relative clause and may instead be expressed via the coreferential subject pronoun in the relative clause:

(55) *Di* def *húman* woman *wé* sub *dɛn* 3pl *fáyn* fine *mɔ́* more *na* loc *América Latina* place *húman* woman *dɛn.* pl

'The women who are the most beautiful are Latin American women.' [ed03sp 025]

Syntactic factors may also constrain plural marking. One of the instances in which plurality is not overtly expressed and left to inferral is in dislocated possessive constructions.

I repeat sentence (31) in (56) below. As is generally the case in dislocated possessive constructions, a personal pronoun coreferential with the possessor (*dɛn* '3pl') links the plural possessor (*dís wán dɛn* 'these ones') and the possessed noun (*mamá* 'mother'). I interpret the linker *dɛn* in these cases as the 3pl pronoun rather than the pluraliser, since singular possessors require the use of the corresponding singular possessive pronoun *in* '3sg.poss' in the same position. Hence the pluraliser remains unexpressed in the construction in order to avoid doubling of the two homophonous forms:

(56) *Na* foc *dé* there *dís* this *wán* one *dɛn* 3pl *mamá* mother *dɛn* 3pl *de* ipfv *mék* make *ɛ́ni* every *tín.* thing 'It is then that these ones' mothers do every thing.' [ab03ay 047]

In (57), we encounter a similar overlap of pl and 3pl. Here, *dɛn* may be interpreted as the pluraliser postposed to the NP or instead, as a resumptive pronoun and the subject of the following verb. In contexts such as these, where a predicate immediately follows

5.1 Determination

a plural-referring NP, the distinction between the pluraliser and a 3pl resumptive pronoun is not possible, since doubling of the form is normally avoided. The distributional characteristics of *dɛn* in these contexts indicate the significant functional overlap of NP and verbal number marking in Pichi:

(57) *Estudiante* student *fɔ* prep *Guinea* place *dɛn* 3pl/pl *de* ipfv *sɔ́fa* suffer *plɛ́nte.* plenty 'Guinean students were suffering a lot.' [ed03sp 086]

Finally, I point out that Pichi has at least two nouns with suppletive plural forms which are occasionally employed instead of the regular plural involving *dɛn* 'pl'. The relevant singular-plural pairs are *gál-gáls* 'girl-girls' and *bɔ́y-bɔ́ys* 'boy-boys'. However, these forms are not suppletive in the true sense, since they feature the segmentable but only marginally productive plural morpheme {-s}, which is only attested with these two nouns. As example (53) above shows, these forms may also be followed by the pluraliser *dɛn*.

### **5.1.4 Genericity**

Generic reference of an NP can be established through the use of bare nouns with or without plural marking, as well as the use of the definite article *di* 'def'. A noun phrase may consist of only a bare noun. The demarcation between count and mass nouns is blurred when they are used as "non-individuated" (Mufwene 1986) nouns in this way, since the number distinction is now irrelevant for both entity types.

Generalisations may be made about a whole class of referents by using the bare form of the corresponding count noun in generic statements like the following ones:


In contrast, the use of the bare form is the normal way of referring to indefinite and non-specific mass nouns like *chɔ́p* 'food' and *pamáyn* 'oil', while definite (and specific by default) mass nouns are preceded by the definite article *di* 'def' like count nouns:


#### 5 The nominal system

In Pichi, weather mass nouns like *brís* 'wind', *tináda* 'thunderstorm', and *rén* 'rain' also have non-specific NP marking and reference when they occur in weather condition clauses like the following one:

(62) *Brís* air *de* ipfv *bló.* blow 'The wind is blowing.' [dj07ae 242]

However, with count nouns, generic reference can also be established by employing a plural noun without a determiner (63):

(63) *Mán* man *dɛn* pl *nó* neg *de* ipfv *bísin* be.busy *fɔ* prep *mék* make *fám* farm *mɔ́.* more 'People are no more into farming.' [ed03sp 053]

Further, the reference of the definite article *di* 'def' may also be construed as generic if it co-occurs with generic TMA marking. In this example, imperfective marking expresses a habitual, generic sense, and the nouns *gabonés* and *guineano* designate the whole class of referents rather than specific ones:

(64) *Pero* but *di* def *gabonés* Gabonese *wé* sub *de* ipfv *tɔ́k* talk *Bata* Fang *wet* with *di* def *guineano* Guinean *wé* sub *de* ipfv *tɔ́k* talk *Bata,* Fang *di* def *sonido* sound *nó* neg *dé* be.loc *di* def *sén.* same

'But the Gabonese who talks Fang and the Guinean who talks Fang, the sound is not the same.' [ma03hm 048]

Example (65) illustrates how generic meaning arises through the interplay of NP marking (the bare NP *tidé pikín* 'children of today'), impersonal use of 2sg, and the habitual reading of the potential modality marker *go*:

(65) *Tidé* today *pikín* child *yu* 2sg *go* pot *gɛ́t* get *bɛlɛ́,* belly *yu* 2sg *púl=an* remove=3sg.obj *yu* 2sg *go* pot *dáy* die *wet* with *bɛlɛ́.* belly 'As for children of today, they get pregnant, they abort it and die because of the pregnancy.' [ab03ay 105]

Bare nouns are also encountered in many idiomatic verb-object collocations involving count nouns such as *mék fám* 'to farm', *gɛ́t bɛlɛ́* 'to be pregnant', or *fála húman* 'to womanise'. Such noun phrases are also characterised by genericity by virtue of their non-specific reference. They equally reflect a general tendency to omit indefiniteness and number marking with non-specific objects (66):

(66) *A* 1sg.sbj *ralla* grate *ín* 3sg.indp *wet* with *rallador.* grater 'I grated it with a grater.' [dj03do 004] 5.2 Noun phrase modification

### **5.2 Noun phrase modification**

Nouns are modified by pre- and post-nominal modifiers and possessive constructions. Postnominal modification via focus and topic markers is treated separately in sections §7.4.2 and §7.5, respectively. Nouns may also be modified through relative clauses (cf. §10.6) and noun complement clauses (cf. §10.5.8).

### **5.2.1 Prenominal modification**

Head nouns of noun phrases may be modified prenominally by other nouns and by verbs in compounds, by nouns in associative constructions, as well as by quantifiers and property items that have been converted to attributive adjectives. In (67), the nouns *mán* 'man' and *húman* 'woman' are modified by the preposed property item *bíg* '(be) big'.

(67) *Bɔt* but *wé* sub *di* def *mán* man *na* foc *bíg* big *mán,* man *di* def *húman* woman *sɛ́f* emp *na* foc *bíg* big *húman,* woman *porque* because *ɔ́l* all *tɛ́n* time *na* foc *húman* woman *dé* be.loc *bɔtɔ́n* under *mán.* man

'But when the man is a big man, the woman, too is a big woman, because it is always the woman who is below the man. [hi03cb 152]

An ordinal numeral or similar quantifier such as *ɔ́da* 'other' immediately follows the article and precedes other modifiers (68):

(68) *Yu* 2sg *pút* put *ɔ́da* other *nyú* new *wán* one *ínsay,* inside *dán* that *wán* one *sé* qot *mék* sbjv *e* 3sg.sbj *nó* neg *smɛ́l.* smell '(Then) you put another one inside, that in order for it not to smell.' [dj03do 048]

Speakers show clear preferences in their use of verbs for prenominal modification in NPs. Firstly, only numerals and other quantifying expressions (e.g. *nɛ́ks* 'next', *plɛ́nte* '(be) plenty') as well as other property items usually function as attributive modifiers.

Secondly, the following more "basic" semantic types of property items have the highest likelihood of occurring as prenominal modifiers to head nouns: dimension (e.g. *bíg* '(be) big' in (67) and *smɔ́l* '(be) small), age (e.g. *ól* '(be) old', cf. 68), value (e.g. *bád* '(be) bad', *bɛ́ta* '(be) very good', *fáyn* '(be) fine, beautiful', *trú* '(be) true', and *(s)trɔ́n* 'be strong, profound (cf. 69), colour (e.g. *blák* '(be) black', *wáyt* '(be) white', and *rɛ́d* '(be) red'):


#### 5 The nominal system

Physical properties (e.g. *swít* '(be) tasty', *évi* '(be) heavy', *hád* '(be) hard', *sáf* '(be) soft') are far less likely to appear in the prenominal position. So are human propensities, be they lexicalised as dynamic (e.g. *krés* '(be) crazy', *jɛ́lɔs* '(be) envious') or inchoative-stative verbs (e.g. *wíkɛd* '(be) wicked'). Further, the corpus contains no instance of a prenominal, modifying use of labile change-of-state verbs like *brók* '(be) broken, break,' *lɔ́s* '(be) lost, lose,' *lɔ́k* 'close, (be) closed', and locative verbs like *sidɔ́n* 'sit, seat'.

Instead, members of the semantic classes listed above preferably occur in other kinds of modifying structures, such as relative constructions (71) and compounds (72):


The few members of the Pichi adjective class (e.g. *fáyn* 'be fine) may appear in the prenominal modifier position like other property items. However, only adjectives may function as complements to the locative-existential copula *dé* in predicate adjective constructions (cf. §7.6.5).

### **5.2.2 Postnominal modification**

Nouns may be modified by postposed elements of two types: focus particles (cf. §7.4.2), the topic marker *náw* 'now', and optionally, by quantifiers like *wán* 'alone' (cf. 93–94), *ɔ́l* 'all' (cf. 96), and *dásɔl* 'only' (cf. 106).

### **5.2.3 Possessive constructions**

Pichi employs four types of possessive constructions through which possessive relations and relations of modification are established between nouns: compounding, the associative construction, the "dislocated possessive construction" (Kouwenberg 1994: 160) and a prepositional phrase construction involving the associative preposition *fɔ*. Compounding shares much of its functional space with the associative construction and both constructions are covered extensively in section §4.4.

Table 5.3 shows that the order of the participating NPs and forms of linkage are relevant for the way in which possessive relations and relations of modification are established. For ease of exposition, these relations are summarily referred to as "possessive" constructions and the participating NPs as "possessor" and "possessed", respectively:

In the associative construction, two nouns are juxtaposed, whereby the "possessor" (the modifier noun) modifies the "possessed" noun (the modified noun). Firstly, this construction is always employed when the possessor is instantiated in a possessive pronoun. 5.2 Noun phrase modification


Table 5.3 Possessive constructions

Secondly, associative constructions express various relations of modification, either exclusively or in complementarity with compounds (cf. §4.4). One relation of modification that is always expressed as an associative construction if the possessor is not a multiconstituent NP is a "measure/entity" relation (73). In such constructions, the modifier noun is the measure (*glás* 'glass') and the modified noun the entity measured (*watá* 'water'):

(73) *Wán glás watá.*

> one glass water

'One glass of water' [dj03do 053]

Unlike the associative construction, which typically instantiates a relation of modification between two noun phrases, the dislocated possessive construction typically serves to express a possessive relation. The possessor is therefore usually animate and human – the data contains no instance of a dislocated possessive construction involving an inanimate possessor.

In the dislocated possessive construction, a possessive pronoun that is co-referential with the possessor intervenes as a linker between the possessor and the possessed noun. With a singular possessor, the 3sg possessive pronoun *in* is therefore chosen, and with a plural possessor the 3pl possessive pronoun *dɛn*:


The dislocated possessive construction requires coreferentiality of the possessive pronoun and the possessor. Hence (76), which involves a 2sg person possessor, is ungrammatical:

(76) \* *Na* foc *yú* 2sg.indp *in* 3sg.poss *hós.* house Intended: 'It's your house.' [ne07fn 231]

#### 5 The nominal system

Recursive possessive relations can be expressed by the juxtaposition of possessive constructions, as in (77):

(77) *Na* foc *dé* there *a* 1sg.sbj *kán* pfv *sabí* know *mi* 1sg.poss *mamá* mother *in* 3sg.poss *papá* father *in* 3sg.poss *fámbul.* family 'It is there that I got to know my mother's father's family.' [fr03ft 044]

In the *fɔ*-prepositional construction, the possessed noun is followed by a prepositional phrase that contains a full noun functioning as a possessor (78) or modifier (79):


Unlike the dislocated possessive construction, the "possessor" in the *fɔ*-construction may be inanimate. This construction therefore typically expresses a relation of modification between a modified ("possessed") and a modifier ("possessor") entity. The construction may express various semantic roles including source (80) and material (81) (cf. §9.1.3 for a complete description of the semantic roles covered by *fɔ* 'prep'):


The *fɔ*-construction is also used to express part-whole relations in the idiomatic expression *pát fɔ* 'part of' (82) or in a partitive construction involving the determiner *sɔn* 'some' (83):


5.3 Quantification

The *fɔ*-construction is also employed to express a possessive relation in the same way as the dislocated possessive construction. There appears to be a stronger likelihood for the use of *fɔ*-prepositional constructions instead of dislocated possessive constructions when the possessed NP is complex and features more than one constituent. This is the case in the following example, in which the possessed noun *pikín* 'child' is additionally modified by the quantifier *lás* 'last':

(84) *Áfta* then *Miguel* name *Ángel* name *wé* sub *na* foc *di* def *lás* last *pikín* child *fɔ* prep *mi* 1sg.poss *antí.* aunt

'Then, there is Miguel Ángel who is the last child of my aunt.' [fr03ft 143]

Another factor that contributes to the use of the *fɔ*-construction is animacy. The resumptive pronoun in the dislocated possessive construction is typically coreferential with an animate, usually human possessor. Therefore, an inanimate possessor is best expressed through the *fɔ*-construction:

(85) *Na* loc *wán* one *prensa* press *internacional* international *wán* one *ministro* minister *fɔ* prep *Gabón* place *kán* pfv *tɔ́k* talk *sé* qot *dán* that *isla* island *na* foc *Gabón.* place

'In an international press [newspaper] a secretary of state of Gabon said that that island is [belongs to] Gabon.' [fr03ft 111]

### **5.3 Quantification**

Quantification is expressed through numerals, as well as a variety of relative, absolute, and negative quantifying expressions.

### **5.3.1 Numerals**

Pichi has a decimal numeral system. Cardinal numerals up to ten are listed in Table 5.4.

In the corpus, no numeral higher than seven was used in natural speech and no speaker except one could list numerals higher than 'ten' without fault. The Spanish numeral system is employed by all speakers and has largely replaced Pichi cardinal numerals above three (cf. §13.3.1 for additional details). Cardinal numerals occur in the prenominal modifier position (86) and may be used independently as pronominals (87). The repetition of cardinal numerals renders a distributive sense (88):


#### 5 The nominal system


Table 5.4 Cardinal numerals

(88) *Yu* 2sg *fít* can *kɛ́r* carry *dɛ́n* 3pl.indp *tú* two *tú.* rep 'You can carry them two by two.' [bo07fn 231]

Pichi has the three lexical ordinal numerals *fɔ́s* 'first' (89), *sɛkɔ́n/sɛ́kɔn* 'second' (90), and *lás* 'last' (91). The first two occur as attributive prenominal modifiers like other property items, while *lás* 'last' preferably occurs in quantifier compounds :


Ordinal numerals except 'first' may also be formed productively through the use of quantifier compounds involving the modifier noun *nɔ́mba* 'number' and a cardinal numeral as the head. Most people also use this construction to express 'second' (92):

(92) *Di* def *nɔmba-tú* number.cpd-two *pikín,* child *e* 3sg.sbj *kán* pfv *tɛ́l* tell *mí* 1sg.indp *di* def *sén* same *tín.* thing '(As for) the second child, she told me the same thing.' [ed03sb 027]

5.3 Quantification

The numeral *wán* has a number of functions that are derived from its cardinality sense. We have seen that it functions as an indefinite deteminer and a pronominal or nominal substitute (cf. 30–31). The adverbialising suffix *-wán* 'adv' is also etymologically related to the cardinal numeral *wán* (cf. also §5.2.1 and §5.4.4). The numeral *wán* also expresses adverbial meanings such as 'alone, single-handedly' with an emphatic nuance, as in (93). When used in this way, *wán* may modify a head noun post-nominally like a postnominal modifier, such as the focus particle *sɛ́f* 'self, emp'. However, *wán* does not modify full nouns by itself. It rather appears after an independent (emphatic) personal pronoun that is coreferential with the full noun in question (94) (cf. also 125–126):


### **5.3.2 Other quantifying expressions**

Non-numeral words express relational, absolute and negative quantification (cf. Table 5.5). Some of these words modify nouns in a way similar to determiners. One of them is the indefinite determiner *sɔn* 'some, a'. Some are only employed attributively with nouns (e.g. *hól* 'whole'). Yet others are only used as pronominals (e.g. *nátin* 'nothing').

The quantifier *ɔ́l* 'all' occurs with count and mass nouns alike. *ɔ́l* is encountered in a pre- (95), and postnominal position (96), yet without any effect on its quantificational properties:

(95) (*…*) *yu* 2sg *de* ipfv *bák* give.back *ɔ́l* all *di* def *mɔní* money (*…*)

'(…) you return all the money (…)' [hi03cb 184]

(96) *Di* def *pikín* child *ɔ́l* all *sé* qot *na* foc *mi* 1sg.poss *yón* own *bikɔs* because *a* 1sg.sbj *dɔ́n* pfv *pé* pay *mɔní.* money '(…) all the children are mine, because I have paid money [the dowry].' [hi03cb 196]

When *ɔ́l* appears immediately before the noun, it is most often found to modify generic nouns like *tín* 'thing', *tɛ́n* 'time', *pɔ́sin* 'person', *mán* 'human being', *plés* 'place', *sáy* 'side, place', and *stáyl* 'manner', as in the two following sentences (cf. §5.4.3. for a complete listing):

(97) *Ɔ́l* all *mán* man *kin* hab *lúk=an,* look=3sg.obj *yu* 2sg *go* pot *sí* see *wi* 1pl *nó* neg *go* pot *mít* meet *nó* neg *bɔ́di* body *na* loc *hós.* house 'Everybody watches it, you'll see, we won't run into anybody in the house.' [ma03ni 038]

#### 5 The nominal system


Table 5.5 Non-numeral quantifiers

(98) *Porque* because *na* foc *mí* 1sg.indp *mí* 1sg.indp *de* ipfv *prepara* cook *ɔ́l* all *tín.* thing 'Because it was me, I was cooking everything.' [dj03do 025]

Rather than seeing syntagmas like *ɔ́l mán* 'everybody' and *ɔ́l tín* 'everything' above as belonging to a word class termed "indefinite pronouns", they are best seen as ordinary NPs involving a quantifer and a generic noun, which may function as equivalents of nominal and adverbial indefinite pronouns in other languages. This analysis is supported by the fact that the generic nouns involved retain their full distributional potential as ordinary nouns; there are no signs of specialisation or grammaticalisation (cf. Haspelmath 1994: 182–183).

The occurrence of plural marking in the quantifier phrase in (99) also illustrates that a distinction between the meanings of 'everybody' and 'all persons/people' is irrelevant in Pichi, since genericity can be expressed through bare "singular" nouns and pluralmarked nouns alike (cf. §5.1.4):

(99) *Mí* 1sg.indp *sɛ́f,* emp *ɔ́l* all *pɔ́sin* person *dɛn* pl *kin* hab *áks* ask *mí* 1sg.indp *sé* qot *yu* 2sg *dɔ́n* pfv *bɔ́n?* give.birth 'As for me, all people ask me, "do you have a child?"' [fr03ft 152]

5.3 Quantification

*Ɔ́l* 'all' may quantify over temporal (100) and locative (101) expressions. This function may also be fulfilled by the attributive quantifier *hól* 'whole' (102). In general, the use of *hól* is, however, rare:


The quantifiers *ónli* 'only' and *sósó* 'only, abundant' have a distribution similar to *hól* above and may appear as prenominal, attributive modifiers to the noun. However, contrary to *hól*, both *ónli* and *sósó* may additionally function as quantifying adverbs. Compare the attributive (a) and adverbial (b) uses of *ónli* (103) and *sósó* (104) in the following two sentence pairs:

(103) a. *Di* def *ónli* only *lángwech* language *wé* sub *dɛn* 3pl *de* ipfv *tɔ́k* talk *fáyn* fine *fáyn,* rep (*…*) 'The only language that they speak really well (…)' [au07se 265] b. *Ónli* only *dɛn* 3pl *wánt* want *hía* hear *Panyá.* Spanish

'They only want to hear Spanish.' [au07se 211]

	- b. *Aa* intj *sósó* only *yandá.* yonder 'Ah, all the way over there.' [ge07ga 050]

In contrast, the relational quantifier *dásɔl* 'only' behaves like the universal relational quantifier *ɔ́l* 'all'. Hence, *dásɔl* may appear to the very left of the reference noun (105) or occur after the reference noun (106). Aside from that, *dásɔl* is used as a sentence adverb and clause linker (cf. §10.7.9):

(105) *Dásɔl* only *wán* one *smɔ́l,* small *wán* one *glas,* glass *yu* 2sg *fúlɔp=an.* fill=3sg.obj 'Only one small, one glass, you fill it up.' [dj03do 052]

#### 5 The nominal system

(106) *Pero* but *di* def *fíba* fever *bin* pst *kɛ́r* carry *wán* one *dé* day *dásɔl.* only 'But the fever lasted only one day.' [ru03wt 062]

The quantifier *ɛ́ni* 'every' quantifies over sets. It therefore has a distributive meaning and can only occur with singular count nouns (107):

(107) *Ɛ́ni* every *dé* day *dɛn* 3pl *de* ipfv *chɔ́p* eat *rɛ́s,* rice *ɛ́ni* every *dé.* day 'Every day they eat rice, every day.' [ed03sp 117]

The quantifier *grén* 'only, exactly' (< *grén* 'grain') only occurs in fixed collocations as a measure word with a preceding cardinal numeral, and followed by a count noun. Like *ɛ́ni* 'every', *grén* therefore quantifies over sets. The resulting quantifier compound functions as an attributive quantifier to the following noun *pikín*:

(108) *Na* foc *yu* 2sg *wan-grén* one.cpd-grain *pikín.* child 'That's your one and only [single] child.' [ge07fn 015]

The relative or partitive quantifiers *sɔn* 'some', *bɔkú* 'much', *plɛ́nte* 'plenty', and *smɔ́l* 'few, a bit' may quantify over count and mass nouns alike. NPs featuring one of these forms may be compared to an implicit standard of comparison, like *smɔ́l* 'few, a bit' in (109) and *sɔn* 'some' in (110):


When the standard of comparison is explicit, the quantifier participates in a partitive construction. Compare *bɔkú* 'much, many' in (111) which precedes the standard *mi kɔntrimán dɛn* 'my countrymen':

(111) *Bikɔs* because *a* 1sg.sbj *gɛ́t* get *bɔkú* much *mi* 1sg.poss *kɔntri-mán* country.cpd-man *dɛn* pl *wé* sub *dɛn* 3pl *húman* woman *kin* hab *dé* be.loc *fɔ* prep *Annobón.* place 'Because I have many of my countrymen whose wives are (usually) in Annobón.' [ed03sb 157]

#### 5.4 Pronouns

The negative quantifier *nó* 'neg, no' is preposed to its referent. This includes the inherently negative indefinite pronoun *nátin* 'nothing'. Additionally, negative quantifier phrases generally appear with support from verb negation. The resulting clause always yields a single negation reading (cf. §7.2.3 for more details). Compare the following sentence:

(112) *Nó* neg *nátín* nothing *nó* neg *dé* be.loc *pantáp=an.* on=3sg.obj 'Nothing is on it [the table].' [li07pe 011]

Some of the quantifiers covered can function as pronominals, as exemplified with *ɔ́l* 'all' in (113) (cf. Table 5.5 for a complete overview). However, a quantifier phrase featuring a generic noun (e.g. *ɔ́l tín* 'all thing' = 'everything') is usually preferred:

(113) *Ɔ́l* all *di* def *tín* thing *wé* sub *yú* 2sg.indp *an* and *dán* that *mán* man *bin* pst *gɛ́t,* get *ɔ́l* all *de* ipfv *lɛ́f* remain *fɔ* prep *dán* that *mán.* man 'All the things that you and that man had, all remains for that man.' [hi03cb 191]

### **5.4 Pronouns**

Pronouns may occur in the syntactic positions of common nouns. At the same time, they fulfil specific grammatical functions and are characterised by distributional preferences and restrictions.

### **5.4.1 Personal pronouns**

Four features are distinguished in the use of personal pronouns: person, number, syntactic (in)dependence, and case (cf. Table 5.6 below). The majority of "dependent pronouns" (with the exception of *mi* '1sg.poss' and *in* '3sg.poss') employed for subject case are also used for the expression of possessive case. Where the "possessive" column has no entry, the corresponding "subject" form is used. None of the forms in the "subject" and "possessive" columns are simultaneously employed as object pronouns.

In addition, there is an overlap in forms for the expression of object case. The "object & emphatic" columns are employed as object pronouns and emphatic pronouns at the same time. However, the 3sg pronouns *=an* and *ín* are suppletive allomorphs. The choice of either of the two forms is phonologically conditioned (cf. §3.2.5). One of these forms, i.e. the clitic *=an* '3sg.obj', is the only dependent object pronoun of Pichi.

The 2pl pronoun *una/unu* is normally invariable throughout the entire paradigm. Both forms are employed with any difference in meaning, but *una* is used in the vast majority of cases. Independent personal pronouns may undergo tonal derivation in order to participate in compound pronouns which express universal and dual number (cf. Table 5.7).

Dependent subject pronouns always occur in finite clauses together with verbs. They may only be separated from the verb by TMA markers, the negator, and preverbal ad-

#### 5 The nominal system


Table 5.6 Personal pronouns

verbs. Only independent personal pronouns may be focused (114), topicalised, modified by postposed elements, and conjoined by the coordinators *an* 'and' or *ɔ* 'or' (115):

(114) *Mí* 1sg.indp *gɛ́t* get *tú* two *brɔ́da.* brother

'I [emp] have two brothers.' [ro07fn 501]

(115) *Bɔt* but *di* def *gɛ́l* girl *nó* neg *kán* pfv *grí* agree *mék* sbjv *e* 3sg.sbj *gí* give *in* 3sg.poss *bɔ́y* boy *frɛ́n* friend *ɔ* or *di* def *pikín* child *ɔ* or *ín* 3sg.indp *sénwe,* self *e* 3sg.sbj *kán* pfv *rɔ́n.* run 'But the girl didn't agree to surrender her boyfriend or the child or herself (and)

A focused or topicalised independent pronoun may be followed by a resumptive dependent pronoun (116). This alternative is not very common in the data:

(116) *Mí* 1sg.indp *a* 1sg.sbj *nó* neg *gɛ́t.* get 'As for me, I don't have (one).' [ma03ni 041]

she ran (away).' [ed03sb 032]

Likewise, only independent personal pronouns occur under focus in cleft constructions involving the focus markers *na* 'foc' (117), and *nóto* 'neg.foc'. The example also shows the use of independent pronouns (i.e. *dɛ́n* '3pl.indp') as regular object pronouns (save the clitic *=an* for 3sg.obj):

(117) *E* 3sg.sbj *wás* wash *dí* this *klós* clothing *dɛn,* pl *e* 3sg.sbj *dráy* dry *dɛ́n,* 3pl.indp *nɔ́* intj *na* foc *mí* 1sg.indp *dráy* dry *dɛ́n.* 3pl.indp 'She washed the clothes, she dried them, no, it is me (who) dried them.' [ru03wt 034]

5.4 Pronouns

The independent form is also selected when a personal pronoun heads a relative clause (118) or is employed as a vocative (119):


Table 5.6 above shows that suppletion and grammatical tone are employed for case and number marking. The following two sentences exemplify the use of tone for pronominal inflection. Sentence (120) is a double-object construction. The object and emphatic pronoun *mí* is high-toned. Hence *mí* must be interpreted as the maleficiary object of the verb *tíf* 'steal', while *ordenador* 'computer' functions as the patient object:

(120) *Dɛn* 3pl *tíf* steal *mí* 1sg.indp *ordenador.* computer 'They stole a computer from me.' [ge07fn 169]

Conversely, (121) is a single object construction. The low-toned ponoun *mi* is a possessive pronoun to the noun *ordenador* 'computer' which functions as a patient NP to the verb *tíf* 'steal':

(121) *Dɛn* 3pl *tíf* steal *mi* 1sg.poss *ordenador.* computer 'They stole my computer.' [ge07fn 170]

The form *=an* '3sg.obj' is exclusively employed to express object case. It functions as a pronominal object to verbs, prepositions, and locative nouns. It is a clitic that forms a single phonological word with the immediately preceding verb, preposition or locative noun. The pronoun *=an* is sometimes employed indiscriminately for singular or plural reference. In such cases, it may be considered to function as a kind of transitivity or verbal agreement marker. In (122), *=an* is coreferential with the plural-referring pronominal *ɔ́l*:

(122) *Mí* 1sg.indp *sénwe* emp *a* 1sg.sbj *mɛ́n=an* care.for=3sg.obj *ɔ́l.* all 'I [emp] myself brought them [the children] all up.' [ma03ni 030]

Dependent possessive pronouns appear before the noun and may in turn be preceded by a demonstrative (123):

5 The nominal system

(123) *Pero* but *dís* this *una* 2pl *baf-rúm.* bath.cpd-room 'But this your [pl] bathroom [look how dirty it is].' [ge07fn 184]

Independent possessive pronouns are formed by placing a possessive pronoun to the left of the pronominal *yón* 'own' (124):

(124) *E* 3sg.sbj *sé* qot *a* 1sg.sbj *gó* go *mɛ́n* care.for *pikín* child *dásɔl* only *ɛf* if *a* 1sg.sbj *dɔ́n* prf *sí* see *yu* 2sg *yón.* own 'She said I will only care for a child when I have seen yours.' [fr03ft 159]

### **5.4.2 Modification of personal pronouns**

Subject and object pronouns can be modified by postposed quantifiers including numerals, focus markers and the topic marker, as well as nouns. Aside from that, the pronominal system may be extended through the formation of compound pronouns.

In (125), the pronoun *yú* '2sg.indp' is modified by *wán* 'one, alone'. *Wán* is semantically compatible with plural referents (126). In (127), the pronoun *ín* '3sg.indp' is modified by *dásɔl* 'only'. Note the obligatory use of independent (emphatic) pronouns with these quantifiers:


Sentence (128) provides an example of modification by a noun. The country name *Camerún* 'Cameroon' modifies the personal pronoun *una* '2pl' by apposition. The modifier noun does not take the pluraliser *dɛn* 'pl':

(128) *A* 1sg.sbj *sé* qot *bikɔs* because *una* 2pl *Camerún* place *una* 2pl *gɛ́t* get (*…*) 'I said because you Cameroonians, you have (…)' [ab03ay 151]

Compound pronouns feature a personal pronoun and the quantifiers *tú* 'two' and/or *ɔ́l* 'all'. They are formed by the same means as other compounds: The lexical H tone of the initial component(s) is erased and replaced by an L tone while the final component

#### 5.4 Pronouns

retains its lexically assigned H tone. Evidence that compounding is indeed at work in the formation of compound pronouns comes from (125–127) above. The presence of the postposed quantifiers *wán* 'alone' and *dásɔl* 'only' in these examples requires the use of H-toned emphatic personal pronouns. In contrast, the 3pl form of the personal pronoun in (129) below is L-toned, although the quantifier *ɔ́l* 'all' is in the same syntactic position as *wán* and *dásɔl* in (125–127) above.

The collocation *dɛn-ɔ́l* 'pl-all' may be employed in order to signal inclusivity of all referents. The use of a resumptive simplex dependent pronoun as in (129) is optional but very common:

(129) *Dɛn-ɔ́l* 3pl.indp.cpd-all *dɛn* 3pl *de* ipfv *salút* greet *dɛn* 3pl *sɛ́f.* self 'They are all greeting each other.' [dj07re 009]

A compound pronoun may also feature the numeral *tú* 'two' as the second component and thereby express dual number (130). Such dual compound pronouns are most frequently formed by additionally incorporating the quantifier*ɔ́l* 'all' into the compound (131). The data contains no trial compound pronouns formed with the numeral *trí* 'three':


Compound personal pronouns are employed in a regular and conventionalised way in order to express dual number with any of the three plural personal pronouns. Note the deletion of the H tones and replacement by L tones over all components of the dual object *wi-ɔl-tú* 'the two of us' save the last one (i.e. *tú* 'two', which bears its original lexical H tone) in (132):

(132) *Lɛk* like *sé* qot *dɛn* 3pl *de* ipfv *hía* hear *wi-ɔl-tú* 1pl.indp.cpd-all.cpd-two *wi* 1pl *de* ipfv *tɔ́k* talk *yét.* yet 'Like if they heard both of us still talking.' [au07se 217]

Examples (130) and (132) also show that dual pronouns are anaphorically referred to (i.e. through the resumptive pronouns *dɛn* '3pl' and *wi* '1pl', respectively) by making use of the corresponding plural pronoun.

The extension of the Pichi pronominal system by compounding is summarised in Table 5.7. Compound object, subject, and emphatic pronouns are identical. For possessive and resumptive pronouns, the regular plural pronouns are employed. Optional elements are in parentheses:

#### 5 The nominal system

Table 5.7 Compound personal pronouns


### **5.4.3 Indefinite pronouns**

In Pichi, the functional equivalents of indefinite pronouns are phrases involving generic nouns preceded by the quantifier and indefinite determiner *sɔn* 'some, a' as well as the quantifiers *ɔ́l* 'all', *ɛ́ni* 'every', and *nó* 'neg'. Table 5.8 provides an overview of 'some' and 'every' indefinites involving the generic nouns *pɔ́sin* 'person', *mán* 'man, person', *tín* 'thing', *sáy* 'side, place', *(káyn) stáyl* '(kind of) style', *tɛ́n* 'time', and *áwa* 'hour, time'. Some examples for their use are provided in (26–27) above as well as (97–99) above.

An extensive treatment of 'no' and 'any' forms, hence negative phrases with the functions of negative indefinites, is provided in §7.2.3.

A few characteristics of the NPs in Table 5.8 are worthy of note. Firstly, Pichi makes no difference between "some" indefinites used in affirmative and realis modality declarative sentences and "free-choice" indefinites (Haspelmath 1997: 48–52) of the "any" type.

Secondly there are a few idiosyncracies in the formation of indefinites: while *sɔn pɔ́sin* 'somebody' is more common than *sɔn mán*, *ɔ́l mán* 'everybody' is favoured over *ɔ́l pɔ́sin*; "manner" is equally often expressed as *sɔn stáyl* as it is involving the modifier substitute *káyn* 'kind'. Finally, note that 'sometimes' is expressed as *sɔn tɛ́n dɛn*, hence a plural NP while *sɔn.tɛ́n* is a lexicalised collocation functioning as an adverb with the meaning 'perhaps'. Also note that *tɛ́n* 'time' is a count noun, hence quantification with *ɛ́ni* 'every' renders the distributive meaning 'every time'.

### **5.4.4 Pronominals**

The pronominals*sɛ́f* 'self', *yón* 'own', and *nátin* 'nothing' occur in the syntactic positions of nouns. At the same time, they are characterised by a preference for specific environments or show distributional restrictions. The anaphoric pronominals *sɛ́f* 'self' and *yón* 'own' are employed to form independent reflexive and possessive pronouns and do not co-occur with determiners either. Instead, they are usually preceded by possessive pronouns. The negative indefinite pronoun *nátin* 'nothing' only occurs in negative clauses.

There is a transition from these more specialised pronominals characterised by restrictions to pronominals like *káyn* 'kind' and *wán* 'one', which favour specific environments,

5.5 Coordination


Table 5.8 Indefinite pronouns

to generic nouns like *mán* 'man, person', *sáy* 'place', *stáyl* 'manner', and *tɛ́n* 'time', which behave like other common nouns but fulfil important functions in the grammatical system of Pichi. For example, *káyn* 'kind' and *wán* 'one' may co-occur with a determiner or a prenominal modifier. *Káyn* appears as a head noun in question words and as a generic noun in the modifier or modified position of certain conventionalised collocations (e.g. *na wán káyn tín* 'foc one kind thing' = 'that's really something').

*Wán* also functions as a generic substitute for any other common noun, and in this function, it may be preceded by prenominal modifiers or determiners (e.g. *di ɔ́da wán* 'def other one' = 'the other one').

### **5.5 Coordination**

The form which is most commonly employed for signalling coordination between two noun phrases is the comitative preposition *wet* 'with' (133). The form *an* 'and' is also used to coordinate noun phrases (134) next to being employed as a sentential coordinator (cf. §10.3). However, most speakers have a clear preference for *wet* rather than *an*:

(133) *Lydia* name *wet* with *Junior,* name *na* foc *dɛ́n* 3pl.indp *a* 1sg.sbj *sabí.* know 'Lydia and Junior, it's them I know.' [fr03ft 134]

#### 5 The nominal system

(134) *Ɔ́l* all *di* def *tín* thing *wé* sub *yú* 2sg.indp *an* and *dán* that *mán* man *bin* pst *gɛ́t* get (*…*) 'All the things that you and that man had (…)' [hi03cb 191]

The disjunctive coordinator is *ɔ* 'or', which alternates in pronunciation between [ɔ̀] and [ò]. This variation in form is likely to be reinforced by the existence of the equivalent Spanish coordinator *o* 'or':

(135) (*…*) *wé* sub *a* 1sg.sbj *tínk* think *sé* qot *na* foc *judías* bean.pl *blancas* white.pl *ɔ* or *rɛ́s.* rice. '(…) of which I think that it is white beans or rice.' [ed03sp 122]

## **6 The verbal system**

Pichi verbs fall into three lexical aspect classes. The verbal system of Pichi is characterised by the use of preverbal particles, which modify the verb for tense, aspect, and modality. These three grammatical categories are interlocked in various ways, which transpire best when larger stretches of discourse are analysed. The system also includes numerous aspectual and modal auxiliary constructions. Verbs, and those denoting properties in particular, may be modified for degree in comparative constructions.

### **6.1 Lexical aspect**

Pichi verbs fall into three lexical aspect classes: stative, inchoative-stative, and dynamic. Most subclasses of inchoative-stative verbs may receive a stative or a dynamic interpretation in the right context, but the inverse is not the case, hence my use of the term "lexical" aspect. In this chapter and others, I employ "situation" as a cover term for events denoted by dynamic verbs as well as states denoted by (inchoative-)stative verbs and predicate adjectives. When a situation is construed as stative, it has no inherent boundaries, e.g. *e dé* '3sg.sbj be.loc' = 's/he/it exists'.

When a situation is construed as inchoative-stative, it encompasses the entry-intostate (inchoative), as well as the ensuing state (stative), e.g. *e chák* '3sg.sbj get.drunk' = 'he got drunk'. Since inchoative-stative verbs may also be read with a stative meaning, the preceding clause may also be translated as 'he is drunk'. Situations denoted by dynamic verbs are conceived of as being bounded; they have an inherent beginning and end (*wi chɔ́p* '1pl eat' = 'we ate') (Sasse 1991b).

I expressly avoid the terms telic (with an inherent endpoint) and atelic (without an inherent endpoint) (Comrie 1976: 44–51) in the description of lexical aspect. The telicatelic distinction blurs the boundaries between lexical aspect (as part of the meaning of the verb), grammatical aspect (expressed e.g. in the perfective-imperfective opposition), and clausal aspect (expressed e.g. by clausal transitivity and temporal adverbs), and is therefore of limited usefulness in this regard.

The inherent temporal structure of Pichi verbs co-determines the meanings that arise when aspect markers co-occur with a verb (cf. §9.2.3 for further valency-related effects of lexical aspect). Therefore, I apply two distributional criteria for delineating the three lexical aspect classes: firstly, co-occurrence with the imperfective marker *de* 'ipfv' and secondly, co-occurrence with the aspectual/phasal verb *bigín* 'begin' in an ingressive aspect auxiliary construction (Sasse 1991b: 8). The latter criterion is particularly useful, because the imperfective marker *de* 'ipfv' optionally intervenes between *bigín* and the following verb.

#### 6 The verbal system

The corpus contains only a handful of verbs that can be classified as stative with sufficient certainty. These are listed in Table 6.1 together with the semantic classes they belong to.


Table 6.1 Stative verbs

Stative verbs do not co-ocur with the imperfective marker *de* 'ipfv'. Secondly, they do not normally appear with the aspectual/phasal verb *bigín* 'begin (to)'. For most speakers, a clause like the following one is therefore ungrammatical:

(1) \* *A bigín (de) hébul dú=an.*

> 1sg.sbj begin ipfv be.capable do=3sg.obj

Intended: 'I began to be capable of doing it.' [to07fn 226]

The two modal verbs *lɛ́k* 'like' and *wánt* 'want' are ambivalent in their lexical aspect. I suggest that *wánt* is ambivalent between a dynamic and a stative sense, while *lɛ́k* vacillates between a stative and an inchoative-stative sense. Most of the time, these two verbs do not co-ocur with de 'ipfv' in imperfective situations. They sometimes do, however, and they are also attested in phasal constructions involving *bigín* 'begin':

(2) *Na* foc *ín* 3sg.indp *a* 1sg.sbj *bigín* begin *de* ipfv *lɛ́k=an.* like=3sg.obj 'That's when I began liking her.' [he07fn 228]

The class of inchoative-stative verbs includes three semantic classes that belong to the large group of labile verbs (cf. §9.2.3 for details): change-of-state verbs, locative verbs, and property items. It also includes two verbs of possession, two verbs of cognition, a verb of perception, and a verb denoting existence in time and space. The class of inchoativestative verbs is therefore much larger than that of stative verbs, which only has a few members.

#### 6.1 Lexical aspect

In this, I concur with analyses that posit a similar distribution of lexical aspect classes in other Afro-Caribbean English-lexifier Creoles (e.g. Winford 1993; Migge 2000). Table 6.2 below lists the relevant (groups of) verbs.


Table 6.2 Inchoative-stative verbs

All inchoative-stative verbs may potentially be interpreted as stative or inchoative in the absence of disambiguating information. This is for example the case when these verbs remain unmarked in basic intransitive clauses (cf. §6.3.1). However, such ambivalence between an ongoing state (stative) and an entry-into-state (inchoative) reading occurs with differing likelihood with the relevant semantic classes.

Within the group of labile verbs, property items are far more likely to be interpreted as stative than inchoative when left unmarked in an intransitive clause. In contrast, most change-of-state verbs and locative verbs may receive a stative and an inchoative interpretation with equal likelihood (cf. §9.2.3). This also holds for inchoative-stative cognition, possession, and perception verbs.

Inchoative-stative verbs are compatible with the imperfective marker *de* 'ipfv' (3) The use of *de* 'ipfv' with these verbs renders an inchoative meaning, which is in the present tense in relation to event time (cf. §6.3.4 for details). Likewise, inchoative-stative verbs may combine with the verb *bigín* 'begin'. The resulting ingressive aspect construction highlights the inchoative, entry-into-state meaning component of the verb (4):


'We began to get to know each other.' [ye07fn 019]

#### 6 The verbal system

The inchoative-stative posture verbs *sidɔ́n* 'sit (down)', *slíp* 'lie down, sleep' and *tínap* 'stand (up)' may co-ocur with the imperfective marker without necessarily acquiring the usual inchoative sense. These verbs appear to vacillate in their lexical aspect between an inchoative-stative and a dynamic sense. Consider the use of *slíp* 'lie, sleep' as an inchoative-stative verb in (5) and as a dynamic verb in (6):


The verb *tínap* 'stand (up)' may also be used as a dynamic verb. However, it is then also usually employed with the different meaning of 'begin to stand (of a toddler)'. Compare the following two uses of this posture verb:


A semantic specialisation of the inchoative vs. the dynamic meanings of the verb is also present with the verb *kɔmɔ́t*. When unmarked, it is left to context to disambiguate the meanings 'come from' (dynamic) and 'hail from' (inchoative-stative) from each other. This is illustrated in (9) and (10), respectively:


q=side 2sg come.from

'Where do you come from?' or 'Where did you exit [e.g. the market]?' [dj050e3 167]

A comparison of (10) and (11) shows that ambiguity does not arise once *kɔmɔ́t* is marked for imperfective aspect:

6.2 The TMA system

(11) *Yu* 2sg *de* ipfv *kɔmɔ́t* come.out *ús=sáy?* q=side 'Where are you coming from?' [dj05ce 170]

(12) *Mí* 1sg.indp *gɛ́t* get *dán* that *problema* problem *wet* with *bɔ́y* boy *dɛn* pl *wé* sub *dɛn* 3pl *kɔmɔ́t* come.out *Bata* Bata *nɔ́,* intj *sé* qot *'no* no *Pichi* Pichi *es* it.is *un* a *dialecto.'* dialect 'I have that problem with guys who are from Bata, right, [they] say "no, Pichi is a dialect [not a language]."' [au07se 219]

The data contains a large number of dynamic verbs from a wide range of semantic classes. Dynamic verbs may appear freely with the imperfective marker *de* 'ipfv' (13) and in ingressive auxiliary constructions featuring the aspectual/phasal verb *bigín* 'begin' (14). The use of the imperfective marker renders a progressive or habitual aspect reading with dynamic verbs. Note that labile inchoative-stative verbs may also be used as dynamic verbs in transitive clauses (cf. §9.2.3 for further details):


### **6.2 The TMA system**

Pichi has a core and a non-core system of tense-mood-aspect (TMA) marking. The core system is constituted by TMA particles which express central TMA notions. These particles (henceforth TMA markers) may be combined with each other, share phonological characteristics, such as monosyllabicity, and form a unit with the verb between which only a small group of preverbal adverbs may intervene. In the non-core system, auxiliary verbs express aspectual and modal notions as minor verbs in serial verb constructions. Besides TMA markers and auxiliary verbs, Pichi also makes use of complementisers in order to express modality.

The markers of the core TMA system and their linear order relative to each other and the verb root are provided in Figure 6.1. The figure shows that all TMA markers are found to the left of the root. The modal complementiser *mék* 'sbjv' is the only TMA marker found to the left of the dependent subject pronoun in a position occupied by clause linkers. It should also be borne in mind that factative TMA is achieved via the bare, unmarked verb, hence involves no overt marker:

#### 6 The verbal system


Figure 6.1 Ordering of TMA markers

The markers that express the two basic aspect categories of imperfective (i.e. *de* 'ipfv') and narrative perfective (i.e. *kán* 'pfv') are closest to the verb root. The marker *kin* 'hab, abl' has the same position when used in its habitual function as with its (marginal) function as a modality marker of ability. The same holds for *fɔ* when it instantiates conditional or obligative mood. When it occurs with the abilitive function it is glossed as 'abl'. There are co-occurrence restrictions for the expression of composite TMA categories (cf. Figure 6.2).

Table 6.3 presents the focal functions of TMA categories that are expressed when markers occur on their own. Factative TMA is included under all relevant categories in recognition of the multiple functions the unmarked verb plays in the TMA system. Factative TMA is indicated by a dash (—) in the column headed by "Marker".

Table 6.3 Functions of TMA markers


Combinations of the TMA markers listed above may render composite TMA categories. All attested combinations are listed in Figure 6.2 below. TMA markers follow

6.3 Aspect

the linear order established in Figure 6.1 where possible. Crossreferences to examples featuring uses of composite categories are provided in the first column.

In the corpus, the maximal number of TMA markers encountered in one clause is three (e.g. 103). The markers *bin* 'pst', *go* 'pot', and *kin* 'hab' are mutually exclusive. The imperfective marker *de* has the widest distribution and co-occurs with all markers except *fɔ*'cond'. In contrast, the narrative perfective marker *kán* has a far more restricted distribution. It only co-occurs with *bin* 'pst', and it does so only in two instances in the corpus. Iterative aspect, expressed by reduplication, is most compatible with the imperfective senses expressed by *de* 'ipfv' and *kin* 'hab', but it is also compatible with the tense/mood markers *bin* 'pst' and *gó* 'pot'. The imperfective aspect harmonises with the habitual aspect. When *de* 'ipfv' and *kin* 'hab' co-occur, there is no additional composite sense. The co-occurrence of subjunctive *mék* 'sbjv' and imperfective *de* 'ipfv' is very rare in the corpus.

The markers *nɛ́a* 'neg.prf' and *mɔs* 'obl' are not attested with any other marker and hence not included. Also unattested is the co-occurrence of *fɔ* 'cond' with the potential mood marker *go* 'pot'.

### **6.3 Aspect**

Sections §6.3.1 to §6.4.4 cover aspect marking by means of TMA markers and auxiliary verbs.

### **6.3.1 The unmarked verb**

Pichi employs factative TMA marking, a phenomenon well known from other languages in the region (cf. Welmers 1973: 348). When the unmarked verb occurs in an intransitive main clause and the clause contains no additional information that may have an effect on the interpretation of TMA, it acquires default interpretations of tense, aspect, and modality in accordance with its lexical aspect. The effect of factative TMA marking (or absence of marking) is summarised in Table 6.4. Bearing in mind that tense is relational, a factative marked (inchoative-)stative verb is interpreted as "present tense" with respect to event time not speech time (cf. §6.5.1).

As Table 6.4 shows, inchoative-stative verbs may receive an imperfective interpretation if focus is on the end-state, i.e. the stative meaning component of the verb. Alternatively, these verbs may receive a perfective interpretation, if focus is on the entry-intostate, i.e. the inchoative meaning component of the verb. The modality reading "realis" indicates that factative TMA in an intransitive clause does not normally render meanings associated with the irrealis domain, i.e. future tense or subjunctive and potential mood.

The unmarked verb also occurs in contexts that are removed from the immediate function of signalling aspect relations. Hence, the unmarked verb occurs in contexts of reduced finiteness (cf. §10.5.3). It occurs in the if-clauses of conditionals (cf. §10.7.11) and with non-initial verbs in clause chaining (cf. §11.4). Equally, verbs in subjunctive clauses

#### 6 The verbal system


Figure 6.2 Composite TMA categories

6.3 Aspect


Table 6.4 Default readings of factative TMA according to lexical aspect class

usually appear devoid of TMA marking (cf. e.g. §10.5.5). The unmarked verb also occurs in singular imperatives (cf. §6.7.3.3).

Perfective aspect via factative TMA marking with dynamic verbs yields an interpretation of the situation as bounded and terminated, hence past by default. Compare *pás* 'pass' and *gó* 'go' in (15):

(15) *A* 1sg.sbj *pás* pass *di* def *domɔ́t* door *bihɛ́n* behind *sáy* side *a* 1sg.sbj *gó* go *fɛ́n* look.for *sigá.* cigarette 'I passed through the entrance at the back, I went to look for a cigarette.' [ro05rt 016]

Since stative verbs have no inherent boundaries, the unmarked stative verb receives an imperfective, i.e. present tense or present state reading (16). Tense is relational in Pichi, so a stative verb like *wánt* 'want' is in the present tense in relation to "event time" (Chung & Timberlake 1985), which is past tense in this example:

(16) *Ɛ́ni* every *sáy* side *wé* sub *pɔ́sin* person *wánt* want *sidɔ́n,* stay *dɛn* 3pl *de* ipfv *sidɔ́n.* stay

'Anywhere that a person wanted to stay, they stayed.' [ma03hm 042]

Given the right context, all factative-marked inchoative-stative verbs may be interpreted as stative or inchoative (hence denoting entry-into-state). While *sabí* '(get to) know' in (17) may be interpreted as either stative or inchoative in the absence of disambiguating cues, an inchoative reading is forced upon the factative marked verb *évi* 'be heavy' in (18). This is due to the presence of a relational element, namely the temporal clause linker *bifó* 'before', which induces an implicit comparison with the prior empty state of the bag:

(17) *A* 1sg.sbj *sabí* know *sɔn* some *kápinta* carpenter *dɛn.* pl

'I know some carpenters.' or 'I got to know some carpenters.' [ro05fe 001]

#### 6 The verbal system

(18) *Bifó* before *wi* 1pl *rích* arrive *fɔ* prep *carretera* road *di* def *bolsa* bag *évi.* be.heavy 'Before we reached the road the bag had become heavy.' [ed03sb 198]

However, when labile inchoative-stative verbs occur in transitive clauses they automatically acquire a dynamic reading, in which case they receive a perfective, bounded and past tense interpretation like any other dynamic verb. Compare the meaning of the labile change of state verb *brók* 'break, be broken' in this example:

(19) *Dán* that *húman* woman *e* 3sg.sbj *brók* break *di* def *plét* plate 'That woman (she) broke the plate.' [au07se 006]

In addition, even in intransitive clauses, adverbials, and preceding tense-aspect marking in the same sentence, paragraph or text will usually disambiguate an inchoative from a stative interpretation. In (20), for example, factative marking with the dynamic verb *ték* 'take' leads to an entry-into-state interpretation of the following factative marked inchoative-stative verb *sidɔ́n* 'sit (down)':

(20) *A* 1sg.sbj *ték* take *di* def *trí* three *chía* chair *dɛn,* pl *dán* that *butaca* elbow.chair *ɔ́p* up *sáy,* side *mí* 1sg.indp *sidɔ́n* sit.down *dé,* there *e* 3sg.sbj *sidɔ́n* sit.down *dís* this *pát.* part 'I took the three chairs, that elbow chair up there, I [emp] sat down there, he sat

down on this side.' [ro05rt 006]

Beyond the expression of aspect taxis, the factative perfective aspect expresses conditional modality in the if-clause of conditionals with dynamic (21) and stative verbs alike (22):


Beyond that, factative marking is encountered in procedural texts in contexts that suggest a habitual reading. In (23), speaker (ge) asks speaker (dj) to explain how *ógi* 'corn porridge' is prepared. In the response in (23b), the dynamic verb *pút* 'put' acquires a habitual sense but remains bare.

(23) a. *Áfta* then *háw* how *fɔ* prep *mék* make *di* def *ógi?* corn.porridge 'Then how do you make the corn porridge?' [ge03do 050]

6.3 Aspect

	- def pot

'Now you can take, a, a small cup, right, you put it on the fire, inside the pot.' [dj03do 051]

### **6.3.2 Perfective and imperfective aspect**

The Pichi system of aspect marking represents a typologically widespread type in which the expression of perfective and imperfective aspect is not fully symmetrical (Dahl 1985: 69–102). The system features a general imperfective aspect marker *de*. Its function is to suppress the inherent boundaries of a situation (Breu 1985; Sasse 1991a,b). Although Pichi has other markers that encode imperfective notions (e.g. *kin* 'hab'), the marker *de* 'ipfv' alone may cover their functions, as well as those of others generally associated with the imperfective domain (e.g. future tense).

At the same time, the expression of perfective aspect is less uniform. On the one hand, perfective aspect is covered by factative TMA for dynamic verbs. Factative marking activates the inherent boundaries of dynamic verbs and thereby expresses perfective aspect by default. However, factative marked (inchoative-)stative verbs do not receive the corresponding perfective reading of entry-into-state by default. Instead, factative marking with stative verbs yields an imperfective reading, namely present or ongoing state, while inchoative-stative verbs are not automatically interpreted with an entry-into-state meaning either.

The narrative perfective marker *kán* 'pfv', rather than factative TMA, is therefore a better candidate for the expression of perfective meanings. As shown in section §6.3.3, the use of *kán* 'pfv' yields typically perfective meanings with all lexical aspect classes. The difference between the marking of dynamic verbs with perfective aspect by factative TMA and by narrative perfective aspect emerges most clearly in their uses in narrative discourse (cf. section §6.8).

Elements like the perfect marker *dɔ́n* and its negative counterpart *nɛ́a*, as well as ingressive, egressive and completive aspect auxiliaries also express various perfective readings. Table 6.5 provides an overview of the formal means of core perfective and imperfective marking and their readings in the three lexical aspect classes. The default tense interpretation of each aspect reading is provided in parentheses (prs = present tense, pst = past tense):

### **6.3.3 Narrative perfective**

The marker *kán* 'pfv' expresses narrative perfective aspect (cf. Jaggar 2006). It encodes perfective aspect and consequently, past tense by default. *Kán* 'pfv' occurs in salient [+high] foreground sequences of narrative discourse, while factative perfective marking is employed for less salient [-high] foreground sequences (cf. §6.8.1). The narrative per-

#### 6 The verbal system

Table 6.5 Perfective and imperfective readings according to lexical aspect class


fective marker therefore shares its functional space with factative TMA, and hence falls short of functioning as a general perfective marker. Although *kán* 'pfv' is homophonous with its lexical source verb *kán* 'come', there is no restriction on its co-occurrence with directional verbs, such as *gó* 'go' (24) or *kán* 'come' (25):


The marker *kán* 'pfv' is largely specialised to use in the foregrounded main line of narrative discourse. Here, it usually marks consecutive and bounded events denoted by dynamic verbs. In this function, the narrative perfective overlaps with perfective marking via factative TMA. But contrary to the latter, narrative perfective marking is employed in foregrounded sequences containing particularly salient, important information. *Kán* is preferred to factative perfective marking when new events unfold. In that, *kán* serves to highlight and focus the event denoted by the verb it refers to.

The three sentences below are an excerpt from a personal narrative. The speaker relates how she went to stay with her paternal uncle during a critical illness. This new information is provided in clauses (26a) and (b), and the relevant verbs (*gó* 'go' and *dé* 'be.loc') are marked by narrative perfective. In (c), the speaker reverts to factative TMA because the sentence now contains given information. Note that the same stative verb *dé* 'be.loc', which occurs with narrative perfective marking in the foregrounded sentence (b), appears with factative TMA in the backgrounded sentence in (c):

6.3 Aspect

	- b. *Mi* 1sg.poss *lét* late *papá* father *in* 3sg.poss *brɔ́da,* brother *a* 1sg.sbj *kán* pfv *dé* be.loc *na* loc *in* 3sg.poss *hós.* house 'My late father's brother, I came to be in his house.' [ab03ay 099]
	- c. *Na* foc *dé* there *a* 1sg.sbj *dé* be.loc *wán* one *hía* year *a* 1sg.sbj *nó* neg *fít* can *dú* do *nó* neg *nátin.* nothing 'It's there that I was (for) one year, I couldn't do anything at all.' [ab03ay 100]

The narrative perfective marker *kán*, even though specialised to narrative discourse, is a typical perfective marker (cf. Table 6.5). Irrespective of the lexical class of the verb, *kán* always activates the potential boundaries of a situation. With dynamic verbs, the situation is bounded and seen as a whole, hence past tense by default (cf. 26 above). The consistent meaning associated with the narrative perfective marker *kán* 'pfv' therefore contrasts with diametrically opposed meanings that arise through factative TMA marking with stative and dynamic verbs respectively.

The use of *kán* with stative (cf. *dé* 'be.loc' in 26b above) and inchoative-stative verbs (cf. 27 below) activates the initial boundary of the situation and focuses the ensuing state. Hence, it yields an inchoative (entry-into-state) meaning with a past tense interpretation in relation to event time. The different meanings that arise when a stative verb like *lɛ́k* 'like, love' is marked for perfective aspect and for factative aspect respectively, is shown by comparison of (28) and (36) further below.


Like factative TMA, the narrative perfective is sometimes employed, albeit rarely, in contexts other than aspect taxis. In (29), *kán* appears in the if-clause of a past conditional (cf. also 21). Maybe this usage reflects a tendency for *kán* to extend its function even further to that of a generalised perfective marker:

(29) *Ɛf* if *yu* 2sg *bin* pst *kán* pfv *bigín* begin *lás* last *semana,* week *yu* 2sg *bin* pst *fɔ* cond *dɔ́n* prf *fínis* finish *tidé.* today

'If you had begun last week, you would have finished by today.' [dj05ae 057]

6 The verbal system

### **6.3.4 Imperfective**

The general imperfective marker *de* 'ipfv' encodes various aspectual readings associated with the imperfective domain (cf. Table 6.5). Imperfective marking may express progressive aspect with dynamic verbs and present tense by default. Compare *smɛ́l* 'smell' and *kúk* 'cook' in (30):

(30) *A* 1sg.sbj *de* ipfv *smɛ́l* smell *di* def *sɛ́nt* scent *fɔ* prep *lɛk* like *háw* how *e* 3sg.sbj *de* ipfv *kúk* cook *plantí.* plantain 'I smell the scent of him cooking plantain.' [dj05ae 025]

Context may force a habitual interpretation on imperfective marked dynamic verbs. In (31), the habitual reading of *chɔ́p* 'eat' is signalled through the presence of the time adverbial *ɛ́ni dé* 'every day':

(31) *Ɛ́ni* every *dé* day *dɛn* 3pl *de* ipfv *chɔ́p* eat *rɛ́s,* rice *ɛ́ni* every *dé.* day 'Every day they eat rice, every day.' [ed03sp 117]

Certain human propensities and body states that may potentially be conceived as stative are expressed as dynamic verbs in Pichi. These include property items such *krés* 'be crazy' and *sík* 'be sick' (cf. §5.2.1 and §7.6.5 for more details). These verbs also take imperfective marking when progressive, continuous, or habitual aspect is to be expressed:


The imperfective marker does not normally co-occur with stative verbs. Yet *de* 'ipfv' is sometimes found with non-dynamic verbs. In (34) and (35) the inchoative-stative verb *gɛ́t* 'get, have' and the stative verb *lɛ́k* 'like' take the imperfective marker without acquiring an inchoative sense. This usage appears limited to modal verbs and verbs of possession like the following two:


6.3 Aspect

The conventional way of expressing imperfective aspect with (inchoative-)stative verbs is, however, by way of factative TMA. In (36) the stative verb *lɛ́k* 'like' remains unmarked, hence is imperfective by default:

(36) *Dɛn* 3pl *nó* neg *lɛ́k* like *pɔ́sin,* person *dɛn* 3pl *tú* too *badhát.* be.mean 'They don't like people, they're too mean.' [ma03hm 012]

In contrast, *de* 'ipfv' is regularly made use of with most inchoative-stative verbs in order to express an inchoative reading with a present tense interpretation in relation to event time. Compare the following two examples, as well as (72) below:


Besides its use for expressing aspectual relations, the functions of *de* reach into the domain of modality and overlap with those of the potential marker *go* 'pot'. The imperfective marker may express future tense in combination with an appropriate time adverbial (39). It can also express conditional modality in then-clauses and hypothetical statements contingent upon inferred conditions (40):


We also encounter the imperfective marker in environments characterised by reduced finiteness. Thus, *de* optionally intervenes between certain aspectual auxiliaries (cf. §6.4.1) and modal verbs and the verbs that follow them (cf. §10.5.3 for more details). Compare the following modal verbs *gɛ́fɔ* 'have to' (41) and *wánt* 'want' (42):


#### 6 The verbal system

### **6.3.5 Habitual**

The central function of the marker *kin* 'hab' is to express an imperfective reading of habitual aspect. Next to that, *kin* is also employed to express iterative aspect (cf. §6.3.6), and it marginally functions as a modal verb of ability (cf. 112). The marker either appears alone in preverbal position (43) or is optionally followed by the imperfective marker *de* if the reference verb is dynamic (44). There is no discernible semantic difference between *kin* and *kin de*. The optional co-occurrence of the two can be seen as a form of aspectual harmony or mutual reinforcement:


Since stative verbs are not normally marked by means of *de* 'ipfv', an important function of *kin* 'hab' is therefore to overtly mark stative verbs for habitual aspect. The habitual marker is therefore compatible with all lexical aspect classes. When used with (inchoative-)stative verbs, *kin* may additionally emphasise the habitual nature of the situation. Examples follow with the stative copula *dé* 'be.loc' (45) and the inchoative-stative verb *nó* '(get to) know' (46):


The habitual marker is also employed in generic statements, such as the following one:

(47) *Dɔ́g* dog *kin* hab *bɛ́t.* bite 'Dogs bite.' [dj07ae 371]

The habitual marker does not co-occur with the tense marker *bin* 'pst' or the potential mood and future tense marker *go* 'pot'. Like the imperfective marker *de* 'ipfv', *kin* 'hab' is itself unspecified for tense. Accordingly, sentence (46) above is translated as past habitual, because the time frame of the corresponding discourse context suggests so.

6.4 Aspectual auxiliaries

### **6.3.6 Iterative**

The reduplication of dynamic verbs yields an imperfective reading of iterative aspect when the reduplicated verb serves as the predicate of a clause. I refer the reader to section §4.5.1 for a detailed treatment of the phonology, morphosyntax, and semantics of reduplication.

Sentence (48) shows a typical context in which an iterative reading of reduplication arises. The reduplicated verb is accompanied by imperfective marking and co-occurs with the plural count noun object *nɔ́mba dɛn* 'numbers':

(48) *Wétin yu de chench-chénch nɔ́mba dɛn*

> what 2sg ipfv red.cpd-change number pl like.that

'Why do you constantly change (telephone) numbers like that?' [ye03cd 131]

*só?*

In a small number of cases in the corpus, the habitual marker *kin* also expresses iterative aspect by itself without additional reduplication. The speaker in the two consecutive sentences in (49) narrates how she repeatedly felt the temperature of her sick grandchild:

(49) a. *Wé* sub *a* 1sg.sbj *kin* hab *mék* make *só,* like.that *a* 1sg.sbj *nó* neg *de* ipfv *fíl* feel *hɔ́t.* heat

'Anytime I would do like this, I wouldn't feel heat.' [ab03ab 065]

b. *Pero* but *wé* sub *a* 1sg.sbj *kin* hab *tɔ́ch* touch *in* 3sg.poss *fút,* foot *in* 3sg.poss *hán* hand *dé,* there *na* foc *só* like.that *dɛn [kó:::l].*

3pl be.cold.emp

'When I would touch his leg, his hand there, that's how terribly cold they were.' [ab03ab 066]

### **6.4 Aspectual auxiliaries**

A specific set of verbs and adverbs functions as auxiliaries in constructions that express aspectual notions. These constructions involve the verbs *bigín* 'begin' (ingressive), *kɔmɔ́t* 'go out' (egressive), *fínis* 'finish' (completive), *wánt* 'want' (prospective), and the Spanishorigin verb *sigue* 'continue' (continuative). The expression of egressive and continuative aspect may also involve the preverbal adverbs *jɔ́s/jís* 'just' and *stíl* 'still', either in conjunction with the corresponding auxiliary verbs or alone.

These auxiliary verbs function as main verbs to complement verbs that are, in turn, specified for an aspect reading by the auxiliary verb. I analyse these aspectual (and modal, cf. §6.7.1) structures as involving complementation rather than verb serialisation. This is because the imperfective marker *de* may optionally intervene between the main and complement verb in some of these structures (cf. §10.5.1). When this the case, *de* 'ipfv' functions as a complementiser while emphasising the continuous nature of the situation denoted by the complement verb. Table 6.6 provides an overview of the functions of aspectual auxiliaries. Optional elements are in parentheses:

#### 6 The verbal system


Table 6.6 Functions of aspectual auxiliaries

Not included in Table 6.6 are constructions involving the verbs *sté* 'stay' and *lás* 'end up'. These verbs participate in adverbial SVCs with a certain degree of aspectual meaning (cf. §11.2.5). However, these constructions are more specialised in their meaning and not as grammaticalised to warrant being seen as aspectual auxiliaries in the same way as the ones covered in this section.

### **6.4.1 Ingressive**

The aspectual verb verb *bigín* 'begin' expresses ingressive aspect. The function of *bigín* as a transitive dynamic verb is exemplified in (50), where it is followed by the object NP *di wók* 'def work' = 'the work':

(50) *A* 1sg.sbj *bigín* begin *di* def *wók* work *wé* sub *yu* 2sg *dɔ́n* prf *gó.* go 'I began the work when you had gone.' [ro05de 024]

Ingressive aspect highlights the crossing of the initial boundary of a situation (51). When employed as an aspectual auxiliary, *bigín* may be immediately followed by a lexical verb (51) or optionally followed by the imperfective marker *de* (52), which stresses the continuous or extended nature of the transition to the situation denoted by the lexical verb:


6.4 Aspectual auxiliaries

The auxiliary *bigín* itself can be marked by tense-aspect markers like any other dynamic verb. In (53), *bigín* cooccurs with the narrative perfective aspect marker *kán* 'pfv'. The auxiliary *bigín* is not attested with stative verbs. But it may combine with inchoativestative verbs, in order to highlight the entry-into-state meaning of verbs from this lexical aspect class (54):


### **6.4.2 Egressive**

The verb *kɔmɔ́t* expresses egressive aspect. The egressive highlights the crossing of the terminal boundary of the situation described by the verb. This auxiliary construction is not attested with stative verbs. The egressive aspect neither carries a connotation of completion like the completive, nor does it establish a relation to reference time like the perfect. The auxiliary *kɔmɔ́t* may optionally be preceded by the preverbal adverb *jís/jɔ́s* 'just' and is immediately followed by the complement verb.

(55) *E* 3sg.sbj *tɛ́l* tell *mí* 1sg.indp *sé* qot *dán* that *papá* father *wé* sub *e* 3sg.sbj *jɔ́s* just *kɔmɔ́t* come.out *cobra* receive *in* 3sg.poss *mɔní* money *fɔ* prep *cacao,* cocoa *salút=an!* greet=3sg.obj 'He said to me "that elderly man that just received the money for his cocoa, greet him!"' [ed03sb 196]

The verb *kɔmɔ́t* has various meanings ranging from more lexical to more grammatical (cf. e.g. uses as a copula verb in §7.6.2 and as a directional verb in motion-direction SVCs in §11.2.1). In the following sentence, *kɔmɔ́t* is used with its presumably focal spatial meaning of 'go/come out':

(56) *Di* def *gɛ́l* girl *kán* pfv *kɔmɔ́t* go.out *dé.* there 'The girl left there [that place].' [ed03sb 030]

In other instances, the meaning of *kɔmɔ́t* is intermediary between a spatial and a more grammatical sense. In (57), it is the presence of the locative question word *ús=sáy* 'where' that creates ambiguity between the literal and the egressive senses of *kɔmɔ́t*. In sentence (58), semantic ambiguity is produced by the presence of *wók* which may mean 'work' (the noun) or 'to work' (the verb). If the former translation is preferred, *wók* is analysed as the (source) object of *kɔmɔ́t*. With the latter translation *wók* is a complement verb:

6 The verbal system

(57) *Ús=sáy* q=side *yu* 2sg *kɔmɔ́t* come.out *chák* get.drunk *só?* like.that 'Where do you come from drunk like this?' [ye07fn 126]

The verb *kɔmɔ́t* may co-occur with any TMA marker compatible with its status as a dynamic verb. Compare its appearance with the habitual marker *kin* 'hab' in (58):

(58) *Wé* sub *e* 3sg.sbj *kin* hab *kɔmɔ́t* come.out *wók,* work *a* 1sg.sbj *kin* hab *mék=an* make=3sg.obj *so,* like.that *lɛk* like *háw* how *mún* month *fínis.* finish 'When he comes from work/ when he has just finished working, I do like this to him [stretches out hand], as soon as the month is over.' [ro05rt 042]

The synonymous and equally common adverbials *jís* and *jɔ́s* can express an egressive notion by themselves when they appear in the preverbal adverb position (59), and thereby be functionally equivalent to egressive *kɔmɔ́t*. The adverb *jís/jɔ́s* may be preceded by a TMA marker and be followed by the lexical verb that it modifies. Note the occurrence of resumptive imperfective marking in (60) (cf. also (1):


I analyse *jís/jɔ́s* as an adverb rather than a preverbal TMA marker or a verb since it occasionally also occurs in the sentence-initial adverbial position with no difference in meaning (61). The adverb *jís/jɔ́s* is also used with no temporal meaning at all (62):


6.4 Aspectual auxiliaries

### **6.4.3 Completive**

The verb *fínis* 'finish' expresses completive aspect. The use of *fínis* as a lexical verb with the meaning 'finish' is exemplified in (63):

(63) *Bɔt* but *dá* that *mɔní* money *de* ipfv *fínis* finish *kwík.* quickly 'But that money used to finish quickly.' [ed03sp 088]

The completive indicates the crossing of the terminal boundary of a situation and adds the nuance of completion. Compare (64):

(64) *E* 3sg.sbj *fínis* finish *bɛ́n* bend *di* def *písis* piece.of.cloth *fáyn.* fine

'She has finished folding the piece of cloth real nice.' [li07pe 043]

The completive may signal a thorough consumption of the subject by the situation (65). This is particularly so when *fínis* co-occurs with the perfect marker *dɔ́n* or with the perfect imperfective *dɔ́n de* 'prf ipfv':

(65) *Náw* now *a* 1sg.sbj *dɔ́n* prf *de* ipfv *fínis* finish *sém* be.ashamed *fɔ* prep *wɛ́r* wear *dán* that *sús,* shoe *ɛf* if *a* 1sg.sbj *bin* pst *nó* know *a* 1sg.sbj *fɔ* cond *kɛ́r* carry *ɔ́da* other *sús.* shoes

'Now I am really ashamed to be wearing those shoes, if I had known I would have brought other shoes.' [ma03hm 021]

The completive auxiliary *fínis* 'finish' also co-occurs with the narrative perfective marker *kán*:

(66) *Di* def *prɔ́blɛm* problem *dɛn* pl *dɔ́n* prf *tú* too *mɔ́ch,* much *kán* pfv *fínis* finish *tɛ́l= àn* tell=3sg.obj *sé* qot *"lɛ́f".* leave 'The problems had become too much, (I) then finally told him "leave".' [ma0313ni 035]

### **6.4.4 Continuative**

The Spanish-origin dynamic verb *sigue* 'continue' expresses continuative aspect. The continuative construction is usually encountered with dynamic verbs and inchoativestative verbs with inherently more dynamic meanings (i.e. with change-of-state verbs but not with property items):

(67) *A* 1sg.sbj *sigue* continue *plé* play *bɔ́l* ball *sóté* until *ívin* evening *tɛ́n.* time

'I continued playing ball until the evening.' [be07fn 189]

#### 6 The verbal system

Alternatively, the preverbal temporal adverb *stíl* 'still' may function as an auxiliary in its own right to express continuative aspect. Contrary to *sigue*, the adverb *stíl* is also found to modify stative verbs like the copula *dé* in (68):

(68) *Mi* 1sg.poss *gran-má* grand-ma *wet* with *mi* 1sg.poss *gran-pá* grand-pa *wé* sub *dɛn* 3pl *stíl* still *dé* be.loc *láyf,* life *dɛn-ɔ́l* 3pl.cpd-all *dɛn* 3pl *dé* be.loc *na* foc *Panyá.* Spain

'My grandmother and my grandfather, when they were still alive, they were all in Spain.' [fr03ft 038]

When *stíl* co-occurs with a dynamic verb, the verb is normally marked for imperfective aspect (69):

(69) *Ɛf* if *yu* 2sg *stíl* still *de* ipfv *smók,* smoke *yu* 2sg *go* pot *sík.* sick

'If you continue smoking, you'll be sick.' [ro05ee 041]

A negative continuative meaning is generally expressed by means of discontinous negation involving the degree and temporal adverb *mɔ́* 'again, more' as in (70):

(70) *E* 3sg.sbj *nó* neg *dé* be.loc *mɔ́.* more 'He's no longer (here/there).' [ye03cd 155]

Like the preverbal adverb *jís* 'just' (cf. §6.4.2), *stíl* may also be preceded by TMA markers. Also like the former adverb, the latter appears with resumptive imperfective marking (71):

(71) *E* 3sg.sbj *de* ipfv *stíl* still *de* ipfv *wáka.* walk 'He's still walking.' [dj05ae 050]

A gradual and inherently comparative nuance of the continuative aspect can be expressed by employing the quantifying adverb *mɔ́-ɛn-mɔ́* 'more and more' (72):

(72) *Dís* this *bɔ́y,* boy *ɛ́ni* every *dé* day *e* 3sg.sbj *de* ipfv *fáyn* be.fine *mɔ́-ɛn-mɔ́.* more-and-more 'This boy, everyday he is getting more handsome.' [ro05ee 046]

### **6.4.5 Prospective**

The lexical verb *wánt/wɔ́nt* 'want' participates in an auxiliary construction that expresses "prospective aspect" (Comrie 1976: 64–65), also referred to as "proximative" (Heine 1994: 36). The prospective aspect denotes imminence of a situation:

6.5 Tense

(73) *Layk* like *háw* how *dɛn* 3pl *wánt* want *kɛ́r* carry *yu* 2sg *na* loc *hospital* hospital *yu* 2sg *dɔ́n* prf *dáy.* die

'As they're about to carry you to hospital, you're already dead.' [ed03sb 100]

The modal readings of desire and intention (cf. §6.7.2.2) and the aspectual reading of prospective are related in their meanings. Hence the difference between modal and prospective *wánt* is not always clear-cut.

For example, (74) is uttered when the speaker looks at a photograph of a father and his daughter, who is very tall for her young age. A desire reading of *wánt* as 'want to' is conceivable if *lɔ́n* 'be long, tall' is seen as a property that can be controlled by the speaker (even if humorously). However, a prospective reading denoting imminence appears more reasonable. Note that the prospective aspect reading of *wánt* triggers an imminent entryinto-state interpretation of the inchoative-stative verb *lɔ́n* 'be long, tall':

(74) *E* 3sg.sbj *wɔ́nt* want *lɔ́n* be.long *lɛkɛ* like *in* 3sg.poss *papá.* father

'She's about to become as tall as her father.' [ma03fn 003]

### **6.5 Tense**

The tense system of Pichi is relational, and in principle, bipartite. There is only one form – the past marker *bin* 'pst' – which has the focal function of a tense marker. Past tense can be expressed by means of *bin* 'pst' with any verb irrespective of its lexical aspect. Next to the past marker, the narrative perfective marker *kán* 'pfv', factative marking and other perfective aspectual readings (i.e. perfect, egressive, and completive) express past tense by default.

In contrast, there is no single form to mark non-past tense. Non-past marking is taken care of by a variety of means, none of which exclusively serves the expression of tense. Hence, the potential mood marker *go* and the imperfective marker *de* express future tense next to their respective modal and aspectual functions. Present tense arises by default through imperfective marking, either via factative TMA with (inchoative-)stative verbs, or through overt marking by markers that express imperfective readings (i.e. *de* 'ipfv' and *kin* 'hab').

Table 6.7 summarises the overt and default basic tense readings that arise through the use of core TMA marking with the three lexical aspect classes. Non-basic, mixed tense-aspect readings (i.e. past/future perfect, past/future progressive) are taken up in the relevant sections (cf. also Table 6.5):

The following sections provide an overview of the general characteristics of tense marking in Pichi (§6.5.1) as well as the expression of past (§6.5.2), present (§6.5.3), and future tense (§6.5.4). The potential mood and future tense marker *go* is covered in §6.7.4.1 in the section on modality. In order to do justice to the workings of relative tense in Pichi (cf. §6.5.1), I use the labels "anterior", "simultaneous", and "posterior" interchangeably with "past", "present", and "future", respectively, where necessary.

#### 6 The verbal system


Table 6.7 Overt and default tense marking

### **6.5.1 Relational tense**

Tense is relational (or "relative") in Pichi. Overt or default tense is assigned in relation to "event time" (Chung & Timberlake 1985) rather than speech time. Relational tense manifests itself in two ways. Firstly, in complex sentences, a subordinate clause is assigned tense in relation to the tense value of the main clause, and there is no need for corresponding overt tense or mood marking in the subordinate clause. Hence, there is no *consecutio temporum* in Pichi.

In (75), the main clause is marked for past tense by *bin* 'pst'. The subordinate clause (which begins with *wé* 'sub'), although simultaneous with the main clause, is not also marked for past. Instead, the factative marked stative verb *dé* 'be.loc' is assigned present tense by default, hence it is interpreted as simultaneous to the main clause verb *sí* 'see':

(75) *A* 1sg.sbj *nó* neg *bin* pst *sí* see *mi* 1sg.poss *gran-má* grand-ma *wé* sub *e* 3sg.sbj *dé* be.loc *láyf.* life 'I didn't see my grandmother while she was alive.' [ro05ee 147]

In (76), the main clause is also marked for past tense. This time, the subordinate clause (which begins with *sé* 'qot') is posterior to the main clause. Posteriority is expressed via the use of the potential marker *go*. Yet there is no additional past tense marking in the subordinate clause, and indeed, it would be ungrammatical. This in spite of the fact that both the main clause and the subordinate clause are set in the past from the vantage point of the speaker. Hence, the event in the main clause, not speech time, is the reference point for the tense assignment of the subordinate clause:

6.5 Tense

(76) *A* 1sg.sbj *bin* pst *chɛ́k* check *sé* qot *rén* rain *go* pot *fɔ́l.* fall 'I thought it might rain.' [ma03hm 022]

A second manifestation of relational tense in Pichi is the absence of explicit tense marking whenever context offers enough information on tense anchoring. Contextual information may be provided by time adverbials as in (77). Here, *yɛ́stadé náyt* 'yesterday night' anchors time reference in the past. Consequently, the imperfective marked verb *kɔ́l* 'call' receives a present tense/simultaneous interpretation in relation to past tense anchoring. Further marking by *bin* is unnecessary, although possible (cf. §6.5.2):

(77) *Yɛ́stadé* yesterday *náyt* night *wé* sub *a* 1sg.sbj *de* ipfv *kɔ́l* call *yú,* 2sg.indp *yɛ́stadé* yesterday *náyt,* night *nɔ́* neg *dís* this *mɔ́nin* morning *náw.* now

'Yesterday night, when I was calling you, yesterday night, not this morning.' [hi03cb 083]

In (78), past tense reference is established through the adverbial *wán ívin tɛ́n* 'one evening' and the factative marked, perfective, hence past tense dynamic verbs *kɔmɔ́t* 'go out' and *gó* 'go'. The imperfective marked verb *rích* 'arrive' in the subsequent clause remains unspecified for tense and receives a simultaneous reading, once more in relation to the past tense anchoring provided by the preceding adverbial and factative-marked dynamic verbs:

(78) *Wán* one *ívin* evening *tɛ́n* time *a* 1sg.sbj *kɔmɔ́t* go.out *mɔ́,* more *a* 1sg.sbj *gó* go *wáka,* walk *wé* sub *a* 1sg.sbj *de* ipfv *rích* reach *na* loc *hós,* house *hía* hear *Djunais* name (*…*)

'One evening, I went out again, I went for a stroll, when I was arriving at the house, [I] hear Djunais [say that…].' [ro05rt 001]

### **6.5.2 Past**

Two types of past tense expression exist in Pichi. The principal means of expressing past tense by default are factative marking (cf. e.g. 15) and the use of the narrative perfective marker *kán* 'pfv' (cf. 24–25). With (inchoative-)stative verbs, factative TMA gives rise to present tense reference by default. This is illustrated in (79) with the stative verb *fíba* 'resemble' and the inchoative-stative verbs *lɛ́k* 'like' and *sabí* '(get to) know' in (79).

(79) *Mí* 1sg.indp *nó* neg *sabí,* know *e* 3sg.sbj *fíba* seem *sé* qot *e* 3sg.sbj *nó* neg *lɛ́k* like *tín* thing *dɛn* pl *fɔ* prep *súp.* soup 'I [emp] don't know, it seems that she doesn't like soupy things.' [ma03hm 059]

#### 6 The verbal system

If, on the contrary, pragmatic context suggests an inchoative reading of inchoativestative verbs, a past tense interpretation is also possible. The change-of-state verb *brók* 'break, be broken' has factative TMA in the following example. Without contextual information the clause *e brók* could mean either 'it broke' or 'it is broken'. However, in this example, factative past tense marking on the preceding dynamic verbs *ték* 'take' and *nák* 'hit' only allows the first translation of *di plét brók*:

(80) *E* 3sg.sbj *ték* take *di* def *háma,* hammer *e* 3sg.sbj *nák* hit *ɔntɔ́p* on *di* def *tébul,* table *di* def *plét* plate *brók.* break 'She took the hammer, she hit [it] on the table, (and) the plate broke.' [ra07se 023]

Factative-marked stative verbs have a default present tense reference in relation to event time. Hence past tense reference can only be established for stative verbs by means of explicit past tense marking (i.e. *via bin* 'pst') or by means of contextual cues in the clause. In (81), the time adverbial *dán tɛ́n* 'that time' anchors time reference in the past, so the stative copula *dé* is interpreted as simultaneous to this tense anchor:

(81) *Dán* that *tɛ́n* time *a* 1sg.sbj *dé* be.loc *fáyn.* fine 'That time, I was fine.' [ru03wt 024]

Secondly, past tense may be explicitly marked by means of the past marker *bin* 'pst', which encodes relational past tense. *Bin* is not obligatory in clauses with past reference. Instead, its use depends on discourse-pragmatic factors. The past marker is generally employed in temporally remote, backgrounded, orienting, and supportive sections of narratives. In this function, *bin* 'pst' is diametrically opposed to the narrative perfective marker *kán* 'pfv'. It should therefore come as no surprise that *bin* has a default imperfective reading next to its function as a past tense marker. Consider sentence (82):

	- pl manicure

'(As for) me, (when) I was there, I only made, this, I used to make fingers, manicure.' [ma03hm 055]

Sentence (82) above is part of an orienting section of a narrative and provides background information to a story. The stative copula *dé* and the first dynamic verb (marked by superscript as <sup>1</sup>*mek* 'make') are overtly marked for past tense with *bin*. Once the use of *bin* with these two verbs has anchored the sentence (and in fact, the entire following narrative) in the past, overt past tense marking is unnecessary with subsequent verbs as is the case with the second dynamic verb (marked by superscript as <sup>2</sup>*mek* in 82).

The fact that *bin* also incorporates imperfective aspect transpires in the TMA marking choices of the sentence. All three verbs denote situations simultaneous to each other, an

6.5 Tense

aspect relation that usually requires imperfective marking with dynamic verbs. However, <sup>1</sup>*mek* is only marked for past tense with *bin*, whereas <sup>2</sup>*mek*, which is devoid of past tense marking, must be marked for imperfective aspect via *de* in order to express simultaneity of the situation.

While past reference may be established by factative TMA alone with dynamic verbs, overt past tense marking is often encountered with stative verbs where the occurrence of the unmarked form would give rise to ambiguity. In (83), *wánt* 'want' is explicitly marked for past tense by *bin*, because the unmarked form would favour a present tense, simultaneous reading. The same holds for the copula verb *dé* in (84):

(83) *Mí* 1sg.indp *dú=an* do=3sg.obj *fɔséko* due.to *sé* qot *a* 1sg.sbj *bin* pst *wánt* want *hɛ́lp=an.* help=3sg.obj

'I [emp] did it because I wanted to help him.' [ro05ee 069]

(84) *A* 1sg.sbj *kán* pfv *kɔmɔ́t* go.out *na* loc *dán* that *hós* house *wé* sub *a* 1sg.sbj *bin* pst *dé.* be.loc 'I left that house where I had been.' [ab03ay 097]

*Bin* can also express past-before-past tense when specifying a situation that is set in the past. In (85a), perfect marking with the dynamic verb *dáy* 'die' anchors time reference in the past. The subsequent clause (85b) featuring the stative copula verb *dé* is marked for *bin.* Hence, the situation referred to by *dé* 'be.loc' is anterior to *dáy* 'die' in the preceding clause.


*Bin* marks past-before-past in the same way in (86). Here, *pás* 'pass' in (b) is anterior to the past tense point of reference provided by *sík* 'be sick' in (a). In this example, we once more witness relational tense at work:

(86) a. *Wán* one *dé* day *wán* one *pikín* child *bin* pst *de* ipfv *sík.* sick 'One day a child was sick.' [fr03cd 071] b. *A* 1sg.sbj *nó* neg *sabí* know *ús=káyn* q=kind *tín* thing *bin* pst *pás.* pass 'I don't know what had happened.' [fr03cd 072]

The past marker also plays an important role as a modal element. *Bin* is used as a conditional modality marker in the if- and then-clauses of past (counterfactual) conditionals (cf. §10.7.11).

#### 6 The verbal system

### **6.5.3 Present**

Present tense is not expressed by means of elements specialised to this function. Instead, present tense reference is established by default through a variety of means. Bare stative verbs (cf. 16) and in the appropriate context inchoative-stative verbs (17) are assigned present tense by default when marked for factative TMA. Present tense reference is also established with inchoative-stative verbs via the use of the imperfective aspect marker *de* (cf. 37) and with both lexical aspect classes by the use of the habitual marker *kin* (cf. 45–46). Dynamic verbs are assigned present tense by default when they appear with the imperfective marker *de* (cf. 30) and the habitual aspect marker *kin* (cf. 43).

### **6.5.4 Future**

Future tense may be expressed explicitly by means of the potential mood marker *go* 'pot'. The marker can be used indiscriminately with stative (87), inchoative-stative (cf. *máred* 'marry, be married' in 91 below) and dynamic verbs (88):

(87) *Mí* 1sg.indp *go* pot *bí* be *dɔ́kta.* doctor 'I'll be doctor.' [ro05ee 025]

(88) *Ín* 3sg.indp *go* pot *chɔ́p=an,* eat=3sg.obj *e* 3sg.sbj *nó* neg *gɛ́t* get *nó* neg *problema.* problem 'He [emp] will/would eat it, he has no problem whatsoever [with this kind of food].' [ro05rt 066]

The expression of future tense is part of a field of interrelated mood and tense-marking functions (cf. §6.7.4.1). I assume that the expression of epistemic possibility is a central function of *go*, which is reflected in the gloss 'pot'. Nevertheless, the function of *go* also leans strongly towards that of a future tense marker in certain contexts.

When a situation is set in a hypothetical frame, hence based on an inferred or explicit condition, the meaning of *go* is modal. When context provides no such frame, the meaning of *go* tilts towards a tense reading. This is particularly the case in the presence of time adverbials (e.g. *tumɔ́ro* 'tomorrow' in 89) or where an intention of the speaker may be deduced from context (90):


#### 6.6 Perfect

Relational tense marking in Pichi allows a future projection from a speaker's vantage point in the past without the tense or mood change characteristic of*consecutio temporum* in languages with absolute tense systems. In (91), the verb in the main clause is marked for past tense. The verb in the subordinate clause introduced by *sé* 'qot' is marked for future, not future-in-the-past:

(91) *A* 1sg.sbj *bin* pst *de* ipfv *chɛ́k* check *sé* qot *a* 1sg.sbj *go* pot *máred.* marry 'I was thinking that I would marry/get married.' [fr03ft 165]

The marker *go* may also combine with *de* 'ipfv' to form a future imperfective, as in (92) below, and with *dɔ́n* 'prf' to form a future perfect (cf. e.g. 101). *Go* may also precede any of the aspectual auxiliaries covered in section §6.4.

(92) *Dɛn* 3pl *tɛ́l* tell *mí* 1sg.indp *sé* qot *ɛf* if *a* 1sg.sbj *pút=an,* put=3sg.obj *a* 1sg.sbj *go* pot *de* ipfv *kéch* catch *Panya.* Spain 'They told me if I put it [the antenna], I will be receiving Spain.' [ma0313ni 047]

Other elements that may express future tense notions are the imperfective marker *de* (cf. e.g. 39) and the prospective auxiliary *wánt* (cf. e.g. 74.

### **6.6 Perfect**

The marker *dɔ́n* expresses the affirmative perfect, while the synonyms *nɛ́a* and *nɔ́ba* express negative perfect. The Pichi perfect is a hybrid category that expresses aspectual and temporal notions simultaneously. The perfect expresses the aspectual notion of completive aspect in combination with the temporal notion of relevance to event time.

The perfect is encountered with dynamic verbs, where it highlights the current relevance of the completed situation (93):

(93) *Di* def *aráta* rat *dɔ́n* prf *kɔmɔ́t* come.out *ínsay* inside *di* def *hól.* hole

'The rat has come out of the hole [it is outside now].' [ro05ee 085]

The combination of perfect marking with an inchoative-stative verb usually yields a resultant state interpretation (94).

(94) *Ɛ,* intj *dán* that *bɔ́y* boy *dɔ́n* prf *kɔ́t* cut *ó.* sp 'Hey, that guy is badly cut.' [dj05ce 226]

In combination with stative verbs, perfect marking may convey a sense of total affectation of the referent by the state. In (95), this sense is reinforced through the presence of the degree adverb *bád* 'extremely':

6 The verbal system

(95) *Dán* That *gál,* girl *e* 3sg.sbj *dɔ́n* prf *lɛ́k=an* like=3sg.obj *bád.* extremely 'That girl, he really loves her.' [bo07fn 232]

Perfect marking is asymmetrical in Pichi. The marker *dɔ́n* 'prf' may not appear next to the negator *nó* 'neg'. The negative affirmative marker is therefore in complementary distribution with the forms *nɛ́a* and *nɔ́ba*, which both function as negative perfect markers. Negative perfect marking often yields the meaning 'not yet':

(96) *E* 3sg.sbj *de* ipfv *fɔgɛ́t* forget *sé* qot *Rubi* name *nɔ́ba* neg.prf *chɔ́p.* eat 'He forgets that Rubi has not yet eaten.' [dj03cd 148]

The negative restriction on *dɔ́n* 'pfv' is suspended when it co-occurs with a tense or mood marker. In that case, the ordering rules applying to TMA markers forestall adjacency of the negator and the perfect marker. Examples follow with *bin* 'pst' (97) and *go* 'pot' (98):


The clause-final adverbial *yét* 'yet' may reinforce the negative perfect without contributing additional meaning (99). A negated factative marked verb in conjunction with *yét* (99) can by itself be functionally very similar to the negative perfect expressed by *nɛ́a/nɔ́ba*:


The perfect marker *dɔ́n* may be combined with other TMA markers. Compare the future perfect in (101) and the past perfect in (102):

(101) *Las* the.pl *cuatro* four *wi* 1pl *go* pot *dɔ́n* prf *dé* be.loc *dé,* there *mí* 1sg.indp *sɛ́f* emp *a* 1sg.sbj *wánt,* want *a* 1sg.sbj *gɛ́fɔ* have.to *gó* go *na* loc *hós.* house 'At four o'clock we will already be there, I myself want, I have to go home.' [ma 03ni 005]

6.6 Perfect

(102) *Di* def *tín* thing *wé* sub *a* 1sg.sbj *bin* pst *dɔ́n* prf *fɔ́s* first *sí* see *wé* sub *a* 1sg.sbj *bin* pst *dɔ́n* prf *tráy=an.* try=3sg.obj 'The thing that I had first seen when I had tried it.' [ed03sb 188]

With dynamic verbs, the combination of *dɔ́n* 'prf' with the imperfective aspect marker *de* renders a perfect progressive meaning. The combination of the notion of current relevance and progressivity in the marker sequence *dɔ́n de* 'prf ipfv' renders an emphatic imperfective with dynamic verbs. It signals that the situation designated by the verb is (already) in full course (103) or on the brink of unfolding (104). Note that the situation in (103) is set in the past, hence the sequence of the three TMA markers *bin dɔ́n de* in (103):


This perfect progressive sense is sometimes additionally reinforced by placing the marker sequence *dɔ́n de* before the completive auxiliary verb *fínis* 'finish':

(105) *Náw* now *a* 1sg.sbj *dɔ́n* prf *de* ipfv *fínis* finish *sém* be.ashamed *fɔ* prep *wɛ́r* wear *dán* that *sús,* shoe *ɛf* if *a* 1sg.sbj *bin* pst *nó* know *a* 1sg.sbj *fɔ* prep *kɛ́r* carry *ɔ́da* other *sús.* shoe

'Now I am completely ashamed to be wearing those shoes, if I had known I would have brought another (pair of) shoes.' [ma03hm 021]

Perfect marking plays an important role in narrative discourse. The marker *dɔ́n* appears in backgrounded, scene-setting and out-of-sequence discourse sections. Sentence (106) begins with an adverbial time clause. It provides background information to the subsequent main clause that is part of the foregrounded main line of the story:

(106) *Wé* sub *a* 1sg.sbj *dɔ́n* prf *jɔ́ch* judge *dɛ́n,* 3pl.indp *a* 1sg.sbj *sé* qot *tumɔ́ro* tomorrow *sénwe* emp *a* 1sg.sbj *de* ipfv *gó* go *mít* meet *in* 3sg.poss *mán.* man

'When I had judged [scolded] them, I said tomorrow [emp] I'm going to meet her husband. [ro05rt 023]

#### 6 The verbal system

### **6.7 Modality**

The modal system of Pichi employs functional elements to express mood, and lexical words to express various types of modality. In my classification of modality into the dynamic, deontic, and epistemic categories, I rely on Palmer (2001).

Pichi has two overtly marked major mood distinctions. The subjunctive mood is employed in the realm of deontic modality. The potential mood serves to express interrelated meanings in the domains of epistemic modality and tense. Two minor moods are the abilitive and obligative moods which are encoded in the preverbal elements *kin* 'abl' (which otherwise expresses habitual aspect) as well as *mɔs* 'obl' and *fɔ* 'prep'. Aside from that, modal verbs and adverbials encode various types of modality. In the Pichi modal system, a number of TMA markers, rather than a single one, therefore share the semantic space of irrealis modality. Subjunctive-indicative and potential-factual are the most general and most systematically applied mood distinctions. Besides that, the imperfective aspect marker *de*, factative TMA, and the past marker *bin* fulfil distinct functions in the modal system of Pichi.

### **6.7.1 Modal elements**

In Pichi, modality is instantiated in adverbs and particles, clause linkers, TMA markers, and modal auxiliary verbs. An overview of the inventory of modal elements according to the modal categories they express is provided in Table 6.8. Elements appearing in the same line co-occur in the corpus, with the exception of the sentential modal elements *ó* 'sp' and *sɛ́f* 'emp', which express assertion and may co-occur freely with other elements in the table.

Conditional modality has been included in the table for the sake of completeness and is covered separately in sections §10.7.11 and §10.7.12 on adverbial clauses and relations. Details on the subjunctive mood are provided in sections §6.7.3.3, §10.5.5, and §10.7.6.

### **6.7.2 Dynamic modality**

Dynamic modality is concerned with the existence of factors internal to the subject with respect to the completion of the situation denoted by the reference verb (Palmer 2001: 76). In the following, the dynamic modality categories of ability, desire and intention are covered. These categories are primarily expressed through modal auxiliary verbs.

#### **6.7.2.1 Ability**

Pichi has a three-way distinction of ability. The modal verb *fít* 'can' expresses ability in a general sense, but it does not normally cover mental ability (107):


167

#### 6 The verbal system

(107) *A* 1sg.sbj *nó* neg *bin* pst *fít* can *tɔ́k,* talk *bikɔs* because *a* 1sg.sbj *nó* neg *fít* can *tɔ́k,* talk *a* 1sg.sbj *kán* come *a* 1sg.sbj *de* ipfv *lúk* look *yú.* 2sg.indp 'I couldn't talk, because I couldn't talk, I came (and) was just looking at you.' [ed03sb 165]

The verbs *hébul* 'be capable' (108) and *mánech* 'be capable, manage' (109) are usually employed to express capacity rather than ability:


The modal verb *fít* 'can' may also express root possibility. It predicates the existence of general (usually social) circumstances that affect the ability of the person involved to perform the situation denoted by the reference verb. The subject of the following sentence has been put to shame by being caught committing a moral offence:

(110) *E* 3sg.sbj *nó* neg *fít* can *dú=an* do=3sg.obj *mɔ́.* more

'He can't do it again [he wouldn't dare do it again].' [ro05rt 041]

The verb *sabí* '(get to) know how to' is used to express mental or learned ability (111). Compare the uses of the modal auxiliaries *fít* and *sabí* with the reference verb *tɔ́k* 'talk' in (107) above and (111):

(111) *Di* def *mán* man *e* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *Panyá.* Spanish

'The man doesn't know how to speak Spanish.' [ye03cd 063]

The corpus features a single instance in which the habitual marker *kin* 'hab' is unequivocally used to express physical ability (112). The use of *kin* as a marker of abilitive mood is marginal and obsolescent. The abilitive *cum* habitual function is, however, still widely attested in Krio. Both functions of *kin* grammaticalised from the English ability modal *can*. The fact that the habitual function alone was retained in Pichi might suggest that continuing contact with English has reinforced the ability function in Krio, while absence of contact with English has led to the erosion of the abilitive sense and expression by the modal verbs *fít* 'can' and *hébul* 'be capable' alone in Pichi.

6.7 Modality

(112) *Bifó* before *a* 1sg.sbj *kin* abl *gráp,* get.up *a* 1sg.sbj *de* ipfv *sí* see *bíg* big *bíg* rep *fáya.* fire 'Before I could get up, I was seeing a huge fire.' [ab03ay 067]

#### **6.7.2.2 Desire and intention**

The modal verb *wánt* 'want' expresses the often indistinguishable notions of desire and intention (113). The verb *mín* 'mean' may also express intention (114). Note the exceptional modal use of the imperfective aspect in (114), in a complement clause introduced by *sé* 'qot' where one would usually find a subjunctive clause introduced by *mék* 'sbjv':


### **6.7.3 Deontic Modality**

Deontic modality is concerned with the existence of factors external to the subject which condition the completion of the situation denoted by the reference verb. The deontic category of obligation is expressed by means of the TMA marker *mɔs* 'obl', obligation and necessity by *gɛ́fɔ* 'have to' or the multifunctional element *fɔ* 'prep' alone. Permission is expressed through the verb *fít* 'can'. Aside from that, the expression of deontic modality is characterised by the use of the subjunctive mood. Directives as well as the entire range of manipulative-directive meanings covered by the complement-taking verbs listed in section §10.5.1 induce the use of subjunctive clauses introduced by the modal complementiser *mék* 'sbjv'.

#### **6.7.3.1 Subjunctive mood**

Subjunctive mood is instantiated in the modal complementiser *mék* 'sbjv' and the specific TMA marking properties of the subjunctive clause. Subjunctive mood appears in directive main clauses (cf. §6.7.3.3). It is also present in the subordinate clauses of deontic modality inducing main verbs (cf. §10.5), i.e. verbs whose meaning contains an element of causation, manipulation, proposal, desire and other affective nuances compatible with deontic modality. Thirdly, subjunctive mood occurs in purpose and consecutive clauses (cf. §10.7.6).

#### **6.7.3.2 Obligation, necessity, and permission**

Obligation denotes the existence of compelling factors in the social world. Both strong and weak obligation are most commonly expressed through the verb *gɛ́fɔ* 'have to' (115).

#### 6 The verbal system

Negative obligation is formed by standard negation of *gɛ́fɔ* and yields a prohibitive meaning (116):


The verb *gɛ́fɔ* (<*gɛ́t fɔ* 'get/have prep') is a lexicalised collocation also attested in Krio (Fyle & Jones 1980) and Cameroon Pidgin (Nkengasong 2016). It was probably calqued from English 'have to' in the protolanguage, but is also likely to have been reinforced by Spanish *tener que* 'have to'. The verb has the distribution of a monorphemic lexeme in contemporary Pichi. It may therefore be followed by *de* 'ipfv' in complement constructions.

Alternatively, Pichi employs the two obligative mood markers *fɔ* 'prep' (117) and *mɔs* 'obl' (cf. 122 below) in order to express obligation. The marker *fɔ* may express both weak and strong obligation. The function of *fɔ* extends further to uses as a TMA marker to indicate counterfactual mood in the then-clause of conditionals, cf. the first and second occurrence of *fɔ* in (117):

(117) *Ɛf* if *dán* that *pikín* child *bin* pst *tɔ́k* talk *trú,* true *dɛn* 3pl *fɔ* cond *púl* remove *dán* that *pikín,* child *dán* that *pikín* child *e* 3sg.sbj *nó* neg *bin* pst *fɔ* cond *dáy.* die

'If that child [girl] had told the truth, the child [foetus] would have been removed, (and) that child [girl] wouldn't have died.' [ab03ay 121]

Impersonalised purposive constructions like (118) are likely to be one point of departure for the occurrence of *fɔ* as a mood marker in finite clauses like (119). The various uses of *fɔ* as a clause linker form part of a web of interrelated functions of this element (cf. §10.2 for an overview):


6.7 Modality

The element *fɔ* 'prep' also appears with a directive tint in non-assertive contexts like direct (120) and indirect (121) questions featuring the question word *háw* 'how':


'Can you help me, I don't know how I should do it/how to do it.' [ro05de 020]

Certain characteristics speak for an analysis of *fɔ* as a TMA marker when it appears in the preverbal position in finite clauses. Like other TMA markers of Pichi, *fɔ* is monosyllabic and low-toned. Equally, it is subject to restrictions. Although *fɔ* 'prep' is attested together with *bin* 'pst' in order to express counterfactual conditional modality (cf. 117 above), it is not encountered with any other TMA marker – unlike modal verbs. Hence, we have *e go gɛ́fɔ pé* {3sg.sbj pot have.to pay} 'she'll have to pay' but not \**e go fɔ pé* {3sg.sbj pot prep pay}.

The same characteristics hold for the element *mɔs* 'obl', which also expresses obligative mood. However, the use of *mɔs* usually renders a strong obligation sense often coupled with a sense of internal compulsion (122). Generally, speakers do not accept the use of *mɔs* 'obl' in syntactic positions which would suggest a verbal status of this element either. For instance, like *fɔ* above, *mɔs* is not attested in conjunction with other TMA markers (123):


Prohibitive clauses featuring *mɔs* 'obl' are formed like regular negative imperatives without a 2sg personal pronoun (124):

(124) *Nó* neg *mɔs* obl *gó* go *dán* that *sáy!* side '(You) must not go to that place!' [ne 07fn 194]

Necessity may be differentiated from obligation by making use of the modal verb *níd* 'need (to)' in affirmative (125) and negative (126) clauses. This modal auxiliary can be employed with same and different subject complement clauses in accordance with the pattern outlined in examples (38–40):

6 The verbal system


Permission is expressed by way of *fít* 'can', a causative/permissive construction involving *lɛ́f* 'leave, allow' (cf. 294 for details) or the main verb *grí* 'agree, allow' and a complement clause (128) (cf. also 35). Note the presence of the imperfective marker *de* in the subjunctive clause in the second example:


'So that's why she allows me to give her a small amount all the time.' [ma03hm 061]

#### **6.7.3.3 Directives**

Directives impose conditions of obligation on the addressee. The central form for expressing this modal category is the modal complementiser and subjunctive marker *mék*. The subjunctive marker may be employed to express directives throughout the entire person-number paradigm, which renders the modal categories traditionally referred to as imperative (2nd person directives) (129) and jussive (1st and 3rd person directives) (130– 131). The addition of the sentence final particle *ó* gives directives an admonitive tinct (129):


6.7 Modality

The subjunctive marker also introduces cohortatives (1st person plural invitations) (139) and optatives (1st, 2nd, 3rd person wishes):

(131) *tín* thing *fɔ* prep *fɔ́s* first *tɛ́n* time *mék* make *e* 3sg.sbj *dé,* be.loc *bikɔs* because *pípul* people *de* ipfv *kán* come *fɔ* prep *kán* come *sí=an.* see=3sg.obj

'(The) thing of the past, let it be, because people come to see it.' [hi03cb 068]

Subjunctive clauses must be employed for all directives except 2sg and 2pl imperatives. With imperatives, subjunctive clauses are optional. There appears to be no difference in meaning between bare and subjunctive marked imperatives. However, singular imperatives must be expressed by the bare verb without a personal pronoun if subjunctive marking is absent (132). Conversely, 2pl imperatives take the corresponding personal pronoun (133):


Negative imperatives (prohibitives) are formed by placing the negator *nó* before the verb (134) or by employing a negative subjunctive clause (135):


All other (i.e. 1st and 3rd person) directives may only be negated by means of a negative subjunctive clause (136):

(136) *Mék* sbjv *e* 3sg.sbj *fɔdɔ́n* fall *ínsay* inside *di* def *glás,* glass *mék* sbjv *e* 3sg.sbj *nó* neg *fɔdɔ́n* fall *na* loc *grɔ́n!* ground 'Let it flow into the glass, don't let it flow onto the floor!' [dj05be 170]

Sequences of imperatives are frequent in discourse. Here, the final verb must be marked for subjunctive mood, while preceding verbs may optionally remain bare. In these circumstances, the subjunctive additionally functions as a marker of consecutive modality:

6 The verbal system

(137) *Tɔ́n=an* turn=3sg.obj *tɔ́n=an* turn=3sg.obj *mék* sbjv *yu* 2sg *nó* neg *para!* stop 'Stir, stir it, and don't stop!' [dj03do 058]

The verb *kán* 'come' (138) may be employed in a way that parallels the use of the subjunctive marker in syntactic position and function (139). However, this usage is restricted to cohortatives:


The force of imperatives can be attenuated. An example follows in (140) of a weakened imperative involving the idiom *a bɛ́g* 'please' and the adverbial *smɔ́l* 'a bit':

(140) *A* 1sg.sbj *bɛ́g,* beg *kán* come *yá* here *smɔ́l!* a.bit 'Please come here a bit [would you please come here?].' [ch07fn 233]

Alternatively, a directive may involve one of the politeness markers *dúya* 'please' (cf. 36 or *plís* 'please' it may be couched in a question featuring the modal verb *fít* 'can' (141), or be formed through circumlocution featuring the verb *tráy* 'try' (142):


### **6.7.4 Epistemic modality**

Epistemic modality serves the expression of a speaker's commitment to asserting a given situation. The epistemic notions of possibility, certainty and assertion are covered in the following four sections. Part of the expression of epistemic possibility accrues to the potential mood marker *go*, which is also employed to express future tense.

6.7 Modality

#### **6.7.4.1 Potential mood**

The central function of the TMA marker *go* 'pot' is the expression of potential mood, hence the epistemic notion of possibility. With this analysis, I follow Essegbey (2008), who analyses a functionally similar morpheme of Ewe as an instantiation of the potential mood. From this point of departure, the marker *go* 'pot' expresses additional related modal and temporal notions like future tense, conditional, hypothetical, and habitual.

The following sentence illustrates the modal use of *go* 'pot'. In the example, speaker (ge) explains what prompted her to leave her teenage daughter in Madrid instead of bringing her along with her to Malabo on vacation. Obviously, speaker (ge) is not making a prediction; this is corroborated by the presence of the experiential verb *fía* 'fear'. Rather, the verb *bɛlɛ́*'impregnate' is marked by *go* 'pot' in order to express an epistemic possibility:

(143) *A* 1sg.sbj *fía* fear *sé* qot *dɛn* 3pl *go* pot *bɛlɛ́* impregnate *mi* 1sg.poss *pikín* child *fɔ* prep *mí.* 1sg.indp 'I feared that my child might be impregnated (on me).' [ge05be 055]

In this example, the potential mood expresses an epistemic possibility, rather than a prediction, in a similar way:

(144) (*…*) *mék* sbjv *yu* 2sg *tɔ́n=an,* turn=3sg.obj *porque* because *bɔtɔ́n* bottom *go* pot *rós.* burn

'(…) turn it, because the bottom might burn.' [dj03do 055]

The marker *go* frequently occurs with the epistemic adverbs *sɔntɛ́n* 'perhaps' and *mebi* 'maybe' in order to indicate a future (145) or a present possibility (146):


Since *go* alone can express potential mood and future tense, the TMA marker sequence *go dɔ́n* 'pot prf' can indicate a future perfect (cf. 101) or a potential perfect. The latter use of potential mood produces a reading of inferred certainty (cf. also §6.7.4.3).

(147) *E* 3sg.sbj *go* pot *dɔ́n* prf *drɔ́ngo,* be.dead.drunk *e* 3sg.sbj *go* pot *dɔ́n* prf *slíp.* sleep 'He should be dead drunk, he should already be sleeping.' [ge07fn 088]

#### 6 The verbal system

Besides its use as a potential mood marker and future tense marker in predictions (cf. §6.5.3), hypothetical statements are among the most common contexts in which *go* 'pot' occurs. A common form of expressive communication in Pichi involves the use of emphatic speech and figurative language and is set within a potential (or hypothetical) modal frame.

The following discourse excerpt involves two speakers who hypothesise about the potential advantage of having a pair of sunglasses that would allow them to see people naked. The use of the linker *if* 'if' signals entry into the realm of potential modality (148a), which is repeatedly marked by *go* in (a), (c) and (e). Note the presence of other modal elements, such as *fít* 'be able, possible' in (a), the imperfective marker *de* instead of *go* in (d), and the use of the factative marked stative verb *wánt* 'want' with a potential meaning once this modal frame has been established (f):

	- b. *Eyé.* intj 'Good gracious.' [ye07ga 008]
	- c. *Dán* that *gafa,* glasses *yu* 2sg *go* pot *slíp* sleep *wet=an.* with=3sg.obj 'Those glasses, you would sleep with them.' [ne07ga 009]
	- d. *A* 1sg.sbj *de* ipfv *slíp* sleep *wet=an* with=3sg.obj *cuñado.* brother-in-law 'I would sleep with them brother.' [ye07ga 010]
	- e. *A* 1sg.sbj *go* pot *púl=an* remove=3sg.obj *na* loc *mi* 1sg.poss *yáy* eye *sé* qot *wétin?* what 'I would remove them [the sunglasses] from my eyes for what?' [ye07ga 011]
	- f. *A* 1sg.sbj *wánt* want *dé* be.loc *flipado* turned.on *ɔ́l* all *áwa,* hour *ɔ́l* all *áwa.* hour 'I would want to be turned on all the time, all the time.' [ye07ga 012]

Potential mood is also systematically exploited to render a habitual reading in narrative discourse anchored in the past (149) and in procedural discourse. Note the presence of the generic phrase *di dé wɛn* '(on) the day that' in (149), which tallies with the nonspecific meaning of the habitual sense of *go* in this example:

(149) *Di* def *dé* day *wɛ́n* sub *mi* 1sg.poss *mamá* mother *go* pot *gɛ́t* get *sɔn* some *faya-wúd* fire.cpd-wood *wé* sub *dɛn* 3pl *brók=an* break=3sg.obj *na* loc *fám,* farm *e* 3sg.sbj *go* pot *tɛ́l* tell *dɛ́n,* 3pl.indp *dɛn* 3pl *go* pot *gó* go *tót=an* carry=3sg.obj

6.7 Modality

## *fɔr=an.*

prep=3sg.obj

'On those days that my mother would get some fire wood that had been broken up at the farm, she would tell them (and) they would go and carry it for her.' [ab03ay 023]

#### **6.7.4.2 Possibility**

The epistemic notion of possibility may be expressed through the use of the potential mood and the epistemic adverbs *sɔntɛ́n* 'perhaps' (cf. 145 above) and *mebi* 'maybe' (cf. 146 above). Besides that, possibility can be signalled when the verb *fít* 'be able, be possible' functions as a modal auxiliary verb (150) or with an expletive subject and a fuller complement clause (151):


Possibility can also be expressed through a construction involving an expletive *fíba* 'seem' (152) or the adverb *sɔntɛ́n* 'perhaps' with or without potential mood marking (153):


#### **6.7.4.3 Certainty**

Inferred certainty, the firmest degree of assertion, can be expressed by way of inferral from obligation with *gɛ́fɔ* 'have to' as in (154). The potential mood marker *go* is also employed in this function, in particular in combination with *dɔ́n* 'prf' (cf. 147 above)):


#### 6 The verbal system

#### **6.7.4.4 Assertion**

The emphatic and focus particle *sɛ́f* 'emp' (cf. §7.4.2) and the sentence particle *ó* (cf. §12.2.4) function as general markers of assertion when they signal clausal focus. Other than that, the verb *trú* 'be true' may be employed as an adverbial, oftentimes repeated for additional force, in order to signal assertion:

(156) *Dɛn* 3pl *bɔ́n* be.born *na* loc *Corisco* place *trú* true *trú.* rep 'They were really born on [the island of] Corisco.' [to07fn 201]

Beyond that, constructions involving cognition verbs (e.g. *sabí* '(get to) know', *nó* 'know', *chɛ́k* 'think, check (out)', *tínk* 'think', *mɛ́mba* 'think, remember', and perception verbs (e.g. *sí* 'see', *hía* 'hear') by themselves also signal different degrees of certainty.

