# A grammar of Choguita Rarámuri

In collaboration with Luz Elena León Ramírez, Sebastián Fuentes Holguín, Bertha Fuentes Loya and other Choguita Rarámuri language experts

Gabriela Caballero

Comprehensive Grammar Library 5

### Comprehensive Grammar Library

Editor: Martin Haspelmath

In this series:


This series grew out of the grammars published in *Studies in Diversity Linguistics*, which are proudly mentioned:


ISSN (print): 2748-971X ISSN (electronic): 2749-7798

# A grammar of Choguita Rarámuri

In collaboration with Luz Elena León Ramírez, Sebastián Fuentes Holguín, Bertha Fuentes Loya and other Choguita Rarámuri language experts

Gabriela Caballero

Gabriela Caballero. 2022. *A grammar of Choguita Rarámuri: In collaboration with Luz Elena León Ramírez, Sebastián Fuentes Holguín, Bertha Fuentes Loya and other Choguita Rarámuri language experts* (Comprehensive Grammar Library 5). Berlin: Language Science Press.

This title can be downloaded at: http://langsci-press.org/catalog/book/324 © 2022, Gabriela Caballero Published under the Creative Commons Attribution 4.0 Licence (CC BY 4.0): http://creativecommons.org/licenses/by/4.0/ ISBN: 978-3-96110-399-7 (Digital) 978-3-98554-057-0 (Hardcover)

ISSN (print): 2748-971X ISSN (electronic): 2749-7798 DOI: 10.5281/zenodo.7189161 Source code available from www.github.com/langsci/324 Collaborative reading: paperhive.org/documents/remote?type=langsci&id=324

Cover and concept of design: Ulrike Harbort Typesetting: Gabriela Caballero, Sebastian Nordhoff Proofreading: Martin Haspelmath Fonts: Libertinus, Arimo, DejaVu Sans Mono Typesetting software: XƎLATEX

Language Science Press xHain Grünberger Str. 16 10243 Berlin, Germany langsci-press.org

Storage and cataloguing done by FU Berlin










# **Abbreviations**


# **Acknowledgments**

I would like to express my deepest gratitude to all the Choguita Rarámuri language experts who patiently and generously invested large amounts of time into teaching me about Rarámuri language and culture. I am especially grateful to elder †Luz Elena León Ramírez, Sebastián Fuentes Holguín, and Bertha Fuentes Loya for their deep involvement in this project and constant encouragement over the years. I am also grateful to María Del Rosario Cervantes Guerrero, Rosa Isela Chaparro Gardea, María Guadalupe Diaz, elder Mateo Espino, Santos Fuentes, †Rosa Fuentes, elder †José María "Chémale" Fuentes, Miguel Angel Fuentes Diaz, Virginia Fuentes Diaz, Angelina Fuentes Diaz, Reyes Fuentes Guerrero, elder †Cornelio Fuentes Hernández, elder †Morales Fuentes Hernández, Carlos Fuentes Moreno, Guillermina Fuentes Moreno, Valentina Fuentes Moreno, Vicente Fuentes Moreno, Giltro Fuentes Palma, Teresa Guerrero Herrera, elder María Dolores Holguín, Javier Holguín Fuentes, Jesusita Loya Guerra, elder †Federico León Pacheco, Francisco Moreno Fuentes and Sebastián Moreno Morales for their contributions as authors, consultants, teachers and creators of language documentation. Many community members also welcomed me into their homes, healing ceremonies, trips to neighboring communities, house parties, graduations, gatherings for races, and harvest celebrations. These relationships contributed immensely to both my understanding of Rarámuri language and culture and my enjoyment of life in Choguita.

All language experts I worked with made crucial contributions to the analyses presented in this grammar. That this grammar is published under my name reflects the fact that these analyses are written in my words. However, this should not be taken to assert authority over the Choguita Rarámuri language itself. The language, the examples of its use quoted in this grammar, and the knowledge expressed therein remain the intellectual property of the Rarámuri people.

Michael Casaus first introduced me to the community of Choguita, and shared local knowledge and advice. I learned a great deal from his high standards in his work, and benefited from the strong ties he developed in Choguita.

Some portions of this grammar were written as part of my doctoral dissertation at UC Berkeley, where I was incredibly fortunate to have been supervised

### Acknowledgments

and guided by Andrew Garrett, Sharon Inkelas and Johanna Nichols. I am grateful to them for playing a crucial role in my understanding of the Choguita Rarámuri language and how it should be best described. Many further improvements were based on feedback and discussions with many colleagues in subsequent years. I am particularly indebted to (in alphabetic order) Andrés Aguilar, Karen Dakin, Marc Garellek, Austin German, Alice Harris, Jason Haugen, Vsevolod Kapatsinski, Enrique Palancar, Bert Remijsen, Sharon Rose, †Enrique Servín, Maziar Toosarvandani and Verónica Vázquez Soto. Johanna Nichols, in particular, gave me enormously valuable feedback in later states of the writing of this grammar. I am also grateful to Lucien Carroll, who contributed greatly to description and analyses of the language as a doctoral research assistant at UCSD, collaborator and co-author. He also developed the web-based user interface that allows access to corpus materials for both community and academic audiences.

I benefited a great deal from Martin Haspelmath's thorough reading of this manuscript, which produced many constructive comments and suggestions that eradicated many errors. I am extremely grateful to have had such a wise, meticulous and kind editor. I am also grateful to all anonymous reviewers for their valuable feedback and to Sebastian Nordhoff for his patience and efficiency in providing support with technical and typesetting matters at Language Science Press.

My research on Choguita Rarámuri was funded by Mexico's Consejo Nacional de Ciencia y Tecnología (CONACYT), the University of California Institute for Mexico and the United States (UCMEXUS), the UC Berkeley Linguistics Department, UC Berkeley's Survey of California and Other Indian Languages, the Endangered Languages Documentation Programme (ELDP) (under awards IGS0042 and IPF0138), and the National Science Foundation/Documenting Endangered Languages program (NSF-DEL) (under award 1160672).

Finally, I would like to thank my family and friends for their unwavering support throughout this process. In particular, I am grateful to Cecilia Hernández Carrillo, †María del Carmen Carrillo Velázquez and †Aristeo Hernández León for their love, encouragement and support since the very beginning. Without them, this book would not be possible. And to Paloma, for being a constant source of joy and for sharing my love of Choguita and the Sierra Tarahumara.

# **1 Introduction**

### **1.1 Linguistic profile of Choguita Rarámuri**

Despite the relevance of the Uto-Aztecan language family in terms of its geographical extension, number of languages, number of speakers and its descriptive tradition (spanning over four centuries), there are still many important gaps in our knowledge of this language family. This book provides the first comprehensive grammatical description of Choguita Rarámuri, a Uto-Aztecan language spoken in the Sierra Tarahumara, a mountainous range in the northern Mexican state of Chihuahua belonging to the Sierra Madre Occidental. The number of speakers of all Rarámuri varieties in the Sierra Tarahumara and diaspora communities in Northern Mexico is estimated to be between 85,000 (INALI 2008) to 100,000 (Embriz Osorio & Zamora Alarcón 2012; Merrill & Burgess 2014). Choguita Rarámuri is the native language of approximately 1,000 people, most of whom use the language in their daily interactions, though continued political violence in the area has led to increased displacement of the Rarámuri people and consequent accelerated language attrition. As documented since the first contact with European settlers, the Rarámuri people continue to resist the encroachment of outsiders on their land and violations of their autonomy as a sovereign nation.

In the remainder of this introductory Chapter, I provide a typological profile of Choguita Rarámuri (§1.1.1), I discuss the genetic relationship between Choguita Rarámuri and other language varieties (§1.1.2), and I survey previous work on the language (§1.1.3). The following sections focus on the geographical (§1.2), historical and socio-political (§1.3) contexts Choguita Rarámuri is spoken within. The history of contact with Europeans is addressed in §1.3.1, §1.3.2 addresses the linguistic ecology and the current socio-political context of the area, while §1.3.3 considers the development of literacy and bilingual education projects sponsored by the Mexican government.

This chapter concludes in §1.4 with a meta-description of this grammar, which includes a description of how this project developed (§1.4.1), theoretical assumptions made in this work (§1.4.2), the data it draws upon and methodologies used to gather it (§1.4.3), the Choguita Rarámuri language experts who contributed to the data and analysis (§1.4.4), and conventions employed in the presentation of

### 1 Introduction

examples (§1.4.5). Finally, this chapter provides an overview of the topics covered in this grammar and the content covered in each chapter (§1.5).

### **1.1.1 Choguita Rarámuri in typological context**

Choguita Rarámuri is a highly synthetic, agglutinating language with a complex morphological system. It displays many of the recurrent structural features documented across Uto-Aztecan, including a predominance of suffixation, headmarking (as defined in Nichols (1986)), and patterns of noun-incorporation and compounding (Sapir 1921; Whorf 1935; Haugen 2008b). Other recurrent Uto-Aztecan features present in Choguita Rarámuri include a complex word prosodic system, a wide range of morphologically-conditioned phonological processes, and exuberant derivational morphological marking, all of which are grammatical phenomena that are of great typological and theoretical interest. Choguita Rarámuri is also of great comparative/historical importance: while several analytical works of Uto-Aztecan languages of Northern Mexico have been produced in the last years (Guerrero Valenzuela 2006, García Salido 2014, Reyes Taboada 2014, Morales Moreno 2016, Villalpando-Quiñonez 2019, *inter alia*), many varieties still lack comprehensive linguistic description and documentation.

In traditional morphological typology, agglutinative morphologies are placed within a scale of agglutination-flexion, with isolating languages on one end of the spectrum and introflexive (or non-linear) languages at the other (isolating > agglutinative > flexive > non-concatenative (or introflexive)). This single scalar hierarchy results from conflating the parameters of phonological fusion, the degree to which individual exponents are phonologically fused to their host (isolating > concatenative > non-concatenative), and flexivity, the degree to which individual exponents exhibit lexically-conditioned variance (suppletive allomorphy) (Bickel & Nichols 2007). Agglutinative languages thus involve concatenative, non-flexive morphological patterns. Uto-Aztecan languages have been described as prototypically agglutinative, with complex verbal morphological systems, a high degree of synthesis, a low degree of phonological cohesion between exponents, and a low degree of cumulation in morphological exponence (see e.g., Langacker 1977: 158). Choguita Rarámuri morphology, which is also highly synthetic and almost exclusively suffixing, displays the following agglutinative-like properties (1):

	- Mostly concatenative, separative exponence
	- Limited flexive exponence
	- Zero exponence

1.1 Linguistic profile of Choguita Rarámuri


While Choguita Rarámuri shares several morphological and morpho-phonological properties and phenomena with other morphologically complex languages that have been characterized as agglutinative, it also crucially departs from the canonical "agglutinative" type in that it has less transparent morpheme boundaries, due to a fair amount of phonological cohesion between exponents closer to the stem, a pattern more frequently attested in flexive morphological patterns (Caballero 2008; Caballero & Kapatsinski 2021). Departures from prototypical agglutinative characteristics are documented elsewhere in Uto-Aztecan (e.g., Cupeño, a Takic language; Hill 2005).

Description of particular word prosodic systems in Uto-Aztecan languages has traditionally received a great deal of attention in the arealist literature, as well as in the typological and theoretical literature, given their complexity. Uto-Aztecan languages have been documented to have a wide variety of stress-accent systems. Some recurrent features of these systems include presence of lexical stress, iterative stress assignment, left-edge alignment, and window restrictions (Munro 1977). One of the most typologically unusual features of the word-prosodic system of Choguita Rarámuri is its initial three-syllable stress window. Stress is assigned within the first three syllables of the word, and there are alternations that maintain this ternarity, such as compounding and multiple affixation (Caballero 2008; 2011b). This kind of stress system is extremely uncommon crosslinguistically, described in the literature in only four other languages of the world (Kager 2012), and predicted to be unattested by some factorial typologies (Elenbaas & Kager 1999).

The stress system of Choguita Rarámuri is also a topic of particular interest given its morphological conditioning. Although stress is lexically governed in some morphemes, there is also emergent default stress assignment. As described for other Uto-Aztecan languages (e.g. Cupeño, Hill & Hill (1968); Alderete (2001)), there is a contrast between underlyingly stressed and unstressed roots in Choguita Rarámuri. Affixes, on the other hand, are either stress-shifting or stressneutral, meaning that they can perturb the root's stress or be neutral regarding stress assignment, respectively. The interaction between roots and affixes of different prosodic types yield complex interactions in verbal paradigms: in words containing no underlyingly stressed roots or stress-shifting suffixes, stress falls by default in the second syllable of the stem; in words containing an underlyingly

### 1 Introduction

unstressed root and a stress-shifting affix, stress falls by default in the third syllable of the word (the immediately adjacent stress-shifting suffix with disyllabic roots). This is exemplified below in Table 1.1 (stressed syllables are in boldface).


Table 1.1: Stress patterns of morphologically complex verbs

These morphological restrictions on stress assignment interact in complex ways with the trisyllabic stress window restriction in compound and other morphological constructions (yielding stress shifts and truncation of nominal roots), and with two subsystems of valency encoded through root allomorphy marked through vocalic alternations and fixed stress for applicative derivation (Caballero 2008).

In addition to stress, the Choguita Rarámuri word prosodic system involves tone. While the development of tonal contrasts has been documented for a number of Uto-Aztecan languages (including Northern Tepehuan (Tepiman; Woo (1970)), Hopi (Manaster-Ramer 1986), Huichol (Corachol; Grimes (1959)), Cahitan (Yaqui (Demers et al. 1999) and Mayo (Hagberg 1989)); and Balsas Nahuatl (Aztecan; Guion et al. (2010)), no variety of Rarámuri, to the best of my knowledge, had been previously described as featuring a tonal contrast. Except for Balsas Nahuatl, tonal Uto-Aztecan languages are located in the Southwest, a linguistic area where tone has also developed in other language families (Caballero & Gordon 2020).

The tone system of Choguita Rarámuri has a restricted distribution, as lexical tone contrasts are exclusively realized on surface stressed syllables (i.e., there is a single lexical tone per prosodic word). This restricted distribution of tone in Choguita Rarámuri is a characteristic shared by all other tonal Uto-Aztecan languages, all of which have "hybrid" word prosodic systems (featuring both stress and tone). The Choguita Rarámuri lexical tone inventory features a threeway contrast (/HL/, /L/ and /H/) in stressed syllables (Caballero & Carroll 2015),

### 1.1 Linguistic profile of Choguita Rarámuri

which contrasts with the binary tone systems documented in neighboring Cahitan languages (Mayo and Yaqui). Stress and tone in Choguita Rarámuri are not only phonologically independent systems, but they are also acoustically distinct: duration and intensity are correlates of stress, whereas fundamental frequency (f0) is the primary correlate of tone (Caballero & Carroll 2015).

Tone also plays a role in the morphological system of Choguita Rarámuri: tone alone may be the exponent of morphological information, it may be associated with specific affixes as a morphologically-conditioned effect, or it may have a morphomic distribution in verbal paradigms (Caballero & German 2021). This type of grammatical tone phenomena has not yet been documented in other tonal Uto-Aztecan languages.

In addition to encoding lexical and morphological contrasts, f0 is also deployed in the Choguita Rarámuri intonation system. Declarative sentences are characterized by a high boundary tone, which stands in contrast to the general crosslinguistic tendency for low boundary tones in these types of sentences (Jun 2014). Analysis of acoustic data reveals that lexical tones may be enhanced in utterancefinal position and that both lexical and grammatical tones take precedence over phrase-level tones if these tones conflict (Caballero et al. 2014, Aguilar et al. 2015, Garellek et al. 2015). In addition to f0, Choguita Rarámuri implements intonational contrasts through duration and non-modal phonation (Aguilar et al. 2015; Caballero et al. 2022). Both tone and intonation are under-studied for Uto-Aztecan languages, except for recent work that addresses Nahuatl language varieties (Guion et al. 2010; Patiño Velázquez 2014; Aguilar 2020).

In terms of its morphological system, Choguita Rarámuri exhibits a case of free affix order where alternative orders are determined by semantic scope, templatic constraints, phonological subcategorization and phonological conditions on stem shape (Caballero 2010). Choguita Rarámuri also exhibits complex patterns of multiple (extended) exponence (ME), a one-to-many mapping between morphological categories and their formal expression. The Choguita Rarámuri case offers an important opportunity to explore the properties of this morphological phenomenon, given that ME in Choguita Rarámuri: i) involves derivational information (*contra* suggestions that ME is exclusively displayed by markers of inflectional categories (Matthews 1972, Stump 2001)); and ii) that it involves categories in specific areas of the layered structure of the verb with characteristic morpho-prosodic properties which make them difficult to parse and prone to be reanalyzed as part of the stem (Caballero 2008, Caballero 2011b). Choguita Rarámuri also provides a relevant testing ground for investigating the potential functional role complex morpho-phonological patterns in morphologically complex languages may have (Caballero & Kapatsinski 2015; 2021).

### 1 Introduction

Finally, Choguita Rarámuri is also typologically relevant given a rich set of valency-changing morphology, and a complex system of case marking and lexical distinctions to refer to spatial and topographic properties of the landscape, also documented for closely-related Guarijío (Miller 1996). These and other aspects of the structure of Choguita Rarámuri are addressed in this reference grammar.

### **1.1.2 Rarámuri language varieties and genetic/genealogical relationships**

### **1.1.2.1 Dialect variation**

Rarámuri dialect diversity has not yet been systematically investigated, but a local-government sponsored dialect survey carried out two decades ago yielded a classification with five dialect areas: North (*Norte*), South (*Sur*), Central (*Centro*), West (*Oeste*) and Highland (*Cumbres*) (Valiñas 2001; see also INALI (2008)).<sup>1</sup> This classification was established on the basis of the assessment of lexical, phonological and syntactic variation.<sup>2</sup> These five dialect areas roughly correspond to the dialect areas proposed by the SIL International published in *Ethnologue*, labeled Central, Western, Northern, Southeastern and Southwestern. There is a high degree of overlap between these classifications, but there is no consensus about the precise boundaries of each dialect: the Ethnologue's Central Tarahumara dialect, for instance, corresponds to an area occupied by two dialects in the local government survey, Central Tarahumara and Northern Tarahumara (Valiñas 2001). The Mexican National Institute for Indigenous Languages (INALI)<sup>3</sup> adopts the classification proposed in Valiñas (2001). Table 1.2 presents the two classifications side by side. The classification adopted by INALI also provides endonyms for each variety (INALI 2008).

An uncontroversial main distinction exists between two major sets of dialects, Tarahumara de la Alta (Rarámuri) and Tarahumara de la Baja (Rarómari), which may include mutually unintelligible varieties. Choguita Rarámuri is part of the Alta Tarahumara dialect continuum within a "Central" dialect area within the Ethnologue's classification ([tar]; Eberhard et al. 2021) and is located in a transitional area between the Central Tarahumara and Northern Tarahumara dialects in (Valiñas 2001).

<sup>1</sup>Merrill & Burgess (2014) translate the term *Cumbres* as "Interior", noting this dialect area does not necessarily correspond with areas that are at a higher altitude, which is implied by the term *Cumbre*.

<sup>2</sup> Some of the phonological parameters include: use of word-initial [g], [k] or zero; initial syllable truncation; word final vowel deletion; height neutralization of /e/; and pre-aspiration of voiceless stops (Valiñas 2001: 122).

<sup>3</sup> *Instituto Nacional de Lenguas Indígenas*.

### 1.1 Linguistic profile of Choguita Rarámuri


Table 1.2: Rarámuri dialect areas

Speakers are aware of dialect differences, but view all Rarámuri varieties as a single language, different from the neighboring languages. The location of several modern Rarámuri communities and closely related Guarijío (Warihío) is shown in Figure 1.1. Choguita is located 35km northwest of Norogachi. More comprehensive assessments are necessary to determine the number and location of all varieties of Rarámuri, degrees of intelligibility amongst these, and number of speakers per variety.

### **1.1.2.2 Alternative names**

*Rarámuri* [raˈɽámuɽi] is the name applied by the Rarámuri people to their own language, as well as the Rarámuri people, land and culture. The phrase [raˈɽámuɽi raˈʔìtʃ͡a] ('Rarámuri language') is used in discourse when disambiguating reference to the language vs. other possible meanings. When referring to people, Rarámuri is the term often used to refer to members of the Rarámuri nation, as opposed to *mestizo* (non-indigenous Mexican) and other non-Indigenous people, but this term has different meanings that are dependent on context. Specifically, Merrill (2001) identifies four levels of self denomination of the term "Rarámuri": (i) all human beings; (ii) indigenous people (vs. *mestizo* and people of European descent); (iii) Rarámuri people (vs. other indigenous people); and (iv) Rarámuri men (vs. Rarámuri women) (2001: 88).

Rarámuri has been mostly known as *Tarahumara* in mass media and previous descriptions and depictions of the language (e.g., the ISO code of the language is [tar]). The term *Tarahumara* was first used in the seventeenth century in the correspondence of Catholic missionaries and the first published works about the language, Tomás de Guadalajara's 1683 grammar and Matthäus Steffel's 1791 dictionary. The term *Rarámuri* was not used in published materials until Miguel de Tellechea's 1826 *Compendio gramatical para la inteligencia del idioma tarahumar* (where the spelling used for the language was *rarámari*) (Merrill 2001: 77).

### 1 Introduction

Figure 1.1: Modern Rarámuri communities and neighboring Guarijío (Warihio) territory (map adapted from Merrill & Burgess 2014: 232).

### 1.1 Linguistic profile of Choguita Rarámuri

The term *Rarámuri* is one of several spellings found in written media, including school textbooks and other texts written in Rarámuri, as well as publications by the Mexican government's National Institute of Indigenous Languages, INALI. In INALI's publications, Rarámuri varieties are referred to by their endonyms, listed in Table 1.3 above. The variation in the spelling of the language is rooted in the phonological inventory of the language: as discussed in Chapter 3 below, Choguita Rarámuri (like other Rarámuri varieties), has two contrastive liquid sounds, an alveolar flap that resembles the Spanish coronal flap, and a lateral flap, which auditorily resembles both a flap and a lateral (with certain phonological environments favoring one or the other impresionistically) (see §3.2). The name of the language (/ɾaˈɽamuɽi/ [raˈɽamuɽi]) in Choguita Rarámuri features an alveolar flap word-initially (allophonically a trill) and two lateral flaps wordmedially, leading to orthographic representations of the lateral flaps as either 'r' or 'l' (*Rarámuri* or *Ralámuli*). It is an open question whether other Rarámuri varieties feature this contrast. In addition to spelling variations concerning the liquid consonants of the language, there is also variation in vocalic segments that reflect dialect differences in terms of phonemic vocalic inventories and processes targeting vowels in each variety.

While the name of the language is represented orthographically in this grammar with a single symbol <r>, lateral flaps are uniformily represented with [l] in the data examples presented in this grammar.

### **1.1.2.3 Genealogical affiliation**

Rarámuri belongs to the Uto-Aztecan (UA) language family, which spans several cultural areas of the North American continent and has a time-depth of between 4000 and 5000 years (Campbell 2000; Fowler 1983; Hill 2010; Silver & Miller 1998). The Uto-Aztecan language family is one of the largest language families in the Americas in terms of geographical extension, ranging from the Great Basin (where Numic varieties are spoken) to El Salvador and Nicaragua (where varieties of Aztecan are spoken) (Miller 1984; Campbell 2000; Mithun 2001). The location of each of the individual branches of the Uto-Aztecan language family is illustrated in the map in Figure 1.2.

Subgrouping within the Uto-Aztecan language family has been the subject of a long debate in the literature (Hill 2011). Rarámuri and closely related Guarijío are dialect continua that form two branches of a larger, uncontested Tarahumara-Guarijío or Tarahumaran branch.<sup>4</sup> The Tarahumaran branch is generally classified within a Taracahitic subgroup (Langacker 1977, Campbell 2000, Mithun 2001,

<sup>4</sup>Referred to as "Tarawarihian" in Merrill & Burgess (2014)

### 1 Introduction

adapted from https://upload.wikimedia.org/wikipedia/commons/e/e4/Uto-Aztecan\_map.svg CC-BY-SA https://commons.wikimedia.org/wiki/User:Noahedits

Figure 1.2: Uto-Aztecan language subgroups

### 1.1 Linguistic profile of Choguita Rarámuri

Opata† Eudeve†

Miller 1984), and as part of a larger Sonoran branch within Southern Uto-Aztecan (Miller 1984, Hill 2001). The place of Rarámuri within a traditionally assumed subgrouping of Uto-Aztecan is shown in Figure 1.3 (Campbell 2000, Mithun 2001, Hill 2011). There is currently no consensus as to the genealogical status of Southern Uto-Aztecan and Taracahitan (for recent discussion see Hill 2011 and Merrill 2013), and this is indicated with parentheses in these branches. Figure 1.4 illustrates the languages within the traditionally assumed Taracahitan branch.

Figure 1.4: Taracahitan branch (adapted from Campbell 2000)

**Rarámuri** Guarijío

### **1.1.3 Previous work**

Yaqui Mayo

Rarámuri varieties have been described since the seventeenth century in the form of grammars, dictionaries, vocabularies and texts. The first known documentation of Rarámuri is by Tomás de Guadalajara, a Jesuit missionary who worked in the missions in the Sierra Tarahumara in 1675, and published a brief grammatical description in 1683, *Compendio del arte de la lengua de los tarahvmares y guazapares*. This small grammatical description was followed more than one hundred years later by Matthäus Steffel's publication in 1791 of a German-Tarahumara dictionary based on German orthography.

After these early grammatical descriptions, most existing documentation of Rarámuri has been produced during the twentieth century. This includes several

### 1 Introduction

grammatical descriptions, dictionaries, vocabularies and texts for several varieties, mainly of Norogachi Rarámuri, a Northern variety (Brambila 1953, Brambila 1976, Brambila 1983, Lionnet 1968, Lionnet 1972, Lionnet 1985). The most comprehensive of these works is David Brambila's (1953) grammar, which is written in the style of colonial grammars, but also provides many examples from texts.<sup>5</sup>

In addition to these works on Norogachi Rarámuri, several linguistic articles, grammars and short manuscripts about diverse aspects of different dialects of the language were published in the second half of the twentieth century, including descriptions of Samachique Rarámuri (a Central variety) (Hilton 1993), and of Western Rarámuri (Burgess 1970, Burgess 1984, Burgess & Merino Mares 2001, *inter alia*), which differs significantly from Northern and Central varieties. More recent linguistic work includes a BA thesis on the morphosyntax of property concepts in Choguita Rarámuri (Islas Flores 2010), a PhD thesis on Urique Tarahumara syntax (Jara 2013), a master's thesis analyzing basic clause structure and other syntactic aspects of Rochéachi Rarámuri (Morales Moreno 2016), and a PhD thesis focusing on the analysis of grammatical aspect of Norogachi Rarámuri (Villalpando-Quiñonez 2019). A pedagogical grammar of Rarámuri was published by the state government of Chihuahua by Enrique Servín Herrera in collaboration with Rarámuri poet and activist Dolores Batista, a native speaker of Ojachichi Rarámuri (Servín 2002).

Other records of Rarámuri are found in ethnographic studies that have documented ethnobotanical and historical knowledge (Bennett & Zingg 1935; Bye et al. 1975; Bye 1976; Merrill 1988; Pintado Cortina 2012a), as well as audio recordings of language and traditional music. Audio recordings are housed at the governmentowned regional radio station, Radio XETAR, where they have been broadcasted since 1982 in Rarámuri, Northern Tepehuan and Pima.<sup>6</sup> To the best of my knowledge, many of the audio recordings that document speech are scripted and do not constitute a representative sample of patterns of spontaneous language use.

In addition to the publications that have a linguistic focus, much work has been carried out in the state of Chihuahua documenting traditional narratives, poetry and other forms of verbal art of the Rarámuri nation by Rarámuri poets and language activists Dolores Batista and Martín Makawi, and by Enrique Servín Herrera, linguist, activist and poet in charge of a Program for Indigenous and Minority Languages hosted by the state government's Cultural Development

<sup>5</sup>A list of published references on the Rarámuri language is provided in Appendix 1 of this grammar.

<sup>6</sup>The Radio XETAR is part of a government office program of indigenous radio stations. The governemnt office is the National Comission for the Development of Indigenous Peoples (*Comisión Nacional para el Desarrollo de los Pueblos Indígenas*, or CDI).

### 1.2 Geographic location and physical environment

Figure 1.5: Location of Choguita Rarámuri and neighboring Uto-Aztecan languages

Office.<sup>7</sup> Their efforts have produced several publications in different Rarámuri varieties and both bilingual and monolingual books and resources for Rarámuri speakers and second language learners (Batista 1994, Servín 2002, Makáwi 2012, inter alia).

### **1.2 Geographic location and physical environment**

CC BY-SA 3.0

Rarámuri is spoken in the southwestern part of the Mexican State of Chihuahua, a rugged area in the Sierra Tarahumara that includes the Copper Canyon. The Sierra Tarahumara is part of the Sierra Madre Occidental, a mountain range that extends from the Southwest United States to Central Mexico, with an area of approximately 50,000 square kilometers (Pintado Cortina 2012b). This grammar describes variety spoken in Choguita (also known as "Choguita de Guachochi" or "Choguita de Norogachi"), in the municipality of Guachochi. Geographically adjacent Uto-Aztecan languages include Guarijío (or Warihó), Yaqui (or Hiaki), Mayo, Northern Tepehuan and O'ob Nook Pima. The location of Choguita Rarámuri and neighboring Uto-Aztecan languages in the Mexican Northwest is shown in the map in Figure 1.5.

<sup>7</sup> *Programa Institucional de Atención a las Lenguas Indígenas y Minoritarias.*

### 1 Introduction

The community of Choguita is part of the *ejido* system, a Mexican land usage system, where rural land plots are devoted for collective use by community members (*ejidatarios*). Choguita is the head community of the *ejido* of Choguita, one of the largest *ejidos* in the Sierra with a total surface area of 285.6 km<sup>2</sup> (28,560 hectares, 110.3 miles<sup>2</sup> ) (Casaus 2008). Figure 1.6 shows the boundaries of the ejido of Choguita, which includes the communities of Bokimoba, Huichachi, Capochi, Basigochi, Coechi, Rayabó, Cochirachi, Ireachi, Rorichi, Rochibo, Upachi, Sehuarachi and Cocohuichi. According to the Mexican government 2000 national census, the community of Choguita has 234 inhabitants, and the entire population of the *ejido* is approximately 1050 (Casaus 2008). With the exception of secondary school teachers and Protestant missionaries, the community is native Rarámuri.

Figure 1.6: Boundary of the ejido of Choguita (Topographic map, Casaus 2008)

### **1.3 Choguita Rarámuri in social and historical context**

This section discusses the historical, social and political contexts in which Choguita Rarámuri is used. First, §1.3.1 addresses the history of contact between the

### 1.3 Choguita Rarámuri in social and historical context

Rarámuri people and Europeans (and after the establishment of the Mexican state, Mexicans of European descent, referred to as *chabóchi* (a Rarámuri word) in the Sierra Tarahumara). This is followed by discussion of Rarámuri language use in §1.3.2, including bilingualism and the political and social factors affecting speech communities since the twentieth century. This section concludes in §1.3.3, which focuses on programs sponsored by the Mexican government to promote so-called "bilingual /bicultural" education in the sierras and standardization of the language, as well as emerging use of the language in social media by native speakers.

### **1.3.1 History of contact with Europeans**

The earliest contact between the Rarámuri and Europeans was in the late sixteenth century (Merrill & Burgess 2014). The first incursion of Jesuit missionaries took place in 1607, with the arrival of Joan Fonte to the Valle de San Pablo, an area in the border between the land of the Rarámuri and the Odami (Tepehuan) nations in what is now known as the district of Balleza (Alegre 1767; Pintado Cortina 2012a).<sup>8</sup>

Ethnographic studies report that at the time of contact the Rarámuri lived in small settlements (villages of 5 to 20 households) relatively spread out across an area of approximately 45,000 square kilometers in the southwestern and central area of what is now the state of Chihuahua (Pennington 1963; Merrill & Burgess 2014). Figure 1.7 shows the location and extension of the Rarámuri territory in the seventeenth century.

European impact has been felt severely in the area for the past 400 years. The campaign of religious conversion and political control that began in the seventeenth century was met with strong resistance from the Rarámuri, who protested and fought back the invasion of their land. Four major rebellion movements are documented between 1648 and 1697, in which many missions were destroyed (Neumann 1991; González Rodríguez 1982; Levi 1999; González Rodríguez et al. 1991). For much of the seventeenth century, the conflict pushed many Rarámuri out of their communities, who took refuge in other areas, including the southwestern region of the Sierra. The Rarámuri population was significantly decimated by the violence inflicted by settlers and also epidemics that afflicted the population across the whole area (Pintado Cortina 2012a).

<sup>8</sup>Historical accounts report that this area was also inhabited by Guazápares, Chínipas, Témoris, Guarijíos, Jovas, Pimas, Conchos, Janos, Julimes, Chinarras, Tobosos, Acoclames, Chizos, Tubares, Tzoes and Cocoymes, and members of the N'dee/N'nee/Ndé (Apache) nation (Neumann 1991: 53, González Rodríguez 1987: 368; cited in Pintado Cortina 2012a).

### 1 Introduction

<sup>©</sup> OpenStreetMap contributors, Open Database License

Figure 1.7: Territory of the Rarámuri nation in the seventeenth century adapted from Merrill & Burgess (2014: 230)

The Rarámuri continued to resist occupation of their land during the following centuries. Population movements to isolated areas to avoid encroaching settler populations became the main form of resistance (Merrill 1983; Levi 1999). Some Rarámuri communities adopted some of the social and religious norms of life in the missions. This process began prior to the expulsion of Jesuits from the Americas in 1767, after which there was another period of relative isolation of Rarámuri communities from colonial institutions (Pintado Cortina 2012a). The Rarámuri from these communities self-describe as *Pagótame* 'baptized ones' (literally 'washed ones'). It should be noted that adoption of foreign religious norms was only a partial process, as *Pagótame* practice a form of syncretic Christianity (Merrill 1983; 1988). Other Rarámuri communities, the *Simalóni* (*cimarrón*) <sup>9</sup> or 'gentile' (unbaptized), resisted adopting any form of Christianity and retreated to the most remote areas of the Sierra.

A new stage in the interaction between the Rarámuri and Europeans began in the nineteenth century with the founding of the Mexican nation state in 1810

<sup>9</sup>The term *Cimarrón* was used during that time period to refer to people from African descent who escaped slavery and founded free Afro-Mexican towns; the missionaries used this term to refer to Rarámuri people who escaped forced labor and religious conversion imposed in the missions (Pintado Cortina 2012a: 53).

### 1.3 Choguita Rarámuri in social and historical context

and the state of Chihuahua in 1823. Shortly after the founding of the state of Chihuahua, a new Law of Colonization (in 1825) allowed Mexican *mestizo* settlers to buy land that had belonged to the missions, causing a new wave of displacement of the Rarámuri, who sought to keep their political autonomy (Pintado Cortina 2012a). Mining became a dominant economic activity in the late nineteenth century for several decades, and an intense and unregulated exploitation of the forests began in the middle of the twentieth century. In addition to this, the Mexican revolution brought about changes to government institutions in the early twentieth century, including reforms in the land ownership system and the creation of *ejidos*, as discussed in §1.2. Together, these factors led to major shifts in the economic and political life of the Sierra in the last century. The effects of these shifts in terms of the attempts to assimilate the Rarámuri through government-sponsored education programs is discussed in §1.3.3 below.

### **1.3.2 Linguistic ecology and sociopolitical context**

At the beginning of the twenty-first century, Choguita Rarámuri was spoken by approximately one thousand speakers (Casaus 2008), a number that includes the inhabitants of all the villages within the *ejido*, of which the community of Choguita is the head village. There are several factors that suggest that the domains of usage of Choguita Rarámuri are contracting. In recent years, an increasing number of speakers are relocating to larger towns within the Sierra Tarahumara and in the capital city of Chihuahua.

As discussed in §1.3.1, the Rarámuri have faced great pressures to assimilate to settler society since the seventeenth century, and their land was reduced to half its original size (Paciotto 1996). Some of the main factors currently threatening the cultural and political autonomy of the Rarámuri nation are: (i) doubling of the *mestizo* population in the Sierra over the last century (Merrill 2013), (ii) increasing forest exploitation, (iii) depletion of water resources, (iv) expansion of road construction, and (v) recurrent violations of indigenous land property, to name only a few. The historic retreat of the Rarámuri to mountainous, isolated areas in order to avoid conflict with the *mestizo* population has had negative repercussions for their economic and political autonomy. Specifically, this displacement into areas highly adverse for maize agriculture is one of the main factors behind the severe marginalization of the Rarámuri (Merrill 1988: 77). Most recently, Choguita has had many young people migrating to urban centers, including *mestizo* towns in the sierras, the state's capital, Chihuahua (Morales Muñoz 2019), and border city Ciudad Juárez, as well as to agricultural fields in the region.

### 1 Introduction

Language decline has been documented in varying degrees in the Sierras. Some communities display interrupted intergenerational transmission of the language, while some others remain completely monolingual. Most communities present an intermediate situation with varying levels of bilingualism (Paciotto 1996). In Choguita, Rarámuri is being learnt by children, who remain monolingual until they attend primary school. Some primary school teachers are native Rarámuri speakers (most of whom speak non-local varieties of the language), but none of the secondary school teachers even know Rarámuri as a second language. In school, Rarámuri is marginally used between first and fourth grade in order to gradually introduce children to Spanish, but children are exposed exclusively to Spanish in the classroom after fifth grade (Severiano González (Choguita primary school director), p.c.).

Choguita has also undergone increasing contact with the *mestizo* population for the past few decades due to the improvement of roads that connect Choguita with *mestizo* enclaves in the Sierra. In their interactions with health promoters, government officials, traders, and religious missionaries, native Rarámuri speakers must switch to Spanish. Rarámuri is used in local administration, traditional ritual contexts, and spoken communication in joint community agricultural activities and drinking parties. Recently, however, Rarámuri speakers have started to switch to Spanish to communicate with each other in these spaces as well, as speakers themselves note and as I have been able to assess during the time I have spent in the community. The advancement of Spanish, thus, can be felt in every sphere of Rarámuri life, and older members of the community express their concern about the proficiency of younger speakers in Rarámuri.

### **1.3.3 Mexican government sponsored "bilingual/bicultural " education and literacy**

In 1989, a Chihuahua local-state office attempted to create a standarized orthographical system for Rarámuri, but the project was never completed (Pintado Cortina 2004), and the existing published materials display a great amount of variation. In Choguita, written materials in Rarámuri play a very limited role. The only written materials in Rarámuri are some sections of the official textbooks used in the local elementary school. The official schooling process is mainly devoted to promoting literacy in Spanish, as has been observed to occur in other indigenous communities in Mexico (Lastra 2001). The official "bilingual/bicultural program", designed by the Mexican Government and argued to reflect a concern about reflecting local cultural characteristics, in actual practice has served only to increase Spanish proficiency among the indigenous population and as a tool of

### 1.4 This grammar

linguistic assimilation in Rarámuri communities. The schooling process, alien to community interests and reality, reinforces stigmatization of native languages.

While there is currently no component of formal education that promotes literacy in Rarámuri in Choguita, speakers nevertheless use a Spanish-based orthography to communicate in Rarámuri in social media and via text messaging. Thus, access to new resources, such as smartphones, allows Choguita Rarámuri speakers to use their language in new contexts. Making the language documentation collection accessible and usable to all interested users therefore involves standing challenges, which also includes anticipating changing needs and possible future agendas, such as language reclamation efforts in diaspora communities.

### **1.4 This grammar**

This grammar provides a comprehensive linguistic description of the phonology, morphology, and syntax of Choguita Rarámuri. This section discusses historical information of the project that gave rise to this grammar (§1.4.1). This is followed by discussion of theoretical assumptions made in this work (§1.4.2) and description of the data sources used and the methodologies employed to elicit this data (§1.4.3). This section also addresses contributions made by individual language experts to this grammar (§1.4.4), followed by an overview of how data examples are presented (§1.4.5).

### **1.4.1 Project development**

This grammar is the product of work carried out together with Choguita Rarámuri language experts, as well as students and other collaborators who became part of the Choguita Rarámuri Language Project. I first became involved with the study of Choguita Rarámuri as a graduate student in 2002, through contacts with researchers and community members who were interested in the language and history of the community. Since my first visit to Choguita, I addressed community members in local assemblies presided over by Choguita's authorities to ask for permission to spend time in the community and learn the language. In these meetings I expressed my interest in writing about the language for my studies and my desire to engage in a long-term relationship with community members interested in language, culture and history documentation. I received approval to study in Choguita, and several people expressed enthusiasm about the idea of having the language documented and interest in working with me.

In every visit to Choguita, I addressed community members to inform them of the activities carried out with individual collaborators and overall progress

### 1 Introduction

with the project and to request their continued support. Initiatives by community members to set their own documentation agenda arose only after several visits. After learning of developments of this project, several community members expressed interest in creating a record of the speech of elders, as well as the community's historical past and receding ritual, cultural, and artistic practices. In particular, the initiative was that a team of local experts (Sebastián Fuentes Holguín, Giltro Fuentes Palma and Francisco Moreno Fuentes) would be in charge of planning and undertaking documentation activities, and requested that video documentation training be made available to community members. A community-developed repository of materials would be created and mobilized with the goal of having younger generations have access to these materials, incentivize their continued transmission and have documentation outcomes available for future generations. In response to this request, several video documentation workshops with adults and high school students were held with the help of a videographer (Jorge Esteban Moreno Romero). Project participants designed and carried out their own video documentation projects, which included interviews with elders, recording of pedagogical materials for children, and recording of different community events. Ritual and traditional events documented include healing ceremonies, races, rain and harvest ceremonies, ritual appointment of local authorities, and Easter celebrations.

The documentary materials produced through this initiative were digitized and copies given to individual creators of materials, as well as local authorities, the *siríame* (governors). Copies of these video recordings, along with recording equipment and a projector, were deposited at the local school, where Mr. Giltro Fuentes Palma would be in charge of using these materials in the school curricula. Mr. Fuentes Palma was appointed by local authorities to have custody of these materials and lead video documentation projects and mobilization of materials. Some of the participants who received the initial training were no longer able to continue carrying out documentation, as some left the community and others acquired time-consuming obligations, including Mr. Fuentes Palma. A standing challenge of this project has thus been to enable a sustainable infrastructure within Choguita for continued community-based language and culture documentation and safe-keeping and use of materials by main stakeholders in this project.

I completed a dissertation on the phonology and morphology of Choguita Rarámuri at the University of California, Berkeley in 2008 (Caballero 2008). Since then, I have continued to revise chapters of the dissertation given continued analysis of the phonology and morphology of the language. This work was also supplemented with data obtained in subsequent field trips which expanded descrip-

### 1.4 This grammar

tion of morphosyntax and syntactic structures to develop the present grammar. I moved to UC San Diego in 2010, where then linguistics graduate student Lucien Carroll joined the project. Carroll primarily carried out annotation and analysis of field data, conducting his own in-situ fieldwork in 2014 and 2018. Other UCSD researchers, both students and faculty members, have collaborated in the analysis of morphological and phonological phenomena. Many Choguita Rarámuri language experts have collaborated as authors, consultants and creators of documentation and analysis. Choguita Rarámuri language experts who have have had a deeper involvement with this project in multiple roles include Rosa Isela Chaparro Gardea, elder †José María "Chémale" Fuentes, elder †Morales Fuentes Hernández, Sebastián Fuentes Holguín, Bertha Fuentes Loya, Guillermina Fuentes Moreno, Giltro Fuentes Palma, elder †Luz Elena León Ramírez, elder †Federico León Pacheco and Francisco Moreno Fuentes (details of their contributions are provided in §1.4.4).

### **1.4.2 Theoretical assumptions**

This grammar aims to provide a comprehensive and careful description and analysis of Choguita Rarámuri without employing formalisms that are likely to date the grammar and make its content uninterpretable or inaccessible in the future (Ameka et al. 2006). Following a long tradition of grammatical description that seeks to characterize patterns and phenomena in individual languages on each language's own terms, each section in this work outlines the language-internal criteria and evidence that motivate postulating the grammatical categories identified in Choguita Rarámuri (see Cristofaro 2006 for discussion).

Nevertheless, this grammar also seeks to identify the ways in which Choguita Rarámuri resembles other languages (related Uto-Aztecan languages, or languages of Northern Mexico and beyond), and thus links description of patterns and phenomena to relevant typological, theoretical and descriptive literature where pertinent. This includes the use of standard conventions of glossing and terminology that would facilitate use by those interested in the typological properties of Choguita Rarámuri and languages of the area. More broadly, this grammar draws from typological and theoretical assumptions about morphological organization, prosodic structure, and the relationship between morphology and phonological processes, areas which display particular complexity in this language.

With respect to the interaction between phonology and morphology, the analysis presented in this grammar is compatible with construction-based approaches

### 1 Introduction

to morphology and phonology, and follows the assumption that morpho-phonological processes are intimately related to the word's layered structure, with phonology being able to apply to nested subconstituents in a word. This is a key assumption made in this work in light of the evidence in Choguita Rarámuri for an organization of the morphological structure of this language in domains, a structure which is exploited to understand the constraints on stress assignment, the domains of application of several phonological processes, and the limited appearance of multiple exponence and variable suffix order in morphologically complex words. I also make the assumption that languages may contain several phonological sub-grammars pertaining to lexical class, morphological categories, or particular morphological constructions. Both layering and construction-specific phonological processes are compatible with frameworks that posit 'cophonologies', phonological sub-grammars associated with individual morphological constructions (Orgun 1996; Anttila 2002; Inkelas & Zoll 2005; see overview and discussion in Inkelas 2014).

With respect to prosodic structure, I adopt the proposal in Selkirk (1980), Selkirk (1996), Nespor & Vogel (1986), and Hayes (1989) for the prosodic hierarchy, where the Prosodic Word is the smaller unit within the hierarchy (further discussed in Chapter 11). I also draw assumptions from property-driven approaches to word-prosodic typology laid out in Hyman (2006) and Hyman (2009), and describe the prosodic system of Choguita Rarámuri in terms of the canonical properties of stress and tone systems, dispensing with the notion of "pitch-accent" found elsewhere in the literature. Finally, in the analysis of stress, tone and intonation, the description and analysis presented here makes reference to principles and assumptions from the Autosegmental-Metrical (AM) framework (Pierrehumbert 1980, Beckman & Pierrehumbert 1986, Ladd 1986) and metrical stress theory (Hayes 1995). These frameworks provide useful tools, including phonological representations, that allow presenting key phenomena in this language. I sought a balance, whereby the notions from any specific theoretical, typological or descriptive approach employed here would not preclude the possibility of alternative analyses that may be more insightful than those presented here.

Given that the data analyzed here comes from a documentary corpus that is heterogeneous in terms of contributing language experts and genres of speech represented, this grammar attempts to describe a single language variety while at the same time addressing and exemplifying some patterns of variation present in the speech community. Reference grammars continue to provide the empirical backbone of developing linguistic theories, research in linguistic typology, and the creation of pedagogical materials for language maintenance, revitalization and reclamation. However, as noted by Evans & Dench (2006), using the

### 1.4 This grammar

metaphor of grammar writing as "catching" language, reference grammars (such as this one) can only aspire to capture static pictures, a very small fraction of a complex linguistic system. Though still limited, documentary corpora provide a more representative window into language as a dynamic system with significant variation and change in progress.

In an effort to articulate the link between this grammar and the documentary corpus in which it is based, I provide details of the content and structure of the Choguita Rarámuri corpus in §1.4.3 below. This grammar is linked to the corpus from which it stems in two other ways. First, each example is provided with source information, which allows users to retrieve the documents from which examples come from in the documentary collections of the language (see more details in §1.4.5). Second, a subset of examples of this grammar are provided with links to sound files; these are intended not only to strengthen the link between grammar and corpus, but also to increase transparency of the analyses presented here, while providing interested users with a deeper level of access to the data on which generalizations are based upon (see Remijsen & Ayoker (2018) for an example of grammatical description that includes sound examples). This more direct representation of sound in the grammar responds to calls for a new paradigm in grammatical description where users are able to form their own conclusions without their being mediated by the analytical interpretation of grammar writers (Rice 2014), enhancing the "reproducibility" of grammatical description. "Reproducible" research is understood here as "research [that] aims to provide scientific accountability by facilitating access for other researchers to the data upon which research conclusions are based" in cases where true replicability (the ability to produce new data by recreating research conditions faithfully) is not possible (Berez-Kroeker et al. 2018: 4). Most importantly, sound in grammar is intended to enable and improve access to original language data by community members who are native speakers or language learners who are interested in language preservation, revitalization and reclamation (see also Rice (2014) for discussion).

While the present grammar aims to provide a glimpse of variation within the speech community, it should be noted that a full account of how lectal variation and bilingualism operate in the community of Choguita and how it may have an impact on Choguita Rarámuri grammar is a topic for future research.

### **1.4.3 Data sources and methodology**

A documentary corpus of Choguita Rarámuri informs the analysis and is the source of the examples presented in this grammar. This corpus comprises data collected during several visits I made to Choguita between 2003 and 2018. The

### 1 Introduction

contexts in which data were elicited are outlined in the following subsections. The corpus consists of over 200 hours of audio and/or video recordings of elicited data, personal, historical, and procedural narratives, conversations, interviews, prayers, and oratory, as well as several hours of sessions where language experts chose the topics to be covered in language teaching sessions.

Recordings and associated annotation materials are available in two archival collections. The first collection (Caballero 2009) is housed at the Endangered Languages Archive,<sup>10</sup> and contains approximately 130 hours of digital audio recordings, 10 hours of video recordings, and a substantial amount of digital transcription and annotations. These materials were obtained from 2003 to 2009. An overview of the background and contents of this collection is provided in Caballero (2017). The second documentary collection (Chaparro Gardea et al. 2019) is housed at the Survey of California and Other Indian Languages at UC Berkeley.<sup>11</sup> This second collection contains over 1,300 digital files, including 115 hours of sound recordings that were primarily obtained between 2011 and 2018.

All of the data examples cited in this work are part of materials where individual contributors consented to be included in this work and other linguistic publications. A subset of the materials of the documentary collections are, however, not cited in this work, as there is a community-wide consensus that certain materials are not to be accessed by outsiders. Specifically, and as decided in local government assemblies, special sensitivities are placed on video recordings of ritual celebrations and activities, as outsiders in the past have recorded them without consulting local authorities or the community at large.

The Choguita Rarámuri documentary corpus comprises a wide sample of different speech genres with different degrees of planning (including conversations, monologues, narratives, myths, ceremonial speeches (*nawésari*), interviews of elders by native speakers, and ritualistic chants and prayers (healing ceremonies). The corpus also includes recordings of elicitation sessions where speakers undertake the role of language teachers, which allows for a great deal of contextualization of the data elicited. Other kinds of elicitation conducted included contextualized- and text-based elicitation, translation, metalinguistic judgements, and elicitation prompted by culturally relevant visual props, as well as and participant observation. The following sections describe the different classes of data collected between the duration of the project, as well as the methodologies employed to this end.

<sup>10</sup>This collection is available on-line at http://elar.soas.ac.uk/deposit/0056.

<sup>11</sup>This collection can be retrieved on-line at http://dx.doi.org/doi:10.7297/X2HH6H70.

1.4 This grammar

### **1.4.3.1 Narratives**

Many examples in this grammar come from narratives, a series of documents that mainly involve monologic speech, of a kind closer to the planned speech event end of a spontaneity continuum. These texts include genres such as general descriptions of events, historically contextualized descriptions, myth narratives, procedural texts, and narratives from visual prompts, recorded (in both audio and video) with single speakers. Audio recordings of texts were typically made in indoor, private spaces, while video-recorded texts generally involve procedural texts and historical descriptions made in situ. These narratives are crucial to linguistic analysis in this reference grammar and are part of the developing record of the history, culture and language of the community as envisaged by community members.

While contributing language experts were sometimes prompted very generally (with questions about how Choguita was during their childhood, questions about pictures of the community, or asking for a traditional story), the topics covered in the narratives were generally selected by each contributing language expert. In some cases, speakers had family members as an audience when recording their text, when this was contextually appropriate. In some cases, a narrative would evolve from a conversation about particular topics or through elicitation of lexical items, blurring the lines between narrative proper, conversation and elicitation. Example sentences that come from narratives are indicated as 'tx' in the source code (see §1.4.5 for more details).

### **1.4.3.2 Conversations**

A subset of recordings in the corpus involve audio and/or video-recorded conversations. In these sessions, speaker participants engaged in conversations, either with other native speakers or with me (as a second language user of Choguita Rarámuri). Conversations recorded can be classified into three types: (i) unstructured conversations that arose in the context of elicitation; (ii) elicited conversations between native speakers; and (iii) semi-structured conversations between monolingual language experts and me for lexical or grammatical elicitation. In conversations arising in the context of elicitation, these mainly involve informal, unstructured exchanges between a language expert and me and focused on lexical items (such as kinship terms), contexts of use of particular structures or linguistically and/or culturally relevant aspects of topics covered in elicitation.<sup>12</sup> In these types of conversation, both Choguita Rarámuri and Spanish are

<sup>12</sup>In a few recordings, the exchanges were exclusively carried out between native speaker participants discussing the meaning of lexical items and constructions elicited using culturally relevant still pictures.

### 1 Introduction

used. The second type of conversations, less frequent, involved eliciting conversations between native speakers in arranged recording sessions, generally with prompted topics of conversation. These types of conversations generally ended up having the structure and tone of a monologic narrative by one speaker and back-channeling by other speakers. Finally, in the third type of conversation, semi-structured conversations, the sessions were designed to elicit particular lexical items or grammatical constructions with monolingual speakers (for example, to elicit tonal minimal pairs).

Since all these interactions took place in the context of elicitation sessions or typically involved topics and constructions that were recorded for the purpose of informing linguistic analysis, they are part of the set of materials from the deposited collections that have an exclusively linguistic focus. None of these conversations can be taken to be representative of Choguita Rarámuri conversational norms, either because the exchanges involved my limited fluency or because the presence of the recorder influenced the interaction (all of the conversational data in the corpus was obtained on-record, with explicit acknowledgement and permission from contributing speakers).

### **1.4.3.3 Interviews**

Some of the examples cited in this grammar are drawn from interviews with elders, which were obtained by community member collaborators as part of the initiative to document the history of the community. These interviews had a special focus on endangered domains of knowledge still possessed by elders, but in attrition or non-existent for younger speakers, such as the complex kinship term system that at the time of recording was not known with its full complexity by speakers in their thirties and younger. Other topics included past agricultural practices, food preparation, ritual practices, as well as personal biographical information of interviewees. The interviews include accounts by interviewees of their experience growing up in Choguita and the differences brought about by recent changes in the community, such as the introduction of the government-run local school or the introduction of electricity. Interviews were video recorded with high quality audio, with additional, separate audio recordings made at the time of the recording with the solid-state recorder. These interactions were almost exclusively dyadic and carried out in Choguita Rarámuri, i.e., no code-switching was attested in these recordings.

### **1.4.3.4 Elicited data**

Much of the data used in the grammar comes from lexical or grammatical elicitation designed to gather specific types of evidence for phonological, morphological, or syntactic analysis. While it is desirable to exemplify patterns and structures drawn from speech events with a lower degree of planning (such as narratives or conversations), some types of constructions are rarely found in other speech genres due to the highly specific semantic contexts they involve. Elicitation was thus critical in assessing the nature and properties of the grammatical constructions discussed in this grammar.

The distinction between conversation, elicitation and narrative was blurred in certain contexts of language work, but data examples are identified as elicited in this work if it meets one of the following criteria: (i) it was obtained by asking speakers to translate a word or phrase from Spanish into Choguita Rarámuri or viceversa (translation elicitation); or (ii) it was uttered by speakers when provided a context and being asked what they may say in such a context; or (iii) it was an answer provided by speakers when offered a Choguita Rarámuri word or clause (either overheard or from annotation of a text) and asked to elaborate about the context in which they would utter it (contextualized and text-based elicitation).<sup>13</sup>

In addition to these strategies, some sessions also involved eliciting metalinguistic judgments or eliciting responses using visual props, such as pictures of rituals or agricultural practices in other Rarámuri communities. Other sessions were video recorded and focused on in-situ descriptions of agricultural terms and the language of space, including deictic terms, topographic terms and landscapebased standardized place names. Finally, one other type of elicitation involved requesting grammaticality judgments of morphologically complex words or clauses in Choguita Rarámuri; the offered forms were either constructed forms with logically possible affix orderings or stress patterns or forms produced by other speakers. I would ask speakers to assess the grammaticality of the offered forms and, if judged grammatical, to discuss their meanings in detail. I have avoided exemplifying any given pattern with this kind of evidence, except for cases where negative evidence (i.e., the ungrammaticality of a particular suffix sequence) is relevant in the discussion. Any data examples that arose through this methodology are indicated as '[pr.]' (for prompted) in the source code.

<sup>13</sup>This type of elicitation proved to be useful for elucidating properties of grammatical constructions or lexical items: particular constructions encountered during an off-record conversation or in the process of annotating individual texts would serve as starting points for exploring grammatical or lexical aspects of the constructions in question.

### 1 Introduction

### **1.4.3.5 Language teaching sessions**

A final type of data examples come from language teaching sessions, where native speaker participants explained lexical items, constructions and expressions, and their cultural contextualization to me. In some cases, I prompted the topics of discussion, which included overheard expressions and terms recorded during elicitation sessions. In most of these sessions, however, speakers were just asked to teach me anything they decided would be appropriate for me to learn as a second language learner and for the purpose of my description and analysis of the language. In these sessions it was frequent that language experts would correct my mistakes when using Choguita Rarámuri, clarifying the form and/or meaning of the expression in question, as well as the appropriate context for its utterance. Other topics covered in these sessions include politeness formulas, vocabulary associated with rituals and traditional agricultural practices, culturally relevant contextualization for the use of some terms and colloquial expressions, and discussions about dialect differences between Choguita Rarámuri and neighboring varieties.

### **1.4.4 Language experts and collaborators**

To date, thirty-four community members have participated in the Choguita Rarámuri language project in different roles, including as authors, consultants, and creators of audio and video recordings, annotation and analysis. All native contributors wanted their utterances attributed to them and to be acknowledged for their expertise and role in this project. A full list of language experts who have contributed to this project is provided below in Table 1.3, with contributors' initials, names and roles in the project. Speakers who contributed their expertise in elicitation sessions are listed as consultants. Primary speakers in monologic texts and interviewees in interviews are listed as authors, while interviewers are listed as interviewers. The initials that precede each contributor's name corresponds to the language expert reference in the source file names listed immediately below example sentences (details of how data examples are represented are provided in §1.4.5).

From the project participants listed in Table 1.3, the main contributors to this grammar are the language experts listed below (in alphabetic order). Below, I briefly describe their role in the project.

**Rosa Isela Chaparro Gardea (RIC)** joined the project in 2011 and quickly became one of the main contributors to the project, authoring several narratives, helping me annotate and translate narratives authored by her and other language experts


Table 1.3: Names, initials and roles of contributing language experts

and participating in elicitation sessions, as well as offering insights about variation between Choguita Rarámuri and closely related Norogachi Rarámuri, where she was born and raised.

Elder **†José María "Chémale" Fuentes [JMF]** was a longstanding leader and authority in Choguita (having served as the head governor several times) who had

### 1 Introduction

a wealth of knowledge about the history of the community, was an accomplished violin player and an expert of the *nawésali* register, the oratory speech style that only a few (those recognized as the wisest in the community) master. He graciously contributed historical and mythical narratives, as well as descriptions of agricultural practices past and present.

Elder **†Morales Fuentes Hernández [MFH]** was a ritual singing shaman (*sikaláme*), violin player who had a great interest in recording the music and rituals of the community. His recordings were always done by his enthusiastic initiative, and with a sense of urgency: ritual music and shamanic singing are some of the domains of local culture that is mainly mastered by elders, with few young people learning this form of art. His own personal style of performance is reflected in the recordings made. He also participated in grammatical elicitation sessions structured as conversations, as he was one of the elders in the community who had very limited knowledge of Spanish.

**Sebastián Fuentes Holguín [SFH]** was one of my key teachers and collaborators since the beginning of the project, and has contributed extensively to every aspect of the project, from authoring several monologic narratives and ceremonial speeches, to being a patient collaborator and teacher in elicitation sessions. During elicitation sessions I benefited from his insights and expertise, and he was always enthusiastic to discuss possible contexts of utterance for constructions elicited. He is a passionate advocate for the language and one of the main leaders of the community-based initiative to document the cultural, historical and linguistic heritage of Choguita.

**Bertha Fuentes Loya [BFL]** was also one of my main teachers and collaborators. She authored narratives and contributed as a consultant for many elicitation questions and was also an excellent language instructor. She also was a key contributor of annotation of texts recorded with other language experts and also shared her expertise in variation between Choguita Rarámuri and neighboring Rarámuri varieties. She contributed to in-situ elicitation focused in agricultural practices and landscape terms. She is also an expert seamstress and authored several narratives and procedural texts about her art.

**Guillermina Fuentes Moreno [GFM]** is a community leader who contributed her knowledge of the history of Choguita in several historical narratives and also shared her vision of the community into the future. She was also a consultant in elicitation sessions and unstructured conversations.

**Giltro Fuentes Palma [GFP]** contributed to the project as an author of narratives, as a consultant in elicitation sessions, with annotation and analysis as well as

creation of video documentation. As part of the team that lead the documentation and mobilization of the history, culture and language of Choguita, he focused on interviews with elders as well as the creation of materials to be used in the local school curriculum. Giltro was appointed by local authorities as the custodian of the documentation materials deposited in Choguita and to head the mobilization of documentation outcomes in the community.

Elder **†Luz Elena León Ramírez [LEL]** was a master storyteller and an invaluable source of linguistic, cultural and historical knowledge. She graciously contributed many narratives to the corpus, including procedural texts, historical narratives (including contact with members of the N'dee/N'nee/Ndé (Apache) nation and sightings of the armed forces of Pancho Villa in the early twentieth century), descriptions of culturally relevant events in the community (such as an epidemic suffered by the community), and personal history from her childhood. She also collaborated in lexical and grammatical elicitation sessions, and helped translate and annotate her own and others' speakers texts. She was a trained midwife and medical assistant and beloved in Choguita and beyond for her generosity and loving service to her community.

Elder **†Federico León Pacheco [FLP]** contributed a wealth of historical and cultural knowledge about Choguita and the Rarámuri nation. One of the few remaining *owirúame* (shaman) in the community, he was one of the bearers of specialized knowledge, including the complex kinship system, which he shared in interviews led by community members. He also contributed his knowledge in interviews focusing on the early history of Choguita, his escape from a boarding school in nearby Norogachi in his childhood, and his path of becoming a healing shaman.

**Francisco Moreno Fuentes [FMF]** was involved in video documentation training and led community-based teams that created records of traditional celebrations and rituals. He also contributed to the project as a consultant in elicitation sessions.

Many other people contributed to this grammar indirectly through informal interactions and other forms of support during my time in Choguita. Though they are not acknowledged by name here, their contribution is deeply appreciated.

### **1.4.5 Representation of examples**

Each glossed example minimally provides, from top to bottom: (i) a broad phonetic transcription in IPA; (ii) a phonemicized transcription (also in IPA) with

### 1 Introduction

morpheme breaks; (iii) glosses; (iv) an English translation; (v) a Spanish translation; and (vi) a source code. Examples may also include underlying phonological representations where relevant. Examples are glossed following the conventions established in the Leipzig Glossing Rules (Comrie et al. 2008). Any departures from this standard are justified where introduced.

A number of examples in the grammar have accompanying audio recordings; these examples are provided with a hyperlink in the source code that directs readers to an open access repository where individual audio files are available for download (Caballero 2022). The goal is to allow readers to access the larger contexts from which the grammatical description is based upon, which enables wider dissemination of the results and the ability of interested community members and academics interested in the language to carefully examine the analyses and description set forth in this grammatical description.

All examples include a Spanish translation (reflecting the local Northern Mexican variety spoken in the Sierra), the language used for translation elicitation and annotation of texts. Spanish translations were given by bilingual language experts of their own utterances or of the words of a different speaker when contributing annotation of texts. When the Spanish translation offered by language experts departs from standard Northern Mexican Spanish, a more standard Spanish translation is supplied and is accompanied by the language expert's verbatim translation, enclosed in double quotation marks. Information included in the translation that is inferred from the context and noted as such by contributing language experts is provided in parentheses. Example sentences are translated by me or by native English speaker student research assistants.

The source code is provided in angled brackets in the last line and provides the contributor's initials and a unique identifier that links the particular example to the documentary corpus. If the source example is taken from handwritten field notes, this unique identifier will reference the year the example was recorded, book and page number (e.g., '05 1:125'). Examples from written field notes also identify the type of document where the data comes from as follows: elicitation ('el'), text ('tx'), interview ('in'), or conversations ('co'). If the source example has been annotated in a digital format, the unique identifier will reference the label of the source document.<sup>14</sup> When available, a hyperlink in the source code enables retrieval of an audio file from an open access repository (Caballero 2022).

<sup>14</sup>This is an automatically generated identifier generated in *Kwaras*, an ELAN corpus management tool created by Russell Horton (Linguistics MA 2012, UCSD) and further developed by Lucien Carroll (Linguistics PhD 2015, UCSD) (Caballero et al. 2019). This unique identifier is derived from the source file name (e.g., 'co12-37') plus the time stamp of the annotation referenced (e.g., '0:49.6').

1.4 This grammar

For examples where a corresponding audio file is available, the source code is provided in curly brackets and gray font. An example of how data is cited is illustrated in (2):

(2) a. *muˈhê taˈmí saˈpâto raˈrèma muˈhê* 2sg.nom *taˈmí* 1sg.acc *saˈpâto* shoes *raˈr-è-ma* buy-appl-fut.sg 'You'll buy shoes from me.' 'Me vas a comprar zapatos (que vendo yo).' < SFH 05 1:74/el > b. *neˈhê ˈpé oˈkwâ raʔiˈtʃ͡âma koriˈmá ˈhîtara neˈhê* 1sg.nom *ˈpé* just *oˈkwâ* couple *raʔiˈtʃ͡â-ma* speak-fut.sg *koriˈmá* fire.bird *ˈhîtara* about 'I'll speak a little about the *korima* (the fire bird).' 'Yo voy a hablar poquito del pájaro *korimá* (el pájaro de fuego).' {LEL

In (2a), the example sentence was uttered by SFH in 2005 in an elicitation context ('el'), and the handwritten notes of that session are located in book 1, page 74. In (2b), the example sentence was uttered by LEL; the code next to the contributing speaker's initials is the unique identifier of that example, derived from the source file name (a text labelled 'tx5') followed by the time stamp of the annotation referenced ('0:22.9').<sup>15</sup> In this example, the source code is represented in curly brackets and grey font since it contains a hyperlink that directs users to an audio file corresponding to this data example available for download.

Where an example presents an excerpt from a dialogue, each speaker is identified by their initials at the start of the transcription line, as shown in (3).

(3) a. [ME] *boniˈlâ ... boni-ˈlâ* be.younger.brother-poss 'Younger brother...'

tx5:00:22.9}

'Hermano menor...'

b. [SFH] *boˈnêsa ba? boˈn-ê-sa* younger.brother--have-cond *ba?* cl 'As if he were a younger brother?' '¿Como si fuera hermano menor?'

<sup>15</sup>In many cases, authors themselves contributed to annotation and transcription of their own speech. In cases where a different language expert collaborated with me in transcribing and translating the texts, only the author of the text referenced is credited in the source code.

### 1 Introduction

c. [ME] *uˈrí, boniˈlâ ˈnísa ˈlá ba ˈni uˈrí,* yes *boni-ˈlâ* be.younger.brother-poss *ˈní-sa* cop-cond *oˈlá* cer *ba* cl *ˈni* emph 'Yes, as if he were a younger brother indeed.' 'Si, si fuera hermano menor, asi es.' {ME in485:02:22.9}

Highly common phrases or expressions are given without a source. These are phrases or expressions that are too common to be ascribed a single source.

All examples provided in this grammar reflect idiolectal variation, which is particularly abundant in terms of pronunciation and various phonological reduction processes. Thus, there is no normalization of transcriptions to reflect any single pronunciation as "standard". Where a particular form illustrates a particular process only attested in the speech of one or few language experts, the divergence is usually illustrated with an underlying phonological representation and a footnote that includes a cross-reference to the chapters/sections addressing those specific processes.

Finally, the reader should not assume that the example sentences in this work accurately reflect the culture, interests, priorities or personalities of the speakers that uttered them: a large amount of examples used to illustrate patterns and phenomena here have been obtained in contexts where the goal was to illustrate structural aspects of the language, and are not necessarily sociologically or ecologically representative.

### **1.5 Overview of the grammar**

Description in this grammar has been organized in a way that aims to facilitate discovery of particular topics in the grammar of Choguita Rarámuri by typologists and linguists interested in Choguita Rarámuri or languages of this region. This book has features of the predominant ascending macrostructure organization of grammars (phonology > morphology > syntax), where description begins with chapters devoted to the sound system to then move to increasingly complex units of analysis (words, phrases, clauses and sentences).

This structure or any organization where domains of linguistic structure are compartmentalized into discrete chapters, however, presents challenges in the organization of the description of linguistics systems, especially with respect to aspects of structure that cross-cut all levels of grammar, such as prosody (Mosel 2006). As shown in this grammar, the prosodic system of Choguita Rarámuri involves complex interactions between lexical, post-lexical and grammatical information that require discussion in its own right. Confronted with this challenge,

### 1.5 Overview of the grammar

this grammar includes a chapter where interactions of prosodic patterns and processes across grammatical domains are addressed in detail with the goal of providing a clear picture of how these complex interactions yield surface forms in this language. This supplements the description that pertains to each individual domain of description, such as description to the lexical role of tone in the chapter devoted to tone and intonation and description of the grammatical role of tone in chapters devoted to morphology. Thus, there is some degree of redundancy built into the description, which seeks to enable different contextualization of the data and the analyses provided.

Like most reference grammars, this volume also adopts a semasiological, formto-function organization, which means that related semantic or conceptual domains may be addressed in different chapters depending on how they are morphosyntactically encoded. There are abundant cross-references provided in the grammar that are intended to help readers identify topics that may be semantically or conceptually related.

This chapter concludes with a summary of each of the following chapters of this grammar.

**Chapter 2** lays out the core structures of the language, spanning phonological patterns and processes, morphological contrasts encoded by pronouns and demonstratives and their function, morphosyntactic and morpho-phonological properties of nouns and verbs, word order patterns, appositive possessive constructions, relative clauses, complement clauses, clause chaining, and complex predicates.

**Chapter 3** addresses the segmental phonology of Choguita Rarámuri. First, it provides the phonological inventory and complex patterns of allophonic variation. This is followed by a description of the various phonological processes, including processes that target labio-velar semi-vowel and voiced bilabial stops, alveopalatal affricates, alveolar fricatives, nasals and rhotics. This chapter includes description of post-consonantal devoicing and voiceless plosive lenition processes.

**Chapter 4** describes the syllabification patterns of Choguita Rarámuri, as several suprasegmental processes make crucial reference to the syllabic structure of words. This chapter includes discussion of underlying syllable structure patterns, as well as surface consonant and vowel sequences.

**Chapter 5** addresses the word-level stress system of Choguita Rarámuri, including acoustic and distributional properties of stress and stress-dependent phenomena. There is also detailed consideration of the lexical stress properties of both

### 1 Introduction

roots and suffixes and description of the typologically unusual initial three syllable stress window that restricts the location of stress in the language.

**Chapter 6** is devoted to tone and intonation. This includes a detailed description of the tonal inventory and its acoustic encoding, tone melodies by root types and interaction with word-level stress, as well as stress-based tonal neutralization. This is followed by description of the intonational characteristics of declarative sentences, tone-specific intonation patterns and non-tonal encoding of intonation.

**Chapter 7** outlines other word-level suprasegmental processes of Choguita Rarámuri that contribute to its prosodic complexity. These include restrictions on the distribution of glottal stops in an initial disyllabic window, minimal word size restrictions and loanword prosodic adaptation patterns.

**Chapter 8** describes the morphology of nouns. First, it addresses the morphotactic generalizations of morphologically complex nouns. This is followed by description of each of the nominal morphological categories, including plural/ pluractionals, case, possessive marking, and derivational morphological processes, including agentive, patientive and experiencer nominalizations and abstract noun nominalizations. This chapter concludes with discussion of adaptation of Spanish loan nouns and tone patterns in morphologically complex nouns.

**Chapter 9** provides a detailed overview of the verbal morphology and morphologically-conditioned phonological processes of Choguita Rarámuri. First, this chapter details the distinctions between verbal root classes in terms of their stress properties in different morphological environments, as well as their transitivity properties. This is followed by a comprehensive description of grammatical tone patterns in morphologically complex words. This chapter also addresses evidence for positing morphological domains in a layered, hierarchical structure of the verb; the evidence reviewed includes phonological processes that apply in specific verbal domains and morphotactic evidence for suffix ordering generalizations and patterns of variable affix order. This chapter concludes with discussion of the verbal complex, which includes clitics and modal particles.

**Chapter 10** describes the morphological properties of minor word classes, including pronouns, demonstratives, adjectives, numerals, quantifiers, definite articles, adverbs and discourse particles and enclitics. The discussion is organized on a classification of these word classes into two groups, depending on whether they may head noun phrases or combine with head nouns in noun phrases and those that cannot.

### 1.5 Overview of the grammar

**Chapter 11** is devoted to prosodic structures and processes that cross-cut the grammar of Choguita Rarámuri, including lexical phonological processes, morphological processes and post-lexical phonological phenomena. First, this chapter outlines the criteria for determining the Prosodic Word in Choguita Rarámuri, and discusses the domains for phonological and morpho-phonological processes below the level of the Prosodic Word. This is followed by discussion of phonological phenomena that are quantity-sensitive in the language vis-à-vis the lack of contrastive vowel length in the language. This chapter also addresses complex prosodic interactions: (i) between stress and tone and between lexical tone and grammatical tone in morphologically complex words, and (ii) between lexical tone and intonation, which includes non-tonal encoding. This chapter concludes with a discussion of prosodic constraints associated with different morphological constructions.

**Chapter 12** examines how nouns and various forms addressed in Chapter 10 combine in forming noun phrases in Choguita Rarámuri. This chapter organizes the description in terms of each of the possible modifiers of nouns, detailing restrictions on co-occurrence, agreement patterns and word order properties. This chapter includes discussion of simple noun phrases, which contain a nominal head and a single modifier, and noun phrases that involve possessive constructions.

**Chapter 13** characterizes basic clause types in terms of their transitivity properties (intransitive, transitive and ditransitive), as well as locative clauses, copular clauses and existential clauses. This chapter includes detailed discussion of the characteristics of clauses headed by postural verbs, and considers the semantics of postural verb themselves and the postural constructions they head.

**Chapter 14** moves on to consider the morphosyntactic properties of different types of sentence types, namely interrogative, negative, imperative and comparative constructions. The prosodic properties of interrogatives are assessed in relation to declarative sentences.

**Chapter 15** details the complex clause structures found in Choguita Rarámuri. This includes discussion of complement clauses, adverbial clauses, relative clauses as well as clausal conjunction, disjunction and adversative coordination. This chapter also provides a description of complex predicate constructions, which includes light verb, auxiliary, serial and V-V incorporation constructions. The properties of constructions with depictive and resultative semantics are also addressed.

# **2 Grammatical overview**

### **2.1 Phonology**

### **2.1.1 Segmental inventory and processes**

The Choguita Rarámuri phonemic consonant and vocalic monophthong inventories are provided in Table 2.1 and Table 2.2, respectively.<sup>1</sup>


Table 2.1: Phonemic inventory of Choguita Raramuri consonants

Table 2.2: Choguita Rarámuri monophthong vowel system


Choguita Rarámuri displays complex patterns of allophonic variation, most of which involve lenition processes. Voiced bilabial plosives in inter-vocalic position may be optionally realized as voiced bilabial fricatives ([β]) (1a) or voiced

<sup>1</sup>The labio-velar approximant is doubly assigned in the bilabial place of articulation column and in the velar place of articulation column in Table 2.1.

### 2 Grammatical overview

approximants ([β]) ( ̞ 1b). Lenition of /b/ in unstressed word-initial syllables (exemplified in (1c)) leads to neutralization of the contrast between /w/ and /b/.

	- a. [ziˈ**β**óò] /si-ˈ**b**ô/ go.pl-fut.pl 'we will go' 'vamos a ir' {MFH tx1133:01:17.2} b. [ˈè**β̞**əma] /ˈè**b**i-ma/ bring-fut.sg 's/he will take it.' 'va a llevarlo.' < BFL 06 6:73/el >
	- c. [waʔˈwí] /**b**aʔˈwí/ 'water' 'agua' < SFH 04 1:17/el >

For some speakers, the voiced bilabial stop is realized as a labio-velar semivowel ([w]) pre-consonantally after post-tonic vowel deletion (2).

(2) [ˈè**w**tiki] /ˈè**b**i-ti-ki/ bring-caus-pst.ego 'I made him/her bring it.' 'Lo/a hice traerlo.' < BFL 06 6:73/el >

Alveopalatal affricates may be produced as alveolar affricates, optionally depalatalizing before low, central vowels. Alveolar affricates may also further reduce, becoming deaffricated in fast speech, a process that frequently targets function words like demonstrative *ˈétʃ͡i*. These processes are exemplified in (3a–b) and (3c–d), respectively.

2.1 Phonology

	- a. [aˈkâ**tʃ**a-la] ~ [aˈkâ**ts**a-la] /aˈkâ**tʃ**a-la/ paternal.grandmother-poss 'her/his paternal grandmother' 'su abuela paterna' {SFH in243:08:09.4}

b. [aˈ**tʃ**â-sa] ~ [aˈ**ts**â-sa] /aˈ**tʃ**â-sa/ to.sit.sg.tr-cond 'if s/he sits him/her down' 'si lo sienta' < SFH 04 1:38/el >

c. [e**s**ˈtá]

```
/ˈétʃ͡i ˈtá/
dem det
'that one'
'ese' < LEL 06 6:141-162/tx >
```
d. [ˈé**s** ˈkútʃ͡i]

/ˈé**tʃ**i ˈkútʃ͡i/ dem small 'those small ones' 'esos pequeños' < BFL 07 frog story\_2/tx >

A striking pattern of segmental reduction involves lenition of voiceless plosives, which are gradiently realized within a continuum ranging from voiceless aspirated stops to complete deletion: [pʰ > p > b > β > β̞> w > Ø]. I analyze this reduction process, exemplified in (4), as a phonetic process that is dependent on rate of speech and position of segments within the Intonational Phrase: productions on the lenis end of the continuum (fricatives and approximants) tend to be produced in utterance-final position, while productions on the fortis end (voiceless and voiced stops) tend to be produced utterance-medially, a factor that may suggest that these alternations are sensitive to phrasal phonological effects.

	- a. *tʃ͡ú riˈká tiˈbúsa ˈlé pa ˈnà kaˈwì βa tʃ͡ú* how *riˈká* that *tiˈbú-sa* take.care-cond *aˈlé* dub *pa]* cl *ˈnà* prox *kaˈwì* land *βa]* cl '(we learnt) how to take care of it, this earth' '(aprendimos) cómo cuidarla, la tierra' {SFH tx977:00:60.0}

2 Grammatical overview

> b. *ˈnà kaˈwì β̞a ˈnà* dem *kaˈwì* earth *β̞a* cl 'this earth' 'este mundo' {SFH tx43:11:11.2}

In addition to these processes, Choguita Rarámuri exhibits cross-linguistically common phonological processes, including palatalization of alveolar fricatives before high vowels. This process may be rendered opaque by post-tonic vowel deletion. This is exemplified in (5).

(5) Surface opaque fricative palatalization [aˈtí**ʃ**-li] /aˈtí**si**-li/ to.sneeze-pst 'S/he sneezed.' 'Estornudó.' < BFL 05 1:111/el >

In addition to consonantal allophonic patterns, vocalic segments undergo various stress-based reduction processes. Specifically, there are three distinct, optional patterns or degrees of vowel reduction: (i) /e/ raises to [i] both pre-tonically and posttonically (6a–b) (a robust process that exhibits optionality only for some speakers); (ii) non-final post-tonic /a/ and /o/ raise to [i] (6c–d); and (iii) high vowels reduce to schwa post-tonically (6e–f).

(6) Stress-based vowel reduction

$$\begin{array}{ll} \text{a.} & \text{[níhé] - [ne/hé]}\\ & \text{/ne} \text{hé/} \\ & \text{'I'} \\ & \text{'yo'} < \text{SFH 07 2:} 63/\text{el} > \\ \text{b.} & \text{[bínè] - [be'né]} \\ & \text{/be'nè/} \\ & \text{'learn'} \end{array}$$

$$\text{'aprander'} < \text{ROF} \text{ 04 1:9/el} > \text{ } \\ \text{c.} \quad \text{['t} \widehat{\text{'} \text{\'\'} \text{\'} \text{ti-li}] \sim [\text{'t} \widehat{\text{'} \text{\'\'} \text{\'} \text{ta-li}]}$$

$$\text{'tf\u0ta-li/}\\\text{'tf\u0ta-li/}\\\text{'It begin-pstr}\\\text{'It began.'}\\\text{'Comenz\u0.'} < \text{SFH 07 in243/in} > \text{'} $$

2.1 Phonology

d. [rono+ˈbâk**i**-ma] ~ [rono+ˈbâk**o**-ma] /roˈno+paˈk**o**-ma/ feet+wash-fut.sg 'S/he will wash their feet.' 'Lavará sus pies.'

```
e. [ˈpòlə-ki] ~ [ˈpòli-ki]
    /ˈpòli-ki/
     cover-pst.ego
    'I covered myself.'
    'Me tapé.' < AHF 05 1:125/el >
```

```
f. [naˈwí-nəla] ~ [naˈwí-nila]
    /naˈwí-nula/
    sing-order
    'They ordered them to sing.'
    'Los mandaron a que cantaran.' < BFL 07 VDB/el >
```
The underlying syllable structure in Choguita Rarámuri is (C)V(ʔ), with optional onsets and optional glottal stop codas. Stress-based vowel deletion yields consonant clusters in surface forms, which include geminates for some speakers, most frequently oral and nasal bilabial ones (e.g., *moˈtépi-po* →*moˈtép-po* 'we will braid her hair').

There is no contrastive vowel length, but long vowel sequences arise at morpheme junctures (e.g., *biʔˈw-à-a* 'they are cleaning') or as a morphologicallyconditioned effect (e.g., the past passive *-ru* suffix conditions lengthening of the stressed stem vowel). There are also diphthongs with falling sonority, which occur morpheme internally and at morpheme boundaries (e.g., *kaiˈnâ-ma* 'they will harvest', *seˈmè-i* 'he used to play the violin').

### **2.1.2 Stress, tone and prosodic structure**

Stress in Choguita Rarámuri is lexically contrastive and assigned to the first, second or third syllable within an initial three-syllable window. Stress in inflected words is determined by the stress properties of both roots and suffixes: roots are stressed or unstressed (with fixed or shifting stress across paradigms, respectively), while suffixes are stress-shifting or stress-neutral. Stress distribution by root and suffix type according to their stress properties is exemplified in Table 2.3:

### 2 Grammatical overview


Table 2.3: Stress patterns of morphologically complex verbs

A three-way lexical tonal contrast between HL (â), L (à) and H (á) tones is exclusively realized on surface stressed syllables: there is only one lexical tone per Prosodic Word and stressless syllables lack lexical tone. The tone-bearing unit is the mora: falling tones have their high target on the stressed syllable, with the fall starting in the tonic and continuing through a post-tonic syllable, if there is one; H tones may spread their high f0 to the post-tonic syllable.

Morphologically governed stress shifts result in lexical tonal alternations: if a stress-shifting suffix is stressed after a stress shift, the stressed suffix syllable will bear the lexical tone of that suffix, either a L tone (7b), a HL tone (7d) or a H tone (7e).

### (7) Root and suffix tones

a. [kiˈmáli] H /kimá-li/ put.on.blanket-pst 'S/he covered with a blanket.' 'Se encobijó.' < BFL el1909 > b. [kimiˈsì] L /kimá-sì/ put.on.blanket-imp.pl 'You all cover yourselves with a blanket!' '¡Encobíjense!' < BFL el1909 > c. [ˈtòli] L /tò-li/ take-pst 'S/he took it.'

'Se lo llevó.' < RIC el921 >

2.1 Phonology

d. [toˈkâ] HL /tò-kâ/ take-imp.sg 'Take it!' '¡Llévalo!' < BFL el1882 > e. [toˈnále] H /tò-nále/ take-desid 'S/he wants to take it.' 'Quiere llevarlo.' < BFL el1882 >

Trisyllabic unstressed roots have second syllable stress and the root's lexical tone is realized on the stressed syllable, e.g., H tone in (8a) or L tone in (8d). When attaching a stress-shifting suffix, these stems have third syllable stress in a root syllable which bears a HL tone regardless of what the lexical tones of the root and suffix(es) are, as shown in (8b–c, e–f).

(8) Tonal neutralization


### 2 Grammatical overview


Tone alone may be the sole exponent of inflectional categories (e.g., L tone is an allomorph of the imperative singular construction, e.g., HL *niˈkâ* 'it barks' vs. L *niˈkà* 'bark!'). In other grammatical tone patterns, stem tones are conditioned by suffixes in certain verb classes.

Phonological and morphological processes and phenomena make reference to the Prosodic Word, which in Choguita Rarámuri can be identified by the following criteria:

	- a. Each prosodic word is assigned a single, main stress within the first three syllables, which bears a lexical tone (§5.1).
	- b. Inflected verbs are minimally bimoraic (§7.2).
	- c. Prosodic words are vowel final (§5.2.2).
	- d. Body part incorporation combines two morphological roots into a single prosodic word, where the new lexical item is assigned a single, main stress in the first syllable of the head of the compound (§9.3.3).
	- e. The glottal stop only emerges within the first two syllables of the prosodic word (§7.1).

### 2.2 Pronouns and demonstratives

### **2.2 Pronouns and demonstratives**

Choguita Rarámuri personal pronouns distinguish two person values (first and second), two numbers (singular and plural) and encode a binary nominativeaccusative case distinction (second person object pronouns do not encode a number distinction). Subject pronouns have full and reduced forms. Table 2.4 summarizes the subject/object distinctions of free pronominal forms.


Table 2.4: Free personal pronouns

Subject pronominal forms may also be encoded though second (Wackernagel) position enclitics, attaching immediately after the first accented phrase or subconstituent of a phrase. Table 2.5 presents the paradigm of enclitic pronominal forms.

Table 2.5: Pronominal enclitic forms


Some representative examples of pronominal forms are provided in (10).

(10) a. *niˈhê ˈmí saʔˈpá aˈwênima*

*neˈhê* 1sg.nom *ˈmí* 2sg.acc *saʔˈpá* meat *aˈwê-ni-ma* grill-appl-fut.sg 'I will grill meat for you.' 'Te voy a asar una carne.' {SFH el685:02:20.5}

### 2 Grammatical overview

	- *ke* neg *taˈmí* 1sg.acc *ˈà-ki* give-pst 'S/he didn't give me any.' 'No me dió.' {JLG co1237:05:48.1}

Third person arguments may be left unmarked (11a), or they may be encoded through demonstratives (e.g., *ˈétʃ͡i* in (11b)) or through an emphatic pronoun (e.g., *biˈnôi* 'himself' in (11c)).


Demonstratives in Choguita Rarámuri may function anaphorically as pronouns or as nominal modifiers and encode two degrees of distance and orientation with

2.3 Discourse particles

respect to the speaker/addressee: (i) *ˈnà* ('this one') is proximal and encodes closeness to the speaker; and (ii) *ˈétʃ͡i* 'that one' is proximal and encodes closeness to either the addressee or the speaker. A third demonstrative (*ˈhê*) is exclusively found in complement clauses of utterance predicates and other complementtaking predicates introducing quoted speech.

### **2.3 Discourse particles**

Choguita Rarámuri possesses a large amount of discourse particles that constitute a set of heterogeneous word classes. These classes are closed and morphologically simple, bearing no inflection or derivation. Each class of particles is composed of fewer than a dozen members per class. A subset of these forms are presented in Table 2.6, with their form, meaning, gloss and cross-reference to the section of the grammar where they are discussed in detail (if available).


Table 2.6: Discourse particles

Discourse particles in Choguita Rarámuri have a wide range of functions and meanings. These include modality particles that encode epistemic stance, such as the particles *(o)ˈla* and *(a)ˈlé*, which encode certainty or doubt, respectively, that speakers may have about the actuality or likelihood that an event takes place. Epistemic particles are highly frequently, found post-verbally with futuremarked verbs. Examples of the contrast between these epistemic markers are provided in (12).

(12) a. *ˈnârma ˈlé*

*ˈnâri-ma* ask-fut.sg *aˈlé* dub '(He) will probably ask.' 'Probablemente va a preguntar.' < BFL 05 1:152/el >

### 2 Grammatical overview

b. *ˈnârmo ˈlá ˈnâri-ma* ask-fut.sg *oˈlá* cer 'S/he will definetly ask.' 'Seguramente que va a preguntar.' < BFL 05 1:152/el >

Another highly frequent discourse marker is *=ko*, a pragmatic enclitic that confers prominence to a word or phrase within discourse. In some contexts it functions as a topic marker, attaching to nouns whose referents have been introduced previously in discourse (13a). In other contexts, *=ko* functions as a pragmatic focus marker (13b).


Finally, the particle *pa* marks syntactic and/or discourse boundaries and is also frequently found in natural discourse (see also (4) above and (13b)).

In addition to these discourse markers, Choguita Rarámuri possesses a large set of negative markers. Two of these forms are basic negative particles, *ke*, a clausal negator (which can be used as interjection), and *ˈkíti*, a prohibitive (negative imperative) particle. Negative particles may combine with other morphemes, yielding morphologically complex negative markers. The set of negative particles and complex negative markers available in the language are provided in Table 2.7, with their gloss, function and approximate translation.

### 2.4 Nouns and noun phrases


Table 2.7: Negative markers

### **2.4 Nouns and noun phrases**

Nouns in Choguita Rarámuri may be inflected for instrumental and locative case and may exhibit complex patterns of possessive marking. A subset of nouns (those referring to animate referents) may also be marked as plural.

Instrumental and locative case markers are productive. Locative case is encoded with two suffixes: *-tʃ͡í* 'on, at', with an adessive reading, and *-rare* 'at, among, between', with an inessive reading, e.g., *saʔpaˈtʃ͡í* 'in the (live, human) flesh' vs.*saʔˈpáriri* 'in the (severed, animal's) meat', reflecting an alienable/inalienable distinction. The choice between locative suffixes for some nouns is however a matter of lexical choice, and some nouns may be optionally marked with either locative suffix (e.g., *maˈtá-riri* ~*mata-ˈtʃ͡í* 'in the metate').

Possessive constructions in Choguita Rarámuri encodes possessors and meronymic (part-whole) relationships. Inalienable (kinship and body-part terms) and alienable nouns may register a possessor with the stress-shifting possessive *-lâ* suffix. Alienable nouns are characterized by being able to appear in non-possessive constructions, while inalienable nouns are exclusively found with possessive marking. The possessive *-lâ* suffix attaches to the head noun in a possessive nominal phrase, following the head-marking profile of the language. Examples of the possessive constructions are shown in (14).

### 2 Grammatical overview

### (14) Possessive marking


```
sandals-poss
'X's sandals'
'sus huaraches' < BFL 09 1:60/el >
```
	- *boʔe-ˈlâ* road-poss 'X's road' 'su camino' < BFL 06 5:128/el >

In addition to this construction, possessives may be marked in a 'double possessive' marking pattern with a *-wa-la* suffix sequence, e.g., *waˈsá-wa-la* 'X's sowing field' ('su campo de cultivo'). For some speakers, the single vs. double marking of possession encodes a singular/plural distinction, e.g., *ne waˈrî-la* 'my palm basket' ('mi canasta de palma') vs. *ne waˈrî-wa-la* 'my palm baskets' ('mis canastas de palma').

Nouns may be modified by demonstratives, adjectives, numerals, definite articles and quantifiers in noun phrases. One common modifier found in noun phrases are definite articles, which encode number and affective stance, either positive or neutral, on the one hand, or negative. The paradigm of definite articles is shown in (15).

(15) Choguita Rarámuri definite articles


### 2.5 Verbs

### **2.5 Verbs**

Verbal morphology in Choguita Rarámuri is largely concatenative, though exponence is also achieved through stress shifts, grammatical tone and other nonconcatenative processes. Furthermore, a large number of morphologically-conditioned phonological process affect the surface form of inflected verbs.

Verbal roots can be divided in three classes depending on their underlying stress and vowel specifications: (i) **Class 1** verbs are stressed (with fixed stress across paradigms) and do not exhibit any vocalic alternations (e.g., *beˈnè* 'to learn'); (ii) **Class 2** verbs are unstressed (exhibiting stress shifts when attaching stressshifting suffixes) and do not exhibit vocalic alternations (e.g., *suˈkú* 'to scratch'); and (iii) **Class 3** verbs are unstressed and undergo final root vowel raising in addition to stress shifts when attaching shifting suffixes (e.g., *raʔˈlá* 'buy'). These verbal classes are illustrated in Table 2.8.


Table 2.8: Choguita Rarámuri verbal root classes

Class 2 and Class 3 roots may undergo valence related alternations through the affixation of transitive and applicative vocalic suffixes that replace the final vowel of the stem. Table 2.9 schematizes the three-way contrast of these stems (syntactic/semantic gaps are symbolized by dashes).

Finite Choguita Rarámuri verbs are marked for tense and/or aspect, mood distinctions (including imperative and reportative) and voice, as well as number and, in the case of the past egophoric suffix, person, conflated in portmanteaux suffixes. Table 2.10 illustrates a subset of inflectional exponents in verbs of different prosodic characteristics.<sup>2</sup>

<sup>2</sup>Note this table does not provide an exhaustive list of possible TAM inflection in Choguita Rarámuri nor the verbs represented constitute a complete sample of prosodic types of verbs.

### 2 Grammatical overview


Table 2.9: Valence stem allomorphy

Table 2.10: Choguita Rarámuri verbal root classes


2.5 Verbs

While most inflectional categories involve non-flexive formatives, some constructions exhibit lexically conditioned allomorphy. This is illustrated by the imperative singular construction, which may be encoded by suffixes (e.g., *toˈkâ!* 'take it!'), a L tone (*iˈsì!* 'urinate!'), or a stress shift (e.g., *raʔiˈtʃ͡â!* 'speak!'). Other TAM exponents exhibit phonologically-conditioned allomorphy: the future singular has an allomorph *-ˈmêa* when attaching to unstressed roots (e.g., *to-ˈmêa* 's/he will take it'), and an allomorph *-ma* when attaching to stressed roots (e.g., *iˈsî-ma*).

This table also illustrates the morphologically-conditioned shifts undergone by unstressed roots (e.g., *ˈtò* 'take' and *raˈʔìtʃ͡a* 'speak'), tonal alternations in verb stems after stress shifts (e.g.,*raˈʔìtʃ͡a-li* 's/he spoke' vs.*raʔiˈtʃ͡â-bo* 'they will speak'), and morphologically-conditioned phonological effects, such as final stem vowel lengthening triggered by the past passive suffix *-ru* (e.g., *ˈtòo-ru* 'it was taken').

Clauses with verbs inflected for past egophoric, encoding an event carried out in the past by a first-person subject in statements and by a second-person subjects in question, have optional nominative-marked pronominal marking. This is exemplfieid in (16).

(16) [ˈpòləki] /ˈpòli-ki/ cover-pst.ego 'I covered it.' 'Lo tapé.' < AHF 05 1:125/el >

In addition to stress changes, grammatical tone patterns, and other morphologically-conditioned phonological effects, Choguita Rarámuri inflected verbs exhibit templatic effects in some morphological constructions. Specifically, syllable truncation is attested in denominal verb forms, body part incorporation and V-V compounding to satisfy requirements on output surface forms. An example is provided below, where disyllabic mood/aspect markers undergo syllable truncation in V-V compounding when attaching an outer inflectional suffix (17a–b), but remain disyllabic otherwise (17c).

(17) Templatic truncation in V-V compounding

```
a. [aˈtʃ͡ènisa]
    /aˈtʃ͡è-nale-sa/
     pour-desid-cond
    'if s/he wants to pour it'
    'si lo quiere echar, verter' < SFH 07 romara/tx >
```
2 Grammatical overview

> b. [tiˈtʃíiksima] /tiˈtʃíi-ki-simi-ma/ comb-appl-mot-fut.sg 's/he will go along making them comb her/him' 'va a ir haciéndola que la peine' < SFH 07 2:67/el > c. [ˈnârisimi] /ˈnâre-simi/ ask-mot 's/he is going along asking' 'va a ir preguntando' < SFH 08 1:148/el >

### **2.6 Word order**

Choguita Rarámuri is a head-final language with canonical SOV word order. In ditransitive clauses with non-pronominal noun phrases, the order of arguments is S-T(heme)-R(ecipient), with variable placement of object noun phrases with respect to the verbal predicate. Word order in ditransitive clauses is exemplified in (18) (other ordering possibilities are exemplified and discussed in §13.1.4).

(18) Ditransitive clauses


Nominal modifiers precede head nouns in noun phrases, as exemplified in (19) (noun phrases are marked in square brackets).

2.7 Appositive possessive constructions and relative clauses

(19) a. *ˈtʃ͡êram ˈsûs ba, ti ˈtʃ͡êram bauˈtîʃ ma ba [ˈtʃ͡êrame* elder *ˈsûs]* Jesús *ba,* cl *[ti* def.sg *ˈtʃ͡êrame* elder *bauˈtîsi]* Bautista *ma* also *ba* cl 'elder Jesús, also elder Bautista' 'Don Jesús, también Don Bautista' {ME in484:07:18.5} b. *biˈlé ariˈmúli, oˈkwâ ariˈmúli ma, biˈkiá ariˈmúli ma [biˈlé* one *ariˈmúli]* decaliter *[oˈkwâ* two *ariˈmúli]* decaliter *ma* or *[biˈkiá* three *ariˈmúli]* decaliter *ma* or 'one decaliter or two decaliters or three decaliters' 'un decalitro o dos decalitros o tres decalitros' {LEL tx68:00:25.8} c. *waʔˈlû kapaˈnî aˈnítʃ͡ini ba [waʔˈlû* big *kapaˈnî]* bell *aˈní-tʃ͡ini* make.sound-ev *ba* cl 'the big bell rang (was heard)' 'se oyó sonar la campana grande' {LEL tx223:03:29.1}

### **2.7 Appositive possessive constructions and relative clauses**

As discussed in §2.4 above, Choguita Rarámuri has possessive constructions that are head-marking. In addition, there is an alternative way of encoding possession in the language through an appositional construction with *ˈníwa*, a grammatically specialized possessive noun used in possessive constructions derived from a verb ('to have'). The root *ˈníwa* as a possessive noun is marked with the possessive *-lâ* suffix and is part of a possessive phrase, exemplified in (20).

(20) *ˈnà ko ˈnè ˈníala ˈlîbro ko ˈnà=ko* prox=emph *ˈnè* 1sg.nom *ˈníwa-lâ* own-poss *ˈlîbro=ko* book=emph 'This here is my book.' 'Este es mi libro.' < BFL 06 4:187-189/el >

Another type of complex noun phrase involves relative clause formation. Choguita Rarámuri exhibits two types of headed relative clauses: (i) those formed via nominalization (21a) and (ii) those that involve finite predicates and subordinators such as *ˈ(n)api* (21c–d).

### 2 Grammatical overview

```
(21) a. ˈétʃ͡i ˈtʃ͡îba muˈkúami ko baˈsûa koˈʔáli
            ˈétʃ͡i
            dem
                 ˈtʃ͡îba
                 goat
                       muˈkú-ami=ko
                       die.sg-ptcp=emph
                                           ma
                                           already
                                                    baˈsû-a
                                                    cook-prog
                                                                koˈʔá-li
                                                                eat-pst
           'that dead goat (goat that is dead) was already eaten cooked by the
            dwellers of (the ones that inhabit) that house'
           'esa chiva (que está) muerta ya se la comieron cociéndola los (que
            habitan) de esa casa' {LEL tx_mawiya:02:43.9}
       b. aʔˈlì ˈétʃ͡i ˈnápu roˈwéma ˈlé ko biˈnôi biˈlá aˈní
            aʔˈlì
            and
                ˈétʃ͡i
                dem
                      [ˈnápi
                      sub
                             roˈwé-ma
                             run.womens.race-fut.sg
                                                        aˈlé]=ko
                                                        dub=emph
                                                                     biˈnôi
                                                                     herself
                                                                             biˈlá
                                                                             indeed
            aˈní
            say.prs
           'and then the one who will run, herself, says'
           'y entonces la que va a correr, ella misma, dice' {LEL tx19:00:39.8}
       c. ripuˈrá...ˈnápu riˈká miˈtʃ͡ípu kuˈʃì ba?
            ripuˈrá
            ax
                   [ˈnápi
                   sub
                           riˈká
                           that
                                miˈtʃ͡ípu
                                carve
                                         kuˈsì]
                                         sticks
                                                ba
                                                cl
           'an ax with which to carve the sticks'
           'hacha con que labrar los palos' {SFH in61:03:30.6}
```
### **2.8 Complement clauses and clause chaining**

Choguita Rarámuri has four major types of complement clauses: (i) finite complement clauses with complementizer (§15.1.1); (ii) interrogative complement clauses (§15.1.2); (iii) asyndetic finite verb complement clauses (§15.1.3); and (iv) reduced complement clauses (§15.1.4). The first type is exemplified in (22), where the complement clause is introduced by the subordinator *ˈ(n)api* (also used to introduce other subordinate clauses, such as relative clauses and adverbial clauses).

(22) *ka ni maˈtʃ͡íki ˈnápu ˈtòoru ba ka=ni* neg=1sg.nom *maˈtʃ͡í-ki* know-pst.ego *[ˈnápi* sub *ˈtò-ru* take-pst.pass *ba]* cl 'I didn't know he had been taken' 'no sabía que se lo habían llevado' < BFL 09 1:39/el >

Choguita Rarámuri also has specialized constructions that involve complementation, including a periphrastic construction encoding indirect causation. In this

### 2.8 Complement clauses and clause chaining

construction, exemplified in (23), a main jussive predicate takes the caused event as a complement and is characterized by the following properties: (i) the lower verb is additionally marked with the jussive verbal affix *nula* 'order, command' deriving a co-lexicalized structure within the complement; and (ii) although there are two causative verbs, the causer is expressed only once.

	- b. *ˈémi taˈmí aˈnèki niˈhê ˈtònula ˈémi* 2pl.nom *taˈmí* 1sg.acc *aˈn-è-ki* tell-appl-pst.ego *[niˈhê* 1sg.nom *ˈtò-nula]* take-order 'You all told me to take it.'

'Ustedes me hicieron que me lo llevara.' < BFL 06 4:94/el >

c. *ˈʔwínula tʃ͡e ˈtâsa riˈké la ˈró [ˈʔwí-nula* harvest-order *tʃ͡e]* again *ˈtâ-sa* ask.for-cond *riˈké* dub *la* perhaps *ˈró* dub 'Maybe we can ask (him/her) to harvest again.' 'A lo mejor le pedimos que vuelva a pizcar.' < BFL 06 4:94/el >

As shown in these examples, only *nula* may appear co-lexicalized with the lower predicate regardless of the specific jussive predicate in the matrix clause (*nula* 'order' (23a), *aˈnè* 'tell' (23b) or *ˈtâ* 'ask' (23c)).

A second type of specialized complementation strategy involves a reportative clause construction that features switch reference marking. These constructions involve a matrix clause with a speech predicate and a complement clause, the content of the reported event. When the notional subjects are coreferential, the dependent verb is marked for tense/aspect and with the epistemic *-o* suffix (24a). When the notional subjects are not coreferential, the dependent verb suffixes the different referent reportative *-la* suffix (24b).

(24) a. *maˈrîa ko ke ʃiˈmíko ˈrú maˈrîa=ko* Maria=emph *[ke* neg *ʃiˈmí-ki-o]* go.sg-pst.ego-ep *ˈrú* say.prs 'Maria says she didn't go.' 'Dice María que no fue.' < BFL 09 3:115/el >

### 2 Grammatical overview

b. *maˈrîa ko ˈhê aˈní hoˈsê ke ʃiˈmíla ˈruá maˈrîa=ko* Maria=emph *ˈhê* it *aˈní* say.prs *[hoˈsê* José *ke* neg *siˈmí-la]* go.sg-rep.dr *ru-ˈwá* say-mpass 'Maria says that José didn't go.' 'Dice María que José no fue.' < BFL 09 3:115/el >

In Choguita Rarámuri clause chaining structures, one of the clauses may be marked with canonical inflection, while other clauses in the chaining structure can only be marked with special inflection (the gerundive suffix *-ká*) and overall involve more restricted structures. This inflection mainly conveys a temporal relation of chronological overlap or chronological sequence (temporal notions that may have extended semantic meanings in some contexts): the marked clause may conveys that two events (drinking and resting) take place simultaneously (25a) or that the events conveyed occur in a temporal sequence (25b).

(25) a. *ˈwé pi ko ne ku iˈsâbika baˈhîba ˈlé ˈwé* int *pi=ko* just=emph *ne* int *ku* rev *[iˈsâbi-ka]* rest-ger *baˈhî-ba* drink-irr.pl *aˈlé* dub 'they need to drink while they rest' 'necesitan tomar descansando' {FLP in243:17:22.2} b. *ku aˈwílitʃ͡i ʃimiˈká, wiˈrómpo ˈkútʃ͡i paˈtʃ͡î ba [ku* rev *aˈwílitʃ͡i* ritual.patio *ʃimi-ˈká],* go.sg-ger *wiˈróm-po* make.blessing-fut.pl *ˈkútʃ͡i* def *paˈtʃ͡î* corn *ba* cl 'having gone back to the ritual patio, we make the blessing (lit. "throw the water") with corn'

'yendo al patio ritual, hacemos la bendición ("echamos el agua") con el maíz' {ME in485:07:11.4}

### **2.9 Complex predicates**

There are four types of constructions in Choguita Rarámuri that may be broadly characterized as involving complex predicates: light verb constructions (§15.6.1), auxiliary verb constructions (§15.6.2), serial verb constructions (§15.6.3) and multipredicate verb constructions involving V-V incorporation (exemplified in (17) above and addressed in §15.6.4).

An example of a light verb construction is provided in (26b). The verb *noˈká* may be used as a main verb form as a change of posture predicate (glossed as 'move') as in (26a) or it may be a semantically bleached verb bearing inflection

### 2.9 Complex predicates

in multi-predicate constructions with activity verbs bearing descriptive content, as in (26b)

(26) a. *ma noˈkáli ma* already *noˈká-li* move-pst 'S/he already moved.' 'Ya se movió.' <BFL 05 1:114/el>

> b. *napaˈwía noˈkáli lé ˈétʃ͡i ˈnà biˈlé riˈhò aʔˈlì biˈlé tʃ͡aˈbôtʃ͡i ˈʃîʔi [napaˈwí-a* get.together-prog *noˈká-li* do-pst *aˈlé]* dub *ˈétʃ͡i* dem *ˈnà* dem *biˈlé* one *riˈhò* man *aʔˈlì* and *biˈlé* one *tʃ͡aˈbôtʃ͡i* mestizo *ˈʃî* also 'A (Rarámuri) man and a *mestizo* (mixed mexican) man got together.' 'Se juntaron un hombre (Rarámuri) y un mestizo.' <SFH 06 choma(2)/tx>

In contrast, in Choguita Rarámuri auxiliary verb constructions, stative and inchoative posture predicates are deployed in auxiliary verb constructions encoding progressive aspect. The auxiliary verb bears tense marking (present, past or future), while the main lexical verb is inflected for present tense regardless of the tense marking on the auxiliary. This is exemplified in (27).

(27) a. *ma nataˈkêa buˈʔíli ˈnà biˈʔà roˈkò ma* already *[nataˈkê-a* faint-prog *buˈʔí-li]* lie.down.sg-pst *ˈnà* then *biˈʔà* early *roˈkò* night 'He had already fainted before dawn.' 'Ya estaba desmayado en la madrugada.' {LEL tx5:04:03.7} b. *"tʃ͡in oˈlá ko ˈétʃ͡i", ˈhê biˈlá ko ˈlàa aˈsáli ˈlé ruˈtûkuri ko ba tʃ͡i=ni* how=1sg.nom *oˈlá=ko* do.prs=emph *ˈétʃ͡i* dem *ˈhê* that *biˈlá=ko* indeed=emph *[ˈlà-a* think-prog *aˈsá-li]* sit.sg-pst *aˈlé* dub *ruˈtûkuri=ko* owl=emph *ba* cl "'That's how I did it to them" that's what he was thinking, the owl." "'Así les hice a esos" eso estaba pensando, el tecolote' {SFH tx152:07:01.9}

As shown in these examples, auxiliary verbs impose no selectional restrictions on the verbs they combine with (e.g., the descriptive verb may be a telic (bounded) predicate like *nataˈkê* 'faint' in (27a) or a stative predicate like *ˈlà* 'think' in (27b)).

# **3 Segmental phonology**

This chapter is devoted to the segmental phonological inventory and phonological processes of Choguita Rarámuri. While many phonological alternations in the language are morphologically conditioned (addressed in Chapter 8, Chapter 9 and Chapter 11), this chapter addresses those segmental processes that are analyzed as being fully phonologically general. This chapter also addresses patterns of phonetic variability of consonant phonemes.

The chapter is laid out as follows. §3.1 provides an overview of the Choguita Rarámuri sound system assumed in this and following chapters. §3.2 introduces the phonemic segmental inventory of the language. §3.3 illustrates the phonemic status of segments with minimal pairs. §3.4 describes the general phonological processes that yield the allophonic variation displayed by obstruents, nasals, rhotics, stops and vowels. The chapter concludes in §3.5 with a description of patterns of phonetic reduction of consonants.

### **3.1 Overview of the Choguita Rarámuri phonological system**

The phonological system of Choguita Rarámuri is characterized by a relatively small phoneme inventory, with eighteen consonants, five contrastive vowels and no contrastive vowel length, and a three-way laryngeal contrast for stops between pre-laryngealized stops (canonically realized as pre-aspirated voiceless stops), plain voiceless stops and voiced stops. This language is also characterized by a simple syllable structure in underlying representations, with no elaborate onsets and only glottal stop as a possible coda. While displaying low elaboration of the consonant inventory and a low level of complexity of syllabic structure (a correlation expected according to Maddieson 2005), Choguita Rarámuri displays a high degree of allophonic variation, as well as a complex word-prosodic system. Choguita Rarámuri syllable structure and syllabic processes are addressed in Chapter 4, while various aspects of its word prosody are addressed in Chapter 5 (stress), Chapter 6 (tone and intonation), and Chapter 7 (other word-level

### 3 Segmental phonology

suprasegmental phonological processes). Interactions between lexical, morphological and phrase-level phonological phenomena are addressed in Chapter 11, dedicated to prosodic interactions.

General phonological processes target almost all segments and involve different reduction processes. Alternations between plain (non-laryngealized) stops and voiced stops involve historically related pairs of segments (p ~ b, k ~ g, t ~ r), some of which have a phonemic status synchronically, but which may also display phonologically, morphologically or lexically conditioned allophonic variation. Many patterns of variation in surface forms are phonetic in nature and display different degrees of inter- and intra-speaker variation, are affected by speech rate and style, are always surface-transparanet and do not interact with phonological rules. Some widespread patterns of variation and optionality are addressed in the analyses presented in this and subsequent chapters.

In terms of its word prosody, Choguita Rarámuri has both stress and a threeway lexical tonal contrast. The complexity of the stress system in this language stems from several factors. First, stress conditions optional vowel reduction and deletion processes. While these processes are complex and, to some extent, gradient, there are at least three different identifiable patterns of vowel reduction targeting different vowel qualities, with more reduction occurring post-tonically than pre-tonically. Syncope yields derived heterosyllabic consonant clusters in coda position word-medially. Thus, surface forms display a moderate level of complexity of syllabic structure. Second, stress is restricted to an initial threesyllable window, a highly marked typological pattern (Kager 2012). Finally, the stress system of this language also features morphological factors that govern stress placement involving stress shift patterns depending on the lexical stress properties of roots and suffixes. The phonological and morphological properties of stress and associated morpho-phonological properties are addressed in Chapter 5.

Tone is exclusively realized in stressed syllables. While tone distribution is dependent on stress, both stress and tone are independent phonologically and encoded through different acoustic means (Caballero & Carroll 2015). Proto-Uto-Aztecan is reconstructed as having a stress-accent system (Munro 1977), with different languages belonging to different branches developing tonal contrasts in addition to stress (e.g., Hopi (Manaster-Ramer 1986), Northern Tepehuan (Tepiman; Woo 1970), Balsas Nahuatl (Aztecan; Guion et al. 2010), Cora (Corachol; McMahon 1967), Huichol (Corachol; Grimes 1959), Yaqui (Cahitan; Demers et al. 1999) and Mayo (Cahitan; Hagberg 1989)). Since the first reports of lexical tone in Choguita Rarámuri (in Caballero 2008 and Caballero & Carroll 2015), Morales Moreno (2016) also reports the existence of lexical tone in Rochéachi Rarámuri,

### 3.2 Phonological inventory

another Central Rarámuri variety, and it is possible that other Rarámuri varieties are also tonal, though no other variety of Rarámuri has been described as being tonal. Choguita Rarámuri also deploys f0 intonationally, resulting in different accommodation strategies when the lexical phonology, the morphology and the intonational phonology assign conflicting tones to the same target tone bearing units. In addition to f0, duration and non-modal phonation encode intonational meaning in the language. Tone and intonation are described in Chapter 6.

In addition to featuring a three-syllable stress window, the word-level prosody of Choguita Rarámuri features a disyllabic window for the distribution of the glottal stop. Distributional restrictions on the glottal stop and a three-syllable stress window have also been documented in the closely related Tara-Guarijío language Guarijío (Miller 1996, Haugen 2014, Haugen 2014), and window-like restrictions on stress placement are described across Rarámuri varieties (including Brambila's (1953) description of Norogachi Rarámuri and Morales Moreno's (2016) description of Rochéachi Rarámuri). In addition to the prosodically restricted distribution of glottal stop, other word-level suprasegmental phonological phenomena in this language include minimal word size restrictions and prosodic loanword adaptation processes. All of these are discussed in Chapter 7.

### **3.2 Phonological inventory**

### **3.2.1 Consonants**

Choguita Rarámuri has a relatively small consonant inventory and a high degree of allophonic variation. The phonemic consonant inventory, presented in Table 3.1, is significantly similar to proposed reconstructions for the Proto-Uto-Aztecan (PUA) consonant system (Voegelin et al. 1962, Miller 1967, Langacker 1977). In contrast to the reconstructed PUA consonant system, Choguita Rarámuri features a three-way laryngeal contrast in stops, with pre-laryngealized, unaspirated and voiced stops. Voicing is only contrastive at the bilabial and alveolar place of articulation (where, as discussed below, the voiceless alveolar plosive alternates with a voiced alveolar flap). Miller (1996) also documents a contrast between pre-laryngealized and plain voiceless stops in closely related Mountain Guarijío, though Mountain Guarijío differs from Choguita Rarámuri in lacking contrastive voicing of stops. In this chart, the voiced labiovelar approximant is represented in the bilabial column; the alveolar and retroflex flaps are represented orthographically as <r> and <l>, respectively. Pre-laryngealized obstruents are represented with pre-aspiration in this chart.

A broad phonemic transcription is used, and the symbols used in this table will be used throughout this grammar to represent these segments. The labiove-

### 3 Segmental phonology

Table 3.1: Phonemic inventory of Choguita Raramuri consonants


lar approximant is doubly assigned in the bilabial place of articulation column and in the velar place of articulation column. This notation indicates that the approximant is both velar and bilabial.

The consonant inventory presented in Table 3.1 includes a retroflex flap that can be characterized as a 'lateral flap', as defined by (Ladefoged & Maddieson 1996: 234). This segment is articulated by making a ballistic contact with the tongue tip in the post-alveolar region, but the sides of the tongue allow air to flow laterally, resulting in a sound that auditorily resembles both a lateral approximant and an alveolar (or, in the case of Choguita Rarámuri, slightly retroflexed) flap. The environments that favor the production/perception of the lateral variant are discussed in §3.4.3. This has led to orthographic representations of this sound as either <r> or <l>, including the name of the language, which is alternatively spelled 'Rarámuri', with word-medial <r>, or 'Ralámuli', with word-medial <l>. The contrast between the two rhotic sounds is ilustrated in (7) in §3.3 below.

As mentioned above, Choguita Rarámuri features a three-way laryngeal contrast of consonants, with pre-laryngealized, unaspirated and voiced consonants. Pre-laryngealized consonants are only attested as onsets of stressed syllables. 'Plain' obstruents are realized with no pre-laryngealization, are not restricted in terms of their distribution and may be realized with post-aspiration as onsets of stressed syllables as part of a process of onset augmentation (see discussion of acoustic correlates of stress in §5.1 and Caballero & Carroll 2015). Prelaryngealized consonants (canonically realized as pre-aspirated stops [ʰp, ʰt, ʰk, ʰtʃ͡]) are analyzed here as specified with a [+spread glottis] feature and 'plain' obstruents as laryngeally unspecified ([p, t, k, tʃ͡]).<sup>1</sup> Given that for some speakers

<sup>1</sup>This analysis is similar to the one provided in DiCanio (2012) for the lenis-fortis contrasts in San Martín Ituñoso Trique, where 'fortis' obstruents are realized with an early adduction (spreading) glottalic gesture, which results in a short period of pre-aspiration before the consonant (2012: 257). See also Arellanes Arellanes (2009) for a similar analysis of the lenis-fortis contrast in San Pablo Güilá Zapotec.

### 3.2 Phonological inventory

pre-aspirated stops are variably realized as pre-aspirated or pre-glottalized, the term pre-laryngealized is used throughout this grammar when referring to these segments.

For some speakers pre-laryngealized consonants can be realized without any laryngeal features in stressed position, neutralizing in the surface with 'plain' consonants, an instance of inter-speaker variation in the realization of these segments. Some examples are shown in (1).<sup>2</sup>

(1) Variable neutralization between pre-laryngealized and 'plain' stops

```
a. [reˈh
         tê] ~ [reˈtê]
     /reˈʰtê/
    'stone'
    'piedra' < GFM 14 1:111 >
b. [riˈʰtʃí] ~ [riˈtʃí]
     /riˈʰt͡ʃí/
    'landslide'
    'reliz' < BFL CR2014-LSC:13 >, < GFM 14 1:111 >, < SFH
     CR2014-LSC:86 >
c. [aˈʰkâ] ~ [aˈkâ]
     /aˈʰkâ/
    'to be sweet or salty'
    'ser dulce o salado' < SFH CR2014-LSC:85 >
```
These patterns of variation in the articulation of pre-laryngealized stops may suggest that the pre-laryngealized contrast may be marginal for some speakers, though most speakers realize the contrast consistently.<sup>3</sup>

There are no voiced coronal or velar stops in the phonological inventory of Choguita Rarámuri. Allophonic alternations show that the voiced counterpart of /t/ is a coronal flap (/ɾ/) (there is no voiced coronal stop /d/ in the Choguita Rarámuri phonological inventory).

### **3.2.2 Vowels**

Choguita Rarámuri makes a phonemic distinction among five cardinal vowels in stressed position. There is no contrastive vowel length. Mid vowels are phoneti-

<sup>2</sup> In (1c), there is post-aspiration as onset augmentation when pre-laryngealization is neutralized in a stressed syllable.

<sup>3</sup>A caveat is that even if pre-aspiration seems to disappear in the speech of some speakers it may leave residual breathy voice on preceding vowels (Marc Garellek p.c.). Whether this is the case in Choguita Rarámuri is a question that remains for future research.

### 3 Segmental phonology

cally open-mid and the only back vowels are rounded. The vocalic inventory is given in Table 3.2.


Table 3.2: Choguita Rarámuri Monophthong Vowel System

Most other Rarámuri varieties are reported as having five cardinal vowels, the qualities of which mostly match the inventory given for Choguita Rarámuri (Brambila 1953, Lionnet 1972, Morales Moreno 2016, inter alia), except for Ojachichi Rarámuri, which is described as having a high, back, *unrounded* [ɯ] vowel (Servín 2002).

### **3.3 Minimal pairs**

### **3.3.1 Consonant minimal pairs**

This section presents minimal pairs that demonstrate the phonemic status of Choguita Rarámuri consonants. Allophonic variation and patterns of neutralization are addressed in §3.4.

The examples in (2) include (near-)minimal pairs that illustrate the contrast between pre-laryngealized and plain (unaspirated) voiceless stops.

(2) Choguita Rarámuri pre-laryngealized obstruent vs. plain obstruent contrast

```
a. [wiˈʰkâ]
    /wiˈʰkâ/
    'many'
    'muchos' {BFL tx191:04:21.5}
```
b. [wiˈ**k**á] /wiˈ**k**á/ 'constellation, plow' 'constelación, barbecho' {ME in484:12:17.5}

3.3 Minimal pairs

c. [ri**ˈʰt**ûli] /ri**ˈʰt**û-li/ freeze-pst 'It froze.' 'Se congeló.' {SFH el1028:00:43.8} d. [riˈ**t**ù] /riˈ**t**ù/ 'encourage' 'animar' {LEL tx19:01:44.2} e. [na**ˈʰp**ô] /na**ˈʰp**ô/ 'prickly pear' 'tuna' {BFL el728:07:09.4} f. [naˈ**p**ò] /naˈ**p**ò/ 'to break, intr.' 'romperse' {MAF el1242:10:13.7} g. [ri**ˈʰtʃ**í] /ri**ˈʰtʃ**í/ 'landslide' 'reliz' < LEL 14 1:9 > h. [riʔˈ**tʃ**ì] /riʔˈ**tʃ**ì/ 'paternal uncle younger than father' 'tío paterno menor que el padre' < LEL 14 1:9 > i. [o**ˈʰk**ó] /o**ˈʰk**ó/ 'pine tree' 'pino' {LEL tx84:05:00.2} j. [ro**ˈk**ò] /ro**ˈk**ò/ 'night'

'noche' {LEL tx68:00:31.3}

### 3 Segmental phonology


The minimal pairs in (2a–f) show that pre-aspirated obstruents contrast with plain stops. This contrast is illustrated in Figure 3.1 (with pre-aspirated [ʰk]) and Figure 3.2 (with the plain (unaspirated) voiceless counterpart [k]).

As shown in (2h, l, n), plain stops may be preceded by a glottal stop. These glottal stop-consonant sequences are analyzed here as (heterosyllabic) consonant clusters ([ʔC]). An alternative (suggested by an anonymous reviewer) is to treat these sequences as pre-glottalized consonants ([ʔC]) that are only licensed in stressed onsets, which would parallel the distribution of pre-aspirated stops in the language (a restriction widely documented in languages with prelaryngelized consonants (Clayton 2010)). An argument against this analysis is that there is evidence that glottal gestures before consonants are not restricted to a stressed onset, e.g. *baʔtʃ͡i-ˈlâ* 'his older brother' (in (2n)) and *aʔwiˈjó* 'germinated' {LEL tx68:00:51.2}. While the glottal stop may be reduced in unstressed position (e.g., in the context of stress shift, where the glottal may be produced as glottalization of an adjacent vowel or sonorant), the glottal gesture is consistently realized regardless of stress conditioning, which contrasts with the realization of pre-aspirated consonants. Thus, [ʔC] consonant sequences largely pattern like other (heterosyllabic) consonant clusters in the language (for discussion of (derived) consonant clusters in the language, see §4.2). The only mismatch between

<sup>4</sup>This root has a L tone for some speakers. The extent to which inter-speaker tonal variation is attested in the Choguita Rarámuri corpus and factors that may govern this variability are questions left for future research.

3.3 Minimal pairs

Figure 3.1: Spectrogram showing realization of pre-aspirated [ʰk] in the word [o**ˈʰk**ól] /oˈʰkó-li/ '(in the) pine tree' {LEL tx84:05:00.2}.

[ʔC] consonant sequences and other consonant clusters result from the restriction that applies to the glottal stop, regardless of whether the glottal stop occurs inter-vocalically or pre-consonantally (see §7.1).

Finally, while these examples show that pre-aspirated stops contrast with plain stops that may be preceded by a glottal stop, for some speakers pre-aspirated stops are variably realized as pre-aspirated or pre-glottalized, e.g., *aˈʰkà ~ a<sup>ʔ</sup> ˈkà* 'sandal' in a pattern that resembles a pattern of optional pre-aspiration or prelaryngealization of fortis consonants in Southern Nevada Northern Paiute (SNNP; Western Numic) (Kataoka 2010).<sup>5</sup>

<sup>5</sup>Different Numic varieties exhibit different reflexes of the reconstructed gradation system of Proto-Numic, with the lenis series involving voicing, spirantization, and rhotacization, with no lenition of the fortis series. In Mono Lake Northern Paiute (MLNP; Western Numic), a system with lenis, voiced fortis and fortis oral stops and affricates, the fortis-lenis contrast is argued to involve closure duration, aspiration duration and voicing during the closure interval (Babel et al. 2013: 234). In the closely related Southern Nevada Northern Paiute (SNNP; Western Numic) the main correlate of the fortis-lenis contrast involves the discontinuity of the acoustic signal for the fortis series, achieved by longer consonant duration and optional co-occurrence of preaspiration or preglottalization (Kataoka 2010).

### 3 Segmental phonology

Figure 3.2: Spectrogram showing realization of plain [k] in the word [ro**ˈkò**] 'night' {SFH el1028:00:43.8}.

The contrast between voiced and voiceless stops at the bilabial and alveolar place of articulation is shown in (3). There are no voiced coronal or velar stops in the phonological inventory of Choguita Rarámuri. The voiced counterpart of /t/ is a coronal flap (allophonic alternations involving these two segments in the phonology of this language are discussed below in §3.4.3).

(3) Voicing contrast for bilabial and alveolar oral stops

$$\begin{array}{ll} \text{a.} & \text{[paítjí]} \\ & \text{/} \text{paítjí} \\ & \text{"corn cob"} \\ & \text{"elote"} \{\text{MDH} \text{ co1}\text{3} \text{6} : \text{03} : \text{36.2}\text{]} \\ \text{b.} & \text{[ba}^{\text{n}} \text{f} \widehat{\text{i}} \text{]} \\ & \text{/} \text{ba}^{\text{n}} \text{f} \widehat{\text{i}} \text{/} \\ & \text{"zucchini"} \\ & \text{"calabacita"} \{\text{LEL t} \text{t} \text{130} : \text{00} : \text{52.5}\text{]} \end{array}$$

3.3 Minimal pairs


Given that the contrast between pre-laryngealized and plain voiceless stops is neutralized in unstressed word-initial position (since pre-laryngealized stops are only attested as onsets of stressed syllables), these minimal pairs reveal a contrast between voiced stops, on the one hand, and either plain voiceless stops or pre-laryngealized voiceless stops, on the other.<sup>7</sup>

The voiced bilabial stop contrasts with the labiovelar semivowel. This is shown in the examples in (4)(for a description of processes targeting labio-velar glides in different positions within the syllable, see §4.4.2 in Chapter 4.)

<sup>6</sup>As attested in this example, the coronal flap is allophonically realized as a trill in word-initial position.

<sup>7</sup>As pointed out by Marc Garellek (p.c.), voiced stops in these examples could potentially be also analyzed as underlyingly plain voiceless, given a process of gradient lenition of these segments which yields voiced segments in fast speech (addressed below in §3.5.1). In these cases, however, we would expect voiced stops to undergo different lenition processes. For instance, word-initial voiced bilabial stops may reduce to labiovelar glides ([**w**aˈkotʃ͡i] /**b**aˈkôtʃ͡i/ 'river'< SFH 04 1:17/el >), as described in more detail in §3.4.5. In other words, lenition of stops in Choguita Rarámuri is generally non-neutralizing, a generalization that appears to hold cross-linguistically (Gurevich 2013).

### 3 Segmental phonology

(4) /w/ vs. /b/


The contrast between the voiced bilabial stop and the labiovelar semivowel in word medial position is shown in (5a–b).

(5) /b/ vs. /w/

a. [ka**ˈb**í] /ka**ˈb**í/ 'to roll something' 'enrollar' < BFL 07 el321/el >

b. [ka**ˈw**í] /ka**ˈw**í/ 'sunrise' 'amanecer' < BFL 07 el321/el >

The contrast between presence and absence of glottal stop before a consonant is shown in (4c–d) and also in (6).<sup>8</sup>

<sup>8</sup>The near-minimal pair *niˈwì* 'marry' and *niʔˈwí* 'to be lightning' differ in terms of absence or presence, respectively, of /ʔ/, in addition to lexical tone differences.

### 3.3 Minimal pairs

(6) /ʔ/ vs. Ø

	- /ra**ʔ**nè/ 'shoot at' 'disparar' {LEL tx221:02:22.7}

The contrast between the alveolar flap (<r> [ɾ]) and the lateral flap (<l> [ɽ]) is shown in (7).

(7) /r/ vs. /l/


### 3 Segmental phonology

```
'silly, distracted'
'tonto, distraído' {MDH co1136:09:42.1}
```

/**r**oˈwé/ 'to run (women's) *ariweta* race' 'correr carrera de *ariweta* {SFH el497:02:56.0}

The contrast between the two rhotic sounds is illustrated in Figure 3.3 (with the alveolar flap phoneme <r> [ɾ] realized as a trill word-initially) and Figure 3.4 (with the retroflex flap phoneme <l> [ɽ] word-initially).

Figure 3.3: Spectrogram showing realization of the alveolar flap phoneme <r> [ɾ] as a trill word-initially in [iˈwé **r**oˈwétia] /iˈwé **ɾ**oˈwéti-a/ 'The girls will run (a women's race)' {LEL tx19:00:26.5}.

3.3 Minimal pairs

Figure 3.4: Spectrogram showing realization of the retroflex flap phoneme <l> [ɽ] word-initially in [we **l**oˈwèma] /we **l**oˈw-è-ma/ 'you stir it well' {BFL tx60:02:34.2}.

The contrast between the bilabial nasal phoneme and the alveolar nasal phoneme is evidenced in the contrasts depicted in the verbal stems in (8).

	- a. [**m**iˈhí] /**m**iˈhí/ 'cook mezcal' 'cocinar mezcal' < SFH 07 2:12/el >
	- b. [**n**iˈhî] /**n**iˈhî/ 'give away, gift' 'dar, regalar' < ROF 04 1:67/el >
	- c. [maˈlála] /**m**aˈlá-la/ daughter.male\_ego-poss

3 Segmental phonology

```
'his daughter'
     'su hija de él' < BFL 05 1:155/el >
d. [naˈlàla]
     /naˈlà-la/
     cry.pot
     's/he can cry'
     'puede llorar' < SFH 05 1:69/el >
```
The next examples show that the alveolar fricative (9a, c) contrasts with the alveo-palatal affricate (9b, d).

(9) /s/ vs. /tʃ͡/


```
'plant'
'sembrar' {LEL tx130:03:07.8}
```
The next set examples show the contrast between pre-aspirated voiceless velar stops and ʔ-voiceless velar consonant sequences (10).

(10) /ʰk/ vs. /ʔk/

a. [aˈ**ʰ**kâ] /aˈ**ʰ**kâ/ 'sweet/salty' 'dulce/salado' < BFL 07 el324/el >

3.3 Minimal pairs

b. [a**ʔ**ˈká] /a**ʔ**ˈká/ 'sandal' 'huarache' < BFL 07 el324/el > c. [oˈ**ʰ**kó] /oˈ**ʰ**kó/ 'pine tree'

'pino' < BFL 07 el323/el > d. [o**ʔ**ˈkô] /o**ʔ**ˈkô/

'pain, to hurt'

'dolor/doler' < BFL 07 el323/el >

Choguita Rarámuri possesses only two semivowels: a voiced labiovelar /w/ and a palatal approximant /j/. The phonemic status of the labiovelar semivowel has already been shown in (5) and (4) above. Some of the examples in (5) and (4) are repeated below in (11), where the voiced bilabial stop (11a, c) contrasts with the labiovelar semivowel (11b, d).

(11) /b/ vs. /w/


### 3 Segmental phonology

The phonemic status of the palatal semivowel, on the other hand, is evidenced in the minimal pairs in (12), where /j/ (12a, c) contrasts word-medially with a glottal stop (12b, d).

```
(12) /j/ vs. /ʔ/
```
	- /koˈ**ʔ**á/ 'eat' 'comer' < SFH 04 1:69/el >

More about the phonemic status of labiovelar and palatal semivowels, as well as processes related to these segments, are discussed in §3.4.

### **3.3.2 Vocalic minimal pairs**

The examples in (13) show minimal pairs involving high vowels (the contrasting vowels are indicated with boldface).

(13) /i/ vs. /u/

	- /t͡ʃ**u**ˈkú-li/ be.bent-pst

### 3.3 Minimal pairs

```
'It was (bent).'
'Estaba (curvado).' {LEL tx177:08:12.8}
```

Minimal pairs involving back vowels are shown in (14).

```
(14) /o/ vs. /u/
```
a. [ˈt**ô**] /ˈt**ô**/ 'bury' 'enterrar' {MAF el1240:3:29.5} b. [ˈt**û**] /ˈt**û**/ 'down' 'abajo' {LEL tx5:01:52.3} c. [ˈk**o**] /ˈk**o**/ 'emphatic clitic (emph)' 'clítico enfático' {SFH tx43:06:58.7} d. [ˈk**ù**] /ˈk**ù**/ 'wood'

'leña' {FLP in61:05:51.9}

Front vowels also create phonemic contrasts, as shown in (15).

(15) /i/ and /e/

a. [ˈw**î**] /ˈw**î**/ 'harvest' 'cosechar' {SFH el549:04:42.0} 3 Segmental phonology

```
b. [ˈwe]
     /ˈwe/
    'very'
    'muy, mucho' {LEL tx84:07:28.3}
c. [ˈti]
     /ˈti/
    'singular definite article, negative evaluation (def.bad)'
    'artículo definido singular, evaluación negativa' {LEL tx5:04:51.7}
d. [ˈté]
     /ˈté/
    'lice'
    'piojos' < SFH 04 1:17/el >
```
Below, the contrast between (16a) (with the verb *biˈtí,* 'stand up up') and (16b) (with the verb *beˈte* 'stay overnight') shows that the contrast between front vowels is not restricted to stressed position (although there is a widespread vowel reduction process that targets unstressed vowels (discussed in detail in §5.2).

(16) /i/ vs. /e/ in pre-tonic position


Finally, the examples in (17–18) show contrasts between [−high] vowels.

(17) /e/ vs. /a/

a. [ˈ**è**] /ˈ**è**/ 'take away' 'llevar' < BFL 07 el336/el >

3.3 Minimal pairs

b. [ˈ**ʔà]** /ˈ**ʔà**/ 'give' 'dar' < BFL 07 el336/el > c. [iˈréli] /iˈr**é-**li/ lock-pst 'It locked it.' 'Lo encerró.' < BFL 07 VDB/el > d. [iˈr**á**ri] /iˈr**á**ri/ 'godparent, godchild' 'padrino, madrina, ahijado, ahijada' (18) /o/ vs. /e/ vs. /a/ a. [ˈm**ó**] /ˈm**ó**/ 'go up, sg.' 'subir' {FLP in243:20:58.8} b. [ˈm**ê**] /ˈm**ê**/ 'win' 'ganar' {MAF el1242:01:56.3} c. [ˈm**à**] /ˈm**à**/ 'run' 'correr' {LEL tx177:05:20.8} d. [ˈt**ò**] /ˈt**ò**/ 'Give it to me!' '¡Dámelo!' e. [ˈt**é**] /ˈt**é**/ 'lice' 'piojo' < SFH 04 1:17/el > f. [ˈt**á**] /ˈt**á**/

3 Segmental phonology

```
'small'
'pequeño' {JLG co1238:01:53.2}
```
The next section lays out details of several general phonological processes targeting both consonantal and vocalic segments.

### **3.4 Processes**

The following is a comprehensive description of phonological processes targeting consonantal segments in Choguita Rarámuri, including palatalization of alveolar fricatives, nasal place assimilation, processes targeting rhotics, postconsonantal devoicing and spirantization of voiced bilabial stops.

### **3.4.1 Palatalization of alveolar fricatives**

Alveolar fricatives palatalize before high vowels. The fricative palatalization rule is schematized in (19).

(19) Fricative palatalization rule s → ʃ / [+ high] V

The degree of palatalization is subject to speaker variation: for many speakers, the allophone is realized as a slightly retroflexed sibilant. Some speakers, however, produce a full-fledged alveopalatal fricative. The following examples show palatalized fricatives in word-medial (20a–c) and word-initial position (20d–f).

(20) Fricative palatalization before high vowels

```
a. [kaˈʃì]
    /kasì/
    'to shatter'
    'quebrar' < ROF 04 1:59/el >
b. [oˈʃì]
    /os-ì/
    write.read-appl 'to read or write for someone'
    'escribirle o leerle a alguien' < JHF 04 1:5/el >
c. [buʃuˈrê]
    /busurê/
    'to wake up'
```
'despertarse' < BFL 05 1:133/el >

3.4 Processes


'Irá.' < BFL 06 tx48(24)/tx >

The rule of fricative palatalization is counterbled by post-tonic vowel deletion. In (21), palatalization overapplies since the trigger high vowel from the underlying representation has been deleted.

(21) Opaque fricative palatalization

a. [aˈtí**ʃ**li] /aˈtísi-li/ to.sneeze-pst 'S/he sneezed.' 'Estornudó.' < BFL 05 1:111/el > b. [miʔˈà**ʃ**nula] /miʔˈà-si-nula/ to.kill-mot-order 'S/he made them kill it.' 'Hizo que lo matara.' < BFL 06 4:145/el > c. [nuˈlè**ʃ**ti-] /nuˈl-è-si-ti-/ to.command-mot-caus- 'S/he makes them go along commanding.' 'Los hace ir mandando.' < BFL 06 4:145/el > d. [wiliˈbá**ʃ**nila]

/wili-ˈbá-si-nula/ stand-inch-mot-order 'S/he makes them go along standing.' 'Lo hace que se vaya parando.' < BFL 06 1:10 > 3 Segmental phonology

The overapplication is schematized in (22):

```
(22) Counterbleeding
      Underlying Representation /atísi-li/
```

There are thus many palatalized fricatives with no surface-apparent conditioning environment.

### **3.4.2 Optional nasal place assimilation**

Nasal phonemes in Choguita Rarámuri display very little allophonic variation. Alveolar nasals display optional nasal place assimilation when they precede bilabial stops. This process, as shown in (23), is fed by post-tonic vowel deletion (the stressed vowels undergo compensatory lengthening, cf. §9.4.3.2).

(23) Optional nasal place assimilation

a. [ˈée**m**pi] /ˈéna-pi/ go.pl-irr.fut.pl 'Maybe they will go.' 'A lo mejor van a ir.' < BFL 07 el35/el >

	- /ˈsû-ni-po/ sew-appl-fut.pl 'They will sew for them.' 'Les van a coser (a ellas).' < SFH 05 1:80/el >

The optionality of nasal place assimilation is evident from examples like (24), where an underlying alveolar nasal does not assimilate to a following bilabial stop after post-tonic vowel deletion (the forms with asterisk indicate unattested forms with vowel deletion and nasal place assimilation).

3.4 Processes

	- a. [ˈtʃóoo**n**po] /ˈtʃ͡ó**n**i-po/ step-fut.pl 'They will step.' 'Van a pisar.' < BFL 07 VDB/el > *\*ˈtʃ͡óompo* b. [ˈjóo**n**po] /ˈjó**n**i-po/ nag-fut.pl 'They will nag.' 'Van a regañar.' < BFL 07 VDB/el > *\*ˈjóompo* c. [ˈlàa**n**po] /ˈlà**n**i-po/ bleed-fut.pl 'They will bleed.' 'Van a sangrar.' < ROF 04 1:82/el > *\*ˈlàampo* d. [roˈmíi**n**po] /roˈmí-**n**a-po/ fold-tr-fut.pl

'They will fold it.' 'Van a doblarlo.' < AHF 05 1:126/el > *\*roˈmíimpo*

There is no evidence that alveolar and bilabial nasals assimilate in place of articulation to other (non-bilabial) following segments. Some relevant examples are shown in (25).

(25) No nasal place assimilation

a. [ˈtâa**n**ko]

/ˈtâ-ni-ki-o/ ask-appl-appl-ep 'S/he asks them'. 'Le pregunta.' < BFL 07 1:31/el >

### 3 Segmental phonology


In (25a–c) post-tonic deletion yields a heterosyllabic cluster with a nasal followed by a voiceless velar; the alveolar nasals, however, are not velarized. Bilabial nasals do not assimilate in place of articulation to following alveolar or velar stops either (as shown in (25d–f)).

Finally, there is no place assimilation when an alveolar nasal precedes a bilabial nasal, as exemplified in (26).

(26) No place assimilation of alveolar nasal [raˈjè**n**ma] /raˈjèni-ma/ be.sunny-fut.sg 'It will be sunny.' 'Va a estar soleado.' < SFH 06 1:128/el >

3.4 Processes

There are no cases in the corpus where an alveolar nasal assimilates in place of articulation to a following bilabial nasal.

### **3.4.3 Processes targeting rhotics**

Recall from §3.2.1 above that there are two rhotics in Choguita Rarámuri, an alveolar flap (/ɾ/ <r>) and a lateral flap (/ɽ/ <l>). The alveolar flap is realized as an alveolar trill word-initially. The rule in (27) represents this exceptionless generalization.<sup>9</sup>

(27) Word-initial allophone of /ɾ/ /ɾ/ → [r] / #

The examples in (28) include an IPA notation (where [r] represents the alveolar trill and /ɾ/ stands for the alveolar flap in the underlying representation). Since alveolar trills and flaps are in complementary distribution, these sounds are represented with <r> in the rest of the grammar.

(28) Word-initial trill

```
a. [raʔˈlá]
     /ɾaʔˈlá/
     'to buy'
     'comprar' < AHF 07 1:74/el >
b. [roˈnò]
     /ɾoˈnò/
     'to boil'
     'hervir' < SFH 04 1:81/el >
c. [reˈká]
     /ɾeˈká/
     'to lay down'
     'acostarse' < SFH 04 1:67/el >
d. [ritiˈwá]
     /ɾitiˈwá/
     'to watch'
     'ver' < JHF 04 1:2/el >
```
<sup>9</sup>An anonymous reviewer asks whether this process may be due to contact with Spanish, where a similar pattern is attested. According to Merrill & Burgess (2014), word-initial [r] in Rarámuri varieties developed from a \*t > r sound change in the early nineteenth century (2014: 242), at a time period where colonial expansion had taken place. This change could thus have involved contact-induced transfer of a phonological pattern, though it may also be an internal development.

3 Segmental phonology

```
e. [ruˈhì]
     /ɾuˈhì/
    'to hail'
    'granizar' < SFH 04 1:123/el >
```
Alveolar trills are not uncommon in word-medial position in loanwords from Spanish: [ba**ˈr**îkatʃ͡i], 'bucket', from *barrica*, and [mo**ˈr**âltʃ͡i], 'bag', from *morral*, are just a couple of examples. For some speakers, there are also trills in wordmedial position in native words: [haˈré], 'some' < LEL 06 5:100/el, BF 06 rec48/tx >, and [wasa**ˈr**á-ma] 's/he will plow' < BFL 06 tx48 nururía/tx >.

The lateral flap, as discussed in §3.2, is a sound that is auditorily both like a lateral approximant and like an alveolar (or retroflex) flap. Choguita Rarámuri back vowels following the flap condition its perception/production as a lateral.<sup>10</sup> While the perceptual salience of the lateral quality of the segment is stronger in word-initial position (29a–e), there are many examples of the lateral flap being unambiguously produced as a lateral word-medially (29f–i).

(29) Lateral allophone of the lateral flap

```
a. [ˈlàna]
     /ˈlàna/
    'bleed'
    'sangrar' < ROF 04 1:82/el >
b. [loˈká]
     /loˈká/
    'drink pinole'
    'tomar pinole' {MDH co1136:10:25.9}
c. [loˈwâ]
     /loˈwâ/
    'stir'
    'revolver' < SFH 04 1:71/el >
d. [laˈké]
     /laˈke/
    'carve (wood)'
    'labrar (madera)' < BFL 04 1:111/el >
```
<sup>10</sup>The same conditioning environment is reported to favor the lateral production of the lateral flap in Naasioi (East Papuan), Barasana (Tucanoan) and Tucano (Tucanoan) (Ladefoged & Maddieson 1996: 243). Lionnet (1972) identified the same conditioning enviroment for the lateral perception of the flap in Norogachi Rarámuri.

3.4 Processes

e. [**l**otʃ͡é] /loˈtʃ͡e/ 'be hungry' 'tener hambre' {JLG co1235:2:24.2} f. [riˈʰtû**l**o] /riˈʰtû-ri-li-o/ freeze-caus-pst-ep 'They froze them.' 'Los congelaron.' < BFL 05 1:112/el > g. [waˈ**l**ó] /waˈló/ 'dry' 'secar' < SFH 04 1:110/el > h. [tʃ͡iˈ**l**ó] /tʃ͡iˈlo/ 'to sizzle' 'chisporrotear' < BFL 05 1:154/el > i. [basaˈ**l**ôwi] /basaˈlowa-i/<sup>11</sup> stroll-impf 'They used to stroll.' 'Paseaban.' < BFL 05 1:162/el >

The flap is also optionally realized as a lateral in cluster with another consonant (after post-tonic syncope), including other rhotics (30a–b), voiceless stops (30c–d), fricatives (30e) and nasals (30f).

(30) Pre- and postconsonantal allophone of /l/

```
a. [ˈnârli]
     /ˈnâri-li/
     ask-pst 'They asked.'
    'Preguntaron.' {MDH co1136:3:13.3}
b. [ˈkôlri]
     /ˈkôli-li/
     visit-pst
    'They visited.'
    'Visitaron.' < BFL 04 1:111/el >
```
<sup>11</sup>A loan from Spanish *pasear*.

### 3 Segmental phonology

c. [tʃ͡iˈkó**l**tia] /tʃ͡iˈkóli-ti-a/ tickle-caus-prog 'They are making them tickle.' 'Los están haciendo haacer cosquillas.' < AHF 05 1:146/el > d. [kaˈpô**l**tia] /kaˈpôli-ti-a/ spherical-caus-prog 'They are making them spherical.' 'Los están haciendo esféricos.' < AHF 05 1:149/el > e. [ˈkó**l**sia] /ˈkóli-si-a/ be.spicy-mot-prog 'It's becoming spicy.' 'Se va haciendo enchiloso.' < ROF 04 1:82/el > f. [kaˈwé**l**nale] /kaˈwéli-nale-/ reject-desid- 'It wants to reject it.' 'Lo quiere rechazar.' < BFL 07 1:155/el >

In all of the examples above, the lateral flap forms a heterosyllabic consonant cluster with a preceding or following rhotic or oral stop after posttonic vowel deletion.

The following examples show the environments which favor the flap (nonlateral) variant of this phoneme. These environments, as shown in (31), overwhelmingly involve front vowels.

(31) Flap allophone of /l/

```
a. [naˈwàɾi]
     /naˈwà-li/
     arrive-pst
    'They arrived.'
    'Llegaron.' < JHF 04 1:1/el >
b. [koˈɾî]
     /koˈlî/
    'chile pepper'
    'chile' {MDH co1137:0:50.1}
```
3.4 Processes

c. [tʃ͡iˈkó**ɾ**ia] /tʃ͡iˈkóli-a/ have.itch-prog 'They are itchy.' 'Tienen comezón.' < AHF 05 1:146/el > d. [kaˈpô**ɾ**itia] /kaˈpôli-ti-a/ be.round-caus-prog 'They are making it round.' 'Lo está haciendo redondo.' < AHF 05 1:149/el > e. [ˈkô**ɾ**i] /ˈkôli/ 'to visit' 'visitar' < BFL 05 1:111/el > f. [nakaˈwé**ɾ**i] /nakaˈwéli/ 'to reject' 'rechazar' < BFL 07 1:155/el > g. [ˈmêni**ɾ**i] /ˈmê-nále/ win-desid 'They want to win.' 'Quieren ganar.'< ROF 04 1:81/el > h. [simiˈná**ɾ**e] /simí-ˈnále/ go.sg-desid 'They want to go.' 'Quieren ir.' < SFH 05 1:86/el >

The flap allophones are produced slightly retroflexed for some speakers. For other speakers, however, the flap allophone is indistinguishable from the alveolar flap phoneme. For each example in (31), however, there is evidence that the flap has a related form with the lateral allophone. For instance, the past suffix /-li/ is realized with a flap allophone in (31a), but with a lateral allophone in (30a) above after posttonic syncope.

### 3 Segmental phonology

### **3.4.4 Post-consonantal devoicing**

Finally, oral stops are also subject to a general phonological rule: without exception, stops devoice post-consonantally. In (32), post-tonic vowel deletion yields an environment in which the onset of the future plural is necessarily voiceless. For instance, as shown in (32a), *ˈnâar-po*, but not *\*ˈnâar-bo*, is unattested after posttonic deletion. There are no examples in the Choguita Rarámuri corpus with a voiced/lenis allophone appearing post-consonantally.

	- a. [ˈnâarpo] /ˈnâri-**p**o/ ask-fut.pl 'They will ask.' 'Van a preguntar.' < SFH 07 in243/in > *\*ˈnaar-bo* b. [desfiˈlârpa] /desfiˈlâr-**p**a/ parade-fut.pl 'They will paarade.' 'Van a desfilar.' < LEL 06 Nov5/el > *\*desfiˈlâr-ba* c. [bamˈpása] /bam-ˈ**p**á-sa/ year-inch-cond 'if they become older' 'si cumple años' < SFH 06 tx12/tx > *\*bam-ˈba-sa* d. [samˈpá] /sam-ˈ**p**á/ be.wet-inch 'to become wet' 'mojarse' < SFH 04 1:113/el > *\*sam-bá*

There is evidence that the onsets of the causative, future plural, and inchoative in (32) have voiced onsets with the same roots in other morphological contexts (this is addressed in detail in Chapter 9). The examples in (33) show that the voiceless allomorphs in (33b) and (33d) are not lexically determined, but

3.4 Processes

phonologically-conditioned. The root *bami* has an inchoative suffix allomorph with a voiced onset in (33a) and a voiceless allomorph in (33b) after pre-tonic vowel deletion; the transitive stem *rapa-na,* 'to split, tr' (*'partir*, tr'), has a future plural suffix allomorph with a voiced onset in (33c) and a voiceless suffix allomorph in (33d) after post-tonic vowel deletion. <sup>12</sup>

	- a. [baˈmíbali] /baˈmí-**b**a-li/ year-inch-pst 'They turned one year older.' 'Cumplieron años.' < SFH 06 tx12/tx >
	- b. [bamˈpása] /bam-ˈ**p**á-sa/ year-inch-cond 'If they turn one year older.' 'Si cumplen años.' < SFH 06 tx12/tx > *\*bam-ˈba-sa*
	- c. [rapaˈnâbo] /rapa-ˈnâ-**b**o/ split-tr-fut.pl 'They will split it (the wood).' 'La van a partir (la leña).' < AHF 05 1:131/el >
	- d. [raˈpámpo] /raˈpá-m-**p**o/ split.appl-tr-fut.pl 'They will split it for her/him.' 'Se la van a partir (la leña).' < AHF 05 1:131/el > *\*raˈpá-m-bo*

These alternations result from application of the rule in (34), which is fed by pre- or posttonic vowel deletion.

(34) Post-consonantal stop devoicing [+ voice] stop → [-voice] / C

<sup>12</sup>Some lexical items have variable pronunciations with bilabial stop and bilabial nasal alternants. The positional verbal predicate for liquids /maˈna/, for instance, has alternative pronunciations with a bilabial nasal stop (*maˈna*) and with a voiced bilabial oral stop (*baˈna*).

### 3 Segmental phonology

### **3.4.5 Spirantization of voiced bilabial stops**

Word-medially, voiced bilabial stops are lenited and realized as voiced bilabial fricatives or approximants intervocalically, and, for some speakers, as labiovelar semivowels pre-consonantally after further weakening. The spirantization rule is schematized in (35a) and the pre-consonantal gliding of bilabial stops is schematized in (35b). Pre-consonantal gliding is fed by post-tonic vowel deletion.<sup>13</sup>

	- a. /b/ → [β] ~ [β] / V ̞ V b. /b/ → [w] / \_C

The following examples show the lenition of /b/ to a voiced fricative ([β]) (36a– b) and a voiced approximant ([β]) ( ̞ 36c–d) intervocalically.

```
(36) Intervocalic lenition of /b/
```

```
a. [ziˈβóò]
    /si-ˈbô/
     go.pl-fut.pl
    'Let's go!'
    '¡Vamos!' {MFH tx1133:01:17.2}
```

<sup>13</sup>As suggested by an anonymous reviewer, the interaction between vowel deletion and preconsonantal gliding may be argued to show that either (i) vowel deletion is phonological, despite being a variable process, or (ii) that both vowel deletion and pre-consonantal gliding are both phonetic processes. Vowel deletion interacts with other variable phonological processes (e.g., nasal place assimilation, described in §3.4.2).

3.4 Processes

'They sat down.' 'Se sentaron.' < BFL 06 6:73/el >

The forms in (37) exemplify the lenition of /b/ to a labiovelar glide ([w]) preconsonantally.

(37) Pre-consonantal lenition of /b/

a. [ˈè**w**tiki] /ˈè**b**i-ti-ki/ bring-caus-pst.ego 'I made him/her bring it.' 'Lo hice traerlo.' < BFL 06 6:73/el >

b. [muˈtʃ͡í**w**po] /muˈtʃ͡í-**b**a-po/ be.siting.pl-inch-fut.pl 'They will sit down.' 'Se van a sentar.' < BFL 06 6:73/el >

More examples of the word-medial labiovelar semivowel allophone of /b/ are provided in (38).

(38) Pre-consonantal [w] allophone of /b/

a. [ˈì**w**ma] /ˈì**b**i-ma/ bring-fut.sg 'S/he will bring it.' 'Lo va a traer.' < BFL 06 6:75/el >


### 3 Segmental phonology

In word-initial position, underlying /b/ may surface as voiced bilabial stops (e.g., (39a)) or undergo lenition and surface spirantized (e.g., (39b)). This latter pattern is the most frequently attested in the Choguita Rarámuri corpus.

(39) Word-initial allophones of /b/

```
a. [biˈnèʃia]
    /beˈnè-simi-a/
    learn-mot-prs
    '(they) go along learning'
    'van aprendiendo' {LEL tx73:02:03.9}
```
b. [**β**enˈtâantʃ͡i] /**b**enˈtânitʃ͡i/ 'window' 'ventana' {SFH tx152:02:14.4}

For some speakers, word-initial underlying voiced bilabial stops undergo gliding (surfacing as [w]), thus neutralizing the phonemic contrast between /w/ and /b/ in word-initial position. Some examples are given in (40).

(40) Word-initial neutralization of /w/ and /b/


3.4 Processes


As these examples show, this neutralization is prevalent before low, central vowels, and marginally attested before high, front vowels (cf. (40e)). There are no examples of this neutralization in word-initial position before mid, front vowels or round, back vowels. For the speakers that display this neutralization, the wordinitial underlying stop is produced as either a full-fledged labiovelar semivowel or a bilabial approximant; when asked to give a careful pronunciation of these words, these speakers produce a bilabial stop.

Underlying labiovelar semivowels, on the other hand, may be neutralized and have surface realizations as voiced bilabial approximants intervocalically. As the examples in (41) show, the neutralization is favored before /a/ (41a–b) and /i/ (41c– d). These are the same environments that favor velarization of /b/ in word-initial position (as shown in (40)).

	- a. [roˈtʃ͡í**β̞**ari] /rotʃ͡í**w**ari/ 'quelite' < SFH 07 6:163-175/el >
	- b. [sa**β̞**aˈróame] /sa**w**aróame/ 'yellow' 'amarillo' < SFH 07 6:163-175/el >
	- c. [wiˈkâ**β̞**i] /wikâ**w**i/ 'to forget, forgive' 'olvidar, perdonar' < SFH 07 1: 183/el >

3 Segmental phonology

> d. [reˈrò**β̞**i] /rerò**w**i/ 'potato' 'papa' < FLP 06 in61/in >

This neutralization is optional and is not attested with the onsets of stressed syllables. This is shown in (42).

(42) No glide hardening in stressed syllables

```
a. [baʔˈwíβa]
     /baʔˈwíwa/
    'icy rain'
    'agua-nieve' < SFH 06 6:74/el >
b. [loˈwâ]
     /loˈwâ/
    'stir'
    'revolver' < SFH 04 1:71/el >
c. [iˈwépi]
     /iˈwépi/
    'wrestle'
    'luchar' < SFH 04 1:96/el >
d. [riˈwè]
     /reˈwè/
    'leave'
```
'dejar' < AFH 05 1:181/el >

### **3.5 Phonetic reduction processes**

This section addresses patterns of reduction/lenition of consonant segments that appear to be phonetic in nature, being affected by speech rate and style, not exhibiting any morphological conditioning and not exhibiting interactions with clearly phonological rules. This phonetic reduction, which is perceptually highly salient to Choguita Rarámuri speakers as a social/regional marker, does not involve contrast neutralization, a crucial difference between this process and the lenition processes operating in the lexical phonology. Several of these patterns exhibit a high degree of inter- and intra-speaker variation.

### 3.5 Phonetic reduction processes

### **3.5.1 Lenition of voiceless plosives**

Plain (non-laryngealized) voiceless stops exhibit a large variety of surface realizations. Specifically, voiceless oral stops are gradiently realized within a continuum that ranges from a voiceless aspirated stop (if stressed) to a labiovelar glide or deletion altogether (schematically: [pʰ > p > b > β > β̞> w > Ø]). Voicing effects in these contexts are so strong that they are clearly perceptible without any instrumental analysis. This continuum for the bilabial place of articulation is exemplified in (43) with instances of the realization of *pa* (in boldface), a monosyllabic function word that marks clause and/or utterance boundaries (square brackets in the examples below indicate clausal boundaries in multiclausal sentences).<sup>14</sup>

### (43) Gradient production of *pa*, clause final marker (cl)


d. *aʔˈlì biˈlá ko waˈbé biˈlá kiˈʔà ˈníla ra pa kuˈrí ke ˈtʃ͡ó me biwaˈtʃ͡êatʃ͡i ˈnà kaˈwì β̞a*

<sup>14</sup>While these processes mostly target weak functional items like clitics and particles, voiceless stops in roots and affixes may also undergo this type of reduction, e.g. (46), (47).

### 3 Segmental phonology

*aʔˈlì* and *biˈlá=ko* indeed=emph *waˈbé* int *biˈlá* indeed *kiˈʔà* long.ago *ˈní-la* cop-rep *ra* say.prs *pa]* cl *kuˈrí* recently *ke* neg *ˈtʃ͡ó* yet *me* almost *biwaˈtʃ͡ê-a-tʃ͡i* solidify-prs-temp *ˈnà* dem *kaˈwì* earth *β̞a]* cl 'And so it was a long time before this earth was solid.' 'Y entonces fue mucho cuando todavía no amacizaba este mundo.' {SFH tx43:11:11.2}

As shown in these examples, the onset of this function word is realized as a voiceless stop (43a,d), a voiced stop (43a,b), a voiced fricative (43c) or a voiced approximant (43d), often displaying different realizations in different positions within the same utterance.

Voiceless velar stops also undergo gradient production, with variable voicing and spirantization. The forms in (44) exemplify this process with the emphatic enclitic *=ko* (in boldface).

	- a. *aʔˈlì biˈlá ko miˈtíira ˈlé riˈhò go, tʃ͡iˈhùna tʃ͡uˈkúlam ba ˈétʃ͡i tʃ͡oˈmá ba aʔˈlì* then *biˈlá=ko* indeed=emph *miˈtí-ra* win.pst.pass-rep *aˈlé* dub *riˈhò=go,* man=emph *tʃ͡iˈhùna* be.disgusted *tʃ͡uˈkúl-ame* be.curved-ptcp *ba* cl *ˈétʃ͡i* dem *tʃ͡oˈmá* snot *ba* cl 'And that is why the Rarámuri person was beaten, because he was disgusted by the snot.'

'Y por eso le ganaron al tarahumar, porque le estuvo teniendo asco al moco.' {SFH tx128:01:26.2}

b. *niˈhê ˈjêla go ˈpâma ˈlé koˈbísi eɾmoˈsîjo ˈka niˈhê* 1sg.nom *ˈjê-la=go* mom-gen=emph *ˈpâ-ma* bring-fut.sg *aˈlé* dub *koˈbísi* pinole *eɾmoˈsîjo* Hermosillo *ˈka* irr 'My mom will bring pinole to Hermosillo.' 'Mi mamá va a traer pinole a Hermosillo.' {SFH el444:00:57.8} c. *taˈmò ko ˈhê riˈgá ˈnòtʃ͡ami ˈhú ˈnà ˈwé iˈsêligam go ... taˈmò=ko* 1pl.nom=emph *ˈhê* like *riˈgá* that *ˈnòtʃ͡-ame* work-ptcp *ˈhú* cop *ˈnà* prox *ˈwé* int

*iˈsêli-g-am=go* be.governor.pl-g-ptcp=emph

'We, that is how we work, the governors.'

'Nosotros los gobernadores así trabajamos.' {JMF tx816:00:00.0}

3.5 Phonetic reduction processes

d. *maniˈké ˈlàr ba ni ˈmátimi aʔˈlá boˈsáli aˈlé baˈʰtâli go ba ne mani=ˈké* to.be.liquid=cop.imp *ˈlàr* think *ba* cl *ni* emph *ˈmá=timi* already=2pl.nom *aʔˈlá* well *boˈsá-li* full-pst *aˈlé* dub *baˈʰtáli=go* corn.beer=emph *ba* cl *ne* emph 'You all got full with the corn beer, I think.' 'Se llenaron ustedes con el tesgüino, yo creo.' {MFH tx1133:00:17.3} e. *tʃ͡iliˈká iˈnásma tʃ͡oˈná ˈhônsa ko ˈtʃ͡ó pa ne tʃ͡iriˈká biˈlá pe tʃ͡iliˈká* like.that *iˈná-s-ma* go.sg-mot-fut.sg *tʃ͡oˈná* that *ˈhônsa=ko* from=emph *ˈtʃ͡ó* again *pa* cl *ne* emph *tʃ͡iriˈká* like.that *biˈlá* indeed *pe* just 'So that you won't be thinking any of that.' 'Para que de eso no vayas pensando nada.' {MFH tx1133:01:06.0} f. *aʔˈlám riˈkátʃ͡imi aʔˈlá iˈwéami raʔaˈmâbi ˈlé ɣo a mi raʔaˈmâmi ˈlé paˈgótami baʔaˈlî aʔˈlám* well *riˈkátʃ͡i=mi* like.that=2.acc *aʔˈlá* well *iˈwé-ame* strong-ptcp *raʔaˈmâ-bi* give.advice-irr.pl *aˈlé=ɣo* dub=emph *a* aff *mi* there *raʔaˈmâ-mi* give.advice-irr.sg *aˈlé* dub *paˈgótami* people *baʔaˈlî* tomorrow 'Perhaps tomorrow people will give you all advice.' 'A lo mejor de aqui a mañana llegan gentes a darles consejos.' {MFH tx1132:00:30.3}

As with bilabial stops, velar stops display a range or surface realizations, ranging from voiceless stops (44a,c,e), voiced stops (44a,b,c,d), and voiced fricatives (44f). These effects are also attested in intervocalic position within morphologically complex words, as shown in (45):

	- a. *aʔˈlì mámi baʔˈwí roˈʔèma oˈhòsa aˈnáuka biˈlé baˈrîka aʔˈlì* and *má=mi* already=2sg.nom *baʔˈwí* water *roˈʔ-è-ma* pour.appl-fut.sg *oˈhò-sa* dekernel-cond *aˈnáu-ka* measure-ger *biˈlé* one *baˈrîka* cask 'When you dekernel it you pour water, having measured a cask.' 'Cuando lo desgranas y ya le echas agua, ya que mides una barrica.' {BFL tx60:00:27.2}

### 3 Segmental phonology

b. *riˈpákina ˈku tʃ͡uˈkúli ti tʃ͡imoˈrí ko iʔˈnèga tʃ͡ú oˈlása riˈpáki-na* above-abl *ˈku* rev *tʃ͡uˈkú-li* stand-pst *ti* dem *tʃ͡imoˈrí=ko* squirrel=emph *iʔˈnè-ga* watch-ger *tʃ͡ú* that *oˈlá-sa* do-cond 'And the squirrel stood watching when he did that to them.' 'Y la ardilla lo estuvo viendo de arriba cuando les hizo eso.' {BFL tx191:03:37.6}

These examples show how the voiceless velar stop (in the simultaneous action  *ka* suffix) emerges as either voiceless (45a) or voiced (45b) in intervocalic position. This effect is part of a phenomenon of variable voicing of velar stops that is also attested within lexical items, as shown in (46):

### (46) Gradient voicing of velar stops


Recall from §3.2.1 above that there is no voicing contrast at the velar place of articulation in Choguita Rarámuri. While variable voicing of velar stops is attested across Choguita Rarámuri speakers regardless of their exposure to other varieties of Rarámuri, voiced velar stops are identified by Choguita Rarámuri speakers as characteristic pronunciations of other dialects (e.g., Norogachi Rarámuri), and seem to function as a highly salient social/regional marker, though no detailed sociolinguistic study has been carried out in this area to date. Crucially, surface voiced velar stops have a more restricted distribution than surface voiced

3.5 Phonetic reduction processes

bilabial stops and coronal flaps in monomorphemic words and do not emerge in any morphological alternations.<sup>15</sup>

Finally, voiceless coronal stops also undergo lenition in fast speech, and this process is attested in intervocalic position. In contrast to bilabial and velar stops, coronal stops do not voice when undergoing lenition; instead, they undergo spirantization and a change in place of articulation. That is, in contrast to the pattern found in other morphological and phonological environments where the voiceless coronal stop alternates with a voiced coronal flap, the lenis counterpart of [t] in these fast speech contexts is a voiceless inter-dental fricative. This effect is exemplified in (47).

(47) Spirantization of voiceless coronal stops in fast speech

	- /maˈ**t**ê**t**ara/ 'thank you' 'gracias' {MFH tx1133:02:38.8}

Preliminary examination of phonetic data reveals these gradient realizations in fast speech reflect a lenition process in which productions on the lenis end of the continuum (fricatives and approximants) tend to be produced in utterance-final position, while productions on the fortis end (voiceless and voiced stops) tend to be produced utterance-medially, a factor that may suggest that these alternations are sensitive to phrasal phonological effects.<sup>16</sup>

While the precise role of post-lexical phonology in conditioning lenition in these environments is still under investigation, there is a clear effect of a constraint that precludes [+continuant] consonants ([β~~w~Ø]) in post-consonantal

<sup>15</sup>Voiced velar stops are only marginally attested in underlying forms in some place names, as in the form *basiˈgotʃ͡i* 'Basigóchi', a toponym containing the root *basiˈko* and the locative suffix *-tʃ͡í* that means 'place where *basiˈko* grows'. The root *basiˈko* is synchronically used to refer to a plant species, and it can be productively derived with the locative suffix, *basiˈko-tʃ͡i*. This last form with the voiceless velar stop in the stem plus the locative suffix has the meaning 'on top of the plant *basikó*'.

<sup>16</sup>Lenition results in the neutralization of the voicing contrast at the bilabial place of articulation, with the voiced bilabial stop also displaying a gradient realization ranging from a stop proper, to a fricative, an approximant, a labiovelar glide or deletion ([b~β~β̞̞~w~Ø]).

### 3 Segmental phonology

position, a constraint that applies across word boundaries, e.g., *oˈwaam pa*, *tʃ͡uˈkúlam ba*, *iˈsêligam go*. In the corpus data examined so far, there are no [-continuant] consonants in these environments, suggesting that a constraint like \*CC[+continuant] is at play. This reduction process does involve contrast neutralization, a crucial difference between this process and the lenition processes operating in the lexical phonology.

### **3.5.2 Depalatalization and deaffrication of alveopalatal affricates**

Alveopalatal affricates can optionally depalatalize before low, central vowels and can be produced as alveolar affricates, as schematized in the rule in (48). This allophonic variation is exemplified in (49), where alveopalatal affricates are optionally depalatalized root-internally (49a–c) or as onsets of suffixes (49d–f). Alternative pronunciations with the alveo-palatal affricate are also shown in (49).

	- a. [aˈkâ**ts**ala] ~ [aˈkâ**tʃ͡**ala] /aˈkâ**tʃ͡**a-la/ paternal.grandmother-poss 'their paternal grandmother' 'su abuela paterna' < FLP 07 in243/tx >
	- b. [aˈ**ts**âsa] ~ [aˈ**tʃ͡**âsa] /aˈ**tʃ͡**â-sa/ to.sit.sg.tr-cond 'if they sit it down' 'si lo sienta' < SFH 04 1:38/el >
	- c. [ˈnò**ts**ari] ~ [ˈnò**tʃ͡**ari] /ˈnò**tʃ͡**ari/ 'work' 'trabajo' < AADB(292)/el >
	- d. [kuˈsú-**ts**ani] ~ [kuˈsú-**tʃ͡**ani] /kuˈsú-**tʃ͡**ani/ make.animal.noise-ev 'It sounds like an animal is making noise.' 'Suena que canta (un animal).' < SFH 05 1:81/el >

3.5 Phonetic reduction processes

e. [oˈsì**ts**ana] ~ [oˈsì**tʃ͡**ana] /oˈsì-**tʃ͡**an-a/ write.read-ev-prog 'It sounds like they're writing/reading.' 'Suena que están escribiendo/leyendo.' < SFH 05 1:88/el > f. [tʃ͡iˈwá**ts**a] ~ [tʃ͡iˈwá**tʃ͡**a] to.tear-tr.pl 'to tear something many times' 'rasgar muchas veces' < ROF 04 1:59/el >

In fast speech, alveopalatal affricates may depalatalize and deaffricate in a high frequency word combination: the distal demonstrative *ˈétʃ͡i* and a following adjective, are pronounced as a single word within nominal phrases, with stress in the second syllable. This involves the deletion of an underlying high vowel, as shown in the examples from text provided in (50).

(50) Depalatalization and deaffrication of *tʃ͡*

a. e**s**ˈtá ˈtʃ͡êlami ko ba *ˈétʃ͡i* dist *ˈtá* det *oˈtʃ͡êlam=ko* old.man=emph *ba* cl 'that old man' 'ese viejo' < LEL 06 6:141-162/tx > b. ˈnè ko aniˈmê oˈlá ˈhê ˈnà e**s** ˈtá raʔˈìtʃ͡iri *ˈnè=ko* 1sg.nom=emph *ani-ˈmêa* say-fut.sg *oˈlá* cer *ˈhê* dem *ˈnà* prox *ˈetʃi* dist *ˈtá* det *raʔˈìtʃ͡i-ri* speak-nmlz 'I am going to say these words.' 'Yo voy a decir esta plática.' < BFL 07 frog story\_2/tx > c. we kaˈní-lam ˈtʃ͡ó ˈníra ˈlé é**s** ˈkútʃ͡i oˈkwâka *we* int *kaˈní-l-ame* love-r-ptcp *ˈtʃ͡ó* also *ˈní-ra* cop-pot *aˈlé* dub *ˈétʃi* dem *ˈkútʃ͡i* small *oˈkwâ-ka* two-coll 'They loved each other, those two, I think.' 'Se querían mucho también, yo creo, ellos dos.' < BFL 07 frog story\_2/tx >

The depalatalization process exemplified in (50a–b) can be analyzed as a reinterpretation of a phonetically ambiguous form, a consonant cluster with an alveopalatal affricate followed by an alveolar voiceless stop. This consonant sequence

### 3 Segmental phonology

would arise across word boundaries in these forms in an intermediate representation after posttonic deletion ({ˈetʃ͡ ˈtá}). The same analysis can be extended to other contexts of depalatalization of affricates: in (51) below, the alveolpalatal affricate undergoes non-local anticipatory assimilation with the following alveolar stop. The posttonic front, high vowel that intervenes between the alveopalatal affricate and the alveolar stop is extra short and can be deleted altogether in fast speech.

(51) Depalatalization and deaffrication of *tʃ͡* ma ˈkárua ˈkú**t**i toˈlí *ma* already *ˈkáru-a* cackle-prog *ˈkútʃi* small *toˈlí* chicken 'The chickens already cackled.' 'Ya cacaraquearon las gallinas.' < SFH 07 el170 (12:25)/el >

# **4 Syllables**

This chapter describes the syllabification patterns of Choguita Rarámuri. Native speakers segment words into syllables in careful speech pronunciation, and several suprasegmental processes make crucial reference to the syllabic structure of words, such as stress assignment (described in §11.3.5) and glottal stop prosody (described in §7.1).

The structure of this chapter is as follows. §4.1 provides an overview of the underlying syllabic structure of Choguita Rarámuri, followed by a description of attested derived consonant and vowel sequences in §4.2 and §4.3, respectively. Finally, §4.4 covers syllable structures involving semi-vowels.

### **4.1 Underlying syllable structure**

The underlying syllable in Choguita Rarámuri is an open syllable ((C)V). Onsets are optional, and never elaborate to a cluster. As in the closely related Taracahitic language, Guarijío (Miller 1996), the only possible (underlying) coda is glottal stop. CV, V, CVʔ, and Vʔ syllables in word-initial position are illustrated in (1–4). These syllable types are exemplified in monomorphemic words.

(1) Word-initial CV syllables

```
a. ta.ˈkí
    'instrumental violin piece'
    'canción instrumental en violín' < SFH 06 6:73/el >
b. pa.ˈkó
    'to wash (dishes)'
    'lavar trastes' < JHF 04 1:3/el >
c. ra.ˈpá
    'to split'
    'partir' < AHF 05 1:131/el >
d. ba.ˈkí
    'to go.in, sg'
    'entrar, sg' {BFL tx191:0:29.9}
```
### 4 Syllables

```
(2) Word-initial (onsetless) V syllables
      a. o.se.ˈrí
          'paper'
          'letter' {JLG co1235:10:21.5}
      b. i.ˈwí
          'breathe'
          'respirar' < JHF 04 1:1/el >
      c. u.ˈtʃ͡ú
          'sting (a bug)'
          'picar (un bicho)' < LEL 07 1:6/el >
     d. a.ˈtʃ͡í
          'to laugh'
          'reirse' < SFH 05 1:98/el >
(3) Word-initial CVʔ syllables
      a. saʔ.ˈpá
          'meat'
          'carne' {SFH tx43:06:07.8}
      b. biʔ.ˈwà
           biʔ.ˈw-à
          'to clean, tr'
          'limpiar, tr' < BFL 04 1:112/el >
      c. waʔ.ˈkó
          'grunt'
          'gruñir' < AHF 05 1:147/el >
     d. raʔ.ˈlá
          'buy'
          'comprar' < SFH 04 1:74/el >
(4) Word-initial Vʔ syllables
      a. uʔ.ˈká
          'choose'
          'escoger' < SFH 07 DB/el >
      b. aʔ.ˈtá
          'arch'
          'arco' < SFH 07 DB/el >
```
4.1 Underlying syllable structure

c. *iʔ.ˈpè* 'to gather small things' 'recoger cosas pequeñas' < BFL 06 4:189/el > d. *oʔ.ˈté.li* 'to burp'

'eruptar' < SFH 05 1:177/el >

The only types of syllables found word-medially in monomorphemic words are CV syllables. These are exemplified in (5). There are no word-medial CVʔ syllables (i.e., with glottal stop coda), since glottal stop is restricted to appear within a disyllabic window (as the coda of a word-initial syllable or the onset of the second syllable of the prosodic word) (for further details about this restriction, see §7.1).

### (5) Word-medial CV syllables

a. *ko.ri.ˈmè.ni* 'bees, honey' 'abejas, miel' {BFL tx191:01:27.6}


Word-medial onsetless (V) syllables are not attested in monomorphemic verbal roots, but can be found in polysyllabic nouns. Some of these nominal roots (e.g. (6a) and (6c)) are lexicalized compounds.

	- a. *a.wa.ˈkó.a.ni* 'scorpion' 'alacrán' < SFH NDB/el >

4 Syllables


Word-medial onsetless (V) syllables are found after morpheme boundaries in morphologically complex words. Examples are given in (7).

	- a. *wi.pi.ˈsó.a wipiˈsó-a* hit-prog 'S/he is hitting.' 'Está pegando.' {LEL tx223:03:03.5}
	- b. *bu.su.ˈrê.a busuˈré-a* wake.up-prog 'S/he is waking up.' 'Se está despertando.' {JMF tx816:00:16.6}
	- c. *su.ˈrí.a suˈrí-a* fight.over.something-prog 'S/he is fighting about something' 'Se está peleando.' < SFH 05 1:100/el >
	- d. *na.ˈkí.o naˈki-o* want-ep 'S/he wants' 'Quiere.' < BFL 04 1:91/el >
	- e. *no.ˈká.o noˈk-á-o* move.tr-ep

4.1 Underlying syllable structure

'S/he moves it.' 'Lo mueve.' < BFL 05 1:114/el >

f. *pi.ˈrê.o piˈrê-o* dwell.pl-ep 'They dwell.' 'Viven.' {LEL tx32:02:51.1}

All phonemic consonants can be onsets. There are no complex onsets. Table 4.1summarizes the types of syllables in Choguita Rarámuri and examples of each syllable type in word-initial and word-medial position.



Taking onset complexity (singleton vs. CC onsets) and presence and elaboration of codas as indices of syllable structure complexity (following Maddieson 2005), we can conclude that Choguita Rarámuri has a simple syllable structure in underlying representations: there are no elaborate onsets beyond a single consonant and no codas, except for glottal stop.

In surface forms, however, posttonic deletion gives rise to heterosyllabic consonant clusters, yielding a moderate level of syllable complexity in surface realizations.<sup>1</sup> This is discussed next.

<sup>1</sup>As discussed in more detail in §5.2 below, Choguita Rarámuri exhibits complex patterns of stress-based vowel reduction and deletion. I assume deletion is a categorical phonological process, with vowel deletion yielding a gestural reorganization that affects the syllabic structure of the resulting word. One pending question, however, is whether these cases involve extreme phonetic vowel reduction instead of a phonological process of deletion. As suggested by an anonymous reviewer, in the latter scenario we may expect the syllable pattern of the target word to remain the same, i.e., with the consonant sequence behaving as if the vowel would have remained in place, as proposed for vowel deletion involving some consonant clusters in Tokyo Japanese (Shaw & Kawahara 2022). I leave this question for future research.

### 4 Syllables

### **4.2 Consonant sequences**

There are very few restrictions as to the types of derived clusters that are possible in Choguita Rarámuri. One of these restrictions involves voiced oral stops. As described in more detail above (§4.1), there is a productive rule of oral stop devoicing in post-consonantal position, fed by stress-governed posttonic vowel deletion. There are, thus, no surface sequences involving a voiced stop following a voiceless consonant in a consonant cluster. This is schematized in (8).

```
(8) Stop devoicing
     *[-voice] C - [+voice] stop
```
Bilabial oral stops in general form heterosyllabic clusters with other voiceless stops and with nasal stops. Nasal stops can form consonant clusters with all other consonants except for other sonorants (although some CC sequences involving identical sonorants are discussed below).

Alveopalatal affricates, on the other hand, may form a heterosyllabic cluster with a following nasal stop (9a–b) or a velar stop (9c).

### (9) tʃ͡N derived clusters

a. [saʔˈmè**tʃm**a] /saʔˈmètʃ͡a-ma/ soak-fut.sg 'It will soak.' 'Se va a mojar' < BFL 05 1:135/el >

b. [rataˈbá**tʃn**i] /rata-bá-tʃ͡a-ni/


Another restriction on derived consonant clusters in Choguita Rarámuri involves alveopalatal affricates: a sequence of an alveopalatal affricate followed by an alveolar stop forms an illicit consonant cluster. Repairs include metathesis (10a–b), and progressive, assimilatory deaffrication of the alveopalatal stop (10c): 4.2 Consonant sequences

```
(10) Repair of tʃ͡t cluster
```

There are thus almost no restrictions as to possible derived consonant clusters in Choguita Rarámuri, except for sequences involving alveopalatal affricates and voiceless stops.

Posttonic vowel deletion also yields sequences of identical stops in morphologically derived environments. The most common type of derived geminates in Choguita Rarámuri involves both oral (11) and nasal (12) stops at the bilabial place of articulation.

(11) Bilabial oral stop geminates

```
a. [naˈtéppo]
    /naˈtépi-po/
    meet.up-fut.pl
    'They will gather.'
    'Se van a reunir.' < BFL 07 el339/el >
```
b. [tʃ͡omaˈî**p**o] /tʃ͡omaˈîpi-po/

cover.facefut.pl 'They will cover their face.' 'Se van a cubrir la cara.' < BFL 07 1:181/el >

	- 'Van a correr.' < JHF 04 1:19/el >

4.3 Vowel sequences

Marginally attested are stop geminates that are shown in (13). These geminates involve alveolar oral stops (13a), velar oral stops (13b), and alveolar nasal stops (13c).

(13) Marginal types of geminates

a. [naˈhî**tt**ipo]

/naˈhîti-ti-po/ become-caus-fut.pl They will make them turn into something.' 'Los van a hacer que se conviertan en algo.' < BFL 07 el339/el >


Fricative geminates ([ss]) and alveopalatal affricate geminates ([tʃ͡tʃ͡]) are not attested. In contexts where these sequences would arise, there are other mechanisms (such as syllable deletion) that block the application of vowel deletion (e.g., /aˈsísi-sa/ → [aˈsíi-sa] 'If she were to wake up', see §9.4.3.1 for more details).

Geminates derived through post-tonic vowel deletion are variably attested across speakers, as there are sequences of syllables with identical onsets that undergo haplology (see §9.4.3.1). The choice between derived geminates and haplology seems to be correlated with idiolects, although both phenomena are documented with all speakers in the Choguita Rarámuri corpus.

### **4.3 Vowel sequences**

While vowel length is not contrastive in Choguita Rarámuri, there are different morphophological and phonological sources for derived long vowels. Two of these sources, compensatory lengthening and passive-induced lengthening, are discussed in Chapter 9. A third source for derived long vowels is found at morpheme boundaries: in (14), the vowel initial progressive *-a* suffix creates a long vowel sequence with roots with final, stressed *a*.

### 4 Syllables

	- a. [huˈrâa] /huˈr**â-a**/ send-prog 'S/he is sending it.' 'Lo está mandando.' < BFL 05 1:151/el > b. [biʔˈwàa]
	- /biʔˈw**-à-a**/ clean-prog 'S/he is cleaning it.' 'Lo está limpiando.' < BFL 05 1:112/el >
	- c. [oˈsàa] /oˈs**à-a**/ read.write-prog 'S/he is writing.' 'Está escribendo.' {JLG co1238:02:40.7} d. [tʃ͡iˈwáa] /tʃ͡iˈw**-**á-a/ rip-prog 'S/he is ripping it.'

'Lo está trozando.' < ROF 04 1:104/el > The imperfective suffix *-e* undergoes vowel reduction: it is realized in the surface as a high front vowel [i] in post-tonic position, and yields a long vowel sequence with roots with final, stressed *i*, as shown in (15).

(15) Derived long high, front vowels

```
a. [suˈwíi]
    /suˈwí-i/
     eat.up-impf
    'S/he used to eat it up.'
    'Se lo acababa' < SFH 04 1:119/el >
```
b. [iˈsîi] /iˈs**î-i**/ urinate-impf 'S/he used to urinate.' 'Orinaba.' < SFH 05 1:80/el >

4.3 Vowel sequences


Finally, there are also minimal pairs developed through *h* deletion (16) and labio-velar semi-vowel deletion (17) in word-medial position. Both of these processes (exemplified in (16b) and (17b) below) yield a long vowel sequence. As pointed out by an anonymous reviewer, the fact that there is stress retraction in (16b) makes it clear that this form is not just a case of vowel hiatus (where /VhV/ → [V.V]), but rather involves a true long vowel.

```
(16) h deletion
```
a. [ˈnâta] /nâta/ 'think' 'pensar' {SFH tx12:01:57.2}

	- a. [ˈnârma] /ˈnâri-ma/ ask-fut.sg 'S/he will ask.' 'Va a preguntar.' b. [naˈwáruma] ~ [ˈn**áà**rma]
	- /na**ˈw**áru-ma/ send-fut.sg 'S/he will command.' 'Va a mandar.'

### 4 Syllables

Forms with derived long vowels in (16b) and (17b) coexist with forms with no *h*-deletion and labio-velar semi-vowel deletion. These forms are subject to a great deal of speaker variation, and are likely to represent a change in progress.

Other vowel sequences involve diphthongs, tautosyllabic V͡V sequences attested in both stressed and unstressed syllables. This analysis is based on how native speakers segment words into syllables in careful speech pronunciation. Crucially, there are no phonological processes in Choguita Rarámuri that allow probing the phonological behavior of these sequences as similar or different than single vowels. Attested diphthongs in Choguita Rarámuri include falling diphthongs (Vi). These are exemplified in (18).

### (18) Falling diphthongs

```
a. ei
     [seˈmèi]
     /seˈmè-i/
     play.violin-impf
    'He used to play the violin.'
    'Tocaba el violín' < SFH 05 1:85/el >
b. oi
     [reˈhòi]
     /reˈhòi/
    'man'
    'hombre' {JLG co1234:16:06.3}
c. ui
     [siˈkúi]
     /siˈkúi/
    'ant'
    'hormiga' < SFH 05 láchimi/tx >
d. ai
     [kaiˈnâma]
     /kaiˈnâ-ma/
     yield.harvest-fut.sg
    'It will yield harvest.'
    'Se va a dar la cosecha.' {SFH tx977:09:36.7}
```
These falling diphthongs occur lexically in word-final syllables (18a-18b), morpheme-internally after stress shifts (18c) and across morpheme boundaries in

4.3 Vowel sequences

word-final syllables (18a). As the examples in (18) show, in these V1V<sup>2</sup> sequences V1 is often but not necessarily stressed.

High front vowels are also attested in rising diphthongs with low central vowel off-glides (19a–b) and mid back vowel off-glides (19c-19d). Crucially, these vowel sequences are parsed as tautosyllabic by native speakers in unstressed syllables: as seen in these examples, these cases involve rising diphthongs in post-tonic position in inflected verbs where no further suffixes are added. These diphthongs are therefore all attested word-finally.

(19) Rising diphthongs


Finally, labio-velar onset semi-vowels turn into labio-velar offglides after posttonic vowel deletion targets the nucleus of the labio-velar semi-vowel onset. This is shown in (20).

### 4 Syllables

	- a. [ˈì.**w**i.li → ˈì**w**.li] /ˈì**w**i-li/ bring.for-pst 'S/he brought it for them.' 'Se lo trajo.' < BFL 06 5:75/el >
	- b. [ku.ˈtʃ͡í**w**i.ma → ku.ˈtʃ͡í**w**.ma] /kuˈtʃ͡í**wi**-ma/ have.kids-fut.sg 'S/he will have children.' 'Va a tener hijos.' < BFL 06 6:74/el >
	- c. [wi.noˈmî.**w**i.pi → wi.no.ˈmî**w**.pi] /winoˈmî**w**i-pi/ have.money-irr.pl 'Maybe they will have money.' 'A lo mejor van a tener dinero.' < BFL 06 6:74/el >

As discussed in §4.4.2, these labio-velar offglides can be weakened and monophthongized with its nucleus.

Finally, Choguita Rarámuri also has a series of heterosyllabic vowel sequences. Attested hiatus sequences across morpheme boundaries involve a stressed vowel followed by a low central vowel (21) or by a mid back vowel, (22). These forms contrast with the forms shown in (19, where parallel vowel sequences in unstressed syllables are parsed as tautosyllabic.

(21) Hiatus sequences with low central vowels

```
a. [ˈmé.a]
    /ˈmé-a/
     bring-prog
    'S/he is bringing it.'
    'Lo está trayendo.' < SFH 04 1:73/el >
b. [re.ˈʔè.a]
    /reˈʔè-a]
     play-prog
    'S/he is playing.'
    'Está jugando.' < SFH 04 1:76-78/el >
```
	- a. [ni.ˈkâ.o] /niˈk**â-o**/ bark-ep 'It barks.' 'Ladra.' < BFL 05 1:114/el >

4 Syllables

> b. [ˈpá.o] /ˈp**á-o**/ throw-ep 'S/he throws it.' 'Lo tira.' < BFL 04 VDB/el > c. [ni.ˈkè.o] /niˈk**-è-o**/ bark-appl-ep 'It barks to them.' 'Les ladra.' < BFL 07 VDB(53)/el > d. [pi.ˈrê.o] /piˈr**ê-o**/ dwell.pl-ep 'They dwell.' 'Viven.' < BFL 05 1:161/el > e. [na.ˈkí.o] /naˈk**í-o**/ want-ep 'S/he wants (it).' 'Quiere.' < BFL 04 1:91/el > f. [bo.ti.ˈwí.o] /botiˈw**í-o**/ sink-ep 'It sinks.' 'Se hunde.' < SFH 05 1:120/el > g. [ʃi.ˈrû.o] /siˈr**û-o**/ hunt-ep 'It hunts.' 'Caza.' < SFH 05 1:136/el > h. [na.ˈrú.o] /naˈr**ú-o**/ exist-ep 'It exists.' 'Existe.' < BFL 04 1:93/el >

So far, the examples presented show verbal roots with final stress followed by the progressive suffix *-a* (21) or by the epistemic suffix *-o* (22). There are

4.3 Vowel sequences

no underlying monomorphemic hiatus sequences with verbal roots (although semi-vowel deletion yields vowel hiatus sequences, as discussed below in §4.4.1). Root-internal vowel hiatus, however, is attested with nominal roots (23). As mentioned above, these root-internal hiatus sequences are only attested with stressed vowels followed by mid, central vowels. To date, no monomorphemic hiatus sequences with final mid back vowels have been recorded in the Choguita Rarámuri corpus.

(23) Nominal root-internal hiatus

$$\text{a. } \quad \begin{array}{l} \text{[t]} \text{\"o\text{'}, \text{k\text{\'e\'}}\text{\"i}]}\\ \text{\text{\'f\'o\'ak\'eri}\text{\'}}\\ \text{\text{\'m}\text{\'omontain dove'}\\ \text{\`a paloma de monte'} \sim \text{SFH NDB/ el } > 1 \end{array}$$


Hiatus sequences, thus, show an interesting asymmetry in the phonological behavior of words of different lexical categories: hiatus sequences are licensed with nominal roots, but are only attested across morpheme boundaries with verbal roots.<sup>2</sup>

<sup>2</sup> In Chapter 9, I discuss how vocalic (onsetless) suffixes may induce deletion of the preceding morpheme's vowel in morphologically defined contexts.

4 Syllables

### **4.4 Semi-vowels**

### **4.4.1 Semi-vowel deletion**

Choguita Rarámuri has word-medial palatal and labio-velar semi-vowels. The example words in (24–25) show that semi-vowels occur independent of any particular vowel quality.

(24) Word-medial palatal semi-vowels

```
a. [kiˈjótʃ͡i]
            /kiˈjótʃ͡i/
            'fox'
            'zorra' < SFH 08 1:103/el >
       b. [koˈjá]
            /koˈjá/
            'squat'
            'estar en cuclillas' < SFH 08 1:103/el >
       c. [naˈjú]
            /naˈjú/
            'to be sick'
            'estar enfermo' {MFH el1318:27:23.9}
       d. [hiˈjé]
            /hiˈjé/
            'to follow a trace'
            'seguir la huella', "huellar" < SFH 05 1:98/el >
(25) Labio-velar semi-vowels
       a. [kuˈwé]
            /kuˈwé/
            'dry season'
            'tiempo de secas' < NDB/el >
       b. [aˈríwi]
            /aˈríwi/
            'to sun-set (dusk)'
            'atardecer' < NDB/el >
       c. [maˈwétʃ͡i]
            /maˈwétʃ͡i/
            'bean cultivating field'
            'campo de cultivo de frijol'3 < NDB/el >
```
<sup>3</sup>This refers to the burnt field where beans are sown.

4.4 Semi-vowels

d. [kaˈ**w**ì] /kaˈ**w**ì/ 'hill, world' 'cerro, mundo' {SFH tx43:11:11.2}

Word-medial semi-vowels are optionally deleted.<sup>4</sup> The palatal semi-vowel, for instance, is optional when preceded by a low central vowel and followed by a stressed, front mid vowel. This is shown in (26).

(26) Optional deletion of palatal semi-vowels

```
a. [raˈjèniri] ~ [raˈèniri]
     /raˈjèniri/
     'sun'
     'sol' {SFH in485:7:04.8}, {SFH in484:10:54.4}
```

There are also examples of optional labio-velar semi-vowel deletion in the Choguita Rarámuri data. The examples in (27) involve a stressed mid, back vowel and a high, front vowel flanking the labio-velar semi-vowel. Optionally, these words are produced with a falling diphthong.

	- a. [siˈnó**w**i] ~ [siˈn**ói**] /siˈnó**w**i/

<sup>4</sup>As observed by an anonymous reviewer, a similar analysis is proposed in (Davidson & Erker 2014) arguing against the claim that glides are inserted in hiatus environments in English.

4 Syllables

```
'snake'
    'víbora' < SFH 04 1:17/el >
b. [reˈròwi] ~ [reˈròi]
    /reˈròwi/
    'potato'
    'papa' < SFH 07 NDB200/el >
```
There are also cases where there is no palatal (28a–c) or labio-velar (28d–f) semi-vowel deletion word-medially. These cases all involve words where the semi-vowel is the onset of a stressed syllable. The unattested forms in the second column in (28) show hypothetical forms with no word-medial semi-vowel.

### (28) No semi-vowel deletion


4.4 Semi-vowels

f. [naˈ**w**à] *\*naˈà* /naˈ**w**à// 'to arrive' 'llegar' < SFH 05 1:73/el >

While semi-vowels delete productively, there is no evidence of productive semi-vowel epenthesis in Choguita Rarámuri. The only forms with apparent epenthetic semi-vowels are shown in (29). In these examples, a monosyllabic root adds the applicative suffix *-è*. As shown here, either a palatal or a labio-velar glide may be epenthesized between the stem and the applicative suffix in a pattern of intra-speaker variation (speaker LEL produced both (29a) and (29b), and speaker ROF produced both (29c) and (29d).

(29) Lexicalized cases of semi-vowel epenthesis

a. [ruˈ**j**è] /ru-ˈè/ tell-appl 'to tell someone' 'contarle' {LEL tx32:8:42.9} b. [ruˈ**w**è] /ru-ˈè/ tell-appl 'to tell someone' 'contarle' {LEL tx32:5:13.9} c. [buˈ**j**è] /bu-ˈè/ wait-appl 'to wait for someone' 'esperarlo' < ROF 04 1:67/el > d. [buˈ**w**è] /bu-ˈè/ wait-appl 'to wait for someone' 'esperarlo' < ROF 04 1:67/el >

Other instances of applicative *-è* suffixation do not involve optional palatal and labio-velar semi-vowel epenthesis. This is exemplified in (30).

4 Syllables

```
(30) No semi-vowel epenthesis
      [ruˈèa]
      /ru-ˈè-a/
      tell-appl-prog
      'They are telling them.'
      'Les están diciendo.' {SFH tx43:04:43.0}
```
There is thus no evidence for a productive, general semi-vowel epenthesis process in the language.

The next section lays out details of several general phonological processes targeting both consonantal and vocalic segments.

### **4.4.2 Semi-vowel monophthongization**

Labio-velar onset semi-vowels turn into labio-velar offglides after posttonic vowel deletion. Some examples are provided in (31).

	- a. [ˈì.**w**i.li] → [ˈì**w**.li] /ˈì**w**i-li/ bring.appl-pst 'S/he brought it for them.' 'Se lo trajo.' < BFL 06 5:75/el >
	- b. [ku.ˈtʃ͡î.**w**i.ma] → [ku.ˈtʃ͡î**w**.ma] /kuˈtʃ͡î**w**i-ma/ have.children-fut.sg 'S/he will have children' 'Va a tener hijos' < BFL 06 6:74/el >
	- c. [wi.noˈmî.**w**i.pi] → [wi.no.ˈmî**w**.pi] /winoˈmî-**w**i-pi/ money-have-irr.pl 'Maybe they will have money' 'A lo mejor van a tener dinero' < BFL 06 6:74/el >

There is a gradient semi-vowel weakening process: after posttonic vowel deletion, labio-velar semi-vowels can range from a fully diphthongal velar rhyme to a completely monphthongized variant. This gradient process is schematized in (32).

4.4 Semi-vowels

(32) Semi-vowel weakening and monophthongization

a. <sup>σ</sup>[w → Vw]<sup>σ</sup> b. Vw]<sup>σ</sup> → V <sup>w</sup>]<sup>σ</sup> c. V <sup>w</sup>]<sup>σ</sup> → V1V<sup>1</sup>

What this scheme illustrates is a process whereby a labio-velar semi-vowel in onset position is re-syllabified as a coda after post-tonic deletion (a); this coda labio-velar semi-vowel may optionally be weakened to a short off-glide of the nucleus vowel (b); the short off-glide may be further weakened by undergoing monophthongization with the nucleus vowel host, yielding a long vowel sequence (c).

This weakening process is exemplified in (33). The segments in question are in bold face. The labio-velar semi-vowel in (33b) is an underlying voiced, bilabial stop (realized as a voiced bilabial approximant in (33a)). (33d) and (33f) have underlying labio-velar semi-vowels.

(33) Gradient weakening and monophthongization of /w/

```
a. [ruruˈtʃ͡íβ̞ia]
     /ruruˈtʃ͡íwi-a/
     sprinkle-prs
     'It's sprinkling.'
     'Está rociando.' < SFH 07 el170/el >
b. [ruruˈtʃ͡íwma]
     /ruruˈtʃ͡íwi-ma/
     sprinkle-fut.sg
     'S/he will sprinkle.'
     'Va a rociar.' < SFH 07 el170/el >
c. [ruruˈtʃ͡íima]
     /ruruˈtʃ͡íwi-ma/
     sprinkle-fut.sg
     'S/he will sprinkle.'
     'Va a rociar.' < SFH 07 el170/el >
d. [tiˈlúwi]
     /tiˈlúwi/
     'Gargle!'
     '¡Haz gárgaras!' < BFL 07 1:43/el >
```


This process is not subject to speaker variation, and the choice between the labio-velar offglide and monophthongization seems to be correlated with rate of speech and care of pronunciation.

# **5 Stress**

The word prosodic system of Choguita Rarámuri is complex, involving both stress and tone. While the distribution of tone is dependent on the location of stress, these are phonologically independent systems that are encoded through independent acoustic means (Caballero & Carroll 2015). These systems, referred to as "hybrid" in the literature, have until relatively recently been understudied (cf. Remijsen 2001, Remijsen 2002, Remijsen & Van Heuven 2005) and are well attested in other Southern Uto-Aztecan languages in addition to Choguita Rarámuri (this aspect of the Choguita Rarámuri word-prosodic system is discussed in more detail in Chapter 6 and Chapter 11). This chapter addresses the main phonological properties of the Choguita Rarámuri stress system, including the stress properties of morphemes and the morphological factors that govern its distribution. The system exhibits a complex interplay between phonological and morphological factors, which are addressed in Chapter 8, Chapter 9 and Chapter 11, in addition to this chapter.

This chapter is structured as follows. §5.1 provides a basic description of the acoustic correlates and distributional properties of stress. §5.2 addresses stressdependent phonological phenomena of vowel reduction and deletion. §5.3 provides a detailed description of the stress properties of roots and suffixes. Finally, §5.4 addresses an overarching phonological restriction on stress distribution to an initial three-syllable stress window and how this restriction, together with morphological stress rules, yields stress alternations in the language.

### **5.1 Acoustic correlates and distributional properties**

Stress is defined cross-linguistically as increased prominence associated with one or more syllables in a word (Gordon & van der Hulst 2020). Choguita Rarámuri exhibits phonetic and phonological properties of languages with word level stress (referred to as "stress-accent" languages in Hyman 2006). These properties include *culminativity* (each lexical word has *at most* one syllable which carries the highest degree of prominence) and *obligatoriness* (each lexical word has *at least* one syllable that carries the highest degree of prominence) (Hyman 1977, Hyman

### 5 Stress

1978, Beckman 1986, Hayes 1995, Hyman 2006). While these two criteria are also used to define languages characterized as "pitch-accent" in the literature, Choguita Rarámuri displays further phonetic and phonological properties that are unique to stress systems, including increased phonetic duration of segments in stressed syllables and reduction of unstressed vowels (fewer vocalic contrasts are realized in unstressed syllables) (the details of which will be discussed in §5.2). For general discussion about the distinction between "stress(-accent)" systems and "pitch-accent" systems, see Poser (1984), Hyman & Wilson (1991), Hyman (1977), Hyman (2001), and Inkelas & Zec (1988). In this description, I follow a property-driven approach to prosodic typology (as advocated in Hyman 2006 and Hyman 2009), which identifies canonical properties of languages and tone languages, but dispenses with the notion of "pitch-accent" as a prosodic type.

In terms of its distribution, stress is underlyingly present for at least some morphemes in this language, and can create lexical contrasts. The examples in (1) show stress minimal pairs.

(1) Choguita Rarámuri stress minimal pairs


As discussed in Chapter 9 and Chapter 11, stress in Choguita Rarámuri may also be the sole exponent of a morphological category (e.g., stress shifts may encode imperative mood, denominalization and applicative derivation). As shown in this chapter, the stress system of Choguita Rarámuri is in part phonologically determined and partly morphologically-conditioned, as attested in other lexical stress systems.

<sup>1</sup>This is a loanword from Northern Mexican Spanish *cochi* 'pig' (which in turn derives from standard Mexican Spanish *cochino* 'pig').

### 5.1 Acoustic correlates and distributional properties

There is no evidence for secondary stress in Choguita Rarámuri. That is, there is no impressionistic evidence of rhythmic prominence, though, as discussed in §11.5, there may be phonological evidence of phenomena sensitive to metrical structure in the language.<sup>2</sup> Stress is left-aligned and assigned in the first (2a), second (2b–c) or third (2d–e) syllable of the word.

(2) Left alignment of stress


Left alignment of stress in stress systems is a pattern that is relatively uncommon cross-linguistically, but attested in all Uto-Aztecan languages with the

<sup>2</sup>A metrical analysis is provided in Chapter 11.

### 5 Stress

exception of Nahuatl language varieties (Munro 1977).<sup>3</sup> Left aligned stress is also an areal trait of Native North American languages (Rice 2010, Caballero & Gordon 2020). Based on the recurrence of both second-mora and second-syllable accentual systems, Munro (1977) reconstructs a second mora stress system for Proto-Uto-Aztecan.

Given that Choguita Rarámuri has both stress and tone, a relevant question to ask is how these systems may be encoded acoustically (Remijsen 2002, Remijsen & Van Heuven 2005). Stress may be encoded through a variety of languagespecific acoustic cues (Gordon & van der Hulst 2020), including increased duration, higher fundamental frequency (f0), greater intensity, and changes in spectral tilt (Roettger & Gordon 2017). Languages with both stress and tone tend to preclude deployment of f0 as an exponent of stress, restricting use of fundamental frequency to encoding of tonal contrasts (Remijsen & Van Heuven 2005, Guion et al. 2010). Analysis of acoustic measures in Choguita Rarámuri reveals that stress has the following phonetic exponents in this language:

	- a. Increased phonetic duration of stressed vowels.
	- b. Augmentation of onsets in stressed syllables.
	- c. Greater intensity of stressed vowels.

The acoustic correlates of stress and tone are examined in Caballero & Carroll (2015) based on the analysis of data obtained through structured elicitation with seven native speakers. The rest of this section summarizes the main results of that study regarding the acoustic encoding of stress in this language (interested readers are referred to this reference for details of data, measures and quantitative results). For the analysis of stress, acoustic correlates were evaluated based on a set of controlled data selected to contrast pretonic (unstressed) vowels with stressed vowels of identical quality in matching segmental contexts. These forms comprise a tonally heterogeneous set and are exemplified in (4).

	- a. *ˈtô-ma* bury-fut.sg *to-ˈmêa* take-fut.sg

<sup>3</sup>Nahuatl varieties have developed penultimate fixed stress as a Mesoamerican areal feature (Munro 1977).

### 5.1 Acoustic correlates and distributional properties


Segment duration of both vowels and consonants is longer in stressed syllables than in unstressed ones: the central duration value for unstressed vowels is 57 ms, with stressed vowels being about 55% longer (at around 86 ms). The durations of onset consonants are similarly longer in stressed syllables than in unstressed syllables (the central duration value for unstressed onsets is 106 ms, and the stressed onsets are 28% longer, around 136 ms). As discussed in Caballero & Carroll (2015), increased segmental duration associated to the encoding of stress in Choguita Rarámuri is related to patterns of vowel reduction and deletion in this language (and described in §5.2 below).

In terms of intensity, stressed vowels have higher average intensity than unstressed vowels: the central intensity value of the stressed vowels is 1.9 dB greater than that of the unstressed vowels.<sup>4</sup>

Higher f0 is one phonetic cue to stress in many languages, but it is not directly associated with stress in Choguita Rarámuri, instead realizing tonal contrats in this language. Thus, and as observed for other languages with both stress and tone, stress in Choguita Rarámuri involves dedicated phonetic exponents for each of these systems, contributing to maximizing the contrast between the two.

Finally, while there was no significant association of f0 with stressed positions, there was also a high degree of variation among speakers in terms of the effect of stress on f0: while some speakers did exhibit a trend toward higher f0 in

<sup>4</sup>The parameters considered to assess the role of intensity were two measures of spectral tilt, low-band H1-H2 and mid-band H1-A2. For both of these measures, a lower value indicates greater vocal effort (indicated through a less shallow decline in spectral intensity), an effect expected for stressed vowels. This expectation was borne out for the Choguita Rarámuri data, though the results were not statistically significant for the mid-band H1-A2 measure, which also exhibited a significant amount of inter-speaker variation. Specifically, the results suggest differences between female and male speakers, but comprehensive investigation of the role of sociolinguistic factors in the acoustic realization of prosody would be necessary in order to determine whether there is a significant pattern in the Choguita Rarámuri data.

### 5 Stress

stressed syllables, no correlation could be detected for other speakers. Speakers also showed differences on how consistently intensity increased as an effect of stress. Speaker-variation in the realization of prosodic contrasts is a topic that needs to be investigated in more depth in this language variety.

### **5.2 Stress-based vowel reduction and deletion**

An important diagnostic of stress in Choguita Rarámuri is patterns of vocalic reduction and deletion. As we have seen in §3.2.2, Choguita Rarámuri contrasts five cardinal vowel qualities in stressed syllables ([a, e, i, o, u]). In unstressed syllables, however, these vowel quality contrasts are often collapsed, reflecting a reduction of the phonetic space. Specifically, vowel height contrasts are neutralized, where mid-front and low-central vowels raise to [i]. Neutralization of vowel height contrasts in unstressed syllables as attested in Choguita Rarámuri is the most common cross-linguistic pattern of unstressed vowel reduction (see Barnes 2002; 2004).

There are three distinct patterns or degrees of vowel reduction. In the first pattern, /e/ optionally raises to [i] both pre-tonically and posttonically. In the second pattern, non-final posttonic /a/ and /o/ optionally raise to [i]. In the third pattern, high vowels optionally reduce to schwa posttonically. These patterns are schematized in (5).

	- /e/ → [i] in pre-tonic and posttonic syllables
	- /a/, /o/ → [i] in non-final, posttonic syllables
	- /i/, /u/ → [ə] in non-final, posttonic syllables

Table 5.1 lays out the surface realization of underlying vowels in pre-tonic, posttonic non-final and posttonic final position. Since all vowel qualities are licensed in stressed position, this chart only considers unstressed vowels.

These patterns of unstressed vowel reduction are addressed below in §5.2.1.1, §5.2.1.3, and §5.2.2.

### **5.2.1 Stress-conditioned vowel reduction patterns**

### **5.2.1.1 Unstressed mid front vowel reduction to [i]**

The first pattern of unstressed vowel reduction in Choguita Rarámuri involves mid front vowels reducing both pre-tonically and posttonically. Pre-tonic vowel

### 5.2 Stress-based vowel reduction and deletion


Table 5.1: Surface realization of unstressed vowels

reduction of mid front vowels is robust. Forms with surface pre-tonic *e* are attested, but these are infrequent. Some examples are presented in (6), where alternative forms with pretonic *i* and pretonic *e* are given. Relevant vowels are in bold face.

(6) Optional pretonic mid, front vowel height neutralization

```
a. [niˈhê] ~ [neˈhê]
     /neˈhê/
    'I'
    'yo' {JLG el1278:3:20.0},{FMF el862:1:42.6}
b. [riˈpòpa] ~ [reˈpòpa]
     /reˈpòpa/
    'back'
    'espalda' < SFH 07 2:65/el >
c. [biˈnè] ~ [beˈnè]
     /beˈnè/
    'learn'
    'aprender' {BFL tx1:0:45.1}, {MDH co1136:9:04.7}
d. [tʃiˈwá] ~ [tʃeˈwá]
     /tʃeˈwá/
    'hit'
    'pegar' < ROF 1:67/el >
e. [miˈhí] ~ [meˈhí]
     /meˈhí/
    'cook mezcal'
    'cocer mezcal' < SFH 07 2:12/el >
```
5 Stress


Other vowel contrasts are preserved pretonically. The examples below show the contrasts licensed: central, front vowels (7a–c); back, high vowels (7d–e); and back, mid vowels (7f–g). The unattested forms listed in (7) show hypothetical forms with pre-tonic, non-initial neutralized vowels. These hypothetical forms would be expected if pretonic vowel reduction would target all vowel qualities.

(7) Pre-tonic vowel contrasts


### 5.2 Stress-based vowel reduction and deletion


As discussed in Chapter 9, there are particular pretonic vowel alternations that are specific of a group of verbal stems. These stems have a root final stressed *a* and a final unstressed, pretonic *i* in specific morphological constructions (e.g. *raʔˈlà-li*, 'buy-pst' vs. *raʔli-ˈmêa*, 'buy-fut.sg'). These vocalic alternations are characteristic of a group of stems where the alternations are morphologicallyconditioned (§5.4), and are not related to the vowel reduction patterns described in this section.

There is inter-speaker variation in terms of reduction patterns, including vocalic reduction patterns. One apparent trend is that older speakers seem to use less mid-vowel neutralization than younger speakers. This is exemplified in the following examples which are part of a recorded conversation. In this interaction, SFH, a 35-year old speaker, uses the verbal stem /bete-/ with pretonic vowel raising (in (8a)); FLP, an 80-year old speaker, on the other hand, responds using the same verbal form with no pretonic mid vowel neutralization (in (8b)). The contrast between the younger speaker's neutralization and lack of neutralization in the older speaker's speech remains constant along the conversation.

(8) Inter-speaker variation in vowel reduction

a. [SFH]: *siˈné roˈkò bitiˈbása? siˈné* one *roˈkò* night *bete-ˈbá-sa* stay.overnight-inch-cond 'He would stay up all night?' 'Quedándose una noche?' {SFH in243:1:29.5}

### 5 Stress

b. [FLP]: *siˈné roˈkò beteˈbása ra ba siˈné* one *roˈkò* night *bete-ˈbá-sa* stay.overnight-inch-cond *ra* rep *ba* cl 'One whole night he would stay up' 'Toda la noche hasta que cumpliera (se quedó sin dormir)' {FLP in243:1:29.5}

Mid front vowels may also be raised to *i* post-tonically. Some examples are shown in (9).

(9) Post-tonic front vowel height neutralization


'Escribía'. < AHF 05 1:127/el >

Pre-tonic and posttonic raising of unstressed *e* is subject to some degree of intra-speaker variation, but this reduction pattern is the one displaying the least amount of variation.

### **5.2.1.2 [**−**high] Unstressed posttonic vowel reduction**

A second unstressed vowel reduction pattern in Choguita Rarámuri involves /a/ and /o/ rising to *i* post-tonically. This reduction process does not take place in

### 5.2 Stress-based vowel reduction and deletion

word-final position. Some examples are provided in (10). Each vowel reduction example is followed by a related form with no vowel reduction.

(10) Post-tonic reduction of low vowels


<sup>5</sup>As described in more detail in §5.3, there are morphologically giverneed stress shifts in complex words.

5 Stress


In (10a–b), (10c–d)<sup>6</sup> and (10e–f), the reduced forms can be contrasted with their stressed, non-reduced counterparts; in (10h) and (10j), on the other hand, reduction does not take place because the vowels in question are in word-final position.

Unstressed vowel reduction is attested in incorporated forms as well. In incorporated forms stress is assigned in the first syllable of the morphological head of the incorporated verb (this morphological stress rule is discussed below in §9.3.3). In the examples in (11), the stress shifts one syllable to the left, yielding reduction of underlying /o/ (11a) and /a/ (11c).

(11) Post-tonic vowel reduction in incorporated verbs

a. [ronoˈbâk**i**ma] /roˈnô+paˈk**ó**-ma/ feet+wash-fut.sg 'S/he will wash her feet' 'Se va a lavar los pies'

<sup>6</sup>The forms in (10a,b) involve a stress shift that is morphologically-conditioned, but the posttonic reduction is subject to speaker variation; this contrasts with the morphologicallyconditioned vocalic alternations to be discussed in Chapter 9 (§9.1), since these do not display speaker variation.

### 5.2 Stress-based vowel reduction and deletion


Not all pre-final, post-tonic underlying /a/ raise to [i]. The potential targets for reduction in (12) share the characteristic of being the first vowel of a vowelinitial suffix. These suffix vowels do not undergo reduction, and hypothetical forms with reduced post-tonic vowels (exemplified in the second column in (12)) are unattested.

(12) Blocked vowel reduction

a. [ˈʃûami] /ˈsû-**a**me/ sew-ptcp 'the one who sews' 'el que cose' < BFL 06 4:168/el > *\*ˈʃû-imi*

b. [saˈwèrami]

/saˈw-è-r-**a**me/ cure-appl-pst.pass-ptcp 'the one who cures' 'el que cura' < BFL 06 4:168/el > *\*saˈw-è-r-imi*

5 Stress

```
c. [baˈjèatʃ͡i]
    /baˈj-è-a-tʃ͡i/
     call.out-appl-prog-loc
    'at the place where they invite, call out'
    'donde llaman, invitan' < BFL 05 2:56/el >
    *baˈ-j-è-i-tʃi
d. [laˈmúami]
    /laˈmú-ame/
     purple-ptcp
    'purple'
    'morado' < LEL 06 6:79/el >
    *laˈmú-imi
e. [ˈpòara]
    /ˈpò-a-ra/
     cover-prog-purp
    'lid'
    'tapadera' < SFH 07 in242/in >
```
*\*ˈpò-i-ra*

Reduction, thus, can be blocked due to morphological restrictions.

### **5.2.1.3 Unstressed high vowel reduction to schwa**

Choguita Rarámuri has a third process of unstressed vowel reduction, where high vowels may reduce to schwa posttonically. The examples below show the target vowels, /i/ (13a–d) and /u/ (13e–f).<sup>7</sup>

### (13) Optional post-tonic reduction to schwa

```
a. [tʃ͡iˈpórəma]
    /tʃ͡iˈpó-ri-ma/
    bounce-caus-fut.sg
    'S/he will make it bounce.'
    'Lo va a hacer rebotar.' < LEL 07Caus_ME >
```
<sup>7</sup>There is only one example where the target of reduction is an underlying low, central vowel: iˈna-n**ə**r-o (/iˈna-n**a**le-o/),'walk-order-ep' < BFL 06 DECW/ el >.

### 5.2 Stress-based vowel reduction and deletion

	- 'Les van a obligar a llevarlo.' < LEL 07 Ind\_Caus/el >

High vowel reduction to schwa is gradient, and favored when preceding or following a back, round vowel or *a*. I have not documented any cases of unstressed high vowel reduction to schwa in pre-tonic position with open class lexical items or in word-final position.<sup>8</sup>

### **5.2.2 Stressed-conditioned vowel deletion**

Unstressed vowels may also undergo syncope in immediately post-tonic syllables. In (14), the deleted vowel is in bold face in the underlying representation.

<sup>8</sup>But it is common to find function words with pre-tonic vowels reducing to schwa in fast speech, e.g. the adverb *biˈlá* 'indeed', is often realized as [b**ə**ˈlá].

### 5 Stress

(14) Post-tonic vowel deletion


As the examples above show, post-tonic vowel deletion targets underlying high (14a), low (14b–c), and mid (14d–e) vowels.

Deletion does not target word-final unstressed vowels. This is shown in (15).

(15) No post-tonic vowel deletion of word-final vowels

a. [naˈkûri] /na-ˈkûri/ pl-help 'They help.' 'Ayudan.' < SFH 05 1:102/el >

### 5.2 Stress-based vowel reduction and deletion

b. [toˈnáli] /tò-nále/ take-desid 'They want to take it.' 'Quieren llevarlo.' < BFL 06 5:148-150/el >

Deletion does not target pretonic vowels as a general phonological process, but instances of pretonic deletion can be found in morphologically specific contexts. For instance, the forms in (16) involve pretonic deletion with root-inchoative sequences. Examples (16b, d, f) show a corresponding form with the same root with no deletion. Other pretonic effects which are stress-conditioned (such as syllable truncation in morphologically specific contexts) are discussed in Chapter 9 and Chapter 11.

(16) Morphologically-conditioned pretonic deletion


5 Stress

> 'if s/he is (not) able to' 'si (no) puede' {JLG co1236:1:49.6} f. [oˈm**è**ri] /omèra/ be.able.prs 'S/he is (not) able.' '(No) puede.' {LEL tx130:5:08.4}

In sum, while pre-tonic vowel deletion is uncommon, posttonic vowels generally reduce or delete. In the latter case, the choice between reduction and deletion results from a gradient process that is affected by rate of speech. This process may also exhibit intra-speaker variation, though this has not yet been systematically examined in the Choguita Rarámuri data.

### **5.3 Stress properties of roots and suffixes**

As described above, stress is part of the underlying representation of at least some morphemes in Choguita Rarámuri. Diachronic loss of contrastive vowel length, reconstructed for Proto-Uto-Aztecan (Campbell & Langacker 1978), is argued to be the likely source of lexically contrastive stress in contemporary Uto-Aztecan varieties (Munro 1977).<sup>9</sup> Choguita Rarámuri stress might have become unpredictable when the Proto-Uto-Aztecan vowel length distinction was lost.

As proposed for other Uto-Aztecan languages (e.g. Cupeño (Cupan; Hill & Hill 1968, Alderete 2001)), Choguita Rarámuri roots fall into two classes: *stressed* and *unstressed*. Stressed roots have fixed stress across morphologically related forms.<sup>10</sup> Unstressed roots, on the other hand, lack underlying lexical specification for stress; stress in these cases is not fixed and its position determined by the phonological and morphological context. Specifically, unstressed roots may receive stress by default when attaching a stress-neutral suffix or shift stress rightward when attaching a stress-shifting suffix (the stress properties of suffixes are described below in §5.3.4). All prosodic words in Choguita Rarámuri, whether they contain a lexically stressed root or an unstressed lexical root, have surface stress, a syntagmatic prominence cued acoustically primarily via intensity and duration (as described in §5.1 above).

<sup>9</sup> Some Uto-Aztecan prosodic systems exhibit weight-based stress systems (e.g., Numic languages, like Southern Paiute (Sapir 1930) and Tümpisa Shoshone (Dayley 1989) are documented to have rhythmic stress systems sensitive to mora count).

<sup>10</sup>Stressed roots are analyzed in Caballero (2008) and Caballero (2011b) as lexically specified with a diacritic mark which is phonetically realized as stress in output forms.

### 5.3 Stress properties of roots and suffixes

The contrast between stressed and unstressed roots is exemplified below with the different stress patterns of nominal roots that add the (stress-shifting) locative suffix *-tʃ͡í*, as exemplified in Table 5.2.


Table 5.2: Stressed and unstressed roots in Choguita Rarámuri

Stressed roots have fixed stress in the first (a–c) or second (d) syllable, whether bare or suffixed with locative *-tʃ͡í*. In unstressed roots, stress falls on the locative suffix, one syllable to the right with respect to their bare counterparts (e–i).

Lexical stress in roots blocks morphologically-conditioned stress shifts; stress in these cases will be fixed on the first, second or third syllable, regardless of what suffixes are attached. On the other hand, words containing unstressed roots have predictable stress assignment. In words containing unstressed roots and stressneutral suffix, stress will be assigned by default on the second syllable of the root or the only syllable of monosyllabic roots. In words containing unstressed roots and stress-shifting suffixes, stress shifts rightward. The default stress assignment algorithm is summarized in (17):

(17) **Default stress assignment in Choguita Rarámuri**: words containing unstressed disyllabic or trisyllabic roots and neutral morphological constructions (affixation and non-concatenative processes) have second syllable stress (Caballero 2011b).