### **6.8 Tense, modality, and aspect in discourse**

In preceding sections, I have provided some examples on the functions of TMA markers in discourse. In the following, I explore these functions further by looking at extracts of narrative discourse. The two relevant, intimately connected discourse-pragmatic notions are sequencing, i.e. the ordering of events along the time axis (Hopper 1982), and grounding, i.e. the distinction between the narrative main line or foreground from the less salient, narratively subordinate background (e.g. Hopper & Thompson 1980; Longacre 1996; Youssef & James 1999).

The picture that emerges from the analysis of the functions of Pichi TMA markers in narrative discourse with respect to grounding and sequencing is presented in Figure 6.3. The distribution of TMA markers in Pichi narrative discourse suggests the existence of a grounding continuum. Figure 6.3 takes this into account by differentiating between a more [+high] and a less salient [-high] foreground, marked by the narrative perfective marker *kán* 'pfv' and the factative marked (hence perfective) dynamic verb, respectively. The feature [+/-sequence] denotes the property of TMA markers to signal successive and discrete events along the narrative time line. Temporal and aspectual characteristics are therefore collapsed in this feature. So [+sequence] typifies consecutive, bounded, and dynamic situations, which may not be reordered without changing the iconic temporal order of the narrative at the same time.

The feature [+/-deixis] allows differentiation between aspect markers without an explicit temporal reference and markers that encode time-deictic reference to a point outside of the predicate. These reference points are event time for *bin* 'pst' and *dɔ́n* 'pfv', and a hypothetical contingency for *go* 'pot' in habitual discourse.

### **6.8.1 Sequencing and grounding**

The beginning of narratives anchored in the past very often features the past marker *bin* 'pst' in the "orientation" section (cf. Labov 1972: 358) characterised by aspect marking

#### 6.8 Tense, modality, and aspect in discourse


Figure 6.3 Functions of TMA markers in narrative discourse

of the imperfective domain (hence imperfective and/or habitual aspect). In this, the past marker is true to its role as a device for backgrounding situations and contributing a sense of temporal remoteness. For similar observations on cognate forms of *bin*, see Winford (2000: 398) for Sranan and Pollard (1989: 63) Jamaican Creole. The marker *bin* 'pst' fulfills this dual function in the orientation section (157b–e) of the excerpt of a personal narrative below. The backgrounding function of *bin* 'pst' correlates with its default aspectual interpretation.

Sentences (157a–d) demonstrate that there is a strong tendency to conceive of situations marked by *bin* as unbounded, hence imperfective by default. The free variation between *bin* 'pst', the imperfective marker *de* 'ipfv', and the marker sequence *bin de* in (b)–(e) with dynamic verbs for the expression of backgrounded, unbounded, and overlapping situations demonstrates the functional similarity of the three marking options:

(157) a. *Dé,* there *ɛ́ni* every *káyn* kind *tín* thing *na* foc *mɔní,* money *yu* 2sg *fít* can *mék* make *ɛ́ni* every *káyn* kind *tín,* thing *yu* 2sg *go* pot *sí* see *mɔní.* money

> 'There, everything is money, you can do anything, you will earn money.' [ma03hm 054]

b. *Mi* 1sg.indp *bin* pst *dé* be.loc *dé* there *a* 1sg.sbj *bin* pst *mék* make *dásɔl,* only *dís,* this *a* 1sg.sbj *de* ipfv *mék* make *fínga* finger *dɛn,* pl *manicura.* manicure

'(When) I was there, I only used to do, I used to do fingers, manicure.' [ma03hm 055]

#### 6 The verbal system


'Apart from that, I used to work in this very job that I do (now), I did it there, too.' [ma03hm 057]

e. *So* so *a* 1sg.sbj *bin* pst *de* ipfv *gɛ́t* get *mi* 1sg.poss *mɔní* money *dé* there *pero* but *yá* here *al* at.the *contrario* contrary *nada.* nothing

'So I used to get my money there but here, on the contrary, nothing.' [ma03hm 058]

In its functions, *bin* 'pst' is therefore antipodal to the narrative perfective marker *kán* 'pfv' (cf. 160–162 below). Like the former, the latter also simultaneously encodes a tense (past tense) and an aspectual value (perfective), and thereby plays an important role in the organisation of narrative discourse. However, the marker *kán* 'pfv' occurs in the most salient, foregrounded sections of the narrative, while *bin* 'pst' appears in backgrounded, supportive, and orienting sections.

Temporal sequence can also be iconically encoded through the linear ordering of bare dynamic verbs as in the "complicating action" (Labov 1972) of the narrative in (158) below. The temporal interpretation of factative marked inchoative-stative verbs hinges on grounding. The inchoative-stative bare verb *slíp* 'lie down' (158)(c) receives an inchoative, dynamic reading as it is foregrounded and forced into sequence in the narrative main line:

(158) a. *E* 3sg.sbj *gó,* go *e* 3sg.sbj *wás* wash *di* def *klós* clothing *dɛn.* pl 'She went off, she washed the clothes.' [ru03wt 033] b. *E* 3sg.sbj *wás* wash *dí* this *klós* clothing *dɛn,* pl *e* 3sg.sbj *dráy* dry *dɛ́n,* 3pl.indp *nó* neg *na* foc *mi* 1sg.indp *dráy* dry *dɛ́n.* 3pl.indp 'She washed the clothes, she dried them, no, I dried them.' [ru03wt 034] c. *Pero* but *di* def *klós* clothing *dɛn* pl *slíp* lie.down *na* loc *dɔ́n* down *ó.* sp 'But the clothes came to lie on the ground.' [ru03wt 035]

6.8 Tense, modality, and aspect in discourse

d. *Mɔ́nin* morning *tɛ́n* time *wé* sub *a* 1sg.sbj *kán* come *lúk,* look *a* 1sg.sbj *de* ipfv *sí* see *sɔn* some *klós* clothing *dɛn,* pl *a* 1sg.sbj *nó* neg *de* ipfv *sí* see *mi* 1pl *yón* own *dɛn.* pl '(In the) morning, when I looked, I saw some clothes, (but) I didn't see mine.' [ru03wt 036]

In contrast, backgrounded and out-of-sequence stative and inchoative-stative verbs, whether bare or marked with *bin* 'pst', receive a stative reading. Sentence (159) below is an orientation section. The stative copula *dé* 'be.loc' has a stative reading in the sentence. The same holds true for the inchoative-stative verb *sidɔ́n* 'sit (down)'. It co-occurs with the past marker *bin* 'pst', which once more not only signals the presence of backgrounded information. The imperfective, unbounded reading of *bin* also resolves the potential ambiguity between an inchoative and a stative interpretation of *sidɔ́n* in favour of the latter:

(159) *Mí* 1sg.indp *bin* pst *dé* be.loc *na* loc *bích* beach *wé* sub *a* 1sg.sbj *bin* pst *sidɔ́n* sit.down *wet* with *mi* 1sg.poss *papá,* father *mi* 1sg.indp *bin* pst *dé* be.loc *na* loc *bích* beach *mɔ́nin* morning *tɛ́n* time *a* 1sg.sbj *gó* go *latrin* latrine *a* 1sg.sbj *gó* go *kaká* defecate (*…*)

'I [emp] was at the beach while I was sitting with my father, I [emp] was at the beach in the morning, I went to the latrine, I went to shit (…)' [ed03sb 171]

Both (inchoative-)stative and dynamic verbs can also be explicitly marked for [+sequence] by the narrative perfective marker *kán* 'pfv'. The boundary-activating function of *kán* propels verbs marked by *kán* into the temporally sequenced narrative main line irrespective of their lexical aspect. With (inchoative-)stative verbs, this invariably induces an inchoative reading. With dynamic verbs, both boundaries of the situation are activated. These two aspect readings, bounded for dynamic verbs and inchoative for stative verbs, make *kán* 'pfv' a typical perfective marker (cf. Sasse 1991b: 11–14), even if its use is specialised to narrative discourse in Pichi.

The orientation section in (160a–b) is followed by a complicating action section in (c), which contains the first foregrounded situation, the inchoative-stative verb *sabí* '(get to) know'. The verb is marked by *kán* 'pfv' and receives an inchoative reading:

(160) a. *Bueno,* good *mi* 1sg.poss *mamá,* mother *mi* 1sg.poss *gran-má* grand-ma *wet* with *mi* 1sg.poss *mamá,* mother *nɔ́,* intj *dɛn* 3pl *kɔmɔ́t* hail.from *na* loc *wán* one *pueblo* village *wé* sub *in* 3sg.poss *ném* name *na* foc *Basakato dé la Sagrada Familia* place

'Well, my mother, my grandmother and my mother, right, they hail from a village whose name is Basakato dé la Sagrada Familia.' [fr03ft 042]

#### 6 The verbal system

b. *Sɔn* some *tɛ́n* time *dɛn* pl *wi* 1pl *kin* hab *de* ipfv *gó* go *dé* there *sɛ́f* emp *fɔ* prep *gó,* go *bueno,* good *fɔ* prep *gó* go *visít* visit *nɔ́,* intj *fɔ* prep *pás* pass *vacaciones* holiday.pl *dɛn.* pl

'Sometimes we even used to go there in order to, well, in order to go visit, in order to spend our holidays.' [fr03ft 043]

c. *Na* foc *dé* there *a* 1sg.sbj *kán* pfv *sabí* know *mi* 1sg.poss *mamá* mother *in* 3sg.poss *papá* father *in* 3sg.poss *fámbul.* family

'That's where I got to know my mother's father's family.' [fr03ft 044]

The following extract illustrates the importance that *kán* 'pfv' has for organising the events of a paragraph with respect to narrative saliency. The verbs in (161a–d) are marked for perfective aspect due to the novel information they contain. Meanwhile, (161) (e) reiterates information already contained in (161) (c) and (d), therefore dispenses with perfective marking and is characterised by the presence of stative, narratively downshifted verbs:

	- b. *A* 1sg.sbj *kán* pfv *kɔmɔ́t* go.out *na* loc *dán* that *hós* house *wé* sub *a* 1sg.sbj *bin* pst *dé.* be.loc 'Then I left that house where I was.' [ab03ay 097]
	- c. *A* 1sg.sbj *kán* pfv *gó* go *na* loc *mi* 1sg.poss *ɔnkúl* uncle *in* 3sg.poss *papá* father *in* 3sg.poss *lét* late *brɔ́da.* brother 'Then I went to my uncle's father's late brother.' [ab03ay 098]
	- d. *Mi* 1sg.poss *lét* late *papá* father *in* 3sg.poss *brɔ́da,* brother *a* 1sg.sbj *kán* pfv *dé* be.loc *na* loc *in* 3sg.poss *hós.* house 'My late father's brother, I came to stay at his house.' [ab03ay 099]
	- e. *Na* foc *dé* there *a* 1sg.sbj *dé* be.loc *wán* one *hía* year *a* 1sg.sbj *nó* neg *fít* can *dú* do *nó* neg *nátin.* nothing 'It is there that I was for one year, I couldn't do anything.' [ab03ay 100]

The use of *kán* 'pfv' in (162) points to the role of the perfective marker in additionally highlighting narratively salient, [+high] foreground information. At the same time, less salient [-high] foreground occurs in the unmarked form of the verb (i.e. the two occurrences of *sɛ́n* 'send' in 162b), which incidentally coincides with a backgrounding passive construction, another downshifting device (i.e. *dɛn sɛ́n mí* (…) 'I was sent (…)'). The introduction of information considered more relevant, and with it the resumption

of the main line, then once more features the perfective marker *kán* 'pfv' with the verb *lɔs* 'lose':

	- b. *E* 3sg.sbj *de* ipfv *ánsa* answer *mí,* 1sg.indp *a* 1sg.sbj *sɛ́n* send *mɔní,* money *dɛn* 3pl *sɛ́n* send *mí* 1sg.indp *sɔn* some *portamonedas* wallet *bɔt* but *e* 3sg.sbj *kán* pfv *lɔs.* lose

'He used to reply to me, I sent money (and) I was sent a wallet but it got lost.' [ed03sb 207]

Like the imperfective marker *de* 'ipfv', the habitual marker *kin* 'hab' marks [-sequence] situations that furnish the background frame for the narrative main line. Next to the habitual marker *kin*, the potential marker *go* also fulfils an important role in expressing habituality with respect to routine procedures. This is shown in the following extract that relates the effect zombification has on its victims. Consider the prolific use of *go* 'pot' to signal (potential) habituality set in a hypothetical frame:

	- d. *Ɛf* if *dɛn* 3pl *go* pot *gó* go *bɛ́r* bury *yú,* 2sg.indp *dɛ́n* 3pl.indp *sénwe* emp *go* pot *gó* go *na* loc *dán* that *bɛ́rin.* burial If they go to bury you, they themselves will go to that burial.' [ed03sb 102]

#### 6 The verbal system

	- die

'In the night, at twelve o'clock they'll come and remove you (and) you haven't died.' [ed03sb 103]

f. *Dɛn* 3pl *go* pot *rɛdí* prepare *yú* 2sg.indp *dɛn* 3pl *go* pot *mék* make *lɛk* like *háw* how *dɛn* 3pl *de* ipfv *mék* make *fɔ* prep *wích,* sorcery *dɛn* 3pl *ték* take *yú* 2sg.indp *dɛn* 3pl *pút* put *yú* 2sg.indp *na* loc *avión* plane *dɛn* 3pl *sɛ́n* send *yú* 2sg.indp *fɔ* prep *ɔ́da* other *kɔ́ntri* country *yu* 2sg *gó* go *wók* work *mɔní.* money

'They'll prepare you the way it's done with sorcery, they'll take you, put you into a plane and send you to another country (and) you'll go earn money (for them).' [ed03sb 104]

Foregrounded sections of sequential action conceived of as particularly tightly-knit may feature clause chaining (cf. §11.4). In chained clauses, tense, aspect and mood marking is overtly expressed with the first initial verb(s) in order to provide orientation and grounding. Subsequent clauses remain bare and occur one after the other without an intonation break or intervening clause linkers. Chained predicates invariably feature resumptive personal pronouns; the subject is repeated with each verb in the series. Verbs that participate in clause chaining are always dynamic, and are hence part of the foregrounded narrative main line. Sequences of chained clauses can be found in (163a) (*yu kán yu dríng*), (e) (*dɛn púl yú*), and (f) (beginning with *dɛn ték yú* until the end of the paragraph).

After a brief interruption by a listener comes a transition to habitual marking via *kin* 'hab' in (164) below. Extracts (163–164) lay bare the difference between habitual discourse centred on *go* 'pot' and *kin* 'hab', respectively. The expression of habituality with *go* rests on the prior establishment of a hypothetical contingency. Hence, paragraph (163) is interlaced with elements characteristic of irrealis modality. The extract begins in (163a) with a conditional clause serving as the referential frame for the *go*-marked discourse up to (f); another conditional clause follows in (d), and the habitual, generic use of *go* coincides with the impersonalised, non-referential use of the 2sg personal pronoun *yu*.

In contrast herewith, habitual discourse centred on *kin* in (164) is introduced by the phrase *e kán bí sé* '3sg.sbj pfv be qot' = 'it came to pass that', a conventionalised opening formula employed in personal accounts and other types of factual narrative. The subjectively high truth value of (163) is underlined by the closure in (g) *a dɔ́n sí, yɛ́s* {1sg.sbj prf see yes} = 'I have seen (this before), yes'.

(164) a. *Dɛn* 3pl *go* pot *púl* remove *dán* that *mán,* man *a* 1sg.sbj *sé* qot *e* 3sg.sbj *kán* pfv *bí* be *sé* qot *dɛn/* 3pl *pípul* people 6.8 Tense, modality, and aspect in discourse

*dɛn* pl *kɛ́r=an,* carry=3sg.obj *dɛn* 3pl *lɛ́f* leave *di* def *cadáver* corpse *dɛn* 3pl *rɔ́n.* run 'They'll remove that man, I say, it came to pass that they/ people carried him, they left the corpse and run away.' [ed03sb 107] b. *A* 1sg.sbj *tínk* think *sɔn* some *fámbul* family *dɛn* pl *wé* sub *dɛn* 3pl *kin* hab *sí* see *sé* qot *dí* this *mi* 1sg.poss *fámbul* family *dé* there *lɛk* like *háw* how *e* 3sg.sbj *dáy* die *e* 3sg.sbj *nó* neg *kɔrɛ́t.* be.correct 'I think some families, when they see that this my family member there, how he died that's not correct.' [ed03sb 108] c. *Dɛn* 3pl *kin* hab *gó* go *na* loc *bɛrin-grɔ́n* burial.cpd-ground *wet* with *gɔ́n.* gun 'They go to the cemetery with a gun.' [ed03sb 109] d. *A* 1sg.sbj *hía* hear *sé* qot *Bata* place *dɛn* pl *kin* hab *sút* shoot *yú.* 2sg.indp 'I heard that the mainlanders (even) shoot you.' [kw03sb 110] e. *Dɛn* 3pl *kin* hab *sút.* shoot 'They shoot (you).' [ed03sb 111] f. *Wé* sub *dɛn* 3pl *sút* shoot *di* def *pɔ́sin,* person *di* def *pɔ́sin* person *kin* hab *sék.* shake 'When they've shot the person, the person shakes.' [ed03sb 112] g. *A* 1sg.sbj *dɔ́n* prf *sí,* see *yɛ́s.* yes 'I have experienced (this), yes.' [ed03sb 113]

The perfect tense-aspect marker *dɔ́n* 'pfv' is employed with [-sequence] situations that digress from the linear narrative main line. The use of this marker prepares terrain for foregrounded and bounded action, a role reserved for functionally equivalent forms in many languages (cf. Anderson 1982; Li et al. 1982; Slobin 1994). The perfect marker may therefore play an important role in signalling the anteriority and causality of a situation immediately relevant to the situations of the narrative main line. Consider (165), which is an excerpt of a narrative about a woman who wants to divorce her husband and is obliged by tradition to pay back the dowry. In this excerpt, the perfect aspect lends itself to use in an "embedded abstract" (Labov 1972), which often occurs in a well-formed Pichi narrative. Through this technique, a speaker steps out of the story line, condenses and adds on to previous foreground material in a series of perfect marked verbs as in (a–c).

Note that the speaker employs some features characteristic of Nigerian (Pidgin) English, since she lived in Nigeria for some time (i.e. *(dé) yɔ́ng* 'be young', *dé frɛ́sh* 'be fresh', *sɛventín* 'seventeen', *etín* 'eighteen', *twɛ́nti* 'twenty', and *yíɛs* 'years'):

#### 6 The verbal system

(165) a. *Yu* 2sg *yɔ́ng,* be.young *yu* 2sg *jɔ́s/* just *sɔntɛ́n* perhaps *yu* 2sg *gɛ́t* get *sɛventín,* seventeen *etín* eighteen *yíɛs* years *ɔ* or *twɛ́nti,* twenty *yu* 2sg *dé* be.loc *yɔ́ng* young *yu* 2sg *dé* be.loc *frɛ́sh,* fresh *yu* 2sg *dɔ́n* prf *kɔmɔ́t,* go.out *yu* 2sg *dɔ́n* prf *bɔ́n* give.birth *fó* four *pikín,* child *yu* 2sg *dɔ́n* prf *bɔ́n* give.birth *fáyf,* five *yu* 2sg *dɔ́n* prf *bɔ́n* give.birth *tɛ́n.* ten

> 'You're young, you just/ perhaps you're seventeen, eighteen years old or twenty, you're young, you're fresh, you've left [the parental home], you've given birth to four children, you've given birth to five, you've given birth to ten.' [hi03cb 187]

b. *Náw* now *wé* sub *yu* 2sg *dɔ́n* prf *de* ipfv *gó* go *yu* 2sg *dɔ́n/* prf

'Now that you're about to leave [the man], you've/ [hi03cb 188]

c. *Dɛn* 3pl *tɛ́l* tell *yú* 2sg.indp *sé* qot *mék* sbjv *yu* 2sg *bák* return *dán* that *mɔní* money *wé* sub *yu* 2sg *dɔ́n,* prf *dán* that *mán* man *dɔ́n* prf *pé* pay *fɔ* prep *yu* 2sg *héd.* head

'They tell you to return that money that you have, that man has paid for you.' [hi03cb 189]

The completive aspect involving the auxiliary *fínis* 'finish' may fulfil a discourse function similar to that of the perfect. The use of the completive aspect in signalling precedence of a situation in relation to reference time in ground-preparing, digressive sequences is illustrated in (166), where it appears together with *dɔ́n* 'prf':

(166) *Kip,* ideo *dɛn* 3pl *dɔ́n* prf *fínis* finish *remata* finish.off *ín* 3sg.indp *dé,* there *Boyé* name *dɔ́n* prf *kán* come *e* 3sg.sbj *púl* remove *wí* 1pl.indp *torí* story *torí.* rep

'(When) they had finished him off there [by hitting him with blunt objects] (and) Boyé had come, he told us the story.' [dj05ce 101]

### **6.9 Comparison**

Pichi employs particles and verbs for expressing comparative, superlative, and equative degree. Sentence (167) exemplifies one of the most common ways of expressing comparative degree. It features the comparee *di tín* 'the thing', the parameter verb *bɔkú* 'be much', the comparative particle *mɔ́*, the standard marker *pás* '(sur)pass', and the standard *di watá* 'the water'. As can be seen, the expression of comparison involves a participantintroducing comparative SVC, in which the V2 *pás* '(sur)pass' functions as the standard marker:

6.9 Comparison

(167) *Pero* but *ɛf* if *di* def *tín* thing *kán* pfv *bɔkú* be.much *mɔ́* more *pás* pass *di* def *watá,* water *e* 3sg.sbj *go* pot *lɛ́f* remain *wán* one *pasta,* paste (*…*) 'But if the thing has become more than the water, a paste will remain (…)' [dj03do 059]

Pichi exhibits a rich variety of constructions for comparison. They include the crosslinguistic types of "Exceed-1" and "Exceed-2" comparatives (Stassen 1985). The "Exceed-1" comparative involves a comparative SVC featuring the V2 *pás* '(sur)pass'. We also find a mixture of a Particle and Exceed comparatives (cf. §6.9.1). Equatives, which express equality of degree between a comparee and a standard, may appear in a construction involving a particle, or alternatively, one involving the verb *rích* 'arrive, equal'.

Table 6.9 provides an overview of Pichi constructions employed for comparison as well as "similatives" (cf §6.9.3). For illustration, it contains elicited variations of the same sentence. The more common constructions are found under the heading "primary", while the column "secondary" features less common ones. Glosses for the Pichi words in the table are: *e* '3sg.sbj', *fɔ* 'prep', *kin* 'hab', *lɛk* 'like', *lɔ́n* 'be long, tall', *mán* 'man, person', *mí* '1sg.indp', *mɔ́*'more', *ɔ́l* 'all', *pás* '(sur)pass', *rích* 'arrive, equal', *sɛ́ns* 'intelligence' and *wáka* 'walk.


Table 6.9 Comparison

In general, relative comparison featuring an explicit standard is less common than absolute comparatives and superlatives, in which the standard must be recovered from discourse context. Speakers often employ the rich inventory of inherently graded verbs, adverbs, particles, phrasal expressions and suprasegmentals for the expression of gradation.

6 The verbal system

### **6.9.1 Comparatives**

A participant-introducing SVC featuring the verb *pás* '(sur)pass' is employed to express comparative degree in an "Exceed-1" comparative (Stassen 1985). The following example features the property item *bíg* 'be big' as the parameter verb:

(168) *Dán* that *gɛ́l,* girl *a* 1sg.sbj *tɛ́l* tell *yú* 2sg.indp *sé* qot *e* 3sg.sbj *chapea* weed *lɛk* like *wán* one *sáy* side *wé* sub *e* 3sg.sbj *bíg* be.big *pás* pass *dí* this *wán.* one 'That girl, I tell you that she weeded like a place that was bigger than this.' [ed03sb 060]

In contexts other than comparison, the verb *pás* occurs as a lexical verb with the meanings '(sur)pass, pass by, move along' as in the following three examples:


An SVC can express comparison (168) on its own. However, the adverb of degree *mɔ́* 'more' is equally often employed in addition to *pás* to form a "mixed comparative" (Stassen 1985). The adverb *mɔ́* 'more' functions as an intensifier, albeit highly conventionalised in its use, rather than being an indispensable element of the comparative construction. It exhibits word order flexibility and may occur after (172) or before (173) the parameter verb:


'Right now, I [emp] think that Libreville and here, here is more expensive than there.' [ma03hm 052]

6.9 Comparison

I assume that preverbal *mɔ́* 'more' is being reinforced by the Spanish comparative construction featuring the adverb *más* 'more'. The comparative constructions of both languages exhibit the same linear structure. Compare (174) in colloquial Spanish with (173) above:

(174) *Aquí* here *es* is *más* more *caro* expensive *que* than *allá.* there 'Here [it] is more expensive than there.'

In the absolute comparative in (175) below, *mɔ́* 'more' occurs as a prenominal modifier to the Spanish noun *énfasis* 'emphasis'. The categorial flexibility of *mɔ́* 'more' is exploited by insertion in a Spanish adjective position in a codemixed collocation. This Pichi-Spanish verb-noun combination is creatively used to render the meaning 'be emphatic':

(175) *Mék* sbjv *e* 3sg.sbj *gɛ́t* get *mɔ́* more *énfasis.* emphasis 'Let it be more emphatic [than usual].' [dj05ce 126]

The corpus also contains an example in which *mɔ́*'more' is employed both in pre- and post-verbal position in order to signal an emphatic absolute comparative:

(176) *E* 3sg.sbj *púl* remove *mɔ́* more *plɛ́nte* be.plenty *mɔ́.* more 'He removed much more.' [au07fn 109]

However, unmixed "Exceed" comparatives are particularly common when the parameter is dynamic, not a property item, and hence semantically neutral as to gradation. The use of *mɔ́* 'more' with such verbs automatically results in a quantity gradation, and *mɔ́* can only occur after the parameter in order to modifiy the predicate in its entirety (177):

(177) *Porque* because *ɔ́da* other *sáy* side *fít* can *dé* be.loc *wé,* sub *a* 1sg.sbj *go* pot *wók* work *só,* like.that *a* 1sg.sbj *go* pot *wín* earn *mɔ́* more *pás* pass *dé.* there 'Because there could be another place where, (if) I worked like this, I might earn more than there.' [dj07ae 495]

When a verb is to be graded as to some defined quantity or some kind of quality, *mɔ́* 'more' is usually omitted. Instead, a degree modifier or an object that specifies the quality or quantity may intervene between the parameter and *pás* '(sur)pass'. Compare the adverbial modifier *fáyn* 'fine' in (178) and the object *Bubɛ* 'Bube' in (179):

(178) *Dí* this *wán* one *dɔ́n* prf *de* ipfv *tɔ́k,* talk *dí* this *wán* one *de* ipfv *tɔ́k* talk *fáyn* fine *pás* pass *in* 3sg.poss *sísta* sister 'This one talks, this one talks [the Bube language] better than her sister.' [ab03ab 010]

6 The verbal system

(179) *Lage* name *de* ipfv *tɔ́k* talk *Bubɛ* Bube *pás* pass *mí.* 1sg.indp 'Lage talks Bube (better) than me.' [fr03ab 012]

When the parameter is a motion verb, the "Exceed" comparative may acquire quite a literal meaning as in (180). The example below also shows that the standard can be modified further by way of a relative clause. Such a relative clause with a locative head noun may be employed in contexts where the parameter is non-gradable and the standard is an entire clause (181):


'I bought more than the money was sufficient for.' [rofn05 001]

The collocation *lɛk háw* 'the way (that), as soon as' may also introduce the standard of complex comparatives like (182), in which the standard is an entire adverbial clause. Note the presence of the standard marker *pás* '(sur)pass':

(182) *Na* foc *lɛk* like *sé* qot *yu* 2sg *wánt* want *tɛ́l* tell *wán* one *pɔ́sin* person *sé* qot *yu* 2sg *dú* do *sɔn* some *tín* thing *pás* pass *lɛk* like *háw* how *yu* 2sg *bin* pst *gɛ́fɔ* have.to *dú=an.* do=3sg.obj

'It's as if you want to tell a person that you've done something more than what you should have done.' [au07ec 049]

The standard clause in (183) is also introduced by *lɛk háw* 'the way (that), as soon as'. The sentence features the locative noun *pantáp* 'on, in addition to' as a standard marker instead of *pás*. The use of *pantáp* in this way is only attested in such complex comparatives:

(183) *Bɔt* but *yu* 2sg *nó* neg *fít* can *tɔ́k* talk *sé* qot *a* 1sg.sbj *chɔ́p* eat *trí* three *spún* spoon *pantáp* on *lɛk* like *háw* how *a* 1sg.sbj *kin* hab *chɔ́p.* eat 'But you can't say that you have eaten three spoons more than you usually eat. [au07ec 045]

A second way of forming comparatives is rare. In "Exceed-2" comparatives (Stassen 1985) the parameter is expressed as a PP, hence a nominal. The marker of comparison, the verb *pás* '(sur)pass', is the only verb of the clause and is employed as an inchoativestative verb.

6.9 Comparison

For these reasons, the construction is more likely to appear with quality-denoting nouns like *sɛ́ns* 'intelligence' in (184) than with property-denoting verbs. Compare (185), where the property *gɛ́t sɛ́ns* 'have brain' = 'be intelligent' is graded in an "Exceed-1" comparative:


In a second, equally rare variant of the "Exceed-2" comparative, the property is expressed as a possessed noun of the comparee (186):

(186) *In* 3sg.poss *sɛ́ns* brain *pás* pass *yu* 2sg *yón.* own 'His intelligence surpasses yours.' [ro05de 040]

Relative comparatives are rivalled in their frequency by absolute comparatives in which the standard of comparison is absent and logically implied. In absolute comparatives, the use of *mɔ́* 'more' as a degree adverbial (187) is the most common option.

(187) *Dí* def *wán* one *na* foc *di* def *hós* house *wé* sub *fáyn* be.fine *mɔ́.* more 'This is the house that's more beautiful.' [nn05fn 011]

In contrast, an SVC with a sentence-final, 'stranded' *pás* as in (188) is not accepted by the majority of speakers who were tested:

(188) *?Dí* def *wán* one *na* foc *di* def *bɔ́y* boy *wé* sub *fáyn* be.fine *pás.* pass ?This is the boy who is more handsome. [to07fn 235]

A sentence-final *pás* is all the same common where it occurs in a clause as the only verb (rather than the V2 of an SVC) with the meaning 'surpass an acceptable limit' (189):

(189) *E* 3sg.sbj *dɔ́n* prf *de* ipfv *pás.* pass 'It's become too much now.' [ro05rr 011]

I should point out that in spite of its apparent categorial flexibility, *mɔ́* 'more' may not be used as a lexical verb meaning 'surpass', unlike the verb *moro* 'surpass' in Sranan Tongo (cf. Blanker & Dubbeldam 2010:139).

#### 6 The verbal system

### **6.9.2 Superlatives**

Superlatives are formed by the same formal means as comparatives. The reference of the standard NP is extended to englobe the entire set of possible referents by means of a standard NP featuring *ɔ́l* 'all' or *ɛ́ni* 'every' and the relevant group of referents. The standard NP often consists of the generic nouns *pɔ́sin* 'person', *mán* 'man, person', *húman* 'woman', and *pípul* 'people' if the comparee is human:

(190) *Boyé* name *stáwt* be.corpulent *pás* pass *ɔ́l* all *mán* man *na* loc *di* def *hós.* house 'Boyé is more corpulent than every person in the house.' [ro05de 060]

However, the most common way of rendering a superlative relation is by means of an absolute superlative without explicit mention of a standard NP. Such constructions are no different from absolute comparatives, and the difference in meaning between the two constructions is inferred from context.

In the following absolute superlative, the Spanish adjective *difícil* 'difficult' is followed by *mɔ́*'more' with a superlative meaning. This sentence was uttered after the speaker had taken us on a tour through a new house and explained the hassles involved in building it:

(191) *Di* def *tín* thing *wé* sub *bin* pst *dé* be.loc *difícil* difficult *mɔ́* more *na* loc *dí* this *hós,* house *fɔ* prep *pút* put *nivel.* level 'The thing that was most difficult [of all the construction work] in this house, (was) to level (the ground).' [ye07fn 065]

Aside from constructions like (191), which involve an implicit standard, there are many absolute superlatives where the standard is even more vague. Such "superlatives" form part of the inventory of intensifying and emphatic devices of the language. They involve lexicalised phrases like *pás mák* 'pass (the) limit' or *nó smɔ́l* 'neg small' = 'not in the least':


Superlative degree may also be signalled by the multifunctional word *óva* 'over, excessively' when used as a verb (194) and an adverbial (195).

(194) *Di* def *chɔ́p* food *óva.* be.excessive 'The food is too much.' [au07ec 042]

6.9 Comparison

(195) *Wɛ́n* sub *dɛn* 3pl *dɔ́n* prf *dríng* drink *óva,* over *nɔ́?* intj 'When they've drunk excessively, right?' [ma03hm 069]

*Óva* may also appear as the first component of a compound verb which expresses an excessive degree of the situation denoted by the verb (cf. §4.4.3 for more details):

(196) *Di* def *hós* house *ova-dɔtí.* over.cpd-be.dirty 'The house is excessively dirty.' [au07ec 027]

Emphatic absolute superlatives may also involve the use of degree adverbs like *bád* 'extremely' (197), *tú (mɔ́ch)* 'too much' (198), or *sóté* 'until, extremely:


Beyond that, Pichi features a number of inherently comparative and superlative words. Like the degree expressions *óva* 'over' covered above, these words are multifunctional and may be employed as adverbs or verbs alike. The words *bɛ́ta* 'be very good', *wos* 'be very bad', *tú mɔ́ch* 'be very/too much', as well as *bɔkú* 'be (very) much' alone may signal an exceptionally high degree of a quality or quantity:


These inherently superlative words may combine with *mɔ́*'more' for additional intensity and emphasis as in the following examples. Note the characteristic syntactic flexibility of *mɔ́* in these sentences:

6 The verbal system


Nuances of superlative degree may also be signalled through the use of emphatic suprasegmental features such as extra-high pitch, pitch range expansion, or vowel lengthening, as well as through other emphatic devices, like ideophones and reduplication.

### **6.9.3 Equatives**

Equative constructions are formed in two ways. The most frequent one involves the preposition *lɛ(kɛ)* 'like' as the standard marker. The preposition is inserted between the parameter and the standard. This construction assigns the same degree of a property to both the comparee and the standard:


Take note of the lexicalised equative construction *bɔkú lɛk nyɔ́ní* 'be many like ants' in (208):

(208) *Yu* 2sg *fít* can *tɔ́k* talk *sé* qot *'mi* 1sg.poss *brɔ́da* brother *dɛn* 3pl *bɔkú* be.much *lɛk* like *nyɔ́ní'.* ant 'You can say "my siblings are many just like ants".' [ro05ee 034]

In constructions featuring an entire equative clause as the standard, the collocation *lɛk háw* 'like how' = 'the way that' is used instead of *lɛk* (209–210). The second example below features a codemixed equative construction featuring the Spanish element *tan* 'as, so'. In unmixed sentences, Pichi does not employ an additional parameter marker like *tan* before the parameter verb:

6.9 Comparison

(209) (*…*) *mék* sbjv *yu* 2sg *nó* neg *para* stop *sóté* until *mék* sbjv *e* 3sg.sbj *tík* be.thick *lɛk* like *háw* how *e* 3sg.sbj *bin* pst *dé* be.loc *só.* like.that

'(…) don't stop until it's (as) thick as it was.' [dj03do 058]

(210) *Mí* 1sg.indp *nóto* neg.foc *tan* as *dɛ́bul* devil *lɛk* like *háw* how *yu* 2sg *de* ipfv *chɛ́k* think *mí.* 1sg.indp 'I'm not as much of a devil as you think I am.' [ye07fn 002]

Pichi speakers employ a second, albeit marginal equative construction, in which the verb *rích* 'arrive' is the only verb. At the same time, the parameter appears as a nominal constituent in a *fɔ*-prepositional phrase. Like the verb *pás* '(sur)pass' in (184) above, the verb *rích* is employed as an inchoative-stative verb in these instances:

(211) *E* 3sg.sbj *nó* neg *rích* arrive *mí* 1sg.indp *fɔ* prep *fáyn.* fine 'He doesn't equal me in beauty.' [ye07fn 134]

Other than that, verb *rích* is employed as an allative motion verb 'reach, arrive (at)'. In addition to its literal sense, *rích* also occurs with the meaning 'equal, be sufficient' (212). *Rích* may also be found as a minor verb in the V2 position of a motion-direction SVC (213):


## **7 The clause**

There are four types of basic, non-complex clause structures in Pichi. Pragmatically marked structures that cut across these four types include negative constructions, questions, as well as focus and topic constructions. The expression of being and having involves a network of functionally overlapping copula and existential verbs, and verbs of possession. Pichi adverbs modify verbs and clauses. The majority of adverbs occupy a clause-initial or a clause-final position, but a small set of time and degree adverbs are also found in preverbal position in the company of TMA markers.

### **7.1 Clause structure**

Four types of clauses can be distinguished by their basic order, as well as the presence and type of the core constituents verb, subject and object: verbal clauses, serial verb clauses, copula clauses and directive clauses.

### **7.1.1 Verbal clauses**

The order of constituents in verbal clauses corresponds to the pattern presented in Figure 7.1. Details on the structure of the noun phrase and the predicate are provided in Figure 5.1 and Figure 6.1 respectively. A few observations on Figure 7.1 follow: Subject NPs (sbj np) may be picked up by a resumptive personal pronoun (pro). They may hence co-occur in the same clause, but such structures involve topicalisation and are therefore pragmatically marked (hence the separation of sbj np and pro with a slash). There are several adverbial slots in a clause, details on the positions of adverb(ials) are covered in detail in §7.7.

Pichi has double object constructions marked by constituent order. The first object NP slot (obj np) is reserved for recipient or beneficiary objects, the second for theme or patient objects (for details, see Table 9.10). There are a clause-initial and a clause-final slot for interjections. The latter may be filled, among other elements, by the sentence-final modal particle and interjection *ó* 'sp' (cf. §12.2.5).


Figure 7.1 Constituent order in verbal clauses

#### 7 The clause

Pichi has a subject-verb word order in intransitive clauses (cf. 3 below), and a subjectverb-object order in transitive clauses (1):

(1) *E* 3sg.sbj *sɛ́n* send *di* def *bɔ́l.* ball 'She threw the ball.' [ra07se 203]

Objects follow the verb. In most double-object constructions, the primary object with the semantic role of recipient or beneficiary is found immediately to the right of the verb. The secondary object encodes the theme or patient and follows the primary object:

(2) *A* 1sg.sbj *sé* qot *"nó* neg *gí=an* give=3sg.obj *leche,* milk *gí=an* give=3sg.obj *wɔtá".* water 'I said "don't give him milk, give him water".' [ab03ab 099]

Full nouns occur on their own as subjects. But a coreferential dependent pronoun may additionally occur in the clause which picks up the definite subject. Such structures may be seen to involve topicalisation by dislocation (cf. §7.5.1). Example (3) features both alternatives:

(3) *Di* def *chía* chair *blák,* be.black *di* def *chía,* chair *e* 3sg.sbj *blák.* be.black 'The chair is black, the chair (it) is black.' [dj05ae 121]

Pronoun resumption is also found with objects. The following two examples illustrate the use of pronominal copying with fronted and topical object NPs. In (4), the full NP *dán mán* 'that man' and in (5) the emphatic 3pl pronoun *dɛ́n* are set off from the rest of the clause by an intonation break and resumed by object pronouns:


An indication that subject pronoun copying may also involve a topic-comment structure comes from examples such as (6). This sentence features the independent, emphatic personal pronoun *dɛ́n* at the beginning of the clause, followed by a coreferential dependent pronoun:

(6) *Dɛ́n,* 3pl.indp *dɛn* 3pl *bin* pst *de,* ipfv *dɛn* 3pl *bin* pst *dɔ́n* prf *sabí* know (*…*)

'As for them, they were, they had already found out (…)' [ma03hm 037]

7.1 Clause structure

Constructions like (6), in which a personal pronoun is fronted for focus or emphasis and immediately followed by a resumptive dependent personal pronoun, are, however, rare. Instead, emphatic personal pronouns appear more often on their own. This pattern suggests that subject pronoun copying is pragmatically less marked than object pronoun copying as encountered in (4) and (5). This observation fits with the high frequency of resumptive pronoun usage in the relativised position of subject relative clauses as compared to the lower frequency in object relative clauses (cf. §10.6.2):

(7) *Mí* 1sg.indp *dɔ́n* prf *sɔ́fa.* suffer 'I [emp] have suffered.' [ab03ab 037]

Quotative clauses introduced by the quotative marker *sé* 'qot' can be found in the syntactic position of the subject or object. A clause introduced by *sé* may also occupy the clause-initial or clause-final adverbial position. Consider the two alternative translations of the following sentence. The first translation renders the function of a quotative complement clause, the second that of an adverbial cause clause:

(8) *A* 1sg.sbj *dɔ́n* prf *de* ipfv *gládin* be.glad *sé* qot *a* 1sg.sbj *dɔ́n* prf *gó.* go

'I was already glad that I was gone.' or 'I was already glad because I was gone.' [ab03ay 091]

In the predicate, the negator *nó*, TMA markers, and preverbal adverbs occur before the verb, in this order. The clitic 3sg.obj pronoun *=an* immediately follows the verb. Apart from the negator *nó* 'neg' and TMA markers, the adverbs of degree *tú* 'too (much)', *tú (mɔ́ch)* 'too (much)', *só* 'so (much)', as well as the temporal adverbs *jís*/*jɔ́s* 'just' and *stíl* 'still' are the only elements that may appear between a subject pronoun or NP and the verb.

In (9), *tú* 'too (much)' occurs before the stative verb *évi* 'be heavy'. In (10), *tú* appears before the locative-existential copula *dé*:


Other adverbs and adverbials are usually found at the clause margins. Compare the clause-final degree adverb *smɔ́l* 'a bit' (< '(be) small') in (11):

(11) *Djunais* name *dɔ́n* prf *dríng* drink *smɔ́l.* a.bit 'Djunais has drunk a bit [of alcohol].' [fr03wt 182]

#### 7 The clause

### **7.1.2 Copula clauses**

Two types of copula clauses should be distinguished. Equative clauses feature the focus markers *na* 'foc' and *nóto* 'neg.foc' in a copula function. I analyse *na*-copula clauses as grammaticalised topic-comment structures, in which the notional subject is topicalised, and the nominal functioning as the copula complement is under focus. These clauses differ from verbal clauses and predicate adjective clauses involving the copula *dé* 'be.loc' in two ways: Pronominal subjects are always from the emphatic series (12), and more often than not, the 3sg and 3pl pronouns remain unexpressed (13) because *na* and *nóto* incorporate 3sg reference by default:


Predicate adjective clauses constitute the second type of copula clause. A small set of property-denoting verbs may also function as predicate adjectives and appear as complements to the locative-existential copula *dé* 'be.loc' (14). Unlike other property items, these adjectives may therefore appear in the same syntactic position as adverbials in this type of copula clause (15):


### **7.1.3 Directive clauses**

The syntax of 2sg directive (imperative) clauses is distinct from other clause types and other directive clauses in that the 2sg subject remains unexpressed (16). However, a 2pl subject must be overtly expressed (17):


7.2 Negation

Moreover, directives are the only type of main clause that feature a TMA marker in the prenominal rather than the preverbal slot; compare the subjunctive marker *mék* 'sbjv' (18):

(18) *Mék* sbjv *a* 1sg.sbj *púl* pull *wán* one *smɔ́l* small *torí?* story 'Should I tell a little story?' [au07se 059]

At the same time, directive subjunctive clauses are structurally no different from other clauses that feature a clause linker at their very left. Compare (18) with the sequential clause introduced by *wé* 'sub' in (19):

(19) *Wé* sub *e* 3sg.sbj *bin* pst *dáy* die *só.* like.that 'And he died just like that.' [ed03sb 126]

### **7.2 Negation**

Pichi negation revolves around the general negator *nó* 'neg', which functions as a negative particle in verb negation and as a negative quantifier in NP negation. Besides *nó*, Pichi features the negative indefinite pronoun *nátin* 'nothing', which is specialised for use in negative clauses. Other than that, Pichi makes use of negative phrases consisting of *nó* and generic nouns that function as negative indefinites and adverbials. Furthermore, clause negation is characterised by negative concord; when the verb is negated, non-specific NPs may also be preceded by *nó* 'neg'.

Finally, negation of the perfect aspect as well as equative clauses and focus constructions is not achieved by the addition of the negator *nó*. Instead, negation in these environments is suppletive or "asymmetrical" (Miestamo 2005: 72). It relies on the use of morphologically distinct elements that incorporate negative polarity as well as the relevant grammatical category.

### **7.2.1 Verb negation**

Table 7.1 below provides an overview of the forms and structures employed to express verb negation. "Standard negation", the negation of declarative clauses (Miestamo 2005) revolves around the general negator *nó* 'neg', see entry 1a in Table 7.1. Verb negation involves "symmetric" (Miestamo 2005) or "additive" negation (Jungraithmayr 1988) with all TMA categories except for perfect tense-aspect and imperative mood. Symmetric negation involves adding the standard negator *nó* without further adjustments to the clause. The negation of perfect tense-aspect is "asymmetric" (Miestamo 2005) or "substitutive" (Jungraithmayr 1988), see entry 1b in Table 7.1. Negation relies on the use of morphologically distinct elements that incorporate negative polarity as well as the relevant grammatical category. The negation of imperatives is also optionally achieved by means of

#### 7 The clause

negative subjunctives and is therefore also asymmetric, see entry 1c in Table 7.1). Further, Pichi makes use of bipolar adverbs to express negative quantification and emphasis, see entry 2 in Table 7.1. The negation of identity-equative copulas is covered in §7.6 and constituent negation is treated in §7.2.4.


Table 7.1 Overview of verb negation

The negation of declarative clauses is symmetrical. They acquire negative polarity when the general negator *nó* is placed before the bare verb or the relevant TMA marker. The position of the negator is canonical. The imperfective-marked verb *gí* 'give' in (20) is negated in (21). A negative existential clause is presented in (22). Note the appearance of negative concord in the latter example:


Sentence (23) contains both an affirmative and a negative clause in the potential mood. The two subsequent examples present an affirmative clause marked for past tense and its negative counterpart (24-25).

(23) *Ho,* intj *dán* that *mán* man *go* pot *dú* do *vɔ́mit* vomit *tidé,* today *e* 3sg.sbj *nó* neg *go* pot *slíp.* sleep 'That man is going to vomit today, he won't sleep.' [ye03cd 143]

7.2 Negation


Imperatives (26) are negated either with a symmetrical structure (27) or with an asymmetrical structure involving a negative subjunctive clause (28):


The negation of the perfect tense-aspect is asymmetrical. While the affirmative features the marker *dɔ́n* 'prf' (29), the negative perfect is formed with a suppletive allomorph, i.e. either of the free variants *nɛ́a* and *nɔ́ba* 'neg.prf' (30):

(29) *Yu* 2sg *dɔ́n* prf *bɔ́n* give.birth *fo* four *pikín,* child (*…*) 'You have given birth to four children, (…)' [hi03cb 187]

(30) *E* 3sg.sbj *nɛ́a* neg.prf *bɔ́n* give.birth *pikín.* child 'She hasn't given birth to a child yet.' [fr03ft 139]

The adverbial *yét* 'still, yet' may appear with the negative perfect without providing an additional meaning besides stressing the nuance of current relevance inherent to the perfect (31). However, the combination *nó – yét* 'not yet' can also express this nuance of the perfect by itself and thereby function as a *de facto* negative perfect marker (32). In an affirmative clause, the adverbial *yét* means 'yet, still', as in (83) in §5.2.3.

(31) *Yu* 2sg *sísta* sister *e* 3sg.sbj *nɔ́ba* neg.prf *máred* marry *yét?* yet 'Your sister isn't married yet?'

#### 7 The clause

(32) *E* 3sg.sbj *nó* neg *máred* marry *yét?* yet 'She isn't married yet?'

The two other combinations of verb negation and a clause-final adverbial are *nó – mɔ́* 'no more, not again' and *nó – sɛ́f* 'not even'. Compare the affirmative use of *mɔ́* 'more' in (26) with (33) below.

(33) *Dɛn* 3pl *nó* neg *go* pot *fláy* fly *na* loc *Bata* place *mɔ́.* more 'They're not going to fly to Bata anymore/again.' [eb07fn 237]

Examples (34) and (35) present the use of *sɛ́f* 'self, emp' in an affirmative and a negative clause, respectively. The negated clause acquires an emphatic negative meaning:


### **7.2.2 Negative concord**

Pichi makes use of negative concord. Verbal and constituent negation co-occur in clauses with negative polarity. Negative concord is pragmatically determined, hence non-strict with lexical nouns, where it only renders emphatic meanings. Negative concord is, grammatically determined, hence strict, with negative indefinite pronouns and phrases. In either case, the negated constituent in constructions featuring negative concord is best interpreted as non-specific.

Pragmatically neutral lexical nouns in subject position are not normally preceded by the general negator *nó* 'neg' in negative clauses:

(36) *Fíba* fever *nó* neg *sube* go.up *ín.* 3sg.indp '(The) fever hasn't risen on him.' [eb07fn 171]

In (37), the plural subject *mán dɛn* 'people' and the singular subject*chɔ́p* 'food' are both not preceded by *nó* 'neg'. The noun *chɔ́p* is the subject of a negative existential clause. Such clauses usually only feature negative concord when extra emphasis is desired (39):

7.2 Negation

(37) *Mán* man *dɛn* pl *nó* neg *de* ipfv *bísin* be.busy *fɔ* prep *mék* make *fám* farm *mɔ́,* more *yu* 2sg *gó* go *fɔ* prep *mákit,* market *chɔ́p* food *nó* neg *dé.*

be.loc

'People don't care about farming anymore, (if) you go to the market there's no food.' [ed03sp 053]

Subject NPs may nevertheless be preceded by *nó*. Such negative clauses featuring negative concord have a single negation reading. Negative concord provides a means of adding an emphatic sense to the negative clause. Compare *dɔ́kta* 'doctor' in (38) and *motó* 'car' in (39):


Object NPs also only feature negative concord when emphasis is intended. Compare the non-emphatic negative clause in (40) with (41). The use of negative concord in (41) gives an emphatic meaning to the object *problema* 'problem'. Also note the presence of the independent emphatic pronoun *ín* '3sg.indp' (41):

(40) *A* 1sg.sbj *nó* neg *gɛ́t* get *pamáyn.* oil 'I don't have (any) oil.' [ab03ay 015]

(41) *Ín* 3sg.indp *go* pot *chɔ́p=an,* eat=3sg.obj *e* 3sg.sbj *nó* neg *gɛ́t* get *nó* neg *problema.* problem 'He [emp] will eat it, he has no problem whatsoever [with this kind of food].' [ro05rt 066]

Often, emphasis comes in combination with other emphatic features, i.e. suprasegmental cues such as increased volume, higher pitch, or reduced speed in the pronunciation of the negator and the negated NP, or the use of emphatic elements. NPs preceded by *nó* in negative clauses can receive an even higher degree of emphasis if the negative quantifier *nó* is followed by the cardinal numeral and indefinite determiner *wán*, as in (42) with the object *wɔ́d* 'word':

(42) *Sóté* until *a* 1sg.sbj *nó* neg *tɔ́k* talk *nó* neg *wán* one *wɔ́d.* word 'Until I didn't say a single word (anymore).' [ab03ay 088]

#### 7 The clause

Negative concord is also found in coordinate NPs featuring the negative coordinator pair *ni – ni*, which is borrowed from Spanish (43). Spanish employs no negative concord in this particular construction (44):


### **7.2.3 Negative indefinite pronouns and phrases**

While negative concord is exploited for pragmatic purposes with lexical nouns, negative concord is strict, and hence grammatically conditioned with negative indefinite pronouns and negative indefinite phrases. Pichi has a single item that can unequivocally be qualified as a polarity sensitive, monomorphemic negative indefinite pronoun, namely *nátin* 'nothing'. The expression *nó bɔ́di* (*<* 'neg body') 'nobody' is partly opaque and may therefore be seen as intermediate between negative indefinite pronoun and negative indefinite phrase: Although *nó bɔ́di* is segmentable, the noun *bɔ́di* is not used as a generic noun with the meaning 'person'. The noun *bɔ́di* also only seldom occurs with the meaning 'body', the regular term for 'body' being *skín.*

Concepts other than 'nobody' and 'nothing' are expressed via segmentable and semantically transparent syntactic phrases featuring the negative quantifier *nó* 'neg' and a generic noun. This mirrors the formation of indefinite phrases, for which there are, however, no non-segmentable exceptions (i.e. *sɔn tín* 'something', *sɔn pɔ́sin* 'somebody', see §5.4.3. Table 7.2 lists Pichi negative indefinite pronouns and negative indefinite phrases:

In verbal clauses, the negative indefinite pronoun *nátin* must be used with a preceding negative quantifier *nó* 'neg' as well as with support from verb negation. This holds for both the subject and object position. Since *nátin* is inherently negative, its use in verbal clauses therefore invariably involves the use of double negative concord. Compare the indefinite NP *sɔn tín* 'something' (45) with the subject and object negative indefinite pronoun *nátin* 'nothing' in (46) and (47) respectively:


7.2 Negation


Table 7.2 Negative indefinite pronouns and negative indefinite phrases

(47) *Mí* 1sg.indp *nó* neg *go* pot *tɛ́l=an* tell=3sg.obj *nó* neg *nátín.* nothing 'I [emp] wouldn't tell him anything.' [bo03cb 138]

In the same vein, the co-occurrence of the negative quantifier *nó* and the negative indefinite pronoun without the simultaneous use of verbal negation is ungrammatical.

(48) *Nó* neg *nátín* nothing *\*(nó)* neg *dé* be.loc *dé.* there 'Nothing is there.'

Strict negative concord also applies to all negative indefinite phrases involving generic nouns including *nó bɔ́di* 'nobody'. Since generic nouns are not inherently negative, verbal clauses featuring negative indefinite phrases involve single negative concord: The generic noun is preceded by the negative quantifier *nó*, and the verb is negated.

(49) *Dís* this *sɔ́nde* Sunday *\*(nó)* neg *bɔ́di* body *\*(nó)* neg *dé* be.loc *na* loc *strít.* street 'This Sunday, nobody is in the streets.' [ro05ee 136]

The negative indefinite phrase *nó mán* 'neg man' = 'nobody' is equally common as *nó*


#### 7 The clause

The affirmative counterparts of the negative indefinite phrases in (49–50) are indefinite (quantifier) phrases involving *pɔ́sin* 'person' and *mán* 'man', which function as indefinite pronouns:


Negative indefinite adverbials are also formed by means of phrasal syntax. The phrase *nó sáy* 'neg place' = 'nowhere' is the most commonly employed expression to negate existence in a place. Compare the affirmative and negative sentences involving *sáy* 'side, place':


The generic noun *sáy* 'side, place' can also be used in a more literal sense to denote 'space, place'. In that case, it is not usually additionally preceded by *nó* in negative clauses unless extra emphasis is intended. Compare the following two examples:


The adverbial concept 'never' is expressed via the phrase *nó wán dé* 'neg one day' (58). Example (57) features the equivalent affirmative phrase *ɔ́l tɛ́n* 'all time' = 'always':

(57) *Di* def *húman* woman *ɔ́l* all *tɛ́n* time *e* 3sg.sbj *dé* be.loc *fáyn.* fine 'The woman is always looking fine.' [dj05ae 155]

7.2 Negation

(58) *E* 3sg.sbj *sé* qot *nó* neg *wán* one *dé* day *e* 3sg.sbj *nó* neg *go* pot *dú=an* do=3sg.obj *mɔ́.* more 'He said he would never do it again.' [ro05ee 134]

The negative pronominal meaning of 'none, any' may be expressed through verb negation and use of the quantifier and indefinite determiner *sɔn* 'some, a', which may refer to count and mass nouns alike. The affirmative clause in (59) features *sɔn* used as pronominal (cf. also 22–23). The negative counterpart of (59) may simply be a negative clause (60):

(59) *Dán* that *banána,* banana *a* 1sg.sbj *gí=an* give=3sg.obj *sɔn.* some 'That banana, I gave him one.' [ab03ab 096]

(60) *A* 1sg.sbj *nó* neg *gɛ́t* get *sɔn.* some 'I don't have some/any.' [eb07fn 303]

Alternatively, the negative indefinite phrase *nó wán*, which features the noun substitute *wán* 'one' may be employed when the referent is a count noun or an individuated entity (61):

(61) *Nó* neg *wán* one *nó* neg *lɛ́f* remain *wet* with *mí.* 1sg.indp 'None (at all) remains with me.' [ye07fn 018]

The use of *nó wán* in such contexts often has emphatic connotations. Accordingly, the cardinal numeral *wán* also appears between the negator *nó* and a noun in emphatic negative phrases like (62) and (63). This usage also corresponds to the use of *wán* as an emphatic indefinite determiner in other contexts (e.g. with nouns under cleft focus in presentatives (cf. 15):


The fixed expression *nó wán dé* 'never' in (58) above is also such an emphatic negative phrase, even if lexicalised.

#### 7 The clause

### **7.2.4 Constituent negation**

Sections §7.2.2 and §7.2.3 have shown that one means of negating nominal constituents is by placing the negator *nó* 'neg' before them. However, this kind of constituent negation does not appear independently of verb negation. A second means available for negating a larger range of constituents is the negative cleft focus construction. An overview of constituent negation is given in Table 7.3.

Table 7.3 Constituent negation


Cleft focus provides a means of negating single constituents and is possible with any constituent that may be focused (cf. §7.4.3.2). In cleft focus constructions, the focused element is fronted to the sentence-initial position and preceded by the negative focus marker *nóto* 'neg.foc'. Compare (64), where the subject NP *ɔ́l húman* 'all women' is singled out for constituent negation:

(64) *Nóto ɔ́l húman fít máred.*

neg.foc all woman can marry

'Not all women can get married.' [ab03ab 196]

Adverbials are negated in the same way as core NPs. Example (65) features the negated time adverbial *tidé* 'today', (66) the reason adverbial *fɔ dán tín*:


'Is it not because of that that you are crying?' [ne05fn 004]

In (67), the speaker abbreviated as (hi) complains about the discrimination of women in wedlock, a condition she likens to slavery. In the example, speaker (hi) first negates the direct quote *e fíba* 'it resembles', the second *nóto* negates the verbal constituent as such:

(67) *Ɛhɛ́,* intj *nóto* neg.foc *"e* 3sg.sbj *fíba,"* resemble *na* foc *esclavitud,* slavery *nóto* neg.foc *"fíba".* seem 'Yes, not "it resembles (slavery)", it's slavery, not "resemble".' [hi03cb 227]

7.3 Questions

Sentences (68) and (69) illustrate how yet larger sentence constituents can be singled out for negation. Both examples are negative factive clauses, in which the existence of the situation of the reference clause is negated:

(68) *Ɛf* if *nóto* neg.foc *yu* 2sg *bay,* buy *dán* that *húman* woman *go* pot *bít* beat *yú* 2sg.indp *sóté* until *yu* 2sg *go* pot *gó* go *lɛ́f=an.*

leave=3sg.obj

'If it wasn't the case that you had bought (it), that woman would beat you until you'd go and leave it there.' [ab03ab 033]

(69) *Nóto* neg.foc *sé* qot *na* foc *hɔ́s* house *dɛn* pl *fɔ* prep *fɔ́s* first *tɛ́n* time *wé* sub *dɛn* 3pl *strɔ́n,* be.strong *e* 3sg.sbj *fɔ* prep *dɔ́n* prf *fɔdɔ́n.*

fall

'(If) it wasn't the case that they were houses of the past that are strong, it would have already collapsed.' [hi03cb 045]

### **7.3 Questions**

This section covers yes-no questions, alternative questions, and content questions, as well as answers to questions. It is useful to refer to §3.4.5 for details on the intonational characteristics of questions.

### **7.3.1 Yes-no and alternative questions**

Yes-no questions have the syntax of declarative clauses and do not involve obligatory question particles. Yes-no questions are therefore distinguished from declarative clauses by intonation (cf. §3.4.5):

(70) *Yu* 2sg *wánt* want *de* ipfv *gó?* go 'Do you want to go?' [eb07fn 202]

However, speakers often employ the interjections *ɛ́n* and *nɔ́*sentence-finally in biased questions in order to channel-check:

(71) *Yu* 2sg *nó=an* know=3sg.obj *ɛ́n?* intj 'You know her, right?' [li07pe 032]

In alternative questions, the first alternative bears question intonation, while the second alternative carries the intonation of a declarative clause:

7 The clause

(72) *Yu* 2sg *sísta* sister *stíl* still *máred* marry *ɔ* or *e* 3sg.sbj *nó* neg *máred* marry *mɔ́?* more 'Is your sister still married or is she no more married?' [ro05ee 050]

### **7.3.2 Content questions**

Content questions are formed by way of a mixed question-word system summarised in Table 7.4. Note that I classify the question element *wétin* 'what' as monomorphemic although it could alternatively be analysed as bimorphemic (i.e. *wé.tín* = \**wé*.thing). However, \**wé*= does not function as a question particle with any other generic noun, and an etymological relation with *wé* 'sub' remains to be proven.


Table 7.4 Question elements (x = questioned noun)

The question word system of Pichi involves three types of both "transparent" and "opaque question elements" (Muysken & Smith 1990): (1) Monomorphemic elements function as question elements or words in their own right. Amongst these, we find the clitic *ús=* 'q', which forms (2) bimorphemic question words with generic nouns in order to render basic concepts like who, what, and when.

Question phrases (3) may consist of a prepositional phrase introduced by *fɔ* 'prep', *fɔséka* 'due to', and *wet* 'with, due to' and contain a mono- or bimorphemic question word (e.g. *fɔ wétin* 'prep what' = 'why'). Alternatively, question phrases may consist of idiomatic clauses featuring the verbs *mék* 'make' or *dú* 'do' and *wétin* or *ús=tín* 'what' in subject position. A second type of question phrase involves constructions featuring

7.3 Questions

the bimorphemic question word *ús=káyn* and a generic or other noun (e.g. *ús=káyn pɔ́sin* 'q=kind person' = 'who', *ús=káyn motó* 'q=kind car' = 'which car').

In Table 7.4, x stands for any noun. Which x and how much x are therefore question noun modifiers and quantifiers, respectively. The table contains all unequivocally accepted question elements and excludes other logically possible but unattested options (e.g. *?ús=plés* 'q=place' = 'where'; *?ús=káyn mán* 'q=kind man' = 'who').

#### **7.3.2.1 Structural issues**

In content questions, any constituent other than the definite article *di*, focus and topic particles, or TMA markers can be questioned through replacement by a question element. Question words show some distributional restrictions when compared to regular nouns.

For instance, question elements are not usually modified by demonstratives and deictic adverbials, or modifier nouns and adjectives. Similarly, only *údat* and *ús=pɔ́sin* 'who' may optionally take the pluraliser *dɛn* (i.e. *\*ús=tín dɛn* 'what pl'):


The pluralisation of 'who' is likely to be a structural borrowing from Spanish, or is at least reinforced by the equivalent Spanish structure. Compare the equivalent Spanish question:

(75) *Quién-es* who-pl *son?* are 'Who are they?'

Question elements also have other distributional characteristics of regular NPs. For example, in the following sentence, *údat* 'who' is found in the possessor position of a dislocated possessive construction, which in turn participates in a presentative clause:

(76) *Na* foc *údat* who *in* 3sg.poss *búk* book *dís?* this 'Whose book (is) this?' [ro05de 055]

Multiple core (77) and adverbial (78) NPs forming part of coordinate structures may also be questioned. These two examples also show that in principle, a sentence may contain several question elements, even if this is rare in natural speech:

#### 7 The clause


Question elements may occur in situ in the original position of the questioned element, or they may be fronted. Questioned subjects naturally occur at the beginning of the clause as shown in (77). They may also optionally be focused in cleft constructions (79):

(79) *Na* foc *údat* who *hambɔ́g* bother *dɛ́n?* 3pl.indp 'Who bothered them?' [ro05de 041]

Objects can be questioned in situ (e.g. *údat dɛn* and *ús=pɔ́sin dɛn* in 73 and 74 above) or be fronted (80). Fronted objects may also optionally be cleft-focused (81):


The objects of prepositions may also be questioned in situ or be fronted. When fronted, either the entire prepositional phrase appears at the beginning of the clause, or the preposition is stranded. However, stranding in questions is only attested with *fɔ* 'prep' (82), *wet* 'with' (83), and *pan* 'on' (84):


#### 7.3 Questions

All constituents that may be questioned in main clauses can also be replaced by question elements in subordinate clauses. Non-subject constituents of subordinate clauses can be questioned in situ (85) or be fronted (86):


Complement or adverbial clauses introduced by *sé* 'qot' are questioned like nominal constituents. The question word is, however, always found in situ as in the rhetorical question in (87). Here a cause clause is questioned by means of the phrase *sé wétin* 'qot what' = 'because of what':

(87) *A* 1sg.sbj *go* pot *púl=an* remove=3sg.obj *na* loc *mi* 1sg.poss *yáy* eye *sé* qot *wétin?* what 'I would remove it [the pair of sunglasses] from my eyes for what?' [ye07ga 011]

#### **7.3.2.2 Questioning subjects and objects**

Questioned subjects naturally occur at the beginning of the question clause, as in (79) above. Questioned objects appear at the beginning of the sentence (88), or in their original position (89). These two examples feature the question word *wétin* 'what', which is used for questioning inanimate entities:


Example (90) illustrates the questioning of a complex object NP. The dislocated possessive construction *údat* in *motó* 'whose car' is the object of *yús* 'use' and under focus with the focus particle *na*. The questioning of a possessor NP is also achieved by circumlocution with the verb *gɛ́t* 'get, have' (91).

Both examples involve the question word *údat* 'who', which is used for questioning human referents. In a minority of cases, the concept 'who' is also expressed by the bimorphemic question words *ús=pɔ́sin* 'q=person' (cf. 74 above) and *ús=mán* 'q=man' (cf. 83 above) in all relevant syntactic positions:

#### 7 The clause


The clitic question element *ús=*q may combine with the pronominal and noun substitute *wán* 'one' in order to render the concept 'which one'. The collocation may be used to selectively question any noun (92). *Ús=wán* is also employed in an idiomatic question clause in order to ask for a person's name (93). The latter usage is conventionalised and very likely to be a calque from the equivalent Spanish phrase *¿cuál es tú nombre?* 'which (one) is your name' = 'what's your name?':


A similar syntactic flexibility is characteristic of the objects of V2 minor (i.e. closed class) verbs in SVCs. The questioned object of *pás* '(sur)pass' in the comparative SVC in (94) and the object of *kɛ́r* 'carry, take' in the motion-direction SVC in (95) may be found in the original syntactic position:


Alternatively, the objects of V2 minor verbs may occur in the sentence-initial, fronted position with or without additional cleft focus marking, with the same liberty as other objects. These constructions leave the V2 of the SVC "stranded" in the sentence-final position. Compare the following two sentences with the two preceding ones above:

(96) *Na* foc *údat* who *dí* this *bɔ́y* boy *bíg* big *pás?* pass 'Who is this boy bigger than?' [lo07he 016]

7.3 Questions

(97) *Ús=sáy* q=side *yu* 2sg *de* ipfv *kɛ́r* carry *di* def *motó* car *gó?* go 'Where are you taking the car to?' [lo07he 018]

At the same time, the questioning of the instrument or material objects of *ték* 'take' in participant-introducing SVCs is characterised by some idiosyncracies. Firstly, speakers seem to prefer to front the questioned object rather than leave it in the original syntactic position between *ték* 'take' and the following major verb (i.e. *bíl* 'build' in the following example). Compare (98):

(98) *Ús=káyn* q=kind *plɛ́nk* board *dɛn* 3pl *ték* take *bíl* build *di* def *hós?* house 'What (kind of) board did they build the house with?' [dj05ce 104]

Secondly, we find double marking of the instrument objects of *ték* 'take' as a rather regular way of questioning these objects. In (99), the object of *ték* (i.e. *ús=tín* 'what') is fronted and focused. The question word and object *ús=tín* is additionally preceded by the instrumental/comitative preposition *wet* 'with' as if the corresponding declarative clause had been something ungrammatical like \**dɛn ték wet plɛ́nk bíl di hós* '3pl take with board build def house' = \*'they took with board to build the house' (cf. also 32–33):

(99) *Na* foc *wet* with *ús=tín* q=thing *dɛn* 3pl *ték* take *bíl* build *di* def *hós?* house 'With what did they build the house?' [dj07ae 479]

However, fronting of the patient object of the major (open class) verb in *ték* SVCs is not accepted (100). Patients are usually questioned in situ in their original syntactic position following the major verb (101):


#### **7.3.2.3 Questioning modifiers**

Modifiers and demonstratives in NPs are questioned via three question elements: the clitic *ús=* 'q, which'; the (marginally employed) phonologically independent question word *wích* 'which', and the bimorphemic question word *ús=káyn* 'q=kind'. Quantifiers are questioned by means of *háw mɔ́ch* 'how much' (cf. 124–126 below). The element *ús=* straddles the boundary of a more functional and a more lexical meaning. Consider the translations of the following two examples, which contrast the rarely used and more lexical *wích* 'which' with the high-frequency question particle *ús=* 'q':

#### 7 The clause


One indication of the more functional status of *ús=* is its cliticisation in the first place (cf. §2.6.3). Secondly, in the majority of instances in the corpus, *ús=* combines with a limited number of generic nouns (e.g. *pɔ́sin* 'person', *mán* 'man, person', *tín* 'thing', *sáy* 'side, place', *tɛ́n* 'time') in order to form general, basic question words with meanings like who, what, where, and when. Yet, *ús=* is nevertheless also used with the meaning 'which' in order to form specific question words questioning modifiers as in the following two examples:


However, questions like (104) and (105) are equally often formed by employing the question word *ús=káyn* 'q=kind' instead of *ús=* alone. The meaning of *ús=káyn* therefore also vacillates between a more literal sense, in which the pronominal and generic noun *káyn* 'kind' retains its lexical meaning of 'kind', and a more functional one, in which the entire question word *ús=káyn* is equivalent to *ús=*, 'q, which'. This ambiguity in the meaning of *káyn* 'kind' is reflected in the translations of the following two examples:


The more functional use of *ús=káyn* is more obvious when it precedes a generic noun as in the following two examples. Here, the phrase *ús=káyn tín* 'q=kind thing' has the

#### 7.3 Questions

same meanings as *wétin* or *ús=tín* 'what'. Note that (108) is a free relative clause and sentence (109) an indirect question. The long forms featuring *káyn* 'kind' are equally common in this position as are the shorter forms *wétin* and *ús=tín:*


The same, more functional use can be observed when *ús=káyn* precedes the generic noun *stáyl* 'style, manner' in order to question an adverbial of manner (cf. 117–118 below). However, *ús=káyn* is not found in conjunction with human-denoting generic nouns like *mán* 'man' or *pɔ́sin* 'person' with the meaning of 'who'.

#### **7.3.2.4 Questioning adverbials**

Adverbials are questioned through mono- and bimorphemic question words as well as through question phrases. Adverbials of time may be questioned with the question word *ús=tɛ́n* 'q=time'. This question word is general in its meaning and may question any time unit:

(110) *Ús=tɛ́n* q=time *yu* 2sg *rích?* arrive 'When [which time/day/month/year] did you arrive?' [dj05ce 154]

Nevertheless, speakers prefer to question time units specifically by using the logically most likely option as in the following questions involving the time units *dé* 'day', *mún* 'month', and *hía* 'year', respectively:


#### 7 The clause

In the same vein, time units of the day are often questioned by the more specific bimorphemic question word *ús=áwa* 'q=hour' (114), which may refer to units of the clock as well as periods of the day (e.g. *mɔ́nin tɛ́n* 'morning', *sán tɛ́n* 'noon', *nɛ́t* 'night'):

(114) *Ús=áwa* q=hour *yu* 2sg *rích?* reach 'When [at what period of the day, at what time] did you arrive?' [dj05ce 153]

The generic nouns*sáy* 'side, place' (pervasive) and *pát* 'part, place' (marginal) combine with *ús=* 'q' in order to render 'where' and question locative adverbials. The question word *ús=sáy* tends to have a more general meaning than *ús=pát* 'q=part, place'. The logical option *ús=plés* 'q=place' is accepted in elicitation but not attested in natural speech. Compare (115) and (116):


The bimorphemic question word *ús=káyn* 'which' is also employed as a modifier of the generic noun *stáyl* 'style' in order to question manner adverbials (117–118). Note the subtle difference in meaning between *ús=káyn stáyl* 'by which means' in the following examples and *háw* 'how' further below:


A second and equally common means of questioning manner adverbials is provided by the monomorphemic question word *háw* 'how'. Sentence (119) involves a main clause, example (120) a main and a subordinate clause:

(119) *Háw* how *e* 3sg.sbj *bin* pst *só,* show *háw* how *e* 3sg.sbj *bigín,* begin *háw* how *e* 3sg.sbj *salút* greet *yú?* 2sg.indp 'How did he show [respect], how did he begin, how did he greet you?' [au07se 134]

7.3 Questions

(120) *Háw* how *yu* 2sg *sabí* know *sé* qot *na* foc *rubio?* blond 'How do you know it's light?' [ab03ab 182]

In addition, *háw* may precede the quantifier *mɔ́ch* 'much' and form an independent question word in order to question a quantity (121) as well as the degree to which the property denoted by the property item applies (122–123):


The collocation *háw mɔ́ch* is also used to question quantifiers of count and mass nouns alike. Compare (124) in which a time quantity (hence duration) is questioned, (125) in which a mass nouns is questioned, and (126) in which the count noun *pikín* 'child' is questioned:


Adverbials of cause can be questioned in a number of ways. Firstly *wétin* and *ús=tín* 'what' regularly occur with the meaning of 'why' (127). The use of these two question words may colour the question with reproach if the subject of the clause is human:

(127) *Wétin* what *yu* 2sg *nó* neg *de* ipfv *wók* work *tidé?* today

'Why [how come] come you're not working today?' [ye05ce 171]

#### 7 The clause

The question word *háw* 'how' is used in a similar way in rhetorical questions that call the legitimacy of an addressee's statement, potential answer or behaviour into question. This type of question clause therefore involves the use of the potential mood:


'How would you be [live like] without having children [what a ridiculous thing to demand]?' [kw03sb 203]

*Wétin* and *ús=tín* 'what' also occur in question phrases as the objects of prepositions that may mark NPs for a cause semantic role (cf. §9.1.3 for details). In (130), *wétin* is the object of *fɔ* 'prep', in (131), *wet* 'with' is followed by *ús=tín* , and in (132), the preposition *fɔséka* 'due to' takes *ús=tín* as its object. The resulting phrases all serve to question adverbials of cause. Note that these phrases can optionally appear under cleft-focus like any other question element:


The preposition *fɔséka* 'due to' may also be employed on its own as a question word in a truncated question phrase of the type presented in (133):

(133) *Fɔséka* due.to *in* 3sg.poss *pikín* child *dáy?* die 'Due to (what) did his child die?' [lo07he 056]

The third way of questioning adverbials of cause is via the idiomatic clauses *wétin mék* 'what make' = 'why, how come' and *ús=tín mék* 'what make' = 'why, how come' (134). *Mék* also occurs with the meaning '(to) cause' as a full verb in questions such as (135):

7.3 Questions


There is some variation in the degree of idiomaticity of *wétin/ús=tín mék* 'what make', which is reflected in the degree of "verbiness" of *mék* 'make'. Example (134) above presents the most common way of employing*wétin mék*.The element *mék* is neither modified for a TMA category nor is it accompanied by other characteristics that would point to its status as a verb.

In contrast, the question in (136) is indicative of a more "verby" status of *mék* than in (134). Here, the questioned situation denoted by *wók* 'work' is the predicate of a quotative clause to the main verb *mék*. The quotative marker and complementiser *sé* 'qot' links the main and subordinate clauses:

(136) *Wétin* what *mék* make *sé* qot *yu* 2sg *nó* neg *wók* work *tidé?* today 'How come you didn't work today?' [dj05ce 174]

Sentence (137) below contains the most verb-like instance of *mék*. Here, *mék* 'make' not only functions as a main verb to the complement verb *wók* 'work'. It also induces a subjunctive mood over the complement clause, because it is employed with its lexical meaning as a deontic causative verb (cf. §9.4.4). Equally, the main verb *mék* is fully finite as can be seen by the presence of the imperfective marker *de* 'ipfv':

(137) *Ús=tín* q=thing *de* ipfv *mék* make *sé* qot *mék* sbjv *yu* 2sg *nó* neg *wók* work *tidé?* today 'What is causing you not to work today?' [ye05ce 173]

### **7.3.3 Answers**

In Pichi, *yɛ́(s*) 'yes' is the central agreement interjection. Both *yɛ́* and *yɛ́s* are employed in formal and informal registers alike. Compare the answer in (138b):

(138) a. *Náw* now *yu* 2sg *fít* can *dríng=an* drink=3sg.obj *nɔ́?* intj 'Now, you're able to drink it, right?' [kw03sp 115] b. *Náw* now *so,* like.that *yɛ́s,* yes *a* 1sg.sbj *fít* can *dríng=an* drink=3sg.obj *fáyn.* fine

'Now, I'm able to drink it [milk] well.' [ed03sp 116]

#### 7 The clause

Stronger degrees of agreement can be signalled by other elements. The interjection ɛhɛ́signals emphatic 'yes'. The focus constructions *na só* 'foc so' = 'that's how it is', *na ín* 'foc 3sg.indp' = 'that's it', and *na di tín* 'foc def thing' = 'that's it' also signal strong agreement.

The elements *nó* and *nɔ́* are used as free variants in order to signal disagreement. The former element is identical in form to the general negator *nó*. Many Pichi speakers agree or disagree with the polarity of the question. Hence agreement with the negative polarity of the question in (139) evokes the use of the agreement marker *yɛ́s*:

```
(139) a. So
           so
               yu
               2sg
                   nó
                   neg
                        go
                        pot
                             chɔ́p?
                             eat
           'So you won't eat?' [chfn05 001]
        b. Yɛ́s.
           yes
           'No (I won't eat).' [lifn05 004]
```
In the same way, disagreement with the positive polarity of the question requires the use of the disagreement marker:

(140) a. *Yu* 2sg *go* pot *chɔ́p?* eat 'Will you eat?' b. *Nó.* neg 'No (I won't eat).' [lifn05 005]

However, *yɛ́s* and *nó* are also used to agree or disagree with the proposition of the utterance, possibly through Spanish influence:

(141) a. *So* so *yu* 2sg *nó* neg *go* pot *kán?* eat 'So you won't come?' b. *Nó, a nó go kán.*

no 1sg.sbj neg pot eat 'No, I won't come.' [lifn05 002]

In sentence-final position, *nɔ́* functions as a question-tag, i.e. a conative interjection. In this function, *nɔ́*is used in rhetorical questions as well as in biased questions, in which the speaker expresses the expectation that the answer will correspond to the polarity of the question (142):

(142) *Yu* 2sg *de* ipfv *fíl* feel *hɔ́t* hot *nɔ́?* intj 'You're feeling hot, aren't you?' [ma03hm 007]

7.3 Questions

Aside from that, *nɔ́* also serves as a phatic interjection in order to solicit attention (cf. §12.2.3). For example, the clause in (143) underlines the speaker's commitment to the truth of a story that he has just narrated:

(143) *Nɔ́,* intj *nó* neg *tɔ́k* talk *ɛ́n!* intj 'No, don't talk [and call into question the truth of my story].' [ed03sb 177]

Strong disagreement can be expressed by the focus construction *nóto só* 'neg.foc so' = 'that's not how it is'. The following two sentences succeed each other in a narrative. The disagreement expressed in (144a) is underlined by sentence (b):

(144) a. *Dɛn* 3pl *tɛ́l=an* tell=3sg.obj *sé* qot *"nóto* neg.foc *só."* like.that 'They said to her "that's not how it was".' [ed03sb 045] b. *Tɛ́l* tell *wí* 1pl.indp *trú!* true 'Tell us (the) truth!' [ed03sb 046]

Even stronger disagreement is expressed through the negative phrases *nó wán dé* 'never' (145b) and *nó wé* 'no way' (146b). The following two sentence pairs illustrate their use in signalling disagreement in response to a question:

	- neg one day neg.foc 1sg.indp 'Never, it's not me.' [lo07he 062]
	- b. *Nó* neg *wé,* way *a* 1sg.sbj *nó* neg *go* pot *kán.* come 'No way, I won't come.' [lo07he 064]

Given the right pragmatic context, the question word *ús=sáy* 'where' may signal strong disagreement as well (147b). The imperative clause *kɔmɔ́t dé* 'get lost' can be employed to express strong and abusive disagreement (148b):

(147) a. *Dɛn* 3pl *dɔ́n* prf *gí* give *yu* 2sg *di* def *mɔní?* money 'Have they given you the money?' [pa07fn 478]

#### 7 The clause

	- b. *Kɔmɔ́t* go.out *dé!* there 'Get lost!' [la07fn 579]

The answer to a content questions may be given in full or truncated sentences consisting of the questioned constituent(s) as in (149b):

	- b. *Nguema* name *Mba.* name 'Nguema Mba' [ed03sp 062]

### **7.4 Focus**

The extensive use of focus structures in sentence formation is a distinctive mark of Pichi. Focus constructions have two principal pragmatic functions in the language. Firstly, they serve to present new information. For this function, I employ the term "presentational focus" (Drubig 2003). Secondly, focus constructions serve to assert previously introduced information that runs counter to the presupposition of an addressee. This function is here referred to as "contrastive focus" (Chafe 1976: 35). Focus is realised through three distinct strategies: suprasegmental focus, particle focus, and cleft focus. Cleft focus may also be applied to verbs in so-called predicate cleft constructions (cf. §7.4.5). The language also employs various other means for emphasis, including presentatives (c.f §7.4.4). The syntactic operation of clefting renders elements under cleft focus pragmatically salient. But it is difficult to determine the semantic differences between cleft focus and other types of focus on the basis of the available data.

### **7.4.1 Suprasegmental focus**

The use of focus constructions is intimately tied to suprasegmental phonology. Firstly, focus at the sentence or clause level may be signalled by emphatic intonation (cf. §3.4.2). Extra-high tone may also be employed to focus individual constituents or groups of constituents (cf. §3.2.5). These forms of suprasegmental focus may be freely combined with the different types of focus constructions presented in the following.

7.4 Focus

### **7.4.2 Particle focus**

Particle focus involves the elements *sɛ́f* 'self, emp', *sénwe* 'emp' and the sentence particles/interjections *ɛ́n* 'intj' and *ó* 'sp' (cf. §12.2.4 for a detailed treatmen of these elements). These elements may signal focus of constituents of varying complexity including entire clauses and sentences. Table 7.5 provides an overview.


Table 7.5 Focus particles

#### **7.4.2.1 Forms and functions**

The reflexive anaphor and emphatic particle *sɛ́f* 'self, emp' is the most frequently used form in particle focus. The following sentence presents the use of *sɛ́f* as a reflexive anaphor (cf. §9.3.5 for a detailed treatment):

(150) *Dán* that *gál* girl *e* 3sg.sbj *kin* hab *fíks* fix *in* 3sg.poss *sɛ́f,* self *pént* paint *in* 3sg.poss *sɛ́f.* self 'That girl, she usually does herself up, paints herself [puts on make-up].' [dj07ae 114]

The two successively uttered sentences (151a) and (151b) exemplify the use of *sɛ́f* 'self, emp' in signalling presentational focus. In (151a), the speaker provides information on the topic *dán mán* 'that man'. In (b), the same speaker fills in the information gap in combination with presentational focus of the entire sentence:

(151) a. *Dán* that *mán* man *e* 3sg.sbj *bin* pst *kán* pfv *gó* go *na* loc *jél* jail *lɔ́n* long *tɛ́n.* time 'That man, he went to jail long ago.' [ma03sh 015] 7 The clause

> b. *Náw* now *e* 3sg.sbj *dɔ́n* prf *dáy* die *sɛ́f.* emp 'Now he is even dead.' [ma03sh 016]

The corpus contains a single occurrence of *sɛ́f* 'self, emp' preceded by a 3sg.poss pronoun which is coreferential with the head noun of the focused NP (152). This structure is a dislocated possessive construction in which *sɛ́f* functions as a nominal in the possessed noun position. In the construction, the low-toned 3sg possessive pronoun and *sɛ́f* together signal emphasis or focus of the preceding noun *di bɔ́y* 'the boy':

(152) *Wé* sub *di* def *bɔ́y* boy *in* 3sg.poss *sɛ́f,* self *wé* sub *e* 3sg.sbj *sí* see *mí,* 1sg.indp *estaba* he.was *contento.* content 'And the boy himself, when he saw me, he was content.' [ab03ay 046]

The construction in (152) is, however, marginal. Note the difference between (152) and the following (153). In the latter example, *sɛ́f* 'self, emp' is used as a regular focus particle, postposed to the high-toned 3sg emphatic personal pronoun:

(153) *Ín* 3sg.indp *sɛ́f* emp *gó* go *na* loc *baf-rúm* bath-room *e* 3sg.sbj *wás.* wash 'He (by) himself went to the bathroom (and) washed.' [ab03ab 148]

Contrastive use of *sɛ́f* 'self, emp' is illustrated in (154). In its function as a focus marker, sɛ́f often assumes a reading of inclusive or exhaustive listing; hence the translation of *sɛ́f* as 'too, also, even' and 'alone, without help'. In fact, a postposed *sɛ́f* most appropriately renders the notion 'too, also' in a sentence like (155).

The following two examples also show that *sɛ́f* has the most flexible scope of all particles. It may signal focus of sentences (154) as well as smaller constituents, such as a personal pronoun (155):


The particle *sénwe* 'emp' is presumably a lexicalised collocation (i.e.*sén.wé* 'same.way'). It is employed in the same way as *sɛ́f* in order to signal presentational and contrastive focus (156). The use of *sénwe* as a clausal focus particle is not attested. In general, sénwe occurs less frequently than *sɛ́f* and is found more often to focus personal pronouns than full nouns. Consider the following example, in which *sénwe* signals presentational focus of the personal pronoun *yú* '2sg.indp':

7.4 Focus

(156) *Dí* this *wán,* one *yú* 2sg.indp *sénwe* emp *yu* 2sg *de* ipfv *gó.* go 'This time, you yourself are going [to die].' [ed03sb 040]

The element *ó* 'intj' may signal presentational or contrastive focus of entire clauses as in (158) below. The particle is a sentence-final element which has scope over all preceding material, which may be a predicate-less sentence (157) or a clause (158). However, modification by means of *ó* also colours the sentence with meanings like warning, assertion, empathy, or emphasis (cf. §12.2.4 for more details):

(157) *Bata* place *tɔ́ng* tongue *ó.* sp 'That's the Fang language for you [see how peculiar it is].' [to03gm 014]

(158) *A* 1sg.sbj *bin* pst *dɔ́n,* prf *a* 1sg.sbj *bin* pst *dɔ́n* prf *blánt* reside *fɔ* prep *Gabón* Gabon *ó.* sp 'I've already, I've already lived in Gabon [contrary to what you think].' [ma03hm 035]

The interjection *ɛ́n* 'intj' is principally employed in sentence-final position as a channel-checking device in order to solicit the attention of an addressee (cf. also §12.2.2). Channel-checking automatically lends prominence to a preceding utterance, hence ɛ́n may function very much like other sentential focus particles (159).

(159) *Djunais,* name *yu* 2sg *badhát* be.mean *ɛ́n.* intj 'Djunais you're really mean.' [fr03wt 032]

Beyond that, *ɛ́n* may also occur in mid-sentence followed by a pause, in order to focus a single constituent. In (160), the Spanish depictive adjective *fresco* 'fresh' is fronted and singled out for focus by *ɛ́n*:

(160) *Fresco* fresh *ɛ́n,* intj *dɛn* 3pl *de* ipfv *gí* give *wí.* 1pl.indp 'Fresh, (that's how) they would give (it) to us.' [ed03sp 103]

#### **7.4.2.2 Eligible constituents**

Any sentence constituent may be subjected to particle focus save dependent personal pronouns, determiners, and TMA particles. Equally, the individual elements of multiconstituent NPs cannot be focused, since an NP must be focused in its entirety. Other than that, constituents of varying degrees of complexity may be focused. Sentence (161) features a prepositional phrase with a single noun under focus, and (162) the complex prepositional phrase and reflexive construction *na yu skín* 'on you(r body)':

7 The clause


In dialogue, verbless, prosodically independent sentences can be found which consist of a focused constituent alone. By singling out particular elements in such a way, a speaker may convey strong emphatic force. Compare the discourse excerpt in the two following examples. In (163a) speaker (hi) emphasises the lack of responsibility of certain mothers by utilising focus with *sɛ́f* 'self, emp'. Her statement is confirmed by speaker (bo) in (b):

(163) a. *Bɔt* but *dán* that *káyn* kind *mamá* mother *dɛn* pl *sɛ́f.* emp 'But these kinds of mother, really.' [hi03cb 113] b. *De* of *verdad.* truth 'Really.' [bo03cb 114]

The corpus contains many examples of focused adverbial phrases, in particular time adverbials, such as *tumɔ́ro* 'tomorrow' in (164):

(164) *Wé* sub *a* 1sg.sbj *dɔ́n* prf *jɔ́ch* judge *dɛ́n,* 3pl.indp *a* 1sg.sbj *sé* qot *"tumɔ́ro* tomorrow *sénwe* emp *a* 1sg.sbj *de* ipfv *gó* go *mít* meet *in* 3sg.poss *mán".* man 'When I had talked them down, I said "tomorrow, I'm going to meet her husband".' [ro05rt 023]

Subordinate clauses may be focused by the same means as other, smaller sentence constituents. The relative clause in (165) is under the scope of the particle *sɛ́f* 'self, emp'. In (166), the clause introduced by *sé* 'qot' is under focus by means of the sentence-final particle *ó*:

(165) *E* 3sg.sbj *lúk* look *di* def *análisis,* analysis *tiene* he.has *paludismo* malaria *de* of *una* one *cruz* cross *wé* sub *kin* hab *kíl* kill *pikín* child *sɛ́f.* emp 'She [the doctor] looked at the analysis "he has malaria of one cross which even kills children".' [ab03ab 120]

7.4 Focus

(166) *Bikɔs* because *dɛn* 3pl *tɔ́k* talk *sé* qot *na* foc *paludismo* malaria *ó.* sp 'Because they said that it's malaria.' [hi03cb 124]

Elements which are part of a coordinate structure can be focused separately (167), and there is no restriction save intelligibility on the number of elements that can be focused in one sentence. Compare (167) which features constituent focus by means of the particle *sɛ́f* and clausal focus by means of a sentence-final *ó*:

(167) *Tú* two *pípul* people *sɛ́f* emp *wet* with *wán* one *pikín* child *dɔ́n* prf *kán* come *ó.* sp 'Even two people and one child have come.'

Example (168) presents clausal focus (or alternatively focus of the object NP *dán* 'convence' *dé*) through *sɛ́f*, as well as focus of the ensuing adverbial phrase *na Pichi* by means of *sénwe*:

(168) *A* 1sg.sbj *bin* pst *wánt* want *tɔ́k* talk *dán* that *"convence"* convince *dé* there *sɛ́f* emp *na* loc *Pichi* Pichi *sénwe,* emp *a* 1sg.sbj *nó* neg *de* ipfv *mɛ́mba.* remember 'I had actually wanted to say that "convence" there in Pichi itself (but) I don't remember [how to say it].' [dj05ae 040]

Constituent and verb negation are compatible with particle focus. When used in combination with negation, particle focus produces emphatic negative readings like 'not at all, not even':

(169) *Nó* neg *mán* man *nó* neg *blánt* reside *yá* here *mɔ́* more *sɛ́f.* emp 'Nobody even lives here anymore.' [ra07fn 064]

Personal pronouns can be focused through the use of the corresponding emphatic, independent form alone instead of resorting to *sɛ́f* or *sénwe* (cf. 155–156 above). Compare subject focus in the rhetorical question in (170):

(170) *Mí* 1sg.indp *wánt* want *dán* that *mán?* man 'Do I [emp] want that man?' [ro05rt 026]

Clausal focus by means of *sɛ́f* is also regularly made use of in combination with the conditional clause linker *ɛf/if* in order to render concessive meaning (cf. §10.7.12).

#### 7 The clause

#### **7.4.2.3 Word order and scope**

Focused constituents may appear in situ, i.e. in the same syntactic position assigned to them in focus-neutral clauses. When this is the case, focus is signalled by the presence of a particle. In (171), the subject NP *in papá* 'her father' is highlighted via presentational focus only by means of the post-posed emphatic particle *sɛ́f* :

(171) *Áfta* then *in* 3sg.poss *papá* father *sɛ́f* emp *kán* come *ték=an.* take=3sg.obj 'Then her father came to take her.' [ab03ab 021]

Focused non-subject NPs may also be found in situ together with a focus particle. Compare the focused PP *fɔ di pikín* in (172):

(172) *Áfta* then *e* 3sg.sbj *nóto,* neg.foc *e* 3sg.sbj *nó* neg *fáyn* fine *fɔ* prep *di* def *pikín* child *sɛ́f.* emp 'Then it's not, it's not good for the child itself.' [fr03ft 199]

When an object NP retains its usual syntactic position after the verb and is followed by a focus particle, discourse context and the presence of suprasegmental focus will usually disambiguate the resulting structure as involving clausal or phrasal focus. In (173) the particle *sɛ́f* 'self, emp' may be construed as having narrow scope over the object NP *dán torí* 'that story', or alternatively, broad scope over the entire sentence:

(173) *Mí* 1sg.indp *nó* neg *sabí* know *us* which *mán* man *dɛn* 3pl *kíl,* kill *a* 1sg.sbj *nɔ́ba* neg.prf *hía* hear *dán* that *torí* story *sɛ́f.* emp 'I don't know who was killed, I haven't even heard that story yet.' or ' I don't know who was killed, I haven't heard that particular story yet.' [ro05de 049]

Adverbials may be be focused by exploiting their syntactic flexibility and placing them at the head of the sentence in combination with a focus particle (174).The corpus contains no instance of an object that has been fronted for focus. We only find focused, sentenceinitial non-subjects occuring in cleft constructions (cf. e.g. 190):

(174) *Lagos* place *sɛ́f,* emp *e* 3sg.sbj *gɛ́t* get *di* def *sáy* side *wé* sub *na* foc *di* def *húman* woman *dɛn* pl *de* ipfv *máred* marry *di* def *mán.* man 'Even in Lagos, there is a place where it's the women (who) marry the men.' [hi03cb 177]

In contrast, examples abound, in which we find dislocated, focused core participants other than subjects simultaneously functioning as clausal topics (cf. §7.5 for more details). The overlayering of focus and topic structures in a single sentence, and the identity of topical and focused constituents in Pichi is only natural, since "given", topical elements often also constitute the most important information in a sentence.

For example, sentence (175) features the dislocated and topical object NP *di róp* 'the rope', followed by the focus particle *sɛ́f*. In contrast to fronting (i.e. in question formation),

#### 7.4 Focus

the use of dislocation comes along with the use of a resumptive pronoun (here the 3sg.obj pronoun *=an*) in the original object position of the left-dislocated constituent:

(175) *Áfta* then *di* def *róp* rope *sɛ́f,* emp *wi* 1pl *nó* neg *sí* see *nó* neg *mán* man *wé* sub *e* 3sg.sbj *híb=an.* throw=3sg.obj 'And the rope, we didn't see anybody who threw it.' [li07pe 005]

Sentence (176) contains a left-dislocated object NP, the emphatic pronoun *mí* '1sg.indp', which is reiterated by the coreferential object pronoun *mí* '1sg.indp'. In this example, too, focus of the dislocated topic is overtly signalled by means of the particle *sɛ́f* :

(176) *Mi* 1sg.indp *sɛ́f,* emp *ɔ́l* all *pɔ́sin* person *dɛn* 3pl *kin* hab *áks* ask *mi* 1sg.indp *sé* qot *'yu* 2sg *dɔ́n* prf *bɔ́n?'* give.birth 'Even me, everybody asks me "have you given birth [do you have a child]?"' [fr03ft 152]

Constructions involving personal pronouns are as well the only ones in which "afterthought" apposition is frequently employed in order to signal focus of personal pronouns. Example (177) contains an appositive *mí* '1sg.indp' within the scope of the focus particle *sénwe* and coreferential with the preceding dependent personal pronoun *a* '1sg.sbj':

(177) *A* 1sg.sbj *go* pot *wás=an* wash=3sg.obj *wet* with *mi* 1sg.poss *hán* hand *mí* 1sg.indp *sénwe.* emp 'I myself would wash it with my hand.' [dj07re 049]

### **7.4.3 Cleft focus**

The two elements *na* (affirmative) and *nóto* (negative) are employed in cleft constructions to signal focus of constituents of all degrees of complexity. The focus phrase *es que* 'it is that' is of Spanish origin and forms an integral part of the Pichi focus system. It is employed to cleft focus entire clauses. Some relevant charateristics of these three elements are given in Table 7.6.

#### **7.4.3.1 Forms and functions**

The form *na* 'foc' signals presentational and contrastive focus, *nóto* 'neg.foc' contrastive focus. It is noteworthy that in the vast majority of instances in the corpus, cleft constructions do not exhibit any overt sign of relativisation. Hence in the following sentence, the subordinator *wé* 'sub' is not present in its potential position (indicated by ∅):

(178) *Áfta* then *na* foc *dán* that *tɛ́n* time ∅ sub *a* 1sg.sbj *kán* pfv *gó* go *na* loc *Alemania.* place 'So it's that time that I went to Germany.' [fr03ft 030]

#### 7 The clause


Table 7.6 Cleft focus particles

The negative focus marker *nóto* is employed instead of *na* to signal negative, contrastive focus. In example (179), *nóto* signals contrastive focus of the object pronoun *ín* '3sg.indp'. Note the use of the emphatic form of the personal pronoun as well as the occurrence of a resumptive *=an* '3sg.obj' at the end of the clause:

(179) *Sé* qot *pero* but *mán* man *mi* 1sg.poss *brɔ́da* brother *dát,* that *ús=tín* q=thing *e* 3sg.sbj *de* ipfv *dú* do *na* loc *yá,* here *nóto* neg.foc *ín* 3sg.indp *wi* 1pl *bɛ́r=an?* bury=3sg.obj 'But man, that's my brother, what's he doing here, isn't it him that we buried?' [ed03sb 139]

The Spanish-origin focusing device *es que* 'it is that' is regularly employed to signal presentational focus with clauses and sentences (180):

(180) *Es* it.is *que* that *e* 3sg.sbj *fáyn* fine *wé* sub *yu* 2sg *nó* neg *sabí* know *sé* qot *e* 3sg.sbj *kɔmɔ́t* come.out *fɔ* prep *di* def *animal.* animal

'It's that it is fine when you don't know that it [the milk] has just come out of the animal.' [ed03sp 105]

Cleft constructions may be employed for signalling presentational and contrastive focus alike. In the following three sentences, speaker (ma) talks about a dog that has been tied to a tree by the neighbours downstairs. After providing circumstantial information in (181a–b), new information is introduced by presentational focus in (c):

(181) a. *Dɛn táy=an.*

> 3pl tie=3sg.obj

'They've tied it [that's why it's barking].' [ma03hm 001]

7.4 Focus

b. *Dɛn* 3pl *gɛ́fɔ* have.to *mín* mean *sé* qot *e* 3sg.sbj *de* ipfv *hambɔ́g* irritate *wí.* 1pl.indp 'They must mean to make it irritate us.' [ma03hm 002] c. *Na fɔ mék nó gó na dɔ́n.*

foc prep make neg go loc down 'That's in order for (us) not to go down.' [ma03hm 003]

The use of contrastive focus is exemplified in the discourse excerpt below. In (182a), speaker (dj) jokingly denies any involvement in the spell that has been cast on speaker (dj). Speaker (ru) retorts by contrastively focusing the 2sg pronoun used in addressing his interlocutor in (b):

	- b. *Na* foc *yú* 2sg.indp *mék=an.* make=3sg.obj 'It's you who made it.' [ru03wt 013]

Both *na* 'foc' and *nóto* 'neg.foc' also function as copula-like elements in clauses like (183b), in which a concrete entity is identified in discourse (cf. §7.6.1 for an extensive treatment of the copula functions of *na/ nóto*). Likewise, *na/ nóto* occur as identity copulas in equative constructions like (184), where we find nominal constituents on both sides of the copula:

	- b. *Na* foc *chak-mán.* drunk.cpd-man 'It's a drunkard.'

'The woman is a strong woman.' [dj05ae 200]

Presumably, the identificational function of *na/nóto* in pragmatic contexts like (183b) is the point of departure for the focus-marking and identity (i.e. equative) functions of *na/nóto* (Heine & Kuteva 2002: 96). The difference between copula clauses and cleft focus has a structural correlate. In focus constructions, the out-of-focus part of the sentence is not normally expressed as a relative clause. Compare the pragmatically neutral clause in (185a) and the corresponding focus construction (b), in which the relativiser *wé* is absent (indicated by ∅):

7 The clause

(185) a. *Dɛn* 3pl *sɛ́n* send *di* def *bɔ́l.* ball 'The ball was thrown.' [au07se 169] b. *Na* foc *pɔ́sin* person ∅ sub *sɛ́n* send *di* def *bɔ́l.* ball 'It's a person/somebody who threw the ball.' [au07se 169]

In copula clauses, however, the use of an overt relative clause introduced by *wé* 'sub' is obligatory if the identified entity is to be modified by a clause. In (186a), new information is introduced. This given information is implicitly referred to by sentence (b), which is therefore best seen to constitute an equative clause rather than a focus construction:

(186) a. *Háw* how *yu* 2sg *kin* hab *kɔ́l=an* call=3sg.obj *wé* sub *pɔ́sin* person *de* ipfv *siente* feel *vergüenza?* shame 'How do you call it, when a person feels ashamed?' [ko0505e3] b. *Na* foc *pɔ́sin* person *wé* sub *de* ipfv *fíl* feel *sém.* shame 'That's a person who feels ashamed.' [ro05fe 028]

The difference between copula predication and a focus structure can also be seen in the use of personal pronouns. In a copula construction, a 3sg independent pronoun may be inserted before *na*/*nóto*:

(187) (*Ín*) 3sg.indp *na* foc *wán* one *mán* man *wé* sub *de* ipfv *plé* play *wet* with *di* def *bɔ́l.* ball '(He/that's) a man who is playing with the ball.' [ra07se 038]

By comparison, the insertion of a 3sg peronal pronoun is ungrammatical in the focus construction in (188), since *na/nóto* is non-referential in these constructions. Likewise, a cleft focus construction cannot be rephrased as a presentative clause (cf. §7.4.4):

(188) (*\*Ín*) 3sg.indp *na* foc *wán* one *Annobón* place *gɛ́l* girl *wích* bewitch *yú?* 2sg.indp '(\*She) a girl from Annobón bewitched you?' [fr03wt 002]

#### **7.4.3.2 Eligible constituents and word order**

Cleft constructions allow the focusing of constituents belonging to most word classes. In cleft constructions, the focused constituents invariably appear sentence-initially, irrespective of their syntactic category.

In the overwhelming majority of cases, focused subjects are neither followed by an out-of-focus relative clause, nor are they anaphorically referred to by a resumptive dependent subject pronoun (the latter is usually the case in subject relative clauses). Cleft focus and particle focus may occur together in the same clause as in this example:

7.4 Focus

(189) *Na* foc *Nguema* name *Mba* name *bin* pst *gí* give *mí* 1sg.indp *dán* that *beca* scholarship *sɛ́f.* emp 'It's Nguema Mba (who) actually gave me that scholarship.' [ed03sp 058]

Cleft-focused non-subjects appear at the beginning of the sentence (190). The use of resumptive pronouns is not attested and the expression of the out-of-focus part of the sentence as a relative clause like in (191) is rare:


There are also numerous instances of focused adverbs. Compare the adverb *só* 'so, like this' in (192), which is often encountered in a cleft construction *na só* 'it's like that, that's how it is', as well as focused *dé* 'there' (193):


'It's there that she's going, you know there [that place]?' [ma03hm 029]

The following two examples are of interest because they each present a focus-neutral clause and constituent focus in one sentence. In (194), the manner adverbial *rɔn-sáy* 'backwards' is first encountered in the clause-final adverbial position, then fronted for presentational focus in a *na*-focus construction. The same applies to *fá* 'be far', which is employed as a locative adverbial in (195):


Example (196) contains an instrumental prepostional phrase featuring the preposition *wet* 'with':

#### 7 The clause

(196) *Na* foc *wet* with *ús=tín* q=thing *dɛn* 3pl *bíl=an?* build=3sg.obj 'It's with what that it was built?' [dj07ae 480]

Sequences of the homophones *na* 'foc' and *na* 'loc' are not attested. Hence, the use of a focused locative prepositional phrase featuring *fɔ* 'prep' as a locative preposition serves as an alternative in (197):

(197) *Na* foc *fɔ* prep *dán* that *área* area *wi* 1pl *sté.* stay 'It's in that area that we stay.' [hi03cb 071]

Entire sentences may also be focused by means of the cleft construction. For one part, sentence clefting may be achieved by means of *na/nóto* optionally followed by the quotative marker and complementiser *sé*.

In (198), we witness the use of *na sé* 'it is that' in order to focus a sentence containing the verb *wánt* 'want' together with its subjunctive complements. Besides cleft focus, this sentence exemplifies other features that characterise emphatic speech in Pichi: The TMA marker sequence *dɔ́n de* 'prf ipfv' is employed instead of *de* 'ipfv' alone, and the repetitive use of verbs with similar meanings serves as a means of emphatic reinforcement:

(198) *Wé* sub *yu* 2sg *dɔ́n* prf *de* ipfv *nák,* hit *na* foc *sé* qot *yu* 2sg *wánt* want *sɔn* some *tín* thing *e* 3sg.sbj *brók,* break *mék* sbjv *e* 3sg.sbj *krás,* crash *mék* sbjv *e* 3sg.sbj *destroza.* destroy 'When you're hitting, it's that you want a thing to break, to crash, to be destroyed.' [au07se 245]

*Nóto sé* 'it's not that' always signals contrastive focus of a clause or sentence (199). In (200), a conditional clause is singled out for focus. Hence, the negative focus marker *nóto* appears after *ɛf* 'if':


'If it's not that [the correct type] you've bought, that woman would beat you until you would go return it.' [ab03ab 033]

The Spanish-derived focus phrase *es que* 'it's that' consists of the 3sg present tense form of the Spanish copula *ser* and the complementiser *que* 'that'. The phrase is firmly

7.4 Focus

entrenched in the Pichi lexicon and signals affirmative focus of entire sentences. The phrase has an equivalent function in Spanish:

(201) *Es* It's *que* that *está* it.is *bien* good *usar* use *el* the *subjuntivo.* subjunctive 'It's that it's good to use the subjunctive (mood).'

In (202), the topical NP *dí káyn pikín* 'this kind of child' is set off from the rest of the sentence by continuative intonation and a pause. The subsequent clause is under presentational focus with *es que* 'it's that', and the topical NP is picked up by the resumptive pronoun *e* '3sg.sbj':

(202) *Entonces* so *dí* this *káyn* kind *pikín,* child *es* it.is *que* that *normalmente* normally *e* 3sg.sbj *go* pot *tɛ́l* tell *yú* 2sg.indp *dán* that *tín,* thing (*…*) So this kind of child, it's that usually it will tell you exactly that (…)' [to03gm 052]

In the example below, *es que* is immediately followed by a locative adverbial, namely the prepositional phrase introduced by *na* 'loc':

(203) *Es* it.is *que* that *na* foc *dán* that *klém* climb *wé* sub *e* 3sg.sbj *de* ipfv *klém,* climb *e* 3sg.sbj *de* ipfv *gó* go *e* 3sg.sbj *de* ipfv *klém.* climb

'It's that in that climb that she's climbing, she's just climbing along.' [au07se 070]

Cleft focus is characterised by a large degree of syntactic flexibility. For example, focusing into a relative clause is permitted. Example (204) presents a subject relative clause featuring focus of a 3sg person (i.e. *na ín* 'it's him'), anaphoric to the preceding head nominal *wán* 'one (person)':

(204) *Bɛt* but *e* 3sg.sbj *fíba* resemble *sé* qot *wán* one *dé* be.loc *wé* sub *na* foc *ín* 3sg.indp *de* ipfv *púl* pull *di* def *ɔ́da* other *wán* one *di* def *torí.* story *Lit.* 'But it seems that one is there that it's him who is telling the other one a story.' [au07se 100]

#### **7.4.3.3 Focus of resumptive elements**

Cleft constructions of the type in (204) above, where a resumptive element is focused, serve an important function in discourse. They serve as anaphors that establish reference to preceding topical material in the sentence or the paragraph. The relevant collocations

#### 7 The clause

involve the focus particle *na* 'foc' followed by the adverbs *yá* 'here', *dé* 'there', *só* 'so, like that', the personal pronoun *ín* '3sg.indp', as well as complex NPs like *dán tɛ́n* 'that time' and *di tín* 'the thing'. Mostly, these collocations function as resumptive adverbials of location, time, or cause, but ín '3sg.indp' may also refer to preceding subjects and objects.

In (205), the topical, clefted adverbial phrase *frɔn in hós* 'from her house' is anaphorically referred to by another clefted adverbial, namely *dé* 'there':

(205) *Na* foc *frɔn* from *in* 3sg.poss *hós,* house *na* foc *dé* there *yu* 2sg *go* pot *ték* take *máred.* marry 'It's from her house, it's there that you'd enter marriage.' [ab03ay 033]

A similar anaphoric relation holds between *di sáy* 'the place' and *na dé* 'it's there' in (206). In fact, the deictic locative adverbs *dé* 'there' and *yá* 'here', as well as the deictic manner adverbial *só* 'like that' need to be clefted in this way, if they are to appear in the clause-initial, rather than their usual clause-final position:

(206) *Di* def *sáy* side *wé* sub *mɔní* money *dé,* be.loc *na* foc *dé* there *yu* 2sg *gɛ́fɔ* have.to *gó.* go 'The place where there's money, that's where you have to go.'

An anaphoric temporal relation may also be established by means of the locative adverbs *yá* 'here' and *dé* 'there'. In (207), the left-dislocated and topical Spanish adverbial *a los quince años completamente* is picked up by the resumptive focus construction *na yá* 'foc here' = 'that's when'. The same principle is at work in (208), where *na dé* refers to a preceding time clause earlier in the paragraph:


The collocation *na ín* features the emphatic 3sg pronoun *ín*, which functions as a "catch-all" anaphora. Hence, it may refer to a preceding subject, object, or time or cause adverbial. The exact nature of the anaphoric relation that holds between *na ín* and its antecedent is therefore determined by context.

In (209), *na ín* refers to the antecedent subject under focus *na di fáyn chɔ́p* 'it's the good food'. Example (210) features a resumptive *na ín* anaphorical to the dislocated, topical object *dís traje fɔ mono* 'this overall-like suit':

7.4 Focus


In (211), *na ín* refers to an antecedent time clause introduced by *wé* 'sub'. When there is a relation of temporal succession like in this example, it is only natural that the *wé*-clause precedes the main clause:

(211) *Wé* sub *e* 3sg.sbj *dɔ́n* prf *dé* be.loc *pan* on *di* def *chía,* chair *na* foc *ín* 3sg.indp *e* 3sg.sbj *strét.* be.straight 'When she was completely on the chair, that's when she straightened up.' [au07se 089]

In turn, cause clauses are more likely to follow their main clauses. As a consequence, sentence-initial cause clauses are in-focus by default, and are therefore quite often additionally marked for focus in a cleft construction.

Whenever this the case, the phrasal expressions *na ín (mék)* 'foc 3sg.indp (make)' = 'that's why' (212) or alternatively, *na di tín (mék)* 'foc def thing (make)' = 'that's why' (213) may refer anaphorically to the preceding cause clause (cf. 134 for the analoguous content question):


Amongst the sentences involving focus of resumptive elements presented so far, we also find focused constituents appearing in the initial position which are not preceded by the focus marker *na* (eg. 207 and 210). There is no reason to see these structures as being fundamentally different from cleft constructions involving the focus marker *na*. The only thing "missing" in these constructions is the focus particle.

### **7.4.4 Presentatives**

Pichi features a presentative construction involving *na*/*nóto* as well as the proximal and distal demonstrative forms *dís* 'this' (214) and *dat* 'that' (215) in sentence-final position.