### 5 Stress

More details about the morphological dimension of stress in nominal and verbal paradigms is provided in Chapter 8 and Chapter 9, respectively. The stress properties of monosyllabic, disyllabic and trisyllabic roots in Choguita Rarámuri are addressed next.<sup>11</sup>

### **5.3.1 Stress properties of monosyllabic roots**

Most roots in Choguita Rarámuri are disyllabic or trisyllabic, but monosyllabic roots are attested. Monosyllabic roots are almost all stressed, but there are attested cases of unstressed monosyllabic roots. This contrast is exemplified in Table 5.3. 12,13


Table 5.3: Stressed and unstressed monosyllabic roots

Stressed monosyllabic roots have fixed stress (Table 5.3a–b), regardless of the type of suffix it may combine with (stress neutral like the past *-li* suffix or stressshifting like the future *-ma/ˈmêa* suffix).<sup>14</sup> Unstressed monosyllabic roots, on the other hand, may shift stress rightward onto the suffix if the suffix is stressshifting (Table 5.3c–d). Unstressed monosyllabic roots have stress in the root (the first syllable of the prosodic word) when attaching stress-neutral suffixes, exemplifying that default stress is assigned in the root (stress neutral suffixes are never stressed).

<sup>11</sup>As discussed in §9.3.3 and §11.3.6, roots longer than three syllables are infrequent. The stress properties of noun incorporation is addressed in §9.3.3

<sup>12</sup>As discussed in §11.2, monosyllabic roots undergo lengthening when bare (e.g., present tense) in order to satisfy a prosodic word bimoraic minimality requirement.

<sup>13</sup>The tone properties of morphologically complex words is addressed in Chapter 8, Chapter 9 and Chapter 11.

<sup>14</sup>The future singular suffix displays an interesting allomorphy: *-ma*, used with stressed roots, and *-ˈmêa*, used with unstressed roots. Consistently, the former is unstressed while the latter is stressed, and root stress seems to be the only parameter that plays a role in allomorph selection. The future singular suffix is the only suffix that displays this stress-conditioned suppletive allomorphy. Guarijio, another Taracahitan language, does not have this allomorphy for the cognate future suffix (in both stressed and unstressed contexts, the future singular suffix is *-ma* Miller 1996).

### 5.3 Stress properties of roots and suffixes

### **5.3.2 Stress properties of disyllabic roots**

Disyllabic roots may be underlyingly stressed or unstressed. Disyllabic stressed roots are exemplified in Table 5.4.


Table 5.4: Stressed disyllabic roots

Most disyllabic roots with first syllable stress are lexically stressed. The great majority of these roots have first syllable stress that remains constant across morphological paradigms, as exemplified in Table 5.4a–e. As shown in these examples, whether these verbal roots combine with neutral constructions (like the past *-li* suffix) or shifting constructions (like the future singular *-ma ~ -mêa* suffix), stress remains fixed on the first syllable.

Unstressed disyllabic roots are exemplified in Table 5.5.

Table 5.5: Unstressed disyllabic roots


### 5 Stress

These roots, with third syllable stress in shifting morphological contexts, increase the proportion of words with third syllable stress in the Choguita Rarámuri corpus.

A set of exceptional disyllabic roots have *first* syllable stress with stress-neutral suffixes, and *second* syllable stress with stress-shifting suffixes. From the Choguita Rarámuri corpus, only eight roots exhibit this behavior. An exhaustive list is given in Table 5.6.


Table 5.6: Disyllabic roots with first and second syllable stress

Comparison of these roots with their cognates in closely related Guarijío (Tara-Guarijío; Miller 1996), reveals that this set of roots is truly exceptional. Specifically, the Guarijío cognates all have three syllables, suggesting that Choguita Rarámuri has innovated initial syllable truncation with these forms. The cognate forms are shown in Table 5.7.

It is not the case that all of the Choguita Rarámuri roots are one syllable shorter than the corresponding Guarijío cognates (e.g. Choguita Rarámuri *raʔìtʃ͡a* – Guarijío *taʔˈitʃ͡a* (M391), 'speak'; Choguita Rarámuri *roʔˈsówa* – Guarijío *tohˈsoá* (M396), 'cough'). The forms in Table 5.7 are derived from trisyllabic roots through a recent diachronic development. This comparative evidence suggests these exceptional forms do not constitute a counterexample to the stress patterns found with unstressed roots elsewhere in the language (where stress shifts are attested), but instead follow the pattern of unstressed trisyllabic roots. I turn to these next.

### **5.3.3 Stress properties of trisyllabic roots**

Most trisyllabic roots are stressed, with fixed second syllable stress or fixed third syllable stress, aas shown in Table 5.8 and Table 5.9, respectively.


Table 5.7: Cognate forms: Choguita Rarámuri and Guarijío roots

Table 5.8: Second-syllable, stressed trisyllabic roots


Table 5.9: Third-syllable, stressed trisyllabic roots


### 5 Stress

There are also unstressed trisyllabic roots. As shown in (18), these roots have shifting stress in morphologically complex constructions.

### (18) Unstressed trisyllabic roots


### 5.3 Stress properties of roots and suffixes

'if s/he endures it' 'si aguanta' {MFH el1318:20:00.0}

h. *anaˈtʃ͡ânale anaˈtʃ͡â-nale* endure-desid 'S/he wants to endure it.' 'Quiere aguantar.' < ROF 04 1:123/el >

The stress shifts that the verbal root *aˈnátʃ͡a,* 'endure' (*'aguntar'*), undergoes in (18) parallels those of unstressed disyllabic roots in different morphological constructions: the verb has second syllable stress when inflected with stress-neutral suffixes in (18a–e), but third syllable stress when inflected with stress-shifting suffixes (18f–i). Further examples of unstressed trisyllabic roots are given in Table 5.10.


Table 5.10: Unstressed trisyllabic roots

### **5.3.4 Stress properties of suffixes**

The stress patterns of unstressed trisyllabic roots in morphologically complex words (shown in (18) and Table 5.10 above) show that stress-neutral suffixes are not pre-stressing, as could have been assumed from the stress pattern of unstressed disyllabic roots. If stress-neutral suffixes were pre-stressing, we would expect third-syllable stress with trisyllabic unstressed roots, immediately preceding the suffixes, instead of the attested second syllable stress. These hypothetical, unattested forms are illustrated in the second column in (19).

### 5 Stress

	- a. *naʔˈsòwili naʔˈsòwa-li* mix-pst 'S/he mixed it.' 'Mezcló.' {JLG co1239:6:12.5} *\*naʔsoˈwâ-li* b. *naʔˈsòwaki naʔˈsòwa-ki* mix-pst.ego 'I mixed it.' 'Mezclé' < ROF 04 1:123/el > *\*naʔsoˈwâ-ki* c. *naʔˈsòwaa naʔˈsòwa-a* mix-prog 'S/he is mixing.' 'Está mezclando.' {JLG co1239:5:58.8} *\*naʔsoˈwâ-a* d. *raˈʔàmali raˈʔàma-li* give.advice-pst 'S/he gave advice.' 'Aconsejó.' < ROF 04 1:64/el > *\*raʔaˈmâ-li* e. *raˈʔàmaa raˈʔàma-a* give.advice-prog 'S/he is giving advice' 'Está aconsejando' < ROF 04 1:64/el > *\*raʔaˈmâ-a*
		- f. *raˈʔàmaki raˈʔàma-ki*

5.3 Stress properties of roots and suffixes

```
give.advice-pst.ego
'I gave advice.'
'Aconsejé.' < ROF 04 1:64/el >
*raʔaˈmâ-ki
```
g. *raˈʔìtʃ͡ili*

*raˈʔìtʃ͡a-li* speak-pst 'S/he spoke.' 'Habló.' {MFH el1318:14:24.5} *\*raʔiˈtʃ͡a-li*

h. *raˈʔìtʃ͡i raˈʔìtʃ͡a-i* speak-impf 'S/he used to speak.' 'Hablaba' {JLG co1239:2:15.2} *\*raʔiˈtʃ͡a-a* i. *raˈʔìtʃ͡aki raˈʔìtʃ͡a-ki*

speak-pst.ego 'I spoke.' 'Hablé' < SFH 05 1:98/el > *\*raʔiˈtʃ͡a-ki*

Finally, stress-neutral suffixes are never stressed, an observation also made in §5.3.1 above. Following the pattern of unstressed trisyllabic roots (in (18) and (19)), we would expect unstressed disyllabic roots adding a stress-neutral suffix (like causative *-ti*) and a stress-shifting suffix (like future singular *-ma*) to have third syllable stress. These verbs, however, have second-syllable stress. This is illustrated in (20).

(20) The unstressability of stress-neutral suffixes

a. *aˈwítisa aˈwí-ti-sa* dance-caus-cond 5 Stress

> 'if s/he makes them dance' 'si lo hace bailar' < SFH 08 1:112/el > *\*awi-ˈtî-sa*

b. *raʔˈlisima*

*raʔˈlì-si-ma* buy-mot-fut.sg 'S/he will go around buying.' 'Va a ir comprando.' < AHF 05 1:130/el > *\*raʔli-ˈsi-ma*

c. *oˈsìsima*

*oˈsì-si-ma*

read.write-mot-fut.sg

'S/he will go around reading/writing.'

'Va a ir leyendo/escribiendo.' < SFH 05 1:78/el >

*\*osi-ˈsì-ma*

d. *tʃ͡oˈníkima*

*tʃ͡oˈní-ki-ma* fist.fight-appl-fut.sg 'They will fist-fight for someone.' 'Van a pelear a chingazos por alguien.' < SFH 05 1:67/el > *\*tʃ͡oni-ˈkî-ma*

The unstressability of stress-neutral suffixes is further evidenced by unstressed monosyllabic roots. Only two of twenty-seven monosyllabic verbal roots are unstressed (*ru* 'say' ('*decir*') and *tó* 'bring' ('*traer*') (exemplified above in §5.3.1 ). These roots shift stress to stress-shifting suffixes, as shown in Table 5.11.

We might have expected the forms in Table 5.11a–d to have second syllable stress when adding stress-neutral suffixes, following the pattern of disyllabic and trisyllabic unstressed roots. Instead, stress in these words is in the root, the first syllable. These cases suggest that the stress rule associated with stress-neutral suffixes must also meet the condition of being assigned *within* the root. Stressneutral suffixes are not part of the stress domain and are thus non-cohering. Cohering suffixes are suffixes that form one prosodic word with the preceding stem (evidenced by their phonological behavior as identical to morphologically simple words), and non-cohering suffixes form prosodic words of their own (Booij 2002).

### 5.4 Initial three-syllable stress window


Table 5.11: Unstressed monosyllabic roots

(For general discussion about cohering and non-cohering affixes, see Dixon 1977; Booij 1977; 1999; 2002).

The morphological conditions for stress and the stress patterns of morphologically complex words are addressed in Chapter 9 and Chapter 11.

### **5.4 Initial three-syllable stress window**

As described for closely related Taracahitan languages (Norogachi Rarámuri (ISO code: tar) (Brambila 1953), Mountain Guarijío (ISO code: var) (Miller 1996), Yaqui (Demers et al. 1999), and Mayo (Hagberg 1989)), stress in Choguita Rarámuri is restricted to an initial stress window. Stress window systems are characterized by alternations that keep stress within a two- or three-syllable margin of the edge of the domain, whether stress is unpredictable or not within this margin (Kager 2012, Everett 1988, Green & Kenstowicz 1995).<sup>15</sup> Window effects are evidenced by alternations in compound, reduplicated and multiple-affixation constructions that maintain stress within a disyllabic margin in the Cahitan languages Yaqui and Mayo, and a trisyllabic margin in Guarijío. Initial three syllable stress windows are uncommon cross-linguistically: outside of the Taracahitan branch, only a handful of languages (most of them in the Americas) have been documented to possess an initial three-syllable window, including the Uto-Aztecan (Numic) language Comanche (ISO code: com) (Smalley 1953: 299) (cited in Kager 2012).

Earlier descriptions of other Rarámuri varieties and Guarijío have documented that stress is left-aligned in these languages and never placed beyond the third

<sup>15</sup>In Pirahã (spoken in the Amazon), for example, stress is assigned to the heaviest syllable within the last three syllables of the word (Everett 1988, Green & Kenstowicz 1995).

### 5 Stress

syllable, with alternations in reduplication and compounding maintaining this three-syllable restriction (see Brambila 1953: 245 for Norogachi Rarámuri and Miller 1996: 49–50 for Mountain Guarijío). Choguita Rarámuri has lost productive prefixation processes, but the three-syllable window is evident in other morphological constructions. Closely related Mountain Guarijío, where prefixing reduplication is still productive, does exhibit systematic alternations to keep stress within the window (Miller 1996).

The forms below show that Choguita Rarámuri stress is clearly left-aligned. Since the longest monomorphemic roots in this language are tetrasyllabic at most, the evidence in (21) involves morphologically complex forms.

### (21) Left alignment of stress


5.4 Initial three-syllable stress window

sneeze-ev-desid 'It sounds like they want to sneeze.' 'Suena que quieren estornudar.' < SFH 07 1:73/el >


While the forms in (21a–e, g) could be ambiguous between second and third syllable stress and penultimate and antepenultimate stress, embedding these forms in further morphology reveals that the correct generalization about stress assignment can only be made with respect to the left edge of the prosodic word. Each pair of morphologically related words ((21a–b), (21c–d), (21e–f) and (21g–h)) shows that stress is constantly on the second or third syllable.

Choguita Rarámuri, like Norogachi Rarámuri and closely related Guarijío, also has constructions that display stress alternations that maintain stress within a left-aligned window margin. There are N-V constructions that are restricted to nouns referring to body parts and bodily fluids. In these constructions, the noun root is fully integrated with the verb morphologically, and both the noun root and the verb root can be used independently. As discussed in Chapter 9, these properties are prototypical of "body part incorporation", a restricted kind of noun incorporation, which is common in languages of the Americas (Baker 1996). Stress in these constructions is actively constrained by the grammar. If the head, the incorporated verb, has second syllable stress in isolation and if the first member, the body-part noun, is two syllables long, stress retracts to the verb's first syllable, the construction's third syllable.<sup>16</sup> This is exemplified in the forms provided in (22).

<sup>16</sup>In (22e–g), the glottal stop associated with the verbal roots does not emerge in the surface incorporated form due to the glottal prosody place restriction described in §7.1

### 5 Stress

	- a. /buˈsí+kaˈsì/ → *busi+ˈkâsi* eye+break 'to become blind' 'volverse ciego'
	- b. /roˈpâ+kaˈsì/ → *ropa+ˈkâsi* stomach+break 'to have a miscarriage' 'abortar'
	- c. /buˈsí+boˈtá/ → *busi+ˈbôta* eye+come.out 'for an eye to come out' 'salirse el ojo'
	- d. /kaˈwá+boˈtá/ → *kawa+ˈbôta* egg+come out 'for an egg to come out' 'salirse el huevo'
	- e. /kuˈtâ+biʔˈrì/ → *kuta+ˈbîri* neck+twist 'to neck-twist' 'torcerse el cuello'
	- f. /tʃ͡oʔˈmá+biˈʔwa/ → *tʃ͡oma+ˈbî*wa mucus+clean 'to mucus-clean' 'limpiar los mocos'
	- g. /tʃ͡eˈréwa+biˈʔwa/ → *tʃ͡ere+ˈbîwa* sweat+clean 'to sweat-clean' 'limpiar el sudor'

All possible interactions of underlyingly stressed and unstressed roots are attested in these forms: unstressed noun plus unstressed verb (22a), stressed noun plus unstressed verb (22b), unstressed noun plus stressed verb (22c), and stressed noun plus stressed verb (22d). Regardless of the underlying stress make-up of the roots of the construction, stress is assigned in the first syllable of the head of the construction, the verbal root. The stress retraction phenomenon involves actual deletion of lexical inherent root stress from the head of the construction. The verbal root *biʔˈw-a* 'to clean', for instance, is a stressed root (with fixed stress when adding stress-shifting suffixes) (e.g. (23)):

### 5.4 Initial three-syllable stress window

	- a. *biʔˈwâma biʔˈw-â-ma* clean-fut.sg 'S/he will clean.' 'Va a limpiar.' < SFH 05 1:72/el > b. *biʔˈwâsa biʔˈw-â-sa* clean-cond 'if s/he cleans' 'si limpia' < SFH 05 1:72/el > c. *biʔˈwâbo*
	- *biʔˈw-â-bo* clean-fut.pl 'They will clean.' 'Van a limpiar.' < SFH 05 1:72/el >
	- d. *biʔˈwânale*
	- *biʔˈw-â-nale* clean-desid 'S/he wants to clean.' 'Quiere limpiar.' < SFH 05 1:72/el > e. *biʔˈwâsi biʔˈwâ-si* clean-imp.pl 'You all clean!' '¡Limpien!' < SFH 05 1:72/el >

In incorporation, however, this verbal root undergoes a stress shift one syllable to the left (e.g., *tʃ͡oma+ˈbiwa* in (22f) above).

Stress in incorporated verbs, therefore, involves both stress deletion and stressreassignment. Brambila (1953) and Miller (1996) interpret similar stress deletion and re-assignment facts in the Taracahitan languages as evidence for a threesyllable stress window. Fourth syllable stress, which would result in the incorporated forms in (22a–f) if there were no stress reassignment, would fall outside this window, and is therefore retracted one syllable to the left.

The stress alternations in the incorporated forms of Choguita Rarámuri can alternatively be attributed to a morphological stress rule specific to incorporated construction rather than a stress window. This morphological stress rule would

### 5 Stress

require stress to be assigned in the first syllable of the head of the incorporated construction. This morphological stress rule is defined in (24).

(24) **Incorporation stress rule:** The head of the incorporation construction (the verbal root) must bear stress in the first syllable

There are, however, further testing grounds for the window hypothesis. The behavior of trisyllabic nouns in incorporation is crucial in this regard. Choguita Rarámuri, like other Uto-Aztecan languages (e.g. Southern Paiute (Sapir 1930) and Kawaiisu (Zigmond et al. 1991)), tends to shorten its trisyllabic nouns to a disyllabic form when incorporated. These truncated forms in incorporation are shown in Table 5.12.

Table 5.12: Noun truncation in incorporation


Truncation of tetrasyllabic nouns in the incorporated forms in Table 5.12 is ambiguously triggered by either an initial three-syllable stress window or a morphological incorporation stress rule defined in (24). However, while most speakers completely reject non-truncated versions of the forms in Table 5.12, for some speakers such forms are in fact interpretable even if never produced spontaneously. These non-truncated forms are shown in (25), where angled brackets indicate that these forms are abstract and not spontaneously produced.

	- a. *tʃ͡ameˈkârepu* <tʃ͡ameˈkâ+repu> tongue+cut 'to cut the tongue' 'cortar la lengua'
	- b. *kutaˈtʃ͡írepu* <kutaˈtʃ͡í+repu> neck+cut 'to cut the neck' 'cortar el cuello'

### 5.4 Initial three-syllable stress window

These forms, with stress in the third syllable, are never spontaneously produced, but their intended meanings can be retrieved. Equivalent non-truncated forms with stress in the fourth syllable, on the other hand, were completely rejected and their intended meaning could not be recovered. These forms are shown in (26).

(26) Uninterpretable, non-truncated incorporated verbs


The incorporation stress rule is violated in the interpretable cases in (25), but the initial three-syllable stress window is violated in the completely rejected forms in (26), suggesting that there is indeed an overarching, exceptionless rule that restricts stress to the first three syllables of the word in Choguita Rarámuri. There is indeed not a single form in the Choguita Rarámuri corpus that has stress outside this three-syllable range.

While the initial three-syllable window is mainly manifested synchronically as a static, exceptionless generalization in the native vocabulary of Choguita Rarámuri, loanwords from Spanish contribute further evidence of this phenomenon, exhibiting alternations related to this restriction. Specifically, Choguita Rarámuri loanwords retain original prominence from the Spanish source (27a–b), in addition to a default HL tone (see also Caballero & Carroll 2013 and Chapter 11 below about prosodic loanword adaptation). Given that Spanish has a final three syllable window, there are cases of loanwords where the prominence would emerge beyond the initial three syllable margin in Choguita Rarámuri. As shown in (27c– e), the repair is to truncate a syllable of the Spanish source to match the prosodic requirement of Choguita Rarámuri (in the examples below, stressed syllables are highlighted with boldface and truncated syllables with underlining in the Spanish source forms).

	- a. *maˈsâna* 'apple' Sp. *manzana*
	- b. *mehoˈrâra* 'acetaminophen' Sp. *mejoral*

5 Stress


Truncation of the source words targets the initial syllable in each of the examples in (27).

While languages with a final ternary stress window – permitting only final, penultimate, or antepenultimate stress – are not uncommon (e.g. Imbabura Quechua, Macedonian, Greek, Hebrew, Spanish, Polish, Zoque, Italian, to name just a few), initial three-syllable stress windows are a typologically highly marked pattern (Kager 2012; see also Caballero 2011b; Martínez-Paricio 2013). Initial three syllable stress windows have been described in languages outside the Uto-Aztecan language family, including Icua Tupi (Tupi), Terena (Arawakan), Wishram Chinook (Chinookan) (Kager 2012), and Azkoitia Basque (Hualde 1998).

<sup>17</sup>Loanword nouns from Spanish are often incorporated in the lexicon with a locative suffix (for more details about loanword adaptation, see §7.3).

# **6 Tone and intonation**

This chapter addresses the Choguita Rarámuri lexical tone and intonation systems. All documented Uto-Aztecan languages are described as having stress,<sup>1</sup> with tonogenesis being independently innovated in several branches, predominantly in the US Southwest/Northern Mexico area (Caballero & Gordon 2020).<sup>2</sup> In addition to Choguita Rarámuri, other Uto-Aztecan languages that have been described as having both stress and tone in their word prosody include Yaqui (Cahitan; Demers et al. 1999), Mayo (Cahitan; Hagberg 1989, Hagberg 1993), Northern Tepehuan (Tepiman; Bascom 1959, Woo 1970), Cora (Corachol; McMahon 1967), Huichol (Corachol; Grimes 1959), Hopi (Manaster-Ramer 1986) and Balsas Nahuatl (Aztecan; Guion et al. 2010).<sup>3</sup>

All tonal Uto-Aztecan languages have a restricted distribution of tonal contrasts to stressed syllables. Restricted tone systems as documented in these languages are similar in some regards to other languages described as 'pitch accent' in the literature (Hyman 2006, van der Hulst 2011). In this grammar, I assume Choguita Rarámuri is better characterized as having both stress and tone as independent systems and that pitch is part of the lexical realization of at least some morphemes, as defined for tonal languages cross-linguistically (Yip 2002; Hyman 2001; Hyman 2006: 229. Choguita Rarámuri is the only Uto-Aztecan language documented to date to feature a three-way tone contrast as opposed to a binary one, suggesting its tone system involves a recent innovation.

Choguita Rarámuri deploys f0 in its intonation system, resulting in different accommodation strategies when the lexical phonology, the morphology and the

<sup>1</sup>Proto-Uto-Aztecan is reconstructed as having a weight-sensitive, rhythmic stress with left alignment (Munro 1977); the synchronic stress systems of daughter languages exhibit are widely diverse, though they exhibit a predominance of left alignment (Caballero 2011a; Chapter 5).

<sup>2</sup>The diachronic development of tone has been linked to the loss of laryngeal features in Hopi (Manaster-Ramer 1986), Northern Tepehuan (Shaul 2000), and varieties of Balsas Nahuatl (Guion et al. 2010).

<sup>3</sup>Recent research argues that some varieties previously analyzed as being tonal exhibit only stress: Reyes Taboada (2014) explicitly rejects a tonal analysis of Santa María de Ocotán Tepehuan (Southern Tepehuan; Tepiman). Work on tone and prosodic documentation more generally in the language family has not yet addressed dialect variation across the different branches.

### 6 Tone and intonation

intonational phonology assign conflicting tones to the same target tone bearing units. Intonation has been understudied for Uto-Aztecan languages: prior work on intonation in the language family includes studies on Nahuatl (Aztecan; Guion et al. 2010, Patiño Velázquez 2014, Aguilar 2020) and Northern Tepehuan (Tepiman; Gil Burgoin & Carrillo Carrillo 2019). This chapter provides a basic description of intonation patterns in declarative and interrogative sentences of this language. Declaratives feature optional H% boundary tones and the presence of optional pitch targets ('rhythmic lead tones') preceding each of the lexical tones: a low target before high and falling tones, and a high target before low tones (Garellek et al. 2015, Caballero et al. 2022). Interrogatives are intonationally characterized by H% boundary tones and raised register across the utterance.

While tone also encodes morphosyntactic information, this chapter addresses only the phonologically predictable aspects of tone in this language, saving the description and analysis of its morphological contribution to the chapters devoted to the verbal morphology (Chapter 9) and nominal morphology (Chapter 8). Interactions between lexical, morphological and phrasal tones are addressed in Chapter 11.

### **6.1 Tone**

### **6.1.1 Tonal inventory**

The Choguita Rarámuri tone system involves a three-way lexical contrast between HL, L and H tones.<sup>4</sup> The HL lexical tone is a phonological unit, a contour tone that is not decomposable into two tonal primitives.

Tone is not conditioned by voicing or laryngeal setting of the preceding consonant, nor any other phonological features. Lexical tonal contrasts in this language are associated exclusively with surface stressed syllables, i.e., there is only one lexical tone per prosodic word and stressless syllables lack lexical tone (see §11.1for discussion of criteria that define prosodic words in the language). Stressed syllables may bear any of the three lexical tones. Tone is also obligatory, i.e., there are no toneless words.

The tone-bearing unit is the mora: falling tones have their high target on the stressed syllable, with the fall starting in the tonic and continuing through a post-tonic syllable, if there is one. H tones may spread their high f0 to the posttonic syllable (Caballero & Carroll 2015, Garellek et al. 2015, Caballero et al. 2022). Figure 6.1 shows the f0 trajectories of the three lexical tones in stressed syllables.

<sup>4</sup>H tones are described as M in Caballero & Carroll (2015).

6.1 Tone

Figure 6.1: Mean f0 tracks (in semitones with base 100 Hz for male speakers, with base 200 Hz for female speakers) over utterance-medial tonic vowels, averaged over vowel thirds (Caballero et al. 2022).

As illustrated in Figure 6.1, lexical tonal contrasts are acoustically encoded both through pitch height and slope. There are relatively small differences in terms of f0 between the three lexical tones in tonic syllables: Caballero & Carroll (2015) report an average difference of 2 semitones between L and H tones, while Caballero et al. (2022) report a 3 semitone difference between L and H tones (illustrated here). There are also some speaker differences, with some speakers showing even narrower f0 distinctions between certain tones.<sup>5</sup> These pitch differences are among the smallest compared to pitch differences for lexical tonal contrasts in other languages with similar tone inventories (Alexander 2010, Silva 2006).<sup>6</sup>

Despite the narrow f0 differences, lexical tonal categories are reliably distinguished by f0 alone and consistent across speakers in the realization of pitch of

<sup>5</sup> It should be noted, however, that the data examined for both studies involved differences in terms of the size of the corpus, number of speakers and types of examples recorded.

<sup>6</sup>Other languages that have been reported with similar narrow pitch differences include Edo, a Volta-Niger language spoken in Nigeria with a binary tone system and Hausa (Chadic), a language with a three-way tone contrast with H, M and HL tones (Maddieson 1979). In the Americas, Guion et al. (2010) report narrow pitch differences in the realization of the two-tone system recently developed in Balsas Nahuatl.

### 6 Tone and intonation

stressed syllables. Furthermore, as reported in Garellek et al. (2015), Caballero et al. (2022) and discussed below, lexical tonal contrasts are further differentiated by f0 changes from the pre-tonic to the tonic and post-tonic syllables (§6.2.2) and by additional acoustic parameters in different phrase positions (§6.2.4).

### **6.1.2 Tonal (near-)minimal pairs**

The (near) minimal pairs in (1) show that tone is contrastive in Choguita Rarámuri in both nouns and verbs,<sup>7</sup> in monosyllabic roots (1a–f), disyllabic roots with second-syllable stress (1g–l), and disyllabic roots with first syllable stress (1m–n) (a circumflex accent indicates a falling HL tone <ô>, a grave accent indicates a L tone <ò>, and an acute accent indicates a H tone <ó>). Only lexical tones are represented in the phonemic transcription.<sup>8</sup>



The examples in (1) show examples of tone in monosyllabic and disyllabic roots with different stress properties. Tonal contrasts are also realized on the stressed syllable of trisyllabic roots, as shown in (2).

<sup>7</sup>The audio examples illustrating verb forms in the set in (1) may be inflected and/or provided within a phrase.

<sup>8</sup>As discussed in §6.2 and Chapter 11, there is evidence of H% boundary tones and post-lexical tonal targets associated with lexical tones in declarative intonation, as well as non-modal phonation and lengthening of specific lexical tones in phrasal boundaries.

```
6.1 Tone
```


### (2) Tonal contrasts in trisyllabic roots

### **6.1.3 Tone patterns by root type and stress position**

As described in §5.3, Choguita Rarámuri roots can be characterized as lexically stressed or unstressed based on their stress properties in morphologically complex words. Both lexically stressed and lexically unstressed roots may have lexical tone realized in the surface tonic syllable. Unstressed roots may bear either H or L tone when bare or in neutral contexts, which can be attributed to a lexical tonal specification. Stressed roots, on the other hand, may bear HL, H or L tone, i.e., there are no unstressed HL-toned roots in the language.

From the logically possible lexical tone melodies by root type (monosyllabic, disyllabic and trisyllabic) and stress position (first, second and third syllable stress), two patterns are not yet attested, namely H and L-toned trisyllabic roots with first syllable stress. This gap in the tonal melody inventory is likely due to the combined effect of the relative infrequency of both trisyllabic roots and firstsyllable stress (from a corpus of 1040 roots, only 15 (1.5%) are trisyllabic roots with first-syllable stress). It is likely that the missing tonal contrasts will be found upon further investigation.

As discussed in Chapter 11, tonal melodies show systematic distributions in terms of phonological characteristics (whether a stem contains a lexically stressed or unstressed root), as well as morphological factors (the type of morphological constructions involved in inflected words).

### **6.1.4 Stress-based tonal neutralization**

As discussed in §5.3, stress distribution is governed by morphological factors. Given that tone distribution is dependent on stress, stress shifts result in tonal alternations in morphologically complex words. Specifically, morphologicallyconditioned stress shifts result in tone neutralization patterns, originally described in Caballero & Carroll (2015). There are two possibilities in terms of surface tone

### 6 Tone and intonation

in contexts where stress shifts. If a stress-shifting suffix is stressed after a stress shift, the stressed suffix syllable will bear the lexical tone of that suffix (this is the case when a monosyllabic or disyllabic unstressed root attaches a stress-shifting suffix). This pattern is shown in the examples provided in Table 6.1.


Table 6.1: Suffix lexical tone after stress shift

What these examples show is that suffixes bear their underlying tone when stressed, e.g., HL in the imperative singular *-ka*̂ (*to-ˈkâ* 'take it!' and *raha-ˈkâ* 'light it up!' in Table 6.1), L in the imperative plural *-sì* (*to-ˈsì* 'you all take it!' and *kimi-ˈsì* 'cover yourselves in blankets!' in Table 6.1), and H in the desiderative *-nále* (*uku-ˈnále* 'it is about to rain' and *kimi-ˈnále* 's/he wants to cover themselves in a blanket' in Table 6.1). The lexical tone of the root, which surfaces when the root attaches neutral morphological contexts, is deleted after the stress shift.

If a trisyllabic unstressed root attaches a stress-shifting suffix, the newly stressed syllable will be a stem syllable. A newly stressed stem syllable after a stress shift will bear a HL tone in these contexts, regardless of what the lexical tone of the root is (H tone in *roʔˈsówa* 'cough' or L tone in *naʔˈsòwa* 'stir' in Table 6.2).

One important aspect of this second pattern is that the surface tonal pattern of these morphologically complex words that have undergone a stress shift is not predictable based on the lexical tonal properties of root morphemes nor the lexical tones of suffixes. An example discussed in Caballero & German (2021) is that of unstressed roots attaching the imperative plural *-sì* suffix. As shown in Table 6.1 above, this suffix bears its lexical L tone when stressed (e.g., *to-ˈsì* 'you


Table 6.2: Stem tone after stress shift

all take it!'), but it will not surface if the stressed syllable after a stress shift is a stem syllable, i.e., the hypothetical forms *\*rosoˈwà-si* (vs. attested *rosoˈwâ-si* 'you all cough!') and *\*naʔsoˈwà-si* (vs. attested *naʔsoˈwâ-si* 'you all stir it') in Table 6.2, with L tone when attaching imperative plural *-sì* are unattested.

This tonal pattern is analyzed in Caballero & German (2021) as resulting from a process of HL tone insertion after a stress shift has deleted the root's lexical tone. This analysis is based on the following assumptions (listed in (3)):

	- Each morpheme has one and only one tone, which is lexically associated with one and only one Tone-Bearing-Unit (a mora within the tonic syllable)
	- In words containing lexically unstressed roots and neutral morphological constructions, stress is assigned within the root; the stressed syllable bears the underlying lexical tone of these roots
	- Stress shifts in shifting environments cause lexical root tones to delete
	- If the newly stressed syllable after a stress shift is a stem syllable, it is toneless and acquires an HL tonal melody

It could be argued that this HL tone insertion process is also at play in Spanish loanwords (Caballero & Carroll 2013; see also §7.3.1 and Chapter 11). Loanwords from Spanish are incorporated into Choguita Rarámuri with faithful prominence

### 6 Tone and intonation

to the stress location of the source words.<sup>9</sup> The stressed syllable in loanwords has a HL tone (no exceptions have yet been documented to this pattern). Relevant examples are given in (4) (the Spanish source words are provided in their orthographic form, where boldface indicates the stressed syllable):<sup>10</sup>



The analysis involving a HL tone insertion process for loanwords would involve the assumption that Spanish loanwords are lexically stressed but toneless in Choguita Rarámuri. Alternatively, the tonal properties of loanwords may be analyzed as involving a reinterpretation of the acoustic properties of Mexican Spanish prominence, which has been argued to include a H\* pitch accent in the stressed syllable in certain intonational contexts (focus-marked words in declarative sentences) (Prieto et al. 1995).

In sum, morphological factors condition stress shifts. Given the dependency tone has on stress for its distribution, the tonal alternations resulting in these contexts are largely predictable based on the lexical tonal properties of the morphemes that make up a morphologically complex word. Further discussion on

<sup>9</sup>As discussed in §8.6, nouns borrowed from Spanish are sometimes incorporated into the Choguita Rarámuri lexicon with the *-tʃ͡í* suffix, which has locative semantics in other contexts. This morphological strategy for loanword nativization is shown in (4g–h).

<sup>10</sup>As shown in (4g–h), some loanwords in Choguita Rarámuri have the same meaning but different source words: in the case of 'bottle', one loanword has as its source word *limeta*, an archaic word no longer in use in Northern Mexican Spanish, which coexists with the loanword originating from 'bottle', *botella*, a contemporary Spanish term.

6.2 Intonation

the mechanism of tonal patterns in morphologically complex words is provided in Chapter 11.

### **6.2 Intonation**

This section presents a basic description of the intonation patterns of declarative and interrogative sentences in Choguita Rarámuri and the realization of lexical tones in different phrasal contexts. The description of declarative intonation presented here builds on results discussed in Caballero et al. (2014), Garellek et al. (2015), and Caballero et al. (2022), based on qualitative and quantitative analysis of instrumental data recorded with four native Choguita Rarámuri speakers (two male, two female) using controlled elicitation.<sup>11</sup> Interactions between lexical tone, grammatical tone and intonation in this language are further addressed in Chapter 11. The intonational characteristics of interrogative constructions are addressed in §6.2.5 and Chapter 14.

This description assumes intonational primitives as laid out in the autosegmental-metrical framework (Pierrehumbert 1980, Beckman & Pierrehumbert 1986, Ladd 1986). These assumptions include the following:

	- Sequences of tonal targets (T) on an autosegmental tier yield intonational contours
	- There are two kinds of tonal targets: (i) tones associating with stressed syllables usually marking focal information or *pitch accents* (T\*) and (ii) tones associating with the edges of phrasal constituents or *boundary tones* (T%)
	- Tonal sequences are licensed by different phonological domains, arranged in a prosodic hierarchy (Selkirk 1986, Nespor & Vogel 1986)

In Choguita Rarámuri declarative intonation there is evidence of (i) a H% boundary tones, (ii) 'lead' (rhythmic) tones associated with lexical tones, and (iii) general and tone-specific non-tonal devices that encode intonation, including vowel rearticulation and lengthening at phrasal boundaries, all of which exhibit both

<sup>11</sup>The role of f0, duration and voice quality in intonation was assessed in these studies through examination of a corpus of data recorded with Choguita Rarámuri speakers. In this corpus, lengths of words and phrases, location of lexical stress and syntactic structure of phrases were manipulated in order to assess the timing and sequencing of tonal (lexical and post-lexical) patterns. Targets were balanced for lexical tone and stress, as well as utterance context (medial vs. final) and elicited in carrier phrases from a written prompt.

### 6 Tone and intonation

inter- and intra-speaker variation. As discussed in Chapter 14, interrogative constructions also involve register manipulation.

The declarative sentences examined all involved canonical word order with intended broad focus. As described in Chapter 13, Choguita Rarámuri is a headfinal language with SOV word order. Examples of this structure are provided in (6), with a clause headed by a ditransitive predicate with both pronominal and NP arguments (6a) and by a transitive predicate with pronominal arguments (6b).

	- a. *ˈmò ˈjêla taˈmí haˈré gaˈjêta ˈàko [ˈmò* 2sg.nom *ˈjê-la]subj* mother-poss *[taˈmí]p. obj.* 1sg.acc *[haˈré* some *gaˈjêta]s. obj.* cookie *ˈà-ki-o* give-pst.ego-ep 'Your mom gave me some cookies.' 'Tu mamá me dio galletas.' < BFL 09 1:89/el > b. *baʔaˈrîni ˈmí ˈàma baʔaˈrî=[ne]subj. [mí]obj. ˈà-ma*

tomorrow=1sg.nom 2sg.acc look.for-fut.sg 'I'll look for you tomorrow.' 'Mañana te voy a buscar.' < LEL 09 1:70/el >

Other Rarámuri varieties are also described as having SOV as their canonical word order, including Western Tarahumara (Burgess 1984), Rarómuri (Urique Tarahumara) (Valdez-Jara 2013), Pahuírachic Rarámuri (Estrada-Fernández & Tona 2013) and Rochéachi Rarámuri (Morales Moreno 2016). It should be noted, however, that SVO order is also frequently attested in Choguita Rarámuri declarative sentences with broad focus.<sup>12</sup>

### **6.2.1 H% boundary tones in declarative sentences**

The highest domain in the prosodic hierarchy assumed here is the Intonation Phrase (IP), which corresponds to a clause. The prosodic hierarchy is schematized in (7) (Selkirk 1980; 1996; Nespor & Vogel 1986; Hayes 1989).

<sup>12</sup>Whether this alternative word order results from contact with Spanish or other factors is a question I leave for future research.

6.2 Intonation

$$
\begin{array}{c}
(7) \quad \text{The procode hierarchy} \\
 \quad \text{Intational Phase} \\
 \quad \mid \\
 \quad \mid \\
 \quad \text{Phonological Phase} \\
 \quad \mid \\
 \quad \mid \\
 \quad \mid \\
 \quad \mid \\
 \quad \text{Procode Words} \\
\end{array}
$$

Counter to a strong cross-linguistic trend, where many languages exhibit a downward trend of pitch over the course of an utterance (Cohen et al. 1982, Ladd 1984), Choguita Rarámuri declaratives generally end with a final rise in f0, suggesting a H% boundary tone at the right edge of the Intonational Phrase (IP). This is shown in Figure 6.2 with a sentence composed of words with lexical L tones, *reˈhòi suˈnù oˈhòli* 'The man dekerneled corn'. As shown in this figure, H% boundary tones accommodate lexical tones: both the lexical L pitch target and the post-lexical H pitch target are clearly differentiated.

Figure 6.2: High boundary tone in declaratives (Garellek et al. 2015). Lexical pitch targets are represented with '\*' in the first tier; stressed syllables are represented with 'S' in the bottom tier.

A final rise in pitch at the IP boundary is not attested if the IP contains a lexical HL tone at the right edge: the rise expected with the presence of a H% is replaced by a pitch fall, i.e., H% boundary tones are overridden by lexical falling tones. This

### 6 Tone and intonation

is shown in Figure 6.3 with a sentence composed of words with lexical HL tones, *Maˈnuêli oˈkwâ koˈlî iʔˈkîli* 'Manuel bit two chili peppers'. As discussed in Garellek et al. (2015), this effect may enhance the distinction between lexical L and HL tones in phrase final position. All three lexical tones are clearly differentiated for all speakers across different intonational contexts (see further discussion in Chapter 11).

Figure 6.3: No high boundary tone in declaratives with lexical HL tones (Garellek et al. 2015). Lexical pitch targets are represented with '\*' in the first tier; stressed syllables are represented with 'S' in the bottom tier.

Finally, there is also evidence that sometimes a L% tone is found at the end of declarative sentences. This is exemplified in Figure 6.4. It is yet an open question what factors may govern the variability in the boundary tone, including interspeaker differences or other variables that have not yet been investigated, though the presence of H% vs. L% boundary tones does not seem to be conditioned by list intonation.

### **6.2.2 Optional rhythmic 'lead tones**

In addition to the presence of boundary tones and pitch raising at prosodic boundaries, Garellek et al. (2015) and Caballero et al. (2022) report on optional pitch targets that precede the lexical tonal targets, referred to as 'lead' tones in these studies. Lead tones are always an opposite pitch target than their lexical counterpart: if the lexical tone is H or HL, the lead tone involves a low pitch target; if

6.2 Intonation

Figure 6.4: L% boundary tone in a declarative sentence with lexical L tones (Caballero et al. 2022). Lexical tones are marked as L\* on the stressed syllables.

the lexical tone is L, on the other hand, the lead tone involves a high pitch target. Low lead tones of lexical H and HL tones are exemplified in Figure 6.5 and Figure 6.6, respectively. A high lead tone preceding a lexical L tone is exemplified in Figure 6.7; this figure also exemplifies the optional nature of lead tones: a high target is attested prior to the lexical L tone of *suˈnù* 'corn', but not prior to *oˈhòli* 'threshed'.

Another property of lead tones is that they are variable in their alignment: while they tend to occur in the pre-tonic syllable, they may be aligned with a syllable prior to the pretonic or the beginning of the tonic syllable (Garellek et al. 2015). In contrast, lexical tones are always consistently aligned with the tonic syllable and the fall of HL tones is always consistently realized with a post-tonic syllable, if there is one (see also Caballero & Carroll 2015).

The fact that optional lead tones are dependent on the pitch height of lexical tones which they precede and that H% boundary tones are not found when a lexical HL tone is at the end of the phrase in order not to override the lexical tone's f0 target suggest that preserving and enhancing the lexical tones of the language are important feature of the tonal grammar of Choguita Rarámuri. As will be discussed in Chapter 11, lexical tones may be replaced by grammatical tones in Choguita Rarámuri, but they appear to resist neutralization in their interaction with post-lexical tones.

### 6 Tone and intonation

Figure 6.5: Lead low tone preceding a lexical H tone (Garellek et al. 2015). Lexical pitch targets are represented with '\*" in the first tier; stressed syllables are represented with 'S' in the bottom tier.

Figure 6.6: Lead low tone preceding a lexical HL tone (Garellek et al. 2015). Lexical pitch targets are represented with '\*' in the first tier; stressed syllables are represented with 'S' in the bottom tier.

### 6.2 Intonation

Figure 6.7: Lead high tone preceding a lexical L tone (Garellek et al. 2015). Lexical pitch targets are represented with '\*' in the first tier; stressed syllables are represented with 'S' in the bottom tier.

### **6.2.3 Intonation patterns of declarative sentences**

The following subsections report on work published in Caballero et al. (2022) and illustrate typical intonation patterns of declarative sentences with sequences of identical tones in contexts intending broad focus.

### **6.2.3.1 Declarative sentences with lexical L tones**

As shown in Figure 6.8, a sentence with a sequence of lexical L tones will typically exhibit alternating low pitch targets for lexical tones and higher pitch targets for the preceding lead tones. This Figure also exemplifies the final high pitch target associated with the H% boundary tone which may be realized in the same stressed syllable as the final lexical L tone, creating a rising pitch contour. A lexical L tone target may be realized with a higher pitch lead tone in the same syllable if followed by another lexical L tone as attested in the word *raˈrài* 'sandals', where the high pitch target is associated with *tʃ͡aˈbè* 'before', the final word in the IP. This example also shows a gradual lowering in f0 of both high-pitched targets of the lead tones as well as the lexical L tones. Thus, sentences with lexical L tones may exhibit declination.<sup>13</sup>

<sup>13</sup>As discussed in Caballero et al. (2022) and §6.2.1 above, there is also evidence that lead tones are optional and that L% boundary tones may also be attested instead of the typical H% boundary tone of declaratives (as exemplified in Figure 6.4).

### 6 Tone and intonation

Figure 6.8: Intonation of a declarative sentence containing only L lexical tones (Caballero et al. 2022). Lexical tones are marked as L\* on the stressed syllables. Non-lexical pitch targets, including the H% boundary tone and high "lead" tones, are marked on the second tier.

As exemplified in Figure 6.7 and Figure 6.4 above, higher pitch targets associated with lead tones are optional. Thus, as discussed in more detail in Caballero et al. (2022), lexical L tones are only associated with a low pitch target in stressed syllables.

### **6.2.3.2 Declarative sentences with lexical H tones**

Declarative sentences with sequences of lexical H tones also show an alternating pattern between the high pitch targets associated with lexical tones and preceding lower pitch targets of lead tones. This is exemplified in Figure 6.9. This example also shows the highest pitch peak at the end of the IP associated to the H% boundary tone.

Furthermore, the example in Figure 6.9 shows that H lexical tones in a sequence exhibit a progressive rise in the f0 of the lexical H targets. The lowpitched lead pitch targets may also be upstepped (represented in the figures as "Lˆ"), given that they are low with respect to the following lexical H tone, but not the preceding one.

This example also shows the spreading of H tones onto following post-tonic syllables, as in *aˈwíame* 'dancers', where the H tone spreads from the tonic to the post-tonic syllable. On the other hand, no spreading is attested after *baˈʰt͡ʃí* 'zucchini'. Instead, a low lead tone is attested immediately preceding the following H tone in *koˈʔáli* 'ate'.

6.2 Intonation

Figure 6.9: Intonation of a declarative sentence containing only H lexical tones (Caballero et al. 2022). Lexical tones are marked as H\* on the stressed syllables. Non-lexical pitch targets, including the H% boundary tone and low "lead" tones, are marked on the second tier.

Finally, there is also evidence of the optionality of low lead tones preceding H lexical tones, as exemplified in Figure 6.10: there is no low pitch target between the two lexical H tones in *riˈʰtê ˈpáli* '(they) threw stones'.

Figure 6.10: Intonation of a declarative sentence containing only H lexical tones with optional low lead tones (Caballero et al. 2022). Lexical tones are marked as H\* on the stressed syllables. Non-lexical pitch targets, including the H% boundary tone and low "lead" tones, are marked on the second tier.

### **6.2.3.3 Declarative sentences with lexical HL tones**

As discussed in §6.2.1 above, in sentences with sequences of lexical HL tones (represented as H\*L, since the low pitch target of this lexical tone may be realized

### 6 Tone and intonation

in a post-tonic syllable), there is no evidence of H% boundary tones. As seen in Figure 6.11, HL lexical tones may have optional low lead tones; the high pitch target of the last lexical HL tone is higher than the preceding two high pitch targets, and the low pitch targets of the pre-final HL tones exhibit declination. This suggests that HL tones in the final position of the IP exhibit pitch range expansion.

Figure 6.11: Intonation of a declarative sentence containing only HL lexical tones (Caballero et al. 2022). Lexical tones are marked as H\*L on the stressed syllables. Non-lexical pitch targets (low "lead" tones) are marked on the second tier.

As discussed in the next section (§6.2.4), utterance fnal HL tones may be optionally rearticulated.

### **6.2.4 Non-tonal encoding of intonation**

Encoding of intonational contrasts in Choguita Rarámuri also involves non-tonal effects. Non-tonal effects associated with Choguita Rarámuri intonational encoding at prosodic boundaries include: (i) vowel rearticulation and (ii) lengthening of L tones. This section summarizes the findings reported on Caballero et al. (2014) and Aguilar et al. (2015). Controlled elicited data was recorded with four native Choguita Rarámuri speakers (two male and two female), and assessment of the role of lengthening and non-modal phonation assessed through quantitive analysis of acoustic and electroglottographic (EGG) data.<sup>14</sup>

<sup>14</sup>Lengthening was analyzed using VoiceSauce (Shue et al. 2011) and phonation was assessed through analysis of CQ (Contact Quotient) calculated in EggWorks.

6.2 Intonation

Vowel rearticulation is exclusively attested with HL tones at the right edge of the IP boundary. This is an effect that is optional for some speakers but very robust for others. Vowel rearticulation may involve glottal closure or glottalization, with frequent devoicing of the final portion of the vowel (VʔV̥ or VV̰V). An example is shown in Figure 6.12.

Figure 6.12: Rearticulation of HL tones in utterance-final position (Caballero et al. 2014)

The following examples show the contrast between rearticulated HL tones and non-rearticulated H and L tones of (near-)minimal pairs elicited in a frame sentence ('*Maˈnuêli riˈwàli oˈkwâ/waʔˈlû* \_' 'Manuel saw two/a big \_').

(8) Rearticulated vs. non-rearticulated tones in phrase-final position

```
a. [naˈʰpóʔò] HL
    /naˈʰpô/
    'prickly pear'
    'tuna' < BFL 11-nahpo >
```

```
b. [oˈʰkó] H
    /oˈʰkó/
    'pine tree'
    'pino' < BFL 12-ohko >
```
6 Tone and intonation

$$\begin{array}{ll} \text{c.} & \text{[na'?i'i]} & \text{HL} \\ & \text{/na'?i'} \\ & \text{'here'} \\ & \text{'aqui'<{BFL 11-naqi}>} \\ \text{d.} & \text{[na'?i]} & \text{L} \\ & \text{/na'?i'} \\ & \text{'five'} \\ & \text{'fivego'<{BFL 12-naqi}>} \end{array}$$

Rearticulation of HL tones is an effect robustly attested for participant female speakers and only optionally attested or marginal for participant male speakers. A systematic study controlling for demographic factors may reveal whether gender plays a role in inter-speaker variation. A follow up study of the role of phonation in tonal encoding in Choguita Rarámuri reported in Caballero et al. (2022) reveals that voice quality improves discrimination of lexical tones, though voice quality measures exhibit patterns that are speaker-dependent. Readers are referred to this reference for more details.

There is also evidence for significant intra-speaker variation in the non-tonal encoding of lexical tones in certain prosodic positions. As shown in the following examples, speakers may deploy additional strategies in utterance final position, including voiceless aspiration. In the example in Figure 6.13, a rearticulated HL tone exhibits modal phonation, followed by creaky phonation, followed by breathy phonation and then aspiration.

Finally, there is evidence to suggest that an additional non-tonal device in the encoding of Choguita Rarámuri intonation is lengthening: Caballero et al. (2014) and Aguilar et al. (2015) report that there is significant lengthening utterancefinally of L tones, where L tones have greater duration than HL and H tones at the right edge of the IP boundary for some speakers. It should be noted, however, that a follow up study did not find a significant role for duration in the encoding of lexical tones in the language using linear discriminant analysis (Caballero et al. 2022).

In sum, preliminary research shows Choguita Rarámuri exhibits the following phonation-lexical tone interactions: (i) rearticulation exclusive to HL tones, and (ii) increased lengthening of L tones. Fundamental frequency (f0) and voice quality interact in many languages in the realization of tonal contrasts (Kingston 2005, Kuang 2013) (e.g., glottalization in Dagbani (Gur; Hyman 1993), and Roman and Tuscan Italian (Di Napoli 2015); breathiness/creakiness in Chickasaw (Muskogean; Gordon 2005); lengthening in Bantu languages (Downing 2008;

6.2 Intonation

Figure 6.13: Voiceless aspiration utterance-final and phasing of the rearticulated vowel for BFL (female speaker) (Caballero et al. 2014).

2010; Hyman & Monaka 2011, among other effects). More detailed examination of the tone-phonation interaction in Choguita Rarámuri will inform a growing body of literature that seeks to understand the interaction of multiple phonetic dimensions in the implementation of prosodic contrasts in a range of typologically diverse languages.

### **6.2.5 Interrogative intonation**

Choguita Rarámuri interrogative constructions are encoded through morphosyntactic and/or intonational devices. Chapter 14 provides a comprehensive description of the morphosyntactic and prosodic encoding of polar interrogatives (§14.2.1) and content questions (§14.2.2). This section summarizes the intonational properties of both types of constructions. As discussed here and in §11.4, the specific intonation patterns attested in interrogatives are shaped by the interaction between lexical tones and intonemes in utterance-final position.

### **6.2.5.1 Polar question intonation**

Polar questions may be classified morphosyntactically into three types: (i) morphosyntactically unmarked polar questions; (ii) polar questions with interroga-

### 6 Tone and intonation

tive particles; and (iii) polar questions with interrogative tags.

In the first type, the interrogative clause is equivalent morphosyntactically to its declarative counterpart, differing only in intonation. A minimal pair between a declarative sentence and its morphosyntactically unmarked polar question counterpart in Choguita Rarámuri is provided in (9). Figure 6.14 and Figure 6.15 show the intonational difference between the declarative sentence in (9a) and the polar interrogative in (9b), respectively.

(9) Declarative vs. morphosyntactically unmarked polar question

	- already bury-pst 'Did s/he bury him/her?' '¿Lo enterró?' < BFL el1307 >

In the declarative sentence exemplified in Figure 6.14, there is a falling f0 contour of the lexical HL tone of the verb (*ˈtô* 'to bury'). As consistently attested in declarative entences with HL lexical tones in utterance-final position, there is no evidence of a H% boundary tone, as the falling lexical tone in the stressed syllable overrides the boundary tone (see §6.2.1 above and §11.4 in Chapter 11). When compared to the declarative sentence in Figure 6.14, the interrogative sentence in Figure 6.15 exhibits raised f0 in the stressed syllable (396 Hz in this example vs. 289 Hz in the declarative in Figure 6.14 for the same female speaker (BFL)). As discussed in §14.2.1.1, this effect may be attributed to the interrogative H% intoneme aligning with the peak of the lexical HL tone in this stressed syllable.

A different intonation pattern is documented in interrogative sentences with a final stressed syllable specified for lexical L tone. This is shown in Figure 6.16.

As shown in the example in Figure 6.16, the lexical tone of the verb root is associated with the stressed syllable and the H% boundary tone docks on a follow-

<sup>15</sup>In this particular example, the suffix vowel surfaces as [o] given an optional round harmony process, where non-round vowels of certain suffixes may become round when preceded by a stressed back stem vowel. This optional process is only attested in the speech of younger speakers. For more details about this process, see §9.4.3.5.

Figure 6.14: Declarative utterance: *ˈmá ˈtôlo* 'S/he buried him/her' (< BFL el1170 >)

Figure 6.15: Morphosyntactically unmarked polar interrogative: *ˈmá ˈtôli?* 'Did s/he bury him/her?' (< BFL el1307 >)

### 6 Tone and intonation

Figure 6.16: Accommodation of L tone and H% boundary tone in *ˈmá ˈnèli?* 'Did s/he see him/her?' (< BFL el1307 >)

ing, unstressed syllable. Lexical L tones are preserved in their interaction with intonemes, as attested in declarative sentences (see §11.4 for further discussion).

A second type of polar interrogative sentence is encoded through a polar interrogative particle (*ˈátʃ͡e* or its reduced form *a*), which occurs in clause initial position. The following minimal pair illustrates the intonational difference between a declarative with a lexical L tone in utterance-final position ((10a) in Figure 6.17) and its polar interrogative counterpart with an interrogative particle ((10b) in Figure 6.18).

(10) Declarative vs. polar interrogative with *ˈátʃ͡e*

```
a. ˈmá naˈwàli
    ˈmá
    already
             naˈwà-li
             arrive-pst
    'S/he already arrived.'
    'Ya llegó.' < SFH-nawa-arrive-L-minimal-sets >
b. ˈátʃe ˈmá naˈwàli
    ˈátʃe
    Q
          ˈmá
          already
                  naˈwà-li
                  arrive-pst
    'Did s/he already arrive?'
    '¿Ya llegó?' < SFH-nawa-arrive-L-minimal-sets >
```
Figure 6.17: Declarative with utterance-final lexical L tone in *ˈmá naˈwàli* 'S/he already arrived.'

Figure 6.18: Polar interrogative with *ˈátʃ͡e* and utterance-final lexical L tone in *ˈátʃ͡e ˈmá naˈwàli* 'Did s/he already arrive?'

As shown in these examples, the intonational encoding of polar questions with interrogative particles involves (i) presence of an obligatory H% boundary tone (recall from §6.2.1 above that declaratives generally end with a H% tone, but that sometimes declaratives end with a L% tone instead), and (ii) raised register across the utterance. As attested in other interrogative and declarative sentences, lexical L tones are preserved: as seen in Figure 6.18, the sharp rise and peak of the H% tone is aligned with the final, post-tonic syllable. Thus, while H% tones are optional in declaratives, these appear to be required in interrogatives. Furthermore, interrogatives exhibit raised register. as shown next, content questions are also characterized intonationally by raised ragister across the utterance.

### 6 Tone and intonation

### **6.2.5.2 Content question intonation**

Content questions in Choguita Rarámuri have a question marker (which may be complex) that appears in clause initial position, as exemplified in (11).

```
(11) Content questions
```

Content (or information) questions are not only morphosyntactically distinct from their declarative counterparts, but are also distinctive in terms of their intonational pattern: as shown in the contrast between a declarative sentence (shown in (12a), Figure 6.19) and a content question (shown in (12b), Figure 6.20).

### (12) Declarative vs. content question

	- who who arrive-pst 'Who arrived?' '¿Quién llegó?' < SFH-nawa-arrive-L-minimal-sets >

The f0 contour of the content question in Figure 6.20 shows register raising across the utterance, as well as a H-toned target in utterance-final position, which can be attributed to a boundary H% tone in the last unstressed syllable of the utterance.

Figure 6.19: Declarative with utterance-final lexical L tone

Figure 6.20: Content question with utterance-final lexical L tone

A H% boundary tone is also attested in content questions where a lexical H tone is associated with the final stressed syllable of the utterance. This is shown in (13), with a contrast between a declarative in (13a) (illustrated in Figure 6.21) and a content question in (13b) (illustrated in Figure 6.22).

(13) Declarative vs. content question: utterance-final, lexical H tone

a. *ˈmá muˈrúli ˈmá* already *muˈrú-li* carry.in.arms-pst 'S/he already carried it in their arms' 'Ya lo cargó en brazos' < BFL-muru-carry-H-minimal-sets >

### 6 Tone and intonation

b. *ˈhêpi ˈkwâ muˈrúli ˈhêpi* who *ˈkwâ* who *muˈrú-li* carry.in.arms-pst 'Who carried it in their arms?' '¿Quién lo cargó en brazos?' < BFL-muru-carry-H-minimal-sets >

Figure 6.21: Declarative with utterance-final lexical H tone

Figure 6.22: Content question with utterance-final lexical H tone

In the declarative sentence in Figure 6.21 there is no evidence of a H% boundary tone, which, as discussed in §6.2.1, is optional in declaratives. In the content question represented in Figure 6.22, on the other hand, there is a clear high pitch target in the final, unstressed syllable of the utterance. This high pitch target is higher than the one associated with the lexical H tone of the stressed syllable (a difference of almost 30Hz in this particular example).

6.2 Intonation

Finally, in content interrogative sentences containing a lexical HL tone in utterance final position, there is no evidence of a H% boundary tone nor any register manipulation when the content question contains a lexical HL tone. This is shown in the contrast between the declarative sentence in (14a) (Figure 6.23) and its content question counterpart in (14b) (Figure 6.24).

(14) Declarative vs. content question: utterance-final, lexical HL tone

```
a. ˈmá iˈsîli
    ˈmá
    already
             iˈsî-li
             pee-pst
    'S/he already peed.'
    'Ya orinó.' < BFL-isi-pee-HL-minimal-set >
b. ˈhêpi ˈkwâ iˈsîli
    ˈhêpi
    who
           ˈkwâ
           who
                 iˈsî-li
                 pee-pst
    'Who peed?'
    '¿Quién orinó?' < BFL-isi-pee-HL-minimal-set >
```
Figure 6.23: Declarative with utterance-final lexical HL tone

In these examples (produced by the same female speaker, BFL), the highest pitch peak associated with the lexical HL tone is comparable in both sentences (270Hz in the declarative in Figure 6.23 and 260Hz in the content question in Figure 6.24), and their pitch countours largely equivalent. Thus, the comparison between the intonational contours of these two sentences suggests that there is no distinctive intonational encoding of a content question where HL tones override

### 6 Tone and intonation

Figure 6.24: Content question with utterance-final lexical HL tone

the H% tone associated with interrogative constructions elsewhere. This stands in contrast to polar questions, where, as shown in Figure 6.15 above, there is significantly raised f0 in the final stressed syllable of the interrogative utterance.

### **6.2.5.3 Summary**

In Choguita Rarámuri, interrogative sentences are characterized by the following intonational characteristics:

	- A boundary H% tone targets the last stressed syllable of the utterance.
	- There is raised register across the utterance.

As shown in this section, interrogative constructions that are morphosyntactically equivalent to declaratives are encoded exclusively through intonation, but prosody also plays a role in those interrogative constructions where there is a morphosyntactic device encoding the interrogative meaning. Cross-linguistically, the existence of morphosyntactic encoding of different utterance types may preclude the use of distinctive intonational structures for the same purposes in some languages (e.g., Navajo (Athabaskan; McDonough 2002)).

Miller (1996) reports that closely-related Mountain Guarijío (Taracahitan; Uto-Aztecan) also has polar questions that are morphosyntactically equivalent to

6.2 Intonation

their declarative counterparts, with the interrogative meaning encoded through ascending intonation ("*[g]eneralmente tienen entonación ascendente*") (1996: 112).<sup>16</sup>

While polar questions show some degree of variation in the realization of the boundary tone and the degree to which register is raised (discussed below and in Chapter 11), they consistently exhibit a high pitch target utterance-finally. As shown in this section, the magnitude of the peak is correlated with the presence/absence of an overt morphological device to encode a question, with the highest pitch excursions attested utterance-finally with morphosyntactically unmarked polar questions.

<sup>16</sup>To the best of my knowledge, no study has addressed the intonation of Mountain Guarijío through assessment of instrumental data.

# **7 Other word-level supra-segmental processes**

In addition to stress and tone, Choguita Rarámuri features other word-level phonological phenomena that involve supra-segmental processes and/or prosodic domains (for further discussion of word-prosodic phenomena in the language, see Chapter 11). These phenomena are addressed in this chapter and include: (i) a disyllabic window for glottal stops (§7.1); (ii) minimality effects in verbs (§7.2); and (iii) loanword prosodic adaptation patterns (§7.3.2).

### **7.1 Glottal stop: an initial disyllabic window**

As discussed in Chapter 4, Choguita Rarámuri syllables are underlyingly CV in shape, with no codas, except for glottal stop. This glottal stop displays strict restrictions on its distribution: glottal stop can only occur between the first and second syllable, either intervocalically (1) or pre-consonantally (2).

(1) Intervocalic glottal stop

```
a. koˈʔá
    'eat'
    'comer' {GFM tx905:0:54.0}
b. raˈʔìtʃ͡a
    'speak'
    'hablar' {JLG co1235:6:09.5}
c. baʔuˈrâ
    'brand livestock'
    'marcar con fierro los animales' < SFH VDB/el >
d. boʔaˈlâ
     boʔa-ˈlâ
     feather-poss
    'feathers, fur'
    'plumas, pelaje' < LEL 5:127-142/el >
```
7 Other word-level supra-segmental processes

```
e. moˈʔè
    'weave'
    'tejer' < SFH 05 1:143/el >
```
	- a. *raʔˈná* 'explode' 'tronar' < SFH 04 1:81/el >
	- b. *aʔˈtá* 'arch' 'arco' < SFH NDB/el >
	- c. *baʔtʃ͡iˈwí* 'sound of water dripping' 'sonido de goteo' < BFL 07 2:31/el >
	- d. *biʔˈrì* 'twist' 'torcerse' < SFH 04 1:114/el >
	- e. *baʔˈwí* 'water' 'agua' {JLG co1234:6:04.4}
	- f. *raʔˈtàtʃ͡i* 'hot/dry season' 'tiempo de calor' {GFM tx904:1:42.9}

The examples in (3) show that glottal stop is associated with the word-initial syllable, as opposed to being associated to the immediately pretonic position: each of the following examples involves forms with a CV**ʔ**VCV syllable structure with stress in the third syllable.

	- a. *boʔoˈbû boʔo-ˈbû* unpluck-tr.imp.sg 'Unpluck it!' '¡Desplúmalo!' < SFH 08 1:51/el >
	- b. *raʔaˈmâsima raʔaˈmâ-si-ma* give.advice-mot-fut.sg

### 7.1 Glottal stop: an initial disyllabic window

'S/he will go along giving advice.' 'Va a ir aconsejándolos.' {JMF tx816:0:39.9}

c. *raʔiˈtʃ͡âsa raʔiˈtʃ͡â-sa* speak-cond 'if she speaks' 'si habla' {LEL tx221:5:04.8}

Finally, when verb roots bearing a glottal stop, incorporate a body-part nominal root, the glottal stop underlying to the verbal root is deleted in the surface form, as exemplified in (4) (in these forms, the compound form has stress in the first syllable of the verb, the head of the compound; for more details about the stress and tone properties of compounds, see §11.3.6).

### (4) Glottal stop deletion in noun incorporation

a. [kutaˈbîri] /kuˈtâ+bi**ʔ**ˈrì/ neck+twist 'to neck-twist' 'torcer el cuello' < BFL 07 1:163/el >


As shown in these examples, the underlying glottal stop of the verbs *biʔrì* 'twist' (4a) and *biʔwá* 'clean' (4b–c) is lost in noun incorporation. This process suggests that glottal stop must not only emerge within an initial disyllabic window within the prosodic word (after the first vowel of this domain), but it must also emerge in the lexical root to which it is underlyingly associated.

Similar distributional restrictions of glottal stop in closely related Mountain Guarijío are analyzed in Caballero (2006) and Haugen (2014) as resulting from

### 7 Other word-level supra-segmental processes

surface glottal stops being underlyingly a floating constricted glottis ([+c.g.]) feature. Alternatively, these distributional facts can be analyzed as resulting from distributional restrictions of a glottal stop segment. Outside of the Uto-Aztecan language family, similar restrictions on laryngeal features have been documented in varieties of Basque, where the glottal fricative ([h]) and aspirated stops are restricted to occur within an initial disyllabic window (Hualde 2018).

Finally, the phonemic glottal stop should be distinguished from glottalization associated with HL tones in utterance final position, a tone-specific effect where a stressed vowel bearing a HL tone in the IP boundary may be rearticulated (/î/ → [íʔì]), as described above in §6.2.4.

### **7.2 Minimality effects**

There is evidence in Choguita Rarámuri for a minimal word size for verbs. Verbal inflectional categories like recent past, present and the imperative singular may be realized by the bare verbal stem. An impressionistic assessment of unsuffixed, monosyllabic verbs inflected for these categories reveals they involve significantly longer vocalic nuclei than their suffixed counterparts. In (5), I present examples of these monosyllabic verb roots when inflected for imperative singular (realized as the lengthened, bare stem), which contrast with suffixed forms that do not involve lengthening (shown in the second column).

(5) Vowel lengthening of monosyllabic roots

```
a. [ˈpáa]
    /ˈpá/
     throw.imp.sg
    'Throw!'
    '¡Tíralo!' {JLG co1234:3:19.8}
b. [ˈpáma]
    /ˈpá-ma/
     throw-fut.sg
    'S/he will throw it.'
    'Lo va a tirar.' {JLG co1234:16:35.6}
c. [ˈmàa]
    /ˈmà/
     run.sg.imp.sg
    'Run!'
    '¡Corre!'
```
7.2 Minimality effects

d. [ˈmàma] /ˈmà-ma/ run.sg-fut.sg 'S/he will run.' 'Va a correr' < BFL 06 el14/el > e. [ˈnóo] /ˈnó/ look.imp.sg 'Look!' '¡Mira!' f. [ˈnóma] /ˈnó-ma/ look-fut.sg 'S/he will look.' 'Va a mirar.' < BFL 06 el38/el > g. [ˈàa] /ˈa/ find.imp.sg 'Find it!' '¡Encuéntralo!' < BFL 05 1:112/el > h. [ˈàma] /ˈà-m/a find-fut.sg 'S/he will find it.' 'Lo va a encontrar.' < BFL 05 1:112/el >

Precise measurements of vowel length of a small sample of monosyllabic roots in unsuffixed and suffixed contexts confirms there is a difference between the vowel durations of unsuffixed, monosyllabic (imperative or present inflected) words and their suffixed (disyllabic) counterparts. As exemplified in (6), the vowel duration of unsuffixed monosyllabic verbs ranges between 140–200ms, while the vowel duration of the suffixed counterpart ranges between 60–120ms.

### 7 Other word-level supra-segmental processes

### (6) Vowel duration of monosyllabic verbal roots


Minimality constraints are asymmetrical in Choguita Rarámuri, as not all morphological classes of words show a minimality condition. As commonly found cross-linguistically, function words are canonically monosyllabic and do not undergo lengthening or undergo any prosodic augmentation process. Furthermore, other major word classes in Choguita Rarámuri are also exempt from minimality requirements. This is exemplified with monosyllabic nouns, which do not exhibit lengthening in unsuffixed contexts like verbal monosyllables do. Measurements of vocalic durations of monosyllabic nouns in unsuffixed and suffixed contexts shows that unsuffixed nouns have comparable durations to the vowels of monosyllabic verb roots when suffixed (7).

(7) Vowel duration of monosyllabic nominal roots


Thus, there is a minimal word constraint which targets verbs. This constraint is defined in (8):

(8) Choguita Rarámuri minimal word constraint for verbs Minimal word (X<sup>0</sup> ) = [[]]<sup>∑</sup>

### 7.3 Loanword prosody

According to this constraint, all verbs in Choguita Rarámuri are at least two moras, where consonants are non-moraic. While vowel length is not contrastive in Choguita Rarámuri, long vowels are derived to satisfy a minimal word constraint, as proposed in this section. Long vowel sequences can also be found after semi-vowel deletion (§4.4.1) and *h* deletion (§4.3), and in morphologically restricted contexts (addressed in Chapter 9, such as compensatory lengthening (§9.4.3.2), and lengthening triggered by a specific suffix (§9.4.3.3)). The role of vowel length and the minimality restriction of Choguita Rarámuri inflected verb stems is discussed further in Chapter 11.

### **7.3 Loanword prosody**

This section presents an overview of prosodic adaptation of Spanish loanwords in Choguita Rarámuri (see also Caballero & Carroll 2013). As discussed in this section, Choguita Rarámuri retains the original prominence from Spanish, which features a stress-accent system, but deploys several repair strategies to make loanwords conform to prosodic restrictions on morphological structure. Specifically, repair strategies are sensitive to strong prosodic restrictions on morphological structure in Choguita Rarámuri, predominance of certain word/morpheme types and acoustic properties of Spanish prominence.

Speakers of Rarámuri varieties have been in contact with Spanish speakers since the seventeenth century (according to Alegre (1956)[1767], the first contact with Spanish speakers in the Sierra Tarahumara was in 1607).<sup>1</sup> As discussed in §1.3.1, while the Rarámuri have historically avoided contact with the *mestizo* (non-indigenous Mexican) population through the retreat to isolated, more mountainous areas (Merrill & Burgess 2014, Pintado Cortina 2012b), Rarámuri speakers in Choguita have had intense contact with Spanish speakers for almost the entirety of the 20th century (e.g., the first road in Choguita was built circa 1920 (Casaus 2008). This intense language contact across the Sierra Tarahumara has brought about a high level of bilingualism and language attrition (see also discussion in §1.3).

Prosodic loanword adaptation processes in Choguita Rarámuri have a bearing on the question of what may be possible patterns of loanword prosody adaptation in contact situations with languages with divergent word prosodic systems

<sup>1</sup>While Rarámuri speakers have been in contact with speakers of other Uto-Aztecan languages (including speakers of Nahuatl (Aztecan)) and other North American Indian languages, before and after colonial times, the focus here is in Rarámuri-Spanish language contact phenomena.

### 7 Other word-level supra-segmental processes

(stress, tone, "pitch accent") (Kubozono 2006; Kuang 2013). In contrast to Choguita Rarámuri, which features a 'hybrid' word prosodic system (with both stress and tone), Spanish has lexically contrastive stress, but no lexical tone. The main features of the Spanish stress systems are the following:

	- a. Primary stress falls within a final three-syllable window (Hualde 2012)
	- b. Stress assignment involves trochaic feet (Lipski 1997: 560)
	- c. There is iterative stress assignment aligned to the right edge of the prosodic word, with a predominance of penultimate stress in the language
	- d. There is no phonological vowel length contrast (Chávez-Peón 2007; Hualde 2012)

The Spanish stress-accent system finds a direct correlate in the stress-accent system of Choguita Rarámuri, while the tonal properties of loanwords can be interpreted as arising from either a default grammatical tone process (as described in §6.1.4) or as a reinterpretation of Spanish post-lexical tonal phenomena.

The focus here is on loanwords that exhibit some degree of phonological adaptation, either segmentally or prosodically. No assumptions are made regarding the degree to which hypothesized loanwords which exhibit no adaptation constitute instead code-switching of bilingual speakers. Given the lack of sociolinguistic data, there are also no assumptions made here regarding the degree of adaptation as a diagnosis for sociolinguistic situation/level of bilingualism (see discussion in Sicoli 1999) nor the relative chronology of incorporation of loanwords. These questions are left for further research.

The Spanish loanword data examined are mostly penultimate stressed Spanish nouns. Some of the loanwords were likely introduced via Nahuatl, e.g., *basa*ˈ*lówa* 'stroll' (from Spanish 'pasear' and the Náhuatl *-oa* impersonal suffix) and *ko*ˈ*máare* 'comadre' and *kom*ˈ*páare* 'compadre' (Nordell 1984).

### **7.3.1 Exceptionless prosodic loanword adaptation patterns**

There are several prosodic loanword adaptation patterns that are exceptionless in Choguita Rarámuri. These are listed in (10).

7.3 Loanword prosody

	- a. Stressed syllables in Spanish are borrowed as the prominent syllables in Choguita Rarámuri via stress and a HL tone.
	- b. Stressed vowels in Spanish retain their original quality when incorporated into Choguita Rarámuri.
	- c. Consonant final source words from Spanish correspond to vowel final loanwords in Choguita Rarámuri.
	- d. Pre-tonic syllable truncation of sourcewords preserves prominence in loanwords within the Choguita Rarámuri initial three syllable stress window.
	- e. Compounds and phrases are borrowed with prominence in the second element.

As shown in (4) in §6.1.4 above, stressed syllables of Spanish source words are retained as the stressed syllables of loanwords in Choguita Rarámuri. Without exception, the stressed syllables of loanwords bear a HL tone. This is exemplified in (11) (stressed syllables in Spanish source words are highlighted in boldface).<sup>2</sup>

### (11) Stress and tone properties of Spanish loanwords


As shown in these examples, while unstressed vowels of source words may undergo adaptation in Choguita Rarámuri (e.g. (11c)), stressed vowels in loanwords retain the original Spanish source word quality when incorporated into Choguita Rarámuri.

<sup>2</sup>*Limeta* in (11a) is an archaic word no longer in use in contemporary Northern Mexican Spanish. As for the loanword *leˈhîdotʃ͡i* in (11i): as discussed in §1.2 above, *ejido* is a rural land plot for collective use by community members (ejidatarios) administered by the federal government.

### 7 Other word-level supra-segmental processes

Another exceptionless prosodic pattern attested in loanwords concerns syllable structure: consonant final source words correspond to vowel-final loanwords in Choguita Rarámuri, given the requirement of all Prosodic Words in Choguita Rarámuri to be vowel-final (see §11.1 for more details about this). This is shown in (12).

(12) Adaptation of consonant-final source words


As shown in these examples, vowel epenthesis word finally repairs loanwords to satisfy the optimal prosodic structure of Choguita Rarámuri. The quality of the epenthesized vowel is either identical to the quality of the stressed vowel if the vowel is low (e.g., 12b–c), or may undergo optional reduction if it is a mid, back rounded vowel (e.g., 12d). If the final vowel is the front, mid vowel [e], the epenthesized vowel is [i] (12a).

Spanish loanwords with final or penultimate stress where stress would fall outside the Choguita Rarámuri initial three syllable stress window undergo pretonic syllable truncation in order to confirm to this strict prosodic requirement of the language. Relevant examples are shown in (13), where truncated syllables from the Spanish source words are highlighted in boldface.

(13) Pre-tonic syllable truncation in Spanish loanwords


### 7.3 Loanword prosody

Spanish compounds and phrases, including very commonly compound proper names, are borrowed into Choguita Rarámuri with the prosodic structure of compounds in Choguita Rarámuri (see §11.3.6), which without exception involves prominence in the second element. As seen in (14), compounds and phrases are incorporated with the original prominence of the second member of the construction.

(14) Loanword prosodic adaptation of compounds and phrases


These examples show that pre-tonic syllable truncation of the source word (e.g., *sus* from *Jesús* in (14c)) keeps stress in the original Spanish stress pattern without violating the strict requirement that stress is placed within the initial three-syllable stress window margin.

### **7.3.2 Optional prosodic loanword adaptation patterns**

In addition to exceptionless loanword prosodic adaptation patterns, there are also patterns where adaptation does not involve repair of illicit prosodic sequences and patterns where an adaptation process applies only optionally. These processes, however, bring the prosodic form of loanwords in conformity with the canonical prosodic structures of Choguita Rarámuri (for discussion of canonical prosodic structures in the language, see §11.3.3 below).

One first case involves loanwords where pre-tonic truncation applies to the source word, despite that there is no violation of a prosodic constraint in Choguita Rarámuri. Relevant examples are given in (15), with the truncated syllable highlighted in boldface in the Spanish sourceword.

(15) Pre-tonic truncation with no violation of Choguita Rarámuri prosody


### 7 Other word-level supra-segmental processes

Without syllable truncation, these forms would conform to the three-syllable stress window. However, syllable truncation in these cases renders these loanwords with the most frequent prosodic forms attested in the native vocabulary, namely stems with second syllable stress (see §11.3.3), which are represented both by forms with lexically specified stress as well as forms that receive stress by default in the second syllable of the root (see §5.3).

There is also optional consonant cluster repair, with pre-tonic consonant clusters being variably repaired by epenthesis (16a–b) or consonant deletion (16c–e). Post-tonic consonant clusters, on the other hand, are generally not repaired (16f– j) (Spanish consonant clusters that are adapted or fail to undergo adaptation are highlighted in boldface).



While the underlying syllable structure of Choguita Rarámuri is CV (with optional glottal stop codas in the first syllable of the Prosodic word), stress-based vowel deletion yields consonant clusters in the surface form (§5.2). However, the distribution of CC clusters in surface form is not homogeneous: being a suffixing language with predominance of disyllabic roots, a highly frequent pattern of morphologically complex words in Choguita Rarámuri is the existence of posttonic consonant clusters at stem-suffix boundaries. Thus, the lower frequency of repaired consonant clusters in post-tonic position in Spanish loanwords is related to this asymmetric distribution of consonant clusters in native Choguita Rarámuri words.

Finally, some of the loanword examples above exhibit long vowel sequences (e.g., (13c), (14e) and (16d)). As discussed in Chapter 3, Chapter 11 and §4 above, there is no contrastive vowel length in Choguita Rarámuri, but there are phonological phenomena that are quantity-sensitive and surface long vowel sequences

### 7.3 Loanword prosody

that are derived through morpheme-specific effects (including suffix-triggered vowel lengthening, §9.4.3.3 and compensatory lengthening, §9.4.3.2). Further examples of loanwords with long vowels are provided in (17).3,4

(17) Long vowel sequences in loanwords


Long vowels in Choguita Rarámuri loanwords are variably attested. This particular pattern may reflect durational properties of Spanish stressed vowels, as discussed in Chávez-Peón (2007). Specifically, long vowels in Choguita Rarámuri loanwords match "semi-long" Spanish stressed vowels, attested in words with penultimate stress before a voiced consonant and in words with final stress before [l]. The phonetic durational properties of Spanish stressed vowels and summarized in Table 7.1.

Most loanwords in Choguita Rarámuri with long vowel sequences also have a post-tonic consonant cluster, derived through post-tonic vowel deletion. As in the native vocabulary, deletion deletion of the post-tonic vowel triggers lengthening of a preceding syllable's stressed vowel (for details about this process, see §9.4.3.2). Examples of this pattern in loanwords are shown in (17d–h). Examples of this pattern in native words is shown in (18) below.

(18) Long vowel sequences in native words and post-tonic consonant clusters

a. [raˈpáampo] /raˈpá-na-po/ tear.appl-tr-fut.pl 'We will tear it.' 'Vamos a romperlo.'

<sup>3</sup> Some of these loanwords also exhibit additional segmental adaptation, including final V epenthesis in (17c).

<sup>4</sup>A ritual dancer.

### 7 Other word-level supra-segmental processes

Table 7.1: Phonetic duration properties of Spanish stressed syllables (adapted from Chávez-Peón 2007)


b. [roˈmíinpo] /romí-na-po/ fold-tr.fut.pl 'We will fold it.' 'Vamos a doblarlo.'

c. [ramuˈwéeltʃ͡ane] /ramuwéli-tʃ͡ane/ play.with.in.laws-ev 'It sounds like they are playing with their in laws.' 'Se oye que están vacilando con sus cuñados.'

In sum, the following properties characterize loanword prosodic adaptation in Choguita Rarámuri:

	- b. There are exceptionless adaptation patterns that repair loanwords to comply with high ranked prosodic requirements (e.g., stress must be within a three-syllable stress window, prosodic words must be vowel final).

# **8 Nouns**

Nouns in Choguita Rarámuri can be broadly defined morphosyntactically as the class of words that may be case-marked. Case marking in this language includes instrumental, locative and oblique case markers (nominative and accusative case is encoded exclusively on pronominal forms, as discussed in Chapter 10). Some classes of nouns do not take all case markers available: kinship and body part terms, two closed classes of inalienable nouns, may only occur in possessive constructions and are not found with locative or instrumental case, but are instead obligatorily head-marked with a suffix encoding a possessive relation and indexing properties of the possessor. Possessive suffixes are also deployed with alienably possessed nouns, though this encoding is not obligatory.

This chapter is concerned with the morphological properties of Choguita Rarámuri nouns, their internal morphological structure, and the derivational and inflectional morphological constructions with which nouns may combine. This chapter begins with an overview of the morphotactic generalizations that hold across morphologically complex nouns §8.1; the following sections describe several nominal morphological constructions, including plural/pluractional marking (§8.2), case marking (§8.3), posessive marking (§8.4), and deverbal noun marking (§8.5). This chapter also addresses how Spanish noun loanwords are incorporated in the language (§8.6). The chapter concludes with a description of tone patterns in morphologically complex nouns (§8.7). The syntactic properties of noun phrases are addressed in Chapter 12.

### **8.1 Morphotactic generalizations**

Productive nominal inflectional morphology in Choguita Rarámuri is mostly restricted to case marking, though core grammatical roles such as subject and objects are not marked in nominal phrases. Pluractional/plural formation is a category that is analyzed as derivational for verbs (a matter of lexical choice) (see Chapter 9, §9.3.1.2), but may be analyzed as inflectional when used with nouns. Nominal stems may be morphologically derived (most frequently, through deverbal processes) or monomorphemic. The bulk of nominal morphology in the

### 8 Nouns

language is devoted to derivational processes, including several morphological constructions to derive nouns from verbal roots; such derivational processes are analyzed here as taking place within a nominal stem level which can then undergo inflectional affixation. Nominal morphology is, however, significantly less complex than verbal morphology, both in terms of agglutination possibilities as well as the sheer number of morphological processes available.

The order of inflectional affixes in nouns is illustrated in (1):

(1) Order of inflectional affixes in nouns


As described below (in §8.4), this morphological structure includes two suffix positions for possessive suffixes, which are ordered sequentially in certain semantic configurations.

Like verbs, the morphological structure of nouns conforms to the general affix ordering principle of derivation ordered within inflection (Bybee 1985), with derivational processes applying within the stem level and inflectional affixes occupying peripheral slots of the morphological structure. There are no documented permutations of affixes in morphologically complex nouns.

The examples in (2) show that the possessive *-lâ* suffix is ordered before the locative case suffix (2a–b) and that the possessive *-wa* suffix is ordered before the possessive *-lâ* suffix (2c). In (2a), the quality of the vowel in the possessive suffix is reduced post-tonically (see §5.1 in Chapter 5).

	- a. *niˈhê moˈʔôlitʃ͡i ˈtʃ͡ó ba neˈhê* 1sg.nom *moˈʔô-lâ-tʃi* head-poss-loc *ˈtʃ͡ó* also *ba* cl 'in my head (memory)' 'en mi cabeza (memoria)' < SFH in61(750)/in >
	- b. *taˈmí biˈlé sioˈrí iˈkîli roˈnôlatʃ͡i taˈmí* 1sg.acc *biˈlé* one *sioˈrí* fly *iˈkî-li* bite-pst *roˈnô-lâ-tʃi* foot-poss-loc 'A fly bit me in the foot.' 'Me picó un mosco en el pie.' < FMF 09 3:61/el >

8.1 Morphotactic generalizations

c. *ˈnè sikoˈríwala ˈnè* 1sg.subj *sikoˈrí-wa-lâ* pot-poss-poss 'my pot' 'mi olla' < BFL 06 4:187 >

Example (3a) shows the relative order between the nominalizer suffix *-li* (within the nominal stem) and the locative *-tʃ͡i* suffix; example (3b) shows the relative order between other derivational suffixes in the nominal stem and the instrumental *-ti* suffix:

### (3) Nominalizer-locative and instrumental suffix order

a. *ˈpé roʔaˈká koˈʔá ˈpé baˈtʃ͡ôkilitʃ͡i ba ˈpé* just *roʔa-ˈká* put.in-ger *koˈʔá* eat *ˈpé* just *baˈt͡ʃôka-li-tʃi* build.with.mud-nmlz-loc *ba* cl 'We eat putting on the bachókilichi.' 'Comemos poniéndolos en los bachókilichi.' {FLP in61:3:16.8} b. *roˈnêamti ronô-ê-ame-ti* foot-have-ptcp-inst 'with an instrument with legs' 'con un instrumento que tiene piernas' {BFL tx1:0:38.1}

Nominal morphological constructions may be either stress-shifting or stressneutral, i.e., triggering stress shifts or causing no stress alternations. As in the rest of the word classes, all lexical tonal contrasts are realized in the stressed syllable of the word. The morphologically-conditioned phonological properties of morphologically complex words are addressed at length in Chapter 5, Chapter 9 and Chapter 11. There are no distinct morpho-phonological properties of nouns in derived environments, so the characterizations below will thus only specify the phonological properties of both roots and suffixes. Morpho-phonological alternations and semantic properties of nominal morphological constructions are addressed in detail in this chapter. Other morphosyntactic and semantic properties of nominal forms are addressed in Chapter 12.

### 8 Nouns

### **8.2 Plural/pluractional marking**

Nouns that refer to animate referents in Choguita Rarámuri are marked plural through a morphological construction that also combines with verbs.<sup>1</sup> With verbs, this construction marks either a plural subject or that an action occurs or is being performed by the same agent several times, or by several agents several times (see §9.3.4). With nouns, the construction exclusively marks plural number. These meanings are related in that they refer to event plurality or 'pluractionality'. Pluractionals have been defined as encompassing meanings that range from iterative and frequentative to distributive and extensive action (Newman 1990, Newman 2012, Wood 2007) .

Choguita Rarámuri pluractional/plural forms are marked through a prefixed element also documented in closely related Rarámuri varieties and Guarijío (Lionnet 1968, Lionnet & Miller 1985, Miller 1996).<sup>2</sup> In Mountain Guarijío, the cognate process (labelled "plural subject, iterative or durative") is more clearly analyzed as reduplication, since the prefixed element is (C)V- (e.g. *saˈé*, *sa-saˈe* 'smell', *iˈsi, i-iˈsi* 'walk' (Miller 1996: 62)).

Plurals/pluractionals in Choguita Rarámuri are marked in one of three ways: (i) through a prefixed vowel (4a-e) (where the vowel quality of the prefix may be harmonized to the root's first syllable vowel, as in (4d) (see also Lionnet 2001b);<sup>3</sup> (ii) through consonant mutation (4f); or (iii) through both consonant mutation and vowel prefixation (4g–j).<sup>4</sup>

<sup>1</sup>This construction deploys the same formal mechanisms of exponence with both nouns and verbs. Plural/pluractional marking in both nouns and verbs is thus treated as at least diachronically related in this grammar. Whether pluractionality as a verbal category and plural number with nouns are indeed part of a single morphosyntactic construction or should be treated as separate morphosyntactic devices is a question left for further research.

<sup>2</sup>Lionnet (2001b) labels the cognate forms of these constructions as "intensive" in Norogachi Rarámuri.

<sup>3</sup>And where plural/pluractional marking involves further phonological changes (as in *iˈwé* 'girls' in (4b)).

<sup>4</sup>There is an example that shows an alternation between morphologically related nouns where the relation is not one of singular/plural, but rather of plural/collective: the deverbal noun *ˈhâwami*, 'authorities', has a pluractional counterpart with prefixation and consonant mutation, *i-ˈhâpiʃi*, which refers the totality of the individual named authorities. This pair is unusual in that it displays a correspondence between the bilabial voiceless stop *p* and the labio-velar glide *w*, a correspondence not found in any of the other plural/pluractional alternations. In addition, the form *ˈhâw-ami* is a nominalization through the participial suffix *-ame*, while the pluractional form has an ending that does not correspond to any productive suffix in the language. Thus, we might conclude that this pair has been lexicalized.

### 8.2 Plural/pluractional marking

### (4) Plural/pluractional marking with nouns<sup>5</sup>


As shown in these examples, consonant mutation involves a voicing toggle, since it produces voicing or lenition of a voiceless stop (4i, l), and devoicing or hardening of a voiced plosive and rhotic (4f–h, j–k, m) in paradigmatically related nominal forms. Plural forms, in addition, may also involve a change in stress placement with respect to the base (see, e.g., (4a, c, f–h)), although there is no phonological predictability as to the location of stress in the derived form.

<sup>5</sup>The form *ˈtôro* in (4l) is a loanword from Spanish and differs from the rest ofthe pluractionalmarked nouns in that the correspondence is between an alveolar voiceless stop in the base stem and an alveolar trill, not an alveolar flap, in the pluractional form. This form also has a long vowel. See §7.3.2 for discussion and examples of Spanish loanwords with long vowels.

### 8 Nouns

This is also attested in cases where the same lexical item will have more than one pluractional form: this is the case with the pluractional of *remaˈrí* 'young man', which is *ˈtémuri* for some speakers (4f), with first syllable stress, but *teˈmári* for others (4g), with second syllable stress. More details on the prosodic and other general phonological properties of plural/pluractional forms in Choguita Rarámuri are provided in §9.3.4.

### **8.3 Case marking**

Case is defined here as a property of noun phrases. Choguita Rarámuri nominals may bear instrumental (§8.3.1) or locative (§8.3.2) case. As mentioned above, core grammatical relations (subject and object) are not encoded morphologically with nouns, though the pronominal system evidences a nominative/accusative system, and accusative markers are attested in other Taracahitan languages, e.g., Yaqui (Lindenfeld 1973, Escalante 1990, Dedrick & Casad 1999).

### **8.3.1 Instrumental case**

Instrumental case is marked with the stress-neutral suffix *-ti*, 6 and is found exclusively marking nouns that bear the semantic role of instruments, i.e., there are no recorded examples where this particular case marking is extended to other semantic roles. An example of instrumental marking on an instrumental NP is provided in (5).

(5) a. *roˈnôti riˈkêka roˈnô-ti* foot-instr *riˈkê-ka* step-ger 'You step on it with the foot.' 'Con el pie se pisa.' {MDH co1136:04:00.2} b. *roˈnô* 'foot'

'pie' < BFL 06 5:127/el >

Further examples of instrumental case marking are given in (6).

<sup>6</sup>As discussed in §5.3, stress neutral suffixes are never stressed and do not trigger any stress changes to the stems they attach to.

8.3 Case marking

(6) Instrumental case

a. *siˈkâti siˈkâ-ti* hand-instr 'with the hand' 'con la mano' < BFL 06 5:128/el > b. *t͡ʃuˈmíti t͡ʃuˈmí-ti* snout-instr 'with the snout' 'con el pico' < SFH 06 6:84/el > c. *koˈbísiti koˈbísi-ti* pinole-instr 'with the pinole' 'con el pinole' < BFL 06 5:128/el > d. *kiˈmáti kiˈmá-ti* kiˈmá-instr 'with the blanket' 'con la cobija' < BFL 06 5:128/el > e. *kuˈʃìti kuˈsì-ti* stick-instr 'with the stick' 'con el palo' {BFL tx1:0:24.9} f. *t͡ʃiˈníti t͡ʃiˈní-ti* cloth-instr 'with the cloth' 'con la tela' < SFH 06 6:84/el > g. *ˈwàsti ˈwàsi-ti* cows-instr 'with the cows'

'con las vacas' {ME in484:1:14.8}

### 8 Nouns


As the examples in (6) show, this construction does not involve any allomorphy or trigger any phonological effects in the bases with which it combines. In-

<sup>7</sup>This example involves a loanword from Spanish. As described in §8.6 below, relatively recent loan nouns from Spanish are adapted with the -tʃ͡i suffix, which is homophonous with the locative suffix.

8.3 Case marking

strumental case marking is very productive,<sup>8</sup> and there seem to be no semantic restrictions of the bases that can combine with this construction (i.e., instrumental case marking is not exclusive to a closed class of nouns).

With inalienably possessed nouns, the base for affixation of the instrumental is the non-possessed, bare root (e.g., (6)). The instrumental suffix may also be added to morphologically complex bases: in (7), for instance, an instrumental suffix is added to a nominal base with the participial *-ame* suffix (the participal suffix is analyzed here as being part of the nominal stem level).

(7) *roˈnêamti ronô-ê-ame-ti* foot-have-ptcp-inst 'with an instrument with legs' 'con un instrumento con piernas' {BFL tx1:0:38.1}

Though instrumental markers commonly display syncretism with comitative markers cross-linguistically (Stassen 2000, Stolz 2001), Choguita Rarámuri comitatives are generally expressed through a different construction, a postpositional phrase, exemplified in (8):


The suffix *-ti* is thus a dedicated instrumental marker in the language.

### **8.3.2 Locative case**

There are two productive locative case markers in Choguita Rarámuri: the stressshifting suffix *-tʃ͡í* and the stress-neutral suffix *-rare*. <sup>9</sup> The locative suffix *-tʃ͡í* means 'on, at' (and may be classified as adessive), while the suffix *-rare* means 'at', and 'among, between' (with an inessive reading). The locative *-tʃ͡i* suffix is

<sup>8</sup>Many Uto-Aztecan languages have a set of instrumental prefixes that, attached to verbs, indicate the instrument with which an activity is carried out. While still synchronically active in the Tepiman and Numic branches of Uto-Aztecan (Dayley 1989), instrumental prefixes only have lexicalized remnants in Choguita Rarámuri (for a discussion of instrumental prefixes in Choguita Rarámuri, see §9.3.2).

<sup>9</sup>Post-tonic vowel reduction yields the surface form *-riri*, as described in §5.2.

### 8 Nouns

exemplified in (9) and the locative *-rare* suffix in (10) (the latter suffix has the surface form *-riri* due to regular rules of post-tonic vowel reduction (see Chapter 3, §5.2).

```
(9) Locative (adessive) -tʃ͡í
      a. wasaˈtʃ͡í
           wasa-ˈtʃ͡í
           land-loc
           'on the land'
           'en la tierra' {SFH tx43:7:44.3}
      b. kawiˈtʃ͡í
           kawi-ˈtʃ͡í
           hill-loc
           'on the hill'
           'en el cerro' {SFH tx977:1:26.6}
      c. koˈbísitʃ͡i
           koˈbísi-tʃ͡i
           corn.meal-loc
           'on the corn meal'
           'en el pinole' < BFL 06 5:128/el >
      d. sipuˈtʃ͡âtʃ͡i
           sipuˈtʃ͡â-tʃ͡i
           skirt-loc
           'on the skirt'
           'en la falda' < BFL 06 5:128/el >
      e. ˈpúratʃ͡i
           ˈpúra-tʃ͡i
           knitted.belt-loc
           'on the knitted belt'
           'en la faja' < BFL 06 5:128/el >
      f. iˈwákatʃ͡i
           iˈwáka-tʃ͡i
           hole-loc
           'in the hole'
           'en el hoyo' {LEL tx177:4:26.3}
```
8.3 Case marking

g. *tiˈjôpitʃ͡i tiˈjôpa-tʃ͡i* church-loc 'at the church' 'en la iglesia' {JLG el1275:0:11.8} h. *baˈtʃ͡ôkilitʃ͡i baˈtʃ͡ôki-li-tʃ͡i* fix.with.mud-nmlz-loc 'on the mud' 'en el barro' {FLP in61:3:16.8} i. *sikoˈrítʃ͡i sekoˈrí-tʃ͡i* olla-loc 'in the pot' 'en la olla'{LEL tx130:5:26.5} j. *saʔpaˈtʃ͡í saʔˈpa-tʃ͡í* meat-loc 'on the meat' 'en la carne' < SFH in61(490)/in > k. *osiˈrítʃ͡i osiˈrí-tʃ͡i* paper-loc 'on the paper' 'en el papel' < FLP in61(739)/in > l. *beˈtôlitʃ͡i beˈtôli-tʃ͡i* plate-loc 'on the plate' 'en el cajete' {LEL tx130:5:41.2} m. *baˈrîkatʃ͡i baˈrîka-tʃ͡i* bucket-loc 'on the bucket' 'en la barrica' {JLG co1234:6:23.3}

### 8 Nouns


<sup>10</sup>In this form, the long stressed vowel is due to compensatory lengthening triggered by posttonic vowel deletion. More details about compensatory lengthening are provided in §9.4.3.2.

<sup>11</sup>This form also undergoes compensatory lengthening of the stressed vowel triggered by posttonic vowel deletion.

8.3 Case marking

c. *kiˈmáriri kiˈmá-riri* blanket-loc 'in the blanket' 'en la cobija' < LEL 06 5:127-9/el > d. *wiˈtʃ͡îriri wiˈtʃ͡î-riri* skin-loc 'in the skin' 'en la piel' < BFL 06 5:128/el > e. *kuˈsìriri kuˈsì-riri* sticks-loc 'in the sticks' 'en los palos' {BFL tx1:0:31.2} f. *reˈʰtêriri reˈʰtê-riri* stone-loc 'in the stones' 'en la piedra' {LEL tx5:2:06.9} g. *kaˈwíriri kaˈwí-riri* hill-loc 'in the hill' 'en el cerro' {LEL tx109:1:17.4} h. *roˈhàriri roˈhà-riri* oak-loc 'in the oak tree' 'en el encino' {SFH tx152:5:57.2} i. *oˈʰkóriri oˈʰkó-riri* pino-loc 'in the pine tree' 'en el pino' {FLP in61:4:07.4}

### 8 Nouns


Miller (1996) describes the cognates of these suffixes in Mountain Guarijío as possessing a similar semantic difference: the Mountain Guarijío locative *-tʃ͡i/ átʃ͡i,* cognate of the Choguita Rarámuri *-t͡ʃi* suffix, is glossed as 'where' and 'at' (1996: 192–193), while the Mountain Guarijío *-tére*/*-rere* suffix, with the short allomorph *-te/-re*, the cognate of the Choguita Rarámuri *-rare* suffix, is glossed as 'below/among' (Spanish *abajo* and *entre*) (1996: 287). In Choguita Rarámuri, however, the choice of suffix in many cases seems to be lexically conditioned. I have recorded alternative forms of the same bases with the two locative suffixes, and speakers have produced them both spontaneously and when prompted during elicitation. Some examples are provided in (11):



8.3 Case marking

In the following pair, however, the choice of locative suffix reflects a semantic distinction between the two suffixes an alienable/inalienable distinction: the noun *saʔˈpá* 'meat' means live flesh in (12a) with the locative *-tʃ͡i* suffix and meat severed from the body in (12b) with the locative *-rare* suffix; that is, the locative *-t͡ʃi* suffix is used when the figure is located on the surface of the ground, while *-rare* is used when the figure is located inside, or within a space defined by the ground, consistent with the description of these suffixes as adessive and inessive, respectively:

(12) Contrast between adessive and inessive locative forms

a. *taˈmí naˈmûti iˈkîli saʔpaˈtʃ͡í taˈmí naˈmûti iˈkî-li saʔpá-tʃ͡í*

> 1sg.acc something bite-pst flesh-loc

'Something bit me in the flesh.' (where *sapa-ˈt͡ʃí* = in the live, human flesh)

'Algo me picó en la carne.' (donde *sapa-ˈt͡ʃí = en la carne del cuerpo*) < FMF 09 3:60/el >

b. *biˈlé sioˈrí t͡ʃuˈkú saʔˈpáriri*

*biˈlé sioˈrí t͡ʃuˈkú saʔˈpá-rare*

one fly stand.four.legs.prs meat-loc

'There is a fly in the meat.' (where *saʔpá-riri* = in the severed cow's meat)

'Hay una mosca en la carne.' (donde *saʔpá-riri* = en el pedazo de carne de vaca) < FMF 09 3:60/el >

In this case, the different locative suffixes seem to have been lexicalized with specific meanings. There are no other documented examples where other bases display a comparable semantic contrast.<sup>12</sup>

In sum, there are some semantic differences between locative *-tʃ͡i* and locative *-rare,* and each construction might be selected depending on the semantic characteristics of each base in some cases ('to be among the trees' vs. 'to be on a rock'), but affix distribution seems to be largely a matter of lexical choice in most cases

<sup>12</sup>A similar difference where the same root has an inalienable meaning with one morphological construction and an alienable meaning with another morphological construction has been reported in other Uto-Aztecan languages: in Cora (Corachol), there is a contrast between *n y iwe'* 'my flesh', and *n y i-weʔ-raʔ*, with possessive suffix *-raʔ,* 'my meat (that I bought)' (Dakin 1991: 302). See §8.4 for details of the behavior of alienable and inalienable nouns in possessive constructions in Choguita Rarámuri.

### 8 Nouns

(i.e., there is no clear semantic property of the referent that requires one locative construction versus the other).

In addition to the basic locative meaning exemplified above, the suffix *-rare* appears on a base where the locative construction seems to involve a collective meaning: in (13), the nominal base *oˈʰkó* 'pine tree' plus the suffix *-rare* encodes a group of pine trees:<sup>13</sup>

(13) *oˈʰkóriri ke mi koˈrína oʔˈná*

*oˈʰkó-rare* pine.tree-loc *ke* cop.impf *mi* dist *koˈrí-na* around-all *oʔˈná* there 'There were many pine trees over there.'

'Había muchos pinos allá.' {FLP in61:4:07.4}

Consider, in contrast, two examples where the same root *oˈʰkó* 'pine tree' plus the locative *-rare* refers to a single tree; the difference between this form and the 'collective' one in (13) is that *oˈʰkó* 'pine tree' is the deictic centre in (14) below, but the figure in (13) above:

### (14) Collective reading of *-rare*


In example (15) below, *oˈʰkó-riri* 'pine-loc' might be interpreted as either meaning that the pine trees are the figure in a deictic center and that the speaker is

<sup>13</sup>This construction is an example of a copular clause headed by a nominal predicate. More information about this type of clause is provided in §13.2.2.

8.3 Case marking

located at a different place ('behind the trees'), or that the speaker is in fact the figure located 'among, between' the trees, the deictic centre.

(15) *aʔˈlì ko ˈnè ˈmá biˈlá ˈnápa ˈnìi oˈʰkóriri ˈnè wiˈlí ˈnà iˈt͡ʃîpika ba neˈká ba aʔˈlì=ko* and=emph *ˈnè* 1sg.nom *ˈmá* already *biˈlá* really *ˈnápa* sub *ˈnì* emph *oˈʰkóriri* pine-loc *ˈnè* 1sg.nom *wiˈlí* stand.sg *ˈnà* prox *iˈt͡ʃîpika* hide-ger *ba* cl *neˈká* see-ger *ba* cl 'So then I stood behind/between the trees, hiding, watching.' 'Entonces ya me paré entre/atrás de los pinos, escondida, viendo.' {LEL tx223:1:29.3}

Locative case also encodes spatial relations relative to specific referents. In these cases, exemplified in (16), locative case is marked on the head noun of the nominal phrase:

(16) a. *naˈsîpa iˈjétʃ͡i naˈsîpa* middle *iˈjé-tʃ͡i* door-loc 'in the middle of the door' 'en medio de la puerta' < FMF 09 3:27/el > b. *suˈwè t͡ʃiˈnírari suˈwé* edge *t͡ʃiˈní-rari* cloth-loc 'on the edge of the cloth' 'por la orilla de la tela' < FMF 09 3:27/el >

Finally, locatives are found in many Choguita Rarámuri toponyms. Place names are derived from combining locative suffixes with nominal roots. Some examples of toponyms derived with these suffixes are provided in (17) (all examples are also documented in Brambila (1976), and the source information from this reference is indicated under each example with the initial "B", year (1976) and page where the data was found).

### 8 Nouns


### **8.4 Possessive marking**

Possessive marking in Choguita Rarámuri nouns encodes possessors and meronymic (or part-whole) relationships. Other semantically related relations, such as the relation 'made of', are encoded through analytic constructions in the language. The function and syntactic properties of possessive constructions are addressed in §12.2.

8.4 Possessive marking

### **8.4.1 Alienable and inalienable possession**

Both inalienable (kinship and body-part terms) and alienable nouns may register a possessor with the stress-shifting possessive *-lâ* suffix.<sup>14</sup>

Following the head-marking pattern of the language, the suffix *-lâ* attaches to the head noun in a possessive nominal phrase, as shown in (18):

```
(18) Possessive marking
      ˈnè waˈsála
      ˈnè
      1sg.nom
               wasá-lâ
               cultivation.land-poss
      'my cultivation land'
      'mi tierra de cultivo' < BFL 09 1:61/el >
```
Further examples of possessive constructions are provided in (19) (as other stress-shifting suffixes, the lexical tone of the suffix surfaces when stressed, (19f)):

(19) Possessive marking

```
a. waˈsîla
    waˈsî-la
    mother.in.law-poss
    'their mother in law'
    'su suegra' < BFL 06 5:127/el >
b. ˈkútʃ͡ala
```

```
ˈkútʃ͡a-la
children-poss
'their little ones, children'
'sus hijos' {SFH tx43:12:39.8}
```
c. *iˈjêla*

```
iˈjê-la
mother-poss
'their mother'
'su mamá' {SFH tx475:7:11.4}
```
<sup>14</sup>A suffix *\*\*-ra*, reconstructed for Proto-Sonoran and possibly for Proto-Uto-Aztecan (Dakin 1991: 299), has cognate forms in Tubar (Lionnet 1978), Lowland (River) Guarijío (Barreras 1988), Highland (Mountain) Guarijío (Miller 1996), Cora (Dakin 1991), and other Tarahumara varieties (Norogachi Rarámuri (Brambila 1953, Lionnet 1972)). Lionnet (1972) describes the suffix *-ra* as a 'determiner' (*"determinado"*), though no definitions of this suffix is provided (1972:18).

### 8 Nouns

d. *aliˈwâla aliˈwâ-la* soul-poss 'their soul' 'su alma' {LEL tx5:0:30.7} e. *apaˈnêrala apaˈnêra-la* wife-poss 'their wife' 'su esposa' {LEL tx5:0:40.9} f. *onoˈlâ ono-ˈlâ* father.male.ego-poss 'their father (male ego)' 'su papá de él' {SFH tx12:2:18.0} g. *kuˈrít͡ʃala kuˈrít͡ʃa-la* uncle-poss 'their uncle (husband of father's sister)' 'su tío (esposo de la hermana del papá)' < FLP in61(510)/in > h. *kuˈmût͡ʃala kuˈmût͡ʃa-la* paternal.older.uncle-poss 'their paternal older uncle' 'su tío paterno mayor (que el papá)' {SFH in485:0:00.2} i. *moʔoˈlîla moʔoˈlî-la* daughter.in.law-poss 'their daughter in law' 'su nuera' {LEL tx19:5:29.8} j. *aʔkaˈlâ aʔka-ˈlâ* sandal-poss 'their sandals' 'sus huaraches' < BFL 09 1:60/el > k. *boʔeˈlâ boʔe-ˈlâ* road-poss 'their road' 'su camino' < BFL 06 5:128/el >

8.4 Possessive marking

l. *siˈpútʃ͡ala siˈpútʃ͡a-la* skirt-poss 'their skirt' 'su falda' < BFL 09 1:60/el > m. *naˈpátʃ͡ala naˈpátʃ͡a-la*

shirt-poss 'their shirt' 'su blusa' < FMF 09 3:32/el >


```
kaˈlí-la
     house-poss
     'their house'
     'su casa' < SFH in61(695)/in >
p. ˈtéla
     ˈté-la
     lice-poss
     'their lice'
     'sus piojos' < FMF 09 3:32/el >
```
are exclusively found with possessive marking.

The examples above show that the suffix *-lâ* attaches to inalienable nouns (19a–i) as well as alienable nouns (19j–p). Alienable nouns are characterized by being able to appear in non-possessive constructions, while inalienable nouns

Choguita Rarámuri has a second possessive construction in which nominal roots are marked by a stress-neutral suffix *-wa* followed by *-lâ* (schematically, V*-wa-lâ*). Haugen (2017) reconstructs *\*-wa* for Proto-Uto-Aztecan as a nominal suffix of attributive possession ("possessed thing") (2017: 172) (see also Langacker 1977 for an alternative proposal).<sup>15</sup> Reflexes of this reconstructed suffix are found

<sup>15</sup>Haugen (2017) makes a distinction between possession in the nominal domain and possession in the verbal domain, a distinction largely equivalent to one of attributive possession and predicative posession, respectively. Both types of posession constructions are documented across the Uto-Aztecan language family. Langacker (1977) reconstructs *\*-wa* as a denominal verb suffix.

### 8 Nouns

in the Numic, Takic, Tepiman, Taracahitan, and Aztecan branches of Uto-Aztecan (Haugen 2017: 186).

Examples of these "double possessive" constructions are given in (20):<sup>16</sup>

(20) Double possessive constructions


Miller (1996) describes the cognate forms of both the *-wa* and *-lâ* suffixes in Mountain Guarijío as 'possessive' and 'absolutive' (non-possessed), respectively (1996: 250–251). The Mountain Guarijío possessive *-wa* suffix is described as being stress-shifting (causing a change of stress in the base), and the Mountain Guarijío absolutive *-la* suffix as stress-neutral, in contrast to the prosodic properties of the cognate forms in Choguita Rarámuri. While the suffix *-wa* is used exclusively in possessive constructions in Mountain Guarijío, the suffix *-la* is also used in this language in partitive and what Miller labels 'determinative' constructions. While in Mountain Guarijío *-wa* may be used as the sole marker of possessors, in Choguita Rarámuri this formative is always followed by *-la* as in the examples shown above.

For some speakers of Choguita Rarámuri the double possessive construction alternates with the *-lâ* possessive construction to mark a plural-singular distinction, respectively, with count nouns. This is exemplified in (21) (all data from < BFL 11/07/09/el >):<sup>17</sup>

<sup>16</sup>The meaning of the possessed forms is compositional, except for the form *waˈsá-wa-la* in (20b) which can be interpreted as 'someone's hill', but is more generally interpreted as 'someone's homeland'. In the case of *baˈt͡ʃâ-wa-la* in (20h), the meaning is 'someone's ancestors' (lit. 'their first ones').

<sup>17</sup>In these examples, the pronominal forms exhibit variation between the full and reduced forms with no apparent grammatical conditioning, e.g., *neˈhê iˈjê-la* {SFH tx475:06:17.3} vs. *ˈnè iˈjêla* {BFL tx1:00:55.5} 'my mother'. This kind of variation is, to the best of my knowledge, not reported in other Rarámuri varieties.

### 8.4 Possessive marking

### (21) A singular-plural distinction of possessed nouns


It is not clear, however, that this singular-plural distinction is found with all count nouns. In (22), the possessive construction is ambiguous between a singular and a plural possessum. The base in (22a) lacks a possessive form with possessive *-la*. In (22b–c), on the other hand, the *-wa-lâ,* 'double possessive', construction in Choguita Rarámuri stands in variation with single-marked possessive nouns in certain forms. In this case, there are no semantic differences between the two alternative constructions.

(22) Optionality between single and double possessive constructions

```
a. ˈnè riˈmêwala
    ˈnè
    1sg.nom
              reˈmê-wa-lâ
              tortillas-poss-poss
    'my tortillas'
    'mis tortillas' < BFL 11/07/09/el >
b. ˈnè waˈsála
    ˈnè
    1sg.nom
              waˈsá-lâ
              sowing.field-poss
    'my sowing fields'
    'mis tierras de cultivo' < LEL 06 5:127-9/el, BFL 09 1:60/el >
c. ˈnè waˈsáwala
    ˈnè
    1sg.nom
              waˈsá-wa-lâ
              sowing.field-poss-poss
    'my sowing fields'
    'mis tierras de cultivo' < LEL 06 5:127-9/el, BFL 09 1:60/el >
```
For some nouns the only available possessive construction is the double possessive *-wa-lâ*. This is exemplified in (23): these nouns, spontaneously given with

### 8 Nouns

the *-wa-lâ* 'double possessive' sequence, were rejected when prompted with the *-lâ* possessive construction (< BFL 09 1:61/el >).

(23) Nouns with obligatory double possessive constructions


For other nouns, the only available possessive construction is the one involving a single possessive marker (*-lâ*), with the 'double possessive' construction *-wa-lâ* being unavailable. All documented cases of nouns lacking a 'double possessive' construction are body-part terms, a closed class of inalienable nouns. Relevant examples are shown in (24),<sup>18</sup> with their spontaneously produced possessive form, as well as the ungrammatical double possessive construction.

(24) Possessive forms of body-part nouns


The exception to this pattern is found with body part nouns that may have alternative readings as either a collection of entities with similar properties (e.g., teeth) or as discrete entities (e.g., a single tooth). For these nouns, the *-lâ* construction refers to the possession of a single, discrete entity (e.g., a single finger, tooth, finger nail, etc.), and the double possessive construction is used when making reference to the possession of multiple fingers, teeth, etc.<sup>19</sup> These forms are exemplified in (25) (all data from < BFL 11/07/09/el >):

<sup>18</sup>I represent body part roots as bound, since they require possessive marking as inalienable nouns.

<sup>19</sup>Body part terms referring to body parts that come in pairs, such as *buˈsí-* 'eye', *kaˈsî-* 'leg', *roˈnô-* 'foot', *tʃ͡oˈkówa-* 'knee', or *siˈkâ-* 'hand', do not have a singular-plural (dual) possessive form alternation as found in (25). Similarly, the term for 'ribs', *waˈt͡ʃíka-*, while composed of multiple, discrete parts, does not have a singular-plural alternation in the possessive (though this might be due to the fact that ribs are all linked or that they are internal (K. Dakin, p.c.)).

### 8.4 Possessive marking

(25) Singular vs. plural distinction in possessed body-part terms

*Form Gloss* a. ˈnè maˈkúsa-la 'my finger' b. ˈnè maˈkúʃu-a-la 'my fingers' c. ˈnè maˈtáka-la 'my molar' d. ˈnè maˈták-wa-la 'my molars' e. ˈnè ʃuˈtû-la 'my finger nail' f. ˈnè ʃuˈtû-wa-la 'my finger nails'

Some kinship terms are also found with the single vs. double possessive construction encoding a singular-plural distinction. These are exemplified in (26):

(26) Singular vs. plural distinction in posessed kinship terms


'mi tío paterno'

d. *neˈhê kuˈmût͡ʃu-wa-la neˈhê* 1sg.nom *kuˈmût͡ʃi-wa-la* paternal.uncle-poss-poss 'my paternal uncles' 'mis tíos paternos'

Other kinship terms, however, do not display this alternation, e.g., *uˈmû-a-la* ambiguously means 'his/her great grandfather' or 'his/her great grandfathers' <

### 8 Nouns

FLP in61(116)/in, BFL 11/07/09/el >.<sup>20</sup> Thus, the distinction between singular vs. plural possessum associated to a single possessive marker vs. two, respectively, seems to be a matter of lexical choice.<sup>21</sup>

Similar to the pattern attested in Choguita Rarámuri, in Lowland (River) Guarijío, the cognate of the *-la* suffix (also *-la*) is found in possessive and non-possessive constructions, and the cognate of the *-wa* suffix is used in addition to *-la* to mark optional possession ("mediated possession") with nouns that are not body part terms, kinship terms or abstract terms (Barreras 1988, cited in Dakin 1991: 301. For Mountain Guarijío, Miller (1996) describes that possessed nouns marked with the *-wa-la* suffix sequence are generally interpreted as plural, though he notes singular readings are also available with this marking pattern (but no examples are provided).

There are inalienable nouns in possessive constructions that are marked with a *-tʃ͡í* suffix, homophonous to the locative *-tʃ͡í* suffix described in §8.3.2, which are added in non-locative contexts. The body-part terms exemplified below are either directly marked with suffix *-t͡ʃí* (27a) or attach this suffix to a base with suffix *-lâ* (27b–c):

(27) Posessive constructions with suffix *-tʃ͡i*

a. *ˈnè t͡ʃa t͡ʃoˈkóbitʃ͡i oʔˈkô ˈnè* 1sg.nom *t͡ʃa* ugly *t͡ʃoˈkówa-tʃ͡i* knees-loc *oʔˈkô* hurt 'My knees hurt.' 'Me duelen mis rodillas.' < BFL 09 1:60/el > b. *niˈhê kaˈwìwalatʃ͡i ˈnèˈhê* 1sg.nom *kaˈwì-wa-la-tʃ͡i* hill-poss-poss-loc 'my hill' 'mi tierra' < BFL 09 1:60/el >

<sup>20</sup>Another kinship term, *paˈpâ-wa-la* 'parents', in (20i), is a borrowed term (from Spanish *papá* 'father') that refers to the two parents.

<sup>21</sup>There is at least one example where a singular/plural distinction in terms of possessors and/or possessed referents is encoded through two possessive constructions involving a concatenation of two distinct formatives: the noun *muˈkî*, 'woman', can be marked as possessed with a *-a-lâ* suffix sequence (*muˈkî-a-la*) or with a *-wa-lâ* suffix sequence (*muˈkî-wa-la*). The two forms were given with different interpretations, with either a singular possessor and possessum in the former case (*muˈkî-a-la* 'his wife') and a plural possessor and possessum in the latter (*muˈkîwa-la* 'their wives') (<BFL 09 4:69/el>). This case also suggests a lexicalization process, given that these possessed forms are exclusively used when the interpretation of the noun *muˈkî* is 'wife'.

8.4 Possessive marking

c. *ˈnè t͡ʃa t͡ʃaˈmékalatʃ͡i*<sup>22</sup> *oʔˈkô ˈnè* 1sg.nom *t͡ʃa* ugly *t͡ʃaˈméka-la-tʃ͡i* tongue-poss-loc *oʔˈkô* hurt 'My tongue hurts.' 'Me duele mi lengua.' < BFL 09 1:60/el >

There are only a few examples with the *-tʃ͡i* suffix with the possessive meaning, suggesting this construction is unproductive. Some of the nouns modified with this suffix in these contexts refer to space or body-parts as possible locations (e.g., for pain), suggesting that these constructions involve a frozen and reanalyzed locative (as in French noun phrases with *chez* and *à*), as suggested by Johanna Nichols (p.c.).

The possessive construction induces truncation of the base noun in at least one construction. In (28), the root *koroˈká*, 'necklace', is shortened to *koˈro* when marked with the possessive:<sup>23</sup>

```
(28) Truncation in possessive construction
      ˈnè koˈróala
      ˈnè
      1sg.nom
               koroˈká-wa-la
               necklace-poss-poss
      'my necklace'
      'mi collar' < BFL 09 1:61/el >
```
Finally, in addition to the head-marking strategy for marking possession, Choguita Rarámuri has an alternative way of expressing a possession relationship through an appositional construction with *ˈníwa*, which may be used predicatively as a verb meaning 'to have', but which behaves as a grammatically specialized possessive noun in possessive constructions (glossed below as 'have'). More details of this syntactic construction are provided in §12.2.3.

### **8.4.2 Meronymic (part-whole) relationships**

Meronymic or part-whole relationships are also expressed morphologically with the possessive *-la* suffix marked on the possessum. The forms in (29) exemplify the meronymic construction, where the possessor precedes the possessed noun:

<sup>22</sup>The prompted form *\*t͡ʃaˈmeka-wa-la* was rejected, but *t͡ʃaˈmék-wa-la,* with root-final vowel deletion, was accepted as grammatical (< BFL 09 1:61/el >).

<sup>23</sup>This resembles a noun truncation process in body part noun incorporation constructions, described in detail in §9.3.3.

### 8 Nouns

(29) Meronymic relations


Further examples of the meronymic construction are given in (30).

### (30) Meronymic possessive marking


As examples (30b) above and (31) below show, the meronymic construction is found with body-part terms used to refer to constituent parts of inanimate nouns:

(31) Meronymic construction with inanimate noun *ˈmêsa roˈnôla ko kaˈpòlo*

8.5 Deverbal nouns

*ˈmêsa* table *roˈnô-la=ko* leg-poss=emph *kaˈpò-li* break-pst 'The table's leg broke.' 'La pata de la mesa se rompió.' < BFL 09 1:33/el >

In (32a–b), both nouns are marked with the possessive suffix, since the whole is a body part term, in turn the part of another whole.

(32) Recursive possessive marking


There are no documented examples where this construction would encode other semantic roles (such as beneficiaries). Also, there are no documented meronymic constructions with the *-wa-lâ* sequence found in forms marking a possessor.

### **8.5 Deverbal nouns**

Choguita Rarámuri possesses several morphological means for deriving nominal stems from verbal and other roots. This section is devoted to describing in detail each of these devices.

### 8 Nouns

### **8.5.1 Agentive, patientive and experiencer nominalizations**

Agentive nominalizations, achieved through the affixation of the participial suffix *-ame* to a transitive base or an intransitive base with an unergative argument, are nominalizations that derive a noun with a meaning 'the one who performs V'. The examples in (33) illustrate agentive derivations, both from bare roots (33a– b) and morphologically complex bases (33c).

### (33) Agentive nominalizations


Verbs that have been nominalized with the participial *-ame* suffix do not take any TAM markers before the participial suffix, a fact that has also been observed for Mountain Guarijío (Miller 1996: 180). There are, however, agentive nominalization constructions with a meaning that involves past tense, i.e. 'the one who has performed V (but no longer does)'. These past agentive nominalizations attach the suffix *-kame*, the cognate of which Miller (1996) analyzes as composed diachronically of a past tense *-ka/-ga* suffix and the participial *-ame* suffix (note that the Mountain Guarijío *-ka/-ga* past tense suffix has no attested cognate in Choguita Rarámuri as a synchronically active past tense marker). Relevant examples are shown in (34):

<sup>24</sup>The example in (33c) also shows how the participial suffix may induce deletion of the previous suffix vowel. There are no recorded examples where the participial suffix deletes vowels of immediately preceding roots (see Chapter 9 for other inflectional suffixes that induce vowel deletion in their bases).

8.5 Deverbal nouns

### (34) Past agentive nominalizations

a. *biˈnèkami beˈnè-kame* learn-pst.ptcp 'one who has studied' 'el estudiado' < BFL 09 1:43/el > b. *ˈnè ˈpé ˈsêstobi kú maˈt͡ʃínikami ka ˈt͡ʃó ba ˈnè* 1sg.nom *ˈpé* just *ˈsêsto=bi* sixth=just *ku* rev *maˈt͡ʃína-kame* go.out-pst.ptcp *ka* cop.irr *ˈt͡ʃó* also *ba* cl 'I also just finished sixth grade' 'yo nomás soy salido de sexto también' {SFH tx12:11:17.2} c. *miˈʔàkami miˈʔà-kame* kill-pst.ptcp 'someone who used to murder, former killer' 'alguien que mataba, antiguo asesino' < BFL 06 el41/el >

Participial suffixes are also involved in deriving adjectives and forming relative clauses, and this is discussed in detail in §10.3 and §15.3, respectively. As discussed in §10.3, the participial suffixes *-ame* and *-kame* do not exhibit tense distinctions and are treated there as suppletive allomorphs.

Nominalizations involving other tense/aspect contrasts (e.g., future, as in 'the one who will become a ritual singer') are expressed through copular predicate constructions. In examples (35a–b), tense distinctions are marked in the copula. In the past tense nominalization, however, there is no copula available; the past tense of the nominalization is encoded by the *-kame* suffix (35c). The nominalization in (35a) encodes imperfective aspect.

(35) Tense/aspect distinctions with participial forms

a. *wikaˈrâame ˈnílo wikaˈrâ-ame* sing-ptcp *ˈní-li-o* cop-pst-ep 'He used to be a ritual singer.' 'Era cantador.' < BFL 06 4:168-171/el > b. *wikaˈrâame ˈníma riˈkó wikaˈrâ-ame* sing-ptcp *ˈní-ma* cop-fut.sg *riˈkó* dub 'He will be a ritual singer '

'Va a ser cantador.' < BFL 06 4:168-171/el >

8 Nouns


Further examples of agentive nominalizations are given in (36):

(36) a. *siˈpâame siˈpâ-ame* raspar.peyote-ptcp 'peyote shaman' 'raspador' < SFH 06 11.06/el > b. *aˈwíami aˈwí-ame* dance-ptcp 'dancer' 'bailador' < SFH 06 11.06/el > c. *kaˈréami kaˈré-ame* lie-ptcp 'liar' 'mentiroso' < SFH 06 11.06/el > d. *ˈwáami ˈwá-ame* be.strong-ptcp 'strong' 'fuerte' < SFH 06 11.06/el > e. *ˈhóami ˈhó-ame* make.hole-ptcp 'the one who makes holes' 'el que hace hoyos' < SFH 06 11.06/el >

8.5 Deverbal nouns

f. *raraˈhîpami raraˈhîp-ame* ball.race-ptcp 'ball race runner' 'corredor de bola (rarajipa)' < SFH 06 11.06/el >

Patientive nominalizations, on the other hand, are formed in Choguita Rarámuri through the participial suffix *-ame* which attaches to a verbal passive base (37a–e) or to an intransitive verb with a theme as subject argument (37f):<sup>25</sup>

(37) Patientive nominalizations

a. *miˈʔàaruami miˈʔà-ru-ame* kill.sg-pst.pass-ptcp 'one who gets killed' 'al que matan' < BFL 09 1:43/el > b. *aˈwèrami aˈwí-è-ru-ame* dance-appl-pst.pass-ptcp 'one who gets danced for' 'al que le bailan' < BFL 09 1:43/el > c. *saˈwèrami saˈwí-è-ru-ame* cure-appl-pst.pass-ptcp 'the healed one' 'el curado' < BFL 06 el41(15)/el > d. *miˈʔàrami miˈʔà-ru-ame* kill.sg-pst.pass-ptcp 'murdered one' 'el matado, asesinado' < BFL 06 el41/el > e. *baˈt͡ʃàrami baˈt͡ʃà-ru-ame* put.inside.sg-pst.pass-ptcp 'imprisoned one' 'encarcelado' < SFH 06 11.06/el >

<sup>25</sup>In these cases, the participial suffix undergoes optional post-tonic vowel reduction in word final position (with raising of [e] to [i]).

### 8 Nouns

f. *baˈkíami baˈkí-ame* go.in.sg-ptcp 'thing that has been put inside somewhere' 'metido' < SFH 06 11.06/el >

Further examples of patientive nominalizations are provided in (38):

	- b. *oˈhòrami oˈhò-r-ame* dekernel-pst.pass-ptcp 'dekerneled' 'desgranado' < SFH 06 11.06/el >
	- c. *ˈèrami*
		- *ˈè-r-ame* take-pst.pass-ptcp 'the thing taken away' 'lo llevado' < SFH 06 11.06/el >
	- d. *paˈkórami paˈkó-r-ame* wash-pst.pass-ptcp 'baptized one' 'bautizado' < SFH 06 11.06/el >
	- e. *riˈwètami riˈwè-t-ame* leave-pst.pass-ptcp 'abandoned one' 'abandonado' < SFH 06 11.06/el >
	- f. *biniˈhîrami biniˈhî-r-ame* denounce-pst.pass-ptcp 'denounced one' 'denunciado' < SFH 06 11.06/el >

g. *saˈwèrami saˈw-è-r-ame* cure-appl-pst.pass-ptcp 'cured one' 'curado' < BFL 06 4:168/el >

h. *t͡ʃoʔˈnárami t͡ʃoʔˈná-r-ame* hit.with.fist-pst.pass-ptcp 'beaten up one' 'golpeado' < BFL 06 4:168/el >

### i. *miˈʔàrami*

*miˈʔà-r-ame* kill.sg-pst.pass-ptcp 'killed one' 'asesinado' < BFL 06 4:168/el >

### j. *koˈʔírame*

*koˈʔí-r-ame* kill.pl-pst.pass-ptcp 'killed ones' 'asesinados' < BFL 06 4:168/el >

### k. *iˈt͡ʃìrame*

*iˈt͡ʃ-ì-r-ame* sow-appl-pst.pass-ptcp 'thing sown' 'lo plantado' < BFL 06 4:168/el >

### l. *ˈʔwîrami*

*ˈʔwî-r-ame* harvest-pst.pass-ptcp 'thing harvested' 'lo cosechado' < BFL 06 4:168/el >

### m. *wiˈt͡ʃôrami wiˈt͡ʃô-r-ame*

wash-pst.pass-ptcp 'washed clothes' 'lo lavado' < SFH 06 4:168/el >

### 8 Nouns


There is variation among speakers in the prosodic make up of agentive and patientive nominalizations. Variation is mostly linked to stress placement in the base, but is also apparent when a given base includes a construction that has more than one possible morphological exponent for a given category. This variation is not exclusively found in nominalization constructions (see further discussion in Chapter 9). Examples of prosodic variation are provided in (39), where both vocalic and stress placement differences are attested.

(39) Prosodic variation of nominalized forms


### f. *iˈt͡ʃiirami*


### h. *niˈʔokirami*

*niˈʔo-ki-r-ame* speak.badly-appl-pst.pass-ptcp 'what has been written' 'lo escrito' < BFL 06 4:168-71/el >

### i. *oˈsírami*

*oˈsí-r-ame* write-pst.pass-ptcp 'what has been written' 'lo escrito' < BFL 06 4:168-71/el >

## j. *osiˈrúami*

	- 'lo escrito' < SFH 06 4:168-71/el >

### 8 Nouns

There is another class of nominalizations, what I will call "theme nominalizations", where the participial marker attaches to a medio-passive base, and the nominalization refers to an inanimate referent that has undergone a change of state. Theme nominalizations are exemplified in (40):

(40) Theme nominalizations *aˈt͡ʃèwami aˈt͡ʃè-wa-ame* throw.in-mpass-ptcp 'the thing that is being thrown in' 'lo que le echan' < BFL 09 1:37/el >

The example in (41) shows a comparison between an agentive and a patientive nominalization; in this case, there is a gap in the theme nominalization, since a clause with the complementizer *ˈnápi* was given instead of a morphologically derived word in (41c) (i.e. the theme nominalization does not contain the participial marker).

	- a. *naˈmûti niˈhîami naˈmûti* thing *niˈhî-ame* gift-ptcp 'the one who gifts things' 'el que regala' < BFL 09 1:43/el >
	- b. *naˈmûti aˈrîwami naˈmûti* thing *a-rîwa-ame* give-mpass-ptcp 'the one who is given things' 'al que le dan cosas' < BFL 09 1:43/el >
	- c. *ˈnápu niˈhíruwa nápi* sub *nihí-riwa* give-mpass 'what is being given away' 'lo que se regala' < BFL 09 1:43/el >

### 8.5 Deverbal nouns

In this case the patientive nominalization is formed from a habitual passive base, instead of a past passive base.<sup>26</sup>

The participial construction also appears in nominalizations that are neither agentive nor patientive. In (42) below, the resulting derivation yields 'experiencer' nominalizations:

(42) Experiencer nominalizations

a. *ˈtû piˈrêami ˈtû* down.river *peˈrê-ame* inhabit.pl-ptcp 'the inhabitants down river' 'los que viven río abajo' < SFH 06 11.06/el > b. *kuˈnêami kuˈnà-ê-ame* husband-have-ptcp 'the ones who have husbands' 'las que tienen marido' < SFH 06 11.06/el >

### **8.5.2 Deverbal nouns with** *-ri*

There are nouns derived from verbal bases through the nominalizing suffix *-ri*. Examples of this suffixes are shown in (43).<sup>27</sup>


<sup>26</sup>One possibility is that this particular base may derive a patientive nominalization from the habitual passive since the verb in question is *à* 'give', a ditransitive verb with a theme argument encoded as a secondary object and a recipient argument encoded as a primary object (Chapter 13 provides an overview of different clause types, including ditransitive clauses (§13.1.4)).

<sup>27</sup>In these examples, the final vowel of the base undergoes post-tonic reduction (raising to [i]). More details about post-tonic vowel reduction are provided in §5.2.

8 Nouns


The form in (44) is an example of a phonologically non-transparent nominalization (where the vowel of the nominalizing suffix undergoes anticipatory asimilation with the following particle).

(44) *ˈpé aʔˈlá amiˈnábi ˈàa busuˈrêro ˈt͡ʃo ˈlé ˈpé* just *aʔˈlá* well *amiˈná=bi* more=just *ˈà* give.prs *busuˈrê-ri* wake.up-nmlz *ˈt͡ʃo* also *aˈlé* dub 'but we must also give more advice (to our children)' 'pero hay que darle más consejo (a nuestros hijos)' < SFH 06 in61(713)/in >

### **8.6 Spanish noun loanwords**

The prosodic properties of loanwords from Spanish are addressed in Chapter 3 and Chapter 7. This subsection addresses the morphological strategies with which borrowed nouns from Spanish are incorporated into the Choguita Rarámuri lexicon. Specifically, relatively recent loan nouns from Spanish appear in constructions with the *-tʃ͡i* suffix. This marker is homophonous with the locative suffix, but in none of the examples below is there any evidence for locative semantics. The formative's sole function in these cases is to nativize the loanwords. This is exemplified in (45).



### 8.7 Tone in morphologically complex nouns

As described in §7.3, there are a few loan nouns from Spanish that do not bear any morphological marking, but display varying degrees of phonological adaptation. Some examples are given in (46):

(46) Varying degrees of adaptation in Spanish loanwords


Finally, there are a few nouns where newer loans coexist with older loanwords, as exemplified in (47). In these pairs, older loans display more phonological adaptation (e.g., *káwi*) than the newer loanword or are based in an archaic form of Spanish (*e.g., limeta*, a noun no longer in use in contemporary Northern Mexican Spanish).

(47) Coexistence of recent and archaic loanwords


There are no other documented morphological strategies to nativize Spanish nouns in the corpus of Choguita Rarámuri.

### **8.7 Tone in morphologically complex nouns**

This section addresses the tonal properties of morphologically complex nouns. While nominal morphology is more limited than verbal morphology, and the interactions between tonal and stress properties of both roots and morphological constructions is largely a matter of general phonological processes, some tonal patterns in morphologically complex nouns exhibit lexical and morphological conditioning.

Nouns exhibit the following word-prosodic properties, which are shared with other major word classes (see also Chapter 11, §11.3.1 and §11.3):

(48) Word-prosodic properties of nouns


### 8 Nouns


From the set of nominal morphological constructions, there are only two suffixes that trigger a stress shift in the bases to which they attach. These are listed in (49). Stress falls on the suffix syllable, where the underlying lexical tone of the suffix is realized.

(49) Nominal stress-shifting suffixes


The examples in (50) and (51) illustrate the stress and lexical tonal properties of these suffixes.


As shown in these examples, stress shifts one syllable rightward to the suffix syllable, the third syllable of the word. In contrast, other nominal suffixes never

### 8.7 Tone in morphologically complex nouns

condition stress shifts with the nouns to which they attach (e.g., the locativeinessive *-rare* suffix never triggers stress shifts; compare, for instance, *saʔˈpá-rare* 'in the meat', with no stress shift and *saʔpa-ˈtʃ͡í* 'on the meat', with stress shift, in (50c) above). There are no other documented cases in the corpus of Choguita Rarámuri where other nominal suffixes trigger stress shifts in the morphologically complex words in which they appear.

While nominal morphological constructions may be classified as stress-shifting or stress-neutral, triggering stress shifts or not, respectively, nominal roots do not exhibit a stressed/unstressed distinction like verbs do. Recall that verbs may be classified as lexically stressed or unstressed, with lexically stressed verbs retaining stress in a fixed position across paradigmatically related words and lexically unstressed verbs consistently shifting stress when attaching a stressshifting construction (see §11.3.8). In words containing unstressed verbal roots and a stress-neutral construction, stress is assigned by default to the second syllable of the root (or the only syllable of monosyllabic roots) (see §5.3). In contrast to this pattern, stress shifts in nouns are not homogenously observed with all suffixes that may be stressed (analyzed here as 'stress-shifting'). As shown in (52) below, some nominal paradigms exhibit stress shifts with the stress-shifting locative-adessive *-tʃ͡í* suffix, but not with the stress-shifting possessive *-lâ* suffix (forms with stress shift are highlighted in boldface).

(52) Stress behavior of morphologically complex nouns


Suffix-triggered stress shifts in nominal forms are thus not homogeneously attested, but rather are a matter of lexical conditioning. However, the distinction between stress-shifting and stress-neutral is kept here for nominal morphological constructions, reflecting the ability to trigger a stress shift and bear stress or not, respectively.

Examples of the three lexical tones in noun roots are provided in (53).

### (53) Lexical tones in nouns


### 8 Nouns


As attested in morphologically complex verbs, the underlying lexical tone of nominal suffixes is revealed when stressed, as shown in (54).


As shown in these examples, the locative *-tʃ͡i* suffix emerges with a H tone when stressed, regardless of the lexical tone of the nominal root (either L (54a), H (54b), or HL (54c)).

These tonal patterns are thus lexical tone patterns, which are contingent on the morphologically-conditioned stress assignment that operates across the language. In addition to these lexical tone patterns, there is also evidence for grammatical tone in nouns. Specifically, the locative-innessive *-rare* [-*riri*] suffix, a stress-neutral suffix that is never stressed, replaces stem L tones with a grammatical H tonal melody in the stressed syllable of the stem. This is shown in (55). As shown in these examples, tonal replacement is only attested when the root bears a lexical L tone (55a–b), but no changes are attested with nouns with other lexical tone patterns (55c–d).

(55) Tonal replacement in morphologically complex nouns


This pattern involves a morphologically-conditioned tonal effect, similar to the one documented with verbs inflected for imperfective or present progressive

### 8.7 Tone in morphologically complex nouns

(see §9.2.3), where morphological constructions condition a tonal change in the stem without bringing about a stress change.

Morphologically-conditioned tonal replacement in nouns, however, exhibits both inter- and intra-speaker variation. Some examples are shown in (56):


For speaker SFH, there is no replacement of the stem lexical L tone when attaching the locative-inessive suffix (56a) (in contrast to (55b) where the same noun undergoes tonal replacement for speaker LEL). Speaker LEL exhibits variable realization of tonal patterns with nouns inflected with this suffix, as in (56b), which contrasts with example (55a), where the same noun undergoes tonal replacement for the same speaker.

There is thus very marginal evidence for grammatical tone patterns in morphologically complex nouns, in contrast to tonal patterns in verbs (§11.3.7). This can be understood as instantiating a common cross-linguistic trend whereby nouns show greater 'phonological privilege' than verbs, where phonological privilege is defined as the ability to support a greater array of phonological contrasts (Smith 2011). Phonological privilege may involve larger number of underlying phonological distinctions, more variety in surface patterns and/or greater resistance to undergoing phonological processes. Lack of grammatical tone in morphologically complex nouns can be also understood as contributing to the phonological privilege of nouns in Choguita Rarámuri.

# **9 Verbs and the verbal complex**

This chapter provides an overview of the verbal morphology, the verbal complex and morphologically-conditioned phonological processes of Choguita Rarámuri and provides evidence for a hierarchical structure of the verb. The chapter is structured along two important topics in the language: (i) phonological and morphological properties of verb roots and morphological constructions (affixes and non-concatenative processes), and (ii) the morphological structure of verbs. Choguita Rarámuri verbal morphology is highly synthetic and agglutinative (where agglutination is understood to involve mostly separative-non-flexive morphology (Bickel & Nichols 2007)). The morphological structure of the language does not instantiate a position-class system (as defined in Simpson & Withgott 1986 and Inkelas 1993),<sup>1</sup> but is rather arranged in verbal zones, in concentric layers that are evidenced by the degrees of morpho-phonological fusion displayed by verbal suffixes.

This chapter provides evidence for twelve suffix positions that are grouped into six verbal domains, schematized in Table 9.1: an Inner Stem, the input to suffixation, a Derived Stem, a Syntactic Stem, an Aspectual Stem, a Finite Verb domain and a Subordinate Verb domain. Each of these domains is semantically, morphotactically, and morpho-phonologically motivated, with morphological exponents closer to the root displaying a higher degree of phonological fusion.

The Inner Stem, described in §9.3, is the domain of body part incorporation, non-concatenative processes and unproductive concatenative markers. The morphological processes taking place at this stage are more tightly fused phonologically to the root than any later morphological process as they may undergo hap-

<sup>1</sup>Position-class or templatic analyses are posited when affix order in a given language is not governed by semantic, syntactic or phonological principles, and every morpheme in the system is assumed to be lexically indexed for a particular fixed position in a total linear arrangement of position classes. In this kind of system, morphemes are rigidly ordered, there are formal dependencies between discontinuous suffixes, inflectional and derivational exponents are interspersed within the verbal structure, and semantically compatible suffixes might be in complementary distribution due to their membership to the same position class (Inkelas 1993; see also Rice 2011). None of these properties characterize the Choguita Rarámuri morphological system.

### 9 Verbs and the verbal complex


Table 9.1: Choguita Rarámuri verbal stem domains

lology (1a), compensatory lengthening (1a),<sup>2</sup> vowel lengthening triggered by the past passive suffix (1b), and round harmony, among other processes (a comprehensive description of the morphologically conditioned phonological processes attested in Choguita Rarámuri verbs is provided in §9.4.3).

(1) Inner Stem examples

```
a. tʃ͡aˈbóopo
    /[tʃ͡aˈbó-pi]InnerStem-po/
    beard-rev-fut.pl
    'They will remove their beards.'
    'Se van a quitar la barba.' < SFH 08 1:5/el >
```
b. *oˈsìiru*

```
/[oˈs-ì]InnerStem-ru/
write-appl-pst.pass
'It was written.'
'Fue escrito.' < SFH 08 1:45/el >
```
<sup>2</sup> In (1a), there is syllable deletion in avoidance of adjacent identical syllable onsets (/tʃ͡aˈbó**-pi-po**/, as well as compensatory lengthening of the stressed syllable as a result of this deletion process ([tʃ͡a**ˈbóo**po]) More details about haplology and compensatory lengthening can be found in §9.4.3.1 and §9.4.3.2, respectively.

The first verbal domain involving suffixation, the Derived Stem, includes inchoative and transitive suffixes (2a), derivational suffixes that are only attested with change-of-state predicates (further discussion of these markers can be found in Appendix A.1). This stem domain is characterized morpho-phonologically by: (i) a non-concatenative morphological process of stress shift to the final syllable of the domain to encode imperative singular; (ii) vowel lengthening induced by the past passive suffix (2b); and (iii) morphologically conditioned stress shifts.

(2) Derived Stem examples

```
a. tʃ͡okoˈbánali
    /[[tʃ͡oko]InnerStem-ˈbá-na]DerivedStem-li/
    be.sour-inch-tr-pst
    'S/he made them go sour.'
    'Hizo que se agriaran.' < SFH 04 1:113/el >
b. rapaˈnâaru
    /[[rapa]InnerStem-nâ]DerivedStem-ru/
    split-tr-pst.pass
    'She was already operated (lit. cut).'
    'Ya la operaron (cortaron).' < SFH 08 1:84/el >
```
The next verbal domain, the Syntactic Stem, is comprised of valence-increasing suffixes that display variable ordering, as well as multiple exponence (3a).<sup>3</sup> In terms of their morpho-phonological properties, these suffixes form a coherent domain since they are stress-neutral and part of the domain of a phonological process of round harmony (3b). Further discussion and examples of these markers can be found in Appendix A.2.

(3) Syntactic Stem examples

a. *noˈkèrtikiri!*

/[[noˈka-è]InnerStem-ri-ti-ki]SyntacticStem-ri/ move.appl-caus-caus-appl-imp.sg 'Move it for someone!' '¡Muéveselo!' < BFL 08 1:28/el >

<sup>3</sup>Two causative markers are attested in this example, instantiating multiple (extended) exponence. See Caballero (2011c) for discussion of this phenomenon in Choguita Rarámuri.

9 Verbs and the verbal complex

> b. *ʃuˈkúkupo* /[[suˈkú]InnerStem-ki]SyntacticStem-po/ scratch-appl-fut.pl 'They will scratch her.' 'La van a arañar.' < BFL 05 1:116/el >

The Aspectual Stem domain is composed of disyllabic modal and aspectual suffixes that are transparently related to independent verbs in the language. As discussed in more detail in §15.6.4, these markers involve a process of V-V incorporation. Aspectual Stem markers may undergo round harmony (4a) and exhibit variable ordering (4b–c). More details about the markers belonging to this stem domain are provided in Appendix A.3.

	- a. *baniˈsútusuma*

/[[[baniˈsú]InnerStem-ti]SyntacticStem-simi]AspectualStem-ma/ pull-caus-mot-fut.sg 'S/he will go along making them pull it.' 'Los va a ir haciendo que lo jalen.' < SFH 07 2:67 rec487 /el >

b. *sutuˈbéetʃ͡-nale*

/[[sutuˈbétʃ͡i]InnerStem-tʃ͡ane-nale]AspectualStem/ trip-ev-desid

'It sounds like they want to trip.'

'Se oye que se quieren tropezar.' < BFL 07 rec300/el >

c. *ˈsûuntʃ͡una*

/[[ˈsû]InnerStem-nale-tʃ͡ane]AspectualStem-a/ sow-desid-ev-prog 'It sounds like she wants to sow.' 'Se oye como que quiere coser.' < SFH 07 1:9/el >

The next two verb domains involve inflectional markers. The Finite Verb domain comprises mood, voice, tense, and aspect suffixes that close verbs used in non-subordinate clauses (5a). Verbs may also add suffixes from a final, Subordinate Verb domain when used in subordinate clauses (5b). Suffixes in these last two domains are largely separable and do not undergo vowel lengthening, rounding harmony or other morpho-phonological effects attested in inner verbal domains.

### (5) Finite and Subordinate Verb examples

```
a. poˈt͡ʃítisima
```
/[[[[poˈt͡ʃí]InnerStem-ti]SyntacticStem-simi]AspectualStem-ma]FiniteVerb/ jump-caus-mot-fut.sg 'I will go along making the dog jump.' 'Voy a ir haciendo que brinque el perro.' < SFH 08 1:72/el > b. *omoˈwáruatʃ͡i* /[[[omoˈwá]InnerStem-riwa-a]FiniteVerb-tʃ͡i]SubordinateVerb/ make.party-mpass-prog-temp 'like with parties, when parties are made' 'así como las fiestas, cuando hacen fiesta' {SFH tx12:5:40.5}

Finally, inflected verbal forms are part of a larger unit, referred to here as the Verbal Complex (discussed in detail in §9.5), which includes person enclitics (6a) and epistemic modality markers (6b). Person enclitics (which encode person, number and a nominative-accusative case distinction) are unrestricted in terms of the syntactic category of the words they may attach to. Epistemic markers (which encode the degree of certainty speakers have towards the actuality of an event) are prosodically independent markers that may only follow inflected verbs. Epistemic markers show some degree of fusion with their bases, as they may condition phonological changes in the inflected verb.<sup>4</sup>

(6) Verbal Complex: person clitics and epistemic markers

a. *kaˈtʃ͡îkini*

/[[[kaˈtʃ͡î]InnerStem-ki]FiniteVerb]=ni]VerbalComplex/

spit-pst.ego=1sg.nom 'I spitted.'

'Escupí.' {BFL el656:1:27.6}

b. *wipiˈsómo ˈlá*

/[[[wipiˈsó]InnerStem-ma]FiniteVerb oˈlá]VerbalComplex/ hit.w.stick-fut.sg cer

'(I) will definitely hit it with a stick.'

'(Le) voy a pegar con palo.' {SFH el1080:3:29.0}

<sup>4</sup> In (6b), the initial vowel of the certainty marker *oˈlá* replaces the final vowel of the future singular suffix *-ma*.

### 9 Verbs and the verbal complex

As argued in this chapter, this organization of the morphology into these verbal domains is critical in understanding the patterns of morphologically-conditioned phonology and variable affix ordering patterns attested in this language.

This chapter also addresses the morphological role of tone, in terms of tonal properties of verb classes, tone alternations as phonological effects imposed by certain morphological constructions, and tone as the exponent of morphological categories. The strictly phonological aspects of the interaction between stress and tone in the word prosodic system of the language are addressed in Chapter 5, while the interaction between tone and intonation is addressed in Chapter 6. Complex prosodic interactions involving stress, tone (lexical and grammatical) and intonation are described in Chapter 11.

This chapter is organized in two parts, which address the verb from the insideout. The first part is concerned with verbal root classes and the prosodic properties of roots and suffixes (in §9.1), grammatical tonal patterns in verbal paradigms (in §9.2), and other non-concatenative morphological processes taking place at the Inner Stem (in §9.3). The second part of the chapter is concerned with the suffixation domain and the clitics and modal particles that occur in the verbal complex (in §9.4 and §9.5). Topics addressed in this second half include the morphotactic evidence for positing different positions in the verbal template (§9.4.2), and the morpho-phonological evidence for positing verbal domains (§9.4.3), including the distribution of stress-shifting and stress-neutral suffixes in the verbal stem. A description of the basic phonological and morphosyntactic properties of verbal suffixes in each verbal domain can be found in Appendix A.1–A.5.

### **9.1 Verbal root classes**

### **9.1.1 The contrast between stressed and unstressed roots**

Choguita Rarámuri verbal roots can be divided in three classes depending on their underlying stress and vowel specifications. Membership in these classes is manifested through presence or absence of stress shifts and vowel quality differences, triggered by specific affixes and non-concatenative morphological processes.

Roots can be first characterized as either underlyingly stressed or unstressed. Stressed roots do not display any stress shifts or vocalic alternations in any morphological context (as exemplified in (7)). All unstressed roots, on the other hand, shift stress one syllable to the right in certain morphological contexts (as exemplified in (8)). Stress syllables are in bold face.

### 9.1 Verbal root classes

### (7) Stressed roots



A partial list of stress-shifting suffixes is provided in Table 9.2.


These suffixes contrast with another class of suffixes, stress-neutral suffixes, that do not trigger any stress shifts on any class of roots, exemplified in Table 9.3.


### 9 Verbs and the verbal complex

Suffixes and non-concatenative processes that condition stress and other morpho-phonological changes on unstressed roots are referred to here as "shifting" morphological constructions. Stress-neutral suffixes and non-concatetative processes, on the other hand, do not trigger any morpho-phonological alternations, and form "neutral" morphological constructions. Here I employ the term "morphological construction" to refer to any morphological process or pattern that combines two sisters into a single constitutent to form a complex word (Inkelas & Zoll 2005: 12). Each individual affix or non-concatenative process thus involves a unique morphological construction. By grouping morphological constructions into "shifting" and "neutral" classes, I suggest that morphological constructions in the language belong to two classes based on their phonological properties. The details of the phonological properties of particular morphological constructions are addressed in §5.3.

Unstressed roots can be further subdivided in two classes: (i) unstressed roots with fully specified vowels (exemplified in (9)); and (ii) unstressed roots with root final unspecified vowels (exemplified in (10)). While the former class of roots (Class 2) has vowels that are fully specified in their underlying representation, I assume that the latter class of roots (Class 3) have a final V slot, whose features are dependent on the morphological construction in which the root takes place. Class 2 roots display no vocalic alternations in both shifting and neutral constructions, and Class 3 roots have final root vowel raising concomitant to the stress shift in shifting constructions.

(9) Unstressed roots with specified vowels (Class 2)


(10) Unstressed roots with final unspecified vowels (Class 3)


### 9.1 Verbal root classes

Stressed roots, with no morpho-phonological changes, are the third type of roots in Choguita Rarámuri (Class 1). The difference between Choguita Rarámuri verbal root classes is summarized in Table 9.4.


Table 9.4: Choguita Rarámuri verbal root classes

Class 3 roots have a final stressed low vowel in neutral constructions. There are Class 2 roots that have a final low vowel, but these roots do not undergo final stem vowel raising in shifting constructions. As shown in (11), there are Class 2 roots with final specified *o* (11a–b), *i* (11c–d), *u* (11e–f) and *a* (11g–h).

### (11) Class 2 roots' final specified vowels


Most Class 3 roots, on the other hand, end in high, front vowels in shifting constructions (but can also end in back, mid vowels (e.g., *noko-ˈmêa* in (12d)). Without exception, these roots end in *a* in neutral constructions. Some examples of these roots with alternating final vowels are provided in (12):

### 9 Verbs and the verbal complex

(12) Class 3 roots' final specified vowels


The generalization is that roots with final *a* are split between Class 2 roots (with no vowel alternation in shifting constructions as in (11g–h)), and Class 3 roots (with vowel alternations in neutral constructions (as in (12)).

### **9.1.2 Stress-shifting and stress-neutral constructions across Uto-Aztecan**

### **9.1.2.1 Morphosyntactic-based accounts**

Some descriptive and historical-comparative works on Uto-Aztecan languages treat the stress shifts and vocalic alternations of roots just described, present in other Uto-Aztecan languages, as stem suppletive allomorphy, where roots with no stress shifts or vocalic alternations encode non-future meanings (like past, perfective, and imperfective), and roots with stress shifts and vocalic alternations encode a future or "unrealized" stem meaning (like irrealis, counterfactual, imperative, and potential) (Langacker 1977: 133). Thus, the morpho-phonological alternations attested in verb roots could be analyzed as being morphosyntactically motivated. There are, however, shortcomings to a classification of shifting and neutral constructions along a future/irrealis scale. Specifically, many of the categories in either shifting or neutral constructions cannot be characterized as having either a non-future or an "unrealized" meaning (such as valency-increasing categories). In addition, the past passive suffix, which is stress-shifting, patterns along with the 'unrealized' or future forms, contrary to the morphosyntacticallybased stem allomorphy proposal.

In Choguita Rarámuri, then, the constructions that trigger a stress shift with certain roots do not themselves encode any constant morphosyntactic features or properties, but rather are a heterogenous class of morphological constructions. This fits in the definition given for *morphomic* stems by Aronoff (1994), where "the mapping from morphosyntax to phonological realization is not direct, but rather passes through an intermediate level" (1994: 25), which is purely morphological (see also Blevins 2003). The two sets of morphological constructions identified (shifting and neutral constructions) cannot be characterized morphosyntac-

9.1 Verbal root classes

tically or semantically, but only phonologically (see also §11.3.7.3 and Caballero & German 2021).

### **9.1.2.2 Conjugation class analysis alternative**

There is an alternative analysis for Choguita Rarámuri verbal roots, and that is to treat the alternations as indicative of conjugation classes (as Brambila 1953 and Lionnet 1972 do for other Rarámuri varieties). Under such an analysis, the stress and vocalic alternation properties of roots would reflect an arbitrary division of the lexicon; that is, the lexically conditioned stem allomorphy would be expressed through item-based alternations (stress shifts and vocalic alternations) when inflected for certain morphological categories.

Under the conjugation class analysis, there would also be three root classes in Choguita Rarámuri.<sup>5</sup> Class 1 roots would not display any stress shift or vocalic alternation when inflected for any morphological category. Class 2 roots would have a "primary stem" and a "secondary stem". For two-syllable stems, the primary stem will have second-syllable stress, while the secondary stem will have third-syllable stress, on a suffix adjacent to the root. Finally, Class 3 roots also have a primary stem and a secondary stem marked by a stress shift and a concomitant vowel alternation: the primary stem has a stem final low vowel (usually *a*, but also *o*), while the secondary stem has a stem final high, front high vowel (*i*). Table 9.5 exemplifies these conjugational classes.



Conjugational classes, as flexive formatives in general, are characterized by displaying item-based variation; this variation, however, is lexically, not morphophonologically, conditioned (Bickel & Nichols 2007). A diagnosis in favor of the

<sup>5</sup>Brambila (1953) and Lionnet (1972) propose three and four conjugation classes, respectively. Lionnet includes irregular forms as additional verbal classes in his classification.

### 9 Verbs and the verbal complex

conjugational class analysis would be the existence of other phenomena that would correlate with the stem types of Choguita Rarámuri besides the stress shift and vocalic alternations identified above. That is, in order to argue that a conjugational class analysis is the correct one for the stem alternations in Choguita Rarámuri, there should be other phenomena correlating with the arbitrary division of the lexicon that could not be explained as arising from morpho-phonological alternations. There are two phenomena in Choguita Rarámuri that could potentially fit this definition. First, Class 2 and Class 3 stems (both of which are unstressed) pattern together in participating in valence-related alternations (described in §9.1.5). However, this valence marking system does not distinguish between any subgroup of the unstressed stems, so there would not be evidence for the distinction between Class 2 and Class 3 stems. Second, there is an apparently lexically-conditioned allomorph of a habitual passive suffix: *-rîwa* for Class 1 stems and *-wá* for the rest.

The evidence for conjugation classes in Choguita Rarámuri is therefore weak. Stress shifts and vocalic alternations of Choguita Rarámuri stems are analyzed here as arising from regular morpho-phonological processes and the existence of two phonological kinds of stems and suffixes: roots with underlying, lexically pre-specified stress, and roots without it; and stress-perturbing suffixes and stress-neutral suffixes. These two kinds of suffixes would be associated with two kinds of morphological constructions. Stress shifts result from regular morphophonological principles that apply to the interactions of the different roots and affixes, rather than from an arbitrary classification of the lexicon into conjugation classes.

While this morpho-phonological analysis of Choguita Rarámuri roots has implications for the analysis of the morphologically-conditioned stress, the conjugational class analysis might be used as a simplified notational device for lexicographic and pedagogical works. Specifically, it is crucial to distinguish three classes of verbal roots in this language in order to predict the roots' behavior in morphological constructions.

### **9.1.3 The interaction of shifting and neutral morphological constructions: stress and vocalic alternations**

The agglutinating nature of the Choguita Rarámuri verbal template allows for the interaction of different kinds of suffixes in terms of their stress properties (as stress-neutral or stress-shifting) in the same word. Given the phonological effects of each type of suffix with unstressed roots, it is necessary to look at unstressed roots that undergo multiple affixation to determine what suffix imposes

9.1 Verbal root classes

its phonological properties on the whole word, i.e., which phonological effects percolate up to the word level.

As we have seen in the previous section, Class 3 roots have a stem with second syllable stress and final *a*) when attaching a stress neutral suffix and a stem with third syllable stress and final unstressed *i* when attaching a stress shifting suffix. In cases of multiple affixation, the stress properties of the word are invariably defined by the first suffix added: in each form exemplified in (13–16), words have third syllable stress when the first suffix is stress shifting (13, 14), and second syllable stress when the first suffix is stress neutral (15–16), regardless of the stress type of outer suffixes. Multiply suffixed forms where the root is immediately followed by a stress neutral suffix, are, however, split with respect to their vocalic qualities: these roots can either have a final stressed *i* vowel (e.g., (15a–b) and (16a–b)) or a final stressed *a* vowel (e.g., (15c–d) and (16c–d)).

(13) Class 3 verb root + stress shifting suffix + stress shifting suffix


(14) Class 3 verb root + stress shifting suffix + stress neutral suffix

a. *osiˈnáliki osi-ˈnále-ki* write.read-desid-pst.ego 'I wanted to read/write.' 'Quería leer/escribir.' < BFL 08 1:18/el > b. *riwiˈwái riwi-ˈwá-i*

find.see-mpass-impf 'It used to be seen/found.' 'Era visto/encontrado.' < BFL 08 1:16/el >

### 9 Verbs and the verbal complex

	- a. *oˈsìsima oˈsì-si-ma* write.read-mot-fut.sg 'S/he will go along reading/writing.' 'Va a ir leyendo/escribiendo.' < SFH 05 1:88/ el >
	- b. *riˈwíwkima riˈwí-wi-ki-ma* find.see-appl-appl-fut.sg 'S/he will find/see it for them.' 'Va a encontrárselo/vérselo.' < BFL 08 1:16/el >

### c. *oˈsàrma*


### (16) Class 3 verb root + stress neutral suffix + stress neutral suffix

a. *oˈsìrili oˈsì-ri-li* write.read-caus-pst 'S/he made them read/write.' 'Lo hizo leer/escribir.' < SFH 08 1:41/ el >

b. *riˈwísio*


*oˈsà-ri-ki* write.read-caus-pst.ego 'I made them read/write.' 'Los hice leer/escribir.' < BFL 08 1:10/el >

9.1 Verbal root classes

d. *raʔˈlártiki raˈlá-ri-ti-ki* buy-caus-caus-pst.ego 'I made them buy it.' 'Los hice comprarlo.' < BFL 08 1:10/el >

Irrespective of the stress type of the suffixes added, then, words built from Class 3 roots will have a stress make-up dependent on the first suffix added, but either a final *i* stem vowel or a final *a* stem vowel.<sup>6</sup>

### **9.1.4 Lexical tone in lexically stressed and unstressed verbs**

In addition to the contrast in terms of lexically specified stress (stressed and unstressed), verbal roots bear lexical tonal contrasts with a distribution that is partly dependent on the lexical stress properties of roots. First, all tonal contrasts (HL, L and H) are realized with lexically stressed roots (those with fixed stress across paradigms). This is exemplified in (17), where stress is fixed with both neutral (e.g., past *-li*) and shifting (e.g., future sg. *-ma*) suffixes:<sup>7</sup>

(17) Tonal properties of lexically stressed roots


<sup>6</sup>The fact that there is variation of stem shape in constructions with multiple suffixes could suggest a "look-ahead" effect of morphology, where variation in vocalic quality is dependent on having multiple suffixation (i.e., the outer suffixes 'see' inside the preceding morphological structure, a violation of bracket erasure). An alternative is to assume that variation in stem selection is exclusively found when Class 3 roots add suffixes of inner layers (e.g., the Syntactic Stem and Aspectual Stem verb domains defined below in §9.4). Given the hierarchical structure of the verb proposed in this grammar, suffixes belong to different verbal domains. The phonological features of certain domains will thus percolate up to the word level and impose its phonological properties to the word form.

<sup>7</sup>As described in §9.2.4 below, a subset of lexically stressed roots have "alternating" tone patterns, or morphologically determined tone in inflection: L tone in neutral morphological contexts and HL tone in shifting morphological contexts. This class differs from other verbal roots since it displays tonal changes in the absence of stress shifts.

### 9 Verbs and the verbal complex

On the other hand, lexically unstressed roots are predominantly H-toned (e.g., (18a–d), though there are also unstressed L-toned roots documented (e.g. (18e– f). Crucially, there are no documented HL-toned unstressed roots. The examples in (18) show how these roots, like all unstressed roots, shift stress in shifting contexts (e.g., when inflected for future singular *-mêa* ~ *-ma*), but keep stress in the stem in neutral contexts (e.g., when inflected for past *-li*). The lexical tone of the root emerges in bare stems and when inflected for neutral constructions.

(18) Tonal properties of lexically unstressed roots


Given the surface tonal distributions in morphologically complex verbs in Choguita Rarámuri, the analysis assumed here is that all lexical tones are specified, in both lexically stressed and lexically unstressed verb roots. The stress and tonal properties of morphologically complex verbs are discussed further in §9.2 below and in Chapter 11.

### **9.1.5 Valence alternations**

In the previous section, we have seen that there are three identifiable verbal root classes in Choguita Rarámuri, which have characteristic stress and vowel alternation properties in morphologically defined contexts. Unstressed roots (Class 2 and Class 3 roots) can, in addition, undergo valence related alternations through the affixation of transitive and applicative vocalic suffixes that replace the final vowel of the stem.

In this valence alternation system, there are three kinds of stems: intransitive, transitive and applicative stems. Intransitive stems end in an unstressed vowel (19). Transitive stems replace the final stem vowel with a stressed *-a* suffix (20).<sup>8</sup>

<sup>8</sup>Heath (1978) provides evidence that an intransitive-transitive contrast marked by *i* for intransitive and *a* for transitive goes back to Proto-Uto-Aztecan.

### 9.1 Verbal root classes

Applicative stems replace the final stem vowel with a stressed, low toned front vowel suffix (*-è* or *-ì*) (21).<sup>9</sup>

### (19) Valence stem allomorphy: Intransitive

a. *niˈhê ˈmá nokoˈmêa niˈhê* 1sg.nom *ˈmá* already *noko-ˈmêa* move.intr-fut.sg 'I will move.'

'Ya me voy a mover.' < SFH 05 1:80/el >

b. *ukuˈmêa uku-ˈmêa* rain.intr-fut.sg 'It will rain.' 'Va a llover.' < SFH 05 1:82/el >

(20) Valence stem allomorphy: Transitive *ˈmá raʔˈláki ˈmá* already *raʔˈl-á-ki* buy-tr-pst.ego '(He) already bought it.'

'Ya lo compró.' < AHF 05 1:130/el >

<sup>9</sup>There is only one example that shows a semantic difference between an applicative stem formed with a final stressed high front vowel and an applicative stem with a final stressed mid front vowel: the verb *raʔˈlá/raʔli-* 'to buy', has two applicative bases: *raʔˈl-è* 'buy from)' (i.a) and *raʔˈlì* 'buy for' (i):


There are no other examples where two applicative stems have been lexicalized with different meanings.

### 9 Verbs and the verbal complex

	- a. *niˈhê ˈmí ˈtrôka noˈkèli niˈhê* 1sg.nom *ˈmí* 2sg.acc *ˈtrôka* truck *noˈk-è-li* move-appl-pst 'I will move the truck for you.'
		- 'Te voy a mover la troca.' < SFH 05 1:80/el >
	- b. *ˈá=mi taˈmí raʔˈlèma ˈá=mi* aff=2sg.nom *taˈmí* 1sg.acc *raʔˈl-è-ma* buy-appl-fut.sg 'Will you buy (it) from me?' '¿Me lo compras (yo lo vendo)?' < SFH 05 1:75/el > c. *mi uˈkèli mi* dem *uˈk-è-li* rain-appl-pst 'It rained for (him).' 'Le llovió.' < SFH 05 1:84/el >

Table 9.6 schematizes the three-way contrast between intransitive, transitive and applicative stems (syntactic/semantic gaps are symbolized by dashes).


Table 9.6: Valence stem allomorphy

Verbs that exhibit this valence alternation may exhibit multiple exponence: the examples in (22–24) show how the applicative is redundantly marked, as there are two applicative markers in (22b), (23b) and (24b), but only one benefactive

### 9.1 Verbal root classes

or malefactive argument introduced. These forms are semantically equivalent to their counterparts with only one applicative exponent (in (22a), (23a), and (24a), respectively) and speakers use the two types of forms interchangeably.

(22) a. *ˈmáni ˈmí suˈwèli reˈmê ˈmá=ni* already=1sg.nom *ˈmí* 2sg.acc *suˈw-è-li* finish-appl-pst *reˈmê* tortillas 'I already finished (ate) up your tortillas.' 'Ya me acabé tus tortillas.' < SFH 05 1:119/el > b. *ˈmáni ˈmí suˈwèkili reˈmê ˈmá=ni* already=1sg.nom *ˈmí* 2sg.acc *suˈw-è-ki-li* finish-appl-appl-pst *reˈmê* tortillas 'I already finished (ate) up your tortillas.' 'Ya me acabé tus tortillas.' <LEL 06 5:123/el > (23) a. *oˈʃìma oˈs-ì-ma* write-appl-fut.sg 'She will write him (a letter).' 'Le va a escribir (una carta).' < BFL 06 2:98/el > b. *oˈʃìkima oˈs-ì-ki-ma* write-appl-appl-fut.sg 'She will write him (a letter).' 'Le va a escribir (una carta).' < BFL 06 2:98/el > (24) a. *roˈnèma roˈn-è-ma* boil-appl-fut.sg 'He will boil it for her.' 'Se la va a hervir.' < BFL 06 2:101/el > b. *roˈnèkima roˈn-è-ki-ma* boil-appl-appl-fut.sg 'He will boil it for her.'

'Se la va a hervir.' < BFL 06 2:101/el >

### 9 Verbs and the verbal complex

Finally, transitive and applicative stems block any stress shift or vocalic alternation conditioned by shifting and neutral morphological constructions. That is, transitive and applicative stems always have stress on their final stem vowel, and block any stress shift or vocalic alternation imposed by shifting morphological constructions.

### **9.1.6 Change-of-state predicates**

In addition to the valence stem allomorphy of unstressed roots, Choguita Rarámuri has a second valency-increasing stem allomorphy system, marked through thematic suffixes and stress alternations in a semantically defined class of verbs, change-of-state predicates. In Choguita Rarámuri, the class of change-of-state predicates is composed of unstressed roots that are marked as intransitive, transitive or applicative through the presence or abscence of transitive suffixes and specific stress patterns. All change-of-state verb roots are disyllabic.

The Guarijío cognates of this verb class are defined as predicates that denote an irreversible change-of-state or condition, generally produced by some kind of contact, with prototypical intransitive-causative pairs 'break', 'twist', 'torn apart', 'shatter', 'spill' (Miller 1996: 153). This class of verbs shares morphological traits accross the Uto-Aztecan language family: Heath (1978) reconstructs a morphological class of verbs for Proto-Uto-Aztecan (PUA) and the intermediate protolanguages, Proto Northern Uto-Aztecan (PNUA) and Proto Southern Uto-Aztecan (PSUA), and gives descriptions of this class of verbs in Southern Paiute, Mono, Luiseño, Cupeño, Cahuilla, Serrano, Hopi, Tepiman, Tarahumara and Aztecan language varieties. This class is composed of disyllabic roots with thematic variation, and it includes verbs denoting events of physical change-ofstate, even though the actual verbs are not cognate from one language to another.

Intransitive stems involve no theme suffixes and have fixed stress: stress is on the second syllable, whether a stress-shifting suffix (future singular in (25a) and conditional in (25c)) or a stress-neutral suffix (past in (25b) and (25d)) is attached.

### (25) Intransitive change-of-state predicates

a. *tʃ͡iˈháma ˈlé ˈtʃ͡îba tʃ͡iˈhá-ma* scatter-fut.sg *aˈlé* dub *ˈtʃ͡îba* goats 'The goats will scatter.' 'Se van a desparramar las chivas.' < SFH 07 1:17/el >

9.1 Verbal root classes

b. *ˈmá tʃ͡iˈháli ˈtʃ͡îba ˈmá* already *tʃ͡iˈhá-li* scatter-pst *ˈtʃ͡îba* goats 'The goats already scattered.' 'Ya se desparramaron las chivas.' < SFH 07 1:17/el > c. *riˈpúsa kuˈsì riˈpú-sa* cut-cond *kuˈsì* wood 'If the wood is cut.' 'Si se corta la leña.' < BFL 08 1:25/el > d. *ˈmá riˈpúli kuˈsì ˈmá* already *riˈpú-li* cut-pst *kuˈsì* wood 'The wood was already cut.' 'Ya está cortada la leña.' < BFL 08 1:25/el >

Transitive stems, on the other hand, are marked by suffixes (the transitive *-ˈnâ* suffix exemplified in (26) and the transitive pluractional *-tʃâ* suffix exemplified in (27c)), and are also sensitive to shifting and neutral constructions: stress is shifted to the transitive suffix, the third syllable, when attaching stress-shifting morphemes ((26a) and (26c)), but have final root (second syllable) stress when attaching stress-neutral suffixes ((26b) and (26d)).

	- a. *tʃ͡ihaˈnâsa naˈpàtʃ͡i tʃ͡iha-ˈnâ-sa* scatter-tr-cond *naˈpàtʃ͡i* blouse 'if she scatters the blouses...' 'si desparrama las blusas' < SFH 07 1:17/el > b. *ˈpîrim oˈlá tʃ͡iˈhánili naˈmûti*
	- *ˈpîri=mi* what=2sg.nom *oˈlá* why *tʃ͡iˈhá-na-li* scatter.tr-tr-pst *naˈmûti* things 'Why did you scatter the things?' '¿Por qué desparramas las cosas?' < SFH 07 1:17/el >
	- c. *ma ripuˈnâma ˈlé kuˈsì ma* now *ripu-ˈnâ-ma* cut-tr-fut.sg *aˈlé* dub *kuˈsì* wood 'S/he'll cut the wood now.'

9 Verbs and the verbal complex

> 'Ya va a cortar la leña.' <RF 04 verbs/el > d. *seˈrûtʃ͡o riˈpúnili kuˈsì seˈrûtʃ͡o* saw *riˈpú-na-li* cut-tr-pst *kuˈsì* wood 'The saw cut the wood.' 'El serrucho cortó la leña.' < BFL 08 1:25/el >

The transitive *-ˈnâ* suffix, thus, has a stressed and an unstressed allomorph. The same is attested for the transitive pluractional *-tʃ͡a* suffix.<sup>10</sup>


Finally, applicative stems are marked through the addition of a transitive suffix and fixed final root (second syllable) stress. As shown in the next examples, the applicative stems of change-of-state predicates have fixed second syllable stress, whether they are in a shifting construction (future singular (28b)) or a neutral construction (past (28d)). Applicative stems are shown with a corresponding nontransitive version of these predicates.

<sup>10</sup>When bearing stress, the transitive suffix often causes the final root vowel to align in color with the first root vowel, as in (27a).

9.1 Verbal root classes

	- a. *ˈpé uˈtʃ͡íin tʃ͡iˈwáli boʔˈnà muˈnî ˈnè ˈpé* just *uˈtʃ͡í=ni* small.space=1sg.nom *tʃ͡iˈwá-li* tear-pst *boʔˈnà* harvest.beans *muˈnî* beans *ˈnè* 1sg.nom 'Just a little bit (in a small piece of land) I tore beans for havest, I.' 'Nomás muy poquito (en un pedazo de tierra chiquito) alcancé a arrancar frijol, yo.' {MDH co1137:9:36.1} b. *niˈhê ˈmí tʃ͡iˈwánima niˈhê* 1sg.nom *ˈmí* 2sg.acc *tʃ͡iˈwá-na-ma* tear.appl-tr-fut.sg 'I'm going to tear it for you.' 'Yo te lo voy a trozar.' < SFH 07 1:21/el > c. *ˈtʃ͡îba tʃ͡iˈháli ˈtʃ͡îba* goats *tʃ͡iˈhá-li* scatter-pst 'The goats got scattered.'

'Las chivas se desparramaron.'

d. *niˈhê ˈmámi tʃ͡iˈhánili naˈpàtʃ͡i niˈhê* 1sg.nom *ˈmá=mi* already=2sg.acc *tʃ͡iˈhá-na-li* scatter-tr-pst *naˈpàtʃ͡i* blouses 'I scattered your blouses.' 'Ya te desparramé las blusas.' < SFH 07 1:21/el >

The differences between the different kinds of stems are schematized in Table 9.7

Table 9.7: Change-of-state predicates and thematic alternations


The effects of neutral and shifting morphological contexts only produce stress shifts with transitive stems of change-of-state predicates.

### 9 Verbs and the verbal complex

### **9.1.7 Summary**

All open class words in Choguita Rarámuri have surface stress, but verbal (and other) roots can be characterized as either lexically stressed or unstressed, a distinction of their underlying phonological representation. This phonological distinction involves significant differences in the prosodic makeup of words, depending on the interaction between roots and different types of suffixes and morphological constructions. Unstressed roots can be further subdivided into two classes if the roots undergo vowel alternations concomitant to the stress shifts. I have presented a morpho-phonological analysis of these three types of roots, and have given arguments in favor of this analysis over a conjugational class analysis.

Unstressed verbs can participate in a valence stem allomorphy system and undergo internal, non-concatenative changes when specified for intransitive, transitive or applicative meanings. Another subset of the unstressed verbs, semantically defined as a class of verbs of change-of-state, participates in a second valence stem allomorphy system. In these systems, final stem vowel changes and fixed stress patterns are the markers of morphological categories (valenceincreasing categories), while any concomitant stress shifts are epiphenomenal of certain morphological constructions, a byproduct of the phonological make up of roots and affixes and their compositional interaction.

In addition to stress shifts and vocalic alternations, morphologically complex verbs are characterized by tonal patterns governed by both phonological and morphological conditioning factors. All lexical tones are analyzed here to be underlyingly specified. Lexically stressed verbs may exhibit any of the three lexical tones of the language (HL, L or H), while lexically unstressed verbs are only specified for H and L tones (i.e., there are no lexically unstressed verbs with lexical HL tones). I turn to the role of tone in morphologically complex verbs in the next section.

### **9.2 The role of tone in verbal morphology**

As described in Chapter 6, Choguita Rarámuri features a three-way lexical tonal contrast. In addition to lexical tone patterns and processes, tone in Choguita Rarámuri has purely morphological functions and properties that require description and analysis independently of its phonological characteristics.

This section is devoted to describing the role of grammatical tone in the verbal morphology of Choguita Rarámuri. Grammatical tone may be defined as a tonal pattern or process that is not general across the phonological grammar 9.2 The role of tone in verbal morphology

of a language, but is instead associated to a specific morpheme or construction, or a natural class of morphemes or constructions (see Rolle 2018).<sup>11</sup>

As discussed in this grammar, stress carries a high functional load in the verbal morphology of the language. While tone appears to carry a smaller functional load than stress, and while lexical tone distribution is itself constrained by stress placement (with morphological distributions of tone being a by-product of the morphological factors that govern stress distribution), tone nonetheless plays an autonomously morphological role in this language. Before addressing each one of these grammatical tone properties, the next section provides an overview of the lexical tonal properties of suffixes in Choguita Rarámuri.

### **9.2.1 Lexical tones of suffixes**

Evidence for three contrastive lexical tones in Choguita Rarámuri is given in Chapter 6 and their distribution in verbal roots was addressed in §6.1.3. Suffixes also have underlying lexical tones, which are evidenced when they are stressed in morphologically complex words. Only those suffixes associated with shifting constructions (stress-shifting) attaching to unstressed roots will reveal their underlying tonal properties in the surface form (as there is no evidence that suffixes belonging to the neutral class bear lexical tones since they are never stressed). The forms in (29) examplify suffixes and their tonal properties when attaching to the lexically unstressed root *sukú* 'to scratch' (Spanish *rascarse*).

(29) Lexical tone of suffixes in partial verbal paradigm


As shown in these examples, suffixes may bear any of the three lexical tones of the language. (30) provides a comprehensive list of shifting suffixes and their underlying tonal properties.

<sup>11</sup>Given that verbal morphology in Choguita Rarámuri is more complex than nominal morphology and the morphology of other word classes, this chapter addresses morphological tone within the verbal domain. The grammatical tonal properties of Choguita Rarámuri nouns are addressed in Chapter 8 (§8.7).

### 9 Verbs and the verbal complex


(30) Lexical tone properties of shifting suffixes

If one of these suffixes is stressed within a complex word, the surface tone pattern of the word may be characterized as involving lexical tone. Lexical tone contrasts with grammatical tone, as the latter involves tonal patterns that are governed directly by morphological factors. I address these next (in §9.2.2 and in §11.3.7).

### **9.2.2 Tone as realizational morphology**

Choguita Rarámuri deploys tone as a morphological exponent in the marking of imperative mood. The imperative singular has four allomorphs, two concatenative exponents (*-kâ* and *-sâ*) and two non-concatenative ones (a floating L tone and a final stress shift).

The low tonal allomorph replaces stem lexical HL tones, neutralizing the contrast between lexical HL and L tones in the imperative mood. The data in (31) exemplify the four allomorphs of the imperative singular construction, contrasting the bare stem (inflected for present tense) and the corresponding imperative singular form.

(31) The imperative singular construction: four allomorphs


9.2 The role of tone in verbal morphology

As shown in these examples, the L tonal allomorph of the imperative singular replaces the HL tone of the stem (31a–b), but tonal replacement does not take place if a suffixal allomorph of the imperative singular is used instead (31c–d). Finally, (31e) shows that a stress shift within the stem may encode the imperative singular inflectional value.

The examples in (32) show a HL-toned verb, *ma*ˈ*tô* 'to carry something in shoulders', inflected for different tense/aspect and person/number values, including the imperative singular (32e).

(32) Partial paradigm of the verb *maˈtó* 'to carry in shoulders'


As shown in the contrast between these inflected forms, a HL-toned verb will keep its lexical tonal melody across derivations without any stress shifts (32), ex-

### 9 Verbs and the verbal complex

cept when inflected for the imperative singular tonal allomorph: in this case, the morphological L tone replaces the HL lexical tone of the stem (32e). In this case, the stem vowel undergoes lengthening in order to accommodate the imperative L-tone allomorph of the imperative singular on the first mora and the H% boundary tone on the second mora of the stem vowel (for more discussion of boundary tones, see §6.2.1).

The following examples show how the imperative singular L tone is a grammatical tone that replaces stem tones, which includes morphologically complex stems composed of roots and stem-forming suffixes, such as the transitive *-na* suffix:

(33) a. *supaˈnàá supa-ˈnà* unravel-tr.imp.sg 'Unravel it!' '¡Deshílalo!' < SFH 2014 1:142> b. *supaˈnâma supaˈ-nâ-ma* unravel-tr-fut.sg '(S)/he will unravel it.' 'Lo va a deshilar.' < SFH 2014 1:142>

As shown in the contrast between (33a) and (33b), a HL tone in a stem containing a stressed transitive *-na* suffix is replaced when inflected for the imperative singular, since the imperative L-tone allomorph associates with the stressed syllable, replacing the stem HL tone.<sup>12</sup>

In contrast, L-toned stems and H-toned stems do not undergo any tonal changes when inflected for the imperative singular. Suffixing allomorphs and stress shifts are available in words containing these stems in order to encode this inflectional construction. This is exemplified in (34).

(34) No tonal replacement


<sup>12</sup>Chapter 11 provides more details about patterns of tonal replacement in morphologically complex words.

9.2 The role of tone in verbal morphology


Replacement of HL tones with grammatical L imperative singular tone thus leads to neutralization of HL and L lexical tones in this construction. Finally, and as shown in (31c–d) above, the L-tone imperative allomorph is incompatible with other imperative singular marking. Thus, there is no multiple exponence involved for this inflectional category.

### **9.2.3 Morphologically-conditioned tone**

The lexical tonal properties of suffixing constructions is addressed in §9.2.1. Recall that these suffixing constructions belong to the shifting class, triggering stress shifts with lexically unstressed roots. There are, however, suffixes that condition tonal alternations onto the stems they attach with without triggering any stress shifts. This is the case of the imperfective *-i* suffix and the present progressive *-a* suffix, two suffixes that impose a L tone on a stressed stem syllable of a complex word containing an underlyingly unstressed root without conditioning any stress shifts: both the imperfective and the present progressive are neutral morphological suffixing constructions that do not trigger stress shifts and are unable to bear stress. The data in (35) exemplify this morphologically-conditioned tone pattern.

(35) Morphologically-conditioned L tone


As the examples in (35) show, the imperfective and the present progressive are associated with a L tone in the stem without triggering stress shifts, in contrast to other neutral constructions such as the past *-li* suffix, which do not trigger any tonal alternations in the stems to which they attach (i.e., the H tone in the stem when attaching the past *-li* suffix is the underlying tone of the verbal root).

As mentioned above, this morphologically-conditioned effect is only attested when the imperfective and progressive suffixes attach to lexically unstressed roots. When the roots exemplified in (35) above attach a shifting construction, the resulting morphologically complex word exhibits a stress shift (in these cases, stress falls on the shifting suffixes). This is shown in (36).

### 9 Verbs and the verbal complex



As in the examples in (36a–c) show, stress shifts with lexically unstressed roots; if stress is on the suffix, the lexical tone associated with that suffix emerges (in the case of the conditional *-sâ* suffix, a falling HL tone). When stress falls on the stem containing these lexically unstressed roots, the surface tone is either the lexical root tone (when attaching the past tense *-li* suffix) or the morphologicallyconditioned L tone (when attaching the imperfective *-i* suffix or the present progressive *-a* suffix). The forms in (36d–f) show that no tonal alternations are attested when the imperfective and progressive suffixes attach to lexically stressed roots (which do not exhibit or any tone changes across morphological paradigms).

Finally, it should be noted that the tonal effect that the imperfective and present progressive suffixes have in the stems to which they attach can be traced diachronically to recent changes in the segmental and prosodic make up of these suffixes: as evidenced in data from Norogachi Rarámuri (Brambila 1953), and the closely related Mountain Guarijío (Miller 1996), both of these suffixes were stressable suffixes that recently lost a palatal glide onset. For more discussion of this recent development, see Chapter 11.

### **9.2.4 Alternating tone stems**

A subset of stressed verbal stems in Choguita Rarámuri (referred to as "alternating" in Caballero & Carroll 2015) exhibit alternating tone patterns that are predictable in terms of the inflectional morphology with which they combine. As exemplified in (37), these stems exhibit a HL tone in shifting morphological contexts (those that condition stress shifts and other morpho-phonological alternations) and a L tone in neutral contexts. There are no stress shifts in either context, and only the tonal alternation is indicative of either the shifting or neutral morphological context. These generalizations hold when considering two-level constructions (stem and only one morphological construction).

9.2 The role of tone in verbal morphology

### (37) Alternating verbs: tonal alternations


The examples in (38) and (39) show the tone alternations of a single verb *biʔwa* 'to clean' (a stem composed of the root *biʔw-* 'to clean' and the transitivizer *-a* suffix), where the surface tonal properties of the word are determined by the type of morphological construction: L if neutral (38) and HL if shifting (39).

### (38) Neutral constructions


### (39) Shifting constructions


In these cases, it is not possible to identify any underlying lexical tone properties of the verb roots and tonal alternations cannot be predicted from the lexical tonal properties of suffixes. These tonal alternations are also independent of stress alternations and other phonological properties. Stress and tone, though closely related, are orthogonal dimensions in Choguita Rarámuri.

### **9.2.5 Summary**

The following is a summary of the different grammatical tone patterns documented Choguita Rarámuri (for further discussion, see Chapter 11, as well as Caballero & German 2021):

### 9 Verbs and the verbal complex

	- Tone may be a morphological exponent of inflection (tone as realizational morphology) (§9.2.2)
	- Suffixes have tonal properties and may induce changes on the stems to which they attach (tone as morphologically-conditioned phonology) (§9.2.3)
	- Tone is systematically distributed across verb classes in terms of neutral vs. shifting inflectional morphology constructions (§9.2.4)

Having addressed the properties of root classes in Choguita Rarámuri, including the tone properties of roots as well as grammatical tone patterns, we can now turn to the non-productive and non-concatenative processes occurring at the next level of the Choguita Rarámuri verb, the Inner Stem.

### **9.3 The Inner Stem: noun incorporation, non-concatenative morphology and unproductive processes**

The Inner Stem, the input to suffixation, is composed of a denominalized noun root, an incorporated verb root, or a verbal root that has optionally undergone a non-concatenative process or a process that is no longer fully productive, including subject number prefixation and pluractional marking.

This first stem domain can also be characterized morpho-phonologically: this is the domain of application of compensatory lengthening (described in §9.4.3.2), and a morphological incorporation stress rule that assigns stress to the first syllable of the head of an incorporated verb (described in more detail in §9.3.3). In general, the morphological processes taking place at this stage are more tightly fused phonologically to the root than any later morphological process.

This sub-section is organized as follows: §9.3.1 describes non-concatenative processes; §9.3.2 gives an overview of instrumental prefixes; §9.3.3 describes the process of body-part incorporation; §9.3.4 describes number suppletion and plural prefixation; and §9.3.5 gives an overview of the unproductive verbalizer suffixes and their phonological effects on their bases.

### **9.3.1 Non-concatenative processes**

Non-concatenative processes in Choguita Rarámuri include conversion (§9.3.1.1), pluractional consonant mutation (§9.3.1.2), stress and tone changes to encode im-

9.3 The Inner Stem

perative singular (§9.3.1.3), and stress shifts encoding a derivational verbalization process (§9.3.1.4).

### **9.3.1.1 Conversion**

Some nominal stems (including some nouns refering to weather) can take verbal morphology with no overt denominal marking. The examples in (41) are listed with the future singular suffix. With some exceptions (like (41a)), most of these zero derived verbs belong to the unstressed stem class in terms of their stress behavior.

### (41) Noun-to-verb conversion


Examples (41c–d) show also that the unstressed vowel of the derived verbal form can harmonize with the stem's first vowel (*moˈri* 'smoke' becomes *morò-* 'to be smoky/foggy' and *uˈki* 'rain' becomes *uku-* 'to rain'). This harmonizing process is absent in cognate forms of zero derivation in Mountain Guarijío (cf. *yuʔkí* 'rain', *yuʔki-má* 'to rain') (Miller 1996: 148).

### **9.3.1.2 Pluractionality: prefixation and consonant mutation**

There is a process that marks plural subject with verbs, or that an action occurs or is being performed by the same agent several times, or by several agents several times. When used with nouns it marks plural number. These meanings are related in that they refer to event plurality or "pluractionality". Pluractionals have been defined as encompassing meanings that range from iterative and frequentative to distributive and extensive action (Newman 1990, Newman 2012, Wood 2007).

Choguita Rarámuri pluractional forms are marked through a prefixed element analyzed in other descriptive works of Rarámuri and Guarijío as reduplication (Lionnet 1968, Lionnet & Miller 1985).<sup>13</sup> In Mountain Guarijío, the cognate process (labelled "plural subject, iterative or durative"), is more clearly analyzed as

<sup>13</sup>Lionnet (2001b) labels these "intensive".

### 9 Verbs and the verbal complex

reduplication, since the prefixed element is (C)V- (e.g. *saé*, *sa-saé* 'smell', *isí, i-isí* 'walk') (Miller 1996: 62).

Pluractionals in Choguita Rarámuri are marked in three ways: (i) through a prefixed vowel (42a–b) (where the vowel quality of the prefix can be harmonized to the root's first syllable vowel); (ii) through consonant mutation (42c–h); or (iii) through both consonant mutation and a prefix element (42i–o).<sup>14</sup>

(42) Pluractionals


Consonant mutation involves a voicing toggle, since it produces voicing or lenition of a voiceless stop (42b–f), and devoicing or hardening of a voiced plosive (42f).

As shown in (43), consonant mutation also targets voiceless velar stops to encode pluractionality.

(43) k ~g alternation in pluractional marking


<sup>14</sup>It has been suggested that the prefix-like element was originally a prefix *i-* that has been leveled in color with the first stem vowel in contemporary Rarámuri varieties (Lionnet 2001b).

9.3 The Inner Stem

The example above shows that the plural of participle *paˈkó-t-ami* (42a), with a word-medial voiceless velar stop, is *paˈgó-t-ami*, with a voiced word-medial velar stop (42b). The data in (44) shows more examples of stop alternations in pluractional constructions.

(44) Stop alternations in pluractional constructions


### **9.3.1.3 Imperative L tone and final stem stress**

As described in §9.2.2, the imperative singular construction has non-concatenative allomorphs, including stress shifts and grammatical tone. Relevant examples are shown in (45).

### (45) Imperative stress shift

a. *raʔaˈmà* give.advice.imp.sg 'Give advice!' '¡Aconseja!'

b. *raʔaˈmâbo raʔaˈmâ-bo* give.advice-fut.pl 'They will give advice.' 'Van a aconsejar.'


9 Verbs and the verbal complex

```
e. raʔiˈtʃ͡âma
     raʔiˈtʃ͡â-ma
     speak-fut.sg
    'S/he will speak.'
    'Va a hablar.' < SFH 2014:60>
f. raˈʔìtʃ͡iki
     raˈʔìtʃ͡a-ki
     speak-pst.ego
    'I spoke.'
    'Hablé.' < SFH 2014:60>
```
Final stem imperative stress of unstressed roots contrasts with the stress pattern of these roots in neutral constructions: in (45), the trisyllabic unstressed roots *raʔáma* 'give advice', and *raʔìtʃ͡a* 'speak', have third syllable stress in the imperative (45a) and (45d)) as well as in shifting constructions ((45b) and (45e)), but second syllable stress in neutral constructions ((45c) and (45f)).<sup>15</sup>

As discussed in further detail in §9.4, the stress shift to mark imperative is restricted to occur within a defined verbal zone, the Derived Stem.

In addition to stress, Choguita Rarámuri deploys tone in the exponence of the imperative singular: in addition to two concatenative suppletive allomorphs, *-sâ* and *-kâ*, a stress shift allomorph, the imperative singular may also be encoded through a L tonal allomorph that replaces stem lexical HL tones. The examples in (32) above, repeated here in (46), show a HL-toned verb, *ma*ˈ*tó* 'to carry something in shoulder', inflected for different tenses/aspect and person/number combinations.

(46) Partial paradigm of the verb *maˈtó* 'to carry in shoulders'

a. *maˈtóa maˈtó-a* carry.shoulders-prs '(S)/he carries it on the shoulders.' 'Lo carga en sus hombros.'

<sup>15</sup>As discussed in more detail in Chapter 11, there are systematic tonal alternations associated with stress shifts on morphologically complex words containing unstressed roots: HL falling tones are associated with stress shifts (e.g., when inflected for the future plural (e.g., (45b)) or future singular, as in (45e)); in the case of the imperative singular encoded through a stress shift, speakers may exhibit variation in the tonal makeup of these forms; the examples above show forms where the imperative singular forms have a L tone, which can be analyzed as the grammatical L tone associated with this construction, an instance of multiple exponence.

9.3 The Inner Stem


As shown in the contrast between these inflected forms, a HL-toned verb will keep its lexical tonal melody across derivations without any stress shifts, except when inflected for the imperative singular tonal allomorph: in this case, the morphological L tone replaces the HL lexical tone of the stem. This construction contributes to the list of non-concatenative processes of the language.

### **9.3.1.4 Stress shift as verbalization**

Some nouns can be denominalized through a stress shift one syllable to the right. The stress shift in these cases cannot be said to be triggered by a shifting construction, since (47a) is inflected with a stress-shifting suffix and (47c) is inflected with a stress-neutral suffix. The stress shift is, thus, only attributable to the verbalizing process.

	- a. *sipuˈtʃ͡âma*

9 Verbs and the verbal complex

```
sipuˈtʃ͡â-ma
     wear.skirt-fut.sg
     'She will wear a skirt.'
     'Se va a poner falda'
b. siˈpútʃ͡a
     'skirt'
     'falda' < BFL 07 Sept 6/el >
c. napaˈtʃ͡âli
     napaˈtʃ͡â-li
     wear.blouse-pst
     'She wore the blouse.'
     'Se puso la blusa.' < BFL 07 Sept 6/el >
d. naˈpátʃ͡a
     'blouse'
     'blusa' < BFL 07 Sept 6/el >
```
Some of the nouns that undergo this stress shift may alternatively undergo other processes to mark verbalization. For instance, the denominal verb in (47a) can be alternatively realized as *sipu-ˈtâ-ma*, 'she will wear a skirt', with denominalizing suffix *-tâ*, which involves a concomitant noun truncation process (see also discussion of this process in §9.3.5). I have not detected any semantic difference between these alternative forms.

### **9.3.2 Instrumental prefixes**

A well-known phenomenon of Uto-Aztecan languages is the presence of a set of instrumental prefixes that indicate the instrument with which a transitive activity is performed, or the manner in which the activity is carried out. These prefixes (approximately 20) are reconstructed for Proto-Uto-Aztecan. A list of reconstructions is given in (48) (from Dayley 1989: 92):


(48) Uto-Aztecan instrumental prefixes

9.3 The Inner Stem


According to Langacker (1977), these instrumental prefixes are morphologically active only in the Numic and Tepiman branches of Uto-Aztecan, and have only lexicalized remnants in the rest of the Uto-Aztecan family. This is true in the case of Choguita Rarámuri. Instrumental incorporation in Choguita Rarámuri is only found in a few lexicalized items, some of which are exemplified in (49).

(49) Choguita Rarámuri instrumental prefixes


In some cases, the incorporated noun does not have a corresponding independent nominal form, as *ma-* (49a). (Lionnet 2001a: 87) identifies several instrumental prefixes in the Norogachi dialect (*ba(ʔ)-* 'water', *ku-* 'wood', *ma-* 'hand', *mo(ʔ)-* 'head', and *na-* 'fire'), but does not provide any examples. The Choguita Rarámuri verbs *moˈtot͡ʃi* 'hit (oneself) in the forehead', *moˈtépia* 'braid hair' are likely lexicalizations of verbal roots plus the instrumental prefix *mo-* 'head', but no alternations are documented to support a synchronic analysis of these forms as morphologically complex.

### **9.3.3 Body-part incorporation**

Another Inner Stem building process in Choguita Rarámuri is noun incorporation. The language has N-V constructions that are restricted to nouns referring to body parts and bodily fluids, with properties that are prototypical of "body part incorporation". This is a restricted kind of noun incorporation, which is common in languages of the Americas, including the Uto-Aztecan language family (Baker 1988). Relevant examples are shown in (50), where the body-part nouns (*moˈʔô* 'head', *buˈsi* 'eye', *roˈpâ* 'belly', etc.) precede a verbal root, the head of the incorporated construction.

(50) Body part incorporation constructions

a. *moʔoˈrêpi moʔo+ˈrêpu* head+cut 'to behead' 'cortar la cabeza' < SFH 07 1:185/ el >

### 9 Verbs and the verbal complex


Some of these constructions (e.g. (50d–f)) are fully compositional and have transparent semantics. The incorporated noun generally fills the syntactic role of object, decreasing the verb's valence. Incorporated constructions can also be externally modified, as example (51) shows.

(51) External modification of incorporated verb *ˈmá koˈsúrti ronoˈrêpili ˈmôno ˈmá* already *koˈsúriti* left *roˈnô+reˈpù-li* leg+cut-pst *ˈmôno* doll 'The doll's left leg already fell' 'Ya se le cayó la pata izquierda al mono' < SFH 07 1:186/el >

Incorporated body-part nouns, however, do not form an open class. Only a few body-part terms are found in these constructions, and not the full range

9.3 The Inner Stem

of lexical items in this semantic class. Body-part terms such as *raˈníka-la* 'heel', *koˈtʃ͡ía-la* 'eyebrow, lashes', and *loˈmíritʃ͡i* 'wrist', among others, have not yet been documented in incorporated constructions.

As for the prosodic make-up of incorporated forms, stress is assigned in the first syllable of the head of the construction, the verbal root, regardless of the underlying lexical stress information the roots might carry (see also (17) on p. 417). In forms where the first element, the noun, is disyllabic and where the second element, the head verb, has second-syllable stress in non-incorporated structures, stress surfaces on the first syllable of the second member (52).

(52) Stress shift in incorporated forms


As shown in these examples, the tonal properties of incorporated verbs are homogeneous: the stressed syllable (the first syllable of the head of the compound) bears a HL tone, regardless of what the underlying tonal properties of the members of the compound are (for more discussion, see §11.3.6).

### 9 Verbs and the verbal complex

Incorporated body-part nouns can also be truncated to fit the prosodic template of incorporated forms. In (53), the trisyllabic nouns *tʃ͡aˈmêka*, 'tongue', and *tʃ͡ereˈwá*, 'sweat', truncate their final syllable in the incorporated form (the prosodic properties of incorporated forms are addressed in more detail in §11.5.4). )

### (53) Truncation of tetrasyllabic nouns in incorporation

```
a. [tʃ͡ameˈrêpu]
    /tʃ͡aˈmêka+reˈpù/
    tongue+cut
    'to cut one's tongue'
    'cortarse la lengua' < SFH 07 1:184/el >
b. [tʃ͡ereˈbîwa]
    /tʃ͡ereˈwá+biʔwá/
    sweat+clean
    'to clean one's sweat'
    'limpiarse el sudor' < SFH 07 1:187/el >
```
Finally, body part incorporation constructions exhibit voicing alternations. In (54), the verb *paˈkó* 'to wash' has a voiced bilabial stop onset on its first syllable in incorporation.

### (54) Stop alternations in incorporated verbs

```
a. [ronoˈbâki-]
    /ronô+pakó-/
    feet+wash
    'to wash one's feet'
    'lavarse los pies' < SFH 04 CS/el >
b. [rameˈbâki-]
```
/rame+**p**akó-/ teeth+wash 'to wash one's teeth' 'lavarse los dientes' < SFH 04 CS/el >

```
c. [kupa+ˈbâki-]
    /kupá+pakó-/
    hair+wash
```
9.3 The Inner Stem

'to wash one's hair' 'lavarse el pelo' < SFH 04 CS/el >

This voicing alternation in incorporated constructions is analyzed as a morphologically-conditioned phonological effect.

### **9.3.4 Suppletion and prefixation in pluractional marking**

Choguita Rarámuri lacks inflectional agreement marking, but displays a few lexically restricted resources to mark plurality and pluractionality on both nouns and verbs. This section describes plural and pluractional marking in verbs.

Like most Uto-Aztecan languages, Choguita Rarámuri has number suppletion for certain roots, mostly positional verbs. Singular and plural agreement is expressed with intransitive patient-like subjects (55a–d) and with transitive objects (55e), in an ergative pattern.

(55) Number verb suppletion


In addition, there are two non-suppletive, semi-productive processes in Choguita Rarámuri to mark plurality. The first one is used exclusively with verbal roots and it marks subject number in an accusative pattern (S and A). When the subject is singular, a prefix *ni-* is added or the root remains unprefixed; when the subject is plural, the prefix *na-* is used (56).

### (56) Number prefixation


### 9 Verbs and the verbal complex

The examples in (56a–d) have an independent root form with no prefix for the singular, but the forms in (56e–f) are not attested alternatively without a subject number prefix.

### **9.3.5 Denominal verbs**

Choguita Rarámuri verbs may be derived through a variety of morphological mechanisms applying to nominal bases. The mechanisms to derive verbs from nominal bases include several productive verbalizing suffixes such as the *-ta* 'make' suffix, the reversive *-bu* suffix, the *-tu* 'gather' suffix and the *-ê* 'have, wear' suffix, among others. Denominal verb constructions of this kind are documented across the Uto-Aztecan language family, with meanings that include notions like 'make', 'put on', 'have', 'remove', 'use', 'get' and 'become', among others (Haugen 2008a; 2017). These constructions are productive in Choguita Rarámuri. This subsection describes each of them in detail.

### **9.3.5.1 The suffix** *-tâ* **'make/become'**

A noun derived with the verbalizer *-tâ (-râ)* suffix, a stress-shifting suffix, becomes a verb meaning 'to make/become N' (57a) or, when used with nouns referring to a piece of clothing, 'to wear N' (57b, c). It follows the nominal root forming a base to which inflectional morphology is added.


9.3 The Inner Stem

Further examples of this denominal constructions are shown in (58).

(58) Denominal verbs


When this suffix is added to trisyllabic nouns, the last syllable of the noun is truncated in order to meet the requirement of this construction to have stress on the denominalizing suffix (59) (the truncated syllable of the noun is highlighted in boldface in the phonemic representation with morpheme breaks).

(59) Noun truncation in verbalizing constructions with the *-ta* suffix *ˈhê ˈnáni siˈpútʃ͡a sipuˈtâmo ˈlá ˈhê* dem *na=ni* prox=1sg.nom *siˈpútʃ͡a* skirt *siputʃa-ˈtâ-ma* skirt-vblz-fut.sg *oˈlá* cer 'I will wear this skirt.' 'Me voy a poner esta falda.' < BFL 07 Sept 6/el >

This truncation process resembles the truncation that trisyllabic nominal roots undergo in incorporation constructions (see §9.3.3 above). Discussion of morphologically-conditioned truncation and templatic effects in Choguita Rarámuri is further discussed in §11.5.

### **9.3.5.2 The reversive suffix** *-bû*

The suffix *-bû*, on the other hand, has a "reversive" meaning and it derives verbs from nouns is a non-productive affix that has the meaning 'remove' or 'undo'.<sup>16</sup> Some examples of this derivational suffix are provided in (60).

(60) a. *niˈhê toˈlí boʔoˈbûma niˈhê* 1sg.nom *toˈlí* chicken *boʔo-ˈbû-ma* feather-rev-fut.sg 'I will pluck the chicken.' 'Voy a desplumar la gallina.' < SFH 08 1:51/el >

<sup>16</sup>A cognate suffix can be found in Guarijío, and Miller identifies the verb *puha* 'to take away', as its source (1996: 151). The Rarámuri cognate of this verb is *bu(ʔ)è* 'to take away'.

9 Verbs and the verbal complex

	- *ˈmá* already *tʃ͡omo-ˈbû-ka!* mucus-rev-imp.sg 'Clean your nose already!' 'Ya límpiate la nariz!' < SFH 08 1:51/el >

### **9.3.5.3 The 'gather' suffix** *-tú*

The suffix *-tú (-rú)*, which means 'to gather' or 'to bring', is another derivational suffix that attaches to nominal roots to form verbs. It is an unproductive suffix restricted to attach to only a few nouns. It is always stressed, it occurs next to the nominal root. As part of the Inner Stem, it is sensitive to the morphophonological effects that characterize this stem domain, including compensatory lengthening. Some examples are given in (61).

	- b. *mati ilaˈrúpo ma=ti* now=1pl.nom *ila-ˈrú-po* cactus-gather-fut.pl 'Let's gather cactus now.' 'Vamos juntando nopales.' < SFH 08 1:52/el >
	- c. *mi riʔˈri naʔiˈrúmisa mi* dem *riʔˈri* there *naʔi-ˈrú-mi-sa* fire-gather-mot.imp-imp.sg 'Go get fire over there!' '¡Ve a traer lumbre allá!' < SFH 08 1:52/el >

9.3 The Inner Stem

d. *ˈhípi oˈmêatʃ͡i rakiˈrúpa ˈlá ˈhípi* today *oˈmêatʃ͡i* Sunday *rakiˈ-rú-pa* palm-gather-fut.pass *oˈlá* cer 'Palms will be gathered today, Sunday.' 'Hoy domingo van a recibir la palma.' (lit. 'las palmas serán juntadas') < SFH 08 1:52/el >

### **9.3.5.4 The 'have'** *-ê* **suffix**

The *-ê* suffix is a derivational suffix that may be characterized as a "denominal verb of possession". I follow Haugen (2017) in defining "denominal verb of possession" as "a construction where an identifiable noun root (or stem) N can be converted into a verb by the [affixation] ... of some affix X such that the N+X complex is treated as a regular verb, able to be inflected with markers of tenseaspect-mood and other categories, and having the meaning of 'to have N'" (2017: 164).

The *-ê* suffix, which has the general meaning of 'to have' or 'to wear', is a replacive suffix. It targets the nominal stem final vowel, and it is always stressed.<sup>17</sup> The examples in (62) show this derivation.<sup>18</sup>

### (62) Verbalizer suffix *-ê*


<sup>17</sup>The cognate of this suffix (*-e*) in Mountain Guarijío is described in Miller (1996) as exhibiting different prosodic properties, yielding variable stress patterns in the morphologically complex words where it is attested (with stress on the stem or the suffix) and not replacing the final stem vowel (e.g., *puhkú-e* ~*puhku-é* 'to have an animal' (1996: 149)).

<sup>18</sup>For discussion of verbless possessive constructions in Yaqui (Taracahitan), see Jelinek & Escalante (2010).

### 9 Verbs and the verbal complex

As shown in these examples, while the suffix is always stressed, there are no homogeneous tonal properties that the morphologically complex verbs containing this verb share (i.e., each of the three lexical tones is attested in these cases, in what appears to be a lexically conditioned pattern).

Further examples of this construction are shown in (63).


This construction is highly productive. There are no apparent restrictions in terms of the semantic properties of the nouns that can undergo this derivation.

### **9.3.5.5 The verbalizer** *-wi* **suffix**

In addition to the verbalizer *-ê* suffix, some nominal stems attach the neutral  *wi* suffix to derive a verb with the meaning 'to own/have N'. Examples of this construction are shown in (64).

### 9.3 The Inner Stem

### (64) Verbalizer suffix *-wi Verb stem Gloss Noun stem Gloss* a. winoˈmí-wi 'to have money' winoˈmí 'money' < BFL 06 5:127/el > b. waˈsá-wi 'to have land' waˈsá 'land for cultivation' < BFL 06 5:127/el > c. ˈkútʃ͡u-wi 'to have children' ˈkútʃ͡i 'little ones' < BFL 06 5:127/el > d. suˈnù-wi 'to have corn' suˈnù 'corn' < BFLL 06 5:127/el > e. muˈnî-wi 'to have beans' muˈnî 'beans' < BFL 06 5:127/el > f. rihiˈmá-wi 'to have relatives' rihiˈmá 'relatives' < BFL 06 5:128/el > g. koˈbîsu-wi 'to have pinole' koˈbîsi 'pinole' < BFL 06 5:128/el > h. aˈʰkà-wi 'to have sandals' aˈʰkà 'sandals' < BFL 06 5:128/el > h. ˈpúru-wi 'to have a knitted ˈpúra 'knitted belt' belt' < BFL 06 5:128/el > i. koˈmâru-wi 'to have a comadre' koˈmâare 'comadre' < BFL 06 5:128/el >

This construction is also highly productive in the language.

A potentially related morphological strategy to derive denoninal verbs with the meaning 'have' or 'own' is through the suffixation of *-i*, which replaces the final vowel of the stem, which may be a reduced form of the verbalizer *-wi* suffix. Relevant examples are shown in (65).



The verbalizer *-wi* suffix and other verbalizing morphological markers may be available with the same nominal stems. In some cases, the choice of marker involves a semantic difference. This is illustrated with the minimal pair in (66) where choice of marker entails a different semantic interpretation.

### 9 Verbs and the verbal complex

### (66) Semantic differences of denominal verbs

```
a. saʔˈpê
    saʔˈp-ê
    meat-vblz
    'to have flesh (one's own)'
    'tener carne (del cuerpo)' < BFL 06 5:129/el >
b. saʔˈpáwi
    saʔˈpá-wi
```

```
meat-vblz
'to have (e.g. cow's) meat'
'tener carne (e.g., de res)' < BFL 06 5:129/el >
```

```
c. cf. saʔˈpá
    'flesh'
    < BFL 06 5:129/el >
```
As these examples show, the choice of verbalizing suffix (*-ê* in (66a) and *-wi* in (66b)) may involve a semantic distinction. Specifically, the choice of verbalizer in this case entails a distinction in terms of an inalienable reading vs. an alienable one, respectively (for more discussion of alienable/inalienable distinctions and their morphological expression in Choguita Rarámuri, see §8.4.1). However, semantic distinctions of this kind are not widespread, which suggests that these derived forms have undergone some degree of lexicalization.

Finally, there are a few constructions that may also encode the meaning 'to own/have N' through a periphrastic construction, in addition to a suffixing construction. This periphrastic construction is exemplified in (67):

```
(67) a. waˈsá
            cultivation.land
                             ˈníwi
                             have
           'to have land'
           'tener tierra de cultivo'
       b. ˈsôda
            soda
                 ˈníwi
                 have
```
'to have soda' 'tener soda'

As shown in these examples, the construction involves a noun followed by the verb *ˈníwi* 'to have'. This auxiliary verb is the likely source of the *-wi* verbalizer suffix (the language also has a related appositive possessive construction, described in §2.7 below).

9.3 The Inner Stem

### **9.3.5.6 The verbalizer** *-pi* **suffix**

Another verbalizer suffixing construction is a derivational suffix, *-pi*, which means 'to remove'. This suffix is exemplified in (68).


This suffix is unproductive in Choguita Rarámuri.

### **9.3.5.7 Non-concatenative verbalizing constructions**

There are a few cases where denominal verbs meaning 'make/wear N' are derived through a stress shift. Specifically, the derived verb will have stress one syllable to the right with respect to the stress location in the related nominal base. This is shown in the example in (69):

(69) Derivational stress shift

a. napaˈtʃ͡â 'to wear a shirt' b. naˈpátʃ͡a 'shirt' < BFL 06 5:128/el >

Denominalization may also involve truncation of the last, unstressed syllable of the base noun even with no overt attachment of any suffix. This is shown in the examples in (70):

(70) Stem truncation in denominalization


These non-concatenative processes to derive verbs from nouns are unproductive in the language.

### 9 Verbs and the verbal complex

### **9.3.6 Summary**

Verbal roots in Choguita Rarámuri may undergo semi-productive and unproductive processes, both concatenative and non-concatenative, before adding any further suffixes. These processes include conversion, pluractional consonant mutation, stress shifts and grammatical tone to mark imperative or verbalization or body-part incorporation. Having described root classes and the processes taking place at the innermost level of the verbal stem, I now turn to the suffixation domain.

### **9.4 Verbal structure and verbal domains**

### **9.4.1 Overview**

The suffix positions and categories expressed in the Choguita Rarámuri verbal structure are schematized in Table 9.8.


Table 9.8: Suffix positions and categories of the Choguita Rarámuri verb

The suffixes in each position do not generally co-occur in the same word, due to their semantic incompatibility (though there are systematic exceptions; these will be discussed in this chapter and the rest of this grammar). A summary of the verbal suffixes and their position in the verbal template is given in Table 9.9. A basic description and examples of the suffixes can be found in Appendix A. The

### 9.4 Verbal structure and verbal domains

"Reference" column in Table 9.9 refers to the section where each individual suffix is described in this Appendix.

This verbal structure does not imply a slot-and-filler, template-like structure, i.e. the labels S1…S12 are not intended to imply a flat structure as in a slot matrix. In the verbal morphology of this language, morphotactic, and morphophonological processes define a hierarchical structure of the verb, with suffixes closer to the Inner Stem displaying less salient morpheme junctures (given by phonological transparency and productivity). In the structure represented in Table 9.10, I identify five verbal zones after the Inner Stem domain: a Derived Stem, a Syntactic Stem, an Aspectual Stem, a Finite Verb domain, and finally a Subordinate Verb domain.

The first identifiable layer in the suffixation domain is the Derived Stem. This layer of the verbal stem includes two kind of semantically restricted, unproductive derivational suffixes (an inchoative suffix and three transitive suffixes). These suffixes are restricted to a semantically defined class of verbs, change-ofstate verbs. There are three morphological and morpho-phonological criteria that allow identifying this stem domain as a defined sub-constituent of the Choguita Rarámuri verb: first, the non-concatenative imperative singular (described in §9.3.1.3) is marked as final stress of this stem domain; second, this level is also the domain of the passive-induced lengthening (discussed below); finally, the Derived Stem undergoes the stress shift that characterizes unstressed stems when combined with shifting suffixes (addressed in §5.3).

The Derived Stem is the input to the next stage of the construction of the Choguita Rarámuri verb, the Syntactic Stem. This next stem domain includes suffixes in S3–S5, suffixes that mark valence-increasing operations. Within this stem domain, suffixes are attested in variable order and display multiple (or extended) exponence. The suffixes in this level also form a coherent domain within the Choguita Rarámuri verb in morpho-phonological terms: these suffixes are stress-neutral, forming a small pocket of unstressable suffixes within a larger, stressable domain. Finally, this stem domain is part of the domain of round harmony, as defined in §9.4.3.

Another layer of the verbal stem is the Aspectual Stem, composed of suffixes in positions S6 to S9, marking desiderative, associated motion, and auditory evidential. These suffixes are formally and semantically related to independent verb forms in the language. Table 9.11 lists these suffixes, their grammaticalized meanings and their independent lexical verb sources.

These aspectual markers exhibit the same properties as constructions described for other Uto-Aztecan languages as "secondary verb" constructions, where grammaticalized formatives derived from independent verbs that encode aspectual-

### 9 Verbs and the verbal complex

### Table 9.9: Choguita Rarámuri verbal suffixes


### 9.4 Verbal structure and verbal domains


### Table 9.10: Choguita Rarámuri verbal stem domains

Table 9.11: Choguita Rarámuri aspectual suffixes and their lexical counterparts


like or adverbial-like meanings (Crapo 1970), a type of V-V incorporation construction analyzed as involving light verbs in Hiaki (Taracahitan; Tubino Blanco et al. 2014) and serialization in Northern Paiute (Numic; Thornes 2011) (the syntactic properties of these constructions are discussed in §15.6.4). These aspectual suffixes in Choguita Rarámuri are disyllabic and have monosyllabic allomorphs and they are integrated prosodically with the stem in a single phonological word. The phonological factors determining the distribution of dysillabic and monosyllabic allomorphs of these suffixes are discussed in Chapter 11 (§11.5). These suffixes are also part of the domain for round harmony and constitute the last layer where this process applies (that is, suffixes to the right of this domain are not targets for spreading of the harmony).

### 9 Verbs and the verbal complex

The final stage in constructing a Choguita Rarámuri inflected verb consists in adding the suffixes in positions S9–S11 in the finite verb, the Finite Verb level suffixes. The grammatical categories marked at this level are mood distinctions (including imperative and reportative), voice, tense, and aspect (and number and person marginally), conflated in portmanteaux suffixes. In this domain there are inflectional affixes that produce an idiosyncratic meaning when combined.

Finally, a finite verb can be the input for another, optional layer of morphology, in order to be used in subordinate clause constructions. These suffixes, in position S12, are the last stage of affixation of the Choguita Rarámuri verb, and are stress neutral. They produce nominalizations, and are at the word boundary.

Table 9.12 summarizes the linear order facts, and the semantic, morphophonological and prosodic properties of affixes that have motivated the verbal zones proposed for the Choguita Rarámuri verb.

The evidence for the proposed structure will be laid out as follows. In §9.4.2, I present the morphotactic evidence for positing the positions in the verbal template, and in §9.4.3, I give the morpho-phonological evidence for positing different verbal domains.

### **9.4.2 Morphotactic evidence for affix ordering generalizations**

This section provides the morphotactic evidence for positing the suffix positions or slots in the verbal template. This evidence comes from the linear ordering properties of the suffixes, as well as their exponence and permutation possibilities. The evidence is presented progressively describing the positions from the Inner Stem towards the outer layers of affixation.

The positions closest to the Inner Stem in the verbal template are occupied by suffixes that are only used with change-of-state predicates: the inchoative suffix (in S1) and a set of transitive suffixes in (S2). Their ordering is illustrated in (71).

(71) inch (S1) – tr.pl (S2)



Table 9.12: Characteristics of the Choguita Rarámuri verb

c. *ˈnè ˈmá aʰkaˈbátʃ͡ili kaˈhê ˈnè* 1sg.nom *ˈmá* already *aʰka-ˈbá-tʃa-li* sweet-inch-tr.pl-pst *kaˈhê* coffee 'I already sweetened the coffee.' 'Ya endulcé el café.' < BFL 08 1:20/el > d. *ˈnè ˈmí baʔˈwí rataˈbátʃ͡kira ˈnè* 1sg.nom *ˈmí* 2sg.acc *baʔˈwí* water *rata-ˈbá-tʃa-ki-ra* heat-inch-tr-appl-pot

'Shall I heat the water for you?'

'¿Te caliento el agua?' < BFL 08 1:21/el >

Position S3 is occupied by a set of applicative suffixes. These applicative suf-

### 9 Verbs and the verbal complex

fixes might have encoded semantic differences in a previous diachronic stage,<sup>19</sup> but synchronically they are lexically selected by the roots to which they attach. For instance, the applicative suffix *-ni* is the only applicative suffix that can be attached to bases derived with transitive suffix *-bû* (e.g., (72a–b)). This applicative suffix is also attested with stems derived through other transitive suffixes (72c).

### (72) tr (S2) – appl (S3)


```
ˈwé
int
    ta
    1pl.nom
              raki-ˈbû-ni-bo
              push-tr-appl-fut.pl
'Let's push it for him.'
```

```
'Vamos empujándoselo.' < SFH 05 1:61/el >
```
c. *riʔoˈnânima*

```
riʔo-ˈnâ-ni-ma
sandpaper-tr-appl-fut.sg
'They will sandpaper (the wood).'
'Se lo va a lijar (la madera).' < SFH 05 1:175/el >
```
The applicative suffixes in slot S3 are in turn followed by a productive causative suffix in slot S4. In (73), applicative suffixes *-ni, -si* and *-wi* preceed the causative suffix *-ti*.

(73) appl (S3) – caus (S4)

a. *gaˈbriêlo suˈwíwtima ˈlé ba gaˈbriêlo* Gabriela *suˈwí-wi-ti-ma* finish.up.appl-appl-caus-fut.sg *aˈlé* dub *ba* cl 'Gabriela will make her finish up (his tortillas).' 'Gabriela va a hacer que ella se las acabe (sus tortillas).' < BFL 08 1:27/el >

<sup>19</sup>Some semantic differences are still retained in cognate suffixes in the closely related Mountain Guarijío (Miller 1996).