#### 7 The clause

Presentatives may be seen as inverted copula clauses with particular deictic force, which direct an addressee's attention to, and identify, an entity. By highlighting an entity in this way, presentatives manifest a functional overlap with (presentational) cleft constructions:


Examples (214–215) may also be expressed with less deictic force as regular copula clauses. The following two equative clauses feature the demonstratives *dí* 'this' and *dán* 'that' in the ordinary prenominal position. When employed in an NP in this way, demonstratives may be realised as the short forms *dí* and *dá* respectively. However, these apocopated forms do not occur in sentence-final position in presentatives like (214–215) above:


The highlighted NP of a presentative construction may be modified by further constituents in the same way as a nominal participant in an equative clause. In (218), the NP *chɔ́p* 'food' has been modified prenominally by *bɛ́ta* 'very good' and post-nominally by a relative clause introduced by *wé* 'sub':

(218) *Na* foc *di* def *bɛ́ta* very.good *chɔ́p* food *wé* sub *mán* man *de* ipfv *chɔ́p* food *dát.* that 'That was the best food that one [I] was eating.' [ed03sp 123]

Content questions may also be formulated as presentatives (219), in which case they may occur without a preposed *na.* This distribution may be linked to the fact that questioned constituents are focused by default, and may optionally co-occur with *na*-focus anyway (cf. §7.3.2):

(219) *Sé* qot *papá* father *gɔ́d* God *ús=káyn* q=kind *trɔ́bul* trouble *dís?* this '(I) said God, what (kind of) trouble (is) this?' [ab03ab 082]

7.4 Focus

### **7.4.5 Predicate cleft**

Besides focus of verbs by means of clausal focus, verbs may be singled out for focus individually in a construction termed "predicate clefting" (e.g. Koopman 1984; Larson & Lefebvre 1991) or "verb fronting" (Muysken 1978).

In Pichi predicate cleft constructions, the focused verb appears twice in the sentence: fronted in the initial focus position directly after the focus marker *na*, and at the same time in its original syntactic position in the out-of-focus part of the sentence. Compare the following example featuring the clefted dynamic verb *gó* 'go'. It is noteworthy that a negative predicate cleft by means of *nóto* 'neg.foc' is not attested:

(220) *Na* foc *gó* go *a* 1sg.sbj *de* ipfv *gó* go *ó.* sp '[Mind you] I'm going.' [ch07fn 151]

Predicate cleft signals presentational or contrastive focus of the predicate and produces intensifying, emphatic meanings. It should therefore be seen as part of the range of emphatic structures that involve iteration in Pichi (i.e. reduplication and repetition, cf. §4.5 and the use of cognate objects, cf. §9.3.3). Neither temporal or causal adverbial meanings, nor factive clauses are expressed through predicate cleft. In natural speech, predicate cleft almost exclusively occurs with dynamic verbs, as in the example above. In fact, the natural speech data in my corpus reveals relatively few instances of predicate cleft constructions in general.

The predicate cleft construction in (221) features a stative verb, the property item *bíg* 'be big'. Like other cleft constructions, predicate cleft does not require marking of the out-of-focus part of the sentence as a relative clause. This is, in fact rejected in unison by all speakers who were asked about this possibility (222):


Sometimes verbs are clefted together with a pronominal object (223). If this is the case, the pronominal object is not repeated with the second verb. The fronting of a subject or adverbial modifier together with the verb is not accepted (224):

(223) *Na* foc *krách=an* scratch=3sg.obj *yu* 2sg *de* ipfv *skrách.* scratch 'You're actually scratching it.' [dj07ae 386] 7 The clause

(224) \* *Na* foc *lúk* look *fáyn* fine *yu* 2sg *lúk.* look Intended: You looked really well. [ne07fn]

However, verbs are not clefted together with TMA markers (225). These always remain in their "original" position with the second verb. The two following examples are of interest because they involve clefting of the major verb of a motion-direction SVC. As these examples show, the minor verb *gó* 'go' remains in its original syntactic position (226):

(225) *Na* foc *wáka* walk *wi* 1pl *bin* pst *de* ipfv *wáka* walk *gó* go *dé.* there 'We actually walked there.' [pa07me 002]

(226) *Na* foc *wáka* walk *wi* 1pl *wáka* walk *gó* go *dé.* there 'We walked there.' [pa07me 003]

The same holds for complements of auxiliaries. In (227), it is once again only the major verb *wáka* 'walk' that gets fronted, while the modal auxiliary verb *wánt* 'want' stays behind:

(227) *Na* foc *wáka* walk *e* 3sg.sbj *wánt* want *wáka* walk *só.* like.this 'He really wants to walk right now.' [pa07me 008]

A few Pichi verbs have homophonous nominal counterparts which are not merely action nominalisations. One of these is *chɔ́p*, which means 'eat' as a verb and 'food' (rather than only 'eating') as a noun. While (228) may be interpreted as involving either predicate or nominal cleft, the cleft construction in (229) is unlikely to be anything else than a nominal cleft construction, since the focused noun *chɔ́p* 'food' is modified by *bɔkú* 'be much':

(228) *Na* foc *chɔ́p* eat/food *e* 3sg.sbj *chɔ́p* eat *yɛ́stadé* yesterday *ó.* sp 'He really ate yesterday.' or 'It's (really good) food that he ate yesterday.' [dj07ae 463]

(229) *Na* foc *bɔkú* much *chɔ́p* food *e* 3sg.sbj *kin* hab *chɔ́p.* eat 'It's a hell of a lot of food that he usually eats.' [dj07ae 462]

7.4 Focus

### **7.4.6 Other means of expressing emphasis**

Focus constructions frequently come along with a variety of other emphatic elements and structures which breathe life into discourse and signal speaker involvement.

For example, the TMA marker sequences *dɔ́n de* 'pfv ipfv' and *dɔ́n de fínis* 'pfv ipfv finish', rather than the imperfective marker *de* alone, may be recruited in order to emphasise that the situation designated by the verb is in full course.

In (230), NP focus (i.e. *dís wán sɛ́f* 'this one emp' co-occurs with a predicate featuring the perfect marker *dɔ́n* and the imperfective marker *de*. Sentence (231) additionally features the completive aspect auxiliary verb *fínis* 'finish (doing something)', which adds even more emphatic force:


'Now I'm really ashamed to be wearing that (pair of) shoes.' [ma03hm 021]

Adverbial modification, for example via the value property items*fáyn* 'be fine' and *bád* 'be bad' or the quantity property item *bɔkú* 'be much' (232), may also express emphasis by itself or in conjunction with other elements and/or focus constructions. The use of the demonstrative determiner *dán* 'that' together with the possessive construction *in yáy* 'his eye' builds up additional emphatic force in (232):

(232) *E* 3sg.sbj *de* ipfv *para* stand *na* loc *dán* that *in* 3sg.poss *yáy* eye *bɔkú* much *bád.* bad 'It [the white spot in his eye] just sits there in that his eye really bad.' [ye03cd 109]

Other means of expressing emphasis and by extension various nuances of sentential focus are the segmental and suprasegmental means outlined in §3.2.5, §3.4.2 and §7.7.3, the various forms of iteration, i.e. repetition (233) and reduplication, predicate cleft, and cognate objects – the latter in combination with the particle *ó* in (234) as well as ideophones (235):


7 The clause

(235) *Dɛn* 3pl *nák=an* hit=3sg.obj *na* loc *in* 3sg.poss *chɛ́s* chest *kip.* ideo 'They hit him (hard) in the chest with a thumping sound.' [dj05ce 100]

### **7.5 Topic**

Topicalisation involves dislocation: The topic appears at the beginning of the sentence and is reiterated in the original syntactic position by a resumptive pronoun. A topic is often set off from the remainder of the sentence by a short pause and a continuative boundary tone. The element *náw* 'now' may optionally function as a post-posed topic marker.

### **7.5.1 Dislocation**

There is a strong tendency for definite subject NPs to be marked as topical by an intonation break, i.e. a short pause and/or continuative intonation, and a resumptive subject pronoun (cf. also §7.1.1). The definite subject in (236) is set off from the rest of the clause by an intonation break, indicated by a comma. At the same time, the following coreferential resumptive pronoun *e* '3sg.sbj' reiterates the topical subject NP *dán skúl* 'that school':

(236) *Dán* that *skúl,* school *e* 3sg.sbj *dé* be.loc *nía* near *bɛrin-grɔ́n,* burial.cpd-ground *nɔ́?* intj 'That school is near the cemetery, right?' [ma03hm 018]

In contrast, the data does not contain a single instance of a resumptive subject pronoun in a clause featuring an indefinite subject. Such clauses are formed in the way of (237) without a resumptive pronoun:

(237) *Wán* one *dé* day *wán* one *pikín* child *bin* pst *de* ipfv *sík.* sick 'One day, a child was sick.' [ye03cd 071]

Non-subject topical NPs also appear at the beginning of the sentence, are normally separated from the rest of the clause by an intonation break, and are referred to by a resumptive element in the clause. The dislocated object *di cartón* 'the carton' in (238) is resumed by the coreferential object pronoun *àn* '3sg.obj':

(238) *Dí* this *cartón,* carton *e* 3sg.sbj *mít=an* meet=3sg.obj *yá?* here 'This cardboard box, did she find it here?' [li07pe 070]

In (239), the topical object NP *ɛ́ni tín* 'everything' is reiterated by the resumptive, coreferential object prounoun *=an* '3sg.obj' after the verb *púl* 'remove':

7.5 Topic

(239) *Ɛ́ni* every *tín,* thing *yu* 2sg *wɔ́nt* want *púl=an* remove=3sg.obj *na* loc *puerto* harbour *yu* 2sg *de* ipfv *pé.* pay 'Everything, you want to remove it from the port, you pay [tax].' [f103fp 002]

Sentence (240) involves the initial, dislocated topical object pronoun *mí* '1sg.indp', which is reiterated in the object position after *sí* 'see' and anaphorically referred to by *a* '1sg.sbj':

(240) *Mí,* 1sg.indp *lɛk* like *háw* how *yu* 2sg *de* ipfv *sí* see *mí,* 1sg.indp *a* 1sg.sbj *dɔ́n* prf *sí* see *plɛ́nte* plenty *tín.* thing 'As for me, as you see me (now), I've seen many things (in life).' [ab03ab 023]

The resumptive pronoun of an antecedent, dislocated topic may also be focused in a cleft construction. Such cross-cutting topic-focus structures are very common in Pichi. In the following sentence, the topical subject NP *Panyá* 'Spain' is picked up by the coreferential 3sg.indp pronoun *ín*, which is, in turn, focused in a cleft construction (cf. also §7.4.3.3):

(241) *Panyá,* Spain *na* foc *ín* 3sg.indp *wɔ́s* be.very.bad *mɔ́.* more 'As for Spain, that's what's really terrible [as a place to live in].' [07fn 040]

Certain types of adverbial clauses are more likely to precede their main clauses than to follow them. When such adverbial clauses do precede their main clauses they usually are topical, and may be set off from the following part of the sentence by an intonation break as well. Compare the purpose clause beginning with *fɔ* 'prep' in (242):

(242) *Fɔ* prep *tɔ́k* talk *Píchi,* Pichi *yu* 2sg *nó* neg *níd* need *fɔ* prep *gó* go *skúl.* school 'In order to talk Pichi, you don't need to go to school.' [au07se 267]

Sentence (243) involves the rather rare case of a right-dislocated, topical (and nominalised hence non-finite) clause namely *fɔ pút nivel* 'to level the ground'. This last example also shows once more that the transition is smooth to focus marking, since (243) may also be seen as an example of pseudo-clefting:

(243) *Di* def *tín* thing *wé* sub *bin* pst *dé* be.loc *difícil* difficult *mɔ́* more *na* loc *dí* this *hós,* house *fɔ* prep *pút* put *nivel.* level 'The thing that was most difficult in [building] this house, [was] to level the ground.' [07fn 065]

### **7.5.2 Topic particle**

It has been shown that dislocation and intonation are by themselves sufficient means of indicating the topicality of a constituent. In addition to dislocation, the adverbial *náw*

#### 7 The clause

'now' may optionally indicate the topicality of a constituent. A particle is, however, not obligatory, often accompanied by an intonation break, and in most cases, by a resumptive element in the clause.

Apart from being used to signal topicality, the particle *náw* 'now' is a time adverbial (244), which may occur in presentational sentences like the following:

(244) *Náw* now *e* 3sg.sbj *tínap* stand *na* loc *grɔ́n.* ground 'Now it's standing on the ground.' [li07pe 093]

Sentence (245) below is a metacomment in which speaker (dj) classifies the term *mɔnt* 'month' as an English word (a more current Pichi term is *mún* 'moon, month'). In this example, the post-posed particle *náw* signals the topicality of *mɔ́nt*.

(245) *"Mɔ́nt"* month *náw,* now *e* 3sg.sbj *dɔ́n* prf *bí* be *inglés.* English 'As for "mɔnt", it's already English.' [dj05ce 030]

Sometimes we encounter sentences in which the topic is not reiterated in a syntagmatic relation within the clause. In such cases, the topic functions like in many topicprominent languages: It is adjoined to the clause and provides a referential frame, within which the precise relation between topic and comment is recovered by pragmatic context (cf. Li & Thompson 1976). For example, in (246), the topicality of *pikín* 'child(ren)' is signalled by *náw* 'now' and an intonation break. However, the "resumptive" pronoun *e* '3sg.sbj' does not refer to the topical syntactic subject *pikín*. Instead, *e* '3sg.sbj' refers to a concept as a whole, namely procreation, which is loosely referred to by the topic *pikín*:

(246) *Pikín* child *náw,* now *e* 3sg.sbj *nó* neg *hád.* hard 'As for [having] kids, that's not difficult.' [hi03cb 162]

Example (247) presents the topical and focused NP *sósó Píchi* 'only Pichi', however without the focus marker *na* 'foc'. The topic is followed by *náw* now' and fronted. The out-of-focus part of the sentence is exceptionally expressed in a relative clause:

(247) *Sósó* only *Píchi* Pichi *náw* now *wé* sub *wi* 1pl *de* ipfv *tɔ́k.* talk '(It was) only Pichi that we used to talk.' [au07se 214]

Example (248) below features the 3sg personal pronoun *ín* '3sg.indp' under assertive focus by means of clefting and additional topic marking by means of a post-posed *náw* 'now':

7.6 being andhaving

(248) *Na* foc *ín* 3sg.indp *náw* now *a* 1sg.sbj *bin* pst *de* ipfv *chɛ́k* check *sé* qot *e* 3sg.sbj *bin* pst *fɔ* cond *dé* be.loc *fáyn* fine *if* if *a* 1sg.sbj *mít* meet *wán* one *pɔ́sin* person (*…*) 'That's why I was thinking it would be fine if I met somebody (…)' [fr03ft 176]

### **7.6 being and having**

The forms employed to express being and having in Pichi form part of a web of interlinked and overlapping functions which extends from the formation of focus structures and copula clauses to the expression of possession and the formation of predicate adjective clauses. An important feature of the expression of both being and having is the notion of time-stability.

Figure 7.2 Expression of being and having

Figure 7.2 maps the linkages between the different elements that participate in the expression of being and having. Time-stable situations are connected with an unbroken, non-time-stable states with a broken line. Glosses for the elements contained in the figure can be culled from the following sections and Table 7.8.

### **7.6.1 Core copulas**

The expression of identity-equation is provided by the elements *na* 'foc', *nóto* 'neg.foc', and *bí* 'be'. The element *dé* 'be.loc' serves as the locative-existential copula. Pichi employs overt copulas in all relevant contexts. The expression of being is characterised by

#### 7 The clause

several asymmetries. Firstly, there is a functional and formal differentiation between the expression of identity (via *na*/*nóto*) and location-existence (via *dé*). Secondly the expression of identity is taken care of by the three suppletive forms *na* 'foc', *nóto* 'neg.foc', and *bí* 'be' which are in complementary distribution with each other. Some relevant characteristics of the distribution of the Pichi core copulas are summarised in Table 7.7.


Table 7.7 Core copulas

Clauses involving the three core copulas *na* 'foc', *nóto* 'neg.foc' and *bí 'be'* feature a subject, the copula and a nominal complement. The functions of the copula include expression of the identity of two participants (249), and classification as member of a group (250):


Further functions are the attribution of a role (251), a name (252), and the expression of a family relationship (253). Note the presence of the verb *tɔ́n* 'turn' which denotes a change of state when used as a copula verb (251):


7.6 being andhaving

(253) *Na* foc *dán* that *tɛ́n* time *a* 1sg.sbj *kán* pfv *sabí* know *sé* qot *mi* 1sg.poss *mamá* mother *na* foc *mi* 1sg.poss *mamá.* mother 'It's then that I came to know that my mother was my mother.' [fr03ft 019]

Equative clauses are characterised by asymmetries and suppletion in the use of personal pronouns, polarity, and TMA marking. These asymmetries derive from the core function of *na*/*nóto* to express identification in presentational sentences like (254) and (255). In these clauses, the identified elements (i.e. *kasára* 'cassava' and *wi Píchi* 'our (kind of) Pichi') are in focus by default. Therefore, I consistently gloss *na*/*nóto* as foc and neg.foc, respectively, in order to render the chiefly pragmatic function of these elements:


In sentences like the two above, *na* has expletive reference and is therefore nonreferential. The core pragmatic function of identification of *na*/*nóto* can be extended to express identity between two full NPs (hence with default 3sg reference) in equative clauses:

(256) *In* 3sg.poss *papá* father *na* foc *chino.* Chinese 'Her father is Chinese.' [ed03sp 028]

However, when identity between a personal pronoun with reference other than 3sg and another NP is expressed, the deeply pragmatic nature of the copula-like element in sentences like (256) above is revealed. Since *na*/*nóto* is not a copula "verb", the subject pronoun cannot come from the dependent series of the pronominal paradigm. Instead, an independent emphatic pronoun must be used:

(257) *Mí* 1sg.indp *na* foc *di* def *wan-grén* one.cpd-grain *pikín.* child. 'I am the only child.' [lo07he 060]

Therefore even equative clauses are best analysed as identificational. These clauses are grammaticalised topic-comment structures, in which the topical subject is followed by an entity identified by *na*/*nóto*. The copula-like element *na*/*nóto* therefore retains its pragmatic, identificational, and focus-marking function even in such "copula clauses".

The two asymmetries in the formation of copula clauses next to negative suppletion (i.e. 255 and 3sg default reference, i.e. 254 and 256) are complemented by a third asymmetry: Whenever overt TMA marking is required or the copula is employed in a context

#### 7 The clause

suggesting reduced finiteness, the copula verb *bí 'be'* is made use of. This complementary distribution is strict. Therefore, a clause like the following one is ungrammatical, since *bí* may not appear in basic identity clauses without overt TMA marking. Compare (256) above and (258) below:

(258) \* *In* 3sg.poss *mamá* mother *bí* foc *rusa.* Russian Intended: 'Her mother is Russian.' [dj07ae 532]

In the following two equative clauses, the presence of the TMA markers *dɔ́n* 'prf' (259) and *go* 'pot' (260) motivates the appearance of the suppletive identity copula *bí* 'be'. In spite of its defective distribution (cf. 258), the copula *bí* behaves much more like a copula verb than *na*/*nóto:* It may take dependent personal pronouns (e.g. in 259) and appear with TMA marking (e.g. 259 and 260).


Sentence (261) below contains two copula clauses. The first one features the copula *bí* marked for past tense by *bin* 'pst'. In contrast, the second clause is not overtly marked for tense, hence the copula cum focus marker *na* is employed. Recall that Pichi employs relational tense. Hence the identity copula *na* may have past tense reference because tense reference has been anchored in the past by the use of *bin* in the preceding clause. In fact, in this example, a past tense reference of *na* is a plausible option because the speaker's mother is deceased (unless the speaker considers reference to her mother to be generic in nature):

(261) *Mi* 1sg.poss *ném* name *bin* pst *bí* be *Francisca* name *Belobe* name *Toichoa,* name *porque* because *mi* 1sg.poss *mamá* mother *in* 3sg.poss *ném* name *na* foc *Belobe* name *Toichoa.* name

'My name was Francicsa Belobe Toichoa because my mother's name is/was Belobe Toichoa.' [fr03ft 090]

A further example involving overt TMA marking in an equative clause follows. Sentence (262) features the narrative perfective marker *kán* 'pfv' followed by *bí* 'be'. Note that the combination of *kán* 'pfv' with the copula *bí* renders a change of state reading of *bí* just like with any other (inchoative-)stative verb (cf. e.g. 27–28):

7.6 being andhaving

(262) *So* so *mí,* 1sg.indp *mi* 1sg.poss *yón* own *e* 3sg.sbj *kán* pfv *bí* be *una* def *desgracia.* disgrace 'So as for me, mine [my matter] came to be a disgrace.' [ab03ay 034]

*Bí* 'be' is also employed instead of *na*/*nóto* in contexts of reduced finiteness. In (263), *bí* occurs as the complement of the modal verb *fít* 'can'. The form *bí* also appears in subjunctive clauses (264). Such clauses are not only inherently future-referring and nonassertive. They also feature reduced tense-aspect marking and are less finite:


Furthermore, *bí* is the only identity copula attested in a context like (265) below. In the example, the copula occurs in a subordinate clause featuring the clause linker *wé* 'sub'. The non-assertive environment of the subordinate clause precludes use of *na*/*nóto* as copulas. This is presumably due to the fact that these particles realise their core function in identificational and presentational sentences, which are assertive structures *par excellence*.

Additionally, tense reference of the subordinate clause is dependent on the main clause, which is set in the past. These factors contribute to the use of *bí* although the context is finite and there is no overt TMA marking in the subordinate clause in (265):

(265) *Frɔn* from *wé* sub *a* 1sg.sbj *bí* be *pikín* child *a* 1sg.sbj *bin* pst *wánt* want *kɔmɔ́t* go.away *na* loc *dís* this *kɔ́ntri.* country 'From when I was a child, I wanted to leave this country.' [ro05ee 027]

A copula clause featuring *bí* 'be' is negated like any other verbal clause. The negator *nó* 'neg' appears in its usual position in the predicate. Compare the following sentence, in which the copula clause is in the potential mood:

(266) *E* 3sg.sbj *nó* neg *go* pot *bí* be *mecánico.* mechanic 'He won't be a mechanic.' [dj05ae 215]

The element *dé* 'be.loc' functions as a locative-existential copula. Accordingly, this form is used to express relatively transient, less permanent existence in space and time, either on its own or when followed by an adverbial complement. The element *dé* also occurs as a copula in predicate adjective constructions (cf. §7.6.5). Hence *dé* may also take adjectives as complements.

#### 7 The clause

The copula *dé* may occur in intransitive clauses without any complement. Such clauses show that *dé* is semantically relatively rich and has a meaning of its own, namely 'exist in a place' or 'exist in a certain manner'. Compare the question in (267a) and the corresponding answer in (267b):

(267) a. *Ebongolo* name *dé?* be.loc 'Is Ebongolo around/his usual self/fine/alright?' [ge07fn 180] b. *Yɛ́s,* yes *e* 3sg.sbj *dé.* be.loc 'Yes, he's here/around/his usual self/fine/alright.' [he07fn 181]

In (268), *dé* takes a locative adverbial phrase introduced by the general locative preposition *na* 'loc' as a complement. The adverbial phrase in (269) involves the locative noun *nía* 'be near':


Locative complements of *dé* 'be.loc' other than locative adverbs like *yandá* 'yonder' in (270) rarely appear without a preposition or a locative noun. Where they do, the absence of the locative noun is usually lexically determined. Compare *dé láyf* 'be.loc life' = 'be alive' in (271). Also note that the copula *dé* receives an imperfective, present tense interpretation like any other unmarked stative verb in Pichi:


Sentence (272) exemplifies how *dé* is used to express existence in time. In contrast to locative complements, time adverbials like *ívin tɛ́n* 'evening' appear as direct complements of the copula *dé* when the intended meaning is 'location in time' (cf. §8.2.2 for other temporal relations):

(272) *Wi* 1pl *dé* be.loc *íbin* evening *tɛ́n.* time 'It's evening.' [dj05ce 249]

7.6 being andhaving

Further, the time is always told in a codemixed Pichi-Spanish construction. The noun phrase employed in telling the time in Spanish appears as a complement of the copula *dé* 'be.loc' which in turn takes a 1pl subject (273). No prepositions are employed in this construction either. Hence here too, there is no formal indication of the adverbial status of the time expression:

(273) *Wi* 1pl *dé* be.loc *las* the.pl *dos* two *y* and *media.* half 'It's two thirty.' [dj05ce 056]

The form *dé* 'be.loc' may also be employed to attribute a relatively transient, non-timestable property to a subject. Hence, *dé* is encountered as a predicator in predicate adjective constructions involving the few adjectives that Pichi has. One of these is *fáyn* '(be) fine' in (274). As explained in detail in §7.6.5, predicate adjective constructions, rather than verbal clauses, are only chosen when the situation is perceived as non-time-stable:

(274) *Dán* that *tɛ́n* time *a* 1sg.sbj *dé* be.loc *fáyn.* fine 'That time I was [feeling] fine.' [ru03wt 024]

Another manifestation of the non-time stable character of the situation predicated by *dé* 'be.loc' is given in the following three sentences. The copula dé is used when an adverbial complement designates a way of being rather than intrinsic being. Adverbial complements can be a simple manner adverb like *só* 'so' (275), a bare noun phrase featuring the generic noun *stáyl* 'manner, style' (276), or a prepositional phrase with the similative and equative preposition *lɛk* 'like' (277):


'I want to be like Miguel Ángel [the way he dresses/acts/looks].' [ye07ga 007]

By extension, *dé* 'be.loc' is also employed whenever an attributed property is questioned directly (278) and indirectly (279), or when a property is attributed to a main clause verb in a free adverbial manner clause (280):

(278) *Háw* how *yu* 2sg *go* pot *dé,* there *yu* 2sg *nó* neg *gɛ́t* get *pikín?* child 'How would you be [feel] (if) you had no child?' [kw03sb 203]

#### 7 The clause


Contrary to the time stable copulas *na*/*nóto* and *bí* described above, *dé* exhibits no irregularities with respect to TMA marking and negation (281). It occurs with the standard negator and any TMA marker compatible with its distribution as a stative verb (282):


### **7.6.2 Copula verbs**

Besides the core system of copula expression covered in the previous section, Pichi recruits a number of stative and dynamic verbs in order to express more specific copula meanings linked to the notions of change-of-state and existence in place and time. Copula verbs and their meanings are provided in Table 7.8.


Table 7.8 Copula verbs

7.6 being andhaving

When employed as a lexical verb, *tɔ́n* means 'turn, stir' (284). In its literal sense, *tɔ́n* is employed as a dynamic verb with an agent subject and a patient object (283), or a locative adverbial (284). The collocation *tɔ́n bák* means 'return' (285):


As a copula verb, *tɔ́n* 'turn' designates a change of state from one identity to another (286):

(286) *Dɛn-ɔ́l* 3pl.cpd-all *dɛn* 3pl *dɔ́n* prf *tɔ́n* turn *europeos* European.pl *dɛn.* pl 'They have all turned into Europeans.' [fr03ft 149]

In contexts other than copula expression and causative formation, *lɛ́f* may be employed as a dynamic verb in transitive clauses with the meaning 'leave (behind)' (287):

(287) *A* 1sg.sbj *lɛ́f* leave *di* def *tín* thing *dɛn* pl *di* def *sáy* side *wé* sub *yu* 2sg *bin* pst *tɛ́l* tell *mí.* 1sg.indp 'I left the things where you told me to.' [ro05de 025]

The verb *lɛ́f* 'leave, remain' also functions as a resultative copula in resultative causative constructions like the following one (cf. §9.4.4 for a thorough treatment):

(288) *Yu* 2sg *go* pot *mék* make *mék* sbjv *di* def *gál* girl *lɛ́f* remain *wet* with *brok-hát.* break.cpd-heart 'You're going to make that girl become broken-hearted.' [ge07fn 103]

Besides the verb *gɛ́t* 'get, have' (cf. §7.6.3), a few other verbs express existence in space and time. When the inchoative-stative verb *lɛ́f* 'leave, remain' occurs in an intransitive clause featuring a comitative or locative adverbial, this verb assumes a copula function with the meaning 'remain (behind), stay temporarily with' (289).

(289) *Machyta* name *lɛ́f* remain *wet* with *in* 3sg.poss *fámbul.* family 'Machyta has remained (temporarily) with his family' [ge07ae 213]

#### 7 The clause

The verb *kɔmɔ́t* 'come out' is employed to indicate a change of state in lexicalised collocations involving associative objects (cf. also §9.3.2). Compare (290):

(290) *A* 1sg.sbj *de* ipfv *trén=an* train=3sg.obj *porque* because *e* 3sg.sbj *go* pot *kɔmɔ́t* come.out *pɔ́sin.* person 'I'm bringing him up because [so that] he will turn out to be a (responsible) person.' [au07se 145]

The dynamic verb *sté* means 'last (a long time)' as in (291). This verb also expresses excessive duration in an adverbial SVC (cf. §11.2.5):

(291) *Bɛ́ta* very.good *tín* thing *nó* neg *de* ipfv *sté.* last 'Good things don't last.' [ra07fn 076]

Finally, the occurrence of *pás* 'pass' and *rích* 'arrive' as inchoative-stative verbs in comparatives like (184) and equatives like (211) may also be seen as manifestations of a copula-like use of these otherwise dynamic verbs.

### **7.6.3 Existentials**

The locative-existential copula *dé* 'be.loc', as well as the verb *gɛ́t* 'have, get, acquire' both participate in existentials, i.e. constructions which predicate the general existence of an entity. Pichi existentials appear in two types of clauses with respect to number and type of participants: Transitive clauses featuring *gɛ́t* 'have' and intransitive clauses featuring *dé* 'be.loc'. Some of the characteristics of these two types of existentials are given in Table 7.9.


Table 7.9 Existential clauses

The *gɛ́t*-existential construction occurs in a transitive clause. The subject position is filled by an expletive 3sg pronoun, the object position by the existing entity (292). This construction exclusively serves the expression of existential meaning and has no locative connotation. None of the other constructions that follow are uniquely employed to express existential meaning in this way:

7.6 being andhaving

(292) *Dís* this *smɔ́l* small *bɔ́tul* bottle *dɛn* pl *Fanta,* name *wé* sub *e* 3sg.sbj *gɛ́t* get *Coca-Cola,* name *e* 3sg.sbj *gɛ́t* get *Fanta,* name *e* 3sg.sbj *gɛ́t* get *limón,* lemon *e* 3sg.sbj *báy=an* buy=3sg.obj *wán.* one

'These small bottles of Fanta, where there is Coca-Cola, there is Fanta, there is Lemon, she bought him one (of them).' [ab03ab 130]

Pichi has other ways of establishing the type of impersonal reference characteristic for *gɛ́t*-existentials besides a 3sg expletive pronoun. The verb *gɛ́t* may also occur with an impersonal 3pl (293) or 2sg (294) pronoun in clauses that are functionally similar to existentials (292):


'But you already know that in Africa you don't have a relationship with a man like in Europe.' [fr03ft 167]

There are no restrictions on the use of *gɛ́t*-existentials in subordinate clauses. In sentence (295), the existential clause appears in a relative clause introduced by *wé* 'sub':

(295) *Bɔt* but *na* foc *dán* that *fɔ́s* first *tɛ́n* time *hós* house *dɛn* pl *wé* sub *e* 3sg.sbj *gɛ́t* get *dá* that *piso* storey *dɛn* pl *fɔ* prep *dán* that *altura* height *dɛn.* pl 'But its those houses of the past where there are those high storeys' [hi03cb 043]

Copula clauses featuring *dé* 'be.loc' typically acquire an existential reading when they lack a copula complement. In these clauses, we find the predicated entity, which may be of varying complexity, in the subject position (296). Since there is no complement to provide further specification, the clause acquires the default locative and manner reading that typifies such *dé*-clauses (297):

(296) *Bueno* good *aunque* although *dé,* there *bɔkú* much *interés* interest *económico* economic *dé.* be.loc 'Alright although there, there's a lot of economic interest.' [fr03ft 110]

#### 7 The clause

(297) *Bueno,* good *mi* 1sg.poss *gran-pá* grand-pa *bin* pst *dé.* be.loc 'Alright, my grandfather was around/fine.' [fr03ft 166]

Hence, constructions featuring *dé* acquire a locative reading when a locative expression is present. In (298), we find the locative adverbial *na sala* 'in the hall':

(298) *Paciente* patient *dɛn* pl *dé* be.loc *na* loc *sala,* hall *yú* 2sg.indp *dɔ́kta* doctor *la* the *una* one *yu* 2sg *de* ipfv *kán?* come 'Patients are in the hall, (and) you doctor, it's (only) at one o'clock that you come?' [ab03ab 118]

Existential clauses featuring *gɛ́t* are not often negated. The data contains only a single negative *gɛ́t*-existential clause, presented in (299). This is probably so because the "true" existential construction featuring *gɛ́t* is subject to an affirmative presupposition:

(299) *Dɛn* 3pl *de* ipfv *kɔ́l* call *dɛn* 3pl *sé,* qot *e* 3sg.sbj *nó* neg *gɛ́t* get *tɔ́k* talk *na* loc *Píchi.* Pichi 'They're called, there is no word (for that) in Pichi.' [dj05be 014]

In contrast, there are many examples of negated *dé*-copula clauses with an existential reading, as in the following two examples. Note the occurrence of negative concord in the first of the two following examples:


Likewise, the corpus does not reveal any instance of a non-finite *gɛ́t* with an existential sense. Conversely, we once more encounter many examples of non-finite *dé* 'be.loc' with an existential reading as in (302):

(302) *Ebanistas* Carpenter.pl *dɛn* pl *gɛ́fɔ* have.to *dé.* be.loc 'Carpenters have to be there/around.' [hi03cb 042]

The same applies to TMA marking. While quite a few *dé*-existentials are found with overt TMA marking as in (303), there is no such example of a *gɛ́t*-existential. The latter type of existential therefore appears to be prototypical in an additional sense – *gɛ́t* existentials typically predicate a generic situation, which is also marked as such by factative tense-aspect:

7.6 being andhaving

(303) *Ɛhɛ́* intj *wán* one *accidente* accident *fɔ* prep *motó* car *bin* pst *dé.* be.loc 'Oh yes, there was a car accident.' [ye03cd 073]

Finally, it is useful to draw attention to the linkages between existential and factive clauses. Factive clauses featuring the copula *dé* are existential clauses with a referentially empty subject position and a complement clause introduced by *sé* 'qot'. The subject is either an expletive *e* '3sg' or a dummy noun like *tín* 'thing', as in this example (cf. eg. 101 for further details on factive clauses):

(304) *Di* def *tín* thing *dé* be.loc *sé* qot *mék* sbjv *e* 3sg.sbj *mék* make *rabia* anger *wet* with *mí.* 1sg.indp 'The thing is that let her be angry with me.' [ye05rr 001]

### **7.6.4 Possessives**

Pichi employs a verbal and a copula strategy in the formation of possessive clauses. The verbs *gɛ́t* 'get, have' and *hól* 'hold, keep' are the principal verbs of possession and express time-stable and non-time-stable possession, respectively. Three collocations involving copulas are also used, albeit less frequently, in order to express possessive relations: *dé fɔ* 'be.loc prep' = 'have' and *na fɔ* 'foc prep' = 'have', as well *dé na/fɔ hán* 'be.loc loc/prep hand' = 'have on'. The use of these collocations may also be differentiated along the criterion of time-stability: *dé fɔ* and *na fɔ* express time-stable, and *dé na hán* transient, non-time-stable possession. Table 7.10 presents some characteristics of possessive clauses.

Table 7.10 Possessive clauses


The verb *gɛ́t* 'get, have' expresses permanent, time-stable possession. When *gɛ́t* occurs in a factative marked clause (305), a lexicalised light verb construction (306), an existential construction (cf. §7.6.3), or other contexts that propose a generic reading, the verb leans towards the stative meaning 'own, be in permanent possession':

(305) *A gɛ́t mɔdɛlɔ́.*

> 1sg.sbj get mother-in-law

'I have a mother-in-law.' [ro05de 009]

#### 7 The clause

(306) *Dí* this *mán* man *gɛ́t* get *líba* liver *ɛ́n,* intj *fɔ* prep *kɔmɔ́t* go.out *wet* with *dís* this *káyn* kind *bíg* big *gɛ́l.* girl 'Than man has guts, right, to go out with such an influential girl.' [dj05ce 291]

Conversely, when *gɛ́t* co-occurs with a TMA marker with a default or explicit perfective reading (307) or sentential aspect suggestive of telicity (i.e. the time clause in 308), an inchoative interpretation of *gɛ́t* as 'acquire, enter into permanent possession' is favoured:


Sometimes we also find the phrases *dé fɔ*'be.loc prep' or *na fɔ*'be.loc prep' expressing time-stable possession. There is no difference in meaning between the two constructions, although *na* 'foc' is employed as a time-stable identity copula in other contexts (310):


The verb *hól* 'hold, keep' expresses non-time-stable, temporary possession in a transitive clause like (311). In such contexts, it is best translated as 'keep'. The temporary nature of possession expressed by *hól* is reaffirmed by the adverbial phrase *durante un mes entero* 'for one whole month', which specifies the period of possession:

(311) *A* 1sg.sbj *fít* can *hól* keep *dán* that *mɔní* money *durante* during *un* def *mes* month *entero.* whole 'I'm able to keep that money for a whole month.' [ro05rt 049]

Speaker (dj) summarises the difference between *gɛ́t* and *hól* in (312). Note the difference in aspect marking with *hól*, *gɛ́t*, and *dráyb* 'drive'. Imperfective aspect is expressed through factative marking with the inchoative-stative verbs *hól* and *gɛ́t*. Meanwhile, it is the presence of *de* 'ipfv' that signals imperfective aspect with the dynamic verb *dráyb*:

7.6 being andhaving

(312) *"Yu* 2sg *hól* hold *wán* one *motó",* car *yu* 2sg *de* ipfv *dráyb=an,* drive=3sg.obj *pero* but *sé* qot *yu* 2sg *gɛ́t,* get *cuando* when *tienes,* you.get *"a gɛ́t wán motó".*

1sg.sbj get one car

' "Yú hól wán motó" (means) you're driving it, but when you possess it, when you have it "a gɛ́t wán motó."' [dj05ae 223]

The notion of temporary possession expressed by *hól* 'hold, keep' may also be applied to a human-possessed NP. A characteristic of West African pedagogy is to confer responsibility for the upbringing of a child to members of the extended family other than the biological parents. Such temporary guardianship is also expressed by *hól*. I leave it to speaker (au) to explain the meaning of *hól* in sentences (313) and (314):


When *hól* 'hold, keep' is employed as a dynamic verb in a transitive clause, it has the literal meaning of 'hold', hence the presence of the imperfective marker *de* in the following example:

(315) *Nó,* neg *na* foc *di* def *húman* woman *de* ipfv *hól* hold *di* def *plét.* plate

'No, it's the woman that's holding the plate.' [ra07se 012]

A second strategy for establishing a non-time-stable possessive relation makes use of the phrasal expression *dé na X hán*/*dé fɔ X hán* 'be in X's hand', where X is the possessor. This phrase is another variant of the copula strategy of possessive clause formation. In such invariably intransitive clauses, the subject instantiates the possessed NP and a prepositional phrase the possessor. In the following example, the transient nature of possession is underscored by the time adverb *náw* 'now':

(316) *George,* name *mi* 1sg.poss *móvil* mobile *nó* neg *dé* be.loc *na* loc *mi* 1sg.poss *hán,* hand *a* 1sg.sbj *nó* neg *gɛ́t* get *móvil* mobile *náw.* now

'George, I don't have my mobile phone on me, I don't have a mobile phone now.' [dj05ae 088]

#### 7 The clause

All possessive clauses covered in this section can be negated by standard verb negation. The negator *nó* 'neg' is inserted between the personal pronoun and the verb:

(317) *Yu* 2sg *sabí* know *sé* qot *yu* 2sg *nó* neg *gɛ́t* get *pikín?* child 'Do you know whether you don't have a child?' [fr03wt 173]

### **7.6.5 Predicate adjectives**

We are concerned here with a few property items that may be employed as predicate adjectives next to their use as inchoative-stative verbs. The fluidity between adjective and verb with these items shows that, notwithstanding its existence, the verb-adjective distinction is weak in Pichi. Adjectives can be identified by their distribution. Only adjectives may appear as complements to the locative-existential copula *dé* in predicate adjective clauses, such as the following one:

(318) *Tidé* today *di* def *húman* woman *dé* be.loc *fáyn.* fine 'Today the woman is fine.' [dj05ae 153]

In (318), *fáyn* 'be fine' is used as an adjective and denotes a physical property, namely a body state in an intransitive clause. The predicate adjective construction featuring the copula *dé* translates as 'be fine, well, healthy'. Contrast this meaning with (319), where *fáyn* is employed as an inchoative-stative verb with the meaning 'be intrinsically fine' hence 'beautiful'. In the latter example, *fáyn* therefore denotes a value:

(319) *Di* def *húman* woman *fáyn.* be.fine 'The woman is beautiful.' [dj05ae 149]

In the corpus, a handful of property items show the potential to function as predicate adjectives. As a general rule, the perceived time stability of the property determines whether it is used as a time-stable inchoative-stative verb or a non-time-stable adjective. The most consistent time-stability distinction is found with the words *bád* 'be bad, ill', *fáyn* 'be fine, beautiful', and *gúd* 'be good, well'. When they occur as adjectives, they denote a body state. When they occur as inchoative-stative verbs, these property items denote a value, an intrinsic property.

Only these three words are unequivocal members of the small adjective class in Pichi. Beyond that, a few more property items are rarely used as predicate adjectives. Table 7.11 lists all property items attested in predicate adjective constructions in the corpus.

The words in the second and third columns of Table 7.11 appear as predicate adjectives in the corpus only rarely. For example, the property item *bráyt* 'be bright, glowing with beauty' is attested as an adjective where it denotes a visible body state as in (320) – the speaker is an elderly lady giving an account of her youth. Compare *frɛ́s* 'be fresh' in the

7.6 being andhaving

*bád* 'ill' *bráyt* 'bright' *frí* 'free' *fáyn* 'fine' *wɔwɔ́* 'messed up' *sló* 'slow' *gúd* 'well' *pyɔ́* 'pure' *spɛ́shal* 'special'

Table 7.11 Predicate adjectives

same sentence, which is used as an inchoative-stative verb to denote a more lasting body state of freshness or youthfulness:

(320) *Moka* place *bɔ́y* boy *dɛn* pl *krés* go.mad *wé* sub *dɛn* 3pl *sí* see *lɛk* like *háw* how *a* 1sg.sbj *frɛ́s,* be.fresh *na* foc *so* like.that *a* 1sg.sbj *dé* be.loc *bráyt.* bright

'The Moka boys went crazy when they saw how fresh I was, that's how bright I looked.' [ab03ay 059]

(321) *Yu* 2sg *skín* body *bráyt* be.bright *ó.* sp

'Your body is really glowing (with beauty).' [dj07ae 165]

The physical property item *wɔwɔ́*'be ugly, messed up' is used by the same speaker as an adjective in (322) and as an inchoative-stative verb in (323). The first example featuring *wɔwɔ́* again expresses a visible state of the street, while the second is more time-stable in its meaning:


Predicate adjective clauses may be marked for TMA like any other copula clause featuring the copula *dé*. Compare the adjective *bád* 'ill' in (324) with a future tense reference:

(324) *Wé* sub *yu* 2sg *go* pot *fɔdɔ́n* fall *yu* 2sg *go* pot *dé* be.loc *bád.* bad 'When you fall you'll be in a bad state.' [ab03ay 114]

Adjectives may also be employed attributively as prenominal modifiers. In this, adjectives behave no differently from other property items (cf. §5.2.1). Below, the adjective *fáyn* 'be fine' appears as a modifier of *gɛ́l* 'girl':

#### 7 The clause

	- '(If) you see a fine girl, you go chat her up.' [au07se 062]

The class of adjectives is closed for words of Pichi origin, since the use of property items as copula complements is lexically restricted. But the predicate adjective construction is a port of entry for Spanish adjectives (cf. §13.2.2).

Finally, I draw attention to the various other means of attributing properties to a noun. Speakers make use of postnominal modification through relative or quotative clauses. Other ways of expressing modification are associative constructions and compounding. Two strategies of modification serve as a productive means of deriving new property items next to the use of Spanish adjectives in the Pichi predicate adjective construction. A *dé*-copula clause with an adverbial complement featuring *wet* 'with' (326), as well as light verb constructions involving *gɛ́t* 'get, have' (327) allow the attribution of a property to a referent:

(326) *E* 3sg.sbj *hád* be.hard *fɔ* prep *mék* make *mék* sbjv *dɛn* 3pl *bíl* build *na* loc *yá* here *só* like.that *bikɔs* because *di* def *grɔ́n* ground *e* 3sg.sbj *tú* too *dé* be.loc *wet* with *stón.* stone It's hard for them to build here because the ground is too stony.' [dj05be 111]

(327) *E* 3sg.sbj *hád* be.hard *fɔ* prep *bíl* build *na* loc *yá* here *bikɔs* because *sé* qot *di* def *grɔ́n* ground *gɛ́t* get *bɔkú* much *sansán.* sand

'It's hard to build here because the ground is very sandy.' [ro05ee 063]

### **7.7 Adverbial modification**

Pichi adverbials modify verbs and clauses. It is useful to distinguish between adverbs proper and adverbials. I employ "adverbial" as a cover term, which includes adverbs, but also encompasses other clause constituents with the functions of adverbs. Adverbs constitute an underived, largely monomorphemic minor word class of their own, and unlike other constituents that may function as adverbials (e.g. common NPs), they do not normally appear in the syntactic positions of other word classes.

Adverbials may occupy a clause-initial, a preverbal, a postverbal and a clause-final position. Some adverbs consist of a single morpheme (e.g. *bambáy* 'gradually', *náw* 'now'), others are lexicalised phrases with idiosyncratic, underivable meanings (for instance. *sɔn.tɛ́n* 'some.time' = 'perhaps'). Other expressions are more or less conventionalised phrases, constituted by means of phrasal syntax (e.g. *bɔkú tɛ́n dɛn* 'many times, often'), but usually not encountered in non-adverbial functions. Often such noun phrase adverbials are fixed collocations involving generic nouns denoting time (*tɛ́n* 'time', *áwa* 'hour'),

7.7 Adverbial modification

manner (*stáyl* 'style', *fásin* 'manner'), and space (*sáy* 'side', *plés* 'place', *pát* 'part'). There is thus a smooth transition from more basic monomorphemic adverbs to more or less lexicalised adverbial phrases.

The expression of degree and manner modification is particularly rich and varied in Pichi and deserves special attention. It should, however, also be pointed out that many adverbial notions are expressed by wholly different means than adverbials. For example, movement verbs may take goal objects, while some spatial and temporal notions may be expressed by motion-direction and adverbial SVCs. Many ideophones function as manner adverbials next to the adverbs of manner covered in this section.

Equally, many clause linkers are not very different in function from the linking adverbs listed in Table 7.12 below (e.g. *bikɔs* 'because', *adɔnkɛ́* 'even if'). Further, modal clauses with expletive subjects (e.g. *e fít bí sé* '3sg.sbj can be qot' = 'it could be that', and *e fíba sé* '3sg.sbj seem qot' = 'it seems that') convey meanings similar to those of sentence adverbs like *sɔntɛ́n* 'perhaps' and *mébi* 'maybe'.

### **7.7.1 Adverbs**

Table 7.12 presents all monomorphemic adverbs found in the corpus and the most common conventionalised phrasal expressions with adverbial functions. The preferred or canonical syntactic positions are also indicated. The table also contains the two most common Spanish-derived adverbs *pero* 'but' and *bueno* 'alright'. Adverbs with multiple meanings are arranged in all the corresponding "adverb type" sections (e.g. *smɔ́ltɛn* 'shortly after' = locative adverb, *smɔ́ltɛn* 'nearly' = modal adverb.

Adverbs that appear at the beginning modify the sentence in its entirety – they have a wide scope. In (328), the linking adverb adverb *pero* 'but', the modal adverb *sɔntɛ́n* 'perhaps', and the time adverb *bambáy* 'gradually' all occur sentence-initially:

(328) *Pero* but *bambáy* gradually *bambáy* rep *sɔntɛ́n* perhaps *yu* 2sg *go* pot *sí* see *di* def *wán* one *wé* sub *go* pot *máred* marry *yú.* 2sg.indp 'But very gradually perhaps you might find the one who will marry you.' [ab03ab 204]

Locative and time adverbs may also occur after the verb, in which case they have narrow scope and modify the meaning of the verb alone. In (329), the repeated locative noun *dɔ́n* 'down' and the locative adverb *yandá* 'yonder' modify the verb *sidɔ́n* 'sit, stay':

(329) *Bɔt* but *ín* 3sg.indp *sidɔ́n* stay *dɔ́n* down *dɔ́n* rep *dɔ́n* rep *yandá.* yonder 'But he stays far down over there.' [ma03ni 026]

The data contains diverse time adverbs. A few of these are monomorphemic, e.g. *bambáy* 'gradually' in (328) above. Others are more or less idiosyncratic phrases containing the time-denoting generic noun *tɛ́n* 'time', as in *bɔkú tɛ́n* 'much time' = 'for a long time'

#### 7 The clause


Table 7.12 Adverbs

(330) or *dé* 'day', as in *wán dé* 'someday' (331). Location-in-time adverbs, like *wán dé* prefer the clause-initial, duration adverbs like *lɔ́n tɛ́n* 'long ago' and *bɔkú tɛ́n* the clause-final position:

(330) *Nó* neg *chɛ́k=an* think=3sg.obj *bɔkú* much *tɛ́n,* time *tɛ́l* tell *mí* 1sg.indp *sé* qot *nó.* neg 'Don't think about it for a long time, tell me "no".' [ye07me 034]

Adverbs with generic time reference like *wán dé* 'someday' and *ɔ́l tɛ́n* 'always' are equally often encountered in the initial as well as the final position:

(331) *Na* foc *ín* 3sg.indp *wán* one *dé* day *a* 1sg.sbj *bin* pst *tɛ́l* tell *wán* one *grand* big *frère* brother *na,* loc *na* loc *mi* 1sg.poss 7.7 Adverbial modification

*colegio* college *dé,* there (*…*) 'That's why one day, I told one of my seniors in, in my secondary school there, (…)' [ye07ga.003]

(332) *A* 1sg.sbj *mɔs* obl *gó* go *Alemania* place *wán* one *dé.* day 'I absolutely have to go to Germany someday.' [to07fn 197]

In clauses featuring double-object constructions, speakers may place a time adverbial between the recipient or beneficiary object and the patient object instead of placing it in the clause-initial or clause final position. This position appears to be focus-induced, since it was encountered more often during the elicitation of adverbials than in natural speech:

(333) *Ebongolo* name *tɛ́l* tell *mí* 1sg.indp *yɛ́stadé* yesterday *in* 3sg.poss *problema.* problem 'Ebongolo told me about his problem yesterday.' [dj07ae 347]

The phrase *e fínis* '3sg.sbj finish' = 'then' is a stand-alone clause, which may function as a linking "adverb" (334). A formal indication of its hybrid status between clause and adverb is that the personal pronoun *e* '3sg.sbj' is sometimes dropped:

(334) *A* 1sg.sbj *gó* go *wás* wash *wet* with *mi* 1sg.poss *hán* hand *mí* 1sg.indp *sénwe* emp *a* 1sg.sbj *dráy=an,* dry=3sg.obj *e* 3sg.sbj *fínis* finish *a* 1sg.sbj *áyɛn=an.* iron=3sg.obj 'I myself went to wash (it) with my own hands, I dried it, then ironed it.' [dj07re 050]

The two modal adverbs *sɔntɛ́n* 'perhaps' (cf. 328 above) and *smɔ́ltɛn* 'nearly' (336) and the evaluative adverb *bádtɛn* 'unfortunately' (335) are lexicalised phrases involving the generic noun *tɛ́n* 'time' as a formative element (cf. 6 above). Modal and evaluative adverbs are normally found in the initial position with scope over the entire clause:

(335) *Bádtɛn* unfortunately *náw,* now *di* def *fɔ́s* first *dɔ́kta* doctor *wé* sub *wi* 1pl *gɛ́t,* get *e* 3sg.sbj *nó* neg *dé* be.loc *ɔ́p* up *na* loc *ɔspítul.* hospital 'Unfortunately, the first doctor that we had wasn't up (there) in the hospital.' [ab03ay 078]

(336) *Smɔ́ltɛn* nearly *a* 1sg.sbj *bin* pst *fɔ* cond *dáy* die *dé.* there 'I nearly died there.' [ed07fn 493]

#### 7 The clause

The adverb *mébi* 'maybe' (337) is not as common as *sɔntɛ́n* 'perhaps'. Note that *smɔ́ltɛn* {small.time} has an entirely opaque sense 'nearly' in the example above, and a more transparent, temporal sense 'shortly (after)' in (338) below.


The L-toned clause-initial linking adverb *so* 'so' (339) differs from the H-toned deictic manner adverb *só* 'like this, like that' (340) in tone alone. The deictic manner adverb *so* 'like this' is often focused and fronted in a *na* cleft construction, in order to establish reference to preceding discourse material (341) (cf. also §7.4.3.3):


The H-toned adverb *só* 'like that' is also found in the conventionalised collocations *(na) yá só* 'right here' (342) and *náw só* 'right now' (343), where its deictic character provides emphasis:


Manner adverbs other than *só* 'like that' and ideophonic adverbs generally occur after the verb, since they directly modify the meaning of the verb. Compare *kwík* 'quickly' and the ideophone *kwáráng* in the two following sentences:

7.7 Adverbial modification


Pichi has a small set of four preverbal adverbs, which appear in the predicate before the verb. The set includes the time adverbs*jís*/*jɔ́s* 'just' and *stíl* 'still', as well as the degree adverbs *só* 'so much', *tú (mɔ́ch)* 'too much'. The use of the preverbal time adverbs *jís*/*jɔ́s* and *stíl* coincides with resumptive imperfective aspect marking – the adverbs are preceded and followed by *de* 'ipfv'. The aspect-marking functions of the time adverbs *jís*/*jɔ́s* 'just' and *stíl* 'still' are covered in §6.4.2 and §6.4.4, repectively (cf. also for a discussion of the position of preverbal adverbs):

(346) *Náw* now *dɛn* 3pl *de* ipfv *jís* just *de* ipfv *kán.* come 'Now, they're just coming.' [ye07je 179]

Preverbal degree adverbs usually occur with gradable property items or light verb constructions which attribute properties as in (347). Hence, sentences like (348), in which a non-gradable verb (i.e. *tɔ́k* 'talk'), and a dynamic one at that, is preceded by a preverbal degree adverb, are very rare:


Non-gradable verbs are more likely to be modified postverbally by the expression *tú mɔ́ch* 'too much' than by preverbal *tú* 'too (much)' (349). The phrase *tú mɔ́ch* includes the quantifying adverb *mɔ́ch.* When a verb is modified in this way for superlative degree, the use of *mɔ́ch* is mandatory. The same applies when *tú mɔ́ch* modifies a nominal (350):


#### 7 The clause

Nonetheless, *tú mɔ́ch* may also be used in preverbal position without any difference in meaning to *tú* 'too (much)'. The following sentence features both possibilities. While the compound property item *smɔl.skín* 'small.body' = 'be thin' is modified preverbally, the property item *dráy* 'be dry, haggard' is modified postverbally by *tú mɔ́ch*:

(351) *Di* def *pikín* child *tú* too *mɔ́ch* much *smɔlskín,* be.thin *e* 3sg.sbj *dráy* be.dry *tú* too *mɔ́ch.* much 'The girl is too thin, she's too lean.' [dj07ae 206]

Somewhat similar to the distribution of *tú (mɔ́ch)* is that of the adverb *só* 'like that, that much'. When *só* occurs in a preverbal position, it implicitly expresses equative degree and means 'that much' (352). However, when *só* appears in the clause-final position, it means 'like that' and therefore retains its central meaning as a manner adverb (cf. 340 above):

(352) *Dɛn* 3pl *nó* neg *de* ipfv *só* like.that *yús=an* use=3sg.obj *mɔ́.* more 'It's not used that much anymore.' [ye07je 009]

The word *mɔ́*'be more, again' also functions as a degree adverb and is characterised by an unusual amount of syntactic flexibility. In contexts other than comparison, *mɔ́* may occur clause-finally as a time adverb with the meaning 'again' (353) and 'still' (354–355):


In negative clauses, *mɔ́*is best translated as 'anymore, no longer, not again'. Compare the following examples with the negated dynamic verb *ánsa* 'answer' (356), and (357) with the negated stative verb and copula *dé* 'be.loc':


7.7 Adverbial modification

In (358) below *mɔ́*may be analysed as occupying the object position of *tɔ́k* 'talk, say' with the meaning 'more'. Alternatively, *mɔ́*may be seen to function as an adverbial and be translated as 'still, again continue to':

(358) *A* 1sg.sbj *nó* neg *de* ipfv *tɔ́k* talk *mɔ́.* more 'I was not talking any longer.' Or 'I was not saying (anything) more/again.' [ab03ay 090]

The scope of *mɔ́*may also be narrower than the clause. In (359), *mɔ́*is in the postnominal position and modifies the preceding NP in a way no different from that of the focus particle *sɛ́f* 'foc' or the quantifier *ɔ́l* 'all'. In (360), *mɔ́*modifies the adverbial *áfta* 'then':


Besides the adverbs treated so far, compounds (361) or constructions featuring generic nouns of place (i.e. *sáy* 'side, place'), time (i.e. *tɛ́n* 'time' and *dé* 'day'), and manner (i.e. *stáyl* 'manner, style') serve as locative, time (362), and manner adverbials (363):


Other than that, Pichi employs noun phrases introduced by prepositions (e.g. *na* 'loc', *fɔ* 'prep', *to* 'to') or locative nouns (e.g. *bifó* 'before', *bɔtɔ́n* 'under', *kɔ́na* 'next to', *míndul* 'middle') to form various types of adverbial phrases which provide modification to clauses:

(364) *A* 1sg.sbj *pút* put *di* def *kí* key *na* loc *pála.* parlour 'I put the key in the parlour.' [to07fn 114]

#### 7 The clause

### **7.7.2 Modification of manner and circumstance**

The corpus contains only few underived manner adverbs (amongst them *kwík* 'quickly, early' in 344 above). Nevertheless, the possibilities for providing manner modification are particularly rich. They encompass the use of adverbials, ideophones, SVCs, secondary predication, compounds, associative constructions, lexicalised iteration, and adverbial clauses of manner.

The value property item *fáyn* 'be fine, nice, correct' is frequently found in clause-final position to provide manner modification. The use of *fáyn* in this way is conventionalised to such an extent that it may be considered an adverb with its own established meaning of 'nicely, properly, in the right way' (a similar case is made for *bád* 'extremely', cf. 385– 386 further below):

(365) *E* 3sg.sbj *fíks* fix *dɛ́n* 3pl.indp *fáyn.* fine 'She has arranged them properly.' [li07pe 069]

Another idiosyncratic way of expressing manner modification is through the lexicalised reduplication *haydháyd* 'secretly' (366):

(366) *Chico,* boy *yu* 2sg *dɔ́n* prf *chɔ́p=an* eat=3sg.obj *haydháyd.* secretly 'Man, you've eaten it secretly.' [ge07fn 333]

Further, Pichi employs the adverb-deriving suffix *-wán* 'adv' to form manner adverbs (367), and the generic noun *stáyl* 'style' (368) in order to form manner-denoting adverbial NPs in clause-final position:


Likewise, prepositional phrases introduced by *fɔ* 'prep' may express manner as in the following example:

(369) *A* 1sg.sbj *wáka* walk *fɔ* prep *fút* foot *wet* with *mi* 1sg.poss *maleta.* suitcase 'I walked by foot with my suitcase.' [ab03ay 075]

For one part, biclausal structures are common in providing modifications of circumstance. Compare the following adverbial clauses introduced by *wé* 'sub' (370) and *sé* 'qot' (371):

7.7 Adverbial modification


Another common way of providing modification to a clause is by means of depictive secondary predication (cf. also §11.3). In the depictive adjunct in (372), the secondary predicate *nékɛd* 'be naked' provides information about the state of the subject *e* '3sg.sbj', while the situation denoted by *kɔmɔ́t* 'come out' unfolds:

(372) *E* 3sg.sbj *kɔmɔ́t* come.out *na* loc *rúm* room *nékɛd.* be.naked 'He left the room naked.' [ra07ve 001]

Modifications of circumstance may also be provided through nominal depictives that come in the guise of prepositional phrases introduced by *wet* 'with' (373) and *lɛk* 'like' (374):


The preposition and clause linker *lɛk* may also introduce a prepositional phrase that indicates sameness of manner. Two examples of such "similatives" (Haspelmath & Buchholz 1998) follow:

(375) *Mí* 1sg.indp *nó* neg *lɛ́k* like *yú* 2sg.indp *bɔt* but *wi* 1pl *fít* can *dé* be.loc *lɛk* like *kɔ́mpin.* friend 'I don't love you but we can be (like) friends.' [ru03wt 029]

The similative collocation *wók lɛk dɔ́kta* functions as a nominal depictive (376) (cf. §11.3 for an extensive discussion of verbal depictives in secondary predications). A similative *lɛk* in (377) translates as 'around':

(376) *Di* def *cubana* Cuban *húman* woman *de* ipfv *wók* work *lɛk* like *dɔ́kta* doctor *na* loc *Malábo.* place

'The Cuban woman works as a doctor in Malabo.' [ro05ee 071]

#### 7 The clause

(377) *Yu* 2sg *fít* can *gí* give *mí* 1sg.indp *lɛk* like *dos* two *mil* thousand *só?* like.that 'Can you give me around two thousand?' [be07fn 311]

Similative clauses are introduced by *lɛk sé* 'like qot' = 'as if', as in the example below:

(378) *E* 3sg.sbj *de* ipfv *dú* do *lɛk* like *sé* qot *e* 3sg.sbj *de* ipfv *fɛ́n* look.for *sɔn* some *tín.* thing 'He's pretending to look for something.' [dj07ae 517]

Some relations of modification that habitually re-occur tend to be expressed through verb-noun compounds. For example, the depictive secondary predication in (379) is more often rendered by (380):


### **7.7.3 Modification of degree**

There are various ways of providing degree modification in Pichi other than by the means covered in §7.7.1. Not all of these involve the use of adverbial constituents. For example, inherently comparative and superlative expressions, cognate objects, some types of focus constructions (i.e. predicate cleft), as well as repetition all provide some form of explicit or implicit modification of degree.

Degree modification may also be realised on the suprasegmental level. Vowel-lengthening and extra-high pitch may indicate a larger amount of intensity, extent, or dimension of a referent, which is generally a property item or an adverbial. The only syllable of the property item *kól* 'be cold' in (381) is pronounced with an extra-high tone and lengthened. The phonetic transcription is provided in squared brackets:

(381) *Pero* but *wé* sub *a* 1sg.sbj *kin* hab *tɔ́ch* touch *in* 3sg.poss *fút,* leg *in* 3sg.poss *hán* hand *dé,* there *na* foc *só* like.that *dɛn* 3pl *kó.ól* [kő::l]*.* cold.emp 'But when I would touch his foot (and) his hand, they were so extremely cold.' [ab03ab 066]

Vowel lengthening and extra-high tone are conventionalised with the preposition *sóté* 'until'. Both phenomena always occur when *sóté* is employed as a clause-final temporal adverb with the meaning 'for a long time' or a degree adverb with the meaning 'extremely' (382):

7.7 Adverbial modification

(382) *Dɛn* 3pl *kéch=an* catch=3sg.obj *dɛn* 3pl *bít=an* beat=3sg.obj *sóté.e* [sőte̋::]*.* until.emp 'They caught him and beat the hell out of him.' [pa07fn 556]

Suprasegmental degree modification is performed in accordance with the syllable structure of the modified word. Monosyllabic words bear an extra-high tone over their H-toned syllable. If the syllable ends in a vowel, liquid, or nasal, it may also be lengthened. Two examples for this pattern are *kól* 'be cold' in (381) above and *fá* 'be far' in (383) below.

The H-toned syllable of a bisyllabic word may also be lengthened if it ends in a vowel or liquid. Compare *fawe* 'be far' in (383) below. Both *fá* and *fáwe* in (383) are additionally emphasised by means of an extra-high tone:

(383) *Wántɛn* suddenly *a* 1sg.sbj *skía,* be.scared *e* 3sg.sbj *sé* qot *"nó* neg *skía,* be.scared *a* 1sg.sbj *kɔmɔ́t* come.out *fá.áwe* [fá:we]*,* far.emp *a* 1sg.sbj *kɔmɔ́t* come.out *fá.á* [fá::]*."* far.emp

'Suddenly, I became scared, he said "don't be scared, I come from very far away, I come from very far".' [ed03sb 176]

In contrast, vowel-lengthening for degree modification is not attested with mono- or bisyllabic words with word-final H-toned syllables that end in plosives or fricatives. With this group of words, we only find emphatic extra-high tone or other types of degree modification. For example, in (384), the property item *bíg* 'be big' is modified for degree by repetition and the H-tone over both iterations is raised:

(384) *Dɛn* 3pl *gɛ́t* get *wán* one *bíg* big *bíg* rep [bıg bı ̋ g]̋ *fám.* farm 'They have a huge farm.' [fr03ft 012]

Property items that do not denote dimension or a physical property and adverbs that do not denote a manner or degree are not usually modified suprasegmentally in this way. One way of providing degree modification to other types of words is by means of the value property item *bád* in clause-final position. For example, in (385) the property item *bád* 'bad' is employed as a degree adverb with the meaning 'extremely'.

(385) *A* 1sg.sbj *de* ipfv *sɔ́ri* feel.sorry *bád.* extremely 'I really feel sorry.' [hi03cb 069]

In (386), *bád* modifies *fáyn* '(be) fine'. The example shows that *bád* retains nothing of its lexical meaning of 'be bad' when employed in this function. It is a true degree adverbial and may also modify a verb which is the antonym of its lexical source:

7 The clause

(386) *E* 3sg.sbj *fáyn* be.fine *bád,* extremely *e* 3sg.sbj *fáyn* be.fine *bád.* extremely 'She is really beautiful, she is really beautiful.' [fr03ft 113]

The sentence-final particle *ó* may also provide degree modification to a sentence (387). The various functions of this particle are covered in detail in §12.2.4:

(387) *E* 3sg.sbj *hád* be.hard *ó.* sp 'It's really difficult.' [ro05fe 037]

## **8 Spatial and temporal relations**

Location in space is expressed by elements from diverse word classes and through a large variety of constructions. Some of the means employed for the expression of spatial relations are carried over into the expression of temporal relations but there are also independent ways of expressing location in time.

### **8.1 Spatial relations**

Prepositions, locative nouns, and locative verbs play a part in expressing spatial relations. Other items involved are motion verbs – verbs whose meanings include a motion component. The relation between "figure" and "ground" may be mediated through various types of structures. The expression of source and goal is of particular interest in the disucssion because it may involve the use of various competing structures.

### **8.1.1 Locative prepositions**

Prepositions are employed to express the location and direction of motion of an entity (the "figure") in relation to a place (the "ground"). Locative prepositions and locative nouns (cf. §8.1.2) belong to separate word classes, but some shared characteristics make the distinction less clear-cut. Table 8.1 contains the Pichi inventory of prepositions. There are no postpositions in Pichi. Non-locative roles expressed by prepositions are covered in §9.1.3. Note that Pichi also has the two temporal prepositions *ápás* 'after' and *síns* 'since' (cf. §8.2).


Table 8.1 Locative uses of prepositions

#### 8 Spatial and temporal relations

Locative prepositions introduce adverbial prepositional phrases. Prepositions differ from locative nouns because they cannot be employed in the syntactic position of nouns. Prepositions require explicit mention of the ground, which is usually a nominal complement (1–2). The prepositions *fɔ* 'prep', *pan* 'on', and *wet* 'with' may however be stranded in questions, cf. (82–83), as well as in relative clauses.


Next to full nouns, locative adverbs may also function as complements to prepositions. Take note of the temporal meaning of the locative adverb *dé* 'there' in (3):

(3) *Wé* sub *in* 3sg.poss *mamá* mother *dáy,* die *na* foc *frɔn* from *dé* there *e* 3sg.sbj *bigín* begin *krés.* be.crazy 'When his mother died, that's when he began to go insane.' [dj07ae 103]

(4) *E* 3sg.sbj *kán* pfv *fɔdɔ́n* fall *sóté* until *yá.* here '(And then) it fell up to here.' [li07pe 090]

The general locative preposition *na* 'loc' and the general associative preposition *fɔ* 'prep' take the locative adverb *yá (só)* '(right) here' as a complement but are not attested with *dé* 'there' or *yandá* 'yonder' as a complement:


It is also common to find the generic noun *sáy* 'side, place' and a demonstrative as a complement to *na* or *fɔ* instead of a deictic locative adverb:

(7) *Na* foc *só* like.that *dɛn* 3pl *de* ipfv *mék* make *café* coffee *na* loc *dí* this *sáy.* side 'That's how they make coffee here.' [ye07ga 038]

8.1 Spatial relations

Personal pronouns do not normally occur as complements to the general (locative) prepositions *na* 'loc' and *fɔ* 'prep'. Pichi employs other means of expressing the relevant notions. For example, the ground may be named more specifically as in (8) or an idiomatic expression may be used, as in (92) further below:

(8) *E* 3sg.sbj *bin* pst *pás* pass *na* loc *mi* 1sg.poss *hós.* house 'She passed by my house [to see me].' [ro05ee 078]

The preposition *to* 'to' is rare. It is employed with a locative function to mark a goal (9). The following sentences represent two of altogether four occurrences of this preposition in the corpus. I point out that in (10), the preposition *to* is used to mark the goal in a motion-direction SVC in the same position as *na* 'loc' or *fɔ* 'prep' (cf. e.g. 58). The use of *to* as a complementiser is even more marginal (cf. 242 for an example involving the main verb of cognition *nó* 'know (how to)) and is not common with the vast majority of speakers:


'When they took her to the doctor, (…)' [ab03ay 121]

The preposition *na* 'loc' expresses location in the most general way. Depending on context, *na* may denote superior (1), interior, proximate, or lateral (7) location. The associative preposition *fɔ* 'prep' is employed as a general locative preposition in ways similar to *na* (cf. e.g. 36, 56, 57 and 84). But compared to *na* 'loc', the preposition '*fɔ* 'prep' is only employed in a minority of instances for the expression of general location.

### **8.1.2 Locative nouns**

Table 8.2 presents the repertoire of locative nouns. The distribution of these elements (cf. Table 8.3 further below) reflects their heterogeneity and intermediary status between noun and preposition. Circumferential location is expressed via the locative verb *ráwn* 'surround' (cf. 38) and distal location by means of the multifunctional word *fá(wé)* '(be) far' (cf. e.g. 383). In Pichi, body part nouns such as *bák* 'back' or *fés* 'face' are not usually employed to express location roles.

Locative nouns have characteristics in common with ordinary nouns. They may occur in the position of NPs, for example as subjects (11) or as goal objects of movement verbs, like *rích* 'arrive' (12). In both cases, an explicit mention of the ground is not required:

(11) (*…*) *mék* sbjv *yu* 2sg *tɔ́n=an,* turn=3sg.obj *porque* because *bɔtɔ́n* bottom *go* pot *rós.* burn

'(…) turn it, because the bottom might burn.' [dj03do 055]

#### 8 Spatial and temporal relations


Table 8.2 Locative nouns

(12) *Yu* 2sg *de klém fɔ rích pantáp.*

ipfv climb prep arrive top

'You're climbing in order to reach the top.' [au07se 086]

In the same vein, a locative noun can appear as the adverbial complement of the locative-existential copula *dé* 'be.loc' (13):

(13) (*…*) *e* 3sg.sbj *dé* be.loc *ɔ́p,* up *gó* go *só!* like.that '(…) it's [farther] up, go this way!' [ma03ni 011]

All locative nouns except *nía* 'near', *kɔ́na* 'next to', and *nadó* 'outside' may also be preceded by the definite article *di* 'def' as in the following example:

(14) *Di* def *dɔ́n* down *na* foc *violeta* violet 'The lower part is violet.' [ma03hm 034]

In addition, all locative nouns except *nía* 'near', *kɔ́na* 'next to', and *nadó* 'outside' may also be preceded by the general locative preposition *na* 'loc' like any ordinary noun. In the data, such constructions are, however, very rare, and none of these locative nouns is preceded by the general associative preposition *fɔ* 'prep' instead of *na* 'loc':

(15) *E* 3sg.sbj *púl=an* remove=3sg.obj *na* loc *pantáp* top *di* def *béd.* bed 'She took him from the bed.' [ab03ab 079]

8.1 Spatial relations

(16) *Na* foc *fɔ* prep *mék* make *nó* neg *gó* go *na* loc *dɔ́n.* down

'It's in order (for us) not to go down.' [ma03hm 003]

The locative nouns *nía* 'near', *kɔ́na* 'next to', *nadó* 'outside', and *bifó* 'before, front' are not normally found as complements to *na* 'loc' in prepositional phrases like the ones above. The peculiar distribution of *nía* and *kɔ́na* may be due to their multifunctionality. *Nía* also functions as a locative verb 'be near' (cf. 40), *kɔ́na* as a common noun 'corner', and *bifó* as a time clause linker 'before' (cf. §10.7.3). In (34) below, *kɔ́na* is employed as a locative noun, in the following example (17), as a common noun:

(17) *E* 3sg.sbj *de* ipfv *sɛ́l* sell *e* 3sg.sbj *de* ipfv *pút* put *smɔ́l* small *smɔ́l* red *wán* one *fɔ* prep *kɔ́na* corner *mék* sbjv *e* 3sg.sbj *fít* can *bák* return *dán* that *mán* man *in* 3sg.poss *mɔní.* money

'(…) she's selling (and) she's putting a bit at the side in order to be able to give that man back his money.' [hi03cb 220]

In turn, *nadó* is a lexicalised collocation, in which the locative preposition *na* already serves as the first component. The second component is the rare noun *do* 'door' (the more current word for 'door' is *do.mɔ́t* 'door.mouth'). Although it is lexicalised, the prepositional phrase which constitutes this collocation therefore has a residual meaning of its own. I assume that this results in the ungrammaticality of a sequence like *\*na nadó* 'loc outside'.

When the locative nouns *bifó* 'before', *bihɛ́n* 'behind', *ɔ́p* 'upperside', and *dɔ́n* 'downside' appear in a nominal position, speakers tend to employ an associative construction featuring the generic place noun *sáy* 'side, place' (19) and sometimes *pát* 'part, place' (20) as a modified noun and the locative noun as a modifier noun. This construction, which serves to derive a nominal structure, is favoured with these nouns when a ground is not mentioned. Compare (18) with an explicit ground (i.e. *di hós* 'the house') and the two sentences thereafter without mention of a ground:


#### 8 Spatial and temporal relations

However, when the ground is explicitly mentioned, most locative nouns participate in a construction that is structurally equivalent to a prepositional phrase featuring a preposition and an object complement. Compare (2) above with (21–23) below:


'He's walking back and forth in front of the door, he's pretending to be looking for something.' [ne07fn 170]

The same holds for the locative nouns *nía* 'near' and *kɔ́na* 'next to', which behave differently from other locative nouns in other contexts:


The ground need not be marked for definiteness as it is in the two examples above. Three sentences follow without overt definiteness marking. In this respect, the same principles of definiteness marking apply as they do for other objects. Note that the locative nouns *ɔntɔ́p* 'top, on' (26) and *pantáp* 'top, on' (15) above are absolute synonyms and equally frequent.


8.1 Spatial relations

(28) *Dán* that *skúl* school *e* 3sg.sbj *dé* be.loc *nía* near *bɛrin-grɔ́n,* burial.cpd-ground *nɔ́?* intj 'That school is near the cemetery, right?' [ma03hm 018]

The locative noun *nadó* 'outside' behaves differently in this respect. The ground may only be expressed in a possessive construction, namely a *fɔ*-prepositional phrase:

(29) *Pɔ́sin* person *dɛn* pl *dé* be.loc *nadó* outside *fɔ* prep *di* def *avión.* plane 'People are outside the plane.' [dj05be 165]

The expression of the ground by way of a *fɔ*-prepositional phrase as in (29) above is not accepted with other locative nouns, i.e. \**míndul fɔ di rúm* {middle prep def room} 'in the middle of the room', \**bihɛ́n fɔ dís hós* {behind prep this house} 'behind this house'. This also holds for the locative associative constructions described further below in (31). Compare the ungrammatical example (30), which involves such a structure:

(30) \* *E* 3sg.sbj *dé* be.loc *bifó* before *sáy* side *fɔ* prep *di* def *hós.* house Intended: 'She's in front of the house.' [ye07de 024]

Furthermore, *dɔ́n* 'down' does not normally occur in clauses with an explicit ground at all. An explicit ground may, however, be included in the clause by making use of another possessive structure, namely an associative construction. *Dɔ́n* enters into a recursive collocation with the generic noun *sáy* 'side, place', which in turn functions as the modifier to the ground in yet another associative construction. Compare the following example:

(31) *Dɔ́n* down *sáy* side *Santa* place *Teresita.* '(At) the lower side (of) Santa Teresita.' [ye07de 021]

All locative nouns except *nadó* 'outside' may be followed by locative adverbs as in the following two examples featuring *dɔ́n* 'down' and *bɔtɔ́n* 'under':


#### 8 Spatial and temporal relations

Moreover, all locative nouns except *nadó* 'outside', *dɔ́n* 'down' and *ɔ́p* 'up' may appear with personal pronouns as the ground in the same way as prepositions like *fɔ* 'prep' and *wet* 'with' (hence prepositions that are not [exclusively] used for the expression of locative roles). This sets the locative nouns to which this applies apart from locative prepositions:


The distribution of the locative nouns discussed is summarised in Table 8.3.


Table 8.3 Distribution of locative nouns

In sum, locative nouns are diverse in nature. All locative nouns differ from prepositions in that they do not require an explicit complement. Some locative nouns cannot be preceded by the determiner or the locative preposition *na* 'loc', and hence lack a decisive diagnostic feature of "nouniness" in Pichi (i.e. *nía* 'near', *kɔ́na* 'next to', and *nadó* 'outside').

Other locative nouns are, in contrast, "nouny". They may not only be preceded by the definite article *di* and the preposition *na*, i.e. *bifó* 'before', *bihɛ́n* 'behind',*ɔ́p* 'up(per side)', *bɔtɔ́n* 'bottom', *dɔ́n* 'down (side)', and *míndul* 'middle'. Many of them may also enter as

8.1 Spatial relations

modifier nouns into associative constructions with the generic place nouns *sáy* 'side, place' and *pát* 'part, place'.

Except *nadó* 'outside' and *dɔ́n* 'down', however, all locative nouns also appear in the same syntactic position as prepositions when relating a figure to an explicitly mentioned ground. In this respect, these two locative nouns are therefore similar in their distribution to the deictic adverbs *yá*, 'here', *dé* 'there', and *yandá* 'yonder'.

### **8.1.3 Locative verbs**

Table 8.4 below provides an overview of the most common locative verbs. These verbs serve to express the manner in which a figure is located with respect to a ground. The column entitled 'manner of location' groups these verbs into three classes (cf. Ameka 2007).


Table 8.4 Locative verbs

With the exception of the locative-existential copula *dé* 'be.loc', all other verbs listed above are labile verbs. Hence they may be used as (inchoative-)stative verbs in intransitive clauses and as dynamic verbs in transitive clauses. In intransitive clauses, the figure is the theme subject (36), and in transitive clauses, the figure is the patient object (37). The ground is expressed as a locative adverb(ial phrase) in both alternations:

(36) *Dɛn líng fɔ dán butaca.*

> 3pl lean prep this armchair

'They're sitting reclined in that armchair.' [befn07 207]

8 Spatial and temporal relations

(37) *E* 3sg.sbj *líng=an* lean=3sg.obj *dé.* there 'He leaned it there.' [li07pe 063]

The copula *dé* 'be.loc' expresses existence in a location or in a manner in its most general sense (cf. §7.6.1). More specific nuances of location are expressed by other locative verbs. Compare the stative use of *ráwn* 'surround' in the intransitive clause in (38):

(38) *Di* def *ríba* river *e* 3sg.sbj *ráwn* surround *di* def *hós.* house 'The river flows around the house.' [dj05be 228]

Next to its use as a locative noun (39), the multifunctional item *nía* 'near' may be employed as a an inchoative-stative (40) or dynamic verb (41) like any other locative verb, although the latter usage is rare:


Some locative verbs select specific figures according to the criterion of animacy. For example, *sidɔ́n* 'sit (down)' generally implies an animate (e.g. *pikín* 'child') and *pín* 'stick (into)' an inanimate (e.g. *stík* 'tree') figure. Consider (42) and (43) respectively:

(42) *E* 3sg.sbj *sidɔ́n* seat *di* def *pikín* child *na* loc *butaca.* armchair 'She seated the child in (the) armchair.' [dj07ae 234]

(43) *E* 3sg.sbj *pín* stick *di* def *stík* tree *na* loc *grɔ́n.* ground 'He stuck the stick in (the) ground.' [li07pe 092]

In contrast, all the other verbs listed in Table 8.4 exhibit no such restrictions. This includes verbs that denote other, typically human postures. For example, *tínap* 'stand (up)' may appear with an inanimate (44) or animate (45) figure as well as in intransitive and transitive (46) clauses alike:

8.1 Spatial relations


Also compare the intransitive use of *slíp* 'sleep, lie, lay' in (47) with the transitive use of *slíp* in (48). Both sentences involve the inanimate figure *bɔ́tul* 'bottle':


'He laid the bottle on the table [in a horizontal position].' [li07pe 072]

The verb *jám* 'make/be in contact' denotes contact between figure and ground. The meaning of *jám* contains no connotation with respect to the type of contact. Hence intransitive *jám* means 'be in contact' in (49). Note the use of the Spanish-derived verb *para* 'stand' as a labile locative verb just like its Pichi equivalent *tínap* 'stand (up)' in (46) above:

(49) *Dɛn* 3pl *para* stand.up *di* def *búk* book *dɛn* pl *sé* qot *dɛn* 3pl *jám* make.contact *dɛn* 3pl *sɛ́f.* self

'The books were stood up [in such way] that they're in contact with each other.' [dj07re 044]

When *jám* is used transitively, context may imply a sudden or forceful contact, as in the following sentence:

(50) *So* so *di* def *mán* man *kán* pfv *pás* pass *nía* near *ín,* 3sg.indp *e* 3sg.sbj *jám=an,* make.contact=3sg.obj *di* def *plét* plate *fɔdɔ́n* fall *na* loc *grɔ́n.* ground

'So the man passed near her, he bumped into her, the plate fell to the ground.' [au07se 013]

#### 8 Spatial and temporal relations

The following two examples involve the stative/dynamic alternation of the verb of adhesion and attachement *hɛ́ng* 'be hung onto, hang onto':


### **8.1.4 Motion verbs**

Besides the locative verbs discussed in §8.1.3, Pichi features verbs of diverse semantic types whose meanings also include a change of location, and hence motion. A selection of the most common ones in the corpus is provided in Table 8.5. Some of these verbs contain the additional meaning components of direction (e.g. *gó* 'go (away)') and/or manner of motion (e.g. *júmp* 'jump'). Further, some verbs denote self-motion of the figure subject, hence are lexically intransitive (e.g. *wáka* 'walk'), or preponderantly appear in intransitive clauses (e.g. *ɛ́nta* 'enter'). Others involve motion caused by the figure subject and are therefore more likely to occur in transitive clauses with an overt ground object than in intransitive clauses without one (e.g. *drɛ́b* 'drive', *pút* 'put').

Moreover, the verbs listed in Table 8.5 differ in the way the ground is expressed as a participant in the clause. Hence we find the ground expressed as prepositional phrases (PP), objects (O), and as objects or prepositional phrases in serial verb constructions (SVCs).

The most commonly employed verbs to simultaneously encode motion and direction are *gó* 'go (away)', *kán* 'come', *kɔmɔ́t* 'go/come out of', and *rích* 'arrive (at)'. These verbs also function as V2 in motion-direction SVCs. With any of these four motion verbs, the ground (i.e. the source or goal of the motion) may be expressed as an object of a transitive clause (53) or as a prepositional phrase in an intransitive clause (54). The second alternative is, however, attested in the majority of cases:


The preference for a prepositional phrase rather than an object also holds when the ground is a named place, such as Malabo, the capital of Equatorial Guinea (55). A PP is



Table 8.5 Motion verbs

also the favoured option when the ground occurs in a motion-direction SVC featuring one of the motion verbs listed above as a V2 (cf. §11.2.1).

(55) *Bɔt* but *wé* sub *e* 3sg.sbj *kán* come *na* loc *Malábo,* Malabo *ɛ́ni* every *nɛ́t* night *in* 3sg.poss *abuela* grandmother *kin* hab *kán* come *hambɔ́g=an.*

bother=3sg.obj

'But when she came to Malabo, every night her grandmother would come bother her.' [ed03sb 042]

In principle, the preposition *fɔ* 'prep' may introduce the inanimate goal of a motion verb instead of *na* 'loc' (56). In practice, the use of the general locative preposition *na*

#### 8 Spatial and temporal relations

instead of *fɔ* 'prep' as in (54) above is pervasive. Nevertheless, *fɔ* must be used when the goal (or any other locative role) is animate (57):


All other verbs in Table 8.5 whose goals may be expressed as a PP, an SVC, and an O exhibit the same pattern with respect to ground marking. This applies to locomotion verbs, such as *wáka* 'walk', *rɔ́n* 'run', or *pás* 'pass (by)', to the caused location verb *pút* 'put', or the caused motion verbs *ték* 'take', *bríng* 'bring', and *kɛ́r* 'carry, take'. The following three examples featuring the verb *kɛ́r* once more present the PP (58), the object (59) and the SVC alternatives (60). Again the PP option is the most common one. Note that the goal object *hospital* 'hospital' in (59) is positioned to the right of the patient object *di pikín* 'the child':


The manner-of-motion verbs *wáka* (also *wɔ́k*) 'walk', *rɔ́n* 'run', and *fláy* 'fly' are intransitive. Speakers univocally reject these verbs in grammaticality judgments featuring an undergoer or goal object (cf. §9.2.1 for more details).

Next to these, we find the manner-of-motion verbs *fɔdɔ́n* 'fall' and *pás* 'pass'. These two verbs allow for the ground to be expressed as an O or a PP without any difference in meaning. Compare *fɔdɔ́n* 'fall' in the following two examples:

(61) *E* 3sg.sbj *fɔdɔ́n* fall *di* def *béd.* bed 'He fell from the bed.' [pa07me 042]

8.1 Spatial relations

(62) *Di* def *bolí* pen *fɔdɔ́n* fall *frɔn* from *di* def *tébul.* table 'The pen fell off the table.' [dj05be 204]

In turn, the use of either the PP or the O strategy of ground marking is accompanied by a change in meaning with the two manner-of-motion verbs *júmp* 'jump' (63) and *klém* 'climb' (64). When the ground is expressed as an object, a clause featuring these two verbs is usually interpreted as involving locomotion (hence motion with a path) of the figure, as in the following two sentences:


When the ground is, however, encoded as a PP, these two verbs may denote motion without a path, or locomotion with a path. Compare the alternative translations of (65), featuring *júmp* 'jump':

(65) *Miguel* name *Ángel* name *de* ipfv *júmp* jump *pantáp* top *di* def *béd.* bed 'Miguel Ángel is jumping on/onto the bed.' [dj07ae 019]

Likewise, speaker (au) finds (66) unacceptable, because he interprets the clause featuring *klém* 'climb' as involving motion without a path on the ground *chía* 'chair':

(66) *Nóto* neg.foc *"e* 3sg.sbj *klém* climb *pantáp* top *di* def *chía."* chair 'Not "he climbed [being] on the chair".' [au07se 085]

The "propulsion verbs" (Longacre 1996: 200) *híb* 'throw' and *flíng* 'fling' are causedmotion verbs without a direction component in their meaning. Here, the ground is preferably expressed as a PP or an equivalent locative adverbial as in the following examples:


#### 8 Spatial and temporal relations

The propulsion verb *sɛ́n* equally involves caused motion without direction when used with the sense 'throw (with aim)'. However, *sɛ́n* additionally involves the notion of aim, hence has a manner component in its meaning:

(69) *E* 3sg.sbj *de* ipfv *sɛ́n* send *di* def *bɔ́l* ball *fɔ* prep *mék* sbjv *e* 3sg.sbj *nák* hit *di* def *cartón.* carton 'He's throwing the ball with aim in order to hit the cardboard box.' [ra07se 175]

In contrast, when sɛ́n occurs as a transfer verb in a double-object construction, it acquires the sense 'throw to, send', and therefore also features a direction component. In such double-object constructions, the ground, a usually animate recipient, is only expressed as an object, not as a PP:

(70) *E* 3sg.sbj *sɛ́n=an* send=3sg.obj *di* def *bɔ́l.* ball 'He threw the ball to him.' [ra07se 093]

Another motion verb which may appear in double-object constructions and has a direction, manner, and causation component is *pút* 'put' (covered in detail in §9.3.4).

### **8.1.5 Expressing source and goal**

The foregoing sections have shown that the prepositions *na* 'loc' and *fɔ* 'prep' have a very general meaning and participate in various types of clauses expressing spatial relations. We have seen that these two prepositions may also mark the ground in clauses with a motion-to and a motion-from component. For example, in (54) above *na* marks the source of *kɔmɔ́t* 'go/come out of', and in (57) above *fɔ* the goal of *gó* 'go'.

In fact, any preposition or locative noun that may serve to express an 'at rest' location role does not contribute any meaning to the motion component of the spatial relation. Instead, these elements specify the part of the ground where the figure is located (cf. Essegbey 2005). Compare the locative nouns *ɔntɔ́p* 'top' (71) and *nía* 'near' (72), which both express 'at rest' location and appear with motion verbs in these two sentences:

(71) *Di* def *pambɔ́d* bird *de* ipfv *fláy* fly *ɔntɔ́p* top *di* def *stík.* tree

'The bird is flying over/above the tree.' [ro05ee 099]

(72) *A* 1sg.sbj *nó* neg *nó* know *wétin* what *mék* make *Anto* name *púl* pull *Reina* name *náw* now *nía* near *Tokobé.* name 'I don't know how come Anto pulled Reina away from Tokobé.' [ab03ab 157]

Hence, when a motion verb lacks a directional sense, it is the combined meaning of the verb, the preposition, and the complement that provides the meaning of the entire construction. The following sentences featuring the prepositions *na* 'loc' and *fɔ* '*prep*' are therefore not interpreted as involving 'at rest' location. Instead, the compositional meaning suggests a goal sense:

8.1 Spatial relations

(73) *Dɛn* 3pl *rɔ́n* run *na* loc *farmacia,* pharmacy *receta* prescription *dé* of *mɛ́rɛsin.* medicine 'They ran to [\*in] the pharmacy, [to get a] prescription for medicine.' [ab03ab 123]

(74) *Dɛn* 3pl *pús* push *di* def *motó* car *na* loc *garaje.* garage 'They pushed the car into [\*in] the garage.'

Sometimes, however, there may be room for ambiguity between a motion and a location reading as in (75), featuring the propulsion verb *sút* 'shoot', which lacks a directional sense. The ground PP introduced by the locative noun *bifó* 'before' may be interpreted as a location (at rest), a source (motion-from) or a goal (motion-to):

(75) *Di* def *soldado* soldier *sút* shoot *bifó* before *di* def *hós.* house 'The soldier shot in front of/at/from the front of the house.' [dj05be 188]

Any potential ambiguity between the goal and source senses of *na* and *fɔ* may be eliminated by employing the directional prepositions *frɔn* 'from, since' (76) and *sóté* 'until, up to' (77):


Alternatively, a motion-direction SVC may be employed to mark a goal with verbs permitting such use as in (78). A biclausal structure featuring a modifying purpose or other adverbial clause may also serve the same end:


Nevertheless, even in clauses featuring inherently directional verbs where no such ambiguity could possibly arise, the goal or source is sometimes additionally marked with a directional preposition. Compare the following example, in which the motion-from sense of *kɔmɔ́t* 'come out of' is reiterated by the ablative motion-from preposition *frɔn* 'from':

8 Spatial and temporal relations

(80) *Olinga* name *kɔmɔ́t* come.out *frɔn* from *bɔtɔ́n.* bottom 'Olinga comes from the bottom [worked himself up from the bottom].' [ye03cd 068]

The general locative preposition *na* 'loc' may also additionally mark the ground when preceded by the directional prepositions *frɔn* 'from' and *sóté* 'until, up to'. This usage is not attested with the associative preposition *fɔ* 'prep':


'(…) so that he should be able to get everywhere (even) up to Riaba.' [fr03cd 070]

The use of the preposition *fɔ* 'prep' may open up another space of ambiguity. *Fɔ* may mark an animate source or beneficiary. Hence, the meaning of clauses featuring verbs which may assign both animate source and beneficiary roles are potentially ambiguous. Compare *recibe* 'receive' and *báy* 'buy' below:


Speakers may resort to other means of expressing these relations in pursuit of disambiguation. Example (83) above and (85) below were both elicited by means of the Spanish sentence *recibió un regalo dé su mamá* 'she received a present from her mother'. In the sentence below, speaker (ro) prefers to employ the transfer verb *dás* 'give as present' which assigns an agent instead of a theme subject:

(85) *Mi* 1sg.poss *mamá* mother *bin* pst *dás* give.as.present *mí* 1sg.indp *sɔn* some *regalo.* present 'My mother gave me a present.' [ro05ee 055]

Speaker (ro) also employs a partitive possessive construction in (86) below in order to render the meaning of Spanish *compré un coche dé mí jefe* 'I bought a car from my boss.' Compare (86) below to (84) above, where speaker (dj) uses the *fɔ*-possessive construction instead (which is structurally similar to the Spanish *de*-possessive construction):

8.1 Spatial relations

(86) *A* 1sg.sbj *bin* pst *báy* buy *wán* one *mi* 1sg.poss *másta* boss *in* 3sg.poss *motó.* car 'I bought one of my boss's cars.' [ro05ee 057]

The manner-of-motion verb *pás* 'pass (by)' is employed to express motion-past a ground. The ground is normally expressed as a PP introduced by a locative preposition (87) or locative noun (88):


The nature of a spatial relation may be specified in detail by making use of the appropriate combination of motion verbs, locative prepositions, locative nouns, and SVCs.

For example, the situation in (89) involves a figure (the theme *pikín* 'child') which undergoes a change-of-location (denoted by *fɔdɔ́n* 'fall') in a motion-from along a path (specified by *frɔn* 'from') out of the specific part (the superior location *ɔ́p* 'upperside') of the ground (the source *stík* 'tree'):

(89) *Di* def *pikín* child *fɔdɔ́n* fall *frɔn* from *ɔ́p* up *di* def *stík.* tree 'The child fell from up in the tree.' [dj05be 201]

In (90), the figure (*wi* '1pl') instigates a motion-from (denoted by *kɔmɔ́t* 'go out') out of the specific part (the anterior location *bifó* 'before') of the ground (the source *chɔ́ch* 'church'):

(90) *Wi* 1pl *kɔmɔ́t* go.out *bifó* before *di* def *chɔ́ch.* church

'We went away from the front side of the church.' [dj05be 179]

Sentence (91) features a change-of-location (denoted by the manner-of-motion verb *fláy* 'fly, rush') in a motion-to (expressed through the V2 *gó* 'go' of a motion-direction SVC) into the specific part (the interior location *ínsay* 'inside') of the ground (the goal *Ela Nguema*, a quarter of Malabo):

(91) *Chico,* boy *a* 1sg.sbj *wánt* want *fláy* fly *gó* go *ínsay* inside *Ela Nguema* place *náw* now *só.* like.that 'Man, I'm about to rush to Ela Nguema right now.' [dj07ae 360]

#### 8 Spatial and temporal relations

Additional dimensions that may add to the complexity of a spatial relation are manner modifications to the clause, reciprocity and animacy. For example, the idiomatic expression *na X hán*, literally 'in X's hand' (where X is the possessor) encodes an animate source as in the following example (cf. §7.6.4 for the use of this idiom in possessive clauses):

(92) *Dɛn* 3pl *púl* remove *di* def *motó* car *na* loc *in* 3sg.poss *hán.* hand 'They seized the car from him.' [to07fn 206]

The locative noun *nía* 'near, next to' expresses various degrees of proximity to the ground including contact with it. Compare the use of *nía* with the verb of adhesion *jám* 'be in/make contact with' in (93). *Nía*, as well as *kɔ́na* 'next to' are also used to express a reciprocal spatial relation, in which figure and ground are ground and figure to each other (94):


Clauses which express spatial relations can be modified further for manner independently of the meaning of the verb.This may be done through adverbial clauses introduced by *sé* 'qot' (cf. §10.7.2), *wé* 'sub' (cf. §10.7.1), or secondary predication (cf. §11.3).

The sentence (95) exhibits a complex spatial relation featuring the figure *e* '3sg.sbj' that has carried out a motion-past (i.e. *pás* 'pass by') the proximity (i.e. *kɔ́na* 'next to') of the ground *chía* 'chair'. The clause is followed by the secondary predicate *dé wáka* 'ipfv walk' which provides information about the manner of movement. The secondary predicate is in turn modified by the compound adverbial *rɔn-sáy* 'backwards':

(95) *E* 3sg.sbj *pás* pass *kɔ́na* next.to *chía* chair *de* ipfv *wáka* walk *rɔn-sáy.* wrong.cpd-side

'She passed by next to (a) chair walking backwards.' [au07se 051]

### **8.2 Temporal relations**

The clause-internal temporal relations of location in time, duration, and iteration are established through adverb(ial)s, quantifiers, prepositions, and lexicalised phrases featuring verbs. The expression of standard time units is characterised by a high incidence of conventionalised codemixing.

8.2 Temporal relations

### **8.2.1 Standard time units**

In Pichi, the two equal halves of the day are split into *dé* 'day' and *nɛ́t* 'night'. The conventionalised associative constructions *mɔ́nin tɛ́n* 'morning time' = 'morning', *sán tɛ́n* 'sun time' = 'midday, noon' (96), *ívin tɛ́n* 'evening time' = 'afternoon, evening', *míndul nɛ́t* 'middle night' = 'midnight' (97) denote the central points of the twenty-four hour day:


The expression *áftanun* 'afternoon' is occasionally heard in the speech of Group 2 speakers (cf. §1.3) in the greeting formula *gúd áftanun* 'good afternoon'. However this word is not usually employed to denote the corresponding period of the day.

The concept 'dawn' may be expressed by means of paraphrase, i.e. via emphatic repetition of the modifier noun *mɔ́nin* 'morning' as in (98) or the use of another emphatic element (here the quantifier *sósó* 'only'), with or without repetition for emphasis (99):

	- '(…) early this morning, very early this morning.' [ye05ce 048]

An additional way of expressing 'dawn' is through a clause featuring the subject *mɔ́nin* and the verb *brék* '(to) dawn' (100), or simply, by way of the Spanish noun *madrugada* 'dawn':


#### 8 Spatial and temporal relations

When telling the time of day, Spanish lexical items are fit into a conventionalised codemixed construction, which does not have an exact equivalent in Spanish (cf. also §13.3.1 on codemixing). There is no other generally accepted way of telling the time:

(102) *So* so *yu* 2sg *wánt* want *dé* be.loc *dé* there *las* the.pl *cuatro,* four *wi* 1pl *dɔ́n* prf *dé* be.loc *las* the.pl *tres* three *y* and *veinte.* twenty 'So you want to be there at four (and) we're already here at three twenty. [ma03ni 005]

The Pichi day names *mɔ́nde* 'Monday', *tyúsde* 'Tuesday', *wɛ́nsde* 'Wednesday', *tɔ́sde* 'Thursday', *fráyde* 'Friday', *satidé* 'Saturday', and *sɔ́nde* 'Sunday' are (falling) out of use. Instead, the vast majority of speakers employ the corresponding Spanish day names *lunes*, *martes*, *miércoles*, *jueves*, *viernes*, *sábado*, and *domingo* at all times. The codemixed sentences in (103) reflect typical usage.

	- 'Go [pl] in the same plane, (on) Saturday!' [fr07se 167]

The Spanish noun phrase *fin de semana* is also usually recruited to express 'weekend':

(104) *A* 1sg.sbj *go* pot *lɛ́f* remain *na* loc *Lubá* place *sóté* until *fin de semana.* weekend 'I'll remain in Luba until the weekend.' [ye05ce 010]

The following Spanish designations for the months of the year are in use: *enero* 'January', *febrero* 'February', *marzo* 'March', *abril* 'April', *mayo* 'May', *junio* 'June', *julio* 'July', *agosto* 'August', *septiembre* 'September', *octubre* 'October', *noviembre* 'November', *diciembre* 'December'. Hence, dates are also exclusively expressed in codemixed structures like the following one:

(105) *El* the *diez* ten *de* of *agosto,* August *bay* by *gɔ́d* God *in* 3sg.poss *páwa,* power *a* 1sg.sbj *go* pot *pás* pass *na* loc *yá.* here '(On) the tenth of August, by the grace of God, I'll pass by this place.' [ab07fn 113]

The two seasons of the year may be designated by the compounds *ren-sísin* {rain.cpdseason} 'rainy season' (106) and *dray-sísin* {dry.cpd-season} 'dry season'. An alternative designation for the rainy season is the phrasal expression *tɛ́n fɔ rén* (107):

(106) *Dís* this *dé* day *dɛn* pl *ren-sísin* rain.cpd-season *go* pot *bigín.* begin 'These days, the rainy season should begin.' [dj05ce 059]

8.2 Temporal relations

(107) *Wi* 1pl *dé* be.loc *tɛ́n* time *fɔ* prep *rén.* rain 'We're in the rainy season.' [ro05ee 116]

The noun *amatán* stands for 'harmattan', the dry and dusty seasonal weather condition throughout West Africa (between November and March):

(108) *Wí* 1pl.indp *de* ipfv *kɔ́l* call *yá* here *só* like.that *amatán* harmattan *dán,* that *lɛk* like *sé* qot *e* 3sg.sbj *kin* hab *dé* be.loc *lɛkɛ* like *niebla.* fog 'Here, we call harmattan that, like it's usually like fog.' [ye05ce 062]

### **8.2.2 Temporal deixis**

Adverb(ial)s, quantifiers, prepositions, and lexicalised phrases featuring verbs are recruited for the expression of temporal deixis within the clause. These means are summarised in Table 8.6 below with respect to the temporal relations of location, duration, and iteration.

In Table 8.6, the letter "X" stands for a compatible time-unit, like *tɛ́n* 'time', *lunes* 'Monday', *tú dé* 'two days', *wán wík* 'one week', *tú mún* 'moon, month', or *wán hía* 'year'. Optional elements are in parentheses.There is considerable flexiblity with regard to TMA marking, the expression of participants, and the use of prepositions or locative nouns in the phrasal expressions in the column entitled "temporal expressions". Therefore, I limit myself to including the most common alternative in the table, and only provide a free translation. Exact glosses of these phrases can be found in the examples further below.

A relation between event time and a point of reference in the present, future, and past can be established by combining an element from Table 8.6 with absolute time reference (i.e. time points like *las dos* 'two o'clock' and *sán tɛ́n* '(after)noon' or calendric units like *viernes* 'Friday') with the appropriate TMA marking. Compare (96), (102), and (103) above.

Some items lexically incorporate time reference to the present, past, or future. Compare the time adverb *náw* 'now' (109) and the temporal nouns*tidé/tudé* 'today' (110). Note that the reference point of *tidé/tudé* is event time, not absolute time. Hence *tidé* in (110) may refer to 'today', the actual day on which the sentence was uttered, or to 'that day', the day on which speaker (ye) conversed with the subject *e* '3sg.sbj':



Table 8.6 Temporal deixis

#### 8.2 Temporal relations

The equally synonymous temporal nouns *tumɔ́ro/tumára* 'tomorrow' incorporate future reference to a day ahead of event time (111). When *tumɔ́ro* is combined with the temporal preposition *ápás* 'after', the resulting collocation means 'the day after tomorrow' and denotes a point of reference two days into the future ahead of event time (112):


The temporal noun *yɛ́stadé* 'yesterday' relates event time to a reference point one day back into the past (113). The temporal preposition *ápás* 'after' also combines with *yɛ́stadé* 'yesterday' in the collocation *ápás yɛ́stadé* 'the day before yesterday' (114):


The temporal nouns *tumɔ́ro/tumára* and *yɛ́stadé* express relative time reference in the same way as *tidé/tudé* above. Depending on context, they may therefore also be translated as 'one day after event time' and 'one day before event time', respectively. Examples (112) and (114) above also show that the preposition *ápás* 'after' may be used to indicate both a posterior and an anterior temporal relation. *Ápás* may therefore be combined with *tumɔ́ro* 'tomorrow' as well as *yɛ́stadé* 'yesterday'. The "spatial frame of reference" (Levinson 2003: 24) of temporal posteriority is characterised by a mirror-like "reflection" (Beller et al. 2005:222) of the speaker's vantage point into both directions of the time stream.

Temporal deixis involving time units other than two days in either direction from event time is achieved through a variety of means. The quantifier *nɛ́ks* 'next' may modify the Pichi nouns *wík* 'week', *mún* 'month', and *hía* 'year' and thereby remove the reference point from event time into the future by one unit. Compare (115) and also note the use of the spatial and temporal preposition *sóté* 'until, up to' which expresses extent:

(115) *A* 1sg.sbj *de* ipfv *lɛ́f* leave *na* loc *Lubá* place *sóté* until *di* def *nɛ́ks* next *wík.* week 'I'm remaining in Luba until the next week.' [ye05ce 014]

#### 8 Spatial and temporal relations

The quantifier *lás* 'last' mirrors the time reference of *nɛ́ks* 'next'. *Lás* 'last' pushes a reference point into the past by one unit from event time as in (116–117). Note the presence of the definite article *di* 'def' in (116):


'If she hadn't gone last year, she'd be with us right here.' [dj05ae 059]

In Pichi, the expression of punctual location in time does not require the use of a locative preposition or locative noun (e.g. *na* 'loc', *fɔ* 'prep', or *ínsay* 'inside') if the temporal expression is inherently time deictic. This is the case in various examples throughout this section featuring relational items like *nɛ́ks* 'next' and *tumára* 'tomorrow' above or *lás* 'last'.

The collocation *lɔ́n tɛ́n* 'long time ago' is also inherently relational. Rather than expressing duration (i.e. \*for a long time), its meaning includes an unspecified reference point in the past:

(118) *E* 3sg.sbj *bin* ipfv *dɔ́n* prf *pás* pass *lɔ́n* long *tɛ́n,* time *nóto* neg.foc *lɔ́n* long *lɔ́n* red *tɛ́n.* time 'It happened long ago, not very long ago.' [ma03sh 001]

The collocation *las-nɛ́t* 'last night' is a compound (119). The lexicalisation of this collocation distinguishes it from other time expressions featuring *lás* 'last' (cf. e.g. *lás hía* 'last year' in 117), which are not usually subjected to the tonal derivation characteristic of compounding:

(119) *Las-nɛ́t* last.cpd-night *a* 1sg.sbj *chakrá* destroy *mi* 1sg.poss *sɛ́ns.* brain 'Last night, I drank myself senseless.' [ra07fn 060]

Spatial expressions are, however, used to encode temporal relations if the temporal expression in the clause is not inherently time deictic. This may apply to temporal location as in (120), where the locative noun *ínsay* 'inside' fulfills this function.

(120) *A* 1sg.sbj *de* ipfv *wét* wait *sé* qot *mék* sbjv *a* 1sg.sbj *gó* go *ínsay* inside *tú* two *dé.* day 'I'm hóping to go in two days.' [dj05ae 033]

8.2 Temporal relations

Neither the associative preposition *fɔ* 'prep', nor the general locative preposition *na* 'loc' are generally employed to mark adverbial phrases with a location-in-time sense. An exception in the data is the presence of *na* 'loc' in the lexicalised collocation *na nɛ́t* 'at night'. All other standard periods of the day are expressed through associative constructions featuring the generic noun *tɛ́n* 'time' (121). In view of the limited number of *tɛ́n* 'time' collocations in Pichi and their often idiosyncratic meanings (cf. 6), even these expressions may be seen as lexicalised structures:

(121) *Bɔ́y* boy *dɛn* pl *dé* be.loc *dé,* there *mɔ́nin* morning *tɛ́n* time *sán* sun *tɛ́n* time *na* loc *nɛ́t,* night *na* foc *Píchi* Pichi *dɛn* 3pl *de* ipfv *tɔ́k* talk *Píchi.*

Pichi

'(The) guys are there, in the morning, at day time, at night, it's only Pichi that they talk.' [au07se 257]

The extension of spatial notions into the temporal domain is also reflected in the means employed to encode the temporal relation of anteriority by means of the locative noun *bifó* 'before'. In contrast to *ápás* 'after', which may express anteriority or posteriority, the use of *bifó* in (122) incorporates an "intrinsic" (Beller et al. 2005: 221) temporal perspective. The intrinsic beginning or end of the time unit itself provides the temporal reference point. Contrary to the "reflection" perspective inherent to *ápás* 'after' (123), a relational linkage with the vantage point of the speaker is not expressed:


'They'll return the week after next [in two weeks].' [he07fn 209]

Duration in time for a specific period is expressed by means of the general associative preposition *fɔ* 'prep' followed by a time expression:

(124) *Yu* 2sg *go* pot *moja* soak *di* def *rɛ́s* rice *na* loc *watá,* water *fɔ* prep *tidé,* today *tú* two *dé* day (*…*)

'You soak the rice in water, for today [one day], (for) two days (…)' [dj03do 019]

An equally common way of expressing duration for a specified period is by means of the verb *kɛ́r* 'carry, take, last'. The "figure" enduring in time is expressed as the subject of the clause and may be inanimate (125) or animate (126), while the specified time period is the object of *kɛ́r*:

	- '(…) but the fever only lasted for a day.' [ru03wt 062]

#### 8 Spatial and temporal relations

[hi03cb 214]

(126) *Háw* how *mɔ́ch* much *tɛ́n* time *yu* 2sg *go* pot *kɛ́r* carry *na* loc *kɔ́ntri?* country 'How long are you going to stay in (your) hometown?' [lo07he 046]

Aside from that, elements that express motion through space are put to use for establishing temporal relations of duration. Firstly, the allative motion-to preposition/clause linker *sóté* 'up to, until' also expresses temporal duration-to (127).

(127) *A* 1sg.sbj *de* ipfv *lɛ́f* remain *na* loc *Lubá* place *sóté* until *wík* week *fínis.* finish 'I'm staying in Luba until the end of the week.' [ro05ee 128]

Secondly, example (128) and (127) illustrate the use of *sóté* together with the lexicalised (factative-marked) clausal structures *mún dɔ́n* 'month done' = 'at the end of the month' and *wík fínis* 'week finish' = 'at the end of the week'. Both expressions establish a punctual and past temporal reference point:

(128) *Mék* sbjv *e* 3sg.sbj *wét* wait *sóté* until *mún* month *dɔ́n,* done *wé* sub *a* 1sg.sbj *gɛ́t* get *di* def *mɔní* money *a* 1sg.sbj *go* pot *báy* buy *di* def *chɔ́p.* food 'Let him wait until the month is over, when I get the money, I'll buy the food.'

The multifunctional item *sóté* 'up to, until' may also introduce finite adverbial extent clauses, in which the subordinate verb may take the full range of TMA and person marking (129). Next to that, *sóté* also appears as a temporal preposition directly followed by a verb as in (130). The resulting combination acquires a resultative sense and means that the situation denoted by the verb has been attained. Since *sóté* is also a preposition, it may also take nominal complements. For example, the complement *táya* 'be tired' in (130) is a non-finite, deverbal noun and appears without TMA or person marking:


The ablative preposition *frɔn* 'from, since' marks a source when used with a spatial sense. In the temporal domain, *frɔn* expresses duration-from a reference point (131). The period of duration may be further specified by employing both *frɔn* 'from' and *sóté* 'until' as in (132). I draw attention to the optional use of another lexicalised clausal structure in the second example, namely *e gó* {3sg.sbj go} 'going to' in order to provide an additional allative sense:

8.2 Temporal relations


The temporal preposition *síns* 'since' is specialised to expressing duration-from but its use is marginal when compared with the frequency of *frɔn* 'from':

(133) *Wi* 1pl *dé* be.loc *yá* here *síns* since *las* the.pl *dos.* two 'We're here since two o'clock.' [ab07fn 242]

The transfer of spatial concepts into the temporal domain is also reflected in the kind of verbs employed. Location in the future features the ablative motion verb *kán* 'come', that of past location and duration the motion verb *pás* 'pass (by)' – hence time is conceived as moving and the reference point as fixed:


The verb *sté* 'stay, be a long time' inherently expresses lengthy duration, so no further specification of the length of the period is required (136). The verb is versatile in its syntactic behaviour. Firstly, it may appear as the only verb of a sentence like (136) or participate as a V1 in an adverbial SVC (cf. §11.2.5 for details):

(136) *Na* foc *wán* one *hós* house *wé* sub *e* 3sg.sbj *dɔ́n* prf *sté* remain *nɔ́?* intj 'It's a house that's been around for a long time, right?' [dj05ae 161]

Secondly, the verb *sté* may also appear with an expletive subject followed by an adverbial time clause which specifies the relevant time period:

	- here

'It's been long, I think it's been long since you [pl] have had insecticide [sprayed] here.' [fr03wt 059]

#### 8 Spatial and temporal relations

The quantifier *ɛ́ni* 'every' expresses iteration of the time unit it refers to (138). Time units are generally conceived as countable, *ɛ́ni* is semantically compatible with any time unit including units of the clock (139):


Table 8.7 contains all locative nouns and prepositions that participate in expressing temporal relations in Pichi. The table complements the inventory of locative and nonlocative prepositions presented in Table 8.1 and Table 9.1, respectively.


Table 8.7 Temporal (uses of) prepositions and locative nouns

## **9 Grammatical relations**

Pichi verbs exhibit a large degree of flexibility in the number and type of nominal participants they may co-occur with. The language has no deeply entrenched lexical contrast between transitive and intransitive verbs – there are only very few verbs that cannot be employed in transitive and intransitive clauses alike (cf. §9.2.1). The vast majority of verbs can act freely as intransitive or transitive verbs. However, with the class of labile verbs, either option has consequences for the semantic role attributed to the subject, the causative reading of the verb, and with most verbs, lexical aspect (cf. §9.2.3). In addition, any transitive verb may also occur in a double-object construction (cf. §9.3.4). Moreover, most verbs may appear with deverbal copies of themselves, so-called cognate objects (§9.3.3). In this way, even verbs unlikely to occur with objects in other contexts can be used transitively.

Pichi has numerous more or less lexicalised verb-noun combinations featuring verbs with general meanings (cf. §9.3.1). Next to these, we also find combinations of verbs and associative objects. These objects may fulfil various non-core semantic roles (cf. §9.3.2). Clauses featuring referentially empty, expletive subjects reflect a need for the subject position to be filled in Pichi clauses (cf. §9.2.4). Reflexivity and reciprocity are largely expressed by the same formal means (cf. §9.3.5–§9.3.6). Verb valency may be adjusted through a rich variety of causative and impersonal constructions involving 3rd person pronouns or human-denoting generic nouns (cf. §9.4). Finally, the expression of weather phenomena (cf. §9.3.7) and body states (cf. §9.3.8) provides good examples for the configuration of semantic roles and grammatical relations in two specific semantic fields.

### **9.1 Expression of participants**

Pichi expresses the relation that holds between a verb and the core participants subject and object(s) by word order with full nouns and a combination of word order and morphological case-marking with personal pronouns. Non-core participants are expressed as prepositional phrases, or in specific cases, as adverbial phrases without prepositions. Besides that, SVCs are recruited to mark participants, even if they are less frequent in terms of general frequency.

### **9.1.1 Subjects**

Verbs usually co-occur with at least one participant, namely a subject. Nonetheless, in certain discourse contexts, subject ellipsis occurs (cf. §9.4.1) and some SVCs allow for

#### 9 Grammatical relations

subjects to remain unexpressed (e.g. in certain types of motion-direction SVCs, cf. §11.2.1). Subjects subsume the actor roles of agent (1) and experiencer (2):


Next to that, subjects may instantiate the undergoer semantic roles of stimulus/body state in certain idiomatic expressions (3), theme in the intransitive alternation of locative verbs (4) and property items (5), and patient in the intransitive alternation of change-ofstate verbs (6):


Beyond that, Pichi also uses semantically empty expletive subjects with certain verbs. These are covered in detail in §9.2.4.

### **9.1.2 Objects**

Objecthood is marked by word order alone with full nouns (7) and by morphological case and word order with pronominal objects (8). Full nouns and pronominal objects may both appear in double-object constructions (cf. §9.3.4). The overt expression of objects is, in principle, optional although in practice prototypically transitive verbs are very likely to occur with an object. In (9), the object of *nák* 'hit' remains unexpressed, but it is coreferential with the suject *e* '3sg.sbj' of the main clause:

9.1 Expression of participants


Objects instantiate undergoer semantic roles such as patient, theme, stimulus, recipient and beneficiary, as well as the actor role of experiencer. Hence the only role that is never instantiated by an object is the agent, which is reserved for subjects.

The goal and source of motion verbs like *gó* 'go' and *kɔmɔ́t* 'go/come out' may also be expressed as objects, although prepositional phrases are more common. Compare the goal object *colegio* 'college' in (10):

(10) *So* so *wé* sub *yu* 2sg *kɔmɔ́t* come.out *colegio* college (*…*)

'So when you came out of college (…)' [ab03ay 029]

Transitive clauses involving movement verbs and their objects may also represent cases of idiomatic transitivity as in (11):

(11) *Di* def *tín* thing *de* ipfv *go* pot *mí* 1sg.indp *bad.* bad

'The matter is going bad for me.' [dj07ae 161]

Cognate objects are referentially empty syntactic objects. They serve the pragmatic function of expressing emphasis:

(12) *Dán* that *torí* story *bin* pst *de* ipfv *swít* be.tasty *mí* 1sg.indp *wán* one *swít.* be.tasty 'I really enjoyed that story.' [ye07ga.006]

Beyond that, a variety of other, non-core semantic roles may be expressed by objects in lexicalised verb-noun collocations involving associative objects (cf. §9.3.2.).

### **9.1.3 Prepositional phrases**

Participants with non-core semantic roles are most commonly expressed through prepositional phrases and in specific cases through SVCs (cf. §9.1.4). Table 9.1 lists the prepositions employed for the expression of non-core semantic roles. Refer to Table 8.1 and Table 8.7 for locative and temporal uses of prepositions and locative nouns.

#### 9 Grammatical relations


Table 9.1 Non-locative uses of prepositions

The semantic roles expressed by the prepositions listed in Table 9.1 are provided in Table 9.2 below. The table reveals a bipartite structure in the marking of semantic roles. The prepositions*fɔ* 'prep' and *wet* 'with' may express virtually all roles listed. In contrast, all other prepositions express a single semantic role. In addition, the prepositions *bay* 'by' and *bitáwt* 'without' are marginal, and in the case of *bay*, only encountered in idiomatic expressions. Given the large range of functions covered by *fɔ* and *wet*, the expression of semantic roles therefore relies just as much on the meaning of the verb as it does on that of the preposition it co-occurs with.

Table 9.2 Expression of non-locative semantic roles by prepositions


The preposition *fɔ* 'prep' may introduce the stimulus NP of a small number of experiential verbs with affected agents. The corpus features five such verbs: *bísin* 'bother, be busy (with)', *gládin* 'be glad (about)', *kɔ́stɔn* 'be used to', *lúkɔt* 'watch out (for)', *sém* 'be ashamed (about)'. Of these verbs, only *lúkɔt* is intransitive; the only non-subject participant this verb may appear with is a stimulus PP (13):

9.1 Expression of participants

(13) *Lúkɔt* look.out *fɔ* prep *tif-mán* steal.cpd-man *dɛn!* pl 'Watch out for thieves!' [dj07ae 096]

In contrast, the stimulus of the verbs *bísin* and *kɔ́stɔn* may either be expressed as a PP in an intransitive clause or an object in a transitive clause. There is no difference in meaning between the two options:


The verb *kɔ́stɔn* 'be used to' is also attested with a third option: it may take a stimulus PP marked by the preposition *wet* 'with'.

(18) *A* 1sg.sbj *dɔ́n* prf *kɔ́stɔn* be.used.to *wet* with *di* def *trɔ́n* strong *láyf.* life 'I've got used to a difficult life.' [dj07ae 102]

The preposition *fɔ* 'prep' may also mark the stimulus of motion of some agent-induced motion verbs like *háyd* 'hide (from)' or *rɔ́n* 'run away (from)' as in the following example (cf. §8.1.5 for the use of *fɔ* in marking locative source roles):

(19) *E* 3sg.sbj *háyd* hide *fɔ* prep *in* 3sg.poss *kɔ́mpin.* friend 'He hid from his friend.' [dj07re 040]

Verbs other than the ones covered above invariably appear with stimulus objects rather than PPs. Compare *lúk* 'look (at)' in (20). Other verbs in this group are *sí* 'see', *hía/yɛ́r* 'hear' and *listin* 'listen':

9 Grammatical relations

(20) *A* 1sg.sbj *lúk=an.* look=3sg.obj 'I looked at him.' [ab03ab 069]

Prepositional phrases introduced by *fɔ*'prep' also denote the semantic roles of purpose (21) and cause (22), the latter in combination with a body state:

(21) *Mí* 1sg.indp *gí* give *dɛ́n* 3pl.indp *diez* ten *mil* thousand *fɔ* prep *transporte.* transport 'I [emp] gave them ten thousand (Francs) for transport.' [fr03cd 005]

(22) *E* 3sg.sbj *dáy* die *fɔ* prep *tɔ́sti.* thirsty 'He died of thirst.' [dj05be 123]

Nevertheless, in the vast majority of cases, a cause of death due to a body state like *hángri* 'hunger', *tɔ́sti* 'thirst' (23) or *sɔfút* 'wound, injury' (24) is marked by *wet* 'with'. Note however that the cause of a sickness is usually expressed as an associative object (cf. §9.3.2):


The prepositions *fɔséka* 'due to' (and its less frequent variant *foséko*) (25) and *bikɔs* 'because, due to' (26) introduce prepositional phrases with the semantic role of cause. However, *bikɔs* is seldom used as a preposition and far more common as a linker of cause clauses (cf. §10.7.7).

Take note of the verb-object phrase *bɔ́n pikín* 'give.birth child' = 'childbirth', which is nominalised in its entirety and involves *bɔ́n* employed as a deverbal noun:


9.1 Expression of participants

The role of instrument is expressed through *wet* 'with' if instruments (27), materials (28), and functions (29) are involved. Instruments and materials can also be expressed by argument-introducing SVCs involving *ték* 'take' (cf. §9.1.4):


Besides that, the preposition *fɔ* is used for an instrument role in a more general sense of 'by means of' (30). Still, the functional overlap of *wet* and *fɔ* may lead to variation in the marking of certain expressions. Compare 'walk by foot' in (29) above with (31) below:


The preposition *bay* 'by (means) of' is only attested in an idiom in the corpus where it marks an instrument NP in a way similar to the general instrument sense denoted by *fɔ* 'prep' in the two preceding examples:

(32) *El* the *diez* ten *de* of *agosto,* August *bay* by *gɔ́d* God *in* 3sg.poss *páwa,* power *a* 1sg.sbj *go* pot *pás* pass *na* loc *yá.* here '(On) the tenth of August, by the grace of God, I'll pass by this place.' [ab07fn 113]

The preposition *wet* 'with' introduces participants with a comitative role (33). A comitative role may also be expressed through an SVC involving *fála* 'follow' if the accompanee is human (cf. e.g. 34). Comitative *wet* 'with' may shade off into general circumstance (34):

(33) *Yu* 2sg *de* ipfv *ɛ́nta* enter *wet* with *sús?* shoe 'You're coming in with shoes?' [ge07fn 092]

#### 9 Grammatical relations

(34) *Yu* 2sg *nó* know *dán* that *tín* thing *wet* with *yun-bɔ́y* young.cpd-boy *nɔ́?* intj 'You know that thing about young guys right?' [au07se 061]

Negative comitative is occasionally expressed through a PP introduced by *bitáwt* 'without' (with the alternative pronunciation w*itáwt*) (35). However, clausal alternatives are preferred to this rare preposition. One means of rendering 'without' is by employing a relative/adverbial clause construction introduced by *wé* 'sub' as in (36):


The use of a PP is only one of numerous means of expressing manner in Pichi (cf. e.g. §7.7.2), in which case the preposition *wet* 'with' usually serves this purpose (37). An equative or similative participant is introduced by *lɛk* 'like' (38):

(37) *Yu* 2sg *nó* neg *de* ipfv *tɔ́k=an* talk=3sg.obj *wet* with *páwa.* power 'You're not saying it forcefully.' [lo07he 065]

(38) (*…*) *wi* 1pl *fít* can *dé* be.loc *lɛk* like *kɔ́mpin.* friend '(…) we can be (like) friends.' [ru03wt 029]

The expression of a beneficiary role by means of *fɔ* 'prep' is covered in detail in §9.3.4 on double-object constructions.

### **9.1.4 Serial verb constructions**

Serial verb constructions (SVC) are utilised to introduce syntactic objects denoting theme, the standard in comparative constructions, instruments and materials, as well as the accompanee in comitative (cf. §11.2). The areally widespread SVC employing a verb meaning 'give' to mark a beneficiary or recipient role does not exist in Pichi. Compare the following SVC, in which the (fronted) object of *ték* 'take' denotes a material:

(39) *Na* foc *ús=káyn* q=kind *tín* thing *dɛn* 3pl *ték* take *mék* make *dís,* this *digamos* let's.say *dí* this *bɔ́tul?* bottle 'What is, let's say this bottle, made of?' [ye05ce 113]

9.2 Verb classes

On the whole, SVCs are not as frequent as other means of marking participants in Pichi – to the exception of the standard in comparison. The use of a comparative construction featuring the verb *pás* 'surpass' is the ordinary way of introducing the standard object:

(40) *Na* foc *dɛ́n* 3pl *bin* pst *de* ipfv *transfiere* transfiere *mɔní* money *mɔ́* more *na* loc *Western* name *Union* name *pás* pass *guineano* Guinean *dɛn.* pl 'It's them who were transferring more money by Western Union than Equatoguineans.' [ye07je 185]

Motion-direction SVCs involving motion verbs like *wáka* 'walk' and *gó* 'go' express locative roles, often in combination with a prepositional phrase, as in the following example:

(41) *Di* def *gɛ́l* girl *wáka* walk *gó* go *na* loc *tɔ́n.* town 'The girl walked to town.' [ne05fn 243]

### **9.2 Verb classes**

Four lexical classes of verbs may be identified in terms of the grammatical relations they specify and with respect to the semantic roles expressed by their subject and object(s). Intransitive verbs occur with no participant other than an actor subject; transitive verbs occur with a subject and may optionally appear with one or two objects; labile verbs take part in a transitivity alternation: in the intransitive clause, labile verbs appear with an undergoer subject. In the transitive clause, they appear with an actor subject and an undergoer object. In addition, most labile verbs exhibit changes in their lexical aspect class and their causative reading in either alternation. Finally, expletive verbs take referentially empty subjects and may be used transitively or intransitively.

### **9.2.1 Intransitive verbs**

Pichi features a small number of intransitive verbs which do not occur with objects. Elicitation with 360 verbs in the corpus revealed the intransitive verbs listed in Table 9.3 below. The group of dynamic intransitive verbs is made up of locomotion verbs as well as verbs denoting other body experiences, weather verbs, verbs of existence in time and space, and an inherently reciprocal "verb of social interaction" (Levin 1993: 201). All these verbs have in common that they involve experiencer and theme subjects, hence actors that are affected by the situation denoted by the verb.

Inchoative-stative and stative intransitive verbs fall into three classes: modal and aspectual verbs (e.g. *fít* 'can' and *dɔ́n* 'be finished'), verbs denoting existence in place or time (e.g. *blánt* 'reside'), and property items, most of which are human propensities (e.g. *badhát* 'be mean', *fúlis* 'be foolish', *ráyt* 'be right') and physical properties (e.g. *hád* 'be hard', *sáful* 'be slow'). One explanation for the intransitivity of verbs from these three classes is the high time-stability of the situations they denote.

#### 9 Grammatical relations


Table 9.3 Intransitive dynamic verbs

Intransitive verbs may only appear with a subject and may not take objects. Participants other than subjects appear in the guise of prepositional phrases. For instance, the locomotion verbs *fláy* 'fly' and *wáka* 'walk' are intransitive. The use of theme (42) or goal objects (43) is rejected as ungrammatical:


9.2 Verb classes


Table 9.4 Intransitive (inchoative-)stative verbs

SVCs and PPs may be employed if the goal is to be made explicit. Compare the following two sentences:


In contrast to *wáka* 'walk' and *fláy* 'fly', other motion verbs like *gó* 'go' can appear in transitive clauses, in which the goal is expressed as an object. This is particularly so when the goal object is a named place. Compare the object *Lubá* '(the town of) Luba' in (46):

9 Grammatical relations

(46) *Dí* this *miércoles* wednesday *a* 1sg.sbj *de* ipfv *gó* go *Lubá.* place 'This Wednesday, I'm going to Luba.' [ro05ee 119]

The transitive motion verb *gó* 'go' and the intransitive motion verb *rɔ́n* 'run' are also found with a meaning other than physical motion through space. Three such cases of idiomatic transitivity follow with *gó* 'go' in (47–48) and *rɔ́n* 'run' in (49):

	- '(…) she was giving me a traditional treatment.'
	- [*Lit.* 'She was running the village thing for me.'] [ab03ay 101]

The intransitive and dynamic body state, body process, and experiential verbs listed in Table 9.3 above require the use of a PP if a participant other than the subject is to be expressed. The stimulus of *lúkɔt* 'look out' needs to be expressed as a *fɔ*-prepositional phrase:


The verb *fɛ́t* 'fight' cannot take an object either (52). A comitative participant needs to be expressed as a prepositional phrase (53) or within a coordinate structure (54):


9.2 Verb classes

(54) *Djunais* name *wet* with *Boyé* name *dɛn* 3pl *de* ipfv *fɛ́t.* fight 'Djunais and Boyé are fighting.' [dj07ae 394]

The ground associated with the intransitive stative verb *blánt* 'reside' may only be expressed as a prepositional phrase (55–56):


Intransitive property items include *gúd* 'be good' (57) and *bɛ́ta/bɛ́tɛ* 'be very good, better' (58). With both property items, a valency-increasing causative construction is required in order to add a participant in addition to the subject (59):


Compare the intransitive verb *gúd* in the examples above to the transitive, causative use of the labile verb *fáyn* 'be fine', which may be used transitively and intransitively with the corresponding changes in the semantic role of the subject.The undergoer (theme) subject of *fáyn* in (60) becomes an actor (agent) subject in (61). Even if this transitive, causativising use of *fáyn* is rather unusual, it is not ungrammatical:


#### 9 Grammatical relations

Nevertheless, most if not all Pichi verbs may take cognate objects, i.e. deverbal copies of themselves. In this way, even verbs unlikely to occur with objects in other contexts can be used transitively. Example (62) involves the intransitive dynamic verb *dáy* 'die' followed by a cognate object:

(62) *Ey,* intj *dán* that *káyn* kind *spɛ́tikul,* glasses *a* 1sg.sbj *dáy* die *dáy.* die 'Hey, that kind of glasses, (if I had it) I would die.' [ne07ga 015]

### **9.2.2 Transitive verbs**

Verbs other than the ones listed in Table 9.3 and Table 9.4 may appear in transitive clauses followed by an object. Syntactic transitivity is therefore not only a feature of highly transitive verbs with prototypical agent subjects like *bít* 'beat', *nák* 'hit', *kíl* 'kill', or *híb* 'throw'. The labile verbs covered in §9.2.3 as well as other (inchoative-)stative and dynamic verbs characterised by a low degree of inherent transitivity may also be followed by objects. For instance, we find verbs denoting body states and body functions amongst this group.

In (63), the verb *swɛla* 'swallow' is followed by the patient object *ín* '3sg.indp', and (64) features the stimulus object *mí* '1sg.indp', object to the body process verb *láf* 'laugh':


There are no restrictions on the transitive use of body function verbs involving "effected objects" (Hopper 1985), objects that come into existence by the situation denoted by the verb. Compare *swɛ́t* 'sweat' featuring the effected object *wɔtá* 'water' = 'sweat' (65). The same holds for experiential (or human propensity) verbs like *jɛ́lɔs* 'envy, be jealous', which may appear with a stimulus object (66):


9.2 Verb classes

Other verbs low on the transitivity scale behave no differently. For instance, typically stative situations denoted by colour-denoting property items may appear in transitive clauses with a patient object. Compare the labile verb *blú* 'be blue, make blue' in (67):

(67) *A* 1sg.sbj *wánt* want *blú* make.blue *dí* this *motó* car *mék* sbjv *e* 3sg.sbj *chénch* change *kɔ́la.* colour 'I want to (paint) this car blue for it to change (its) colour.' [dj07ae 150]

In the same vein, neither a physical property like *hɔ́t* 'be hot' (68), nor a value concept like *día* 'be expensive' (69) is barred from appearing in a transitive clause. Note the causativising effect of the transitive use of these inchoative-stative labile verbs in (68) and (69) below as well as in (67) above:


Likewise, motion verbs other than the intransitive locomotion verbs listed in §9.2.1 freely alternate between transitive and intransitive uses. The following sentence presents the non-literal use of the manner-of-motion/caused-motion verb *sube* 'go/bring up, rise/ raise', the body state *fíba* 'fever', and the animate experiencer *ín* '3sg.indp':

(70) *Fíba* fever *nó* neg *sube* go.up *ín.* 3sg.indp 'His fever hasn't risen.' [*Lit.* 'The fever hasn't risen on him.'] [eb07fn 171]

### **9.2.3 Labile verbs**

A large number of Pichi verbs are labile; they alternate in their meaning depending on whether they occur with an object in a transitive clause or without an object in an intransitive clause. Labile verbs participate in a transitivity alternation that causes a covariation of the semantic macro-role of the subject (undergoer vs. actor), the causation reading of the verb (non-causative vs. causative), and with most verbs, the stativity value (inchoative-stative vs. dynamic).

Five subclasses of labile verbs can be identified in semantic terms: change-of-state verbs, locative verbs, property items, experiential verbs, and aspectual verbs. In formal terms, only two broad classes, however, need to be distinguished. With the first three subclasses, the intransitive-transitive alternation is accompanied by an inchoativestative/dynamic alternation. With experiential and aspectual verbs, however, both the intransitive and the transitive alternants are dynamic.

#### 9 Grammatical relations

Table 9.5 lists the relevant features of labile verbs in accordance with the two formal and five semantic classes. An additional co-variation feature not included in the table is the tense interpretation of the inchoative-stative and dynamic variants of class (a). The unmarked inchoative-stative variants of class (a) verbs receive a present tense interpretation when they are used as stative verbs. Alternatively, they receive a past tense interpretation if they are used as inchoative verbs.

In turn, the unmarked dynamic variants of class (b) behave like other dynamic verbs and receive a past tense interpretation (cf. §6.3.1 for an extensive treatment). The abbreviation ista stands for inchoative-stative.

Table 9.5 Characteristics of labile verbs


Class (a) labile verbs are employed as inchoative-stative verbs in intransitive clauses and as dynamic verbs in transitive clauses. Either use co-varies with the "role of the subject": Intransitive clauses have an undergoer subject, while transitive clauses feature an actor subject and an undergoer object.

In the corpus, change-of-state verbs constitute the largest subclass of labile verbs. Some representative change-of-state verbs are provided in (71). With some verbs, the change of state of the subject is more likely to have been caused by (a) an external (usually animate and unmentioned) agent, with others (b) by a cause internal to the subject (cf. Croft 1990; Haspelmath 1993; Levin & Hovav 1995). This difference is reflected in the glosses given. Group (a) verbs are rendered with their dynamic meanings, group (b) with their stative meanings. The verbs are also loosely grouped along semantic criteria, such as 'destruction' (e.g. *brók* 'break', *chakrá* 'destroy'), 'material transformation' (e.g. *bɛ́n* 'bend', *bwɛ́l* 'boil'), 'body states' (e.g. *bɛlfúl* 'be satiated', *táya* 'be tired'), and 'natural states' (e.g. *rɔ́tin* 'be rotten', *sók* 'be wet'), 'other human states' (e.g. *wɛ́r* 'wear', *máred* 'marry'):

9.2 Verb classes



In the intransitive clause in (72) below, the change-of-state verb *chák* 'be drunk, get drunk' takes an undergoer subject (with the specific role of patient). In the transitive clause in (73), *chák* now takes an actor subject (with the specific role of agent) and an undergoer (patient) object. In the intransitive clause, the verb has a non-causative meaning, while the verb in the transitive clause has a causative meaning.

At the same time, the aspectual reading of the bare factative change of state verb is adjusted. When the verb is employed as a bare inchoative-stative verb in a basic intransitive clause, as in (72) below, it normally receives a present tense interpretation – the situation holds at reference time. In turn, the dynamic variant of *chák* receives a default past tense interpretation in (73).


'They got the guard drunk first in order for them to be able to steal.' [dj07ae 052]

When used intransitively with factative TMA, there is generally a stronger tendency for change-of-state verbs from group (b) to receive a stative interpretation, as in (72) above. In contrast, many group (a) verbs are more likely to receive an inchoative interpretation focussing on the change-of-state, since most of these verbs feature an implicit agent or (natural) force. When verbs with implicit agents appear in intransitive clauses, there is therefore a higher tendency for speakers to employ the perfect tense-aspect rather than factative TMA in order to indicate a change-of-state. The use of perfect marking via *dɔ́n* 'prf' focuses the end-state of the change of state.

#### 9 Grammatical relations

Compare *fráy* 'fry', an "agentive" group (a) verb in the intransitive and transitive clause, respectively. The combination of perfect marking and "agentive" verb renders a resultative meaning very close to passive voice in (74):


Change-of-state verbs also differ with respect to their likelihood to occur in intransitive or transitive clauses. The higher "agentivity" of group (a) verbs like *fráy* 'fry' makes it less likely for these verbs to appear in agentless, intransitive clauses than group (b) verbs like *bɛlfúl* 'be satiated' or *táya* 'be tired'.

Two further semantic classes of labile verbs are locative verbs and property items. These two subclasses alternate between inchoative-stative and dynamic uses. The two following examples involve the intransitive (76) and transitive (77) use of the locative verb *lé* 'lie, lay'. The latter example also features the transitively used locative verb *slíp* 'sleep, lie, lay'. A more extensive listing of locative verbs and a detailed treatment of their distribution is given in §8.1.3:

(76) *Di* def *kasára* cassava *lé* lie *míndul* middle *tú* two *stík.* tree

'The cassava is lying between two branches.' [li07pe 080]

(77) *E* 3sg.sbj *lé* lay *di* def *bɔ́tul* bottle *pantáp* top *di* def *tébul,* table *e* 3sg.sbj *slíp* lay *di* def *bɔ́tul* bottle *pantáp* top *di* def *tébul.* table 'He laid the bottle on the table, he brought the bottle into a horizontal position on the table.' [li07pe 074]

Property items behave no differently from change-of-state and locative verbs. Consider the intransitive/transitive and stative/dynamic uses of the physical property denoting verb *lɔ́n* 'be long, lengthen' in the two following examples:


9.2 Verb classes

Property items of all other semantic types may be used in the same way as *lɔ́n* '(be) long' (cf. §4.1.2 for a listing of relevant semantic types). Compare the intransitive meaning of 'be small' of the dimension concept *smɔ́l* '(be) small' in the intransitive clause in (80) with the causative meaning 'make small, shrink' in the transitive clause in (81). The imperfective marker *de* 'ipfv' specifies *smɔ́l* in (81) just like any dynamic verb with simultaneous taxis:


A value concept like *fáyn* '(be) fine, beautiful' may also be subjected to the intransitive/transitive alternation characteristic of labile verbs. Compare the intransitive, stative use of this property item:

(82) *Libreville* place *fáyn.* be.fine 'Libreville is (a) nice (place).' [ma03sh 009]

Now consider the transitive use of *fáyn* in the following two sentences. Note that a transitive use may also lead to an idiosyncratic meaning of *fáyn*. Sentence (83) presents the regular, derived transitive meaning of 'make beautiful', while (84) represents a case of idiomatic transitivity with a "dative of interest" reading of the experiencer object pronoun of *fáyn*. Such a meaning is also recorded for cases of idiomatic transitivity with other verbs low on the transitivity scale, e.g. the motion verbs *gó* 'go' (47), *rɔ́n* 'run' (49), and *sube* 'rise, raise' (70):

(83) *Nóto* neg.foc *klós* clothing *go* pot *fáyn* make.fine *yú* 2sg.indp *sino* but *que* that *na* foc *yú* 2sg.indp *gɛ́fɔ* have.to *fáyn* make.fine *yu* 2sg *sɛ́f.* self

'It's not clothes that would make you beautiful, it's rather you that has to make yourself beautiful.' [dj07ae 176]

(84) *Dán* that *bɛ́lps* babe *de* ipfv *fáyn* be.fine *mí.* 1sg.indp 'I find that babe gorgeous.' [*Lit*. 'That babe is fine to me.'] [dj07ae 174]

Although there are no restrictions on the transitive use of property items, such usage is rare in non-elicited language data. There is a pronounced preference by speakers to employ other means to render causative meaning with property items.

#### 9 Grammatical relations

For instance, in the following two examples, the property items *fáyn* '(be) fine' and *blák* '(be) black' are employed as secondary predicates. Sentence (85) features a resultative causative construction, and (86) involves a resultant state resultative construction:


Labile experiential and aspectual verbs in class (b) of Table 9.5 differ from class (a) verbs in that they remain dynamic in both the intransitive and transitive alternation. However, the features of "role of subject" and "causation reading" provided in Table 9.5 co-vary in the same way with class (b) verbs as they do with class (a) verbs.

Labile experiential verbs constitute a smaller group than change-of-state verbs. I give a complete listing of experiential verbs in the corpus with glosses of intransitive meanings in (87). Experiential verbs comprise (a) body movements and processes, as well as (b) mental states denoting various types of affective conditions:


Consider the use of the group (b) dynamic experiential verb *krés* 'be crazy, drive crazy' in the following intransitive (88) and transitive (89) clauses respectively:


9.2 Verb classes

The following two sentences illustrate the use of the group (a) body movement verb *háyd* 'hide, conceal'. In both the intransitive (90) and transitive (91) clauses the imperfective marker *de* 'ipfv' is present, so experiential verbs do not exhibit the stativity alternation that characterises the other semantic classes covered so far:


The final class of labile verbs are aspectual verbs (also known as phasal verbs), i.e. verbs with largely temporal semantics, which usually occur in constructions with lexically fuller verbs. These verbs remain dynamic in transitive and intransitive clauses as well. Hence they do not alternate in their stativity value either.

Aspectual verbs serve to highlight the crossing of the left boundary (inception), the middle (continuation) or the right boundary (completion) of the situation denoted by the verb they specify. The four labile aspectual verbs of inception (a) and completion (b) found in the corpus are listed in (92):


I give an example for the intransitive and transitive uses of the verb of completion *fínis* 'finish' in the following two examples. The verbs *fínis* and *bigín* 'begin' are also employed as aspectual auxiliary verbs in completive and ingressive auxiliary constructions (cf. §6.4.1 and §6.4.3, respectively):


The discussion in this section has shown that labile verbs may be classified into five semantic and two form classes. I have also mentioned that the different semantic classes appear in their intransitive and transitive variants with differing likelihood. The factor that determines to a great part the distribution of labile verbs over the two clause types is "agentivity". On one end we find property items, change-of-state verbs denoting body

#### 9 Grammatical relations

states (e.g. *táya* 'be tired') and natural states (e.g *dráy* 'be dry'), experiential verbs denoting body processes and movements (e.g. *rɛ́s* 'rest'), mental state verbs (e.g. *gládin* 'be glad'), as well as aspectual verbs. In natural speech, these semantic (sub)classes share a higher likelihood of occuring in intransitive clauses rather than transitive ones.

In contrast, "agentive" change-of-state verbs denoting "destruction" and "material transformation" (e.g. *brék* 'break'), experiential verbs denoting physical movement (e.g. *múf* 'move'), and the entire class of locative verbs (e.g. *slíp* 'sleep, lie') generally occur with equal likelihood in both transitive and intransitive clauses.

### **9.2.4 Expletive verbs**

Expletive verbs take the dependent pronoun *e* '3sg.sbj' or a generic noun as an expletive subject. However, none of the verbs covered in the following exclusively occur with expletive subjects. The expletive subject is a core participant in syntactic terms, but it has no referential quality and appears in constructions which require the subject position to be filled.

Such dummy subject (pro)nouns are found with verbs with copula functions, with evaluative verbs, with Spanish-origin verbs which take expletive subjects in Spanish, and a weather verb (cf. §9.3.7 for a separate treatment of the weather verb *fɔ́l* '(to) rain'). All elements in the corpus which may take expletive subjects are listed in Table 9.6. The copula verbs *bí* 'be', *dé* 'be.loc' and the focus markers *cum* copulas *na/nóto* occur in copula clauses with expletive subjects.

Sentence (95) illustrates the expletive use of the locative-existential copula *dé* in the factive clause *e dé sé* 'it's that'. The second occurrence of *dé* also shows that when the copula *dé* functions as the predicate of an existential clause, the existing entity (i.e. *sɔn wích* '(some) witches') must be expressed as the subject of the clause. Hence, existential clauses featuring *dé* have no expletive subjects:

(95) *E* 3sg.sbj *dé* be.loc *sé,* qot *yu* 2sg *sabí* know *sé* qot *yá* here *só* like.that *sɔn* some *wích* sorcerer *dé* be.loc *nɔ́.* intj 'It is that, you know that here there are sorcerers, right.' [ed03sb 093]

There is no difference in meaning between the use of *dé* 'be.loc' and the identity copulas *bí* and *na*/*nóto* in factive clauses like the following two. However, contrary to other elements with expletive subjects, *na*/*nóto* may never occur with the dummy subject *e* '3sg.sbj' in factive clauses, nor be preceded by an emphatic pronoun as is the case in equative clauses (cf. §7.6.1). This is so because the focus markers/identity copulas *na*/*nóto* incorporate 3sg reference (cf. §7.6.1). The identity copula *bí* may also appear with an expletive subject in factive clauses in the place of *dé* 'be.loc' (96):

(96) *E* 3sg.sbj *fít* can *bí* be *sé* qot *na* foc *paludismo.* malaria 'It could be that it's malaria.' [fr03wt 058]

9.2 Verb classes


Table 9.6 Expletive verbs

In these functions, *bí* and *dé* are also used as introductory formulas of narratives with the meaning 'it came to pass that' (97):

(97) *E* 3sg.sbj *kán* pfv *bí* be *sé* qot *mi* 1sg.poss *abuela,* grandmother *wé* sub *a* 1sg.sbj *bin* pst *smɔ́l,* be.small *e* 3sg.sbj *gó* go *ríba* river (*…*)

'It came to pass that my grandmother, when I was small, she went to the river (…)' [ed03sb 015]

Factive clauses can alternatively be formed with the help of two semantically empty dummy nouns, the generic noun *tín* 'thing' and the noun *kés* 'matter':


#### 9 Grammatical relations

Also compare the cleft focus construction in (100) featuring *di tín* 'the thing' with the functionally equivalent construction in (101), featuring the expletive pronoun *e*:


When the verb *fíba* 'resemble' occurs in a transitive clause, the 3sg.sbj pronoun is not expletive (102):

(102) *E* 3sg.sbj *fíba* seem *dɛ́bul.* devil 'He resembles a devil.' [ra07fn 072]

In contrast, when used intransitively, *fíba* is best translated as 'seem' (103) and may take a complement clause (104). In such contexts, *fíba* also takes an expletive 3sg.sbj pronoun:


The verb *lɛ́f* 'leave, remain' occurs as a copula verb with an expletive subject in clauses like the following one:

(105) *E* 3sg.sbj *lɛ́f* leave *wán* one *pɔ́sin* person 'There is one person remaining.'

The verb *sté* 'stay, last (a long time)' also functions as a copula element in intransitive clauses (106). Both verbs occur with expletive *e* in their copula function. Also consider *níd* 'need, be necessary' (107):

(106) *E* 3sg.sbj *dɔ́n* prf *sté,* be.long *a* 1sg.sbj *tínk* think *sé* qot *e* 3sg.sbj *dɔ́n* prf *sté* last *wé* sub *una* 2pl *bin* pst *gɛ́t* get *insecticida* insecticide *yá.* here 'It's been long, I think it's been long since you [pl] have had insecticide [sprayed] here.' [fr03wt 059]

9.2 Verb classes

(107) *E* 3sg.sbj *níd* need *sé* qot *mék* sbjv *a* 1sg.sbj *gó* go *dé* there *tumɔ́ro.* tomorrow 'It is necessary that I go there tomorrow.' [dj07ae 512]

Evaluative verbs also take expletive subjects. Examples follow with the property items *hád* '(be) hard' (108) and *fáyn* '(be) fine' (109):


The two verbs *falta* 'lack' (< Sp. *faltar* 'lack') and *sigue* 'continue, follow' (< Sp. *seguir* 'continue, follow') are established loans of Spanish origin which have been borrowed together with their selectional properties. Like their Spanish etymons these two verbs require expletive subjects. Contrary to Spanish, subject pronouns are not normally dropped in Pichi. Examples (110) and (111) therefore feature the expletive pronoun *e* '3sg.sbj':


Yet *falta* and *sigue* may also take referentially full subjects in intransitive clauses. Compare *falta* in (110) and (112), and *sigue* in (111) and (113):


At this point, a word is in order on the raising properties of expletive verbs. In (114) below, the 3sg.sbj pronoun *e* anaphorically refers to *dɛn yón* 'theirs', so the object of the complement clause beginning with the non-finite complementiser *fɔ* 'prep' has been raised into subject position in the main clause. For other speakers, however, raising is not accepted with evaluative verbs (115):

#### 9 Grammatical relations


The verb *fíba* 'resemble' takes full complement clauses introduced by the complementiser *sé* (116). Neither reduced *fɔ*-complement clauses nor SVCs are accepted in clause linkage. Hence, an SVC like (117) is ungrammatical:


However, the subject of the complement clause may be raised into subject position of the main clause without any structural change. The result is an idiosyncratic structure, in which the coreferential subjects of the main and subordinate clauses are both overtly expressed (118):

(118) *Boyé* name *fíba* seem *sé* qot *e* 3sg.sbj *gɛ́t* get *mɔní.* money 'Boyé seems to have money.' [*Lit.* 'Boyé seems that he has money.'] [dj07ae 256]

Existential constructions featuring expletive subjects and the verb *gɛ́t* 'get, exist' are covered in detail in section §7.6.3.

### **9.3 Valency**

I now turn to describing valency in select types of Pichi constructions. I cover the grammatical relations mediated by verbs in these constructions as well as the semantic roles assigned to the core participants subject and object. I also treat valency in two semantic fields, namely in the expression of weather phenomena and body states. These fields are of particular interest due to the variety of valency configurations found in the clauses used to express them.

9.3 Valency

### **9.3.1 Light verb constructions**

Pichi features numerous more or less conventionalised collocations that involve verbs with a fairly general meaning followed by undergoer objects. Many of these collocations appear to be light verb constructions, in which the bulk of semantic content is carried by the object rather than the verb. The most common of these light verb constructions are provided in Table 9.7. The most common constructions involve the verbs *gɛ́t* 'get, have', *gí* 'give', *mék* 'make' and *púl* 'pull, remove'. At the lower end of the table, we find constructions involving verbs which are only found in a single collocation.

A good number of the constructions listed above constitute borderline cases between ordinary verb-noun collocations assembled by phrasal syntax and conventionalised or lexicalised verb-noun collocations. Two criteria may be useful in determining which of these construction are conventionalised to the point of qualifying as light verb constructions. Firstly, the object in more conventionalised collocations has a tendency to occur bare. Secondly, there is a relatively stringent restriction on pronominalising light verb objects. Some salient characteristics of light verb constructions are explored in the following by means of constructions involving the verb *gɛ́t* 'get, have'.

The verb *gɛ́t* is an inchoative-stative transitive verb, which occurs with a stative (119), and at other times, an inchoative reading (120). The verb also has various functions as an existential and modal verb and expresses possession (cf. e.g. 121 below).

	- 'He got (himself) another woman, many problems, (…)' [ma03ni 025]

Ordinary objects of *gɛ́t* may occur bare or be preceded by determiners depending on pragmatic circumstance. In (121), the non-specific noun *bɔ́y* 'boy' is preceded by the indefinite determiner wán. In contrast, non-specific objects of light verbs have a strong tendency to occur devoid of any definiteness marking. In (121), the noun *mán* 'man' of the light verb construction *gɛ́t mán* 'have a man/husband' remains bare:

(121) *Smɔ́l* small *gál,* girl *ɛf* if *e* 3sg.sbj *nó* neg *gɛ́t* get *wán* one *bɔ́y,* boy *e* 3sg.sbj *nó* neg *gɛ́t* get *mán,* man *pero* but *di* def *húman,* woman *di* def *bíg* big *wán* one *dɛn* pl *sɛ́f.* emp

'As for young girls, if they don't have a boy-friend, if they don't have a man [they feel worthless], but even women, the grown ones themselves.' [hi03cb 154]

Objects of *gɛ́t* may be pluralised with a post-posed *dɛn* (122) and may occur with prenominal modifiers like *bɔkú* 'much' or the 1sg possessive pronoun *mi* (122). The verb *gɛ́t* may also take pronominal objects (123), or occur with no overt object at all where reference has been established earlier on (124).

#### 9 Grammatical relations


Table 9.7 Light verb constructions

9.3 Valency

	- '(…) maybe I have children or maybe I don't have [children].' [hi03cb 158]

We have seen that non-specific objects of light verbs tend to occur as bare nouns. Nonetheless, specific objects of *gɛ́t* in light verb constructions may occur with determiners if so required. Compare *di fɔ́s bɛlɛ́* 'first pregnancy' in (125):

(125) *Dásɔl* only *wé* sub *a* 1sg.sbj *dɔ́n* prf *bíg* big *wé* sub *a* 1sg.sbj *fɔ́s* first *gɛ́t* get *di* def *fɔ́s* first *bɛlɛ́* belly (*…*) 'Then when I was grown, when I first had the first pregnancy (…)' [ed03sb 017]

The NP *di fɔ́s bɛlɛ́* the first pregnancy' in (125) above also shows that objects of light verbs are encountered with prenominal modifiers. Likewise, object NPs in light verb constructions may be placed under focus (126). Although there are not many instances of pronominalised light verb objects in the data, these also occur. In (127), the object pronoun *=an* substitutes for *torí* 'story':


Table 9.8 presents a frequency analysis of *gɛ́t* in verb-object collocations in a subcorpus of 30,000 words. The verb *gɛ́t* enjoys a total number of 345 tokens, of which 136 tokens involve *gɛ́t* as a modal verb and *gɛ́t* without an overt object. In line (a) of the table, I give the remaining 209 tokens which represent uses of *gɛ́t* in collocations involving full noun objects.

In line (b), I provide the total number of verb-noun collocations that do not qualify as light verb constructions according to the distributional criteria introduced above. Line (c) gives the total number of constructions that should be considered light verb constructions. I also list the four most frequent constructions with the corresponding tokens. I take care to distinguish cases in which a collocation like *gɛ́t pikín* is employed with the

#### 9 Grammatical relations


Table 9.8 Frequency of *gɛ́t* collocations

general meaning of 'have children' from ones in which the collocation is used with a specific meaning like 'have one, two, etc. children'. The corresponding percentages in relation to the total number of collocations in line (a) are given in the rightmost column.

Table 9.8 reveals that light verb constructions proper represent 33 per cent of the total number of occurrences of collocations involving *gɛ́t* and an object. Of all *gɛ́t*-constructions contained in the corpus, *gɛ́t pikín* 'have children' is the most frequent one and accounts for 16 per cent of the total of light verb constructions. An additional information of interest is that the total number of types (different constructions) of light verb construction amounts to eleven, seven of which occur only once each. In view of these facts, I assume that the functional load of *gɛ́t* as a light verb is only moderate.

Next to the borrowing of Spanish verbs, verb-noun collocations consisting of a Pichi verb and a Spanish noun are an important means of extending the lexicon. codemixed constructions allow speakers to tap into the nominal lexicon of Spanish in order to derive new "verbal" meanings. These constructions are characterised by a high degree of structural equivalence between Pichi and Spanish. Not only is the order of constituents in verb-noun collocations the same in both languages, the meanings of the light verbs employed in the respective languages are also highly compatible with each other. There is therefore a strong tendency towards convergence in codemixed collocations. Accordingly, the verbs in these collocations may have the selectional characteristics of the Pichi verb in one instance, while in another, the Pichi verb may select its complement as if it were the synonymous Spanish verb (cf. Muysken 2000: 184)

For instance, none of the nouns in the collocations *gɛ́t rabia* 'be angry' (128), *gɛ́t novio* 'have a boyfriend' (129) and *gí permiso* 'give permission' (130) are encountered with a determiner in the corpus. The meanings of the verbs and the distribution of nouns in these constructions are identical to those in the Spanish equivalents *tener rabia* 'be angry', *tener novio* 'have a boyfriend', and *dar permiso* 'give permission'.

9.3 Valency


On the other hand, there are established mixed collocations which feature a determiner. One of these is *gí wán vuelta* 'take a walk'. Like the Pichi verb *gí* 'give', the Spanish verb *dar* 'give' selects a determined object in the expression *dar una vuelta* 'take a walk':

(131) *E* 3sg.sbj *de* ipfv *gí* give *wán* one *vuelta* round *kwík* quickly 'She's doing a round quickly.' [dj05be 120]

Other codemixed collocations are further removed from the pole of light verb constructions. The collocation *gí beca* 'give a scholarship' (132–133) occurs with or without determiners in accordance with the referential properties of the NP:


In sum, we can observe that next to a few "proper" light verb constructions, Pichi makes use of less tightly integrated collocations featuring Pichi or Spanish nouns by means of ordinary phrasal syntax. These constructions are flexible, allow the insertion of functional elements and modifiers, as well as object substitution by pronouns.

### **9.3.2 Associative objects**

In Pichi, syntactic objects can denote various less central semantic roles which may alternatively be expressed through prepositional phrases. Accordingly, associative objects appear to the right of patient objects in double-object constructions (cf. §9.3.4), in a position usually reserved for adverbial adjuncts. An associative object is an instantiation

#### 9 Grammatical relations

of some entity typically associated with the situation denoted by the verb. Associative objects in Pichi are reminiscent of inherent object constructions as found in the Kwa languages of West Africa (see Essegbey 1999 for Ewe). Contrary to inherent objects, however, associative objects are not obligatory and may remain unexpressed at all times. Equally, associative objects usually only occur with specific verbs (cf. e.g. 141). The verbobject collocations described in this section therefore appear to involve specialisation or lexicalisation. The use of associative objects can therefore only serve as a productive means of increasing verb valency with the verbs listed in Table 9.9.

Here follows an example with the verb *wás* 'wash (oneself)' and its associative object *wɔtá* 'water'. The pragmatic context coerces a semantic role of instrument or means on the associative object:

(134) *A* 1sg.sbj *de* ipfv *wás* wash *wɔtá.* water 'I'm washing (myself with) water.' [dj07ae 274]

All verb-noun collocations involving associative objects in the corpus are listed in Table 9.9. In most cases, the verb-noun combination given in the table is the preferred means of expressing the corresponding semantic relation between the verb and object listed.

Associative objects are assigned a content role by the labile change-of-state verbs*fúlɔp* 'fill up' (135) and *pák* 'pack, fill up' (136):


Content objects can be replaced by a corresponding prepositional phrase without a change in meaning. Compare the PPs introduced by *wet* 'with' in (137) and (138):


Instrument is among the most common semantic roles expressed by associative objects (139). The instrument role may also be expressed by a *wet*-prepositional phrase (140):


Table 9.9 Associative objects


#### 9 Grammatical relations

It is noteworthy that many other verbs that assign an instrument role to a participant do not seem to take instrument associative objects; for example, *kɔ́t* 'cut' is not attested with an associative object and requires the instrument to be expressed as a prepositional phrase:


Sentences (143) and (144) provide examples for the use of associative objects with the semantic role of purpose. These may equally be expressed through a prepositional phrase introduced by the associative preposition *fɔ* (144):


The source of the motion verb *fɔdɔ́n* 'fall' may be realised as an associative object (145). Alternatively, the source may be indicated via the preposition *frɔn* 'from' when it marks the ground (146). Note the possibility of additionally using the "at rest" locative noun *ɔ́p* 'up(perside)' to mark the ground in (146):


The semantic role of the objects of *smɛ́l* 'smell' can only be disambiguated by context. In (147) the associative object *chɔ́p* 'food' denotes the source of the sensation, in (148), *chɔ́p* denotes the stimulus:

(147) *E* 3sg.sbj *de* ipfv *kúk,* cook *áfta* then *e* 3sg.sbj *de* ipfv *smɛ́l* smell *chɔ́p.* food 'He's cooking, afterwards he'll smell of food.' [dj07ae 013]

9.3 Valency

(148) *Yu* 2sg *de* ipfv *smɛ́l* smell *chɔ́p,* food *dɛn* 3pl *de* ipfv *fráy* fry *ɛ́ks* egg *dé.* there 'You smell food, they're frying eggs there.' [dj07ae 016]

Non-locative goal is the semantic role of objects associated with the verbs *kɔmɔ́t* 'come out' and *tɔ́n* 'turn (into)'.

(149) *E* 3sg.sbj *de* ipfv *trén* train *yú* 2sg.indp *sé* qot *yu* 2sg *go* pot *kɔmɔ́t* come.out *pɔ́sin.* person 'She is bringing you up to become a responsible person.' [au07se 131]

(150) *E* 3sg.sbj *tɔ́n* turn *pɔ́sin* person *wán* one *tɛ́n.* time 'He turned into a human-being at once.' [ma03sh 006]

The objects of *sík* 'be sick' denote the cause of the sickness that the subject is suffering from. The verb *sík* is not attested with a prepositional phrase alternative in the data; the use of an associative object appears to be the conventional way of expressing this state of affairs:

(151) *E* 3sg.sbj *de* ipfv *sík* sick *fíba.* fever 'She's sick with fever.' [dj07ae 273]

Another instance of an associative object with the semantic role of cause is *mɔní* 'money', the object of *kráy* 'cry' in (152):

(152) (*…*) *dán* that *papá* father *de* ipfv *kráy* cry *in* 3sg.poss *mɔní.* money.

'(…) that elderly man was crying over his (lost) money.' [ed03sb 200]

An associative object may be fronted for emphasis (153). However, unlike patient or beneficiary objects, associative objects may not be questioned with *wétin* 'what' or *údat* 'who'. Instead, associative objects must be questioned with the corresponding adverbial question phrase or with the selective question element *ús=káyn* 'which', which questions modifiers.

Hence the clause *e de sík fíba* '3sg.sbj ipfv be sick fever' = 'she's sick (with) fever' cannot be questioned as \**wétin e de sík* 'what 3sg.sbj ipfv be.sick' = 'what is she sick (with)?' Rather, the question must be phrased as in (154):


#### 9 Grammatical relations

### **9.3.3 Cognate objects**

In Pichi, "cognate objects" (Baron 1971) are deverbal nouns derived from themselves. Firstly, deverbal nouns occur with a few particular verbs in a non-emphatic, non-specific context and contribute little if nothing at all to the meaning denoted by the verb.

For example, the objects of *sík* 'be sick' and verbs of sound and speech-emission like *síng* 'sing' and *tɔ́k* 'talk, say' may occur with speech- or sound-denoting cognate objects in non-emphatic contexts. The cognate objects of these verbs have in common that they are not simply the corresponding action nominal of the verb. Instead, they have slightly idiosyncratic meanings:


'I want to say that small word there.' [dj05ae 037]

Aside from that, the use of cognate objects provides an important means of expressing emphasis in pragmatically marked, emphatic contexts such as (157–158). Emphatic cognate objects are frequently preceded by the indefinite determiner *wán* 'one, a' which provides emphasis in other contexts as well (e.g. in the context of negative indefinite phrases, cf. 63):


The cognate object *dáy* 'death' also appears as a cognate object to the verb *dáy* 'die' in emphatic contexts like (159):

(159) *Éy,* intj *dán* that *káyn* kind *spɛ́tikul,* glasses *a* 1sg.sbj *dáy* die *dáy.* death 'Hey, [if I had] that kind of glasses, I would surely die.' [ne07ga 015]

There is good reason to assume that the fronted "verb" in predicate cleft constructions like the following one is also a deverbal noun. One indication for this is that the verb is never fronted with predicate constituents like TMA markers. In this view, clefted verbs may also be seen as types of cognate objects:

9.3 Valency

(160) *Na* foc *gó* go *a* 1sg.sbj *de* ipfv *gó* go *ó!* sp '[Mind you] I'm going now!' [ch07fn 151]

### **9.3.4 Double-object constructions**

The bulk of Pichi verbs can occur with one as well as two objects. The primacy of the object next to the verb – which is usually animate and has the role of recipient or beneficiary – is evident in double-object constructions involving two object pronouns. The presence of two pronominal objects is ungrammatical if the clitic object pronoun *=an* is preceded by the low-toned personal pronoun *una/unu* '2pl' or another 3sg.obj pronoun *=an* (for details, see §3.3, on tone-conditioned suppletive allomorphy). In such cases, it is the patient object that remains unexpressed. Compare the double-object construction in (161) with the ungrammatical example (162) and sentence (163). In the latter example, the 3sg.obj theme object *=an* remains unexpressed:


Double-object constructions can be divided into three types according to relevant semantic and syntactic properties. Table 9.10 provides an overview of the semantic roles


Table 9.10 Syntax and semantics of double-object constructions

#### 9 Grammatical relations

of objects involved in double-object constructions and their syntactic positions as primary objects immediately to the right of the verb or secondary objects following the primary objects. Some semantic roles associated with the position of primary and secondary objects may alternatively be expressed by prepositional phrases or SVCs. Where such alternatives exist, they are provided in the two rightmost columns.

In the type 1 double-object construction, the primary object to the right of the verb occupies the recipient role, while the secondary object that follows the recipient invariably takes on a patient role. This kind of construction is found with verbs expressing the transfer of an entity or an act of communication from the subject to a recipient. All ditransitive communication and transfer verbs encountered in the corpus are listed in (164):


Pichi has no SVCs of the give type in order to mark a recipient or beneficiary. In doubleobject constructions featuring transfer verbs, the primary object next to the verb always has the semantic role of recipient (165). With transfer and communication verbs, a beneficiary is usually expressed in a PP introduced by *fɔ* 'prep' (166). Hence, double-object constructions are the only means of expressing the grammatical relation between the ditransitive verb, its subject, and its recipient and theme objects:


The following example features transfer verb *gí* 'give' and a prepositional phrase introduced by *fɔ* 'prep'. The PP can only denote a beneficiary with the recipient remaining unexpressed. Hence the second translation is ungrammatical, since the recipient object cannot alternatively be expressed as a prepositional phrase:

(167) *Dɛn* 3pl *gí* give *di* def *mɔní* money *fɔ* prep *mí.* 1sg.indp 'They gave the money (to someone) for me.' [lo07fn 555] but not 'They gave me the money.'

9.3 Valency

The following two double-object constructions involve the transfer verb *gí* 'give' (168) and the verb of communication *púl (torí)* 'narrate (a story)' (169):


The verb *sɛ́n* 'send, throw' denotes a situation in which both a transfer and a motion event co-occur. When *sɛ́n* is used in a double-object construction, the primary object is always a recipient (170).

(170) *E* 3sg.sbj *gɛ́fɔ* have.to *sɛ́n* send *mí=an.* 1sg.indp=3sg.obj 'He has to send/throw it to me.' [ye07de 001]

Like with other transfer verbs, the recipient of *sɛ́n* may not be expressed as a prepositional phrase. Where we do find a prepositional phrase (usually introduced by *fɔ* 'prep'), it can therefore only denote a beneficiary or a goal but not a recipient (171):

(171) *E* 3sg.sbj *gɛ́fɔ* have.to *sɛ́n=an* send=3sg.obj *fɔ* prep *yú.* 2sg.indp

> 'He has to send it to (where) you (are).' or 'He has to send it for you.' [ye07de 003] but not 'He has to send it to you.'

Type 2 double-object constructions are best understood in terms of syntactic promotion. A participant that is more commonly expressed as a prepositional phrase is promoted to object status. In contrast to type 1, the use of type 2 constructions is therefore optional. We find the type 2 double-object construction with two kinds of verbs. First, it is encountered with any Pichi transitive verb save transfer verbs and verbs of communication (type 1). With these verbs, which form the vast majority of Pichi verbs, the primary object has the semantic role of beneficiary. The secondary object is assigned a patient role.

Sentence (172) features two type 2 double-object constructions. The verb *dú* 'do' takes the primary, beneficiary object *mí* '1sg.indp' and the patient object *sɔn fébɔ* 'a favour'. The verb *wás* 'wash' also takes *mí* '1sg.indp' as the beneficiary object while *klós dɛn* 'clothing' functions as the patient object:

(172) *A* 1sg.sbj *wánt* want *mék* sbjv *yu* 2sg *dú* do *mí* 1sg.indp *sɔn* some *fébɔ* favour *mék* sbjv *yu* 2sg *wás* wash *mí* 1sg.indp *sɔn* some *klós* clothing *dɛn.* pl

'I want you to do me a favour (and) wash some clothes for me.' [ru03wt 030]

#### 9 Grammatical relations

The semantic role of beneficiary may subsume a maleficiary, i.e. the affected party of a socially unacceptable action. In (173), a worried mother explains why she has left her teenage daughter in Spain instead of bringing her along with her to Malabo. Also, compare the first object of *tíf* 'steal', the maleficiary *mí* '1sg.indp' in (174):


We have seen that a recipient must be expressed as an object in type 1 double-object constructions. In contrast to type 1 constructions, type 2 constructions alternate freely with constructions in which the beneficiary is expressed as a prepositional phrase introduced by the associative preposition *fɔ* 'prep'. In fact, the alternative involving a prepositional phrase is more common than the corresponding double-object construction. Compare the type 2 double-object construction (175) involving the verb *bay* 'buy' with the PP alternative (176):


The second type of type 2 construction involves the caused location verb *pút* 'put'. Here, the primary object has the semantic role of goal, while the secondary object fulfills a theme role. In (177), the primary object of *pút* is the goal object *=an* '3sg.obj', while the secondary object *saldo* '(mobile phone) credit' is the theme. Sentence (178) also features the goal object *=an* '3sg.obj', while the theme object is *cacahuete* 'groundnut':


9.3 Valency

However, the corpus contains many more examples of *pút*-constructions, in which the goal role is expressed through a locative construction rather than a primary object. Likewise, there is no sentence in the data in which the goal object of *pút* is a full noun. The locative construction may be a PP (179) or involve a locative noun (180). Unlike a few other verbs with a motion component (cf. §8.1.4), the goal of *pút* cannot be expressed as a complement of the V2 of a motion-direction SVC (e.g. *\*a pút=an gó na glás* {1sg.sbj put=3sg.obj go loc glass} = 'I put it into the glass'):


There is a preference to interpret a PP introduced by *fɔ* 'prep' as a beneficiary in *pút*double-object constructions, particularly where an object pronoun theoretically allows for both interpretations as in (181). A sentence like (180) above, which involves a locative noun (i.e. *ínsay* 'inside') is therefore preferred to avoid ambiguity. Nevertheless, an alternative with a prepositional phrase involving the general locative preposition *na* may also be exploited to the same end (182):


Note, however, that *pút* 'put' may also appear in a type 2 double-object construction, in which the primary object is a beneficiary – just like any other transitive verb:

(183) *Yu* 2sg *pút=an* put=3sg.obj *wán* one *sardina* sardine *ɔntɔ́p.* top '(Then) you put a sardine on top for him.' [ro05rt 064

Type 3 double-object construction involve verbs that may take associative objects (cf. Table 9.9 above). Type 3 constructions differ from type 1 and type 2 constructions in that the primary object occupies the semantic role of patient. The secondary object is an associative object which may alternatively be expressed without any syntactic rearrangement through the mere insertion of a preposition, serial verb, or any other element

#### 9 Grammatical relations

between the two objects. The associative objects in type 3 constructions may therefore be paraphrased with the same means as associative objects in single-object constructions. Compare the double-object construction in (184) with the single-object construction involving a PP in (185):


### **9.3.5 Reflexivity**

In the majority of cases, reflexivity is expressed through an object NP consisting of the pronominal and reflexive anaphor *sɛ́f* 'self' and a preceding possessive pronoun with the same person and number as the subject. Sometimes, the body part nouns *skín* 'body', *bɔ́di* 'body', and *héd* 'head' are also employed as reflexive anaphors in the same syntactic position as *sɛ́f*. A clause featuring a reflexive object NP indicates that the subject does something to her- or himself. The corpus only contains clauses in which subjects serve as antecedents to the reflexive anaphor, cf. (186):

(186) *Dán* that *gál* girl *e* 3sg.sbj *kin* hab *fíks* fix *in* 3sg.poss *sɛ́f,* self *pént* paint *in* 3sg.poss *sɛ́f.* self 'That girl habitually fixes herself up, paints herself [puts on make up].' [dj07ae 114]

Aside from that, reflexive constructions also form part of idiomatic expressions with little reflexive meaning but characterised by low transitivity. I give a sentence featuring the idiom *sék in sɛ́f* 'shake 3sg.poss self' = 'make an effort':

(187) *E* 3sg.sbj *sék* shake *in* 3sg.poss *sɛ́f* self *bɔkú* much *fɔ* prep *tɔ́n* turn *general.* general 'He made a big effort to turn general.' [ur07ae 498]

The nouns *skín* 'body', *bɔ́di* 'body', and *héd* 'head' are far less commonly used than *sɛ́f* as reflexive anaphors. Equally, these three nouns usually occur as reflexive anaphors with verbs, whose meanings imply an actual physical effect on the body. The following three sentences illustrate this usage:

(188) *A* 1sg.sbj *de* ipfv *sí* see *mi* 1sg.poss *skín* body *na* loc *lukinglás.* mirror 'I'm seeing myself/my body in the mirror.' [dj07ae 496]

9.3 Valency


A reflexive relation within an NP is expressed through the use of a possessive pronoun in conjunction with the pronominal *yón* 'own' as a modifier to a head noun:

(191) *Bɔt* but *fɔ* prep *Bata* place *dɛn* 3pl *de* ipfv *ték/* take *dán* that *wán* one *sí* if *que* that *dɛn* 3pl *yón* own *máred* marriage *día,* be.expensive *dɛn* 3pl *yón* own *máred* marriage *de* ipfv *kári* carry *mɔní* money *ɛ́n.* intj 'But as for the mainlanders, they take/ as for that one, their marriage is expensive, their marriage costs money.' [hi03cb 010]

Besides that, Pichi has a number of inherently reflexive verbs. For most of these verbs, the use of a reflexive anaphor is optional. Such verbs denote situations involving body or mental processes and physical movements which imply volition and instigation by the actor subject rather than a spontaneous event.

Compare *wɛ́r* 'dress (up)' in an explicitly reflexive clause (192) and a clause in which reflexivity remains unexpressed (193):


### **9.3.6 Reciprocity**

Next to its use as a reflexive anaphor, the pronominal *sɛ́f* 'self' also serves as a reciprocal pronominal with plural referents. In sentence (194), the reciprocal NP is an object to the verb *sláp* 'slap', in (195) to the locative noun *bifó* 'before':


#### 9 Grammatical relations

Reflexive and reciprocal meaning may be disambiguated through contextual factors, i.e. verb meaning and the presence of plural referents. The occurrence of compound personal pronouns indicating dual number (i.e. *dɛn-ɔl-tú* 'the two of them') as in (196) or universal inclusivity (*dɛn-ɔ́l* 'all of them') as in (197) is also quite common in reciprocal contexts:


Reciprocal relations within the NP find expression through the pronominal *yón* as illustrated in (198):

(198) *Dɛn* 3pl *lúk* look *dɛn* 3pl *yón* own *fés.* face 'They looked at thier own faces.' [eb07fn 313]

Pichi also has inherently reciprocal verbs, many of which preferably do not occur with the anaphor *sɛ́f* (cf. §9.4.3).

### **9.3.7 Weather phenomena**

Pichi has three types of constructions for expressing weather phenomena. The first type of construction consists of an intransitive clause with the weather phenomenon in the subject position. The verbs used in the first type of construction have a general meaning and also occur in other contexts including transitive clauses. Three sentences follow featuring the two weather verbs *bló* 'blow' (199–200) and *bráyt* 'be bright' (201) and the weather nouns *tináda* 'thunderstorm', *brís* 'air', and *sán* 'sun':


9.3 Valency

Sentence (202) exemplifies the transitive usage of *bló* 'blow', sentence (203) that of *bráyt*, here with the meaning 'brighten, light up':


In expressions where reference is made to the general atmospheric condition, the noun *dé* 'day, weather' appears in the subject position instead of a specific natural element. This usage is exemplified in the following three sentences and also in (203) above:


The second type of construction also involves an intransitive clause but it features the expletive subject pronoun *e* '3sg.sbj' rather than a weather noun. This construction is limited to a single intransitive verb, namely *fɔ́l* 'rain', which exclusively functions as a weather verb (207). The verb *fɔ́l* may, however, also occur in the first type of construction, together with the weather noun *rén* 'rain' in subject position (208):


The third type of construction involves existential clauses featuring the possessive and existential verb *gɛ́t* 'get, exist' or the locative-existential copula *dé* 'be.loc' (cf. §7.6.3 for details on the syntax of these clauses). This construction is only attested in codemixed utterances involving a Spanish atmospheric phenomenon:

9 Grammatical relations


### **9.3.8 Body states**

Body states are expressed in constructions involving transitive (cf. 1a–1c in Table 9.11) and intransitive (2a–2c) clauses. Type 1 constructions in the table involve transitive clauses. In type 1a constructions, the affected body part is found in the subject position, while the experiencer is in the object position. This construction is the preferred one for expressing pain and hurt. The verb is either of the dynamic experiential verbs *hát* 'hurt' or *pén* 'pain' (211–213).


Table 9.11 Expressing body states

(211) *Mi* 1sg.poss *bɛlɛ́* belly *de* ipfv *hát* hurt *mí.* 1sg.indp 'My stomach is hurting me.' [dj07ae 312]

9.3 Valency


In type 1b constructions, the subject of the transitive clause is a deverbal noun denoting the experience, while the object instantiates the experiencer. Instead of an experiential verb, we find an idiomatically used dynamic verb *kéch* 'catch'. The body states of hunger, thirst, and sleep(iness) may be expressed in this way, usually combined with a sense of suddenness or unexpectedness. Compare the following three examples:


Type 1c constructions in Table 9.11 are the mirror-image of type 1b constructions. The experiencer is in the subject position, while the body state or sensation is expressed as a deverbal noun in the object position. Hunger, thirst, and sleep(iness) can be expressed by this construction with the dynamic body state verbs *fíl* 'feel' (217–218). Hunger and thirst can also be expressed in combination with the verb *sɔ́fa* 'suffer, endure' (219):


#### 9 Grammatical relations

Proof for the nominal status of the body state in the constructions above is provided by sentences (218) and (219). In the former example, we find a dislocated possessive construction in the object position of *fíl*. In the latter example, the indefinite determiner *wán* 'one, a' precedes *hángri* 'hunger', the object of *sɔ́fa* 'endure'.

The type 1c construction also serves to express the body states 'feel hot' and 'feel cold'. Compare the following two examples:


Type 2 constructions involve intransitive clauses. In type 2a, the experiencer appears in the subject position. The body state is instantiated in a dynamic verb. Once more, the basic body states of hunger and thirst can be expressed in this way (222–223). However, other transient body states like *sík* 'be sick' also appear in this construction (224):


In type 2b constructions, the body state verb is inchoative-stative. Compare *táya* 'be tired' (225) *wɛ́l* 'be well' (226), *bɛlfúl* 'be satiated' (227), and *hát* 'be hurt' (228):


9.4 Valency adjustments

(228) *Di* def *gál* girl *hát.* hurt 'The girl is hurt.' [dj05be 006]

Type 2c constructions are intransitive copula clauses. The body state verb appears as an adjective complement to the locative-existential copula *dé* 'be.loc' (229).The property items *gúd* 'be well', *bád* 'be bad', and *fáyn* 'be fine' appear in such predicate adjective constructions when they express a transient body state rather than an (intrinsic) value (cf. §7.6.5):

(229) *Dán* that *tɛ́n* time *a* 1sg.sbj *dé* be.loc *fáyn.* fine 'That time, I was fine.' [ru03wt 024]

The two body state expressions *sík* 'be sick' and *bɛlfúl* 'be satiated' may also appear in transitive clauses involving associative objects (cf. §9.3.2):


### **9.4 Valency adjustments**

Verb valency is adjusted in three ways. For one part, the omission of the core participant subject (cf. §9.4.1) or object (cf. §9.4.2) reduces verb valency by one. Object omission is also at play when reflexive and reciprocal object pronouns remain unexpressed (cf. §9.4.3). Second, a notional patient object may be added to a clause by employing a causative construction (cf. §9.4.4). Causative constructions involve biclausal structures and secondary predication. They are therefore a means of increasing valency periphrastically.

Thirdly, an agent can be backgrounded, though not wholly removed, by employing as the subject the 3pl dependent pronoun *dɛn* or a generic human-denoting noun with impersonal reference (cf. §9.4.5). In that, agent backgrounding is functionally similar to passive voice in other languages.

### **9.4.1 Unexpressed subjects**

Subjects are normally expressed overtly but subject omission (indicated by ∅) occasionally occurs with verbs with impersonal reference, as with *fít* 'can' in an excerpt from a procedural text (232):

#### 9 Grammatical relations

(232) ∅ 2sg *fít* can *sifta* sieve *ín* 3sg.indp *sóté* until *tú* two *tɛ́n* time *mék* make *mék* sbjv *dán* that *smɔ́l* small *smɔ́l* rep *watá* water *dɛn* pl *nó* neg *lɛ́f.*

remain

'(You) can sift it up to two times to make none of that little water remain.' [dj03do 008]

In another context, we find something similar to subject omission. The quotative marker *sé* may appear at the beginning of an independent prosodic unit, rather than within a prosodically integrated sentence. In such contexts, the element *sé* straddles the boundary of a verbal meaning 'say' and its function as a quotative marker and introducer of direct discourse. Hence, the "absence" of a subject may be seen as a form of omission (cf. also §10.4).

The following two sentences are uttered in sequence by the same speaker. Compare the ambiguous function of *sé* in (233b), which is introduced by *sé*, with (a) where *sé* is firmly integrated into the sentence as a quotative marker:

(233) a. *E* 3sg.sbj *tɛ́l=an* tell=3sg.obj *sé* qot *"papá* father *mí* 1sg.indp *nɛ́va* neg.prf *chɔ́p* eat *mi* 1sg.indp *sénwe".* emp 'He told him "please, I myself haven't eaten yet".' [ye03cd 149] b. *Sé chico, dí tín nó go dú mí.*

> qot boy this thing neg pot do 1sg.indp '(He said) "man, this won't do for me".' [ye03cd 150]

A final form of subject omission occurs when the particles *na* 'foc' and *nóto* 'neg.foc' incorporate 3sg reference by default in their function as identity copulas. When pronominal reference is to be overtly established, *na*/*nóto* must be preceded by independent (emphatic) personal pronoun (cf. also §7.6.1). Dependent pronouns may not precede these two particles.

### **9.4.2 Unexpressed objects**

In principle, objects need not be overtly expressed. In practice, highly transitive verbs are unlikely to appear without a patient object, even if the object is non-specific. The verb *bló* 'give a blow' in (234) denotes a situation which implies a high degree of volition and instigation by an agent. Equally, the situation involves no notion of affectedness of the agent (cf. Næss 2007):

(234) *A* 1sg.sbj *bló* give.blow *dí* this *pikín.* child 'I gave this child [guy] a blow.' [dj07ae 031]

When *bló* occurs without an object it is understood to be the homophonous *bló* 'rest, relax' (235), a verb which is lower on the transitivity scale, but may also be used transitively (236), due to its status as a labile experiential verb:

9.4 Valency adjustments


When highly transitive verbs are used in a context of non-specificity, they usually occur with generic nouns as objects. Compare the non-specific object *sɔn tín* 'something' of the highly transitive verb *híb* 'throw (away)' (237) and *pɔ́sin* 'person', object of *nák* 'hit' (238):

(237) (*…*) *yu* 2sg *híb* throw *sɔn* some *tín* thing *fɔ* prep *grɔ́n* ground (*…*)

'(…) (if) you throw something on the ground (…)' [hi03cb 028]

(238) (*…*) *na* foc *ín* 3sg.indp *e* 3sg.sbj *de* ipfv *nák* hit *pɔ́sin.* person

'(…) that's why she's hitting somebody.' [au07se 191]

The omission of objects is more common with verbs characterised by a lower degree of semantic transitivity, in particular where the objects are non-specific. Object omission is therefore principally found with "effected-object verbs" (Hopper 1985) and "affectedagent verbs" (Tenny 1994; Næss 2007).

The objects of effected-object verbs come into existence through the situation denoted by the verb. They are not affected or changed by the situation denoted by the verb like the patient objects of more prototypically transitive verbs. The non-specific effected objects of verbs of speech and sound emission often occur without a speech- and sound-denoting noun or pronoun. Consider the following use of *tɔ́k* 'say, talk' in a transitive (239) and in an intransitive clause (240):


Another verb that may be used in this way is *síng* 'síng' (241):

(241) *E* 3sg.sbj *de* ipfv *síng* síng *na* loc *Píchi.* Pichi 'He sings in Pichi.' [au07se 233]

#### 9 Grammatical relations

Likewise, the effected non-specific objects of verbs denoting a process of production may remain unexpressed. Compare *só* 'sew' (242–243) and *kúk* 'cook' (244–245) in the transitive and intransitive sentence pairs below:


Affected-agent verbs are also lower on the scale of semantic transitivity than prototypical transitive verbs, because the actors are themselves affected by the situation in addition to the undergoer. In this group, we find transitive motion verbs like *rích* 'reach, arrive' (246–247) and *gó* 'go (away)' (248–249), whose goal objects may remain unexpressed:


Typical affected-agent verbs are the ingestive verbs *chɔ́p* (250) 'eat' and *dríng* 'drink' (251). These two transitive verbs are usually encountered without a patient object when its reference is non-specific. Note that object omission with *dríng* in combination with a habitual reading renders the idiomatic meaning 'habitually drink alcohol':

9.4 Valency adjustments

(250) *A* 1sg.sbj *kán* pfv *chɔ́p.* eat '(Then) I ate.' [ed03sb 016]

(251) *Dí* this *pɔ́sin* person *de* ipfv *dríng,* drink *na* foc *chak-mán.* drink.cpd-man 'This person drinks, he's a drunkard.' [dj07ae 363]

A final group of affected-agent verbs denote sensory perception, as well as mental and physical activities. Verbs belonging to this group that regularly occur without an overt non-specific object are *lúk* 'look', *hía* 'hear, understand', *sabí*/*nó* 'know', and *sí* 'see'.

When *sí* 'see, perceive' occurs without an object, its non-specific reading may translate as 'understand' or 'witness' (cf. e.g. 240). However, *sí* is also very often encountered in a non-specific context with a 3sg object pronoun (252) or an object NP *di tín* 'the thing' (253). Both of these objects are only faintly referential and therefore appear to function as dummy objects in very much the same way as non-referential subjects with expletive verbs (cf. §9.2.4):


The cognition verb *mɛ́mba* often appears without an explicit object with its meaning of 'remember' (254):

(254) *A* 1sg.sbj *nó* neg *de* ipfv *mɛ́mba.* remember 'I don't remember.' [fr03ft 047]

However, when *mɛ́mba* occurs in a transitive clause, it is best translated as 'think of', both with a specific object (255) and a non-specific one (256):


#### 9 Grammatical relations

Likewise, verbs denoting physical activities often occur with unexpressed objects. Consider *plé* 'play' in (257):

(257) *Bɔt* but *wi* 1pl *fít* can *de* ipfv *plé* play *a* 1sg.sbj *jám* make.contact *yú* 2sg.indp *yu* 2sg *fɔdɔ́n.* fall 'But we could be playing [football], I hit you (and) you fall.' [au07se 178]

The non-specific objects of verbs denoting the characteristic property of an agent often remain unexpressed. A sense of non-specificity permeates the following example featuring the verb *bɛt* 'bite'. It manifests itself in the use of the bare noun *dɔ́g* 'dog', the presence of the habitual aspect marker *kin* and the absence of an overt object:

(258) *Dɔ́g* dog *kin* hab *bɛ́t.* bite 'Dogs bite.' [dj07ae 371]

### **9.4.3 Unexpressed reflexive and reciprocal nominals**

Pichi speakers may make use of the reflexive anaphor *sɛ́f* or a body part noun in order to express reflexivity and reciprocity (cf. §9.3.5 and §9.3.6). There are also verbs that allow a reflexive interpretation but do not generally occur with a reflexive pronoun. Verbs whose reflexive pronouns usually remain unexpressed instantiate "middle voice" (Kemmer 1993) and denote situations that imply volition and instigation by the agent, involve physical action of the agent upon her/himself, or imply movement of the body.

The following examples involve the "body care" verbs *wás* 'wash' (259), *báf* 'bathe' (260), and *wɛ́r* 'dress (up)' (193). Note that *wɛ́r* takes an object in (262) and still implies reflexivity:


#### 9.4 Valency adjustments

In principle, these verbs may also occur with a reflexive pronoun, although they do so less frequently. Compare the usage of *wás* 'wash (oneself)' and *wɛ́r* 'dress (up)' in the following sentences:


The basic posture verbs *slíp* 'lie (down), sleep', *tínap* 'stand (up)' and *sidɔ́n* 'sit (down)' are never encountered with a reflexive pronoun in the corpus (cf. §8.1.3 for an extensive treatment). In contrast, verbs denoting less prototypical postures, e.g. *líng* 'lean over' and *bɛ́n* 'bend (over)' in (265–266), as well as those denoting other types of body-related events, e.g. *háyd* 'hide' in (267–268) are found with or without reflexive pronouns:


Other verbs in this group that occur with or without reflexive pronouns are the synonymous verbs *bló* 'rest' or *rɛ́s* 'rest':


9 Grammatical relations

(271) *A* 1sg.sbj *wánt* want *gó* go *bló* relax *mi* 1sg.poss *sɛ́f.* self 'I want to go rest.' [dj07ae 380]

Verbs with an inherently reciprocal meaning may appear with or without the reflexive and reciprocal anaphor *sɛ́f* 'self'. Consider the use of reciprocal *sɛ́f* with the sexual act denoting verbs *nák* 'knock' (272) and *slíp* 'sleep with' (273), as well as the unexpressed reciprocal pronoun in (274). These examples also illustrate that sexual act denoting verbs, including highly transitive ones like *nák*, do not imply a male agent in Pichi:


Conversely, the inherently reciprocal verbs of social interaction *mít* 'meet' and *mítɔp* 'meet' do not normally occur with the anaphor *sɛ́f* (275–276):


Nevertheless, like other inherently reciprocal verbs, *mít* and *mítɔp* may take part in a reciprocal alternation (cf. also *fɛ́t* 'fight' in 52). The two participants may be expressed as coordinate subjects in an intransitive clause while reciprocity is understood. Compare the transitive use of *mít* 'meet' in (277), with its intransitive use with two coordinate subjects in (278):


9.4 Valency adjustments

A further example for this alternation is provided with *fíba* 'resemble' in the following transitive and intransitive sentences:


### **9.4.4 Causative constructions**

A lexically restricted means of expressing causation in Pichi is the use of labile verbs in transitive clauses (cf. §9.2.3). Pichi also features inherently causative verbs like *kíl* 'kill', which pairs with *dáy* 'die' in a semantic relation of causation. In this section, we are, however, only concerned with fully productive means of causative expression in Pichi.

Pichi causative constructions are periphrastic and involve the use of subordinate predication. Hence, the causative verb is realised as a main verb to a subordinate predicate of effect. Table 9.12 summarises the majority patterns of causative formation in Pichi. Minor variations to these patterns are discussed below.


Table 9.12 Causative constructions

Causative and permissive constructions are formed with the two verbs *mék* 'make' and *lɛ́f* 'leave, permit'. Examples (281–282) present their use in non-causative transitive clauses:


#### 9 Grammatical relations

Two types of causative constructions can be distinguished on formal grounds (cf. Yakpo 2012b, Yakpo 2017b). The most common type of causative construction in Pichi inolves a "balanced" structure (Cristofaro 2003). The causative event is expressed in two finite clauses and the causative verb and the verb-of-effect are linked in a relation of subordination. Sentence (283) below features the (inanimate) causer NP *lotería* 'lottery', the causative main verb *mék* 'make', the causee NP *mi mɔní* 'my money', and the subordinate verb-of-effect *bɔkú* 'be much'. The subordinate status of the effect situation is evident through its appearance in a subjunctive clause introduced by the modal complementiser and subjunctive marker *mék* 'sbjv':

(283) *Lotería* lottery *dɔ́n* prf *mék* make *mék* sbjv *mi* 1sg.poss *mɔní* money *bɔkú.* be.much 'The lottery has made my money become a lot.' [dj07ae 198]

The second type of causative construction involves a "deranked" (Cristofaro 2003) or "reduced" (Lehmann 1988) structure and argument sharing. The causee (here *=an* '3sg.obj') is the syntactic object of the causative main verb *mék* and at once the notional subject of the subordinate verb-of-effect *gó* as in (284). This construction is marginal in terms of frequency, and only attested with Group 1 speakers (cf. §1.3). I could not identify any semantic differences between the two types of caustive constructions:

(284) *A* 1sg.sbj *go* pot *mék=an* make=3sg.obj *gó* go *tumɔ́ro.* tomorrow 'I'll make him go tomorrow.' [to05fn 030]

Both transitive and intransitive verbs may be causativised. Example (285) features a causative construction with the intransitive verb of effect *bɛ́lch* 'belch' and (286) one with the transitive verb *wích* 'bewitch'. Like all complement clauses, the subjunctive clause in these constructions can optionally be introduced by the quotative marker *sé* 'qot' in addition to *mék* 'sbjv' (286):


Sentence (287) illustrates the two options for rendering causative meaning with labile verbs. Before the comma, the verb *drɔ́ngo* 'be/get drunk' is used as a transitive and causative verb followed by the patient object pronoun *=an* '3sg.obj'. In the second half of the sentence, causative meaning is expressed periphrastically through the *mék* causative construction. When the second option is used, the speaker may want to express that causation is less direct. Meanwhile, the use of the transitive variant of a labile verb implies a direct, possibly even physical implication of the causer:

9.4 Valency adjustments

(287) *A* 1sg.sbj *drɔ́ngor=an,* get.drunk=3sg.obj *a* 1sg.sbj *mék* make *mék* sbjv *e* 3sg.sbj *drɔ́ngo.* be.drunk 'I got him drunk, I made him drunk.' [dj07ae 053]

The following example illustrates the causative use of the ditransitive transfer verb *gí* 'give' in a double-object construction:

(288) *E* 3sg.sbj *bin* pst *mék* make *mék* sbjv *a* 1sg.sbj *gí* give *di* def *gɛ́l* girl *di* def *plantí.* plantain 'She made me give the girl the plantain.' [dj05be 003]

There are no restrictions on negation in causative constructions. The causative verb in the main clause (289) as well as the verb of effect in the subordinate clause (290) may be negated:


remain

'(You) can sift it up to two times to make none of that little water remain.' [dj03do 008]

There are instances in which TMA marking in the subjunctive clause of effect is not reduced as it usually is in a subjunctive clause (cf. §10.5.1). These instances involve the idiomatic expressions *na ín mék*/*na di tín mék* 'that's why' and the question phrase *wétin mék* 'why'.

Hence, the subordinate clauses in (291) and (292) feature regular TMA marking via *dɔ́n* 'prf' and *de* 'ipfv', respectively, instead of subjunctive marking. Nonetheless, even these idioms are occasionally conceived of as regular causative constructions with the reduced TMA marking characteristic of subjunctive subordinate clauses (cf. 286 above):


#### 9 Grammatical relations

The subjunctive marker *mék* also introduces the complement clauses of other main verbs, which – like the causative verb *mék* 'make' – induce deontic modality over their subordinate clauses. One such main verb is *wánt* 'want' (293) (cf. §10.5.5 for a full treatment of the functions of *mék* 'sbjv' in subordinate clauses):

(293) *Ús=sáy* q=side *yu* 2sg *wánt* want *mék* sbjv *di* def *smók* smoke *kɔmɔ́t?* come.out 'Where do you want the smoke to come out?' [ye07fn 123]

Besides that, *mék* 'sbjv' introduces purpose and certain types of consecutive clauses (cf. §10.7.6) as well as imperatives and other types of directive main clauses (cf. §6.7.3.3). The conflation of these functions in the element *mék* represents a case in which the semantic linkages within a functional domain are actually instantiated in a single form (cf. Bybee et al. 1994: 213–30; Song 2001: 25–33).

The verb *lɛ́f* 'leave, remain' is employed as a causative verb in the formation of permissive causatives. This type of causative is usually formed differently from the causative proper, i.e. constructions featuring the causative verb *mék* 'make'. The effect situation is also expressed in a subjunctive clause. Yet, it is commonplace to express the causee as the object of *lɛ́f* and reiterate it as the subject of the subordinate subjunctive clause.

Consider the following two permissives and compare them with a causative construction like (285) above. In (285), the causative verb *mék* takes no object pronoun *mí* '1sg.indp' that is co-referential with the subject *a* '1sg.sbj' of the subjunctive clause.


[ab03ab 151]

The verb *lɛ́f* 'leave, remain' is also employed in the formation of resultative causatives. Resultative causative constructions serve to causativise stative situations denoted by property items, as well as stative situatons denoted by the identity copulas *na/nóto* and *bí* and their complements in equative clauses. Resultative causative constructions do not feature a subordinate clause. Instead, the effect situation is expressed as a resultative complement to the causative verb *lɛ́f* (cf. §11.3 for resultative adjuncts in secondary predicate constructions).

Sentence (296) features the property item *yún* 'be young'. The resultative causative equivalent in (297) features the causer *e* '3sg.sbj' = 'it' (i.e. 'the clothing'), the causee *yú* '2sg.indp', which is an object to *lɛ́f* 'leave', as well as the resultative complement *yún*

#### 9.4 Valency adjustments

'young'. The verb *lɛ́f* in these constructions may either be used as an inchoative-stative verb, as in (297), or a dynamic verb, as in (301) below, where *lɛ́f* is specified by *de* 'ipfv':


Example (298) presents a non-causative predication involving the inchoative-stative property item *kɔrɛ́t* 'be correct'. The resultative causative counterpart in (299) features the force causer *gɔ́d* 'God', the causative verb *lɛ́f* 'leave', and the resultative complement *kɔrɛ́t* '(be) correct':


Sentence (300) is an equative clause featuring the identity copula/focus marker *na* 'foc'. The causative equivalent in (301) once more features the resultative causative verb *lɛ́f*, as well as the compound noun and resultative complement *yun-bɔ́y* 'young.cpd-boy':


An interesting semantic aspect of the use of resultative causatives is that they are not attested with human causers occupying the agent role. All recorded instances of resultative causatives feature inanimate force causers in the subject position. I assume that speakers prefer to employ causative constructions featuring *mék* 'make' where the causer is human, or where they intend to convey a notion of strong agency on the part of the causer even if it is inanimate (e.g. sentence (283) above with the force causer *lotería* 'lottery' and the property item *bɔkú* 'be much' as a verb of effect).

The verb *pút* 'put' is also used as a causative verb in a few instances in the corpus. In (302) below, *pút* is employed like *lɛ́f* in (297) and (301) above in order to express

#### 9 Grammatical relations

the resultative causative equivalent of a non-causative equative clause. The sentence contains the non-causative equative clause *yu húman na bíg húman* 'your wife is an important woman' and the causative equivalent *pút yu sɛ́f bíg mán* 'make yourself an important man':

(302) *Ɛf* if *yu* 2sg *húman* woman *na* foc *bíg* big *húman,* woman *e* 3sg.sbj *hád* hard *fɔ* prep *pút* put *yu* 2sg *sɛ́f* self *bíg* big *mán.* man 'If your wife is an important woman, it is difficult to make yourself an important man.' [ma03hm 083]

### **9.4.5 Impersonal constructions**

A backgrounding passive may be formed by using impersonal *dɛn* '3pl' in the subject position. To begin with, the 3pl personal pronoun *dɛn* may be used generically to refer to a loosely specified collective. Example (303) features the generic, impersonal use of *dɛn* in a transitive clause:

(303) *Dɛn* 3pl *de* ipfv *wɛ́r* wear *wáyt* white *ɔ́p* up *violeta* violet *dɔ́n.* down 'They [the pupils] wear white up (and) violet down.' [ma03hm 032]

The pronoun *dɛn* is also used impersonally with verbs characterised by a higher degree of semantic transitivity. In clauses with verbs that presuppose a volitional, instigating, and animate agent and an affected patient, impersonal use of *dɛn* serves to background a non-specific agent:

(304) *Esto* this *na* loc *wán* one *ɔ́da* other *kɔ́ntri,* country *dɛn* 3pl *go* pot *púl* remove *yú* 2sg.indp *inmediatamente,* immediately *dɛn* 3pl *de* ipfv *púl* remove *yú* 2sg.indp *wók.* work

'This in another country, they would remove you immediately, they would remove you from your job.' [ye03cd 077]

The following two sentences exemplify the pragmatic and syntactic rearrangements which go along with the use of the labile property item *strét* 'be straight, straighten' in an intransitive (305) and a transitive clause (306), respectively. In the intransitive clause, the subject *ród* 'road' is patient to the inchoative-stative verb *strét*. In the transitive clause, impersonal *dɛn* in subject position denotes the backgrounded agent, while the patient *ród* is now in object position:

(305) *Di* def *ród* road *strét.* be.straight 'The road is straight.' [dj07ae 122]

9.4 Valency adjustments

(306) *Dɛn* 3pl *dɔ́n* prf *strét* straighten *di* def *ród.* road 'The road has been straightened.' [dj07ae 123]

Impersonal *dɛn* always refers to an unspecified group of animate, usually human agents. The lower the agent is on the animacy scale, and hence its capacity of volition and instigation, the less likely it is to be referred to by impersonal *dɛn*. For example, (307) sounds awkward, since the backgrounded agent is construed as animate and human. A situation involving a non-human agent like *snék* 'snake' is therefore more likely to be expressed through an 'active' clause with a foregrounded agent in subject position (308):


However, impersonal *dɛn* does not retain its plural reference by default. Sentence (309) was elicited by means of the "caused positions" video clip series of the Language and Cognition Group of the Max Planck Insitute for Psycholinguistics in Nijmegen. In all preceding clips, the agent of a series of actions had been a single individual. Nonetheless, the following sentence was given in response to a still image showing a pot lying upside down on a table:

(309) *Dɛn* 3pl *pút=an* put=3sg.obj *mɔ́t* mouth *dɔ́n* down *fɔ* prep *di* def *tébul.* table 'It has been put mouth-down on the table.' [li07pe 089]

Impersonal *dɛn* is subject to some morphosyntactic restrictions inherent to the nonspecific nature of the pronoun. Impersonal *dɛn* may not be focused, relativised, or subjected to other operations which require specific reference.

Agent-backgrounding may also be achieved via the use of generic, non-specific, and non-referential nouns like *pɔ́sin* 'person' and *mán* 'man, human-being'. The generic noun *pɔ́sin* 'person, human-being' may occur as an agent subject in transitive clauses and function like impersonal *dɛn* '3pl'. The noun *pɔ́sin* refers to a backgrounded non-specific human agent. Compare the use of *pɔ́sin* and *dɛn* in these two near-identical sentences:

(310) *Pɔ́sin* person *go* pot *entiende* understand *bɔt* but *e* 3sg.sbj *nó* neg *dé* be.loc *bien.* good 'One would understand but it's not correct.' [dj05be 043]

#### 9 Grammatical relations

(311) *Dɛn* 3pl *go* pot *hía* hear *ín* 3sg.indp *bɔt* but *e* 3sg.sbj *nó* neg *só* like.that *dé* be.loc *claro.* clear 'It would be understood but it's not so clear.' [ye0502e2 050]

In addition to *dɛn* '3pl', other personal pronouns are also sometimes used with weak reference. Example (312) features the use of *wi* '1pl' in the idiom which serves as a response to the enquiry 'how are you?'. Also compare the use of *wi* in (313):


'It's like with us now, (if) we spoke Spanish, we would know that this person, is Nigerian.' [ma03hm 045]

Likewise, the impersonal backgrounded use of *yu* '2sg' is common in procedural texts (314):

(314) *Dé,* there *ɛ́ni* every *káyn* kind *tín* thing *na* foc *mɔní,* money *yu* 2sg *fít* can *mék* make *ɛ́ni* every *káyn* kind *tín* thing *yu* 2sg *go* pot *sí* see *mɔní.* money

'There, everything is money, you can do anything (and) you'll earn money.' [ma03hm 054]

Finally, the copula and focus marker *na* 'foc' may be used to construct purpose-like clauses with impersonal reference with an obligation reading in combination with the prepositions *fɔ* 'prep' or *to* 'to' and a subsequent verb without person-marking (315):


## **10 Clause linkage**

Relations between clauses may be established in various ways in complex clauses consisting of more than one verb. A relation between clauses can be expressed by using linking adverbials and anaphoric pronouns (§10.1). Adjacent clauses may also be linked by continuative intonation alone, or in combination with the other means available (§10.8). Clause linkers may be employed in order to form complex coordinate (§10.3), complement (§10.5), relative (§10.6), and adverbial clauses (§10.7).

The resulting constructions are syntactically integrated to varying degrees. For instance, subjunctive clauses introduced by *mék* 'sbjv' (§10.5.5), purpose clauses introduced by *fɔ* 'prep', and complement clauses introduced by *fɔ* 'prep' or *de* 'ipfv' are less finite and arguably syntactically subordinate to their main clauses. At the same time, it is not very useful to posit a relation of syntactic subordination between clauses in many (other) adverbial relations. In these structures, the linked clauses retain their full potential for the expression of person, tense, aspect, and modality (e.g. the various types of adverbial clauses introduced by *wé* 'sub', cf. §10.7.1). Clauses may also be linked in multiverb constructions, which are covered separately in section §10.8.

### **10.1 Linking adverbs and anaphor**

Linking adverbs occur at the beginning of a clause and ensure referential continuity with a preceding clause, often in combination with continuative intonation. Recurrent linking elements are *áfta* 'then, afterwards', *bɔt* 'but', the phrasal adverbial *dán tɛ́n* '(at) that time', as well as the anaphoric phrase *na ín* 'foc 3sg.indp'.

The adverb *áfta* 'then, afterwards' relates a situation with a previous one. It can be employed in ways very similar to that of certain clause linkers in prosodically more integrated constructions involving the clause linker *wé* 'sub' (cf. §10.7.1). In the following sentence, *áfta* and *wé* both establish a link of temporal succession with the preceding clause. Both elements are preceded by continuative intonation (indicated by a comma):

(1) *Yu* 2sg *gó* go *yu* 2sg *pé,* pay *siete* seven *mil* thousand *yu* 2sg *baja,* go.down *áfta* then *yu* 2sg *fínis* finish *yu* 2sg *sube,* go.up *wé* sub *yu* 2sg *de* ipfv *pák* pack *mɔ́* more *siete* seven *mil,* thousand *wé* sub *yu* 2sg *sube.* go.up

'You go, you pay, seven thousand, you go down, then you finish, you go up and take seven thousand again and go up.' [f203fp 012]

Example (2) shows how the sequential meaning of *áfta* can be read as a result relation in combination with continuative intonation:

#### 10 Clause linkage

(2) *A* 1sg.sbj *nó* neg *sabí* know *ús=tín* q=thing *bin* pst *kán* pfv *pás,* pass *áfta* then *e* 3sg.sbj *gó* go *na* loc *hospital.* hospital 'I don't know what happened that he went to (the) hospital.' [ye03cd 074]

The adverb *áfta* may also introduce the then-clause of reality conditionals in which the if-clause is introduced by *lɛk* 'like' (3):

(3) *Lɛk* like *náw,* now *lɛk* like *Boyé* name *só* like.that *na* foc *mi* 1sg.poss *mán,* man *áfta* then *mi* 1sg.poss *sísta* sister *go* pot *kɔ́l=an* call=3sg.obj *sé,* qot *wé* sub *e* 3sg.sbj *go* pot *kán,* come *"ús=sáy* q=side *mi* 1sg.poss *brɔda-lɔ́* brother.cpd-law *dé?"* be.loc 'Suppose now, suppose Boyé here were my husband, then my sister would call him, if she came, "where's my brother-in-law?"' [ro05de 005]

Example (4) shows how the sequential meaning of *áfta* can be read as a reason relation:

(4) *Ɛf* if *yu* 2sg *sí* see *sé,* qot *sɔn* some *sáy* side *di* def *plés* place *klín,* be.clean *áfta* then *dɛn* 3pl *de* ipfv *dú* do *di* def *tín* thing *dɛn* pl *fáyn,* fine *yu* 2sg *nó* neg *go* pot *bísin* be.busy *ɛf* if *yu* 2sg *gasta* spend *mɔní.* money

'If you see that, somewhere the place is clean and/ because things are done well, you don't bother if you spend money.' [ma03hm 066]

The phrasal adverbial *dán tɛ́n* 'at that time' also relates a situation to a preceding one. In (5b), *dán tɛ́n* indicates a temporal relation of simultaneity with the preceding clause (a):

	- that time 1sg.poss aunt get belly

'At that time my aunt was pregnant.' [fr03ft 087]

The phrase *na ín*, consisting of the focus marker *na* and the emphatic 3sg pronoun *ín* establishes various types of anaphoric relationships (cf. also §7.4.3.3). In (6), a temporal interpretation is favoured due to the presence of the adverbial *las doce* 'twelve (o'clock)':

(6) *Bikɔs* because *ín* 3sg.indp *de* ipfv *sé,* qot *ɛ́ni* every *las* the.pl *doce* twelve *na* foc *ín* 3sg.indp *in* 3sg.poss *abuela* grandmother *kin* hab *kán* come *kɔ́l=an.* call=3sg.obj

'Because she would say, always at twelve o'clock, that's when her grandmother used to come and call her.' [ed03sb 150]

10.2 Clause linkers

### **10.2 Clause linkers**

Next to the use of anaphors, intonation, and SVCs, Pichi employs a large array of clause linkers to express relations between clauses. Linkers that serve to introduce adverbial clauses more specialised in their meanings are dealt with in §10.7. At the same time, most types of relations, including adverbial ones, can be expressed by one, or a combination of, the multifunctional elements *wé* 'sub', *sé* 'qot', *mék* 'sbjv', and *fɔ* 'prep'.

These four linkers have multiple, partially overlapping functions, which are mapped in Figure 10.1. The ways in which these four linkers introduce different types of clauses are covered in the following sections of this chapter.

Serial verb construction

Figure 10.1 Functions of *fɔ*, *mék*, *wé*, and *sé* by clause type

### **10.3 Coordination**

Coordinate clauses may be linked by way of intonation as well as the linkers *wé* 'sub' and *an* 'and'. In (7), *bús* 'forest' bears a continuative boundary tone, which links the clause to the following one after the comma.

10 Clause linkage

(7) *Só* so *e* 3sg.sbj *gó* pot *na* loc *bús,* forest *e* 3sg.sbj *sé* qot *e* 3sg.sbj *de* ipfv *gó* go *kíl* kill *bíf.* wild.animal 'So he went to the forest, (and) he said he was going to kill wild game.' [ma03sh 004]

The clause linker *wé* 'sub' can, amongst its other uses, link coordinate clauses. The preposition *wet* 'with' may only conjoin NPs (cf. §5.5), hence an important function of *wé* is to serve as a clausal connective that can be translated as 'and (then)'. The formal differentiation between NP and clausal coordination in Pichi corresponds to an areal (West) African pattern (Mithun 1988: 349–353).

In the following excerpt from a personal narrative, the first *wé* 'sub' in (b) establishes a link (b) to the preceding clause (a) after a clause-final declarative intonation (indicated by the full stop). At the same time, context suggests a more temporal meaning of 'when' of the second *wé* in (b). Clause (c) resumes the narrative after declarative intonation at the end of (b):

(8) a. *Áfta* then *na* foc *mi* 1sg.poss *gran-má* grand-ma *a* 1sg.sbj *bin* pst *de* ipfv *kɔ́l* call *mamá.* mother 'So it's my grandmother that I used to call mother.' [fr03ft 016] b. *Wé* sub *wi* 1pl *kán* pfv *kán* come *na* loc *tɔ́n,* town *wé* sub *a* 1sg.sbj *bigín* begin *gó* go *skúl,* school *wé* sub *a* 1sg.sbj *bin* pst *gɛ́t,* get *a* 1sg.sbj *tínk* think *sé* qot *seis* six *años.* years And then we came to town, and then I began to go to school, when I was, I think six years old.' [fr03ft 017] c. *A* 1sg.sbj *bigín* begin *gó* go *skúl* school

'I began going to school.' [fr03ft 018]

The sequential and temporal meanings of *wé* 'sub' in clauses like (8b) above may extend into contiguous meanings such as adversative (9). The various related meanings of *wé* in these contexts may blur beyond recognition the demarcation between the coordinate clauses described in this section and the adverbial clauses covered in §10.7.1.

(9) a. *Frijoles* bean.pl *yɛ́s* yes *frijoles.* bean.pl '[The Cubans call them] frijoles, yes frijoles.' [ed03sp 119] b. *Wé* sub *yá* here *só,* like.that *frijoles* bean.pl *na* foc *haricots* beans *na* loc *yá.* here 'While here, frijoles is haricot here.' [ed03sp 120]

The quotative marker *sé* 'qot' also functions as a sequential connective and clause coordinator in ways very similar to *wé* 'sub' when it signals inner speech or "internal

10.4 Quotation

awareness" (Güldemann 2008: 422) and thereby often occurs without an overt subject as in (10):

	- b. *Sé* qot *mɔ́nin* morning *tɛ́n* time *a* 1sg.sbj *go* pot *gó,* go *a* 1sg.sbj *báy,* buy *a* 1sg.sbj *ték* take *tú* two *peso* peso (*…*) 'So in the morning, I would go and buy (it), I would take two pesos (…)' [ed03sp 083]

The element *an* 'and' may link NPs as well as coordinate clauses. Its use is, however, exceedingly rare, and speakers overwhelmingly favour coordinate structures linked by means of *wé* 'sub' or reduced clauses involving secondary predication (cf. §11.3):

(11) *E* 3sg.sbj *nák* hit *di* def *tébul* table *an* and *di* def *stáyl* style *wé* sub *e* 3sg.sbj *nák* hit *di* def *tébul* table *strɔ́n,* be.strong *e* 3sg.sbj *kán* pfv *sék* shake *di* def *plét* plate *an* and *di* def *plét* plate *kán* pfv *brók.* break 'He hit the table and the way that he hit the table in a strong way, he shook the

plate, and the plate broke.' [au07se 014]

The disjunctive coordinator *ɔ* 'or' may also link coordinate clauses, cf. (54) for an example.

### **10.4 Quotation**

The element *sé* 'qot' is characterised by an exceptional polyfunctionality that includes use as a lexical verb 'say' and use as quotation marker for direct speech and naming, renders inner speech and internal awareness, introduces adverbial clauses of manner, circumstance, and purpose, and reaches into the domain of clausal complementation. Following Güldemann (2008), I assume that the function as an index of direct reported speech lies at the heart of the functional versatility of *sé* 'qot'.

The element *sé* occurs with a more lexical meaning of 'say'. It may take TMA marking and at the same time predicate a quotative construction. In the following example, *sé* is employed as a speech verb. It is marked for potential mood by means of *go* 'pot' and introduces a direct quote:

10 Clause linkage

(12) *Di* def *dé* day *wé* sub *yu* 2sg *go* pot *níd=an,* need=3sg.obj *yu* 2sg *go* pot *sé* qot *"a* 1sg.sbj *nó* neg *gɛ́t* get *pamáyn",* oil *yu* 2sg *go* pot *kɔ́t* cut *gadinɛ́ks.* egg-plant

'The day that you will need it, you are going to say "I don't have oil," (and) you will cut egg-plants.' [ab03ay 015]

In the example below, the use of *sé* as a lexical verb 'say' coincides with the presence of habitual marking (i.e. *kin* 'hab'). However, in the overwhelming majority of instances, *sé* remains bare, and hence marked for factative TMA, since quotative constructions by their very nature occur in reported, past-time discourse:

(13) *E* 3sg.sbj *kin* hab *sé* qot *"kán* come *wi* 1pl *gó* go *na* loc *Barca* place *wi* 1pl *gó* go *dríng."* drink 'He usually says "come let's go to Barca and drink".' [ro05rt 029]

The transition from a more lexical reading of *sé* to a more functional one is far from clear-cut (which is why I have opted for a unitary gloss of 'qot' in all contexts). First, distributional restrictions set *sé* apart from the true speech verbs *tɔ́k* 'talk, say' and *tɛ́l* 'tell'. For instance*,* sé does not normally take a nominal object, as does *tɔ́k*. Compare (14a) and (14b).

(14) a. \* *Mék* sbjv *a* 1sg.sbj *sé* qot *wán* one *wɔ́d.* word Intended: 'Let me say one word.' [to07fn 219] b. *A* 1sg.sbj *tɔ́k* talk *wán* one *wɔ́d.* word 'I said one word.' [to07fn 220]

Beyond that, adverbials do not usually modify *sé* 'qot' (15a). Adverbials only appear as quoted complements indexed by *sé* (b). Again, there is no restriction on adverbial modification of the speech verb *tɔ́k* 'talk, say' (c):

(15) a. \* *A* 1sg.sbj *sé=an* qot=3sg.obj *kwík.* quickly Intended: 'I said it quickly.' [to07fn 221] b. *A* 1sg.sbj *sé* qot *"kwík".* quickly 'I said "quickly".' [to07fn 222] c. *A tɔ́k=an kwík*

1sg.sbj talk=3sg.obj quickly 'I said it quickly.' [to07fn 223]

10.4 Quotation

Secondly, *sé* 'qot' is not normally encountered as a verbal complement. Hence below, the speech verb *tɔ́k* 'talk, say' appears as verbal complement to the modal verb *fít* 'can'. The appearance of *sé* in this position is not attested.

(16) *Yu* 2sg *fít* can *tɔ́k* talk *"a* 1sg.sbj *de* ipfv *fíl* feel *di* def *sɛ́nt* scent *fɔ* prep *lɛk* like *háw* how *e* 3sg.sbj *de* ipfv *kúk* cook *di* def *plantí"* plantain *ɔ* sp *"a* 1sg.sbj *de* ipfv *siente* feel *di* def *sɛ́nt* scent *sé* qot *pɔ́sin* person *de* ipfv *kúk* cook *plantí* plantain *dé".* there 'You can say "I smell the scent of him cooking the plantain", or "I smell the scent that somebody is cooking plantain there".' [dj05ae 026]

Note that I do not analyse *sé* as a V2 of a complementation SVC when it functions as a complementiser to a verb like *siente* 'feel' above (cf. also §10.5.6). The peculiar distribution of *sé* as a speech "verb" and its broad functional domain, which extends far beyond complementation, may point to the fact that *sé* 'qot' did not start out as a speech verb in the first place. Instead, it is conceivable that the use of *sé* as a speech "verb" is derived from quotation just like its many other functions (cf. Güldemann 2008: 272–275). In this view, the resemblance of *sé* with a purported English etymon *sáy* may be due either to chance or to the convergence of diverse etymologies and functions in one form.

The recurrent use of quotative clauses introduced by *sé* 'qot' with or without a preceding subject in order to render direct and inner speech is a conspicuous feature of longer stretches of narrative discourse. Direct speech in Pichi rarely serves the sole aim of giving neutral reports of utterances. One of its crucial functions is the creation of an atmosphere of vivacity and authenticity that builds up tension and draws listeners into the narrative. Compare (17a–17e), in which speaker (ed) recalls his difficulty in distinguishing a transsexual man from a woman:

(17) a. *A* 1sg.sbj *sé* qot *"na* foc *mán* man *dís?"* this 'I said "this is a man?"' [ed03sb 222] b. *E* 3sg.sbj *sé* qot *"na* foc *mán."* man 'He said "it's a man".' [ed03sb 223] c. *A* 1sg.sbj *sé* qot *"yu* 2sg *de* ipfv *krés* be.crazy *mán."* man 'I said "you're crazy, man".' [ed03sb 224] d. *E* 3sg.sbj *sé* qot *"na* foc *mán* man *dís."* this 'He said "this is a man".' [ed03sb 225] e. *Sé* qot *na* foc *mán?* man '(You) say it's a man?' [ed03sb 226]

#### 10 Clause linkage

Example (18) shows that the absence of overt subjects in this type of discourse opens up a grey area in which there is ample room for both a more functional and a more lexical reading of a subject-less, clause-initial *sé*. Compare the unambiguous use of *sé* as a speech verb in (18a) with the alternative translations of the subject-less *sé* in (18b):

(18) a. *E* 3sg.sbj *go* pot *sé* qot *e* 3sg.sbj *de* ipfv *fíɛ,* fear *e* 3sg.sbj *nó* neg *go* pot *gí* give *mí* 1sg.indp *di* def *tín* thing *wé* sub *a* 1sg.sbj *de* ipfv *sɛ́n=an.* send=3sg.obj

> 'He would say, he was afraid (and) he wouldn't give me the thing that I was sending him for.' [ab03ab 041]

b. *Sé* qot *ín* 3sg.indp *nó* neg *wánt* want *in* 3sg.poss *abuelo* grandfather *skrách=an.* scratch=3sg.obj '(He'd) say he [emp] doesn't want his grandfather to scratch him.' or 'Because he doesn't want his grandfather to scratch him.'

Reported discourse also renders inner speech at important narrative junctures. In such a context, reported discourse may serve to express the intention of referents as in the sentences below:

(19) *In* 3sg.poss *brɔ́da* brother *dɛn* pl *ɔ́l* all *kɔmɔ́t* go.out *na* loc *tɔ́n* town *yá* here *só* like.that *dɛn* 3pl *sé* qot *dɛn* 3pl *de* ipfv *kán* come *ték=an.*

take=3sg.obj

'His brothers all left town, (so) they said they came to take her.' [ab03ay 142]

(20) *E* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *ni* even *Panyá,* Spanish *e* 3sg.sbj *sé* qot *e* 3sg.sbj *wánt* want *muchachita* young.girl *de* of *diecisiete* seventeen *años.* year.pl

'He doesn't even know how to talk Spanish (and) he says he wants a young girl of seventeen years.' [ye03cd 053]

Speakers may use 3rd person pronouns in reported speech as in (20) above or insert direct quotations as in (21) below. These elements together constitute some of the conspicuous characteristics of Pichi narrative discourse, in which the already weak boundary between direct and indirect speech in Pichi is often deliberately blurred as part of a performance-oriented narrative technique:

(21) *Tidé* today *e* 3sg.sbj *kán* come *e* 3sg.sbj *sé,* qot *"a* 1sg.sbj *tínk* think *sé* qot *a* 1sg.sbj *go* pot *fínis* finish *ɔ́l* all *di* def *resto".* rest

'Today he came, he said "I think I am going to finish all the rest".' [ye03cd 147]

10.4 Quotation

A further facet of the quotative function is the use of *sé* in a naming construction which serves to identify a nominal element by name and introduce members of a list (cf. Güldemann 2008: 398). The named or listed items appear as nominal objects of *sé*.

(22) *Krío* Krio *mamá* mother *dɛn* pl *wé* sub *dɛn* 3pl *de* ipfv *tɔ́k* talk *Píchi* Pichi *dɛn* 3pl *kin* hab *tɔ́k* talk *sé* qot *grín.* green 'The elderly Krio women, when they talk Pichi, they usually say green.' [as opposed to 'verd' like younger people] [dj05ce 257]

In combination with the verb *kɔ́l* 'call', the naming construction translates as 'be in a kinship relation with X':

(23) *Na* foc *fada-lɔ́,* father.cpd-law *na* foc *di* def *papá* father *wé* sub *e* 3sg.sbj *bɔ́n* beget *mí,* 1sg.indp *na* foc *ín* 3sg.indp *mi* 1sg.poss *mán* man *go* pot *kɔ́l* call *sé* qot *suegro.* father-in-law 'That is the father-in-law, that is the father who begat me, it is him that my

husband would call father-in-law.' [ro05de 007]

Sentence (24) exemplifies the use of *sé* in listing. In these examples, the name or members of the list appear as nominal complements of *sé*:

(24) *A* 1sg.sbj *fít* can *tɛ́l* tell *yú* 2sg.indp *sé* qot *morera,* mulberry *teca,* teak *kalabo.* kalabo

'I can tell you mulberry, teak, kalabo [listing types of wood].' [ro05de 051]

The use of *sé* to identify a nominal element represents the only context in which the quotative marker does not introduce a clause. Through this characteristic, the naming construction may be structurally identical to a copula construction involving the focus marker and identity copula *na* 'foc'. Compare the two consecutive sentences in (a) and (b) below:

	- 'Namely my mum.' [dj05ce 037]

The data also contains examples in which the use of *sé* as a deictic identifier of a nominal entity has been taken to its logical conclusion. In (26), *sé* expresses identity in combination with the copula and focus marker *na:*

(26) *Di* def *pikín* child *ɔ́l* all *sé* qot *na* foc *mi* 1sg.poss *yón* own *bikɔs* because *a* 1sg.sbj *dɔ́n* prf *pé* pay *mɔní.* money 'The children are all mine because I have paid money [the dowry].' [hi03cb 196] 10 Clause linkage

Aside from the functions covered in this section, the element *sé* 'qot' is employed as a general clausal complementiser (cf. §10.5).

### **10.5 Complementation**

This section covers complex clauses featuring subordinate clauses with the syntactic function of complements. In the following, such clausal participants are referred to as complement clauses. Five strategies of integration of main and subordinate verbs are used next to each other, and sometimes they overlap (cf. Table 10.1). These strategies are covered in the following sections.

### **10.5.1 Finiteness**

Finiteness is an indicator of the degree of integration of Pichi complement clauses with main clauses. Main verbs vary with respect to how syntactically independent their complement predicates may be. Main verbs differ with respect to the complementiser they occur with, the time reference they project over their complement predicates, the person and TMA marking potential they accord their complement verbs, and the potential they confer on their complement verbs to be negated. In this vein, complement clauses consisting of a verb alone constitute the non-finite pole and complement clauses, in which the verb retains its full syntactic potential and constitutes the finite pole of complement clauses. Table 10.1 checks the four principal complementation strategies in Pichi against five diagnostics of finiteness. "Complement clause" is abbreviated as "CC" in the table, "main clause" as "MC".

The complementation strategies in Table 10.1 form part of a continuum of complement clauses. The cline from non-finiteness to finiteness encompasses four complementation strategies, featuring the three overt complementisers *fɔ* 'prep', *mék* 'sbjv', and *sé* 'qot', and a "zero" strategy. At the left end of the continuum, we find the highest number of syntactic restrictions in CCs linked to main verbs via the associative preposition *fɔ* 'prep'. These are aspectual and modal auxiliary constructions. The subject of the CC verb must be co-referential with that of the main verb, is dependent on the temporal specification provided by the main verb, and may not be marked independently for person, negative polarity, or TMA.