9.4 Verbal structure and verbal domains

b. *ˈnè a ˈmí ˈʃûntikisa ˈró ˈnè* 1sg.nom *a* aff *ˈmí* 2sg.acc *ˈsû-ni-ti-ki-sa* sew-appl-caus-appl-cond *ˈró* q 'What if I made you sew her a skirt?' '¿Qué tal si te hago coserle una falda?' < BFL 08 1:28/el > c. *to jaˈdîra ˈpáʃtiri bo! to* exh *jaˈdîra* Yadira *ˈpási-ti-ri* throw-appl-caus-imp.sg *bo!* exh 'Let's see, throw it to Yadira!' '¡A ver, tíraselo a Yadira!' < BFL 08 1:28/el > d. *oˈʃìwtimo ˈlá ˈnè ˈjéla oˈsì-wi-ti-ma* write-appl-caus-fut.sg *oˈlá* cer *ˈnè* 1sg.nom *ˈjé-la* mom-poss 'She'll make him write my mom (a letter).' 'Va a hacer que le escriba a mi mamá.' < BFL 08 1:28/el >

The productive causative suffix *-ti* appears ordered before the productive applicative suffix *-ki* (as shown in (74)), motivating a slot S5 for the applicative suffix.

```
(74) caus (S5) – appl (S6)
```



### 9 Verbs and the verbal complex

While the Causative-applicative (*-ti-ki*) order is the most commonly attested, these suffixes can also permutate their order. In the examples in (75) the applicative suffix *-ki* is ordered before the causative suffix *-ti*.

	- a. *tòo, ˈjêni ˈdûlse ˈìwkitiri jadîra tòo,* go! *ˈjêni* Yeni *ˈdûlse* candy *ˈìwi-ki-ti-ri* bring.appl-appl-caus-imp.sg *jadîra* Yadira 'Make Yeni bring candy for Yadira!' '¡Ve, haz que Yeni le traiga dulces a Yadira!' < BFL 07 1:62/el >
	- b. *to miˈtʃ͡íktiri bo to* exh *miˈtʃ͡í-ki-ti-ri* carve-appl-caus-imp.sg *bo!* exh 'Carve it for him' '¡Lábraselo!' < BFL 08 1:107/el >
	- c. *ˈnè tʃ͡oʔˈmá biʔˈwìktimo ˈlá tiˈwé ˈnè* 1sg.nom *tʃ͡oʔˈmá* mucus *biʔˈwì-ki-ti-ma* clean-appl-caus-fut.sg *oˈlá* cer *tiˈwé* girl 'I'll make her clean the girl's nose.' 'La voy a hacer que le limpie los mocos a la niña.' < BFL 08 1:55/el >

The applicative-causative suffix order, as exemplified in (75), is marginally attested in the Choguita Rarámuri corpus (the factors that may condition this affix permutation pattern are discussed in Caballero 2010). The overwhelming preference for the causative-applicative order motivates positing a separate slot, S5, for the productive applicative suffix *-ki*, separate from the rest of the applicative suffixes in S3. An additional argument in favor of keeping the two applicative positions distinct comes from the difference in productivity between the inner applicatives and the later, productive suffix *-ki*.

The applicative suffix is in turn followed by the desiderative suffix *-nále* (in slot S6). This suffix is exemplified in (76).

(76) appl (S5) – desid (S6)

a. *ˈnè mi biˈlé ˈwàsi miʔˈríkinili muˈhê oˈmáwaratʃ͡i ˈnè=mi* 1sg.nom=2sg.acc *biˈlé* one *ˈwàsi* cow *miʔˈrí-ki-nale* kill-appl-desid *muˈhê* 2sg.nom *oˈmáwaratʃ͡i* party 'I want to kill one cow for you, for your party.' 'Quiero matar una vaca para ti (para tu fiesta).' < SFH 07 2:65-66/el > 9.4 Verbal structure and verbal domains


The desiderative suffix *-ˈnále* is then followed by the associated motion *-simi*, as exemplified in (77). Further discussion and analysis of the nature of the shortlong alternation for this and other suffixes is provided in Chapter 11 (§11.5.4).

(77) desid (S6) – mot (S7)


### 9 Verbs and the verbal complex

d. *ˈmán ˈnè t͡ʃaˈkéna ˈwánsia iˈnâli riˈʰtê ˈmá=n* already *ˈnè* 1sg.nom *t͡ʃaˈkéna* aside *ˈwá-nale-simi-a* throw-desid-mot-prog *iˈnâli* go *riʰˈtê* stone 'I go along wanting to throw away the stones.' 'Ya me dan ganas de ir quitando las piedras.' < BFL 08 1:88/el >

The desiderative and associated motion suffixes are also attested in the inverse order, as shown in (78).

### (78) mot-desid affix order


*aˈwí-simi-nale-i* dance-mot-desid-impf 'She wanted to go along dancing.' 'Quería irse bailando.' < SFH 07 2:72-73/el >

c. *ˈnèn naˈlàsnila iˈnâli ˈnè=n* int=1sg.nom *naˈlà-simi-nale-a* cry-mot-desid-prog *iˈnâli* go 'I'm going along feeling like crying.' 'Voy queriendo llorar.' < BFL 08 1:89/el >

d. *ˈá biˈlá taˈmí ˈjòrsinira ˈruá*

*ˈá* aff *biˈlá* really *taˈmí* 1sg.acc *ˈjò-ri-simi-nale-ra* mad-caus-mot-desid-rep *ru-ˈwá* say-mpass 'He says he wants to go along making me mad.' 'Dice que me quiere ir haciendo enojar.' < SFH 08 1:72/el >

The next position is occupied by the auditory evidential suffix *-tʃ͡ane*, another disyllabic suffix that is transparently related to an independent verb in the language, *(a)tʃ͡ane*, 'say, sound like'. The following examples show the evidential suffix ordered after the associated motion suffix (79) and the desiderative suffix (80).

(79) mot (S7) – ev (S8)

a. *wikuˈwâstʃ͡ina wikuˈwâ-simi-tʃane-a* whistle-mot-ev-prog 'It sounds like they are going around whistling.' 'Se oye que van chiflando.' < SFH 07 2:74, el492/el > b. *ˈwîstʃ͡ane ˈwî-simi-tʃane* harvest-mot-ev 'It sounds like they are going along harvesting.' 'Se oye que van pizcando.' < SFH 08 1:132/el > c. *ˈá biˈlá uˈbârstʃ͡ani ˈá* aff *biˈlá* really *uˈbâ-ri-simi-tʃane* bathe-caus-mot-ev 'It sounds like they are going along bathing them.' 'Se oye que van bañándolos.' < SFH 08 1:150/el > d. *ˈá biˈlá ˈwé ˈàastʃ͡ani wikoˈki uʔˈpá ˈá* aff *biˈlá* really *ˈwé* int *ˈà-simi-tʃane* look.for-mot-ev *wikoˈki* mushrooms *uʔˈpá* back 'It sounds like they are going along looking for mushrooms back there.' 'Se oye que van buscando hongos atrás.' < SFH 08 1:145/el > (80) desid (S6) – ev (S8) a. *wikaˈrântʃ͡ane wikaˈrâ-nale-tʃane* sing-desid-ev

'It sounds like they want to sing.'

'Se oye como que quieren cantar.' < SFH 07 1:9/el >

b. *ˈmá koʔ-ˈnáltʃ͡ani*

*ˈmá* already *koʔ-ˈnále-tʃane* eat-desid-ev 'It seems they already want to eat.' 'Como que ya quieren comer.' < SFH 08 1:124/el >

### 9 Verbs and the verbal complex

c. *roˈkò mi boˈʔóri ˈnà haˈré moʔiˈnáaltʃ͡ini roˈkò* night *mi* there *boˈʔóri* up.there *ˈnà* there *haˈré* some *moʔi-ˈnále-tʃane-i* enter.pl-desid-ev-impf 'Last night it sounded like theey wanted to go inside up there.' 'Anoche se oía que querían entrar allá arriba.' < SFH 08 1:124/el >

The desiderative and evidential suffixes, too, can appear in the inverse order, as shown in (81), due to factors discussed in more detail in Caballero (2010). 20

(81) Evidential-desiderative affix order

a. *oʔˈpéstʃ͡analo oʔˈpés-tʃane-nale-o* vomit-ev-desid-ep 'It sounds like they want to vomit.' 'Se oye como que quieren vomitar.' < BFL 07 rec300/el > b. *paraˈértʃ͡analo paraˈér-tʃane-nale-o* dance.paraeri-ev-desid-ep 'It sounds like they want to dance paraéri.' 'Se oye como que quieren bailar paraéri.' < BFL 07 1:182/el > c. *ˈá biˈlá ˈt͡ʃîkle ˈkétʃ͡itʃ͡anili ˈkûruwi ˈá* aff *biˈlá* really *ˈt͡ʃîkle* gum *ˈkétʃ͡i-tʃane-nale* chew-ev-desid *ˈkûruwi* kids 'It seems like the kids want to chew gum.' 'Se oye que los niños quieren mascar chicle.' < SFH 08 1:146/el > d. *ˈnápi ˈlé aˈtístʃ͡anala ˈnápi* rel *aˈlé* dub *aˈtísi-tʃane-nale-a* sneeze-ev-desid-prog 'It is like somebody wants to sneeze (it sounds like it).'

'Como que se oye que quieren estornudar.' < SFH 08 1:122/el >

<sup>20</sup>In (81a–b), the last vowel of the desiderative suffix is replaced by the epistemic suffix *-o*, that encodes an event is the result of a psychological or mental state, in this case, 'want'. Epistemic marking in desiderative-marked verbs is attested in the speech of some speakers (e.g., BFL in (81a–b), but not others (e.g., SFH in (81c–d)). More details about this suffix can be found in Appendix A.5.2.

### 9.4 Verbal structure and verbal domains

The desiderative, associated motion and evidential suffixes precede a set of stress-shifting suffixes, posited to occupy slot (S9). All of the suffixes in this position are suffixes encoding both voice and tense/aspect. A list is provided in (82) (more details about each individual suffix can be found in the Appendix in the cross-referenced sections).<sup>21</sup>



As shown in the next examples, the desiderative suffix (83), the associated motion suffix (84) and the evidential suffix (85) all precede these inflectional suffixes, which are more peripheral in the verbal template.

### (83) desid (S6) – fut.sg (S9)


<sup>21</sup>The motion imperative suffix and irrealis sigular suffix are homophonous, but distinct suffixes in the language.

### 9 Verbs and the verbal complex

### (84) mot (S7) – fut.sg (S9)


(85) ev (S8) – mpass (S9) *ˈnè iˈtʃ͡ìirtʃ͡unua /ˈnè* 1sg.nom *itʃ͡-ì-ri-tʃane-wa* plant.appl-appl-ev-mpass 'It sounds like (corn) is being planted for me.' 'Se oye como que me están sembrando maíz.' < SFH 07 1:10/el >

The evidential suffix can also appear ordered after these suffixes, under circumstances described in Caballero (2010). The examples below show the evidential suffix preceeded by the stressed allomorph of the future singular suffix (86a–b) and by the habitual passive suffix (86c).

### (86) Future/Habitual Passive – Evidential affix order


9.4 Verbal structure and verbal domains

*ˈnápi* rel *aˈlé* dub *naˈkó-ma-tʃane-a* fight-fut.sg-ev-prog *ˈwàsi* cows 'It sounds like the cows are going to fight.' 'Se oye como que las vacas se van a pelear.' < SFH 07 1:140/el > c. *ˈnápi riˈménuwatʃ͡ana ˈnápi* rel *riˈmé-nu-wa-tʃane-a* make.tortillas-appl-mpass-ev-prog 'It sounds like they are making him tortillas.' 'Como que se oye que le están haciendo tortillas.' < SFH 07 2:69/el >

Finally, there is another slot of stress-shifting affixes that mark mood (potential (pot) *-ra*, imperative singular (imp.sg) suffixes *-kâ* and *-sâ*, and imperative plural (imp.pl) *-sì*). In (87), the potential suffix and the imperative singular suffix *-sa* are ordered after the motion imperative suffix (in S8):<sup>22</sup>

(87) mot.imp (S9) – Mood (S10)

```
a. ˈjurka osiˈmêra ˈlé
     ˈjuri-ka
     take-imp
              osi-ˈmê-ra
              write-mot.imp-pot
                                    aˈlé
                                    dub
    'Go, take him to see if he writes.'
    'Ve y llévalo a ver si escribe.' < BFL 08 1:94/el >
b. ˈâamsa
```

```
ˈâ-me-sa
give-mot.imp-imp.sg
'Go give it to her!'
'¡Ve y dáselo!' < ROF 04 1:112/el >
```
As shown so far, there is a fair amount of morphotactic evidence for positing the positions of a complex verbal template in Choguita Rarámuri. The evidence laid out in this subsection involves attested linear ordering of suffixes. There are, however, two other important morphotactic phenomena in the Choguita Rarámuri verb, namely variable order of suffixes and multiple (or extended) exponence.

Several of the examples above show that several suffixes do not have a fixed order with respect to other suffixes, in interactions that are specific to defined

<sup>22</sup>There is no example that demonstrates the relative ordering between the evidential suffix (posited in S9) and the mood suffixes in (S10). The evidential is semantically incompatible with at least the imperative mood.

### 9 Verbs and the verbal complex

pairs of suffixes belonging to the Syntactic Stem and Aspectual Stem domains of the verb. Suffixes that might switch their order include: causative and applicative; desiderative and associated motion; and desiderative and evidential.

In addition, the suffixes in the Syntactic Stem can display multiple exponence, i.e., they can be multiply marked without an equivalent semantic recursivity. This is exemplified in (88).

(88) Causative and Applicative Multiple Exponence



Variable suffix ordering and multiple exponence might render the verbal structure proposed in Table 9.8 a highly abstract representation. This structure, however, will be retained as a descriptive device since variable orders of suffixes are restricted to specific pairs of suffixes. The applicative, for instance, while variably ordered with respect to the causative, has a fixed position preceding the desiderative suffix and the associated motion suffix, among others. Moreover, variable suffix ordering and multiple exponence are only found with suffixes that belong to particular layers or domains in the verb.

### 9.4 Verbal structure and verbal domains

There is other evidence showing that the concatenation of the Choguita Rarámuri verb is not unidimensional, and that there is an internal organization or hierarchy of processes. The next section will address the morphologicallyconditioned phonology that make suffixes in the inner layers of the verb more tightly fused with the root than outer, inflectional suffixes.

### **9.4.3 Phonological transparency and morpheme boundary strength**

The Choguita Rarámuri verb displays a nested structure that can be characterized in terms of the semantics and overall function of clusters of suffixes (valenceincreasing, aspectual, etc.), and the morphologically-conditioned phonology that yields different degrees of morpho-phonological fusion between suffixes. This sub-section is concerned with the latter phenomena.

The phonological phenomena discussed in this section are root-suffix haplology (§9.4.3.1), compensatory lengthening (§9.4.3.2), past-passive induced lenthening (§9.4.3.3), imperative stress shift (§9.4.3.4), round harmony (§9.4.3.5), and the distribution of stress-shifting and stress-neutral suffixes in the verb (§9.4.3.6).

### **9.4.3.1 Stem-suffix haplology**

We have seen that stress-conditioned vowel deletion results in derived consonant clusters and geminates in Choguita Rarámuri (see §5.2.2 and §4.2). As highlighted above, derived geminates are subject to inter-speaker (and to a lesser extent, intra-speaker) variation. The alternative to having a derived geminate is to have syllable deletion in avoidance of adjacent identical syllable onsets. In this language, haplology between a final syllable in the Inner Stem and a following suffix syllable with identical onsets takes place in morphologically complex constructions. In (89a,c,e), the root's underlying final, unstressed syllable and the immediately adjacent suffix syllable have identical onsets, leading to deletion of one syllable. The examples in (89b,d) show how the root's final syllable is not deleted in other morphological constructions. The forms in (89a,c,e) show unattested, hypothetical forms with adjacent root and suffix syllables with identical onsets.

(89) Root-suffix haplology

a. [aˈsíisa] /aˈsísi-sa/ wake.up-cond

### 9 Verbs and the verbal complex

```
'if s/he wakes up'
     'si se despierta' < BFL 08 1:1/el >
     *aˈsísi-sa
b. [aˈsísma]
     /aˈsísi-ma/
     wake.up-fut.sg
     's/he will wake up'
     'se va a despertar' < BFL 08 1:1/el >
c. [sutuˈbétʃ͡ini]
     /sutuˈbétʃ͡i-tʃ͡ane/
     trip-ev
     'it sounds like they are tripping'
     'suena que se tropiezan' < SFH 07 1:143/el >
     *sutuˈbetʃ͡itʃ͡ini
d. [sutuˈbétʃ͡inili]
     /sutuˈbétʃ͡i-nale/
     trip-desid
     's/he is about (wants) to trip'
     'se quiere tropezar' < BFL 07 1:138/el >
e. [sikoˈráanili]
     /sikoˈrána-nale/
     eye.secrete-desid
     's/he is about (lit. wants) to have an eye secretion'
     'quiere lagañear' < BFL 08 1:1/el >23
     *sikoˈrananale
```
Example (90a) shows how haplology also targets suffixes belonging to the Inner Stem domain:

(90) Inner Stem suffix-suffix haplology

a. [tʃ͡aˈbóopo] /tʃ͡aˈbó-**pi**-po/

<sup>23</sup>The Spanish translation of this verb form is 'le quieren salir lagañas, quiere lagañear', which is translated to English as 'it is imminent that s/he will have eye secretion'.

9.4 Verbal structure and verbal domains

beard-rev-fut.pl 'They will remove their beards.' 'Se van a quitar la barba.' < SFH 08 1:5/el > b. [tʃ͡aˈbópili] /tʃ͡aˈbó-**pi**-po/ beard-rev-pst

'They removed their beards.' 'Se quitaron la barba.' <SFH 08 1:5/el >

No haplology takes place between identical syllable sequences within roots (e.g. the form *\*aˈsí-ma* /aˈsísi-ma/, with the expected reading 'S/he will wake up' ('wake.up-fut.sg') in (89b), is unattested). Syllables with identical onsets belonging to two suffixes can optionally undergo deletion, as shown in (91)).

### (91) Optional suffix-suffix haplology

a. [raʔaˈmânkiki]

/raʔaˈmâ-na-ki-ki/ advise-desid-appl-pst.ego 'I wanted to advise them.' 'Quise aconsejarlos.' < BFL 06 5:132/el >

b. [miˈtʃ͡íiki] /miˈtʃ͡í-**ki**-ki/ carve-appl-pst.ego 'I carved it for them.' 'Se lo labré.' < BFL 08 1/el >

Similar cases in the literature are treated as instances of a "Repeated Morpheme Constraint" (Menn & MacWhinney 1984), a phenomenon where sequences of morphemes that are homophonous are prohibited. The Choguita Rarámuri case could be analyzed to instantiate this phenomenon, even though it involves only phonologically similar sequences of morphemes (for an overview of this phenomenon and theoretical implications, see Inkelas (2014)).

### **9.4.3.2 Compensatory lengthening**

As discussed in Chapter 3 and further addressed in Chapter 11 below, there is no evidence of contrastive vowel length in Choguita Rarámuri. Surface long vowel sequences, however, are not uncommon and are salient acoustically in this language. There are several processes that yield these vowel sequences in

### 9 Verbs and the verbal complex

surface representations. One of such processes is a word minimal size constraint affecting verbs (§11.1). Another source for surface vowel length is compensatory lengthening (CL), the phenomenon whereby the deletion of one element triggers a corresponding lengthening of another element.<sup>24</sup> I address this process in this section.

The more widespread CL pattern in Choguita Rarámuri involves deletion of a vowel that triggers lengthening of a preceding syllable's stressed vowel. This is exemplified in (92).

(92) ˈCVCV > ˈCV:C


'vacilar con cuñados'

<sup>24</sup>CL has been treated as the transfer or preservation of a phonological unit, i.e. a mora, within a prosodic unit in the phonological literature (Hyman 2003, McCarthy & Prince 1986, inter alia), or as a phonetically-based process that results from isochrony, the preservation of phonetic duration (Timberlake 1983, Barnes & Kavitskaya 2000). I assume that while this process was based phonetically, it is now part of the lexical phonology of the language.

<sup>25</sup>This verb is translated into Spanish by Choguita Rarámuri speakers as 'vacilar con los cuñados'. This verb more accurately refers to a very specific kind of social interaction that involves joking playing around with a sister- or brother-in-law, a register documented also in Mountain Guarijío Miller (1996).

### 9.4 Verbal structure and verbal domains

Vowel deletion in these contexts occurs due to posttonic syncope in derived environments (described in more detail in §5.2.2). Some of the examples in (92) are given in context in (93) below. In these cases, CL takes place when the intervening consonant is a sonorant. Lengthened vowels are underlined in the surface form and deleted vowels are in bold face in the underlying representation.

### (93) CL with intervening sonorant



Cases of CL triggered by deletion of a whole syllable have not been reported or even mentioned, to the best of my knowledge, as a logical type of CL. A second pattern of apparent vowel CL in Choguita Rarámuri, however, involves precisely the deletion of a syllable. In (94a), the tetrasyllabic root *nabiˈsûri* truncates the final syllable when attaching the disyllabic desiderative suffix *-nále*. The result is a stem with a long stressed vowel. There are no other potential sources for lengthening in this case (such as passive-conditioned lengthening or vowel loss), so the lengthening must be attributed to syllable deletion. CL takes place with an intervening voiceless affricate (94c), and an intervening voiceless fricative (94e). Below each example of syllable-triggered CL includes a related form with no deletion.

9 Verbs and the verbal complex

	- a. [nabiˈsûunili] /nabiˈsû**ri**-nale/ form.line-desid 'They want to form a line.' 'Se quieren formar.' < BFL 07, SF 08 1:83 /el > b. [nabiˈsûrima] /nabiˈsûri-ma/ form.line-fut.sg 'S/he will form a line.' 'Se va a formar.' < BFL 07 VDB/el > c. [sutuˈbéetʃ͡-nale] /sutuˈbé**tʃ͡i**-tʃ͡a-nale/ trip-ev-desid 'It sounds like they want to trip.' 'Se oye que se quieren tropezar.' < BFL 07 rec300/el > d. [sutuˈbétʃ͡ima] /sutuˈbétʃ͡i-ma/ trip-fut.sg 'S/he will trip.' 'Se va a tropezar.' <LEL 06 5:35/el > e. [aˈsíisa] /aˈsí**si**-sa/ wake.up-cond 'if s/he wakes up' 'si se despierta' < SFH 08 1:82/el > f. [aˈsísima] /aˈsísi-ma/ wake.up-fut.sg 'S/he will wake up.' 'Se va a despertar.' < SFH 08 1:82/el >

We could alternatively analyze CL triggered by syllable deletion as CL triggered by consonant deletion after cyclically applied posttonic syncope. That is, deletion would not target the syllable as a unit. Instead, the consonant, after

### 9.4 Verbal structure and verbal domains

being syllabified as coda of the preceeding syllable, would be the target of a phonetic weakening process to a semi-vowel and subsequent monophthongization (as proposed for other cases of CL by De Chene & Anderson 1979). This can be represented schematically as in (95):


There is no evidence, however, that all derived coda consonants can glide, except for /b/ (cf. §4.4.2). Furthermore, not all labio-velar semi-vowels undergo monophthongization, as shown in the examples in (96). Their existence makes it hard to posit a special set of semi-vowels that would not weaken and monopthongize with the syllable nucleus.

(96) Non-monophthongized labio-velar semi-vowels


Whether we analyze this last set of cases as instances of CL or not, the cases of vowel lengthening shown above are uncontroversially a case of CL triggered by V loss. CL is seemingly restricted to targeting stressed vowels of roots or derivational suffixes in the Inner Stem, delimitating this stem domain.

### **9.4.3.3 Past passive-conditioned lengthening**

Another morphologically-conditioned phonological effect involves vowel lengthening triggered by the past passive construction. The past passive suffix *-ru* is a stress-shifting affix with a stressed and an unstressed allomorph. The unstressed allomorph has the property of triggering lengthening of the final stem stressed vowel. This is exemplified in (97).

### 9 Verbs and the verbal complex

	- a. *naˈʔî oˈsìiru naˈʔî* here *oˈs-ì-ru* write-appl-pst.pass 'Something was written here' 'Aqui escribieron' < SFH 08 1:45/el >
	- b. *ˈkani bahuˈréero ba ka=ni* ˈneg=1sg.nom *bahuˈré-ru* invite.to.drink-pst.pass *ba* cl 'I wasn't invited to drink' 'No me invitaron al tesgüino' < BFL 07 2:33/el >
	- c. *ˈtòoru grabaˈdôra ˈtò-ru* take-pst.pass *grabaˈdôra* recorder 'The recorder was taken.' 'Se llevaron la grabadora.' < SFH 08 1:45/el > d. *naˈʔî iˈtʃ͡íiru*
		- *naˈʔî* here *itʃ͡i-ru* sew-pst.pass 'It was sewn here.' 'Aquí sembraron.' < SFH 08 1:45/el >

This effect, which cannot be predicted from the prosodic or phonological properties of the affix, can be considered as an instance of dominance. 'Dominant' affixes (as opposed to 'recessive' affixes) have been defined as affixes which delete or neutralize contrasts in the base to which they attach (Kiparsky et al. 1982, Inkelas 1998). Although dominant affixes are typically described as involving the deletion of accentual or tonal information from the base, there are also cases of dominant affixes that neutralize vowel length in the base (such as Mam Maya (Willard 2004)). I argue that the past passive suffix is a dominant suffix which imposes lengthening in a preceding stressed syllable.

There are instances where the vowel quality of the past passive suffix (a high, back round vowel) is neutralized in height in posttonic position. This yields a suffix form that is homophonous with the active voice past suffix (*-li*). In (98–99), the passive constructions would thus be homophonous with past active constructions, except that the lengthening in the stressed root vowel is a clear index of

### 9.4 Verbal structure and verbal domains

the passive construction. It is possible that there is a change in progress where the lengthening is being reanalyzed as the marker of past passive.

(98) Neutralized vowel quality of past passive suffix


While the lengthening in (98a) could be alternatively analyzed as compensatory lengthening triggered by syncope, the lengthening in (98b) cannot be attributed to compensatory lengthening, since there is no posttonic syncope in this form.

Passive-induced lengthening targets the root or a derivational suffix in the Inner Stem. (99a) shows the derivational suffix *-rú* undergoing legthening. The same base does not undergo vowel lengthening with the active past suffix *-li*.

(99) Past passive induced lengthening of Inner Stem suffix

	- *haˈsînto* Jacinto *raki-ˈrú-li* palm-gather-pst 'Jacinto gathered palms.' 'Jacinto juntó palmas.' < SFH 08 1:97/el >

The target of passive-induced lengthening includes the transitive suffixes in position S2, in the Derived Stem domain, as shown in (100).

### 9 Verbs and the verbal complex

	- a. *miˈgêl tʃ͡aʔiˈbúuru siˈkâla miˈgêl* Miguel *tʃ͡aʔi-ˈbú-ru* get.stuck-tr-pst.pass *siˈkâ-la* hand-poss 'Miguel's hand got stuck (by somebody else).' 'Le atoraron la mano a Miguel.' < SFH 08 1:97/el >
	- b. *ˈmá tʃ͡ihanâari napátʃ͡i ˈmá* already *tʃ͡iha-nâ-ru* scatter-tr-pst.pass *naˈpátʃ͡a* blouses 'The blouses were thrown around.' 'Ya desparramaron las blusas.' < SFH 07 1:17-21/el >
	- c. *ˈmá rapaˈnâaru ˈmá* already *rapa-nâ-ru* split-tr-pst.pass 'She was already operated (lit. cut).' 'Ya la operaron (cortaron).' < SFH 08 1:84/el >

On the other hand, suffixes on the Syntactic Stem (and any later morphological stem domains) block lenghtening of the stressed syllable in past passive constructions. In each example in (101), the past passive suffix does not trigger lengthening of an immediately preceding applicative suffix.

(101) No past passive induced lengthening of Syntactic Stem suffixes

a. *ˈnè amaˈtʃ͡îkiru*

*ˈnè* 1sg.nom *amaˈtʃ͡î-ki-ru* pray-appl-pst.pass 'They were praying for me' (lit. 'I was being prayed for').' 'Me rezaron.' < SFH 05 2:105/el >

b. *ˈmáni baʔiˈrúkuru baʔˈwí ˈmá=ni* already=1sg.nom *baʔi-ˈrú-ki-ru* water-gather-appl-pst.pass *baʔˈwí* water 'They already brought me water.' 'Ya me trajeron agua.' < SFH 08 1:84/el >

9.4 Verbal structure and verbal domains

c. *niˈhê koˈbísi ˈpásiru niˈhê* 1sg.nom *koˈbísi* pinole *ˈpá-si-ru* throw-appl-pst.pass 'They threw my pinole.' 'Me tiraron el pinole.' < SFH 08 1:85/el >

There are no constructions in the corpus where the past passive suffix imposes lengthening on a base including the suffixes in positions S3–S4 either, which implies that the domain of lengthening is restricted to the suffixes up to position S2, the Derived Stem domain. As shown in (102), for at least some speakers, the past passive suffix is *-liru* when attaching to V-final stems, with the liquid flap syllabifying as coda of the preceding syllable (due to general post-tonic vowel deletion and reduction process described in §5.2, the suffix surfaces as [-lri] in the examples below).


<sup>26</sup>The Spanish translation given for this verb is 'franquear', which likely has a limited use in Northern Mexico among native Spanish speakers; the drinking referred to is ritual drinking in Rarámuri communities associated with community-based work, where the host of the community work offers invitees corn beer to share.

### 9 Verbs and the verbal complex

In these examples, the past passive suffix with the extra consonant not found with other speakers also triggers lengthening in the stressed syllable of the stem. This form of the suffix appears to reflect a more conservatve form of this suffix and reveals the source of morphologically-conditioned lengthening in compensatory lengthening. The cognate suffixes in Mountain Guarijío and Norogachi Rarámuri are disyllabic *-reru* ~ *-riru* and *liri*, respectively (Miller 1996: 143; Brambila 1953). What this suggests is that the source of vowel lengthening associated with the Choguita Rarámuri past passive arose through compensatory lengthening upon vowel (and eventually syllable) deletion of the suffix.

### **9.4.3.4 Imperative singular stem formation: final stem stress shift and tonal alternations**

As described in §9.3.1, the imperative may be marked as final stem stress. This marking is restricted to be realized in a domain that includes the Derived Stem domain. Transitive stems of change-of-state predicates (described in §9.1.6) have an imperative with stress on the transitive suffix. This is shown in (103a).

### (103) Imperative stress shift

a. *Imperative, transitive kasiˈnâ kasi-ˈnâ* break-tr.imp.sg 'Break it!' '¡Rómpelo!' b. *Transitive stem + shifting suffix kasiˈnâma kasi-ˈnâ-ma* break-tr-fut.sg 'S/he will break it.' 'Lo va a romper.' c. *Transitive stem + neutral suffix kaˈsìnali kaˈsì-na-li* break-tr-pst 'S/he broke it.'

'Lo rompió.'

### 9.4 Verbal structure and verbal domains

Stress on the transitive suffix is characteristic of transitive stems in shifting constructions (103b), and contrasts with second syllable stress of the same transitive stems in neutral constructions (103c). The imperative of intransitive changeof-state predicates will, on the other hand, involve fixed second syllable stress plus an imperative singular suffix (as in (104)).

(104) Imperative of change-of-state predicates

```
a. kaˈsíka
     kaˈsí-ka
     break-imp.sg
    'Break yourself!'
    '¡Rómpete! < SFH 08 1:98/el >
b. waˈtʃ͡íka
    waˈtʃ͡í-ka
     be.straight-imp.sg
    'Straighten up!'
    '¡Enderézate!' < SFH 08 1:98/el >
```
Since the transitive suffixes of change-of-state predicates are part of the Derived Stem, we can identify final stem stress to mark imperative as a process restricted to this verbal zone.

### **9.4.3.5 Round harmony**

Choguita Rarámuri has a round harmony process,<sup>27</sup> where non-round vowels of certain suffixes may become round when preceded by a stressed back stem vowel. The following examples show the role of stem stressed vowels as triggers of the rounding of the following suffix vowels: in (105a) and (105c), a stem final high, back vowel triggers rounding in the vowels of the causative, applicative and associated motion suffixes; in (105b) and (105d), on the other hand, there is no rounding of applicative suffix vowels with a stem final high, front vowel.

### (105) Round harmony triggers

a. *Round harmony baniˈsútusuma baniˈsú-ti-si-ma* pull-caus-mot-fut.sg 'S/he will go along making them pull it.' 'Los va a ir haciendo que lo jalen.' < SFH 07 2:67 rec487 /el >

<sup>27</sup>I refer to this process as round harmony, although this process is gradient rather than categorical.

### 9 Verbs and the verbal complex


While round harmony in Choguita Rarámuri resembles other vowel harmony systems in its perseveratory, root-controlled nature, there is also evidence that harmony can be blocked or favored by the vocalic quality of an inflectional suffix following the target vowels (these suffixes are themselves outside of the domain of harmony). In (106a) and (106c), the applicative suffix *-ki* and the causative suffix *-ti* are realized with a round vowel after a stem with a final back vowel *if* the following inflectional suffix has a back vowel as well. The role of the final inflectional suffix in the harmony can be appreciated in (106b) and (106d), where the applicative and causative suffixes do not undergo round harmony when followed by an inflectonal suffix with a high, front vowel.

	- a. *kupuˈrókuma kupuˈró-ki-ma* blink-appl-fut.sg 'S/he will blink to her.' 'Va a parpadearle.' < BFL 05 2:22/el >

9.4 Verbal structure and verbal domains


This shows, then, than while clearly root-controlled, this process is also partially anticipatory. It also shows that back harmony is restricted to a subconstituent of the hierarchical structure of the verb. We have seen examples of round harmony targeting the vowels of causative *-ti* (S4), applicative *-ki* (S5), and associated motion *-simi* (S7). The examples in (107) show an array of suffixes undergoing rounding harmony. These suffixes belong in the Derived Stem, Syntactic Stem and Aspectual Stem domains.

(107) Round harmony in the Derived Stem and the Syntactic Stem

a. Applicative *-si* (S3): Round harmony *ˈpáʃuru* ˈpá-si-ru throw-appl-pst.pass 'It was thrown in my direction (for me).' 'Me lo tiraron (hacia mi).' <RF 04 1:82/el > b. Applicative *-si* (S3): No harmony *ˈpásiki* ˈpá-si-ki throw-appl-pst.ego 'I threw it for him.' 'Se lo tiré.' < ROF 04 1:82/el >

### 9 Verbs and the verbal complex

	- catch-ev-prog 'It sounds like they are catching (mice).' 'Se oye como que andan atrapando ratones.' < SFH 07 1:10/el >

There are cases where harmony appears to be blocked: in (108) there is no round harmony with the applicative and evidential suffixes, despite the presence of the trigger (a back, stressed vowel in the stem) and the following inflectional suffix with a back vowel. Instead, the back vowel of the evidential suffix has undergone height neutralization.

(108) Blocked harmony

*miˈsúkitʃ͡ina miˈsú-ki-tʃ͡ane-a* catch-appl-ev-prog 'It sounds like they are catching (some mice) for somebody.' 'Se oye como que le están atrapando ratones.' < SFH 07 1:10/el >

We have seen that the [+round] feature can spread over more than one vowel (e.g. (107a)), so it cannot be argued that harmony is limited in its rightward (or

<sup>28</sup>There is evidence that a following inflectional suffix with a front high vowel blocks the rounding harmony process: *ˈsû-n-tʃ͡ an-i* /ˈsû-ni-tʃ͡ane-i/ 'sow-appl-ev-imp' 'It used to sound like they were sowing stuff for her' 'Se oía como que le cosían' (< SFH 07 1:9/el >)

9.4 Verbal structure and verbal domains

leftward) spreading. Instead, it is possible that this pattern has been rendered opaque by posttonic vowel height neutralization.

Finally, back round vowels also favor reduction of posttonic unstressed vowels to schwa. Posttonic vowel reduction to schwa occurs frequently when preceded by a back, stressed vowel. The examples below, however, show that reduction to schwa takes place after central (109a) and front, mid vowels (109b) as well.

(109) /i/ > ə / á, é


The examples in (109) thus shows that the gradient process of unstressed vowel reduction to schwa and round harmony do not appear in the same vocalic environments.

In sum, rounding harmony in Choguita Rarámuri is stem-controlled but is simultaneously sensitive to outer inflectional suffixes, which are in turn out of the harmony domain. The targets of rounding harmony include the root and Inner Stem processes, as well as suffixes up to position S9. There is no evidence that other potential targets occurring in outer positions of the stem undergo rounding harmony. Thus, rounding harmony constitutes another phenomenon that contributes to creating less salient junctures between suffixes of certain inner domain of the verbal stem.

### **9.4.3.6 Stress and the morphologically complex verb: stress-shifting and stress-neutral suffixes**

It has been shown so far that the agglutinating structure of the Choguita Rarámuri verb is not uniform with respect to its morpho-phonological properties. There is yet another important property of the verbal stem that suggests an internal, layered organization: the characterization of suffixes as stress-shifting and stress-neutral. This section discusses how suffixes are grouped into layers in the

### 9 Verbs and the verbal complex

verb according to their stress properties. Stress-shifting suffixes combine with the Inner Stem and comform the stress domain, while non-shifting suffixes are outside the stressable domain. The stress properties of Choguita Rarámuri suffixes align with other properties that define the stem levels or domains, since stress-shifting and stress-neutral suffixes are grouped in interleaved layers in the morphological structure of the stem.

It has been proposed that accent systems where the interaction between prespecified information and word formation processes yields competing lexical accents, prosody is determined by morphology: a "headmost" accent wins, and the phonological properties of this morphological head percolate to the word level (Revithiadou 1998: 3–4). Under this account, "heads" are characterized as derivational morphemes (not inflectional ones). In Choguita Rarámuri there is no correlation between the suffixes' prosodic properties and their status as derivational or inflectional morphology.

### **9.5 The verbal complex: clitics and modal particles**

As described in Chapter 10, pronominal forms have corresponding enclitic forms, which are prosodically dependent on their host that do not carry any restrictions about the syntactic category of the words they attach to (Bickel & Nichols 2007). Table 9.13 lists the clitic pronominal forms (free pronouns are given in parenthesis). Third person is marked with a demonstrative *mi*, both as a free form and as an enclitic.


Table 9.13: Pronominal enclitic forms

Choguita Rarámuri also has epistemic modality markers (the morphological characteristics of these markers are addressed in Chapter 10). Epistemic modality, or the expression of the degree of certainty speakers have towards the actuality of an event, is marked in Choguita Rarámuri through two modal particles that follow inflected verbs: *aˈlé*, which expresses doubt and uncertainty (110a), and *oˈlá*,

### 9.5 The verbal complex: clitics and modal particles

which marks certainty, and often volition (110b). Forms lacking such particles have a neutral interpretation with respect to the speaker's commitment to the truth value of the proposition.

### (110) Epistemic modality markers

```
a. ˈnârma ˈlé
    ˈnâri-ma
    ask-fut.sg
                 aˈlé
                 dub
    '(He) will probably ask.'
    'Probablemente va a preguntar.' <BL 05 1:152/el >
b. ˈnârmo ˈlá
    ˈnâri-ma
    ask-fut.sg
                 oˈlá
                 cer
    'S/he will definetly ask.'
    'Seguramente que va a preguntar.' <BL 05 1:152/el >
```
As the examples in (110) show, these particles have the phonological effect of inducing vowel deletion of the final vowel of the singular future suffix. This phenomenon has led some to describe these epistemic elements as "suffixes with independent stress" in other Rarámuri dialects (Burgess 1984). These particles, however, show their independent-word status through their prosodic independence, and their ability to appear after person clitics. In a strong hypothesis of syntax-phonology interactions, cliticization follows syntax, which predicts that clitics are able to attach to other clitics, but affixes cannot attach to clitics (Zwicky & Pullum 1983).

Finally, deletion between the final future suffix and the epistemic particles takes place with an intermediate enclitic that has lost its vowel. This is exemplified in (111).

(111) Vowel deletion induced by epistemic markers

```
a. tʃ͡aʔiˈmêon ˈlá
     tʃ͡aʔi-ˈmêa=ni
     stuck-fut.sg=1sg.nom
                              oˈlá
                              cer
    'I will get stuck.'
    'Me voy a atorar.' < BFL 05 1:133/el >
```
### 9 Verbs and the verbal complex

b. *baˈhîmon oˈlá baˈhî-ma=ni* drink-fut.sg=1sg.nom *oˈlá* cer 'I will drink.' 'Voy a tomar.' < AHF 05 2:101/el > c. *riˈkùman ˈlé riˈkù-ma=ni* be.drunk-fut.sg=1sg.nom *aˈlé* dub 'I will probably get drunk.' 'Probablemente me voy a emborrachar.' < BFL 05 2:120/el > d. *riˈkùmon oˈlá riˈkù-ma=ni* be.drunk-fut.sg=1sg.nom *oˈlá* cer 'I will get drunk.' 'De seguro me voy a emborrachar.' < BFL 05 2:120/el >

Vowel deletion takes place post-lexically after the intermediate person clitic loses its vowel.

### **9.6 Summary**

The verbal structure of Choguita Rarámuri displays morphotactic, prosodic and morpho-phonological properties that define a concentric organization of suffixes, with more fused suffixes closer to the root and more separable suffixes in the outer layer of the verb. The verbal structure scheme proposed in this chapter is repeated in Table 9.14.

Despite having mostly a fixed position, the ordering of suffixes is not arbitrary and conforms to general principles. There are no discontinuous dependencies across suffix positions, as is frequent in position class morphologies. The structure proposed, instead, fits Bybee's lexical-derivational-inflectional continnum, and generally conforms to the universal principles of relevance, derivation within inflection and scope. This is a property attributed to layered morphologies vs. templatic or position class morphologies, where general semantic and syntactic principles do not determine the whole range of affix ordering facts (Bickel & Nichols 2007, Stump 1993). In the analysis proposed here, the layered structure of the Choguita Rarámuri verb accounts for a zone of variable order, multiple exponence in the Syntactic Stem, and the fixed order in the rest of the zones.


Table 9.14: Choguita Rarámuri verbal stem domains

# **10 Minor word classes**

This chapter addresses the properties of Choguita Rarámuri minor word classes. Minor word classes can be classified into two groups, depending on whether they may head noun phrases or combine with head nouns in noun phrases and those that cannot. The first group includes pronouns (§10.1), demonstratives (§10.2), adjectives (§10.3), numerals (§10.4), quantifiers (§10.5), and definite articles (§10.6). The second class of minor word classes includes adverbs (§10.7) and a set of particles and enclitics that bear a range of discourse pragmatic functions (§10.8).

From these word classes, adjectives and adverbs are the ones that have the largest sets of forms, though they both have significantly fewer lexical items than major word classes. As documented for other Uto-Aztecan languages (e.g., Cupeño (Takic; Hill 2005), underived adjectives in Choguita Rarámuri comprise a limited set, with most property concepts encoded by words derived by regular derivational mechanisms from verbs (addressed in Chapter 9). In contrast, there is an elaborate system of adverbial free and bound morphemes, and a particularly developed system encoding direction/location, as documented in closely related Mountain Guarijío (Taracahitan; Miller 1996). This chapter provides an overview of the morphological properties of these small word classes, while their syntactic behavior is addressed in other chapters.

### **10.1 Pronouns**

In contrast to nouns, free personal pronouns and pronominal enclitics are casemarked. Four different sets of pronouns are described: free personal pronouns in §10.1.1, pronominal enclitics in §10.1.2, emphatic pronouns in §10.1.3 and interrogative pronouns in §10.1.4.

### **10.1.1 Personal pronouns**

Choguita Rarámuri personal pronouns distinguish two person values (first and second) and two numbers (singular and plural). In addition to these distinctions, pronouns may encode a binary nominative-accusative case distinction. Nominative pronouns encode subjects of both matrix clauses and subordinate clauses,

### 10 Minor word classes

as well as possessors in noun phrases. Accusative pronominal forms are used to encode single objects of transitive predicates and both primary and secondary objects of ditransitive predicates.<sup>1</sup> A subset of pronominal forms, namely first person forms and the second singular form, are morphologically complex and composed of person forms reconstructed from Proto-Uto-Aztecan (Langacker 1977) and a *ˈhê* formative that appears to function as a demonstrative in a limited set of contexts. This *ˈhê* formative may be lacking in reduced forms of these pronominal forms.<sup>2</sup> The second singular and first plural subject pronominal forms also exhibit a change in the final vowel of the reduced form (from [u] to [o]), a change also attested in the first plural object pronominal form. The paradigm of Choguita Rarámuri free personal pronouns is shown in Table 10.1.

Table 10.1: Free personal pronouns


Reduced pronominal forms (*ˈnè*, *ˈmò* and *taˈmò*) are prosodically independent: they are stressed and may host enclitics. Reduced pronouns for first and second singular pronouns are monosyllabic and do not undergo lengthening as attested with monosyllabic inflected verbs, i.e., these pronominal forms are exempt from the minimality requirement described in §11.2 that applies to verbs stems. Examples of free personal pronouns in main clauses, both unreduced and reduced, are provided in (1).

<sup>1</sup>Nominative/accusative case distinctions in pronominal forms are documented for other Rarámuri varieties, including Norogachi Rarámuri (Brambila 1953), and Rochéachi Rarámuri (Morales Moreno 2016). For closely related Mountain Guarijío, Miller (1996) reports a binary distinction between subject pronominal forms and oblique pronominal forms, with the latter employed to encode objects of transitive and ditransitive clauses, subjects of subordinate clauses and nominal possessors (1996:230).

<sup>2</sup>Contexts where *ˈhê* appears to function as a demonstrative are described in §10.2 below. Villalpando-Quiñonez (2019) documents a similar pattern in Norogachi Rarámuri free pronouns (2019: 37).

10.1 Pronouns

### (1) Free personal pronouns

a. *neˈhê ˈpé oˈkwâ raʔiˈtʃâma koriˈmá ˈhîtara neˈhê* 1sg.nom *ˈpé* just *oˈkwâ* couple *raʔiˈtʃâ-ma* speak-fut.sg *koriˈmá* fire.bird *ˈhîtara* about 'I'll speak a little about the *korima* (the fire bird). ' 'Yo voy a hablar poquito del pájaro *korimá* (el pájaro de fuego).' {LEL tx5:00:22.9}

### b. *"ˈnè ʃiˈmêo ˈlá naʔˈpôʃia", ˈhê aˈní*

*ˈnè* 1sg.nom *siˈ-mêa* go-fut.sg *oˈlá* cer *naʔˈpô-si-a* weed-mot-prog *ˈhê* dem *aˈní* say.prs "'I'm going to weed", it is said like that.'

"'Voy a escardar", así se dice.' {JLG el1274:17:17.5}

c. *noˈrînima ˈétʃ͡i biˈléara tʃ͡oˈkêami taˈmí ruˈwèʃia noˈrîni-ma* arrive-fut.sg *ˈétʃ͡i* dem *biˈléara* another *tʃ͡oˈkêami* bet.settler *taˈmí* 1sg.acc *ruˈwè-si-a* tell-mot-prog 'The other bet settler (of *ariweta* race) arrives to tell me.' 'Viene la otra chokéami (apuntadora de carrera de *ariweta*) a

decirme.' {LEL tx19:01:17.9}

d. *aʔˈlì tamuˈhê ˈmá … ˈmá aʔˈlì raˈwè ˈmá ti napaˈwí ... hiˈrâmia aʔˈlì* and *tamuˈhê* 1pl.nom then *ˈmá* then *aʔˈlì* later *raˈwè* day *ˈmá=ti* already=1pl.nom *napaˈwí* gather *hiˈrâ-mi-a*

bet-mov-prog

'And then we gather that day to bet.'

'Entonces nosotros ese día ya nos juntamos para apostar.' {LEL tx19:01:32.3}

e. *ˈkíti ku taˈmò raˈlàmuli ˈpé kuˈrì oˈtʃêrikam ˈhú taˈmò ko ˈkíti* because *ku* emph *taˈmò* 1pl.nom *raˈlàmuli* Rarámuri *ˈpé* just *kuˈrì* recently *oˈtʃêra-kame* grow-pst.ptcp *ˈhú* cop *taˈmò=ko*

1pl.nom=emph

'Because us the Rarámuri have grown up just recently, our people.' 'Porque hace poco que crecimos nosotros los rarámuri, la gente como nosotros.' {FLP in243:18:57.8}

### 10 Minor word classes

f. *aʔˈlì ˈmá ke iˈtêo ˈlá ˈnà ˈháp taˈmò tʃ͡iriˈká ruˈjèma ko aʔˈlì* and *ˈmá* already *ke* neg *iˈtê* not.exist *oˈlá* cer *ˈnà* prox *ˈhápi* sub *taˈmò* 1pl.acc *ˈétʃ͡i* dem *riˈká* like *ruˈ-è-ma=ko* say-appl-fut.sg=emph 'And then there is nothing for us to be told.' 'Y luego ya no hay para que nos digan a nosotras.' {GFM tx905:02:04.1}

Second person pronouns do not encode a number distinction in their accusative form (this is also reported in closely related Mountain Guarijío (Miller 1996)). Examples of second person pronominal forms are provided in (2).

	- a. *muˈhê ˈmá ke ˈtʃ͡i ˈbíri rimuˈrú ˈhípi ko ba? muˈhêˈmá* 2sg.nom *ke* anymore *ˈtʃ͡i* neg *ˈbíri* which *rimu-ˈrú* kinds *ˈhípi=ko* dream-prs *ba?* today=emph cl 'And you don't dream many things anymore?' '¿Y ahora ya no sueñas muchas cosas?' {MDH co1136:16:38.4} b. *ˈmò ˈwé biˈnè ˈétʃ͡i ˈmò* 2sg.nom *ˈwé* int *biˈnè* know.prs *ˈétʃ͡i* dem "'You know a lot about that..."' "'Tu sabes mucho de eso..."' {LEL tx5:05:18.9} c. *baʔaˈrîni ˈmí ˈàma baʔaˈrî=ni* tomorrow=1sg.nom *ˈmí* 2sg.acc *ˈà-ma* look.for-fut.sg 'I'll look for you tomorrow.' 'Mañana te busco.' <LEL 09 1:70/el> d. *aʔˈlì ˈétʃ͡i ˈmín aˈnèma aʔˈlì aʔˈlì* and *ˈétʃ͡i* dem *ˈmí=ni* 2sg.acc=1sg.nom *aˈn-è-ma* say-appl-fut.sg *aʔˈlì* later

"'And then I'll tell you"' "'Y entonces te digo"' {LEL tx19:01:13.5}

10.1 Pronouns

e. *kuˈrí oˈtʃêrirami ˈémi ko ba kuˈrí* recently *oˈtʃêri-r-ame* grow-pst.pass-ptcp *ˈémi=ko* 2pl.nom=emph *ba* cl 'You all who have recently grew up.' 'Ustedes los crecidos hace poco' {FLP in61:00:38.3}


'Nomás de ese tanto les digo, ustedes, mis hijos.' {SFH tx12:12:46.1}<sup>3</sup>

Third person arguments may be left unmarked (3a), or they may be encoded through demonstratives (e.g., the demonstrative *ˈétʃ͡i* in (3b)) or through an emphatic pronoun (e.g., *biˈnôi* 'himself' in (3c)).


c. *ˈápi aʔˈlì biˈnôi wikaˈrâ ko ˈhê aˈní ˈrú*

<sup>3</sup>As discussed in Chapter 12 (§12.2.2), Choguita Rarámuri has a dedicated pronominal form, namely *ˈkîni*, which encodes first person possessors (singular or plural) of kinship terms.

### 10 Minor word classes

*ˈápi* sub *aʔˈlì* then *biˈnôi* himself *wikaˈrâ=ko* sing.prs=emph *ˈhê* dem *aˈní* say.prs *ˈrú* say.prs 'when he sings he says this' 'cuando canta él así dice' {LEL tx71:03:15.7}

The use of demonstratives to refer to third person arguments is discussed in more detail in §10.2 below.

### **10.1.2 Pronominal enclitics**

Free subject pronouns have corresponding enclitic forms, phonologically bound formatives that are prosodically dependent on their host and are unrestricted regarding the syntactic category of the words they attach to, two properties that may serve as diagnostics of clitics cross-linguistically (Bickel & Nichols 2007). Choguita Rarámuri pronominal enclitics are unstressed, monosyllabic forms with high front vowels, a trait that may be attributed to general processes of post-tonic vowel reduction (as described in §5.2). Table 10.2 illustrates the clitic pronominal forms (free subject and object pronouns are given in parenthesis).

Table 10.2: Pronominal enclitic forms


Choguita Rarámuri person enclitics can attach to verbs and hosts of virtually any category, and, like many other Uto-Aztecan languages (Steele 1976) and other Rarámuri varieties (Morales Moreno 2016), are generally in what is traditionally called the Wackernagel position, immediately after the first accented phrase or sub-constituent of a phrase (Bickel & Nichols 2007). The following examples illustrate the distribution of person enclitics hosted by a wide range of word classes in a variety of syntactic contexts: (i) subordinating morphemes in subordinate clauses (4a); (ii) demonstratives (including definite articles) within Noun Phrases (4b); (iii) preposed particles (4c); (iv) negative adverbs (4d); (v) epistemic particles (4e); (vi) nouns (4f); and (vii) free person pronouns (4g). As shown in these examples, pronominal enclitic forms may undergo vowel deletion. Clitic hosts are underlined.

	- a. Subordinator *riˈmùini ˈnáptim noˈkáo riˈmù-i=ni* dream-impf=1sg.nom *[ˈnápi=timi* sub=2pl.acc *noˈká-o]* move-ep 'I used to dream that you all were moving.' 'Yo soñaba que ustedes se movían.' < BFL 05 1:114/el > b. Demonstratives (within a noun phrase) *ˈtin toˈrí siʔˈrítimo ˈlá ˈti=ni* def.bad=1sg.nom *toˈrí* chicken *siʔˈrí-ti-ma* drown.intr-caus-fut.sg *oˈlá* cer 'I will drown the chicken.' 'Voy a ahogar al pollo.' < BFL 05 2:49/el > c. Preposed particles *aʔˈlì ˈkun noˈrînima aʔˈlì* later *ˈku=ni* rev=1sg.nom *noˈrîni-ma* return-fut.sg 'I will come back later.' 'Al rato vuelvo.' < BFL 05 2:49/el > d. Negative adverbs *ˈkeni ˈtâʃi ˈtʃ͡ó maˈnâ baˈʰtâri ˈke=ni* neg=1sg.nom *ˈtâsi* neg *ˈtʃ͡ó* yet *maˈnâ* make.beverage *baˈʰtâri* corn.beer 'I haven't made corn beer yet.' 'No he hecho tesgüino todavía.' < BFL 05 2:56/el > e. Epistemic particles *noˈkèli ˈlén ˈmáo noˈk-è-li* move-pst *aˈlé=ni* dub=1sg.nom *ˈmá-o* maybe-ep 'Maybe I moved him.' 'A lo mejor lo moví.' < BFL 05 1:114/el >

### 10 Minor word classes

f. Nouns *napaʔˈlì noˈkáli ronoˈtʃ͡íni oʔˈkô ˈnápi* sub *aʔˈlì* later *noˈká-li* move-pst *[ronoˈtʃ͡í=ni* legs=1sg.nom *oʔˈkô]* hurt 'When I moved, my legs hurt.' 'Cuando me moví me dolieron las piernas.' < BFL 05 1:114/el > g. Full pronouns *ˈpé taˈmòm naˈhâta iˈʃì ˈpé* just *taˈmò=mi* 1pl.acc=dem *naˈhâta* follow *iˈsì* do.prs 'It went like that, following us around.' 'Así anduvo siguiéndonos.' < BFL 05 text 2/tx>

Although the list of possible hosts in (4) is not exhaustive, it illustrates clearly the unrestrictedness of possible hosts for the person enclitics in Choguita Rarámuri.

### **10.1.3 Emphatic pronouns**

There are two emphatic pronominal forms in Choguita Rarámuri, listed in (5).

	- a. *biˈnôi* singular
	- b. *aˈbôi* plural

These pronominal forms focus attention on the participants encoded as subjects in contexts where other potential arguments could be subjects. This is exemplified in (6).

(6) a. *aʔˈlì ˈétʃ͡i ˈnápu roˈwéma ˈlé ko biˈnôi biˈlá aˈní "ˈjénan ˈá saˈjèrima" ˈá aˈní*

> *aʔˈlì* and *ˈétʃ͡i* dem *ˈnápu* sub *roˈwé-ma* women.race-fut.sg *aˈlé=ko* dub=emph *biˈnôi* emph.sg *biˈlá* really *aˈní* say.prs *"ˈjéna* yes *a=ni* indeed=1sg.nom *ˈá* indeed *saˈjèri-ma"* take.on-fut.sg *ˈá* indeed *aˈní* say.prs 'And then the one who will run, herself, says: "yes, I will take on the challenge".'

'Y entonces la que va a correr ella misma dice "sí le voy a entrar".' {LEL tx19:00:39.8}

10.1 Pronouns

b. *ˈnápu riˈká ˈne aʔˈlá beˈnèrpo ˈtʃ͡ó ˈkîni ˈkûtʃ͡uwa ˈtʃ͡ó ˈkûtʃ͡uwa ˈtʃ͡ó ˈémi aˈbôi ba niˈbí ˈnápu* sub *riˈká* like *ˈne* int *aʔˈlá* well *beˈnè-ri-po* learn-caus-fut.pl *tʃ͡o* also *ˈkîni* my *ˈkûtʃ͡uwa* children *ˈtʃ͡ó* also *ˈkûtʃ͡uwa* children *tˈʃo* also *ˈémi* 2pl.nom *aˈbôi* emph.pl *ba* cl *niˈbí* nibí 'So that we can teach well our children, children, you all.' 'Para enseñarles bien a nuestros hijos, los hijos, ustedes mismos.' {SFH tx12:05:30.0}

In example (6a), the singular emphatic pronoun *biˈnôi* makes clear that the runner, and not other potential actors, is the source of the quoted speech. In (6b), the plural emphatic pronoun *aˈbôi* is used to emphasize the addressees, the children of the speaker who he is giving advice to.

### **10.1.4 Interrogative pronouns and phrases**

Choguita Rarámuri has a set of interrogative pronouns and phrases, most of which are morphologically complex. Of this set, only four forms are morphologically simplex. This inventory also includes the interrogative pronoun *(he)ˈkwâ*, where the *ˈhê* formative present in pronominal forms (see Table 10.2) is optional in the interrogative pronoun. These are provided in Table 10.3.


Table 10.3: Choguita Rarámuri interrogative pronouns: basic forms

A recurrent pattern in morphologically complex question words and phrases involves the interrogative word *tʃ͡ú* followed by another morpheme, whether bound (e.g., *-ˈrupi* in *tʃ͡ú ˈrupi* 'How much?') or free (e.g., a demonstrative (e.g., *(ˈnà-ti* 'that-with' in *tʃ͡ú ˈnà-ti* 'with what?)). The set of morphologically complex question words is shown in Table 10.4.

### 10 Minor word classes


Table 10.4: Choguita Rarámuri interrogative pronouns

In contrast to free pronominal forms, interrogative pronouns are not case marked, except for the forms *ˈpîri ˈnà-ti?* and *ˈtʃ͡ú ˈnà-ti?* 'with what', where the question words (*ˈpîri* and *ˈtʃ͡ú* are followed by the proximal demonstrative *ˈnà* bearing the *-ti* instrumental case marker.

In contrast to closely related Mountain Guarijío (Miller 1996), Choguita Rarámuri does not deploy interrogative pronouns as indefinite pronouns.

### **10.2 Demonstratives**

Demonstratives are defined here as a set of deictic elements that may refer or restrict reference to referents in a situational (exophoric) usage, identifying entities in a surrounding physical situation, but also may be used endophorically in discourse and recognitional deixis (Himmelmann 1996, Enfield 2003). Demonstratives in Choguita Rarámuri may function anaphorically as pronouns (§10.2.1) or modifying a nominal element (§10.2.2). Both sets involve the same set of forms, and are distinguished in the following sections in terms of their function. Other

10.2 Demonstratives

potential uses of demonstratives in discourse are yet to be examined and are left out of the scope of this grammar.

### **10.2.1 Demonstrative pronouns**

As described in §10.1.1, reference to third person arguments may be achieved through demonstrative pronouns. Two degrees of distance and orientation with respect to the speaker/addressee is encoded by these forms, listed in (7):

	- a. *ˈnà* 'this one' proximal, close to the speaker
	- b. *ˈétʃ͡i* 'that one' proximal, close to the addressee or the speaker

The use of both proximal demonstrative pronouns is frequently attested in the Choguita Rarámuri corpus. Examples of these demonstrative pronouns is provided in (8).

	- a. *ˈpé ˈnàbi tʃ͡e ˈt͡ʃéti koˈlì biˈtíam ku ˈpé* just *ˈnà=bi* this.one=just *tʃ͡e* also *ˈt͡ʃéti* det.pl *koˈlì* over.there *biˈtí-ame* lie.down.pl-ptcp *ku* rev 'Just these ones (the dead people) who are lying down over there also.'

'Nada más estos que están (acostados) ahí por de aquel lado (los muertos).' <FLP 07 in243(511)/in>

b. *"ˈnà ko ˈpâalichi paˈkótami" ˈhê biˈlá aˈní ba aˈní ˈnà=ko ˈpâalichi paˈkó-t-ame" ˈhê biˈlá aˈní*

this.one=emph priest wash-pacient-ptcp dem really say.prs *ba* cl *aˈní*

say.prs

"'This one is baptized by a priest" that's what they say.'

"'Este es bautizado por padre" así dicen.' {SFH tx475:08:03.9}

	- *ˈétʃi=ko* those.ones=emph *tʃ͡oˈmí* there *ˈtʃ͡ó* also *muˈtʃûwi* sit.pl.prs

'They were (sitting) also over there.'

'Ellos también estaban allá.' <FL 06 in61(302)/in>

### 10 Minor word classes

d. *ˈétʃ͡i ko ˈwé aˈnátʃ͡a ˈétʃi=ko* that.one=emph *ˈwé* int *aˈnátʃ͡a* endure.prs 'That one really endures (to run).' 'Ese aguanta mucho (correr).' {JLG el1278:03:59.6}

In addition to these demonstrative pronouns, there is a third, more restricted demonstrative pronoun *ˈhê*. This demonstrative pronoun is exclusively found as the pronominal complement of utterance predicates ('say', 'tell'), as well as other predicates that express a positive attitude regarding the truth of the proposition expressed as their complement ('think', 'believe'). Crucially, this formative appears in contexts where there is quoted speech, and its function appears to be to index the direct speech. Relevant examples are provided in (9).

	- a. *aʔˈlì ˈhê aˈníli: "ˈnè ko a maˈtʃ͡í ˈkúmi ˈjéna biˈtê ˈétʃ͡i oˈhí" aʔˈlì* and *ˈhê* dem *aˈní-li* say-pst *ˈnè=ko* 1sg.nom=emph *a* indeed *maˈtʃ͡í* know.prs *ˈkúmi* where *ˈjéna* indeed *biˈtê* live.prs *ˈétʃ͡i* that *oˈhí"* bear 'And they said: "I do know where the bear lives".' 'Y entonces dijeron: "yo sí se dónde vive el oso".' {LEL tx32:02:11.0} b. *aʔˈlì ˈhê aˈní ˈtîo: "a riˈwèʃi, ˈkíti biˈlé tʃ͡apiˈʃì ˈétʃ͡i" aʔˈlì* and *ˈhê* dem *aˈní* say.prs *ˈtîo:* uncle *"a* indeed *riˈwè-si,* leave-imp.pl *ˈkíti* neg.imp *biˈlé* one *tʃ͡api-ˈsì* grab-imp.pl *ˈétʃ͡i"* that 'And my uncle said: "leave it, don't touch it".' 'Y dijo mi tío: "déjenlo no lo tienten".' {LEL tx84:05:43.6}

The cognate of the *ˈhê* demonstrative in Norogachi Rarámuri is described as a proximal demonstrative in (Villalpando-Quiñonez 2019). Examples provided of this formative in Brambila (1953) and Villalpando-Quiñonez (2019) appear in the same environments as in Choguita Rarámuri, namely in the context of utterance and other predicates that introduce quoted speech.

10.2 Demonstratives

### **10.2.2 Adnominal demonstratives**

The two demonstrative forms that may be used pronominally may also be used as adnominal demonstratives: *ˈnà* 'this, speaker-proximate' and *ˈétʃ͡i* 'that, speaker or addressee proximate'. As shown in the examples in (10), adnominal demonstratives precede the head noun (further details relating the syntactic order of elements within noun phrases is addressed in Chapter 12).

(10) a. *"ˈpîri ˈtʃêtimi oˈlá taˈmí ke biˈlé ˈpé ta ukuˈwèami u pa?" ˈhê riˈká biˈlá iˈjòani ˈétʃ͡i Patricio, ˈá biˈlá ko ˈpîri* why *ˈtʃê=timi* why=2pl.nom *oˈlá* why *taˈmí* 1sg.acc *ke* neg *biˈlé* one *ˈpé* just *ta* little *ukuˈwè-ame* ukuwea-ptcp *ˈhu* cop *pa?* cl *ˈhê* dem *riˈká* like *biˈlá* really *iˈjòani* angry.prs *[ˈétʃi* that *Patricio]* Patricio *ˈá* indeed *biˈlá=ko* really=emph "'Why didn't you do that ukuweruwa thing?" says sometimes Patricio.' "'¿Porqué ustedes no me hicieron eso del ukuwéruwa?" a veces nos dice Patricio enojados.' {SFH tx475:09:16.8} b. *aʔˈlì ko ˈmá wiˈlí siˈkôtʃ͡i ˈúmiri ˈétʃ͡i koriˈmá aʔˈlì=ko* and=emph *ˈmá* already *wiˈlí* stand.sg.prs *siˈkô-tʃ͡i* corner-loc *ˈúmiri* long.time *[ˈétʃi* this *koriˈmá]* fire.bird 'And then that *korima* (fire bird) had been standing in the corner already for a long time.' 'Y entonces ya andaba mucho rato ahí por el rincón el *korimá* (pájaro de fuego).' {LEL tx5:01:28.4} c. *ˈnà leˈhîdotʃ͡i ˈtʃ͡ó ˈmáti waˈná ˈnà roˈhàna ˈtʃ͡ó [ˈnà* this *leˈhîdotʃ͡i]* ejido *ˈtʃ͡ó* also *ˈmá=ti* already=1pl.nom *waˈná* aside *ˈnà* this *roˈhà-na* separate-tr.prs *ˈtʃ͡ó* also 'This *ejido* also, we were separated.' 'Este ejido también, ya nos apartaron.' {JMF tx817:01:03.0} d. *tiˈwé ˈnà ˈtʃîkle koˈʔáli tiˈwé* girl *[ˈnà* this *ˈtʃîkle]* gum *koˈʔá-li* eat-pst 'The girl chewed this gum.' 'La niña se comió este chicle.' {SFH el1028:04:15.8}

### 10 Minor word classes

As shown in these examples, the "addressee/speaker-proximate" demonstrative *ˈétʃ͡i* 'that' indexes an object or person accessible to the addressee and/or the speaker: in the case of (10a), the speaker's son, known also to me, the addressee; in the case of (10b), the topic of the narrative, the *korima* (fire bird) that has been previously introduced in discourse. The 'speaker-proximate' demonstrative *ˈnà* 'this' is used to index an entity accessible to the speaker, e.g., the *ejido* of Choguita in (10c).

Cross-linguistically, demonstratives encode more than aspects of the spatial configuration, and their use is determined by interactional, cognitive and discourse pragmatic factors (Enfield 2003, Hanks et al. 2005). Future research will reveal in which ways these factors may determine the use and function of demonstratives in Choguita Rarámuri.

### **10.3 Adjectives**

Adjectival forms in Choguita Rarámuri are encoded through a small, close set of words, primary adjectives (described in §10.3.1) and a set of morphologically derived forms from verbs through participial suffixes (described in §10.3.2) (Islas Flores (2010) presents a description and analysis of adjectives and morphologically derived forms that encode property concepts in Choguita Rarámuri).

### **10.3.1 Primary adjectives**

Primary adjectives in Choguita Rarámuri are monomorphemic (underived) lexical items that denote property concepts and may appear as modifiers of head nouns in noun phrases (see Chapter 12). Primary adjectives are not derived from other morphemes. A comprehensive, though non-exhaustive list of primary adjectives is provided in (11). As discussed in §10.7, *waʔˈlû* 'big, sg.' may be use adverbially to modify predicates.

	- a. *ˈkútʃ͡i* 'little, small, pl.'
	- b. *ˈtàa* 'little, small, sg.'
	- c. *waʔˈlû* 'big, sg.'
	- d. *(i/o)ʔˈwéri* 'big, pl.'
	- e. *ˈtʃáti* 'ugly'
	- f. *waˈrìna* 'fast, light'

10.3 Adjectives


As shown in these examples, some primary adjectives have suppletive forms for singular and plural (e.g., *ˈkútʃ͡i* 'little, small, pl.' (11a) vs. *ˈtàa* 'little, small, sg.' (11b)), but not all primary adjectives encode number distinctions (e.g., *naˈsina* 'lazy' (11g)). Among a number of morphological mechanisms to encode singularplural contrasts, number-sensitive suppletion is attested across Uto-Aztecan for both nouns (Hill & Hill 2000) and verbs (Haugen & Everdell 2015). As discussed in Chapter 9 and Chapter 8, number suppletion is attested in Choguita Rarámuri verbs and nouns in limited instances. <sup>4</sup>

Primary adjectives are morphologically characterized for their ability to take on degree morphology: as shown in (12), primary adjectives may attach the suffix *-bê*, a stress-shifting, HL-toned suffix that encodes a high degree of the property encoded by the adjective.

(12) Degree morphology with primary adjectives


*ka* because.neg *tʃ͡è* because.neg *wiʰka-ˈbê* many-more *riˈhò-li* inhabit.pl-pst *ba* cl 'Because there were not that many more people living here.' 'Porque no había mucha más gente.' {JMF tx817:00:36.9}

b. *ˈníma be ˈlào ˈpé a wiliˈbê iˈnáma ˈlé ˈmò ko ba ˈní-ma* cop-fut.sg *be* be *ˈlà-o* think-ep *ˈpé* just *a* aff *wili-ˈbê* long-more *iˈná-ma* go.along.sg-fut.sg *aˈlé* dub *ˈmò=ko* 2sg.nom=emph *ba* cl 'I think so, you will be around (live) for a longer time.'

'Yo creo que si, tu si vas a andar mucho tiempo.' <FL 06 in61(704)/in>

Primary adjectives with and without degree morphology may undergo further affixation, as shown in (13), where adjectives marked with *-bê* (13a–b) or without degree morphology (13c) may additionally attach the progressive *-a* suffix followed by a nominalizer suffix (stress-shifting *-ra*) in superlative forms.

<sup>4</sup>Closely related Mountain Guarijío has a small class (about a dozen lexical items) of primary adjectives that include cognate forms of the Choguita Rarámuri primary adjectives, e.g., *werumá* 'long', *weré* 'wide' (Miller 1996: 238). In contrast to Choguita Rarámuri, primary adjectives in Mountain Guarijío do not exhibit number suppletion, and instead encode number contrasts through a productive process of prefixing reduplication (1996: 238).

### 10 Minor word classes

	- a. *aʔˈlì ˈétʃ͡i taˈbêara biniˈlâ ko ˈá riˈpîli aʔˈlì* and *ˈétʃ͡i* dem *ta-ˈbê-a-ra* small-more-prog-nmlz *bini-ˈlâ=ko* small.sister-poss=emph *ˈá* aff *riˈpî-li* stay-pst 'And then the youngest sister stayed.' 'Y entonces la hermana menor se quedó.' {LEL tx32:04:05.0} b. *sereˈbêrio be ko riˈwèi taˈbêara ko sereˈbêrio* Silverio *be=ko* be=emph *riˈwè-i* be.named-impf *ta-ˈbê-a-ra=ko* small-more-prog-nmlz=emph 'Silverio was the name of the youngest one.' 'Se llamaba Silverio el más chico.' <FL 06 in61(286)/in> c. *baʔaˈrîna riˈkó ˈmá waʔˈlûara kotʃ͡iˈlâ ko ʃiˈmíli ˈtʃîba ˈmèra baʔaˈrî-na* tomorrow-distr *riˈkó* emph *ˈmá* already *waʔˈlû-a-ra* big-prog-nmlz *kotʃ͡i-ˈlâ=ko* daughter-poss=emph *siˈmí-li* go.sg-pst *ˈtʃîba* goat *ˈmèra* herd.prs 'The next dat the oldest daughter went to take care of the goats.' 'Al otro día la hija más grande se fue a cuidar a las chivas.' {LEL tx32:03:57.6}

I posit that primary adjectives in superlative forms undergo a zero conversion process, which allows them to take tense morphology before being nominalized. Alternatively, and given that very few forms present this morphological pattern, it could be hypothesized that these forms are lexicalized and not a productive sequence of morphemes available to all primary adjectives. These questions remains open for further research.

### **10.3.2 Property concepts derived from verbs**

Most property concepts in Choguita Rarámuri are derived from verbs through participial suffixes (*-ame* and *-kame*), which are also attested in the derivation of nouns from verbal bases (see §8.5.1). Examples of derived adjectives are provided in (14).

	- a. *roˈsâ-kame* 'white, sg.'

10.4 Numerals


The cognate forms of these participial suffixes (*-me* and *-kame*) are documented in Mountain Guarijío (Miller 1996), and are analyzed by Miller as lacking tense marking in the case of *-me* and as encoding past tense in the case of *-kame* (1996: 180). As discussed in §8.5.1, there are clear instances of Choguita Rarámuri participial forms derived through *-kame* that encode a past tense meaning and forms derived with *-ame* that lack a similar tense specification. However, in the case of property concepts, such TAM contrasts are not identifiable, which suggests that  *ame* and *-kame* function as suppletive allomorphs in the derivation of these forms. The role of participial suffixes in deriving relative clauses in Choguita Rarámuri is discussed further in §15.3.

As described in §8.2 in Chapter 8, Choguita Rarámuri verbs encode event plurality or pluractionality though a number of morphological mechanisms, including prefixation, consonant mutation or both prefixation and consonant mutation in a pattern of multiple exponence. This is exemplified in the derived adjectives above, where the contrast between (14a) and (14b) shows a number distinction marked through mutation of the first stem consonant (/r/-/t/), while the contrast between (14c) and (14d) shows the number distinction marked through both prefixation and mutation of the second root consonant.

### **10.4 Numerals**

Choguita Rarámuri numerals constitute a separate word-class defined morphologically by the ability to take multiplicative/frequentative, inclusive and collective morphology. Syntactically, numerals may modify head nouns in noun phrases or head noun phrases as described in Chapter 12. The numeral system can be characterized as an arithmetic base-10 system. The basic numerals are shown in Table 10.5. From the lower numbers in this set (one to ten), two forms are morphologically complex, namely 'eight' (*oˈsá naˈó* 'two times four')<sup>5</sup> and 'nine' ('not nine'), while the rest are monomorphemic. Higher numbers are all

<sup>5</sup>The word for number 'four' exhibits inter-speaker variation in its pronunciation with an optional palatal glide as an onset of the stressed syllable (*naˈó* ~ *naˈjó*)

### 10 Minor word classes

morphologically complex and involve derivation of the lower numerals in the expression of the tens-series. Speakers exhibit a strong tendency to use Spanish loanwords for numbers above ten, including adoption of the Spanish words *ciento* and *mil* for 'one hundred' and''one thousand', respectively.


Table 10.5: Choguita Rarámuri numerals

Numeral bases may be derived with a multiplicative/frequentative suffix (stressshifting *-sá*), which encodes the meaning 'x-times'. This suffixing construction is exemplified in (15).

	- a. *beˈsá naˈhùura ˈruá ˈá ˈrú be-ˈsá* three-mltp *naˈhù-ra* fall-rep *ru-wá* say-mpass *ˈá* aff *ˈrú* say.prs 'They say (the earth) fell down three times.' 'Dicen que se cayó (la tierra) tres veces.' {FLP in243:18:51.7} b. *ˈmá naˈósa ku aˈwíli ˈnè ˈmá* already *naˈó-sa* four-mltp *ku* again *aˈwí-li* dance-pst *ˈnè* 1sg.nom 'I already danced four times.' 'Ya bailé cuatro veces yo.' {MDH co1136:12:40.6}

### 10.4 Numerals

Numerals may also attach an 'inclusive' suffix (H-toned, stress-shifting suffix *-ná*), which encodes the meaning 'all of N', where N equals the total number of entities referred to by the head noun. Examples of this construction are shown in (16).

	- a. *okoˈná siˈkâla oko-ˈná* two-incl *seˈkâ-la* hand-poss 'in both hands' 'en las dos manos' < BFL 09 1:34/el >
	- b. *naʔˈsòoka koʔˈpôo okoˈná koʔˈpô biˈlá ba naʔˈsò-wi-ka* mix-mpass-ger *koʔ-ˈpô* eat-fut.pl *oko-ˈná* two-incl *koʔ-ˈpô* eat-fut.pl *biˈlá* indeed *ba* cl

'One must eat mixing it, one must eat the two foods.'

'Hay que comer revolviendo, hay que comer de las dos comidas.' {MDH co1136:00:29.2}

c. *ˈnè bikiˈná maˈkúsarani tʃ͡iˈwí ko ˈnè* 1sg.nom *biki-ˈná* three-incl *maˈkúsa-ra=ni* fingers-poss=1sg.nom *tʃ͡iˈwí* hit *ko* emph 'I hit myself in the three fingers.'

'Me pegué en los tres dedos.' < BFL 09 1:34/el >

d. *uˈsànna ˈá maˈní baʔˈwí uˈsàni-na* six-incl *ˈá* aff *maˈní* be.liquid *baʔˈwí* water 'There is water at six places.' 'En seis partes hay agua.' < BFL 09 1:35/el >

Numeral bases may also attach the stress-neutral collective suffix (a toneless *-ka* suffix) to mean that an activity is performed jointly by a given number of participants. This construction is exemplified in (17).

### (17) Collective numerals

a. *oˈkwâka ˈtúpisi baʔˈwí oˈkwâ-ka* two-coll *ˈtú-pi-si* bring.imp-mot.imp .pl *baʔˈwí* water 'Bring water between the two of you!' '¡Traigan agua entre los dos!' < BFL 09 1:34/el >

### 10 Minor word classes

b. *ˈmá biˈláti aˈbôi oˈkwâka ka aˈní tamuˈhê tʃ͡oˈkêami ˈpîri raˈwé roˈwéma umuˈkî ˈmá* already *biˈlá=ti* indeed=1pl.nom *aˈbôi* emph.pl *oˈkwâ-ka* two-coll *ka* emph *aˈní* say.prs *tamuˈhê* 1pl.nom *tʃ͡oˈkêami* bet.settlers *ˈpîri* which *raˈwé* day *roˈwé-ma* race.women-fut.sg *umuˈkî* women 'Then us two, the bet settlers, we say which day the women will race.' 'Ya entre las dos decimos, nosotras las chokéami que día van a correr las mujeres.' {LEL tx19:01:23.6}

Table 10.6 provides the multiplicative, inclusive and collective derivation of basic numerals (data on inclusive derivations is from < BFL 09 1:34/el >). Some derived forms are unattested in the corpus, and these are marked with "-".


Table 10.6: Numerals and derived forms

Finally, numeral bases may undergo multiple suffixation and attach both the inclusive and collective suffixes. The inclusive *-ná* suffix attaches directly to the numeral root and the collective *-ka* suffix attaches to the inclusive-marked numeral base. Relevant examples are shown in (18).

	- a. *ˈápi iˈsêlikami ˈká ˈlé biˈkiánika suˈwâba maˈjôra ma ˈápi* sub *i-ˈsêrikami* pl-governor.pl *ˈká* cop.irr *aˈlé* dub *biˈkiá-ni-ka* three-incl-coll *suˈwâba* all *maˈjôra* mayor

10.5 Quantifiers

*ma* also 'Like the governors, all three of them, all of the mayors, too.' 'Así como los gobernadores, los tres, todos los mayores también.' {JMF tx816:00:36.7} b. *ˈétʃ͡o ˈnà biˈtʃ͡èti ʃiˈmíbiki naˈjónka ka ˈétʃ͡i* dem *tʃ͡oˈnà* there *biˈtʃ͡è=ti* there=1pl.nom *siˈmíbi-ki* go.pl.pst-pst.ego *naˈjó-na-ka* four-incl-coll *ka* emph 'Over there all four of us went.'

'Allí fuimos los cuatro.' {LEL tx84:04:47.9}

In the case of (18a), the numeral marked with the inclusive and the collective refers to the three governors of the local, traditional government performing an event collectively. In (18b), the numeral refers to a group of four people that includes the speaker, a referent established previously in discourse.

### **10.5 Quantifiers**

There is a small set of quantifiers in Choguita Rarámuri. They are listed in (19). These quantifiers do not take any morphological marking, though they may be morphologically complex themselves. From this list, two quantifiers are derived from numerals: *oˈkwâ* 'few' is derived from the numeral 'two', while *iˈbiri* 'some, a few' is derived from the pluractional form of the numeral one (*biˈlé*) (see §9.3.1.2 for details on pluractional marking). In contrast to their numeral sources, numeral quantifiers do not take any inclusive, collective or other morphology that numerals may take (as described in §10.4 above).

(19) Choguita Rarámuri quantifiers


As shown in (20), these quantifiers can function as modifiers of head nouns in noun phrases, and are compatible with demonstrative pronouns used as definite articles, as in (20b) (the syntactic behavior of these quantifiers is further addressed in Chapter 12).

### 10 Minor word classes

(20) a. *aʔˈlì ˈhípi ko biˈlá ˈmá ˈbéti wiʰˈkâ riˈhòi ˈhípi ko aʔˈlì* and *ˈhípi=ko* today=emph *biˈlá* really *ˈmá* already *ˈbé=ti* emph=1pl.nom *wiʰˈkâ* many *riˈhòi* people *ˈhípi=ko* today=emph 'And today there is a lot of us people.' 'Ahora ya habemos mucha gente.' {SFH tx12:01:00.5} b. *aʔˈlì ˈétʃ͡i ʃiˈnêami raˈlàmuli ko ˈhê aˈníli aʔˈlì* and *ˈétʃ͡i* dem *siˈnêami* everyone *raˈlàmuli=ko* men=emph *ˈhê* dem *aˈní-li* say-pst 'And then all of the men said.' 'Entonces todos los hombres dijeron.' {LEL tx32:06:32.5} c. *suˈwâba naˈmûti raʔlaˈká ˈká ruˈwá biˈlá tʃ͡oˈnà? suˈwâba* all *naˈmûti* things *raʔla-ˈká* buy-ger *ˈká* cop.irr *ru-ˈwá* say-mpass *biˈlá* really *tʃ͡oˈnà* there 'Everything is bought there?' '¿Todo es comprado allá?' {MDH co1137:12:20.8}

Quantifiers may also be used pronominally and fill an argument slot subcategorized for by the verb. This is shown in (21), where quantifiers may fill the agent-subject role in (21a–b) or the object-theme in (21c).


10.6 Definite articles

c. *ˈmáti suˈwâba ruˈjè ˈétʃ͡i ˈtîo ba ˈmá=ti* already=1pl.nom *suˈwâba* all *ru-ˈè* say-appl *ˈétʃ͡i* dem *ˈtîo* uncle *ba* cl 'And we told everything to the uncle.' 'Y ya le dijimos todo al tío.' {LEL tx84:04:29.6}

### **10.6 Definite articles**

Choguita Rarámuri possesses a set of definite articles that encode a definite referent in a noun phrase. These definite articles encode number (singular vs. plural) and affective stance (positive/neutral vs. negative). The Choguita Rarámuri definite articles are listed in (22).

(22) Choguita Rarámuri definite articles


As stated above, definite articles in this language may encode affective stance. "Affect" is defined here as: "... a broader term than emotion, to include feelings, moods, dispositions and attitudes" (Ochs & Schieffelin 1989: 7); see also Neely (2019). In the case of Choguita Rarámuri definite articles, choice of the definite article involves a positive/neutral or negative evaluation of the referent of the head noun of the noun phrase. The positive/neutral article forms are derived from the adjective 'small', which also encodes number distinctions and is also used to mean literal small size. The contrast between positive and negative affective stance in definite articles is exemplified in (23).

	- a. context: the speaker announces the arrival of an acquaintance with whom they have a good/neutral relationship. *ˈtá muˈkî naˈwàli ˈtá* def.good *muˈkî* woman *naˈwà-li* arrive-pst 'The woman arrived.' 'La mujer llegó.'

### 10 Minor word classes

b. context: the speaker announces the arrival of an acquaintance who they dislike. *ˈtí muˈkî naˈwàli ˈtí* def.bad *muˈkî* woman *naˈwà-li* arrive-pst 'The woman arrived.' 'La mujer llegó.'

In this pair of examples, the choice of definite article encodes affective stance toward the referent, either a positive or neutral one with the definite article *ta* (23a) or a negative one with the definite article *ti* (23b).

Positive/neutral evaluation articles are further exemplified in (24).

### (24) Positive/neutral evaluation definite articles


b. *ˈwé ˈá ritiˈwî oˈlá ˈkútʃi uˈmûri ko ˈwé* int *ˈá* aff *ritiˈwî* see.mpass *oˈlá* cer *ˈkútʃi* def.good *uˈmûri=ko* great.grand.parents=emph 'One indeed gets to see (meet) the great grandparents.' 'Uno si ve (conoce) a los bisabuelos.' {ME in485:05:17.0}

In (24a), from the Frog Story narrative, the singular definite article *ta* (from the noun phrase *ta koˈtʃ͡î* 'the dog') refers to an argument that was already introduced in discourse where the speaker expresses a neutral stance. In (24b), the plural definite article *ˈkútʃ͡i* is used with a positive stance, encoding respect and affection towards elders (*uˈmuri*, the great grand parents).

Negative evaluation articles are further exemplified in (25).

(25) Negative evaluation definite articles

a. *ke biˈlé wiˈtʃ͡íami ˈhú es ˈtí ritʃ͡aˈní riˈpári moˈéna ba ke* neg *biˈlé* indeed *wiˈtʃ͡í-ame* fall-ptcp *ˈhú* cop.prs *etʃ͡i* dem *ˈtí* def.bad *ritʃ͡aˈní* giant *riˈpá-ri* above-all *moˈén-a* go.up-prs *ba* cl 'He wouldn't fall the giant (richaní) when he would climb up.'

10.7 Adverbs

'No se caía el gigante (richaní) cuando se iba para arriba.' {SFH tx43:02:52.7}

b. *ˈnàa akiˈná noˈrînima ˈlé ˈtʃéti naˈkôam pa ˈnà* then *akiˈná* over.here *noˈrîni-ma* arrive-fut.sg *aˈlé* dub *ˈtʃéti* def.bad *naˈkô-ame* fist.fight-ptcp *pa* cl 'Then the fierce ones will arrive.' 'Entonces van a llegar los que pelean.' {LEL tx221:03:17.8}

In (25a), the singular article *ti* is used derogatorily in reference to *ritʃ͡aˈni*, a giant in a myth of creation that kidnapped a woman and would eat children. In (25b), the plural definite article *ˈtʃ͡éti* encodes a negative evaluation of a group of people who are fierce and combative.

There are also instances of *ti* and *ˈtʃ͡éti* attested in the corpus where these forms do not convey a negative stance, but rather seem to be deployed in a stanceneutral way. Examples of this are provided in (26).

(26) *aʔˈlì biˈlá ˈhípi biˈlá ˈmá noˈkí maˈkò maˈrí bamˈpáma ˈlé ti tiˈwé niˈhê ˈkútʃ͡ara baˈtʃ͡áwara ba aʔˈlì* and *biˈlá* indeed *ˈhípi* now *biˈlá* indeed *ˈmá* already *noˈkí* almost *maˈkò* ten *maˈrí* five *bam-ˈpá-ma* have.birthday-inch-fut.sg *aˈlé* dub *[ti* def.sg *tiˈwé]* girl *niˈhê* 1sg.nom *ˈkútʃ͡a-la* child-poss *baˈtʃ͡á-wa-la* first-vblz-poss *ba* cl 'And now the girl, my daughter, will turn fifteen, my oldest child.'

'Pues ahora ya mi hija ya va a cumplir quince años mi hija la más grande.' {SFH tx43:04:05.6}

### **10.7 Adverbs**

Choguita Rarámuri has an elaborate system of adverbial elements, which are formally and semantically heterogeneous. This word class is identified primarily by the function of modifying predicates. Formally, they are characterized by a relatively unrestricted distribution and, in a number of cases, the boundary between adverbs and particles is tenuous (see, e.g., discussion of evidential and epistemic particles, addressed in §10.8). Given the formal and semantic diversity that adverbs exhibit, this section addresses the morphological properties of individual adverbial subclasses in terms of their meaning. These subclasses include spatial adverbs (§10.7.1), temporal adverbs (§10.7.2), and manner adverbs (§10.7.3).

### 10 Minor word classes

### **10.7.1 Spatial adverbs**

The spatial adverb subclass is the most elaborate in Choguita Rarámuri, both in terms of the number of semantic contrasts encoded and the different morphological status of individual forms. This subclass also exhibits formal and semantic similarities with demonstratives (described in §10.2 above), though the diachronic developments that led to these similarities is left out of the scope of this grammar. Spatial adverbs can be further classified into deictic (§10.7.1.1) and directional (§10.7.1.2).

### **10.7.1.1 Deictic adverbs**

The subclass of deictic adverbs is defined as the set of elements that situate an event with respect to the deictic center. The set of Choguita Rarámuri deictic adverbs is provided in (27).

### (27) Deictic adverbs


The adverbs *miˈká* 'far' (27c) and *aʔˈmi, waʔˈmi* 'over there, far away' (27d) appear to be morphologically derived from *mi* 'there' (27b). None of the deictic adverbs take any morphology. Examples of these deictic elements are provided in (28).


10.7 Adverbs

'esos los que vivieron (crecieron) mucho primero allá lejos en kanochi, dicen' {SFH tx43:10:45.1}

b. *ˈnápu riˈká haˈré ˈtʃ͡ó aʔˈmí raˈbô piˈrêami ˈnápu* sub *riˈká* like *haˈré* some *ˈtʃ͡ó* also *aʔˈmí* over.there *raˈbô* Rayebó *piˈrê-ami* live.pl-ptcp 'Like others who live there in Rayebó.' 'Así como otros los de allá de Rayebó.' {SFH tx43:00:37.6}

c. *aʔˈmi biˈlá, aʔˈmí ˈwé miˈká siˈmíbira, aʔˈmí biˈléana, biˈléana aʔˈmí* over.there *biˈlá,* really *aʔˈmí* over.there *ˈwé* very *miˈká* far *siˈmíbira,* go.pl.pst *aʔˈmí* over.there *biˈléana,* another.place *biˈléana* another.place 'Over there very far they went, over there to another place, another place.' 'Allá muy lejos se fueron, allá a otra parte, a otra parte.' {LEL tx88:01:30.1}

As shown in these examples, deictic adverbs exhibit a very unrestricted distribution, and some may be combined with one another (e.g., *mi kaˈʔé* in (28a)) and/or with other adverbs (e.g., *we miˈká* in (28c)), but possible combinations and their collocations are restricted, i.e., while *mi* and *kaˈʔé* are frequently compounded, there are no attestations in the corpus of these adverbs with an alternative order.

### **10.7.1.2 Directional adverbs**

Choguita Rarámuri possesses a geomorphic spatial reference system, a type of system where the anchor is defined by a feature or gradient of the environment, e.g., rivers, mountains or other inclinations (O'Meara & Pérez Báez 2011: 842).<sup>6</sup> As noted below, frames of reference where the anchor is the speaker and/or addressee (egocentric intrinsic) are also available to Choguita Rarámuri speakers. As in other geomorphic systems, the orientation of the anchor is relevant, e.g., whether an object is located upstream or downstream from a river depends on which direction the river is flowing. In the case of Choguita Rarámuri, one central anchor is a creek that cross-cuts the valley where Choguita is located. In the following examples (in (29)), the speaker instructs a young child to pull a door lock in opposite directions.

<sup>6</sup>Geomorphic spatial reference systems are subsumed under the 'intrinsic' category of other spatial reference systems classifications (see Levinson 1996).

### 10 Minor word classes

(29) a. *kaˈʔó rakiˈbú! kaˈʔó* upstream *rakiˈbú!* pull.imp.sg 'Pull it upstream!' '¡Jálalo río arriba!' <BFL 09 2:93/el > b. *ˈtû-na rakiˈbú! ˈtû-na* downstream-toward *rakiˈbú!* pull.imp.sg 'Pull it downstream!'

'¡Jálalo río abajo!' <BFL 2011 1:3/el >

Speakers anchor the events which they describe in narratives and conversations through a set of directional and locative roots and a set of directional bound markers that encode movement through a path or from a source. The Choguita Rarámuri directional system bears some resemblances with that of Mountain Guarijío, described in Miller (1996), a language spoken in a similar ecological context (the rugged, mountainous areas of the Sierra Madre Occidental in Northern Mexico).

A non-exhaustive list of directional and locative roots is provided in (30). This list includes terms that involve an egocentric frame of reference, e.g., *boˈʔó* 'opposite bank' (from the speaker and/or addressee perspective) in (30e).

### (30) Directional and locative roots


<sup>7</sup> From the speaker and/or addressee perspective.

### 10.7 Adverbs

The directional vocabulary and grammar of Choguita Rarámuri includes landscape terms that reflect the local ecological context of its speakers. A non-exhaustive list of these terms is provided in (31), with both English and Spanish approximate translations.

(31) Landscape vocabulary


c. *koˈlì*

'around the corner' (diagonally placed in reference to an object and out of sight)

'a la vuelta (diagonalmente ubicado en referencia a un objecto y fuera de vista)'

d. *naˈpûtʃ͡i*

'slight land depression (descent followed by a rise)' 'pequeña depresión en el terreno'


### 10 Minor word classes

Directional and landscape roots may combine with a set of directional bound morphemes that encode motion toward a goal or away from a source. An example is provided in (32), where the *-ki* suffix 'on top of, on the surface of' attaches to the directional root *riˈpá* 'up'.

(32) *"riˈpáki risoˈt͡ʃí biˈtê" aˈníli riˈpá-ki* up-supe *risoˈt͡ʃí* cave *biˈtê"* inhabit.prs *aˈní-li* say-pst "'He lives up on top of a cave", he said.' "'Vive por arriba en una cueva", dijo.' {LEL tx32:02:39.2}

Directional suffixes encode allative (motion towards), superessive (motion or location on the surface of), and ablative (motion away from) meanings, as well as overall orientation of referents (here labelled as adessive). Table 10.7 lists the directional suffixes identified in the Choguita Rarámuri corpus with relevant examples.

The following examples (in (33)) illustrate the allative *-mi, -ri* and *-na* suffixes.

```
(33) Allative suffixes
```



### Table 10.7: Directional suffixes

### 10 Minor word classes

d. *ˈmí riʔˈléna iˈnâlo biˈlé reˈhòi ˈmí* dist *reʔˈlé-na* down-all *iˈnâ-li-o* go-pst-ep *biˈlé* one *reˈhòi* man 'A man is going down there (downhill).' 'Un hombre va para abajo (cerro abajo).' < FMF 09 2:65/el >

The example in (34) shows that the *-ri* suffix can also be used to encode motion away from a source, an ablative meaning.

```
(34) reʔˈréri iˈnâlo
      reʔˈré-ri
      down-abl
                  iˈnâ-li-o
                  go.sg-pst-ep
      'He comes from below.'
      'Viene de abajo.' < FMF 09 2:65 >
```
The following pair illustrates the difference between the description of a static location (in this instance morphologically unmarked) (35a) and description of motion toward a goal, marked with the allative suffix (35b).

(35) Static location vs. motion toward a goal

```
a. maˈt͡ʃí wiˈlí toˈwí
     maˈt͡ʃí
     outside
             wiˈlí
             stand.prs
                        toˈwí
                        boy
    'The boy stands outside.'
    'El niño está parado afuera.' < FMF 09 3:26/el >
b. maˈt͡ʃími
     maˈt͡ʃí-mi
     outside-all
    'toward outside'
    'hacia afuera' < FMF 09 3:26/el >
```
The superessive *-ki* suffix and the allative/ablative *-na* suffix may co-occur in some forms where motion involves motion on top of a surface and towards a goal or away from a source. This is exemplified in (36).

```
(36) a. ˈtûkuna
           ˈtû-ki-na
           downriver-supe-all
           'Going downward (downriver)'
           'De bajada (río abajo)' < FMF 09 2:65/el >, {ME in485:07:31.4}
```
10.7 Adverbs

b. *riˈpákina ˈku tʃ͡uˈkúl ti tʃ͡imoˈlí ko iʔˈnèka tʃ͡ú oˈlása riˈpá-ki-na* up-supe-all *ˈku* rev *tʃ͡uˈkú-li* to.be.curved-pst *ti* def.bad *tʃ͡imoˈlí=ko* squirrel=emph *iʔˈnè-ka* watch-ptcp *t͡ʃú* what *oˈlá-sa* do-cond 'From up above the squirrel was watching, what will it do.' 'Desde por arriba estaba la ardilla viendo, a ver qué estaba haciendo.' {BFL tx191:03:37.6}

As shown in (36a), the superessive *-ki* suffix may undergo round vowel harmony in specific phonological contexts (details of the process of round vowel harmony is discussed in §9.4.3.5).

### **10.7.2 Temporal adverbs**

Temporal adverbs can be classified as deictic or non-deictic. Deictic temporal adverbs are defined as those that situate the event in time with respect to the speech event or another reference point. A list of deictic temporal adverbs is provided in (37).

### (37) Deictic temporal adverbs


In contrast to deictic temporal adverbs, non-deictic temporal adverbs situate events within periods of time, whether concerning the diurnal cycle (38) or the seasonal cycle (39).<sup>8</sup>

(38) Non-deictic temporal adverbs - diurnal cycle


<sup>8</sup>The form *biˈʔà roˈkò* 'early morning' (lit. 'early night') in (38d) refers to the time of the day before sunrise.)

### 10 Minor word classes


Some temporal adverbs are morphologically related to other word classes: this is the case with *baˈrá* 'rainy season' (39a), which is related to the verb *baˈrâmi* 'be thirsty'. None of the temporal adverbs takes any morphological marking.

### **10.7.3 Manner adverbs**

Choguita Rarámuri has a small set of manner adverbs, some of which are exemplified in (40). As described in §10.3, the root *waʔˈrû* may function both as an adverb modifying predicates ('greatly, a lot') or as an adjective modifying nouns ('big, sg.').<sup>9</sup>


Manner adverbs do not take any inflectional or derivational morphology, nor do they appear to be morphologically derived from other word classes. As discussed in Chapter12 (§12.1.5), adverbs may also modify adjectives in noun phrases.

### **10.8 Discourse particles and enclitics**

From all the minor word classes, discourse particles and enclitics constitute a set of heterogeneous word classes that are closed and morphologically simple,

<sup>9</sup> Some of the forms in this list are homophonous with forms in other word classes, such as *ˈtʃ͡áti*, which is homophonous with adjective *ˈtʃ͡áti* 'ugly'.

10.8 Discourse particles and enclitics

bearing no inflection or derivation. Each class of particles is composed of fewer than a dozen members per class. Particles have a wide range of functions and meanings. Some, but not all, particles are also characterized by being phonologically reduced, and not being subject to the minimality requirement that holds for verbs (see §11.2 for discussion of minimality effects in Choguita Rarámuri).

### **10.8.1 Interjections**

Interjections are characterized by their ability to be uttered in isolation as a complete utterance, and as not belonging to any other word class. The inventory of interjections attested in the Choguita Rarámuri corpus is provided in (41) with approximate English and Spanish translations (some forms from this set are loanwords from Spanish, e.g. (41f)).



Interjections tend to appear as independent clauses and appear separated from the rest of a larger utterance through a pause (indicated as <...> pr <,>). This is exemplified in (42) (which represents a section of a dialogue) and (43).

(42) a. [ME]: *boniˈlâ... boni-ˈlâ* be.younger.brother-poss 'younger brother' 'hermano menor' b. [SFH]: *boˈnèsa ba?*

*boˈnè-sa* be.younger.brother-cond *ba?* cl 'As if he were a younger brother?' '¿Como si fuera hermano menor?'

### 10 Minor word classes

c. [ME]: *uˈrí, boniˈlâ ˈnísa ˈlá ba ˈni uˈrí,* yes *boni-ˈlâ* be.younger.brother-poss *ˈní-sa* cop-cond *oˈlá* cer *ba* cl *ˈni* emph 'Yes, as if he were a younger brother indeed.' 'Si, si fuera hermano menor, asi es.' {ME in485:02:22.9}

(43) *ˈmá ˈpé tʃ͡iˈkí ˈhú pa ... maˈtêtala ba ˈmá* already *ˈpé* just *tʃ͡iˈkí* that.much *ˈhú* cop.prs *pa* cl *maˈtêtala* thank.you *ba* cl 'That is all (lit. that is the little bit there is) thank you.' 'Ya es todo ... gracias.' {LEL tx88:03:47.4}

In both of these examples, the interjections are separated from the following (42c) or preceding (43) clause by a pause and an intonational break.

### **10.8.2 Connective particles**

Connective particles are defined as particles that serve the function of connecting sentences within discourse. A list of monomorphemic connective particles in Choguita Rarámuri is provided in (44).

(44) Monomorphemic connective particles


These connective particles do not take any morphology and are analyzed here as basic, but other connectives may be morphologically complex. Among the basic connectives, *aʔˈlì* 'and, then, so' (44a) (derived from the temporal adverb *aʔˈlì* 'late in the day', 'tarde en el día') may combine with other morphemes to form complex connectives. These and other morphologically complex connectives are exemplified in (45).

(45) Complex connectives


Connectives, whether monomorphemic or complex, occur in between the two sentences they connect. This is illustrated in the examples in (46), (47) and (48).

10.8 Discourse particles and enclitics

(46) a. *ˈnà napaˈwía noˈkáli aˈlé ˈet͡ʃi ˈnà biˈlé reˈhòi aʔˈlì biˈlé biˈlé t͡ʃaˈbôt͡ʃi ˈnà* dem *napaˈwí-a* get.together-prog *noˈk-á-li* move-tr-pst *aˈlé* sub *ˈet͡ʃi* dem *ˈnà* dem *biˈlé* one *reˈhòi* man *aʔˈlì* and *biˈlé* one *biˈlé* one *t͡ʃaˈbôt͡ʃi* mestizo *sî* also 'A (Rarámuri) man and a mestizo man got together.' 'Se juntaron un hombre (rarámuri) y un mestizo.' < SFH tx choma (2) > b. *aʔˈlì beˈlá ko beˈlá ko maˈníìla ruˈámi ka aʔˈlì* and *beˈlá=ko* really=emph *beˈlá=ko* really=emph *maˈní-ì-la* be.liquid-appl-rep *ru-ˈá=mi* say-mpass=dem *ka* cop.irr 'And then it is said they were given something (a liquid).' 'Y entonces dicen que les pusieron algo (un líquido).' < SFH tx choma (3) > (47) a. *ˈkíti beˈlá we ˈkíti* that.is.why *beˈlá* really *we* int 'that is why' 'por eso' < SFH tx choma (24) > b. *aʔˈlì tʃ͡iˈhônsa ˈét͡ʃi reˈhòi baˈhîsa ˈká ˈét͡ʃi t͡ʃoʔˈmá aʔˈlì* and *tʃ͡iˈhônsa* thene *ˈét͡ʃi* dem *reˈhòi* man *baˈhî-sa* drink-cond *ˈká* irr *ˈét͡ʃi* dem *t͡ʃoʔˈmá* snot 'and if the man had drunk the snot...' 'y si el hombre hubiera tomado el moco...' < SFH tx choma (25) > (48) a. *biˈlé naˈmûti tʃ͡iˈká noˈrîno, waʔˈlû roˈsâkame biˈlé* one *naˈmûti* thing *tʃ͡iˈká* over.here *noˈrîno,* arrive *waʔˈlû* big.sg *roˈsâ-kame* be.white-pst.ptcp 'One thing came, a white, big thing.' 'Una cosa vino, una cosa blanca, grande.' {ROF tx\_rosakame:00:13.7} b. *aʔˈlì tamuˈhê ˈpé maˈhâli ko aʔˈlì* then *tamuˈhê* 1pl.nom *ˈpé* little *maˈhâ-li=ko* get.scared-pst=emph 'And then we got scared.'

### 10 Minor word classes

'Y entonces nosotros nos asustamos.' {ROF tx\_rosakame:00:17.1}

The interpretation and function of connectives in complex clauses is discussed in Chapter 15.

### **10.8.3 Negative particles**

Choguita Rarámuri possesses two negative particles, *ke*, a clausal negator (which can be used as interjection), and *ˈkíti*, a prohibitive (negative imperative) particle. Negative particles may combine with other morphemes, yielding morphologically complex negative markers: in Choguita Rarámuri, the clausal negator *ke* combines with other morphemes. The set of negative particles and complex negative markers available in the language are provided in Table 10.8, with their gloss, function and approximate translation.


Table 10.8: Negative markers

The syntactic properties and functions of each of these negative particles and negative constructions are discussed in further detail in §14.3.

### **10.8.4 Epistemic particles and enclitics**

As discussed in §9.5, Choguita Rarámuri possesses modality particles that encode epistemic stance. Epistemic modality is defined here as the expression of the degree of certainty speakers have towards the actuality of an event or its potential

10.8 Discourse particles and enclitics

of occurring at a time later than the speech event. A list of particles that encode epistemic stance is provided in (49).

(49) Epistemic modality particles


The dubitative *aˈlé* (with reduced form *lé*) and certainty *oˈlá* (with reduced form *lá*) markers are frequently attested post-verbally. Their different epistemic functions can be appreciated in the contrast in meaning they create when associated to predicates inflected for future tense: in (50a), the dubitative particle encodes lack of knowledge on the speaker's part, while in (50b), the certainty marker conveys a high degree of commitment from the speaker about the likelihood the event encoded by the verbal predicate will take place in a future time.<sup>10</sup>

(50) Epistemic modality markers


While frequently attested with verbs inflected with future tense, epistemic markers are not obligatory in these contexts. Lack of an epistemic marker involves a neutral interpretation with respect to the speaker's commitment to the likelihood the event will take place in these contexts. Examples of this are shown in (51) (relevant predicates lacking epistemic markers are highlighted in boldface).

<sup>10</sup>As described in §9.5, epistemic particles are vowel initial and may coalesce with the final vowel of the future suffix.

### 10 Minor word classes

(51) a. *"ˈmán ku ʃiˈmêa ˈhípi ko, ˈmá ˈwé miˈká iˈnârtʃ͡ani" ˈhê aˈníli ˈétʃ͡i tiˈwé ko "ˈmá=ni* already=1sg.nom *ku* rev *si-ˈmêa* go.sg-fut.sg *ˈhípi=ko,* now=emph *ˈmá* already *ˈwé* int *miˈká* far *iˈnâr-tʃ͡ani"* go.around-ev *ˈhê* dem *aˈní-li* say-pst *ˈétʃ͡i* dem *tiˈwé=ko* young.woman=emph "'I'm already going now, it already sounds like they're far away" said the young woman.' "'Ya me voy a ir, ya se oye muy lejos" dijo la muchacha.' {LEL tx32:10:34.5} b. *aʔˈlì ˈnè ko ˈmá bitiˈtʃ͡í ˈá buˈjèa aˈtí, noˈrînima ˈétʃ͡i biˈléara tʃ͡oˈkêami*

*taˈmí ruˈwèʃia aʔˈlì* and *ˈnè=ko* 1sg.nom=emph *ˈmá* already *bitiˈtʃ͡í* house *ˈá* aff *buˈjè-a* wait-prog *aˈtí,* cop.sg *noˈrîna-ma* arrive-fut.sg *ˈétʃ͡i* dem *biˈléara* another *tʃ͡oˈkê-ame* settle.bet-ptcp *taˈmí* 1sg.acc *ru-ˈwè-si-a* tell-appl-mot-prog

'And then I wait for her in the house, the other bet settler arrives to tell me.'

'Y entonces yo ya la espero en la casa, viene la otra chokéami a decirme.' {LEL tx19:01:17.9}

c. *aʔˈlì tʃ͡iˈhônsa ˈmám aˈtʃèma aˈsûkar aʔˈlì aˈrîna aʔˈlì* and *tʃ͡iˈhônsa* then *ˈmá=mi* already=2sg.nom *aˈtʃ͡-è-ma* add-appl-fut.sg *aˈsûkar* sugar *aʔˈlì* and *aˈrîna* flour 'And then you add to it sugar and flour.'

'Y entonces ya le echas la azúcar y la harina.' {BFL tx60:02:08.1}

In (51a), the main predicate inflected for future tense that lacks an epistemic marker is reported speech, while in (51b) and (51c), the relevant future predicates with no epistemic markers are part of procedural texts, a description of the procedure for settling a bet for a women's race and a recipe for making corn beer, respectively.

The dubitative marker is also attested with verbs inflected for other tense/ aspect/mood specifications, and may have specific modal interpretations when combined with certain inflection values. When combined with past or present progressive inflected verbs, the dubitative particle encodes lack of certainty from the speaker about a past event (52a–b) or a current event (52c).

10.8 Discourse particles and enclitics

(52) a. *ah! ˈmí ruˈwèli aˈlé ko ah* oh *ˈmí* 2sg.acc *ru-ˈè-li* say-appl-pst *aˈlé=ko* dub=emph 'Oh! They told me, I think.' 'Ah! Me dijeron, creo.' {FLP in243:08:07.9}


When combined with a verb inflected for conditional mood, the dubitative particle confers a deontic reading: in (53a), the speaker states that an event (giving water to the horse) is an obligation, while in (53b), the speaker encourages his children to not forget rituals and carry them to future generations.

(53) a. *baʔˈwí maˈnèsa aˈlé ra ba baʔˈwí* water *maˈn-è-sa* be.located.water-appl-cond *aˈlé* dub *ra* ra *ba* cl 'We must give water to the horse.' 'Hay que darle agua al caballo.' {JLG co1236:06:21.3} b. *ke biˈlá ˈtâsi wikaˈwâsa ˈlé naˈlìna ˈhípi tamuˈhê... ke* neg *biˈlá* really *ˈtâsi* neg *wikaˈwâ-sa* forget-cond *aˈlé* dub *naˈlìna* but *ˈhípi* today *tamuˈhê* 1pl.nom '(we) should not forget, but nowadays we...' 'no debemos olvidar nosotros, nomás que ahora nosotros ...' {SFH tx475:08:33.0}

Like the dubitative marker, the certainty particle *oˈlá* is attested with predicates inflected with a range of tense/aspect/mood values. This is exemplified in (54).