A significant number of modal and aspectual auxiliary verbs take clausal complements without an intervening complementiser, indicated by the column headed by "∅ (none)" in Table 10.1. A small sub-group of these verbs may, however, optionally be followed by the imperfective marker *de*, which may then be seen to function as complementiser. However, the presence of *de* 'ipfv' also adds an aspectual nuance by emphasising the continuous nature of the situation denoted by the CC verb. Such structures are therefore slightly more finite. On the one hand, the CC verb may be marked for aspect. On the other hand, the time reference of the CC verb is determined by the taxis relation projected by the MC verb over the complex clause; for example the CC verb *chɔ́p* 'eat' is necessarily


Table 10.1 Complementation and finiteness

10.5 Complementation

#### 10 Clause linkage

in a relation of simultaneous taxis with the MC verb *bigín* 'begin to' in a complement construction like *a bigín de chɔ́p* 'I began to eat.'

Subjunctive complement clauses are, again, more finite. They may be same or different subject, always feature person marking, and must be negated independently of the main verb to signal negative polarity. They are, however, restriced in their TMA marking potential and depend on the main verb in their time reference (they are invariably futureprojecting). At the right end of the continuum we find fully-fledged biclausal structures introduced by the quotative marker *sé* 'qot', which therefore functions as a typical finite complementiser. Not included in Table 10.1 are the clause linkers *ɛf(ɛ)* and *íf* 'if', which may function as complementisers in indirect question clauses (cf. §10.6.5).

### **10.5.2 Complement-taking verbs and complementisers**

Table 10.2 lists approximately sixty frequent Pichi main verbs that may take different types of complement clauses. The table sorts these verbs according to the type of complement clause linkage these verbs are attested with. The feature "semantic class" correlates strongly with the complementiser provided in the "linkage type" column. Beginning from the top of the table, the clause "linkage types" increase in finiteness as they descend towards the bottom. Verbs that may take complements introduced by *sé* 'qot' are not fully listed, since that would make the list unduly long. Equally, some of the verbs listed with complementisers other than *sé* 'qot' may nevertheless take complements introduced by *sé* when these are statements of fact and have independent time reference, e.g. *a de sɔ́ri sé e dɔ́n kán* 'I'm sorry that he has come.' Conversely, speech verbs take quotative complements introduced by *sé* but subjunctive complements when these are indirect commands, e.g. *a hála sé "kán"* 'I hollered "come"' vs. *a hála sé mék e kán* 'I hollered for him to come.'

Some verbs are listed twice under two types of clause linkage where the functions of complement clauses differ correspondingly. For example, *wánt* usually appears without an overt complementiser (∅) in prospective aspect constructions. However, *wánt* takes ∅ and *de* 'ipfv' complements in same-subject (desire) modal auxiliary constructions, and must take *mék* 'sbjv' with different-subject complements. Likewise, the general subordinator *wé* 'sub' is not listed in Table 10.2, since its function as a complementiser is marginal. Table 10.2 does not capture many other distributional complexities of complementisers and idiosyncracies of complementation, including negation in complement constructions. Details are provided in the corresponding sections of this chapter.

### **10.5.3** *De* **'ipfv'**

The aspectual and modal verbs *bigín* 'begin', *wánt/wɔ́nt* 'want, be about to', *fít* 'can', and *gɛ́fɔ* 'have to' feature complements introduced by the zero strategy or complement verbs preceded by the imperfective marker *de* 'ipfv'. *Bigín* is particularly likely to occur with *de* 'ipfv' when used as an ingressive auxiliary (cf. §6.4.1 for examples). The use of the imperfective marker de emphasises the continuous nature of the situation dennoted by the verb. Compare the following constructions.


Table 10.2 Complement-taking verbs, semantic class, and type of clause linkage




Note that both verbs in the constructions above are always co-referential; they have a subject in common. Aspect-marking for simultaneous taxis via imperfective aspect is also found with depictive secondary predicates (cf. §11.3).

10.5 Complementation

### **10.5.4** *Fɔ* **'prep'**

The multifunctional element *fɔ* 'prep' is, amongst its many other uses, employed to mark the citation form of verbs (e.g. *fɔ rós* 'to burn', *fɔ espia* 'to spy on'). As a clause linker, *fɔ* introduces nominal, hence non-finite complements. Hence, when *fɔ* is used as a complementiser, the complement verb may not take an overtly expressed subject and the main and complement verbs have the same subject by default.

Some aspectual and modal verbs are characterised by variation in their occurrence with *fɔ*-complements. For instance, *grí* 'agree' and *tráy* 'try' are attested with the zero strategy of complementation and with complements introduced by *fɔ*. The modal verb *tráy* 'try' appears without the element *fɔ* in (30) and with it in (31):


As a complementiser, *fɔ* introduces the complements of aspectual and modal verbs that may not occur without an overt complementiser. One of these verbs is *lɛ́f* 'leave, stop to' (32), a verb that expresses the aspectual notion of cessation:

(32) *Mék* sbjv *e* 3sg.sbj *lɛ́f* leave *fɔ* prep *dríng.* drink 'She should leave drinking.' [ra07fn 033]

The element *fɔ* 'prep' also introduces the complements of a number of experiential and body state verbs, which are also not attested in any other type of construction. These verbs predetermine a simultaneous time reference of their complements. An example follows, in which *fɔ* introduces the complement of the experiential verb *sém* 'be ashamed' (33):

(33) *Náw* now *a* 1sg.sbj *dɔ́n* prf *de* ipfv *fínis* finish *sém* be.ashamed *fɔ* prep *wɛ́r* wear *dán* that *sús,* shoe *ɛf* if *a* 1sg.sbj *bin* pst *nó* know *a* 1sg.sbj *fɔ* cond *kɛ́r* carry *ɔ́da* other *sús.* shoe

'Now I am completely ashamed to be wearing those shoes, if I had known I would have brought another (pair of) shoes.' [ma03hm 021]

Furthermore, *fɔ* introduces complements of a number of verbs whose meaning contains an element of proposal, desire, evaluation, and similar affective nuances compatible with deontic modality. I regroup these verbs under the label "weak deontic". The deontic meaning of these verbs is also compatible with the modal meanings of *fɔ* itself

10 Clause linkage

(cf. §6.7.3.2). When main and complement verbs have the same subject, the complement clause may be introduced by *fɔ*. Compare the verbs *ísi* 'be easy' (34) and *grí* 'agree' (35):


Any weak deontic verb may alternatively take a subjunctive clause complement introduced by the subjunctive marker and modal complementiser *mék* 'sbjv' if the main verb is understood to induce a posterior time reference over the complement verb. For example, the complements of the weak deontic verb *mɛ́mba* 'remember to' may be introduced by *fɔ* 'prep' (36) or by *mék* 'sbjv' (37). In both sentences below, the main and complement clauses share the same subject. However, the subjunctive clauses is more finite – it requires an overt subject. In contrast, the use of a *fɔ*-complement does not permit the occurrence of an overt subject.


With weak deontic verbs, the subjunctive marker *mék* 'sbjv' may not only be employed instead of *fɔ* 'prep'. A subjunctive clause may also immediately follow *fɔ*. Hence all weak deontic verbs may feature the complementiser series *fɔ mék* 'prep sbjv' as in (38) below:

(38) *So* so *wé* sub *yu* 2sg *dɔ́n* prf *lán* learn *yu* 2sg *lángwech* language *ɛ́n,* intj *e* 3sg.sbj *dɔ́n* prf *hád* hard *fɔ* prep *mék* sbjv *yu* 2sg *lán* learn *Panyá.* Spanish 'So when you've learned your (home) language, it is hard for you to learn

Spanish.' [to03gm 020]

The use of subjunctive complement clauses is, however, required with weak deontic verbs whenever the main and complement clauses do not have the subject in common. Compare (31) above with (39) below. Both sentences feature the main verb *tráy* 'try':

10.5 Complementation

(39) (*…*) *a* 1sg.sbj *go* pot *tráy* try *mék* sbjv *e* 3sg.sbj *báy* buy *mí* 1sg.indp *dán* that *káyn* kind *gafas* glasses *por* by *dios.* God '(…) I will try that she buys me that kind of glasses, by God.' [ye07ga 003]

A subjunctive complement is also necessary if the complement verb is negated. This is so because non-finite verbs – including those that appear in *fɔ*-complements – are not normally negated in Pichi. Compare the negated complement clause introduced by *mék* 'sbjv' in (40) with the affirmative complement clause introduced by *fɔ* 'prep' in (36) above. Both sentences involve the main verb *mɛ́mba* 'remember':

(40) *Na* foc *ín* 3sg.indp *a* 1sg.sbj *mɛ́mba* remember *mék* sbjv *a* 1sg.sbj *nó* neg *gó* go *dé.* there 'That's when I remembered not to go there.' [bo05fn 021]

The evaluative verbs *fáyn* 'be fine', *hád* 'be hard', *ísi* 'be easy', *bád* 'be bad', and *gúd* 'be good' may be followed by a *fɔ*-complement when the subject of the main clause is expletive, i.e. refers to no specific person or entity as in (41). Complements of evaluative main verbs with expletive subjects function as the notional subject of the main clause:

(41) *E* 3sg.sbj *fáyn* fine *fɔ* prep *dríng* drink *smɔ́l-wán.* small-adv 'It's good to drink little.' [ma03hm 071]

Once the complement situation has a fully referential subject (which is necessarily not co-referential with the expletive subject of the main clause), a subjunctive complement clause is required (42):

(42) *Wé* sub *yu* 2sg *de* ipfv *dríng,* drink *e* 3sg.sbj *dé* be.loc *fáyn* fine *sé* qot *mék* sbjv *yu* 2sg *nó* neg *chák.* get.drunk 'When you drink, it's good not to get drunk.' [ur07fn 288]

Note the presence of the quotative marker and general complementiser *sé* 'qot' in (42) above. Any subjunctive clause may additionally be preceded by *sé* (cf. §10.5.5):

### **10.5.5** *Mék* **'sbjv'**

Verbs expressing the strong deontic notion of manipulation are only attested with subjunctive complements introduced by *mék* 'sbjv'. These complements have a dependent time reference; the complement situation is always posterior to that of the main verb. Strong deontic verbs invariably express a strong degree of manipulation, a notion that is compatible with the use of subjunctive mood in directives (cf. §6.7.3.3). Below follow subjunctive complements of the manipulative verbs *fɔ́s* 'force' (43) and *tún* 'tune, persuade' (44):

10 Clause linkage


The class of manipulative complement-taking verbs also includes the verbs *mék* 'make, cause to' (45) and *lɛ́f* 'leave, permit' in their respective functions as causative and permissive-causative verbs (cf. §9.4.4):

(45) *E* 3sg.sbj *mék* make *mék* sbjv *in* 3sg.poss *húman* woman *dríng* drink *di* def *cerveza.* beer 'He made his woman drink the beer.' [dj05be 001]

The manipulative verb of desire *wánt* 'want' is very versatile. It may occur in same subject complement clauses (cf. e.g. 113) or may take complements featuring the imperfective marker *de* 'ipfv' (cf. e.g. 27). When the subjects of the main and complement clauses are not the same, a subjunctive *mék*-complement is required:

(46) *E* 3sg.sbj *nák* hit *di* def *plét* plate *pan* pan *di* def *tébul* table *bikɔs* because *e* 3sg.sbj *wánt* want *mék* sbjv *di* def *plét* plate *brók.* break

'He hit the plate on the table because he wanted the plate to break.' [au07se 194]

Speech verbs employed as verbs of ordering and manipulation always take subjunctive complements. Examples of such indirect imperatives are provided in the following two sentences involving the verbs *tɛ́l* 'tell (to)' (47) and *bɛ́g* 'ask to' (48):


In a few instances in the data, the complements of strong deontic verbs are not introduced by *mék* 'sbjv'; the subjunctive marker is absent. I give two examples featuring the main verbs *lɛ́f* 'leave, permit' (49) and *wánt* 'want' (50):

(49) *Na* foc *ín* 3sg.indp *mi* 1sg.poss *gran-má* grand-ma *bin* pst *kán* come *tɔ́k* talk *sé* qot *in* 3sg.indp *nó* neg *go* pot *lɛ́f* leave *mi* 1sg.indp *a* 1sg.sbj *gó.* go 'That's when my grandma said that she [emp] wouldn't let me go.' [fr03ft 078]

10.5 Complementation

(50) *Sé* qot *ín* 3sg.indp *nó* neg *wánt* want *in* 3sg.poss *abuelo* grandfather *skrách=an.* scratch=3sg.obj '(He) said, he [emp] didn't want his grandfather to scratch him.' [ab03ab 042]

Notwithstanding the absence of the subjunctive marker, I analyse the clauses in bold in (49) and (50) above as subjunctive clauses. Evidence comes from the reduced TMA marking that characterises these clauses. Although both subordinate clauses are futurereferring, they are not marked by *go* 'pot' as they would if they occurred in main clauses or clauses with independent time reference (e.g. in quotative clauses introduced by *sé* 'qot'). Instead, the subordinate verbs *gó* 'go' and *skrách* 'scratch' appear stripped of any TMA marking as do subjunctive complements introduced by *mék* 'sbjv'.

This shows that the reduction of TMA marking, or "deranking" (Stassen 1985: 76–86; cf. also Cristofaro 2003) of the subjunctive subordinate clause is just as much a diagnostic of subjunctive mood as is the presence of the modal complementiser *mék* 'sbjv'.

### **10.5.6** *Sé* **'quot'**

We saw in the preceding two sections that the quotative marker *sé* 'qot' can optionally introduce any subjunctive complement featuring the modal complementiser *mék* 'sbjv'. This distribution is in line with the function of the quotative marker as a general complementiser.

The quotative marker *sé* 'qot' introduces the finite complement clauses of speech (51), cognition (52), and perception verbs (53). Complement clauses introduced by *sé* have independent time reference and are not reduced; they are finite and may occur with the full range of TMA marking as in the following examples:

(51) *Yɛ,* yeah *a* 1sg.sbj *kán* pfv *tɛ́l=an* tell=3sg.obj *sé* qot *'chica,* girl *mí* 1sg.indp *nó* neg *lɛ́k* like *yú* 2sg.indp *bɔt* but *wi* 1pl *fít* can *dé* be.loc *lɛk* like *kɔ́mpin'.* friend

'Yeah, I told her "girl, I don't love you but we can be like friends".' [ru03wt 029]


When *sé* 'quot' introduces the complements of speech verbs, the difference between direct and indirect speech hinges on pronominal reference. For instance, the sentence in quotes in (51) above is a direct speech complement of *tɛ́l* 'tell', because reference to *chica*

#### 10 Clause linkage

'girl' switches from *=an* '3sg.obj' in the main clause to the object pronoun *yú* '2sg.indp' in the complement clause.

With cognition and perception main verbs, the perceived situation can also be expressed as an adverbial time clause introduced by *sé* 'qot' (54) (cf. also 115 further below) or *wé* 'sub' (cf. 109), and an adverbial time clause introduced by *lɛk háw* 'the way that' (54). The adverbial clause is marked for imperfective aspect, since it is simultaneous with the main clause situation:

(54) *A* 1sg.sbj *de* ipfv *hía* hear *ín* 3sg.indp *sé* qot *e* 3sg.sbj *de* ipfv *nák* hit *di* def *gitá* guitar *ɔ* or *a* 1sg.sbj *de* ipfv *hía* hear *ín* 3sg.indp *lɛk* like *háw* how *e* 3sg.sbj *de* ipfv *nák* hit *di* def *gita.* guitar

*Lit*. 'I hear him that he's playing the guitar.' or 'I hear him how he's playing the guitar.' [dj05ae 053]

The quotative marker also introduces the complements of copula verbs in statements of facts. In such factive clauses, the copula verb takes a dummy noun like *tín* 'thing', *kés* 'matter', or the expletive subject pronoun *e* '3sg.sbj'. Factive main clauses like the one in (55) are very common as introductory formulas in narrative discourse (cf. also 98):

(55) *E* 3sg.sbj *dé* be.loc *sé* qot *dán* that *gál* girl *e* 3sg.sbj *bin* pst *de* ipfv *kán* come *yá.* here 'It's that/it came to pass that that girl used to come here.' [ru03wt 019]

Evaluative verbs like *fáyn* 'be fine', *gúd* 'be good', or *bád* 'be bad' can induce either an indicative or a subjunctive mood over their complements. Evaluative verbs are followed by indicative complements when these are intended to convey factual information about present or past situations (56):

(56) *E* 3sg.sbj *fáyn* fine *sé* qot *e* 3sg.sbj *kán* come *yɛ́stadé.* yesterday 'It's good that he came yesterday.' [dj07ae 260]

A subjunctive complement (albeit with the usual optional *sé* 'qot') is required when the evaluative main verb refers to a potential situation (57). By expressing a preference, it harmonises with the deontic sense associated with the subjunctive mood in Pichi:

(57) *E* 3sg.sbj *fáyn* fine *sé* qot *mék* sbjv *e* 3sg.sbj *kán* come *tumɔ́ro.* tomorrow 'It's good for him to come tomorrow.' [dj07ae 257]

Interrogative complements of speech, cognition, and perception verbs are no different from headless, free relative clauses and are covered in §10.6.5.

10.5 Complementation

### **10.5.7** *Wé* **'sub'**

The multifunctional linker *wé* 'sub' is employed as a subordinator in relative clauses, an adverbial clause linker and a clausal coordinator. In a small minority of complement relations in the corpus, it is also used as a complementiser.

The *wé-*clause in (58) is a borderline case that may either be analysed as an adverbial clause, i.e. a modifying time clause, or a subject complement clause:

(58) *E* 3sg.sbj *dɔ́n* prf *sté,* be.long *a* 1sg.sbj *tínk* think *sé* qot *e* 3sg.sbj *dɔ́n* prf *sté* be.long *wé* sub *una* 2pl *bin* pst *gɛ́t* get *insecticida* insecticide *yá.* here

'It's long ago, I think that it's long ago that you people had insecticide here/ when you people last had insecticide here.' [fr03wt 060]

The same holds for the *wé-*clause in (59), which can be interpreted as the complement clause of *hía* 'hear' or an indirect interrogative clause, although the presence of a subsequent *sé-*complement clause favours the latter interpretation:

(59) *Yu* 2sg *nó* neg *hía* hear *wé* sub *a* 1sg.sbj *tɛ́l* tell *Mario* name *sé* qot *quiero* I.want *cocinar?* cook 'You didn't hear that I told Mario that I want to cook?' or 'You didn't hear when I told Mario that I want to cook?' [ye03cd 124]

In turn, the complement status of the *wé-*clause in (60) featuring the experiential main predicate *sɔ́ri* 'feel sorry' is unequivocal. The functional equivalence of *wé* and the general complementiser*sé* in such complement clauses is illustrated by way of the analogous example in (61). However, the data contains no examples of *wé*-complement clauses to speech verbs:


The following example involving *wé* 'sub' is also a straightforward case of complementation involving an experiential main predicate:

(62) *Tɛnk* thank *gɔ́d* God *wé* sub *yu* 2sg *dɔ́n* prf *kán!* come 'Thank God that you have come!'

#### 10 Clause linkage

### **10.5.8 Complements of nouns**

The elements*fɔ*'prep' and *sé* 'qot' may also introduce purposive complements of nouns and modify a head noun in a way very similar to a relative clause. Below, *fɔ*introduces the non-finite (hence nominal) complement *pás* 'pass' of the head noun *sáy* 'place'. The same function may be fulfilled by *sé* 'qot'. In the second half of the, the *sé*-clause attributes a finite complement clause to the head noun *sáy* 'place', and thereby, introduces a *quasi* relative clause:

(63) *E* 3sg.sbj *gɛ́t* get *ɔ́da* other *sáy* side *fɔ* prep *pás,* pass *bɔt* but *a* 1sg.sbj *de* ipfv *fɛ́n* look for *di* def *sáy* side *sé* qot *yu* 2sg *nó* neg *go* pot *gɛ́t* get *hambɔ́g* irritation *fɔ* prep *pípul* people *dɛn.* pl 'There is another place to pass (through), but I am looking for the place where you wouldn't be bothered by people.' [ma03ni 009]

In the first example below, a *sé-*clause specifies the matter of the abstract noun *fúlis* 'foolishness'. In the second example, the anaphoric demonstrative pronominal *dá wán* 'that one' is modified by a subjunctive marked purpose clause introduced by *sé* 'qot':

(64) *A* 1sg.sbj *sé* qot *bikɔs* because *una* 2pl *Camerún,* place *una* 2pl *gɛ́t* get *di* def *fúlis* foolishness *sé,* qot *wé* sub *náw* now *wé* sub *yu* 2sg *ték=an,* take=3sg.obj *yu* 2sg *go* pot *sɛ́l=an.* sell=3sg.obj

'I say because you Cameroonians, you have the foolish habit that, when now, when you take it, you will sell it.' [ab03ay 151]

(65) *Yu* 2sg *trowé=an,* pour=3sg.obj *yu* 2sg *pút* put *ɔ́da* other *nyú* new *wán* one *ínsay,* inside *dá* that *wán* one *sé* qot *mék* sbjv *e* 3sg.sbj *nó* neg *simɛ́l.*

smell

'You pour it away, (then) you put another new one [water] inside, that (is) so that it does not smell.' [dj03do 048]

### **10.6 Relativisation**

In Pichi, subjects, objects, and PPs, as well as possessor and possessed nouns may be relativised. The most common means of forming relative clauses involves the use of the morphologically invariant subordinator *wé* 'sub' as a relative clause linker. Next to *wé* 'sub', the linkers *sé* 'qot' and *fɔ* 'prep' marginally fulfil the function of relative clause linkers when they introduce noun complements (cf. §10.5.8).

In the second strategy of relative clause formation, no relative clause linker is employed and the relative clause simply follows the main clause. Hence, there is a "gap" between the two clauses. However, resumptive pronouns may optionally refer back to the

10.6 Relativisation

relativised head noun in most types of relative clauses. Aside from that, restrictive and non-restrictive relative clauses are not systematically distinguished on formal grounds.

The use of resumptive pronouns is nearly general in subject relative clauses with [+specific] head nouns, fairly common in object relative clauses, and rare in the relativisation of PPs. The frequency of resumptive pronouns with subject relative clauses runs counter to the predictions of the relativisation accessibility hierarchy (cf. Keenan & Comrie 1977), and it should be worthwhile investigating whether it constitutes an areal West African phenomenon (see, however, a similar distribution of resumptive pronouns in Tok Pisin (Sankoff & Brown 1976) and popular Brazilian Portuguese (Tarallo 1983).

In the example sentences in this section, relative clauses are set in squared brackets. Table 10.3 summarises important features of the different types of relative clauses that Pichi has (RC = relative clause).


Table 10.3 Features of relative clauses

### **10.6.1 General characteristics**

The linker *wé* 'sub' introduces relative clauses as well as adverbial and coordinate clauses. Since the use of resumptive subject pronouns is very common (but still optional) in subject relative clauses with [+specific] head nouns (cf. Table 10.3 above), some subject relative clauses may therefore have the same constituent order as an adverbial clause introduced by *wé* 'sub'. Consider the alternative relative and adverbial translations I provide for sentence (66):

(66) *Ɔ́l* all *dí* this *mán* man *dɛn* pl *[wé* sub *dɛn* 3pl *gɛ́t* get *mɔní],* money *na* foc *di* def *tín* thing *wé* sub *dɛn* 3pl *de* ipfv *mék.* make 'All these men who have money, that's what they do.' or 'All these men, when they have money, that's what they do.' [ed03sb 133]

#### 10 Clause linkage

However, the meaning of the sentence above is not as ambiguous as it may appear. Relative clauses are never separated from their main clauses by a prosodic break; relative constructions form single prosodic units. In contrast, adverbial clauses are very often separated from their main clauses by a prosodic break: The main clause bears continuative intonation, and the subordinate clause is separated from the main clause by a pause. The adverbial clause then begins with the high pitch onset that is characteristic for independent utterances (cf. also §3.4.4). An adverbial interpretation of the clause introduced by *wé* 'sub' in (66) above would therefore only be possible if a comma were inserted between *mán dɛn* 'men' and *wé* 'sub'.

In contrast, pronoun resumption, even if possible, is not very often seen in object relative clauses, even if the head noun is [+specific]. In the object relative clause below, *gɛ́t* 'get' is not followed by an object pronoun co-referential with the head noun *mɔní*:

(67) *Mék* sbjv *e* 3sg.sbj *bák* give.back *yú* 2sg.indp *di* def *mɔní* money *[wé* sub *e* 3sg.sbj *gɛ́t].* get 'Let him give you back the money that he got.' [fr03cd 027]

The possibility of abstaining from pronoun resumption in Pichi relative clauses, such as (67) (for a subject relative clause without a resumptive pronoun, cf. 78 below) and the prosodic unity of relative constructions are good arguments for viewing relative clauses as embedded clauses.

Relative clauses always follow the head NP that they refer to. The head NP and its relative clause can be separated by quantifiers (68), as well as topic and focus particles (69). The examples in this section and the following ones also show that TMA and person marking in relative clauses is "balanced" (Stassen 1985); hence it is not reduced in comparison with that of declarative clauses:


Headed restrictive and non-restrictive relative clauses cannot be distinguished on formal grounds. In (70), the commas in squared brackets in the translation indicate the nonrestrictive alternative interpretation of the sentence. Note the presence of the English loan *apart from* in this example:

(70) *Apart* apart *from* from *mi* 1sg.poss *antí* aunt *[wé* sub *e* 3sg.sbj *dé* be.loc *yá],* here *ɔ* or *di* def *pikín* child *dɛn* pl *fɔ* prep *mi* 1sg.poss *gran-má* grand-ma *wet* with *mi* 1sg.poss *gran-pá* grand-pa *[wé* sub *dɛn* 3pl *stíl* still *dé* be.loc *láyf],* life *dɛn-ɔ́l* 3pl.indp.cpd-all *dɛn* 3pl

10.6 Relativisation

*dé* be.loc *na* foc *Panyá.* Spain

'Apart from my aunt [,] who is here, or the children of my grandmother and grandfather [,] who are still alive, they are all in Spain.' [fr03ft 038]

If the head noun has plural reference, the pluraliser *dɛn* 'pl' appears immediately after the head noun and before the subordinator *wé* 'sub' (71). Note the presence of the resumptive subject pronoun *dɛn* '3pl' in the relative clause, which is co-referential with the head noun *tín dɛn* 'things':

(71) *Porque* because *dán* that *tín* thing *na* foc *tín* thing *dɛn* pl *[wé* sub *dɛn* 3pl *dɔ́n* prf *sté* stay *dán* that *tɛ́n* time *dɛn* pl *wé* sub *esclavitud* slavery *dé].*

be.loc

'Because those are things that have stayed (from) those times when there was slavery.' [hi03cb 228]

Pichi exhibits generous possibilities of relative clause formation (cf. Keenan 1985: 148). For example, the relativisation of a nominal that is part of a coordinate structure is permitted (72). Equally, a relative clause may contain a focused resumptive pronoun (73):

(72) *Bikɔs* because *mí* 1sg.indp *dé* be.loc *sɔn* some *stáyl,* style *layk* like *dán* that *gɛ́l* girl *[wé* sub *mí* 1sg.indp *wet=an* with=3sg.obj *bin* pst *gó* go *dé],* there *a* 1sg.sbj *tɛ́l=an* tell=3sg.obj *sé a* 1sg.sbj *wɔ́nt* want *sí* see *háw* how *dɛn* 3pl *de* ipfv *mék.* make 'Because I was (feeling) a way, like that girl with whom I went there, I told her that I wanted to see how it is done.' [ed03sb 149]

(73) *Bɔt* but *di* def *pé* pay *wé* sub *yu* 2sg *gɛ́fɔ* have.to *pé,* pay *if* if *yu* 2sg *nó* neg *de* ipfv *gí* give *mí* 1sg.indp *yu* 2sg *fɔ́s* first *mán* man *[wé* sub *na* foc *in* 3sg.indp *gí* give *yú* 2sg.indp *dí* this *bɛlɛ́],* belly *yu* 2sg *de* ipfv *gí* give *mi* 1sg.indp *di* def *pikín* child *[wé* sub *de* ipfv *kɔmɔ́t].* come.out

*Lit*. 'But the price that you have to pay (is), if you don't give me your first man, who it is him who gave you the first pregnancy, you will give me the child that will come out.' [ed03sb 020]

Likewise, there is no restriction on the relativisation of the subject or object of a complement clause (74) or of an indirect question clause (75):

(74) *Na* foc *dán* that *bɔ́y* boy *[wé* sub *a* 1sg.sbj *tɛ́l* tell *yú* 2sg.indp *sé* qot *in* 3sg.poss *mamá* mother *dɔ́n* prf *gó* go *na* loc *Panyá].* Spain 'It is that boy (of) who I told that his mother has gone to Spain.' [he07fn 253]

10 Clause linkage

(75) *A* 1sg.sbj *gɛ́t* get *sɔn* some *kɔ́mpin,* friend *sɔn* some *Ghana-mán* Ghana.cpd-man *[wé* sub *a* 1sg.sbj *nó* neg *sabí* know *ús=sáy* q=side *dán* that *mán* man *dé].* be.loc

*Lit*. 'I have a friend, a Ghanaian who I don't know where that man is.' [ed03sb 188]

Relative constructions are also made use of to express adverbial relations of time, location, and manner through the relativisation of generic nouns like *áwa* 'time, hour', *tɛ́n* 'time' (119), *dé* 'day' (120) and *stáyl* 'manner, style (§10.7.4).

### **10.6.2 Subjects and objects**

Subject relative clauses normally feature a resumptive subject pronoun that is co-referential with the [+specific] relativised noun. Relative clauses featuring a relativised subject pronoun also usually contain a resumptive pronoun if the head NP is not a 3rd person pronoun (77):


Sentence (77) exemplifies the relativisation of subjectNPs without resumptive pronominal marking. Although the head nouns *gabonés* and *guineano* are preceded by the definite article *di* 'def', these nouns have [-specific], generic reference, hence they are not reiterated by a resumptive subject pronoun in the relative clause (cf. also §5.1.4):

(78) *Pero* but *di* def *gabonés* Gabonese *[wé* sub *de* ipfv *tɔ́k* talk *Bata]* Fang *wet* with *di* def *guineano* Guinean *[wé* sub *de* ipfv *tɔ́k* talk *Bata],* Fang *di* def *sonido* sound *nó* neg *dé* be.loc *di* def *sén.* same

'But the Gabonese who talks Fang and the Guinean who talks Fang, the sound is not the same.' [ma03hm 048]

"Gap" subject relative constructions without the subordinator are not attested. However, object relative clauses formed by means of the gap strategy are sometimes heard. The relativised cognate object in (79) is a patient object. Note the absence of the subordinator *wé* 'sub' as well as that of a resumptive object pronoun in the relative clause after the verb *wánt* 'want':

10.6 Relativisation

(79) *Mék* sbjv *e* 3sg.sbj *bít* beat *yú,* 2sg.indp *mék* sbjv *e* 3sg.sbj *dú* do *yú* 2sg *di* def *dú* do *[e* 3sg.sbj *wánt]* want (*…*) 'Let him beat you, let him do to you [what he wants] (…)' [bo03cb 135]

Object relative clauses involve the use of the subordinator *wé* 'sub' in the vast majority of cases. Take note of the absence of a resumptive object pronoun with reference to the non-specific head noun *bloques dɛn* 'blocks':

(80) *Sɔn* some *bloques* block.pl *dɛn* pl *lɛ́f* remain *[wé* sub *dɛn* 3pl *gɛ́fɔ* have.to *monta]* mount *nɔ́?* intj 'Some blocks remain that have to be mounted, right?' [ye03cd 114]

A resumptive pronoun may also refer to a recipient head noun in a double-object construction (81). Recipient resumptive pronouns are optional and may therefore be omitted as in (82):


### **10.6.3 Prepositional phrases**

There are no formal constraints on the relativisation of PPs. However, this type of relativisation is rather rare compared to that of subjects and objects. The following relative constructions involve relativised prepositional phrases introduced by the prepositions *fɔ* 'prep' and *pan* 'on'. These two prepositions, as well as the preposition *wet* 'with', can also be stranded, in other words they may remain in their original position, while the relativised NP appears at the beginning of the sentence. Pied-piping of prepositions, i.e. the appearance of the preposition at the beginning of the relative clause, is not attested:

(83) *Di* def *béd* bed *[wé* sub *e* 3sg.sbj *de* ipfv *slíp* sleep *pan],* on *e* 3sg.sbj *dé* be.loc *na* loc *di* def *rúm.* room 'The bed that she sleeps on, it's in the room.' [tr05fn 047]

In the more common alternative to stranding, a resumptive pronoun fills the original position of the relativised noun. Compare *wet=an* 'with her' in (72) above. Alternatively, a resumptive pronoun need not be used at all. The exact meaning of the sentence is then provided by pragmatic context. In such instances of "prepositional phrase chopping" (Tarallo 1983; 1985) disambiguation is left to pragmatic context.

In (84), there is no *wet* 'with' in the relative clause to point to the semantic role of instrument of the relativised head noun *gɔ́n* 'gun':

#### 10 Clause linkage

(84) *Dɛn* 3pl *de* ipfv *gó* go *wet* with *dán* that *gɔ́n* gun *[wé* sub *dɛn* 3pl *de* ipfv *kíl* kill *bíf]* wild.animal *ɔ* or *pistola.* pistol 'They go with that gun which they kill wild animals (with) or a pistol.' [ed03sb 114]

Similarly, the *wé-*clause in (85) induces a locative, that in (86b) an instrumental interpretation. It is also of interest that (86b) is an example for the use of *fɔ* as an introducer of a noun complement that is very similar in function to the preceding relative clause (cf. §10.5.8):

	- b. *Dán* that *tín* thing *[wé* sub *dɛn* 3pl *de* ipfv *mék* make *súp],* soup *fɔ* prep *mék* make *fufú.* fufu 'That thing they make soup (with), in order to make fufu (with).' [dj05ce 099]

In a similar vein, the *wé-*clauses in (87) and (88) allow that a causal meaning is inferred:

(87) *So* so *na* foc *di* def *tín* thing *[wé* sub *e* 3sg.sbj *rɔ́n],* run *e* 3sg.sbj *kɔmɔ́t.* go.out 'So that is why [*lit*. the thing that] she fled, (and) she left.' [ed03sb 041]

(88) (*…*) *e* 3sg.sbj *go* pot *sé* qot *e* 3sg.sbj *de* ipfv *fíɛ* fear *e* 3sg.sbj *nó* neg *go* pot *gí* give *mí* 1sg.indp *di* def *tín* thing *[wé* sub *a* 1sg.sbj *de* ipfv *sɛ́n=an].* send=3sg.obj '(…) he would say he is afraid, he would not give me the thing that I had sent him (for).' [ab03ab 041]

Such constructions are structurally no different from those involving objects, and, like the latter, they may involve "gap" constructions. Note the absence of the subordinator *wé* 'sub' in the following example. The head noun of the relative clause *sáy* 'side, place' is the syntactic object of *sidɔ́n* 'sit (down), stay':

(89) *A* 1sg.sbj *de* ipfv *gó* go *nía* near *di* def *sáy* side *[Paquita* name *sidɔ́n].* stay 'I am going near where Paquita stays.' [dj05be 147]

10.6 Relativisation

Prepositional phrase chopping should be differentiated from instances in which the goal of a verb may be expressed as an object, as is the case in double-object constructions involving *pút* 'put' in (90) (cf. §9.3.4 for more details). Once more, note the occurrence of a "gap" relative clause in this example:

(90) *A* 1sg.sbj *ték* take *tú* two *peso* peso *a* 1sg.sbj *báy* buy *dán* that *dís-tín,* this-thing *sɔn* some *smɔ́l* small *pépa* paper *[dɛn* 3pl *de* ipfv *pút=an* put=3sg.obj *cacahuete].* groundnut 'I took two pesos (and) I bought this whatsit, a small paper (into which)

groundnuts are put.' [ed03sp 083]

Example (91) shows how the resumption of the entire relativised noun in the position of relativisation can be an alternative to stranding or chopping. Anaphoric NP reiteration is accompanied by a deictic element, the demonstrative *dís* 'this' in (91):

(91) *Bikɔs* because *wán* one *isla* island *dé* be.loc *[wé* sub *e* 3sg.sbj *fíba* seem *sé* qot *petroleo* oil *dé* be.loc *na* loc *dís* this *isla].* island

'Because there is an island of which it seems that there is oil on this island.' [fr03ft 109]

In sentence (92) below, the direct object*sɔn fáyn* 'a beauty' is resumed through another full NP, namely the demonstrative pronominal *dá wán* 'that (one)':

	- one

'I say black girls have a beauty which white women do not have (that one).' [ed03sp 046]

Full NP anaphora can also be observed in the complex relative construction in (75) above, where *dán mán* 'that man' in the relative clause refers to the head noun *Ghanamán* 'Ghanaian'. All these structures are reminiscent of correlative constructions found in other languages and demonstrate the diversity of relativisation strategies in Pichi.

### **10.6.4 Possessors**

When a possessor noun is relativised, a co-referential possessive pronoun and the possessed noun immediately follow the subordinator *wé* 'sub' (93):

10 Clause linkage

(93) (*…*) *dɛn* 3pl *de* ipfv *kɔmɔ́t* hail.from *na* loc *wán* one *pueblo* village *[wé* sub *in* 3sg.poss *ném* name *na* foc *Basakato dé la Sagrada Familia].* place '(…) they come from a village whose name is Basakato dé la Sagrada Familia.' [fr03ft 042]

The preceding example features a possessor head noun that functions as the subject of the relative clause. When the possessor head noun functions as the object of the relative clause, it is relativised by way of a structure in which the head noun and the relative clause function as the topic. The remainder of the main clause functions as the comment, and is set off from the topic by a pause, while a possessive pronoun anaphorically refers to the head noun (94):

(94) *Dán* that *húman* woman *[wé* sub *a* 1sg.sbj *só* show *yú],* 2sg.indp *in* 3sg.poss *motó* car *dé* be.loc *na* loc *strít.* street 'That woman which I showed you, her car is in the street.' [dj05ae 068]

Possessed nouns are relativised like core participants. Reference is upheld due to the juxtaposition of the possessed noun and the relative clause (95):

(95) *A* 1sg.sbj *ték* take *di* def *stík* tree *in* 3sg.poss *kandá* bark *[wé* sub *a* 1sg.sbj *sí* see *dé],* there *a* 1sg.sbj *rós=an.* burn=3sg.obj 'I took the bark<sup>i</sup> of the tree<sup>j</sup> that<sup>i</sup> saw there, I burnt it<sup>i</sup> .' [bo05n 001]

### **10.6.5 Free relatives and indirect questions**

Free relative clauses do not feature an overt head noun and are introduced by a question word. In free relative constructions featuring question words, the relative clause is formally identical with the corresponding content question (cf. §7.3.2). The subordinator *wé* 'sub' is not employed to introduce free relative clauses. Free relative clauses often function as objects of verbs of cognition, perception, asking, or speaking.

Below, we find a free subject relative clause, which is introduced by the question word *wétin* 'what':

(96) *A* 1sg.sbj *dɔ́n* prf *tɛ́l* tell *yú* 2sg.indp *[wétin* what *pás* pass *na* loc *nɛ́t],* night *dán* that *nɛ́t.* night 'I've already told you what happened in the night, that night.' [ab03ab 043]

Free relatives introduced by the question words *údat* 'who', *ús=mán* 'who', and *ús=pɔ́sin* 'who' question human referents. The following two examples are free object relative clauses:

(97) *Dɛn* 3pl *nó* neg *nó* know *[údat* who *hambɔ́g=an].* bother=3sg.obj 'They don't know who disturbed her.' [dj05ce 127]

10.7 Adverbial relations

(98) *Mí* 1sg.indp *nó* neg *sabí* know *[ús=mán* q=man *dɛn* 3pl *kíl],* kill *a* 1sg.sbj *nɔ́ba* neg.prf *hía* hear *dán* that *torí* story *sɛ́f.* emp 'I don't know which man they killed, I haven't even heard that story.' [ro05de 049]

The corresponding question words also introduce the free variants of relative clauses with generic head nouns like *tɛ́n* 'time' and *sáy* 'side' which function as adverbial clauses of time and place. Compare (99).

(99) *E* 3sg.sbj *nɛ́a* neg.prf *tɛ́l* tell *mí* 1sg.indp *[ús=tɛ́n* q=time *e* 3sg.sbj *go* pot *rích* arrive *dé].* there 'He hasn't told me when he is going to arrive there.' [eb07fn 582]

The question word *háw* 'how' introduces free relatives and indirect questions that question a property (100), quantity, or degree; the latter two in the collocation *háw mɔ́ch* 'how much' (101):


Indirect yes-no question clauses may be introduced by the clause linker *ɛf(ɛ)* or *if* 'if' which then functions as a complementiser in combination with sentence-final question intonation. Alternatively, such question clauses may be introduced by *sé* 'qot' if phrased as a question in the type of direct speech that characterises the use of quotative *sé* 'qot' in many contexts:


### **10.7 Adverbial relations**

The clause linkers *wé* 'sub' and *sé* 'qot' together have the potential to participate in the expression of most types of adverbial relations that we find in Pichi. Additionally, Pichi features an array of adverbial clause linkers with more specific meanings. These are

#### 10 Clause linkage

summarised in Table 10.4 below. The following sections provide an overview of adverbial clause formation in Pichi. Purpose clauses are covered in §10.7.6.

The first column in Table 10.4 below provides an overview of the types of adverbial clauses attested. The second column contains the linkers that introduce these types of clauses in Pichi. Alternative means of formation are given in the remaining three columns: The third column indicates whether a clause introduced by *wé* 'sub' or*sé* 'qot' can be used instead of the linker in the second column in order to express the same adverbial relation.

The fourth column provides other alternatives for expressing the corresponding adverbial relation. Independent sentences may also be linked through adverbials. These are contained in the last column on the right. A blank space indicates that the corresponding means is not available.

### **10.7.1** *Wé* **'sub'**

The subordinator *wé* 'sub' may introduce adverbial clauses of time, condition, and concession. Although *wé* is most commonly used to express temporal relations the other uses are frequent as well. A *wé*-clause may precede (104) or follow (cf. 110 below) its main clause and is often set off from preceding and following material by a prosodic break (cf. also §10.6.1). In this function, *wé* is best translated as 'when':

(104) *Wé* sub *a* 1sg.sbj *go* pot *fínis* finish *skúl,* school *a* 1sg.sbj *go* pot *tɔ́n* turn *dɔ́kta.* doctor

'When I finish school, I'll become a doctor.' [ro05ee 023]

The expression of time relations by means of *wé*-clauses cannot be divorced from the function of *wé* 'sub' of introducing sequences of coordinate clauses. Compare the time clause in (104) with the multiple occurrences of *wé* here:

(105) *Pero* but *wé* sub *a* 1sg.sbj *kán* pfv *mít* meet *dís* this *mán,* man *wé* sub *wi* 1pl *bigín* begin *bɔ́n* beget *in* 3sg.poss *yón* own *pikín* child *dɛn.* pl 'But then/when I met this man, and then we began to have his own children.' [ab03ab 214]

Time clauses introduced by *wé* are interpreted as being in a relation of temporal overlap with the main clause if both clauses contain imperfective readings (106) or are unspecified with respect to aspect like the two clauses in (104) above containing the potential mood marker *go* 'pot':

(106) *Wé* sub *e* 3sg.sbj *kin* hab *kɔmɔ́t* come.out *wók* work *a* 1sg.sbj *kin* hab *mék=an* make=3sg.obj *só,* like.that *lɛk* like *háw* how *mún* month *fínis.* finish

'When he leaves work, I do to him like this [stretches out hand in a gesture that indicates that her husband's salary should be handed over to her], as soon as the month is over.' [ro05rt 042]


Table 10.4Adverbial relations

10.7 Adverbial relations

#### 10 Clause linkage

The relation between a main clause and a dependent clause introduced by *wé* can also be one of temporal succession rather than overlap. The interpretation of the temporal relation between the clauses depends on the lexical aspect class of the verbs involved as well as on aspect-marking. For example, in (107) perfective marking with the dynamic verbs *rích* 'reach' and *sé* 'say, qot' implies succession, however brief the interval:

(107) *Wé* sub *a* 1sg.sbj *rích* reach *na* loc *hós* house *dé,* there *a* 1sg.sbj *sé* qot *'yu* 2sg *go* pot *tɛ́l* tell *mi* 1sg.indp *di* def *sáy* side *wé* sub *unu* 2pl *kin* hab *gó* go *mítɔp.'* meet 'When I reached the house, I said "you're going to tell me where you usually meet.' [ro05rt 018]

Temporal succession can be rendered more explicit through the use of the perfect marker *dɔ́n* 'prf' in the main or dependent clause. Hence, the main clause in (108) is posterior to the time clause introduced by *wé* 'sub':

(108) *A* 1sg.sbj *go* pot *firma* sign *wé* sub *a* 1sg.sbj *go* pot *dɔ́n* prf *chɔ́p.* eat 'I will sign when I have finished eating.' [ye03cd 038]

The boundary is fuzzy between temporal and other adverbial meanings of clauses introduced by *wé*. In (109), the temporal sense of the *wé*-clause shades off into a manner or circumstance sense. Context may also give rise to a concessive meaning of the subordinate clause (110):


sisters know, that will be difficult.' [ro05rt 034]

The relation between the first clause in (111a) and the clause introduced by *wé* is best interpreted as adversative. This is illustrated by the follow-up clause in (111b):

(111) a. *Yu* 2sg *nó* neg *bin* pst *dé* be.loc *na* loc *mákit,* market *wé* sub *a* 1sg.sbj *tɛ́l* tell *yú* 2sg.indp *sé* qot *mék* sbjv *yu* 2sg *bríng* bring *mi* 1sg.indp *watá?* water 'Weren't you at the market although I had told you to bring me water?' [ye0503e? 069]

10.7 Adverbial relations

b. *Wétin* what *yu* 2sg *kán* come *sin* without *watá?* water 'Why did you come without water?' [ye0503e? 070]

Finally, in (112b), we find two wholly independent clauses separated by an intonation break, with the second one being introduced by *wé*. The *wé-*clause is contrasted with the implicitly understood concessive proposition in squared brackets. Clause (112b) may be interpreted as being in a causal relationship to clause (a):

(112) a. *Sɔn* some *mamá* mother *dɛn,* pl *dɛn* 3pl *bád.* bad 'Some mothers, they are bad.' [ab03ay 109] b. *Wé* sub *yu* 2sg *pikín,* child *yu* 2sg *nó* neg *aconseja* advise *ín* 3sg.indp *frɔn* from *doce* twelve *años.* years 'Because as for your child, you didn't advise her from twelve years on.' [although you know about the dangers of early pregnancy].' [ab03ay 109]

The linker *wé* 'sub' is also encountered in the temporal source clause introducers *frɔn wé* and *síns wé,* both of which mean 'since' (cf §10.7.10).

### **10.7.2** *Sé* **'quot'**

The quotative marker *sé* 'qot' may provide adverbial modifications of purpose and result, cause, manner and circumstance, time and condition. The answer to (a) in (b) below can be interpreted as a cause clause. The *sé-*clause in this example once more vividly illustrates the diversity of meanings of *sé*, particularly in contexts like this one, where it straddles the boundary between quotation proper and other, related functions:

(113) a. *Wétin* what *yu* 2sg *de* ipfv *wét?* wait 'What [why] are you waiting?' [fr03wt 048] b. *Sé* qot *in* 3sg.poss *mamá* mother *go* pot *dráyb=an* drive=3sg.obj

'(He) says/because his mother will chase him away at first.' [dj03wt 049]

*fɔ́s.* first

The codemixed example (114) features a *sé*-clause that permits a temporal or conditional interpretation. These interpretations are favoured due to the sentence-initial position of the *sé*-clause.The sentence is also instructive because the speaker uses the Spanish temporal conjunction *cuando* 'when' in order to render Pichi *sé* 'qot' when reiterating the clause in Spanish:

#### 10 Clause linkage

(114) *"Yu* 2sg *hól* hold *wán* one *motó",* car *yu* 2sg *de* ipfv *dráyb=an,* drive=3sg.obj *pero* but *sé* qot *yu* 2sg *gɛ́t,* get *cuando* when *tienes,* you.get *"a gɛ́t wán motó".*

1sg.sbj get one car

' "Yú hól wán motó" (means) you're driving it, but if you possess it, when you have it "a gɛ́t wán motó".' [dj05ae 223]

A *sé*-clause that follows a main clause and is marked for temporal overlap with the main clause by means of imperfective aspect may function as a modification of manner or circumstance in the same way as a *wé*-clause. Compare (115) with (109) above:

(115) *Dɛn* 3pl *púl* remove *di* def *motó* car *na* loc *garaje* workshop *sé* qot *dɛn* 3pl *de* ipfv *pús=an.* push=3sg.obj 'They removed the car from the workshop by pushing it.' [pa05fn 024]

Such clauses also lend themselves to a concessive interpretation if suggested so by pragmatic context. Compare the concessive *wé*-clause in (110) with the following *sé*clause in (116):

(116) *E* 3sg.sbj *dú* do *di* def *ejercicio* exercise *sé* qot *e* 3sg.sbj *táya.* be.tired 'She did the exercise while/although she was tired.' [ra07ve 021]

Finally, *sé* is optionally attested with many adverbial clause linkers, among them *bikɔs* (*sé*) 'because'. *Sé* is obligatory when prepositions take clausal, rather than nominal complements, e.g. *fɔséko sé* 'due to, because', and *lɛk sé* 'as if' (117):

(117) *"A* 1sg.sbj *hól* hold *wán* one *motó"* car *na* foc *lɛk* like *sé* qot *yu* 2sg *de* ipfv *dráyb* drive *wé* sub *yu* 2sg *de* ipfv *wók.* work '"A hól wán motó" is like you drive (a car temporarily) while you work.' [dj05ae 225]

### **10.7.3 Time clauses**

I have shown that temporal relations between clauses may be established in various ways through the polyfunctional linker *wé* 'sub'. The following clause linkers express adverbial relations of time with more specific meanings.

Relative clauses featuring the generic head nouns *áwa* 'time', *tɛ́n* 'time', and *dé* 'day' function as time clauses. The nature of the temporal relation between the main and the relative clause situations is determined by lexical and clausal aspect marking:

(118) *Di* def *húman* woman *kán* come *na* loc *hós* house *di* def *áwa* hour *[wé* sub *a* 1sg.sbj *de* ipfv *kúk].* cook 'The woman came to the house when I was cooking.' [ro05de 022]

10.7 Adverbial relations


The clause-linker and collocation *lɛk háw* 'as soon as' introduces time clauses. Time clauses introduced by *lɛk háw* precede their main clauses and establish a relation of anteriority with the main clause.This linker may also introduce adverbial manner clauses (cf. §10.7.4 below):

(121) *Tumɔ́ro,* tomorrow *lɛk* like *háw* how *yu* 2sg *tɔ́k* talk *wet* with *Buehú,* name *yu* 2sg *kɔ́l* call *mí,* 1sg.indp (*…*) 'Tomorrow, as soon as you've talked to Buehu, you call me, (…)' [fr03cd 111]

The linker *bifó* 'before' introduces time clauses that are in a relation of posteriority to the main clause. *Bifó*-clauses are preferably sentence-initial, though they are also found in sentence-final position in after-thought apposition, as in (123):


It is interesting that the corpus contains no instance of an after-relation expressed by *áfta* 'after' in analogy with *bifó* in (123) above. Apparently, *áfta* may only serve as an 'and then' clausal connective and does not mean relational 'after'. Hence, after-relations must be constructed as iconical 'and then' relations with the proadverbial *áfta* as in (124):

(124) *Lɛ́f=an,* leave=3sg.obj *a* 1sg.sbj *go* pot *chɔ́p,* eat *áfta* then *a* 1sg.sbj *go* pot *dríng.* drink 'Leave it, I will eat, then I will drink.' [ye03cd 079]

Alternatively, the after-relation can be expressed by an initial *wé-*clause accompanied by perfect marking, as in (125):

10 Clause linkage

(125) *Sifta,* sift *wé* sub *a* 1sg.sbj *dɔ́n* prf *sifta* sift *ín,* 3sg.indp *e* 3sg.sbj *de* ipfv *lɛ́f* leave *wet* with *di* def *watá.* water 'Sift (it), when I have sifted it, it'll be left with the water.' [dj03do 007]

### **10.7.4 Manner clauses**

Manner clauses may be expressed through a relative construction featuring the generic head noun *stáyl* 'style, manner':

(126) *A* 1sg.sbj *bin* pst *chɔ́p* eat *di* def *plantí* plantain *di* def *stáyl* style *[wé* sub *pɔ́sin* person *dɛn* pl *fɔ* prep *Malábo* Malabo *dɛn* 3pl *de* ipfv *chɔ́p=an]* food=3sg.obj 'I ate the plantain the way Malabo people eat it.' [dj05ae 069]

Manner clauses may also be formed by way of adverbial clauses introduced by the collocation *lɛk háw* 'like how' = 'the way that'. Compare the near-identical sentence above with the two following ones:


'I don't know why they won't restore it the way it was back then.' [hi03cb 038]

Manner clauses introduced by *lɛk háw* 'like how' are also often employed to denote the perceived situation of a main clause verb of sensory perception like *hía* 'hear' (129), *sí* 'see', *lúk* 'look', *smɛ́l* 'smell' (130), and *fíl* 'feel'. Such clauses vacillate between readings denoting manner and temporal overlap:


The collocation *lɛk háw* also forms part of the idiomatic phrase *lɛk háw yu (de) sí X* (X referring a person) which means something along the lines of 'when looking at X you should also know'. Compare the following example:

10.7 Adverbial relations

(131) *Mí,* 1sg.indp *lɛk* like *háw* how *yu* 2sg *de* ipfv *sí* see *mí* 1sg.indp *a* 1sg.sbj *dɔ́n* prf *sí* see *plɛ́nte* plenty *tín.* thing '(As for) me, when you looking at me you should also know that I have seen many things [in life].' [ab03ab 023]

Manner clauses introduced by *lɛk háw* may shade off into a temporal reading and viceversa. Manner clauses generally follow their main clauses as in the preceding examples. In contrast, time clauses introduced by *lɛk háw* normally precede their main clauses (cf. 121 above in the previous section).

However, we also sometimes find manner clauses introducecd by *lɛk háw* in a sentenceinitial, topical position. When such a clause is marked for an imperfective reading, it is likely to be interpreted as a manner clause. *Lɛk háw* then means 'the way that' (132):

(132) *Lɛk* like *háw* how *e* 3sg.sbj *de* ipfv *wáka,* walk *e* 3sg.sbj *butú,* stoop *e* 3sg.sbj *nó* neg *bɛ́n.* bend 'The way he's walking (now), he's stooped over, he's not bent over.' [au07se 082]

On the other hand, if a sentence-initial clause introduced by *lɛk háw* is marked for a perfective reading, it is very likely to be interpreted as a time clause. *Lɛk háw* then translates as 'as soon as'. In (133), the subordinate clause contains the factative marked (hence perfective) dynamic verb *pút* 'put'. Compare the temporal interpretation of this sentence with the manner reading of (132) above. Also compare the temporal interpretation of the factative-marked verb *pút* 'put' in the previous section in (121) above:

(133) *Lɛk* like *háw* how *e* 3sg.sbj *pút* put *dán* that *mɔní* money *na* loc *mi* 1sg.poss *hán,* hand *nó* neg *wét* wait *mɔ́!* more

'As soon as he has put that money into my hand, no time to waste!' [ro05rt 043]

If a manner interpretation is nevertheless desired for a clause featuring a situation marked for a perfective reading, a relative construction featuring the head noun *stáyl* 'style, manner' is chosen. In (134), the manner relation is expressed via a relative construction. This option is chosen because the subordinate dynamic verb *nák* 'hit' is marked for factative TMA, hence it is perfective and bounded:

(134) *E* 3sg.sbj *nák* hit *di* def *tébul* table *an* and *di* def *stáyl* style *[wé* sub *e* 3sg.sbj *nák* hit *di* def *tébul* table *strɔ́n],* be.strong *e* 3sg.sbj *kán* pfv *sék* shake *di* def *plét,* plate *an* and *di* def *plét* plate *kán* pfv *brók.* break 'He hit the table and the way that he hit the table in a strong way, he shook the plate, and the plate broke.' [au07se 014]

Other means of providing manner modification by clauses are adverbial SVCs and the use of adverbial clauses introduced by *wé* 'sub' and *sé* 'qot'. Note that equative clauses – manner clauses which serve as the standard in a comparison – are also introduced by the collocation *lɛk háw* (cf. 209–210).

10 Clause linkage

### **10.7.5 Locative clauses**

The formation of locative clauses involves the relativisation of the generic head nouns *sáy* 'side' and less frequently *plés* 'place'. Locative adverbial relations can only be expressed via such relative constructions, because the linker *wé* 'sub' does not introduce headless locative relative clauses:


### **10.7.6 Purpose and result clauses**

The clause linkers *fɔ* 'prep' and *sé* 'qot', as well as the subjunctive marker *mék* are employed to introduce purpose clauses. A purpose relation typically involves a willful and animate subject that intentionally performs a main clause action aimed at the completion of the situation in the subordinate clause. There are no semantic restrictions on the type of main verb that purpose clauses may modify in Pichi. Neither is there any formal difference between "realised" (i.e. that the purpose is achieved) and "unrealised" purpose clauses (cf. Bickerton 1981: 59).

Thus below, we find purpose clauses modifying main clauses with verbs as diverse as *ol* 'be old' or *wét* 'wait':


The motion verbs *gó* 'go' (139) and *kán* 'come' (140 below) may optionally reinforce the purposive sense of the subordinate clause:

(139) *Dɛn* 3pl *kán* pfv *kɛ́r* carry *mí* 1sg.indp *na* loc *Madrid* place *fɔ* prep *mék* sbjv *dɛn* 3pl *gó* go *opera* operate *mí.* 1sg.indp 'They took me to Madrid in order to operate on me.' [fr03ft 026]

When the subjects of the main and subordinate clauses are identical, the purpose clause may be introduced by the non-finite clause linker *fɔ* 'prep' alone (140):

10.7 Adverbial relations

(140) *Mi* 1sg.poss *papá* father *bin* pst *kán* come *yá* here *fɔ* prep *kán* come *wók.* work 'My father came here in order to work.' [fr03ft 063]

When the main and subordinate clauses have different subjects, the purpose clause is expressed as a more finite subjunctive clause. Such purpose clauses are marked in the same way as other types of different-subject subordinate clauses that involve a form of deontic modality. The subjunctive marker may optionally be preceded by *fɔ* 'prep' as in (141):

(141) *Layk* like *háw* how *dɛn* 3pl *go* pot *pút* put *yú* 2sg.indp *na* loc *tébul* table *yu* 2sg *dɔ́n* prf *de* ipfv *rɔ́tin,* rot *fɔ* prep *mék* sbjv *dɛn* 3pl *gó* go *bɛ́r* bury *yú* 2sg.indp *kwík.* quickly

'As soon as they put you on the table you are already rotting away for you to be buried quickly.' [ed03sb 101]

However, a very frequent alternative is for both different- (142) and same-subject (143) purpose clauses to be introduced by the subjunctive marker alone:


'I'll watch it at a neighbour's in order to look at the (first) ladies.' [ma03hm 074]

Negation of the subordinate situation obligatorily entails the use of subjunctive purpose clauses, even where the subjects of the main and subordinate clauses are identical, as in (144):

(144) *A* 1sg.sbj *dríng* drink *di* def *mɛ́rɛsin* medicine *fɔ* prep *mék* sbjv *a* 1sg.sbj *nó* neg *sík.* be.sick 'I drank the medicine in order not to fall sick.' [ro05de 021]

When the purpose clause is fronted for emphasis, it is not usually introduced by *mék* alone. Instead, the purpose clause is normally introduced by *fɔ* 'prep' or *sé* 'qot' and then followed by *mék* 'sbjv'. This is probably so because a sentence-initial *mék* 'sbjv' signals the presence of a subjunctive-marked directive main clause:

10 Clause linkage

(145) *Fɔ* prep *mék* sbjv *yu* 2sg *fít* can *ɛ́nta* enter *yu* 2sg *gɛ́fɔ* have.to *bísin* be.busy *na* loc *wán* one *pɔ́sin* person *fɔ* prep *di* def *fámbul* family (*…*)

'In order to be able to enter, you have to be involved with a person of the family (…)' [ed03sb 077]

Different subject purpose clauses may also additionally feature the quotative marker *sé* 'qot' like any other subjunctive subordinate clause. In such cases, the purpose clause is also usually marked for subjunctive mood. Compare the sentence below; it contains a purpose clause introduced by *sé mék* as well as one introduced by *mék* alone:

(146) *A* 1sg.sbj *bin* pst *lás* end.up *gó* go *a* 1sg.sbj *de* ipfv *fɛ́n* look.for *bíg* big *bíg* rep *mamá* mother *dɛn,* pl *sé* qot *mék* sbjv *dɛn* 3pl *bí* be *mi* 1sg.poss *gɛ́l* girl *frɛ́n,* friend *mék* sbjv *dɛn* 3pl *de* ipfv *gí* give *mi* 1sg.indp *chɔ́p.* food 'I finally went to look for mature/established women for them to be my girlfriends, for them to give me food.' [ed03sp 079]

The following two sentences featuring clauses introduced by the quotative marker *sé* can be interpreted as purposive although they are not followed by subjunctive clauses. These sentences are further evidence for the polyfunctionality of the quotative marker. Here, the expression of speaker intention through inner speech rendered in a quotative construction acquires a purposive reading.

This is the case in the 1st person statement of intention in direct speech in (147), in which *sé* functions more like a clause linker as well as in the 3rd person indirect speech, in which *sé* behaves like a lexical verb (148):


Finally, a *sé*-clause may acquire a result reading when it features non-modal TMA marking or when a modal complementiser is absent. Compare the following example:

(149) (*…*) 3sg.sbj *e* shoot=3sg.obj *sút=an* qot *sé* 3sg.sbj *e* prf *dɔ́n* want *wɔ́nt* go *gó* take=3sg.obj *ték=an,* 3sg.sbj see *e* def *sí* thing *di* prf *tín* ipfv *dɔ́n* turn *de* person *tɔ́n pɔ́sin.* '(…) he shot it [the animal] and was about to go take it, (when) he saw the thing

turning into a human-being.' [ma03sh 005]

10.7 Adverbial relations

### **10.7.7 Cause clauses**

Cause relations may be expressed through *sé*-clauses and adverbial clauses introduced by the linkers *bikɔs (sé)* 'because', *foséka* 'due to, for the sake of', and *ás* 'as'. Speakers also employ the Spanish-origin linkers *porque* 'because' and *como* 'as', which have been borrowed into Pichi and form an integral part of the Pichi system of clause linkage (cf. §13.2.3 for a more detailed treatment in the context of codemixing). Compare *bikɔs (sé)* below:

(150) *A* 1sg.sbj *drɛ́b* drive *mi* 1sg.poss *mán* man *bikɔs* because *sé* qot *a* 1sg.sbj *nó* neg *wánt=an* want=3sg.obj *mɔ́.* more 'I drove my husband away because I didn't want him anymore.' [ro05de 015]

Cause clauses introduced by *bikɔs* may appear at the beginning of the sentence (151). When this is so, the cause clause is focused with *na* 'foc' and reiterated by means of one of the resumptive expressions *na ín* 'foc 3sg.indp' and *na di tín* 'foc def thing', both of which mean 'that's why' in this particular context:

(151) *Na* foc *bikɔs* because *e* 3sg.sbj *bɔ́n* give.birth *pikín,* child *na* foc *di* thing *tín* make *mék* 3sg.sbj *e* die *dáy.*

'It is because she gave birth (to a child), that's why she died.' [dj05be 051]

Cause clauses introduced by *ás* 'as' (152) and *como* 'since' (153) precede their main clauses:


The linkers *bikɔs* 'because' and *porque* 'because' may be found in the initial position in sentences (i.e. in prosodically independent utterances) with a weak causal link with preceding sentences. In such instances, these linkers function as discourse markers that introduce elaborations to preceding material. Compare the use of *porque* in (154):

	- b. *Porque* because *mi* 1sg.indp *sɛ́f,* emp *fɔ́s* first *tɛ́n* time *a* 1sg.sbj *bin* pst *de* ipfv *sidɔ́n* stay *dásɔl* only *wet* with *húman* woman *dɛn.* pl 'Because me, formerly I was staying only with women.' [fr03ft 173]

#### 10 Clause linkage

The preposition *fɔséka* (and its free variant *fɔséko)* 'due to, for the sake of' takes nominal, not clausal, complements (cf. e.g. 25). However, when *fɔséka* is followed by the quotative marker and complementiser *sé* 'qot', the resulting collocation may introduce a cause clause like the other linkers treated in this section (155):

(155) *Mí* 1sg.indp *dú=an* do=3sg.obj *fɔséko* due.to *sé* qot *a* 1sg.sbj *bin* pst *wánt* want *hɛ́lp=an.* help=3sg.obj 'I [emp] did it because I wanted to help her.' [ro05ee 069]

### **10.7.8 Extent and result clauses**

Speakers make use of the linker *sóté* 'until' in order to express a relation of temporal extent, as in the first example below. Such clauses may also be interpreted as result clauses in the appropriate context (157). *Sóté* 'until' is a multifunctional word that is also used as a preposition (cf. §9.1.3), as a degree adverbial (cf. §7.7.3), and in the expression of spatial extent (cf. e.g. 4):


Extent clauses introduced by *sóté* are marked for subjunctive mood when the speaker expresses an anticipated outcome as in (158). This usage may be due to transfer from Spanish. The equivalent Spanish conjunction *hasta que* 'until (that)' is also used with the subjunctive mood. Compare the subjunctive-marked *llegue* 'arrive' in (159).


However, the appearance of subjunctive marking in a clause like (158) above also harmonises with deontic notions like preference and desire that also underlie the use of subjunctive in similar clause types, e.g. purpose clauses:

10.7 Adverbial relations

### **10.7.9 Limit clauses**

Limit clauses are formed by using the quantifying adverb *dásɔl* 'only' before the appropriate adverbial clause linker. Below, *dásɔl* 'only' collocates with *fɔ* 'prep', which in turn, introduces a non-finite purpose clause:

(160) *A* 1sg.sbj *bin* pst *mék=an* make=3sg.obj *dásɔl* only *fɔ* prep *hɛ́lp.* help 'I did it only in order to help.' [dj05be 129]

The following example illustrates the use of *dásɔl* followed by *sé* 'qot', which introduces a finite complement clause:

(161) *Wi* 1pl *de* ipfv *sí* see *dásɔl* only *sé* qot *di* def *písis* piece.of.cloth *dɔ́n* prf *hɛ́ng.* hang 'We only see that the piece of cloth is already hanging.' [li07pe 059]

The quantifying adverb *ónli* 'only' may be employed in the same way as *dásɔl* and occurs equally often in limit clauses. In this sentence, *ónli* 'only' precedes a cause clause introduced by *bikɔs* 'because':

(162) *Ónli* only *bikɔs* because *yu* 2sg *de* ipfv *tɔ́k* talk *só,* like.that *yu* 2sg *de* ipfv *salút* greet *só,* like.that *yu* 2sg *de* ipfv *ánsa* answer *só.*

like.that

'Only because you talk like that, you greet like that, you respond like that.' [au07se 158]

### **10.7.10 Source clauses**

Temporal source clauses may be introduced by the collocations *frɔn wé* {from sub} 'since' and *síns wé* 'since sub' = 'since'. Both collocations require the subordinator because they involve prepositions that take nominal complements. Compare the following examples:


The preposition *síns* 'since' is one of two dedicated temporal prepositions of Pichi (the other one being *ápás* 'after', cf. §8.2.2) and may introduce source clauses in combination with the subordinator *wé* 'sub', cf. (165) below:

#### 10 Clause linkage

(165) *Síns* since *wé* sub *a* 1sg.sbj *bí* be *pikín,* child *a* 1sg.sbj *de* ipfv *mɛ́mba* think.of *fɔ* prep *kɔmɔ́t* go.out *na* loc *dí* this *kɔ́ntri.* country 'Since I was a child, I think about leaving this country.' [li07fn 303]]

### **10.7.11 Conditional clauses**

Table 10.5 summarises the most common ways of expressing conditional relations in Pichi. It features the three functionally identical if-clause introducers *ɛf*, *ɛfɛ,* and *if*, all of which mean 'if' as well as the various types of TMA marking attested in the if- and then-clauses. I comment on the relative frequency of the different constructions below:



For one part, a conditional relation can be expressed by the juxtaposition of clauses and a prosodic break at the margin of the first clause (indicated by commas). In such sentences, the order of clauses is iconical; the if-clause(s) come(s) first, as in (166):

(166) *Yu* 2sg *kɔmɔ́t* go.out *dɔ́n,* down *yu* 2sg *wánt* want *ɛ́nta* enter *mɔ́,* more *yu* 2sg *gɛ́fɔ* have.to *gó* go *pé* pay *ɔ́da* other *quinientos.* five.hundred

'(If) you come out from below and you want to enter again, you have to go pay five hundred again.' [f203fp 005]

Secondly, a conditional relation may be signalled overtly through the use of the equative preposition and clause linker *lɛk (sé)* {like qot} 'as if, supposing that' (167). The use of *lɛk (sé)* is not attested with counterfactuals:

(167) *Lɛk* like *sé* qot *yu* 2sg *de* ipfv *dríng* drink *nɔ́,* intj *dán* that *pɔ́sin* person *wé* sub *dé* be.loc *yandá,* yonder *e* 3sg.sbj *de* ipfv *kán* come *sube* go.up *wí* 1pl.indp *wet* with *glás,* glass *na* foc *di* def *tín* thing *wé* sub *mék* make *mék* sbjv *yu* 2sg *nó* neg *dríng* drink *nó* neg *nátin* nothing *wet* with *glas.* glass 'Supposing that you were (out) drinking, right, (and) that person who is over

there comes up to us with a glass, that's what would make you not drink anything from a glass.' [ed03sb 097]

10.7 Adverbial relations

The linker*lɛk* may also introduce the then-clauses of conditional sentences. In the few cases attested, the if-clause is then always explicitly marked by the conditional clause introducer *ɛf* or *if*. This constellation renders a form of bipartite and discontinuous conditional clause marking. Compare the following sentence:

(168) *Ɛf* if *yu* 2sg *bin* pst *bigín* begin *lás* last *wík,* week *lɛk* like *yu* 2sg *dɔ́n* prf *fínis* finish *di* def *wók.* work 'If you had begun last week, you would have finished the work.' [ro05de 029]

The third way of expressing a conditional relation is the most frequent one in the data and involves one of the conditional clause linkers *ɛf, ɛfɛ,* or *if* 'if'. These forms are equivalent in meaning and occur in free variation. However, *ɛf* is the most frequent form. Any of these linkers may introduce the if-clause of reality, potential, and counterfactual conditionals. Sentence (169) is a reality conditional:

(169) *Pero* but *ɛf* if *na* foc *húman* woman *na* foc *bíg* big *húman* woman *yu* 2sg *mán* man *nó* neg *de* ipfv *tɔ́n* turn *bíg* big *mán.* man 'But if it's the wife who's an influential woman, your [her] husband doesn't [automatically] turn into an influential man.' [ma03hm 079]

Sentence (170) features a potential conditional relation. The most common type of potential conditional features factative TMA in the if-clause, while the then-clause features the potential marker *go*. Sometimes, the imperfective marker *de* 'ipfv' comes to mark conditional modality in the then-clause instead of *go* 'pot' (cf. e.g. 174)

(170) *Ɛf* if *yu* 2sg *chɔ́p* eat *ɔ́l* all *dís* this *chɔ́p* food *wé* sub *e* 3sg.sbj *nó* neg *dɔ́n,* done *tumɔ́ro* tomorrow *yu* 2sg *go* pot *sík.* sick 'If you eat/ate all this food that is not done, you'll/'d be sick tomorrow.' [ro05ee 045]

The markers *go* 'pot' (171) and *de* 'ipfv' (172) are also found to mark conditional modality in hypothetical statements contingent upon inferred conditions. The two following sentences are not preceded by an overt if-clause. The "condition" is deduced from context:


Although the verb in the if-clause of potential conditionals usually apears with factative TMA, a minority of conditionals also feature *go* 'pot' or *de* 'ipfv' in the if-clause and in the then-clause, as in (173) and (174). I interpret this use as instances of modal harmony between the two hypothetical situations:

10 Clause linkage


*Lit*. 'But the price that you have to pay (is), if you don't give me your first man who it is him who gave you the first pregnancy, you will give me the child that will come out.' [ed03sb 020]

Counterfactual conditionals feature the past marker *bin* in the if-clause. In the thenclause, we either find the marker sequence *bin fɔ* 'pst cond' (175) or the conditional mood marker *fɔ* 'cond' alone (176) irrespective of past or present tense reference of the situation. Also note the occurrence of potential mood marking in the complement clause introduced by *sé* 'qot' in (176):


The marker(s) *(bin) fɔ* are also encountered in counterfactual statements contingent upon inferred conditions (177). Sentence (178) illustrates that *fɔ* may fulfil the latter function by itself, without explicit tense marking by *bin*, if a past tense temporal frame has been set by prior discourse:


In the vast majority of cases, the if-clause precedes the then-clause in Pichi conditionals. Nevertheless, the corpus contains a few instances of initial then-clauses (179). These types of conditionals are pragmatically marked and usually involve focus of the preposed

10.7 Adverbial relations

then-clause. This example is also of interest, because it reflects some of the residual obligation meaning that the preposition *cum* modal particle *fɔ* may have in counterfactual conditionals (cf. also §6.7.3.2):

	- die

'I should/would have married at twenty-one if Maura's father hadn't died.' [ab03ab 210]

### **10.7.12 Concessive clauses**

Concessive meaning may be expressed by clauses introduced by *wé* 'sub' (cf. 110) and *sé* 'qot' (cf. 116). Alternatively, concessive meaning may be expressed through conditional clauses in conjunction with clausal focus by means of the focus particle *sɛ́f* 'emp', or by way of the Spanish-derived clause linker *aunque* 'although'. In (180), the conditional relation is not signalled overtly. The presence of the focus particle *sɛ́f* 'emp' alone is sufficient to signal concession:

(180) *Yu* 2sg *nó* know *sɛ́f,* emp *yu* 2sg *jɔ́s* just *kán* come *yu* 2sg *nó* neg *go* pot *sabí,* know *yu* 2sg *nó* neg *go* pot *tɔ́k* talk *lɛkɛ* like *dɛ́n.* 3pl.indp

'Even if you know, if you have just come, you wouldn't know, you wouldn't talk like them.' [ma03hm 044]

A concessive clause may also be introduced by the linkers *ɛf(ɛ)* and *if*, just like a conditional clause. TMA marking is also the same as in conditional clauses:

(181) *Ɛf* if *yú* 2sg.indp *na* foc *smɔ́l* small *húman* woman *sɛ́f,* emp *dɛn* 3pl *go* pot *kɔ́l* call *yú* 2sg.indp *dama.* lady

'Even if you [emp] are an insignificant woman, they'll call you lady.' [ma03hm 076]

Concessive clauses are sometimes also introduced by the Spanish clause linker *aunque* 'although' (182):

(182) *Aunque* although *nóto* neg.foc *paludismo,* malaria *if* if *dɛn* 3pl *gív* give *tratamiento* treatment *yu* 2sg *nó* neg *go* pot *dáy.* die 'Even if it is not malaria, if they give you a treatment, you won't die [of the treatment].' [fr03wt 061]

The linker *adɔnkɛ́* 'no matter if' also introduces concessive clauses. *Adɔnkɛ́* is often part of a disjoint structure, namely *adɔnkɛ́* — *wáns*, 'even if — once'. The concessive clause is introduced by the first, and the main clause by the second element (183):

10 Clause linkage

(183) *Adɔnkɛ́* even.if *e* 3sg.sbj *nó* neg *sí* see *yú* 2sg.indp *wán* one *hól* whole *dé,* day *e* 3sg.sbj *nó* neg *bísin,* be.busy *wáns* once *yu* 2sg *bríng di pamáyn.*

bring def oil

'Even if she didn't see you the whole day, she didn't care, if only you brought the oil.' [ab03ab 036]

### **10.8 Intonation**

Continuative intonation accompanies various types of clause linkage (cf. also §3.4.4). For example, it may be found at the boundary between coordinate clauses and the main and subordinate clauses in conditionals. Continuative intonation also occurs on its own without any other linker to signal a relation between adjacent clauses. By definition, serial verb construction do not, however, involve continuative intonation. They form single prosodic units. The main and subordinate clauses of relative constructions are not normally linked by continuative intonation either.

The deictic manner adverb *só* 'like that' in the example below bears a continuative boundary tone. Such a non-final intonation at the boundary of the first clause signals that it is linked with the subsequent one. The nature of the relation between the clauses is determined by context. In this case, a cause relation reading is favoured:

(184) *Bɔkú* much *motó* car *dɛn* pl *dé* be.loc *yá* here *só,* like.that *a* 1sg.sbj *nó* neg *nó* know *sé* qot *Pancho* name *mék* make *lɛkɛ* like *sé* qot *e* 3sg.sbj *de* ipfv *sube* go.up *bihɛ́n* behind *wí* 1pl.indp *e* 3sg.sbj *baja* go.down *mɔ́.* more

'(Because) a lot of cars were just there, I didn't know that Pancho pretended to go up behind us and went down again.' [ye03cd 176]

Conditional relations are also frequently signalled by means of continuative intonation alone instead of clause linkers (185):

(185) *Yu* 2sg *mék=an* make=3sg.obj *in* 3sg.poss *fray-rɛ́s,* fry.cpd-rice *in* 3sg.poss *banána* banana *dé,* there *e* 3sg.sbj *go* pot *chɔ́p=an.* eat=3sg.obj '(if/when) you make him his fried rice (and) his banana, he will eat it.' [ro05rt 059]

## **11 Multiverb constructions**

I employ the term "multiverb constructions" (MVCs) as a generous cover term for serial verb constructions (SVCs), secondary predication, and clause chaining in Pichi. Multiverb constructions all have in common that there is some form of semantic dependence of one or more predicates with another, which is reflected in some form of reduction, restriction, or merging of elements of one or several predicates. Nonetheless, some of the constructions described in this chapter exhibit varying degrees of resemblance with some of the multiclausal structures covered in chapter 10. As a consequence, the classification as "multiverb" or "multiclausal" is sometimes difficult to make.

### **11.1 General characteristics**

Multiverb constructions form a continuum of structures involving predicates that are strung together in various ways. The area covered by MVCs stretches from tightly integrated verb strings to clause chains, i.e. structures that can barely be distinguished from a series of fully finite clauses. In the middle range of the continuum, we find secondary predication, which is characterised by more flexibility than SVCs, both in the types of verbs that may enter the construction as well as in the ways of paraphrasing them. SVCs are the most integrated MVCs. I use the term SVC only for constructions where "[o]ne verb is from a relatively large, open, or otherwise unrestricted class, and another from a semantically or grammatically restricted (or closed) class" (Aikhenvald 2006: 21).

The verb from the restricted class in SVCs is henceforth referred to as the "minor verb" and the open-class verb as the "major verb" (Durie 1997). The relative position of verbs in SVCs is indicated by V1, V2, VX irrespective of their function as minor or major verbs. Whether (and which) SVCs constitute monoclausal or multiclausal structures in Pichi is left to future research to determine.

SVCs are less central to event integration in Pichi than the variety of constructions might suggest. SVCs constitute somewhere between five to twenty per cent of the clause linkage types in a given text. Equally, older (50+ years) speakers tend to use SVCs more frequently. Maybe increased language contact between Pichi and the non-serialising languages Spanish and Bube has led to the reinforcement of already existing, non-serialising strategies of clause linkage in Pichi (see Hajek 2006 on contact-induced "deserialisation"). This observation concerns in particular argument-introducing SVCs, a prominent type of SVC in serialising languages of the region. Equally, there is a tendency towards the lexicalisation of SVCs involving particular verbs. This characteristic warrants analysing at least some of these SVCs as compound verbs.

#### 11 Multiverb constructions

### **11.2 Serial verb constructions**

Table 11.1 lists all types of SVCs identified in the corpus. The table lists the minor verbs of each construction. The semantic class of the major verb is indicated in the V1 or V2 column, e.g. "dynamic verb". Where there is no significant semantic restriction on the semantic class of the major verb, the row simply contains the entry "verb". The possibility of using switch-function (pro)nouns is listed in the table and discussed below where it applies.


Table 11.1 Serial verb constructions

Not included in Table 11.1 are structures involving the following words with highly grammaticalised functions: *go* 'pot' (<*gó* 'go'), *kán* 'pfv' (<*kán* 'come'), *mék* 'sbjv' (<*mék* 'make'), and *sé* 'qot' (< *sé* ?'say'). Comparative SVCs featuring the minor verb *pás* '(sur)pass' are covered in detail in §6.9.

Verbs that participate in SVCs may not be separated by juncture markers such as declarative intonation, pauses, and continuative intonation, nor adverbial clause linkers and complementisers. Equally, the V2 may not be negated separately from the V1, while the negation of V1 has scope over the entire construction. Compare the following examples involving a motion-direction SVC:

11.2 Serial verb constructions


Further, the V2 of an SVC does not appear with TMA marking, since it acquires its TMA specifications from the V1. Only the V1 is marked for tense, mood, and aspect. Hence, the second translation of (3) as a motion-direction SVC is rejected. Instead, the construction may only be interpreted as involving a depictive secondary predicate (cf. §11.3.2). The inability to be independently marked for TMA also distinguishes SVCs from the verbal complements of aspectual and modal auxiliaries, some of which may be preceded by the imperfective aspect marker as well (cf. §10.5.3).

(3) *Yu* 2sg *de* ipfv *kɛ́r=an* carry=3sg.obj *de* neg *gó* go *hospital.* hospital 'You're carrying him while going to the hospital.' [pa07me 009] but not 'You're taking him to the hospital.'

### **11.2.1 Motion-direction SVCs**

Motion-direction SVCs involve one of the four motion verbs listed in Table 11.1 as minor verbs and V2s. These verbs contribute direction to the motion expressed by the V1. The construction is only attested with a total of eight motion verbs in the major verb, V1 position (cf. Table 8.5 for a summary of some of their semantic and syntactic characteristics). Of these verbs, four denote locomotion (i.e. *wáka* 'walk', *rɔ́n* 'run', *fláy* 'fly', and *fála* 'follow'), while the remaining four (*ték* 'take', *kɛ́r* 'take, carry', *bríng* 'bring', and *sɛ́n* 'send') include direction, manner, and causation as part of their meaning.

The V1 position is therefore not open to other potential candidates with similar meanings (e.g. *drɛ́b* 'drive', *ɛ́nta* 'enter', or *pús* 'push'), and the use of other motion verbs usually involves non-serial strategies of expressing direction. Indeed, the lexical specialisation of this SVC may justify an analysis of the construction as involving compound verbs rather than more open structures created by syntactic processes.

The following example presents a motion-direction SVC involving the V1 *rɔ́n* 'run' and the V2 *gó* 'go', which expresses motion away from the ground. §8.1.5 contains an extensive treatment of goal and source-marking in combination with motion-direction SVCs and other constructions involving spatial relations:

(4) *E* 3sg.sbj *sé* qot *"mɔ́mi* mum *mɔ́mi,* mum *yu* 2sg *nó* neg *de* ipfv *sí* see *dán* that *mán* man *wé* sub *e* 3sg.sbj *rɔ́n* run *gó* go *abuela* grandmother *in* 3sg.poss *rúm?"* room 'He said "mum, mum, don't you see that man who ran into grandmother's

room?"' [ab03ab 053]

#### 11 Multiverb constructions

The goal of the motion may be expressed as an object of the V2 motion verb as in (4) above. The goal may also be instantiated in a prepositional phrase introduced by *na* 'loc' (5). Motion-direction SVCs featuring a transitive V1 can involve a "switch-function" (pro)noun (Aikhenvald 2006: 14–15), in which case the object *=an* '3sg.obj' of the V1 *kɛ́r* 'carry' may be analysed as the subject of the V2 *gó* 'go' in the following example:

(5) *A* 1sg.sbj *kɛ́r=an* carry=3sg.obj *gó* go *na* loc *comedor.* dining-room 'I carried him to the dining-room.' [ab03ab 091]

A string of two verbs may be followed by additional serial verbs. Example (6) illustrates multiple serialisation with the verb string *kɛ́r-gó-wáka* 'carry-go-walk'. The construction is an overlap of a motion-direction SVC (*kɛ́r-gó*) and a motion-action SVC (*gó-wáka*):

(6) *Di* def *bíg* big *wán,* one *a* 1sg.sbj *bin* pst *de* ipfv *kɛ́r=an* carry=3sg.obj *gó* go *wáka* walk *na* loc *nɛ́t* night *wet* with *Tokobé.* name 'As for the big one, I was carrying it off travelling by night with Tokobé.' [ab03ab 006]

The V2 *kán* 'come' expresses motion towards a ground (7). Strings involving the verb *kɔmɔ́t* 'go out' as the V2 express evacuation, i.e. motion out of a ground (8). Note the presence of the prepositions *fɔ* 'prep' and *na* 'loc' which mark goal and source, respectively:


The notion of 'movement up to' is formed with the verb *rích* 'arrive' in the V2 position as in (9). This construction is, however, rare:

(9) *A* 1sg.sbj *tínk* think *sé* qot *e* 3sg.sbj *gɛ́t* get *treinta* thirty *y* and *ocho* eight *años* years *náw* now *wé* sub *e* 3sg.sbj *dɔ́n* prf *de* ipfv *gó* go *rích* reach *cuarenta.* forty 'I think that he's thirty-eight years old now and is already going towards forty.' [fr03ft 146]

The situation expressed by motion-direction SVCs is more often expressed in nonserial structures featuring prepositional phrases as in (10). In these constructions, context and common sense disambiguate the potentially locative (i.e. 'in the pharmacy') and goal ('to the pharmacy') meanings of the PP introduced by the general locative preposition *na* 'loc':

11.2 Serial verb constructions

(10) *Dɛn* 3pl *rɔ́n* run *na* loc *farmacia,* pharmacy *receta* prescription *de* of *mɛ́rɛsin.* medicine

'They ran to the pharmacy, (to get a) prescription for medicine.' [ab03ab 123]

Motion-direction SVCs and alternative ways of expressing the events they denote are also treated extensively in section §8.1.5.

### **11.2.2 Motion-action SVCs**

Motion-action SVCs involve the motion verbs *gó* 'go' and *kán* 'come' as minor verbs in the V1 position. This SVC denotes movement and subsequent action. It often has an underlying purposive meaning best translated as 'go/come in order to'. The construction is the most frequent SVC in the corpus and involves a large variety of minor verbs in the V2 position.

The construction may involve another motion verb as V2 (11), or any other dynamic verb (12). Motion-action SVCs are only attested with a dynamic V2:

(11) *Di* def *pikín* child *dɔ́n* prf *gɛ́t* get *sɛ́ven* seven *hía,* year *e* 3sg.sbj *go* pot *wánt* want *gó* go *wáka,* walk *"hɛ́,* intj *nó* neg *kɔmɔ́t* go.out *na* loc *hós!"*

house

'(When) the child is seven years old, she will want to go walk [roam around], [then you tell her], "don't you leave the house!"' [ab03ay 115]

(12) *Ápás* after *tumɔ́ro* tomorrow *a* 1sg.sbj *go* pot *gó* go *sí* see *mi* 1sg.poss *mamá.* mother 'After tomorrow, I will go see my mother.' [ro05ee 131]

Below follow motion-action SVCs involving the minor verb *kán* 'come' as the V1. Like *gó*-SVCs, *kán*-SVCs are encountered with (13) and without (14) resumptive subject marking with the V2:


SVCs involving the use of *kán* as a verb in a motion-action SVC like (14) need to be distinguished from the use of *kán* as a narrative perfective aspect marker in a sentence like (15). There are two ways of making the distinction. Firstly, in (14), the lexical verb *kán* 'come' may be marked by TMA markers like any other Pichi verb. On the contrary, the

#### 11 Multiverb constructions

narrative perfective marker kán 'pfv' is subject to co-occurrence restrictions. For example, the TMA marker sequence \**de kán* 'ipfv pfv' in (14) above would be ungrammatical in (15).

(15) *Na* foc *ín* 3sg.indp *e* 3sg.sbj *kán* pfv *vɛ́ks,* be.angry *e* 3sg.sbj *kán* pfv *gó.* go 'That's why he got angry, (and) he left.' [fr03ft 190]

Secondly, speakers may employ resumptive subject marking with the V2 in sentences like (13) above in order to avoid the potential ambiguity between a motion-action SVC and a verb marked for narrative perfective aspect (i.e. *yu kán púl=an* '2sg come remove=3sg.obj' = '(then) you removed it)'. The same strategy is employed in (16) below. In both examples, the bare lexical verb *kán* 'come' is likely to be interpreted as the narrative perfective marker *kán* 'pfv' if the sequence were not interrupted by the personal pronoun *yu* '2sg'. That said, these two uses of *kán* are often very similar and appear to be diachronically related:

(16) *Porque* because *if* if *yu* 2sg *mék,* make *yu* 2sg *sí* see *dán* that *polvo,* powder *e* 3sg.sbj *de* ipfv *pút=an* put=3sg.obj *ínsay,* inside *yu* 2sg *kán* come *yu* 2sg *dríng,* drink (*…*) 'Because if you make, you see that powder, he's putting it inside, you come and drink (…)' [ed03sb 099]

Motion-action SVCs frequently involve the use of resumptive *gó* and *kán*. In (17), the verb string is interrupted by the adverbial phrase *na peluquería* 'to the hairdresser's', after which we find a resumptive *gó*. Example (18) features resumptive *kán* after the adverbial phrase *wán dé '*one day':


A more literal motion meaning may give way to a purposive meaning. In (19), movement to the speakers hometown has already occurred before the motion-action SVC *a gó bɔ́n* 'I went to give birth' follows. There is no prosodic juncture between the two clauses:

(19) *A* 1sg.sbj *gó* go *fɔ* prep *kɔ́ntri* country *a* 1sg.sbj *gó* go *bɔ́n.* give.birth 'I went to my home town in order to give birth.' 11.2 Serial verb constructions

In (20), the literal meaning of the V1 *gó* recedes further behind a purposive sense. In this example, we see how motion through space instantiated by *kɛ́r* 'bring', the motion metaphor of the purpose clause introduced by *fɔ* 'prep', and the motion/purpose reading of *gó* itself harmonise:

(20) *Dɛn* 3pl *kán* pfv *kɛ́r* carry *mí* 1sg.indp *na* loc *Madrid* place *fɔ* prep *mék* sbjv *dɛn* 3pl *gó* go *opera* operate *mí.* 1sg.indp 'They took me to Madrid in order to go and operate me.' [fr03ft 026]

The motion-action SVC in (21) does not involve directed motion through space either. The SVC *a gó a púl di trɔsis* 'I (went and) removed the trousers' involves no motion other than removing the pair of trousers:

(21) *A* 1sg.sbj *púl* remove *in* 3sg.poss *camiseta,* singlet *a* 1sg.sbj *pút=an* put=3sg.obj *pantáp* top *béd,* bed *a* 1sg.sbj *gó* go *a* 1sg.sbj *púl* remove *di* def *trɔsis* trousers *a* 1sg.sbj *híb=an* throw=3sg.obj *ínsay* inside *di* def *bañera.* bath.tub

'I removed his singlet, I put him on the bed, I (went and) removed his trousers, I heaved him into the bath tub.' [ab03ab 083]

Example (21) also points towards a difference in meaning that may arise between motion-action serialisation without resumptive subject marking (cf. e.g. 20 above) and motion-action SVCs, in which the V2 has an overt subject pronoun (cf. e.g. 21). While the former type tends to extend metaphorically into the expression of purpose relations, the latter tends to focus the action designated by V2. Motion-action SVCs involving *kán* also lend themselves to less literal interpretations. Compare (16) above, where the V1 *kán* also focuses the following V2 *dríng* 'drink'.

### **11.2.3 Participant-introducing SVCs**

In participant-introducing SVCs, a noun appears as the syntactic object of the minor verb, and this object may occupy diverse semantic roles. One type of participant-introducing SVC involves the verb *ték* 'take'. *Ték*-SVCs may in turn be divided into two types.

In the first type, the object of the V1 *ték* 'take' is the instrument or means used for performing V2. Compare *wán blák lapá* 'a black cloth' in (22). The instrument may also be an abstract noun like *páwa* 'power' (23) or *papá gɔ́d* 'God' in the idiom in (24):


#### 11 Multiverb constructions

(24) *A* 1sg.sbj *ték* take *papá* father *gɔ́d* God *bɛ́g=an.* ask=3sg.obj 'I implored him in the name of God.' [sa07fn 297]

In the second type, the object of the V1*ték* 'take' is the theme of the V2. This type of *ték*-SVC is far more frequent than the one involving an instrument role. Equally, in this type, the theme is always reiterated by a resumptive object pronoun following V2, and very frequently it additionally involves resumptive subject marking. These two characteristics may make such *ték*-SVCs difficult to distinguish from clause chaining when the first subevent of the situation denoted by the SVC may actually involve "taking" in a literal sense (cf. §11.4).

Compare the alternative translations of (25) and (26). Note the use of a resumptive object pronoun alone in the first example, and the use of both a resumptive object and subject pronoun in the second one:


However, a theme object of *ték* need not be an entity that can be "taken" in a literal sense. The following example once more involves resumptive object and subject pronouns. With an object like *yáy* 'eye', no literal interpretation of *ték* as 'take' is possible here:

(27) *A* 1sg.sbj *tɛ́l* tell *yú* 2sg.indp *sé* qot *mi* 1sg.poss *mán* man *ték* take *ín* 3sg.poss *yáy* eye *e* 3sg.sbj *pút=an* put3sg.obj *bɔtɔ́n* bottom *grɔ́n* ground *só.* like.that 'I tell you that my husband diverted his eye [gaze] down like this.' [ro05rt 011]

When the theme object of *ték* is human, it may also receive a comitative 'together with' interpretation. This occurs with the object *di gɛ́l* 'the girl' in the relative construction in (28):

(28) *Porque* because *e* 3sg.sbj *fíba* seem *sé* qot *di* def *gɛ́l* girl *[wé* sub *e* 3sg.sbj *bin* pst *de* ipfv *ték* take *kɔmɔ́t],* go.out *e* 3sg.sbj *bin* pst *gɛ́t* get *bɔkú* much *bɔ́y* boy *dɛn.* pl 'Because it seems that the girl that he was going out with, she had many boyfriends.' [fr03ft 127]

11.2 Serial verb constructions

Example (28) above is also noteworthy because it shows what happens when the object of *ték* is relativised. The object *di gɛ́l* 'the girl' is placed in the head noun position, while the relativised position may remain empty, which leads to V1 and V2 occurring next to each other. Contiguity of *ték* and the V2 is also found when the object of *ték* is fronted in content questions. Sentence (29) features the questioned concrete noun *plɛ́nk* 'board' and (30) the abstract noun *stáyl* 'manner':


SVCs involving *ték* are less frequent than equivalent combinations of verbs and prepositions. A PP involving *wet* 'with' is more commonly employed to express the semantic role of instrument (31). Comitative *ték*-serialisations are even less common. Speakers usually resort to a PP introduced by the preposition *wet* 'with' as in (36) further below:

(31) *Dɛn* 3pl *bíl* build *di* def *strít* street *wet* with *caterpillar.* caterpillar 'The street was built with a caterpillar.' [dj05be 078]

The competition between the serial and prepositional strategies of participant-marking is manifest in the rather exceptional sentences (32) and (33) elicited from two different speakers. Here, the questioning of the instrument noun produced redundant marking of the question phrase *ús=káyn tín* 'q=kind thing' = 'what' with both a preposition and a *ték*-SVC. Non-interrogative double uses of this kind were not found, however:


The verb *fála* 'follow, accompany' participates as a V1 in the expression of a comitative role. The object of *fála* is the accompanee of the situation denoted by the V2. The object of *fála* is usually human and placed between V1 and V2:

(34) *Yɛ́s,* yes *Concha* name *fála* follow *Princess* name *gó* go *viaje.* voyage

'Yes, Concha went on the voyage together with Princess.' [dj05be 097]

#### 11 Multiverb constructions

Once more, most speakers prefer to express accompaniment through non-serial alternatives. One possibility is the use of the verb *jwɛ́n* 'join', followed by the nominalised reference verb as in (35). The most common means involves a comitative prepositional phrase introduced by *wet* 'with' (36):


A final type of participant-introducing SVC is the comparative construction featuring the verb *pás* '(sur)pass' (37). The object of *pás* is the standard of comparison. Comparative SVCs are covered in detail in section §6.9.1:

(37) *Lage* name *de* ipfv *tɔ́k* talk *Bubɛ* Bube *pás* pass *mí.* 1sg.indp 'Lage speaks Bube better than me.' [fr03ab 012]

### **11.2.4 Complementation SVCs**

This type of SVC features a verb of immediate perception as a minor verb and V1. In the corpus, this construction is attested with *sí* 'see' and *hía* 'hear' as V1. The construction features a switch-function (pro)noun. In (38), the object of *sí* 'see', i.e. *sɔn wáyt pambɔ́d '*a white bird', functions as the notional subject of the V2 *kán* 'come':

(38) *A* 1sg.sbj *sí* see *sɔn* some *wáyt* white *pambɔ́d* bird *de* ipfv *kán.* come 'I saw a white bird coming.' [ed03sb 174]

Apart from participant overlap via switch-function, a defining feature of complementation SVCs is the temporal overlap between V1 and V2. Hence, in the example above, the dynamic verb *kán* 'come' is marked for imperfective aspect, which signals simultaneity with the situation denoted by the factative marked V1 *sí* 'see.' The appearance of imperfective aspect to indicate simultaneity is also found with depictive secondary predicates (cf. §11.3.2). Complementation SVCs are, however, syntactically more integrated; they involve switch-function (pro)nouns while secondary predication may not.

When the V2 in a complementation SVC is an inchoative-stative property item, the V2 may appear with an overt subject *e* '3sg.sbj', which is coreferential with the preceding object pronoun *=an* '3sg.obj', as in the example below. Without the V2 subject, the property item *fáyn* would be interpreted as an adverbial modifier of *sí* 'see'. This structure is now in fact identical to some of the depictive secondary predications covered in section

11.2 Serial verb constructions

§11.3, e.g. (62). Complementation structures are therefore not so clear-cut cases of SVCs, and it is debatable whether they should not be seen as "overlapping clauses" (Ameka 2006), hence multiclausal structures.

(39) *If* if *yu* 2sg *gó* go *fɔ* prep *di* def *máred,* marry *yu* 2sg *sí=an* see=3sg.obj *e* 3sg.sbj *fáyn.* be.fine 'If you go to the marriage, you see it (to be) nice.' [*Lit*. '(…) it is nice.'] [hi03cb 006]

The more common alternative to complementation SVCs is for the perceived situation to be expressed as a complement clause introduced by *sé* 'qot', as in the following example:

(40) *Yu* 2sg *jɔ́s* just *hía* hear *sé* qot *pɔ́sin* person *dɛn* pl *bin* pst *de* ipfv *tɔ́k* talk *bɔt* but *yu* 2sg *nó* neg *lístin.* listen 'You just heard that people were talking but you didn't listen.' [au07se 109]

### **11.2.5 Adverbial SVCs**

Two verbs in the corpus appear as minor verbs in adverbial SVCs. In these structures the V1 provides a modification that is temporal in nature. The verb *lás* 'be the last to, end up' enters into an adverbial SVC as a minor verb (41). Proof for the verbal status of *lás* comes from (42): *lás* may not appear in the postverbal adverbial position. In contrast, the word *fɔ́s* 'first' which also expresses temporal meanings may, since it is an adverb (43):


The dynamic verb *sté* 'stay' is employed as the V1 in an SVC in order to express (excessive) duration. This SVC is frequently used in a context of current relevance, where it commonly appears together with the perfect marker *dɔ́n*:

(44) *Yu* 2sg *dɔ́n* prf *sté* stay *kán?* come 'Did you come long ago?' [ge07fn 164]

#### 11 Multiverb constructions

Many speakers instead prefer to express duration through a biclausal structure with co-referential subjects (45) or an expletive subject to *sté* (46). The latter use is once more similar to secondary predication covered below in section §11.3:


### **11.3 Secondary predication**

Pichi deploys reduced clauses as adjuncts to clauses fully specified for person and TMA. In the following, I refer to the predicator of the reduced clause as the secondary predicate, and to that of the full clause as the primary predicate (Schultze-Berndt & Himmelmann 2004; Himmelmann & Schultze-Berndt 2006). Secondary predicates may range in complexity from fully-fledged clauses to reduced clauses consisting of the secondary predicate alone. In Pichi, there is therefore no clear-cut distinction between structures involving secondary predication and some of the time, manner, and result clauses covered in section §10.7.

There are two types of secondary predication in Pichi, namely depictives (§11.3.2) and resultatives (§11.3.3). The difference between the two types is both semantic and formal. Resultative secondary predicates instantiate an end-state and can therefore be seen to stand in a relation of temporal sequentiality or posteriority to the primary predicate. In formal terms, only inchoative-stative property items can function as resultative secondary predicates.

Depictive secondary predicates are in a temporal relation of simultaneity to the primary predicate and therefore contribute manner or temporal readings to the primary predicate. It makes little sense to distinguish further in Pichi between secondary predicates commonly referred to as depictives and those known as circumstantials (cf. Himmelmann & Schultze-Berndt 2006). The semantic and formal differences that we find between individual constructions are due to differences in the lexical aspect class, degree and type of transitivity, and other semantic features (e.g. animacy) of the primary and secondary predicates. These features also co-determine whether a secondary predicate is subject- or object-oriented. With depictives, the lexical aspect class of the secondary predicate also determines whether the secondary predicate is marked for imperfective aspect, by factative TMA (i.e. with inchoative-stative verbs) or by the use of *de* 'ipfv' (i.e. with dynamic verbs).

11.3 Secondary predication

### **11.3.1 Secondary predication vs. serial verb constructions**

Secondary predicates can be distinguished from SVCs on formal grounds. For one, the secondary predicate is connected to the primary predicate in a loose way, i.e. via adjunction. The secondary predicate can therefore be paraphrased by fuller clauses with sometimes only slight modifications to the sentence (cf. 51).

A second distinguishing feature is that secondary predicate constructions do not involve switch-function (pro)nouns. In the following motion-direction SVC, *=an* '3sg.obj', the object of the V1 *kɛ́r* 'carry', simultaneously functions as the notional subject of the V2 *kán* 'come'. In fact, the overt expression of a subject pronoun with the V2 would be ungrammatical (i.e. \**a kɛ́r=an e gó na comedor* 'I carried him, he went to the dining-room').

(47) *A* 1sg.sbj *kɛ́r=an* carry=3sg.obj *gó* go *na* loc *comedor.* dining-room 'I carried him to the dining-room.' [ab03ab 091]

The following secondary predicate construction is therefore rejected. The object *mí* '1sg.indp' of the primary predicate *mít* 'meet' may not simultaneously serve as the subject of the secondary predicate *kúk* 'cook':

(48) \* *E* 3sg.sbj *mít* meet *mí* 1sg.indp *de* ipfv *kúk.* cook Intended: 'He came across me while (I was) cooking.' [pa07me 017]

In such cases of object-subject identity involving an animate participant, the secondary predicate must rather have an explicit subject, even if the primary predicate object and the secondary predicate subject are co-referential:

(49) *E* 3sg.sbj *mít* meet *mí* 1sg.indp *a* 1sg.sbj *de* ipfv *kúk* cook *sɛ́f.* emp 'He came across me while I was actually cooking.' [ro05de 023]

Further, the V2 of an SVC acquires its TMA specification from the V1; the V2 may not be independently marked for tense, mood, and aspect (cf. 3). In contrast, depictive secondary predicates must be marked for simultaneity by imperfective aspect, either via factative TMA or via *de* 'ipfv'. Compare the imperfective-marked secondary predicate *chɔ́p* 'eat' in this example:

(50) *Yu* 2sg *pikín* child *sidɔ́n* sit *de* ipfv *chɔ́p* eat *dɛn* 3pl *tú* two *brɛ́d.* bread

'Your child was sitting (there) eating those two loaves of bread.' [ab03ab 128]

Many secondary predicates in the data do not feature overt subjects either and in that, they resemble the V2s of SVCs like (47) above. However, contrary to the SVC in (47), the notional subject of the secondary predicate may optionally be expressed. Secondary predicates may therefore be expanded into fuller clauses.

#### 11 Multiverb constructions

The following sequence of near-identical resultative constructions graphically shows the progression from the reduced clause typical of secondary predication to a biclausal structure involving overt clause linkage:

	- b. *A* 1sg.sbj *lɛ́f* leave *di* def *domɔ́t* door *e* 3sg.sbj *ópin.* be.open 'I left the door open.' [pa07me 030]
	- c. *A* 1sg.sbj *lɛ́f* leave *di* def *domɔ́t* door *sé* qot *e* 3sg.sbj *ópin.* be.open 'I left the door open.' [pa07me 031]

### **11.3.2 Depictives**

In formal terms, there are two types of depictive secondary predicates. One type features a bare verb with a stative interpretation, the other a dynamic verb marked for imperfective aspect. Both types are therefore marked for simultaneous taxis with the primary predicate – the bare inchoative-stative verb by default via factative TMA, and the dynamic verb via explicit imperfective aspect marking. Further, depictive secondary predications can be differentiated according to their participant orientation. Subject-oriented predicates predicate a situation relating to the subject, object-oriented ones relate a situation relating to the object.

Transitive verbs denoting various types of use or manipulation are prone to occuring with object-oriented depictive predicates. For example, affected-agent verbs like the verbs of ingestion *dríng* 'drink' and *chɔ́p* 'eat' appear with object-oriented secondary predicates with a depictive function:


Another group that appears with object-oriented depictives are verbs of handling and manipulation (e.g. *bay* 'buy', *kɛ́r* 'carry', *sɛ́l* 'sell', *yús* 'use'). The following example illustrates this usage by means of *kɛ́r* 'carry' and the secondary predicate *ɛ́nti* 'be empty':

(54) *A* 1sg.sbj *kɛ́r* carry *di* def *bokit-pán* bucket.cpd-pan *ɛ́nti.* be.empty 'I carried the bucket empty.' [pa07me 039]

11.3 Secondary predication

Subject-oriented depictives occur in intransitive clauses with various types of intransitive or low-transitivity primary predicates. A prominent group of primary predicates encompasses locomotion verbs like *kɔmɔ́t* 'go/come out' as in this example:

(55) *E* 3sg.sbj *kɔmɔ́t* go.out *na* loc *rúm* room *nékɛd.* be.naked 'He left the room naked.' [ra07ve 001]

Some depictive secondary predications may alternatively be expressed through nominal depictives. One strategy involves the use of a prepositional phrase introduced by the multifunctional preposition *wet* 'with' (cf. also 376 and 373–374 in §7.7.2):

(56) *E* 3sg.sbj *kɔmɔ́t* go.out *na* loc *wók* work *wet* with *hángri.* hunger 'He left work hungry.' [ra07ve 073]

A common subject-oriented depictive construction in the data involves the expression of "associated posture" (Enfield 2002): The secondary predicate denotes a situation that holds while the subject assumes a posture denoted by the primary predicate. The secondary predicate is therefore both participant- (the subject) and event-oriented (the primary predicate). When associated posture verbs co-occur with a dynamic secondary predicate, temporal simultaneity is marked overtly by imperfective marking. This is the case in (57) where the posture verb *sidɔ́n* 'sit (down)' is followed by the imperfective marked dynamic verb *chɔ́p* 'eat':

(57) *Yu* 2sg *pikín* child *sidɔ́n* sit *de* ipfv *chɔ́p* eat *dɛn* 3pl *tú* two *brɛ́d.* bread 'Your child was sitting there eating those two loaves of bread.' [ab03ab 128]

The secondary predicate in an associated posture construction may also be another locative verb that elaborates on the type of posture taken by the subject. In (58), the posture verb *sidɔ́n* 'sit (down)' is followed by the inchoative-stative locative verb *ráwn* 'form a circle'. Since *ráwn* is not dynamic, the situation of temporal overlap is not marked by means of the imperfective aspect. It is rather marked by factative TMA, hence the bare verb (cf. §6.1). The use of a co-referential subject pronoun with the V2 (the second *dɛn* '3pl' in the example) is the norm if the secondary predicate is not dynamic:

(58) *Dɛn* 3pl *sidɔ́n* sit *dɛn* 3pl *ráwn* surround *di* def *fáya.* fire

'They're sitting around the fire.' or 'They sat down around the fire.' [ro05ee 115]

The following example also involves associated posture, this time featuring the locativeexistential copula *dé* 'be.loc' serving as a primary predicate. The general locative meaning of the copula allows various interpretations of associated posture. The use of *dé* 'be.loc' together with the adverbial complement *dé* 'there' in such a construction also conveys affective nuances like negligence or irritation with the situation denoted by the secondary predicate:

11 Multiverb constructions

(59) *Di* def *pikín* child *dé* be.loc *dé* there *de* ipfv *kráy.* cry

'The child is just (standing/sitting/lying) there crying.' [pa07me 027]

A second, equally common subject-oriented secondary predicate features a dynamic locomotion verb as the primary predicate. The secondary predicate provides information about the subject as well as the event denoted by the primary predicate itself. In the example below, both verbs are dynamic, hence, imperfective marking is again used to express the temporal overlap of the two predicates. Note the optional use of a resumptive subject pronoun with the secondary predicate:

(60) *Dɛn* 3pl *de* ipfv *fála* follow *dɛn* 3pl *sɛ́f* self *dɛn* 3pl *de* ipfv *rɔ́n.* run

'They're following each other running.' [dj07re 005]

The construction in (61) features the locomotion verb *wáka* 'walk' as primary predicate and the idiomatic reflexive construction *ópin in sɛ́f* '(to) boast' as secondary predicate. Note the presence of the resumptive subject pronoun *e* '3sg.sbj' in this example as well:

(61) *E* 3sg.sbj *nó* neg *gɛ́t* get *mɔní,* money *wétin* what *e* 3sg.sbj *de* ipfv *wáka* walk *e* 3sg.sbj *de* ipfv *ópin* open *in* 3sg.poss *sɛ́f* self *so?* like.that

'He doesn't have money, why does he go around boasting like that?' [ye07je 132]

Animacy provides additional cues to the meaning of constructions involving secondary predication. When the object of a transitive verb has an animate object as in (62) (i.e. *mí* '1sg.indp'), the secondary predicate may be interpreted as either subject- or objectoriented. In such cases, the secondary predicate requires a subject pronoun in order to establish reference with either of the two participants:

(62) *Pero* but *dɛn* 3pl *kán* pfv *dú* do *mí* 1sg.indp *a* 1sg.sbj *de* ipfv *sté* stay *na* loc *Móka,* place *dɛn* 3pl *kán* pfv *dú* do *mí* 1sg.indp *na* loc *Móka.* place

'But they did it to me while I was staying in Moka, they did it to me in Moka.' [ab03ay 071]

### **11.3.3 Resultatives**

Resultative secondary predicates express resultant states, hence, they also involve stative(ly interpreted) property items. Resultative meaning arises in sentences featuring highly transitive effected-object verbs as primary predicates and property items as secondary predicates. Resultatives are invariably object-oriented.

11.3 Secondary predication

In (63), the verb of production *pént* 'paint' is followed by the patient object *hós* 'house' and the secondary predicate *blák* 'black'. The secondary predicated is an object-oriented resultative adjunct that denotes the resultant state of the event denoted by the primary predicate:

(63) *E* 3sg.sbj *pént* paint *di* def *hós* house *blák.* black 'He painted the house black.' [pa07me 037]

Sentence (64) features the effected-object verb *mék* 'make, prepare' as a primary predicate. Note that the secondary predicate takes the subject pronoun *e* '3sg.sbj', which is co-referential with the primary predicate object *café* 'coffee'. The overt subject pronoun is not necessary here because the resultative predicate is clearly object-oriented (unlike the primary predicate presented in 49 above). I assume that an explicit subject pronoun is nevertheless employed because of the presence of the preverbal degree adverb *tú* 'too (much)'. This makes the secondary predicate more complex and motivates the use of a finite resultative clause featuring an overt subject:

(64) *Dɛn* 3pl *mék* make *di* def *café* coffee *e* 3sg.sbj *tú* too *swít.* be.sweet 'They prepared the coffee (it's) too sweet.' [ra07ve 064]

Resultatives may be paraphrased by employing a nominal strategy. The resultative secondary predicate in (65), i.e. the property item *wɔwɔ́* 'be ugly, messed up' may be vaguely paraphrased via an NP in which it appears as a prenominal modifier to the generic noun *stáyl* 'manner' (66). The generic noun *stáyl* 'manner' is also used in modifications of manner (cf. e.g. 368 in §7.7.2) and in manner question words (cf. e.g. 117 in §7.3.2), hence in (66), it is ambiguous between a participant-oriented resultative reading and an event-oriented manner reading:

(65) *Dɛn* 3pl *bíl* build *di* def *ród* road *wɔwɔ́.* be.ugly 'They built the road (and it's) shoddy.' [ra07ve 059]

(66) *Dɛn* 3pl *bíl=an* build=3sg.obj *wɔwɔ́* ugly *stáyl.* style

'They built it (and it's) shoddy.' or 'They built it shoddily.' [ra07ve 060]

Pichi resultative constructions are object-oriented and require the secondary predicate to be an inchoative-stative property item. Neither inchoative-stative verbs from other semantic classes nor dynamic verbs are employed as resultative secondary predicates. In contrast, Pichi's West African sister languages have object-oriented resultative SVCs featuring dynamic secondary predicates as in the Krio example below, and subject-oriented resultatives featuring change-of-state secondary predicates as in Ghanaian Pidgin English.

11 Multiverb constructions

(67) Krio *Di* def *wúmán* woman *kúk* cook *rɛ́s* rice *sɛ́l.* sell 'The woman cooked rice and sold it.' (Finney 2004: 72)

(68) Ghanaian Pidgin English *A* 1sg.sbj *chɔ́p* eat *táya.* be.tired 'I ate (until I was) tired (of it).' (Own knowledge)

Accordingly, Pichi also does not have a resultative completive aspect construction featuring the dynamic verb *fínish* 'finish' as a secondary predicate, as in the following example. In Pichi, completive aspect is instead expressed via an auxiliary construction and a verbal complement (cf. §6.4.3):

(69) Ghanaian Pidgin English *A* 1sg.sbj *chɔ́p* eat *fínish.* finish 'I've finished eating/I'm done eating.' (Own knowledge)

Resultant situations like the ones above must therefore be expressed through fuller clauses in Pichi. When the secondary predicate is not a property item and subject-oriented, a clause linker like *sóté* 'until' may be sufficient. When the secondary predicate is not a property item and object-oriented, a chained clause with person-marking is required:


### **11.4 Clause chaining**

Clause chaining is utilised in narrative discourse to describe tightly-knit situations that take place in sequence. In chained clauses, speakers use one predicate after the other without pausing or placing clause linkers between them. However, chained predicates invariably feature resumptive personal pronouns, and the subject is repeated with each verb in the series. Verbs that participate in clause chaining are always dynamic and form part of foregrounded sections of narrative discourse (cf. §6.8.1).

TMA marking is reduced in chained clauses. Tense, aspect, and mood marking is overtly expressed with the initial one or two verb(s) in order to provide orientation

11.4 Clause chaining

and grounding. Subsequent verbs, however, remain bare. Clause chaining is therefore different from linkage involving fuller clauses through the absence of prosodic juncture marking and the reduction of TMA marking. At the same time, chained clauses differ from SVCs because they exhibit overt person marking.

The clause chain below features the initial verbs *rɛdí* 'be/make ready' and *mék* 'make', which are both fully finite and marked for potential mood. The verbs following *mék*, i.e. *ték* 'take', *pút* 'put', *sɛ́n* 'send', and *gó* 'go' are all left bare without TMA marking. Instead, they form part of a clause chain, in which the initial two verbs alone provide the temporal, aspectual, and modal frame of reference. Note that the bare verbs in the clause chain cannot be interpreted as being marked for factative TMA, since the temporal and modal frame of the paragraph is provided by the potential mood marked on *rɛdí* 'prepare' and *mék* 'make':

(72) *Dɛn* 3pl *go* pot *rɛdí* prepare *yú* 2sg.indp *dɛn* 3pl *go* pot *mék* make *lɛk* like *háw* how *dɛn* 3pl *de* ipfv *mék* make *fɔ* prep *wích,* sorcery *dɛn* 3pl *ték* take *yú* 2sg.indp *dɛn* 3pl *pút* put *yú* 2sg.indp *na* loc *avión* plane *dɛn* 3pl *sɛ́n* send *yú* 2sg.indp *fɔ* prep *ɔ́da* other *kɔ́ntri* country *yu* 2sg *gó* go *wók* work *mɔní.* money

'They would prepare you like the way it's done by sorcery, they'll take you, put you into a plane, and send you to another country (and) you'll go earn (them) money.' [ed03sb 104]

The following example illustrates how the difference between clause chaining and the linkage of fully finite clauses may hinge on intonation when a series of dynamic verbs are marked for factative TMA. In (73), the verbs *ték* 'take'*, pé* 'pay', *kɔmɔ́t '*go out', *rích* 'arrive', and *pé* 'pay' are iconically ordered along the time axis and describe successive events. However, they are separated by pauses. Additionally, the last constituent of each clause bears continuative intonation (indicated by a comma), which alerts the hearer to the existence of a clausal boundary. For these reasons, (73) does not involve clause chaining:

(73) *Lúk=an,* look=3sg.obj *di* def *dé* day *wé* sub *dís* this *Paquita* name *in* 3sg.poss *papá* father *bin* pst *kán* come *ték=an,* take=3sg.obj *e* 3sg.sbj *pé* pay *avioneta,* small.aircraft *kɔmɔ́t* go.out *Alemania,* place *rích* reach *na* loc *Douala,* place *pé* pay *avioneta,* small.aircraft *e* 3sg.sbj *kán* come *na* loc *yá* here *só.* like.that

'Look at her, the day that Paquita's father came to take her, he paid (a ticket for) a small plane, left Germany, got to Douala, paid (a ticket for) a small plane, (and) came here.' [ab03ay 140]

## **12 Pragmatic elements and routines**

The pragmatically oriented elements covered in this chapter form part of a range of performative and expressive devices which extend into the realm of gesture and body posture. In this chapter, sections are dedicated to ideophones, interjections, and two other elements that have much in common with interjections but defy rigid demarcation: the sentence particle *ó* 'sp' as well as 'suck teeth', a phonetic realisation with important pragmatic functions.

Like interjections, ideophones represent an expressive dimension of communicative interaction. Many members of both word classes do not enter into grammatical constructions with other word classes, and are fit into discourse by adjunction. Equally, many interjections and ideophones manifest some degree of sound symbolism, hence the relation between form and meaning is not entirely arbitrary. However, while interjections can occur as utterances on their own, ideophones (unless they are interjectional) occur as parts of utterances.

The two final sections cover two important manifestations of the socio-pragmatic domain of politeness, namely the address system and greeting routines.

### **12.1 Ideophones**

Ideophones are words with expressive semantics and particular structural characteristics (e.g. Westermann 1930: 187–189; Doke 1935: 118–119; Dingemanse 2017). Ideophones are found in different syntactic categories in Pichi, but they may be argued to constitute a coherent semantic class. Ideophones are not very prominent in the corpus and tend to be employed more by older, Group 2 (cf. §1.3) speakers. All ideophones encountered in the data are listed in Table 12.1.

Many of the ideophonic manner adverbs given in Table 12.1 only occur once in the corpus. It is therefore difficult to ascertain how widespread the use of these ideophones is, and whether some of them are sound symbolic ad hoc creations, whether they are carried over from other languages used by the speaker, or whether they form part of the lexicon of Pichi (e.g. *bwa, fwífwífẃi* and *wówó/wéwé*).

Ideophones differ in formal terms from other Pichi lexemes in at least three respects: (1) most ideophones listed in Table 12.1 belong to minor tone classes (cf. Table 3.2); (2) many ideophones feature marginal (sequences of) phonemes (cf. Table 2.2 and Table 2.6) only attested with this word class, e.g. /gb/ and /kp/ (*gbin*, *gbogbogbo*, *kpu*), or otherwise rare clusters like /fw/ (*fwífwífwí* 'sound of the wind blowing'). Many ideophones also feature otherwise rare CV syllable structures (e.g.*súkútúpampa*, *kutuku*, *wéwé*), or a single vowel type (e.g. /a/ in *kata*-) and identical or similar consonants (e.g. /w/ and /r/ in *wuru*-). At

#### 12 Pragmatic elements and routines

Table 12.1 Ideophones


least one ideophone, namely *bwa*, is also pronounced with a breathy voice. About half of the ideophones listed in Table 12.1 represent cases of lexicalised full or partial iteration, i.e. duplication and triplication. The importance of iteration with ideophones reflects the sound symbolic dimension of this word class.

In formal terms, some ideophones appear to have undergone the morpho-tonological process of reduplication (e.g. *kata.kátá* 'be hectic') while others seem to have been subjected to the syntactic process of repetition (e.g. *fwífwífwí* 'sound of the wind blowing') (cf. §4.5 for a detailed description of the formal and functional aspects of both types of iteration).

Ideophones that involve reduplication feature a suprasegmental structure of the type that we find with the ideophonic property item *kata.kátá* 'be (hyper-)active, hectic' in (46) below: Two phonetic L tones over the first two syllables of the reduplicant are followed by a succession of two phonetic H tones over the base (with the last H tone result-

12.1 Ideophones

ing from raising of the lexical L to H, which is phonologised and does not vary). Another ideophone belonging to this group is the ideophonic noun *pɔtɔ.pɔ́tɔ́*:


The most commonly used ideophonic (and generally quite frequent) property item is *wɔ.wɔ́* 'be ugly, messy, in disorder'. This property item also belongs to the group of ideophones with a tone configuration that suggests the operation of reduplication rather than repetition.

A second group of ideophones involves repetition. Some words of this group may be encountered as simplex forms, (i.e. *kutuku* 'sound of the heart beating', *kakara* 'be restless') and may optionally be repeated in order to express meanings associated with repetition, such as emphasis or duration. Iterations of such ideophones therefore do not involve lexicalisation proper, even if there is a strong tendency for them to be repeated in discourse.

Hence, the ideophonic property item *kakara* 'be restless' is employed as a dynamic verb in (9), preceded by the imperfective marker *de* 'ipfv' and repeated for emphasis. The comma after the first *kakara* signals the presence of a short pause, which indicates the presence of (syntactic) repetition rather than (morpho-tonological) reduplication:

(3) *Ín* 3sg.indp *de* ipfv *kakara,* be.restless *kakara* rep *kakara.* rep 'He [emp] was all restless.' [ab03ab 047]

The ideophone *kutuku* 'sound of the heart beating' may also optionally be repeated for emphasis, as in the following sentence:

(4) *Na* foc *só* like.that *in* 3sg.poss *hát* heart *mék* make *kutuku* ideo *kutuku* ideo *kutuku.* ideo

'That's how his heart was going "kutuku kutuku kutuku".' [ab03ab 070]

Other ideophones that formally involve repetition are not usually encountered as simplex forms. Therefore, the ideophonic adverb *gbo.gbo.gbo* which expresses haste or precipitous hurry has no attested simplex form \**gbo*. The ideophone only occurs as a triplicated iteration, as in this example:

(5) *Tokobé* name *dɔ́n* prf *wɛ́r* wear *klós* clothing *gbogbogbo.* ideo 'Tokobé had put on (her) clothes in haste.' [ab03ab 111]

#### 12 Pragmatic elements and routines

Likewise, the ideophonic adverb *fwí.fwí.fwí* 'sound of the wind blowing' only occurs as a triplicated lexicalised repetition (and is therefore written as one word). In the example below, this ideophone modifies the preceding clause headed by the Spanish-derived verb *sopla* '(to) fan, (to) blow':

(6) *Na* foc *só* like.that *a* 1sg.sbj *de* ipfv *wáyp=an,* wipe=3sg.obj *a* 1sg.sbj *de* ipfv *sopla* blow *ín* 3sg.indp *fwífwífwí.* ideo 'I was wiping him like that, I was fanning him.' [ab03ab 068]

Both groups of ideophones, i.e. those involving lexicalised reduplication and lexicalised repetition, can be contrasted with ideophones like *gbin* 'sound of a hard, sudden blow' in (7). This ideophone is not encountered with any form of iteration in the corpus:

(7) *E* 3sg.sbj *gí* give *mí* 1sg.indp *gbin.* ideo 'He hit me hard and suddenly.' [ne07fn 008]

Some other combinations of verbs and ideophonic manner adverbs that are not encountered with iteration in the data are: *nák kip* 'hit=3sg.obj ideo' = 'hit and produce a dull thud', *mék nɔ́ys tík* 'make noise ideo' = 'make a cracking noise'.

The reduplication or repetition of ideophones therefore evokes the same type of "disaggregation" of the reference situation that we find with iterated non-ideophones.This may explain why ideophones like *gbin*, *kip*, and *tík* are not iterated. These ideophones denote sudden and inherently terminative situations, which are not normally associated with the typically cyclic, repetitive, disaggregated events depicted by iterated ideophones.

The following sentence is illustrative of the cyclic and repetitive semantics of iterated ideophones. The ideophone *kamú.kamú* depicts the countermovement of a pair of buttocks as their owner strides along:

(8) *Yu* 2sg *sí* see *lɛk* like *háw* how *in* 3sg.poss *bata* buttocks *dɛn* pl *de* ipfv *sék* shake *kamúkamú?* ideo

'Do you see her buttocks moving to-and-fro (as she walks along)?' [ye07fn 231]

Most ideophones in the corpus are employed as adverbs and modify the verb with respect to manner. A few ideophones preferably function as (property-denoting) verbs or nouns, and one is used as an interjection (*kɔ́ngkɔ́ngkɔ́ng* 'seek permission to enter'

Ideophonic verbs are found in the syntactic positions available to other property items in Pichi. Hence, the ideophone *kakara* 'be restless' is employed as a dynamic verb in (9), and preceded by the imperfective marker *de* 'ipfv'. Note the repetition of the ideophone for emphasis:

(9) *Ín* 3sg.indp *de* ipfv *kakara,* be.restless *kakara* rep *kakara.* rep 'He [emp] was all restless.' [ab03ab 047]

12.1 Ideophones

Ideophonic property items also appear in the prenominal modifier position. Compare *katakátá* 'be (hyper-)active, hectic' in the following sentence:

(10) *Na* foc *wán* one *katakátá* hectic *mán.* man 'He's a hectic man.' [tr07fn 229]

The most commonly used ideophonic (and generally quite frequent) verb is *wɔwɔ́* 'be ugly, messy, in disorder'. This verb, too, is attested as a stative verb (11), and in a prenominal position as an attributive modifier (12). Some speakers also employ *wɔwɔ́* as an adjective, i.e. a complement to the locative-existential copula *dé* 'be.loc' (13). Another indication of the ideophonic nature of *wɔwɔ́*besides its segmental structure is that it is often pronounced with reduced speed and overarticulation, and is accompanied by a facial expression suggestive of disapproval:


Ideophonic nouns appear in the same syntactic position as other nouns. In the following sentence, *pɔtɔpɔ́tɔ́* 'mud' is the head of an object NP featuring the quantifier *bɔkú* 'much':

(14) *Dán* that *sáy* side *gɛ́t* get *bɔkú* much *pɔtɔpɔ́tɔ́* mud *ó.* sp '[Mind you] that place is very muddy.' [ne07fn 230]

Ideophonic adverbs usually modify verbs in the clause-final position. Sentence (15) illustrates the depictive power of an ideophonic adverb like *kwáráng* when used to express the sensory experience connected to playing the African board game Oware. Example (16) presents the ideophonic adverb *fwí* 'sound of the wind blowing', which modifies the preceding Spanish-derived verb *sopla* '(to) fan, (to) blow':

	- 3pl ipfv play=3sg.obj ideo

'It is played with this hollow sound (of the seeds falling into the pits of the wooden Oware board).' [ro07fn 519]

12 Pragmatic elements and routines

(16) *Na* foc *só* like.that *a* 1sg.sbj *de* ipfv *wáyp=an,* wipe=3sg.obj *a* 1sg.sbj *de* ipfv *sopla* blow *ín* 3sg.indp *fwífwífwí* ideo 'I was wiping him, I was fanning him just like that.' [ab03ab 068]

In the following sentence, speaker (ro) uses the ideophonic adverb *súkútúpampa* in order to depict the supposedly cheap and mean manner in which sex workers in Malabo offer their services for sale:

(17) *Dɛn* 3pl *de* ipfv *sɛ́l* sell *dɛn* 3pl *skín* body *súkútúpampa.* ideo 'They barter their bodies away.' [ro05fn 240]

Ideophonic manner adverbs sometimes occur in what appears to be a nominal position as in the following two sentences. Actually, the ideophones do not enter syntactic constructions in these examples either. Instead, the preceding generic verbs *mék* 'make' and *gí* 'give' function as a quotative index followed by a syntactically independent utterance consisting of the ideophonic adverb:


Some other combinations of verbs and ideophonic manner adverbs encountered in the data are: *nák kip* 'hit=3sg.obj ideo' = 'hit and produce a dull thud', *mék nɔ́ys tík* 'make noise ideo' = 'make a cracking noise', *kráy wówó wówó* 'cry ideo rep' = 'cry bitterly'.

### **12.2 Interjections**

In the following, I employ the term "interjection" liberally as a cover term for words, phrases, and clauses that index physical and discursive entities (Kockelman 2003), cognitive and emotional states (Ameka 1992b), and social relations. Interjections are pragmatically oriented elements that appear at the beginning or end of an utterance or constitute utterances onto themselves.

In (20), the initial interjection *ɛ́* 'intj' (cf. Table 12.4) functions as an attention-getter and by doing so, indexes the following utterance. The sentence-final element *ɛ́n* functions as a channel checker, i.e. 'have you heard what I've just said?' and thereby solicits a preferably affirmative response. The example also shows that interjections are set off from the rest of the utterance by a prosodic break (indicated by the comma). This indicates that they function as co-text rather than forming an integral part of the clause:

12.2 Interjections

(20) *Ɛ́,* intj *dí* this *mán* man *gɛ́t* get *líba,* liver *ɛ́n.* intj 'Hey, this man has guts, you know.' [dj05ce 290]

Following Ameka (1992a,b), I classify Pichi interjections along three functions: expressive, conative, and phatic. Many interjections are "primary" (Ameka 1992b) and constitute a micro word class of mostly monosyllabic "small words" which do not occur in contexts other than those described here. Some primary interjections are also phonologically deviant. For example, interjections constitute the only word class in which vowel length may be distinctive (i.e. *a* '1sg.sbj' vs. *aa* 'expression of insight'). Other interjections are "secondary" and also employed as members of other word classes, and they may enter into grammatical constructions with other constituents.

In the following, I cover the most commonly used interjections. Some interjections are cross-classified and may therefore be members of more than one of the three functional types (e.g. *mamá* 'mother' which is employed as an expressive and a phatic interjection).

### **12.2.1 Expressive**

Expressive interjections reflect the emotional and cognitive state of the speaker, but they also serve a communicative purpose by drawing the attention of potential listeners to the mental state of the utterer. Consider the expressive interjections in Table 12.2.


Table 12.2 Expressive interjections

An exemplary primary interjection with an expressive meaning is *chɛ́*or *cháy*, which conveys the feeling of exasperation in the face of a difficult task. In (21),*chɛ́*is the reaction of speaker(dj) to a particularly ungrammatical sentence that I (ko) submit to him for a grammaticality judgement:

12 Pragmatic elements and routines

(21) a. *Na* foc *di* def *púl* remove *di* def *motó* car *fɔ* prep *di* def *mecánico* mechanic '[Can you say:] "It's the removal of the car from the mechanic" [ko0502e2 045] b. *Chɛ́!* intj 'Phew [now this is too much]!' [dj05be 045]

An extreme physical sensation is expressed by the primary interjection *áy*. As indicated by the two following examples, the sensation may be pain or pleasure, in particular the pleasure of good food or sexual delight. There are therefore overlaps in meaning with the (near-)identical interjection *ay* in Spanish:


The interjection *ékié* is an established loan from Fang. It expresses counterexpectation, amazement, and (often playful) indignation. In the corpus, *ékié* is mainly used by female speakers. Sentence (24) is a humorous comment by speaker (ge) addressed to her female friend. The latter has just said that she finds a white European acquaintance of hers attractive. Speaker (ge) teases her friend by pretending to be outraged and calls her *buscablanco* 'look.for-white.male' = 'female sex worker specialised on white men':

	- intj look.for.cpd-white.male

'Good gracious, (you're a) prostitute.' [ge07fn 077]

Pichi also features expressive "secondary interjections" (Bloomfield 1935: 176; Ameka 1992b) which function as members of other word classes besides their use as deicticpragmatic elements. One group of secondary interjections stems from religious terminology. The lexicalised collocation *papá gɔ́d* is a Pichi term for 'God' (25). As an interjection, *papá gɔ́d* is used to implore the help of God during prayer and inner speech (26), or to express self-pity and exasperation (27). Note that *papá gɔ́d* in (26) is preceded by the conative interjection *oó*, which introduces an emphatic vocative (cf. Table 12.4):

(25) *Papa* father *gɔ́d* God *go* pot *mék* make *mék* sbjv *e* 3sg.sbj *chénch,* change *mék* sbjv *e* 3sg.sbj *chénch* change *fásin.* manner 'God will make him change, change (his) habits.' [dj07ae 160]

12.2 Interjections


A number of expressive secondary interjections in the corpus are Spanish-derived and used in similar ways in European Spanish. The interjections *Señor mío* 'my Lord' and *Dios mío* 'my God' express sentiments like surprise, irritation, and frustration (28):

(28) *Señor* Lord *mío,* my *tɛ́l* tell *mí,* 1sg.indp *mi* 1sg.poss *mán* man *e* 3sg.sbj *de* ipfv *kɔmɔ́t* go.out *wet* with *yú?* 2sg.indp 'Good Lord tell me, is my husband going out with you?' [ro05rt 009]

A second group of expressive secondary interjections includes kinship terms and other human-denoting nouns. These nouns are intermediary in their function. On the one hand, these nouns resonate with a strong emotive component when used as interjections. However, by the very nature of their meaning as kinship terms and terms of address, they also index the social relation which they stand for and thereby convey a phatic message to interlocutors.

The Spanish noun *chico* 'boy' is one of the most frequently employed secondary interjections and covers a large range of expressive meanings. It conveys real, playful, or mock surprise (29a–29b), shock and amazement (29c), awe and admiration (29d):

	- d. *Chico,* boy *Jibril* name *trɔ́n* be.strong *ó!* sp 'Wow, Jibril is really great.' [ye05ce 023]

The following excerpt renders reported discourse of a conversation, in which speaker (ro) is taking her husband to task for cheating on her. The husband tells (ro) that he and his lover would meet up in a car. An incredulous (ro) repeats what her husand has just told her in (30a), and then cries out *mamá* 'mother' in shock (b). Her mental state at that

#### 12 Pragmatic elements and routines

moment is reflected by (c). The kinship term *papá* 'father' is employed as an expressive interjection in a similar way to *mamá* (cf. 34 below):

(30) a. *Ínsay* inside *di* def *motó,* car *na* foc *dé* there *unu* 2pl *de* ipfv *slíp* sleep *unu* 2pl *sɛ́f?* self *'*In the car, that's where you sleep with each other?' [ro05rt 020] b. *Mamá.* mother

'Good gracious.' [ro05rt 021]

c. *A* 1sg.sbj *krés.* be.crazy 'I went mad.' [ro05rt 022]

The interjection *mamá* therefore expresses the emotional stress that speaker (ro) was experiencing at that moment. But beyond that, *mamá* is also a meta-comment on the amorality of the husband's act, a performative element embedded in reported discourse, directed at us, the listeners of the narrative. This type of "rhetorical underlining" (Longacre 1996: 39), in which the narrator steps out of the narrative and addresses her audience is a significant element of Pichi narrative technique. The use of *mamá* in this way sheds light on the communicative dimension of expressive interjections in Pichi.

The interjection *bió bió* '(lo and) behold' expresses surprise. By doing so, this interjection also has a strong phatic component to its meaning:

(31) *Bió* behold *bió,* rep *dɛn* 3pl *dɔ́n* prf *kán.* come 'Lo and behold, they've (finally) come.' [pa05fn 456]

The Spanish noun *mierda* 'shit' is used as an expressive interjection for anger and annoyance. The Pichi equivalent *kaká* 'faeces' is not used in this way. However, the Pichi compound *kaka-rás* {faeces.cpd-arse} 'shitty arse' is used as an insult.

(32) *Mierda* shit *mierda,* rep *ús=sáy* q=side *e* 3sg.sbj *pás?* pass 'Shit, shit, which way did she go?' [ro05rt 002]

### **12.2.2 Phatic**

Phatic interjections and phrases are embedded in the verbal interaction between interlocutors. These elements are interactional and are aimed at constructively maintaining the communicative situation. Table 12.3 lists the phatic interjections encountered in the corpus. The functions of the phatic elements and agreement markers *yɛ́(s*) 'yes' and *nó/nɔ́* 'no' are covered in detail in §7.3.3.

The phatic interjection *aa* expresses sudden insight into a proposition or real-world fact. In this, its meaning is similar to *o.k.* (33b).

12.2 Interjections


Table 12.3 Phatic interjections

	- b. *Aa* intj *o.k.* intj 'Alright.' [hi03cb 076]

The interjection *e* is usually uttered with a compressed voice and an extra-high tone. It is also lengthened to up to three beats. It is best translated as 'good gracious' and expresses dismay and empathy with a deplorable event or fact. In (34), the expressive meaning of *e* is underlined by the presence of the interjection *papá* 'father':

(34) a. *E* 3sg.sbj *gɛ́t* get *bɛlɛ́,* belly *wé* sub *e* 3sg.sbj *wɔ́nt* want *púl* remove *di* def *bɛlɛ́.* belly 'She was pregnant and wanted to abort the pregnancy.' [ko03cb 099] b. *É* intj *papá!* father

'Good gracious!' [bo03cb 100]

The kinship terms *papá* 'father' and *mamá* 'mother' are also employed as phatic interjections in appealing for consideration, empathy, and compassion by evoking the nature of the kinship relation that holds between a parent and a child, a provider and a dependent. Consider (35), where speaker (ye) relates how Rubi appeals to the person represented by =*an* '3sg.obj' to leave him his fair share of the remaining food:

12 Pragmatic elements and routines

	- b. *E* 3sg.sbj *tɛ́l=an* tell=3sg.obj *sé* qot *"papá* father *mí* 1sg.indp *nɛ́va* neg.prf *chɔ́p* chop *mí* 1sg.indp *sénwe."* emp '(So) he [Rubi] told him "please, I also haven't eaten myself." ' [ye03cd 149]

The interjection *dúya* 'please' and the less frequent *plís* 'please' play an important role as politeness markers. Both interjections are used in polite imperatives like the following one:

(36) *Pút=an* put=3sg.obj *mɔ́* more *dúya!* please 'Put [play] it again, please!' [au07se 095]

### **12.2.3 Conative**

Conative interjections solicit a verbal or kinetic response from listeners. By their imperative nature, they are used in calling and responding, seeking approval and confirmation, constraining and restraining the interlocutor. Table 12.4 lists common conative interjections.

One of the numerous functions of the sentence-final particle *ó* is its use as a vocative marker in combination with a personal name (cf. §12.2.3 for more). An emphatic, imploring vocative is formed by preposing the interjection *oó* to the name or term of address of the person called (cf. 26 above).

The appropriate way of responding to the call of a social superior is by calling out the term of address of the caller (37b). If the caller is a peer, the person called may also simply respond with *yɛ́s* 'yes':

(37) a. *Pancho!*

name

b. *Mamá!* mother

'(Yes) mum.'

Alternatively, a person can respond with a response call involving the vowels /e/ and /o/ with different degrees of lengthening and in slightly varying pitch configurations over the lengthened vowel. A response call may simply follow the call or additionally feature the caller's name (38b):

	- name
	- b. *Yéé* intj *Paquita!* name

12.2 Interjections


Table 12.4 Conative interjections

The interjection *hɛ́*is employed as a remonstrative when a grown-up or social superior scolds a child or a socially inferior. It is used shortly before, or in the very moment a person commits a transgression in order to warn and rebuke them:

(39) *Di* def *pikín* child *dɔ́n* prf *gɛ́t* get *sɛ́ven* seven *hía,* year *e* 3sg.sbj *go* pot *wánt* want *gó* go *wáka,* walk *"hɛ́,* intj *nó* neg *kɔmɔ́t* go.out *na* loc *hós!"*

house

'(When) the child is seven years old, she will want to roam the streets, (then you say to her) "don't you dare leave the house!"' [ab03ay 115]

The interjection *aa* (homonymous with the phatic *aa* in 33 above) expresses negligence. In that sense, it may communicate to an interlocutor not to worry or bother about a situation. In the appropriate context, negligence may shade off into impatience and serve to express irritation with a person's insisting or nagging stance. In the latter case, *àá* is often pronounced with a rising contour and supported by 'suck teeth' (41).

The uses of this interjection point towards an area of transition between phatic interjections aimed at constructively maintaining a communicative situation and conative interjections with their imperative nature:

#### 12 Pragmatic elements and routines


The interjections and phrases *ɛ́n*, *nɔ́*, *hía* 'hear', and *yu de hía?* '2sg ipfv hear' = 'do you hear?' are employed as channel checking devices in seeking feedback or approval from discourse participants. Thus, they always bear the boundary tone associated with question intonation. Compare *ɛ́n*, which occurs in sentence-final position, sometimes after a pause in order to increase dramatic effect, as well as *nɔ́*(cf. 143 further above):

(42) *Di* def *trú* true *comedor* dining.room *dé* be.loc *fɔ* prep *soja,* soy *Manolete,* name *Corrobes,* name *ɛ́n.* intj 'The true dining room had soy (oil), Manolete (oil), Corrobes (oil), you hear?' [ab03ab 031]

A sentence-final *hía* 'hear' may require explicit approval, so it is used in addressing a listener of equal or inferior social rank. The undertone of authority is stronger with the question phrase *yu (de) hía?* in (44a), which always requires explicit approval, usually with the idiom *a hía* 'I have heard' (44b):

	- 'I've heard.' [ru03wt 057]

Other phrasal interjections are employed to seek attention, approbation and confirmation. The phrases *nó tɔ́k* 'don't talk', which appears together with *nɔ́* and *ɛ́n* in (143) further above, and *nó láf* 'neg laugh' = 'I'm not kidding' (45b) underline the credibility of the speaker's proposition or story. A similar function is fulfilled by *a tɛ́l yú* 'I tell you' in (46):

12.2 Interjections

(45) a. *Djunais* name *tɔ́k* talk *sé* qot *"nɔ́* intj *Rubi,* name *dí* this *gɛ́l* girl *lɛ́k* like *yú,* 2sg.indp *dí* this *gɛ́l* girl *lɛ́k* like *yú,* 2sg.indp *náw* now *bigín* begin *mék=an* make=3sg.obj *só."* like.that '(So) Djunais said "really Rubi, this girl likes you, this girl likes you, now go about it like this with her."' [ru03wt 021] b. *Nó* neg *láf!* laugh 'I'm not kidding.' [ru03wt 022]

(46) *A* 1sg.sbj *tɛ́l* tell *yú,* 2sg.indp *dɛn* 3pl *nó* neg *lɛ́k* like *pɔ́sin.* person 'I tell you, they don't like people.' [ma03hm 010]

The interjections *dí bɔ́y* 'this boy' and *dí gál/gɛ́l* 'this girl' are used as vocatives for calling social equals or inferiors whose names are unknown. These interjections of address are rather informal but not necessarily impolite. They are frequently heard on the streets of Malabo, where they are employed particularly by the youth:

(47) *Dí* this *bɔ́y,* boy *ús=sáy* q=side *dɛn* 3pl *de* ipfv *sɛ́l* sell *brɛ́d* bread *na* loc *yá?* here 'Hey you, where is bread sold around here?' [nn07fn 241]

### **12.2.4 Suck teeth**

The terms "suck teeth", or "kiss teeth", (transcribed as "chip" and glossed as skt) describe a phonetic realisation whose various meanings are determined through pragmatic context. "Suck teeth" is produced by closing the mouth and creating a suction in the oral cavity. The lips are then opened while keeping the teeth closed. The influx of air through the teeth in order to fill the vacuum of the oral cavity produces a release sound followed by a short hiss. "Suck teeth" can be modulated in sound and meaning by manipulating the amount of suction and pursing the lips in varying degrees while the air rushes through. "Suck teeth" therefore covers a range of phonetic realizations situated between the bilabial click [ʘ] and the dental click [ʇ].

"Suck teeth" is employed as a signifier of "negative affect" (Figueroa 2005) throughout large parts of Africa and in African-descended communities of the Americas (Rickford & Rickford 1976). The large range of functions and meanings of "suck teeth" in Pichi corresponds to those recorded for the entire cultural area.

"Suck teeth" is employed as an expressive interjection to convey negatively loaded sentiments ranging from annoyance, irritation, and frustration to exasperation, fatigue, and weariness. In (48), speaker (ed) remembers the hard times he went through as a student in Cuba when the Equatoguinean government stopped paying him his stipend. "Suck teeth" underlines his feeling of exasperation as he delves into his memories:

#### 12 Pragmatic elements and routines

(48) a. *Sɔ́fa* suffer *dán* that *mɔní* money *bin* pst *de* ipfv *dú* do *mí* 1sg.indp *fɔ* prep *ús=tín* q=thing *"chip"?* skt '(The) suffering that [lack of] money caused me for what [skt]?' [ed03sp 099] b. *Tɛ́l* tell *yú,* 2sg.indp *a* 1sg.sbj *sɔ́fa* suffer *dé.* there '(I) tell you, I suffered there.' [ed03sp 100]

In (49), the negative affect associated with "suck teeth" is downgraded to signal a frustrated effort and an ensuing change of heart. After her grandchild has fallen sick in the night, speaker (ab) is at a loss about the appropriate treatment. "Suck teeth" expresses her indecision:

(49) *A* 1sg.sbj *wánt* want *ték* take *solwatá* saltwater *mék* sbjv *a* 1sg.sbj *gí=an,* give=3sg.obj *a* 1sg.sbj *sé* qot *"chip",* skt *nɔ́.* intj 'I wanted to take salt-water and give it to him (and) I said (to myself) [skt] no.' [ab03ab 094]

"Suck teeth" is rarely used to exclusively render inner speech. Rather, there is a smooth transition from expressive to phatic meanings. Sentence (50) is coloured by reproach. Speaker (ye) employs "suck teeth" – albeit with a humorous undertone – to indicate his irritation with the fact that he has not been invited to go eat at Marathon (a restaurant), while his interlocutors had:

(50) *Náw* now *só,* like.that *e* 3sg.sbj *falta,* lack *una* 2pl *dɔ́n* prf *gó* go *na* loc *Marathon,* place *mí* 1sg.indp *nó* neg *gó* go *na* loc *Marathon* place *"chip".* skt 'Right now it remains, you [emp] have gone to Marathon, while I haven't gone to Marathon [skt].' [ye05ce 303]

In sentence (51), speaker (ma) recollects the circumstances of the separation from her husband. "Suck teeth" not only expresses the negative feelings that she recalls. The interjection also communicates to the interlocutor that (ma) attaches a negative moral judgment to the fact that her husband *gɛ́t ɔ́da húman* 'got (himself) another woman':

(51) *E* 3sg.sbj *gɛ́t* get *ɔ́da* other *húman* woman *"chip",* skt *bɔkú* much *problema,* problem *dán,* that *mí* 1sg.indp *bin* pst *dɔ́n* prf *sté,* be.long *wi* 1pl *bin* pst *gɛ́t* get *bɔkú* much *problema* problem *dé* there *áfta/* then

'He got another woman [skt], many problems, that, I [emp] stayed (and) we had many problems at that time, then/' [ma03ni 031]

"Suck teeth" in sentence (52) below combines expressive and phatic meanings in a similar way as in (51) above. Speaker (ed) relates that he had not intended to marry a

12.2 Interjections

woman from his place of origin, *Pagalú*, the island of Annobón, until his mother arranged a marriage for him. "Suck teeth" aptly summarises the negligent disinterest that speaker (ed) states to have had for women from *Pagalú*:

	- b. *"Chip"* skt *a* 1sg.sbj *nó* neg *bin* pst *bísin.* be.busy '[skt] I couldn't care less.' [ed03sp 006]

Feigned disinterest and playful insubordination colour the use of "suck teeth" by female Pichi speakers in flirtatious male-female interaction. A simple "suck teeth" by Beatrice in (53b) is sufficient to ward off the humorous advances of speaker (ye) in (53a). The remonstrative character of "suck teeth" in (b) points towards an area of transition from expressive and phatic to conative functions of "suck teeth":

	- b. *"Chip".* skt [skt]

The conative function of "suck teeth" is brought to its conclusion in (54)(a), where *chip* accompanies an insult. The pungency of the insult, and by extension the gesture of "suck teeth" itself, is commented by speaker(dj) in (54b):

	- 'You take that person for a dog.' [dj07ae 368]

### **12.2.5 The particle** *ó* **'sp'**

The sentence-final particle *ó* plays an important pragmatic role. It is employed for degree modification (cf. e.g. 387 in §7.7.3), may signal clausal focus (cf. §7.4.2), is used as a vocative marker, and provides a means of modifying a sentence with various expressive

#### 12 Pragmatic elements and routines

[ye05ce 079]

and emphatic meanings . The function of the particle also extends into the domain of modality. I analyse the element *ó* as a "sentence particle", because it is never set off by a prosodic break from the utterance it modifies. Instead, *ó* forms a prosodic unit with the preceding utterance. One indication for this is that *ó* normally forms a syllable with the final consonant of the preceding word, i.e. *e bád ó* [é bá tó] '3sg.sbj be.bad sp' = 'it's really bad'.

The particle *ó* serves as a vocative marker in combination with personal names in order to call people from a distance (55) or get their attention during conversation (56). Presumably, it is this function of alerting which lies at the heart of the other uses that follow:


Urgency, advise, and warning colour the sentences in which this particle is used. The following, successively spoken sentences (57a–57c) are characterised by an air of urgency and warning as speaker (ab) relates a near-death experience:

(57) a. *A* 1sg.sbj *dɔ́n* prf *tɛ́l* tell *mi* 1sg.poss *sísta* sister *sé* qot *"na* foc *di* def *pikín* child *dát* that *ó."* sp 'I had already told my sister "mind you, this is the [my] child".' [ab03ay 081] b. *Mék* sbjv *yu* 2sg *mɛ́n=an* care.for=3sg.obj *ó!'* sp 'Be sure to take good care of her [because I'm going to die].' [ab03ay 082] c. *A* 1sg.sbj *dɔ́n* prf *de* ipfv *gó* go *ó,* sp *a* 1sg.sbj *dɔ́n* prf *de* ipfv *gó* go *ó.* sp

'I'm going [dying], I'm going.' [ab03ay 083]

Further gradations of the meanings of *ó* are found in the following sentence. In the example, *ó* assumes the function of a modal particle, a marker of assertion, which signals commitment by the speaker to the truth of the proposition:

(58) *Yɛ́s,* yes *a* 1sg.sbj *sabí* know *dé* there *yɛ́s,* yes *bɔt* but *a* 1sg.sbj *nɔ́ba* neg.prf *ɛ́nta* enter *ínsay* inside *ó.* sp 'Yes, I know that place, yes. But mind you, I haven't entered the place before.' [ma03hm 016]

12.3 Terms of address

### **12.3 Terms of address**

Often, the African and European given names of individuals are only known to relatives and close friends. Peers tend to address each other by nicknames which may be conferred on an individual during interaction with family members, friends, the neighbourhood, and the wider community. Nicknames may also change in the passage of time as new events come to mark a person's daily life.

I list three nicknames in (59) that are used by peers most of the time in addressing their bearers. The bearer of the first name is female, the second and third names are borne by men. As can be seen, peer nicknames tend to be characterised by an air of informality:


*Lage* was born aboard a flight from Madrid to Malabo, operated by the now defunct National Airline of Equatorial Guinea, in Spanish *Líneas Aéreas de Guinea Ecuatorial* (abbrev. LAGE). Her birth back then was the talk of the town and the name stuck for a life time. *Boyé Loco*'s name is composed of his Bube given name *Boyé* and the Spanish adjective *loco* 'crazy' due to his reputation as a charismatic *bon vivant*. The byname *Johnson* originates in the brand name *Johnson Power Systems*. Due to the unreliablity of power supply in Malabo, generators produced by *Johnson* are ubiquitous in Malabo. The nickname is a humorous allusion to the bearer's supposed sexual prowess.

Figure 12.1 presents the degree of formality from informal (the –pole) to very formal (the +pole) attached to the terms of address covered in the following (Mühleisen 2005: 209). The corresponding kinship terms can be taken from Figure 12.2 further below.


Figure 12.1 Degree of formality of terms of address

Spanish honorifics are employed for the most formal degrees of relationships between interactants. Without doubt, this circumstance is intimately tied to the status of Spanish itself as a language of dominance, distance, and social asymmetry. The address terms *señor* (male) and *señora* (female) are in use with first names (FN) or last names (LN). The latter option follows Spanish usage (i.e *Señora Belobe Toichoa* 'Ms Belobe Toichoa') and is commonly employed in symmetrical or asymmetrical relations in institutional or work contexts in the formal sector of the economy.

The former option, *señor* or *señora* with an FN (i.e. *Señora Maura*, *Señor Javier*), is not common in Spanish. In Pichi, it is a means of respectfully addressing an already familiar,

#### 12 Pragmatic elements and routines

social superior in less formal situations than the ones appropriate for *señor/a* and LN. The use of *señor*/*a* and FN parallels that of the Spanish honorifcs *don* (male) and *doña* (female) followed by FN, for elderly and respected members of the communty, i.e *Don Samuel* and *Doña Cristina*. The combination *don*/*doña* and FN is, however, current in Spanish.

Two Pichi address terms of a high degree of formality are, also in use, namely *sá* 'sir' and *má* 'madam, mother'. These two terms are used as address terms and sentence-final address tags when interacting with an elder of higher social rank, usually without an FN or LN. For example *sá* 'sir' can be found in the respectful speech of a well-behaved child or youngster when replying to an enquiry by an elder. Compare the following answers by a child to a when enquiry by a female elder (60) and a yes-no question by a male elder who is not a family relative (61):


Kinship-based terms of address are situated in the middle range of formality and may be used in addressing familiar persons or strangers. The dimension of age naturally relates to the degree of formality in so far as senior members of society are more likely to be addressed by one of the more formal terms of address in Figure 12.1. Under normal circumstances, the use of an FN presupposes the use of a kinship term if the addressee is older than oneself (i.e. *Mamí Rose* 'mother Rose' = 'respectful address term for Rose, who is of my mother's generation'). The use of a first name alone for an older person is highly inappropriate. For people of the same age group, and young people in particular, kinship terms are, however, not required as terms of address. Social equals may refer to each other by their first names or their nicknames alone.

Figure 12.2 provides the kinship-based address terms referred to in Figure 12.2 arranged along the dimension of age.

As a general principle, any of the address terms listed may be combined with an FN. In practice, an FN hardly ever follows the same generation terms *cuñado*/*a* 'brother/sisterin-law', *sísta '*sister' or *brɔ́da* 'brother'. At the same time, the use of an FN with a kinship term for an addressee one or two generations older tends to be avoided as well unless there is a high degree of familiarity and/or an actual kinship relation between the interlocutors. Compare the following combination of address term and FN:

(62) *Tía* aunt *Tokó,* name *ús=sáy* q=side *yu* 2sg *de* ipfv *gó?* go 'Auntie Tokó, where are you going?' [ye07fn 213]

#### 12.3 Terms of address


Figure 12.2 Kinship-derived terms of address

I should point out that Spanish kinship terms form an integral part of the address system of Pichi. The Spanish terms *cuñado*/*a* 'brother/sister-in-law' – with *cuñado* invariably being pronounced as [kùnjáò] – have been appropriated and changed in their meaning. In Pichi, these two kinship terms function as markers of aknowledgment and solidarity amongst peers. They are therefore used to address any person of the same generation, whether related or not. In this function, *cuñado*/*a* are far more common than the equivalent *sísta* 'sister' and *brɔ́da* or *brá* 'brother':

(63) *Cuñado,* brother-in-law *mí* 1sg.indp *gɛ́fɔ* have.to *gó* go *fɛ́n* look.for *dán* that *mi* 1sg.poss *prima* cousin *ó,* sp *Cristina.* name 'Brother(-in-law), I [emp] really have to go look for that my (female) cousin, Cristina.' [ge07ga 045]

In the same vein, the Spanish kinship terms *tía* 'aunt' and *tío* 'uncle' are equally common as *antí* 'aunt' and *ɔnkúl* 'uncle' as terms of address. The same holds for the Spanishderived terms *abuela* 'grandmother' and *abuelo* 'grandfather' as opposed to *gran-má* 'grandmother' and *gran-pá* 'grandfather'. However, the Pichi words *antí* 'aunt' and *ɔnkúl* 'uncle' are more often used to denote the kinship relation as such (64):

(64) *E* 3sg.sbj *mít* meet *mi* 1sg.poss *antí.* aunt 'He met my aunt.' [fr03ft 084]

Conversely, the Spanish words *abuela* 'grandmother' and *abuelo* 'grandfather' are more common as terms of address and at least as common as *gran-má* 'grandmother' and *gran-pá* 'grandfather' in denoting the kinship relation as such:

(65) *Abuela,* grandmother *Guinea* Equatorial.Guinea *fít=an?* fit=3sg.obj 'Grandmother, (so) Equatorial Guinea is good for him?' [fr03ab 171]

Since first names are not normally used to refer to social superiors, including next of kin, a kinship term will normally be used to refer to an older common kin. In (66), 12 Pragmatic elements and routines

speaker (ro) is conversing with her nephew. She refers to her own husband as *yu ɔnkúl* 'your uncle':

(66) *Yu* 2sg *ɔnkúl* uncle *nó* neg *gɛ́t* get *nó* neg *hambɔ́g* bother *fɔ* prep *chɔ́p.* food 'Your uncle [my husband] is not picky about food.' [ro05rt 058]

### **12.4 Greetings and other routines**

A general greeting routine is normally initiated by addressing an individual with the phrase in (67) and a group of people by (68). These phrases may be reformulated at will to enquire after the health of partners, children, or other relatives (69). A general observation is that conventional Spanish greeting routines are widely used together with Pichi routines (i.e. *buenos días* 'good morning'):


The enquiry is usually replied to by one of the phrases in (70–72):


The most wide-spread greeting formula amongst the youth or peers and in relaxed and informal social settings is featured in (73). This greeting involves the element *fá*, not found in any other context in Pichi, but almost certainly derived from the English

12.4 Greetings and other routines

word *fashion.* It is also attested in Krio, Nigerian Pidgin and Cameroon Pidgin. Notably, it is also found in Maroon Spirit Language (Jamaica, Bilby 1983: 50), as well as in Sranan and the other creoles of Suriname in the almost identical form *o fa* 'which fashion, how' (Wilner 1994: 50). In Pichi, a common reply to the idiom is (74):


Longer exchanges of greetings are usually initiated by employing the property item *gúd* 'be good' together with the noun that denotes the period of the day in which the greeting takes place. The resulting collocations constitute greeting formulas by themselves but are very often followed by one of the general greeting formulas in (67–69) above. The collocation *(gúd) mɔ́nin* 'good morning' or a simple *mɔ́nin* 'morning' is used from sunrise to noon (75):

(75) *'Gúd* good *mɔ́nin'* morning *na* foc *sóté* until *las* the.pl *doce.* twelve 'Good morning is until twelve o'clock.' [ye07je 015]

*Gúd ívin* is used from noon to sunset (76). The collocation *gúd áftanun* 'good afternoon' is sometimes used by Group 1 (cf. §1.3) speakers instead of *gúd ívin*, but it is virtually absent from the speech of Group 2 speakers:

(76) *Frɔn* from *las* the.pl *doce,* twelve *sóté* until *e* 3sg.sbj *gó* go *las* the.pl *seis,* six *na* foc *"gúd* good *ívin".* evening 'From twelve to six o'clock, its "good evening".' [ye07je 011]

The collocation *gúd náyt* 'good night' is used after night has fallen. The presence of the otherwise rare variant *náyt* 'night' in the greeting instead of *nɛ́t* 'night' is indicative of the formulaic, lexicalised character of the collocation.

Also note the apposition of the 2pl pronoun *una* when a greeting is directed to more than one person. The use of *yu* '2sg' in the same position as *una* in greetings directed at an individual is not attested. Responses to greetings usually involve the repetition of the corresponding phrase by the interlocutor:

(77) *Una* 2pl *gúd* good *náyt.* night 'Good night to you [pl].' [ye07je 045]

#### 12 Pragmatic elements and routines

Other greetings are issued on specific occasions rather than periods of the day. On the occasion of imminent travel, the most common way of bidding farewell is by saying *wáka fáyn* 'walk fine'. Upon arrival, the traveler is greeted by *wɛ́lkɔm* 'welcome'.

The greeting formula *kúsɛ́* (< Yoruba *kuṣẹ*, cf. Abraham 1958) is said as a token of encouragement and empathy towards one or more people engaged in physically strenuous work (e.g. a group of construction workers working on the road). *Kúsɛ́* is also used to congratulate a person for their good work:

(78) *Una* 2pl *kúsɛ́* good.job *ó!* sp

'(We) encourage you [pl] in your good work!' [ye07je 028]

Gratitude is expressed by means of *tɛ́nki* 'thank you' (79a). Reply options are provided in (b) and (c). Note that *fɔ nátin* (b) and *na nátin* (c) are calques from Spanish *dé nada* 'of nothing' = 'you're welcome':

	- c. *Lɛ́f,* leave *na* foc *nátin.* nothing 'Don't mention it, it's nothing.' [ye07fn 503]

## **13 Pichi and Spanish in contact**

The integration of Spanish elements into Pichi discourse is thoroughly conventionalised, and encompasses borrowing, calquing and codeswitching. Many of the mixing phenomena that can be observed are not "interactionally meaningful" (Auer 1998: 20) and point towards codeswitching as an "unmarked choice" (Myers-Scotton 1993), i.e. the normal way of speaking Pichi (cf. Yakpo 2015; 2017a). I summarily refer to the Pichi-Spanish contact phenomena described in this chapter by the cover term "codemixing" (Muysken 2000). This implies patterned and "sedimented" (Auer 1999) uses of non-native elements in multilingual interactions. Codemixing therefore forms an integral part of the grammar and pragmatics of Pichi (cf. Yakpo 2009b, Yakpo 2018). In this chapter, all Spanish elements are set in bold.

### **13.1 Patterns of contact**

Codemixing systematically affects different areas of Pichi grammar and lexicon and it does so with differing frequency and depth.The use of certain lexical items and structures involving Pichi and Spanish material is so conventionalised that they can be said to constitute an integral part of the grammatical system and lexicon of Pichi. Table 13.1 summarises some of the most conventionalised patterns of Pichi-Spanish codemixing.

Muysken (2000) identifies three patterns of codemixing that accommodate cross-linguistic mixing phenomena: insertion, alternation, and congruent lexicalisation. All three of these patterns are operative in Pichi-Spanish codemixing. But the type of back-andforth switching characteristic of much of Pichi discourse points towards a prominent role of congruent lexicalisation: Material from either language is grafted on grammatical structures common to both languages. Consider the following example:

(1) *A* 1sg.sbj *kɔmɔ́t* leave *colegio,* high.school *a* 1sg.sbj *dé* be.loc *fuera* outside *con* with *mi* 1sg.poss *mísis* matron *cuatro* four *años* year.pl *a* 1sg.sbj *nó* neg *ték* take *bɛlɛ́,* belly *a* 1sg.sbj *nó* neg *lɛ́f* leave *mi* 1sg.poss *vájin.* virginity 'I came out of high school, I was outside with my guardian for four years, I didn't become pregnant, I didn't give up my virginity.' [ab03ay 132]

While the noun *colegio* 'college' looks more like an insertion into a Pichi grammatical structure (the noun is left unmarked like a Pichi noun in this position), the switch *fuera con mi* 'outside with my' is best understood as an instance of congruent lexicalisation. Each element could be replaced by the corresponding Pichi elements *nadó wet*

#### 13 Pichi and Spanish in contact

#### Table 13.1 Patterns of Pichi-Spanish contact


*mi*. In this context the possessive pronoun *mi* '1sg.poss' is of particular interest. It is a homophonous diamorph, a morpheme that is identical in form and function in both languages including its suprasegmental feature of low tonedness in Pichi and Equatoguinean Spanish. Besides that, *mi* functions as a possessive pronoun through juxtaposition with the possessed noun in both languages.

I subjected a smaller section of the corpus consisting of a total of 22,059 words (or tokens, i.e. occurrences of words, irrespective how many times they occur) to a thorough analysis. The subcorpus contains 1475 types (different words). The analysis reveals that the presence of Spanish types and tokens in the Pichi texts varies with word classes in the ways listed in Table 13.2.

With respect to types, Table 13.2 shows that a total of 50% of all nouns and approximately 28% of all verbs that occur are Spanish. Property items (or "adjectives" in Spanish) were counted separately and amounted to a total of 44% of Spanish types. For numerals, the Spanish percentage stands even higher at 62%.

However, the percentage of Spanish tokens (i.e. total instances of occurrences even if the same word occurs several times) reveals a different picture. Numerals still top the list (47%). But they are followed by a much lower percentage of Spanish nouns (19%) and adjectives (18%). This shows that the frequency with which Spanish words are used is considerably lower than the absolute number of Spanish words in Pichi discourse.

#### 13.2 Specific constituents


Table 13.2 Type-token analysis of Spanish words in Pichi discourse

With the exception of numerals, the Spanish ratio of tokens stands at roughly 20% of an average text.

### **13.2 Specific constituents**

The following four sections describe the specifics of codemixing involving noun phrases, verbs and adjectives, functional elements, and other constituents.

### **13.2.1 Noun phrases**

Inserted Spanish constituents belong to various word classes, but the insertion of content words, and nouns in particular, prevails. Thus we find *novio* 'fiancé' and *pueblo* 'village' in (2). Note that both Spanish nouns are objects of Pichi elements, the first of a verb, the second of a preposition:

(2) *Mék* sbjv *yu* 2sg *nó* know *sé* qot *yu* 2sg *dɔ́n* prf *gɛ́t* get *novio* boyfriend *na* loc *pueblo,* village *na* loc *kɔ́ntri.* country

'You should know that you already have a fiancé in the village, in the hometown. [ab03ay 010]

When Spanish nouns are inserted as in (2), they usually remain bare where Pichi nouns do so, or are accompanied by Pichi determiners and the pluraliser *dɛn* 'pl' in the same way as Pichi nouns are. In (3), the definite Spanish noun *paciencia* 'patience' is preceded by the Pichi definite article *di*:

(3) *Porque* because *fɔ́s,* first *di* def *paciencia,* patience *yu* 2sg *nó* neg *go* pot *gɛ́t=an.* get=3sg.obj 'Because first, the patience, you wouldn't have it.' [fr03ft 189]

When a specific Spanish plural noun is inserted, there is a strong likelihood that it will be additionally marked with the postposed Pichi pluraliser *dɛn*, in accordance with the

#### 13 Pichi and Spanish in contact

pattern that applies to Pichi count nouns (4). Conversely, Spanish nouns exhibit a strong tendency to occur devoid of Pichi number and definiteness marking where the noun is non-specific as with *rallador* 'grater' in the second example:


The occurrence of *pruebas* 'proofs' in (6) demonstrates that Spanish nouns may well be devoid of Pichi noun phrase marking, but not necessarily so of the Spanish plural morpheme {*-s*}:

(6) *Yu* 2sg *go* pot *gɛ́t* get *prueba-s.* proof-pl 'You will have proof.' [ma03sh 013]

This is not surprising however, since in Spanish, determiner-less plural count nouns may have non-specific reference. The semantic overlap between Spanish plural nouns and Pichi bare nouns in codemixing can be seen in (7). Here the Pichi bare nouns *pía* 'avocado' and *sadín* 'sardine' are functionally equivalent to the Spanish plural noun *tomates* 'tomatoes':

(7) *Mí* 1sg.indp *wet* with *Rubi* name *wi* 1pl *mék* make *jwɛn-jwɛ́n,* red.cpd-join *wi* 1pl *báy* buy *pía,* avocado *wi* 1pl *báy* buy *sadín,* sardine *wi* 1pl *báy* buy *tomates,* tomatoes *wi* 1pl *desayuna.* have.breakfast

'Me and Rubi, we teamed up and bought avocados, we bought sardines, we bought tomatoes, we had breakfast.' [ye03cd 152]

The insertion of larger nominal groups as opposed to single nouns is rarer. In fact, most of the Spanish adjective-noun combinations we encounter are collocations that are somewhat lexicalised in Spanish. Compare *traducción directa* 'direct translation' in (8):

(8) *Na traducción directa e mék.*

> foc translation direct 3sg.sbj make

'It's a direct translation that she made.' [to03gm 042]

The order of constituents normally remains unchanged when Spanish elements are inserted into a Pichi NP. In (9), the Pichi quantifier*lás* 'last' is used in prenominal position with the inserted Spanish noun *semana* 'week'. However, note that Spanish also features a quantifier + noun order in NPs (i.e. *la última semana* '(the) last week'):

13.2 Specific constituents

(9) *Ɛf* if *yu* 2sg *bin* pst *kán* pfv *bigín* begin *las* last *semana* week *yu* 2sg *bin* pst *fɔ* cond *dɔ́n* prf *fínis* finish *tidé.* today

'If you had begun last week you would have been finished today.' [dj05ae 057]

I would assume that the inverse NP constituent order (noun + adjective in the majority of cases) of Spanish NPs blocks the admixture of single Spanish attributive adjectives into Pichi NPs (cf. Sankoff & Poplack 1981). This is largely borne out by the data.

There is, however, some variation, although it is not all that frequent. In (10), the Spanish adjective *directo* 'direct' occurs after the Pichi noun *ɔnkúl* 'uncle' in a Pichi NP and thereby follows the constituent order of a Spanish NP:

(10) *Na* foc *wán* one *ɔnkúl* uncle *directo,* direct *fɔ* prep *mi* 1sg.poss *mamá* mother *in* 3sg.poss *papá* father *in* 3sg.poss *fámbul* family *pát.* part

'He's a direct uncle on the part of my mother's father's family.' [fr03ft 051]

In (11), we find the opposite situation. The Spanish adjective *especial* 'special' is in a prenominal position, hence in the syntactic slot of attributively used Pichi property items:

(11) *E* 3sg.sbj *bríng* bring *fís,* fish *e* 3sg.sbj *kúk* cook *sɔn* some *especial* special *fís,* fish *e* 3sg.sbj *gí* give *mí* 1sg.indp *mék* sbjv *a* 1sg.sbj *chɔ́p.* eat 'She brought (a) fish, she cooked a particular fish and gave it to me in order to eat.' [ed03sb 015]

There are other instances of Spanish adjectives that follow Pichi nouns in Pichi NPs. But in these cases, the function of the Spanish words parallels that of some Pichi value property items that are used as adverbials in the same syntactic position. The Spanish adjective *serio* 'serious' in (12) may be likened to the Pichi manner adverb *fáyn* 'well, really' in (13):


13 Pichi and Spanish in contact

### **13.2.2 Verbs and adjectives**

The low ratio of Spanish verbs as opposed to nouns in the type and token count may be striking at first glance. However, this tendency may stem from the fact that a small number of high frequency Pichi verbs (e.g. *mék* 'make', *gɛ́t* 'get, have', *gí* 'give') participate in conventionalised verb-noun collocations, in which a Pichi verb is followed by a Spanish noun (cf. §9.3.1 for an extensive treatment). Some of these are *gí permiso* 'give permission', *mék rabia* 'be annoyed', *gɛ́t novio/novia* 'have a boy/girlfriend'. The collocations also include calques from Spanish. Compare *gí wán vuelta* 'give one round' = 'take a walk' which is a one-to-one calque of Spanish *dar una vuelta*:

(14) *E* 3sg.sbj *de* ipfv *gí* give *wán* one *vuelta* round *kwík.* quickly 'She's taking a walk quickly.' [dj05be 120]

The admixture of Spanish verbs follows established rules. Spanish verbs are always inserted into Pichi clauses in an invariant form of the 3sg person of the Spanish present tense paradigm. This insertion rule is valid without exception across the three regular Spanish verb inflection classes. Due to its frequency, the 3sg present tense form is also the default form found in most contact scenarios involving Spanish (Clements 2009: 20– 21). Examples follow with *controla* 'control' (<*controlar*) in (15), *entiende* 'understand' (<*entender*) in (16), and *sufre* 'suffer' (<*sufrir*) in (17):


The 3sg invariant form is combined with Pichi TMA markers like any Pichi verb as can be seen by the presence of *kin* 'hab' in (15) and *go* 'pot' in (16) above. Inserted Spanish verbs may also be reduplicated by the same derivational process that applies to Pichi verbs. Compare *pica-píca* 'red.cpd-cut.up' = 'repeatedly cut up (into small pieces)' in (18):

(18) *A* 1sg.sbj *bigín* begin *de* ipfv *pica-píca,* red.cpd-cut.up *wi* 1pl *fráy* fry *patata,* potato *wi* 1pl *fráy* fry *plantí.* plantain 'I began to (repeatedly) snip [the trimmings], we fried potatoes, we fried plantain.' [ye03cd 172]

13.2 Specific constituents

Pichi exhibits a phonologically conditioned suppletive allomorphy in the pronominal system. The lexical pitch configuration of a verb determines the choice of allomorph used for the expression of 3sg pronominal object case (cf. §3.2.5). Vowel-final verbs with a word-final low tone take the object pronoun *ín* '3sg.indp' – this group includes a few Pichi verbs and all inserted Spanish verbs (19). This is because the 3sg invariant form of the Equatoguinean Spanish verb always features a word-final L-toned vowel:

(19) *Fíba* fever *nó* neg *sube* go.up *ín.* 3sg.indp 'The fever hasn't risen on him.' [eb07fn 171]

The form *sigue* (<*seguir*) 'follow, continue' is highly conventionalised in its use. It is also employed as an auxiliary verb to indicate continuative aspect in a complement construction:

(20) *A* 1sg.sbj *go* pot *sigue* continue *chɔ́p.* eat 'I'll continue eating.' [be05 057]

In a similar vein, the verbs *sube* (<*subir*) 'go up' and *baja* (<*bajar*) 'go down' are far more frequent than their Pichi counterparts *gó ɔ́p* and *gó dɔ́n* (21):

(21) *Bɔkú* much *motó* car *dɛn* pl *dé* be.loc *yá* here *só,* like.that *a* 1sg.sbj *nó* neg *nó* know *sé* qot *Pancho* name *mék* make *lɛk* like *sé* qot *e* 3sg.sbj *de* ipfv *sube* go.up *bihɛ́n* behind *wé* sub *e* 3sg.sbj *baja* go.down *mɔ́.* more 'So many cars were there, I didn't know Pancho pretended to go up behind us (and) went down again.' [ye03cd 178]

Spanish adjectives do not only occur as attributes to Pichi nouns. They are systematically inserted into Pichi predicate adjective clauses as complements to the locativeexistential copula *dé* 'be.loc' (22).

(22) *Wán* one *yáy* eye *dé* be.loc *blanco* white *e* 3sg.sbj *nó* neg *de* ipfv *sí.* see 'One eye is white, it doesn't see.' [ye03cd 106]

Neither adjectives nor past participles usually exhibit Spanish-style gender agreement with the subject and are normally inserted in the masculine form. However, past participles always come along with the regular Spanish adjective-deriving morphology (23).

(23) *A* 1sg.sbj *wánt* want *dé* be.loc *flipa-do* turned.on-adj *ɔ́l* all *áwa,* hour *ɔ́l* all *áwa.* hour 'I want to be turned on all the time, all the time.' [ye07ga 012]

#### 13 Pichi and Spanish in contact

I have shown that a handful of Pichi property items may be employed as adjectives and inchoative-stative verbs alike (cf. §7.6.5). When used as adjectives, these property items denote a non-time-stable body state and may appear as complements to the copula *dé.* When used as inchoative-stative verbs, these property items denote a time-stable value. The property item *bád* 'be bad' displays this kind of behaviour. Hence, *bád* means '(intrinsically) bad' (24) when used as an inchoative-stative verb and 'ill' when it appears as a complement to the copula *dé* (25):


Spanish also exhibits a distinction based on time-stability with respect to property items. In contrast to Pichi, the distinction may, however, be applied to almost any adjective of the language. Examples (26) and (27) involve the 3sg present form of the timestable identity copula *ser* and the 2sg present of the non-time-stable locative-existential copula *estar,* respectively. A comparison of the Pichi examples in (24) and (25) above with the two sentences below show the functional overlap of the relevant constructions in the two languages:

(26) Spanish *Es* He.is *malo.* bad 'He is bad.' (Own knowledge)

(27) Spanish *¿Estás* You.are *mal* bad *hoy?* today 'Do you feel bad today?' (Own knowledge)

Despite the similarities between the *dé* + property item construction and the Spanish *estar* + adjective construction, all predicatively used Spanish adjectives always appear as complements to the Pichi locative-existential copula *dé*, regardless of whether the denoted property is non-time-stable or time-stable.

Hence the time-stable property denoted by the Spanish adjective *blanco* 'white' appears as a complement to the copula *dé* in (22) above, while the Pichi colour term *wáyt* 'be white' can only be employed as a inchoative-stative verb as in (28):

```
(28) Di
      def
          mán
          man
               wáyt.
               be.white
      'The man is white.' [ed05fn 077]
```
13.2 Specific constituents

Why is the time-stability distinction not maintained with predicatively used Spanish adjectives? An explanation is that the Pichi construction involving the copula *dé* and an adjectival complement is more compatible with congruent lexicalisation than the use of Spanish adjectives as (inchoative-)stative verbs. With the former pattern, the phrasal syntax of adjectival predication remains identical in both languages. This allows speakers to graft such codemixed constructions onto a common grammatical structure (cf. Meechan & Poplack 1995). Pichi-Spanish contact in the predicate adjective construction has therefore led to the generalisation of a rather marginal structure specialised to a handful of Pichi property items. The obligatory use of a copula in these mixed collocations may also be seen as a case of structural interference from Spanish where a copula verb *must* be used in predicate adjective constructions.

### **13.2.3 Functional elements**

The most frequently used Spanish functional elements are the cause clause linkers *como* 'since' (29) and *porque* 'because' (30). Both linkers form an integral part of the Pichi system of clause linkage and are best seen to have been borrowed into the language:


The linkers *como* and *porque* are employed in the same syntactic position as the Pichi equivalents *as* 'as' (31) and *bikɔs* 'because' (32), respectively:

(31) *'As* as *in* 3sg.poss *sísta* sister *dɛn* 3pl *bin* pst *de* ipfv *kɔ́l* call *in* 3sg.poss *mamá* mother *sé* qot *sísta,* sister *in* 3sg.indp *de* ipfv *kɔ́l* call *in* 3sg.poss *mamá* mother *sé* qot *sísta.* sister

'As her sisters would call her mother sister, she [emp] would call her mother sister.' [ab03ay 145]

(32) *Bɛt* but *a* 1sg.sbj *dɔ́n* prf *nó* neg *wétin* what *yu* 2sg *níd,* need *bikɔs* because *wi* 1pl *gɛ́t* get *sɔn* some *prɔ́blɛm* problem *wé* sub *wi* 1pl *de* ipfv *tɔ́k* talk *Pichi* Pichi *na* loc *Malábo.* place

'But I already know what you need, because we have a problem when we talk Pichi in Malabo.' [au07se 005]

#### 13 Pichi and Spanish in contact

Table 13.3 shows the frequency with which the Spanish linkers *como* and *porque* occur in Pichi sentences in relation to *as* and *bikɔs*. The table indicates that in the overwhelming majority of cases (89% for*como* and 91% for *porque*) both conjunctions occur as single constituents in Pichi clauses rather than in clausal switches in which the following material is also in Spanish. The second line of Table 13.3 shows that these two Spanish function words are established loans. In 76% of all occurrences, 'since' is expressed as *como*, hence only 24% is expressed with the Pichi equivalent *as*. In 41% of all cases 'because' is expressed as *porque*, so Pichi *bikɔs* occurs as the causal conjunction in 59% of all cases.

Table 13.3 Distribution and frequency of *como* and *porque*


The clause linker *aunque* 'although' occurs so frequently that it is best seen to be fully integrated into the Pichi lexicon as well. In Spanish too, *aunque* is used both as a concessive or adversative conjunction as in (33) and as a a similative adverbial as in (34):


'When you make it at home, just take it (and) put it into a pot, like one cup or so.' [dj03do 010]

The Spanish time clause linker *mientras* 'while' occurs less systematically, but it still provides an optional resource for combining clauses:

(35) *Mientras* while *yu* 2sg *de* ipfv *sí* see *sé* qot *di* def *tín* thing *de* ipfv *transforma* transform *pɔ́sin* person *yu* 2sg *de* ipfv *kɔ́t* cut *wán* one *tín* thing *fɔ* prep *in* 3sg.poss *fínga.* finger

'While you see that the thing is turning into a human-being you cut off a part of its finger.' [ma03sh 012]

The Spanish coordinator pair *ni – ni* 'neither – nor, not even' can express negative disjunction in Pichi utterances. Like in Spanish, *ni* can be used alone (36) or in discon-

13.2 Specific constituents

tinuous negation (37). Unlike in Spanish, however, subject disjunction in Pichi requires the kind of negative concord characteristic of other negative clauses in Pichi (37):

(36) *E* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *ni* neither *Panyá,* Spanish *e* 3sg.sbj *sé* qot *e* 3sg.sbj *wánt* want *muchachita* young.girl *de* of *diecisiete* seventeen *años.* year.pl

'He doesn't even know how to speak Spanish, (and) he says he wants a girl of seventeen years.' [ye03cd 053]

(37) *Ni* neither *ín* 3sg.indp *ni* neither *in* 3sg.poss *brɔ́da* brother *dɛn* 3pl *nó* neg *lán.* learn 'Neither him nor his brother studied.' [ro05de 145]

In (38), we find the cardinal numeral *wán* 'one' in a peculiar construction with the meaning 'around' in combination with quantity expressions. When *wán* is employed in this way, it usually modifies NPs containing numerals (38) and time units (39):

(38) *Yu jɔ́s gɛ́t wán diecisiete años.*

> 2sg just get one seventeen year.pl

'You're just about seventeen years old.' [ab03ay 105]

(39) *Tumɔ́ro* tomorrow *mɔ́nin* morning *tɛ́n,t* time *wán* one *las* the.pl *siete* seven *só,* like.that *a* 1sg.sbj *go* pot *gó* go *dé.* there 'Tomorrow in the morning, around seven or so, I will go there.' [ye03cd 011]

I attribute this particular usage of the numeral *wán* to structural borrowing from Spanish. In Spanish, the plural indefinite articles *unos, unas* serve the same function (40).

(40) Spanish

*Me* Me *faltan* they.lack *unos* one.pl *dos* two *mil* thousand *francos.* franc.pl

'I am short of some 2000 francs CFA.' (Own knowledge)

### **13.2.4 Other constituents**

Spanish discourse markers and adverbs frequently occur at the beginning of a sentence. Speakers often use Spanish material that is not syntactically integrated into a Pichi clause structure. This includes the high frequency adverbs *bueno* 'well' (41) and *pero* 'but' (42). Conversely, the interjection *chico* 'boy' (42) is not common in European Spanish. It might have developed in Equatoguinean Spanish and Pichi through mutual reinforcement and calquing of other person-denoting interjections in Pichi and other Equatoguinean languages.

13 Pichi and Spanish in contact


The interjection *chico* 'boy' in (42) above is more common than other human-denoting Pichi equivalents such as *mán* 'man', *papá* 'father', or *mamá* 'mother'*.* The Spanish noun *mierda* 'shit' is very common as a deprecative interjection (43):

(43) *Mierda* shit *mierda,* rep *ús=sáy* q=side *e* 3sg.sbj *pás?* pass 'Shit, shit, which way did she go?' [ro05rt 002]

Whole adverbial phrases are also admixed in this way. Like discourse markers, these occur at the beginning or the end of a clause:

(44) *A* 1sg.sbj *fít* can *hól* hold *dán* that *mɔní* money *durante* during *un* one *mes* month *entero.* entire 'I can keep that money during an entire month.' [ro05rt 049]

Alternation may also involve larger syntactically independent chunks of Spanish up to a clause boundary:

(45) *A* 1sg.sbj *bɔ́n* be.born *nayntín* nineteen *twɛnti-fó,* twenty.cpd-four *por lo tanto* therefore *ahora* now *tengo* I.have *ochenta* eighty *años.* year.pl 'I was born in 1924, therefore I am now eighty years old.' [ab03ay 007]

The Spanish focus syntagma *es que* 'it is that' may also be seen as a peripheral element which constitutes an independent syntactic unit (46). However, *es que* is so much an integral part of the Pichi system of focus marking that it seems like a holophrastic borrowing (cf. §7.4.3 for more). Also note the interesting switch to Spanish at the clausal boundary between the relative clause and the following main clause.

(46) *Es* It.is *que* that *húman* woman *wé* sub *e* 3sg.sbj *gɛ́t* get *bɛlɛ́* belly *siempre* always *suele* usually *ser* be *así.* like.that 'It's that women who are pregnant are always like that.' [ro03rr 008] 13.3 Specific semantic fields

### **13.3 Specific semantic fields**

Some semantic fields are more regularly affected by codemixing than others. Numerals and other similarly tightly interwoven semantic fields like the expression of time or colour are characterised by the extensive use of Spanish words and structures. In many instances, the corresponding Pichi expressions are no longer used or are falling out of use. The corresponding Spanish words and structures have been borrowed into Pichi.

### **13.3.1 Numerals, days, and dates**

In natural speech, the occurrence of Pichi cardinal numerals drops rapidly after*trí* 'three'. The percentages of attributive cardinal numerals of Pichi and Spanish provenance in the corpus are presented in Table 13.4. Borrowing has had a profound impact on the Pichi numeral system, where Spanish numerals have substituted all but the basic Pichi numerals below eight. Note that this table only lists the usage of *wán* 'one' as a cardinal numeral and does not include *wán* in its use as an indefinite determiner with the meaning 'a'.


Table 13.4 Use of Pichi numerals

The attributive use of Spanish numerals goes along with the insertion of Spanish head nouns – there is no instance of a mixed combination of a Spanish numeral and a Pichi noun:

(47) *Lɛ́f=an* leave=3sg.obj *mék* sbjv *e* 3sg.sbj *rích* reach *a* to *los* the.pl *quince* fifteen *años.* year.pl 'Leave her, let her reach [the age of] fifteen years.' [ab03ay 138]

When telling the time, Spanish lexical items are fit into a conventionalised mixed construction which does not have an exact equivalent in Spanish. In the Pichi construction, the clock time is an adverbial complement to the locative-existential copula *dé* 'cop'. The copula, in turn, takes the 1pl subject *wi* (49). In the Spanish construction, the clock time functions as the subject of the identity copula *ser* 'be' (49):

13 Pichi and Spanish in contact


Equally, the majority of speakers employ Spanish dates. One of the few tokens of a date featuring Pichi numerals was produced by a lady of more than eighty years of age (51). I assume this instance and the few other similar ones in the corpus to be holophrastic insertions. This view is supported by the fact that the date in (51) is the speaker's date of birth and perhaps just as significantly, she was married to a Nigerian in her youth. Other than that, this speaker's use of numerals parallels the one outlined in Table 13.4 above:


Most speakers are not familiar with Pichi day names and employ the Spanish day nomenclature (52). Even older speakers rarely if ever use the corresponding Pichi day names *mɔ́nde* 'Monday', *tyúsde* 'Tuesday', *wɛ́nsde* 'Wednesday', *tɔsde* 'Thursday', *frayde* 'Friday', *sátidé* 'Saturday', and *sɔ́nde* 'Sunday' (53):


The elicitation of Pichi day names with two speakers below twenty-eight years was unsuccessful save *sɔ́nde* 'Sunday', certainly because of its social importance for religious practice. A speaker above fifty-five years experienced considerable difficulties in retrieving Pichi day names (54a–c). *Wɛ́nsde* 'Wednesday' was only retrieved after an external input (b) and the elicitation of 'Thursday' and 'Friday' produced the misnomers *tyúsde* 'Tuesday' (c) and *wɛ́nsde* 'Wednesday' (d), respectively:

(54) a. *Miercoles* wednesday *na,* foc *áy,* intj *pero* but *a* 1sg.sbj *sabí=an.* know=3sg.obj '"Wednesday" is, ah [pause], but I know it.' [ro05ee 123] 13.3 Specific semantic fields


In contrast, Pichi designations for the seasons of the year are fully in use, as shown by the use of the compound noun *ren-sísin* 'rainy season' (55) and *amatán* 'harmattan' in (56):

(55) *Dís* this *dé* day *dɛn* pl *ren-sísin* rainy.cpd-season *go* pot *bigín.* begin

'These days, the rainy season should begin.' [dj05ce 059]

	- fog

'Here, we call harmattan that, like it's usually like fog.' [ye05ce 062]

### **13.3.2 Colours**

Colour terminology was elicited with three speakers between the ages of twenty-one and twenty-seven and with two speakers above the age of fifty-five. The exercise revealed the apparent-time differences in colour terminology contained in Table 13.5. Pichi terms are in normal font, variants are indicated by a semicolon. Spanish terms are in italics. Table 13.5 indicates that the younger speakers employ the basic Pichi colour terms *blák* 'black' and *wáyt* 'white' consistently. The colours 'red' and 'yellow' are more frequently referred to by the Spanish terms *rojo* and *amarillo*, respectively, but the Pichi terms *rɛ́d* 'red' and *yɛ́lo* 'yellow' are also used. All other colours are uniquely referred to by Spanish terms. The older group consistently makes use of Pichi *rɛ́d* 'red' in addition to the basic colours *blák* and *wáyt*. Meanwhile 'blue' and 'green' are referred to by the Pichi terms *blú* and *grín* , respectively, or by their Spanish equivalents *azul* and *verde*.

At least in apparent time, the range of Pichi colour terms appears to have been reduced from the six colours *blák*, *wáyt*, *rɛ́d*, *yɛ́lo*, *blú*, and *grín* with the older group, to the two basic colours *blák* and *wáyt*, supplemented by the less frequent *rɛ́d* and *yɛ́lo* (Table 13.5).

Many West African languages, including basilectal Nigerian Pidgin (Faraclas 1996: 286) express colours and hues other than 'black' and 'white' through periphrasis, suprasegmentals and ideophones. We also find the expression of colours through periphrasis in Pichi, as in (57) and (58).

#### 13 Pichi and Spanish in contact


Table 13.5 Apparent-time differences in the use of colour terms


'That thing is yellow like a banana.' [i.e. It has a bright yellow colour]

The rarity of Pichi colour terms beyond the basic ones of *blák* and *wáyt* with the younger group may therefore be indicative of a departure from the West African composite system of colour denomination towards a European simplex system in which nonbasic colours are denoted by specific property items.

When Spanish colour terms are used attributively, they occur with Spanish head nouns (59). The corpus contains no examples of mixed collocations involving a Spanish colour denoting property item and a Pichi head noun:

(59) *A* 1sg.sbj *tínk* think *sé* qot *na* foc *judías* bean.pl *blancas* white.pl *o* or *no* neg *sé.* I.know 'I think they're white beans or so.' [eb03sp 122]

Spanish colour terms also occur as predicate adjectives in the specific type of mixed copula clause involving Spanish adjectives covered in §13.2.2 above. In contrast, Pichi colour terms are only lexicalised as inchoative-stative verbs.

### **13.3.3 Other semantic fields**

Other semantic fields characterised by a high incidence of codemixing involve formalised, institutional domains. One of the few Pichi country names in use is *Panyá* 'Spain', the

13.3 Specific semantic fields

designation for the former colonial power. Spanish lexemes are exclusively employed for country names like *Guinea (Ecuatorial)* 'Equatorial Guinea', *Gabón* 'Gabon' (60), ethnonyms like *europeo* 'European' or *cameruneses* 'Cameroonians' (61), as well as terms belonging to the state domain such as *problema diplomático* 'diplomatic problem' (60):

(60) *Entonces* so *wán* one *problema* problem *diplomático* diplomatic *kán* pfv *dé* be.loc *entre* between *Guinea* place *wet* with *Gabón.* place

'So a diplomatic problem came to be between Guinea and Gabon.' [fr03ft 007]

(61) *Cameruneses,* Cameroonians *yɛ́s* yes *dɛn* 3pl *plɛ́nte* be.plenty *yá.* here 'Cameroonians, yes they are many here.' [ma07fn 607]

Also compare the Spanish terms *registro* '(civil) registry' and *registra* '(to) register' in (62):

(62) *A* 1sg.sbj *bin* pst *gɛ́fɔ* have.to *chénch* change *in* 3sg.poss *ném* name *na* loc *registro* register *a* 1sg.sbj *registra* register *ín.* 3sg.indp 'I had to change her name in the register, I registered her.' [ab03ay 162]

The Pichi lexemes *skul* 'school'*, gɔ́bna* 'government' (63), and *chɔ́ch* 'church' (64) designate these institutions in their general sense and are favoured over their Spanish equivalents *escuela, gobierno*, and *iglesia*:


Meanwhile, the incidence of Spanish lexemes rises with the degree of specificity of words within the semantic fields designated by these superordinates. Thus, we have *catedral* 'cathedral' in (64) above, *bolí* 'pen' and *cuaderno* 'exercise book' (65), as well as *profe(sor)* 'teacher' – though *tícha* 'teacher' is also common, however less so beyond primary school.

(65) *Wé,* sub *yu* 2sg *wánt* want *báy* buy *cuaderno,* exercise.book *bolí* pen *ɔ́l* all *dán* that *tín* thing *dɛn* pl *na* foc *wet* with *dólar.* dollar 'While, if you want to buy exercise books, pens, all those things are with the dollar.' [ed03sp 096]

13 Pichi and Spanish in contact

(66) *Di* def *profesor,* teacher *na* foc *bɛ́ta* very.good *profe.* teacher

'The (secondary school) teacher is a very good teacher.' [dj05be 172]

The preponderance of Spanish lexemes in other semantic fields reflects the asymmetric power relation that holds between Pichi and Spanish in a different way. For example, semantic fields relating to illness and medical treatment that are highly differentiated in other languages of the region (e.g. Yoruba, see Adegbite (1993)) probably did not assert itself in Pichi due to the marginalisation of African medical science with the advent of colonialism. In (67), we therefore find *placenta* 'placenta' and *matriz* 'womb' for which only the general term *bɛlɛ́*'belly, womb' is recorded and Spanish *membrano* 'membrane' which has no equivalent in Pichi:

(67) *Wé* sub *dɔ́kta* doctor *ópin,* open *wé* sub *dɛn* 3pl *bigín* begin *drɔ́* draw *di,* def *sɔn* some *tín* thing *we* sub *e* 3sg.sbj *kin* hab *dé* be.loc *bihɛ́n* behind *placenta,* placenta *na* foc *membrana,* membrane *sɔn* some *kán* pfv *lɛ́f* remain *bifó* before *di* def *matriz,* womb *so* so *di* def *matriz* womb *nó* neg *kán* pfv *lɔ́k.* lock

'When the doctor opened (the womb), they began to draw out the, a certain thing that is usually behind the placenta, it's a membrane, some remained in front of the womb, so the womb didn't close.' [ab03ay 084]

The systematic use of Spanish items also occurs in semantic fields that designate aspects of material and non-material culture of external origin. In (68), a car mechanic explains the disadvantages of an Opel ignition cable. Note the Spanish technical terms in the sentence:

(68) *Hɛ́,* intj *a* 1sg.sbj *go* pot *fála* accompany *yú* 2sg.indp *bikɔs* because *sɔn* some *cable* cable *dé* be.loc *wé* sub *na* foc *fɔ* prep *Opel,* name *yu* 2sg *intenta* try *bríng* bring *Opel* name *in* 3sg.poss *yón* own *na* foc *corriente,* electricity *Opel* name *de* ipfv *kɛ́r* take *bɔkú* much *corriente* electricity *só* like.that *e* 3sg.sbj *nó* neg *go* pot *fít* can *ɛ́nta* enter *na* loc *dán* that *bujía,* ignition.plug *yu* 2sg *go* pot *wánda* wonder *sɛ́f.* emp

'Hey, I'll accompany you because there's a cable which is an Opel (cable), (and if) you try to connect the Opel one with electricity, Opel takes a lot of electricity, so it won't be able to enter that ignition plug, (and) you'll be very surprised.' [f103fp 017]

Spanish kinship terms have also left their mark on the language (cf. also §12.3). In (69), we find *primo* '(male) cousin', a kinship concept that is only rarely expressed by the Pichi term *kɔsín.*

13.3 Specific semantic fields

(69) *A* 1sg.sbj *tínk* think *sé* qot *dɛn* 3pl *papá* father *na* foc *mi* 1sg.poss *mamá* mother *in* 3sg.poss *primo.* cousin 'I think that their father is my mother's cousin.' [fr03ft 059]

Conversely, the incidence of Spanish words is low in semantic fields characterised by the use of autochthonous technology, such as farming and with designations for locallygrown foodstuffs and other flora. Thus, in (70), we have *díg grɔ́n* 'dig ground' = 'plough up the ground', *plánt chɔ́p* '(to) plant food', *gádin* 'small field, garden', *jakató* 'bitter tomato' and *kíp* 'grow, rear', as well as *pamáyn* 'oil', and *gadinɛ́ks* 'egg-plant' (71):


'The day when you would need it, you would say "I don't have oil", (and) you would harvest egg-plants.' [ab03ay 015]

## **14 Typological summary of Pichi**

Pichi is an Afro-Caribbean English-lexifier creole language spoken on the island of Bioko, Equatorial Guinea. With somewhere between 100–150,000 speakers, Pichi is one of the most widely spoken languages of the country. Pichi is an offshoot of 19th century Krio (Sierra Leone) and shares many characteristics with its sister languages Krio, Aku (Gambia), and Nigerian, Cameroonian, and Ghanaian Pidgin. However, insulation from English and intense contact with Spanish, the colonial and official language of Equatorial Guinea, have given Pichi a character distinct from the other West African English-lexifier creoles and pidgins.

Pichi has a nominative-accusative alignment, SV(O) word order and adjective-noun order, prenominal determiners, and prepositions. Pichi has a seven-vowel system and twenty-two consonant phonemes, including two labio-velar plosives. The language has a two-tone system with tonal minimal pairs, morphological tone for the marking of pronominal case distinctions, and numerous tonal processes. The morphological structure of Pichi is largely isolating. However, there is some inflectional and derivational morphology in which affixation and tone are put to use. Pichi is characterised by a weak verb-adjective distinction.

The categories of tense, modality, and aspect are primarily expressed through preverbal particles. Pichi is an aspect-prominent language in which aspect, rather than tense, plays a dominant role in expressing temporal relations. Besides that, the modal system includes an indicative-subjunctive opposition. The copula system employs various suppletive forms and is differentiated along the semantic criterion of time-stability.

Pichi verbs fall into three lexical aspect classes: dynamic, inchoative-stative, and stative. Content questions are formed by way of a mixed question-word system which involves transparent and opaque question elements. Clause linkage is characterised by a large variety of strategies and forms, in which a subordinator, a quotative marker, and two modal complementisers stand out as multifunctional elements with overlapping functions. The language also features various types of multiverb constructions. These include secondary predication, clause chaining, and serial verb constructions. Amongst the latter figure instrumental serial verb constructions involving the verb *ték* 'take' as well as comparative constructions featuring the verb *pás* '(sur)pass'.

Many of the typological characteristics summarised above align Pichi closely with the Atlantic-Congo languages spoken in the West African littoral zone and beyond. At the same time, characteristics like the prenominal position of adjectives and determiners show a typological overlap with English. There are also numerous structural and lexical parallels with the Afro-Caribbean English-lexifier creoles of the (Circum-)Caribbean, such as, for example, Jamaican, Creolese (Guyana), and the creole languages of Suriname.

## **15 Texts**

The following six Pichi texts represent four types of genre: narrative, routine procedure, elicitation, and conversation. Each sentence is provided with its text codes (placed above the sentence it refers to). This allows comparison with the analysis of examples in the grammar section. In conversations, speakers can be identified by the two-letter speaker code at the beginning of the text code. All texts contain Spanish material ranging from single words to whole sentences. An interlinear gloss of Spanish material is provided where it occurs in the same utterance along with Pichi material. For sentences entirely in Spanish, only a free translation is provided. There are only a few Bube elements in the text, all of which stem from speaker (ab). The presence of Bube material is indicated in squared brackets. Bube was not transcribed due to the absence of a scientific grammar and comprehensive dictionary at the time of field research.This has been partly remedied by the publication of Bolekia Boleká (2009), but the description of Bube still leaves much to be desired.

### **15.1 Narrative and conversation: Miguel falls sick**

The main narrator in the following text is Abuela 'grandmother' (ab). Other discourse participants are Francisca (fr) and myself (ko). The text begins with a conversation between (ab) and (fr) on the latter's competence in the Bube language. From (023)–(038), the conversation gives way to a brief story by (ab), in which she relates the hardship she endured living as an adolescent away from her family with a *mísis* 'matron'. In (039)– (042), (ab) then draws a comparison between the style of upbringing back then and her grandson Miguel's behaviour towards grown-ups today.

This leads (ab) to the main narrative from (043)–(134), in which (ab) gives an account of how her grandson Miguel came down with malaria a few nights before the recording took place, and how he was brought to hospital. The protagonists of this personal narrative are (ab) herself, her grandson Miguel, and his mother Tokobé. The narrative is characterised by extensive codemixing between Pichi and Spanish, as well as Pichi and Bube.

(1) ab03ab 008

*Hɛ́* intj *a,* 1sg.sbj *yu* 2sg *nó* neg *de* ipfv *tɔ́k* talk *Bubɛ,* Bube *a* 1sg.sbj *wɔ́nda* wonder *náw* now *lɛk* like *háw* how *e* 3sg.sbj *dɔ́n* prf *fɔgɛ́t* forget *Bubɛ* Bube *wé* sub *e* 3sg.sbj *gó* go *Panyá.* Spain

'Hey I, you don't speak Bube, I wonder now how she had forgotten Bube when she went to Spain.'


'This one (here) already speaks it, this one talks better than her sister.'

*sísta.* sister

(4) fr03ab 011

*Nó,* neg *nóto* neg.foc *trú* be.true *abuela.* grandmother 'No, that's not true grandmother.'

(5) fr03ab 012

*Lage* name *de* ipfv *tɔ́k* talk *Bubɛ* Bube *pás* pass *mí.* 1sg.indp 'Lage speaks Bube better than me.'

(6) ab03ab 013

*E* 3sg.sbj *de* ipfv *tɔ́k* talk *Bubɛ* Bube *pás* pass *yú?* 2sg.indp 'She speaks Bube better than you?'

(7) ab03ab 014 *Dís* this *wán* one *sɛ́f,* foc *yu* 2sg *dɔ́n* prf *de* ipfv *tráy.* try

'Even this one [you], you're making an effort.'

(8) ko03ab 015 *Bɔt* but *yu* 2sg *bin* pst *de* ipfv *tɔ́k* talk *Bubɛ* Bube *bifó?* before 'But you were speaking Bube before?'

(9) ab03ab 016

*E* 3sg.sbj *bin* pst *de* ipfv *tɔ́k=an,* talk=3sg.obj *e* 3sg.sbj *nó* neg *bin* pst *de* ipfv *hía* hear *ɔ́da* other *lángwej.* language 'She was speaking it, she didn't understand any other language.'

(10) fr03ab 017

*Wé* sub *a* 1sg.sbj *bin* pst *smɔ́l,* be.small *a* 1sg.sbj *bin* pst *de* ipfv *tɔ́k* talk *Bubɛ.* Bube 'When I was small, I was speaking Bube.'

(11) ab03ab 018

*Wé* sub *yu* 2sg *kɔmɔ́t* come.out *sík* be.sick *dán* that *sík* be.sick *na* loc *Panyá,* Spain *wé* sub *yu* 2sg *bin* pst *sík,* be.sick *náw* now *yu* 2sg *bigín tɔ́k Panyá.*

begin talk Spain

'When you had just been sick in Spain, when you were sick, then you began speaking Spanish.'

(12) ab03ab 019

*Áfta,* then *yu* 2sg *dé* be.loc *hía,* here *¿cuántos* how.many *años* year.pl *estuviste* you.were *aquí?* here

'Then, you were here, how many years were you here?'

(13) fr03ab 020

*Medio* half *año,* year *seis* six *meses.* month.pl 'Half a year, six months.'

(14) ab03ab 021

*Áfta in papá sɛ́f kán ték=an.*

then 3sg.poss father self come take=3sg.obj

'Then her father himself came to take her [away from here].'

(15) ab03ab 022

*Áfta* then *es* it.is *la* def *respuesta.* answer 'Then that's the answer.'

(16) ab03ab 023

*Mí,* 1sg.indp *lɛk* like *háw* how *yu* 2sg *de* ipfv *sí* see *mí,* 1sg.indp *a* 1sg.sbj *dɔ́n* prf *sí* see *plɛ́nte* be.plenty *tín.* thing 'As for me, as you see me, I've seen many things (in life).'

(17) ab03ab 024

*A* 1sg.sbj *nó* know *di* def *tɛ́n* time *wé* sub *yu* 2sg *de* ipfv *smɛ́l* smell *pamáyn,* oil *swit-ɔ́yl.* tasty.cpd-oil

'I know the time when you'd smell oil, sweet oil.'

(18) ab03ab 025

*Yu* 2sg *mísis* matron *sɛ́n* send *yú* 2sg.indp *gó* go *na* loc *shɔ́p,* shop *sé* qot *gó* go *báy* buy *mí* 1sg.indp *swit-ɔ́yl.* tasty.cpd-oil

'The matron [head (f.) of the household that speaker (ab) was staying in] of the house would send you to the shop, saying "go buy some sweet oil for me".'

(19) ab03ab 026 *Yu* 2sg *gó* go *yu* 2sg *bríng* bring *sɔn* some *ɔ́yl.* oil 'You would go (and) bring some oil.' (20) ab03ab 027 *Wé* sub *e* 3sg.sbj *lúk=an* look=3sg.obj *só* like.that [exclamation]. 'And she'd look at it like this [exclamation in Bube].' (21) ab03ab 028 *Dís?* this 'This?' (22) ab03ab 029 *Dís* this *nóto* neg.foc *Manolete.* name 'This is not Manolete (oil).' (23) ab03ab 030 *Gó* go *lɛ́f=an,* leave=3sg.obj *gó* go *lɛ́f=an!* leave=3sg.obj 'Go leave it, go leave it [bring it back]!' (24) ab03ab 031 *Di* def *trú* be.true *comedor* dining-room *dé* be.loc *fɔ* prep *soja,* soya *Manolete,* name *Cordobés.* name 'The real dining-room has soy bean oil, Manolete, Cordobes [vegetable oil brands], right?' (25) ab03ab 032 *Na* foc *di* def *bɛ́tɛ* very.good *bɛ́tɛ* rep *swít* tasty *ɔ́yl,* oil *pyɔ́* pure *pyɔ́* rep *uvas.* grapes 'That's the very best sweet oil, (made from) purest grapes.' (26) ab03ab 033 *Ɛf* if *nóto* neg.foc *yu* 2sg *báy,* buy *dán* that *húman* woman *go* pot *bít* beat *yú* 2sg.indp *sóté* until *yu* 2sg *go* pot *gó* go *lɛ́f=an.* leave=3sg.obj

'If it weren't the case that you had bought (the right oil), that woman would beat you until you would go leave it [bring it back].'

(27) ab03ab 034

*Wé* sub *yu* 2sg *dɔ́n* prf *gó* go *lɛ́f=an,* leave=3sg.obj *yu* 2sg *go* pot *gí* give *di* def *mán* man *wé* sub *e* 3sg.sbj *de* ipfv *sɛ́l* sell *di* def *funda,* receptacle *e* 3sg.sbj *nó* neg *go* pot *ték=an* take=3sg.obj *fɔ* prep *yú.* 2sg.indp

'When you've gone to leave it, you would give (the oil) to the man who is selling the receptacle (with the oil) and he wouldn't take it (back) from you.'

(28) ab03ab 035

*Yu* 2sg *gɛ́fɔ* have.to *gó* go *fɔ* prep *yu* 2sg *fámbul.* family

'You would have to go to your (own) family.'

(29) ab03ab 036

*Yú* 2sg.indp *gí* give *dɛ́n* 3pl.indp *dán* that *smɔ́l* small *pamáyn* oil *mék* sbjv *dɛn* 3pl *gí* give *yú* 2sg.indp *mɔní* money *yu* 2sg *go* pot *báy* buy *di* def *wán* one *wé* sub *yu* 2sg *mísis* matron *dé,* there *adɔnkɛ́* even.if *e* 3sg.sbj *nó* neg *sí* see *yú* 2sg.indp *wán* one *hól* whole *dé,* day *e* 3sg.sbj *nó* neg *bísin* care *wáns* once *yu* 2sg *bríng* bring *di* def *pamáyn.* oil

'You would give them [your family] that little bit of oil so that they gave you money (so that) you would go buy the one [the correct oil] that your matron there, even if she didn't see you for a whole day, she wouldn't care once you brought the [correct] oil.'

```
(30) ab03ab 037
```
*Mí* 1sg.indp *dɔ́n* prf *sɔ́fa.* suffer 'I have suffered.'

(31) ab03ab 038

*A dɔ́n sí bihɛ́n* [continues in Bube].

1sg.sbj prf see behind

'I have seen behind (…)'

(32) ab03ab 039

*Pero* but *pikín* child *tidé,* today *náw* now *yu* 2sg *sɛ́n=an,* send=3sg.obj *dí* this *pikín,* child *wé* sub *a* 1sg.sbj *de* ipfv *sɛ́n=an,* send=3sg.obj *e* 3sg.sbj *nó* neg *de* ipfv *gó* go *mɔ́.* more

'But a child today, (if) you send it (for something) now, this child, when I'm sending him, he doesn't go anymore.'

(33) ab03ab 040

*Ay, todo el día de hoy tú me vas a mandar, ¡vete tú misma!*

'[My grandson would say] "Oh, the whole day today you're going to send me around, you go yourself!"'

(34) ab03ab 041

*Náw* now *náw* rep *mék* sbjv *a* 1sg.sbj *sɛ́n=an* send=3sg.obj *na* loc *gran-pá* grand-pa *in* 3sg.poss *rúm,* room *e* 3sg.sbj *go* pot *sé* say *e* 3sg.sbj *de* ipfv *fíɛ,* fear *e* 3sg.sbj *nó* neg *go* pot *gí* give *mí* 1sg.indp *di* def *tín* thing *wé* sub *a* 1sg.sbj *de* ipfv *sɛ́n=an.*

send=3sg.obj

'Right now, let me (try) send him to grandfather's room, he [my grandson] would say that he's afraid, (that) he wouldn't give me the thing I'm sending him for.'

(35) ab03ab 042

*Sé* qot *ín* 3sg.indp *nó* neg *wánt* want *in* 3sg.poss *abuelo* grandfather *skrách=an.* scratch=3sg.obj

'Because he [emp] doesn't want his [deceased] grandfather to scratch him.'

(36) ab03ab 043

*A* 1sg.sbj *dɔ́n* prf *tɛ́l* tell *yú* 2sg.indp *wétin* what *pás* happen *na* loc *nɛ́t,* night *dán* that *nɛ́t.* night 'I've already told you what happened at night, that night.'

(37) ab03ab 044 *Yɛ́stadé.*

yesterday

'Yesterday.'

(38) ab03ab 045

*Mí* 1sg.indp *gó* go *na* loc *mi* 1sg.poss *béd,* bed *a* 1sg.sbj *bigín* begin *de* ipfv *mɛ́mba* remember *mi* 1sg.poss *yón* own *trɔ́bul.* trouble 'I [emp] went to bed, I began thinking about my own problems.'

(39) ab03ab 046

*Dɛ́n* 3pl.indp *slíp* sleep *dɛn* 3pl *de* ipfv *ronca.* snore

'They [emp] [the others in the house] had lied down and were snoring.'

(40) ab03ab 047

*Ín* 3sg.indp *de* ipfv *kakara,* ideo *kakara* rep *kakara.* rep 'He [Miguel] was all fidgety.'

(41) ab03ab 048 *E* 3sg.sbj *de* ipfv *costumbre.* habit 'He's was getting used to it.' (42) ab03ab 049 *Di* def *wé* way *in* 3sg.poss *áwa* hour *nɔ́ba* neg.prf *rích* arrive *fɔ* prep *slíp,* sleep *e* 3sg.sbj *go* pot *bigín* begin *de* ipfv *hala-hála* red.cpd-shout *mí.* 1sg.indp 'Since his time for sleeping hadn't come yet, he was going to begin shouting for me.' (43) ab03ab 050 *Smɔ́ltɛn* shortly.after *slíp* sleep *kéch=an.* catch=3sg.obj 'Shortly after, he became sleepy.' (44) ab03ab 051 *E* 3sg.sbj *sé* say *"áy"* intj *a* 1sg.sbj *hía* hear *di* def *hála.* shout 'He said "ay", I heard the shout.' (45) ab03ab 052 *In* 3sg.poss *mamá* mother *sé* say *wétin* what *pás,* happen *wétin* what *pás?* happen 'His mother said what happened, what happened?' (46) ab03ab 053 *E* 3sg.sbj *sé* say *"mɔ́mi* mum *mɔ́mi* mum *yu* 2sg *nó* neg *de* ipfv *sí* see *dán* that *mán* man *wé* sub *e* 3sg.sbj *rɔ́n* run *gó* go *abuela* grandmother *in* 3sg.poss *rúm?"* room 'He said "mum, mum don't you see that man who ran into grandmother's room?"' (47) ab03ab 054 *"E* 3sg.sbj *dɔ́n* prf *pás,* pass *e* 3sg.sbj *dɔ́n* prf *pás,* pass *e* 3sg.sbj *dɔ́n* prf *pás."* pass '"He has just passed by, he has just passed by, he has just passed by."' (48) ab03ab 055 *E* 3sg.sbj *bigín* begin *de* ipfv *trímbul.* tremble 'He began to tremble.' (49) ab03ab 056 *Náw* now *e* 3sg.sbj *sé/* say *in* 3sg.poss *mamá* mother *tɛ́l=an* tell=3sg.obj *sé* qot *nɔ́,* neg *abuela* grandmother *fue* went *a* to

*la* the *cocina* kitchen *a* to *beber* drink *agua.* water

'Now he said/ his mother told him that "no, grandmother went to the kitchen to drink water".'

(50) ab03ab 057

*E* 3sg.sbj *sé* say *pero* but *es* it.is *un* def *hombre.* man 'He said "but it's a man".'

(51) ab03ab 058

*E* 3sg.sbj *nó* neg *kán* pfv *slíp* sleep *mɔ́* more *ó.* sp 'He actually didn't sleep again.'

(52) ab03ab 059

*Sɔn* some *káyn* kind *fíba* fever *kán* pfv *kéch=an,* catch=3sg.obj *Tokobé* name *nó* neg *kán* pfv *sabí.* know

'He got a serious fever (and) Tokobé didn't get to know (about it).'

(53) ab03ab 060

*Mɔ́nin* morning *tɛ́n* time *e* 3sg.sbj *gráp* get.up *e* 3sg.sbj *sé* say *Miguel* Miguel *vete* go *a* to *hacer* do *pipí* wee-wee *y* and *vete* to *a bañarte, hay clase.*

bathe there.is class

'In the morning she got up (and) she said "Miguel go do a wee-wee and go take a bath, you have classes".'

(54) ab03ab 061

*E* 3sg.sbj *kán,* come *e* 3sg.sbj *sé* qot *"abuela,* grandmother *llevame* bring.me *al* to *hospital".* hospital

'He came, he said "grandmother take me to hospital".'

(55) ab03ab 062

*E* 3sg.sbj *sé* qot *"no* neg *puedo* I.can *parar".* stand 'He said "I can't (even) stand".'

(56) ab03ab 063 *E* 3sg.sbj *dé* be.loc *na* loc *grɔ́n.* ground 'He was (lying) on the ground.'

(57) ab03ab 064 *Na* foc *só* so *e* 3sg.sbj *de* ipfv *swɛ́t.* sweat 'He was sweating just like that.' (58) ab03ab 065 *Wé* sub *a* 1sg.sbj *kin* hab *mék* make *só,* so *a* 1sg.sbj *nó* neg *de* ipfv *fíl* feel *hɔ́t.* hot 'When I would do like this [places her hand on her forehead], I wasn't feeling heat.' (59) ab03ab 066 *Pero* but *wé* sub *a* 1sg.sbj *kin* hab *tɔ́ch* touch *in* 3sg.poss *fút,* leg *in* 3sg.poss *hán* arm *dé,* there *na* foc *só* so *dɛn* 3pl *kól* be.cold *[ko::l].* 'But when I would touch his leg (and) his arm there, they were so incredibly cold.' (60) ab03ab 067 *A* 1sg.sbj *ték=an* take=3sg.obj *pút=an* put=3sg.obj *pantáp* on *mi* 1sg.poss *bɛlɛ́.* belly 'I put him onto my stomach.' (61) ab03ab 068 *Na* foc *só* so *a* 1sg.sbj *de* ipfv *wáyp=an,* wipe=3sg.obj *a* 1sg.sbj *de* ipfv *sopla* blow *ín* 3sg.indp *fwífwífwí.* ideo 'I was wiping him, I was fanning him just like that.' (62) ab03ab 069 *A* 1sg.sbj *lúk=an.* look=3sg.obj 'I looked at him.' (63) ab03ab 070 *Na* foc *só* so *in* 3sg.poss *hát* heart *mék* make *kutuku* ideo *kutuku* rep *kutuku.* rep 'His heart was racing just like that.' (64) ab03ab 071 *A* 1sg.sbj *kɔ́l* call *Tokobé* name *a* 1sg.sbj *sé* qot *"mɔ́mi",* mum *a* 1sg.sbj *sé* qot *"kán".* come 'I called Tokobé, I said "mother", I said "come".' (65) ab03ab 072 *A* 1sg.sbj *bɛ́g,* ask.for *lúk* look *dís* this *pikín,* child *dí* this *pikín* child *nó* neg *dé* be.loc *gúd.* good 'Please, look at this child, this child is not well.' (66) ab03ab 073

*Na* foc *ín* 3sg.indp *e* 3sg.sbj *de* ipfv *kán* come *púl* remove *mí* 1sg.indp *dán* that *torí.* story 'That's when she was coming to tell me that story.'

(67) ab03ab 074 *E* 3sg.sbj *sé* qot *"na* loc *nɛ́t",* night *e* 3sg.sbj *sé* qot *"na* foc *só* like.that *yu* 2sg *bin* pst *hía* hear *ín* 3sg.indp *hála".* shout 'She said "at night", she said "that's how you heard him shout".' (68) ab03ab 075 *E* 3sg.sbj *sé* qot *frɔn* from *dán* that *hála* shout *di* def *pikín* child *nó* neg *slíp* sleep *mɔ́.* more 'She said "since that shout the child didn't sleep again".' (69) ab03ab 076 *E* 3sg.sbj *kán* pfv *gɛ́t* get *fíba.* fever 'He got a fever.' (70) ab03ab 077 [A sentence in Bube]. (71) ab03ab 078 *"Vete, a bañar."* '"Off you go, go have a bath".' (72) ab03ab 079 *E* 3sg.sbj *púl=an* remove=3sg.obj *na* loc *pantáp* on *di* def *béd.* bed 'She pulled him from the bed.' (73) ab03ab 080 *Na* foc *só* so *e* 3sg.sbj *de* ipfv *swɛ́t.* sweat 'He was sweating just like that.' (74) ab03ab 081 *A* 1sg.sbj *púl* remove *in* 3sg.poss *klós,* clothing *a* 1sg.sbj *híb=an* heave=3sg.obj *pantáp* on *di* def *béd.* bed 'I removed his clothes, I heaved him onto the bed.' (75) ab03ab 082 *Sé* qot *"papá* intj *gɔ́d,* God *ús=káyn* q=kind *trɔ́bul* trouble *dís?"* this '(I) said (to myself) "oh God, what kind of trouble is this?"' (76) ab03ab 083 *A* 1sg.sbj *púl* remove *in* 3sg.poss *camiseta,* singlet *a* 1sg.sbj *pút=an* put=3sg.obj *pantáp* on *béd* bed *a* 1sg.sbj *gó* go *a* 1sg.sbj *púl* remove *di* def *trɔsís* trousers *a* 1sg.sbj *híb=an* heave=3sg.obj *ínsay* inside *di* def *bañera.* bathtub 'I removed his singlet, I put him on the bed (and) I removed the trousers (and) I

heaved him inside the bathtub.'

(77) ab03ab 084 *A* 1sg.sbj *ték* take *wán* one *kɔ́p* cup *watá,* water *a* 1sg.sbj *ték=an* take=3sg.obj *a* 1sg.sbj *mék* make *bwa* ideo *bwa* rep *bwa* rep *bwa.* rep 'I took a cup of water, I took it (and) splushed him all over with water.' (78) ab03ab 085 *A* 1sg.sbj *sé* qot [continues in Bube]. 'I said (…)' (79) ab03ab 086 *Smɔ́ltɛn* shortly.after *e* 3sg.sbj *mék* make [imitates exhalation]. 'Shortly he made [imitates exhalation].' (80) ab03ab 087 *A* 1sg.sbj *sé* qot *"¿cómo* how *sientes?"* you.feel 'I said "how do you feel?"' (81) ab03ab 088 *E* 3sg.sbj *sé* qot *"abuela* grandmother *ya* already *siento* I.feel *bien."* good 'He said "grandmother, I already feel fine".' (82) ab03ab 089 *E* 3sg.sbj *kɔmɔ́t* come.out *na* loc *bañera,* bathtub *ín* 3sg.indp *sɛ́f* self *kán* pfv *gó.* go 'He came out of the bathtub, he himself left (it).' (83) ab03ab 090 *A* 1sg.sbj *gí=an* give=3sg.obj *di* def *haf-táwɛl.* half.cpd-towel 'I gave him the [his] little towel.' (84) ab03ab 091 *A* 1sg.sbj *kɛ́r=an* carry=3sg.obj *gó* go *na* loc *comedor.* dining-room 'I carried him to the dining-room.' (85) ab03ab 092 *Sé* qot *Tokobé,* name *kɛ́r* carry *di* def *pikín* child *na* loc *ɔspítul.* hospital '(I) said Tokobé, bring this child to hospital.'

(86) ab03ab 093 *Mí* 1sg.indp *nó* neg *sé* qot *di* def *pikín* child [continues in Bube]. 'I know that the child (…)' (87) ab03ab 094 *A* 1sg.sbj *wánt* want *ték* take *solwatá* saltwater *mék* sbjv *a* 1sg.sbj *gí=an,* give=3sg.obj *a* 1sg.sbj *sé* qot *"chip"* skt *nɔ́.* neg 'I wanted to take saltwater and give it to him, I said [skt] no.' (88) ab03ab 095 *E* 3sg.sbj *sé* qot *na* foc *hángri.* hunger 'He said "it's hunger" [that's worrying me].' (89) ab03ab 096 *Dán* that *banána,* banana *a* 1sg.sbj *gí=an* give=3sg.obj *sɔn.* some 'That banana [points to a stalk lying in the corner], I gave him one.' (90) ab03ab 097 *E* 3sg.sbj *sé* qot *"abuela,* grandmother *e* 3sg.sbj *nó* neg *kɛ́r".* carry 'He said "grandmother, it wasn't enough."' (91) ab03ab 098 *Mí* 1sg.indp *sé* qot [continues in Bube]. 'I [emp] qot (…)' (92) ab03ab 099 *A* 1sg.sbj *sé* qot *nó* neg *gí=an* give=3sg.obj *leche,* milk *gí=an* give=3sg.obj *wɔtá!* water 'I said "don't give him milk, give him water!"' (93) ab03ab 100 *A* 1sg.sbj *gí=an.* give=3sg.obj 'I gave him (the water).' (94) ab03ab 101 *E* 3sg.sbj *sé* qot *e* 3sg.sbj *nó* neg *kɛ́r.* carry

'He said it wasn't enough.'

(95) ab03ab 102

*Lɛk* like *háw* how *Tokobé* name *púl* remove *di/* def *e* 3sg.sbj *nɔ́ba* neg.prf *púl* remove *di* def *glás* glass *e* 3sg.sbj *wánt* want *mɔ́.*

more

'As soon as Tokobé removed the/ she hadn't yet removed the glass (and) he wanted more.'

(96) ab03ab 103

*Mí* 1sg.indp *gó* go *dɔ́n.* down

'I went down(stairs).'

(97) ab03ab 104

*Wé* sub *a* 1sg.sbj *kɔmɔ́t* come.out *dɔ́n,* down *a* 1sg.sbj *gó* go *sidɔ́n* sit *bifór=an,* before=3sg.obj *a* 1sg.sbj *sé* qot *"¿cómo* how *sientes?"*

feel.2sg

'When I came back from downstairs, I went to sit before him (and) I said "how do you feel?"'

(98) ab03ab 105

*E* 3sg.sbj *sé* qot *"abuela,* grandmother *siento* I.feel *mal,* bad *quiero* I.want *ir* go *al* to *hospital".* hospital

'He said "grandmother, I feel bad, I want to go to the hospital".'

(99) ab03ab 106

*E* 3sg.sbj *sé* qot *"cuando* when *una* a *persona* person *está* is *enferma* sick *los* the *demás* others *no* not *deben* must *estar* be *con* with *ella sentada".*

him seated

'He said "when a person is sick, the others are not supposed to be sitting with him".'

(100) ab03ab 107

*Porque cuando se va a vomitar, se va a mojar con vómito*. 'Because when he vomits they will get wet with vomit.'

(101) ab03ab 108

*A sé "has vomitado?"*

1sg.sbj qot have.you vomited

'I said "did you vomit?"'

(102) ab03ab 109

*E* 3sg.sbj *sé* qot *"sí,* yes *abuela* grandmother *yo* I *siento* I.feel *a* to *vomitar."* vomit 'He said "yes, grandmother I feel like vomiting."

(103) ab03ab 110

*A* 1sg.sbj *sé* qot *"Tokobé* name *kán* come *ó!"* sp 'I said "Tokobé come, please!"'

(104) ab03ab 111

*Tokobé* name *dɔ́n* prf *wɛ́r* wear *klós* clothing *gbogbogbo* ideo *"nó* neg *fɔ* prep *fɛ́n* look.for *cuaderno".* exercise.book

'Tokobé had already worn her clothes in a rush, "no we have to look for the patient's logbook".'

(105) ab03ab 112

*E* 3sg.sbj *mít* meet *wán* one *ól* old *ól* rep *cuaderno,* exercise.book *di* def *tɛ́n* time *fɔ* prep *Niumbɛ,* name *na* foc *ín* 3sg.indp *e* 3sg.sbj *bin* pst *rɔ́n* run *wet=an* with=3sg.obj *ɔspítul.* hospital

'She found a very old patient's book, from the time of Niumbe, that's when she ran off to the hospital with him.'

(106) ab03ab 113

*Dɛn* 3pl *gó* go *na* loc *ɔspítul.* hospital 'They went to the hospital.'