### 10 Minor word classes

(54) a. *"ˈnè ko ˈjéna ˈá ˈwé kaˈlé, ˈnè ko ˈá riˈká ˈwé tiˈbúmo ˈlá, ˈá oˈtʃ͡êrima ˈlé" ˈnè=ko* 1sg.nom=emph *ˈjéna* aff *ˈá* aff *ˈwé* int *kaˈlé* love.prs *ˈnè=ko* 1sg.nom=emph *ˈá* aff *riˈká* like *ˈwé* int *tiˈbú-ma* take.care-fut.sg *oˈlá* cer *ˈá* aff *oˈtʃ͡êri-ma* grow-fut.sg *aˈlé* dub "'I do love him very much, I will always take care of him, he will grow." "Yo si lo quiero mucho, yo siempre lo voy a cuidar mucho, sí va a crecer." {LEL tx32:14:11.1} b. *benˈtûra riˈwè oˈlá ˈro, kaˈlîstro onoˈlâ ro benˈtûra* Ventura *riˈwè* be.named.prs *oˈlá* cer *ˈru,* say *kaˈlîstro* Calixto *ono-ˈlâ* father-poss *ro* say.prs 'They say Ventura was his name, Calixto's father.' 'Dicen que se llamaba Ventura el papá de Calixto.' {ME in484:02:19.1} c. *ˈí biˈláni ˈnà ruˈmê oˈlá ˈnápu riˈkám ˈnâriani ˈnâria ˈnà tʃ͡ú riˈká bukuˈwêruwa ˈí* here *biˈlá=ni* indeed=1sg.nom *ˈnà* this *ru-ˈmêa* say-fut.sg *oˈlá* cer *ˈnápu* sub *riˈká=mi* like=2sg.nom *ˈnâri-a=ni* ask-prog=1sg.acc *ˈnâri-a* ask-prog=1sg.nom *ˈnà* this *tʃ͡ú* how *riˈká* how *bukuˈwêruwa* make.bukuweruwa 'Here I will tell because I was asked about how the bukuwéruwa ritual works.' 'Aqui voy a contar porque me preguntaste sobre el bukuwéruwa.' {SFH tx475:00:27.6}

The certainty marker *oˈlá* may also convey deontic mood when combined with future tense inflected predicates. This is shown in (55).

(55) *naˈlìna aʔˈlá ˈnâtika nokiˈbôo oˈlá ˈnà naˈlìna* but *aʔˈlá* well *ˈnâti-ka* think-ger *noki-ˈbô* do-fut.pl *oˈlá* cer *ˈnà* then 'But you all should think well.' 'Nomás que tienen que pensar bien.' {SFH tx12:11:31.4}

The particles *a* and *biˈlá* have a different distribution than the dubitative and certainty markers, and generally appear preceding the predicate or other particles in the beginning of sentences or clauses. These markers appear in contexts

10.8 Discourse particles and enclitics

where speakers claim epistemic authority about the propositional content of the utterance. This is exemplified in the following examples: in (56a), the speaker, an expert seamstress in the community, begins a narrative in which she describes a procedure of making a skirt; in (56b), the speaker recounts an experience from her childhood.

(56) a. *ˈnè ko biˈlá aniˈmêa oˈlá tʃ͡ú reˈká niwaˈrîa biˈlé ˈpúra ba ˈnè=ko* 1sg.nom=emph *biˈlá* really *ani-ˈmêa* say-fut.sg *oˈlá* cer *tʃ͡ú* Q *reˈká* how *niwaˈrîa* make-mpass-prog *biˈlé* one *ˈpúra* belt *ba* cl 'I am going to say how a faja (belt) is made.' 'Yo voy a decir como se hace una faja.' {BFL tx1:00:20.7} b. *aʔˈlì ko ˈtʃ͡ó ˈnà ˈhônsa ko ˈmá biˈlán ˈá riˈwá ˈtʃ͡ó ʃi ˈtʃ͡ó ba aʔˈlì=ko* and=emph *ˈtʃ͡ó* dem *ˈnà* prox *ˈhônsa=ko* since=emph *ˈmá* already *biˈlá=ni* indeed=1sg.nom *ˈá* aff *riˈwá* see *ˈtʃ͡ó* also *ʃi* also *ˈtʃ͡ó* also *ba* cl 'And then from there I also saw him.' 'Y entonces de ahí yo también ya vi.' {LEL tx71:04:36.3}

Other markers encoding epistemic stance in Choguita Rarámuri include evidential markers. This includes the reportative markers -*la* (reportative, different subject) and *-lo* (reportative, same subject),<sup>11</sup> as well as the auditory evidential *ˈtʃáne* suffix.

### **10.8.5 Pragmatic enclitic**

Choguita Rarámuri also possesses a pragmatic marker that confers prominence to a word or phrase within discourse, the emphatic enclitic *=ko*. <sup>12</sup> This morpheme is fairly unrestricted in terms of the word classes and syntactic constituents it can attach to, which is taken as evidence of its status as a clitic. In terms of

<sup>11</sup>Miller (1996) describes a reportative enclitic *=ra* in Mountain Guarijío (1996:312), though no alternations are discussed that would reflect a contrast between same vs. different subject as in Choguita Rarámuri.

<sup>12</sup>The term 'emphatic' is used in the description of cognate forms of other Rarámuri and Guarijío varieties, including Western Tarahumara (Burgess 1984), Urique Rarómari (Valdez-Jara 2013), Rochéachi Rarámuri (Morales Moreno 2016), and Mountain Guarijío (Miller 1996), among others.

### 10 Minor word classes

its function, this emphatic enclitic is described in other varieties as encoding pragmatic statuses compatible with focus (Valdez-Jara 2013) or both focus and topic, as proposed by Morales Moreno (2016) for Rochéachi Rarámuri, and as suggested by Miller (1996) for Mountain Guarijío (1996: 313–315).

The following examples illustrate the distribution of this enclitic, which may attach to nouns (57a–c), pronouns (57b, d–e), adverbs (57f) and conectives (57g–h). As shown in (57h), the emphatic enclitic may also appear in final clause position.

(57) a. *aʔˈlì tʃ͡aˈbôtʃ͡i ko ˈwé biˈlá raˈʔìla baˈhîla ruˈá ˈétʃ͡i tʃ͡oʔˈmá ba aʔˈlì* and *tʃ͡aˈbôtʃ͡i=ko* mestizo=emph *ˈwé* int *biˈlá* truly *raˈʔì-la* like-rep *baˈhî-la* drink-rep *ru-ˈwá* say-mpass *ˈétʃ͡i* dem *tʃ͡oʔˈmá* snot *ba* cl 'And the Mexican man they say he really enjoyed drinking it.' 'Y el mestizo dicen que se lo tomó muy a gusto.' {SFH tx128:01:47.8} b. *"hierbas" ko ba, ke beti ... ke beni maˈtʃí ˈnè ko ba hierbas=ko* grasses=emph *ba,* cl *ke* neg *be=ti* really=1pl.nom *ke* neg *be=ni* really=1sg.nom *machí* know.prs *ˈnè=ko* 1sg.nom=emph *ba* cl '(They call them) "grasses", we don't ... I don't know, me.' '(Les dicen) "hierbas", nosotros no ... yo no se, yo.' {GFM tx785:02:22.0} c. *amiˈná haˈré ko ˈá ˈmá biˈlá amiˈná* later *haˈré=ko* some=emph *ˈá* aff *ˈmá* already *biˈlá* indeed 'later the other ones already did' 'después los otros ya' {SFH tx475:09:52.9} d. *ˈtâsi biˈlé maˈtʃí ˈnè ko ba ˈtâsi* neg *biˈlé* one *maˈtʃí* know.prs *ˈnè=ko* 1sg.nom=emph *ba* cl 'I don't know.' 'Yo no se.' {ME in485:08:05.6}

10.8 Discourse particles and enclitics

e. *aʔˈlì biˈlá ˈá maˈtʃía ruˈwá biˈnôi ko ˈnà ... aʔˈlì* and *biˈlá* indeed *ˈá* aff *maˈtʃí-a* know-prog *ru-ˈwá* say-mpass *biˈnôi=ko* emph.sg=emph *na* then 'and they say that he himself knows when' 'y dicen el mismo sabe cuando' {LEL tx71:02:50.2} f. *ʔwîbo ˈmá biˈlá ˈhípi ko ba ʔwî-bo* harvest-fut.pl *ˈmá* already *biˈlá* truly *ˈhípi=ko* now=emph *ba* cl 'Now we will harvest (it is time to harvest).' 'Ahora vamos a pizcar (es tiempo de pizca).' {MDH co1136:02:29.9} g. *aʔˈlì ko biˈláti ˈmá ˈpé a iˈwêami ˈnòtʃ͡ika koʔˈpô aˈlé ˈmá aʔˈlá aʔˈlì=ko* and=emph *biˈlá=ti* really=1pl.nom *ˈmá* already *ˈpé* just *a* aff *iˈwê-ame* strong-ptcp *ˈnòtʃ͡i-ka* work-ger *koʔ-ˈpô* eat-fut.pl *aˈlé* dub *ˈmá* already *aʔˈlá* well 'In that time we will work hard to eat.' 'Ya en ese tiempo ya vamos a trabajar duro para comer.' {SFH tx12:08:47.7} h. *aʔˈlì ko tamuˈhê ˈpé maˈhâli ko aʔˈlì=ko* then=emph *tamuˈhê* 1pl.nom *ˈpé* little *maˈhâ-li=ko* get.scared-pst=emph

'Then we got scared.'

'Entonces nosotros nos asustamos.' {ROF tx\_rosakame:00:17.1}

In (57a–c), the emphatic enclitic attaches to nouns whose referents have been introduced previously in discourse: a protagonist of a narrative, a mestizo man in (57a); the topic of a conversation, medicinal plants in (57b); and a group of people in a historical narrative in (57c). The examples in (58) show that the emphatic enclitic *=ko* may also attach to interrogative pronouns, which may be taken as evidence of the status of the emphatic enclitic as a marker of pragmatic focus (see Morales Moreno (2016) on discussion of the distribution of the emphatic enclitic in Rochéachi Rarámuri).

(58) a. *ma ˈpîri ko naˈʔèbo muˈnî baʔaˈlî pa? ˈmá* so *ˈpîri=ko* what=emph *naˈʔ-è-bo* fire-make-fut.pl *muˈnî* beans *baʔaˈlî* tomorrow *pa?* cl 'So what are we going to make the fire with (to cook) the beans tomorrow?'

### 10 Minor word classes

'¿Entonces qué le vamos a echar de lumbre a los frijoles (para cocinar) mañana?' {MDH co1136:01:12.8}

b. *ˈétʃ͡i oˈtʃêrami ke ˈlé pa, ˈpîri ko ˈhú aˈlé? ke piˈlá=m matʃ͡i-ˈsâa ˈníla ba ˈétʃ͡i* dem *oˈtʃêrami* old.people *ke* neg *aˈlé* dub *pa* cl *ˈpîri=ko* what=emph *ˈhú* cop.prs *aˈlé* dub *ke* neg *piˈlá=mi* really=dem *matʃ͡i-ˈsâ* know-cond *ˈní-la* cop-pot *ba* cl

"Well, I think its the old people, or what might it be? Because I don't know what it might be.'

"Pues pienso que son los viejitos, o qué pueda ser? Yo no se que sea. {LEL tx223:04:09.2}

In the following example, the emphatic enclitic confers pragmatic focus to a second person pronominal form in a question (59b), as part of a conversational exchange.

(59) a. [MDH]*ˈwé riˈmùli roˈkò ˈwé* int *riˈmù-li* dream-pst *roˈkò* night 'He did dream it a lot last night.' 'Si lo soñó mucho en la noche.' {MDH co1136:18:43.0} b. [MDH]: *ˈmò ko ba? ˈmò=ko* 2sg.nom=emph *ba* cl 'And you?'

'¿Y tu?' {MDH co1136:18:49.1}

The emphatic enclitic may undergo a post-lexical process of lenition, with the voiceless plosive onset surfacing as voiced (pos-lexical lenition of voiceless plosives is addressed in §3.5.1). Examples of the lenited production of *=ko* are shown in (60).

(60) a. *taˈmò ko ˈhê riˈgá ˈnòtʃ͡ami hu ˈnà iˈsêligam go ... taˈmò=ko* 1pl.nom=emph *ˈhê* dem *riˈká* like *ˈnòtʃ͡-ame* work-ptcp *hu* cop.prs *ˈnà* dem *iˈsêli-kame=ko* be.governor.pl-pst.ptcp=emph 'We, that is how we work, the ones who have been governors...' 10.8 Discourse particles and enclitics

'Nosotros los que hemos sido gobernadores así trabajamos...' {JMF tx816:00:00.0}

b. *aʔˈlám riˈkátʃ͡imi aʔˈlá iˈwéami raʔaˈmâbi ˈlé fɣo a mí raʔaˈmâmi ˈlé paˈgótami baʔaˈlî aʔˈlá=mi* well=2pl.acc *riˈkátʃ͡i=mi* like.that=2pl.acc *aʔˈlá* well *iˈwé-ame* strong-ptcp *raʔaˈmâ-bi* give.advice-irr.pl *aˈlé=ko* dub=emph *a* aff *mí* there *raʔaˈmâ-mi* give.advice-irr.sg *aˈlé* dub *paˈgótami* people *baʔaˈlî* tomorrow 'Perhaps tomorrow people will give you all advice.' 'A lo mejor de aqui a mañana llegan gentes a darles consejos.' {MFH tx1132:00:30.3}

The cognate forms of this enclitic in Guarijío (*=ka* in River Guarijío and *=ga, =go* in Mountain Guarijío) are described with similar functions as the ones attested in Choguita Rarámuri. Miller (1996) proposes this enclitic is diachronically derived from a copular predicate, *=ga*, plus a subordination suffix *-o* (see §13.2.1 for description of copular predicates in Choguita Rarámuri).

### **10.8.6 Final particles**

The particle *pa* (the onset of which may also undergo post-lexical lenition, as described in §3.5.1), may be categorized as a discourse particle (glossed here as cl) that has the main function of marking syntactic and/or discourse boundaries. It is frequently attested in natural discourse (narratives and conversations) in clause and sentence final position, generally followed by a pause and/or an intonation break (see also Morales Moreno 2016 for a description of the function and distribution of this particle in Rochéachi Rarámuri). Examples of the final particle *pa* are provided in (61). In these examples, clauses are marked with square brackets.

(61) a. *aʔˈlì ˈétʃ͡i biˈláti beˈnèli tamuˈhê ba tʃ͡ú riˈká tiˈbúsa ˈlé pa ˈnà kaˈwì ba aʔˈlì* and *ˈétʃ͡i* dem *biˈlá=ti* really=1pl.nom *beˈnè-li* learn=pst *tamuˈhê* 1pl.nom *ba]* cl *tʃ͡ú* how *riˈká* how *tiˈbú-sa* care-cond *aˈlé* dub *pa]* cl *ˈnà* dem *kaˈwì* land *ba* cl 'So this is how we learnt, how to tend for it, the land.' 'Entonces así aprendimos nosotros, cómo cuidarla, la tierra.' {SFH tx977:00:60.0}

### 10 Minor word classes

b. *ˈnápu koˈlìki biˈtí ba ˈnápu* sub *koˈlì-ki* around.the.side-loc *biˈtí* lie.pl.prs *ba]* cl 'Like the ones who lie in that other side (by the graveyard).' 'Como los que están (acostados) de aquel lado (del panteón).' <GFM co1136[17\_430-17\_445]> c. *aʔˈlì biˈlá ko waˈbé biˈlá kiˈʔà ˈníla ra pa kuˈrí ke ˈtʃ͡ó me biwaˈtʃ͡êatʃ͡i ˈnà kaˈwì β̩a aʔˈlì* and *biˈlá=ko* indeed=emph *waˈbé* int *biˈlá* indeed *kiˈʔà* long.ago *ˈní-la* cop-rep *ra* rep *pa]* cl *kuˈrí* recently *ke* neg *ˈtʃ͡ó* yet *me* almost *biwaˈtʃ͡ê-a-tʃ͡i* solidify-prog-temp *ˈnà* this *kaˈwì* earth *ba]* cl 'And so it was a long time before this earth was solid.' 'Y entonces fue mucho cuando todavía no amacizara este mundo.' <tx43[11\_112-11\_182]> d. *baˈhîm baʔˈwí pa? baˈhî=mi* drink=2sg.nom *baʔˈwí* water *pa?* cl 'Did you drink water?' '¿Tomaste agua?' {MDH co1140:12:10.8}

As shown in these examples, the final particle *=pa* is not only attested at the end of declaratives, but may also close interrogative clauses (61d).

# **11 Prosody: domains and interactions**

This chapter is devoted to prosodic structures and processes that cross-cut the grammar of Choguita Rarámuri, including lexical phonological processes, morphological processes and post-lexical phonological phenomena. Choguita Rarámuri is a prosodically complex language, with both phonological and morphological factors affecting the realization of prosodic phenomena. Many aspects of the description addressed in this chapter are already provided in other parts of the grammar, including Chapter 6 and Chapter 14. This chapter is devoted to examining in detail the interactions of prosodic patterns and processes across grammatical domains in this Uto-Aztecan language with the goal of addressing how these complex interactions yield surface forms in the language.

The structure of this chapter is as follows. §11.1 presents the criteria for determining the Prosodic Word in Choguita Rarámuri, and discusses the domains for phonological and morpho-phonological processes below the level of the Prosodic Word. §11.2 discusses the relationship between different phenomena in Choguita Rarámuri that are quantity-sensitive and the lack of contrastive vowel length in the language. §11.3 addresses the interaction between stress and tone and between lexical tone and grammatical tone in morphologically complex words. §11.4 summarizes the tonal and non-tonal encoding of intonation in Choguita Rarámuri and the interaction between lexical tone and intonation in this language. Finally, §11.5 addresses prosodic constraints associated to different morphological constructions and their relationship to the general prosodic organization of the language.

### **11.1 Defining the Prosodic Word and other prosodic domains in Choguita Rarámuri**

In this grammar, I adopt the proposal in Selkirk (1980; 1996); Nespor & Vogel (1986) and Hayes (1989) for the prosodic hierarchy schematized in (1), where the Prosodic Word is the smaller unit within the hierarchy and the Intonational Phrase its largest constituent.

11 Prosody: domains and interactions

```
(1) The prosodic hierarchy
    Intonational Phrase
           ⎜
    Phonological Phrase
           ⎜
    Clitic Group
           ⎜
    Prosodic Word
```
The Prosodic Word in Choguita Rarámuri can be identified through phonotactic constraints and as the domain of application of several phonologically-general processes. These diagnostic criteria include:

	- a. Each prosodic word is assigned a single, main stress within the first three syllables, which bears a lexical tone (§5.1).
	- b. Inflected verbs are minimally bimoraic (§7.2).
	- c. Prosodic words are vowel final (§5.2).
	- d. Body part incorporation combines two morphological roots into a single prosodic word, where the new lexical item is assigned a single, main stress in the first syllable of the head of the compound (§9.3.3).
	- e. The glottal stop only emerges within the first two syllables of the prosodic word (§7.1).

Taken together, these diagnostic criteria motivate the identification of prosodic words in Choguita Rarámuri. While the domain of stress assignment and of the distribution of the glottal stop are coextensive with the left edge of the prosodic word, the right edge boundary of the prosodic word is characterized by avoidance of vowel reduction and deletion.<sup>1</sup> Thus, identifying the Prosodic Word in Choguita Rarámuri involves a convergence of evidence that characterizes different edges of this prosodic unit. The minimality condition, which requires all Prosodic Words in the language to be minimally bimoraic, and the requirement of having a single stress per Prosodic Word, are the only criteria that hold for entire Prosodic Words in this language.

<sup>1</sup>This restriction of the right edge may have a functional motivation, since TAM inflectional values are often encoded by single vowel exponents that may replace the final vowel of the stem to which they attach (see Chapter 9).

### 11.1 Defining the Prosodic Word and other prosodic domains

There is also evidence for phonological domains below the word level. As discussed in detail in Chapter 9, Choguita Rarámuri exhibits a large number of phonological processes that apply to restricted subconstituents of complex verbs. The notion of morphological domains for phonological processes is thus an important one in this language, as well as the notion of mismatches between phonological and morphological domains, as analyzed in Bantu languages (Hyman 1987; 1998; 2008; Odden 1996; Myers 1998; Downing 2001; Bickmore 2007), and Athabaskan languages (McDonough 2000), among other language families (see also Inkelas 2014 for an overview). These domains are motivated in Chapter 9, and can be summarized as follows in terms of the phonological processes attested in the language.

	- a. **Haplology**: in a sequence of two unstressed, syllables with identical onset consonants at the Inner Stem plus suffix boundary, one syllable gets deleted.
	- b. **Compensatory lengthening**: deletion of a vowel or syllable in a suffix triggers compensatory lengthening of an adjacent Inner Stem stressed vowel.
	- c. **Passive-conditioned lengthening**: the past passive *-ru* suffix triggers lengthening of an adjacent Inner Stem stressed vowel.
	- d. **Imperative singular stress shift**: the imperative singular may be encoded morphologically via a rightward stress shift within the Derived Stem.
	- e. **Round vowel harmony**: non-round vowels of certain suffixes may become round when preceded by a stressed round stem vowel within the Aspectual Stem.

An additional domain below the word level is a constituent composed by the Stem (root plus any stem-forming suffixes), plus the first morphological layer that attaches to the Stem (see Chapter 3 and Chapter 9), whatever this morphological layer may be. Furthermore, in terms of stress assignment, the classification of constructions in shifting vs. neutral distinguishes suffixing constructions that are either *cohering*, forming part of the stress domain, or *non-cohering*, being outside of the stress domain, respectively (Chapter 3). The cohering vs. non-cohering distinction represents a structural mismatch between phonological domains and morphological subconstituents: cohering suffixes form a single phonological domain with the base they attach to, while non-cohering form a separate domain.

### 11 Prosody: domains and interactions

Finally, this chapter addresses whether some phonological and morphophonological processes in the language are better analyzed as being constrained by metrical structure (see §11.5.4).

### **11.2 Vowel length, stress and minimality effects**

Recall from Chapter 3 that there is no evidence for contrastive vowel length in Choguita Rarámuri, yet there is evidence that quantity plays a role in the phonological grammar of the language:

	- a. A minimal prosodic word is bimoraic (§11.2).
	- b. Vowel deletion may trigger compensatory lengthening (§9.4.3.2).
	- c. The past passive suffix conditions vowel lengthening in the stressed stem syllable (§9.4.3.3).

While quantity distinctions are referenced by different phonological and morphophonological processes, vowel length is not relevant for stress assignment in Choguita Rarámuri. As discussed in Caballero (2008) and Caballero (2011b) (and in more detail in §11.3.1 below), a metrical analysis of stress assignment in Choguita Rarámuri involves an unbounded, weight-insensitive iambic system. While typologically rare and claimed to be universally dispreferred (Hayes 1995), other similar cases have been documented in the literature (see, e.g., Graf & Ussishkin 2003). As discussed in Chapter 3, Proto-Uto-Aztecan has been reconstructed as having a weight-sensitive, stress-accent system (Munro 1977). Choguita Rarámuri stress might then have become lexicalized when vowel length distinctions from the proto-language were lost.

That vowel length in Choguita Rarámuri is relevant for some, but not all, phonological phenomena in the language resembles a case of what Fitzgerald (2012) terms *prosodic inconsistency* in Tohono O'odham, a related Uto-Aztecan language from the Tepiman branch. In contrast to Choguita Rarámuri, Tohono O'odham does have contrastive vowel length. Quantity, however, exhibits an asymmetry in terms of its role in the phonological system of the language, with a quantity-insensitive stress system and a quantity-sensitive prosodic morphological patterns.

11.3 Prosodic properties of morphologically complex verbs

### **11.3 Prosodic properties of morphologically complex verbs**

Stress systems are defined as those in which there is an increased prominence associated with one or more syllables in a word (Gordon & van der Hulst 2020). Choguita Rarámuri displays both phonological and phonetic properties that are characteristic of stress systems, including culminativity (each lexical word has at most one syllable which carries the highest degree of prominence) and obligatoriness (each lexical word has at least one syllable that carries the highest degree of prominence) (Hyman 1977; 1978; 2006; Beckman 1986; Hayes 1995), as well as unstressed vowel deletion and reduction (fewer vocalic contrasts are realized in unstressed syllables) (Caballero 2008; 2011b). Choguita Rarámuri is also a tonal language, fitting the definition of a "… language … in which an indication of pitch enters into the lexical realization of at least some morphemes", i.e., tone is the output of lexical phonology (Hyman 2006: 229). Stress and tone in Choguita Rarámuri are not only phonologically distinct systems, but they are also encoded through independent acoustic means (Caballero & Carroll 2015).

The surface form of morphologically complex verbs in this language results from the stress and tone properties of root morphemes and morphological constructions and complex interactions between phonologically general processes and morphologically governed phonological processes. Stress and tone assignment apply within a morphological domain, namely the Stem. This section synthesizes the different components necessary in deriving the surface prosodic forms of inflected verbs.

### **11.3.1 Stress patterns and metrical feet**

The Choguita Rarámuri stress-accent system can be analyzed in metrical terms (Caballero 2008; 2011b). The generalizations about the distributional properties of word-level stress in Choguita Rarámuri are summarized in (5) (see also Chapter 5):

	- a. Each prosodic word has a single main stress.
	- b. There is no secondary stress (i.e., stress is non-iterative).
	- c. Roots can be lexically stressed (and have fixed stress across paradigms) or lexically unstressed (with shifting stress in certain morphological contexts).

### 11 Prosody: domains and interactions


While there is no secondary stress assignment in the language, a metrical analysis can nevertheless be posited. Default stress assignment, second syllable stress within the root, can be modeled using a single iambic foot aligned to the left edge of the prosodic word, as exemplified in (6), with words containing the stressneutral past *-li* suffix:

	- a. (tʃ͡aˈpí)li tʃ͡aˈpí-li grab-pst 'S/he grabbed it.' b. (raˈʔì)tʃ͡ali raˈʔìtʃ͡a-li speak-pst 'S/he spoke.'

The metrical analysis of morphologically complex words containing unstressed roots and stress-shifting constructions (with third syllable stress), as well as noun incorporation constructions (stress in the first syllable of the construction's head, the third syllable), may be analyzed through a ternary metrical constituent, an iambic foot with a single left-adjoined syllable (Zoll 2004; see also Selkirk 1980; Dresher & Lahiri 1991; Rice 1993; Itô & Mester 1992, and Blevins & Harrison 1999). This ternary constituent is schematized in Figure 11.1, where F = Foot and Ad-Foot = Adjoined Foot).

The metrical analysis of morphologically complex words containing unstressed roots and stress-shifting suffixes (exemplified with the stress-shifting conditional *-sâ* suffix) is provided in (7). The metrical analysis of noun incorporation constructions is provided in (8).

11.3 Prosodic properties of morphologically complex verbs

Figure 11.1: Ternary metrical constituent with single left-adjoined syllable (Zoll 2004)

	- a. (<tʃ͡a>piˈsâ) /tʃ͡api-ˈsâ/ grab-cond 'if s/he were to grab it' 'si lo agarra'
	- b. (<ra>ʔiˈtʃ͡â)sa /raʔiˈtʃ͡â-sa/ speak-cond 'if s/he were to speak' 'si hablara' {JLG el1274:13:52.1}
	- a. (<si>waˈbô)ta /siwa-ˈbôta/ tripe-come.out 'Its tripes come out.' 'Se destripa.'
	- b. (<bu>siˈkâ)si /busi-ˈkâsi/ eye-break 'It becomes blind.' 'Se vuelve ciego.'

In sum, a metrical analysis of Choguita Rarámuri involves positing a single metrical constituent, a binary iambic foot or an extended ternary iambic foot, at the left edge of the prosodic word. Given that there is no role for contrastive vowel length in stress assignment, this is a weight-insensitive iambic system.

### 11 Prosody: domains and interactions

The distribution of feet is morphologically governed: binary feet in neutral morphological environments and ternary metrical feet in shifting morphological environments. Further discussion of the use of multiple types of metrical feet in different morphological environments can be found in Bennett (2012) and Bennett (2013). Further discussion of the use of ternary feet include Martínez-Paricio (2013) and Martínez-Paricio & Kager (2015).

### **11.3.2 Lexical tone patterns**

As discussed in Chapter 6, Choguita Rarámuri has three lexical tones: falling (/HL/, <ô>), low (/L/, <ò>), and high (/H/, <ó>). Lexical tones are exclusively realized on surface stressed syllables, i.e., there is only one lexical tone per prosodic word and stressless syllables lack lexical tone. The examples in (9) exemplify the three-way tonal contrast (Caballero & Carroll 2015: 465) in both nouns and verbs.


Tone is not dependent on voicing of preceding consonants nor any other phonological features, and there is no evidence of tone spreading. The Tone-Bearing-Unit (TBU) is the mora: falling tones have their high target on the stressed syllable, with the fall continuing through a post-tonic syllable, if there is one (H\*L) (Caballero & Carroll 2015). As described above, stressless syllables lack lexical tones: the surface tonal properties of these syllables vary depending on intonation (Garellek et al. 2015).

### 11.3 Prosodic properties of morphologically complex verbs

### **11.3.3 Canonical prosodic shapes of roots and suffixes**

Most roots in Choguita Rarámuri are disyllabic or trisyllabic: from a corpus of 1,004 nominal and verbal roots, 47% percent are disyllabic and 40% trisyllabic. There are tetrasyllabic roots (10%), but most of these are, with different degrees of transparency, internally complex, so a maximally tetrasyllabic prosodic size cannot be established. Finally, there are also monosyllabic roots. While a mere 3% of the corpus, the existence of these roots make it impossible to establish a disyllabic minimality restriction for open class lexical morphemes.<sup>2</sup>

As for suffixes, most are monosyllabic, but there are a few disyllabic suffixes, including the Desiderative *-ˈnále* suffix, the Associated Motion *-simi* suffix, the Habitual Passive *-ˈrîwa* suffix, the Future Singular *-ˈmêa* suffix,<sup>3</sup> the Auditory Evidential *-tʃ͡ane* suffix, and the Indirect Causative *-nula* suffix. Almost all of these disyllabic suffixes are related to synchronically active free lexical morphemes with varying degrees of transparency (Desiderative *-nále* and the verb *naki*, 'want'; Associated Motion *-simi* and the verb *simi*, 'go singular'; Auditory Evidential *-tʃ͡ane* and the verb *tʃ͡ane*, 'make noise, say'). Suffixes, like roots, are thus not restricted to a canonical prosodic shape. <sup>4</sup>

While no generalizations in terms of maximal size applies to roots and no generalization about minimal or maximal size applies for suffixes, there are certain formal characteristics that usually apply to roots vs. suffixes. Given the stress properties of roots (with mainly second or third syllable stress) and the post-tonic vocalic reduction and deletion processes operating in the language (described above in Chapter 5) the boundaries between roots and suffixes are often the target of posttonic syncope. This, in turn, yields derived stems with shared formal properties, such as final consonants. Roots derived with the productive causative suffix *-ti* are a good example of this, since posttonic syncope often targets the vowel of the causative suffix, generating a class of causative stems ending in a lateral flap. These causative stems are then recursively suffixed with the allomorph *-ti* of the causative suffix. This is shown in the examples provided in (10):

<sup>2</sup>When considering the proportions of roots separately for nouns and verbs, there are some differences: most monosyllables are verbs (5% of verbs are monosyllabic and only 2% of nouns are monosyllabic), and most tetrasyllables are nouns (19% of nouns are tetrasyllabic, while only 5% of verbs are tetrasyllabic).

<sup>3</sup>Miller (1996) has proposed that the disyllabic future singular suffix *-mêa*, also present in Guarijío, derives from the reconstructed verb *\*mi(l)a* 'run, go' (Miller 1996: 133) from "Proto-Sonoran". Miller defines "Sonoran" as a sub-branch of Uto-Aztecan located in Mexico's northwest which would include Tepiman and Taracahitic languages.

<sup>4</sup>Althhough, as discussed in §11.5.3 below, a set of disyllabic suffixes undergo truncation in certain contexts.

### 11 Prosody: domains and interactions

	- a. *ˈnè uˈbârtipo ˈnè* 1sg.nom *uˈbâ-ri-ti-po* bathe-**caus**-caus-fut.pl 'I'll be forced to bathe.' 'Van a hacer que me bañe.' < LEL 05 ECME(40)/el > b. *ˈnè ˈmí ˈmêrtima oˈlá ˈnè ˈmí ˈmê-ri-ti-ma oˈlá*
	- 1sg.nom 2sg.acc win-**caus**-caus-fut.sg cer 'I will make you win.' 'Te voy a hacer ganar.' < LEL 05 ECME(3)/el > c. *koˈʔártinale koˈʔá-ri-ti-nale* eat-**caus**-caus-desid 'She wants to make him eat.'

```
'Quiere hacerlo comer.' < BFL 06 ECDW(55)/el >
```
Syncope, thus, participates in the creation of a new paradigm of causative stems with particular word-internal (stem-delimiting) codas.

### **11.3.4 Prosodic properties of roots and morphological constructions**

As discussed in Chapter 5, root morphemes belong to two classes depending on their stress behavior: (lexically) stressed roots retain stress in a fixed syllable in different morphological environments, while in (lexically) unstressed roots the location of stress is morphologically governed. The analysis proposed in this grammar is that this asymmetric behavior in terms of stress results from an underlying phonological difference between the two root classes, with stressed roots being lexically prespecified with a diacritic that is phonetically realized as stress in surface forms, and unstressed roots lacking this diacritic and receiving stress by default (Caballero 2011c). All prosodic words in Choguita Rarámuri, whether they contain a lexically stressed root or an unstressed lexical root, have surface stress, a syntagmatic prominence cued acoustically via intensity and duration (Caballero & Carroll 2015).

Morphological constructions (affixes and non-concatenative morphological processes) also exhibit an asymmetric phonological behavior. They are classified as belonging to one of two classes in terms of their stress properties: they may be stress-shifting, triggering stress shifts when attaching to unstressed roots, or

### 11.3 Prosodic properties of morphologically complex verbs

stress-neutral, not conditioning any stress changes in the bases with which they combine.

Table 11.1 illustrates the contrast between stressed and unstressed roots in neutral and shifting morphological environments and the tonal patterns associated to stressed syllables in different morphological environments (Caballero 2008; 2011b). Shading highlights cells with morphologically complex words undergoing stress shifts.

Table 11.1: Morphologically-conditioned stress shifts and tonal alternations in verbs (Caballero 2008; Caballero & Carroll 2015)


This table exemplifies how stressed roots retain stress in a fixed location across morphological contexts (in the first, second or third syllable), while unstressed roots exhibit a rightward stress shift in shifting environments (in this case, exemplified with the conditional *-sâ* suffix).

Unstressed roots receive stress by default on the second syllable of the root in neutral environments (i.e., monosyllabic roots bear stress on the root). The default stress assignment algorithm is summarized in (11).

(11) **Default stress assignment in Choguita Rarámuri**: words containing unstressed disyllabic or trisyllabic roots and neutral morphological constructions have second syllable stress (Caballero 2011b).

The two classes of morphological constructions are not only distinguished by their stress effects, but also in terms of other morpho-phonological effects they may trigger: shifting constructions may induce vocalic alternations or other morpho-phonological changes, a phenomenon documented across the Uto-Aztecan language family (Heath 1977; 1978). In Choguita Rarámuri there are vocalic

### 11 Prosody: domains and interactions

alternations that unstressed roots may undergo, which motivates a characterization of Choguita Rarámuri verbs into three classes, if both stress and vocalic alternations are considered (for full discussion, see Chapter 9). Lexically stressed roots constitute the first class, Class 1, and undergo no morpho-phonological changes. Lexically unstressed roots can be divided into two sub-classes: (i) unstressed roots with fully specified vowels (Class 2 verbs) and (ii) unstressed roots with root final unspecified vowels (Class 3 verbs). Class 3 roots have a final V slot, whose features are dependent on the morphological construction that combines with the root, with final root vowel raising in shifting constructions. The three verbal root classes are summarized in Table 11.2.


As shown in Table 11.2, Class 2 roots have a final stressed low vowel in neutral constructions. Most Class 3 roots, on the other hand, end in high, front vowels in shifting constructions (such as the future singular *-ˈmêa*), although some verbs of this class can also end in back, mid vowels (e.g., *noko-ˈmêa* in (12d). Without exception, these roots end in *a* when attaching a stress neutral suffix (such as the past *-li*). Some examples of these roots with alternating final vowels are provided in (12):

(12) Class 3 roots: vowel alternations


### 11.3 Prosodic properties of morphologically complex verbs

Finally, the analysis of stress and tone presented here assumes that all lexical tones are underlyingly specified (see also Caballero & German 2021). Unstressed roots may bear either H or L tone when bare or in neutral contexts, which can be attributed to a lexical tonal specification. Stressed roots, on the other hand, may bear HL, H or L tones. There is thus an asymmetry between stressed and unstressed roots in terms of lexical tone, as there are no unstressed HL-toned roots in the language.

When stressed, shifting suffixes may bear any of the three lexical tones. Given that tonal contrasts are only realized in stressed syllables, there is no evidence that stress neutral suffixes have underlying lexical tone, since they are never stressed. The tonal properties of stress shifting suffixes are shown in a fragment of the paradigm for the root *suˈkú* 'to scratch' in (13).

(13) Neutral and shifting suffixing constructions


Words containing an unstressed root like *suˈkú* and a stress neutral suffix (e.g., the past *-li* suffix (13a)), bear stress on the second syllable, given the default stress assignment rule given in (11). Words containing an unstressed root and a stress shifting suffix (e.g., the future singular *-mêa* suffix (13e)), bear stress on the suffix, the third syllable of the prosodic word. More details about the tonal properties of roots and morphological constructions (affixes and non-concatenative morphological exponents) is provided in §11.3.

### **11.3.5 Stress and lexical tone**

As reported in Chapter 5 and Chapter 9, all prosodic words have a surface stressed syllable where lexical tonal contrasts are realized. This generalization may be stated as follows:

(14) Lexical tone is a property of morphemes associated to the surface stressed syllable.

### 11 Prosody: domains and interactions

All three lexical tones (H, L, and HL) may be associated with the stressed position, whether the first, second or third syllable. As stated in (11) above, unstressed roots receive stress by default on the second syllable of the root in stress neutral environments (i.e., monosyllabic roots bear stress on the root).

Morphologically-conditioned stress shifts result in tone neutralization patterns. As shown in (13) above, if a stress-shifting suffix is stressed after a stress shift, the stressed suffix syllable will bear the lexical tone of that suffix (this is the case when a monosyllabic or disyllabic unstressed root attaches a stress-shifting suffix). This pattern is further exemplified in Table 11.3.


Table 11.3: Suffix lexical tone after stress shift

As shown in (13) and Table 11.3, shifting suffixes bear their underlying tone when stressed (e.g., HL in the imperative singular *-kâ* in forms (a–b) in Table 11.3, L in the imperative plural *-sì* in forms (c–d) in Table 11.3. The lexical tone of the root, which surfaces in neutral morphological contexts, is deleted after the stress shift. On the other hand, if a trisyllabic unstressed root attaches a stress-shifting suffix, the newly stressed syllable will be a stem syllable. As shown in Table 11.4, a newly stressed stem syllable will bear a HL tone in these contexts, regardless of what the lexical tone of the root is (H tone in *roʔˈsówa* 'cough' (examples (a–b) in Table 11.4), L tone in *naʔˈsòwa* 'stir' (examples (c–d) in Table 11.4).

One important aspect of this second pattern is that the surface tonal pattern of these morphologically complex words that have undergone a stress shift is

### 11.3 Prosodic properties of morphologically complex verbs


Table 11.4: Stem tone after stress shift

not predictable based on the lexical tonal properties of root morphemes nor the lexical tones of suffixes. An example discussed in Caballero & German (2021) is that of unstressed roots attaching the imperative plural *-sì* suffix. As shown in examples (c–d) in Table 11.3 above, this suffix bears its lexical L tone when stressed, but it will not surface if the stressed syllable after a stress shift is a stem syllable, i.e., the hypothetical forms *\*rosoˈwà-si* (for example (b) in Table 11.4 and *\*naʔsoˈwà-si* (for example (d) in Table 11.4) with L tone when attaching imperative plural *-sì*, are unattested.

This tonal pattern is analyzed in Caballero & German (2021) as resulting from a process of default HL tone insertion after a stress shift has deleted the root's lexical tone. This analysis is based on the following assumptions:

	- b. In words containing lexically unstressed roots and neutral morphological constructions, the stressed syllable bears the underlying lexical tone of these roots
	- c. Stress shifts in shifting environments cause lexical root tones to delete
	- d. If the newly stressed syllable after a stress shift is a stem syllable, it is toneless and acquires a default HL tonal melody

It could be argued that this default HL tone insertion process is also at play in Spanish loanwords (Caballero & Carroll 2013) (see also §7.3). Loanwords from

### 11 Prosody: domains and interactions

Spanish are incorporated into Choguita Rarámuri with faithful prominence to the stress location of the source words. The stressed syllable in loanwords has a HL tone (no exceptions have yet been documented to this pattern). Relevant examples are given in Table 11.5 (the Spanish sourcewords are provided in their orthographic form, where boldface indicates the stressed syllable).

**Spanish Loanword Tone Gloss sourceword** a. To**más** [toˈmâʃi] HL 'Thomas' b. man**za**na [maˈsâna] HL 'apple' c. **sá**bado [ˈsâbuto] HL 'Saturday' d. **Juan** [ˈhuâni] HL 'Juan' e. Da**niel** [raˈniêri] HL 'Daniel' f. li**me**ta [liˈmêta-ʧi] HL 'bottle' g. pa**sear** [basaˈlôwa] HL 'to take a stroll'

Table 11.5: Tone patterns in Spanish loanwords

The analysis involving a default HL tone insertion process for loanwords would involve the assumption that Spanish loanwords are lexically stressed but toneless in Choguita Rarámuri. Alternatively, the tonal properties of Choguita Rarámuri loanwords may be analyzed as involving a reinterpretation of the acoustic properties of Mexican Spanish prominence, which has been argued to include a H\* pitch accent in the stressed syllable in certain intonational contexts (focusmarked words in declarative sentences) (Prieto et al. 1995).

In sum, morphological factors condition stress shifts. Given the dependency tone has on stress-accent for its distribution, the tonal alternations resulting in these contexts are largely predictable based on the lexical tonal properties of the morphemes that make up a morphologically complex word. Further discussion on the mechanism of default tonal assignment is provided in §11.3.7.3.

### **11.3.6 Stress and tone properties of compounds**

Body part incorporation constructions involve a single Prosodic Word, with a construction specific rule assigning a single stress to the first syllable of the head of the compound. As addressed in Chapter 5 and Chapter 9, in these N-V constructions the noun root is fully integrated with the verb morphologically, and both the noun root and the verb root are attested as independent roots in the language. As described in §9.3.3, stress in these constructions is actively constrained

### 11.3 Prosodic properties of morphologically complex verbs

by the grammar: if the head, the verb, has second syllable stress in isolation and if the first member, the incorporated body-part noun, is two syllables long, stress retracts to the verb's first syllable, the construction's third syllable. The stressed syllable in body-part incorporation forms bears a HL tone. Relevant examples are given in (16).

	- a. *busiˈkâsi busi+ˈkâsi* eye+break 'to become blind'
	- b. *ropaˈkâsi ropa+ˈkâsi* stomach+break 'to have a miscarriage'
	- c. *busiˈbôta busi+ˈbôta* eye+come.out 'for eyes to come out'
	- d. *kutaˈbîri kuta+ˈbîri* neck+twist 'to neck-twist'
	- e. *tʃ͡omaˈbîwa tʃ͡oma+ˈbîwa* mucus+clean 'to mucus-clean'

As discussed in Chapter 9, the stress properties of incorporated forms can be analyzed as resulting from a mophological stress rule specific to these constructions that requires stress to be assigned in the first syllable of the head of the incorporated construction:

(17) Incorporated verb stress rule

The head of the incorporation construction, the verbal root, must bear stress in the first syllable

This morphological stress rule is a process that is analyzed here as involving both stress deletion and stress re-assignment. Thus, HL tone assignment in verbal

### 11 Prosody: domains and interactions

compounds can be analyzed as resulting from the process of default HL tone insertion that is attested in other morphological contexts where a stress shift has deleted a lexical tone (e.g., Table 11.4).

### **11.3.7 Grammatical tone**

Tone patterns that play a morphological role in any linguistic system have been defined in a variety of ways in the literature (for an overview, see Hyman 2016). This grammar adopts the term *grammatical tone* and assumes the following definition (Rolle 2018; see also Caballero & German 2021):

(18) **Grammatical tone**: A tonal pattern or process that is not general across the phonological grammar of a language, but is instead associated to a specific morpheme or construction, or a natural class of morphemes or constructions.

While several Uto-Aztecan languages have developed lexical tone, there is little information in the literature as to whether these languages feature grammatical tone patterns. Exceptions include Milpa Alta Nahuatl, where tone is reported to distinguish present from imperfect verb forms among other contrasts (Whorf et al. 1993) and Wixárika (Huichol; Corachol)), where tone encodes tense distinctions (present vs. past) (Banerji 2014).

Both derivational and inflectional morphological constructions may exhibit an underlying lexical tone and/or trigger a tonal alternation in the stems to which they attach. This section provides an overview of grammatical tone patterns as documented in inflectional constructions and as attested in morphologically complex words composed of a stem plus one inflectional morphological construction, as described in Caballero & German (2021). In terms of the presence/absence of affixation concomitant to grammatical tone and the type of tonal alternation, there are at least three patterns of grammatical tone: (i) tone as an exponent of morphological information; (ii) tone as a morphologically-conditioned tonal effect and (iii) tone that is determined by the type of morphological construction (stress-shifting or stress-neutral) in a class of verbs. Each of these patterns is described next. Grammatical tone patterns reveal that stress and tone, though intricately related in their distribution, are separate dimensions in the classification of verbal root morphemes into different classes.

### **11.3.7.1 Tone as a morphological exponent**

A first pattern of grammatical tone attested in Choguita Rarámuri involves tone as an exponent of inflection (also referred to as a "tonal morpheme" in Welmers

### 11.3 Prosodic properties of morphologically complex verbs

1959 and Hyman 2016), a type of non-concatenative morphology that may also be referred to as *realizational morphology*, where a morphological category is encoded by a phonological process other than concatenation of segmental morphemes (see Inkelas 2014 for discussion). Tonal exponence in Choguita Rarámuri is attested in the imperative singular construction, which features four lexically conditioned suppletive allomorphs, two of which are concatenative ((i) the *-ka*̂ suffix and (ii) the *-sa*̂ suffix), and two of which are non-concatenative ((iii) a stress shift and (iv) a L tone realized in the stressed syllable of HL-toned stems). The L imperative singular morphological construction is exemplified in Table 11.6.


Table 11.6: Imperative singular allomorphs

The L tone allomorph of the imperative singular overwrites the lexical HL tone of the root (examples (a–b) in Table 11.6): HL *ni'kâ* 'it barks' vs. L *ni'kà* 'bark!'. The L tone allomorph does not combine with a verb stem with a suffixal allomorph of the imperative singular (examples (c–d) in Table 11.6, i.e., there is no multiple exponence of the imperative singular. The second non-concatenative allomorph of the imperative singular is exemplified in example (e) in Table 11.6, where a rightward stress shift within the stem encodes the inflectional value: the verb has second syllable stress (and lexical L tone) when inflected for present (*naʔ'sòwa* 's/he stirs it'), but third syllable stress (and HL tone assigned by default) when inflected for imperative singular (*naʔso'wâ* 'stir it!').

As stated above and exemplified in Table 11.6, only HL stem tones are replaced by the imperative singular L tone. In the case of L-toned stems, the L imperative singular allomorph vacuously applies (examples (a–c) in Table 11.7), but it is blocked from applying to H-toned stems (examples (d–f) in Table 11.7).

### 11 Prosody: domains and interactions


Table 11.7: Tone in imperative singular forms

As exemplified here, tonal overwritting of HL tones by the grammatical L tone neutralizes the contrast between lexical HL and L tones in the imperative singular.

A subset of inflected verbs displays inter-speaker variation in terms of grammatical tone: unstressed roots (those that exhibit stress shifts when attaching a stress-shifting suffix or non-concatenative process) may surface with one of two tonal patterns: (i) a grammatical L tone that replaces stem HL tone or (ii) a HL stem tone associated with the shifting morphological class. This variation in surface forms is illustrated in Table 11.8.

Table 11.8: Tonal variation in imperative singular forms


### 11.3 Prosodic properties of morphologically complex verbs

As shown in these examples, all speakers produced a form with a rightward stress shift in the imperative singular, but they show different tonal patterns in the derived imperative singular: one speaker (RIC) produces a L tone in the stressed syllable (example (a) in Table 11.8), the expected L grammatical tone of the imperative singular; in contrast, two other speakers (BFL and LEL) produce a HL tone in the stressed syllable (examples (b–c) in Table 11.8), the expected pattern if lexical tone has been neutralized and a default HL tone assigned. This latter pattern results from a process of stress-based tonal neutralization, where a stress shift deletes a lexical tone and a default HL tone is assigned to the stressed stem syllable (§11.3.5). This kind of inter-speaker variation suggests that speakers may have more than one generalization available regarding the tonal properties of these inflected forms. This kind of variation is not yet documented in other morphological contexts.

### **11.3.7.2 Morphologically-conditioned tone**

In addition to tonal exponence, Choguita Rarámuri possesses grammatical tone patterns that are associated with specific suffixes when attaching to a prosodicallydefined class of verbs. Specifically, unstressed, H-toned roots surface with a L tone when attaching the imperfective *-i* suffix and the progressive *-a* suffix. The imperfective and progressive suffixes are stress-neutral, which means the tonal changes they condition with unstressed, H-toned roots take place without any concomitant stress shifts nor any other (morpho-)phonological effects. The examples in Table 11.9 show the contrast between the imperfective and progressive constructions, on the one hand, and the past tense *-li* suffix, a stress-neutral suffix that does not impose a L tone on the stems to which it attaches. In the latter case, the lexical tone of the root (H or L) is associated with the stressed syllable.

As shown in examples (d–e) in Table 11.9, L-toned unstressed roots do not exhibit any tonal changes when inflected for imperfective or progressive, showing the process vacuously applies with these verbs.

The contrast between unstressed roots (examples (a–c) in Table 11.10) and stressed roots (examples (d–f) in Table 11.10) demonstrates that only unstressed roots exhibit a L tone when attaching these suffixes.

As discussed previously and seen in these examples, the rightward stress shift in shifting environments is only attested with unstressed verbs (Table 11.9a–c). Stressed roots, on the other hand, do not undergo stress shifts (Table 11.9d–f). Stressed roots do not replace their lexical tones when attaching the imperfective and progressive suffixes.


Table 11.9: Morphologically-conditioned L tone

Table 11.10: Morphologically-conditioned L tone: stressed vs. unstressed verbs


### 11.3 Prosodic properties of morphologically complex verbs

These morpheme-specific phonological effects originated through recent diachronic changes. Comparative evidence from closely related Norogachi Rarámuri (Brambila 1953) and Mountain Guarijío (Miller 1996) shows that the progressive *-a* suffix and the imperfective *-i* suffix have recently developed from suffixes that were shifting.<sup>5</sup> The shifting allomorphs of these suffixes are attested in the speech of Choguita Rarámuri native speakers with command of the closely related Norogachi Rarámuri (NR). For these speakers, the Choguita Rarámuri forms co-exist with the NR forms that exhibit the conservative form of these suffixes, which are stressed, include a palatal glide onset and have a H tone. These forms are exemplified in Table 11.11.

Table 11.11: Norogachi Rarámuri cognate suffixes: imperfective and progressive


The morphologically-conditioned tonal effects exhibit by the progressive and imperfective suffixes are thus the result of a recent innovation of morphologicallyconditioned tone.

### **11.3.7.3 Grammatical tone distributed by morphological class**

A third type of grammatical tone pattern in Choguita Rarámuri is attested in a class of lexically stressed verbs referred to as "alternating" in Caballero & Carroll (2015). In these verbs, surface tonal patterns are conditioned by the type of inflection the verb exhibits: a HL tone in shifting morphological environments and a L tone in neutral morphological environments. Stress-shifting constructions do not trigger stress shifts when combining with lexically stressed roots, but cause a rightward stress shift with lexically unstressed roots (§5.3). In the alternating verb class, stress-shifting suffixes condition specific tone patterns on the stem without triggering any stress shifts. These tonal alternations attested in alternating verbs are exemplified in Table 11.12.

<sup>5</sup> In Mountain Guarijío, Miller describes a progressive *-a* suffix that may bear stress (e.g., *yuʔkuá=ga* 'it is windy' (1996: 140)), the cognate form of the Choguita Rarámuri progressive *-a* suffix.

### 11 Prosody: domains and interactions

Table 11.12: Alternating verbs: tonal alternations


The examples below show the tone alternations of a single verb *biʔˈw-a* 'to clean' (a stem composed of the root *biʔw-* 'to clean' and the transitivizer*-a* suffix), where the surface tonal properties of the word are determined by the type of inflection: L if the construction is stress-neutral (19) and HL if the construction is stress-shifting (20).

### (19) Stress-neutral constructions



Alternating verbs contrast with other lexically stressed verbs: as shown below, these verbs (like the verb *raʔˈnè* 'to shoot' exemplified below) have fixed stress

### 11.3 Prosodic properties of morphologically complex verbs

across paradigms and there are no tonal alternations conditioned by the type of morphological construction: there is fixed stress and constant tone values, regardless of whether the verb combines with stress-neutral (21) or stress-shifting (22) markers.

(21) Stress neutral constructions


(22) Stress-shifting constructions


For the alternating verb class there is thus no evidence of underlying lexical tone nor are tonal alternations predictable based on the lexical tonal properties of suffixes. These tonal alternations are also independent of stress alternations and other phonological properties of roots and suffixes and non-concatenative processes. Stress and tone, though closely related, are thus orthogonal dimensions in Choguita Rarámuri.

Finally, the tonal patterns of alternating verbs just discussed are relevant to an alternative analysis where stress-shifting and stress-neutral constructions constitute a morphosyntacatically motivated class distinction: as discussed in §9.1.2, it is assumed in the Uto-Aztecanist literature that neutral constructions are posited to realize "non-future" morphosyntactic categories, including realis inflectional categories such as past, perfective and imperfective, while shifting constructions are posited to realize "future" or "unrealized" categories, which map onto irrealis categories, such as future, imperative and potential (Langacker 1977: 133). The consequence of such an analysis for grammatical tone patterns would involve positing two tonal morphemes, a L tone for 'non-future'(realis) categories and a HL tone for "future" (irrealis) categories.

The arguments against a morphosyntactic account of the shifting vs. neutral distinction are provided in §9.1.2 above. As noted, a realis/irrealis distinction is not amenable for all shifting and neutral constructions. One relevant example

### 11 Prosody: domains and interactions

involves the past passive construction (exemplified above in (20d), *biʔ'wâar*u 'it was cleaned'), which cumulatively encodes passive voice and past tense. The past passive may be categorized as a "realis" (or "non-future") construction in a morphosyntactically-based account, but it nevertheless patterns morphophonologically with the stress-shifting class, contrary to expectation if the distinction were morphosyntactically motivated.

The shifting/neutral distinction in Choguita Rarámuri tonal alternations conforms to the predictions made on a *morphomic* account: the inflected forms of lexemes as shifting or neutral are not determined by morphosyntactic inflectional features, nor any semantic or phonological principles, but rather by purely (or autonomously) morphological properties. The stress-shifting vs. stress-neutral distinction thus involves morphosyntactically heterogeneous classes of morphological constructions.

### **11.3.8 Stress and tonal properties of inflected verbs**

The previous subsections (§11.3.1–§11.3.7) have addressed each of the factors that shape the surface prosodic properties of inflected verbs in Choguita Rarámuri, including the phonological size of roots, the lexical stress and tone properties of roots and morphological constructions, as well as the different kinds of interactions between lexical and grammatical tone which may lead to lexical tone replacement and grammatical tone assignment. In terms of lexical-grammatical tone interaction, Choguita Rarámuri grammatical tone patterns can be classified as belonging to one of the following classes:

	- a. **Tone as a morphological exponent**: the imperative singular L allomorph is a tone-replacing construction that involves no affixation and replaces stem HL tone with a L tone.
	- b. **Morphologically-conditioned tone**: the imperfective *-i* and progressive *-a* suffixes replace H tone of stems containing an unstressed root with a L tone.
	- c. **Tone determined by morphological class ('alternating verbs')**: tone is determined by the type of inflection construction the root combines with (HL if shifting, L if neutral) in morphosyntactically heterogeneous classes; there is no evidence of lexical tone.

Patterns (23a) and (23b) involve tonal overwriting (i.e., a grammatical tone pattern replaces lexical tone), while the pattern in (23c) involves the assignment of a grammatical tone to a class of underlyingly toneless verbs.

### 11.3 Prosodic properties of morphologically complex verbs

The surface tonal patterns of two-level constructions (roots plus one layer of inflection) is summarized in Table 11.13, based on a classification in terms of: (i) the stress properties of roots (unstressed or unstressed); (ii) the tonal properties of constructions (HL, L or H); and the ability of morphological constructions to trigger morpho-phonological changes or not onto stems (shifting or neutral, respectively). Shading represents cells in the paradigm that exhibit a grammatical tone pattern.

An analysis of this system based on tonal underspecification is proposed in Caballero & German (2021). This analysis is based on the following assumptions:

	- a. All tones (HL, L and H) are lexically specified
	- b. Stress and tone are orthogonal dimensions in the phonological grammar of Choguita Rarámuri
	- c. Verb roots are either specified or unspecified lexically for tone

In this analysis, morphologically complex words which contain tonally specified verbs may exhibit interactions between lexical and grammatical tones, given that specific constructions may be associated with grammatical tone melodies that trigger the replacement of base stem tone. In contrast, in morphologically complex words containing tonally unspecified verbs (alternating roots and unstressed roots where lexical tone has been deleted after a stress shift), tonal patterns realize inflectional morphological information.<sup>6</sup>

Based on their lexical specification for stress and tone, Choguita Rarámuri verbs may thus be characterized as follows:

<sup>6</sup>Caballero & German (2021) refer to the latter kind as *paradigmatic tone*. Rolle (2018) defines paradigmatic grammatical tone as follows: "[i]n a grammatical paradigm consisting of grammatical categories, tonal values to the root/stem which (i) show extensive inconsistency within grammatical categories (no paradigmatic consistency across rows or columns), and (ii) show extensive inconsistency across roots/stems in parallel paradigms (no 'transparadigmatic' consistency across rows or columns), and (iii) there being little positive evidence for determining the underlying tone of the root/stem" (2018: 109). While surface tonal patterns of Choguita Rarámuri alternating verbs exhibit consistent association within grammatical categories (shifting vs. neutral), there is little positive evidence of the underlying tone of the stems. The term *paradigmatic tone* here refers to the fact that, in these cases, cells in the paradigms are associated with tonal values that are not predictable from lexical tone properties of individual morphemes.

### 11 Prosody: domains and interactions

Table 11.13: Surface tonal melodies in inflected verbs (root + inflectional construction) (Caballero & German 2021)


### 11.4 The interaction between lexical tone and intonation

	- a. Lexically stressed with lexically specified tone: stress is fixed across paradigms; underlying lexical tone is realized on the stressed stem syllable; grammatical tone may replace lexical tone (in the imperative singular).
	- b. Lexically unstressed with lexically specified tone: stress shifts in shifting environments; the lexical tone of the stem is associated to the stressed stem syllable in neutral contexts; the lexical tone of the suffix is associated to the stressed suffix syllable in shifting contexts; grammatical tone may replace lexical tone (in the imperfective and progressive).
	- c. Lexically stressed with no lexical tone (alternating roots): stress is fixed across paradigms; surface tone is dependent on the type of inflection, L in neutral environments and HL in shifting environments.
	- d. Lexically unstressed with no lexical tone (tonal neutralization): lexically unstressed verbs have a lexically specified tone that is deleted after a stress shift. A HL tone is associated to the newly stressed, toneless stem syllable.

### **11.4 The interaction between lexical tone and intonation**

This section provides details of how lexical tone and intonation interact in the tonal grammar of Choguita Rarámuri. As discussed in Chapter 6, lexical tonal contrasts are implemented through a variety of acoustic means, some of which are speaker-dependent and some of which are dependent on the intonational context. As discussed in this section, there is also robust evidence that both lexical and grammatical tones are preserved in tone-intonation interactions (Caballero et al. 2014; Aguilar et al. 2015; Garellek et al. 2015).

### **11.4.1 Tone-intonation interactions in declaratives**

As discussed in §6.2, Choguita Rarámuri does not only deploy f0 to encode lexical tonal contrasts, but also to encode intonation: declarative sentences exhibit a High boundary tone (H%) (Caballero et al. 2014; Garellek et al. 2015). This is shown in Figure 11.2 (Figure 6.2 in Chapter 6) with a sentence composed of words with lexical L tones, *reˈhòi suˈnù oˈhòli* 'The man dekerneled corn' (lexical pitch

### 11 Prosody: domains and interactions

targets are represented with "\*" in the first tier; stressed syllables are represented with "S" in the bottom tier).

Figure 11.2: High boundary tone in declaratives (Garellek et al. 2015)

As shown here, in a sentence with lexical L tones, H% boundary tones accommodate lexical tones: both the lexical L pitch target and the post-lexical H pitch target are clearly differentiated.

On the other hand, the rise expected with the presence of a H% is replaced by a pitch fall if the Intonational Phrase (IP) contains a lexical HL tone at the right edge. This is shown in Figure 11.3 (Figure 6.3 in Chapter 6) with a sentence composed of words with lexical HL tones, *Maˈnuêli oˈkwâ koˈlî iʔˈkîli* 'Manuel bit two chili peppers'.

Thus, H% boundary tones are overridden by lexical falling tones, an effect which may enhance lexical/morpho-lexical tones in phrase final position: lexical tones are clearly differentiated for all speakers across different intonational contexts (see further discussion in Garellek et al. 2015).

### **11.4.2 Tone-intonation interactions in interrogatives**

The morphosyntactic and prosodic properties of different types of interrogative constructions is addressed in §14.2. This section addresses the patterns that emerge in contexts where a lexical tone and an intonational tone conflict in terms of their association in different types of interrogative constructions. Lexical tonal contrasts are prioritized in these constructions.

### 11.4 The interaction between lexical tone and intonation

Figure 11.3: No high boundary tone in declaratives with lexical HL tones (Garellek et al. 2015)

Different types of interrogative constructions share the following intonational characteristics:

(26) Intonational properties of Choguita Rarámuri interrogative constructions


Consider first the behavior of lexical tone in phrase-final position in unmarked polar interrogative constructions. In these cases, the boundary H% tone aligns with a lexical HL tone in the stressed syllable. Figure 11.4 and Figure 11.5 (repeated from §14.2.1.1) show a minimal intonational pair between a declarative utterance and its morphosynactically equivalent, polar interrogative counterpart, respectively. A lexical HL tone is associated with the penultimate syllable of the utterance. As shown in the contrast between the two Figures, f0 is significantly raised in the final stressed syllable of the polar question in Figure 11.5 due to the association of an interrogative H% intoneme aligning with the pitch peak of the lexical HL tone in the stressed syllable.

In contrast, when an interrogative utterance has a lexical L tone in the final stressed syllable, the lexical tone takes precedence and is preserved in the stressed syllable, and the H% boundary tone docks on the following, post-tonic syllable. Figure 11.6 illustrates this pattern.

### 11 Prosody: domains and interactions

Figure 11.4: Declarative utterance: *ma ˈtôlo* 'S/he buried him/her' (< BFL el1170])

Figure 11.5: Morphosyntactically unmarked polar interrogative: *ma ˈtôli?* 'Did s/he bury him/her?' < BFL el1307 >

Figure 11.6: Accommodation of L tone and H% boundary tone in *ma ˈnèli?* 'Did s/he bury him/her?' < BFL el1307 >)

Finally, in interrogative constructions where the last stressed syllable of the utterance bears a lexical H tone, an interrogative H% tone associates in the final, unstressed syllable, with a pitch target that is higher than the one associated with the lexical H tone in the stressed syllable. This is illustrated in Figure 11.7, with an example of a content question (see also §14.2.2).

Figure 11.7: Content question with utterance-final lexical H tone in *ˈhêpi ˈkwâ muˈrú-li* 'Who did s/he carry?'

In sum, as has been documented in declarative utterances, the three-way lexical tonal contrast of Choguita Rarámuri is preserved in interrogative constructions.

### 11 Prosody: domains and interactions

### **11.4.3 Summary**

The cross-linguistic study of intonational encoding of tonal languages document several possibilities in terms of the phonological relationship between lexical tonal systems and intonational phonology. These include:

	- Manipulation of register (Hausa (Chadic; Nigeria) (Inkelas et al. 1987))
	- Prosodic rephrasing (Chitumbuka (Bantu; Malawi) (Downing 2007))
	- Different tonal accommodation strategies (e.g., intonational tones override lexical tones in Coreguaje (Tukanoan; Colombia) (Gralow 1985))
	- No use of f0 (Navajo (Athabaskan) (McDonough 2001)).
	- Non-tonal encoding of intonation, including lengthening (Shekgalagari (Bantu; Botswana) (Hyman & Monaka 2011)).

Preservation of lexical and morphological tones in tone-intonation interactions as documented in Choguita Rarámuri have been reported for other languages, which include Serbo-Croatian (Godjevac 2006), Stockholm Swedish (Riad 2006) and Curacao Papiamentu (Remijsen & Van Heuven 2005).

In addition to boundary tones interacting with lexical tones, there are intonational pitch targets, "lead" tones, that are dependent on the lexical tones of Choguita Rarámuri (as discussed in §6.2.2, there are optional post-lexical tones that are opposite to the lexical tones they precede, namely a L target before H and HL lexical tones, and a H tone before a lexical L tone). Lead tones are thus a kind of tone-dependent intonational phenomenon, that may also contribute to preserving lexical tone contrasts and/or their enhancement in different intonational contexts (Garellek et al. 2015).

In addition to tone-intonation interactions concerning f0 effects, Choguita Rarámuri also exhibits the following tone-specific and general non-tonal effects at prosodic boundaries (see §6.2.4; Caballero et al. 2014; Aguilar et al. 2015):

	- a. Rearticulation exclusive to HL tones
	- b. Increased lengthening of L tones

One important question yet to be addressed in depth in Choguita Rarámuri is the role that multiple dimensions of phonetic realization of tones may be playing

### 11.5 Prosodic constraints on morphological shapes

in the enhancement of the lexical tone contrast of the language: as reported in Caballero & Carroll (2015), the pitch differences between H, HL and L tones in Choguita Rarámuri are reliable, but very narrow. These results are comparable to those found for other Uto-Aztecan tonal languages (e.g., Balsas Nahuatl (Guion et al. 2010), though there are still many gaps in our knowledge in the phonetic implementation of prosodic contrasts in Uto-Aztecan languages. While Choguita Rarámuri is not a "high-density" tonal language, the functional load of tone may require additional cues to increase the dispersion of the tonal inventory.

### **11.5 Prosodic constraints on morphological shapes**

This section addresses phenomena in Choguita Rarámuri that may be classified under the rubric of morphologically-determined prosodic templatic effects, where specific morphological constructions impose output prosodic shape constraints. As described in Chapter 9, some verbal morphological constructions induce truncation of the bases to which they attach. These constructions may be analyzed as imposing prosodic restrictions where surface forms must conform to an output templatic form. As shown below, the prosodic constituents referenced in these templatic constraints are attested and deployed in different areas in the Choguita Rarámuri grammar and are thus independently motivated.

The Choguita Rarámuri constructions associated with truncation are body part incorporation (described in §11.5.1), the denominal suffix construction 'make, wear' (described in §11.5.2), both of which induce truncation of the final syllable in trisyllabic bases, as well as aspect/mood suffixes with alternating long (disyllabic) and short (monosyllabic) prosodic forms. In the analysis proposed here, allomorphy of aspect/mood suffixes that results in the long/short alternation in morphologically complex words results from a general morpho-phonological syllable truncation process, the same mechanism that is deployed in body-part incorporation and denominal constructions to enforce a prosodic templatic shape in output forms. We address each of these constructions next.

### **11.5.1 Truncation in body-part incorporation**

As described in Chapter 9, a body-part term may be incorporated to a verb stem, the head of the construction. The result is a single prosodic word, with a single stress in the first syllable of the head of the compound. Disyllabic body part terms do not undergo any prosodic changes in incorporation, but trisyllabic body part

### 11 Prosody: domains and interactions

terms undergo syllable truncation when combined with verbs in this construction. The contrast between the two types of bases is shown in the examples in (29):

	- a. /buˈsi+kaˈsì/ → *busi-ˈkâsi* eye+break 'to become blind (lit. to eye-break)' 'quedarse ciego (lit. romperse-ojo)' < SFH 06 1:112/el >
	- b. /tʃ͡aˈmèka+reˈpu/ → *tʃ͡ameˈrêpu* tongue+cut 'to cut one's tongue' 'cortarse la lengua' < SFH 07 1:187/el >
	- c. /tʃ͡eˈrewa+biʔˈwa/ → *tʃ͡ere-ˈbîwa* sweat+clean' 'to clean sweat' 'limpiarse el sudor' < SFH 07 1:187/el >

Truncation of nominal bases in noun incorporation ambiguously results from a noun incorporation stress rule, repeated in (30) below, or the initial threesyllable stress window, which requires stress to be assigned to the first three syllables of the prosodic word (see Chapter 3 and Chapter 9).<sup>7</sup>

(30) **Body part incorporation stress rule**: the head of the construction must have the stress in the first syllable.<sup>8</sup>

### **11.5.2 Truncation in denominal verb constructions in** *-ta*

A second construction that involves truncation of nominal bases is the deverbal *-tâ* 'make, wear' suffix, addressed in §9.3.5.1. As with body-part terms in noun incorporation, disyllabic nouns attaching this denominal suffix undergo no changes (31a, c), but trisyllabic nouns truncate their final syllable when combining with this construction (31e).

<sup>7</sup>On the other hand, there is evidence that syllable truncation may also be a device to satisfy the three-syllable stress window: as described in §7.3, Spanish loanwords may exhibit truncation of the source word in order to keep the original prominence within the initial three syllable stress window (e.g., *nauguˈraripo* 'let us inaugurate it', from Sp. *inauguˈrar*, with a truncated first syllable).

<sup>8</sup>This stress rule is formalized in Caballero (2011b) as ACC-TO-HEAD(1), a language-specific constraint.

### 11.5 Prosodic constraints on morphological shapes

(31) Truncation in denominal verb forms in *-tâ*

```
a. aʰkaˈrâsa
     aʰka-ˈrâ-sa
     sandal-vblz-cond
    'if they wear sandals'
    'si se enhuaracha' < SFH 08 1:47/el >
b. cf. aˈʰkà
    'sandal'
    'huarache'
c. noriˈrâma
     nori-ˈrâ-ma
     cloud-vblz-fut.sg
    'It will become cloudy.'
    'Se va a nublar.' < BFL 04 1:92/el >
d. cf. noˈrí
    'cloud'
    'nube'
e. sipuˈtâma
     sipu-ˈtâ-ma
     skirt-vblz-fut.sg
    'S/he will wear a skirt.'
    'Se va a poner falda.' < LEL 06 4:185/el >
f. cf. siˈpútʃ͡a
    'skirt'
    'falda'
```
Like body-part incorporation, this construction imposes an output shape constraint in the derived forms: the derived forms must have third syllable stress, with stress assigned in the derivational suffix. Syllable truncation in this construction thus fulfills the same prosodic function documented in body part incorporation. The prosodic template involved here is analyzed as an extended iambic foot aligned to the left edge of the prosodic word in Caballero (2011b).

### **11.5.3 Truncation in aspect/mood marking constructions**

Another phenomenon in Choguita Rarámuri where a complex interaction between phonological and morphological factors takes place is one involving pro-

### 11 Prosody: domains and interactions

sodic constraints on morphological shapes in aspectual/mood marking. A general description of these suffixes within the context of the verbal morphological structure is provided in Chapter 9, where they are described as aspectual/mood suffixes that are clustered in the Aspectual Stem, a domain within the verbal structure of Choguita Rarámuri comprising suffix positions S6 to S9. As discussed in §15.6.4, these constructions result from V-V incorporation.

The relevant aspectual/mood markers, which encode desiderative, associated motion, and auditory evidential meanings, exhibit an allomorphy pattern contrasting "long" (disyllabic) forms and "short" (monosyllabic) ones. The correspondences between allomorphs is provided in (32):

(32) Long and short allomorphs of aspect/mood suffixes


As shown here, the short (monosyllabic) allormphs correspond transparently to the first syllable of the long (disyllabic) allomorphs. These suffixes are in turnrelated to independent verbal predicates, as shown in (33):

(33) Aspect/mood suffixes and corresponding independent verbs


As can be appreciated from this comparison, the relationship between the suffixes and the corresponding independent verbs is also highly transparent: there is a perfect match in terms of segmental phonological material in the case of the associated motion and the auditory evidential, and a high degree of phonological similarity with the desiderative, where the first syllable of the suffix corresponds to the first syllable of the corresponding, independent verb. In contrast to the independent verb forms, which are prosodically independent and bear stress, the suffixes are prosodically dependent on a verb host.

A single, morphologically complex verb, may exhibit a recursion of attachment of these constructions. Their relative order with respect to each other and other constructions in the complex verb is governed by several factors, including semantic, phonological and morphological factors. An overview of the principles

### 11.5 Prosodic constraints on morphological shapes

governing affix order is provided in Caballero (2010). A separate mechanism governs the form and distribution of allomorphs of aspect/mood markers in Choguita Rarámuri. This is addressed next.

### **11.5.3.1 Allomorph distribution**

The distribution of long and short allomorphs is predictable on the basis of the existence of outer suffixes attached to a base containing aspect/mood markers: short allomorphs are always followed by other suffixes, while long allomorphs are aligned to the right edge of the prosodic word. These two distributions are shown in (34) and (35).

### (34) Short allomorph distribution


/reʔe-ˈbû-ri-simi-li/ stone-gather-caus-mot-pst 's/he went along making them gather stones' 'va a querer ir juntando piedras' < SFH 07 2:63/el >

d. [tiˈtʃ͡iksima]

/tiˈtʃ͡i-ki-simi-ma/

comb-appl-mot-fut.sg

'S/he will go along making them comb her/him.'

'Va a ir haciéndola que la peine.' < SFH 07 2:67/el >

	- a. [ˈsûrnili] /ˈsû-ri-nale/ sew-cause-desid 'S/he wants to make them sew.' 'Quiere hacerlos coser.' < BFL EDCW(52/el > b. [poˈlâptinili] /poˈlâ-pi-ti-nale/ cover-rev-refl-desid 'S/he wants to uncover herself.' 'Se quiere destapar.' < BFL 08 1:56/el > c. [ˈnârisimi] /ˈnâri-simi/ ask-mot 'S/he is going along asking.' 'Va a ir preguntando.' < SFH 08 1:148/el > d. [toˈrétʃ͡ani] /toˈré-tʃ͡ane cackle-ev 'It sounds like cackling.' 'Suena a que están cacareando.' < SFH 07 1:7/el > e. [uˈbástʃ͡ani] /uˈbá-simi-tʃ͡ane/ bathe-mot-ev 'It sounds like they are going along bathing.' 'Suena que van bañándose.' < SFH 08 1:150/el >

One exception to this distribution is found in morphologically complex words containing an unstressed root that immediately attaches a desiderative suffix: in

### 11.5 Prosodic constraints on morphological shapes

these cases, the long allomorph of the suffix is selected, regardless of whether there are outer suffixes or not. This pattern results from the stress properties of the suffix and the bases with which it may combine: the desiderative suffix is the only stress shifting suffix among the aspect/mood markers described here (the associated motion and the evidential are stress neutral). The long, stressed allomorph may undergo post-tonic syncope of the final suffix vowel when combining with unstressed roots, but is always attested in its long form. The relevant examples are shown in (36).

(36) Stressed allomorph of the desiderative


As (36b) and (36c) show, the last vowel of the long allomorph may undergo deletion (36b), but there is no evidence for syllable truncation, regardless of the presence or absence of outer suffixes.

Outside of this pattern, long and short allomorph distribution is not dependent on the stress properties nor the stress position within the stem. This is shown in (37) and (38).

### 11 Prosody: domains and interactions

	- a. [noriˈwísimi] /noriˈwí-simi/ vanish-mot 'It goes along vanishing.' 'Se va desapareciendo.' [FLP in61(482)/in > b. [wikaˈrâsika]
		- /wikaˈrâ-si-ka/ sing-mot-ger 'It was going along singing.' 'Iba de pasada cantando.' < BFL 06 EJP(10)/el >
	- a. [ˈkétʃ͡isimi] /ˈkétʃ͡i-simi/ chew-mot 'S/he goes along chewing.' 'Va queriendo mascar.' < SFH 08 1:145/el > b. [poˈlâptisio] /poˈlâ-p-ti-si-o/ cover-rev-refl-mot-ep 'It goes along uncovering itself.' 'Va destapándose.' < BFL 08 1:56/el >

Long and short allomorphs are not lexically conditioned by their bases, either. This is shown in (39).

(39) a. [oˈpés**tʃane**] /oˈpési-**tʃane**/ vomit-ev 'It sounds like somebody is throwing up.' 'Se oye que vomitan.' < BFL 07 rec300/el >

### 11.5 Prosodic constraints on morphological shapes

b. [oˈpés**tʃi**nilo] /oˈpési-**tʃane**-nale-o/ vomit-ev-desid-ep 'It sounds like somebody wants to throw up.' 'Se oye que quieren vomitar.' < BFL 07 rec300/el > c. [ˈnâri**simi**] /ˈnâri-**simi**/ ask-mot 'He goes along asking.' 'Va preguntando.' < SF 08 1:148/el > d. [ˈnâr**si**ma] /ˈnâri-**simi**-ma/ ask-mot-fut.sg 'S/he will go along asking.'

'Va a ir preguntando.' < SFH 08 1:148/el >

Long allomorphs of aspect/mood markers may be followed by vocalic TAM suffixes. These vocalic suffixes replace the final vowel of the allomorph, as shown in (40).

(40) Replacement of final vowels of long allomorphs

a. [kaˈtʃ͡ísnili] /katʃ͡í-simi-nale-i/ spit-mot-desid-impf 'S/he wants to go along spitting.' 'Quiere ir escupiendo.' < SFH 08 1:75/el >

b. [toˈnáltʃ͡ino]

/tò-nále-tʃ͡ane-o/ take-desid-ev-ep 'It sounds like they want to take it.'

'Se oye que se lo quieren llevar.' < BFL 06 5:148/el >

As shown in (41), short allomorphs do not have their vowel replaced by these vocalic suffixes.

### 11 Prosody: domains and interactions

	- a. [kotʃ͡iˈnálsia] /kotʃ͡i-ˈnále-simi-a/ sleep-desid-mot-progr 'S/he is going along wanting to sleep.' 'Va queriéndose dormir.' < BFL 08 1:60/el >

The hypothetical and unattested form \**kotʃ͡iˈ-nal-s-a*, where the progressive *-a* suffix would replace the vowel of the short allomorph of the associated motion (*-si*), is ungrammatical with the intended reading of example (41).

A single morphologically complex word may contain more than one aspect/ mood marker, given that these constructions are semantically compatible and do not exhibit any blocking effects. In these cases, the main generalizations about their distribution is the same as the one stated above: if a base containing two aspect/mood markers attaches more suffixes, then both aspect/mood markers will be short. This is shown in (42).

(42) a. [ˈpák**sini**mi] /ˈpáki-**simi-nale**-mi/ brew-mot-desid-irr.sg 'Perhaps s/he may want to go along brewing.' 'A lo mejor va a querer ir colando.' < SFH 08 1:147/el > b. [iˈtʃí**nsi**ma] /iˈtʃí-**nale-simi**-ma/ sow-desid-mot-fut.sg 'S/he will go along wanting to sow.' 'Va a querer ir sembrando.' < LEL 06 EDCW123/el >

### **11.5.3.2 A prosodically motivated morpho-phonological alternation**

The differences in prosodic shape between long and short allomorphs may be analyzed as resulting from a phonological process of syllable truncation, albeit one that is morphologically-conditioned, as the syllable truncation processes attested in body part incorporation and denominal verb constructions with the *-ta* suffix. The claim here is that syllable truncation, whether deleting a stem syllable or an affix syllable, is deployed in specific morphological constructions in order to satisfy the prosodic shape output forms must have.

### 11.5 Prosodic constraints on morphological shapes

The prosodic shape of aspect/mood markers in Choguita Rarámuri was treated in Caballero (2008) as instantiating a suppletive allomorphy pattern. The distinction between a suppletive allomorphic pattern, a morpho-lexical distinction, vs. a morpho-phonological alternation, where allomorphs are derived via phonological processes from a single underlying form, is not a trivial one (Kiparsky 1996; Paster 2006).<sup>9</sup> The proposal here is that short allomorphs are derived from long allomorphs via a morphologically specific process of syllable truncation.

Choguita Rarámuri aspect/mood markers form a single morphological construction, which can be analyzed as involving V-V incorporation. The 'long' allomorphs are disyllabic, matching the canonical shape of stems in the language. These constructions are analyzed here as associated to a prosodic template, a bimoraic foot (σμσμ, σμμ) which matches the minimal prosodic word in Choguita Rarámuri (more details are provided in §11.1 below) aligned to the *right* edge of the prosodic word. Relevant examples showing the proposed metrical structure of output forms is provided in (43):

(43) A bimoraic foot aligned to the right edge of the prosodic word


<sup>9</sup>Two criteria discussed in Kiparsky (1996) are (i) whether the allomorphy pattern is specific to lexical items (suppletion) or general in the language (morpho-phonology), and (ii) whether the allomorphy pattern involves more than one segment (suppletion) or a single segment (morphophonology). The argument being that it is expected that morpho-phonological alternations are predictable based on general phonological processes and that general phonological processes generally target a single segment.

### 11 Prosody: domains and interactions

TAM vocalic suffixes that replace the stem final vowel do not alter the templatic requirement, allowing these vocalic suffixes to attach to disyllabic suffixes without inducing truncation. The parsing of these suffixes into the metrical structure is as follows:

(44) Parsing of vocalic TAM suffixes with long allomorphs


TAM suffixes with the CV shape do induce truncation of aspect/mood suffixes in order to satisfy the prosodic template requirement associated with these constructions. This is shown in (45), where deleted syllables are highlighted in boldface in the underlying representation.

(45) Truncation with TAM suffixes with CV shape


Other aspects of the phonological shape of long and short allomorphs of aspect/mood suffixes is predictable based on fully general phonological processes of the language, namely the vowel quality of individual forms and the possibility of vowel deletion dependent of stress placement and position within the prosodic word. Specifically, long and short allomorphs undergo post-tonic vowel reduction or deletion, which are widespread in the language (and as described in §5.2):

### 11.5 Prosodic constraints on morphological shapes

non-final post-tonic vowels may be neutralized in terms of their height (e.g., / tʃ͡**a**ne/ [-tʃ͡ine] 'ev', /-n**a**le/ [-ni] 'desid') (e.g., (46a–b) or they may be deleted (e.g., /-nal**e**/ [-nal] 'desid', /-s**i**/ [-s] 'mot') (e.g., (46c–d)).

	- a. [rosoˈwâtʃ**i**no] /rosowâ-tʃ**a**ne-o/ cough-ev-ep 'It sounds like coughing.' 'Se oye que tosen.' < BFL 07 rec301/el >
	- b. [naˈlàn**i**ma]

/naˈlà-n**a**le-ma/

cry-desid-fut.sg

'S/he will want to cry.'

'Va a querer llorar.' < SFH 08 1:125/el >

c. [koʔˈnált**i**ma]

/koʔa-ˈnál**e**-ti-ma/

eat-desid-caus-fut.sg

'She will make him want to eat.'

'Lo va a querer hacer comer.' < SFH 07 EDCW(30)/el >

d. [wikuˈbastʃ͡ani] /wikuˈba-s**i**mi-tʃ͡ane/ whistle-mot-ev 'It sounds like someone is going along whistling.' 'Se oye como que van chiflando.' < SFH 08 1:158/el >

There are alternative analyses for the form and distribution of long and short allomorphs of aspect/mood suffixes in Choguita Rarámuri. One alternative would involve positing syllable truncation as a strategy to reduce the number of unparsed syllables in Prosodic Words. This hypothesis would be similar to proposals made for Hopi (Southwest US) (Gouskova 2003) and Southeastern Tepehuan (Tepiman; Mexico) (Kager 1997).

While Choguita Rarámuri lacks secondary stress, a possible alternative analysis of syllable truncation would involve optimization of metrical structure if truncation is assumed to take place when affixation contributes unparsed syllables post-tonically. Such an analysis requires the following assumptions: (i) there

### 11 Prosody: domains and interactions

is iterative syllable parsing; (ii) feet are binary; and (iii) the conditioning environment for syllable truncation is rendered opaque after truncation. This hypothetical analysis can be exemplified with a comparison between the posited metrical structure of actual forms and corresponding hypothetical abstract forms with no syllable truncation. It would thus be possible to posit that syllable truncation in these contexts is due to a mechanism that reduces unparsed syllables within Prosodic Words in Choguita Rarámuri (a mechanism enforced by a constraint such as Parse-σ or another similar rule or constraint).

This hypothesis would explain syllable truncation patterns in morphologically complex words with more than one aspect/mood disyllabic suffix (recursive compounding), where truncation would result in a form with hypothetical exhaustive parsing. This is exemplified with the analyzed form in (47).

	- b. cf. /simi/ → [-s]

This hypothesis would also predict that truncation would apply to every eligible disyllabic suffix if the output form would be exhaustively parsed into metrical feet. However, and as shown in the examples in (48), truncation does not take place every time there is recursive affixation of aspect/mood markers, even if this would optimize the surface phonological form (these examples are shown with a foot aligned to the right edge of the Prosodic word, but an alternative parsing does not affect the main argument):

	- b. \*(raraˈhîp)Ft(ti-tʃ͡i)Ft
	- c. /aˈwí-ri-si-nale/ → (aˈwí-r)Ft-si-(nili)Ft dance-caus-mot-desid 'S/he wants to make her/him go long dancing.' 'Quiere ir haciéndola bailar'

### 11.5 Prosodic constraints on morphological shapes

As shown in these examples, the attested forms have an unparsed syllable, while the hypothetical but unattested words with recursive syllable truncation would yield Prosodic Words with syllables exhaustively parsed into metrical feet. Thus, a single mechanism of syllable truncation to reduce unparsed syllables does not predict the attested surface forms in these contexts in Choguita Rarámuri.

In sum, the alternation that yields monosyllabic allomorphs from disyllabic ones for aspect/mood suffixes in Choguita Rarámuri does not result in global phonological optimization in the language.

### **11.5.4 Prosodic templates in Choguita Rarámuri**

The proposal set forth here is that aspect/mood markers are a single type of construction associated to a syllable truncation process, a phenomenon also associated to body part incorporation constructions and deverbal constructions with the *-ta* 'wear/make' suffix (addressed in §11.5.1 and §11.5.2, above).

Syllable truncation satisfies different prosodic templates: as discussed in (X), a morphological stress rule is proposed for body part incorporation, requiring the head of the construction, the second member of the compound, to have stress in its first syllable. The first member of the compound may undergo truncation to satisfy this requirement. The same prosodic output shape is involved in the deverbal construction with the *ta* 'make/wear' suffix. In both cases, the resulting surface form has an extended iambic foot aligned to the left edge of the prosodic word. This prosodic shape is schematized in (49):

(49) Prosodic template in body part incorporation and deverbal constructions (<σ>σˈσ)PrWd

In the case of aspect/mood suffixes, the proposal above is that the prosodic template associated with these constructions involves a bimoraic foot (σµµ, σµσµ). This prosodic template is aligned to the right of the Prosodic Word. This prosodic shape is schematized in (50):

(50) Prosodic template in aspect/mood markers (σµσµ)PrWd (σµµ)PrWd

Given their transparent relationship with independent verbs, these aspect/ mood markers may be analyzed as involving *verbal compounding* as part of a process of V-V incorporation. Their association with a minimal Prosodic Word

### 11 Prosody: domains and interactions

would thus be consistent with the observation that stems are canonically disyllabic, while affixes are canonically monosyllabic. In Mountain Guarijío, Miller (1996) analyzes some verbal suffixes as straddling the boundary between verbal compounds and suffixes, including the associated motion form *-si*, the cognate form of Choguita Rarámuri associated motion *-simi* suffix, which he analyzes as an intermediate form between a compounded verbal stem and a suffix.

The allomorphy pattern documented with aspect/mood suffixes provides some clues about the prosodic organization of Choguita Rarámuri. There are two morphoprosodic restrictions in morphologically complex words in Choguita Rarámuri:


# **12 Noun phrases**

This chapter addresses the internal structure of noun phrases in Choguita Rarámuri. While the development of syntactic arguments for an in-depth characterization of these constructions lies beyond the scope of this grammar, there is evidence for the existence of a noun phrase constituent broadly defined as a syntactic constituent that may function as an argument of verbal predicates (Dryer 2007b). As described in Chapter 10, a subset of minor word classes can be characterized by their ability to head noun phrases (pronouns) or combine with head nouns in noun phrases (demonstratives, adjectives, numerals, definite articles and quantifiers). Noun phrases which contain a head noun and one of these modifiers are described in §12.1. Noun phrases that involve genitive constructions or other complex modifiers are addressed in §12.2.

### **12.1 Simple noun phrases**

Word order in other Rarámuri varieties (Rochéachi Rarámuri (Morales Moreno 2016: 64)) and closely related varieties (Mountain Guarijío (Miller 1996: 73)) is described as being verb final, with order of constituents exhibiting both syntactic and pragmatic conditioning. While the ordering of constituents in the Choguita Rarámuri clause likewise appears to exhibit syntactic restrictions and pragmatically-determined order variation, word order within the noun phrase can be described in terms of a template, which is schematized in (1).

(1) Word order in noun phrases Demonstrative – Definite article – [Numeral/Quantifier] – Adjective – Noun

Noun phrases, replicating the basic word order of clauses, are head final, i.e., the order of constituents in noun phrases involves modifiers preceding the head noun. In contrast to nouns, pronouns are case marked (see §10.1) and occur as the single constituent in noun phrases, though there are examples attested in the Choguita Rarámuri corpus where pronouns occur adpositionally with other nouns, as shown in (2) (see also Miller 1996:229) for description of the same phenomenon in Mountain Guarijío).

### 12 Noun phrases

(2) *ˈkíti ˈbéli ti ˈwé riˈsóati ka ruˈwá tamuˈhê raˈlàmuli ko ba ˈkíti* that.is.why *ˈbéli=ti* indeed=1pl.nom *ˈwé* int *riˈsó-a=ti* struggle-prog=1pl.nom *ka* cop.irr *ru-ˈwá* say-mpass *tamuˈhê* 1pl.nom *raˈlàmuli=ko* people=emph *ba* cl 'That is why it is said that we are very poor (struggle a lot), us the Rarámuri people.' 'Por eso dicen que estamos muy pobres (batallamos mucho) nosotros los tarahumaras.' {SFH tx128:01:58.8}

With the exception of numerals and quantifiers, which do not co-occur, all other modifiers may potentially co-occur with each other and head nouns. The discussion below addresses the behavior of each of the different elements that may modify a head noun in noun phrases and the relative ordering of modifiers when they co-occur in noun phrases.

### **12.1.1 Demonstratives**

As discussed in §10.2.2, demonstratives may be use adnominally. Demonstratives are attested preceding the head noun in noun phrases.<sup>1</sup> This is exemplified in (3). Noun phrases are indicated with square brackets.

(3) a. *aʔˈlì ˈétʃi muˈkî ko ˈwé saˈpù ˈwé siˈnàka na "ˈwé saˈpù aˈsísi!" aˈnèli aʔˈlì* and *[ˈétʃ͡i* dem *muˈkî]=ko* woman=emph *ˈwé* int *saˈpù* fast *ˈwé* int *siˈnà-ka* scream-ger *na* then *ˈwé* int *saˈpù* fast *aˈsísi* get.up.imp.sg *aˈn-è-li* say-appl-pst

> 'And then the woman with a hurry shouted: "hurry up, get up", she said to him.'

> 'Y entonces la mujer apurada gritando le dijo: "pronto levántate" le dijo.' {LEL tx5:01:21.3}

b. *aʔˈlì biˈlá ko waˈbé biˈlá kiˈʔà ˈníla ra pa, kuˈrí ke ˈtʃ͡ó me biwaˈtʃ͡êatʃ͡i ˈnà kaˈwì ba*

<sup>1</sup> For Mountain Guarijío, Miller (1996) reports that demonstratives may follow nouns in noun phrases, though he notes this is rarely attested (1996: 235). No similar cases have been documented so far in Choguita Rarámuri.

12.1 Simple noun phrases

*aʔˈlì* and *biˈlá=ko* really=emph *waˈbé* long.ago *biˈlá* really *kiˈʔà* before *ˈní-la* cop-rep.sg *ra* say.prs *pa* cl *kuˈrí* recently *ke* neg *ˈtʃ͡ó* yet *me* almost *biwa-ˈtʃ͡ê-a-tʃ͡i* hard-vrbl-prog-loc *[na* this *kaˈwì]* world *ba* cl 'And then they say it was a long time ago when this world hadn't consolidated (become solid) yet.' 'Y entonces dicen que fue mucho antes, cuando todavía no amacizaba bien este mundo.' {SFH tx43:11:11.2}

The following examples (in (4)) show that demonstratives appear in initial position when co-occurring with other modifiers within the noun phrase.<sup>2</sup>

(4) a. *ˈápu riˈká raˈpâko ˈétʃi biˈlé muˈkî ke naˈwàlo la ˈro ˈnápi* sub *riˈká* like.that *raˈpâko* yesterday *[ˈétʃ͡i* dem *biˈlé* one *muˈkî]* woman *ke* neg *naˈwà-lo-la* arrive-pst-rep.ds *ˈru* say.prs 'like yesterday that one woman didn't arrive, they say' 'así como ayer esa mujer no llegó, dicen' {JLG co1237:03:31.6} b. *ˈwé maˈtʃína iˈnáli ˈés ti koriˈmá ˈwé* int *maˈtʃína* see.prs *iˈnáli* go.sg-ep *[ˈétʃ͡i* dem *ti* def.bad *koriˈmá]* fire.bird 'It is very noticeable when the *korima* (fire bird) is passing by.' 'Se ve muy bien cuando va (pasa) el *korimá* (pájaro de fuego).' {LEL tx5:04:51.7}

In (4a), the demonstrative occurs in phrase initial position, followed by a numeral (*biˈlé* 'one') and the head noun (*muˈkî* 'woman'). In (4b), the demonstrative precedes a definite article (*ti* 'the, negative stance') and the head noun (*koriˈmá* 'fire bird') in the noun phrase.

### **12.1.2 Definite articles**

Definite articles, which encode number and affective stance of a definite referent (§10.6), precede head nouns in noun phrases, as exemplified in (5).<sup>3</sup>

<sup>2</sup>As described in §3.5.2, alveopalatal affricates may depalatalize and deaffricate in fast speech in high frequency word combinations that include the demonstrative *ˈétʃ͡i*, as exemplified in (4b). <sup>3</sup>As described in §10.6, while the definite articles encoding negative stance are generally attested when the speaker conveys a negative attitude towards the referent, they can also be attested with a neutral connotation, as in (5c).

### 12 Noun phrases

(5) a. *ke ˈpé ˈtâʃi ˈhú pa, ˈwé beti ˈwé wiˈtʃ͡íwami ˈtʃéti erˈmâno ba ke* neg *ˈpé* little *ˈtâsi* neg *ˈhú* cop.prs *pa* cl *ˈwé* int *be=ti* emph=1pl.nom *ˈwé* int *wiˈtʃ͡í-w-ame* believe-mpass-ptcp *[ˈtʃ͡éti* def.pl.bad *erˈmâno]* evangelical.brothers *ba* cl 'Well no, we believe the evangelical brothers very much.' 'Pues no, les hacemos mucho caso a los hermanos evangélicos.' {FLP in243:02:08.4} b. *ka ˈtʃ͡è ko niˈrú ... ˈnàri ... ˈtʃéti ˈâba tʃ͡aˈbèi ko ba ka* because *ˈtʃ͡è=ko* because=emph *niˈrú* cop.impf *ˈnàri* then *[ˈtʃ͡éti* def.pl.bad *ˈâba]* broad.bean *tʃ͡aˈbèi=ko* before=emph *ba* cl 'because there were no broad beans before.' 'porque no había habas antes.' {FLP in61:02:24.4} c. *aʔˈlì biˈlá ˈhípi biˈlá ˈmá noˈkí maˈkò maˈrí bamˈpáma ˈlé ˈtí tiˈwé niˈhê ˈkútʃ͡ara baˈtʃ͡áwara ba aʔˈlì* and *biˈlá* indeed *ˈhípi* now *biˈlá* indeed *ˈmá* already *noˈkí* almost *maˈkò* ten *maˈrí* five *bam-ˈpá-ma* have.birthday-inch-fut.sg *aˈlé* dub *[ˈtí* def.sg *tiˈwé]* girl *niˈhê* 1sg.nom *ˈkútʃ͡a-la* child-poss *baˈtʃ͡á-wa-la* first-vblz-poss *ba* cl 'And now the girl, my daughter, will turn fifteen, my oldest child.' 'Pues ahora ya la muchacha ya va a cumplir quince años, mi hija la más grande.' {SFH tx43:04:05.6}

Definite articles may co-occur with other modifiers in nominal phrases, including demonstratives, as shown above in (4b), and adjectives, as shown in (6).

(6) *ˈtʃêram ˈsûs ba, ti ˈtʃêram bauˈtîʃ ma ba [ˈtʃêram* elder *ˈsûs]* Jesús *ba,* cl *[ti* def.sg *ˈtʃêram* elder *bauˈtîʃ]* Bautista *ma* also *ba* cl 'elder Jesús, also elder Bautista' 'don Jesús, también don Bautista' {ME in484:07:18.5} 12.1 Simple noun phrases

### **12.1.3 Numerals**

Numerals, like other modifiers, are attested preceding the head noun in nominal phrases. This is exemplified in (7).

	- *[biˈlé* one *ariˈmûli]* decaliter *[oˈkwâ* two *ariˈmûli]* decaliter *ma* or *[biˈkiá* three *ariˈmûli]* decaliter *ma* or 'one decaliter or two decaliters or three decaliters' 'un decalitro o dos decalitros o tres decalitros' {LEL tx68:00:25.8}

The following examples show the relative ordering of numerals and other modifiers, including demonstratives.


'cuando se hace una fiesta grande en una casa' {GFP tx1009:00:44.9}

As shown in these examples, numerals follow demonstratives (8a) but precede adjectives (8b).

The behavior and distribution of numerals in noun phrases in Choguita Rarámuri contrasts with that of closely related Mountain Guarijío, where the most frequent order in noun phrases is that of head nouns followed by numerals (Miller 1996: 244).

### 12 Noun phrases

### **12.1.4 Quantifiers**

Quantifiers in Choguita Rarámuri must also precede the head noun in noun phrases, as exemplified in (9).

(9) a. *aʔˈlì ko ˈtʃ͡ó ˈnà ˈhônsa ko ˈá biˈláni ˈwé ... ˈwé ˈtʃ͡ó ˈam tʃ͡i riˈká wiˈtʃ͡íwi ˈtʃ͡ó ˈnè ˈá ˈmá wiˈʰkâ naˈmûti ˈa aʔˈlì=ko* and=emph *ˈtʃ͡ó* dem *ˈnà* dem *ˈhônsa=ko* since=emph *ˈá* aff *biˈlá=ni* truly=1sg.nom *ˈwé* int *we* int *ˈtʃ͡ó* also *ˈam* aff=dem *tʃ͡i* dem *riˈká* like.that *wiˈtʃ͡íwi* believe.prs *ˈtʃ͡ó* also *ˈnè* 1sg.nom *ˈá* aff *ˈmá* already *[wiˈʰkâ* many *naˈmûti]* things *ˈa* indeed 'and then from there I knew how to believe in all of these things' 'entonces de ahí ya supe cómo creer yo muchas cosas' {LEL tx71:04:48.6} b. *ˈwé biˈlá ... riˈsóa iˈkí wiˈʰkâ paˈkótam pa ˈwé* int *biˈlá* indeed *riˈsó-a* suffer-prog *iˈkí* happened *[wiˈʰkâ* many *paˈkót-ame]* wash-ptcp *pa* cl 'Indeed many people died (they endured suffering).' 'En verdad murieron muchas personas (les pasó sufrimiento).' {LEL tx372:04:51.5} c. *ˈápi iˈsêlikami ˈká ˈlé, biˈkiánika, suˈwâba maˈjôra ma ˈápi* sub *[i-ˈsêri-kami]* pl-governor.pl-ptcp *ˈká* cop.irr *aˈlé* dub *biˈkiá-ni-ka* three-incl-coll *[suˈwâba* all *maˈjôra]* manager *ma* also 'those who are governors, the three of them, all of the managers, too' 'los que son gobernadores, los tres (gobernadores), todos los mayores también' {JMF tx816:00:36.7}

As described in §10.5, some quantifiers in Choguita Rarámuri are derived from numerals (e.g., *oˈkwâ* 'few' is derived from the numeral 'two'). There are no recorded instances of numerals and quantifiers co-occurring in noun phrases. Quantifiers may nevertheless co-occur with other modifiers, including demonstratives, as shown in (10).

12.1 Simple noun phrases

(10) *aʔˈlì ˈétʃi ʃiˈnêami raˈlàmuli ko ˈhê aˈníli aʔˈlì* then *[ˈétʃ͡i* dem *siˈnêami* everyone *raˈlàmuli]=ko* men=emph *ˈhê* it *aˈní-li* say-pst 'then all the men said' 'entonces todos los hombres dijeron" {LEL tx32:06:32.5}

### **12.1.5 Adjectives**

Adjectives in Choguita Rarámuri may be basic (a small class of underived adjective roots) or may be derived through morphological means (§10.3). Adjectives, both basic and derived, may also be used predicatively in copular clauses (§13.2.3). A comprehensive description of adjectives in Choguita Rarámuri is provided in Islas Flores (2010). This section addresses the distribution of basic adjectives in noun phrases. As shown in (11), adjectives precede head nouns.

(11) a. *aʔˈlì ne maˈhâali es ˈkútʃi koˈtʃî] ko aʔˈlì* and *ne* int *maˈhâ-li* fear-pst *es* dem *[ˈkútʃ͡i* little *koˈtʃ͡î]=ko* dog=emph 'And the dog got very scared.' 'Y se asustó mucho el perrito.' {BFL tx191:04:24.7} b. *waʔˈlû kapaˈnî aˈnítʃ͡ini ba [waʔˈlû* big *kapaˈnî]* bell *aˈní-tʃ͡ini* make.sound-ev.sound *ba* cl 'The big bell rang (was heard).'

'Se oyó sonar la campana grande.' {LEL tx223:03:29.1}

The following example (shown above in (8b) and repeated here in (12)) shows the relative ordering of adjectives with respect to other modifiers within noun phrases, namely following numerals and preceding the head noun.

(12) *biˈlé waʔˈlû fiêsta oriˈwábatʃ͡i biˈléna bitiˈtʃ͡í [biˈlé* one *waʔˈlû* big *fiêsta]* party *oriˈwába-tʃ͡i* make.mpass-temp *biˈlé-na* one-incl *bitiˈtʃ͡í* house 'when one big party is made at one house' 'cuando se hace una fiesta grande en una casa' {GFP tx1009:00:44.9}

Adjectives may be modified by adverbs, as shown in (13), where a degree adverb, *we* 'very', modifies the adjective *ˈtʃ͡áti*, which in turn modifies the head noun (the deverbal noun *nawiˈri* 'disease').

### 12 Noun phrases

(13) *apʔaˈlì nawiˈrí naˈwà ˈsíli we ˈtʃáti nawiˈrí ˈhêmi naˈʔî puêblo ˈnápi* sub *aʔˈlì* and *nawi-ˈrí* get.sick-nmlz *naˈwà* arrive *ˈsí-li* go.sg-pst *[we* int *ˈtʃ͡áti* ugly *nawi-ˈrí]* get.sick-nmlz *ˈhêmi* this *naˈʔî* here *puêblo* town 'when a very dangerous (ugly) disease arrived in this town' 'cuando llegó una enfermedad muy peligrosa en este pueblo' {LEL tx372:00:24.0}

The distribution of adjectives in noun phrases in Choguita Rarámuri differs from that of adjectives in Mountain Guarijío noun phrases, where adjectives can be attested post-nominally or may appear in discontinuous constructions (Miller 1996). With adjectives and other types of modifiers, Choguita Rarámuri exhibits stricter ordering restrictions within the noun phrase than those attested in Mountain Guarijío.

### **12.2 Complex noun phrases: Possessive constructions**

Choguita Rarámuri possessive constructions involve a noun modified by a noun phrase encoding a range of meanings, including part-whole relationships and possession or ownership. The term 'possessive' is used here to refer to adnominal constructions, even if the semantic relations encoded by the construction do not involve possession. Examples of possessive constructions with different meanings are exemplified in (14).

	- a. *Meronymic (part-whole) ˈmêsa roˈnôla ˈmêsa* table *roˈnô-la* leg-poss 'the table's leg' 'la pata de la mesa' < BFL 09 1:33/el >

b. *Possession raˈniêli uˈpîla raˈniêli* Daniel *uˈpî-la* wife-poss 'Daniel's wife' 'la esposa de Daniel' {LEL tx5:03:03.8}

### 12.2 Complex noun phrases: Possessive constructions

As discussed in §8.4.1, Choguita Rarámuri exhibits possessive classification, where there is a contrast between two types of possession, alienable and inalienable, determined by the lexical properties of possessed nouns. Inalienable nouns (kinship terms, body parts and other nouns) are obligatorily possessed. Both alienable and inalienable nouns may be used in a possessive construction. The schema in (15) shows the possessive construction template.<sup>4</sup>

(15) [[possessor: noun/pronoun]N1 [possessum: noun-poss]N2]NP

The possessum takes the possessive suffix (poss) *-la* (optionally preceded by a suffix *-wa* with some nouns) that encodes a possessor, with no person marking on the possessed noun.

### **12.2.1 Nominal possessors**

Examples of possessive constructions with nominal possessors are provided in (16), which follow the template schematized in (15): a possessor noun precedes the possessum marked with the possessive suffix.

(16) a. *tʃ͡omaˈlî aʔˈwâla tʃ͡omaˈlî* deer *aʔˈwâ-la* horn-poss 'deer's horn' 'cuerno de venado' {SFH tx152:07:40.4} b. *teˈwé maiˈlâ teˈwé* girl *mai-ˈlâ* father.female.ego-poss 'girl's father' 'papá de la muchacha' {LEL tx32:13:56.7} c. *reˈhòi ariˈwâla reˈhòi* man *ariˈwâ-la* soul-poss 'man's soul'

<sup>4</sup>Mountain Guarijío (Miller 1996) and Norogachi Rarámuri (Brambila 1953;1976) are documented to possess a classifier for domesticated animals (*puhkú* in Mountain Guarijío, *bukú* in Norogachi Rarámuri), marked with the possessive suffix and followed by the specific animal name. This construction has not been documented in Choguita Rarámuri, which instead encodes possession of domesticated or mythical animals through a specialized denominal verb of possession, as described in §9.3.5.4 (a strategy also documented in Mountain Guarijío (Miller 1996: 149)).

### 12 Noun phrases

'alma del hombre' {LEL tx5:01:11.3}

There are no attested instances of possessive constructions with nominal possessors with alternative order of possessor and possessum, nor instances of intervening elements between possessor and possessum in the Choguita Rarámuri corpus (a separate appositional construction that encodes a possession relationship through a specialized possessive noun is described in §12.2.3 below).

### **12.2.2 Pronominal possessors**

The Choguita Rarámuri possessive construction with pronominal possessors exhibits the same structure as the one with nominal possessors: the possessor precedes the possessum, which is marked with the suffix *-la* to encode a possessor. Examples of pronominal possessors are provided in (17). As discussed in §10.1, personal pronouns encode case in Choguita Rarámuri. Pronominal possessors are encoded with the nominative-marked pronoun series.

(17) a. *ˈnè waˈsála ˈnè* 1sg.nom *waˈsá-lâ* cultivation.land-poss 'my cultivation land' 'mi tierra de cultivo' < BFL 09 1:61/el > b. *ˈmò uˈmûala ˈmò* 2sg.nom *uˈmûa-la* grandfather-poss 'your great grandfather'

'tu bisabuelo' {FLP in61:04:55.2}

c. *ˈémi oˈtʃ͡îpala ˈémi* 2pl.nom *oˈtʃ͡îpa-la* paternal.grandfather-poss 'your (pl.) grandfather' 'abuelo de ustedes' {SFH tx12:09:13.8} 12.2 Complex noun phrases: Possessive constructions

d. *ˈnè waˈrîla ˈnè* 1sg.nom *waˈrî-la* basket-poss 'my basket' 'mi canasta' < BFL 11/07/09/el >

Pronominal possessors exhibit variation between the full and reduced forms in possessive constructions with no apparent grammatical conditioning, as shown by the examples in (18), which are equivalent in meaning.<sup>5</sup>

(18) a. *neˈhê iˈjêla neˈhê*

1sg.nom *iˈjê-la* mother-poss 'my mother'

'mi madre' {SFH tx475:06:17.3}

b. *ˈnè iˈjêla ˈnè* 1sg.nom *iˈjê-la* mother-poss 'my mother' 'mi madre' {BFL tx1:00:55.5}

There are cases attested where the pronominal possessor is omitted, if the possessor is a third person argument. This is exemplified in (19), where the possessor is established previously in the discourse.

(19) *aʔˈlì ko ˈkútʃala biˈlá onoˈkáli ba ˈkîni ˈtîo ba*

*aʔˈlì=ko* and=emph *ˈkútʃ͡a-la* children-poss *biˈlá* indeed *onoˈká-li* do-pst *ba* cl *ˈkîni* 1.poss *ˈtîo* uncles *ba* cl

'And then later the children do it (plow), my uncles.'

'Y luego ya después lo hacen (barbechan) los hijos, mis tíos.' {LEL tx130:05:13.3}

Choguita Rarámuri also has a dedicated pronominal form encoding first person possessors (singular or plural) of kinship terms. This is exemplified above in (19) (*ˈkini ˈtîo* 'my uncles'), and further exemplified in (20).

<sup>5</sup>This kind of system is also reported in Rochéachi Rarámuri, though Morales Moreno (2016) describes restrictions in the distribution of reduced pronominal forms in Rochéachi Rarámuri (2016: 72). In the Choguita Rarámuri corpus the long vs. short forms appear to be interchangeable, with a higher frequency of the short forms, though no detailed counts have been carried out controlling for discourse contexts.