(107) ab03ab 114

*Sé* qot *nɔ́* neg *bifó* before *di* def *dɔ́kta* doctor *de* ipfv *kán* come *wé* sub *a* 1sg.sbj *de* ipfv *kán* come *fɔ́s.* first

'Then before the doctor was coming I was already coming first [had come to the hospital from home as well].'

(108) ab03ab 115

*Dɛn* 3pl *gó* go *sé* qot *análisis.* analysis

'They went for an analysis.'

(109) ab03ab 116

*Dɛn rɔ́n gó mék análisis.*

3pl run go make analysis

'The rushed off to make an analysis.'

(110) ab03ab 117

*Lɛk* like *háw* how *e* 3sg.sbj *de* ipfv *bríng* bring *di* def *análisis,* analysis *wi* 1pl *sí* see *di* def *dɔ́kta* doctor *dɔ́n* prf *de* ipfv *kán,* come *ya* here *era la una y algo.*

was def one and something

'As soon she [Tokobé] brought the analysis, we saw the doctor coming (when) it was already past one o'clock.'

(111) ab03ab 118

*Paciente* patient *dɛn* pl *dé* be.loc *na* loc *sala,* hall *yú* 2sg.indp *dɔ́kta* doctor *"la* def *una* one *yu* 2sg *de* ipfv *kán?"* come 'Patients are in the waiting room, (and) you doctor, "you're [only] coming at one o'clock?"'

(112) ab03ab 119

*Dɔ́kta* doctor *dɛn* pl *nó* neg *dé* cop *na* loc *dís* this *kɔ́ntri* country *na* foc *mék* make *pípul* people *dɛn* pl *de* ipfv *dáy* die *plɛ́nte.* plenty

'There are no doctors in this country, that's what's making people die a lot.'

(113) ab03ab 120

*Wántɛn* at.once *wé* sub *e* 3sg.sbj *lúk* look *di* def *pikín,* child *e* 3sg.sbj *lúk* look *di* def *análisis,* analysis *"tiene* he.has *paludismo* malaria *de* of *una* one *cruz* cross *wé* sub *kin* hab *kíl* kill *pikín* child *sɛ́f."* foc

'At once, when he looked at the child, he looked at the analysis, "he has malaria of one cross [degree of intensity] that can even kill a child".'

(114) ab03ab 121

*Yu* 2sg *de* ipfv *mɛ́mba* remember *sé* qot *e* 3sg.sbj *de* ipfv *slíp.* sleep

'You would think that he [the boy] was sleeping.'

(115) ab03ab 122 *Dɛn* 3pl *gí=an* give=3sg.obj *mɛ́rɛsin.* medicine 'He was given medicine.'

(116) ab03ab 123

*Dɛn* 3pl *rɔ́n* run *na* loc *farmacia,* pharmacy *receta* prescription *de* of *mɛ́rɛsin.* medicine

'They rushed to the pharmacy [to get a] prescription.'

(117) ab03ab 124

*Dɛn* 3pl *bin* pst *gí=an* give=3sg.obj *di* def *receta* prescription *fɔ* prep *kán* come *báy=an.* buy=3sg.obj 'They had given her [Tokobé] the prescription in order to come buy it.' (118) ab03ab 125 *Sé* qot *mɔ́mi,* mum *e* 3sg.sbj *sé* qot *"siento* I.feel *hambre."* hunger '(He) said, "mum, I feel hungry".' (119) ab03ab 126 *Mɔ́mi,* mum *gó* go *báy* buy *tú* two *bocadillo!* bun [I told his mum] 'Mum, go buy two buns!' (120) ab03ab 127 *Tú* two *brɛ́d.* bread 'Two (loaves) of bread.' (121) ab03ab 128 *Yu* 2sg *pikín* child *sidɔ́n* sit *de* ipfv *chɔ́p* eat *dɛn* 3pl *tú* two *brɛ́d.* bread 'Your child [directed at the listener [fr]] was sitting (there) eating those two (loaves of) bread.' (122) ab03ab 129 *E* 3sg.sbj *sé* qot *"a* 1sg.sbj *wánt* want *Fanta".* name 'He said "I want Fanta".' (123) ab03ab 130 *Dís* this *smɔ́l* small *bɔ́tul* bottle *dɛn* pl *Fanta,* name *wé* sub *e* 3sg.sbj *gɛ́t* get *Coca-Cola,* name *e* 3sg.sbj *gɛ́t* get *Fanta,* name *a* 3sg.sbj *gɛ́t* get *limón,* lemon *e* 3sg.sbj *báy=an* buy=3sg.obj *wán.* one 'These small bottles of Fanta, of which there is (also) Coca-Cola, there is Fanta, there is lemon, she bought one for him.' (124) ab03ab 131 *E* 3sg.sbj *nák=an.* hit=3sg.obj 'He gulped it down.' (125) ab03ab 132

*Náw* now *e* 3sg.sbj *dɔ́n* prf *wánt* want *bigín* begin *de* ipfv *fɛ́t* fight *wet* with *di* def *chía,* chair *di* def *say* side *wé* sub *dɛn* 3pl *sidɔ́n.* sit 'Now he wanted to begin fighting with the chair, where they were sitting [due to his delirium].'

(126) ab03ab 133 *Sé* qot *"nɔ́,* neg *dɔ́kta* doctor *wi* 1pl *dɔ́n* prf *fít* can *gó?"* go '(We) said "doctor, can we go now?"' (127) ab03ab 134

*E* 3sg.sbj *sé* qot *"una* 2pl *dɔ́n* prf *fít* can *gó."* go 'He said "you can already go now".'

### **15.2 Narrative and conversation: Annobón sorcery**

The following text begins with a conversation between Francisca (fr), Rubi (ru), and Djunais (dj) in which (fr) tries to persuade (ru) to give an account of how he was bewitched. Speaker (fr) manages to coax (ru) into telling the story by jokingly threatening to report to the police (015) and to bring the matter into the Equatoguinean reality TV show "Vivencias" (016)–(017). Speaker (ru) then relates in (018)–(044) how he was bewitched by a fling of his from the island of Annobón, which has caused him to fall sick with fever. The protagonists are (ru), (dj), and (ru)'s fling "the girl from Annobón". In the remainder of the text (057ff.), (fr) tries to convince (ru) and (dj) of the importance of malaria prevention.

(1) ru03wt 001

*Wán* one *Annobón* place *gɛ́l* girl *wích* bewitch *mí* 1sg.indp *mán.* intj

'A girl from Annobón bewitched me, man.'

(2) fr03wt 002

*Na* foc *wán* one *Annobón* place *gɛ́l* girl *wích* bewitch *yú?* 2sg.indp 'It's a girl from Annobón that bewitched you?'

(3) fr03wt 003

*Na* foc *fɔ* prep *dán* that *tín* thing *mék* sbjv *yu* 2sg *gó* go *dɔ́kta.* doctor 'That's why you should go to the doctor.'

(4) fr03wt 004

*Údat* who *tɛ́l* tell *yú* 2sg.indp *sé* qot *e* 3sg.sbj *wích* bewitch *yú?* 2sg.indp 'Who told you that she bewitched you?'

(5) fr03wt 005 *Na* foc *torí* story *a* 1sg.sbj *de* ipfv *hía* hear *ó!* sp 'I'm hearing the story [come on let's hear the story]!' (6) fr03wt 006 *Yu* 2sg *sabí* know *ús=káyn* q=kind *tín* thing *na* foc *wích* bewitch *nɔ́?* neg 'You know what sorcery is, right? (7) fr03wt 007 *Annobón?* place '(And) Annobón?' (8) fr03wt 008 *Yu* 2sg *sabí* know *ús=tin* q=thing *na* foc *Annobón* place *sɛ́f.* foc 'You even know what Annobón is.' (9) ko03ft 009 *Yu* 2sg *fɔgɛ́t* forget *sé* qot *a* 1sg.sbj *dɔ́n* prf *gó* go *dé.* there 'You forgot that I had already gone there.' (10) fr03wt 010 *E* 3sg.sbj *bin* pst *dé* there *na* loc *Annobón* place *yɛ́stadé.* yesterday 'He was in Annobón yesterday.' (11) fr03wt 011 *Djunais,* name *na* foc *ín* 3sg.indp *mék* make *sé* qot *mék* sbjv *dɛn* 3pl *wích=an.* bewitch=3sg.obj '(It's) Djunais, it's him who made them bewitch him.' (12) dj03wt 012 *Nó* neg *mete* put *mí* 1sg.indp *ínsay* inside *dí* this *tɔ́k* talk *a* 1sg.sbj *bɛ́g!* ask.for 'Don't involve me in this matter, please!' (13) ru03wt 013 *Na* foc *yú* 2sg.indp *mék=an.* make=3sg.obj 'It's you who made it [laughter].'


*If* if *mí* 1sg.indp *kɛ́r* carry *dís* this *plába* trouble *náw,* now *ɛ́n,* intj *na* loc *comisaría,* police.station *una* 2pl *sabí* know *sé* qot *dɛn* 3pl *nó* neg *lɛ́k* like *dís* this *tin,* thing *nó* neg *nátin* nothing *fɔ* prep *wích,* bewitch *dí* this *wán* one *go* pot *tɔ́n* turn *plába* trouble *serio.* serious 'If I take this matter, right, to the police-station, you [pl] know that they don't like this thing, nothing concerning sorcery, this would turn into serious trouble.'

(16) fr03wt 016

*Áfta* then *dɛn* 3pl *go* pot *kɛ́r* carry *una,* 2pl *na* loc *Vivencias* name *fɔ,* prep *ús=wán* q=one *na* foc *in* 3sg.poss *ném?* name 'Then they'd take you [pl] to Vivencias" to, what's his name?'

(17) fr03wt 017

*Fɔ* prep *Olinga,* name *wé* sub *e* 3sg.sbj *go* pot *gó* go *chám* chew *in* 3sg.poss *Panyá* Spanish *dé.* be.loc 'To Olinga and he would go speak his bad Spanish there.'

(18) fr03wt 018

*Una* 2pl *púl* remove *di* def *torí!* story 'Tell [pl] the story!'

(19) ru03wt 019

*E* 3sg.sbj *dé* be.loc *sé* qot *dán* that *gɛ́l* girl *e* 3sg.sbj *bin* pst *de* ipfv *kán* come *yá.* here

'It's that that girl used to come here.'

(20) ru03wt 020

*Mí* 1sg.indp *nó* neg *bin* pst *de* ipfv *lúk=an* look=3sg.obj *ó.* sp

'Mind you, I [emp] wasn't looking at [paying attention to] her.'

```
(21) ru03wt 021
```
*Djunais* name *tɔ́k* talk *sé,* qot *nɔ́* neg *Rubi* name *dí* this *gɛ́l* girl *lɛ́k* like *yú,* 2sg.indp *dí* this *gɛ́l* girl *lɛ́k* like *yú,* 2sg.indp *náw* now *bigín mék=an só.*

begin make=3sg.obj like.that

'Djunais said, no Rubi, this girl likes you, this girl likes you, now begin doing it like this.'


*A* 1sg.sbj *gó,* go *a* 1sg.sbj *lúk* look *di* def *gɛ́l,* girl *wi* 1pl *bigín* begin *tɔ́k,* talk *wi* 1pl *bigín* begin *tɔ́k,* talk *wi* 1pl *bigín* begin *tɔ́k* talk *tal* so *tal.*

so

'I went, I had a look at the girl, we began to talk and talk and talk, and so on.'

(26) ru03wt 026

*Tumɔ́ro* tomorrow *di* def *gɛ́l* girl *wánt* want *sé* qot *mék* sbjv *wi* 1pl *slíp.* sleep

'The next day the girl wanted us to sleep (with each other).'

(27) ru03wt 027

*E* 3sg.sbj *insiste* insist *sóté* until [click].

'She insisted until [clicks with his fingers].'

(28) fr03wt 028 *Una slíp?*

2pl sleep

'You slept (with each other)?'

(29) ru03wt 029

*Yɛ,* yes *a* 1sg.sbj *kán* pfv *tɛ́l=an* tell=3sg.obj *sé* qot *'chica,* girl *mí* 1sg.indp *nó* neg *lɛ́k* like *yú* 2sg.indp *bɔt* but *wi* 1pl *fít* can *dé lɛk kɔ́mpin'.*

be.loc like friend

'Yeah, I eventually told her "girl, I [emp] don't love you but we can be like friends".'

(30) ru03wt 030 *"A* 1sg.sbj *wɔ́nt* want *mék* sbjv *yu* 2sg *dú* do *mí* 1sg.indp *sɔn* some *fébɔ,* favour *mék* sbjv *yu* 2sg *wás* wash *mí* 1sg.indp *sɔn* some *klós* clothing *dɛn."* pl '"I want you to do me a favour and wash some clothes for me."' (31) fr03wt 031 *Ɔ́l* all *dán* that *tɛ́n* time *Djunais* name *de* ipfv *gív=an* give=3sg.obj *di* def *acción,* action *e* 3sg.sbj *de* ipfv *pút* put *calor.* heat 'All that time Djunais was causing commotion, he was fanning the flames.' (32) fr03wt 032 *Djunais* name *yu* 2sg *badhát* be.mean *ɛ́n.* intj 'Djunais, you're mean, you know.' (33) ru03wt 033 *E* 3sg.sbj *gó,* go *e* 3sg.sbj *wás* wash *di* def *klós* clothing *dɛn.* pl 'She went (and) she washed the clothes.' (34) ru03wt 034 *E* 3sg.sbj *wás* wash *di* def *klós* clothing *dɛn,* pl *e* 3sg.sbj *dráy* dry *dɛ́n,* 3pl.indp *nɔ́,* neg *na* foc *mí* 1sg.indp *dráy* dry *dɛ́n.* 3pl.indp 'She washed the clothes, she dried them, no, it was me who dried them.' (35) ru03wt 035 *Pero* but *di* def *klós* clothing *dɛn* pl *slíp* lie *na* loc *dɔ́n* down *ó.* sp 'But the clothes came to lie down [on the ground].' (36) ru03wt 036 *Mɔ́nin* morning *tɛ́n* time *wé* sub *a* 1sg.sbj *kán* pfv *lúk* look *a* 1sg.sbj *de* ipfv *sí* see *sɔn* some *klós* clothing *dɛn,* pl *a* 1sg.sbj *nó* neg *de* ipfv *sí* see *mi* 1sg.poss *yón* own *dɛn.* pl 'In the morning, when I came to look, I saw some clothes (but) I didn't see mine.' (37) ru03wt 037 *Áfta* then *a* 1sg.sbj *de* ipfv *mít=an* meet=3sg.obj *nía* near *di* def *klós* clothing *dɛn* pl *di* def *mɔ́nin* morning *mɔ́nin* rep *tɛ́n.* time

'Then I find her next to the clothes early in the morning.'

(38) ru03wt 038

*A* 1sg.sbj *áks=an* ask=3sg.obj *sé* qot *"ús=say* q=side *di* def *klós* clothing *dɛn* pl *dé?"* be.loc 'I asked her "where are the clothes?"'

(39) ru03wt 039

*E* 3sg.sbj *sé* qot *"nó,* neg *a* 1sg.sbj *de* ipfv *sí* see *lɛk* like *sé* qot *dɛn* 3pl *dɔ́n* prf *tíf* steal *sɔn".* some 'She said "no, it seems to me like some have been stolen".'

(40) ru03wt 040

*Ús=say* q=side *mi* 1sg.poss *klós* clothing *dɛn* pl *dé,* be.loc *di* def *ívin* evening *tɛ́n,* time [click] *fíba,* fever *fíba* fever *sóté* until *a kɔ́l=an.*

1sg.sbj call=3sg.obj

'Where were my clothes, in the evening [clicks with his fingers], fever, fever until finally I called her.'

(41) ru03wt 041

*E* 3sg.sbj *dé,* be.loc *e* 3sg.sbj *nó* neg *de* ipfv *ánsa* answer *mí* 1sg.indp *mɔ́,* more *e* 3sg.sbj *de* ipfv *pás* pass *só* so *lɛk* like *sé* qot *e* 3sg.sbj *nó* neg *nó* know *mí* 1sg.indp *mɔ́.* more

'She was there and wasn't responding to me anymore, she was passing by as if she didn't know me anymore.'

(42) ru03wt 042

*A* 1sg.sbj *tɛ́l=an* tell=3sg.obj *sé* qot *"chica,* girl *sóté* until *yu* 2sg *de* ipfv *kán* come *na* loc *mi* 1sg.poss *drim* dream *dɛn* pl *ɛ́n,* intj *na* foc *só* so *só* rep *tín* thing *yu* 2sg *mék* make *mí,* 1sg.indp *tráy* try *reduce* reduce *ín".* 3sg.indp

'I told her "girl, you even come into my dreams, you know, it's this and that you did to me, try to reduce that".'

(43) ru03wt 043

*"¿Tú piensas eso de mí?"*

'[She replied] "You think that of me?"'

(44) ru03wt 044

*A* 1sg.sbj *dɔ́n* prf *explica* explain *Boyé* name *dɛn,* pl *sé* qot *na* foc *só* so *mí* 1sg.indp *de* ipfv *mɛ́mba,* remember *ɔ́l* all *tín.* thing 'I've already explained to Boyé and the others, that's how I remember everything.'

(45) fr03wt 045 *Yu* 2sg *dɔ́n* prf *gó* go *sí* see *yu* 2sg *mamá?* mother 'Have already gone to see your mother?' (46) ru03wt 046 *Nɔ́.* neg 'No.' (47) fr03wt 047 *Wétin* what *yu* 2sg *de* ipfv *wét?* wait 'What are you waiting (for)?' (48) dj03wt 048 *Sé* qot *in* 3sg.poss *mamá* mother *go* pot *dráyb=an* drive=3sg.obj *fɔ́s.* first 'Because his mother would chase him away first.' (49) fr03wt 049 *In* 3sg.poss *mamá* mother *go* pot *dráyb=an* drive=3sg.obj *fɔ́s* first *pero* but *in* 3sg.poss *mamá* mother *na* foc *di* def *pɔ́sin* person *wé* sub *e* 3sg.sbj *fít* can *gó* go *wáka* walk *wet=an,* with=3sg.obj *mí* 1sg.indp *nó* neg *sabí* know *wáka.* walk 'His mother could chase him away first but his mother is the only person that could go walk with him [i.e. take care of his spiritual protection], I don't know how to walk.' (50) ru03wt 050 *Annobón* place *mɛ́rɛsin* sorcery *nó* neg *de* ipfv *tɔ́n* turn *mi* 1sg.poss *héd.* head 'Annobón sorcery doesn't turn my head [have an effect on me].' (51) dj03wt 051 *Annobón* place *mɛ́rɛsin,* sorcery *e* 3sg.sbj *nó* neg *de* ipfv *gó* go *bihɛ́n.* behind 'As for Annobón sorcery, it doesn't go behind [have a profound effect].' (52) fr03wt 052 *No obstante,* nonetheless *a* 1sg.sbj *bɛ́g* ask.for *gó* go *sí* see *dɔ́kta* doctor *fɔ́s,* first *hía?* hear 'Nonetheless, please go see the doctor first, (you) hear?'

*ónli* only (53) fr03wt 053

*Na* foc *fɔ* prep *dán* that *tín* thing *yu* 2sg *nó* neg *de* ipfv *gó* go *dɔ́kta* doctor *porque* because *yu* 2sg *de* ipfv *chɛ́k* think *sé* qot *na* foc *wích?*

bewitch

'Is that why you're not going to the doctor because you think it's witchcraft?'

(54) fr03wt 054

*Gó* go *dɔ́kta* doctor *fɔ́s,* first *wé* sub *di* def *dɔ́kta* doctor *go* pot *gí* give *yu* 2sg *sɔn* some *tín* thing *mék* sbjv *yu* 2sg *fíl* feel *smɔ́l* a.bit *fáyn,* fine *yu* 2sg *bigín* begin *mék* make *di* def *ɔ́da* other *tín* thing *dɛn.* pl

'Go to the doctor first, when the doctor will give you something for you to feel a fine a bit, you begin to do the other things.'

(55) fr03wt 055

*Yú* 2sg.indp *de* ipfv *hía?* hear 'Do you hear?'

(56) ru03wt 056 *A* 1sg.sbj *hía.* hear 'I hear.'

(57) fr03wt 057

*E* 3sg.sbj *fít* can *bí* be *sé* qot *na* foc *paludismo.* malaria 'It could be that it's malaria.'

(58) fr03wt 058

*Ús=tɛ́n* q=time *una* 2pl *lás* be.last *impregna* impregnate *una* 2pl *mosquiteros* mosquito.nets *dɛn?* pl

'When did you [pl] last impregnate your [pl] mosquito nets?'

(59) fr03wt 059

*E* 3sg.sbj *dɔ́n* prf *sté,* last *a* 1sg.sbj *tínk* think *sé* qot *e* 3sg.sbj *dɔ́n* prf *sté* last *wé* sub *una* 2pl *bin* pst *gɛ́t* get *insecticida* insecticide *yá.*

here

'It's been a long time, I think that it's been a long time that you had insecticide here.'

(60) fr03wt 060

*Dán* that *bíg* big *bíg* rep *mosquito* mosquito *dɛn* pl *wé* sub *dɛn* 3pl *fíba* resemble *aviones* plane.pl *dɛn.* pl

'Those huge mosquitos that resemble airplanes.'

(61) fr03wt 061

*Aunque* even.if *nóto* neg.foc *paludismo,* malaria *if* if *dɛn* 3pl *gív* give *yú* 2sg.indp *tratamiento* treatment *yu* 2sg *nó* neg *go* pot *dáy.* die 'Even if it's not malaria, if they give you a treatment you won't die.'

### **15.3 Conversation: Dinner for four**

The text that follows is an extensive conversation involving four people: Boyé (ye), Djunais (dj), Francisca (fr), and sporadically myself (ko). The conversation was recorded during a dinner hosted by (fr). A relaxed and cheerful atmosphere reigns during the conversation and the discourse participants, who are members of the same extended family, joke and tease each other on numerous occasions (e.g. in (015)–(019), (091)–(94) and the entire section from (130)–(143)). The conversation also contains many instances of Pichi-Spanish codemixing (e.g. (001)–(008)).

The text features three themes between which the speakers switch to and fro.The main theme is the ongoing construction of a family house commissioned by (fr) and overseen by (ye). This discussion is contained in sections (001)–(038), (99)–(120), (154)–(164), and (173)–(178) and is chiefly concerned with problems in a cement delivery ordered from two protagonists named Buehu and Gabriel. The sections on the construction works are driven by (fr), who repeatedly brings the conversation topic back to this issue of great importance to her.

A second theme revolves around eating. In (080)–(097), (dj) and (ye) comment on each other's cooking abilities, in (121)–(127), an exchange ensues about the effect of the pepper in the food, and in (132)–(143), (ye) teases (dj) because the latter has just drunk tap water (which is not without risk in Malabo). In (144)–(153) and (164)–(172), both (dj) and (ye) complain about the eating habits of Pancho (pa) who is not present at the table. Both (dj) and (ye) live in one place with (pa) and the account of (ye) in (173)–(178) shows that (pa) was also supposed to run an errand for (fr) as part of the building activities. A third theme is the interlude in (051)–(078) in which (fr) and (ye) scoff at Olinga, the TV presenter of "Vivencias", a popular Equaotoguinean TV reality show.

(1) ye03cd 001

*Pues* so *hemos* we.have *estado* been *ahí,* there *a* 1sg.sbj *tínk* think *sé* qot *wán* one *las* the.pl *cuatro* four *wé* sub *di* def *chɛ́f* boss *kɔmɔ́t* go.out *e* 3sg.sbj *nó* neg *aparece* appear *yet.* yet

'So we were there, I think around four o'clock that the boss went out (and) he hadn't appeared yet.'

(2) ye03cd 002

*Di* def *ɔ́da* other *mán* man *tɛ́l* tell *mí* 1sg.indp *sé* qot *dɛn* 3pl *dɔ́n* prf *báy* buy *veinte* twenty *sacos.* bags 'The other man told me that they had bought twenty bags.'

(3) ye03cd 003

*E* 3sg.sbj *lɛ́f* remain *doce.* twelve

'Twelve remain.'

(4) ye03cd 004

*E* 3sg.sbj *falta* lack *mɔní* money *fɔ* prep *púl* remove *saco* sack *dɛn* pl *dé* there *fɔ* prep *kɛ́r=an* carry=3sg.obj *na* loc *hós.* house 'The money is lacking to remove the bags there in order to bring them to the house.'

(5) fr03cd 005

*Me* me *van* they.will *a* to *tocar* touch *los* the.pl *cojones* testicle.pl *porque* because *mí* 1sg.indp *gí* give *dɛ́n* 3pl.indp *diez* ten *mil fɔ transporte.*

thousand prep transport

'They're going to get me really annoyed because I gave them ten thousand for transport.'

(6) ye03cd 006

*Na* foc *só* so *ín* 3sg.indp *de* ipfv *tɛ́l* tell *mí.* 1sg.indp 'That's what he [emp] told me.'

(7) fr03cd 007

*Mék* sbjv *dɛn* 3pl *transporta* transport *di* def *cemento* cement *na* loc *Ela Nguema* place *porque* because *no* neg *estaba* was *dicho* said *que* that *dɛn* 3pl *go* pot *gó* go *lɛ́f* leave *di* def *cemento.* cement

'Let them transport the cement to Ela Nguema because it hadn't been agreed that they would go leave the cement [lying there].'

(8) fr03cd 008

*Di* def *cemento,* cement *estaba* was *dicho* said *que* that *na* foc *fɔ* prep *kɛ́r=an* carry=3sg.obj *directamente* directly *na* loc *Ela Nguema.*

place

'The cement, it had been agreed that it is to be taken directly to Ela Nguema.'

(9) fr03cd 009

*Dát* that *min* mean *sé* qot *Buehu* name *nó* neg *kán* come *e* 3sg.sbj *nó* neg *gí* give *nó* neg *mɔní* money *nó* neg *nátin.* nothing 'That means that Buehu didn't come (and) he didn't give (them) any money at all.'


*Tumɔ́ro* tomorrow *mɔ́nin* morning *tɛ́n,* time *wán* one *las* the.pl *siete* seven *só* like.that *a* 1sg.sbj *go* pot *gó* go *dé.* there 'Tomorrow morning, around seven o'clock or so I'll go there.'

(12) ye03cd 012

*Ɔ* or *bɔkú* much *mán* man *dɛn* pl *bin* pst *de* ipfv *fɛ́n/* look.for 'Or many people were looking for/'


*Ɔ́l* all *dí* this *compromiso* agreement *dɛn* pl *fɔ* prep *escribiendo* writing *dán* that *bɔ́y* boy *in* 3sg.poss *apellido,* surname *wétin* what *mék* make *yu ráyt mi ném?*

2sg write 1sg.poss name

'All these agreements writing that guy's surname, how come you've written my name?'

(15) ye03cd 015

*Mék* sbjv *nó* neg *mi* 1sg.poss *caligrafía* handwriting *gó* go *na* loc *dán* that *pépa!* paper

'None of my handwriting should go on that paper!'


*Na* foc *fɔ* prep *ús=káyn* q=kind *tín,* thing *explica* explain *mí!* 1sg.indp 'It's for what, explain to me!'

(18) fr03cd 018 *A* 1sg.sbj *níd* need *wán* one *lista* list *de* of *participantes.* participants 'I need a list of participants.' (19) ye03cd 019 *Na* foc *compromiso* agreement *dát* that *ó.* sp 'That's actually an agreement.' (20) ye03cd 020 *Dán* that *ficción,* fiction *Bata-mán* place-man *go* pot *tɔ́n=an* turn=3sg.obj *rɔn-say.* wrong.cpd-side 'That fiction [fictitious agreement], the Fang [the person delivering the cement] will turn it upside down.' (21) ye03cd 021 *E* 3sg.sbj *sé* qot *ɔ́l* all *tidé* today *e* 3sg.sbj *bin* pst *de* ipfv *kɔ́l* call *yú* 2sg.indp *yu* 2sg *nó* neg *ték* take *teléfono.* telephone 'He said the whole of today, he was calling you (and) you didn't pick the phone.' (22) fr03cd 022 *"Ɔ́l* all *tidé* today *e* 3sg.sbj *bin* pst *de* ipfv *kɔ́l* call *mí",* 1sg.indp *e* 3sg.sbj *kɔ́l* call *mí* 1sg.indp *wán* one *tɛ́n* time *dásɔl.* only '"All of today he was calling me", he called me only once.' (23) fr03cd 023 *Bueno,* alright *a* 1sg.sbj *bɛ́g* ask.for *tumɔ́ro,* tomorrow *gó* go *recupera* recover *di* def *mɔní* money *wé* sub *yu* 2sg *lɛ́f.* leave 'Alright, please tomorrow, go recover the money that you left.' (24) fr03cd 024 *Náw* now *só* so *a* 1sg.sbj *gɛ́fɔ* have.to *pé* pay *mɔní* money *mɔ́* more *fɔ* prep *gó* go *kɛ́r* carry *di* def *cemento* cement *na* loc *hós.* house 'Now I have to pay money again in order to bring the cement to the house.' (25) fr03cd 025 *Nɔ́,* neg *yu* 2sg *sabí* know *di* def *tín* thing *wé* sub *yu* 2sg *go* pot *tɛ́l=an?* tell=3sg.obj 'No, you know what you're going to tell him?' (26) fr03cd 026 *Yu* 2sg *go* pot *tɛ́l=an* tell=3sg.obj *sé* qot *wi* 1pl *nó* neg *de* ipfv *ték* take *di* def *cemento,* cement *mék* sbjv *e* 3sg.sbj *bák* give.back *yú* 2sg.indp *di* def *mɔní* money *wé* sub *e* 3sg.sbj *gɛ́t,* get *ɛ́n?* intj 'You'll tell him that we're not taking the cement (and) that he should give you

back the money that he has, right?'

(27) fr03cd 027 *Mék* sbjv *e* 3sg.sbj *bák* give.back *yú* 2sg.indp *di* def *mɔní* money *wé* sub *e* 3sg.sbj *gɛ́t.* get 'Let him give you back the money that he has.' (28) ye03cd 028 *Di* def *wán* one *wé* sub *e* 3sg.sbj *dɔ́n* prf *sobra* remain *ín.* 3sg.indp 'The one [amount] that has remained with him.' (29) fr03cd 029 *Di* def *wán* one *wé* sub *e* 3sg.sbj *dɔ́n* prf *sobra* remain *ín.* 3sg.indp 'The one that has remained with him.' (30) fr03cd 030 *Áfta,* then *como* because *ín* 3sg.indp *níd* need *cemento,* cement *mék* sbjv *e* 3sg.sbj *gó* go *gí* give *yú* 2sg.indp *di* def *mɔní,* money *nóto* neg.foc *tumɔ́ro* tomorrow *e* 3sg.sbj *go* pot *gí* give *di* def *mɔní.* money 'Then, since he [emp] needs cement [as well], let him go give you the money, it won't be tomorrow that he'll give (you) the money.' (31) fr03cd 031 *Pero* but *di* def *tɛ́n* time *wé* sub *ín* 3sg.indp *go* pot *gɛ́fɔ* have.to *báy* buy *cemento,* cement *mék* sbjv *e* 3sg.sbj *ték* take *dán* that *cemento* cement *dé.* there 'But when he himself has to buy cement, let him take that cement (there).' (32) fr03cd 032 *Áfta,* then *bueno,* alright *tɛ́l=an* tell=3sg.obj *sé* qot *mék* sbjv *e* 3sg.sbj *bák* give.back *yú* 2sg.indp *di* def *mɔní,* money *mí* 1sg.indp *go* pot *tɔ́k* talk *wet=an.* with=3sg.obj 'Then, alright, tell him to give you back the money, I myself will talk to him.' (33) fr03cd 033 *Mék* sbjv *e* 3sg.sbj *ték* take *dán* that *cemento.* cement 'Let him take that cement.' (34) fr03cd 034 *Áfta,* then *e* 3sg.sbj *go* pot *ték* take *di* def *cemento.* cement 'Then he'll take the cement.'

	- 'No, no, no, I'm not into that (kind of) style.'

*Dɛn* 3pl *go* pot *só=an* show=3sg.obj *na* loc *Vivencias,* name *na* foc *di* def *tín* thing *dɛn* 3pl *de* ipfv *tɔ́k* talk *dé.* there 'They'll show it on Vivencias [a TV show], that's the (kind of) thing they talk (about) there.'

(52) ye03cd 052

*Wán* one *mán* man *wé* sub *e* 3sg.sbj *nó* neg *gɛ́t* get *mɔní,* money *e* 3sg.sbj *dɔ́n* prf *gɛ́t* get *sesenta* sixty *años.* year.pl

'[In Vivencias there was] a man who doesn't have money, he's already sixty years old.'

(53) ye03cd 053

*E* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *ni* even *Panyá,* Spanish *e* 3sg.sbj *sé* qot *e* 3sg.sbj *wánt* want *muchachita* young.girl *dé* of *diecisiete años.*

seventeen years

'He didn't even know how to speak Spanish, he said he wants a young girl of seventeen years.'

(54) ye03cd 054

*E* 3sg.sbj *sé* qot *"yo* I *quiero* I.want *una* one *muchachita* young.girl *de* of *diecisiete* seventeen *años".* year.pl 'He said "I want a young girl of seventeen years".'

(55) ye03cd 055

*E* 3sg.sbj *nó* neg *wánt* want *ni* neither *treinta* thirty *ni* neither *cuarenta,* forty *mm* intj *mm,* intj *diecisiete* seventeen *años.* year.pl 'He neither wanted thirty nor forty years, no no, seventeen years.'

(56) ye03cd 056

*Na* loc *Vivencias* name *dɛn* 3pl *de* ipfv *só* show *dán* that *tín.* thing

'In Vivencias they show that (kind of) thing.'

(57) ye03cd 057

*Ús=tín* q=thing *dɛn* 3pl *kin* hab *de* ipfv *gó* go *fɛ́n* look.for *mán* man *dɛn* pl *wé* sub *nó* neg *sabí* know *tɔ́k* talk *Panyá?* Spanish 'Why do they always go look for people who don't know how to talk Spanish?'

(58) ye03cd 058

*Dí* this *Olinga.* name 'This Olinga.'

(59) fr03cd 059

*A* 1sg.sbj *nó* neg *sabí* know *lɛk* like *háw* how *e* 3sg.sbj *dé* be.loc *in* 3sg.poss *wók,* work *bɔt* but *e* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *Panyá.* Spanish

'I don't know how he is at his work, but he doesn't know how to speak Spanish.'

(60) ko03cd 060 *Údat?* who

'Who?'

(61) fr03cd 061

*Olinga* name *na* foc *wán* one *presentador* presenter *fɔ* prep *wán* one *programa.* programme 'Olinga is a presenter of a programme.'

(62) fr03cd 062

*Na* foc *reportero,* reporter *el* the *programa* programme *más* most *popular* popular *de* of *este* this *país,* country *in* 3sg.poss *ném* name *na* foc *Vivencias.*

name

'He's a reporter, the most popular programme of this country, its name is Vivencias.'

(63) fr03cd 063

*Di* def *mán* man *e* 3sg.sbj *nó* neg *sabí* know *tɔ́k* talk *Panyá.* Spain

'The man doesn't know how to speak Spanish.'

(64) fr03cd 064

*E* 3sg.sbj *de* ipfv *chɛ́r* tear *wán* one *káyn* kind *chɛ́r* tear *min* mean *sé* qot *e* 3sg.sbj *de* ipfv *mék* make *fallos* error.pl *dɛn,* pl *faltas.*

mistake.pl

'He "tears one kind of Spanish" means that he makes errors, mistakes.'

(65) fr03cd 065

*E* 3sg.sbj *de* ipfv *chɛ́r* tear *wán* one *káyn* kind *chɛ́r* tear *wé* sub *mí* 1sg.indp *yón* own *Panyá/* Spanish 'He makes such serious mistakes where my own Spanish/'

(66) ye03cd 066

*E* 3sg.sbj *tɔ́k* talk *sé* qot *ín* 3sg.indp *na* foc *poeta.* poet

'He said he's a poet.'

(67) ye03cd 067

*E* 3sg.sbj *kin* hab *de* ipfv *híb* throw *sɔn* some *poesía* poetry *dɛn,* pl *chico* intj *nɔ́.* neg 'He kicks some poetry, man really.'

(68) ye03cd 068

*Olinga* name *kɔmɔ́t* come.out *frɔn* from *bɔtɔ́n.* bottom

'Olinga comes from the bottom [has very modest origins].'

(69) ko03cd 069

*Bɔt* but *na* foc *ín* 3sg.indp *wé* sub *pípul* people *layk=an* like=3sg.obj *nɔ́.* neg 'But that's why people like him, right.'

(70) fr03cd 070

*Áfta* then *primera* first *dama* lady *báy=an* buy=3sg.obj *wán* one *motó,* car *wán* one *todo terreno,* cross-country.vehicle *wán* one *cuatro por cuatro,* four-wheel.drive *mék* sbjv *e* 3sg.sbj *fít* can *de* ipfv *rích* arrive *ɔ́l* all *say* side *sóté* until *na* loc *Riaba.* place

'Then the first lady bought him a car, a cross-country vehicle, a four-wheel drive so that he could reach all places even up to Riaba.'

(71) fr03cd 071

*Wán* one *dé* day *wán* one *pikín* child *bin* pst *de* ipfv *sík.* be.sick 'One day a child was sick.'

(72) fr03cd 072

*A* 1sg.sbj *nó* neg *sabí* know *ús=káyn* q=kind *tín* thing *bin* pst *pás.* happen 'I don't know what had happened.'

(73) fr03cd 073

*Nó,* neg *na* foc *wán* one *mán,* man *ɛhɛ́* exactly *wán* one *accidente* accident *fɔ* prep *motó* car *bin* pst *dé.* be.loc 'No, it was a man, oh yes, there had been a car accident.'

(74) fr03cd 074

*A* 1sg.sbj *nó* neg *sabí* know *ús=tín* q=thing *bin* pst *kán* pfv *pás* happen *áfta* then *e* 3sg.sbj *gó* go *na* loc *hospital.* hospital 'I don't know what had happened for him to arrive at the hospital.'

(75) fr03cd 075

*Di* def *bɔ́y* bɔ́y *dé* be.loc *dé* there *e* 3sg.sbj *dɔ́n* prf *de* ipfv *dáy.* die

'The guy [a casuality] was already dying.'

(76) fr03cd 076

*E* 3sg.sbj *pút* put *micrófono* microfone *así,* so *e* 3sg.sbj *sé* qot *"los* the.pl *últimos* last.pl *suspiros,* sigh.pl *de* of *un* one *momento* moment *al otro se va a morir".*

to other refl go to die

'He put the microfone like this, he said "the last sighs, from one moment to another he'll die".'

(77) fr03cd 077


*púl* remove *yú* 2sg.indp *wók.* work

'This is in another country, they would remove you immediately, they would remove you from work.'

(78) ye03cd 078 *Mán* man *dɔ́n* prf *diaboliza.* diabolise 'People have become devilish.'


*Lɛ́f=an,* leave=3sg.obj *a* 1sg.sbj *go* pot *chɔ́p,* eat *áfta* then *a* 1sg.sbj *go* pot *dring.* drink

'Leave it, I'll eat, then I'll drink.'

(81) ye03cd 081 *Yu* 2sg *nó* neg *sabí* know *na* foc *mí* 1sg.indp *kúk?* cook 'You don't know it's me who cooked?'


(85) fr03cd 085 *Ús=káyn* q=kind *tín* thing *e* 3sg.sbj *kúk,* cook *fray-rɛ́s?* fry.cpd-rice 'What did he cook, fried rice?'


*Di* def *tɔ́p,* top *di* def *nota* mark *máxima* highest *na* foc *diez,* ten *entonces* so *yu* 2sg *de* ipfv *gí* give *mí* 1sg.indp *cuatro* four *con* with *cinco.*

five

'The top, the highest mark is ten, and you give me four over five.'

(94) ye03cd 094 *Nó* neg *wi* 1pl *de* ipfv *conversa,* converse *nó* neg *vɛ́ks* be.angry *Djunais.* name 'No, we're conversing, don't be angry Djunais.' (95) dj03cd 095 *A* 1sg.sbj *nɔ́* neg *de* ipfv *vɛ́ks.* be.angry 'I'm not angry.' (96) ye03cd 096 *Wi* 1pl *de* ipfv *conversa* converse *na* loc *tébul.* table 'We're conversing at the table.' (97) ye03cd 097 *A* 1sg.sbj *nó* neg *fít* can *kɛri* carry *yú* 2sg.indp *restaurante* restaurant *bikɔs* because *sé* qot *yu* 2sg *go* pot *fɛ́t* fight *wet* with *sɔn* some *pɔ́sin* person *dé.* there 'I can't take you to a restaurant because you would fight with somebody there.' (98) fr03cd 098 *Wétin* what *yu* 2sg *de* ipfv *tɔ́k/* talk 'What are you talking/ [music from below drowns the recording for a few minutes] (99) ye03cd 099 *Gabriel* name *e* 3sg.sbj *gɛ́t* get *jege.* ? Gabriel has a "jege".' (100) fr03cd 100 *Ús=káyn* q=kind *tín* thing *na* foc *jege?* ? 'What's a "jege"?' (101) ye03cd 101 *E* 3sg.sbj *gɛ́t* get *sɔn,* some *sɔn* some *smɔ́l* small *jege* ? *ínsay* inside *in* 3sg.poss *yáy* eye *só.* like.that 'He has a, a small "jege" inside his eye like that.' (102) fr03cd 102 *Ús=káyn* q=kind *tín* thing *na* foc *jege?* ? 'What's "jege"?'

(103) dj03cd 103 *Dán* that *wáyt* white *tín* thing *wé* sub *e* 3sg.sbj *dé* be.loc *na* loc *in* 3sg.poss *yáy.* eye 'That white thing that's in his eye.' (104) ye03cd 104 *Sɔn* some *tín* thing *de* ipfv *tapa* cover *in* 3sg.poss *retina.* retina 'Something covers his retina.' (105) dj03cd 105 *Nɔ́,* neg *wán* one *accidente* accident *wé* sub *e* 3sg.sbj *bin* pst *gɛ́t.* get 'No, it's an accident that he had.' (106) ye03cd 106 *Wán* one *yáy* eye *dé* be.loc *blanco,* white *e* 3sg.sbj *nó* neg *de* ipfv *sí.* see 'One eye is white, it doesn't see.' (107) fr03cd 107 *Mí* 1sg.indp *nó* neg *fija* notice *ín.* 3sg.indp 'I didn't notice it.' (108) ye03cd 108 *Yu* 2sg *gɛ́fɔ* have.to *fija* notice *ín.* 3sg.indp 'One has to notice it.' (109) ye03cd 109 *E* 3sg.sbj *de* ipfv *para* stand *na* loc *dán* that *in* 3sg.poss *yáy* eye *bɔkú* much *bád.* extremely 'It sits there in his eye real bad.' (110) ye03cd 110 *Ɛhɛ́,* exactly *ús=tín* q=thing *wi* 1pl *go* pot *tɔ́k* talk *fɔ* prep *Gabriel?* name 'So, what are we going to say to Gabriel?' (111) fr03cd 111 *Tumɔ́ro,* tomorrow *lɛk* like *háw* how *yu* 2sg *tɔ́k* talk *wet* with *Buehu,* name *yu* 2sg *kɔ́l* call *mí,* 1sg.indp *if* if *yu* 2sg *tɔ́k* talk *wet=an* with=3sg.obj *ɔ* or *yu* 2sg *nó* neg *tɔ́k* talk *wet=an.* with=3sg.obj Tomorrow, as soon as you've talked to Buehu, you call me, whether you talk to him or you don't talk to him.'

#### (112) fr03cd 112

*If* if *yu* 2sg *nó* neg *tɔ́k* talk *wet=an,* with=3sg.obj *yu* 2sg *kɔ́l* call *mí,* 1sg.indp *dán* that *tɛ́n* time *yu* 2sg *go* pot *gɛ́fɔ* have.to *kán* come *na* loc *wók,* work *mék* sbjv *a* 1sg.sbj *gí* give *yú* 2sg.indp *wán* one *cheque,* cheque *mék* sbjv *yu* 2sg *gó* go *ték* take *mɔní* money *fɔ* prep *báy* buy *di* def *bloques* brick.pl *dɛn* pl *wé* sub *dɛn* 3pl *lɛ́f.* remain

'If you don't talk to him, you call me, then you'll have to come to work, so that I give you a cheque, in order for you to go get money to buy the remaining bricks.'

(113) fr03cd 113

*Áfta* then *yu* 2sg *fɔ* prep *pé* pay *dɛ́n.* 3pl.indp 'Then you have to pay them.'

(114) fr03cd 114

*Sɔn* some *bloques* brick.pl *dɛn* pl *lɛ́f* remain *wé* sub *dɛn* 3pl *gɛ́fɔ* have.to *monta* mount *nɔ́?* neg 'Some bricks remain that they have to build up, right?'

(115) fr03cd 115

*Pero* but *e* 3sg.sbj *bin* pst *tɛ́l* tell *mí* 1sg.indp *sé* qot *mék* sbjv *a* 1sg.sbj *báy* buy *cuarenta* forty *bloques.* brick.pl 'But he told me to buy forty bricks.'

(116) ye03cd 116

*Yɛ́s* yes *cuarenta.* forty 'Yes forty.'

(117) fr03cd 117

*Entonces* so *fɔ* prep *monta* mount *ɔ́l* all *di* def *baf-rúm,* bathe.cpd-room *e* 3sg.sbj *bin* pst *tɛ́l* tell *mí* 1sg.indp *sé* qot *na* foc *cuarenta* forty *mil* thousand *fɔ* prep *di* def *wók* work *wet* with *di* def *bloques* brick.pl *dɛn* pl *wé* sub *dɛn* 3pl *lɛ́f.* remain 'So in order to build the whole bathroom, he had told me that it's forty thousand for the work with the remaining bricks.'

(118) fr03cd 118

*Dán* that *tɛ́n* time *tumɔ́ro,* tomorrow *ɛf* if *yu* 2sg *nó* neg *kán* pfv *sí* see *dán* that *mán,* man *mék* sbjv *a* 1sg.sbj *kán* come *mék* sbjv *a* 1sg.sbj *gí yú di mɔní.*

give 2sg.indp def money

'By that time tomorrow, if you don't see that man, let me come and give you the money.'

(119) ye03cd 119 *Mék* sbjv *a* 1sg.sbj *gí* give *yú* 2sg.indp *di* def *cheque* cheque *mék* sbjv *yú* 2sg.indp *gó* go *na* loc *banco* bank *yu* 2sg *gó* go *cobra.* receive 'Let me give you the cheque so that you go to the bank and receive (the money).' (120) ye03cd 120 *Áfta* then *una* 2pl *báy* buy *di* def *bloques* brick.pl *dɛn* pl *tumɔ́ro.* tomorrow 'Then you [pl] buy the bricks tomorrow.' (121) dj03cd 121 *Na* loc *in* 3sg.poss *fés,* face *na* loc *in* 3sg.poss *héd,* head *sí=an!* see=3sg.obj [Comments on the effects of the pepper in the food (ye) has just tried] 'In his face, in his head, look at him!' (122) ye03cd 122 *Tɔ́k* talk *bifó* before *di* def *[unintelligible]* 'Talk in front of/ (…)' (123) ye03cd 123 *¿Sí* yes *o* or *no?* no 'Yes or no?' (124) ye03cd 124 *Yu* 2sg *nó* neg *hía* hear *wé* sub *a* 1sg.sbj *tɛ́l* tell *Pancho* name *sé* qot *quiero* I.want *cocinar?* cook 'Didn't you hear when I told Pancho that I wanted to cook?' (125) ye03cd 125 *Djunais* name *tɔ́k* talk *trú!* true 'Djunais tell the truth!' (126) ye03cd 126 *Sóté* until *a* 1sg.sbj *tɛ́l* tell *Djunais* name *sé* qot *pút* put *mí* 1sg.indp *wet* with *Pancho,* name *wi* 1pl *go* pot *chɔ́p* eat *wán* one 'I even told Djunais to put [dish the food] for me and Pancho, we'll eat in one

*say.* side

place.'

(127) ye03cd 127

*Porque* because *a* 1sg.sbj *chɛ́k* think *sé/* qot 'Because I thought that/'

(128) fr03cd 128

*Porque* because *ɛ́ni* every *tɛ́n* time *wé* sub *mí* 1sg.indp *de* ipfv *kɔ́l/* call *e* 3sg.sbj *nó* neg *gɛ́t* get *móvil* mobile *mɔ́?* more 'Because anytime that I call/ doesn't he have a mobile-phone anymore?'

(129) ye03cd 129

*Ús=nɔ́mba* q=number *yu* 2sg *gɛ́fɔ* have.to *dán* that *móvil?* mobile

'Which number do you have in that [your] mobile?'

(130) ye03cd 130

*Yu* 2sg *nó* neg *sí* see *dán* that *gyál* girl *de* ipfv *chénch,* change *e* 3sg.sbj *de* ipfv *chénch-chénch* red.cpd-change *dán* that *nɔ́mba* number *dɛn* pl *lɛk* like *terrorista* terrorist *wé* sub *e* 3sg.sbj *nó* neg *wánt* want *mék* sbjv *dɛn* 3pl *kéch=an.* catch=3sg.obj

'Don't you see that girl [referring to speaker (fr)] changes, she constantly changes those numbers like a terrorist who doesn't want to be caught.'

(131) ye03cd 131

*Wétin* what *yu* 2sg *de* ipfv *chench-chénch* red.cpd-change *nɔ́mba* number *dɛn* pl *só?* like.that 'Why are you constantly changing numbers like that?'

(132) ye03cd 132

*Nó* neg *drink* drink *watá,* water *nó* neg *drink* drink *watá,* water *yu* 2sg *go* pot *siente* feel *ín* 3sg.indp *bád,* extremely *a* 1sg.sbj *tɛ́l* tell *yú.*

2sg.indp

[Addresses speaker (dj) who is drinking tap water] 'Don't drink water, don't drink water, you'll feel it real bad, I tell you.'

(133) ye03cd 133

*A* 1sg.sbj *de* ipfv *tɛ́l* tell *yú,* 2sg.indp *yu* 2sg *go* pot *sí* see *náw* now *yu* 2sg *nó* neg *go* pot *fínis* finish *dán* that *watá.* water 'I'm telling you, you'll see now you won't finish that water.'

(134) ye03cd 134

*A bin wánt intenta dríng watá.*

1sg.sbj pst want intend drink water

'I had wanted to try to drink water.'

(135) dj03cd 135 *Mí* 1sg.indp *nóto* neg.foc *yú.* 2sg.indp 'I'm not you.' (136) ye03cd 136 *Sí,* see *sí,* see *e* 3sg.sbj *fíba* resemble *vɔ́mit,* vomit *yu* 2sg *de* ipfv *sí?* see 'See, see, it [the water] seems like vomit, you see?' (137) ye03cd 137 *A* 1sg.sbj *de* ipfv *tɛ́l* tell *yú,* 2sg.indp *e* 3sg.sbj *fíba* resemble *vɔ́mit* vomit *ínsay* inside 'I'm telling you, it seems like vomit inside his mouth now.' (138) dj03cd 138 *Yu* 2sg *dé* be.loc *bád* bad *ɛ́n.* intj 'You're mean, really,' (139) ye03cd 139 *E* 3sg.sbj *dé* be.loc *lɛkɛ* like *sé* qot *yu* 2sg *de* ipfv *drink* drink *ácido.* acid 'It's as if you're drinking acid.' (140) ye03cd 140 *Háw* how *yu* 2sg *siente* feel *dán* that *watá?* water 'How does that water feel to you?' (141) dj03cd 141 *E* 3sg.sbj *kin* hab *táyt* be.tight *mi* 1sg.poss *bɛlɛ́* belly *náw* now *só,* so *chakrá* destroy *dán* that *wín* wín *mí.* 1sg.indp 'It tightens my stomach like this, shakes up that pap [which is being served for dinner] a bit, you won't defeat me.'

*in* 3sg.poss

*pasta* paste *smɔ́l,* a.bit

*yu* 2sg *nó* neg *go* pot

*mɔ́t* mouth *náw.* now

(142) dj03cd 142 *Yu nó fít.*

2sg neg can 'You can't.'

(143) ye03cd 143

*Hó,* intj *dán* that *mán* man *go* pot *dú* do *vɔ́mit* vomit *tidé,* today *e* 3sg.sbj *nó* neg *go* pot *slíp.* sleep 'Ho, that man (dj) will vomit today, he won't sleep.'


*báy* buy *tomates,* tomatoes *wi* 1pl *desayuna.* breakfast

'Me and Rubi, we teamed up, we bought avocados, we bought sardine, we bought tomatoes (and) we had breakfast.'


*Mɔní* money *nó* neg *dé* be.loc *dɔ́n* down *mɔ́?* more

'Is there no money left down (there) [for your daily expenses]?'

(155) ye03cd 155

*E* 3sg.sbj *nó* neg *dé* be.loc *mɔ́.* more 'None is left.'

(156) fr03cd 156

*Veinte* twenty *mil* thousand *wé* sub *bin* pst *dɔ́n* prf *fínis?* finish

'Twenty thousand (that) have already finished?'

(157) ju03cd 157

*Dɛn-ɔ́l* 3pl.cpd-all *fínis.* finish

'They've already finished.'

(158) fr03cd 158

*Pero* but *apenas* barely *dos* two *semanas* week.pl *wé* sub *yu* 2sg *bin* pst *tɛ́l* tell *mí* 1sg.indp *sé* qot *yu* 2sg *níd* need *a* 1sg.sbj *pút* put *veinte mil dɔ́n.*

twenty thousand down

'But (it's) barely two weeks (ago) that you told me that you needed me to put twenty thousand down for you.'

(159) ye03cd 159

*Na* foc *yú* 2sg.indp *bin* pst *tɛ́l* tell *mí.* 1sg.indp 'It was you who told me.'

(160) ye03cd 160

*Bueno,* alright *una* 2pl *bríng* bring *mí* 1sg.indp *di* def *pépa.* paper 'Alright bring me the paper.'

(161) ye03cd 161

*Porque* because *en* in *dos* two *semanas* weeks *mék* make *veinte* twenty *mil* thousand *fínis.* finish

'Because to make twenty thousand finish in two weeks.'

(162) fr03cd 162

*Háw* how *mɔ́ch* much *una* 2pl *de* ipfv *ték* take *por* by *día?* day

'How much do you take [spend] per day?'

	- *A* 1sg.sbj *go* pot *bríng* bring *di* def *pépa.* paper 'I'll bring the paper.'

*"Mék* sbjv *yu* 2sg *tɛ́l* tell *dɛ́n* 3pl.indp *sé* qot *fɔ* prep *mí,* 1sg.indp *ɛ́ni* every *dé* day *ɛf* if *yu* 2sg *de* ipfv *ték* take *un* one *kilo,* kilo *e dú."*

3sg.sbj do

[Continues quoting Pancho] '"Tell them that for me, every day, if you take one kilo, it's enough."'

(165) dj03cd 165

*'Di* def *dé* day *wé* sub *yu* 2sg *sí* see *bɛ́ta* very.good *chɔ́p* food *yu* 2sg *de* ipfv *chɔ́p* food *fáyn.'* fine

[Quotes his inner speech to Pancho] 'The day [when] you find good food, you eat well.'

(166) dj03cd 166

*Di* def *dé* day *wé* sub *pɛ́pɛ* pepper *nó* neg *dé* be.loc *ínsay* inside *pɔ́t* pot *"a* 1sg.sbj *nó* neg *de* ipfv *chɔ́p* food *dí* this *porquería!"*

mess

[Continues quoting his inner speech to Pancho] 'The day [when] there is no pepper in the pot (you say) "I won't eat this mess."'

(167) ye03cd 167

*A* 1sg.sbj *tɛ́l* tell *yú* 2sg.indp *sé* qot *una* 2pl *de* ipfv *pík* pick *pɛ́pɛ* pepper *aunque* even *na* loc *bús,* forest *e* 3sg.sbj *go* pot *chɔ́p ɔ́l káyn tín ɛf e gɛ́t pɛ́pɛ.*

eat all kind thing if 3sg.sbj get pepper

'I tell you, you could pick pepper like in the forest, he would eat any kind of thing if it has pepper.'

(168) ye03cd 168 *Yɛ́stadé* yesterday *a* 1sg.sbj *kúk* cook *mí* 1sg.indp *sénwe,* foc *al final* finally *a* 1sg.sbj *gó* go *chɔ́p.* eat 'Yesterday I myself cooked (and) then I ate.' (169) ye03cd 169 *Na* foc *Pancho* name *dɛn* pl *bin* pst *de* ipfv *combate* fight *ín* 3sg.indp *dé* there *mɔ́nin* morning *tɛ́n.* time 'It's Pancho they were having an argument with there in the morning.' (170) ye03cd 170 *Sí,* see *na* foc *só* so *mí* 1sg.indp *sɛ́f* foc *kin* hab *dé* be.loc *wé* sub *a* 1sg.sbj *kin* hab *kúk.* cook '(You) see that's how I am, too, when I cook.' (171) ye03cd 171 *Bɔt* but *wé* sub *pɔ́sin* person *de* ipfv *kúk* cook *ín* 3sg.indp *sénwe* foc *"chip".* skt 'But when somebody himself cooks, "chip".' (172) ye03cd 172 *A* 1sg.sbj *bigín* begin *de* ipfv *pica-píca,* red.cpd-cut.up *wi* 1pl *fráy* fry *patata,* potato *wi* 1pl *fráy* fry *plantí.* plantain 'I began to cut up (the trimmings), we fried potatoes, we fried plantain.' (173) fr03cd 173 *Una* 2pl *bin* pst *tɔ́k* talk *wet* with *Pancho?* name 'Did you talk to Pancho?' (174) ye03cd 174 *Wi* 1pl *dɔ́n* prf *tɔ́k* talk *wet=an.* with=3sg.obj 'We've talked to him.' (175) fr03cd 175 *Bɔt* but *wétin* what *a* 1sg.sbj *bin* pst *gí* give *yú* 2sg.indp *dán* that *fax?* fax 'But (then) what did I give you that fax for?' (176) ye03cd 176 *Dán* that *dé* day *a* 1sg.sbj *bít* beat *Pancho,* name *a* 1sg.sbj *bít=an* beat=3sg.obj *a* 1sg.sbj *tɛ́l=an* tell=3sg.obj *sé,* qot *sóté* until *a* 1sg.sbj *tɛ́l=an* tell=3sg.obj *sé* qot *"ɛf* if *yu* 2sg *wánt,* want *a* 1sg.sbj *de* ipfv *alquila* rent *yú* 2sg.indp *taksí,* taxi

*yu* 2sg *sube* go.up *ɔ́p,* up *e* 3sg.sbj *sáful."* be.slow

'That day I beat Pancho, I beat him and told him that, I even told him that "if you want I'll rent you a taxi, you drive up, (and) it's cool (like that)".'

(177) ye03cd 177

*"A* 1sg.sbj *de* ipfv *gí* give *yú* 2sg.indp *quinientos",* fifteen *a* 1sg.sbj *tɛ́l* tell *di* def *taksi-mán* taxi.cpd-man *sé* qot *"arriba a mi casa".*

up to 1sg.poss house

'[I told Pancho] "I'll give you five hundred", I told the taxi driver "up to my house".'

(178) ye03cd 178

*Bɔkú* much *motó* car *dɛn* pl *dé* be.loc *yá* here *só,* so *a* 1sg.sbj *nó* neg *nó* know *sé* qot *Pancho* name *mék* make *lɛkɛ* like *sé* qot *e* 3sg.sbj *de* ipfv *sube* go.up *bihɛ́n* behind *wé* sub *e* 3sg.sbj *baja* go.down *mɔ́.* more

'(Because) there were many cars there, I didn't know that Pancho pretended to go up behind and then went down again.'

### **15.4 Conversation: On sun glasses**

The text below is the transcription of a brief conversation captured on video. It features the discourse participants Boyé (ye), Nenuko (ne), and Lage (ge). The style is informal and jovial. It involves peer-to-peer communication and is decidedly male in its orientation. The text opens with an anecdote by (ye) from his secondary school time (001)–(005). Having heard from a classmate that the President of Equatorial Guinea (Obiang Nguema) could supposedly see people naked through the pair of dark sunglasses that he wore in public (002), (ye) decides to ask his mother to get him such a pair on one of her trips abroad (003).

In what follows, (ne) and (ye) carry the idea further. Of course, the implicit idea is that it would allow them to see the opposite sex naked in the streets. The ensuing conversation is of particular interest because it contains a number of linguistic forms that serve to express emphatic, emotionally involved speech in Pichi. It involves the generous use of emphatic prosodic features such as extra-high pitch, indicated by double acute accents in the text (*bla̋k* 'really dark' (001); *slıp̋* 'sleep' (010), *pen̋* 'pain' (015) and the entire sentence (012)), vowel lengthening (*eyé* 'intj' (008), *ɔ́l* 'all' (012)), and increased volume (sentences (009)–(010), (015), (017)–(018)).

At the segmental level, we find additional defining elements of emphatic speech like interjections (*por Dios* 'by God' (003), *eyé* 'good gracious' (008), the term of address and interjection *cuñado* 'brother(-in-law)' (010), the sentence particle *ó* 'sp' (010)). Further, the conversation features two cognate objects (*swít* 'be tasty' (006) and *dáy* 'die' (016)). The emphatic style of the text also transpires in the use of irrealis modality marking signalled by *go* 'pot' in (009), (011) and (015); *de* 'ipfv' in (010) and factative marking in (012) and (016)–(017). The hypothetical frame provides a backdrop to the boastful self-expression that characterises the conversation from (007) onwards.

The video recording also reveals specific kinetic events that are characteristic for emphatic and self-expressive peer-to-peer communication in Pichi speech culture. For example, (ye) accompanies his interjection in (008) by a movement of the head and torso away from the speaker (ne). Equally, (ne) underlines his comment in (009) by getting up, walking briefly past (ye), and returning to sit on his stool, while laughing intensely. Both motion events are variations of what I assume to be an areal West African kinetic figure employed in certain genres of informal, interactional communication. In this figure, a person abruptly turns aways from the group during a communicative peak (i.e. after the punch line of a joke or an anecdote), describes a circular movement away from the group and joins it again after a brief moment, usually accompanied by laughing.

(1) ye07ga 001

*A* 1sg.sbj *sé,* qot *wán* one *mi* 1sg.poss *kɔ́mpin* friend *nɔ́,* neg *e* 3sg.sbj *bin* pst *de* ipfv *tɛ́l* tell *mí* 1sg.indp *sé/* qot *yu* 2sg *sí* see *Obiang* name *Nguema,* that *dán* time *tɛ́n e* 3sg.sbj *de* ipfv *wɛ́r* wear *sɔn* some *gafas* glasses *dɛn* pl *wé* sub *dɛn* 3pl *bla̋k.*

be.black

'I say one of my friends, right, he was telling me that/ you see Obiang Nguema, that time he was wearing some glasses that were really dark.'

(2) ye07ga 002

*E* 3sg.sbj *sé* qot *dɛn* 3pl *bin* pst *tɛ́l* tell *mí* 1sg.indp *sé,* qot *wé* sub *e* 3sg.sbj *kin* hab *dé* be.loc *na* loc *estadio* stadion *só,* like.that *yu* 2sg *dé* be.loc *na* loc *estadio,* stadion *dɛn* 3pl *de* ipfv *mék* make *Copa de su Excelencia,* President's.Cup *dɛn* 3pl *sé* qot *dán gafa, e de sí ɔ́l mán nékɛd'.*

that glasses 3sg.sbj ipfv see all man be.naked

'He [my friend] said when he's in the stadion like that, (when) you're in the stadion (and) they're doing the President's Cup, they say (with) those glasses, he sees everybody naked.'

(3) ye07ga 003

*Na* foc *ín* 3sg.indp *wán* one *dé* day *a* 1sg.sbj *bin* pst *tɛ́l* tell *wán* one *grand* big *frère* brother *na,* loc *na* loc *mi* 1sg.poss *colegio* college *dé,* there *a* 1sg.sbj *tɛ́l=an* tell=3sg.obj *sé* qot *"mi* 1sg.poss *mamá* mother *de* ipfv *viaja* travel *bɔkú,* much *a* 1sg.sbj *go* pot *tráy* try *mék* sbjv *e* 3sg.sbj *báy* buy *mí* 1sg.indp *dán* that *káyn* kind *gafas* glasses *por* by *Dios".* God 'That's why one day, I told one of my seniors [in French: 'big brother'] in, in my secondary school there, I told him "my mother travels a lot, I'll try to have her buy that kind of glasses for me by God".'

(4) ye07ga 004 *A* 1sg.sbj *wánt* want *de* ipfv *sí* see *ɔ́l* all *mán* man *nékɛd.* be.naked [laughter] 'I want to be seeing everybody naked.' (5) ye07ga 005 *A* 1sg.sbj *wánt* want *dé* be.loc *lɛk* like *Obiang* name *Nguema.* 'I want to be like Obiang Nguema.' (6) ye07ga 006 *Dán* that *torí* story *bin* pst *de* ipfv *swít* be.tasty *mí* 1sg.indp *wán* one *swít.* be.tasty 'I was really enjoying that story.' (7) ne07ga 007 *A* 1sg.sbj *fít* can *sé* qot *if* if *yu* 2sg *consigue* obtain *gafas* glasses *wé,* sub *yu* 2sg *go* pot *wɔ́k* walk *na* loc *ród.* road 'I can tell you if you obtained glasses which, you would walk on the road.' (8) ye07ga 008 *Eyé* intj [éjé::]. 'Good gracious.' (9) ne07ga 009 *Dán* that *gafa* glasses *yu* 2sg *go* pot *slıp̋* sleep *wet=an.* with=3sg.obj 'Those glasses, you would sleep with them.' (10) ye07ga 010 *A* 1sg.sbj *de* ipfv *slíp* sleep *wet=an* with=3sg.obj *cuñado.* brother-in-law [laughter] 'I would sleep with them brother.' (11) ye07ga 011 *A* 1sg.sbj *go* pot *púl=an* remove=3sg.obj *na* loc *mi* 1sg.poss *yáy* eye *sé* qot *wétin?* what 'I would remove them from my eyes for what?' (12) ye07ga 012 *A* 1sg.sbj *wa̋nt* want *dé* be.loc *flipado* turned.on *ɔl* all *áwa,* hour *ɔl* [ɔ::l] all *áwa.* hour 'I would want to be turned on all the time, all the time.'


*A* 1sg.sbj *sé,* qot *na* foc *fɔ* prep *tɔ́k* talk *fɔ* prep *dán* that *(…)*

'I say, one has to talk about that (…)

(15) ne07ga 015

*Yu* 2sg *go* pot *lás* end.up *sí* see *sɔn* some *nékɛd* be.naked *wé* sub *na* foc *ín* 3sg.indp *go* pot *mék* make *mék* sbjv *yu* 2sg *yáy* eye *pen̋ ó.*

pain sp

'You'll end up seeing some (kind of) nakedness that will really make your eyes pain.'

(16) ne07ga 016

*Ey,* intj *dán* that *káyn* kind *spɛ́tikul* glasses *a* 1sg.sbj *dáy* die *dáy.* death

'That kind of glasses, I would really die.'

(17) ne07ga 017

*Wé* sub *yu* 2sg *tɛ́l* tell *húman,* woman *"lúk* look *di* def *wán,* one *yu* 2sg *wánt* want *tɔ́k* talk *wet* with *mí?"* 1sg.indp 'And you would say to women, "look at this one, you (actually) want to talk to me [now that I have seen all of you]?"'

(18) ye07ga 018 *Yú, yú?*

2sg.indp 2sg.indp

'You, you?' [laughter]


### **15.5 Routine procedure: Preparing corn-porridge**

Below follows a procedural text in which Djunais (dj) explains to me (ko) and Lage (ge) how to prepare *ógi* 'corn porridge'. The text features the type of TMA marking characteristic for this narrative genre. Procedural texts may exhibit more than other genres the regular use of factative TMA marking (bare verbs) in order to describe routine procedures and when giving instructions (e.g. (001)–(005)). Likewise the text contains many instances of bare, non-initial verbs typical of clause chaining (e.g. *trowé=an* 'pour=3sg.obj' (040) *bigín* (043) and *pút=an* 'put=3sg.obj' (051).

A second way of expressing (hypothetical) routines appears in (018)–(020). Here the potential mood marker *go* 'pot' is used when (dj) briefly digresses to compare the preparation of *ógi* with that of rice porridge. The text also contains a few instances of unexpressed subjects (*sifta* 'sift' (007), *fít* 'can' (008)) as well as a brief conversation (021)–(034) after which (dj) quickly turns back to describing the cooking:

(1) ko03do 001 *Djunais* name *a* 1sg.sbj *bɛ́g* ask.for *explica* explain *mí.* 1sg.indp

'Djunais, please explain to me [how to prepare maize porridge].'

(2) dj03do 002

*A* 1sg.sbj *ralla* grate *di,* def *di,* def *di* def *maíz.* corn 'I grate(d) the, the corn.

(3) ge03do 003

*Yu* 2sg *ték* take *di* def *maíz* corn *yu* 2sg *hól=an.* hold=3sg.obj 'You take the corn and hold it.'

(4) dj03do 004

*A* 1sg.sbj *ralla* grate *in* 3sg.indp *wet* with *rallador.* grater 'I grate it with a a grater.'


*Sifta,* sift *wé* sub *a* 1sg.sbj *dɔ́n* prf *sifta* sift *ín,* 3sg.indp *e* 3sg.sbj *de* ipfv *lɛ́f* remain *wet* with *di* def *watá.* water 'Sift (it), when I have sifted it, it remains with the water.'

(8) dj03do 008

*Fít* can *sifta* sift *ín* 3sg.indp *sóté* until *tú* two *tɛ́n* time *mék* make *mék* sbjv *dán* that *smɔ́l* small *smɔ́l* rep *watá* water *dɛn* pl *nó* neg *lɛ́f.*

remain

'(You) can sift it up to two times to make that little bit of water not remain.'

(9) dj03do 009

*Sɔn* some *dé* be.loc *yet* yet *sɛ́f* foc *wé* sub *a* 1sg.sbj *nó* neg *mék,* make *entonces* so *dán* that *wán* one *wé* sub *lɛ́f,* remain *una* 2pl *fít* can *kɛ́r=an* carry=3sg.obj *gó* go *aunque* like *ínsay* inside *wán* one *bɔ́tul* bottle *fɔ* prep *wán* one *mineral* mineral *una* 2pl *pút=an, na congelador.*

put=3sg.obj loc fridge

'Some still remains that I didn't make, so that one that remains, you [pl] can put it inside a mineral (water) bottle and put it into the fridge.'

(10) dj03do 010

*Wé* sub *yu* 2sg *de* ipfv *mék=an* make=3sg.obj *na* loc *hós,* house *jɔ́s* just *ték=an* take=3sg.obj *pút=an* put=3sg.obj *na* loc *pɔ́t* pot *aunque wán tasa só.*

like one cup like.that

'When you make it at home, just take it and put it into a pot, approximately one cup or so.'

(11) dj03do 011

*If* if *yu* 2sg *de* ipfv *mék=an* make=3sg.obj *só* like.that *e* 3sg.sbj *go* pot *bɔkú* become.much *pero* but *na* foc *só* like.that *e* 3sg.sbj *gɛ́fɔ* have.to *dé.* be.loc

'If you do it like that it will be(come) much but that's how it has to be.'

(12) dj03do 012 *Wé* sub *a* 1sg.sbj *ralla* grate *ín,* 3sg.indp *a* 1sg.sbj *mék=an,* make=3sg.obj *pút* put *di* def *pɔ́t* pot *na* loc *fáya* fire *wet* with *smɔ́l* small *watá,* water *a* 1sg.sbj *bigín* begin *de* ipfv *pút* put *dán* that *mezcla* mixture *dé* there *sóté* until *e* 3sg.sbj *dé* be.loc *só* like.that *sénwe.* foc 'When I grated it, I make it, (I) put the pot on the fire with a bit of water, I begin to put that mixture in there until it is just like this.' (13) dj03do 013 *Igual* equal *sɛ́f* foc *wet* with *di* def *wán* one *fɔ* prep *rɛ́s.* rice 'The same with the one (made) with rice.' (14) ko03do 014 *So* so *wán* one *dé* be.loc *fɔ* prep *rɛ́s* rice *sɛ́f?* foc 'So there's one made with rice, too?' (15) ko03do 015 *Na* foc *di* def *sén* same *fásin* manner *fɔ* prep *dú=an?* do=3sg.obj 'Is it done the same way?' (16) dj03do 016 *Yu/* 2sg *rɛ́s,* rice *yu* 2sg *de* ipfv *bít=an.* beat=3sg.obj 'You/ as for rice, you beat it.' (17) dj03do 017 *Bít=an* beat=3sg.obj *yu* 2sg *mék=an* make=3sg.obj *só* like.that *sɛ́f.* foc '(You) beat it (and) make just like this.' (18) ko03do 018 *So* so *yu* 2sg *go* pot *bít* beat *di* def *rɛ́s?* rice 'So you beat the rice?' (19) dj03do 019 *Yu* 2sg *go* pot *moja* soak *di* def *rɛ́s* rice *na* loc *watá,* water *fɔ* prep *tidé,* today *tú* two *dé,* day *lɛk* like *háw* how *yu* 2sg *wánt* want *nɔ́,* neg *di* def *dé* day *yu* 2sg *de* ipfv *calcula* calculate *sé* qot *yu* 2sg *wánt* want *chɔ́p=an.* eat=3sg.obj

'You soak it in water, for today [one day], two days, as you want, right, the (number of) days you calculate that you want to eat it.'

(20) dj03do 020

*Yu* 2sg *wánt* want *chɔ́p=an* eat=3sg.obj *tú* two *dé* day *áfta,* then *yu* 2sg *go* pot *mék=an* make=3sg.obj *mék* sbjv *e* 3sg.sbj *dé* be.loc *na watá.*

loc water

'(If) you want to eat it two days afterwards, you make it be in the water [for that time].'

(21) ko03do 021

*Ús=say* q=side *yu* 2sg *lán* learn *fɔ* prep *kúk?* cook 'Where did you learn to cook?'

(22) dj03do 022

*A* 1sg.sbj *gó* go *skúl.* school 'I went to school.'

(23) dj03do 023

*A* 1sg.sbj *gó* go *skúl* school *pero* but *ɔ́l* all *di* def *smɔ́l* small *tín* thing *dɛn* pl *yá* here *só* like.that *na* foc *tín* thing *dɛn* pl *wé* sub *mí* 1sg.indp *de* ipfv *mék=an* make=3sg.obj *na* loc *hós.* house

'I went to school but all the small things here are things that I make at home.'

(24) dj03do 024

*Pero,* but *Sita* name *bin* pst *dé* be.loc *nɔ́,* neg *mamá.* mother

'But Sita was (still) around [alive], right, mother.'

(25) dj03do 025

*Porque* because *na* foc *mí* 1sg.indp *mí* 1sg.indp *de* ipfv *prepara* prepare *ɔ́l* all *tín.* thing 'Because it's me, I [emp] prepare everything.'

(26) ko03do 026

*Yu* 2sg *húman* woman *go* pot *gládin.* be.glad 'Your wife will be happy.'

### (27) ko03do 027

*Na* foc *Djunais* name *go* pot *kúk* cook *fɔ* ass *in* 3sg.poss *fámbul.* family 'It's Djunais who'll cook for his family.'


*E* 3sg.sbj *gó* go *wet* with *in* 3sg.poss *mamá?* mother 'Did he go with his mother?'


*A* 1sg.sbj *sé* qot *dís* this *tín* thing *yá* here *só,* like.that *ɛf* if *di* def *kɔ́n* corn *bin* pst *bɔkú* be.much *lɛk,* like *di* def *watá* water *náw* now *só,* like *di* that *watá/* def water

'I say this thing right here, if the corn was a lot like, the water now, the water/'

(36) ge03do 036

*Dán* that *tín* thing *na* foc *di* def *pasta.* paste 'That is the paste.'

(37) dj03do 037

*Di* def *pasta* paste *yɛ́s.* yes 'The paste, yes.'

(38) dj03do 038

*Na* foc *di* def *tín,* thing *na* foc *ín* 3sg.indp *a* 1sg.sbj *níd* need *fɔ* prep *mék* make *di* def *pasta* paste *porque* because *dɛn* 3pl *de* ipfv *sɛ́l=an simple só.*

sell=3sg.obj simple like.that

'That's it, that's what I need to make the paste because it [the flour] is sold simple like that.'

(39) dj03do 039

*Yu* 2sg *fɔ* prep *trowé* pour *di* def *watá* water *yá* here *só,* like.that *na* foc *háw* how *só* like.that *di* def *tín* thing *bin* pst *fɔ* cond *lɛ́f bɔtɔ́n.*

remain bottom

'You have to pour this water here away, that's how the thing should have remained at the bottom.'

(40) dj03do 040

*Pero* but *e* 3sg.sbj *bin* pst *fɔ* cond *lɛ́f* remain *bɔkú* much *sé* qot *de* of *tal* so *forma* form *que* that *sí,* yes *a* 1sg.sbj *fít* can *ték* take *di* def *wɔtá* water *a* 1sg.sbj *trowé=an,* pour=3sg.obj *lɛ́f* leave *di* def *pán* pan *na* loc *sán,* sun *e* 3sg.sbj *dráy* be.dry *e* 3sg.sbj *lɛ́f lɛkɛ garí náw.*

remain like gari now

'But enough should have remained in such way that, yes, I can take the water and pour it away, leave the pan in the sun, (and then) it dries and remains like gari now.'

(41) ko03do 041

*Na* foc *só* like.that *a* 1sg.sbj *sabí=an* know=3sg.obj *sɛ́f.* foc

'That's how I know it, too.'

(42) dj03do 042

*Lɛk* like *háw* how *dɛn* 3pl *de* ipfv *mék=an* make=3sg.obj *yu* 2sg *de* ipfv *sí* see *na* foc *kɔsta* custard *nɔ́,* neg *wán* one *kɔsta,* custard *sɔntɛ́n* perhaps *na* foc *só* like.that *dɛn* 3pl *de* ipfv *mék* make *ɔ* or *dɛn* 3pl *de* ipfv *pút* put *dán* that *colorante* colourant *ínsay* inside *wé* sub *e de chénch.*

3sg.sbj ipfv change

'The way it's done, you see it's a custard, a (kind of) custard, it may be done like that or that colourant that changes (the colour) is put inside.'

#### (43) dj03do 043

*E* 3sg.sbj *tɔ́n* turn *arena,* sand *dán* that *água* water *dé* be.loc *a* 1sg.sbj *fít* can *ték* take *wán* one *spún,* spoon *a* 1sg.sbj *bigín* begin *de* ipfv *mék=an* make=3sg.obj *normal.* normal

'(When) it turns into sand [farina], that water over there, I can take a spoon (of it) and begin to make normally.'

(44) dj03do 044

*Pero* but *como* because *di* def *arena* sand *tú* too *lílí-lí,* little-rep *kɔ́n* corn *tú* too *smɔ́l* be.small *náw,* now *a* 1sg.sbj *mezcla* mix *ín ɔ́l.*

3sg.indp all

'But since the sand [farina] is too little, the corn is too little now, I mixed all of it [in making the porridge].'

(45) dj03do 045

*Wɛ́n* sub *a* 1sg.sbj *go* pot *kliár=an,* clear=3sg.obj *sɔn,* some *bɔtɔ́n* bottom *mɔ́,* more *e* 3sg.sbj *go* pot *kán* come *gɛ́t* get *dí* this *tín.*

thing

'When I clear it, some, at the bottom again, it will come to have this thing.'

(46) dj03do 046

*Pero* but *dí* this *watá,* water *una* 2pl *nó* neg *trowé=an* pour=3sg.obj *lo* def *que* that *sí,* yes *una* 2pl *sí* see *sé* qot *e* 3sg.sbj *dɔ́n* prf *slíp e dɔ́n slíp fáyn, dí watá dɔ́n baja.*

sleep 3sg.sbj prf lie fine this water prf go.down

'But the water, you [pl] don't pour it away, rather, you [pl] see that it has settled, it has settled nicely, the water has gone down.'

(47) dj03do 047

*Dí* this *tín* thing *dɔ́n* prf *baja,* go.down *wé* sub *di* def *watá* water *una* 2pl *dɔ́n* prf *de* ipfv *sí=an* see=3sg.obj *ɔ́p* up *lɛkɛ* like *sé* qot *na watá normal.*

foc water normal

'The thing [farina] has gone down, and as for the water, you [pl] see it above as if it were normal water.'

(48) dj03do 048

*Ɛf* if *yu* 2sg *ték* take *dán* that *watá* water *dé,* there *yu* 2sg *trowé=an,* pour=3sg.obj *yu* 2sg *trowé=an* pour=3sg.obj *pero* but *mék* sbjv *e,* 3sg.sbj *yu* 2sg *fít* can *ték* take *dán* that *watá* water *yu* 2sg *trowé=an* pour=3sg.obj *yu* 2sg *pút* put *ɔ́da* other *nyú* new *wán* one *ínsay,* inside *dán* that *wán* one *sé* qot *mék* sbjv *e* 3sg.sbj *nó* neg *smɛ́l.* smell

'If you take that water, you pour it away, but let it, you can take that water and you pour it away and you put another new one [water] inside, that is in order for it not to smell.'

(49) dj03do 049

*Porque* because *e* 3sg.sbj *de* ipfv *sigue* follow *wán* one *bád* bad *smɛ́l.* smell

'Because (otherwise) a bad smell follows.'

(50) ge03do 050

*Áfta* then *háw* how *fɔ* prep *mék* make *di* def *ógi?* corn.porridge

'Then how do you make the corn porridge?'

(51) dj03do 051

*Yu* 2sg *fít* can *ték* take *náw,* now *wán,* one *wán* one *smɔ́l* small *kɔ́p* cup *nɔ́,* neg *yu* 2sg *pút=an* put=3sg.obj *na* loc *fáya,* fire *ínsay* inside *di* def *pɔ́t.*

pot

'Now you can take, a, a small cup, right, you put it on the fire, inside the pot.'

(52) dj03do 052

*Dásɔl,* only *wán* one *smɔ́l,* small *wán* one *glás,* glass *yu* 2sg *fúlɔp=an.* fill=3sg.obj 'Only, one small, one glass, you fill it up.'

(53) ge03do 053

*Wán* one *glás* glass *watá.* water 'A glass of water.'

(54) dj03do 054

*Ɛhɛ́,* exactly *wán* one *glás* glass *watá* water *aparte,* apart *yu* 2sg *pút=an* put=3sg.obj *ínsay,* inside *dán* that *wán* one *dé* there *yu* 2sg *fít* can *ték* take *medio* half *fɔ* prep *dán* that *sén* same *glás,* glass *fɔ* prep *dí* this *tín* thing *yá.* here

'Exactly, a glass of water apart, you put it inside, as for that one you can take half in that same glass, in this thing here.'

(55) dj03do 055

*Yu* 2sg *de* ipfv *tɔ́n=an,* turn=3sg.obj *yu* 2sg *nó* neg *fít,* can *yu* 2sg *nó* neg *para* stop *así,* like.this *mék* make *yu* 2sg *tɔ́n=an* turn=3sg.obj *porque* because *bɔtɔ́n* bottom *go* pot *rós.* burn

'You turn it, you can't, you don't stop like that, turn it because the bottom might burn.'

(56) dj03do 056

*E* 3sg.sbj *go* pot *rós* burn *e* 3sg.sbj *go* pot *lɛ́f* remain *lɛkɛ* like *pan-kék.* pan.cpd-cake

'It might burn and become like pancake.'

(57) dj03do 057

*Yu* 2sg *gɛ́fɔ* have.to *de* ipfv *tɔ́n=an.* turn=3sg.obj 'You have to be turning it.'

(58) dj03do 058

*Tɔ́n=an* turn=3sg.obj *tɔ́n=an,* turn=3sg.obj *mék* sbjv *yu* 2sg *nó* neg *para* stop *sóté* until *mék* sbjv *e* 3sg.sbj *tík* be.thick *lɛk* like *háw* how *e* 3sg.sbj *bin* pst *dé* be.loc *só.* so

'Turn it, turn it, don't stop until it is thick, just the way it was (here).'

(59) dj03do 059

```
Pero
but
     ɛf
     if
        di
        def
             tín
             thing
                    kán
                    pfv
                         bɔkú
                         be.much
                                    mɔ́
                                    more
                                           pás
                                           pass
                                                 di
                                                 def
                                                      watá,
                                                      water
                                                               e
                                                               3sg.sbj
                                                                        go
                                                                        pot
                                                                             lɛ́f
                                                                             remain
                                                                                      wán
                                                                                      one
pasta,
paste
       e
       3sg.sbj
                go
                pot
                     lɛ́f
                     remain
                              lɛkɛ,
                              like
                                      pan-kék
                                      pan.cpd-cake
                                                      wán
                                                      one
                                                            tín
                                                            thing
                                                                   só,
                                                                   like.that
                                                                             e
                                                                             3sg.sbj
                                                                                      go
                                                                                      pot
                                                                                           tú
                                                                                           too
tík.
```
become.thick

'But if the thing has become more than the water, a paste will remain, it will become like a kind of pancake, it will become too thick.'

### **15.6 Elicitation: Caused positions**

The text below results from the elicitation of "caused positions" with the help of the corresponding set of video clips that form part of the "Manual for the field season 2001" of the Language and Cognition Group of the Max-Planck-Institute for Psycholinguistics in Nijmegen. Like most elicitations in the corpus, this one was conducted with two (or more) speakers – Lindo (li) and Djunais (dj) – simultaneously. The elicitation shows in an exemplary way the use of the intransitive/inchoative-stative vs. transitive/dynamic variants of Pichi locative verbs. It features numerous other verbs with a spatial meaning component as well (e.g. *pút* 'put' and *dé* 'be.loc').

	- 3pl just ipfv throw=3sg.obj then 3sg.sbj hang

'It's just being thrown, then it hangs.'

(10) li07pe 010 *Nó* neg *nátin* nothing *nó* neg *dé* be.loc *na* loc *di* def *tébul.* table 'Nothing is on the table.' (11) li07pe 011 *Nó* neg *nátin* nothing *nó* neg *dé* be.loc *pantáp=an.* on=3sg.obj 'Nothing is on it.' (12) li07pe 012 *Náw* now *sɔn* some *tín* thing *dɔ́n* prf *dé* be.loc *pan* on *di* def *tébul* table *wé* sub *na* foc *haricots* beans *dɛn.* pl 'Now something is on the table that's beans [<French 'haricots'].' (13) li07pe 013 *Di* def *húman,* woman *e* 3sg.sbj *bríng* bring *di* def *tú* two *bɔ́l* ball *dɛn* pl *pan* on *di* def *tébul.* table 'The woman, she brought the two balls onto the table.' (14) li07pe 014 *E* 3sg.sbj *kán* come *mék* make *di* def *sén* same *tín* thing *nɔ́.* neg 'She did the same thing, right?' (15) dj07pe 015 *Fɔ́s* first *e* 3sg.sbj *fíba* resemble *sé* qot *dɛn* 3pl *bin* pst *dɔ́n* prf *dɔ́n.* be.done 'First, it seemed that they [the beans] were done [cooked].' (16) dj07pe 016 *Náw* now *só* so *dɛn* 3pl *nó* neg *dɔ́n,* be.done *yu* 2sg *sí?* see 'Right now they aren't done, you see?' (17) dj07pe 017 *Náw* now *fɔ́s* first *haricots* beans *dɛn* pl *bin* pst *dɔ́n* prf *kúk.* cook 'Now first, the beans were cooked.' (18) li07pe 018 *A* 1sg.sbj *nó* neg *tínk.* think 'I don't think (so).'

(19) li07pe 019 *Na* foc *di* def *sén* same *tín.* thing 'It's the same thing [in both video clips].' (20) li07pe 020 *E* 3sg.sbj *bríng* bring *haricots* beans *na* loc *hán,* hand *e* 3sg.sbj *lɛ́f* leave *dɛ́n* 3pl.indp *pan* on *di* def 'She brought beans in her hand (and) she left them on the table.' (21) li07pe 021 *Di* def *róp* rope *dé* be.loc *pantáp* on *di* def *tébul.* table 'The rope is on the table.' (22) li07pe 022 *Di* def *róp* rope *nó* neg *fít* can *slíp.* sleep 'The rope can't lie.' (23) li07pe 023 *Na* foc *pɔ́sin* person *de* ipfv *slíp.* sleep 'It's a person that lies down.' (24) ko07pe 024 *E* 3sg.sbj *lé* lie *pantáp* on *di* def *tébul?* table '[So can I say] it's lying on the table?' (25) li07pe 025 *Nó,* neg *e* 3sg.sbj *dé* be.loc *pantáp* on *di* def *tébul.* table 'No, it's on the table.' (26) li07pe 026 *Ɛf* if *e* 3sg.sbj *lé* lie *na* foc *lɛk* like *sé* qot *e* 3sg.sbj *de* ipfv *slíp.* lie 'If it's lying it's like it's lying.' (27) li07pe 027 *Na* foc *pɔ́sin* person *de* ipfv *lé.* lie 'It's a person that lies.'

*tébul.* table

*di* def *tébul.* table

(28) li07pe 028 *Na* foc *kasára.* cassava 'That's cassava.' (29) li07pe 029 *E* 3sg.sbj *bríng* bring *di* def *kasára* cassava *na* loc *in* 3sg.poss *hán.* hand 'She brought the cassava in her hand.' (30) li07pe 030 *Di* def *cartón* carton *dé* be.loc *pantáp* on *di* def *tébul.* table 'The carton is on the table.' (31) li07pe 031 *E* 3sg.sbj *pút* put *di* def *kasára* cassava *ínsay* inside *di* def *cartón* carton *wé* sub *dé* be.loc *pantáp* on 'She put the cassava into the carton that is on the table.' (32) li07pe 032 *Yu* 2sg *nó=an* know=3sg.obj *ɛ́n?* intj 'You know her, right?' (33) li07pe 033 *Yu* 2sg *nó* neg *nó?* know 'You don't know (her)?' (34) li07pe 034 *E* 3sg.sbj *hɛ́ng=an* hang=3sg.obj *míndul* middle *tú* two *stík* tree *dɛn.* pl 'He hung it up between two branches.' (35) li07pe 035 *Hɛ́ng=an* hang=3sg.obj *na* foc *colgar.* hang '"Hɛ́ng=an" is "colgar" [in Spanish].' (36) li07pe 036 *Ɛf* if *e* 3sg.sbj *kwís=an,* squeeze=3sg.obj *e* 3sg.sbj *go* pot *spwɛ́l.* spoil 'If he squeezes it, it will spoil.'


*E* 3sg.sbj *pút* put *dɛ́n* 3pl.indp *pan* on *di* def *tébul.* table 'He put them on the table.'

(41) dj07pe 041

*Tú* two *dífrɛn* different *bɔ́tul* bottle *dɛn* pl *fɔ* prep *vino.* wine 'Two different bottles of wine.'

(42) dj07pe 042

*Di* def *tú* two *bɔ́tul* bottle *dɛn* pl *fít* can *slíp* lie *pantáp* on *tébul* table *sɛ́f.* foc

'The two bottles can (actually) even lie on the table.'

(43) li07pe 043

*E* 3sg.sbj *fínis* finish *bɛ́n* bend *di* def *písis* piece.of.cloth *fáyn.* fine

'He has finished folding the piece of cloth nicely.'

(44) li07pe 044

*E* 3sg.sbj *pút* put *wán* one *smɔ́l* small *stík* tree *nía* near *di* def *stík* tree *wé* sub *e* 3sg.sbj *para.* stand 'She put a small stick next to the tree that's standing.'

(45) li07pe 045

*E* 3sg.sbj *apoya* lean *wán* one *háf* half *stík* tree *fɔ* prep *wán* one *stík.* tree 'She leaned a branch on a tree.'


(55) li07pe 055 *E* 3sg.sbj *de* ipfv *kwís* squeeze *di* def *bɔ́l* ball *fɔ* prep *mék* sbjv *di* def *bɔ́l* ball *fít* can *hɛ́ng* hang *fáyn.* fine 'He's squeezing the ball in order for the ball to be able to be suspended just right.' (56) dj07pe 056 *E* 3sg.sbj *pút* put *di* def *bɔ́tul* bottle *pantáp* on *di* def *tébul* table *pero* but *di* def *mɔ́t* mouth *dé* be.loc *dɔ́n.* down 'He put the bottle on the table but with the mouth down.' (57) li07pe 057 *E* 3sg.sbj *pút* put *di* def *bɔ́tul* bottle *pan* on *di* def *tébul* table *wet* with *di* def *mɔ́t* mouth *dɔ́n* down *ɔ* or *rɔn-sáy.* wrong.cpd-side 'He put the bottle on the table with the mouth down or upside-down.' (58) li07pe 058 *Di* def *písis* piece.of.cloth *hɛ́ng* hang *na* loc *di* def *stík,* tree *bikɔs* because *nó* neg *mán* man *nó* neg *pút=an.* put=3sg.obj 'The piece of cloth is hanging from the tree, because nobody has put it (there).' (59) li07pe 059 *Wí* 1pl.indp *de* ipfv *sí* see *dásɔl* only *sé* qot *di* def *písis* piece.of.cloth *dɔ́n* prf *hɛ́ng.* hang 'We only see that the piece of cloth is now hanging.' (60) li07pe 060 *E* 3sg.sbj *dɔ́n* prf *cuelga* hang *na* loc *di* def *stík.* tree 'It's hanging from the tree.' (61) li07pe 061 *Dís* this *wán* one *dé* be.loc *sé* qot *a* 1sg.sbj *mít* meet *wán* one *bɔ́tul* bottle *wé* sub *e* 3sg.sbj *dé* be.loc *míndul* middle *tú* two *stík* tree *dɛn.* pl 'This one [still image] is like I've come across a bottle that's between two trees.' (62) li07pe 062 *E* 3sg.sbj *bríng* bring *di* def *kasára* cassava *e* 3sg.sbj *pút=an* put=3sg.obj *nía* near *di* def *stík.* tree 'She brought the cassava (and) she put it next to the tree.' (63) li07pe 063 *E* 3sg.sbj *líng=an* lean=3sg.obj *dé.* there 'She abutted it there.'

(64) li07pe 064 *Dán* that *húman* woman *lɔ́n* be.long *bad.* extremely 'That woman is really tall.' (65) li07pe 065 *Chico,* intj *Dios* my *mío.* God 'Wow, my God.' (66) dj07pe 066 *E* 3sg.sbj *bríng* bring *escalera,* ladder *e* 3sg.sbj *líng=an* lean=3sg.obj *nía* near *di* def *stík.* tree 'She brought a ladder, she leaned it against the tree.' (67) dj07pe 067 *E* 3sg.sbj *bríng* bring *trí* three *kasára,* cassava *e* 3sg.sbj *lé* lay *dɛ́n* 3pl.indp *pantáp* on *di* def *tébul.* table 'She brought three cassavas, she laid them on the table.' (68) dj07pe 068 *E* 3sg.sbj *lé=an* lay=3sg.obj *pantáp* on *di* def *tébul.* table 'She laid them on the table.' (69) li07pe 069 *E* 3sg.sbj *fíks* fix *dɛ́n* 3pl.indp *fáyn.* fine 'She arranged them nicely.' (70) li07pe 070 *Dí* this *cartón,* carton *e* 3sg.sbj *mít=an* meet=3sg.obj *yá?* here 'The carton, did she find it [lying] here?' (71) li07pe 071 *E* 3sg.sbj *pút* put *di* def *róp* rope *ínsay* inside *di* def *cartón* carton *wé* sub *e* 3sg.sbj *dé* be.loc *pantáp* on *di* def 'She put the rope inside the carton that's on the table.' (72) li07pe 072 *E* 3sg.sbj *slíp* lay *di* def *bɔ́tul* bottle *pantáp* on *di* def *tébul.* table 'She laid the bottle down on the table.'

*tébul.* table

(73) li07pe 073 *Di* def *bɔ́tul* bottle *lé* lie *náw* now *pantáp* on *di* def *tébul.* table 'The bottle is now lying on the table.' (74) li07pe 074 *E* 3sg.sbj *lé* lay *di* def *bɔ́tul* bottle *pantáp* on *di* def *tébul,* table *e* 3sg.sbj *slíp* lay *di* def *bɔ́tul* bottle *pantáp* on *di* def *tébul.* table 'She laid [*le*] the bottle on the table, she laid [*slíp*] the bottle on the table.' (75) li07pe 075 *Di* def *bɔ́tul* bottle *slíp* lie *pantáp* on *di* def *tébul* table *bikɔs* because *di* def *bɔ́tul* bottle *lé* lie *dé.* there 'The bottle is lying [*slíp*] on the table because the bottle is lying [*lé*] there.' (76) li07pe 076 *Náw* now *e* 3sg.sbj *ték* take *róp,* rope *e* 3sg.sbj *hɛ́ng* hang *di* def *róp* rope *na* loc *di* def *stík* tree *wet* with *kasára.* cassava 'Now he took a rope, he hung the rope from the tree with a cassava.' (77) li07pe 077 *Nóto* neg.foc *só,* so *a* 1sg.sbj *tɔ́k=an* talk=3sg.obj *bád,* bad *Djunais?* name 'Isn't it so, did I say that wrong, Djunais?' (78) dj07pe 078 *E* 3sg.sbj *táy* tie *di* def *kasára* cassava *wet* with *róp* rope *áfta* then *e* 3sg.sbj *hɛ́ng=an.* hang=3sg.obj 'He tied the cassava with a rope, then he hung it up.' (79) li07pe 079 *Dís* this *stík,* tree *e* 3sg.sbj *slíp* lie *pan* on *di* def *tébul.* table 'This stick, it's lying on the table.' (80) dj07pe 080 *Di* def *kasára* cassava *lé* lie *míndul* middle *tú* two *stík.* tree 'The cassava is lying between two trees.' (81) li07pe 081 *Di* def *kasára* cassava *tínap* stand *míndul* middle *tú* two *stík.* tree 'The cassava is standing between two trees.'



*E* 3sg.sbj *pín* stick *di* def *stík* tree *na* loc *grɔ́n.* ground 'She stuck the stick into the ground.'

### (93) dj07pe 093

*Náw* now *e* 3sg.sbj *tínap* stand *na* loc *grɔ́n.* ground

'Now it's standing (upright) in the ground.'

#### (94) dj07pe 094

*Di* def *pɔ́t* pot *náw* now *só* so *e* 3sg.sbj *slíp* lie *pan* on *di* def *tébul.* table 'Right now, the pot is lying on the table.'

#### (95) dj07pe 095

*E* 3sg.sbj *slíp* lay *di* def *escalera* ladder *na* loc *grɔ́n.* ground 'She laid the ladder on the ground.'

## **16 Word lists**

The following two sections contain a Pichi–English–Pichi word list. It includes all of the approximately one thousand Pichi roots contained in the corpus. The vocabulary also features words of Spanish, French, Bube, and Fang origin that occur with a high frequency in the corpus. The English–Pichi section is useful for cross-reference to the Pichi–English section. The latter section contains additional information on variation, usage, morphological structure and the source language of the entry where applicable.

Property items with an entry like *blák* '(be) black' occur as verbs and as attributive adjectives, i.e. *blák motó* 'black car'. Property items with an entry like *sík* 'be sick' are normally only employed as verbs, i.e. *e de sík* 's/he is sick' (?*sík pɔ́sin* 'sick person'). Commonly employed Spanish-derived words are also included. Spanish-derived words are written following Spanish orthographic conventions and without tone-marking. The following abbreviations are used in addition to those listed on page xi:


### **16.1 Pichi–English**

### **A - a**

**a** *pron.* 1sg subject dependent pronoun. **aa**<sup>1</sup> *intj.* expression of insight. **aa**<sup>2</sup> *intj.* expression of impatience. **abuela** *n.* grandmother; term of address. See: gran-má. From: Spanish. **abuelo** *n.* grandfather; term of address. See: gran-pá. From: Spanish. **aburre** *v.* be bored. From: Spanish. **accidente** *n.* accident. From: Spanish. **adopta** *v.* adopt, act as a guardian to a child. From: Spanish. **adɔnkɛ́***link.* even if, no matter if. **adváys** *v.* advise. **aeropuerto** *n.* airport. From: Spanish. **affaire** *n.* affair. From: French. **africana** *pn; adj.* African f. From: Spanish. **africano** *pn; adj.* African m. From: Spanish. **Áfrika** *n.* Africa. **áfta** *adv.* then; afterwards. **áftanun** *n.* afternoon. **aguanta** *v.* bear, persevere. From: Spanish. **ajáajá** *n.* trickery. **áks**<sup>1</sup> *n.* axe. Variant: háks. **áks**<sup>2</sup> *v.* ask. Variant: háks. **almacén** *n.* store. From: Spanish. **alquila** *v.* rent (a taxi, house). From: Spanish.

### **B - b**

**bába** *v.* have a hair cut, shave beard. **bád** *v.* (be) bad. — *adj.* ill, sick. — *adv.* extremely. **badhát** *v.* be mean. Morph: bád-hát. **bádtɛn** *adv.* unfortunately. Morph: bád-tɛ́n. **bañera** *n.* bathtub. From: Spanish.

**altar** *n.* altar. From: Spanish. **amatán** *n.* harmattan. **ambiente** *v.* party, live it up, have fun. Usage: informal. See: pachá. From: Spanish. **an**<sup>1</sup> *link.* and. **=an**<sup>2</sup> *pron.* 3sg enclitic object pronoun. **animal** *n.* animal. From: Spanish. **Annobón** *pn.* the island of Annobón. From: Spanish. **ánsa** *v.* answer. Variant: hánsa. **antí** *n.* aunt; term of address. **ápás** *prep.* after (temporal). **apellido** *n.* surname. From: Spanish. **aráta** *n.* rat. **aráwn** *adv.* around. **arena** *n.* sand. See: sansán. From: Spanish. **ás** *link.* as, because. See: como. **aunque** *link.* even if, although. — *adv.* approximately, like; even. From: Spanish. **avión** *n.* plane. From: Spanish. **áwa** *n.* hour, time. **áy** *intj.* expression of pain or pleasure. From: Spanish. **áyɛn** *v.* iron. Variant: háyɛn.

— *n.* iron.

**báf** *v.* bathe. **bág** *n.* bag. **baja** *v.* go down, take down. See: gó dɔ́n. From: Spanish. **bák** *n.* back (body part). — *v.* give back. **bambáy** *adv.* gradually. **bambú** *n.* bamboo.

**banána** *n.* banana. **banco** *n.* bank. From: Spanish. **banfá** *v.* disfigure, be sickly. **bangá** *n.* palm tree, palm kernel. **baptáys** *v.* baptise. **bat** *adv.* but. See: bɛt; bɔt. **Bata**<sup>1</sup> *pn.* 1) capital of the continental part of Equatorial Guinea, second largest city of the country; the continental part of Equatorial Guinea. 2) a Fang person, the Fang language; any person or language from the continental part of Equatorial Guinea. **bata**<sup>2</sup> *n.* buttocks. **bautiza** *v.* baptise. See: baptáys. From: Spanish. **báy** *v.* buy. **bay** *prep.* by. Usage: rare. **bebí** *n.* 1) baby. 2) sweetheart, babe. **béd** *n.* bed. **bɛ́g** *v.* ask for, implore. **bɛ́lch** *v.* belch. **bɛlɛ́***n.* belly, womb; pregnancy. — *v.* impregnate. **bɛlɛ-húman** *n.* pregnant woman. **gɛ́t bɛlɛ́**be pregnant. **gí bɛlɛ́**impregnate. **púl bɛlɛ́** abort. **bɛlfúl** *v.* be satiated, full. **bɛ́lps** *n.* sweetheart, babe. Usage: informal. **bɛ́lt** *n.* belt, strap. **bɛ́n** *v.* bend (over); fold. **bɛ́r** *v.* bury. **bɛ́rin** *n.* burial. **bɛ́t** 1 *v.* bite. **bɛt**<sup>2</sup> *adv.* but. **bɛ́ta** *v.* (be) very good, excellent. Variant: bɛ́tɛ. **bí** *v.* identity copula. **bía**<sup>1</sup> *n.* beer. **bía**<sup>2</sup> *v.* bear. **bích** *n.* beach. **bíf** *n.* wild animal, meat.

**bifó** *loc.* (in) front (of), before. — *link.* before (temporal). **bíg** *v.* (be) big. **bigdé** *n.* festivity, party; public holiday. Morph: big-dé. **bigín** *v.* begin. — *aux.* ingressive aspect auxiliary. **bihɛ́n** *loc.* rear, behind, after (also temporal). **bikɔs** *link.* because. Variant: bikɔ́s. **bíl** *v.* build. **bilíf** *v.* believe. **bin** *TMA.* past tense marker. **bió** *intj.* expression of surprise. **bis** *v.* repeat. From: Spanish. **bísin** *v.* bother (about); be busy (with). **bísnɛs** *n.* business. **bít** *v.* beat. **bitáwt** *prep.* without. Variant: witáwt. Usage: rare. See: sin. **blák** *v.* (be) black, of dark colour. **blánt** *v.* reside. **bláy** *n.* bag, basket. **bló**<sup>1</sup> *v.* blow (air). **bló**<sup>2</sup> *v.* relax, rest. **bló**<sup>3</sup> *n.* blow. — *v.* give a blow. **bló wín** (to) fart. **bloque** *n.* brick. From: Spanish. **blɔ́d** *n.* blood. **blɔkɔ́s** *n.* male genitals. **blú** *v.* (be) blue. **bocadillo** *n.* bun. From: Spanish. **boîte** *n.* discotheque. From: French. **bókit** *n.* bucket. **bolí** *n.* pen. From: Spanish. **bón** *n.* bone. **bɔbí** *n.* breast. **bɔ́di** *n; prom.* body. **bɔ́ks** *n.* box. **bɔkú** *v; quant.* (be) much. **bɔ́l** *n.* ball. **bɔ́n** *v.* give birth; be born.

**bɔ́s** *v.* burst (open). **bɔt**<sup>1</sup> *adv.* but. **bɔ́t** 2 *v.* hit with the head; (cause to) rebound. **bɔtɔ́n** *loc.* bottom, under(neath). **bɔ́tul** *n.* bottle. **bɔ́y** *n.* boy. **bráket** *v.* coincide, be on par with. **bráyt** *v.* be bright, radiant. **brék** *v.* dawn. **brɛ́d** *n.* bread. **bríng** *v.* bring. **brís** *n.* air; wind. **brók** *v.* break, be broken.

### **C - c**

**cacao** *n.* cocoa. From: Spanish. **café** *n.* coffee. From: Spanish. **Camerún** *pn.* Cameroon. From: Spanish. **camión** *n.* lorry. From: Spanish. **camiseta** *n.* singlet. From: Spanish. **cemento** *n.* cement. From: Spanish. **cielo** *n.* sky. From: Spanish. **clase** *n.* class. From: Spanish. **cobra** *v.* charge, receive. From: Spanish. **cobra mɔní** receive money, salary. **colegio** *n.* college. From: Spanish. **comisaría** *n.* police station. From: Spanish.

**como** *link.* because, since. See: as. From:

**brɔ́da** *n.* brother; term of address. **brɔdalɔ́** *n.* brother-in-law. Morph: brɔ́da-lɔ́. Variant: bralɔ́. **Bubɛ** *pn.* Bube person, the Bube language. **bueno** *adv.* alright. From: Spanish. **búk** *n.* book. **bús** *n.* forest. **butaca** *n.* armchair. From: Spanish. **butú** *v.* stoop over. **bwa** *ideo.* sound of gushing water. **bwɛ́l** *v.* boil. **bya** *ideo.* sound of coughing. **byabyá** *n.* beard.

Spanish. — *adv.* like. **congelador** *n.* fridge. From: Spanish. **Corisco** *pn.* island off the coast of mainland Equatorial Guinea. **cruz** *n.* cross. From: Spanish. **cuñada** *n.* sister-in-law; term of address. **cuaderno** *n.* exercise book. From: Spanish. **cuñado** *n.* brother-in-law; term of address. From: Spanish. **cuenta** *v.* narrate. See: púl torí; tɛ́l. From: Spanish.

### **Ch - ch**


**chico** *n.* boy. — *intj.* expression of surprise, amazement, admiration. From: Spanish. **chíf** *n.* chief, boss; term of address. **chík** *v.* insult; provoke. **chikilís** *v.* tickle. **"chip"** *intj.* 'suck teeth', marker of negative

### **D - d**

**dák** *v.* be dark. **dán** *det.* that (distal demonstrative modifier). See: dát. Variant: dá. **dáns** *v; n.* dance. **dás** *v.* give as a present, for free. **dásɔl** *adv.* then. — *quant.* only. See: ónli. **dát** *det; prom.* that (distal demonstrative modifier and pronominal). **dáy** *v.* die; death. **day-mán, day-pɔ́sin** *n.* corpse. **dé**<sup>1</sup> *adv.* there. **dé**<sup>2</sup> *v.* locative-existential copula. **dé**<sup>3</sup> *n.* day; weather. **de** *TMA.* imperfective aspect marker. — *link.* complementiser-like function with a small number of auxiliary verbs. **desayuna** *v.* have breakfast. From: Spanish. **dɛ́bul** *n.* devil. — *v.* be devilish. **dɛ́n** *pron.* 3pl independent pronoun. **dɛn** *pron.* 3pl dependent prounoun; postnominal plural marker. **dí** *det.* this (proximal demonstrative modifier). Variant: dís. **di** *det.* definite article. **día**<sup>1</sup> *n.* deer. **día**<sup>2</sup> *v.* be expensive. **dífrɛn** *v.* (be) different.

affect. **chɔ́ch** *n.* church. **chɔ́p** *v.* eat. — *n.* food. **chúk** *v.* pierce, stab, sting. **chúk nɛ́f** stab with a knife. **chukchúk** *n.* thorn. Morph: chúk-chúk. **chupete** *n.* nipple. From: Spanish.

**díg** *v.* dig. **dináy** *v.* deny, refuse. **Dios mio** *intj.* my God. From: Spanish. **díp** *v.* be deep. **dís** *det; prom.* this (proximal demonstrative modifer and pronominal). Variant: dí. **discoteca** *n.* discotheque. From: Spanish. **disfruta** *v.* enjoy (oneself). From: Spanish. **dógo** *n.* fool. **dominó** *n.* domino. **domɔ́t** *n.* door. **dote** *n.* dowry. From: Spanish. **dɔ́g** *n.* dog. **dɔ́kta** *n.* doctor. **dɔ́n** <sup>1</sup> *TMA.* perfect tense-aspect marker. **dɔ́n** 2 *loc.* lowerside, down. **dɔ́n** 3 *v.* be done, finished. **dɔtalɔ́** *n.* daughter-in-law. Morph: dɔ́ta-lɔ́. **dɔtí** *v.* be dirty. **dráy** *v.* be dry; be haggard, thin. **dráyva** *n.* driver. **drɛ́b** *v.* drive (a vehicle); chase away. Variant: dráyb. **dríng** *v.* drink. **drɔ́,** *v.* 1) draw, remove; be sticky (of an okro soup). 2) draw (a drawing). **drɔ́ngo** *v.* be dead drunk. **dú** *v.* do, make; be enough. **dúya** *intj.* please.

### **E - e**

**é** *intj.* expression of dismay, empathy. **e** *pron.* 3sg dependent subject pronoun. **éch** *n.* age, age group. **ékié** *intj.* expression of counterexpectation, amazement. From: Fang. **éks** *v.* bounce. **Ela Nguema** *pn.* popular quarter in the eastern part of Malabo.

### **Ɛ - ɛ**

**ɛ́***intj.* attention getter. See: ey. **ɛf** *link.* if, whether. See: if. Variant: ɛfɛ. **ɛhɛ́***intj.* yes (strong agreement). **ɛ́ks** *n.* egg. **gadinɛ́ks** 'garden-eggs' = eggplant.

**ɛ́n** *intj.* channel check.

### **F - f**

**fá** *v.* be far. **fadalɔ́** *n.* father-in-law. Morph: fada-lɔ́. **fála** *v.* accompany, follow. **fám** *n.* farm. **mék fám** (to) farm. **fámbul** *n.* (extended) family, family member. **Fang** *pn.* Fang person, the Fang language. **fásin** *n.* manner, habit. **bád fásin** bad manners, habits. **gúd fásin** good manners, habits. **fát**<sup>1</sup> *v.* (be) fat. **fát**<sup>2</sup> *v.* fart. **fáwe** *v.* (be) far. See: fá. **fáya** *n.* fire. **fáyf** *quant.* five. **fáyn** *adj.* (be) fine, beautiful. **fés** *n.* face. **févɔ** *n.* favour. Variant: fébɔ. **fɛ́da** *n.* feather.

**espia** *v.* spy on, tail somebody. From: Spanish. **estrella** *n.* star. From: Spanish. **ét** *quant.* eight. **évi** *v.* 1) be heavy. 2) be impressive. Variant: (h)ébi. **exactamente** *adv.* exactly. From: Spanish. **extranjero** *n.* foreigner. From: Spanish. **éy** *intj.* attention getter. See: ɛ́.

**ɛ́ni** *quant.* every. **ɛ́nta** *v.* enter. **ɛ́nti** *v.* be empty. **ɛskyús** *v.* excuse. **ɛsplén** *v.* explain.

**fɛ́n** *v.* look for. **fɛ́t** *v.* fight. **fía** *v.* fear. Variant: fíɛ. **fíba**<sup>1</sup> *v.* resemble; seem. **fíba**<sup>2</sup> *n.* fever. **fija** *v.* notice, remark, pay attention. From: Spanish. **fíks** *v.* fix, arrange; repair. **fíl**1 *v.* feel. **fíl hángri** feel hungry. **fíl tɔ́sti** feel thirsty. **fíl slíp** feel sleepy. **fíl**<sup>2</sup> *n.* field. **fin dé semana** *n.* weekend. From: Spanish. **fínga** *n.* finger. **finga-nél** *n.* finger nail, claw. **fínis** *v.* finish. — *aux.* completive aspect auxiliary. **fís** *n.* fish. See: físin. **físin** *v.* fish. See: fís.

**fisin-mán** fisherman. **Fistɔn Nɔmba-wán** *pn.* Pichi-speaking village close to Malabo. **fít** *v.* can, be able. **fityáy** *v.* cheek, offend. **fláy** *v.* 1) fly. 2) rush (to a place). **flíng** *v.* fling, throw with force. **fó** *quant.* four. **fotó** *n.* photo. **fɔ** *prep.* for; due to; by; in order to; at; in; to; from; general associative preposition (expresses Beneficiary, Cause, Manner, Purpose, Location, Source, Goal roles). — *link.* non-finite complementiser, purpose clause introducer. — *TMA.* **fɔdɔ́n** *v.* fall. **fɔgɛ́t** *v.* forget. **fɔ́k** *v.* fuck. Usage: informal. **fɔ́l** 1 *v.* rain. **fɔ́l** <sup>2</sup> *n.* fowl. **mán fɔ́l** *n.* cock. **húman fɔ́l** *n.* hen. **fɔrɛ́va** *adv.* (for) ever. Variant: fɔrɛ́ba. **fɔ́s** 1 *quant; adv.* first.

### **G - g**

**Gabón** *pn.* Gabon. From: Spanish. **gádin** *n.* garden, small farming plot. **gafas** *n.* glasses. From: Spanish. **gál** *n.* girl. See: gáls. Variant: gyal; gɛ́l, gyɛ́l. **gáls** *n.* girl-pl. See: gál. **garí** *n.* gari. **gasolina** *n.* petrol. From: Spanish. **gɛ́l** *n.* girl. See: gál. **gɛ́l frɛ́n** *n.* girlfriend. **gɛ́t** *v.* get, acquire, have, be in permanent possession; existential verb. **gí** *v.* give. Variant: gív. **gitá** *n.* guitar.

**fɔ́s tɛ́n** formerly. **fɔ́s** 2 *v.* force. **fɔséka** *link.* due to. Variant: fɔséko. **frase** *n.* sentence. From: Spanish. **fráy** *v.* fry. **fray-rɛ́s** *n.* fried rice. **fráyde** *n.* friday. **frɛ́n** *n.* friend. Usage: rare. See: kɔ́mpin. **frɛ́s** *v.* (be) fresh. **frɛskól** *n.* mucus. Morph: frɛ́s-kól. **frí** *v; adj.* be free. **fritámbo** *n.* antilope. **frɔn** *prep.* since. **frɔn bɔkú tɛ́n** since long. **frɔn - sóté** *prep.* from - until. **frút** *n.* fruit. **fruta** *n.* fruit. From: Spanish. **fufú** *n.* fufu. **bit-fufú** *n.* pounded fufu. **fúl** *v.* be foolish. — *n.* fool. **fúlis** *v.* (be) foolish. **fúlɔp** *v.* fill, be full. **fút** *n.* foot, leg. **kobofút** *n.* bowlegs, bowlegged person. **fwífwífwí** *ideo.* sound of wind blowing.

**gládin** *v.* be happy. **glás** *n.* glass. **gó** *v.* go, leave. **gó dɔ́n** *v.* go down. **gó ɔ́p** *v.* go up. **go** *TMA.* potential mood marker. **gɔ́d** *n.* God. **gɔ́n** *n.* gun. **gɔ́vna** *n.* governor; government. Variant: gɔ́bna. **graba** *v.* record. From: Spanish. **gran-** *adj.* grand (only used in compounds). **gran-má** *n.* grandma.

**gran-pá** *n.* granpa. **gran-pikín** *n.* grandchild. **gran-mɔ́da** *n.* grandmother. **granát** *n.* groundnut. **grand frère** *n.* big brother. From: French. **gráp** *v.* get up. **grás** *n.* grass. **grén** *n.* grain. **wan-grén pikín** single child. **greví** *n.* gravy. **grí** *v.* agree, allow. **grídi** *v.* be greedy, stingy. Variant: gridin.

### **Gb - gb**

**gbin** *ideo.* sound of a hard and sudden blow.

### **H - h**

**hád** *v.* be hard. **háf** *quant.* half. **háf áwa** half an hour. **hála** *v.* shout. **háma** *v.* hammer. — *n.* hammer. **hambɔ́g** *v.* bother. **hán** *n.* hand, arm. **hángri** *v.* be hungry. Variant: hángrin. **haricot** *n.* beans. From: French. **hásis** *n.* ashes. **hát** *n.* heart. **háw** *inter.* how. **háw mɔ́ch** *inter.* how much. **háyd** *v.* hide. **haydháyd** *adv.* secretely. **hébul** *v.* be capable, able. **héd** *n.* head. **hés** *v.* lift. **hɛ́***intj.* expresses warning, rebuke. **hɛ́lp** *v.* help. Variant: hɛ́p. **hɛ́ng** *v.* hang (onto). **hía**<sup>1</sup> *adv.* here. See: yá. Variant: híɛ.

**grín** *v.* (be) green. **grís** *n.* fat. **gró** *v.* grow. **grɔ́n** *n.* ground; plot. **báy grɔ́n** buy land. **bɛrin-grɔ́n** *n.* burial-ground. **gúd** *v; adj.* (be) good. **Guinea** *pn.* Equatorial Guinea. Variant: Guinea Ecuatorial. From: Spanish. **guineana** *pn.* Equatoguinean f. **guineano** *pn.* Equatoguinean m.

**gbogbogbo** *ideo.* in haste.

**hía**<sup>2</sup> *n.* hair. **hía**<sup>3</sup> *n.* year. Variant: yía. **las-hía** *n.* last year. **nɛ́ks hía** next year. **hía**<sup>4</sup> *v.* hear, understand. Variant: híɛ, yɛ́r. **híb** *v.* heave; throw. **híl** *n.* mountain, hill. **hó** *intj.* expression of mockery and ridicule. **hól**<sup>1</sup> *v.* hold; keep, be in temporary possession of. **hól**<sup>2</sup> *n.* hole. **hól**<sup>3</sup> *quant.* whole. **hóm** *n.* home. **hóm trénin** *n.* good upbringing. **hós** *n.* house, building; home. See: hóm. **hospital** *n.* hospital. From: Spanish. **hɔ́lidé** *n.* holiday. Variant: hɔlidé. **hɔ́n** *n.* horn. **hɔ́nti** *v.* hunt. Variant: hɔ́ntin. **hɔnti-mán** *n.* hunter. **hɔ́ri** *v.* hurry, rush. Variant: hɔ́rin. **hɔ́t** *v.* be hot, warm.

**húk** *v.* hook, hook arms. — *n.* hook.

### **I - i**

**if** *link.* if, whether. See: ɛf. **ín** *pron.* 3sg independent and object pronoun. **in** *pron.* 3sg possessive pronoun. **Ínglis** *pn.* English(-speaking) person, the English language.

### **J - j**

**jakató** *n.* bitter tomato. **jám** *v.* make contact, be in contact. **jél** *n.* jail. **jɛ́lɔs** *v.* envy, be jealous. **jɛ́ntri** *n.* riches. **jɛntri-mán** rich man, rich person. **jís** *aux; adv.* just; egressive aspect

### **K - k**

**kágo** *n.* goods. **kaká** *v.* defecate. — *n.* faeces. **kakara** *ideo.* be restless. **kamúkamú** *ideo.* countermovement of buttocks when walking. **kán**<sup>1</sup> *v.* come. **kán**<sup>2</sup> *v.* (narrative) perfective aspect marker. **kandá** *n.* skin, bark, outer layer. **káp** *n.* cap. Variant: kyáp. **kápinta** *n.* carpenter. **kapú** *v.* fight over; seize. **kasára** *n.* cassava. **katakátá** *ideo.* (be) (hyper-)active, hectic. **katakátá mán** hyper-active, hectic man. **káyn** *prom.* kind.

**húman** *n.* woman. Variant: wúman.

**ínsay** *loc.* inside, amongst; in (temporal). **intenta** *v.* intend, try. From: Spanish. **ísi** *v.* (be) easy. Variant: ísin. **isla** *n.* island. From: Spanish. **ívin** *n.* evening. Variant: íbin. **ívin tɛ́n** *n.* evening.

auxiliary. Variant: jɔ́s. **jɔ́b** *n.* job. **jɔ́ch** *v.* judge. **jɔmba** *n.* lover. **mék jɔmba** make love. **júmp** *v.* jump. **jwɛ́n** *v.* join.

**kéch** *v.* catch. **kechɔ́p** *v.* realise. **kék** *n.* cake. **pan-kék** *n.* pancake. **kenú** *n.* canoe. **kés** *n.* matter. **kɛ́r** *v.* carry; take; last. Variant: kɛ́ri; kyɛ́r(i); kári. **kí** *n.* key. **kíchin** *n.* kitchen. **kík** *v.* kick. **kíl** *v.* kill. **kilo** *n.* kilo. From: Spanish. **kin** *TMA.* habitual aspect marker; abilitive mood marker (marginal). **kíp** *v.* keep. **kip** *ideo.* sound of a dull thud.

**klém** *v.* climb. **klɛ́va** *v.* be clever. **klía** *v.* be clear. **klín** *v.* be clean. **klós** *n.* clothing. **kokó** *n.* cocoa yam. **kokonát** *n.* coconut. Morph: koko-nát. **kól** *v.* be cold. **kóla** *n.* kola nut. **Kómbe** *pn.* Kombe person, the Kombe language. **kót** *n.* coat. **kɔ́ba** *v.* cover. — *n.* cover. **kɔ́f** *v.* cough. **kɔfí** *n.* coffee. **kɔ́l** *v.* call. **kɔ́la** *n.* colour. **kɔ́lech** *n.* college. **kɔmɔ́t** *v.* 1) go out, come out. 2) turn out, become. Variant: kɔ́mɔ́t. — *aux.* egressive aspect auxiliary. **kɔ́mpani** *n.* (commercial) company. **kɔ́mpin** *n.* friend. **kɔmplít** *v.* be complete. **kɔ́n** *n.* corn. **kɔ́na** *n.* corner. — *prep.* next to. **kɔnfyús** *v.* confuse. **kɔ́ngkɔ́ngkɔ́ng** *ideo; intj.* sound of knocking, employed to seek permission to enter. **kɔ́nk** *n.* snail. **kɔ́nt** *v.* count. **kɔ́ntri** *n.* country, village, hometown. **kɔ́p** *n.* cup.

#### **kɔ́s** 1 *v.* cost. **kɔ́s** 2 *v.* insult, offend. **kɔsín** *n.* cousin. **kɔ́sta** *n.* custard. **kɔ́stɔn** *v.* be used to, be accustomed to. **kɔ́t** *v.* cut; have sex. — *n.* cut, gash, (open) wound. **kɔ́tlas** *n.* cutlass. **krás** *v.* crash. **kráy** *v.* cry. **krés** *v.* be crazy, mad. **kres-mán** *n.* madman. **kres-húman** *n.* mad woman. **Krió** *pn.* Krio (Fernandino) person, the Krio language (i.e Pichi as spoken by Fernandinos). **Krió mamá, mamá Krió** *n.* elderly woman of the Fernandino community. **krɔ́b** *v.* scrub; have sex. Usage: informal. **krɔ́s** *v.* cross. — *n.* cross. **kúk** *v.* cook. **kúsɛ́***intj.* expresses encouragement and empathy to person working or for good work done. **kutuku** *ideo.* sound of the heart beating. **kwáráng** *ideo.* sound of round and hard object(s) falling into a receptacle. **kwári** *n.* quarry. **kwáta** *n.* quarter (of a town). **kwɛ́nch** *v.* die (off). **kwɛ́sɔn** *n.* question. Variant: kwɛ́syɔn.

**kwík** *adv.* quickly.

**kɔrɛ́t** *v.* be correct; correct, discipline.

### **Kp - kp**

568

**kpu** *ideo.* sound of impact on a soft matter.

### **L - l**

**láf** *v.* laugh. **lámp** *n.* lamp; electricity. lámp nó dé 'lámp neg be.loc' = 'there's a power-cut' **lán** *v.* learn, teach. **lángwech** *n.* language, one's native language. **lapá** *n.* cloth. **lás** *v.* be last; end up; endure. — *quant.* last. **lata** *n.* can. From: Spanish. **latrín** *n.* latrine; any place of defecating. **láyf** *n.* life. **dé láyf** be alive. **layk** *adv.* See: lɛk. **láyn** *n.* line. **láyt** *v.* 1) light, be lit. 2) be tipsy. — *n.* light. **lé** *v.* lie, lay. **lési** *v.* be lazy. Variant: lésin. **lét** *v.* be late; late (deceased). **lɛ́f** 1 *v.* 1) remain; leave (tr.), stop. 2) resultative copula. 3) allow (causative-permissive verb). **lɛ́f** <sup>2</sup> *n.* left (side). **lɛ́f-hán** *n.* left-hand (side). **lɛ́k** *v.* like. **lɛk** *adv.* like, as (standard marker in equative constructions). Variant: lɛkɛ,

### **M - m**

**má** *n.* 1) mother; term of address. 2) madam; term of address. See: mamá, mamí, mɔ́mi. **mák** *v; n.* mark. **pás mák** pass the limit (expresses superlative degree). **mákit** *n.* market. **Malábo** *pn.* capital of Equatorial Guinea, largest town on Bioko island. From: Bube.

layk. **lɛk háw** *link.* as soon as; the way that. **lɛk sé** *link.* as if. **lɛ́ta** *n.* letter. **líba** *n.* liver. **gɛ́t líba** have guts. **líf**<sup>1</sup> *n.* leaf. **líf**<sup>2</sup> *v.* live; reside. See: láyf. **líka** *n.* alcohol. **lílí** *v.* (be) little, tiny. **lílíbit** *adv.* a (little) bit. Morph: líli-bít. See: lílí. **líng** *v.* lean against, be reclined; lounge. **lónson** *v.* be lonely, miss (a person). **lɔ́k** *v.* close; lock. — *n.* lock. **lɔ́k** *n.* luck. See: lɔ́ki. **bad-lɔ́k** bad luck. **lɔ́ki** *v.* be lucky. Variant: lɔ́kin. **lɔ́n** *v.* (be) long, tall. **lɔ́n tɛ́n** long time ago, since long. **lɔ́s** *v.* loose. — *n.* louse. **Lubá** *pn.* second largest town of Bioko. From: Bube. **Lubá ród** the road to Luba. **lúk** *v.* look.

**malanga** *n.* taro. From: Spanish. **malérya** *n.* malaria. **sík malérya** be sick with malaria. **mamá** *n.* mother; term of address. See: mamí, mɔ́mi, má. — *intj.* expression of surprise or shock. **mamí** *n.* mother; term of address. See: mamá, mɔ́mi, má. **mán** *n.* man; person, human-being. — *intj.* expression of surprise or

amazement. **mánech** *v.* manage. **máred** *v.* marry. **más** *v.* mash; tread on. **mása** *n.* boss, master; term of address. **másta** *n.* boss, master; term of address. See: mása, chíf. **matapenso** *n.* pestle. From: Spanish. **máyn** *v.* mind, care for. Variant: mɛ́n. **mék**<sup>1</sup> *v.* make, do; causative verb. **mék**<sup>2</sup> *link.* subjunctive mood marker, modal complementiser; in order to. **mekés** *v.* hurry. **mélk** *n.* milk. **ményéményé** *ideo.* whine; nag in a childlike fashion. **mɛ́mba** *v.* remember, remind; think of, think about. **mɛ́n** *v.* 1) care for, mind. 2) cure. Variant: máyn. **mɛ́rɛsin** *n.* medicine; sorcery. **mí** *pron.* 1sg independent and object pronoun. **mi** *pron.* 1sg possessive pronoun. **mierda** *n; intj.* shit. From: Spanish. **míks** *v.* mix. **mín** *v.* mean (to), intend. **miná** *n.* penis. **míndul** *loc.* (in the) middle (of), amongst. See: ínsay. **ministerio** *n.* ministry. From: Spanish. **mísis** *n.* Mrs; head of the household (f.),

### **N - n**

**na**<sup>1</sup> *part.* focus marker; identity copula (affirmative). **na**<sup>2</sup> *prep.* general locative preposition. **nadó** *loc.* outside. **nák** *v.* hit; make love to; gulp down a drink. **nátin** *prom.* nothing. **náw** *adv.* now.

matron; term of address. **mít** *v.* meet; find. **mítɔp** *v.* meet. **Moka** *pn.* village in the interior of Bioko. From: Bube. **mosquitero** *n.* mosquito net. From: Spanish. **mosquito** *n.* mosquito. From: Spanish. **motó** *n.* car, vehicle. Variant: moto. **móvil** *n.* mobile phone. From: Spanish. **mɔ́** *adv.* more (comparative particle); again. — *v.* be more. **mɔ́ch** *adv.* much. **tú mɔ́ch** *adv.* too much. See: tú. **mɔ́-ɛn-mɔ́** *adv.* more and more. **mɔ́mi** *n.* mother; term of address. See: mamá, mɔ́mi, ma. **mɔ́nde** *n.* monday. **mɔní** *n.* money. **mɔ́nin** *n.* morning. **gúd mɔ́nin** good morning. **mɔ́nin tɛ́n** *n.* morning. **mɔnkí** *n.* monkey. **mɔs** *TMA.* must (obligative mood marker). **mɔ́t** *n.* mouth. **múf** *v.* move. **mulata** *n.* African European f. From: Spanish. **mulato** *n.* African European m. From: Spanish. **mún** *n.* moon, month.

**náw só** *adv.* right now. **nawá** *intj.* expression of exasperation and (self) pity. **náyn** *quant.* nine. **náys** *v.* (be) nice. **nékɛd** *v.* be naked. **nél** *n.* nail. **ném** *n.* name.

**nésɔn** *n.* nation(ality), (a) people. Variant: nésyɔn. **nétif** *v.* (be) customary. **nɛ́a** *TMA.* negative perfect tense-aspect marker. Morph: nɛ́va. Variant: nɛ́ba, nɔ́ba. **nɛ́f** *n.* knife. **nɛ́k** *n.* neck. **nɛ́ks** *quant.* next. **nɛ́t** *n.* night. **na nɛ́t** *adv.*at night, in the night. **ní** *n.* knee. **ni** *link.* neither. — *adv.* even. From: Spanish. **nía** *loc.* near, next to. **níd** *v.* need. **nídul** *n.* needle. **Nigeria** *pn.* Nigeria. From: Spanish. **nít** *n.* nit. **nó**<sup>1</sup> *part.* negative particle.

### **O - o**

**o** *link.* or. See: ɔ̀. From: Spanish. **ó** *intj.* sentence particle. **ógi** *n.* corn porridge. **ól** *v.* (be) old. **ónli** *quant.* only. See: dásɔl. **ópin** *v.* be open. Variant: hópin.

### **Ɔ - ɔ**

**ɔ** *link.* or. **ɔ́da** *quant.* other. **ɔf** *prep.* of. Usage: rare. **ɔ́fis** *n.* office. **ɔ́l** *quant.* all, every. **ɔndastán** *v.* understand. **ɔnkúl** *n.* uncle; term of address. **nó**<sup>2</sup> *v.* know. **normal** *adj.* normal. From: Spanish. **nós** *n.* nose. **nóto** *part.* focus marker (negative); identity copula (negative). **novia** *n.* girlfriend. From: Spanish. **novio** *n.* boyfriend. From: Spanish. **nɔ́** *part.* negative particle; question tag. See: nó. **nɔ́ba** *TMA.* negative perfect tense-aspect marker. See: nɛ́a. **nɔ́mba** *n.* number. **nɔ́ys** *n.* noise. **nube** *n.* cloud. From: Spanish. **nyangá** *v.* put on airs, coquet. **nyɔ́ní** *n.* ant. **nyú** *v.* (be) new. **Nyúmbili** *pn.* most populous and densely-populated quarter of Malabo. **nyús** *n.* news.

**opinyáy** *v.* be enlightened, cultivated. Morph: ópin-yáy. **opó** *n.* vagina. **óva** *adv.* over, excessively. Variant: óba. — *v.* be over; be excessive. **ova-hɔ́t** *v.* be overhot, overheat.

**ɔntɔ́p** *loc.* top, on. **ɔ́p** *loc.* up(perside), above. **ɔráyt** *adv.* alright. **ɔspítul** *n.* hospital. **ɔ́t** *v.* extinguish. **ɔ́yl** *n.* oil.

### **P - p**

**pá** *n.* father; term of address for man of one's father's generation. See: papá. **paciente** *n.* patient. From: Spanish. **pachá** *v.* party, live it up, have fun. Usage: informal. See: ambiente. From: Spanish. **Pagalú** *pn.* Annobón island, person from Annobón, the language of Annobón. **pák** *v.* pack; fill, be full. **pála** *n.* parlour; sitting room. **paludismo** *n.* malaria. From: Spanish. **pamáyn** *n.* oil. **pambɔ́d** *n.* bird. **pán** *n.* pan. **dɔtí pán** rubbish bin. **pan** *prep.* on. **pantáp** *loc.* top; on. Variant: pantɔ́p. **Panyá** *pn.* Spain, Spaniard, the Spanish language. **papa** *n.* potato. From: Spanish. **papá** *n.* father; term of address. See: papí. — *intj.* expression of surprise or shock. **papá gɔ́d** *n.* father God. — *intj.* expression of exasperation, self pity, address of God. **papí** *n.* father; term of address. See: papá. **para** *v.* stop. From: Spanish. **paraláys** *n.* paralyse. **pareja** *n.* couple. From: Spanish. **pás** *v.* 1) pass (by); standard marker in comparative constructions. 2) happen. **pát** *n.* part; place. **patata** *n.* potato. From: Spanish. **páwa** *n.* power, strength; potency. **páwda** *n.* powder. **pé** *v.* pay. **peluqueria** *n.* hairdresser. From: Spanish. **pén** *v.* pain. **pént** *v.* paint. **pépa** *n.* paper. **petróleo** *n.* oil (crude ~). From: Spanish.

**pɛ́pɛ** *n.* pepper. **pía**<sup>1</sup> *n.* avocado. **pía**<sup>2</sup> *n.* pair. **pía sús** *n.* pair of shoes. **Píchi** *pn.* the Pichi language. Variant: Pichínglis. **pík** *v.* pick (up). **pikín** *n.* child; member of a specified group. **Guinea pikín** *n.* person of Equatoguinean stock; typically Equatoguinean person. **píl** *v.* peel. **pín** *v.* stick in(to). **pínch** *v.* pinch. Variant: spínch. **pipí** *v.* urinate. — *n.* urine. **pípul** *n.* people. **písis** *n.* piece of cloth, rag. **piso** *n.* storey. From: Spanish. **písul** *n.* pistol. **plába** *n.* trouble, problem, matter. **plánt** *v.* 1) plant. 2) plait (hair). **plantí** *n.* plantain. **bwɛl-plantí** *n.* boiled plantain. **fray-plantí** *n.* fried plantain. **grin-plantí** *n.* green, unripe plantain. **plástik** *n.* plastic. **plataforma** *n.* oil rig. From: Spanish. **plaza** *n.* square. From: Spanish. **plé** *v.* play. — *n.* game. **plés** *n.* place. **plét** *n.* plate. **plɛ́nk** *n.* board, plank. **plɛ́nte** *v; quant.* (be) plenty, a lot. **plís** *intj.* please. **pó** *v.* (be) poor. **policía** *n.* police. From: Spanish. **polís** *n.* police. **por Dios** *intj.* by God. From: Spanish.

**porcería** *n.* mess. From: Spanish. **porque** *link.* because. See: bikɔs. From: Spanish. **pɔ́mp** *v; n.* pump. **pɔ́nis** *v.* punish. **pɔ́sin** *n.* person, human-being. Variant: pɛ́sin. **pɔ́t** *n.* pot. **pɔtɔpɔ́tɔ́** *n; ideo.* mud; any mushy substance. **práwd** *v.* (be) boastful, proud. **prík** *n.* penis. See: miná. **prima** *n.* cousin f. From: Spanish. **primo** *n.* cousin m. From: Spanish. **príng** *ideo.* sound of ringing. **problema** *n.* problem. From: Spanish. **profesor** *n.* teacher. Variant: profe. From:

### **R - r**

**rás** *n.* arse. Usage: informal. **kaka-rás** *n.* arse (also used as an insult). **ráwn** *v.* surround. **ralla** *v.* grate. From: Spanish. **ráyt**<sup>1</sup> *v.* write. **ráyt**<sup>2</sup> *n.* right. **gɛ́t ráyt** be right. *n.* **rayt-hán** *n.* right hand (side). **Rebola** *pn.* town on the east coast of Bioko. **refyús** *v.* refuse. **relámpago** *n.* lightning. From: Spanish. **rén** *n.* rain. **restaurante** *n.* restaurant. From: Spanish. **reunión** *n.* meeting. From: Spanish. **rɛ́d** *v.* (be) red, orange. **rɛdí** *v.* prepare. **rɛ́p** *v.* be ripe. **rɛ́s** 1 *v.* rest.

Spanish. **prɔ́blɛm** *n.* problem. **prɔ́mis** *v.* promise. **prɔpatí** *n.* property. **prúf** *v.* prove; disclose. **pueblo** *n.* village. From: Spanish. **puerto** *n.* harbour. From: Spanish. **pues** *link.* so. From: Spanish. **púl** *v.* pull, remove. **púl bɛlɛ́** abort. **púl brís** breathe, inhale. **púl fotó** take a picture. **púl torí** tell a story, converse. **pús** *v.* push. **pút** *v.* put. **pyɔ́** *v.* (be) pure.

**rɛ́s** <sup>2</sup> *n.* rice. **rɛspɛ́t** *v.* respect, admire. **Riaba** *pn.* town on the east coast of Bioko. **rích** *v.* arrive; reach; (to) equal. **ríng** *n.* ring. **ríva** *n.* river. Variant: ríba. **ród** *n.* road. **róp** *n.* rope. **rós** *v.* roast, burn. **rɔ́b** *v.* rub. **rɔ́n** 1 *v.* (be) wrong. **rɔn-sáy** *n.* reverse, upside down, inside out. **rɔ́n** 2 *v.* run. **rɔ́tin** *v.* rot. **rubio** *adj.* light. From: Spanish. **rúm** *n.* room. **baf-rúm** *n.* bath-room.

### **S - s**

**sá** *n.* sir; term of address. **sabí** *v.* know; know how to. **saco** *n.* sack. From: Spanish. **sadín** *n.* sardine. **sáf** *v.* (be) soft. **sáful** *v.* be careful; slow; smooth, cool. **sala** *n.* hall. From: Spanish. **saldo** *n.* units (mobile phone ~). From: Spanish. **gɛ́t saldo** have units (on the mobile phone). **salút** *v.* greet. **sán** *n.* sun. **sán tɛ́n** *n.* (after)noon. **sansán** *n.* sand, soil. **sastre** *n.* tailor. From: Spanish. **sátidé** *n.* saturday. **sáwa** *v.* shower. **sáy** *n.* side; place. **sáyn** *v.* sign. **sé** *v.* say. — *link.* quotative marker and clause linker with a large range of functions. **séb** *v.* share, divide. Variant: syéb. **sék** *v.* shake; dance. Variant: syék. **sém** *v.* be ashamed; put to shame. Variant: syém. **sén** *prom.* same. Variant: syén. **sénwe** *part.* focus marker; also, too, even. Morph: sén-wé. Variant: syénwe. **serie** *n.* series, soap opera. From: Spanish. **sɛ́f** *prom.* self. — *part.* focus marker; even, too, also. **sɛkɔ́n** *quant.* second. Variant: sɛ́kɔn. **sɛ́l** *v.* sell. **sɛ́n** *v.* send, throw with aim. **sɛ́ns** *n.* brain, mind. **gɛ́t sɛ́ns** be intelligent. **sɛ́nt** *n.* scent. **sɛ́ven** *quant.* seven. Variant: sɛ́ben. **sí** *v.* see; find.

**síd** *n.* seed. **sidɔ́n** *v.* sit (down); reside. Variant: sídɔ́n. **sífta** *v.* sift. — *n.* sieve. **sigá** *n.* cigarette. **sigue** *v.* continue, follow. From: Spanish. — *aux.* continuative aspect auxiliary. **sík** *v.* be sick. — *n.* sickness. **síkrit** *n.* secret. **síks** *quant.* six. **sin** *prep.* without. See: bitáwt. From: Spanish. **síng** *v.* sing. — *n.* song. **sinimá** *n.* movie; cinema. **síns** *prep.* since (temporal). **síryɔs** *v.* (be) serious. **sísin** *n.* season. **ren-sísin** *n.* rainy season. **dray-sísin** *n.* dry season. **sísɔs** *n.* scissors. **sísta** *n.* sister; term of address. **sistalɔ́** *n.* sister-in-law. Morph: sísta-lɔ́. Variant: sistɛlɔ́. **skía** *v.* be scared, scare. **skín** *n.* body. **smɔlskín** *v.* be thin, emaciated. Morph: smɔ́l-skín. **skrách** *v.* scratch. Variant: krách. **skúl** *n.* school. **gí skúl** give classes. **gó skúl** go to school. **skwís** *v.* squeeze. Variant: kwís. **slák** *v.* be loose, loosen. **sláp** *v.* slap. **slím** *v.* (be) slim. **slíp** *v.* lie, lay; sleep; have sex. **slipás** *n.* slipper. **sló** *adj.* (be) slow. **smát** *v.* (be) fast.

**smɛ́l** *v.* smell. — *n.* smell. **smók** *v; n.* smoke. **smɔ́l** *v.* (be) small. — *quant.* a bit, few. **smɔ́ltɛn** *adv.* shortly after; nearly. Morph: smɔ́l-tɛ́n. **snék** *n.* snake. **só**1 *v.* show. **só**<sup>2</sup> *v.* sew. **só**<sup>3</sup> *adv.* like that, like this. **so** *adv.* so, thus, hence. **sobrina** *n.* niece. From: Spanish. **sobrino** *n.* nephew. From: Spanish. **sók** *v.* be wet. **soldado** *n.* soldier. From: Spanish. **solwatá** *n.* the sea, saltwater. Morph: sɔ́l-watá. **sólya** *n.* soldier. **sonido** *n.* sound. From: Spanish. **sosáyti** *n.* association, confraternity. **sósó** *quant.* only; abundant. **sóté** *prep.* until (temporal, locative). — *adv.* for a long time; even, extremely. **sɔ́fa** *v.* suffer. **sɔfút** *n.* wound, injury. Morph: sɔ́-fút. **sɔ́l** *n.* salt. **sɔn** *quant.* some; indefinite determiner. **sɔ́nde** *n.* sunday. **sɔnilɔ́** *n.* son-in-law. Morph: sɔni-lɔ́. **sɔntɛ́n** *adv.* perhaps, maybe. Morph: sɔn-tɛ́n. **sɔ́p** *n.* shop. Variant: syɔ́p. **sɔ́ri** *v.* feel sorry. Variant: sɔ́rin. **sɔ́t** 1 *v.* (be) short. Variant: syɔ́t. **sɔ́t** <sup>2</sup> *n.* shirt. Variant: syɔ́t. **spɛ́n** *v.* spend. **spɛ́tikul** *n.* glasses. **spírit** *n.* spirit. **spít** *v.* spit. — *n.* spit. **spót** *v.* be stylish, dressed up; exhibit the

self-confident demeanor of a well-dressed person. **spotinbɔ́y** *n.* well dressed, stylish guy. **spún** *n.* spoon. **spwɛ́l** *v.* spoil; use up. **stá** *n.* star. Usage: rare. See: estrella. **stát** *v.* start. **stáwt** *v.* (be) corpulent. **stáyl** *n.* style, manner. **ús=stáyl** *inter.* how. **sté** *v.* stay; last (a long time). **stík** *n.* tree, branch, stick, wood. Variant: tík. **stíl** *adv.* still. — *aux.* continuative aspect auxiliary. **stíma** *n.* ship. **stón** *n.* 1) stone. 2) testicle. — *v.* throw stones at. Variant: tón. **stɔ́p** *v.* stop. **strét** *v.* (be) straight; respectable; sincere. Variant: trét. **strít** *n.* street. Variant: trít. **strɔ́n** *v.* (be) strong; (be) hard; (be) difficult; (be) profound. See: tránga. Variant: trɔ́n. **styú** *n.* stew. **súb** *v.* shove, push. **sube** *v.* go up, take up. See: gó ɔ́p. From: Spanish. **súga** *n.* sugar. **súkútúpampa** *ideo.* in a cheap and mean fashion. **súp** *n.* soup. **bangá súp** *n.* palm nut soup. **granát súp** *n.* groundnut súp. **sús** *n.* shoe. **sút** *v.* shoot. **swɛ́la** *v.* swallow. **swɛ́t** *v.* sweat. — *n.* sweat. **swín** *v.* swim. **swít** *v.* 1) (be) tasty. 2) (be) sweet.

### **T - t**

**taksí** *n.* taxi. **tal** *adv.* so. From: Spanish. **tamátis** *n.* tomato. **tan** *adv.* as. From: Spanish. **tánap** *v.* stand (up). See: tínap. Variant: stánap. **tarjeta** *n.* card. From: Spanish. **táwɛl** *n.* towel. **táy** *v.* tie. **táya** *v.* be tired. **táyt** *v.* be tight, tighten. **tébul** *n.* table. **ték** *v.* take. See: kyɛ́r. **teléfono** *n.* telephone. From: Spanish. **televisión** *n.* television. From: Spanish. **tɛ́l** *v.* tell, narrate. **tɛ́n** *n.* time. **tɛ́nki** *intj.* thanks. **tia** *n.* aunt; term of address. See: antí. From: Spanish. **tích** *v.* teach. **tícha** *n.* teacher. **tidé** *n.* today. Variant: tudé. **tíf** *v.* steal. **tif-mán** *n.* thief. **tifoidea** *n.* typhoid fever. From: Spanish. **sík tifoidea** be sick with typhoid fever. **tík** *v.* (be) thick. **tík** *ideo.* cracking sound. **tín** *n.* thing. **tinada** *n.* thunderstorm. **tínap** *v.* stand (up), put into an upright position. See: tánap. Variant: stínap. **tínk** *v.* think. **tio** *n.* uncle; term of address. See: ɔnkúl. From: Spanish. **tít** *n.* tooth. **tití** *n.* girl. **to** *prep.* to (locative preposition and complementiser). Usage: rare. **tomate** *n.* tomato. From: Spanish.

**topé** *n.* palm-wine. **torí** *n.* story. **tót** *v.* carry. **totó** *n.* vagina. See: opó. **tɔ́ch** *v.* touch. **tɔ́k** *v.* talk. — *n.* word, speech, language. **kɔntri-tɔ́k** *n.* the language of one's home town, native language. **tɔ́n** <sup>1</sup> *n.* town. **tɔ́n** 2 *v.* 1) turn. 2) become. **tɔ́n bák** *v.* return, come back. **tɔ́ng** *n.* tongue; language. **Bata tɔ́ng** *n.* the Fang language; any language of the continental part of Equatorial Guinea. **tɔ́p** *n.* top. **tɔ́sti** *v.* be thirsty. Variant: tɔ́stin. **tradicional** *adj.* traditional. From: Spanish. **tránga** *v.* (be) strong; (be) hard; (be) difficult. See: strɔ́n. **transporte** *n.* transport. From: Spanish. **trávul** *v.* travel. **tráy** *v.* try; make an effort. **trén** *v.* train; educate, bring up. See: trénin. **trénin** *n.* training; upbringing. See: trén. **trí** *quant.* three. **trímbul** *v.* tremble. Variant: strímbul. **tripas** *n.* intestines. From: Spanish. **trót** *n.* throat. **trowé** *v.* 1) throw (away). 2) pour. **trɔ́bul** *n.* trouble, hardship. **trɔsís** *n.* trousers. **trú** *v.* (be) true. **tú**1 *quant.* two. **tú**<sup>2</sup> *adv.* too (much). **tudé** *n.* See: tidé. **tumára** *n.* tomorrow, the next day. Variant: tumɔ́ro. **tumbú** *n.* worm.

**tumɔ́ro** *n.* See: tumára. **tún** *v.* 1) tune. 2) persuade, chat up.

### **U - u**

**údat** *inter.* who. **una** *pron.* 2pl pronoun (emphatic and non-emphatic). Variant: unu. **ús=** *inter.* which; clitic interrogative particle in wh- question words. **ús=áwa** *inter.* what time, when.

**V - v**

**vájin** *n.* virgin(ity). Variant: bájin. **vecina** *n.* neighbour f. From: Spanish. **vecino** *n.* neighbour m. From: Spanish. **verdura** *n.* vegetables. From: Spanish. **vɛ́ks** *v.* be angry. **viaja** *v.* travel. From: Spanish.

### **W - w**

**wách** *v.* watch. **wahála** *n.* trouble; problem. **wáka** *v.* walk. Variant: wɔ́k. **wán** *quant.* 1) one; a (indefinite determiner). 2) only, alone, single-handedly. 3) approximately. — *prom.* one (noun substitute). **-wán** *aff.* adverbialising suffix. **wánda** *v.* wonder. Variant: wɔ́nda. **wán-ɛn-háf** *quant.* one and a half. **wáns** *link.* once. **wánt** *v.* want. — *aux.* prospective aspect auxiliary. Variant: wɔ́nt. **wántɛn** *adv.* suddenly; at once. **wás** *v.* wash. **watá** *n.* water. Variant: wɔtá. **hɔt-watá** *n.* hot, warm water. **kol-watá** *n.* cold, cool water.

**ús=káyn** *inter.* which (kind of). **ús=pɔ́sin** *inter.* who. **ús=say** *inter.* where. **ús=stáyl** *inter.* how. **ús=tɛ́n** *inter.* when. **ús=tín** *inter.* what.

**tyúsde** *n.* tuesday.

**vílech** *n.* village. **vino** *n.* wine. From: Spanish. **visít** *v.* visit. **vɔ́mit** *v.* vomit. **vuelta** *n.* round. From: Spanish. **gí wán vuelta** take a walk.

**wáyf** *n.* wife. **wayó** *n.* cunning. — *v.* be cunning. **wáyp** *v.* wipe. **wáyt** *v.* (be) white, light, clear. **wé**<sup>1</sup> *link.* introduces relative clauses; coordinate clauses; time clauses; complement clauses (marginal). Variant: wɛ́n. **wé**<sup>2</sup> *n.* way. **wék** *v.* wake (up). **wékɔp** *v.* wake (up). **wés** *n.* buttocks, genital area; bottom part of an entity. **wét** *v.* wait. **wet** *prep.* with (expresses comitative, instrument, cause, manner, circumstance roles); and. **wétin** *inter.* what.

**wéwé** *ideo.* sound of crying and wailing. **wɛ́l** *v.* be well. **wɛ́lkɔm** *intj.* welcome. **wɛ́nsde** *n.* wednesday. **wɛ́r** *v.* wear; dress up. **wí** *pron.* 1pl independent pronoun. **wi** *pron.* 1pl dependent pronoun. **wích**<sup>1</sup> *inter.* which. **wích**<sup>2</sup> *v.* bewitch. — *n.* witch, sorcerer; sorcery. **wík** *n.* week. **wíkɛd** *v.* (be) wicked. **wín** *v.* wín; defeat. **wínda** *n.* window. Variant: wíndo.

### **Y - y**

**yá** *adv.* here. **yá só** *adv.* right here. **yabás** *n.* onion. **yandá** *adv.* yonder, over there. Variant: nyandá. **yáy** *n.* eye. **yé** *intj.* response to a call. **yés** *n.* ear. **yét** *adv.* yet, still.

**wók** *v; n.* work. **wós** *v.* be very bad, worse. **wówó** *ideo.* of crying. **wɔ́d** *n.* word. **wɔ́k** *v.* walk. Usage: rare. See: wáka. **wɔ́l** *n.* world. **wɔ́ri** *v.* worry. Variant: wɔ́rin. **wɔtá** *n.* water. Variant: watá. **wɔwɔ́***ideo; v.* (be) ugly; (be) messy, in disorder. **wúd** *n.* wood. **faya-wúd** *n.* fire wood. **wuruwúrú** *ideo; n.* disorderliness; confusion.

**yɛ́r** *v.* hear, understand. See: hía. **yɛ́s** *intj.* yes. Variant: yɛ́. **yɛ́stadé** *n.* yesterday. **yón** *prom.* own, noun substitute. **yú** *pron.* 2sg independent pronoun. **yu** *pron.* 2sg dependent pronoun. **yún** *v.* (be) young. Variant: nyún. **yús** *v.* use. Variant: nyús.

### **16.2 English–Pichi**

### **A - a**

**a** *quant.* sɔn; *quant.* wán (1). **abilitive mood marker** *TMA.* kin. **able** *v.* fít. **abort** *v.* púl bɛlɛ́, see: púl. **above** *loc.* ɔ́p. **abundant** *quant.* sósó. **accident** *n.* accidente. **accompany** *v.* fála. **acquire** *v.* gɛ́t. **active (hyper~)** *ideo.* katakátá. **admire** *v.* rɛspɛ́t. **adopt** *v.* adopta. **adverbialising affix** *aff.* -wán. **advise** *v.* adváys. **affair** *n.* affaire. From French. **Africa** *n.* Áfrika. **African European f.** *n.* mulata. **African European m.** *n.* mulato. **African f.** *pn; adj.* africana. **African m.** *pn; adj.* africano. **after** *loc.* bihɛ́n; *prep.* ápás. **afternoon** *n.* áftanun; *n.* sán tɛ́n, see: sán. **afterwards** *adv.* áfta. **again** *adv.* mɔ́. **age** *n.* éch. **agree** *v.* grí. **air** *n.* brís. **airport** *n.* aeropuerto. **airs (put on ~)** *v.* nyangá. **alcohol** *n.* líka. **alive (be ~)** *v.* dé láyf, see: láyf. **all** *quant.* ɔ́l. **allow** *v.* grí; *v.* lɛ́f 1 (1). **alone** *quant.* wán (2). **alright** *adv.* bueno;

*adv.* ɔráyt. **also** *part.* sɛ́f. **altar** *n.* altar. **although** *link.* aunque. **amongst** *loc.* ínsay; *loc.* míndul. **and** *link.* an<sup>1</sup> ; *link.* wé<sup>1</sup> ; *prep.* wet. **angry** *v.* vɛ́ks. **animal** *n.* animal. **animal (wild ~)** *n.* bíf. **answer** *v.* ánsa. **ant** *n.* nyɔ́ní. **antilope** *n.* fritámbo. **approximately** *adv.* aunque; *quant.* wán (3). **arm** *n.* hán. **around** *adv.* aráwn. **arrange** *v.* fíks. **arrive** *v.* rích. **arrogant** *v.* práwd. **arse** *n.* rás; *n.* kaka-rás, see: rás. **as** *adv.* lɛk; *link.* as. **as if** *link.* lɛk sé, see: lɛk. **as (much)** *adv.* tan. **as soon as** *link.* lɛk háw, see: lɛk. **ashamed** *v.* sém. **ashes** *n.* hásis. **ask** *v.* áks<sup>2</sup> . **ask for** *v.* bɛ́g. **association** *n.* sosáyti. **at** *prep.* fɔ; *prep.* na<sup>2</sup> . **at night** *adv.* na nɛ́t, see: nɛ́t. **at once** *adv.* wántɛn. **aunt** *n.* antí;

*n.* tia. **auxiliary (completive aspect ~)** *aux.* fínis. **auxiliary (continuative aspect ~)** *aux.* sigue; *aux.* stíl. **auxiliary (egressive aspect ~)** *aux.* kɔmɔ́t;

### **B - b**

**babe** *n.* bebí (2); *n.* bɛ́lps. **baby** *n.* bebí (1). **back (body part)** *n.* bák. **bad** *v.* bád. **bad (very ~)** *v.* wós. **bag** *n.* bág; *n.* bláy. **ball** *n.* bɔ́l. **bamboo** *n.* bambú. **banana** *n.* banána. **bank** *n.* banco. **baptise** *v.* baptáys; *v.* bautiza. **bark** *n.* kandá. **basket** *n.* bláy. **bathe** *v.* báf. **bathtub** *n.* bañera. **be** *part.* na<sup>1</sup> ; *part.* nóto; *v.* bí. **be at** *v.* dé<sup>2</sup> . **beach** *n.* bích. **beans** *n.* haricot. **bear** *v.* aguanta; *v.* bía<sup>2</sup> . **beard** *n.* byabyá. **beat** *v.* bít. **beautiful** *adj.* fáyn; *v.* bráyt; *v.* frɛ́s. **because** *link.* ás; *link.* bikɔs; *link.* como;

*aux; adv.* jís. **auxiliary (ingressive aspect ~)** *aux.* bigín. **auxiliary (prospective aspect ~)** *aux.* wánt. **avocado** *n.* pía<sup>1</sup> . **axe** *n.* áks<sup>1</sup> .

*link.* porque. **become** *v.* kɔmɔ́t (2); *v.* tɔ́n 2 (1). **bed** *n.* béd. **beer** *n.* bía<sup>1</sup> . **before (temporal)** *link.* bifó. **begin** *v.* bigín. **behind** *loc.* bihɛ́n. **belch** *v.* bɛ́lch. **believe** *v.* bilíf. **belly** *n.* bɛlɛ́. **belt** *n.* bɛ́lt. **bend (over)** *v.* bɛ́n. **between** *loc.* míndul. **bewitch** *v.* wích<sup>2</sup> . **big** *v.* bíg. **bird** *n.* pambɔ́d. **bit (a ~)** *adv.* lílíbit; *quant.* smɔ́l. **bite** *v.* bɛ́t 1 . **bitter tomato** *n.* jakató. **black** *v.* blák. **blood** *n.* blɔ́d. **blow** *n.* bló<sup>3</sup> . **blow (air)** *v.* bló<sup>1</sup> . **blue** *v.* blú. **board** *n.* plɛ́nk. **boastful** *v.* práwd. **body** *n.* skín; *n; prom.* bɔ́di. **boil** *v.* bwɛ́l. **bone** *n.* bón. **book** *n.* búk. **bored** *v.* aburre.

**born, be ~** *v.* bɔ́n (1). **boss** *n.* chíf; *n.* mása; *n.* másta. **bother** *v.* hambɔ́g. **bother (about)** *v.* bísin. **bottle** *n.* bɔ́tul. **bottom** *loc.* bɔtɔ́n. **bottom part** *n.* wés. **bounce** *v.* éks. **bowlegs** *n.* kobofút, see: fút. **box** *n.* bɔ́ks. **boy** *n.* bɔ́y; *n.* chico. **boyfriend** *n.* novio. **brain** *n.* sɛ́ns. **branch** *n.* stík. **bread** *n.* brɛ́d. **break** *v.* brók. **breakfast** *v.* desayuna. **breast** *n.* bɔbí. **breathe** *v.* púl brís, see: púl. **brick** *n.* bloque. **bright** *v.* bráyt. **bring** *v.* bríng. **bring up** *v.* trén.

### **C - c**

**cake** *n.* kék. **call** *v.* kɔ́l. **can** *n.* lata; *v.* fít. **canoe** *n.* kenú. **cap** *n.* káp. **capable** *v.* hébul. **car** *n.* motó. **card** *n.* tarjeta. **care for** *v.* máyn; *v.* mɛ́n (1). **careful (be ~)** *v.* sáful. **carpenter** *n.* kápinta. **carry** *v.* kɛ́r;

**brother** *n.* brɔ́da. **brother, big ~** *n.* grand frère. **brother-in-law** *n.* brɔdalɔ́; *n.* cuñado. **Bube** *pn.* Bubɛ. **bucket** *n.* bókit. **build** *v.* bíl. **building** *n.* hós. **bun** *n.* bocadillo. **burial** *n.* bɛ́rin. **burial-ground** *n.* bɛrin-grɔ́n, see: grɔ́n. **burn** *v.* rós. **burst (open)** *v.* bɔ́s. **bury** *v.* bɛ́r. **business** *n.* bísnɛs. **busy (with)** *v.* bísin. **but** *adv.* bat; *adv.* bɛt<sup>2</sup> ; *adv.* bɔt<sup>1</sup> . **buttocks** *n.* bata<sup>2</sup> ; *n.* wés. **buy** *v.* báy. **by** *prep.* bay; *prep.* fɔ; *prep.* na<sup>2</sup> ; *prep.* wet.

*v.* tót. **cassava** *n.* kasára. **catch** *v.* kéch. **cement** *n.* cemento. **chair** *n.* chía. **chair, (arm~)** *n.* butaca. **change** *v.* chénch. **charge** *v.* cobra. **chase away** *v.* drɛ́b. **chat up** *v.* tún (2). **check** *v.* chɛ́k. **cheek** *v.* fityáy. **chest** *n.* chɛ́s. **chew** *v.* chám.

**chief** *n.* chíf. **child** *n.* pikín. **chop** *v.* cháp. **church** *n.* chɔ́ch. **cigarette** *n.* sigá. **cinema** *n.* sinimá. **class** *n.* clase. **claw** *n.* finga-nél, see: fínga. **clean** *v.* klín. **clear** *v.* klía. **clever** *v.* klɛ́va. **climb** *v.* klém. **close** *v.* lɔ́k. **cloth** *n.* lapá. **cloth (piece of ~)** *n.* písis. **clothing** *n.* klós. **cloud** *n.* nube. **coat** *n.* kót. **cock** *n.* mán fɔ́l, see: fɔ́l 2 . **cocoa** *n.* cacao. **cocoa yam** *n.* kokó. **coconut** *n.* kokonát. **coffee** *n.* café; *n.* kɔfí. **cold** *v.* kól. **college** *n.* kɔ́lech. **colour** *n.* kɔ́la. **come** *v.* kán <sup>1</sup> . **come across** *v.* mít. **come back** *v.* tɔ́n bák, see: tɔ́n 2 . **come out** *v.* kɔmɔ́t (1). **company (commercial ~)** *n.* kɔ́mpani. **comparative particle** *adv.* mɔ́. **complementiser** *link.* de; *link.* fɔ; *link.* mék<sup>2</sup> ; *link.* sé; *link.* wé<sup>1</sup> . **complete** *v.* kɔmplít. **completive aspect auxiliary** *aux.* fínis. **conditional mood marker** *TMA.* fɔ. **confuse** *v.* kɔnfyús. **confusion** *ideo; n.* wuruwúrú. **contact (make, be in ~)** *v.* jám.

**continuative aspect auxiliary** *aux.* sigue; *aux.* stíl. **continue** *v.* sigue. **converse** *v.* púl torí, see: púl. **cook** *v.* kúk. **copula (affirmative identity ~)** *part.* na<sup>1</sup> . **copula (existential ~)** *v.* dé<sup>2</sup> . **copula (identity ~)** *v.* bí. **copula (locative ~)** *v.* de<sup>2</sup> . **copula (negative identity ~)** *part.* nóto. **copula (resultative ~)** *v.* lɛ́f 1 (1). **coquet** *v.* nyangá. **corn** *n.* kɔ́n. **corn porridge** *n.* ógi. **corner** *n.* kɔ́na. **corpse** *n.* day-mán, day-pɔ́sin, see: dáy. **corpulent** *v.* stáwt. **correct** *v.* kɔrɛ́t. **cost** *v.* kɔ́s 1 . **cough** *v.* kɔ́f. **count** *v.* kɔ́nt. **country** *n.* kɔ́ntri. **couple** *n.* pareja. **cousin** *n.* kɔsín. **cousin f.** *n.* prima. **cousin m.** *n.* primo. **cover** *n.* kɔ́ba; *v.* kɔ́ba. **crash** *v.* krás. **crazy** *v.* krés. **cross** *n.* cruz; *n.* krɔ́s; *v.* krɔ́s. **cry** *v.* kráy. **cunning** *n.* wayó. **cunning (be ~)** *v.* wayó. **cup** *n.* kɔ́p. **custard** *n.* kɔ́sta. **customary** *v.* nétif. **cut** *n.* kɔ́t; *v.* kɔ́t. **cut (hair, beard)** *v.* bába. **cutlass** *n.* kɔ́tlas.

### **D - d**

**dance** *v.* sék; *v; n.* dáns. **dark** *v.* blák; *v.* dák. **daughter-in-law** *n.* dɔtalɔ́. **dawn** *v.* brék. **day** *n.* dé<sup>3</sup> . **deceased** *v.* lét. **deep** *loc.* dɔ́n 2 ; *v.* díp. **deer** *n.* día<sup>1</sup> . **defeat** *v.* wín. **defecate** *v.* kaká. **demonstrative modifier** *det.* dán; *det.* dí; *det; prom.* dát; *det; prom.* dís. **demonstrative pronominal** *det; prom.* dát; *det; prom.* dís. **deny** *v.* dináy. **destroy** *v.* chakrá. **determiner** *det.* dán; *det.* dí; *det; prom.* dís. **determiner (definite)** *det.* di. **determiner (indefinite)** *quant.* sɔn; *quant.* wán (1). **devil** *n.* dɛ́bul. **devilish (be ~)** *v.* dɛ́bul. **die** *v.* dáy. **death** *n.* dáy.

### **E - e**

**ear** *n.* yés. **easy** *v.* ísi. **eat** *v.* chɔ́p. **egg** *n.* ɛ́ks. **eggplant** *n.* gadinɛ́ks, see: ɛ́ks. **egressive aspect auxiliary** *aux.* kɔmɔ́t; *aux; adv.* jís.

**die (off)** *v.* kwɛ́nch. **different** *v.* dífrɛn. **dig** *v.* díg. **dirty** *v.* dɔtí. **discipline** *v.* kɔrɛ́t. **disclose** *v.* prúf. **discotheque** *n.* boîte; *n.* discoteca. **disfigure** *v*. banfá **disorderliness** *ideo; n.* wuruwúrú. **do** *v.* dú; *v.* mék<sup>1</sup> . **doctor** *n.* dɔ́kta. **dog** *n.* dɔ́g. **domino** *n.* dominó. **done** *v.* dɔ́n 3 . **door** *n.* domɔ́t. **down** *loc.* dɔ́n 2 . **dowry** *n.* dote. **draw** *v.* drɔ́(1). **dress up** *v.* wɛ́r. **drink** *v.* dríng. **drive (a vehicle)** *v.* drɛ́b. **driver** *n.* dráyva. **drunk (be, get ~)** *v.* drɔ́ngo. *v.* chák. **dry** *v.* dráy. **dry season** *n.* dray-sísin, see: sísin. **due to** *link.* fɔséka; *prep.* fɔ; *prep.* wet.

**eight** *quant.* et. **electricity** *n.* lámp. **empty** *v.* ɛ́nti. **end up** *v.* lás. **endure** *v.* lás. **English** *pn.* Ínglis. **enjoy (oneself)** *v.* disfruta. **enlightened (be ~)** *v.* opinyáy. **enough (be ~)** *v.* dú; *v.* rích. **enter** *v.* ɛ́nta. **envy** *v.* jɛ́lɔs. **equal (to ~)** *v.* rích. **Equatoguinean f.** *pn.* guineana, see: Guinea. **Equatoguinean m.** *pn.* guineano, see: Guinea. **even** *adv.* aunque; *adv.* ni<sup>2</sup> ; *adv.* sóté; *part.* sénwe; *part.* sɛ́f. **even if** *link.* adɔnkɛ́; *link.* aunque.

### **F - f**

**face** *n.* fés. **faeces** *n.* kaká. **fall** *v.* fɔdɔ́n. **family** *n.* fámbul. **family member** *n.* fámbul. **Fang** *pn.* Fang. **Fang (person, language)** *pn.* Bata<sup>1</sup> (2). **far** *v.* fá; *v.* fáwe. **farm** *n.* fám. **fart** *v.* bló wín, see: bló<sup>3</sup> ; *v.* fát<sup>2</sup> . **fast** *v.* smát. **fat** *n.* grís; *v.* fát<sup>1</sup> . **father** *n.* pá; *n.* papá; *n.* papí. **father-in-law** *n.* fadalɔ́. **favour** *n.* févɔ. **fear** *v.* fía. **feather** *n.* fɛ́da. **feel** *v.* fíl<sup>1</sup> .

**evening** *n.* ívin; *n.* ívin tɛ́n, see: ívin. **every** *quant.* ɛ́ni; *quant.* ɔ́l. **exactly** *adv.* exactamente. **excessive (be ~)** *v.* óva. **excessively** *adv.* pás mák, see: mák; *adv.* óva. **excuse** *v.* ɛskyús. **exercise book** *n.* cuaderno. **expensive** *v.* día<sup>2</sup> . **explain** *v.* ɛsplén. **extinguish** *v.* ɔ́t. **extremely** *adv.* bad; *adv.* sóté. **eye** *n.* yáy.

**festivity** *n.* bigdé. **fever** *n.* fíba<sup>2</sup> . **fever (typhoid ~)** *n.* tifoidea. **few** *quant.* smɔ́l. **field** *n.* fíl<sup>2</sup> . **fight** *v.* fɛ́t. **fight over** *v.* kapú. **fill** *v.* fúlɔp; *v.* pák. **find** *v.* mít; *v.* sí. **fine** *adj.* fáyn. **finger** *n.* fínga. **finger nail** *n.* finga-nél, see: fínga. **finish** *v.* fínis. **finished** *v.* dɔ́n 3 . **fire** *n.* fáya. **fire wood** *n.* faya-wúd, see: wúd. **first** *quant; adv.* fɔ́s 1 . **first time (the ~)** *adv.* fɔ́s tɛ́n, see: fɔ́s 1 **fish** *n.* fís; *v.* físin. **fisherman** *n.* fisin-mán, see: físin.

.

**five** *quant.* fáyf. **fix** *v.* fíks. **fling** *v.* flíng. **fly** *v.* fláy. **focus marker** *part.* na<sup>1</sup> ; *part.* nóto; *part.* sénwe; *part.* sɛ́f. **fold** *v.* bɛ́n. **follow** *v.* fála; *v.* sigue. **food** *n.* chɔ́p. **fool** *n.* dógo; *n.* fúl. **foolish (be ~)** *v.* fúl; *v.* fúlis. **foot** *n.* fút. **for** *prep.* fɔ. **(for) ever** *adv.* fɔrɛ́va. **force** *v.* fɔ́s 2 . **foreigner** *n.* extranjero. **forest** *n.* bús.

### **G - g**

**Gabon** *pn.* Gabón. **game** *n.* plé. **garden** *n.* gádin. **gari** *n.* garí. **gash** *n.* kɔ́t. **genital area** *n.* wés. **genitals (male ~)** *n.* blɔkɔ́s. **get** *v.* gɛ́t. **get up** *v.* gráp. **girl** *n.* gál; *n.* gɛ́l; *n.* tití. **girlfriend** *n.* gɛ́l frɛ́n; *n.* novia. **girls** *n.* gáls. **give** *v.* gí. **give a blow** *v.* bló<sup>3</sup> . **give as a present** *v.* dás.

**forget** *v.* fɔgɛ́t. **formerly** *adv.* fɔ́s tɛ́n, see: fɔ́s 1 . **four** *quant.* fo. **fowl** *n.* fɔ́l 2 . **free** *v; adj.* frí. **fresh** *v.* frɛ́s. **friday** *n.* fráyde. **fridge** *n.* congelador. **friend** *n.* frɛ́n; *n.* kɔ́mpin. **from** *prep.* fɔ; *prep.* frɔn. **from - until** *prep.* frɔn - sóté, see: frɔn. **front** *loc.* bifó. **fruit** *n.* frút; *n.* fruta. **fry** *v.* fráy. **fuck** *v.* fɔ́k. **fufu** *n.* fufú. **full** *v.* bɛlfúl; *v.* fúlɔp; *v.* pák.

**give back** *v.* bák. **give birth** *v.* bɔ́n (1). **give classes** *v.* gí skúl, see: skúl. **glad** *v.* gládin. **glass** *n.* glás. **glasses** *n.* gafas; *n.* spɛ́tikul. **go** *v.* gó. **go down** *v.* baja; *v.* gó dɔ́n, see: gó. **go out** *v.* kɔmɔ́t (1). **go smoothly** *v.* sáful. **go to school** *v.* gó skúl, see: skúl. **go up** *v.* gó ɔ́p, see: gó; *v.* sube. **God** *n.* gɔ́d. **God (father ~)** *n.* papá gɔ́d. **God (my ~)** *intj.* Dios mio.

16 Word lists

**good** *v; adj.* gúd. **good evening** *intj.* gúd ívin, see: ívin. **good morning** *intj.* gúd mɔ́nin, see: mɔ́nin. **good (very ~)** *v.* bɛ́ta. **goods** *n.* kágo. **government** *n.* gɔ́bna. **gradually** *adv.* bambáy. **grain** *n.* grén. **grand** *adj.* gran-. **grandchild** *n.* gran-pikín, see: gran-. **grandfather** *n.* abuelo; *n.* gran-pá, see: gran-. *n.* gran-papá, see: gran-. **grandmother** *n.* abuela; *n.* gran-má, see: gran-;

### **H - h**

**habit** *n.* fásin. **habitual aspect marker** *TMA.* kin. **haggard** *v.* dráy. **hair** *n.* hía<sup>2</sup> . **hairdresser** *n.* peluqueria. **half** *quant.* háf. **half an hour** *quant.* háf áwa, see: háf. **half (one and a ~)** *quant.* wán-ɛn-háf. **hall** *n.* sala. **hammer** *n.* háma; *v.* háma. **hand** *n.* hán. **hang** *v.* hɛ́ng. **happen** *v.* pás (2). **happy** *v.* gládin. **harbour** *n.* puerto. **hard** *v.* hád; *v.* strɔ́n; *v.* tránga. **hardship** *n.* trɔ́bul. **harmattan** *n.* amatán. **have** *v.* gɛ́t; *v.* hól<sup>1</sup> . **have sex** *v.* fɔ́k; *v.* kɔ́t;

*n.* gran-mɔ́da, see: gran-. **grass** *n.* grás. **grate** *v.* ralla. **gravy** *n.* greví. **greedy** *v.* grídi. **green** *v.* grín. **greet** *v.* salút. **ground** *n.* grɔ́n. **groundnut** *n.* granát. **groundnut soup** *n.* granát súp, see: súp. **grow** *v.* gró. **guitar** *n.* gitá. **gulp down** *v.* nák. **gun** *n.* gɔ́n. **guts (have ~)** *v.* gɛ́t líba, see: líba.

*v.* krɔ́b; *v.* nák; *v.* slíp. **have units** *v.* gɛ́t saldo, see: saldo. **he** *pron.* e; *pron.*ín. **head** *n.* héd. **hear** *v.* hía<sup>4</sup> ; *v.* yɛ́r. **heart** *n.* hát. **heave** *v.* híb. **heavy** *v.* évi (1). **hectic** *ideo.* katakátá. **help** *v.* hɛ́lp. **hen** *n.* húman fɔ́l, see: fɔ́l 2 . **her** *pron.* =an<sup>2</sup> ; *pron.* ín; *pron.* in. **here** *adv.* hía<sup>1</sup> ; *adv.* yá. **here (right ~)** *adv.* yá só, see: yá. **hide** *v.* háyd. **hill** *n.* híl. **him** *pron.* =an<sup>2</sup> ; *pron.* ín.

**his** *pron.* in. **hit** *v.* nák. **hit with the head** *v.* bɔ́t 2 . **hold** *v.* hól<sup>1</sup> . **hole** *n.* hól<sup>2</sup> . **holiday** *n.* hɔ́lidé. **home** *n.* hóm. **hometown** *n.* kɔ́ntri. **hook** *n.* húk; *v.* húk. **horn** *n.* hɔ́n. **hospital** *n.* hospital; *n.* ɔspítul. **hot** *v.* hɔ́t.

### **I - i**

**I** *pron.* a. **ideophone** *ideo.* bwa; *ideo.* bya; *ideo.* fwífwífwí; *ideo.* gbin; *ideo.* gbogbogbo; *ideo.* kakara; *ideo.* kamúkamú; *ideo.* katakátá; *ideo.* kip; *ideo.* kutuku; *ideo.* kwáráng; *ideo.* kpu; *ideo.* ményéményé; *ideo.* príng; *ideo.* súkútúpampa; *ideo.* tík; *ideo.* wéwé; *ideo.* wówó; *ideo; intj.* kɔ́ngkɔ́ngkɔ́ng; *ideo; n.* wuruwúrú; *ideo; v.* wɔwɔ́; *n; ideo.* pɔtɔpɔ́tɔ́. **if** *link.* ɛf; *link.* if. **ill** *adj.* bád.

**hour** *n.* áwa. **house** *n.* hós. **how** *inter.* háw; *inter.* ús=stáyl, see: stáyl. **how much** *inter.* háw mɔ́ch, see: háw. **human-being** *n.* mán; *n.* pɔ́sin. **hungry** *v.* hángri. **hunt** *v.* hɔ́nti. **hunter** *n.* hɔnti-mán, see: hɔ́nti. **hurry** *v.* hɔ́ri; *v.* mekés. **hurt** *v.* pén.

**imperfective aspect marker** *TMA.* de. **implore** *v.* bɛ́g. **impregnate** *v.* bɛlɛ́. **impressive** *v.* évi (2). **in** *prep.* fɔ; *prep.* na<sup>2</sup> . **in front of** *loc.* bifó. **in order to** *link.* mék<sup>2</sup> ; *prep.* fɔ. **in (temporal)** *loc.* ínsay. **ingressive aspect auxiliary** *aux.* bigín. **injury** *n.* sɔfút. **inside** *loc.* ínsay. **inside out** *n.* rɔn-sáy, see: rɔn<sup>1</sup> . **insult** *v.* chík; *v.* kɔ́s 2 . **intelligent (be ~)** *v.* gɛ́t sɛ́ns, see: sɛ́ns. **intend** *v.* intenta; *v.* mín. **interjection** *ideo; intj.* kɔ́ngkɔ́ngkɔ́ng; *intj.* aa<sup>1</sup> ; *intj.* aa<sup>2</sup> ; *intj.* áy<sup>2</sup> ; *intj.* bió; *intj.* cháy; *intj.* chico;

*intj.* "chip"; *intj.* Dios mio; *intj.* dúya; *intj.* é; *intj.* ékié; *intj.* éy; *intj.* ɛ́; *intj.* ɛhɛ́; *intj.* ɛ́n; *intj.* hɛ́; *intj.* hó; *intj.* kúsɛ́; *intj.* mamá; *intj.* mán; *intj.* nawá; *intj.* ó; *intj.* papá;

### **J - j**

**jail** *n.* jél. **jealous** *v.* jɛ́lɔs. **job** *n.* jɔ́b. **join** *v.* jwɛ́n.

### **K - k**

**keep** *v.* kíp. **key** *n.* kí. **kick** *v.* kík. **kill** *v.* kíl. **kilo** *n.* kilo. **kind** *prom.* káyn. **kitchen** *n.* kíchin. **knee** *n.* ní.

### **L - l**

**lámp** *n.* lámp. **language** *n.* lángwech; *n.* tɔ́k; *n.* tɔ́ng.

*intj.* papá gɔ́d; *intj.* plís; *intj.* por Dios; *intj.* yé; *n; intj.* mierda. **interrogative particle** *inter.* ús=. **intestines** *n.* tripas. **iron** *n.* áyɛn; *v.* áyɛn. **iron, corrugated** *n.* chapa. **island** *n.* isla. **it** *pron.* =an<sup>2</sup> ; *pron.* e; *pron.* ín. **iterative aspect marker** *TMA.* kin. **its** *pron.* in.

**judge** *v.* jɔ́ch. **jump** *v.* júmp. **just** *aux; adv.* jís.

**knife** *n.* nɛ́f. **know** *v.* no<sup>2</sup> ; *v.* sabí. **know how to** *v.* sabí. **kola nut** *n.* kóla. **Kombe** *pn.* Kómbe. **Krio** *pn.* Krió.

**language (native ~)** *n.* lángwech; *n.* kɔntri-tɔ́k, see: tɔ́k. **last** *quant.* lás; *v.* kɛ́r;

*v.* lás; *v.* sté. **late** *v.* lét. **latrine** *n.* latrín. **laugh** *v.* láf. **lay** *v.* lé; *v.* slíp. **lazy** *v.* lési. **leaf** *n.* líf. **lean against** *v.* líng. **learn** *v.* lán. **leave** *v.* gó; *v.* lɛ́f 1 (1). **left** *n.* lɛ́f 2 ; *n.* lɛf-hán, see: lɛ́f 2 . **leg** *n.* fút. **letter** *n.* lɛ́ta. **lie** *v.* láy; *v.* lé; *v.* slíp. **life** *n.* láyf. **lift** *v.* hés. **light** *n.* láyt; *v.* láyt (1). **light (in colour)** *adj.* rubio; *v.* wáyt. **lightning** *n.* relámpago. **like** *adv.* como; *adv.* lɛk. *v.* lɛ́k. **like that** *adv.* só<sup>2</sup> . **line** *n.* láyn. **little** *v.* lílí.

### **M - m**

**madam** *n.* má (2). **madman** *n.* kres-mán, see: krés. **make** *v.* dú; *v.* mék<sup>1</sup> . **make an effort** *v.* tráy. **malanga** *n.* malanga. **malaria** *n.* malérya;

**live** *v.* líf. **live it up** *v.* ambiente; *v.* pachá. **liver** *n.* líba. **locative noun** *loc.* bifó; *loc.* bihɛ́n; *loc.* bɔtɔ́n; *loc.* dɔ́n 2 ; *loc.* ínsay; *loc.* míndul; *loc.* nadó; *loc.* nía; *loc.* ɔntɔ́p; *loc.* ɔ́p; *loc.* pantáp. **lock** *n.* lɔ́k; *v.* lɔ́k. **lonely** *v.* lónson. **long** *v.* lɔ́n. **long (temporal)** *adv.* sóté. **long time ago** *adv.* lɔ́n tɛ́n, see: lɔ́n. **look** *v.* lúk. **look for** *v.* fɛ́n. **loose** *v.* lɔ́s. **loose(n)** *v.* slák. **lorry** *n.* camión. **lounge** *v.* líng. **louse** *n.* lɔ́s. **lover** *n.* jɔmba. **low** *loc.* dɔ́n 2 . **lower chest** *n.* chekó. **luck** *n.* lɔ́k. **lucky** *v.* lɔ́ki.

*n.* paludismo. **man** *n.* mán. **manage** *v.* mánech. **manner** *n.* fásin; *n.* stáyl. **mark** *v; n.* mák. **market** *n.* mákit.

**marry** *v.* máred. **mash** *v.* más. **master** *n.* mása; *n.* másta. **matron** *n.* mísis. **matter** *n.* kés; *n.* plába. **maybe** *adv.* sɔntɛ́n. **me** *pron.* mí. **mean** *v.* badhát. **mean (to)** *v.* mín. **meat** *n.* bíf. **medicine** *n.* mɛ́rɛsin. **meet** *v.* mít; *v.* mítɔp. **meeting** *n.* reunión. **mess** *n.* porcería. **messy** *ideo; v.* wɔwɔ́. **middle** *loc.* míndul. **milk** *n.* mélk. **mind** *n.* sɛ́ns; *v.* máyn; *v.* mɛ́n (1). **ministry** *n.* ministerio. **miss (a person)** *v.* lónson. **mix** *v.* míks.

### **N - n**

**nail** *n.* nél. **naked** *v.* nékɛd. **name** *n.* ném. **narrate** *v.* cuenta. **nation** *n.* nésɔn. **near** *loc.* nía; *prep.* kɔ́na. **nearly** *adv.* smɔ́ltɛn. **neck** *n.* nɛ́k. **need** *v.* níd. **needle** *n.* nidul. **negative particle** *part.* nó<sup>1</sup> ; *part.* nɔ́. **neighbour f.** *n.* vecina.

**monday** *n.* mɔ́nde. **money** *n.* mɔní. **monkey** *n.* mɔnkí. **month** *n.* mún. **moon** *n.* mún. **more** *adv.* mɔ́. **more and more** *adv.* mɔ́-ɛn-mɔ́. **more (be ~)** *v.* mɔ́. **morning** *n.* mɔ́nin. **mosquito** *n.* mosquito. **mosquito net** *n.* mosquitero. **mother** *n.* má (1); *n.* mamá; *n.* mamí; *n.* mɔ́mi. **mountain** *n.* híl. **mouth** *n.* mɔ́t. **move** *v.* múf. **movie** *n.* sinimá. **much** *adv.* mɔ́ch; *v; quant.* bɔkú; *v; quant.* plɛ́nte. **mucus** *n.* frɛskól. **mud** *n; ideo.* pɔtɔpɔ́tɔ́. **must** *TMA.* mɔs. **my** *pron.* mi.

**neighbour m.** *n.* vecino. **neither** *link.* ni. **nephew** *n.* sobrino. **new** *v.* nyú. **news** *n.* nyús. **next** *quant.* nɛ́ks. **next to** *loc.* nía; *prep.* kɔ́na. **nice** *v.* náys. **niece** *n.* sobrina. **night** *n.* nɛ́t. **nine** *quant.* náyn. **nipple** *n.* chupete. **nit** *n.* nít.

**no** *part.* nó<sup>1</sup> ; *part.* nɔ́. **noise** *n.* nɔ́ys. **noon** *n.* sán tɛ́n, see: sán. **normal** *adj.* normal. **nose** *n.* nós.

### **O - o**

**obligative mood marker** *TMA.* mɔs. **of** *prep.* ɔf; *prep.* wet. **offend** *v.* fityáy; *v.* kɔs<sup>2</sup> . **office** *n.* ɔ́fis. **oil** *n.* ɔ́yl; *n.* pamáyn. **oil (crude ~)** *n.* petróleo. **oil rig** *n.* plataforma. **old** *v.* ól. **on** *loc.* ɔntɔ́p; *loc.* pantáp; *prep.* na<sup>2</sup> ; *prep.* pan. **on par (be ~)** *v.* bráket. **once** *link.* wáns. **once (at~)** *adv.* wán tɛ́n, see: wán. **one** *prom.* wán;

### **P - p**

**pack** *v.* pák. **pain** *v.* pén. **paint** *v.* pént. **pair** *n.* pía<sup>2</sup> . **pair of shoes** *n.* pía sús, see: pía<sup>2</sup> . **palm kernel** *n.* bangá. **palm nut soup** *n.* bangá súp, see: súp. **palm tree** *n.* bangá. **palm-wine** *n.* topé. **pan** *n.* pán. **pancake** *n.* pan-kék, see: kék. **paper** *n.* pépa.

**nothing** *prom.* nátin. **notice** *v.* fija. **now** *adv.* náw. **now (right ~)** *adv.* náw só, see: náw. **number** *n.* nɔ́mba.

*quant.* wán (1). **one and a half** *quant.* wán-ɛn-háf. **onion** *n.* yabás. **only** *quant.* dásɔl; *quant.* ónli; *quant.* sósó; *quant.* wán (2). **open** *v.* ópin. **or** *link.* o; *link.* ɔ. **orange** *v.* rɛ́d. **other** *quant.* ɔ́da. **outer layer** *n.* kandá. **outside** *loc.* nadó. **over** *adv.* óva. **over (be ~)** *v.* óva. **overheat** *v.* ova-hɔ́t, see: óva. **own** *prom.* yón.

**paralyse** *n.* paraláys. **parlour** *n.* pála. **part** *n.* pát. **party** *n.* bigdé; *v.* ambiente; *v.* pachá. **pass** *v.* pás (1). **past tense marker** *TMA.* bin. **patient** *n.* paciente. **pay** *v.* pé. **peel** *v.* píl. **pen** *n.* bolí.

**penis** *n.* miná; *n.* prík. **people** *n.* nésɔn; *n.* pípul. **pepper** *n.* pɛ́pɛ. **perfect tense-aspect marker** *TMA.* dɔ́n **perfect tense-aspect marker (negative ~)** *TMA.* nɛ́a; *TMA.* nɔ́ba. **perfective aspect marker (narrative ~)** *v.* kán<sup>2</sup> . **perhaps** *adv.* sɔntɛ́n. **persevere** *v.* aguanta. **person** *n.* mán; *n.* pɔ́sin. **persuade** *v.* tún (2). **pestle** *n.* matapenso. **petrol** *n.* gasolina. **photo** *n.* fotó. **Pichi** *pn.* Píchi. **pick (up)** *v.* pík. **pierce** *v.* chúk. **pinch** *v.* pínch. **pistol** *n.* písul. **place** *n.* pát; *n.* plés; *n.* sáy; *pn.* Guinea. **plait** *v.* plánt (2). **plane** *n.* avión. **plank** *n.* plɛ́nk. **plant** *v.* plánt (1). **plantain** *n.* plantí. **plastic** *n.* plástík. **plate** *n.* plét. **play** *v.* plé. **please** *intj.* dúya; *intj.* plís. **plenty** *v; quant.* plɛ́nte. **plot** *n.* grɔ́n. **plural marker** *pron.* dɛn. **police** *n.* policía; *n.* polís.

1 . **police station** *n.* comisaría. **poor** *v.* pó. **pot** *n.* pɔ́t. **potato** *n.* papa; *n.* patata. **potency** *n.* páwa. **potential mood marker** *TMA.* go. **pour** *v.* trowé (2). **powder** *n.* páwda. **power** *n.* páwa. **pregnancy** *n.* bɛlɛ́. **pregnant** *adj.* gɛ́t bɛlɛ́, see: bɛlɛ́. **pregnant woman** *n.* bɛlɛ-húman, see: bɛlɛ́. **prepare** *v.* rɛdí. **preposition** *prep.* ápás; *prep.* bay; *prep.* bitáwt. *prep.* fɔ; *prep.* frɔn; *prep.* kɔ́na; *prep.* na<sup>2</sup> ; *prep.* ɔf; *prep.* pan; *prep.* sin; *prep.* síns; *prep.* sóté; *prep.* to; *prep.* wet. **problem** *n.* plába; *n.* problema; *n.* prɔ́blɛm. **profound** *v.* strɔ́n. **promise** *v.* prɔ́mis. **pronoun** *pron.* a; *pron.* =an<sup>2</sup> ; *pron.* dɛ́n; *pron.* dɛn; *pron.* e; *pron.* ín; *pron.* in; *pron.* mí; *pron.* mi; *pron.* una;

*pron.* wí; *pron.* wi; *pron.* yú; *pron.* yu. **property** *n.* prɔpatí. **prospective aspect auxiliary** *aux.* wánt. **proud** *v.* práwd. **prove** *v.* prúf. **provoke** *v.* chik. **public holiday** *n.* bigdé.

### **Q - q**

**quarrel** *v.* chakrá. **quarry** *n.* kwári. **quarter (of a town)** *n.* kwáta. **question** *n.* kwɛ́sɔn. **quickly** *adv.* kwík. **quotative marker** *link.* sé.

### **R - r**

**radiant** *v.* bráyt. **rag** *n.* písis. **rain** *n.* rén; *v.* fɔ́l 1 . **rainy season** *n.* ren-sísin, see: sísin. **rat** *n.* aráta. **reach** *v.* rích. **realise** *v.* kechɔ́p. **rear** *loc.* bihɛ́n. **rebound** *v.* bɔ́t 2 . **receive** *v.* cobra. **record** *v.* graba. **red** *v.* rɛ́d. **refuse** *v.* dináy; *v.* refyús. **relax** *v.* bló<sup>2</sup> . **remain** *v.* lɛ́f 1 (1). **remember** *v.* mɛ́mba. **remind** *v.* mɛ́mba. **remove** *v.* drɔ́(1); *v.* púl. **rent** *v.* alquila. **repair** *v.* fíks. **repeat** *v.* bis.

**pull** *v.* púl. **pump** *v; n.* pɔ́mp. **punish** *v.* pɔ́nis. **pure** *v.* pyɔ́. **purpose clause introducer** *link.* fɔ; *link.* mék<sup>2</sup> . **push** *v.* pús; *v.* súb. **put** *v.* pút.

**resemble** *v.* fíba<sup>1</sup> . **reside** *v.* blánt; *v.* líf; *v.* sidɔ́n. **respect** *v.* rɛspɛ́t. **respectable** *v.* strét. **rest** *v.* bló<sup>2</sup> ; *v.* rɛ́s 1 . **restaurant** *n.* restaurante. **restless (be ~)** *ideo.* kakara. **reverse** *n.* rɔn-sáy, see: rɔ́n 1 . **rice** *n.* rɛ́s 2 . **rich person** *n.* jɛntri-mán, see: jɛ́ntri. **riches** *n.* jɛ́ntri. **right** *n.* ráyt<sup>2</sup> . **right (be ~)** *v.* gɛ́t ráyt, see: ráyt<sup>2</sup> . **right (side)** *n.* rayt-hán, see: ráyt<sup>2</sup> . **ring** *n.* ríng. **ripe** *v.* rɛ́p. **river** *n.* ríva. **road** *n.* ród. **roast** *v.* rós. **room** *n.* rúm. **rope** *n.* róp.

**rot** *v.* rɔ́tin. **round** *n.* vuelta. **rub** *v.* rɔ́b.

### **S - s**

**sack** *n.* saco. **salt** *n.* sɔ́l. **saltwater** *n.* solwatá. **same** *prom.* sén. **sand** *n.* arena; *n.* sansán. **sardine** *n.* sadín. **satiated** *v.* bɛlfúl. **saturday** *n.* sátide. **say** *v.* sé. **scared (be ~)** *v.* skía. **scent** *n.* sɛ́nt. **school** *n.* skúl. **school (secondary ~)** *n.* colegio. **scissors** *n.* sísɔs. **scratch** *v.* skrách. **scrub** *v.* krɔ́b. **sea** *n.* solwatá. **season** *n.* sísin. **second** *quant.* sɛkɔ́n. **secret** *n.* síkrit. **secretely** *adv.* haydháyd, see: háyd. **see** *v.* sí. **seed** *n.* síd. **seem** *v.* fíba<sup>1</sup> . **seize** *v.* kapú. **self** *prom.* sɛ́f. **sell** *v.* sɛ́l. **send** *v.* sɛ́n. **sentence** *n.* frase. **series** *n.* serie. **serious** *v.* síryɔs. **seven** *quant.* sɛ́ven. **sew** *v.* só<sup>3</sup> . **shake** *v.* sék. **share** *v.* séb. **she** *pron.* e.

**run** *v.* rɔ́n 2 . **rush** *v.* hɔ́ri. **rush (to a place)** *v.* fláy (2).

**ship** *n.* stíma. **shirt** *n.* sɔ́t 2 . **shit** *n.* kaká; *intj.* mierda. **shoe** *n.* sús. **shoot** *v.* sút. **shop** *n.* sɔ́p. **short** *v.* sɔ́t 1 . **shortly after** *adv.* smɔ́ltɛn. **shout** *v.* hála. **shove** *v.* súb. **show** *v.* só<sup>1</sup> . **shower** *v.* sáwa. **sick** *adj.* bád. **sick (be ~)** *v.* sík; *v.* banfá. **sickness** *n.* sík. **side** *n.* sáy. **sieve** *n.* sífta. **sift** *v.* sífta. **sign** *v.* sáyn. **since** *prep.* frɔn. **since long** *adv.* frɔn bɔkú tɛ́n, see: frɔn; *adv.* lɔ́n tɛ́n, see: lɔ́n. **since (temporal)** *prep.* síns. **sincere** *v.* strét. **sing** *v.* síng. **singlet** *n.* camiseta. **sir** *n.* sá. **sister** *n.* sísta. **sister-in-law** *n.* cuñada; *n.* sistalɔ́. **sit** *v.* sidɔ́n. **six** *quant.* síks. **skin** *n.* kandá. **sky** *n.* cielo. **slap** *v.* sláp.

**sleep** *v.* slíp. **sleepy (be ~)** *v.* fíl slíp, see: fíl<sup>1</sup> **slim** *v.* slím. **slipper** *n.* slipás. **slow** *adj.* sló. **slow (be ~)** *v.* sáful. **small** *v.* smɔ́l. **smell** *n.* smɛ́l; *v.* smɛ́l. **smoke** *v; n.* smók. **snail** *n.* kɔ́nk. **snake** *n.* snék. **so** *adv.* so; *adv.* tal; *link.* pues. **soap opera** *n.* serie. **soft** *v.* sáf. **soil** *n.* sansán. **soldier** *n.* soldado; *n.* sólya. **some** *quant.* sɔn. **song** *n.* síng. **son-in-law** *n.* sɔnilɔ́. **sorcerer** *n.* wích<sup>2</sup> . **sorcery** *n.* mɛ́rɛsin; *n.* wích<sup>2</sup> . **sorry** *v.* sɔ́ri. **sound** *n.* sonido. **soup** *n.* súp. **Spain** *pn.* Panyá. **Spaniard** *pn.* Panyá. **Spanish** *pn.* Panyá. **spend** *v.* spɛ́n. **spirit** *n.* spírit. **spit** *n.* spít; *v.* spít. **spoil** *v.* spwɛ́l. **spoon** *n.* spún. **spy on** *v.* espia. **square** *n.* plaza. **squeeze** *v.* skwís. **stab** *v.* chúk. **stand (up)** *v.* tánap;

.

*v.* tínap. **standard marker** *v.* pás (1). **star** *n.* estrella; *n.* stá. **start** *v.* stát. **stay** *v.* sté. **steal** *v.* tíf. **stew** *n.* styú. **stick** *n.* stík. **stick in(to)** *v.* pín. **sticky (of an okro soup)** *v.* drɔ́(1). **still** *adv.* stíl; *adv.* yét. **sting** *v.* chúk. **stingy** *v.* grídi. **stone** *n.* stón (1). **stoop (over)** *v.* butú. **stop** *v.* lɛ́f 1 (1); *v*. para; *v.* stɔ́p. **store** *n.* almacén. **storey** *n.* piso. **story** *n.* torí. **straight** *v.* strét. **strap** *n.* bɛ́lt. **street** *n.* strít. **strength** *n.* páwa. **strong** *v.* strɔ́n; *v.* tránga. **style** *n.* stáyl. **stylish (be ~)** *v.* spót. **subjunctive mood marker** *link.* mék<sup>2</sup> . **subordinator** *link.* wé<sup>1</sup> . **suddenly** *adv.* wántɛn. **suffer** *v.* sɔ́fa. **sugar** *n.* súga. **sun** *n.* sán. **sunday** *n.* sɔ́nde. **surname** *n.* apellido. **surpass** *v.* pás (1). **surround** *v.* ráwn. **swallow** *v.* swɛ́la. **sweat** *n.* swɛ́t;

*v.* swɛ́t. **sweet** *v.* swít (2). **sweetheart** *n.* bebí (2);

### **T - t**

**table** *n.* tébul. **tailor** *n.* sastre. **take** *v.* kɛ́r; *v.* ték. **take a picture** *v.* púl fotó, see: púl. **take down** *v.* baja. **take up** *v.* sube. **talk** *v.* tɔ́k. **tall** *v.* lɔ́n. **taro** *n.* malanga. **tasty** *v.* swít (1). **taxi** *n.* taksí. **teach** *v.* lán; *v.* tích. **teacher** *n.* profesor; *n.* tícha. **tear** *v.* chɛ́r. **telephone** *n.* teléfono. **telephone (mobile ~)** *n.* móvil. **television** *n.* televisión. **tell** *v.* tɛ́l. **tell a story** *v.* púl torí, see: púl. **term of address** *n.* antí; *n.* brɔ́da; *n.* cuñada; *n.* cuñado; *n.* chíf; *n.* gran-má, see: gran-; *n.* gran-pá, see: gran-; *n.* grand frère; *n.* má (1); *n.* má (2); *n.* mamá; *n.* mamí; *n.* mása; *n.* másta; *n.* mísis;

*n.* bɛ́lps. **swim** *v.* swín.

*n.* mɔ́mi; *n.* ɔnkúl; *n.* pá; *n.* papá; *n.* papí; *n.* sá; *n.* sísta; *n.* tia; *n.* tio. **testicle** *n.* stón (2). **thanks** *intj.* tɛ́nki. **that** *det.* dán; *det; prom.* dát; *link.* sé; *link.* wé<sup>1</sup> . **the** *det.* di. **them** *pron.* dɛ́n. **then** *adv.* áfta; *adv.* dásɔl. **there** *adv.* dé<sup>1</sup> . **there (over ~)** *adv.* yandá. **they** *pron.* dɛ́n; *pron.* dɛn. **thick** *v.* tík. **thief** *n.* tif-mán, see: tíf. **thin** *v.* dráy; *v.* smɔlskín, see: skín; *v.* slím. **thing** *n.* tín. **think** *v.* chɛ́k; *v.* tínk. **think of** *v.* mɛ́mba. **thirsty** *v.* tɔ́sti. **this** *det.* dí; *det; prom.* dís. **thorn** *n.* chukchúk. **three** *quant.* trí.

**throat** *n.* trót. **through** *prep.* wet. **throw** *v.* flíng; *v.* híb; *v.* sɛ́n; *v.* stón; *v.* trowé (1). **throw away** *v.* trowé (1). **throw stones at** *v.* stón. **thunderstorm** *n.* tinada. **tickle** *v.* chikilís. **tie** *v.* táy. **tight(en)** *v.* táyt. **time** *n.* áwa; *n.* tɛ́n. **tiny** *v.* lílí. **tipsy** *v.* láyt (2). **tired** *v.* táya. **TMA marker** *link.* mék<sup>2</sup> ; *TMA.* bin; *TMA.* de; *TMA.* dɔ́n 1 ; *TMA.* fɔ; *TMA.* go; *TMA.* kán<sup>2</sup> ; *TMA.* kin; *TMA.* mɔs; *TMA.* nɛ́a; *TMA.* nɔ́ba. **to** *prep.* fɔ; *prep.* to. **today** *n.* tidé. **tomato** *n.* tamátis; *n.* tomate.

### **U - u**

**ugly** *ideo; v.* wɔwɔ́. **uncle** *n.* ɔnkúl; *n.* tio. **under** *loc.* bɔtɔ́n. **understand** *v.* hía<sup>4</sup> ; *v.* ɔndastán;

**tomorrow** *n.* tumára. **tongue** *n.* tɔ́ng. **too** *part.* sénwe; *part.* sɛ́f. **too much** *adv.* tú mɔ́ch, see: mɔ́ch; *adv.* tú<sup>2</sup> . **tooth** *n.* tít. **top** *loc.* ɔntɔ́p; *loc.* pantáp; *n.* tɔ́p. **touch** *v.* tɔ́ch. **towel** *n.* táwɛl. **town** *n.* tɔ́n 1 . **traditional** *adj.* tradicional; *v.* nétif. **train** *v.* trén. **training** *n.* trénin. **transport** *n.* transporte. **travel** *v.* trávul; *v.* viaja. **tree** *n.* stík. **tremble** *v.* trímbul. **trickery** *n.* ajáajá. **trouble** *n.* plába; *n.* trɔ́bul. **trousers** *n.* trɔsís. **true** *v.* trú. **try** *v.* intenta; *v.* tráy. **tuesday** *n.* tyúsde. **tune** *v.* tún (1). **turn** *v.* tɔ́n 2 (1). **turn out** *v.* kɔmɔ́t (2). **two** *quant.* tú<sup>1</sup> .

*v.* yɛ́r. **unfortunately** *adv.* bádtɛn. **units (mobile phone ~)** *n.* saldo. **until** *prep.* sóté. **up** *loc.* ɔ́p. **upbringing** *n.* trénin.

**upside down** *n.* rɔn-sáy, see: rɔ́n 1 . **urinate** *v.* pipí. **urine** *n.* pipí. **us** *pron.* wí.

### **V - v**

**vagina** *n.* opó; *n.* totó. **vegetables** *n.* verdura. **vehicle** *n.* motó. **village** *n.* kɔ́ntri;

### **W - w**

**wait** *v.* wét. **wake (up)** *v.* wék; *v.* wékɔp. **walk** *v.* wáka; *v.* wɔ́k. **want** *v.* wánt. **warm** *v.* hɔ́t 1 . **wash** *v.* wás. **watch** *v.* wách. **water** *n.* watá; *n.* wɔtá. **water (cold, cool ~)** *n.* kol-watá, see: watá. **water (hot, warm ~)** *n.* hɔt-watá, see: watá. **way** *n.* wé<sup>2</sup> . **way (the ~ that)** *link.* lɛk háw, see: lɛk. **we** *pron.* wí; *pron.* wi. **wear** *v.* wɛ́r. **weather** *n.* dé<sup>3</sup> . **wednesday** *n.* wɛ́nsde. **weed** *v.* chapea. **week** *n.* wík. **weekend** *n.* fin dé semana. **welcome** *intj.* wɛ́lkɔm. **well** *v.* wɛ́l. **well-dressed (be ~)** *v.* spót. **wet** *v.* sók.

**use** *v.* yús. **use up** *v.* spwɛ́l. **used to (be ~)** *v.* kɔ́stɔn.

*n.* pueblo; *n.* vílech. **virgin(ity)** *n.* vájin. **visit** *v.* visít. **vomit** *v.* vɔ́mit.

**what** *inter.* ús=tin, see: ús=; *inter.* wétin. **when** *inter.* ús=áwa, see: ús=; *inter.* ús=tɛ́n, see: ús=; *link.* wé<sup>1</sup> . **where** *inter.* ús=sáy, see: ús=. **whether** *link.* ɛf; *link.* if. **which** *inter.* ús=; *inter.* wích<sup>1</sup> . **which (kind of)** *inter.* ús=káyn, see: ús=. **while** *link.* wé<sup>1</sup> . **whine** *ideo.* ményéményé. **white** *v.* wáyt. **who** *inter.* údat; *inter.* ús=pɔ́sin, see: ús=. **whole** *quant.* hól<sup>3</sup> . **wicked** *v.* wíkɛd. **wife** *n.* wáyf. **win** *v.* wín. **wind** *n.*brís. **window** *n.* wínda. **wine** *n.* vino. **wipe** *v.* wáyp. **witch** *n.* wích<sup>2</sup> . **with** *prep.* wet. **without** *prep.* bitáwt; *prep.* sin.

**woman** *n.* húman. **womb** *n.* bɛlɛ́. **wonder** *v.* wánda. **wood** *n.* stík; *n.* wúd. **word** *n.* tɔ́k; *n.* wɔ́d. **work** *v; n.* wók.

### **Y - y**

**year** *n.* hía<sup>3</sup> . **yes** *intj.* yɛ́s. **yes (strong)** *intj.* ɛhɛ́. **yesterday** *n.* yɛ́stadé. **yet** *adv.* yét.

**world** *n.* wɔ́l. **worm** *n.* tumbú. **worry** *v.* wɔ́ri. **wound** *n.* kɔ́t; *n.* sɔfút. **write** *v.* rayt<sup>1</sup> . **wrong** *v.* rɔ́n 1 .

**you (pl)** *pron.* una. *pron.*unu. **you (sg)** *pron.* yú; *pron.* yu. **young** *v.* yún.

## **References**


#### References


Clements, J. Clancy. 2009. *The linguistic legacy of Spanish and Portuguese: Colonial expansion and language change* (Cambridge Approaches to Language Contact). Cambridge: Cambridge University Press.

Comrie, Bernard. 1976. *Aspect*. Cambridge: Cambridge University Press.


#### References


*West African linguistics: Descriptive, comparative, and historical studies in honor of Russell G. Schuh*, 100–133. London: J.M. Dent & Co.


Kemmer, Suzanne. 1993. *The middle voice*. Amsterdam: John Benjamins.


#### References


#### References


Yip, Moira. 2002. *Tone*. Cambridge: Cambridge University Press.


## **Name index**

Abraham, Roy Clive, 466

Adegbite, Wale, 484 Aikhenvald, Alexandra Y., 423, 426 Alleyne, Mervyn C., 6, 37, 45 Ameka, Felix, 448–450 Ameka, Felix K., 287, 433 Anderson, Lloyd B., 185 Auer, Peter, 467 Baker, Philip, 6 Baron, Naomi, 344 Beller, Sieghard, 303, 305 Berry, Jack, 1, 45 Bickerton, Derek, 412 Bilby, Kenneth M., 465 Blanker, J. C. M., 191 Bloomfield, Leonard, 450 Boersma, Paul, 12 Bolekia Boleká, Justo, 4, 489 Bordal Steien, Guri, 37, 42, 45 Breu, Walter, 145 Brown, Penelope, 395 Buchholz, Oda, 275 Bybee, Joan L., 368 Cantús, Dolores García, 11 Castillo-Rodríguez, Susana, 4 Chafe, Wallace L., 226 Chung, Sandra, 143, 158 Clements, J. Clancy, 472 Comrie, Bernard, 81, 135, 156, 395 Coomber, Ajayi, 11 Corcoran, Chris, 10 Criper, Lindsay, 45 Criper-Friedman, Lindsay, 45 Cristofaro, Sonia, 366, 391 Croft, William A., 324

Cusic, David D., 93 Dahl, Östen, 145 Devonish, Hubert, 37 Di Cristo, Albert, 61 Dillard, John L., 6 Dingemanse, Mark, 443 Dixon, R. M. W., 76 Doke, Clement Martyn, 443 Downing, Laura J., 93 Drubig, Hans Bernhard, 226 Dubbeldam, Jaap, 191 Durie, Mark, 423 Enfield, N. J., 437 Essegbey, James, 175, 294, 340 Faraclas, Nicholas, 6, 37 Faraclas, Nicholas G., 45, 86, 481 Farquharson, Joseph, 77 Fennig, Charles D., 1 Figueroa, Esther, 457 Finney, Malcolm Awadajin, 10, 11, 440 Fyfe, Christopher, 1 Fyle, Clifford N., 7, 11, 170 Good, Jeffrey C., 37, 45 Görlach, Manfred, 6 Güldemann, Tom, 377, 379, 381 Gussenhoven, Carlos, 45 Hajek, John, 423 Hammarström, Harald, 1 Hancock, Ian F., 6 Haspelmath, Martin, 124, 132, 275, 324 Heine, Bernd, 89, 156, 235 Himmelmann, Nikolaus P., 434

#### Name index

Hirst, Daniel, 61 Holm, John A., 1 Hopper, Paul J., 178, 322, 359 Hovav, Malka Rappaport, 324 Huber, Magnus, 1, 6 International African Institute, 12 Jaggar, Philip J., 145

James, Winford, 178 Jones, Eldred Durosimi, 7, 11, 170 Jungraithmayr, Herrmann, 201

Keenan, Edward L., 395, 397 Kemmer, Suzanne, 362 Kockelman, Paul, 448 Koopman, Hilda Judith, 243 Kouwenberg, S., 118 Kuteva, Tania, 235

Labov, William, 178, 180, 185 Larson, Richard, 243 Lefebvre, Claire, 243 Lehmann, Christian, 366 Levin, Beth, 317, 324 Levinson, Stephen C., 303 Ley Núm. 5/2.007, 2 Li, Charles N., 185, 248 Lipski, John M., 1, 2, 4, 42 Longacre, Robert E., 178, 293, 452 Lynn, Martin, 4

Martín del Molino, Amador, 11 Meechan, Marjory, 475 Michaelis, Susanne Maria, 17 Miestamo, Matti, 201 Migge, Bettina, 137 Mithun, Marianne, 376 Morgades Besari, Trinidad, 4, 11 Morgenthaler García, Laura, 4 Mufwene, Salikoko S., 115 Mühleisen, Susanne, 461 Muysken, Pieter, 6, 212, 243, 338, 467 Myers-Scotton, Carol, 467

Næss, Åshild, 358, 359 Ngefac, Aloysius, 2 Nkengasong, Nkemngong, 170

Odden, David, 93 Olo Fernandes, Lucas, 2

Palmer, Frank Robert, 166 Pierrehumbert, Janet Breckenridge, 61 Pollard, Velma, 179 Poplack, Shana, 471, 475 Post, Marike, 2

Rickford, Angela E., 457 Rickford, John R., 6, 457 Rivera Castillo, Yolanda, 37 Rivera-Castillo, Yolanda, 37 Rountree, S. Catherine, 37

Sala, Bonaventura Mbiydzenyuy, 2 Sankoff, David, 471 Sankoff, Gillian, 395 Sasse, Hans-Jürgen, 135, 145, 181 Schultze-Berndt, Eva, 434 Schultze-Berndt, Eva F., 434 Simons, Gary F., 1 Singler, John Victor, 6 Slobin, Dan I., 185 Smith, Dandeson, 9 Smith, Norval, 6, 7, 212 Song, Jae Jung, 368 Stassen, Leon, 187, 188, 190, 391, 396 Suzuki, Keiichiro, 58

Tarallo, Fernando, 395, 399 Tenny, Carol L., 359 Thompson, Sandra A., 81, 178, 248 Timberlake, Alan, 143, 158

Udofot, Inyang, 45 UNESCO, 12

Weenink, David, 12 Welmers, William Everett, 141 Westermann, Diedrich, 443

Name index

Wilner, John, 465 Winford, Donald, 6, 137, 179

Yakpo, Kofi, 1, 2, 4, 6, 11, 12, 37, 42, 52, 92, 366, 467 Yillah, Mohamed Sorie, 10 Yip, Moira, 56 Youssef, Valerie, 178

Zarco, Mariano de, 1, 2, 11

## **Language index**

Aku, 6, 487

English, 2, 6, 8, 9, 12, 45, 487, 559 Fa d'Ambô, 1, 12 Fang, 1, 4, 12, 559 French, 2, 12, 559 German, 12 Jamaican, 487 Kombe, 12

Bube, 1, 4, 12, 489, 559

Krio, 1, 2, 6–11, 440, 465, 487

Lunga Ngola, 2 Lungwa Santome, 2, 12

Mende, 45

Nigerian, 4

Portuguese, 1, 2

Saramaccan, 45

Spanish, 1, 2, 4, 9, 11, 12, 17, 20, 45, 59, 67, 75, 97, 121, 170, 194, 206, 213, 224, 229, 233, 234, 238, 240, 255, 289, 300, 333, 338, 339, 416, 421, 450–452, 461– 464, 466, 467, 474, 477, 487, 489, 559

Sranan, 179

Twi, 12

Yoruba, 466, 484

abilitive mood, 168 adjectives, 475 adjuncts, 339, 340, 345, 350, 438 adverbial phrases, 287, 298, 316, 343 adverbs, 278, 301 agent, 326, 357, 362, 369, 370 alternative question, 72 alternative questions, 212 anaphora, 105, 108, 110, 112, 131, 247, 374 animacy, 121, 288, 292, 298, 371 answers, 226, 462 article,106,107,111,115,116, 282, 468, 469 aspect, 9, 75, 139, 141, 151, 178, 186, 271, 324, 373, 407, 408, 428, 432, 435, 437, 440, 441, 473 aspectual verbs, 151 assimilation of segments, 49 associative, 285 associative constructions, 75, 90, 119, 213 associative objects, 258, 343 associative preposition, 280, 291, 296, 314, 315, 346, 389, 426 auxiliaries, 157, 168, 169, 171, 172, 186, 329, 473 basic vocabulary, 7 beneficiary, 198, 269, 296, 312, 343, 348, 349 beneficiary SVC, 346 body states, 314, 320, 322, 323, 334, 357, 385, 387, 474 borrowing, 75, 107, 206, 213, 216, 234, 333, 338, 415, 468, 476, 478, 479, 482

cardinal numerals, 79, 80, 106, 113, 122, 479 case, 130 causative constructions, 223, 321, 323, 325, 327, 370 cause, 314 cause clauses, 199, 215, 241, 407, 416, 417, 422 circumstantial clauses, 408 clause chaining, 141, 430, 441, 539 clause structure, 201 cleft constructions, 107, 214, 222, 239, 242 cliticisation, 34, 112, 127, 129, 212, 218 codemixing, 467, 485 cognate objects, 81, 245, 311, 322, 398, 535 cohortatives, 173, 174 colour terminology, 482 comitative, 432 comitative SVC, 424 comparative constructions, 156, 195 comparative degree, 91, 191 comparative SVC, 317, 424, 432 complement, 169 complement clause, 261 complement clauses, 171, 332, 334, 368, 394, 433 complementisers, 281, 383 finite, 393 non-finite, 389 complements, 118, 200, 250, 253, 280, 294, 306, 338, 357 copula complements, 255, 259, 282 completive aspect, 152, 155, 157, 165, 186, 385

compounding, 118, 119, 272 compounds nouns, 90 concessive clauses, 406, 476 conditional clauses, 144, 147, 149, 167, 170, 184, 421 consonant clusters, 27, 33 content questions, 63, 223, 226, 242 continuative aspect, 146, 151, 152, 156, 385, 473 continuative intonation, 72, 239, 246, 373, 396, 424, 441 contrastive focus, 226, 227, 234, 235, 238 conversion, 82 coordination clauses, 377 noun phrases, 134 copula identity, 253, 262 locative-existential, 259, 260, 262, 288, 330, 357, 473, 479 copula verb, 258 core participant, 357 core participants, 75, 232, 309, 330, 334, 402 declarative intonation, 63, 69, 70, 100, 376, 424 emphatic, 71 definite article, 213 definiteness, 109, 284, 470 degree modification, 92, 273, 276, 277, 459 deletion of segments, 27, 28 demonstratives, 112, 129, 213, 217, 242, 280 deontic modality, 201, 385, 391 dependent object pronoun, 127 derivation, 82 affixation, 79 conversion, 82 tonal, 79 verbs, 92 direct speech, 377, 379, 403, 414

directives, 167, 174, 201, 389 dislocation, 247 double-object construction, 367 double-object constructions, 198, 350 downdrift, 47, 50 downstep, 47, 50, 56 dummy noun, 261 dummy nouns, 330, 361, 392 effected objects, 360 egressive aspect, 151, 152, 154, 157, 385 emphasis, 57, 64, 73, 91, 107, 112, 150, 193, 199, 204, 205, 226, 246, 276, 344, 445, 446, 460 emphatic intonation, 67 emphatic pronouns, 128, 130 emphatic stress, 64 equative clauses, 251, 330, 411 equative degree, 195 experiencer, 310, 311, 317, 323, 327, 354 expletive, 353, 389, 392 factative TMA,145,146,157,158,160,162, 179, 180, 326, 378, 411, 419, 436 finiteness, 141, 252, 253, 384 focus, 241, 459 force, 325, 369 free relative clauses, 403 fronting, 217, 243, 248 future perfect, 175 future perfect tense-aspect, 164 future tense, 146, 158, 175, 265 general, 282, 291, 294, 426 general locative preposition, 238, 296 generic noun, 132 generic nouns, 86,133, 209, 212, 280, 330, 359 goal, 318, 343, 348, 349, 360, 401 habitual aspect, 9, 116, 139, 140, 144, 146, 150, 151, 154, 162, 168, 175, 176, 179, 185, 362, 378 headless relative clauses, 402

ideophones, 23, 45, 77, 96, 194, 245, 268, 274, 446, 448, 481 imperatives, 8, 143, 171, 172, 200, 225, 368, 454 imperfective aspect, 9, 137, 139–141, 143, 146, 149–152, 154, 156, 157, 161– 163, 166, 169, 180, 245, 386, 404, 408, 411, 419, 432, 437 inalienability, 87 independent possessive pronouns, 130 independent pronouns, 128, 130 indirect questions, 402 indirect speech, 380, 414 infinitive, 146 ingressive aspect, 135, 137, 139, 151–153, 329 insertion of segments, 21, 25, 28 instrument, 399 instrument SVC, 424, 429 interjections, 457 irrealis modality, 141, 184, 535 iterative aspect, 151 jussives, 172 kinship terminology, 84, 381, 451–453, 461, 463, 484 labile verbs, 82, 136, 139, 144, 287, 330, 358, 365, 366, 370 lexical aspect, 135, 139 light verb constructions, 339 list intonation, 67, 68 loan intonation, 67 loan words, 239, 299, 300, 415, 450, 473, 475, 479, 480 locative adverbials, 111, 220, 237, 240, 293 locative clauses, 190, 405, 412 locative nouns, 35, 129, 190, 273, 294, 297, 298, 308, 342, 349 locative prepositions, 281, 297 locative verbs, 136, 290, 310, 323, 326, 330, 437, 547

maleficiary, 129, 348 manner clauses, 255, 409, 411 manner of motion, 293 modal verbs, 167 modality, 9, 75, 141, 166, 185, 373 mood, 139, 164, 178, 435 motion verb, 307 motion verbs, 153, 190, 195, 244, 294, 297, 311, 313, 317, 319, 323, 327, 342, 349, 360, 412, 425 motion-action SVCs, 429 motion-direction SVCs, 427 named place, 290 narrative perfective aspect, 145 negation, 55, 211, 231, 251, 253, 256, 264, 367, 413, 424, 477 negative cleft constructions, 211 negative phrases, 209 non-locative prepositions, 311 noun phrase adverbials, 274, 290, 301 noun phrase structure, 103 number nouns, 115, 116, 119 pronouns, 127, 131 objects, 75, 76, 81, 198, 214, 215, 217, 222, 281, 290, 316–318, 345, 362, 381, 402, 469 obligative mood, 171 order of objects, 346 ordinal numerals, 117, 122 participant-introducing SVCs, 432 partitive, 108, 120, 126, 296 passive, 326, 371 past perfect tense-aspect, 164 past tense, 147, 157, 158, 161, 163, 180, 181, 202, 252, 324, 325, 420 past-before-past tense, 158, 161 patient, 96, 129, 198, 257, 269, 287, 310, 311, 322, 323, 325, 339, 343, 346, 357, 360, 366, 370, 398, 439

perfect progressive, 165 perfect tense-aspect, 157, 165 perfective aspect, 143, 146, 147, 155, 157, 178, 180, 183, 406, 411, 427 pluraliser, 35, 77, 110–112, 130, 213, 397, 468, 469 possession verbs, 136, 197, 264 possessive constructions, 118, 121, 213, 245, 285, 296, 356 possessive pronominal, 130, 352 possessive pronouns, 127, 129 posture verbs, 138, 289, 438 potential mood, 184, 419 predicate cleft, 55, 344 predicate structure, 140 prepositional phrases, 274, 275, 280, 283, 290, 309, 311, 314, 316, 318, 319, 339, 346, 394, 395, 399, 426 present tense, 137, 141, 149, 157, 158, 161, 162, 254, 324, 325, 472 presentational focus, 226, 227 preverbal adverbs, 76–78, 139, 151, 153, 154, 156, 199, 272 progressive aspect, 10 prohibitives, 171, 173 pronouns, 133 propulsion verbs, 293 prosodic focus, 56, 226 prospective aspect, 151, 152, 156, 163 pseudo-cleft constructions, 247 purpose clauses, 169, 170, 247, 295, 368, 372, 377, 394, 407, 416, 417 quantifiers, 130, 131, 201, 209, 213, 217, 221, 298, 299, 301, 396, 470 question intonation, 73, 211 question words, 35, 171, 212, 225, 402 quotative clauses, 199, 223, 377, 382, 391, 414 quotative marker, 384, 389, 392, 414, 416 realis modality, 143 recipient, 198, 269, 311, 316, 345, 347, 348, 399

reciprocity, 95, 298, 317, 352, 365 reduplication, 92, 93, 151, 194, 243, 445 reflexivity, 95, 227, 351, 365 relative clauses, 114, 117, 129, 190, 199, 219, 235, 236, 239, 243, 248, 394, 402, 412 relativisation accessibility, 395, 398 repetition, 55, 80, 92, 93, 100, 243, 245, 446 result clauses, 414, 416 resultative constructions, 306, 326, 328, 365, 436, 440 resultative SVC, 10, 439 resumptive 'go', 428 resumptive 'kan', 428 resumptive imperfective marking, 154, 156, 271 resumptive pronouns, 11, 114, 119, 121, 131,199, 237, 246, 247, 394, 396, 397, 399, 427, 429, 430, 438, 441 secondary predicates, 275, 298, 328, 438 sentence particle, 172 serial verb constructions, 5, 317 similatives, 276, 316, 476 stimulus, 310–312, 314, 320, 322 stranding, 76, 191, 214, 216, 280, 395, 399, 401 subject, 357 subject omission, 357, 377, 380 subjects, 184, 215, 310, 317, 318, 321, 323– 325, 328, 330, 346, 352–355, 369, 370, 386, 387, 390, 399, 412, 413, 427–429, 432, 435, 437, 439, 440 subjunctive mood, 166, 169, 174, 201, 203, 223, 238, 253, 368, 391, 394, 414, 416 subordination, 366, 373 subordinator, 233, 393, 407, 417 superlative degree, 194 switch-function (pro)nouns, 432

tense, 9, 75, 139–141, 163, 164, 166, 178, 253, 324, 373, 425, 435, 440 theme, 198, 287, 296, 297, 310, 311, 317, 318, 321, 345, 346, 348, 430 topic marker, 118, 130, 249 transitivity, 323 valency, 357 valency adjustments, 373 verb classes, 334 vocatives, 129, 227, 450, 454, 460 word class, 279 word classes, 77, 80, 82, 92, 96, 97, 124, 236, 266, 443, 449, 450 word order, 188, 198, 236, 309, 310 word orderes, 77

yes-no questions, 62

# Did you like this book?

This book was brought to you for free

Please help us in providing free access to linguistic research worldwide. Visit http://www.langsci-press.org/donate to provide financial support or register as a community proofreader or typesetter at http://www.langsci-press.org/register.

## A grammar of Pichi

Pichi is an Afro-Caribbean English-lexifier Creole spoken on the island of Bioko, Equatorial Guinea. It is an offshoot of 19th century Krio (Sierra Leone) and shares many characteristics with West African relatives like Nigerian Pidgin, Cameroon Pidgin, and Ghanaian Pidgin English, as well as with the English-lexifier creoles of the insular and continental Caribbean. This comprehensive description presents a detailed analysis of the grammar and phonology of Pichi. It also includes a collection of texts and wordlists. Pichi features a nominative-accusative alignment, SVO word order, adjective-noun order, prenominal determiners, and prepositions. The language has a seven-vowel system and twenty-two consonant phonemes. Pichi has a two-tone system with tonal minimal pairs, morphological tone, and tonal processes. The morphological structure is largely isolating. Pichi has a rich system of tense-aspect-mood marking, an indicativesubjunctive opposition, and a complex copular system with several suppletive forms. Many features align Pichi with the Atlantic-Congo languages spoken in the West African littoral zone. At the same time, characteristics like the prenominal position of adjectives and determiners show a typological overlap with its lexifier English, while extensive contact with Spanish has left an imprint on the lexicon and grammar as well.