### 12 Noun phrases

(20) a. *"ˈnà ko ʃiʔˈrînuwala paˈkóala ˈkám pa" ˈhê biˈlá aˈní ˈétʃ͡i ˈkîni ˈwênuwa tʃ͡aˈbèe ˈna=ko* this.one=emph *siʔˈrînuwala* shirínuwala.ritual *paˈkó-a-la* baptize-prog-purp *ˈká=mi* cop.irr=dem *pa* cl *ˈhê* it *biˈlá* indeed *aˈní* say.prs *ˈétʃ͡i* dem *ˈkîni* 1.poss *ˈwênuwa* parents *tʃ͡aˈbè* before "'This is the one that baptizes with shirínuwala", that's what our parents used to say long ago.' "'Este es el que bautiza de shirínuwala", eso decían nuestros padres antes.' {SFH tx475:07:57.9} b. *ˈnápu ˈriˈká ke naˈtâsa ˈkini ˈwênuwala ˈro ba ˈni ˈnápu* like *ˈriˈká* that *ke* neg *naˈtâ-sa* think-cond *ˈkini* 1.poss *ˈwênuwa-la* parents-poss *ˈru* say.prs *ba* cl *ˈni* indeed 'Like when they say our parents don't think.' 'Como cuando dicen que no piensan nuestros padres.' {SFH

tx475:08:17.7}

The use of the possessive suffix appears to be optional when the pronominal possessor is *ˈkini*, as exemplified above, where the possessive suffix is omitted in (20a) but attested in the same lexical item in (20b). As shown in (21), both a nominative-marked pronoun and a possessive pronoun can be employed in pronominal possessive constructions involving kinship terms.

	- b. *niˈhê apaˈlótʃ͡ala niˈhê* 1sg.nom *apaˈlótʃ͡a-la* maternal.grandfather-poss 'my maternal grandfather' 'mi abuelo materno' {LEL tx109:00:46.8}

While the possessive suffix is omitted with the possessive pronominal in (21a), it is attested, and in fact, required, with the nominative-case marked pronominal possessor in (21b).

In other Rarámuri varieties, a full set of dedicated possessive pronouns for kinship terms are documented: in Norogachi Rarámuri, Brambila (1976) reports

### 12.2 Complex noun phrases: Possessive constructions

possessive pronouns that contrast both number and person for speech act participants (*kene* '1sg.poss', *kemu* '2sg.poss', *keti* '1pl.poss', and *ketumu* '2pl.poss'), with a single form encoding third person possessors (*kepu* '3poss') (1976: 246– 248). In contrast, only the first person possessor is encoded with a dedicated pronominal form in Choguita Rarámuri.

### **12.2.3 Appositive possessive constructions**

In addition to the head-marking strategy for marking possession, Choguita Rarámuri has an alternative way of expressing a possession relationship through an appositional construction with *ˈníwa*, which may be used predicatively as a verb meaning 'to have', but which behaves as a grammatically specialized possessive noun in possessive constructions *ˈníwa* (glossed below as 'have'). The root *ˈníwa* as a possessive noun is marked with the possessive *-lâ* suffix and is part of a possessive phrase. These constructions are exemplified in (22):

	- a. *niˈhê ˈníala ˈét͡ʃi koˈbísi neˈhê* 1sg.nom *níwa-lâ* have-poss *ˈét͡ʃi* dem *koˈbísi* pinole 'It is mine that pinole.'

'Es mío ese pinole.' < FMF 09 3:32/el >

b. *ˈkúmi buˈʔí ˈnè ˈníala wiˈt͡ʃá ˈkúmi* where *bu'ʔi* lie.sg *ˈnè* 1sg.nom *ˈníwa-lâ* have-poss *wit͡ʃá* needle 'Where is my needle?'

'¿Dónde está mi aguja?' < BFL 11/07/09/el >

c. *nà ko ˈnè ˈníala ˈlîbro ko na=ko* prox=emph *ˈnè* 1sg.nom *ˈníwa-lâ* have-poss *ˈlîbro* book *ko* emph 'This here is my book.' 'Este es mi libro.' < BFL 06 4:187-189/el >

The nouns in these constructions with the possessive noun *ˈníwa* have not been documented with the head-marking posessive strategy, which suggests this constitutes a class of non-possessible nouns in Choguita Rarámuri. Further evidence of the difference between the head-marking possessive vs. possessive noun strategy is shown in the following example, where the possessed noun *saʔˈpá*

### 12 Noun phrases

'flesh' appears with head-marking when inalienably possessed (23a), but appears with the possessive noun construction in the alienable possession reading (23b) (where 'meat' refers to meat severed from a cow).

(23) Inalienable vs. alienable possession

```
a. niˈhê saʔˈpála
    neˈhê
    1sg.nom
              saʔˈpá-lâ
              meat-poss
    'my flesh'
    'mi carne (de mi cuerpo)' < FMF 09 3:32/el >
b. niˈhê ˈníala saʔˈpá
    neˈhê
    1sg.nom
              ˈníwa-lâ
              have-poss
                         saʔˈpá
                         meat
    'my meat (to eat)'
    'mi carne (para comer)' < FMF 09 3:32/el >
```
These constructions resemble minimal appositive possessive constructions as documented in Ainu (Bugaeva et al. 2021; Johanna Nichols, p.c.).

# **13 Basic clause types**

This chapter describes several specialized basic clause types in Choguita Rarámuri in terms of their internal structure and the type of predicate heading each clause type. A first distinction is made between clauses headed by verbal predicates and those headed by nonverbal predicates. Clauses headed by verbal predicates are described in terms of their argument structure, including the distinction between intransitive and transitive clauses. Non-verbal clauses are described in terms of whether they are headed by nominal or locative predicates. The basic Choguita Rarámuri clause consists of a predicate and the arguments it subcategorizes for. Core arguments take the form of noun phrases, whether nouns (with or without modifiers), or personal pronouns (see Chapter 12). From these, only pronominal forms encode core grammatical relations (i.e., there is no case marking in nominals). Given that noun phrases can be elided, a minimal clause in Choguita Rarámuri consists of an inflected verbal predicate. The classification of clauses into intransitive, transitive and ditransitive is thus based on their potential syntactic configuration, even if core arguments are not overtly expressed in the clause.

This chapter is organized according to both structural and functional characteristics of clause types. Clauses headed by verbal predicates are described in §13.1. The morphosyntactic properties of core grammatical relations in basic verbal clauses is discussed in §13.1.1. The basic properties of intransitive clauses are described in §13.1.2, those of monotransitive clauses are addressed in §13.1.3, while ditransitive clauses are described in §13.1.4. Clauses headed by non-verbal predicates are addressed in §13.2. A brief description of different types of copulas is given in §13.2.1; copular clauses involving nominal predicates are described in §13.2.2; clauses headed by locative predicates are described in §13.2.3. The chapter concludes in §13.2.4 with a description of existential clauses headed by predicates of possession.

### **13.1 Verbal clauses**

This section describes the properties of intransitive, monotransitive and ditransitive clauses. Before turning to each type of predicate, a description of the mor-

### 13 Basic clause types

phosyntactic criteria that allow identifying core grammatical relations in Choguita Rarámuri is provided first.

### **13.1.1 Basic clause types and transitivity properties**

Clauses are structurally distinguished by predicate type and valency. Determining a particular predicate's valency is, however, not trivial, since there is no case marking of core grammatical functions in nouns, and only pronominal forms exhibit a distinction for subject and object marking. Some morphosyntactic criteria that help identify core grammatical relations are summarized in (1):

	- a. First and second person subjects are encoded by subject-marked pronouns.
	- b. First and second person objects are encoded by object-marked pronouns.
	- c. The subject of a basic intransitive clause corresponds to the direct object of a derived transitive clause when the verb undergoes a causative (valence increasing) operation.
	- d. The new participant introduced through a causative (valence increasing) operation is encoded as the subject.
	- e. A valence decreasing operation (passive) will demote the subject, direct object or recipient argument (primary object in ditransitive clauses) to a peripheral status (the adjunct argument appears in a postpositional phrase or is completely omitted).
	- f. Benefactive participants in ditransitive clauses (basic or derived through applicativization) may be encoded by object pronouns.

Table 13.1 and Table 13.2 below summarize the subject/object distinctions of free pronominal forms and their enclitic counterparts (free subject and object pronouns are given in parenthesis). For the description of the morphological properties of pronominal forms, see Chapter 10.

The forms in (2) illustrate properties (1a–d): in the causative construction in (2c), the object corresponds to the subject of its basic, non-causative counterpart (2a) and its applicative, non-causative counterpart (2b) (property (1c)); the causer argument introduced through the valence increasing derivation is case marked as subject (property (1d)).

### 13.1 Verbal clauses

Table 13.1: Free personal pronouns


Table 13.2: Pronominal enclitic forms


(2) Morphosyntactic marking in a basic clause and related causative construction


### 13 Basic clause types

c. *Causative construction ˈmín ˈnè onoˈlâ riˈmêentima ˈmí=ni* 2sg.acc=1sg.nom *ˈnè* 1sg.nom *ono-ˈlâ* father-poss *riˈmê-ni-ti-ma* make.tortillas-appl-caus-fut.sg 'I will make you make tortillas for my dad.' 'Te voy a hacer que le hagas tortillas a mi papá.' < BFL 08 1:161/el >

Example (3) illustrates properties (1e–f): in the basic construction in (3a), the benefactive argument of an active ditransitive verb is encoded by an object pronoun (property (1f)); in the passive correlate of this basic construction, the agent argument of the ditransitive verb is demoted and omitted (property (1e)).

(3) a. *Basic construction ˈmò ˈjéla taˈmí haˈré gaˈjêta ˈàko ˈmò* 2sg.nom *ˈjé-la* mother-poss *taˈmí* 1sg.acc *haˈré* some *gaˈjêta* cookie *ˈà-ki-o* give-pst.ego-ep 'Your mom gave me some cookies.' 'Tu mamá me dió unas galletas.' < BFL 09 1:89/el > b. *Passive construction aˈjéna, ˈpé ˈtáa kan ˈhônsa ˈtʃ͡íri ˈàari aˈjéna* yes *ˈpé* little *ˈtá* small *ka=ni* cop.irr=1sg.nom *ˈhônsa* since *ˈétʃ͡i* dem *ˈjíri* kind *ˈà-ru* give-pst.pass 'Yes, since I was very small I was given that.' 'Sí, desde chiquito me dieron eso.' < FLP 06 in61(441)/in >

The next subsections review the basic properties of intransitive (§13.1.2), transitive (§13.1.3) and ditransitive (§13.1.4) clauses.

### **13.1.2 Intransitive clauses**

Choguita Rarámuri intransitive verbs are monovalent verbs that subcategorize for a single, nominative case argument. As mentioned above, morphological case marking of core grammatical relations is only evident in pronominal forms, with dedicated subject and object forms. The following examples illustrate intransitive clauses with pronominal arguments (4a–c) and noun phrase subject arguments (4d–e). Subject arguments are indicated in square brackets.

13.1 Verbal clauses

### (4) Intransitive clauses


'And now today we don't believe anymore.'

'Y ahora ya no creemos.' {GFM tx905:1:47.1}

c. *aʔˈlì ˈhípi ko biˈlátimi ke biˈlé ˈpé ˈtâʃi tʃ͡iriˈká ˈéni ˈémi ko ba aˈní bí aʔˈlì* and *ˈhípi=ko* now=emph *beˈlá=[timi]subj* really=2pl.nom *ke* neg *biˈlé* one *ˈpé* just *ˈtâsi* neg *ˈétʃ͡i* dem *riˈká* like *ˈéna* go.pl *ˈémi=ko* 2pl.nom=emph *ba* cl *aˈní* say.prs *bi* emph

'And now you all do not live (go around) like that, you all.'

'Y ahora ustedes no andan así, ustedes.' {SFH tx12:3:42.7}

d. *tʃ͡aˈbè kiˈʔà ná biˈlé koriˈmá naˈwà-li biˈlé-na bitiˈtʃ͡í*

*tʃ͡aˈbè* before *kiˈʔà* long.ago *ná* there *[biˈlé* one *koriˈmá]subj* fire.bird *naˈwà-li* arrive-pst *biˈlé-na* one-incl *bitiˈtʃ͡í*

house

'Long time ago, one fire bird (*korima*) arrived there at one house.' 'Hace mucho tiempo llegó un pájaro de fuego (*korimá*) en una casa.' {LEL tx5:0:26.4}

e. *we tʃ͡aˈbèe kiˈʔà ko ne biˈlá raˈsíra naˈtêami ˈníli ˈétʃ͡i ˈnàri raˈʔìtʃ͡iri aˈlé we* int *tʃ͡aˈbè* before *kiˈʔà=ko* long=emph *ne* int *beˈlá* indeed *raˈsíra* more *naˈtê-ame* have.value-ptcp *ˈní-li* cop-pst *[étʃi* dem *ˈnàri* this *raˈʔìtʃa-li]subj* speak-nmlz

'Long time ago, this talk was very important.'

'Antes era más valiosa la plática.' {JMF tx817:0:00.0}

### 13 Basic clause types

Intransitive clauses may be headed by verbs that have been derived through a valence-decreasing operation, such as the passive construction exemplified in (5).

(5) *ˈtòoru grabaˈdôra ˈtò-ru* take-pst.pass *[grabaˈdôra]subj* recorder 'The recorder was taken.' 'Se llevaron la grabadora.' < SFH 08 1:45/el >

Intransitive clauses tend to display a verb final order, but, as shown in (5), there are no strict restrictions about order of verb and noun phrase in these clauses.

### **13.1.3 Transitive clauses**

Predicates heading transitive clauses subcategorize for a subject noun phrase (marked as subject if pronominal) and a single object noun phrase (marked as object if pronominal). The forms in (6) illustrate transitive clauses where the single object noun phrase is indicated in square brackets.

	- a. *baʔaˈrîni ˈmí ˈáma baʔaˈrî=ni* tomorrow=1sg.nom *[ˈmí]obj* 2sg.acc *ˈá-ma* look.for-fut.sg 'I'll look for you tomorrow.' 'Mañana te busco.' < LEL 09 1:70/el >
	- b. *aʔˈlì tʃ͡iˈhônsa ˈétʃ͡i riˈhò baˈhîsa ˈká ˈétʃ͡i tʃ͡oʔˈmá aʔˈlì* and *tʃ͡iˈhônsa* then *ˈétʃ͡i* dem *riˈhòi* man *baˈhî-sa* drink-cond *ˈká* irr *[ˈétʃi* dem *tʃoʔˈmá]obj* snot 'and if the man would have drank the snot'

'y si el hombre hubiera tomado el moco' < SFH 06 choma(21)/tx >

c. *matʃ͡iˈbô biˈláti ˈlá tamuˈhê na ˈétʃ͡i muˈhê ˈnâtara ˈnápumi riˈká ˈhônsa oˈtʃérili muˈhê*

*matʃ͡i-ˈbô* know-fut.pl *beˈlá=ti* really=1pl.nom *oˈlá* cer *tamuˈhê* 1pl.nom *na* dem *[ˈétʃi* dem *muˈhê* 2sg.nom *ˈnâta-ra* think-poss *ˈnápi=mi* sub=2sg.nom *reˈká* like *ˈhônsa* since *oˈtʃéri-li* grow-pst *muˈhê]obj* 2sg.nom 'We will know your thoughts of the time when you were growing up.'

13.1 Verbal clauses

'Nosotros vamos a saber tus pensamientos desde que tu creciste.' {SFH in61:0:22.1}


'Y cuando está sembrado el maíz era llevado, dicen.' {LEL tx32:1:34.7}

Transitive clauses may be derived through valence increasing operations, such as causative, transitive and applicative, applying to intransitive predicates, as exemplified in (2) above. Transitive predicates may also take clausal complements (for a discussion of the syntax of complementation cross-linguistically, see Cristofaro 2005 and Noonan 2007). Examples of complement-taking transitive verbs are given in (7) (in these examples, complement-taking verbs are in boldface and complement clauses in square brackets).

(7) Complement-taking transitive verbs

a. *aʔˈlì a ˈjá saˈjèli naˈwàatʃ͡i ˈétʃ͡i koriˈmá paˈtʃ͡á bitiˈtʃ͡í baˈkíli aʔˈlì* and *a* aff *ˈjá* soon *saˈjè-li* feel-pst *[naˈwà-a-tʃ͡i* arrive-prog-loc *ˈétʃ͡i* dem *koriˈmá* fire.bird *paˈtʃ͡á* inside *bitiˈtʃ͡í* house *baˈkí-li]* go.in.sg-pst

'And then they felt when the fire bird arrived and when he went inside the house.'

'Y luego sintieron cuando llegó el *pájaro de fuego* y cuando entró adentro a la casa.' {LEL tx5:00:42.8}

b. *ˈnè ko ˈtʃ͡é ˈémi raˈpâko ku ʃiˈmêo maˈjê*

### 13 Basic clause types

*ˈnè=ko* 1sg.nom=emph *[ˈtʃ͡é* again *ˈémi* 2pl.acc *raˈpâko* yesterday *ku* rev *si-ˈmêa]* go.sg-fut.sg *maˈjê* believe 'I thought you guys would come back yesterday.' 'Yo creía que iban a llegar ayer.' < BFL 09 1:57/el >

Complement clauses in Choguita Rarámuri are described in §15.1.

### **13.1.4 Ditransitive clauses**

Ditransitive clauses in Choguita Rarámuri subcategorize for three core arguments, one subject noun phrase (case-marked as nominative if pronominal), and two objects, a recipient-like ("R") argument and a theme-like ("T") argument (with Theme defined as "something which undergoes a change in location or to which a location is attributed" (Dryer 2007a)). As the following examples show, recipient (R) arguments are marked as objects, patterning together with single object arguments of monotransitive verbs: in (8a–b), the recipient argument is encoded through the object pronoun (*taˈmí* '1sg.acc'). In contrast, the theme (T) argument, if pronominal, is encoded with the subject pronoun (*ˈémi* '2pl.nom').<sup>1</sup>

	- a. *riwiˈrítʃ͡i biˈtêami tʃ͡e taˈmí ˈémi ˈàki [riwiˈrítʃ͡i* Tewirichi *biˈte-ame]subj* inhabit.sg-ptcp *tʃ͡e* tʃ͡e *[taˈmí]<sup>R</sup>* 1sg.acc *[ˈémi]<sup>T</sup>* 2pl.nom *ˈà-ki* give-pst.ego 'People from Tewirichi gave you guys to me (for adoption).' 'Unos de Tewirichi me los dieron a ustedes (en adopción).' < BFL 09 1:89/el >
	- b. *ˈmò ˈjêla taˈmí haˈré gaˈjêta ˈàko [ˈmò* 2sg.nom *ˈjê-la]subj* mother-poss *[taˈmí]<sup>R</sup>* 1sg.acc *[haˈré* some *gaˈjêta]<sup>T</sup>* cookie *ˈà-k-o* give-pst.ego-ep 'Your mom gave me some cookies.' 'Tu mamá me dió unas galletas.' < BFL 09 1:89/el >

This pattern, where recipient arguments of ditransitive predicates and objects of monotransitive predicates pattern together as being object-marked, suggests

<sup>1</sup>This contrasts with the case marking pattern found in closely-related Yaqui (Taracahitan), where some ditransitive verbs license accusative marking on both a theme and beneficiary arguments (Guerrero Valenzuela & Van Valin 2004).

13.1 Verbal clauses

that Choguita Rarámuri is a primary object language, as argued for other Uto-Aztecan languages, e.g., Huichol (Corachol; Comrie 1982) (see also Dryer 1986 and Dryer 2007a). Other Uto-Aztecan languages that have been described as primary object languages include Cora (Corachol; Vázquez Soto 2002), Southeastern Tepehuan (Tepiman; García Salido 2007a), and Yaqui (Taracahitan; Felix Armendáriz 2000, Guerrero Valenzuela & Van Valin 2004).

Ditransitive clauses with three non-pronominal core argument noun phrases are very infrequently attested in the Choguita Rarámuri corpus. Documented ditransitive clauses with non-pronominal noun phrases suggest that the order of arguments is S-T(heme)-R(ecipient), with a high degree of flexibility regarding the placement of object noun phrases with respect to the verbal predicate, as exemplified in (9).

(9) Order of nominal noun phrases in ditransitive clauses


*kumuˈtê ˈlâmina ˈèbili raˈlàmuli [kumuˈtê]subj* sheriff *[ˈlâmina]<sup>T</sup>* tin.roof *ˈèbi-li* bring.appl-pst *[raˈlàmuli]<sup>R</sup>* people 'The sheriff brought tin roof for the people.' 'El comisariado les trajo lámina a la gente.' < SFH 09 3:51/el >

```
c. S-V-T-R
```
*ˈétʃ͡i muˈkî ko ka tʃ͡e ˈàli liˈmôsna ta roˈwéami ba [ˈétʃ͡i* dem *muˈkî]subj=ko* woman=emph *ka* neg *tʃ͡e* neg *ˈà-li* give-pst *[liˈmôsna]<sup>T</sup>* present *[ta* det.sg *roˈwé-ame]<sup>R</sup>* run.womens.race-ptcp *ba* cl 'Those women did not give a present to the (winning) runner.'

'Esas mujeres no le dieron limosna a la corredora.' < BFL 09 1:89/el >

d. V-T-R *ˈpé aʔˈlá amiˈnábi ˈàa busuˈrêro ˈtʃ͡ó ˈlé* [*ˈkîni ˈkútʃ͡uwa*] *ˈlîna ˈnòtʃ͡a koˈʔá ˈtʃ͡ó*

### 13 Basic clause types

*ˈpé* just *aʔˈlá* well *amiˈnábi* more *ˈà-a* give-prog *[busuˈrê-ro]<sup>T</sup>* wake.up-nmlz *ˈtʃ͡ó* also *aˈlé* dub *[ˈkîni* our *ˈkútʃ͡uwa]<sup>R</sup>* children *ˈlîna* so.that *ˈnòtʃ͡a* work.prs *koˈʔá* eat.prs *ˈtʃ͡ó* also 'But we must give our children more advice so that they will eat (sustain themselves) from their work.' 'Pero hay que darle más consejo a nuestros hijos, para que trabajen y coman (de su trabajo).' < SFH 06 in61(713)/in > e. S-V-T-R *ˈmí muˈkî ˈwé naˈtêami ˈàli aˈsûkar ˈkûruwi [ˈmí* dist *muˈkî]subj* woman *ˈwé* int *naˈtê-ame* cost-ptcp *ˈà-li* give-pst *[aˈsûkar]<sup>T</sup>* sugar *[ˈkûruwi]R/* children 'That woman gave (sold) the sugar to the children for very expensive.' 'Esa mujer les dió (vendió) azúcar muy cara a los niños.' < SFH 09 3:52/el >

As shown in these examples, the theme argument may appear pre-verbally (9a–b), or both theme and recipient may appear post-verbally (9c–e). On the other hand, in clauses with both pronominal and nominal noun phrases, pronominal arguments tend to be expressed pre-verbally, while non-pronominal arguments tend to be expressed post-verbally: in (9a–b, e) above, both theme and recipient noun phrases are non-pronominal and post-verbal, while in (10) below, pronominal objects are placed pre-verbally (the recipient argument in (10a–b), and the theme argument in (10c).

(10) Order of nominal and pronominal elements in ditransitive clauses

```
a. S-R-V-T
```
*ˈétʃ͡i muˈkî ko ka tʃ͡e taˈmí àa liˈmôsna ba [ˈétʃ͡i* dem *muˈkî]subj=ko* woman=emph *ka* neg *tʃ͡e* neg *[taˈmí]<sup>R</sup>* 1sg.acc *à-a* give-prog *[liˈmôsna]<sup>T</sup>* present *ba* cl 'Those women didn't give me a present.' 'Esas mujeres no me dieron limosna.' < BFL 09 1:89/el > b. S-R-V-T *ˈmí muˈkî ˈwé naˈtêami taˈmí raʔliˈkí aˈsûkar [ˈmí* dist *muˈkî]<sup>S</sup>* woman *ˈwé* int *naˈtê-ame* cost-ptcp *[taˈmí]<sup>R</sup>* 1sg.acc *raʔliˈkí* sell.pst *[aˈsûkar]<sup>T</sup>* sugar 'That woman sold me the sugar at a very high price.' 'Esa mujer me vendió muy cara el azúcar.' < SFH 09 3:51/el >

13.2 Locative, copula and existential clauses

c. S-T-V-R *niˈhê ˈmí ˈâma miˈgêl [neˈhê]<sup>S</sup>* 1sg.nom *[ˈmí]<sup>T</sup>* 2sg.acc *ˈâ-ma* give-fut.sg *[miˈgêl]<sup>R</sup>* Miguel 'I will give you to Miguel (so he can be your godfather).' 'Te voy a dar a Miguel (para que sea tu padrino).' < SFH 09 3:51/el >

Guerrero Valenzuela & Van Valin (2004) propose that, while predominantly a primary/secondary object language, Yaqui also has ditransitive predicates with a direct/indirect object pattern. Whether Choguita Rarámuri also has ditransitive clauses with a direct/indirect object pattern is a question left for further research.<sup>2</sup>

### **13.2 Locative, copula and existential clauses**

Cross-linguistically, clauses with nonverbal predicates whose properties exhibit variation across languages include those involving nominal predicates, adjectival predicates and locative predicates (Dryer 2007a). Given that the class of basic, underived adjectives in Choguita Rarámuri is very small (see Chapter 10, §10.3), this section focuses on clauses headed by nominal and locative predicates. This section also addresses existential clauses headed by predicates of possession. Before turning to these clauses, a brief description of types of copulas in Choguita Rarámuri is provided next.

### **13.2.1 Types of copulas**

Copular clauses are a clause type in which the contentful predicate is not a verb, but some other category like an adjective, a noun or a preposition. Copular clauses in Choguita Rarámuri have an internal structure that includes a copula verb, a subject nominal phrase and an unmarked complement. There are four suppletive stems of copula verbs corresponding to different tense/aspect distinctions. These are: (i) the copula *hu* for present tense; (ii) the copula *ke* for past imperfective; (iii) the copula *ka* for irrealis; and (iv) the copula *ni-* for all other tense/aspect distinctions. Specifically, the present, past imperfective and irrealis copula verbs do not attach any tense/aspect or mood suffixes, but the copula *ni-* is a bound root that requires suffixation. Table 13.3 summarizes the set of Choguita Rarámuri suppletive copular verbs, their gloss and approximate translation.

<sup>2</sup>As discussed in Guerrero Valenzuela & Van Valin (2004), Yaqui exhibits a direct/indirect object pattern given the existence in the language of two different case markers that allows this marking. An anonymous reviewer suggests that given the fact that Choguita Rarámuri lacks case marking in nouns, a similar pattern may not be possible.

### 13 Basic clause types



Examples of the bound root *ni-* copula with different TAM suffixes is provided in (11).

	- a. *ˈníli ˈní-li* cop-pst 'it was'
	- b. *ˈníra ˈní-ra* cop-rep 'they say it was'
	- c. *ˈníma ˈní-ma* cop-fut.sg '(it/she/he) will be'
	- d. *ˈníbo ˈní-bo* cop-fut.pl '(they/we) will be'
	- e. *ˈnísa*
		- *ˈní-sa* cop-cond 'if it were'

The list of inflected forms of the copula verb *ni-* in (11) is not exhaustive, since there are no restrictions as to which TAM markers can attach to this verb. Examples of these copula verbs in context are provided in the sections below.

13.2 Locative, copula and existential clauses

### **13.2.2 Clauses headed by nominal predicates**

In copular clauses headed by nominal predicates, the characteristic of the subject is a noun phrase. The examples in (12) illustrate copular clauses headed by nominal predicates.

(12) Copular clauses headed by nominal predicates

```
a. enferˈmêra ˈkéʔe?
    enferˈmêra
    nurse
                ˈke
                cop.impf
    'Did she use to be a nurse?'
    '¿Era enfermera?' < BFL 09 1:85/el >
```
	- dem prox plant=emph chuchupate cop.prs 'This plant is chuchupate.'
		- 'Esta planta es chuchupate.' < GMF 09 3:63/el >

Copula verbs may precede the complement in copular clauses, as exemplified in (13), but examples of this kind are marginally attested in the corpus (the only attested examples showing this pattern were obtained through translation elicitation using Spanish, but are so far unattested in natural discourse).

(13) a. *ˈétʃ͡i ˈgîltro ˈhú komiˈsârio ˈétʃ͡i* dem *ˈgîltro* giltro *ˈhú* cop.prs *komiˈsârio* sheriff 'Giltro is sheriff.' 'Giltro es comisario.' < FMF 09 3:57/el >

### 13 Basic clause types

b. *ˈétʃ͡i tiˈwé ˈníma ˈlé sukuˈrûame ˈétʃ͡i* dem *tiˈwé* girl *ˈní-ma* cop-fut.sg *aˈlé* dub *sukuˈrû-ame* shaman-ptcp 'That girl will be a shaman.' 'Esa niña va a ser curandera.' < FMF 09 3:57/el >

As described in Chapter 10 (§10.3) and Chapter 15 (§15.3), property concepts in Choguita Rarámuri are encoded through verbal predicates, which may be nominalized through a participial morpheme. Copular clauses headed by these nominalized forms, exemplified in (14), exhibit the same internal structure as other copular clauses headed by nominal predicates.

### (14) Copular clauses with nominalized predicates

a. *miˈná biˈlátimi ˈmá oˈtʃérami ˈníbo ˈlé amiˈná* later *beˈlá=timi* really=2pl.nom *ˈmá* already *oˈtʃéra-ame* grow-ptcp *ˈní-bo* cop-fut.pl *aˈlé* dub 'Later on when you all will be adults.' 'Después ya van a ser crecidos ustedes.' {SFH tx12:12:36.7} b. *aʔˈlì tamuˈhê biˈlá ko naˈlìna ˈwé winoˈmîwam ˈníbira ˈrái ba aʔˈlì* then *tamuˈhê* 1pl.nom *beˈlá=ko* really=emph *naˈlìna* then *ˈwé* int *winoˈmî-w-ame* money-have-ptcp *ˈní-bi-la* cop-irr.pl-rep *ru-ˈwá* say-mpass *ba* cl 'Then we would have been the wealthy ones.' 'Entonces nosotros hubiéramos sido los del dinero.' < SFH 06 choma(25)/tx > c. *á biˈlá ˈwé ˈkáira ro ˈwé baˈhîbam ˈká á* aff *beˈlá* really *ˈwé* int *ˈkáira* happy *ro* ro *ˈwé* int *baˈhî-wa-ame* drink-mpass-ptcp *ˈká* cop.irr 'It feels really good when one is drinking.' 'Se siente uno muy agusto cuando anda tomando.' < SFH 06 in61(407)/tx > d. *we tʃ͡aˈbèe kiˈʔà ko ne biˈlá raˈsíra naˈtêami ˈníli étʃ͡i ˈnàri raˈʔìtʃ͡ili aˈlé we* int *tʃ͡aˈbè* before *kiˈʔà=ko* before=emph *ne* int *beˈlá* really *raˈsíra* more *naˈtê-ame* cost-ptcp *ˈní-li* cop-pst *ˈétʃ͡i* dem *ˈnàri* this *raˈʔìtʃ͡a-li* speak-pst *aˈlé* dub 'Long ago these words were more valuable.' 'Antes era más valiosa la plática.' {JMF tx817:0:00.0}

### 13.2 Locative, copula and existential clauses

While most property concepts are encoded through verbal predicates, there is a small set of true adjectives. Copular clauses headed by these adjectival predicates have the same internal structure than copular clauses headed by nominal predicates. This is shown in (15).

	- a. *ˈétʃ͡i tiˈwé ˈwé seˈmáti ˈhú ˈétʃ͡i* dem *tiˈwé* girl *ˈwé* int *seˈmáti* pretty *ˈhú* cop.prs 'This girl is very pretty.' 'Esta niña está muy bonita'.
	- b. *ˈkíti ˈwé ˈrîko ˈhú haˈré ba ˈkíti* because *ˈwé* int *ˈrîko* wealthy *ˈhú* cop.prs *haˈré* some *ba* cl 'And that is why some are really wealthy.' 'Y por eso unos son muy ricos.' < SFH 06 choma(30)/tx >

Copular clauses with nominal predicates may express a partitive semantic relation ('made of'), a relation that is sometimes cross-linguistically encoded with genitive case. This is shown in (16).

### (16) Copular clauses with partitive meaning


The next section addresses a second type of nonverbal clauses, namely those headed by locative predicates.

### 13 Basic clause types

### **13.2.3 Clauses headed by locative predicates**

Clauses headed by locative predicates in Choguita Rarámuri deploy one of a set of positional/postural verbs. Positional/postural verbs in this language denote positions of entities (both animate and inanimate) dependent of the physical properties of the referent in terms of their inherent or transitional shape, though the use of postural predicates also reveals a grammatical system of noun classification (see Ameka & Levinson 2007 for a discussion of the semantic and grammatical properties of postural verbs cross-linguistically).

In their basic (morphologically unmarked form), posturals in Choguita Rarámuri are stative, intransitive verbs that denote a particular posture or position of their single argument (which is encoded as the subject). These verbs can be derived to function as change of state predicates, which may be either intransitive or transitive. When intransitive, the change of state into a particular posture is brought about by the entity undergoing the change itself (encoded as the single argument subcategorized for by the verb), and when transitive, the change of posture of the object argument is brought about by a second participant, an agent-like argument encoded by the subject. Derivation from a stative postural verb to a change-of-state predicate generally involves an intransitive/transitive morphological alternation: a stem final *-i* suffix for intransitives and a stem final *-a* suffix for transitives, as exemplified in (17) (see §9.1.5 for more details on valence changing processes in Choguita Rarámuri).<sup>3</sup>

(17) a. *wiˈlí*


stand-tr 'S/he stands someone up.' 'Lo para.'

The stative postural predicates and their corresponding intransitive and transitive change-of-state counterparts may also be encoded through suppletive forms. These are exemplified in (18).

<sup>3</sup>Cognate suffixes to the Choguita Rarámuri intransitive *-i* and transitive *-a* suffixes are widely attested across the Uto-Aztecan language family, including in the Aztecan branch, in Cahitan, Southern Paiute, as well as in Cupan language varieties (Heath 1977: 32).

### 13.2 Locative, copula and existential clauses

	- a. *aˈtí* to.be.sitting.sg 'S/he is sitting.' 'Está sentada.'

b. *aˈsí aˈsí* to.sit.sg.intr 'S/he sits.' 'Se sienta.' c. *aˈtʃ͡â aˈtʃ͡â* to.sit.sg.tr

'S/he sits someone down.' 'Lo sienta (e.g., al niño).'

The contrast encoded through suppletion involves a state (18a), an intransitive, change of state (18b) and a transitive change of state (18c). The consonantal changes in the stem in this set is not predictable through any synchronically productive phonological processes, and are thus analyzed here as involving a lexical contrast.

The following examples illustrate the stative (19), intransitive change of state (20) and transitive change of state (21) uses of postural verbs (a full inventory of postural predicates in Choguita Rarámuri is provided in Table 13.4):

	- a. *ˈétʃ͡i ko tʃ͡oˈmí ˈtʃ͡ó muˈtʃûwi aʔˈlì ˈétʃ͡i=ko* dem=emph *tʃ͡oˈmí* there *ˈtʃ͡ó* also *muˈtʃûwi* sit.pl *aʔˈlì* and 'They were (sitting) also over there, and...' 'Ellos también estaban allá y...' < FLP 06 in61(302)/in > b. *aʔˈlì ˈwé biˈlá kiˈrì ke biˈlé saˈjèka tʃuˈkúli wiˈtʃ͡ôa ˈétʃ͡i ko*
		- *aʔˈlì* and *ˈwé* int *beˈlá* really *kiˈrì* still *ke* neg *biˈlé* one *saˈjè-ka* realize-ger *tʃuˈkú-li* bent.sg-pst *wiˈtʃ͡ô-a étʃ͡i=ko*

wash.clothes-prog dem=emph

'And then she was still without realizing anything, washing clothes, her.'

### 13 Basic clause types

'Y entonces estaba sin pensar nada, sin darse cuenta de nada, lavando, ella.' {LEL tx32:4:11.9}

c. *ˈpé ˈnábi tʃ͡e ˈtʃéti koˈlìi biˈtíam ku ˈpé* just *ˈna=bi* prox=just *tʃé* also *ˈtʃéti* def.pl.bad *koˈlì* around.corner *biˈtí-ame* be.lying.down.plptcp *ku* rev 'They say that those that are lying down over there around the corner...' 'Dicen que estos que están ahí por de aquel lado a la vuelta...' {FLP

in243:20:07.8}

### (20) Intransitive change of state postural verbs

a. *aʔˈlì ˈhê aˈnèli: "ˈwé saˈpù aˈʃíska, ˈpîri tʃ͡uˈkú naˈʔî?" aʔˈlì* and *ˈhê* dem *aˈnè-li* say-pst *ˈwé* int *saˈpù* hurry *aˈsí-si-ka* sit.up-mot-ger *ˈpîri* what *tʃ͡uˈkú* four.legs *naˈʔî* here 'And then she told him: "Hurry up getting up! What is that sitting here?'

'Y luego le dijo: "¡Levántate pronto! ¿Qué está aquí?"' {LEL tx5:0:52.7}

b. *ˈnápu riˈká ˈnà muˈrípi roˈkáata tʃuˈkúbuka biˈlá ko aʔˈwâli ba ˈnápi* sub *riˈká* like *ˈnà* then *muˈrípi* close *roˈkáta* cliff *tʃuˈkú-ba-ka* four.legs-inch-ger *beˈlá=ko* indeed=emph *aʔˈwâ-li* throw.pl-pst *ba* cl

'Because he (the deer) stood (on four legs) close to the cliff and then threw them.'

'Así como se paró (el venado, en cuatro patas) cerca de la orilla del barranco y los tiró.' {SFH tx152:8:49.1}

c. *wiˈlíbura ruˈwá ˈétʃ͡i ˈnà maˈtʃí riˈtériri wiˈlí-ba-la* stand-inch-rep *ru-ˈwá* say-mpass *ˈétʃ͡i* dem *ˈnà* dem *maˈtʃí* outside *reˈté-riri* rock-loc 'Se paró allá afuera en una piedra.' 'He stood there outside in a rock.' {LEL tx5:2:06.9}

13.2 Locative, copula and existential clauses

(21) Transitive change of state postural verbs

a. *ˈpé biˈlé riˈʰtê ˈá riʔˈpí tʃ͡oˈnà ko ˈnápo ˈnàti muˈtʃûwili ˈpé* just *biˈlé* one *riʰˈtê* stone *ˈá* aff *riʔˈpí* remain *ˈétʃ͡i* dem *ˈná=ko* prox=emph *ˈnápi* where *ˈnà=ti* dem=1pl.nom *muˈtʃûwi-li* sit.tr.pl-pst 'Just one stone remains there where we put them.' 'Queda una piedra ahí donde las pusimos.' {LEL tx19:2:35.0} b. *siˈné ˈkátʃ͡i ˈmáti ˈá ˈhâwamti*<sup>4</sup> *á ˈníbo siˈné* some *ˈkátʃ͡i* times *ˈmá=ti* already=1pl.nom *ˈá* aff *ˈhâwa-ma=ti* stand.tr.pl-fut.sg=1pl.nom *ˈá* aff *ˈní-bo* cop-fut.pl 'Where we live, perhaps we will some times be elected as authorities.' 'Donde vivimos a lo mejor en veces vamos a ser autoridades.' {SFH tx12:11:54.2} c. *ˈmá ˈbéni aʔˈlá maˈtʃ͡í ˈnè ko ˈmá ke oˈmèrili ˈtʃ͡ó ˈmêa ˈtʃ͡ó ˈnápi niˈhê wiˈláli ko ba ˈmá* then *ˈbé=ni* because=1sg.nom *aʔˈlá* well *maˈtʃ͡í* know *ˈnè=ko* 1sg.nom=emph *ˈmá* then *ke* neg *oˈmèra-li* be.able-pst *ˈtʃ͡ó* either *ˈmê-a* win-prog *ˈtʃ͡ó* either *ˈnápi* sub *neˈhê* 1sg.nom *wiˈlá-li=ko* stand.tr**-**pst=emph *ba* cl 'Because I know very well that she couldn't win, the one I chose (lit. the one I stood).' 'Porque ya se que no pudo ganar la que puse yo (lit. 'la que yo paré).' {LEL tx19:4:28.0}

Postural predicates may also have suppletive forms encoding singular-plural distinctions or these contrasts may be derived through the plural/pluractional construction, which may also involve stem consonant mutation plus prefixation (cf. §8.2). This is further exemplified in (22).

<sup>4</sup> In this case, the verb 'make stand' means 'appoint' (in Spanish, literally 'parar autoridades').

### 13 Basic clause types

(22) Plural/pluractional form of positional/postural verb


b. *u-ˈtʃ͡úwi* 'bent, curved, on four legs, pl'

Table 13.4 depicts the set of Choguita Rarámuri postural predicates, with transitivity and singular-plural/pluractional contrasts.


Table 13.4: Postural/positional predicates in Choguita Rarámuri

Clauses with stative postural predicates may also have the function of encoding the existence of the noun referent. This is shown in (23), where the postural predicate in this context can be interpreted as non-specific with respect to the specific posture of the referent (sitting, standing, etc.).

(23) Postural/positional predicates in existential clauses *noˈrátʃ͡i ˈben ko aˈtí tʃ͡ó ba noˈrátʃ͡i* Norogachi *ˈbe=ni=ko* be=1sg.nom=emph *aˈtí* sit.sg *ˈtʃ͡ó* emph *ba* cl 'I was in Norogachi.' 'Yo estaba en Norogachi.' < FLP 06 in61(264)/in >

The predicate in these cases specifies both the existence of the referent the position in which entities are configured, making reference to their semantic

### 13.2 Locative, copula and existential clauses

properties, or provides an indication of their grammatical classification. In terms of semantic categorization, the postural verb set in Choguita Rarámuri involves a partitioning of the "etic" space, defined as the "structure of the natural world … [an] objective description of the domain which makes maximal discriminations", which may be contrasted with an 'emic' space where languages group in abstract way these discriminations that is based on the semantic properties of the referents (Levinson & Wilkins 2006: 8). Nouns have a canonical posture depending on the physical properties (e.g., volume, shape, dimensions, etc.) of their referents, and these characteristics may determine the choice of a postural predicate: fourlegged (e.g., a dog, a truck) vs. two-legged (e.g., a person, a ladder) vs. a recipient with a base (e.g., a pot) vs. a cylindric and thin object (e.g., a stick, a pencil), etc. Choice of postural predicates may also reflect circumstantial or transitional physical characteristics of the referent at the time of utterance (e.g., a person washing clothes or grinding corn in a *metate* is *tʃ͡uˈkú* 'bent/curved'.) The semantic categorization properties of postural verb is exemplified in (24) (Table 13.5 below provides examples of the classification system dependent on the postural verb nouns combine with in locative constructions.).

### (24) Positional/postural predicates: semantic categorization

a. *aʔˈlì ˈhê aˈnèli: "ˈwé saˈpù aˈʃíska! ˈpîri tʃuˈkú naˈʔî?" aʔˈlì* and *ˈhê* dem *aˈn-è-li* say-appl-pst *ˈwé* int *saˈpù* hurry *aˈsí-si-ka* sit.up-mot-ger *ˈpîri* what *tʃuˈkú* be.curved *naˈʔî* here

'And then she told him: "hurry up getting up, what is that standing (in four legs) here?'

'Y luego le dijo: "¡Levántate pronto! ¿Qué está aquí?"' {LEL tx5:0:52.7}

b. *ˈí oˈnátʃ͡i tʃukuˈbáma ˈlé toˈwí*

*ˈí* here *oˈnátʃ͡i* here *tʃuku-ˈbá-ma* be.curved-inch-fut.sg *aˈlé* dub *toˈwí* boy 'The boy will bend here.' 'Aquí se va a agachar el niño.' < BFL 11/03/09/el > c. *ˈétʃ͡i tʃ͡oʔˈmá biˈlá maˈníira ˈráa iˈbíli beˈtôli ˈétʃ͡i* dem *tʃ͡oʔˈmá* snot *beˈlá* really *maˈní-ri-la* sit.container-caus-rep *ru-ˈwá* say-mpass *i-ˈbíli* pl-one.pl *beˈtôli* plates 'But it is said that they were given snot to drink in some plates.'

### 13 Basic clause types

'Pero dicen que les dieron moco en unos platos.' < SFH 06 choma(6)/tx >

d. *ˈí aˈmáni ˈbôte*

*ˈí* here *aˈmáni* pl-sit.container *ˈbôte* cans

'Here are the cans.'

'Aquí están los botes.' < BFL 11/03/09/el >

e. *ˈpé ˈnàbi tʃ͡é t͡ʃéti koˈlì biˈtíam ku ˈpé* just *ˈnà=bi* prox=just *tʃé* also *tʃ͡éti* det.pl *ko'lì* around.there *biˈtí-ame* lie.down.pl-ptcp *ku* rev 'These ones (the dead people) who are lying down over there, also.' 'Estos que están (acostados) ahí por de aquel lado (los muertos).' < FLP 07 in243(511)/in >

In (24a), the noun referent about which the location is being predicated is in sight and stands in four legs, and in (24b) the referent is a child about to crawl, requiring the use of the predicate *tʃ͡uˈkú* 'to be curved'. The referents in (24c–d) involve containers with bases, for which the predicate *mani* 'to sit (a container)' is required. In (24c), the referent requires the use of the predicate *biti* 'to lie down, plural',<sup>5</sup>

While the categorization of postural predicates closely reflects fine-grained distinctions about postures of referents, there are constructions where the assignment of a particular entity to a particular postural verb seems arbitrary, revealing their classificatory nature. Consider, for instance, example (25), where the existence of a ring in the hand or blood on the floor is described with the verb *tʃ͡uˈkú*, which typically describes the position of entities that are on four legs, bent or curved.

(25) Postural predicates as grammatical classificatory devices


<sup>5</sup>This is a euphemism used to refer to dead people in this narrative.

13.2 Locative, copula and existential clauses

*ˈétʃ͡i* dem *ˈnà* prox *ˈá* aff *tʃuˈkú* be.bent *ˈlá* blood *maˈtʃí* outside *ˈnápi=ko* sub=emph *ˈnà* prox *ˈhônsa* from *ˈtò-li* take-pst *ˈétʃ͡i* dem *aliˈwâ-la* soul-poss "There is the blood outside where he took the soul from." "Ahí se ve la sangre afuera de donde se llevó el alma." {LEL tx5:3:59.2}

Thus, locative constructions not only reflect semantic categorization but also a classification system that is grammatical, as it may be applied to abstract nouns or concrete nouns where their physical properties are not considered for their classification (e.g., space objects like sun, stars, moon, etc. are also assigned a postural predicate in existential clauses). An overview of postural predicates and the semantic categorization of nouns in other Uto-Aztecan languages (including Yaqui (Taracahitan), Mayo (Taracahitan), Guarijío (Taracahitan) and Lower Pima (Tepiman)) is provided in O'Meara & Guerrero Valenzuela (2015). Nouns in the language are grouped into different classes depending on the postural verb they combine with in locative constructions. Table 13.5 gives a sample of how nouns are classified depending on the postural verb they combine with in locative constructions.

In addition to exhibiting valency-related morphological changes, postural predicates may also attach the movement *-ro* suffix to encode that the event depicted by the predicate is a process that is incomple and may involve motion. This is exemplified in (26).

(26) Use of motion *-ro* suffix with postural/positional verbs

a. *ˈpeta naˈʔî ˈá ˈhârpo ˈrú*


### 13 Basic clause types

men and everyone, too, and all the women'<sup>6</sup> 'todo lo que estamos (parados) sabiendo (recordando), dar consejos a los hombres y a todos también y a todas las mujeres' {JMF tx816:00:10.3}

The cognate form of this motion suffix in Mountain Guarijío is the stressshifting suffix *-to/ro*, described in (Miller 1996: 165) as a verbal suffix that encodes

<sup>6</sup>The emphatic particle *so* is only attested in the speech of older speakers and often in the context of *nawesari*, ceremonial speeches. Younger speakers use the emphatic particle *ko* in the same contexts where *so is attested.*


Table 13.5: Noun classes and locative constructions

### 13.2 Locative, copula and existential clauses

motion with a "distributive" sense ("'ir a ... en más de un lugar' [o] 'en más de un tiempo'" ["'to go to more than one place' or 'in more than one occasion'"].<sup>7</sup> Further examples of this construction with Choguita Rarámuri postural predicates are exemplified in (27).

(27) Postural/positional predicates in motion constructions


Postural verbs often occur in complex predicate constructions in which a descriptive predicate qualifies the way in which an entity came to be in a particular position (complex predicate constructions are addressed in Chapter 15). In the examples below, the descriptive predicate (which precedes the postural/positional predicate) qualifies that the position (in this instance lying down) came about by fainting (28a), dying (28b) or sleeping (28c).

### (28) Postural verbs in complex predicate constructions

a. *ˈmá nataˈkêa buˈʔíli biˈʔà roˈkò ˈmá* already *nataˈkê-a* faint-prog *buˈʔí-li* lie.down.sg-pst *biˈʔà* early *roˈkò* night 'He was already lying down unconscious early in the morning.' 'Ya estaba desmayado en la madrugada.' {LEL tx5:04:03.7}

<sup>7</sup>This suffix contrasts with another "motion" suffix in Mountain Guarijío, *-si* (cognate of the Choguita Rarámuri associated motion *-si(mi)* suffix), which is described as bearing a more aspectual meaning.

### 13 Basic clause types


In negative locative clauses, a predicate exclusively found in negative polarity copular clauses (*iˈtê*), is found. The negative locative clause in (29a) is contrasted with its positive polarity locative counterpart in (29b).

	- a. *naˈʔî ke iˈtê ˈbôte naˈʔî* here *ke* neg *iˈtê* be.neg *ˈbôte* can 'The can is not here.' 'Aquí no está el bote.' < LEL 09 1:74/el > b. *naˈʔî maˈní ˈbôte naˈʔî* here *maˈní* sit.container *ˈbôte* can 'Here is (sits) the can.' 'Aquí está el bote.' < LEL 09 1:74/el >

Further examples of negative locative clauses with the predicate *iˈtê* in negative locative clauses are shown in (30).

### (30) Negative polarity predicate *iˈtê*: further examples

a. *amiˈná ke iˈtêrimi ˈlé wiˈkarâami amiˈná ko ba amiˈnábi* afterwards *ke* neg *iˈtê-ri-mi* be.neg-ri-irr.sg *aˈlé* dub *wikaˈrâ-ame* sing-ptcp *amiˈnábi=ko* afterwards=emph *ba* cl 'Afterwards there won't be any ritual singers.' 'Después ya no va a haber cantador.' {FLP in243:12:45.3} 13.2 Locative, copula and existential clauses

b. *ˈmá ˈkátʃ͡i iˈtêli biˈlé muˈkî aˈpâtʃ͡e ba amiˈnábi ko ˈá wiˈká riˈhòrimi tʃ͡ó aˈlé ke ˈnà ˈmá* already *ˈká* neg *tʃ͡è* neg *iˈtê-li* be.neg-pst *biˈlé* one *muˈkî* woman *aˈpâtʃ͡e* apache *ba* cl *amiˈnábi=ko* afterwards=emph *ˈá* aff *wiˈká* many *riˈhò-ri=mi* man-vblz=dem *tʃ͡ó* anymore *aˈlé* dub *ke* neg *ˈnà* then 'because there weren't any single Apache women, (otherwise) afterwards there would have been more (Apache) people' 'porque ya no había ninguna mujer apache, (si no) hubiera habido mucha gente (apache)' {LEL tx110:2:41.8}

Postural verbs may also be used in negative polarity constructions with other negative particles. Examples of positional predicates in negative constructions are provided in (31).

	- a. *ˈká ˈtʃ͡è ˈwêsi moˈtʃí bitiˈt͡ʃí ko ba ˈká* neg *ˈtʃ͡è* neg *ˈwêsi* nobody *moˈtʃí* sit.pl *bitiˈt͡ʃí=ko* house=emph *ba* cl 'Nobody stays (sits) at home.' 'No se está nadie en la casa.' < BFL 09 1:65/el > b. *ˈí oˈnátʃ͡i ˈká ˈtʃ͡è wiˈlí esˈkwêla ba ˈi* here *oˈnátʃ͡i* here *ˈká* neg *ˈtʃ͡è* neg *wiˈlí* stand.sg *esˈkwêla* school *ba* cl

'The school is not here.'

'Aquí no está la escuela.' < BFL 09 1:65/el >

Postural verbs in negative constructions may be used when the particular posture is in the scope of negation: in (32), for instance, it is predicated that nobody is sitting (i.e., but there might be people standing).

(32) Positional/postural predicates in the scope of negation *tʃ͡iˈnà ˈká ˈtʃ͡è ˈwêsi muˈtʃûi ba ˈétʃ͡i* dem *ˈnà* dem *ˈká* neg *ˈtʃ͡è* neg *ˈwêsi* nobody *muˈtʃûwi* sit.pl *ba* cl 'There is nobody sitting there.' 'No hay nadie sentado ahí.' < BFL 09 1:65/el >

### 13 Basic clause types

A relevant question to ask in any language with positional verbs is which verb may be used as a "default", as languages with these systems are documented to exhibit a default collocation, which may be subject to pragmatically-induced variation (Ameka & Levinson 2007). In the case of Choguita Rarámuri, a good candidate for a default postural copular verb in locative constructions for human beings and other referents is the verb *atí* 'sit, sg' (and plural suppletive verb *mutʃúwi*). In the following narrative, a woman is scared by a coyote sitting outside (33a); in the next clause, the verb *mutʃúwi* 'sit, pl' is used to describe how the scared woman, together with her friends, locks herself inside the house (33b); the postural verb in this case describes that the women are located inside of the locked house. The referents are not in sight and there is nothing inherent in their configuration that would preclude a description with, for instance, the verb *háwi* 'stand, pl'.

	- a. *aʔˈlì ˈwé maˈhâli ˈétʃ͡i muˈkî basaˈt͡ʃî riwiˈsâ aʔˈlì* and *ˈwé* int *maˈhâ-li* get.scared-pst *ˈétʃ͡i* dem *muˈkî* woman *basaˈt͡ʃî* coyote *riwi-ˈsâ* see-cond 'So she got very scared when she saw the coyote.' 'Entonces se asustó mucho la mujer cuando vió al coyote.' {LEL tx\_mawiya:1:02.6}
	- b. *we ˈérika muˈtʃîli patʃ͡á we* int *ˈéri-ka* close-ger *muˈtʃûwi-li* sit.pl-pst *paˈtʃá* inside 'She closed the door shut and they were inside.' 'Cerró fuerte la puerta y estuvieron adentro.' < LEL tx\_mawiya:1:07.5 >

Thus, the default collocation when referring to human beings is 'sit'. This is also shown in example (31) above, where the use of positional predicates in negative existential and negative locative clauses provides a test for default collocations. In contrast to human beings 'sitting' by default, the canonical position for buildings is to 'stand', as shown in (31b) above. This means that the default collocation involves presupposing, not asserting, the properties of the figure (Ameka & Levinson 2007: 859).

Finally, there are recorded examples where postural verbs may also have grammaticalized functions as aspectual markers, a common development cross-linguis-

### 13.2 Locative, copula and existential clauses

tically (Ameka & Levinson 2007).<sup>8</sup> The following examples show a contrast between a construction where the postural predicate *tʃ͡uˈkú*, 'be on four legs/bent/ curved', has a habitual aspect meaning (34a) and a construction where the postural predicate *aˈtí*, 'sit', has a non-habitual reading (34b) (There is no evidence in the Choguita Rarámuri corpus that other postural predicates have grammaticized as aspectual markers.).

(34) Positional predicates with aspectual marking function


b. *ˈétʃ͡i ˈwé ˈnâtili ˈàa aˈtí ˈétʃ͡i* dem *ˈwé* int *ˈnâta-li* think-nmlz *ˈà-a* give-prog *aˈtí* sit.prs 'He is giving advice (for only a while).' 'Dando consejos (sólo un rato).' < FMF 09 3:57/el >

In sum, the system of positional predicates in Choguita Rarámuri may be characterized as featuring properties characteristic of languages with small sets (3–5) of contrastive locative verbs, which contrast with languages with large-set positional languages (which may feature between 9 to up to several hundred predicates). The properties of small-set positional languages are described in (Ameka & Levinson 2007: 858–9), some of which are summarized in (35).

	- There is a central role of human posture verbs (e.g., 'sit', 'stand', 'lie down', etc.).
	- There is a semantic categorization component, with reference to the spatial configuration of the referent (abstract axial and/or geometric).
	- The default collocation use of positionals involves presupposing (vs. asserting) the properties of the figure.

<sup>8</sup> "It is cross-linguistically common for postural verbs such as 'sit', 'stand', and 'lie' to have grammaticalised functions (e.g. as copula verbs, aspect markers, etc.). The regularity with which these postural verbs operate as a linker in locative constructions, in particular, led Ameka & Levinson (2007) to coin the label "postural-type language" (Gaby 2006: 461).

### 13 Basic clause types

• There is a grammatical classification of nominal concepts which may be independent.

As has been shown in this section, Choguita Rarámuri exhibits these properties in its positional predicate system.

### **13.2.4 Existential clauses expressing predicate possession**

In contrast to existential clauses containing locative predicates and a positional verb, where the existence of a given referent is given in function to a particular spatial location, 'true' existential clauses state the existence of a referent without providing any further information about the position of the referent or their grammatical classification. The latter kind of existential clauses in Choguita Rarámuri involve the verb *niˈlú* (with a suppletive stem *naˈlú* appearing in free variation), a lexical verb meaning 'exist' or 'to be in existence'. The examples in (36) illustrate this construction:

	- a. *ˈpé kuˈríbi ni biˈlé buˈrîto niˈlú pa ˈpé* just *kuˈrí=bi* recently=just *ni* nor *biˈlé* one *buˈrîto* donkey *niˈlú* exist *pa* cl 'Just barely, there were no donkeys either.' 'Nomás apenas, ni burros había tampoco.' < FLP 06 in61(230)/in > b. *tʃ͡i ˈbíri ˈnápi ˈkim naˈlúa* **ˈlé ... kuˈtʃâra** *ˈétʃ͡i* dem *ˈbíri* kind *ˈnápu* sub *ˈki=mi* sub=dem *naˈlú-a* exist-prog *aˈlé* dub *kuˈtʃâra* spoons 'There are a lot of things of that kind... spoons.' 'Muchas cosas que hay de ese tipo...cucharas.' < SFH 06 in61(725)/in > c. *aˈpítʃ͡iri aʔˈlì ˈpîri ˈʃî ˈtʃé ˈá niˈlú*? *aˈpítʃ͡iri* great.grandchildren *aʔˈlì* and *ˈpîri* what *ˈsî* else *ˈtʃé* also *ˈá* aff *niˈlú* exist.prs 'Greatgrandchildren and then, what else is there?' 'Tataranietos y luego, ¿qué más hay?' < SFH 06 in63(143)/in > d. *napaʔˈlì ke ˈtʃ͡ó niˈlúi ko sekunˈdâria ba ˈnápi* sub *aʔˈlì* later *ke* neg *ˈtʃ͡ó* yet *niˈlú-i=ko* exist-impf=emph *sekunˈdâria* secondary *ba* cl 'when there wasn't any secondary school yet' 'cuando todavía no había secundaria' {SFH tx12:1:22.2}

### 13.2 Locative, copula and existential clauses

Existential verbs may bear the full range of inflectional marking attested in other verbs. Some examples of the predicate *niˈlú* with several tense/aspect distinctions are provided in (37).

	- a. *we niˈlúmi ˈlé ˈlóali ba we* int *niˈlú-mi* exist-irr.sg *aˈlé* dub *ˈlówa-li* be.hungry-nmlz *ba* cl 'Perhaps there will be a lot of hunger.'

'A lo mejor va a haber mucha hambre.' {SFH in243:20:23.8}

b. *ˈpé kuˈríi ˈhônsa niˈlúuli ˈnápi riˈká ˈmá niˈlúsa ˈétʃ͡i rupuˈrá ba? ˈpé* just *kuˈrí* recently *ˈhônsa* since *niˈlú-li* exist.sg-pst *ˈnápu* sub *riˈká* like *ˈmá* already *niˈlú-sa/* exist.sg-cond *ˈétʃ͡i ripuˈrá ba*

dem ax cl

'Just recently, when there were axes already?'

'¿Hace poquito cuando ya hubo hacha?' {SFH in242:0:33.2}

	- *ˈká* because *ˈtʃ͡è* neg *ka* ka *niˈlú-i* exist-impf *tiˈjôpa* church *ˈí* here *kiˈʔà=ko* before=emph *noˈráatʃ͡i* Norogachi

*be=ko* but=emph *naˈlìna* because *oˈmáwa* party *noki-ˈwá-i* make-mpass-impf *pa* cl

'Because there was no church here before, they would make the celebrations in Norogachi.'

'Porque antes aquí no había aquí iglesia, en Norogachi era donde hacían las fiestas.' < FLP 06 in61(424)/in >

d. *we aʔˈlá rutuˈbúri oraˈkáli ba baˈhîtʃ͡i ˈbéli niˈlúami baʔˈwí pa we* int *aʔˈlá* well *rutuˈbúri* rutubúri *oraˈká-li* make-pst *ba* cl *baˈhîtʃ͡i* well *beˈlá* indeed *niˈlú-ame* exist-ptcp *baʔˈwí* water *pa* cl 'Yes, a well done *rutubúri*, with water from the well.'

'Sí, bien hecho el *rutubúri*, con agua del aguaje.' {FLP in243:0:32.1}

As seen in this examples, the predicate *niˈlú* may be inflected for irrealis singular (37a), conditional (37b), imperfective (37c) and participial (37d), among others.

Negative existential constructions may also be expressed through the predicate *niˈlú* and negative adverbs, as shown in (38).

### 13 Basic clause types

(38) *ˈkátʃ͡i niˈlúuli ba ripuˈrá pa ˈká* because *ˈtʃé* neg *niˈlú-li* exist-pst *ba* cl *ripuˈrá* axes *pa* cl 'There were no axes.' 'No había hachas.' {FLP in243:19:46.9}

Alternatively (and more frequently), negative existential clauses may be expressed through the negative predicate *iˈtê*, as shown for negative existential clauses involving positional/postural predicates, and as shown in (39).

(39) The negative predicate *iˈte* in existential clauses


cuando todavía no era ejido aquí.' {JMF tx817:0:08.6}

# **14 Sentence types**

This chapter is concerned with non-basic, non-complex clause types that encode different illocutionary acts and other pragmatically-marked structures, including interrogatives, negatives and imperatives. These constructions are achieved by a number of syntactic and morphosyntactic strategies. This chapter also addresses the properties of comparative constructions. This description is organized in terms of morphosyntactic and syntactic properties of each construction, as well as the prosodic properties of a subset of these constructions. Each construction is described in terms of a comparison to the morphosyntactic (and, where relevant, prosodic) characteristics of declarative sentences.

This chapter is organized as follows. The basic properties of declarative sentences are summarized in §14.1, to serve as a baseline for the description of nonbasic sentence types. Interrogatives are described in §14.2, including the distinction between polar questions (§14.2.1) and content questions (§14.2.2). Negative clauses are addressed in §14.3. Imperatives are addressed in §14.4. Finally, comparative constructions are reviewed in §14.5.

### **14.1 Declarative sentences**

Choguita Rarámuri is a highly agglutinative, head-final language, where word order is canonically SOV, a pattern documented across Uto-Aztecan languages (Langacker 1977). The basic word order is exemplified below, with a clause headed by a ditransitive predicate with both pronominal and NP arguments (1a) and by a transitive predicate with pronominal arguments (1b).

	- a. *ˈmò ˈjêla taˈmí haˈré gaˈjêta ˈʔàko [ˈmò* 2sg.nom *ˈjê-la]agent* mother-poss *[taˈmí]recipient* 1sg.acc *[haˈré* some *gaˈjêta]theme* cookie *ˈʔà-k-o* give-pst.ego-ep 'Your mom gave me some cookies' 'Tu mamá me dió unas galletas' < BFL 09 1:89/el >

### 14 Sentence types

b. *baʔaˈrîni ˈmí ˈàma baʔaˈrî=[ni]agent* tomorrow=1sg.nom *[ˈmí]theme* 2sg.acc *ˈà-ma* look.for-fut.sg 'I'll look for you tomorrow' 'Te voy a buscar mañana' < LEL 09 1:70/el >

Declarative sentences can be described morphosyntactically and syntactically in terms of the basic clause types as described in Chapter 13 (§13.1.1). The main morphosyntactic properties of basic clauses are summarized in (2):

	- The basic clause consists of a predicate and the arguments it subcategorizes for.
	- Pronominal forms encode core grammatical relations (subject and object), but there is no case marking in noun phrases.
	- Given that noun phrases can be elided, a minimal clause in Choguita Rarámuri consists of an inflected verbal predicate.

Declaratives can also be characterized by the following prosodic properties (see §6.2; Caballero et al. 2014, Garellek et al. 2015):

	- There is evidence of H% boundary tones
	- Boundary H% tones may be overridden by lexical or grammatical HL tones
	- There are "lead" (rhythmic) tones associated with lexical tones: lexical H and HL tones are optionally preceded by a L lead tone, while lexical L tones are optionally preceded by a H lead tone (Garellek et al. 2015)
	- There are general and tone-specific non-tonal devices that encode intonation, including vowel re-articulation and lengthening at phrasal boundaries (Aguilar et al. 2015; Caballero et al. 2014)

Figure 6.2 and Figure 6.3 from Chapter 3 (repeated here as Figure 14.1 and Figure 14.2 for convenience) show the presence of H% boundary tones and their absence in sentences with a final lexical or grammatical HL tone, respectively.

The morphosyntactic and prosodic properties of declaratives provide the baseline for comparison when describing interrogative and imperative constructions. These are addressed next.

### 14.1 Declarative sentences

Figure 14.1: High boundary tone in declaratives (Garellek et al. 2015)

Figure 14.2: No high boundary tone in declaratives with lexical HL tones (Garellek et al. 2015)

### 14 Sentence types

### **14.2 Interrogative constructions**

Different languages exploit a variety of morphosyntactic means to encode different speech acts, including assertions, commands and requests. While it is not uncommon that assertions or declarative sentences are unmarked, requests for information, or interrogative clauses, are cross-linguistically encoded through several grammatical devices. This section presents an overview of the conventionalized grammatical means Choguita Rarámuri exploits to express such requests for information.

Interrogative constructions in Choguita Rarámuri include a distinction between polar questions (discussed in §14.2.1) and content questions (discussed in §14.2.2), and each type in turn can be classified into different subtypes depending on their morphosyntactic and prosodic properties.

As discussed below, interrogative constructions that are morphosyntactically equivalent to declaratives are encoded exclusively through prosodic means, but prosody also plays a role in those interrogative constructions where there is a morphosyntactic device encoding the interrogative meaning. The existence of morphosyntactic mechanisms encoding different utterance types may preclude the use of distinctive intonational structures for the same purposes in some languages (e.g., Navajo (Athabaskan; McDonough 2002)). In Choguita Rarámuri, both polar and content interrogative constructions are characterized by two main prosodic properties. These are listed in (4):

	- A boundary H% tone targets the last stressed syllable of the utterance.
	- There is raised register across the utterance.

While polar questions show some degree of variation in the realization of the boundary tone and the degree to which register is raised (discussed below and in Chapter 11), they consistently exhibit a high pitch target utterance-finally. As shown below, the magnitude of the peak is correlated with the presence/absence of an overt morphological device to encode a question, with the highest pitch excursions attested utterance-finally with morphosyntactically unmarked polar questions.

14.2 Interrogative constructions

### **14.2.1 Polar questions**

Polar questions may be classified morphosyntactically into three types:

	- Morphosyntactically unmarked polar questions.
	- Polar questions with interrogative particles.
	- Polar questions with interrogative tags.

Each type is addressed in the following subsections.

### **14.2.1.1 Morphosyntactically unmarked polar questions**

Polar interrogative clauses may be equivalent morphosyntactically to their declarative counterparts, differing only in their prosodic make-up. Some examples of unmarked polar questions are provided in (6) from conversational data.

	- a. *pe uˈsànabi? pe* just *uˈsàni-na=bi?* six-distr=just 'Just in six places? 'Nomás en seis partes?' {SFH in61:04:37.8}
	- b. *ke biˈlé pe ˈtâʃi iˈtêeli ˈònam t͡ʃaˈbè ko? ke* neg *biˈlé* one *pe* neg *ˈtâsi* neg *iˈtê-li* exist.neg-pst *ˈòna-ame* cure-ptcp *t͡ʃaˈbè=ko* before=emph 'There was no medicine before?' 'No había medicina antes?' {SFH in61:06:39.6}

As discussed in Chapter 6, closely-related Mountain Guarijío (Taracahitan; Uto-Aztecan) is reported to also have polar questions that are morphosyntactically equivalent to their declarative counterparts, with the interrogative meaning encoded through ascending intonation ("*[g]eneralmente tienen entonación ascendente*") (Miller 1996: 112). A minimal pair between a declarative sentence and its morphosyntactically unmarked polar question counterpart in Choguita Rarámuri is provided in (7). Figure 14.3 and Figure 14.4 show the intonational difference between the declarative sentence in (7a) and the polar interrogative in (7b), respectively.

### 14 Sentence types

	- a. *ˈmá ˈtôlo ˈmá* already *ˈtô-li* bury-pst 'S/he buried him/her.'<sup>1</sup> 'Lo enterró.' < BFL el1170 > b. *ˈmá ˈtôli? ˈmá ˈtô-li?*
		- already bury-pst 'Did s/he bury him/her?' '¿Lo enterró?' < BFL el1307 >

The declarative sentence in Figure 14.3 features the falling pitch contour of the lexical HL tone of the verb (*ˈtô* 'to bury') (note there is no evidence of a H% boundary tone found in declaratives in this utterance since the falling lexical tone in the stressed syllable overrides the boundary tone, as discussed in §11.4).

<sup>1</sup> In this particular example, the suffix vowel surfaces as [o] given an optional round harmony process, where non-round vowels of certain suffixes may become round when preceded by a stressed back stem vowel. for more details about this process, see §9.4.3.5.

Figure 14.4: Morphosyntactically unmarked polar interrogative: *ˈmá ˈtôli?* 'Did s/he bury him/her?' (< BFL el1307 >)

The comparison between the declarative utterance in Figure 14.3 and the interrogative one in Figure 14.4 shows raised register and significantly raised f0 in the stressed syllable in the interrogative (from 289 Hz to 396 Hz in this particular example for the same female speaker (BFL)). The high pitch target in the stressed syllable may be attributed to the interrogative H% intoneme aligning with the peak of the lexical HL tone in this stressed syllable.

A different intonation pattern is documented in interrogative sentences with a final stressed syllable specified for lexical L tone. This is shown in Figure 14.5.

As shown in this Figure, the lexical tone of the verb root is associated to the stressed syllable, while the H% boundary tone docks on a following, unstressed syllable. Thus, and as documented in declarative contexts, lexical tones are preserved over intonational ones (see §11.4 for further discussion).

### **14.2.1.2 Polar questions with interrogative particles**

Polar questions may also be encoded through polar interrogative particle *ˈátʃ͡e* or its reduced form *ˈá*, which occur in clause initial position. Some examples of polar questions are provided in (8) (in these examples polar interrogative words are highlighted in **boldface**).

Figure 14.5: Accommodation of L tone and H% boundary tone in *ˈmá ˈnèli?* 'Did s/he see him/her?' (< BFL el1307 >)

	- a. *ˈátʃe wiˈtʃ͡iara hu?*
		- *ˈátʃe wiˈtʃ͡iara hu?*

Q truth cop.prs

'Is it true?'

'¿Es verdad? < SFH 06 in61(580)/in >

b. *ˈátʃi ˈmí riˈwèki?*

```
ˈátʃe
Q
     ˈmí
     2sg.nom
                riˈwè-ki
                leave-pst.ego
```

```
'Did you leave it?'
```

```
'¿Lo dejaste?' < GFP 09 3:14/el >
```
c. *ˈám ku ˈpàa saˈmîra?*

*ˈá=mi ku ˈpàa saˈmîra?*

	- '¿Ya trajiste a Samira?' < JLG co1237[0\_496-0\_506 >

The following minimal pair illustrates the contrast between declarative sentences (e.g., (9a)) and their polar interrogative counterparts (e.g., (9b)).

14.2 Interrogative constructions

	- a. *ˈmá naʔˈpôli naˈʔî ˈmá* already *naʔˈpô-li* weed-pst *naˈʔî* here 'S/he already weeded here.' 'Ya escardó aqui.' < SFH el1586 > b. *ˈátʃe ˈmá naʔˈpôli naˈʔî?*
		- *ˈátʃe* Q *ˈmá* yet *naʔˈpo-li* weed-pst *naˈʔî?* here 'Did s/he already weed here?'

'¿Ya escardó aqui?' < SFH el1586 >

As seen in this minimal pair, there is no change in the order of constituents in the interrogative construction.

The following minimal pair illustrates the prosodic difference between a declarative with a lexical L tone in utterance-final position ((9a) in Figure 14.6) and its polar interrogative counterpart with an interrogative particle ((9b) in Figure 14.7).

(10) Declarative vs. polar interrogative with *ˈátʃ͡e*

```
a. ˈmá naˈwàli
    ˈmá
    already
             naˈwà-li
             arrive-pst
    'S/he already arrived.'
    'Ya llegó.' < SFH-nawa-arrive-L-minimal-sets >
b. ˈátʃe ˈmá naˈwàli?
    ˈátʃe
    Q
          ˈmá
          already
                   naˈwà-li
                   arrive-pst
    'Did s/he already arrive?'
```
'¿Ya llegó?' < SFH-nawa-arrive-L-minimal-sets >

As attested in polar interrogatives that are morphosyntactically unmarked, the intonational encoding of polar questions with interrogative particles also involves the presence of a H% boundary tone, raised register and the preservation of lexical L tone: as seen in Figure 14.7, the sharp rise and peak of the H% tone is aligned with the final, post-tonic syllable.

Figure 14.6: Declarative with utterance-final lexical L tone in *ˈmá naˈwàli* 'S/he already arrived.'

Figure 14.7: Polar interrogative with *ˈátʃ͡e* and utterance-final lexical L tone in *ˈátʃ͡e ˈmá naˈwàli?* 'Did s/he already arrive?'

14.2 Interrogative constructions

### **14.2.1.3 Polar questions with interrogative tags**

A third type of polar question involves the use of a negative interrogative tag (formally a phrase) at the end of the sentence. The negative interrogative tag characterizes sentences as questions, but also contributes an expectation that the answer will be positive. An example of this construction is provided in (11):

(11) Polar questions with interrogative tags

a. *ˈmá ku siˈmíli maˈnûel, we ra ˈkeo? ˈmá* already *ku* rev *siˈmí-li* go.sg-pst *maˈnûel,* Manuel *we ra ˈke-o?* or or neg-ep 'Manuel already left, or not?' 'Manuel ya se fue, ¿o no?' < BFL 09 1:45/el >

The tag question involves a disjunctive conjunction and a negative particle. In this particular example, the speaker expects a positive answer.

### **14.2.2 Content questions**

In the canonical structure of content questions in Choguita Rarámuri, the question word appears in clause initial position. This is exemplified in (12).

	- a. *ˈpîri iʔˈkîli koˈtʃ͡î*? *ˈpîri* what *iʔˈkî-li* bite-pst *koˈtʃ͡î?* dog 'What did the dog bite?' '¿Qué mordió el perro?' < BFL 09 el725/el >
	- b. *ˈhêpi ˈkwâmi ˈʔâbo maˈjêi winoˈmî? ˈhêpi ˈkwâ=mi* who=2sg.nom *ˈʔâ-bo* give-fut.pl *maˈjê-i* think-impf *winoˈmî?* money 'Who did you think they were going to give the money to?' '¿A quién creías que le iban a dar el dinero?' < BFL 09 1:12/el >

As mentioned above, content (or information) questions are also characterized by a distinctive intonational contour: as shown in the contrast between a declarative sentence (shown in (13a), Figure 14.8) and a content question (shown in (13b), Figure 14.9).

### 14 Sentence types

	- a. *ˈmá naˈwàli ˈmá* already *naˈwà-li* arrive-pst 'S/he already arrived.' 'Ya llegó.' < SFH-nawa-arrive-L-minimal-sets > b. *ˈhêpi ˈkwâ naˈwàli?*
	- *ˈhêpi* who *ˈkwâ* who *naˈwà-li* arrive-pst 'Who arrived?' '¿Quién llegó?' < SFH-nawa-arrive-L-minimal-sets >

Figure 14.8: Declarative with utterance-final lexical L tone

The pitch track of the content question in Figure 14.9 shows a boundary H% tone in the last unstressed syllable of the utterance and register raising across the utterance.

A H% boundary tone is also present in content questions where a lexical H tone is associated to the final stressed syllable of the utterance. This is shown in (14), with a contrast between a declarative in (14a) (illustrated in Figure 14.10) and a content question in (14b) (illustrated in Figure 14.11).

(14) Declarative vs. content question: utterance-final, lexical H tone

a. *ˈmá muˈrúli ˈmá* already *muˈrú-li* carry.in.arms-pst 'S/he already carried it in their arms' 'Ya lo cargó en brazos' < BFL-muru-carry-H-minimal-sets >

### 14.2 Interrogative constructions

Figure 14.9: Content question with utterance-final lexical L tone

b. *ˈhêpi ˈkwâ muˈrúli?*

*ˈhêpi ˈkwâ muˈrú-li*

who who carry.in.arms-pst

'Who carried it in their arms?'

'¿Quién lo cargó en brazos?' < BFL-muru-carry-H-minimal-sets >

Figure 14.10: Declarative with utterance-final lexical H tone

In the declarative sentence represented in Figure 14.10 there is no evidence of a H% boundary tone, which is optional in declaratives. In the content question represented in Figure 14.11, on the other hand, there is a clear pitch target in the final, unstressed syllable of the utterance, which is higher than the one associated with the lexical H tone of the stressed syllable (a difference of almost 30Hz in this particular example).

### 14 Sentence types

Figure 14.11: Content question with utterance-final lexical H tone

There is, however, no evidence of a H% boundary tone nor any register manipulation when the content question contains a lexical HL tone in utterance final position. This is shown in the contrast between the declarative sentence in (15a) (Figure 14.12) and its content question counterpart in (15b) (Figure 14.13).

(15) Declarative vs. content question: utterance-final, lexical HL tone

a. *ˈmá iˈsîli ˈmá* already *iˈsî-li* pee-pst 'S/he already peed.' 'Ya orinó.' < BFL-isi-pee-HL-minimal-set > b. *ˈhêpi ˈkwâ iˈsîli? ˈhêpi* who *ˈkwâ* who *iˈsî-li* pee-pst 'Who peed?' '¿Quién orinó?' < BFL-isi-pee-HL-minimal-set >

In this pair of examples (produced by the female speaker BFL), the highest pitch peak associated with the lexical HL tone is comparable (270Hz in the declarative in Figure 14.12 and 260Hz in the content question in Figure 14.13), and their pitch countours largely equivalent. Thus, the comparison between the intonational contours of these two sentences shows that there is no evidence of any distinctive intonational encoding of a content question where HL tones override the H% tone associated with interrogative constructions elsewhere. This stands in contrast to polar questions, where, as shown in Figure 14.4 above, there is significantly raised f0 in the final stressed syllable of the interrogative utterance.

Figure 14.12: Declarative with utterance-final lexical HL tone

Figure 14.13: Content question with utterance-final lexical HL tone

Content or information questions use one of a set of interrogative pronouns and phrases (see also §10.1.4), which mark the clause as an interrogative one and fulfill the function of indicating which information is being requested. Table 14.1 displays the question words and phrases encountered in the Choguita Rarámuri corpus.

While the canonical position of interrogative words in content questions is in clause initial position (as in (16a)), they may also appear *in situ* (as in (16b)).

(16) Position of interrogative words in content questions

a. *Fronted ˈpîri iʔˈkîli koˈt͡ʃî?* 14 Sentence types

> *ˈpîri* what *iʔˈkî-li* bite-pst *koˈt͡ʃî?* dog 'What did the dog bite?' '¿Qué mordió el perro?' < BFL 09 el725/el > b. *In situ koˈt͡ʃî ˈpîri iʔˈkî-li? koˈt͡ʃî* dog *ˈpîri* what *iʔˈkî-li* pst 'What did the dog bite?' '¿Qué mordió el perro?' < BFL 09 el725/el >

As shown in (16b), the question word for the object appears pre-verbally and after the Subject NP, as in the unmarked SOV word order in Choguita Rarámuri (as described in §14.1).


Table 14.1: Choguita Rarámuri interrogative words and phrases

### 14.2 Interrogative constructions

The examples so far have shown constituent interrogatives in simple clauses. In complex clauses, it is possible to ask questions where the question word stands for an argument of a complement clause. In these cases, the interrogative word is fronted in the matrix clause. This is shown in the contrast between a declarative sentence with a subordinate clause (in (17a)) and the content question counterpart of this declarative (in (17b)). The question word is highlighted in boldface.

(17) Constituent interrogative: argument of subordinate clause


'Who did you think they were going to give the money to?'

'¿A quién pensaste que le iban a dar el dinero?' < BFL 09 1:12/el >

As can be seen in these examples, constituent order is not altered beyond placement of the question word in sentence initial position: both sentences share the property of having the theme argument of the complement clause in sentence final position (*winoˈmî* 'money'), and the dependent verb (*ˈʔà* 'give') preceding the matrix verb (*maˈjê* 'think').

Further examples of constituent interrogatives of arguments of complement clauses are provided in (18).

(18) Constituent interrogative: argument of subordinate clause


### 14 Sentence types

These examples show a content question where the question word stands for an argument of a complement clause, an object argument in (18a) and a recipient argument (18b) of the ditransitive predicate *ˈʔà* 'to give' (the complement clauses are indicated with square brackets). Further details on the syntax of complement clauses are provided in §15.1.

### **14.3 Negation**

Negation may be expressed in Choguita Rarámuri through a number of mechanisms. First, there are several negative free forms, addressed in §14.3.1 below, which are antonyms of positive free forms (*aˈjena* and *ˈtio* 'yes'). Some of these forms are also deployed in clausal negation, as discussed in §14.3.2. Negative forms that function as constituent negators are described in §14.3.3. Finally, negative existential and negative locative clauses are addressed in §14.3.4.

The set of negative particles and complex negative markers available in Choguita Rarámuri are provided in Table 14.2, with their gloss, function and approximate translation (for more details about the morphological properties of negative particles, see §10.8.3).<sup>2</sup>

In addition to these negative particles and complex negative markers, there are morphologically complex negative quantifiers in the language which involve negative particles, including *ke/ˈtâsi (biˈlé) naˈmûti* 'none, nothing' and *ke/ˈtâsi ˈwêsi* 'nobody'.

### **14.3.1 Negative free forms**

A range of negative forms, including *ke*, and *ke ˈtâsi*, are frequently found in answers to questions, whether the questions are negative (as in (19)) or positive (as in (20) and (21)) indifferently (in contrast to languages that have specialized markers for answers contradicting negative questions, e.g. French *si*, German *doch*, inter alia).

<sup>2</sup>As discussed in Chapter 10 (§10.8), 'particles' are defined as a set of heterogeneous word classes that are characterized by being closed and morphologically simple, bearing no inflection or derivation and, in some cases, being phonologically reduced. Each class of particles is composed of fewer than a dozen members per class and may have a wide range of functions and meanings.

14.3 Negation


### Table 14.2: Negative markers

	- a. [GCH]: *ke me apaˈrûame ˈétʃ͡i ke* neg *me* almost *apaˈrûame* fierce *ˈétʃ͡i?* dem 'Is he not that fierce that one? '¿No es tan bravo ese?' < GCH co1136 >
	- b. [MDH]: *B: ke, ke me oˈbáta ˈlé ke,* neg *ke* neg *me* almost *oˈbáta* fierce.pl *aˈlé* dub 'No, they are not that fierce' 'No, casi no son bravos (no se enojan)' {MDH co1136:14:41.4}
	- a. [SFH]: *naˈlìna ke pe ˈá ˈtʃ͡étimi itʃ͡aˈkámi koˈʔáa onoˈkáli aˈlé tʃ͡aˈbèewami ko ˈnápu riˈká ˈnàa ... ˈnápu riˈká reˈpôjo ˈmí itʃ͡iˈwáwamti? naˈlìna* so *ke* neg *pe* just *ˈá* aff *ˈtʃ͡e=timi* also=2pl.nom *itʃ͡aˈ-kámi* plant-ptcp *koˈʔá-a* eat-prog *onoˈká-li* do--pst *aˈlé* dub *tʃ͡aˈbè-w-ami=ko* before-nmlz-ptcp=emph *ˈnápu* sub *riˈká* like *ˈnà* dem *ˈnápu* sub *riˈká* like *reˈpôjo* cabbage

14 Sentence types

> *ˈmí* there *itʃ͡iˈ-wá-w-am-ti?* plant-mpass-w-ptcp-com 'And did you all indeed also plant (for eating) like ... like the planting of cabbage? '¿Y sí sembrarían antes (cosas) como ... así como la siembra de repollo?' {SFH in242:3:42.0}, {in242:3:52.8}

b. [FLP]: *ke ˈtâsi ke* neg *ˈtâsi* neg

'No.' {FLP in242:03:55.7}

	- a. [SFH]: *tʃ͡ú riˈká? ˈnè waˈjéla ˈjûa aʃiˈsâ? tʃ͡ú* Q *riˈká* how *ˈnè* 1sg.nom *waˈjé-la* younger.sister.male.ego-poss *ˈjûa* with *asi-ˈsâ?* sit-cond 'How? As if he were (together) with my younger sister?' '¿Cómo? ¿Cómo si estuviera (emparejado) con mi hermana menor? {SFH in485:02:32.0} b. [ME]: *ke!*
		- *ke!* neg 'No!' {ME in485:02:33.8}

These negative forms are also used to contradict statements assumed by the speaker to be incorrect, as in (22).

	- a. [GCH]: *ˈmá naˈwàli miˈkêli ˈmá* already *naˈwà-li* arrive-pst *miˈkêli* Michael 'Michael already arrived' 'Ya llegó Miguel'
	- b. [MDH]: *ke ˈtâsi, pe ke ˈtʃ͡ó naˈwàli ke* neg *ˈtâsi* neg *pe* just *ke* neg *ˈtʃ͡ó* yet *naˈwà-li* arrive-pst 'No, he hasn't arrived yet' 'No, todavía no llega'

14.3 Negation

The form *pe ke biˈlé* is an emphatic interjection, adding emphasis to a question or statement to express a strong disagreement. Examples of this emphatic interjection are provided in (23) and (24).

	- a. [FLP]: *ˈnápi sibiˈrîko ˈdîas aˈtí ˈká ˈtʃ͡è wiliˈbê wasaˈrúi ba ˈnápi* sub *sibiˈrîko* Federico *ˈdîas* Diaz *aˈtí* be.sitting.prs *ˈká* neg *ˈtʃ͡è* neg *wiliˈbê* be.lying.prs *wasa-ˈrú-i* plow-pst.pass-impf *ba* cl 'Because where Federico Díaz is (lives) it was barely plowed (the land).'

'Porque donde está Federico Díaz casi no estaba barbechado (no estaba muy ancha la tierra).' {FLP in61:04:21.8}

	- a. [SFH]: *ke biˈlé kiliˈsântemi ˈòuka riˈká ke* neg *biˈlé* one *kiliˈsân=timi* fertilizer=2pl.nom *ˈòwi-ka* cure.fertilize-ger *riˈká* like 'Didn't you all cure (fertilize) with fertilizer?' '¿No curaban (fertlizaban) con fertilizante?' {SFH in484:01:50.0} b. [ME]: *ˈpé ke biˈlé! ˈpé ke biˈlé! ˈpé kuˈríbi ko ˈhônsa niˈlú aˈlé ˈpé* just *ke* neg *biˈlé!* one *ˈpé* just *ke* neg *biˈlé!* one *ˈpé* just *kuˈrí=bi=ko* recently=just=emph *ˈhônsa* since *niˈlú* exist.prs *aˈlé* dub 'Not at all! Not at all! It just recently started (fertilizing crops)'

'No, nada! No, nada! Hace poco que empezó (lo de fertilizar)' {ME in484:01:50.6}

The emphatic negative interjection *ˈpé ke biˈlé* is not found with other constituents in clauses. The negative interjections *ke* and *ke ˈtâsi*, on the other hand, may also occur in phrases and clauses and are involved in clausal negation, as described next.

### 14 Sentence types

### **14.3.2 Clausal negation**

Clausal negation in Choguita Rarámuri is achieved using the forms *ke*, *ke ˈtâsi*, *ka tʃ͡è* and *ke/ˈtâsi ˈtʃ͡ó*. Negative forms typically occur before the predicate, following a cross-linguistic trend noted in (Dryer 2007a: 105).<sup>3</sup> Negative forms may also occur in clauses where the predicate has been elided. This is shown in (25a) and (25b–c), respectively.

(25) Clausal negation


*ke* neg

*ke* neg

'Long ago we didn't have electricity and no water.' 'Antes no teníamos luz ni agua.' < SFH 09 4:2/el >

Further examples of clausal negation are provided in (26).

(26) Clausal negation

a. *ke ˈtâsi ˈʃíli ke* neg *ˈtâsi* neg *ˈsí-li* come.pl-pst 'They didn't come.' 'No vinieron.' {MDH co1136:03:12.0}

<sup>3</sup> In clausal negation in closely related Mountain Guarijío, Miller (1996) reports that the negative form *kaʔí* appears in clause-initial position (1996: 119).

14.3 Negation

b. *ˈká ˈtʃè kaiˈnâma ˈlé ke naʔˈpôʃuwa ˈká ba ˈká* neg *ˈtʃè* anymore *kaiˈnâ-ma* yield.harvest-fut.sg *aˈlé* dub *ke* neg *naʔˈpô-suwa* weed-cond.pass *ˈká* cop.irr *ba* cl 'There won't be any (harvest) yield if there is no weeding.' 'No se da (la cosecha) si no se escarda.' {MDH co1136:04:39.5} c. *ˈkíti ke tʃo riˈhòoli ˈmá naˈʔî ba ˈkíti* because *ke* neg *tʃo* yet *riˈhòo-li* inhabit.pl-pst *ˈmá* already *naˈʔî* here *ba* cl 'Because there was almost no people (living) here yet.' 'Porque casi no había gente aqui todavía.' {JMF tx817:00:36.9} d. *"ka biˈlé tʃo aˈwí ba" aˈnè ka* neg *biˈlé* one *tʃo* yet *aˈwí* dance *ba* cl *aˈn-è* say.imp.sg-appl ' "They haven't danced yet" you (should) tell them.' ' "Todavía no bailan", diles.' < JLG co1237[9\_156-9\_167] >

Clausal negation with *ka tʃ͡è*, as in (26b), is attested in conditional sentences where *ka tʃ͡è* appears in the main apodosis (consequence) clause, preceding the protasis clause. Other kinds of uses of *ka tʃ͡è* in clausal negation are attested in the context of a sequence of clauses where the negative form involves a negative consequence (roughly translated here as 'because not'). These uses are exemplified in (27) and (28).

(27) Clausal negation in clause chaining contexts


### 14 Sentence types

b. *ˈká ˈtʃè ko uluˈbê buʔuˈrúi tʃ͡aˈbèi ko ˈká* neg *ˈtʃè=ko* because=emph *waʔlu-ˈbê* big-more *buʔu-ˈrú-i* road-vblz-impf *tʃ͡aˈbèi=ko* before=emph 'Because there was no road before.' 'Porque antes no había camino grande.' {ME in485:06:50.6}

### (28) Clausal negation in clause chaining contexts


*ˈká ˈtʃè* neg because *wiʰka-ˈbê* many-more *riˈhò-li* inhabit.pl-pst *ba* cl *oˈkwâ* few *riˈhò-r-ame* inhabit.pl-pst-ptcp *ba* cl *ˈpé* just *be-ˈsá* three-times *maˈkòi* ten *riˈhò-li* inhabit.pl-pst *aˈlé* dub 'Because there were not that many people, (just) a few people, just about thirty people lived here, I think.'

'Porque casi no había gente, había muy poquita como treinta yo creo.' {JMF tx817:00:36.9}

As shown in the following example (29), the elements of morphologically complex negative markers need not be contiguous (negative forms are highlighted in boldface).

(29) Discontiguous negative forms

*aʔˈlì ˈmá ke ˈnà ˈtòli ˈtʃó ˈét͡ʃi riˈhò aliˈwâla ko aʔˈlì* and *ˈmá* already *ke* neg *ˈnà* that *ˈtò-li* take-pst *ˈtʃó* anymore *ˈét͡ʃi* dem *riˈhò* man *aliˈwâ-la=ko* soul-poss=emph 'And then he didn't take the man's soul anymore.' 'Y entonces ya no se llevó el alma del señor.' {LEL tx5:1:38.7}

### **14.3.3 Constituent negation**

Constituent negation in Choguita Rarámuri is encoded through a series of negative constructions, involving *ke* 'no', *ke ˈtâsi biˈlé* 'not one', *ni biˈlé* 'nor anything',<sup>4</sup>

<sup>4</sup>The form *ni* is a borrowing from Spanish, the negative conjunction.

14.3 Negation

*ke ˈtâsi naˈmûti* 'not a thing', and *ke ˈtâsi ˈwêsi* 'nobody'. These negative forms appear immediately preceding the modified constituent. Examples of constituent negation are provided in (30).

(30) Constituent negation *"ˈká ˈtʃè biˈlé kaˈsè maˈtʃ͡íni ba ˈni" ˈká* neg *ˈtʃè* anymore *biˈlé* one *kaˈsè* place *maˈtʃ͡íni* come.out.sg *ba* cl *ˈni* ni "'It's not coming out in any place."' "'No sale en ninguna parte."' {SFH tx152:6:37.5}

As described above, Choguita Rarámuri also expresses negation through two negative quantifiers: *(ke) ˈtâsi (biˈlé) naˈmûti* 'none, nothing' (where the noun *naˈmûti* means 'thing') and *(ke) ˈtâsi ˈwêsi* 'nobody' (where *ˈwêsi* means 'nobody, no one', as it is not attested outside of negative contexts). These negative quantifiers are exemplified in (31).

(31) Negative quantifiers


In cases where there is both clausal negation and constituent negation using negative quantifiers, the clausal negation form appears in clause initial position, preceding the negative quantifiers. This is shown in (32). The negative particles and quantifiers are highlighted in boldface.

14 Sentence types

(32) *ˈtâsi=ni biˈlé ˈwêsi ˈnè-nali ˈtâsi=ni* neg=1sg.nom *biˈlé* one *ˈwêsi* nobody *iʔˈnè-nale* see-desid 'I don't want to see anybody.' 'No quiero ver a nadie.'

In this particular example, the 1st person singular subject is encoded through the enclitic *=ni*, which attaches after the first constituent, following the pattern for Wackernagel position clitics.

Finally, there are also cases where complex negative expressions are discontinuous in clauses with constituent negation. This is shown in (33).

(33) Discontinuous negative forms in constituent negation *ˈpé ˈwé kaʔˈlá kaˈjèni kiˈʔàa ko ... ˈwé aʔˈlá ke bi ko biˈlé witaˈt͡ʃí naˈkí ba ˈpé* just *ˈwé* int *kaʔˈlá* well *kaˈjèni* harvest.prs *kiˈʔà=ko* before=emph *ˈwé* int *aʔˈlá* well *ke=bi=ko* neg=just=emph *biˈlé* one *witaˈt͡ʃí* fertilizer *naˈkí* need *ba* cl 'Crops would turn out well, there wasn't any need for fertilizer.' 'Se daba muy bien antes la cosecha, no se necesitaba abono.' {FLP in61:02:46.8 0:02.0}, {FLP in61:2:49.8}

While discontiguous, the complex negative marker still precedes the modified constituent (the noun *witaˈt͡ʃí* 'fertilizer').

### **14.3.4 Negative existential and locative clauses**

Existential negation is encoded through a dedicated negative predicate of existence used in conjunction with the negative particle *ke*, which contrasts with the positive polarity postural predicates deployed in locative clauses (see §13.2.3 and §13.2.4). Examples of negative existential predication are provided in (34).

(34) Negative existential predication

a. *ˈmá ke iˈtê ku, ˈmá oˈkwâ biˈtí ko ˈmá* anymore *ke* neg *iˈtê* be.neg *ku* wood *ˈmá* already *oˈkwâ* two *biˈtí=ko* be.lying.prs=emph 'There is no more wood, there is only a little left.' 'Ya no hay leña, ya nomás quedan unos pocos.' {MDH co1136:01:09.9}

14.4 Imperatives


Negative locative predication employs the same negative morphemes. This is shown in (35).

(35) Negative existence markers *ke iˈtê* in negative locative clauses *naˈʔî ke iˈtê ˈbôte naˈʔî* here *ke* neg *iˈtê* be.neg *ˈbôte* can 'The can is not here.' 'Aqui no está el bote.' < LEL 09 1:74/el >

### **14.4 Imperatives**

Imperatives are defined as constructions that encode directive speech acts (including orders, requests, warnings, invitations, etc.), usually directed at addressees (second persons) König & Siemund (2007); in some cases the term "imperative" is extended to constructions where commands, requests, etc. are addressed to the first or third person. This section addresses the multiple constructions that encode directive speech acts in Choguita Rarámuri, including positive imperatives, prohibitives, hortatives and motion imperatives.

### **14.4.1 Positive imperative**

Imperatives in Choguita Rarámuri are headed by an imperative-marked verb (for a description of imperative morphology, see Chapter 9). Imperative suffixes en14 Sentence types

code subject number distinctions, with a set of concatenative and non-concatenative allomorphs encoding a single addressee and a suffix encoding that the directive speech act is addressed to a group.

As discussed in §9.2.2, there are four allomorphs of the imperative singular: the stress-shifting *-kâ* suffix (36a), the stress-shifting *-sâ* suffix (36b–c), a rightward stress shift (36d) and a L tonal exponent (36f).

(36) Singular addressee imperatives


b. *koˈsâ!*

*ko-ˈsâ!* eat.imp.sg 'Eat!' '¡Come!'


14.4 Imperatives

g. *cf. hiˈrâ* 'S/he bet.' 'Apuesta.'

The stress-shift imperative allomorph is phonologically-conditioned, and is selected by unstressed trisyllabic roots (e.g., *naʔ'sòwa* 'to stir' in (36d)). The L tone allomorph of the imperative singular, on the other hand, is realized on the stressed syllable of HL-toned stems (e.g., *hiˈrâ* 'to bet' in (36f)) (see §11.3.7 in Chapter 11). The distribution of the allomorphs is otherwise lexically determined, though there are stems where the same lexical root may be attested with more than one allomorph in an apparent case of free variation (i.e., with no apparent semantic differences) and without any multiple exponence (i.e., one or another allomorph will surface, but not both at the same time), as shown in (37) and (38). This availability of several imperative morphemes may be related to strategies to attenuate directive speech acts. Such distinctions are not apparent in the available corpus of the language, though a larger corpus with greater contextualization of social contexts may reveal differences in these terms for the different imperative devices available.<sup>5</sup>

(37) Free variation in imperative allomorph selection


(38) Free variation in imperative allomorph selection

a. *tòˈkâ tòˈ-kâ* take-imp.sg 'Take it!' '¡Llévatelo!' < BFL el1882 >

<sup>5</sup>This seconds a suggestion made in (Miller 1996: 107) about potential distinctions between imperative constructions in closely-related Mountain Guarijío, where several imperative suffixes appear to be in free variation.

14 Sentence types

> b. *tòˈsâ tò-ˈsâ* take-imp.sg 'Take it!' '¡Llévatelo! < BFL el1882 > c. *ˈtòo ˈtòo* take.imp.sg 'Take it!' '¡Llévatelo!' < BFL el1882 >

As shown in these examples, variation in allomorph selection may involve a choice between two allomorphs (a suffix allomorph and a tonal allomrph in (37)), or it may involve three morphological marking options (two suffixal allomorphs and the tonal allomorph in (38)). In the latter example, the L tone allomorph of the imperative singular vacuously applies in the case of a L-toned stem like *tò* 'to take'.

In contrast to the imperative singular, the imperative plural involves no allomorphy, but a single productive suffix (stress-shifting *-sì*), exemplified in (39).

### (39) Plural addressee imperative

a. *aʔˈlì ˈmá ˈhê aniˈmêa Lola "ku baˈhîsi ne ko!" aniˈmêa aʔˈlì* and *ˈmá* already *ˈhê* it *ani-ˈmêa* say-fut.sg *Lola* Lola *"ku* rev *baˈhî-si* drink-imp.pl *ne=ko!"* exh=emph *ani-ˈmêa* say-fut.sg 'And Lola will promptly say: "Go on, drink up!", she will say." 'Y ya va a decir Lola: "¡Ya tómenle pues!", va a decir." {JLG co1234:18:28.2} b. *aʔˈlì ˈmá wiˈrôsa ko ˈhê aniˈmêa "ja baˈhîsi ba!" ˈhê aniˈmêa aʔˈlì* and *ˈmá* already *wiˈrô-sa=ko* throw.up.in.air-cond=emph *ˈhê* it *ani-ˈmêa* say-fut.sg *"ja* already *baˈhî-si* drink-imp.pl *ba!"* cl *ˈhê* it *ani-ˈmêa* say-fut.sg 'And when they throw it up in the air (the corn beer), they say "You all drink already!", that's how it is said.' 'Y ya cuando lo tiran (el tesgüino) para arriba, se dice "¡Ya tomen!",

así se dice.' {JLG co1234:12:52.2}

### 14.4 Imperatives


muchos.' {MFH tx1133:02:30.5}

As described above, the plural imperative is used when the speaker is addressing a group with a directive command. Crucially, the group excludes the speaker, as directive speech forms including the speaker deploy a different morphological construction, namely the hortative construction, described below in §14.4.3. 6

A special form is used in positive imperatives, where the speaker conveys encouragement, possibly as a politeness strategy, in the directive speech. This form, involving the imperative-inflected verb plus an exhortative particle *ne* and the emphatic enclitic *ko*, is attested in a variety of contexts involving a polite encouragement. This is exemplified in the commonplace expression in (39a) and (39b) above and (40).

<sup>6</sup>Miller (1996: 110) describes that closely related Mountain Guarijío does not make a distinction between single and multiple addresses in imperative morphology, but the future plural suffix ( *po/bo*, cognate with the Choguita Rarámuri future plural *-pô* suffix) is used in conjunction with imperative suffixes to attenuate the force of the directive command. There are no equivalent forms in the Choguita Rarámuri corpus.

14 Sentence types

(40) *ja koˈʔá ne ko! ja* already *koˈʔá* eat.imp.sg *ne=ko!* exh=emph 'Go on, eat!' '¡Come, pues!'

This construction is also deployed in affirmative answers to polite directive speech acts, as exemplified in another commonplace type of exchange (41b).

```
(41) a. A: ja baˈhîsi kaˈhêke!
            ja
            already
                    baˈhî-si
                    drink-imp.pl
                                  kaˈhêke!
                                  coffee
           'You all drink coffee!'
           '¡Ya tomen café!'
       b. B: hoʔu ne ko!
```
*hoʔu* yes *ne=ko!* exh=emph 'All right!' '¡Órale pues!'

### **14.4.2 Prohibitive**

Negative imperatives or prohibitives in Choguita Rarámuri are encoded through a construction involving a dedicated prohibitive morpheme and an imperativemarked verb. The imperative morphology in the inflected verb is the same as in positive imperative clauses (see §14.4.1). Prohibitive clauses contain the prohibitive marker *ˈkíti* in clause initial position. This is shown in (42).

(42) Prohibitive clauses

```
a. ˈkíti naˈlàka!
     ˈkíti
     proh
           naˈlà-ka
           cry-imp.sg
    'Don't cry!'
    '¡No llores!' <BFL 05 2:89/el >
b. ˈkíti riʔˈná bitiˈʃí!
     ˈkíti
     proh
           riʔˈná
           on.back
                    biti-ˈsí
                    lie.down.pl-imp.pl
    "'Don't sleep on your back!"'
    "'No se acuesten boca arriba!" {LEL tx5:06:00.1}
```
14.4 Imperatives

c. *"ˈkíti ˈpáka loˈkási!" ˈkíti* proh *ˈpá-ka* throw-ger *loˈká-si!"* drink.pinole-imp.pl "Don't drink pinole spilling it!' "¡No tomen pinole tirándolo!" {MDH co1136:10:25.9}

The marker *ˈkíti* is also used in subordinate clauses introducing reason clauses, as exemplified in (43) (see Chapter 15, §15.2.3).


As shown in these examples, the clause introduced by *ˈkíti* may be either positive (43a–b) or negative (43c).

14 Sentence types

### **14.4.3 Exhortative**

Imperative constructions in Choguita Rarámuri with an exhortative meaning are marked with the clause initial particle *to* and may optionally have a closing exhortative particle *bo*. This construction is exemplified in (44).

(44) a. *ˈtoo ˈjêni ˈdûlse ˈìwkitili jaˈdîra ˈto* exh *ˈjêni* Yéni *ˈdûlse* candy *ˈìwi-ki-ti-li* bring.appl-appl-caus-imp.sg *jaˈdîra* Yadira 'Go on, make Yeni bring candy for Yadira!' '¡Ve, haz que Yeni le traiga dulces a Yadira!' < BFL 07 1:62/el > b. *ˈtoo miˈtʃíktili bo ˈto* exh *miˈtʃí-ki-ti-li* carve-appl-caus-imp.sg *bo* exh 'Go on, carve it for him!' '¡Anda, lábraselo!' < BFL 08 1:107/el > c. *niˈbi! ˈtoo iˈʔné bo! niˈbi* look *ˈto* exh *iˈʔné* see.imp.sg *bo* exh 'Look! Take a look at it!' '¡Mira! ¡Míralo!' < BFL 06 Nov04/el >

In these examples, the speaker encourages the addressee to carry out an activity in a polite fashion. These types of polite requests are frequent in every day conversation.

### **14.4.4 Motion Imperatives**

One type of imperative construction in Choguita Rarámuri involves suffixes that, alone or in combination with imperative mood suffixes, encode a motion associated with the main verb of the event plus a command ('Go and do X!'), directed at a single or plural addressee. The stress-shifting suffix *-me* is used for single addressees. When there are multiple adressees, the Motion Imperative involves the stress-shifting suffix *-pi* (with stress-shifting allomorph *-bô*, followed by the imperative plural suffix. Examples of the motion imperative construction are provided in (45).

14.5 Comparatives

(45) a. *ˈétʃ͡i ˈhùrimi na! ˈétʃ͡i* dem *ˈhùri-mi* send-mot.imp *na* there 'Go and send (the fire bird) to him!' '¡Ve y mándale (el pájaro de fuego) a ese!' < LEL 06 tx5(80)/tx > b. *taˈmí ku ˈákipisi taˈmí* 1sg.acc *ku* rev *ˈá-ki-pi-si* look.for-appl-mot.imp.pl-imp.pl 'You all go and look for it for me!' 'Vayan a buscármelo!' < BFL 08 1:164/el >

As described in Chapter 9, there is an associated motion marker used in declarative sentences (for discussion about associated motion in cross-linguistic perspective, see Guillaume 2016).

### **14.5 Comparatives**

The final type of specialized sentence type addressed in this chapter is comparative constructions. Across languages, comparative constructions involve a predicate that encodes a predicative scale, and two noun phrases, one which encodes the object of comparison (the Comparee NP), and another which encodes the standard of comparison (the Standard NP) (for an overview of the typological parameters of variation of comparative constructions, see Stassen 1985 and Haspelmath & the Leipzig Equative Constructions Team 2017). The Choguita Rarámuri comparative construction involves a gradable predicate expressed through an adjective, a deverbal nominalized form or a verbal predicate. If the gradable predicate is encoded through an adjective or a nominalized form, the comparative also features an associated copular verb. Degree is optionally expressed through an adverb or a comparative suffix in the predicate. The object of comparison is a nominal phrase optionally followed by an emphatic particle. Finally, the standard of comparison is introduced through the postpositions*ˈ(j)ua* 'with' or*ˈkítara* 'about'. The schema in (46) shows the structure of Choguita Rarámuri comparative constructions.

(46) Structure of comparative constructions [Comparee NP] [(Adverb)] [Gradable predicate] [(Copula)] [Standard NP] [Postposition]

The examples in (47) illustrate the structure of comparative constructions.

### 14 Sentence types

(47) a. *siˈkâ watoˈná tʃ͡uˈkúam ˈká ˈwé raˈsíra ˈwáami ˈhú kohiˈná ˈjûa siˈkâ* arm.hand *watoˈná* right *tʃ͡uˈkú-ame* be.bent-ptcp *ˈká* cop.irr *ˈwé* int *raˈsíra* more *ˈwá-ame* be.strong-ptcp *ˈhú* cop.prs *kohiˈná* left *ˈjûa* than 'The right arm/hand if flexed is stronger than the left one.' 'El brazo derecho si está flexionado es más fuerte que el izquierdo.' < BFL 09 1:30/el > b. *ˈmò ko wiʔlí ˈhú taˈmí ˈjûa ˈmò=ko* 2sg.nom=emph *wiʔˈlí* tall *ˈhú* cop.prs *taˈmí* 1sg.acc *ˈjûa* than 'You are taller than me.' 'Tu estás más alto que yo.' < LEL 09 2:6/el > c. *ˈmò ko ˈtéeri ˈhú taˈmí ˈjûa ˈmò=ko* 2sg.nom=emph *ˈtéri* tall *ˈhú* cop.prs *taˈmí* 1sg.acc *ˈjûa* than 'You are shorter than me.' 'Tu estás más chaparro que yo.' < GFM 09 3:101/el >

All three examples in (47) involve an adjective or nominalized gradable predicate and the associated present tense copula *hu* 'is'. In (47a), the object of comparison noun phrase includes a relative clause, while the gradable predicate, the nominalized form *ˈwáami* 'strong', is modified by two degree adverbs (*we* 'very' and *raˈsíram* 'more'). In all of the examples in (47), the object of comparison appears in the initial position in the construction, followed by the gradable predicate, the copula verb, the standard of comparison and the postposition.

Negative comparisons ('less than') are encoded through negative particles that have under their scope the gradable predicate. Negative comparison constructions are exemplifed in (48).

(48) a. *ˈét͡ʃi tiˈwé ko ke me waˈríni loˈrêna ˈjûa ˈét͡ʃi* dem *teˈwé=ko* girl=emph *ke* neg *me* almost *waˈrína* light *loˈrêna* Lorena *ˈjûa* than 'That girl is less light (fast) than Lorena.' 'Esa muchacha es menos ligera que Lorena.' < LEL 09 2:6/el > b. *ˈhê ˈnà aˈsêite ko ke me natiˈkí ˈhê ˈnà ˈkítara ˈhê* dem *ˈnà* prox *aˈsêite=ko* oil=emph *ke* neg *me* almost *natiˈkí* expensive *ˈhê* dem *ˈnà* prox *ˈkítara* than 'This oil is less expensive than this other one.' 'Este aceite es menos caro que este otro.' < LEL 09 2:6 Steffel/el >

14.5 Comparatives

c. *wasaˈtʃ͡î ko ke me apaˈrûame ˈhú wasaˈtʃ͡î waʔˈlûara ˈjûa basaˈtʃ͡î=ko* coyote=emph *ke* neg *me* almost *apaˈrû-ame* angry-ptcp *ˈhú* cop.prs *basaˈtʃ͡î* coyote *waʔˈlû-a-ra* big-a-nmlz *ˈjûa* than 'The coyote is less dangerous than the wolf.' 'El coyote es menos bravo que el lobo.' < SFH 09 3:85/el >

Comparative constructions may also be encoded via asyndetic conjunction or parataxis, the juxtaposition of clauses with no overt conjunction, with verbal ellipsis in the second clause. This is exemplified in (49).

(49) *naˈʔî ko ke me ruruˈwá, niˈhê biˈtêritʃ͡i ko we [naˈʔî=ko* here=emph *ke* neg *me* almost *ruruˈwá]* cold *[niˈhê* 1sg.nom *biˈtêritʃ͡i=ko* house=emph *we]* int 'It is less cold here than were I live' (lit. 'It is not very cold here. It is very (cold) where I live.') 'Hace menos frío aquí que en donde vivo' (lit. 'Aquí casi no hace frío. Donde yo vivo, mucho.') < SFH 09 3:85/el >

Comparative constructions may also include a gradable predicate which is morphologically marked as comparative through the suffix *-bê* 'more/surpass'. The examples in (50) illustrate this morphological marker:

(50) a. *ˈká ˈt͡ʃè wikaˈbê riˈhòo-li ba oˈkwâ riˈhòorame ba ˈpé be-ˈsá maˈkòi riˈhòo-li ˈlé ˈká* neg *ˈt͡ʃè* neg *wikaˈ-bê* far-more *rihói-li* inhabit.pl-pst *ba* cl *oˈkwâ* few *rihói-li-ame* inhabit.pl-pst-ptcp *ba* cl *ˈpé* just *be-ˈsá* three-times *maˈkòi* ten *riˈhòi-li* inhabit.pl-pst *aˈlé* dub 'Because there were almost no people living here, more far away, just like thirty lived here.' 'Porque casi no había gente aquí, había muy poquita como treinta yo creo.' < JMF 09 tx817(6)/tx > b. *ˈní-ma be ˈlà-o ˈpé a wili-ˈbê iˈná-ma ˈlé ˈmò ko ba ˈní-ma* cop-fut.sg *be* be *ˈlà-o* think-ep *ˈpé* just *a* aff *wili-ˈbê* long-more *iˈná-ma* go.along.sg-fut.sg *aˈlé* dub *ˈmò=ko* 2sg.nom=emph *ba* cl 'I think so, you will be around (live) for a longer time.'

### 14 Sentence types

'Yo creo que si, tu si vas a andar mucho tiempo.' < FLP 06 in61(704)/in >


This morphological construction also marks superlative constructions, as shown in (51) below.

	- 'Se llamaba Silverio el más chico.' < FLP 06 in61(286)/in >

Comparatives may also involve a non-copular verbal predicate, as the following examples in (52) show. The example in (52a) contains no gradable predicate.

(52) a. *ˈˈnè ko ˈwé maˈt͡ʃí ʃimè-a bioˈlîn ˈmò ˈkítara ˈnè=ko* 1sg.nom=emph *ˈwé* int *maˈt͡ʃí* know *seˈmè-a* play-prog *bioˈlîn* violin *ˈmò* 2sg.nom *ˈkítara* than 'I know how to play the violin better than you.' 'Yo se tocar más el violín que tu.' < GFM 09 3:101/el >

14.5 Comparatives

b. *nè ko ˈwé raˈsíra ˈnót͡ʃiri oˈlá beˈnè-a muˈhê ˈjûa ˈnè=ko* 1sg.nom=emph *ˈwé* int *raˈsíra* more *ˈnót͡ʃi-li* struggle-pst *oˈlá* cer *beˈnè-a* learn-prog *muˈhê* 2sg.nom *ˈjûa* than 'I struggle more to learn than you.' 'Batallo más para aprender que tu.' < GFM 09 3:101/el >

The Choguita Rarámuri comparative construction is of a type that may be characterized as a 'locational/adverbial comparative' Stassen (1984), as the markers of the standard of comparison are postpositions in the language. The construction consists of a single clause and the comparee NP can fulfill any grammatical function. In contrast, other Uto-Aztecan languages have been analyzed as having a 'particle comparative', including Comanche (Numic; Smalley 1953), Tohono O'odham (Tepiman; Zepeda 1983), and Tümpisa Shoshone (Numic; Dayley 1989) (see discussion in Stassen 2013).

# **15 Complex clauses and complex predication**

This chapter is concerned with the ways in which basic clauses may be extended through subordination, coordination and other devices in Choguita Rarámuri. The constructions addressed in this chapter are assumed to involve complex relationships between distinct events. These complex relationships are generally encoded cross-linguistically through subordination and coordination structures.

Three main types of subordinate clause constructions are discussed in the first section of this chapter: complement clauses in §15.1, adverbial clauses in §15.2, and relative clauses in §15.3). This classification depends on the main functions that subordinate clauses may have: to serve as the complement of a verbal predicate (complementation) or to modify a verb phrase or a whole clause in the periphery of a matrix clause (adverbial subordination). Choguita Rarámuri also exhibits other specialized complex clause constructions, including indirect causative constructions (§15.1.5) and complements of speech verbs (§15.1.7). A class of verbs, namely reportative verbs, exhibit switch reference marking in the complement clauses they head; these types of clauses are addressed in §15.1.6.

This chapter also addresses clausal coordination constructions, the type of syntactic constructions where two or more constituents that have equivalent status are combined into a larger syntactic unit and still have the same semantic relations with other surrounding elements (Haspelmath 2007). Choguita Rarámuri clause coordination constructions include conjunction (§15.4.1), disjunction (§15.4.2) and adversative conjunction (§15.4.3).

Another type of complex construction attested in the language addressed in this chapter involves verbal chaining structures (§15.5), where two or more predicates may convey that some events take place simultaneously or in a temporal sequence. Additionally, these constructions may imply associated semantic meanings related to chronological overlap and succession, such as cause and effect, result, and manner.

Finally, this chapter addresses constructions that may be analyzed as involving complex predicates, including light verb constructions (§15.6.1), auxiliary verb 15 Complex clauses and complex predication

constructions (§15.6.2), serial verb constructions (§15.6.3) and V-V incorporation constructions(§15.6.4).

### **15.1 Complement clauses**

Complementation involves syntactic configurations where a notional sentence or predication is an argument of a matrix predicate (Noonan 2007). Complement clauses in Choguita Rarámuri are identified as such if the complement clause fulfills a syntactic argument of the matrix clause. In most instances, the complement clause is a syntactic argument of the matrix predicate; in the case of Choguita Rarámuri, complement clauses involve a matrix transitive predicate whose object is a complement clause. Four major types of complement clauses can be identified on the basis of their morphosyntactic properties:

	- a. Finite complement clauses with complementizer (§15.1.1).
	- b. Interrogative complement clauses (§15.1.2).
	- c. Asyndetic finite verb complement clauses (§15.1.3).
	- d. Reduced complement clauses (§15.1.4).

In addition to these four types of complement clauses, Choguita Rarámuri has specialized constructions that involve complementation, including an indirect causative construction (§15.1.5), a reportative clause construction that features switch reference marking (§15.1.6), and complement clauses of speech verbs (§15.1.7).

### **15.1.1 Finite complement clauses with complementizer**

This type of complement clause has almost all of the properties of an independent clause and is introduced by a complementizer, the subordinating particle *ˈnápi* or its reduced form *ˈ(n)ápi*, which also is used in other types of subordinate clause constructions (§15.2) and relative clauses (§15.3). This type of complement clause is exemplified in (2). The matrix clause always precedes the complement clause (represented in square brackets in the examples below).

### 15.1 Complement clauses

(2) a. *ˈnè ko ˈá biˈlá ˈwé ˈtʃ͡ó ˈá maˈjêa ˈníli ˈápi ˈhípi roˈkò ˈá noˈrînima ˈnè=ko* 1sg.nom=emph *ˈá* aff *biˈlá* indeed *ˈwé* int *ˈtʃ͡ó* also *ˈá* aff *maˈjê-a* think-prog *ˈní-li* cop-pst *[ˈápi* sub *ˈhípi* today *roˈkò* night *ˈá* aff *noˈrîni-ma]* return-fut.sg 'I also think that he will return tonight.' 'Yo también pienso que va a venir hoy en la noche.' {LEL tx32:11:36.7} b. *ˈká ni maˈtʃ͡íki ˈnápu ˈtòoru ba ˈká=ni* neg=1sg.nom *maˈtʃ͡í-ki* know-pst.ego *[ˈnápi* sub *ˈtò-ru* take-pst.pass *ba]* cl 'I didn't know he had been taken.' 'No sabía que se lo habían llevado.' < BFL 09 1:39/el >

The finite verb of the complement clause appears in clause final position and is unrestricted in terms of TAM markers available in matrix clauses. In these particular examples, the complement verb is marked as future (2a) or may be passivized (2b). The matrix clause and complement clause can also be negated independently from each other: in (2b), only the main predicate is negated, but the complement is in the positive polarity, which points to the relative low degree of integration between the complement and the matrix clause in this type of complement clause structure.

### **15.1.2 Interrogative complement clauses**

Another type of complement clause involves a subordinate finite clause introduced by an interrogative expression (see §10.1.4). Several verbs of speech and cognition can take this type of complement, as shown in (3). The interrogative word occurs in clause initial position as in matrix interrogative clauses (§14.2).

(3) a. *ˈétʃ͡i riˈká biˈlá ko matʃ͡iˈnália ˈí ˈnàʔ ko ˈnáp ... tʃ͡ú riˈkám ˈká riˈwèeli tʃ͡aˈbèe ˈétʃ͡i* dem *riˈká* like *biˈlá=ko* indeed=emph *matʃ͡i-ˈnáli-a* know-desid-prog *ˈí* here *ˈnà=ko* here=emph *[tʃú* q how=dem *riˈká=mi* cop.irr *ˈká* was.like-pst *riˈwè-li* before *tʃ͡aˈbè]* 'She wanted to know how it (life) was here before.' 'Ella quería saber cómo era (la vida) antes aquí.' {SFH tx475:00:54.4}

### 15 Complex clauses and complex predication

b. *taˈmò ˈkútʃ͡uala ˈtʃ͡ó beˈnèma ˈtʃ͡ó ˈnà ˈtʃ͡útimi riˈká ˈnàwa noˈká ˈnà oˈtʃ͡êla noˈká tʃ͡aˈbè kiˈʔà ba taˈmò* 1pl.nom *ˈkútʃ͡ua-la* children-poss *ˈtʃ͡ó* also *beˈnè-ma* learn-fut.sg *ˈtʃ͡ó* also *ˈnà* then *[ˈtʃú=timi* how=2pl.nom *riˈká=m* indeed=dem *ˈnàwa* arrive.prs *no-ˈká* do-ger *ˈnà* then *oˈtʃ͡ê-la* grow.old-rep *no-ˈká* do-ger *tʃ͡aˈbè* before *kiˈʔà* long.ago *ba]* cl 'Our children are also going to learn how you all lived, how you grew up.' 'Nuestros hijos también van a aprender cómo crecieron antes ustedes, como vivían.' {SFH in61:00:46.7}, {SFH in61:00:49.2} c. *ˈmá ˈmí baʔaˈrîna ˈmí riˈkáatʃ͡i ˈmá aʔˈlá matʃ͡iˈsâa ˈká ˈtʃ͡i ˈníria ˈétʃ͡i ˈtáa ba ˈmá* already *ˈmí* dem *baʔaˈrîna* next.day *ˈmí* dem *riˈkátʃ͡i* perhaps *ˈmá* already *aʔˈlá* well *matʃ͡i-ˈsâ* know-cond *ˈká* cop.irr *[ˈtʃú* how *ˈnír-a* feel-prog *ˈétʃ͡i* def *ˈtá* small.one *ba]* cl 'And then already the next morning I think he knows how the child is feeling.' 'Ya a la mañana siguiente yo creo que va a saber qué siente el niño.' {SFH tx475:02:44.7}

As described above for declarative complement clauses with complementizers (§15.1.1), the matrix clause always precedes the complement clause. All properties identified for declarative finite complement clauses with complementizers hold for embedded questions, including the lack of restrictions in terms of TAM marking in the complement clause. As shown in the following example, negation also applies independently in the matrix clause and the complement clause: in (4), the negative particle in sentence initial position does not have scope over the complement clause.

(4) *ke piˈláti biˈlé matʃ͡iˈká moˈtʃ͡íli taˈmò ˈká tʃ͡è wiˈlé ˈtʃ͡ó matʃ͡iˈwá aʔˈlì ko ba ˈpîri ˈnà ta naˈjúka suwiˈjá ba ke* neg *biˈlá=ti* indeed=1pl.nom *biˈlé* one *matʃ͡i-ˈká* know-ger *moˈtʃ͡í-li* sit.pl-pst *taˈmò* 1pl.nom *ˈká* because *ˈtʃ͡è* neg *biˈlé* then *ˈtʃ͡ó* also *matʃ͡i-ˈwá* know-mpass *aʔˈlì=ko* and=emph *ba* cl *[ˈpîri* what *ˈnà* dem *ta* 1pl.nom *naˈjú-ka* be.sick-ger

15.1 Complement clauses

*suwi-ˈjá* finish.off-prog *ba]* cl 'We were without knowing, us, because then it wasn't known yet what it was (that was) making us sick and dying (lit. 'being finished off').' 'Pues nosotros estábamos sin saber, porque en ese tiempo todavía no se sabía nada de qué nos enfermábamos y nos moríamos.' {LEL tx372:00:41.4}, {LEL tx372:00:47.4}

### **15.1.3 Asyndetic finite verb complement constructions**

Another type of complement clause in Choguita Rarámuri involves asyndetic subordination, where the matrix and subordinate clause are juxtaposed without any overt complementizer. These constructions are exemplified in (5).


### 15 Complex clauses and complex predication

In asyndetic finite verb complementation, the complement clause may have a different subject from the matrix clause. This is exemplified in (5b–d).

Utterance predicates may also encode complements through the asyndetic finite verb construction, as exemplified in (6).

(6) a. *waˈbê biˈlá kiˈʔà ˈníla ... ˈhê biˈlá riˈkám aˈnía aˈní biˈsá ˈtʃ͡é tʃ͡oˈkêami naˈhùla ˈrá ˈnà kaˈwì waˈbê* int *biˈlá* indeed *kiˈʔà* long.ago *ˈní-la* cop-rep *ˈhê* it *biˈlá* indeed *riˈká=mi* like=dem *aˈní-a* say-prog *aˈní* say.prs *[bi-ˈsá* three-mltp *ˈtʃ͡é* again *tʃ͡oˈkê-ame* begin-ptcp *naˈhù-la* fall.down-rep *ru-ˈwá* saympass *ˈnà* this *kaˈwì]* world 'They say that long, long ago the world began to fall apart three times.'<sup>1</sup> 'Dicen que mucho antes tres veces se empezó a caer este mundo.' {SFH tx43:11:18.2} b. *aʔˈlì ˈhê aˈníli: "ˈnè ko ˈá raˈè ˈwé biˈlá ˈnà ... ˈwé biˈlá wiʰˈkâ ˈnà ... aʔˈlì* and *ˈhê* it *aˈní-li:* say-pst *[ˈnè=ko* 1sg.nom=emph *ˈá* aff *raˈè* know *ˈwé* int *biˈlá* indeed *ˈnà* dem *ˈwé* int *biˈlá* indeed *wiʰˈkâ* many *ˈnà]* dem 'and then he said: "I do know (where he lives), (with) many ..."' 'y entonces dijo: "Yo sí conozco (donde vive), (con) muchos..." {LEL tx32:02:29.0} c. *aʔˈlì ˈhe aˈnèli: "ˈwé saˈpù aˈsísika! ˈpîli tʃ͡uˈkú naˈʔî?" aʔˈlì* and *ˈhe* it *aˈn-è-li* say-appl-pst *[ˈwé* int *saˈpù* hurry *aˈsísi-ka]* get.up-imp.sg *[ˈpîli* what *tʃ͡uˈkú* to.be.bent.prs *naˈʔî]* here 'And then she told him: "Hurry, get up! What is that sitting here?'

'Y luego le dijo: "¡Levántate pronto! ¿Qué hay (está) aquí?"' {LEL tx5:00:52.7}

As shown in these examples, both indirect (6a) and direct (6b–c) speech is introduced through this type of construction. More details about the properties of complements of speech verbs are provided in §15.1.7.

<sup>1</sup> In this sentence, the adverb *kiˈʔà* 'long ago' modifies the complement clause.

15.1 Complement clauses

### **15.1.4 Reduced complement clauses**

A fourth type of complement clause involves a reduced complement: there is argument sharing, and complement-taking verbs with this complement clause type take only same subject complements. These verbs are modal and phase verbs and other verbs with similar semantic characteristics. There appears to be no constraint on the transitivity or valence of the complement verb, and any type of subject is eligible for argument sharing. The complement clause is also restricted in terms of the TAM specification of the complement clause, and marked only as progressive or medio-passive. This type of complement clause is exemplified with complements of the phase verbs *ˈtʃ͡ôta* 'begin' (7) and *suˈní* 'finish' (8).

(7) a. *ˈmá ˈtʃ͡ôtali aˈwìa ˈmá* already *ˈtʃ͡ôta-li* begin-pst *[aˈwì-a]* dance-prog 'They started dancing.'

'Ya empezaron a bailar.' {JLG co1237:08:12.1}

	- b. *suˈnísa naʔˈpôa paˈtʃî ... muˈnî ko ˈro suˈní-sa* finish-cond *[naʔˈpô-a* weed-prog *paˈtʃî]* corn *muˈnî=ko* beans=emph *ˈru* say.prs 'When they finish weed for (planting) corn, then the beans.' 'Cuando terminan de escardar el maíz, después el frijol.' {MDH co1136:03:33.2}
	- c. *suˈnísa ˈʔwîa ˈmá ˈtʃ͡étʃ͡o ku wiˈlíram ke wasaˈráa suˈní-sa* finish-cond *[ˈʔwî-a]* harvest-prog *ˈmá* already *ˈtʃ͡é* again *ˈtʃ͡ó* again *ku* rev

### 15 Complex clauses and complex predication

*wiˈlí-r-ame* stand-pst.passs-ptcp *ke* cop.impf *wasaˈrá* plow.prs 'When they finish harvesting, then he'd start plowing again.' 'Cuando terminaban de pizcar luego se ponía a barbechar otra vez.' {LEL tx130:04:16.7}

Both *ˈtʃ͡ôta* 'begin' and *suˈní* 'finish' precede the complement clause, as in the other complement types surveyed so far. The verb in the complement clause is marked present progressive (7a, 8) or medio-passive (7b).

Constructions with the modal verb *oˈmèra* 'can' also involve a reduced complement clause with the same properties observed for phase predicates like 'begin' and 'finish' exemplified above. As shown in (9), the ordering main verb – complement clause is also attested in complements of this modal verb, and the subject argument is shared by the matrix and complement verbal predicates.


### 15.1 Complement clauses

The example in (9d) shows that the main verb and the subordinate clause need not be adjacent in these complement clauses, as the main verb may be followed by an adverb modifying it (*ˈtʃ͡ó* 'also'). The examples below show reduced complement clauses where the shared subject argument (10a), other adverbs (10b), and object noun phrases (10c) may be realized inside the complement clause.

(10) a. *aʔˈlì biˈlá ˈnà ˈá riˈká ke oˈmèaki ti roˈplân riˈkîina aʔˈlì* and *biˈlá* really *ˈnà* then *ˈá* aff *riˈká* like *ke* neg *oˈmèa-ki* be.able-ego *[ti* plane *roˈplân* plane *riˈkî-na]* down-vblz.prs 'Ya no pudo bajar el avión.' 'The plane could not go down." {SFH tx12:02:10.7} b. *aʔˈlì ke me oˈmèra paˈtʃ͡ána siˈmíra aʔˈlì* and *ke* neg *me* almost *oˈmèra* can.prs *[paˈtʃ͡á-na* inside-all *siˈmíra]* pass.prs 'But he couldn't get all the way inside (lit. 'pass inside')' 'Pero no pudo meterse todo adentro (lit. 'pasar adentro')' {LEL tx177:00:45.2} c. *aʔˈlì ke oˈmèrali ˈnà ku matʃ͡iˈbûa moˈʔôla aʔˈlì* and *ke* neg *oˈmèra-li* can-pst *[ˈnà* prox *ku* rev *matʃ͡i-ˈbû-a* outside-vblz-prog *moˈʔô-la]* head-poss

'And he couldn't take his head out.'

'Y no pudo sacar la cabeza.' {LEL tx177:01:48.8}

There are also instances where the shared subject argument of the main and complement clause may be postposed, as shown in (11).<sup>2</sup>

(11) *ke ˈtâsi ˈtʃ͡ó oˈmèri biˈhí raˈʔìtʃ͡a ˈétʃ͡i ko ke* neg *ˈtâsi* neg *ˈtʃ͡ó* also *oˈmèra* can.prs *[biˈhí* yet *raˈʔìtʃ͡a]* speak.prs *ˈétʃ͡i=ko* dem=emph 'She can't talk yet that one.' 'Todavía no puede hablar esa.' {JLG co1235:06:03.1}

Finally, these phase and modal complement-taking predicates are also attested without an overt complement, i.e., as intransitive predicates. This is shown in (12).

<sup>2</sup>Morales Moreno (2016) analyzes similar constructions in Rochéachi Rarámuri as instances of right-hand dislocation to encode a topical argument. Given (11) involves a third person argument, an expected co-referential pronominal form in the clause is not attested.

### 15 Complex clauses and complex predication

As shown in these examples, the lexical meaning of these predicates is the same, whether they host a complement clause or not.

	- *ˈmá* already *biˈlá=ni* indeed=1sg.nom *suˈní-ma* finish-fut.sg *ˈa* aff *riˈké* would *pa* cl 'I was about to finish.'

'Ya iba a acabar yo.' {MDH co1137:09:43.0}

	- aff

'Who knows if he'd be able to (take him on)? I think he can (take him on), take him.'

'¿Quién sabe si lo podrá? Yo creo que si lo puede, a lo mejor si puede llevárselo.' {MDH co1136:08:57.2}

Villalpando-Quiñonez (2019) describes several auxiliary verbs in Norogachi Rarámuri, which he proposes have been recently grammaticalized from matrix predicates and encode aspectual meanings in auxiliary verb constructions. Two of such verbs have cognates in Choguita Rarámuri that are identified here as involving complementation instead: (i) constructions with the verb *ˈtʃ͡ôta* 'begin'; and (ii) constructions with the verb *suˈní* 'finish'. It is argued here that the phase and modal verbs in Choguita Rarámuri described in this section still exhibit the properties of full lexical predicates that take reduced complement clauses and have not yet undergone any process of auxiliation. Choguita Rarámuri auxiliary verb constructions are described in §15.6.2.

### **15.1.5 Indirect causative construction**

Choguita Rarámuri possesses a distinct class construction encoding indirect causation. The contrast between direct and indirect causation in Choguita Rarámuri

### 15.1 Complement clauses

is encoded morphosyntactically: a co-lexicalized structure entails physical contact between the causer and the causee (13a), while a periphrastic construction (exemplified in (13b)) entails that the causee may act on its own, with varying degrees of volition.

(13) a. *niˈhê toˈwí koˈʔátili niˈhê toˈwí*

1sg.nom boy *koˈʔá-ti-li* eat-caus-pst 'I fed the boy' (forcing the spoon into his mouth).' 'Hice comer al niño (dándole con una cuchara en la boca).'

b. *niˈhê toˈwí koˈʔánula nuˈlèki niˈhê* 1sg.nom *toˈwí* boy *koˈʔá-nula* eat-order *nuˈl-è-ki* order-appl-pst.ego 'I ordered the boy to eat.' 'Hice comer al niño (le ordené que comiera).'

Indirect causative constructions in Choguita Rarámuri involve a periphrastic construction in which a main jussive predicate takes the caused event as a complement. Some properties that characterize this structure are: (i) the lower verb is additionally marked with the jussive verbal affix *nula* 'order, command' deriving a co-lexicalized structure within the complement; (ii) although there are two causative verbs, the causer is expressed only once, as a core argument of the matrix predicate, i.e. the two jussive verbs share the actor. This is exemplified in (14).<sup>3</sup>

(14) a. *ˈémi taˈmí aˈnèki niˈhê ˈtònula ˈémi* 2pl.nom *taˈmí* 1sg.acc *aˈn-è-ki* tell-appl-pst.ego *[niˈhê* 1sg.nom *ˈtò-nula]* take-order 'You all told me to take it.' 'Ustedes me hicieron que me lo llevara' b. *ˈhuâni ko taˈmí toˈlí ˈpónula nuˈlè ˈhuâni=ko* Juan=emph *taˈmí* 1sg.acc *[toˈlí* chicken *ˈpó-nula]* pluck-order *nuˈl-è* order-appl.prs 'Juan makes me pluck the chicken.' 'Juan me hace que desplume pollos.'

<sup>3</sup> In (14d), there are three clauses: the matrix clause headed by *maˈjê* 'believe', a complement clause headed by the jussive predicate *nuˈlè* 'to order', and the lower clause headed by the predicate *miˈʔà* 'to kill'.

### 15 Complex clauses and complex predication


In these constructions, the causee may be expressed once, e.g. as an accusative argument in the main clause in (14b), or twice, as an accusative in the main clause and nominative within the lower clause in (14a). The lower verb cannot bear any inflection marker (though derivational morphology is possible as in (14c)) nor can it be modified by temporal adverbs or negation. The position of the dependent unit also varies, i.e. it may appear extraposed to the right (14a), but the sentence can display the canonical word order of main clauses, when the unit is embedded and followed by the matrix predicate or causee in clause-final position, as in (14b– c).

While any jussive predicate may be the matrix predicate in indirect causatives, giving different shades to the manipulative force, only the predicate *nula* may appear co-lexicalized with the lower predicate. This is exemplified in (15).


15.1 Complement clauses

c. *aʔˈlì tʃ͡iˈhônsa ko ˈmá ˈpé oˈt͡ʃérisa ko nuluˈrîa baˈʔwí tuˈwúnula aʔˈlì* and *tʃ͡iˈhônsa=ko* then=emph *ˈmá* already *ˈpé* little *oˈt͡ʃéri-sa=ko* grow-cond=emph *nulu-ˈrîa* order-hab.pass *[baˈʔwí* water *tuˈwú-nula]* bring-order 'And then when [they, the children] already grow a little, they were sent to bring water.' 'Y ya cuando crecen un poquito los mandan a traer agua.' {BFL tx48:00:41.4}, {BFL tx48:00:45.7}

The following example (in (16)) shows that a clause headed by a jussive predicate may have more than one subordinate clause, each of which will have *ˈnula* colexicalized with each lower predicate.

(16) *ˈwàas ˈmá niˈsènula ˈtʃ͡îba ˈmá boˈrêko ˈmá niˈsènula nuluˈrîa [ˈwàsi* cows *ˈmá* also *niˈsè-nula]* shepherd-order *[tʃ͡îba* goats *ˈmá* also *boˈrêko* sheep *ˈmá* also *niˈsè-nula]* shepherd-order *nulu-ˈrîa* order-hab.pass 'They are sent to shepherd cows, too, to take care of goats and sheep.' 'Los mandan a cuidar vacas también, a cuidar chivas, borregos también.'

{BFL tx48:00:54.8}

The periphrastic structure is often used as an imperative, where the speaker verbally commands the addressee to instruct a third participant (present or absent) to perform an action, as below. Speakers use this formula often in every-day interactions, and they report it as a formula of polite request. This is exemplified in (17).

(17) *gaˈbriêla nuˈlè biˈlé tʃ͡imoˈrí ʃiˈrûnula gaˈbriêla* Gabriela *nuˈl-è* order-appl.imp *[biˈlé* one *tʃ͡imoˈrí* squirrel *ʃiˈrû-nula]* hunt-order 'Tell Gabriela to hunt a squirre.l' 'Dile a Gabriela que cace una ardilla'.

Mountain Guarijío is also documented to have indirect causative structures, including a periphrastic structure with double encoding of the causative (Miller 1996), as shown in (18c).

### 15 Complex clauses and complex predication

	- a. *noʔó koʔkónulani teurusio noʔó* 1sg.acc *koʔkó-nula-ni* eat-order-pres *teurusio* Tiburcio 'Tiburcio commands me to eat.' 'Tiburcio manda que yo coma.'
	- b. *nulani teurusio noʔó koʔkómi ruhka nula-ni* order-pres *teurusio* Tiburcio *noʔó* 1sg.acc *koʔkó-mi* eat-adv *ruhka* adv 'Tiburcio commands that I eat.' 'Tiburcio manda que yo coma.'
	- c. *nulani teurusio noʔó koʔkónurega nula-ni* command-pres *teurusio* Tiburcio *noʔó* 1sg.acc *koʔkó-nur-e-ga* eat-order-appl-ptc.prs 'Tiburcio commands that I eat (lit. that/while I am eating).' 'Tiburcio manda que yo coma.'

As shown in these examples, the predicate *nula/nuˈle*, cognate of Choguita Rarámuri *nuˈla*, may appear: (i) co-lexicalized with a matrix predicate (18a); (ii) as the matrix predicate in a complementation construction (18b); or (iii) both as the matrix predicate and co-lexicalized with the lower verb deriving a co-lexicalized structure within the complement (18c). Only the latter strategy is attested in Choguita Rarámuri.

### **15.1.6 Switch reference in reportative clauses**

Switch-reference is a phenomenon where a set of morphemes associated with the juncture of two clauses encodes whether a prominent argument in each clause is co-referent. Typically, co-referent arguments are subject arguments, although in some switch-reference systems the tracked co-referent arguments are not the subjects of their clauses (McKenzie 2015). A phenomenon with an areal diffusion, switch reference is attested in the West and Southwest of North America (Jacobsen 1967; Jacobsen Jr 1983; McKenzie 2015). Within the Uto-Aztecan language family, switch reference has been documented in Southern Paiute (Numic; Sapir 1930), Comanche (Numic; Charney 1993), Hopi (Jeanne & Hale 2019), Tohono O'odham (Tepiman; Hale 1983; Hale 1992), Cupeño (Takic; Hill 2005), and closely related Mountain Guarijío (Tara-Guarijío; Miller 1996).

### 15.1 Complement clauses

In Choguita Rarámuri, switch reference is attested in the reportative construction, an evidential construction that indicates that the speaker's source of information is not direct, but hearsay. Reportative constructions involve a matrix clause with a speech predicate and a complement clause, the content of the reported event. It involves a stress-neutral productive marker, and is added to the verb of the complement clause. When the notional subjects are correferential, the dependent verb is marked for tense/aspect and the epistemic *-o* suffix (19a) (this suffix is a replacive suffix that triggers vowel deletion on the base onto which it attaches). When the notional subjects are not correferential, the dependent verb suffixes the reportative *-la* suffix, which indicates disjoint reference (19b).


More examples of reportative constructions with same and different referents are provided in (20) and (21), respectively.

(20) Switch reference: same reference

a. *ˈá biˈlá ko aˈní ˈmâgre neˈhê amaˈtʃíkoro ˈruá ˈá* aff *biˈlá=ko* really=emph *aˈní* say.prs *ˈmâgre* nuns *neˈhê* 1sg.nom *amaˈtʃí-ki-li-o* pray-appl-pst-ep *ru-ˈwá* say-mpass 'The nuns<sup>i</sup> say that (they<sup>i</sup> ) prayed for me.' 'Las monjas<sup>i</sup> dicen que (ellas<sup>i</sup> ) me rezaron.' b. *maˈnuêli ko ˈwé biˈlá riˈkúlo ˈrú maˈnuêli=ko* Manuel=emph *ˈwé* int *beˈla* really *riˈkú-li-o* get.drunk.sg-pst-ep *ˈrú* say.prs

'Manuel<sup>i</sup> says (he<sup>i</sup> ) got drunk.'

'Manuel<sup>i</sup> dice que (él<sup>i</sup> ) se emborrachó.'

### 15 Complex clauses and complex predication

### (21) Switch reference: different reference

a. *á biˈlá oˈkám tʃ͡aˈnía ne ka ˈhém isiˈmâtara ruˈa tʃ͡aˈbè á* aff *beˈla* really *oˈká=mi* many=dem *tʃ͡aˈní-a* sound-prs *ne* int *ka* ka *ˈhémi* here *i-siˈmâta-ra* pl-pass.pl-rep.ds *ru-ˈwá* say-mpass *tʃ͡aˈbè* before 'Many people<sup>i</sup> say that they<sup>j</sup> used to pass through here long time ago.' 'Muchas personas<sup>i</sup> dicen que por aquí pasaban<sup>j</sup> mucho antes.' {LEL tx223:04:17.6} b. *tʃ͡iˈnà ba ˈétʃ͡i biˈlá tòola ˈruá aliˈwâ-la ba ˈétʃ͡i* dem *ˈnà* there *ba* cl *ˈétʃ͡i* dem *biˈlá* indeed *tò-la* take.pst.pass-rep.ds *ˈru-wá* say-mpass *aliˈwâ-la* soul-poss *ba* cl 'They<sup>i</sup> say that that one<sup>j</sup> got his soul stolen there.'

'Cuentan<sup>i</sup> que a ese<sup>j</sup> ahí le llevó el alma.' {BFL tx\_muerto:0:35.8}

This switch-reference system is restricted to reportative constructions, as it is not generalized to all constructions involving dependent clauses in Choguita Rarámuri. For Western Tarahumara, Burgess (1984) argues that switch reference is attested in adverbial clauses with the suffixes *-sa* and *-so*, glossed as 'when/having' (1984: 137) and cognate with the Choguita Rarámuri conditional *-sa* suffix. There is no evidence of switch reference in adverbial clauses of this type in Choguita Rarámuri (see §15.2.1 below).

### **15.1.7 Direct speech complements**

Direct speech complementation involves a matrix clause with a verb of speaking (e.g., 'say') and a complement clause with reported speech as a direct quote, where the verb is finite and the clause is introduced without a complementizer. The following examples illustrate the properties of direct speech complements, where the complement may follow (22a) or precede (22b–c) the matrix clause. Prosodically, the reported clause constitutes its own Intonational Phrase (IP), and has the canonical intonation of the sentence type of the complement clause, e.g., declarative (22a–b) or interrogative (22c) (for details of the prosodic properties od dedclaratives and interrogatives, see §6.2, as well as §14.1 and §14.2).

15.2 Adverbial clauses

(22) a. *ˈhê beˈla aˈnía ruˈwá tʃ͡aˈbôtʃ͡i: "ah, si aʔˈlá ka ˈmí raˈʔìa ka ˈmí pa ˈnàri baˈhîwa ba" ˈhê* dem *beˈla* really *aˈní-a* say-prog *ru-ˈwá* say-mpass *tʃ͡aˈbôtʃ͡i* mestizo *[a* ah *si* int *ne* int *aʔˈlá* good *ka=mi* emph=dem *raˈʔì-a* like-prog *ka=mi* emph=dem *pa* cl *ˈnàri* that *baˈhî-wa* drink-mpass *ba]* cl 'And the *mestizo* was saying: "ah, it is so good and tasty to drink this".' 'Así dijo el mestizo: "¡Ay, cómo está rico tomar esto!".' < SFH 06 choma(10) > b. *"á si ˈném raˈʔìa ˈkam pa ˈnàri tʃ͡oʔˈmá ko ba" aˈnía ˈrá [a* ah *si* int *ˈné=mi* int=dem *raˈʔìa* tasty *ˈka=mi* emph=dem *pa* cl *ˈnàri* prox *tʃ͡oʔˈmá=ko* snot=emph *ba]* cl *aˈní-a* say-prog *ru-ˈwá* say-mpass "'This snot is really tasty", he would say.' "'Está muy rico este moco", así decía.' < SFH 06 choma(15) >

c. *"okoˈká ropaˈtʃ͡í ba?" ˈhê biˈlá ko aˈnèla ˈrá ˈétʃ͡i ko ˈkôasa aˈlé ba [oko-ˈká* hurt-ger *ropaˈtʃ͡í* belly *ba]* cl *ˈhê* it *biˈlá=ko* indeed=emph *aˈn-è-la* say-appl-rep *ru-ˈwá* say-mpass *ˈétʃ͡i=ko* dem=emph *ˈkôa-sa* feed-cond *aˈlé* dub *ba* cl "'Does your belly hurt?" asked those who were feeding him.' "'Te duele el estómago?" le preguntaban los que le dieron de comer.' {SFH tx43:10:01.5}

Indirect speech complements are encoded through the reportative clause construction described in §15.1.6 above.

### **15.2 Adverbial clauses**

This section addresses adverbial clauses, which constitute a major type of nonmatrix clause, together with complement clauses and relative clauses. While a complement clause functions as an argument of a higher predicate (Cristofaro 2005, Noonan 2007), adverbial clauses provide contextual information about the event described in the main clause, syntactically acting as a modifier of verb phrases or entire clauses (Thompson et al. 1985). The following subsections describe the different sub-types of adverbial clauses in Choguita Rarámuri in terms

### 15 Complex clauses and complex predication

of their function and morphosyntactic properties. Six main types of adverbial clause types are identified in this language: (i) conditional clauses (§15.2.1); (ii) purpose clauses (§15.2.2); (iii) reason clauses (§15.2.3); (iv) locative clauses (§15.2.4); (v) temporal clauses (§15.2.5); and (vi) manner clauses (§15.2.6).

### **15.2.1 Conditional clauses**

Conditional clauses express a conditional relationship between two events and share the following structural properties in Choguita Rarámuri: (i) the protasis (condition clause) generally precedes the apodosis (description of the actual/ potential outcome); (ii) the verb of the protasis clause is obligatorily marked with the stress-shifting conditional suffix *-sâ*; 4 and (iii) the protasis is optionally marked by *ka* in clause final position, a morpheme that is used as an irrealis copula in other contexts, but in conditional clauses marks the conditional clause as not yet realized. This is exemplified in (23) (protasis clauses are marked with squared brackets).

(23) a. *aʔˈlì ˈmá aʔˈlì ko ˈmá aʔˈlá waʃiˈsâa ka ku sipiˈbáti ka ˈnà [aʔˈlì* and *ˈmá* already *aʔˈlì=ko* afterwards=emph *ˈmá* already *aʔˈlá* well *wasi-ˈsâ* cook-cond *ka]* irr *ku* rev *sipi-ˈbá-ti-ka* be.cold-inch-caus-ger *ˈnà* dem 'And when it is cooked, you cool it off again.' 'Y ya cuando se cose se enfría otra vez.' {LEL tx68:01:30.6} b. *aʔˈlì ˈmá biˈkiá roˈkò ˈnà tʃ͡ereˈbása ka ˈmá kosˈtâaltʃ͡i muˈtʃ͡ûwuka [aʔˈlì* and *ˈmá* already *biˈkiá* three *roˈkò* night *ˈnà* dem *tʃ͡ereˈbá-sa* spend.night-cond *ka]* irr *ˈmá* already *kosˈtâaltʃ͡i* sac *muˈtʃ͡ûwu-ka* put.in.pl-ger 'And when three nights have passed, you put it in a sac.' 'Y ya cuando pasan tres noches, se pone en un costal.' {LEL tx68:00:38.0}, {LEL tx68:00:40.4} c. *niʔˈwísa ka ˈmá ukuˈmêa aˈlé [niʔˈwí-sa* to.be.lightning-cond *ka]* irr *ˈmá* already *uku-ˈmêa* rain-fut.sg *aˈlé* dub 'When there is lightning, then it will rain.' 'Cuando relampaguea, ya empieza a llover.' {MFH el1318:05:56.4}

<sup>4</sup> Felix Armendáriz (2006) reports thar closely-related River Guarijío has a conditional suffix *-so*, while Miller (1996) describes that Mountain Guarijío has adverbial suffixes *-sa*, *-sao* that may be translated as 'when'.

15.2 Adverbial clauses

d. *aʔˈlì ˈétʃ͡i kaiˈnâsa ˈmá tʃ͡oˈnà moˈʔíram ˈkúutʃ͡i ba ˈne [aʔˈlì* and *ˈétʃ͡i* dem *kaiˈnâ-sa]* finish-cond *ˈmá* already *ˈétʃ͡i* dem *ˈnà* there *moˈʔí-r-ame* go.in.pl-pst.pass-ptcp *ˈkútʃ͡i* children *ba* cl *ˈne* ne 'And when that one was finished, the children went there.' 'Y ya cuando se terminó esa, ya ahí entraron los niños.' {SFH tx12:02:40.6}

While the verb in the protasis clause is always marked with the conditional mood suffix *-sâ*, the verb of the apodosis clause may be marked with a limited set of TAM and other markers: in (23a–b), as part of a procedural text, the suffix *-ká* in the apodosis has a gerundive meaning (see §15.5 for discussion of constructions where the suffix *-ká* is used in chaining structures); alternatively, the apodosis may be marked with future tense (23c) or it may be nominalized with a particpial marker (23d).

The examples in (24) show that, while the protasis generally precedes the apodosis, the opposite order is also attested.

(24) a. *suˈnù niˈhîbo ˈlá, waˈrî niˈhîsa ba suˈnù* corn *niˈhî-bo* give-fut.pl *oˈlá* cer *[waˈrî* palm.basket *niˈhî-sa* give-cond *ba]* cl 'We give corn when they give palm baskets.' 'Damos maíz cuando dan waris (canastas de palma).' {MDH co1140:00:20.6} b. *ˈpé ˈnabi ˈlá awiˈmêa ˈlé, ˈhípi ˈmá baˈtùsa ka ba ˈpé* just *ˈna=bi* prox=just *oˈlá* cer *awi-ˈmêa* dance-fut.sg *aˈlé* dub *[ˈhípi* now *ˈmá* already *baˈtù-sa* grind-cond *ba]* cl

> 'This one will dance, I think, now that we grind the corn (for the corn beer).'

'Este va a bailar yo creo ahora que ya muélamos (el tesgüino).' {MDH co1140:15:26.8}

c. *uˈmûa ruˈwá naˈmûti ˈhêmi ˈsírsa, ˈwàsi, kaˈbâjo uˈmûa* all.kinds *ru-ˈwá* throw-mpass *naˈmûti* things *[ˈhêmi* over.there *ˈsíri-sa]* pass.pl-cond *ˈwàsi* cows *kaˈbâjo* horses 'All kinds of things would be thrown when they would pass by over there, to the cows, the horses.'

*ka* irr

### 15 Complex clauses and complex predication

'De ahí les tiraban a los animales cuando pasaban por alˈla, a las vacas, los caballos.' {LEL tx109:01:43.8}

While the order of the protasis and apodosis is relatively flexible, the apodosis surfaces with a subset of TAM suffixes available in the language. As shown above, the apodosis may also be marked as resultative or it may be nominalized, pointing to the morphosyntactic integration of clauses in the conditional construction. In addition to this, the two clauses are prosodically and semantically integrated: protasis and apodosis have a single prosodic contour, and a clause bearing conditional marking and having the conditional function may not stand as a complete utterance.

### **15.2.2 Purpose clauses**

Adverbial clauses that express purpose are coded by a clause in which the verb is the minimally required element. The main clause encodes an event that is performed in order to bring about the event depicted in the purpose clause. The subordinate clause may be used with a purposive function even in the absence of dedicated morphology to indicate this function. Purpose clauses may appear at the right periphery of the main clause and are exemplified in (25).

(25) a. *aʔˈlì tamuˈhê ˈmá … ˈmá aʔˈlì raˈwé ˈmáti napaˈwí hiˈrâmia aʔˈlì* and *tamuˈhê* 1pl.nom *ˈmá* already *aʔˈlì* later *raˈwé* day *má=ti* already=1pl.nom *napaˈwí* gather *[hiˈrâ-mi-a]* bet-mot-prog 'And then we gather that day to bet.' 'Entonces nosotros ese día ya nos juntamos para apostar.' {LEL tx19:01:32.3} b. *kiˈsâra koʔˈpôo ba! kiˈsâra* cook.imp.sg *[koʔ-ˈpô* eat-fut.pl *ba]* cl 'Cook so you all eat!' '¡Guisa para que coman!' {MDH co1136:00:13.4} c. *ku napaˈbûa ˈmá ku kaˈtêupo aˈlé pa ku* rev *napaˈbû-a* gather-prog *[ˈmá* already *ku* rev *kaˈtêwi-po* put.away-fut.pl *aˈlé* dub *pa]* cl 'They are going to gather them to put them away.' 'Ya lo van a juntar para guardarlos.' {JLG co1234:16:22.1}

15.2 Adverbial clauses

d. *aʔˈlì ˈnà ˈwé biˈláni aˈnè ˈnà riˈká aˈsâ ˈtʃ͡ó koˈbísi ˈkôbia ˈtʃ͡ó roˈwéam ba ˈwé iˈwêri ˈnà ˈmàmam pa aʔˈlì* and *ˈnà* dem *ˈwé* int *biˈlá=ni* really=1sg.nom *aˈn-è* say-appl *ˈnà* dem *riˈká* like.that *aˈsâ* sit.sg.tr *ˈtʃ͡ó* also *koˈbísi* pinole *ˈkôbi-a* give.pinole-prog *ˈtʃ͡ó* also *roˈwé-ame* women.race-ptcp *ba* cl *[we* int *iˈwêri* strong *ˈnà* dem *ˈmà-m-ame* run.sg-m-ptcp *pa]* cl

'And then I tell them that I will give pinole to the runner, so that she will run faster.'

'Y luego les digo que le estoy dando pinole a la corredora para que corra más recio.' {LEL tx19:04:02.7}

Purpose clauses may be marked with a subset of TAM specifications, including present (25a) or future (25b–c), or they may be nominalized (25d).

While purpose clauses are generally placed at the right edge of the main clause as in (25), there are also purpose clauses that precede the main clause, as shown in (26).

(26) *onoˈlâ ku ˈáʃia aˈkíbo [ono-ˈlâ* father.male.ego-poss *ku* rev *ˈá-si-a]* look.for-mot-prog *aˈkíbo* left 'He left to look for her dad.' 'Se fue a buscar a su papá.' < GFM 09 2:25/el >

Purpose clauses may also be introduced by a subordinator, *ˈnápu* or *ˈkíti*, as exemplified in (27). These subordinating morphemes, as discussed in this chapter, may introduce different types of subordinate clauses.

(27) a. *aʔˈlì biˈláti ˈʔá ˈjéntʃ͡o ˈá miʔˈlípi ˈlé pa, biˈlé toˈlí ba aʔˈlì tʃ͡iˈhônsa reˈmênipi ˈléti ˈûa ba ˈnápu riˈkáti ˈâbo ˈétʃ͡i aʔˈlì* and *biˈlá=ti* indeed=1pl.nom *ˈá* aff *ˈjén* aff *ˈtʃ͡ó* also *ˈá* aff *miʔˈlí-pi* kill.pl-irr.pl *aˈlé* dub *pa* cl *biˈlé* one *toˈlí* chicken *ba* cl *aʔˈlì* and *tʃ͡iˈhônsa* then *reˈmêni-pi* make.tortillas-irr.pl *aˈlé=ti* dub=1pl.nom *ˈûa* with *ba* cl *[ˈnápu* sub *riˈká=ti* like.that=1pl.nom *ˈâ-bo* give-fut.pl *ˈétʃ͡i]* dem 'And we'll also kill a chicken and then make some tortillas so we can give it to them.'

### 15 Complex clauses and complex predication

'Y también vamos a matar una gallina y luego hacerle unas tortillas para que le demos.' {SFH tx475:04:20.9}

b. *aʔˈlìmi noˈkèema aʔˈlá ˈnàri ˈkíti ko ke muˈjâma aʔˈlì=mi* then=2sg.nom *noˈk-è-ma* move-appl-fut.sg *aʔˈlá* well *ˈnàri* then *[ˈkíti=ko* sub=emph *ke* neg *muˈjâ-ma]* become.rotten-fut.sg 'And then you will stir it well (often) so that it won't rot.' 'Y lo vas a mover bien (cada rato) para que no se pudra.' {BFL tx60:00:43.1}

As with other subordinate clause constructions, a single intonation contour applies to the complex as a whole.

### **15.2.3 Reason clauses**

While purpose clauses encode a motivating event that is unrealized at the time of occurrence of the main event, reason clauses encode a motivating event that is realized at the time of occurrence of the event encoded by the main clause (Thompson et al. 1985). In Choguita Rarámuri, reason clauses are marked as past tense (28a), mediopassive (28b) or other TAM distinctions that encode that the proposition encoded by the verbal predicate is realized. As shown in the examples in (28), reason clauses also appear at the right periphery of the main clause and are optionally introduced by the subordinator *ˈkíti* (28c).

(28) a. *ˈá ˈjén ˈá, aʔˈlá kaˈjèna tʃ͡aˈbè, ˈwé ˈpé ˈwé aʔˈlá uˈkúli ˈlìna tʃ͡aˈbè ko ba ˈá* aff *ˈjén* aff *ˈá* aff *aʔˈlá* well *kaˈjèna* yield.harvest.prs *tʃ͡aˈbè* before *[ˈwé* int *ˈpé* just *ˈwé* int *aʔˈlá* well *uˈkú-li* rain-pst *aʔˈlìna* because *tʃ͡aˈbè=ko* before=emph *ba]* cl 'Yes, there was indeed a good yield (harvest) because it did rain a lot before.' 'Sí, sí se daba muy bien (la cosecha) porque sí llovía mucho antes.' {ME in484:00:38.5} b. *ˈmá biˈlá ko oʔˈmôna tiˈbîli ˈlé ˈétʃ͡i ˈwênala ko ˈmá toˈwáa ko biˈlé ba ˈmá biˈlá=ko oʔˈmôna tiˈbî-li aˈlé ˈétʃ͡i*

15.2 Adverbial clauses

*ˈwêna-la=ko* parents-poss=emph *[ˈmá* already *to-ˈwá=ko* take-mpass *biˈlé* one *ba]* cl

'The parents remained very sad because one of them (their children) was taken away.'

'Ya los papás se quedaron muy tristes porque se llevaron a uno de ellos (de sus hijos).' {SFH tx152:11:11.5}

c. *ˈpé aˈwêni be ko riˈhòwili tʃ͡aˈbè ˈkíti ˈpé koˈlì tiˈjôpatʃ͡i ˈníla rá siˈnéwi ko ba*

*ˈpé* just *aˈwêni* alone.pl *be=ko* just=emph *riˈhòwi-li* live.people-pst *tʃ͡aˈbè* before *[ˈkíti* sub *ˈpé* just *koˈlì* side *tiˈjôpatʃ͡i* church *ˈní-la* cop-rep *ru-ˈwá* say-mpass *siˈnéwi=ko* first.time=emph *ba]* cl

'There were just a few people living (here) before, that is why it is said it was over there by the church the first time.'

'Vivían poquitos antes, por eso dicen que fue allá por aquel lado de la iglesia la primera vez.' {SFH tx12:00:52.0}

As shown above, reason clauses may involve no subordinating morpheme (28a–b) or they may involve the subordinating morpheme *ˈkíti* (28c). Reason clauses may also be introduced by the subordinating marker *ˈká ˈtʃ͡è* (a complex marker often surfacing in a reduced form as*ˈkátʃ͡i*). For some speakers, this marker is exclusively attested in negative polarity clauses (e.g., (29a)), where there is no additional negative polarity morpheme). For other speakers, however, this morpheme may introduce reason clauses that have positive polarity, as shown in (29b).

(29) a. *ˈmáti ... tʃ͡iˈlàa muˈtʃ͡í ba weti ˈmá a riˈgá nosoˈwípi ˈlé ˈkátʃ͡i biˈlé oˈwâami niˈlú kiˈʔà ko ˈmá ba*

> *ˈmá=ti* already=1pl.nom *ˈétʃ͡i* dem *ˈlà-a* think-prog *muˈtʃ͡í* think *ba* cl *we=ti* int=1pl.nom *ˈmá* already *a* aff *riˈká* like.that *nosoˈwí-pi* be.finished-irr.pl *aˈlé* dub *[ˈká* because *ˈtʃè* neg *biˈlé* one *oˈwâ-ame* cure-ptcp *niˈlú* exist *kiˈʔà=ko* before=emph *ˈmá* also *ba]* cl

> 'And then we thought perhaps we will die (be finished) because there was no medicine before.'

'Y entonces pensábamos a lo mejor nos vamos a morir porque antes no había medicina.' {LEL tx372:02:00.0}

### 15 Complex clauses and complex predication

	- cl

'And then we were very scared because other people lived very far away.'

'Y ya teníamos mucho miedo nosotras porque otros vivían muy lejos de allí.' {LEL tx84:07:20.2}

### **15.2.4 Locative adverbial clauses**

Locative adverbial clauses in Choguita Rarámuri are introduced with the subordinating morpheme *ˈnápi* plus the proximal demonstrative *ˈna*. Locative clauses are exemplified in (30).

(30) *ˈwé mikaˈbê niˈlá ko ˈnápo ˈnà tʃ͡aˈbôtʃ͡i ˈmí riˈhòoa riˈwèli ko sekunˈdâria ba ˈwé* int *mikaˈbê* far-more *niˈlá=ko* cop-pot=emph *[ˈnápi ˈnà* sub prox *tʃ͡aˈbôtʃ͡i* mestizos *ˈmí* dist *riˈhòwi-a* live.sg-prog *riˈwè-li=ko]* remain-pst=emph *sekunˈdâria* secondary.school *ba* cl 'It must have been very far away, where the mestizos lived, the secondary school.' 'Sería muy lejos, donde vivían los mestizos, la secundaria.' {SFH tx12:1:29.4}

The main clause may contain a demonstrative (in boldface) that is coreferential with the locative adverbial clause. This is shown in (31).

(31) a. *ˈnápo ˈnà ti roˈplâno buˈjèi ˈétʃ͡o ˈnà biˈláni má tʃ͡ukuˈká neˈhê [ˈnápi* sub *ˈnà* prox *ti* def.bad *roˈplâno* plane *buˈj-è-i]* road-have-impf *ˈétʃi ˈnà* dem prox *biˈlá=ni* really=1sg.nom *ˈmá* already *tʃ͡uku-ˈká* crawl-ger *neˈhê* 1sg.nom 'There, where there was a road for the plane, I was going there.' 'Ahí donde era el camino del avión, ahí iba yo.' {SFH tx12:02:14.7}

15.2 Adverbial clauses

b. *ˈpé biˈlé riˈʰtê a riˈpî tʃ͡oˈnà ko ˈnápo ˈnàti muˈtʃ͡ûwili ˈpé* just *biˈlé* one *riˈʰtê* stone *a* aff *riˈpî* remain *tʃoˈnà=ko* there=emph *[ˈnápi* sub *ˈnà=ti* prox=1pl.nom *muˈtʃ͡ûwi-li]* sit.tr.pl-pst 'Just one stone remains there where we put them.' 'Nomás queda una piedra ahí donde las pusimos.' {LEL tx19:02:35.0}

Locative adverbial clauses may precede (31a) or follow (31b) the main clause.<sup>5</sup>

### **15.2.5 Temporal clauses**

There are two strategies in Choguita Rarámuri for encoding temporal clauses: (i) through a subordinating suffix *-tʃ͡i* that attaches to the dependent verb conjugated for present tense (32a); and (ii) through the morphologically complex subordinator *(n)apaˈlì* (a contraction from the subordinating particle *ˈnápi* and the time adverb *aʔˈlì* (32b).

(32) a. *aʔˈlì aˈjá saˈjèli naˈwàatʃ͡i ˈétʃ͡i koriˈmá paˈtʃ͡á bitiˈtʃ͡í baˈkíli aʔˈlì* and *aˈjá* soon *saˈjè-li* feel-pst *[naˈwà-a-tʃi* arrive-prog-loc *ˈétʃ͡i* dem *koriˈmá korima paˈtʃ͡á]* inside *bitiˈtʃ͡í* house *baˈkí-li* go.in.sg-pst 'And then they felt when the *korima* arrived and went inside the house.' 'Y luego sintieron cuando llegó el *korima* y entró a la casa.' {LEL tx5:00:42.8} b. *aʔˈlì ˈmá ku apaˈlì ˈmá ke ˈtékili moˈtʃ͡íʃili ˈétʃ͡i haˈréara ko aʔˈlì* and *ˈmá* already *ku* rev *[apaˈlì* when *ˈmá* already *ke* neg *ˈtéki-li]* be.drunk.pl-pst *moˈtʃ͡í-si-li* get.up.pl-mot-pst *ˈétʃ͡i* def *haˈréara=ko* some=emph 'And then when they were no longer drunk, the others (another group of people) got up.' 'Y entonces ya cuando ya no estaban borrachos, los demás (otros) se levantaron.' {LEL tx5:02:35.7}

<sup>5</sup>Clauses headed by locative predicates are addressed in §13.2.3 above.

### 15 Complex clauses and complex predication

These two examples encode different semantic relationships between the temporal clauses and the main clause in terms of the temporal sequence between the two events they encode. Thus, in (32a), the subordinate clause marked with the subordinating suffix *-tʃ͡i* encodes a proposition that is simultaneous to the proposition encoded by the main clause. In contrast, the subordinate clause introduced by the morpheme *napaʔˈlì* in (32b) encodes an event that occurs before the event encoded by the main clause.

### **15.2.6 Manner clauses**

Another type of adverbial clause attested in the Choguita Rarámuri corpus are manner subordinate clauses, which encode notions that can be replaced by a single word expressing manner relationships (Thompson et al. 1985). As shown in (33), manner subordinate clauses in Choguita Rarámuri may involve finite verbal predicates (33a, c) or verbs with the gerundive *-ká* suffix (33b) (constructions with clauses marked with this suffix in clause chaining are discussed in §15.5).


15.3 Relative clauses

c. *ˈwé bi ˈko riˈsóa riˈká ˈmá koʔˈwá ba tʃ͡aˈbèi ko ˈwé* int *bi=ˈko* bi=emph *[riˈsó-a* suffer-prog *riˈká]* like *ˈmá* already *koʔ-ˈwá* eat-mpass *ba* cl *tʃ͡aˈbèi=ko* before=emph 'Before people would eat struggling.' 'Antes comían batallando.' {SFH tx12:04:08.2}

As shown in these examples, manner adverbial clauses involve parataxis.

### **15.3 Relative clauses**

This subsection addresses the structural properties of relative clauses in Choguita Rarámuri. Relative clauses are defined in the typological literature as subordinate clauses that restrict or delimit the reference of a noun phrase (Andrews 2007). Strictly speaking, this description applies to *headed relative clauses*. In contrast to headed relative clauses, *headless relative clauses* do not combine syntactically with a head noun but rather act like referential or quantificational nominals themselves (Caponigro 2021).

Choguita Rarámuri exhibits two types of headed relative clauses, namely those that involve nominalization (§15.3.1) and those that involve finite predicates and subordinators (§15.3.2). The question of whether there are instances of headless relative clauses in Choguita Rarámuri is left out of the scope of this grammar. Other subordination strategies in the language are addressed in Chapter 15.

### **15.3.1 Relative clauses via nominalization**

Choguita Rarámuri employs a nominalization strategy to form relative clauses, as documented in closely related Mountain Guarijío (Miller 1996: 179) and Yaqui (Álvarez González 2012, Guerrero Valenzuela 2012) and other Uto-Aztecan languages (for a summary, see García Salido 2021). The nominalization strategy involves attaching participial suffixes (*-ame* or *-kame*) to the verb of the relative clause. This is exemplified in (34), with both relativized subjects and objects. As shown in these examples, the relative clause may precede the noun it modifies (34a), but most frequently the relative clause follows the noun it modifies (34b–e).

(34) a. *aʔˈlì ko ˈétʃ͡i ko ˈhê biˈlá aniˈwá tʃ͡aˈbè kiˈʔà ˈétʃ͡i ... ˈétʃ͡i ˈtʃêlame reˈhòi aʔˈlì=ko* and=emph *ˈétʃ͡i=ko* dem=emph *ˈhê* it *biˈlá* really *ani-ˈwá* say-mpass *tʃ͡aˈbè* before *kiˈʔà* long.ago *ˈétʃ͡i* dem *oˈtʃ͡êl-ame* grow.old-ptcp *reˈhòi* man 'and then they say that long ago that old man (the man who is old)'

### 15 Complex clauses and complex predication

'y allí dicen que antes ese señor viejito (el señor que es viejito)' {LEL tx71:01:43.9}

b. *aʔˈlì ˈétʃ͡i ˈtʃîba muˈkúami ko ma baˈsûa koˈʔáli ˈétʃ͡o ˈnà peˈrêami aʔˈlì* and *ˈétʃ͡i* dem *ˈtʃîba* goat *muˈkú-ame=ko* die.sg-ptcp=emph *ma* already *baˈsû-a* cook-prog *koˈʔá-li* eat-pst *ˈetʃ͡i* dem

*ˈnà peˈrê-ame*

there inhabit.pl-ptcp

'That dead goat (goat that is dead) was eaten cooked by the dwellers of (the ones that inhabit) that house.'

'Esa chiva (que está) muerta se la comieron los (que habitan) de esa casa.' {LEL tx\_mawiya:02:43.9}

c. *taˈmò ˈhípi oˈtʃêlami ke me aʔˈlá matʃ͡iˈká uˈtʃ͡úwi taˈmò* 1pl.nom *ˈhípi* today *oˈtʃ͡êl-ame* grow.old-ptcp *ke* neg *me* almost *aʔˈlá* well *matʃ͡i-ˈká* know-ger *uˈtʃ͡úwi*

lie.down.pl.prs

'Those of us growing older now almost don't know things well.'

'Los que crecemos ahora, casi no estamos sabiendo bien.' {SFH in61:00:36.2}

d. *iʔˈwîra ˈtʃ͡ó ˈkîni ˈnà … kompaˈɲêro ˈtʃ͡ó ˈí tʃ͡oˈgîta piˈrêami ˈtʃ͡ó ba iʔˈwîra* help.pl *ˈtʃ͡ó* also *ˈkîni* 1poss *ˈnà* dem *kompaˈɲêro* friends *ˈtʃ͡ó* also *ˈí* here *tʃ͡oˈgîta* Choguita *piˈrê-ame* dwell.pl-ptcp *ˈtʃ͡ó* also *ba* cl

'helping out out friends, the ones that live here in Choguita' 'ayudándoles a nuestros compañeros, los que viven aqui en Choguita' {SFH tx12:12:10.1}

e. *ˈétʃ͡i koliˈmî aniˈrîame biˈlá ˈwîʃula ruˈwá ˈétʃ͡i* dem *koliˈmî* rainbow *ani-ˈrîwa-ame* say-mpass-ptcp *biˈlá* indeed *ˈwî-si-la* take-mot-rep *ru-ˈwá* say-mpass 'It is said that the one that is called rainbow (*korimí*) goes along taking it.' 'Se dice que ese que le dicen arcoiris (*korimí*) lo va llevando.' {BFL tx\_muerto:01:22.3}

Relative clauses formed via nominalization may lack an antecedent, as shown in (35), where the relative clause encodes the nominal predicate in a copular clause.

15.3 Relative clauses

(35) a. *ˈwé sukuˈrûame ˈú ku ˈwé* int *sukuˈrû-ame* do.evil.witchcraft-ptcp *ˈhú=ko* cop.prs=emph 'He is very evil.' 'Es muy hechicero.' {CFH tx\_korimaka:00:22.3} b. *ˈápi iˈsêlikami ka ˈlé, biˈkiánika, suˈwâba maˈjôra ma ˈnápi* sub *i-ˈsêli-kame* pl-governor.pl-ptcp *ka* cop.irr *aˈlé* dub *biˈkiá-ni-ka* three-incl-coll *suˈwâba* all *maˈjôra* manager *ma* also 'those who are governors, the three of them, all of the managers, too' 'los que son gobernadores, los tres (gobernadores), todos los

mayores también' {JMF tx816:00:36.7}

As described for Mountain Guarijío (Miller 1996: 180), it is possible to analyze derived adjectives as relative clauses with an overt antecedent, and deverbal nouns like *sukuˈrûame* in (35) as relative clause without an overt antecedent.

### **15.3.2 Relative clauses via finite clauses**

A second productive strategy to encode relative clauses in Choguita Rarámuri involves a finite predicate and a subordinating particle *ˈnápi* (or its reduced form *ˈápi*) and no co-referencing morphology within the relative clause (a case of "gapping" as defined in the typological literature Lehmann 1986). As discussed in §15.1 and §15.2, this subordinating morpheme introduces complement clauses and certain classes of adverbial clauses. This strategy is exemplified in (36). Relative clauses are indicated with square brackets.

(36) a. *siˈrínala ˈbél ˈá aniˈrîa ˈétʃ͡i ˈnápu tʃ͡oˈpém ʃiˈmèla noˈká, ˈétʃ͡i ˈnápu ukuˈwéa noˈká siˈrína-la* god.father-poss *beˈlá* indeed *ˈá* aff *ani-ˈrî-a* say-mpass-prog *ˈétʃ͡i* dem *[ˈnápi* sub *tʃ͡oˈpé=mi* Montezuma.pine=dem *siˈmè-la* pass-rep *noˈká]* do.prs *ˈétʃ͡i* dem *[ˈnápi* sub *ukuˈwéa* ukuwéa.ceremony *noˈká]* do.prs

### 15 Complex clauses and complex predication

'They call him sirínala (godfather) the one that passes the Montezuma pine (over someone's head), the one that blesses with Ukuwéa.'

'Le dicen sirínala (padrino) a ese que hace con el ocote pasándolo, al que hace el ukuwéa.' {SFH tx475:07:33.3}

b. *aʔˈlì ˈétʃ͡i ˈápu roˈwéma ˈlé ko biˈnôi biˈlá aˈní: "ˈjénan ˈá saˈjèrima" ˈá aˈní aʔˈlì* and *ˈétʃ͡i* dem *[ˈnápi* sub *roˈwé-ma* run.womens.race-fut.sg *aˈlé]=ko* dub=emph *biˈnôi* herself *biˈlá* indeed *aˈní* say.prs *"ˈjéna=ni* aff=1sg.nom *ˈá* aff *saˈjèri-ma"* take.on-fut.sg *ˈá* aff *aˈní* say.prs

'And then the one who will run, herself, says: "yes, I will take on the challenge".'

'Y entonces la que va a correr ella misma dice "sí le voy a entrar".' {LEL tx19:00:39.8}

The examples in (36) involve relativized subject arguments. Relative clauses encoded via finite predicates and subordinators may also encode other core arguments, such as objects (37).

(37) a. *aʔˈlì tʃ͡oˈnà ˈhônsa ko ti ˈnà ˈmáti ˈnâri ˈápu roˈwéma ˈlé ˈátʃ͡i kaˈnílsa ˈníli roˈwéa aʔˈlì* and *tʃ͡oˈnà* then *ˈhônsa=ko=ti* from=emph=1pl.nom *ˈnà* dem.prox *ˈmá=ti* already=1pl.nom *ˈnâri* ask.prs *[ˈnápi* sub *roˈwé-ma* run.womens.race-fut.sg *aˈlé* dub *ˈátʃ͡i* if *kaˈníli-sa* be.happy-cond *ˈní-li* cop-pst *roˈwé-a* run.womens.race-prog 'And then we ask the one (woman) who will run if she would like to race.' 'Y entonces de ahí ya le preguntamos a la que va a correr a ver si se

siente a gusto.' {LEL tx19:00:33.2}

Relative clauses encoded with finite predicates and the subordinating particle *ˈnápi* can also encode adjunct arguments, including a time (38a), an instrument (38b), a location (38c) and a comitative argument (38d).

(38) a. *ˈpé ˈá ˈmá buˈàli ˈtʃ͡ó ˈétʃ͡i ˈnà rupuˈlá tʃ͡aˈbèi ˈnáp aʔˈlì muˈhê ˈnâtatʃ͡i ko ba?*

15.3 Relative clauses

*ˈpé* just *ˈá* aff *ˈmá* recently *buˈà-li* come.out-pst *ˈtʃ͡ó* also *ˈétʃ͡i* dem *ˈnà* that *rupuˈlá* ax *tʃ͡aˈbèi* before *[ˈnápi* **sub** *aʔˈlì* then *muˈhê* 2sg.nom *ˈnâta-tʃ͡i]=ko* think-temp=emph *ba* cl 'They had recently started using (appeared, come out) axes before, since the time you remember?' '¿Ya apenas habían salido antes las hachas antes, cuando tu te acuerdas? {SFH in484:03:23.8} b. *ripuˈlá...ˈnápu riˈká miˈtʃ͡ípu kuˈʃì ba? ripuˈlá* ax *[ˈnápi* sub *riˈká* that *miˈtʃ͡ípu* carve *kuˈsì]* sticks *ba* cl 'an ax with which to carve the sticks' 'hacha con que labrar los palos' {SFH in61:03:30.6} c. *ˈnápi ˈnà ˈá piˈrê aˈbôni siˈné ˈkátʃ͡i ˈmáti ˈá ˈhâwamti ˈá ˈníbo ko [ˈnápi* sub *ˈnà* dem.prox *ˈá* aff *piˈrê]* inhabit.pl *aˈbôni* emph.pl *siˈné* some *ˈkátʃ͡i* times *ˈmá=ti* already=1pl.nom *ˈá* aff *ˈhâwa-ame=ti* stand.pl-ptcp=1pl.nom *ˈá* aff *ˈní-bo=ko* cop-fut.pl=emph 'Where we live, perhaps we will some times be elected as authorities.' 'Donde vivimos a lo mejor en veces vamos a ser autoridades.' {SFH tx12:11:54.2} d. *ˈá biˈlá ˈtʃ͡ó biˈhí iˈjêna ˈtʃ͡ó ˈétʃ͡i apaˈnêrala ko ˈhípi ˈrú ˈnápi ˈûa iˈjênili ˈrú ba ˈá* aff *biˈlá* indeed *ˈtʃ͡ó* still *biˈhí* still *iˈjêna* go.around.sg.prs *ˈtʃ͡ó* still *ˈétʃ͡i* dem *apaˈnêra-la=ko* partner-poss=emph *ˈhípi* today *ˈrú* say.prs *[ˈnápi* sub *ˈjûa* with *iˈjêni-li* go.around.-pst *ˈrú]* say.prs *ba* cl 'He is still together with his wife now, with the one he used to live with before.' 'Todavía anda con su mujer ahora, con la que vivía (andaba) antes.' {SFH tx43:04:34.0}

In the case of the relativized time clause in (38a), the relativized clause makes reference to an adverb, *tʃ͡aˈbèi* 'before'. In the relativized locative clause in (38c), there is no antecedent, nominal or otherwise (notionally, the relative clause refers to a place, Choguita). The relative clauses in (38b) and (38d), on the other hand,

### 15 Complex clauses and complex predication

have nominal antecedents, *ripuˈlá* 'ax' and *apaˈnêrala* 'partner, wife', respectively. The head of the relative clause precedes the relative clause in every example, whether the relativized clause encodes a core argument or an adjunct.

The structure of Choguita Rarámuri relative clauses encoded through finite predicates has the same properties as the equivalent constructions in Rochéachi Rarámuri (Morales Moreno 2016: 38). The subordinating particle in Rochéachi Rarámuri is *ˈmapu* (cognate of Choguita Rarámuri *ˈnápi*), and also found in the speech of elderly Choguita Rarámuri speakers to introduce other subordinate clauses (addressed in Chapter 15).

### **15.4 Coordination**

This section describes clausal coordination in Choguita Rarámuri, where two or more clauses are combined to form a larger syntactic unit. Three main types of coordination strategies are identified, namely conjunction (§15.4.1), disjunction (§15.4.2) and adversative conjunction (§15.4.3). Within each section, the description is organized in terms of the morphosyntactic properties of the different constructions of each coordination type.

### **15.4.1 Conjunction**

### **15.4.1.1 Conjunction marked with** *aʔˈlì*

Phrasal and clausal conjunction in Choguita Rarámuri may involve syndetic coordination with the connective *aʔˈlì* 'and (then)', that is placed before the second conjunct. Syndetic coordination marked with *aʔˈlì* is exemplified in (39).


15.4 Coordination

c. *ˈétʃ͡i baˈtʃ͡á hiˈrâsa aʔˈlì ˈmá ˈhúmiʃi ˈétʃ͡i* dem *baˈtʃ͡á* first *hiˈrâ-sa* bet-cond *aʔˈlì* and *ˈmá* already *ˈhúmasi* take.off.pl.prs 'First they bet and then they take off.' 'Primero apuestan y ya arrancan.' {LEL tx19:01:40.3} d. *ˈwé aˈnè ˈhùria aʔˈlì ˈwé ˈá ʃiˈmí ˈétʃ͡i roˈwéami ko ˈwé* int *aˈn-è* say-appl *ˈhùri-a* send.off-prog *aʔˈlì* and *ˈwé* int *ˈá* aff *siˈmí* go.sg *ˈétʃ͡i* def *roˈwé-ami=ko* women.race-ptcp=emph 'They tell her to run, and the runner then goes faster.' 'Le dicen que corra y la corredora va corriendo más recio.' {LEL tx19:03:40.9} e. *ˈmá aʔˈlá nataˈkêa aˈkíbi aʔˈlì ˈmá ke aˈnítʃ͡ani ˈtʃ͡ó ˈmá aʔˈlá nataˈkê-a aˈkíbi aʔˈlì ˈmá ke aˈní-tʃ͡ani*

already well faint-prog went and already neg make.noise-ev *ˈtʃ͡ó* also 'She fainted and then it wasn't making noise anymore.' 'Ya se desmayó y ya no se oía.' {LEL tx71:04:43.6}

As shown in these examples, coordination marked with *aʔˈlì* involves a temporal relationship between the conjuncts where the event described in the second clause temporally follows the event described in the first clause. As shown in these examples, coordination involves full clauses with independent predicates and arguments, even if arguments may be co-referential (e.g., (39c)).

Clausal conjunction may involve ellipsis of the predicate of the second clause. This is exemplified in (40), where angled brackets <> indicate the ellipsis site in the second clause.

(40) a. *aʔˈlì reˈhòi ko ˈmá buˈʔíli aʔˈlì muˈkî ko ke ˈtʃ͡ó aʔˈlì* and *reˈhòi=ko* man=emph *ˈmá* already *buˈʔí-li* lay.down.sg-pst *aʔˈlì* and *muˈkî=ko* woman=emph *ke* neg *<>* <> *ˈtʃ͡ó* yet 'Y el señor ya estaba acostado y la mujer todavía no.' 'And the man was already asleep but not the woman.' {LEL tx5:00:48.3}

### 15 Complex clauses and complex predication

As shown in (40), the connective may appear immediately preceding the first of two connected clauses or at the beginning of a clause or a sequence of clauses, as exemplified in (41).

(41) a. *aʔˈlì biˈlá ko ˈnà ˈétʃ͡i muˈkî ko ˈèbiʃuwa ˈétʃ͡i saʔˈpá oʔˈwí ˈpé biˈlá ko ˈá riˈká koˈʔáa ruˈá ˈrám pa ˈnápu riˈká ˈnàa aʔˈlì* and *biˈlá=ko* indeed=emph *ˈnà* then *ˈétʃ͡i* dem *muˈkî=ko* woman=emph *ˈèbi-suwa* bring-cond.pass *ˈétʃ͡i* dem *saʔˈpá* meat *oʔˈwí* raw *ˈpé* just *biˈlá=ko* indeed=emph *ˈá* aff *riˈká* like.that *koˈʔá-a* eat-prog *ru-ˈwá* say-mpass *ru-ˈwá=mi* say-mpass=dem *pa* cl *ˈnápi* that *riˈká* like *ˈnà* dem 'And then the woman had been brought raw meat and she would eat it like that, they say.' 'Y entonces a la mujer le traía carne cruda y tenía que comer así.' {SFH tx43:06:17.3}

Details of the function of this and other connectives in Choguita Rarámuri discourse are left out of the scope of this grammar.

### **15.4.1.2 Asyndetic conjunction**

Clausal conjunction in Choguita Rarámuri may also be encoded through asyndetic conjunction, i.e., through juxtaposition of finite clauses with no overt conjunction marker. This is exemplified in (42). Conjoined clauses exhibit prosodic unification.

(42) a. *ˈpé niˈhê aˈníami ˈú: "ku rohoˈnâsa, roˈhàʃi ba"*

*ˈpé* just *niˈhê* 1sg.nom *aˈní-ame* say-ptcp *ˈhú* cop *ku* rev *roha-ˈnâ-sa* separate-tr-imp.sg *roˈhà-si* separate-impl.pl *ba*

cl

'I just tell them (the women betting with me): "Separate it (the bet), and separate yourselves".'

'Yo nomás digo: "Aparten (la apuesta) y apártense (las que apuestan conmigo)".' {LEL tx19:04:24.9}

b. *banaˈká ˈpáli ripiˈjá, ˈnè ku ˈtèaki aʔˈlì ˈlé ba banaˈká* over.there *ˈpá-li* throw-pst *ripiˈjá,* knife *ˈnè* 1sg.nom *ku* rev *ˈtèa-ki* find-pst.ego *aʔˈlì* later *aˈlé* dub *ba* cl 'He threw the knife there and I found it later.'

15.4 Coordination

'Tiró el cuchillo allá y yo lo encontré más tarde.' {JLG co1234:03:29.4}

As in conjunction constructions that involve a connective, the ordering of clauses is iconic, with the first clause describing an event that precedes temporally the event described by the second clause, as shown in (42). In other cases, however, asyndetic coordination does not involve temporal succession. This is illustrated in (43).

(43) *ˈkárka naˈpònili, ˈkárka kaˈsìnili ˈká* everything *riˈká* like.that *naˈpò-na-li* break-tr-pst *ˈká* everything *riˈká* like.that *kaˈsì-na-li* tear.apart-tr-pst 'He broke it all and tore it all.' 'Lo quebró todo y lo desmoronó.' {MDH co1137:00:36.2}

**15.4.2 Disjunction**

Disjunctive clausal coordination is achieved through a variety of morphosyntactic means, including: (i) a construction with the borrowed Spanish disjunctive conjunction *o* 'or'; (ii) a disjunctive prepositive conjunction *wera* 'or'; (iii) a postpositive enclitic *=ma* 'or' that attaches to each disjunctive clause; or (iv) through parataxis. Each of these strategies is described next.

### **15.4.2.1 Disjunction marked with** *o* **'or'**

A frequently attested strategy for encoding disjunction in the Choguita Rarámuri corpus involves the borrowed Spanish conjunction *o* 'or'. Its use in clausal disjunctive coordination is exemplified in (44), which shows a fragment of a narrative.

	- b. *ˈétʃ͡i ˈhápi ke ˈmí kaˈléa ˈníli, o biˈlé ˈtʃó͡ ˈmá ˈhê aˈníame ka ruˈwá: "ˈnà ˈétʃ͡i ˈhùrimi ˈnà*"

### 15 Complex clauses and complex predication

*ˈétʃ͡i* dem *ˈnápi* sub *ke* neg *ˈmí* dem *kaˈlé-a* like-prog *ˈní-li* cop-pst *o* or *biˈlé* one *ˈtʃ͡ó* also *ˈmá* already *ˈhê* it *aˈní-ame* say-ptcp *ka* cop.irr *ru-ˈwá* say-mpass *ˈnà* prox *ˈétʃ͡i* dem *ˈhùri-mi* send-mot.imp.sg *ˈnà* prox '(He sends it) to someone he doesn't like or someone else tells him:"go and send it to that one".' '(Se lo manda) a uno que no le cae bien u otro le dice: "mándaselo a ese".' {LEL tx5:05:11.8}, {LEL tx5:05:15.5}

The use of the borrowed Spanish disjunctive conjunction *o* is also attested in coordination of interrogative clauses, as shown in (45).

(45) *naʔˈpôma? o ˈí ʃiˈkâ-la kaˈpòma?*

*naʔˈpô-ma* weed-fut.sg *o* or *ˈí* here *siˈkâ-la* hand-poss *kaˈpò-ma?* break-fut.sg

'(You mean) to weed? Or to break one's hand here?'<sup>6</sup>

'¿(Quieres decir) escardar? ¿O quebrarse aqui en la mano?' {JLG co1236:00:06.4}

### **15.4.2.2 Disjunction marked with** *wèra* **'or'**

A second construction that employs syndetic coordination to encode disjunction is one that employs the marker *ˈwèra*, which has the same distribution as the borrowed disjunctive conjunction *o*. This construction is exemplified in (46).

(46) a. *ˈkúmi ˈátʃ͡i ˈtôola ˈlé, ˈwèra ˈpé risoˈtʃ͡í roˈʔáli ˈlé ˈkúmi* where *ˈátʃ͡i* Q *ˈtô-la* take.pst.pass-rep *aˈlé* dub *ˈwèra* or *ˈpé* just *risoˈtʃ͡í* cave *roˈʔ-á-li* lay.down.sg-tr-pst *aˈlé* dub 'They don't know if they would bury him or they would lay him down in a cave.' 'No saben si lo enterraban o lo llevaban a una cueva.' {LEL tx109:03:02.7}

<sup>6</sup>The speaker points at her own hand.

15.4 Coordination

### **15.4.2.3 Disjunction marked with** *=ma* **'or'**

Clausal disjunction may also be achieved through cliticization of the disjunctive connective *=ma*, a postpositive marker. This is exemplified in (47).

(47) a. *koˈbísi loˈkása ba, baʔˈwí ma, baˈhîrʃia ˈma ba koˈbísi* pinole *loˈká-sa* take.pinole-cond *ba* cl *baʔˈwí=ma* water=or *baˈhî-ri-si-a=ma* drink-caus-mot-prog=or *ba* cl 'When she takes pinole, or water, or giving her to drink some water.' 'Cuando toma pinole o agua o dándole de tomar agua.' {LEL tx19:04:08.4}

Disjunction of phrases is also attested with the enclitic *=ma*. This is exemplified below, where the disjunction morpheme appears after each phrase in a monosyndetic construction (48a), or the two last phrases in a construction with multiple coordinands (48b).

	- b. *biˈlé ariˈmûli, oˈkwâ ariˈmûli ma biˈkiá ariˈmûli ma biˈlé* one *ariˈmûli,* decaliter *oˈkwâ* two *ariˈmûli=ma* decaliter=or *biˈkiá* three *ariˈmûli=ma* decaliter=or 'One decaliter, two deacaliters or three decaliters.' 'Un decalitro o dos decalitros o tres decalitros.' {LEL tx68:00:25.8}

### **15.4.2.4 Disjunction through parataxis**

Disjunction may also be achieved through parataxis. This is exemplified in the following text fragment, where (49b) is interpreted involving disjunction within the larger context.

(49) a. *ˈkúm ko ˈtòli ba? ˈkúmi=ko* where=emph *ˈtò-li* take-pst *ba* cl 'Where did he take it?' '¿Dónde lo llevaría?' {LEL tx84:07:15.2}

### 15 Complex clauses and complex predication


### **15.4.3 Adversative conjunction**

Adversative conjunction may be encoded through asyndesis or through syndetic coordination with a dedicated adversative conjunction. Each of these constructions is addressed next.

### **15.4.3.1 Asyndetic adversative conjunction**

Choguita Rarámuri possesses an adversative conjunction construction, which may be encoded through parataxis, like other forms of clausal coordination. This is shown in (50).

(50) a. *wiˈʰkâ napaˈwíka raˈʔìtʃ͡ili ˈmí riʔˈlé-ki esˈkwêlitʃ͡i ko, ˈnè ke ʃiˈné iˈjêna wiˈʰkâ* many *napaˈwí-ka* gather-ger *raˈʔìtʃ͡i-li* speak-pst *ˈmí* dist *reʔˈlé-ki* below-supe *esˈkwêla-tʃ͡i=ko,* school-loc=emph *ˈnè* 1sg.nom *ke* neg *siˈné* once *iˈjêna* go.sg.pst 'Many (people) got together to speak over there at the school, but

I've never gone.' 'Se juntaron muchos allá en la escuela a platicar, pero yo nunca he

ido.' {MDH co1137:07:34.8}

b. *ˈmá biˈlá rikiˈnâmo oˈlá riˈké ti roˈplâno ke biˈlá oˈmêaki riˈkîna ˈmá* already *biˈlá* really *rikiˈnâ-mo* go.down-fut.sg *oˈlá* cer *riˈké* perhaps *ti* def.bad *roˈplâno* plane *ke* neg

15.4 Coordination

*biˈlá* really *oˈmêaki* be.able-pst *riˈkîna* go.down 'And the plane was about to go down, but it couldn't go down.' 'Ya se iba a bajar el avión, nomás que no pudo bajar.' {SFH tx12:01:58.8}

c. *ke biˈlé muˈrúli ˈhípi ko, ˈétʃ͡i biˈlá ˈá muˈrúma aˈlé ba ke* neg *biˈlé* one *muˈrú-li* carry.with.arms-pst *ˈhípi=ko* today=emph *ˈétʃ͡i* dem *biˈlá* indeed *ˈá* aff *muˈrú-ma* carry.with.arms-fut.sg *aˈlé* dub *ba* cl 'He didn't carry any (wood), but he will bring it.' 'Ahora no trajo (leña), pero él la va a acarrear.' {JLG co1236:03:21.1}

### **15.4.3.2 Adversative conjunction marked with** *naˈlîna* **'but'**

Adversative conjunction may also be marked through the connective *naˈlîna* 'but' (or its reduced form *aˈlîna* or *ˈlîna*)). As the following examples show, in these constructions the adversative conjunction appears between the two conjoined clauses (e.g., (51)).

(51) a. *aʔˈlì ˈnà kotʃ͡iˈká buˈʔílo maˈjêli, ˈlîna ke ˈtâsi kotʃ͡iˈká buˈʔíli ˈétʃ͡i reˈhòi ko aʔˈlì* and *ˈnà* then *kotʃ͡i-ˈká* sleep-ger *buˈʔí-l-o* lay.down.sg-pst-ep *maˈjê-li* think-pst *ˈlîna* but *ke* neg *ˈtâsi* neg *kotʃ͡i-ˈká* sleep-ger *buˈʔí-li* lay.down-pst *ˈétʃ͡i* dem *reˈhòi=ko* man=emph 'And then he thought he was asleep, but the man was not asleep.' 'Nomás que pensó que estaba dormido (lit. acostado durmiendo), nomás que no estaba dormido el señor.' {LEL tx5:00:35.0}, {LEL tx5:00:38.3} b. *moˈʔêa tʃ͡uˈkúrili ˈtʃ͡ó ˈáa ba, paˈtʃ͡á ˈsàbia ke ˈlé ba, naˈlîna be ke ˈtâsi riˈwáli ˈtʃ͡ó ˈá riˈká ba moˈʔ-ê-a* head-vblz-prog *tʃ͡uˈkúri-li* become.bent-pst *ˈtʃ͡ó* also *ˈá-a* look.for-prog *ba* cl *paˈtʃ͡á* inside *ˈsàwi-a* smell-prog *ke* perhaps *aˈlé* dub *ba,* cl *naˈlîna* but *be* be *ke* neg *ˈtâsi* neg *riˈwá-li* see-pst *ˈtʃ͡ó* also *ˈá* aff *riˈká* indeed *ba* cl 'He got his head in it (the bottle) looking for it, smelling inside, but

### 15 Complex clauses and complex predication

he didn't see him.'

'Metió la cabeza buscando, oliendo para buscarlo, pero de todas formas no lo vió.' {SFH tx152:01:32.6}, {SFH tx152:01:36.1}

c. *ˈpé ˈtáa biˈlá ko riˈhòi ˈníla ˈrá aʔˈlì ˈétʃ͡i ko naˈlîna ˈwé naˈlîna ˈwé hiˈwêlami ˈtʃ͡ó ˈníla ˈlá ˈnápu riˈká ˈpé* just *ˈtá* small.sg *biˈlá=ko* indeed=emph *riˈhòi* man *ˈní-la* cop-rep *ru-ˈwá* say-mpass *aʔˈlì* and *ˈétʃ͡i=ko* dem=emph *naˈlîna* but *ˈwé* int *naˈlîna* but *ˈwé* int *hiˈwê-l-ame* be.strong-pst-ptcp *ˈtʃ͡ó* also *ˈní-la* cop-rep *ru-ˈwá* say-mpass 'He was very small, but he was a very strong man.' 'Era muy chiquito, pero nomás que era un hombre muy fuerte.' {SFH tx43:02:38.9}

### **15.5 Verbal chaining structures**

576

Choguita Rarámuri has a construction that involves a sequence of clauses, where one of the clauses may be marked with canonical inflection (and show no restrictions in terms of TAM marking), while the remaining clauses can only be marked with special inflection (the gerundive suffix *-ká*) and show overall more restricted structures. This inflection mainly conveys a temporal relation of chronological overlap or chronological sequence, though, as discussed below, these temporal notions may have extended semantic meanings in some cases. In the following example, the first clause, marked with the stress-shifting suffix *-ká*, glossed as gerundive (ger), conveys that two events (drinking and resting) take place simultaneously (52).

(52) *ˈwé pi ko ne ku iˈsâbika baˈhîba ˈlé, ˈmá ˈòwisa ˈnà paˈtʃ͡í pa ˈwé* int *pi=ko* just=emph *ne* int *ku* rev *iˈsâbi-ka* rest-ger *baˈhî-ba* drink-irr.pl *aˈlé* dub *ˈmá* already *ˈòwi-sa* fertilize-cond *ˈnà* dem *paˈtʃ͡í* corn *pa* cl 'They need to drink while they rest when they (are done) fertilizing this corn.' 'Necesitan tomar descansando hasta que (terminen de) fertilizar este maíz.' {FLP in243:17:22.2}

15.5 Verbal chaining structures

In contrast, in (53) the events conveyed occur in a temporal sequence, such that the event described in the finite clause (performing a ritual blessing) occurs after a series of events, described by the *ka*-marked clauses, have taken place.

(53) *baʔˈwí rataˈbátʃ͡ika, tʃ͡oˈpé ˈjûa ʃiˈmèrika, ku aˈwílitʃ͡i ʃimiˈká, wiˈrónipo ˈkútʃ͡i paˈtʃî ba*

*baʔˈwí* water *rata-ˈbá-tʃ͡i-ka* be.hot-inch-tr-ger *tʃ͡oˈpé* Montezuma.pine *ˈjûa* with *siˈmèri-ka* pass.on.top-ger *ku* rev *aˈwílitʃ͡i* ritual.patio *simi-ˈká* go.sg-ger *wiˈró-ni-po* make.blessing-appl-fut.pl *ˈkútʃ͡i* def *paˈtʃî* corn *ba* cl 'Having heated up the water, and having passed on top with a (lit) Montezuma pine, and having gone back to the ritual patio, we make the blessing with corn.'

'Calentando agua, pasando por arriba con un ocote (prendido), yendo al patio ritual, hacemos la bendición ("echamos el agua") con el maíz.' {ME in485:07:08.4}, {ME in485:07:11.4}

This construction is extremely frequent and attested extensively across speech genres in the Choguita Rarámuri corpus. The following examples are from procedural texts, where the temporal relations involve sequences of events (54) (the finite clauses in these examples are marked as medio-passive).

	- b. *suˈnù ku aˈnáaga aˈrîo oˈlá waˈrîtʃ͡i ko suˈnù* corn *ku* rev *aˈnáwi-ka* measure-ger *a-ˈrîwo* give-mpass *oˈlá* cer *waˈrî-tʃ͡i=ko* palm.basket-loc=emph 'Having measured the corn, it is given to them, in the *wari* (palm basket).'

'Se mide el maíz y se les da, en el wari (canasta de palma).' {MDH co1136:05:44.8}

c. *aʔˈlì ˈmá ˈnà ... ˈmá oˈtʃ͡é ku ˈnà baʔˈwêtʃ͡i ˈnà ... ˈnà muˈtʃ͡ûuka ku ˈnà paˈkóka ˈnà batuˈʃíwa mataˈtʃ͡í*

### 15 Complex clauses and complex predication

*aʔˈlì* and *ˈmá* already *oˈtʃ͡é* once.again *ku* rev *ˈnà* dem *baʔˈwê-tʃ͡i* water-loc *ˈnà* dem *muˈtʃ͡ûwi-ka* put-ger *ku* rev *ˈnà* dem *paˈkó-ka* wasah-ger *ˈnà* dem *batuˈsí-wa* grind-mpass *mata-ˈtʃ͡í* grinding.stone-loc 'And then having soaked (placed) it again in the water, and having washed (rinsed) it, then it is ground in the grinding stone.' 'Y luego se remoja otra vez en agua y se enjuaga ya para molerlo en

el metate.' {LEL tx68:00:58.8}

Clauses marked with *-ká* are non-finite, while the final clause in each of these examples (the 'unmarked' clause) may be inflected like an independent sentence. Furthermore, marked clauses are not introduced by any overt subordinator. These properties are characteristic of clause chaining structures (see Longacre 2007 for an overview), though the formal and semantic properties of the Choguita Rarámuri construction with *ka*-marked clauses blur the lines between clause chaining and other types of related constructions in this language.<sup>7</sup>

As exemplified in (52) above, the *ka* construction in Choguita Rarámuri often involves two verbs which are covalent, i.e. they share one or more arguments, a property also present in serial verb constructions (see §15.6.3 below). As shown in (53) and below in (55), the *ka*-marked clauses may contain an overt object argument not shared with the unmarked clause.

	- *ˈtá* small *tʃ͡iˈkîto* small *muˈtû-ka* hold.on.arms-ger *ku* rev *naˈwà-li* arrive-pst *ˈtʃ͡ó* also 'She arrived holding a small one in her arms too.' 'Llegó con un chiquito en los brazos también.' {LEL tx32:09:03.3}

<sup>7</sup>Morales Moreno (2016) analyzes cognate constructions with the suffix *-ká* as involving secondary predication in Rochéachi Rarámuri. This suffix is described as a resultative state suffix in Norogachi Rarámuri (Villalpando-Quiñonez 2019). The cognate *-ka/ga* suffix is described as a past participle (Miller 1996) in closely-related Mountain Guarijío, as a participial in River Guarijío (Felix Armendáriz 2006) and as a participial also in Yaqui (Dedrick & Casad 1999, Guerrero Valenzuela & Van Valin 2004, Guerrero Valenzuela & García Salido 2019).

15.5 Verbal chaining structures

c. *tʃ͡aˈbèe go, siˈnêam ˈká biˈlá ko ˈwé awiˈwái tʃ͡aˈbèe go ba, napaˈbûka ba suˈnù tʃ͡aˈbè=ko* before=emph *siˈnê-ame* all-ptcp *ˈká* cop.irr *biˈlá=ko* indeed=emph *ˈwé* int *awi-ˈwá-i* dance-mpass-impf *tʃ͡aˈbè=ko* before=emph *ba,* cl *napa-ˈbû-ka* get.together-tr-ger *ba* cl *suˈnù* corn 'Long time ago, everybody used to dance a lot long ago, while gathering corn.' 'Hace mucho, todos bailaban mucho antes, juntando maíz.' {MDH co1140:15:46.3}

In each of the examples in (55), the finite clause is headed by an intransitive verbal predicate. In (56) below, the finite clause is headed by a transitive verb, and the *ka*-marked clause contains an overt object (*biˈlé baˈrîka* 'one cask') that is non-coreferential with the object argument of the finite clause (*baʔˈwí* 'water'). Both clauses share an external argument, the second singular subject argument marked in the main, finite clause.

(56) *aʔˈlì ˈmámi baʔˈwí roˈʔèma oˈhòsa aˈnâuka biˈlé baˈrîka aʔˈlì* and *ˈmá=mi* already=2sg.nom *baʔˈwí* water *roˈʔ-è-ma* pour.appl-fut.sg *oˈhò-sa* dekernel-cond *aˈnâwi-ka* measure-ger *biˈlé* one *baˈrîka* cask 'When you dekernel it you pour water, having measured one cask.' 'Cuando lo desgranas y ya le echas agua, ya que mides una barrica.' {BFL tx60:00:27.2}

In this example (as in (55c)), the clause marked with *-ká* appears after the finite clause. Clause chaining structures are often described as 'medial-final' clause chaining in the typological literature (Longacre 2007), but in Choguita Rarámuri the linear order of the marked and unmarked clauses is variable (as also documented in Northern Paiute (Numic) (Toosarvandani 2016)). Another example of this is shown in (57).

(57) *noˈkí naˈwàko ˈlá poˈtʃ͡íka noˈkí* almost *naˈwà-ki* arrive-pst.ego *oˈlá* cer *poˈtʃ͡í-ka* jump-ger 'He almost arrived jumping ("with a single jump").' 'Casi llegó a la puerta brincando ("de un brinco").' {LEL tx84:03:31.1}

### 15 Complex clauses and complex predication

In (57), the *ka*-marked clause depicts an event that temporally precedes the event described in the finite clause, and can be interpreted as encoding manner (the event, arriving to a location, was accomplished through jumping). Other examples where the *ka*-marked clauses may be interpreted in a manner sense are shown in (58).

(58) a. *kuˈsìtini wipiˈsóka miˈʔàki kuˈsì-ti=ni* stick-inst=1sg.nom *wipiˈsó-ka* hit.w.stick-ger *miˈʔà-ki* kill.sg-pst.ego 'I hit it to death with a stick.' 'La maté a palos (a la rata).' < BFL 09 1:88/el > b. *baˈtʃ͡ókati niˈwâbo kaˈlí baˈtʃ͡ó-ka=ti* fix.with.mud-ger=1pl.nom *niˈwâ-bo* make-fut.pl *kaˈlí* house 'We made the house by fixing the walls with mud.' 'Enzoquetando hicimos la casa.' < BFL 09 1:88/el > c. *weˈpáka miˈʔàlo basaˈtʃ͡î weˈpá-ka* hit-ger *miˈʔà-li* kill.sg-pst *basaˈtʃ͡î* coyote 'They hit the coyote to death.' 'Mataron a golpes al coyote.' < BFL 09 1:88/el > d. *naˈhâtika miˈʔàa ruˈwáo tʃ͡imoˈrí tʃ͡aˈbèi naˈhâti-ka* chase-ger *miˈʔà-a* kill.sg-prog *ru-ˈwá-o* say-mpass-ep *tʃ͡imoˈrí* deer *tʃ͡aˈbèi* before 'They say they would chase deer to death before.' 'Dicen que persiguiéndolo lo mataban al venado antes.' < BFL 09 1:88/el > e. *baniˈʃúkani moʔoˈbûko ˈkâha baniˈsú-ka=ni* pull-ger=1sg.nom *moʔo-ˈbû-ki* go.up-tr-pst.ego *ˈkâha* box 'I got the box up by pulling it.' 'Jalando subí la caja.' < BFL 09 1:88/el > f. *"niˈhê ˈmá ku ˈmètika naˈwàki" aˈní niˈhê* 1sg.nom *ˈmá* already *ku* rev *ˈmèti-ka* drive-ger *naˈwà-ki* arrive-pst.ego *aˈní* say.prs "'I arrived driving" it is said.' "'Llegué manejando" se dice.' {JLG el1278:03:47.5}

15.5 Verbal chaining structures

In these examples, the marked clauses precede temporally the event described in the finite clause. In other cases, as shown in (59), the sense of manner is conveyed also in cases where the temporal relationship would be one of chronological simultaneity.

(59) a. *ˈétʃ͡i riˈká biˈlám wiˈróka riˈká biˈlá ku baˈjèa ˈrú ba ˈni ˈétʃ͡i* dem *riˈká* like *biˈlá=m* indeed=dem *wiˈró-ka* throw.up-ger *riˈká* like *biˈlá* indeed *ku* rev *baˈjè-a* call-prog *ˈrú* say.prs *ba* cl *ˈni* emph 'Like that, throwing water up (blessing), that is how they would call it (the rain).' 'Así tirando para arriba (bendiciendo) así era como le llamaban (a la lluvia).' {ME in484:05:29.8} b. *ˈpé itʃ͡aˈká ko koˈʔáa-li ˈpé* just *itʃ͡a-ˈká=ko* sow-ger=emph *koˈʔá-li* eat-pst 'They ate from what they planted.' 'Sembrando comían ("comen sembrado").' {SFH tx12:04:09.9} c. *ˈétʃ͡i riˈgá naʔˈsòoka ˈbel koʔˈpôo ba ˈétʃ͡i* dem *riˈká* like.that *naʔˈsòwi-ka* mix-ger *ˈbel* indeed *koʔ-ˈpô* eat-fut.pl *ba* cl 'One has to eat like that, by mixing (the foods).' 'Hay que comer así revuelto.' {MDH co1136:00:52.8} d. *ˈmèti ˈbél aˈlé ba, ˈmèka ˈbél isiˈmípili aˈlé ku ba ˈmèti* drive.prs *beˈlá* indeed *aˈlé* dub *ba* cl *ˈmè-ka* drive-ger *beˈlá* indeed *i-siˈmípi-li* pl-go.pl-pst *aˈlé* dub *ku* rev *ba* cl 'They drive, they go back (in their trucks) by driving.' 'Manejan, se van (en las camionetas) otra vez manejando.' {MDH co1136:07:57.2} e. *ˈnâli ˈpé beˈláni ko ˈnè iˈjêla iʔˈnèka biˈnè ˈnè ko tʃ͡ú reˈká oˈlá ˈétʃ͡i ba ˈnâli* then *ˈpé* just *beˈlá=ni=ko* really=1sg.nom=emph *ˈnè* 1sg.nom *iˈjê-la* mom-poss *iʔˈnè-ka* watch-ger *biˈnè* learn.pst *ˈnè=ko* 1sg.nom=emph *tʃ͡ú* how *reˈká* like *oˈlá* do.prs *ˈétʃ͡i* dem *ba* cl 'Then I learned by watching my mom, how would she make it.'

### 15 Complex clauses and complex predication

'Entonces yo aprendí viendo a mi mamá a ver cómo le hacía.' {BFL tx1:00:55.5}

In the following example (60), the temporal sequence between the clauses has an interpretation of cause and effect.

(60) *baniˈwîtika, witʃ͡iˈmêa ˈlémi! baniˈwîti-ka* entangle-ger *witʃ͡i-ˈmêa* fall-fut.sg *aˈlé=mi* dub=2sg.nom

'You will trip and you will fall!'

'¡Te vas a enredar y te vas a caer!' {MDH co1136:12:18.6}

According to Longacre (2007: 400), the temporal relations encoded in clause chaining may give rise to associated semantic readings, such as result, cause and effect, among others. The sentences in (58) and (59) may thus be analyzed as clause chaining.<sup>8</sup>

As exemplified so far, none of the *ka*-marled clauses are introduced by an overt subordinator. There are, however, cases, where an overt coordinator may appear in a clause chain. This is shown in (61), where the coordinator occurs sentenceinitially, as well as in the clause chain.

(61) *aʔˈlì baʔˈwí ˈhûa naˈsòo-ka aʔˈlì ˈhê riˈká taˈmò uˈtʃ͡-è-i aʔˈlì* and *baʔˈwí* water *ˈhûa* with *naˈsòwi-ka* mix-ger *aʔˈlì* and *ˈhê* it *riˈká* like.that *taˈmò* 1pl.acc *uˈtʃ-è-i* put-appl-impf 'And so mixing it with water, and then he would put it on us like that.'

'Y entonces mezclándola con agua y entonces así nos lo ponía.' {GFM tx785:02:49.6}

As discussed in Toosarvandani (to appear), other Uto-Aztecan languages that are documented to have clause chaining allow overt coordinators in chaining structures.

Finally, there are cases where negation only has scope over the *ka*-marked clause. This is exemplified in (62).

(62) *ke piˈláti biˈlé matʃ͡iˈká moˈtʃ͡íli taˈmò ˈká ˈtʃ͡è wiˈle ˈtʃ͡ó matʃ͡iˈwá aʔˈlì ko ba*

<sup>8</sup>Alternatively, the gerundive *-ka* suffix may be deployed in a different type of construction that may be better analyzed as involving adverbial modification, with *ka*-marked clauses better analyzed as sentence margins in a co-ranking structure.

15.6 Complex predicates

*ke* neg *biˈlá=ti* indeed=1pl.nom *biˈlé* one *[matʃ͡i-ˈká* know-ger *moˈtʃ͡í-li]* be.sitting.pl-pst *taˈmò,* 1pl.nom *ˈká* because *ˈtʃ͡è* neg *biˈlé* one *ˈtʃ͡ó* yet *matʃ͡i-ˈwá* know-mpass *aʔˈlì=ko* then=emph *ba* cl

'Well, we were without knowing, because then nothing was known yet.'

'Pues nosotros estábamos sin saber porque en ese tiempo todavía no se sabía nada.' {LEL tx372:00:41.4}

The Choguita Rarámuri *-ká* suffix is cognate with a *-ka* marker in Hiaki (Taracahitan) that is analyzed as a medial verb marker (Martínez Fabián 2006), which may convey both a simultaneous and a sequential interpretation.<sup>9</sup> Elsewhere in the Taracahitan branch, Mountain Guarijío is also described as possessing a system of medial markers, five morphemes, which may encode same vs. different subject distinctions, in addition to sequential and simultaneous distinctions (Miller 1996: 200). Miller (1996) remarks that this morphological system of medial verb markers is a recent innovation in the language.

### **15.6 Complex predicates**

Choguita Rarámuri has complex constructions that can be broadly defined as involving complex predicates. While variously characterized in the literature depending on the particular morphosyntactic properties of individual languages under consideration, complex predicate constructions are nevertheless defined as involving a complex argument structure and a grammatical functional structure of a single predicate (Butt 1995: 108). Complex predicates thus involve a single 'inflectional domain' containing two or more distinct predicates, each of which selects for at least one argument in its argument structure (Baker 1997: 247) and each of which contributes to the predicate information that in other constructions is associated with a head (Butt & Geuder 2001).

Complex predicate constructions may be further classified into sub-types, including light verb constructions and serial verb constructions, based on their syntactic and semantic properties (Butt 1995, Svenonius 2008). The status of auxiliary verb constructions as a type of complex predicate construction varies within the literature: Butt (1995), for instance, excludes auxiliary verb constructions from the class of complex predicates in her typology, but the boundaries between this

<sup>9</sup> It should be noted, however, that other studies of Yaqui syntax (Dedrick & Casad 1999, Guerrero Valenzuela & Van Valin 2004, Guerrero Valenzuela & García Salido 2019) do not analyze this marker as a medial verb marker.

### 15 Complex clauses and complex predication

type of construction and others uncontroversially defined as involving complex predicates may be blurry (Svenonius 2008). There are four types of constructions in Choguita Rarámuri that may be broadly characterized as involving complex predicates: light verb constructions (§15.6.1), auxiliary verb constructions (§15.6.2), serial verb constructions (§15.6.3) and multi-predicate verb constructions involving V-V incorporation that have been analyzed as serial verb constructions or light verb constructions in other Uto-Aztecan languages (§15.6.4).

### **15.6.1 Light verb constructions**

While the term "light verb" is typically employed in the literature to characterize a wide range of structures cross-linguistically, it is typically used to refer to a class of verbs that are semantically bleached and which appear in a syntactically formed complex predicate (Brugman 2001, Brugman 2001, Butt & Geuder 2001, Bowern 2004). According to Butt (1995), one key defining characteristic of light verbs is that they have a main verb use, which contrasts with its light counterpart in bearing descriptive content (see also Svenonius 2008; cf. Tubino Blanco et al. 2014). In addition to these characteristics, light verbs may impose selectional restrictions on the types of complements and arguments they combine with (e.g., light verbs may require their complement be of a certain transitivity value), in contrast to auxiliary verbs and serial verbs, which do not impose selectional restrictions. For discussion of light verb constructions in the closely related Hiaki language (Taracahitan), see Tubino Blanco et al. (2014).

A set of constructions in Choguita Rarámuri fit the definition of light verb constructions. These constructions consist of two predicates, the first of which (the heavy predicate) contributes most of the lexical meaning, while the second (the light verb) contributes finite verbal categories, stance or valence. Choguita Rarámuri light verb constructions include the *noˈká* 'do' construction (§15.6.1.1), the *ˈní-* 'do' construction (§15.6.1.2), and the *iˈsì* 'do' construction (§15.6.1.3).

### **15.6.1.1 The** *noˈká* **'do' construction**

Like light verb constructions documented cross-linguistically, the verb *noˈká* has a main verb use and a light verb use. As a main verb form, *noˈká* means 'move', a predicate that describes a change in posture and which may undergo valence related alternations through the affixation of transitive and applicative suffixes that replace the final vowel of the stem (see §9.1.5). Examples of this predicate as a main, free standing verbal predicate are given in (63).

15.6 Complex predicates

(63) a. *ˈmá noˈkáli ˈmá* already *noˈká-li* move-pst 'S/he already moved.' 'Ya se movió.' <BFL 05 1:114/el> b. *niˈhê ˈmá nokoˈmêa niˈhê* 1sg.nom *ˈmá* already *noko-ˈmêa* move-fut.sg 'I will move.' 'Ya me voy a mover.' <SFH 05 1:80/el> c. *niˈhê ˈmí ˈtrôka noˈkèli niˈhê* 1sg.nom *ˈmí* 2sg.acc *ˈtrôka* truck *noˈk-è-li* move-appl-pst 'I will move the truck for you.'

'Te voy a mover la troca.' <SFH 05 1:80/el>

This predicate has a semantically bleached version found in multi-predicate constructions, acting as a light verb bearing inflection and following a verbal predicate that bears descriptive content (the descriptive (heavy) verb always precedes the light verb). This is exemplified in (64), where the heavy-light verb sequence is shown in square brackets.

(64) a. *napaˈwía noˈkáli ˈlé ˈétʃ͡i ˈnà biˈlé riˈhò aʔˈlì biˈlé tʃ͡aˈbôtʃ͡i ˈʃîʔi [napaˈwí-a* get.together-prog *noˈká-li* do-pst *aˈlé]* dub *ˈétʃ͡i* dem *ˈnà* dem *biˈlé* one *riˈhò* man *aʔˈlì* and *biˈlé* one *tʃ͡aˈbôtʃ͡i* mestizo *ˈsî* also 'A (Rarámuri) man and a *mestizo* (mixed mexican) man got together.' 'Se juntaron un hombre (Rarámuri) y un mestizo.' <SFH 06 choma(2)/tx>

b. *ˈnáp aʔˈlì ˈí biˈlá ... napaˈbûa noˈkíla ˈétʃ͡i tʃ͡aˈbôtʃ͡i aʔˈlì ˈétʃ͡i riˈhò ˈûa ba ˈnápi* rel *aʔˈlì* then *ˈí* here *biˈlá* really *[napaˈbû-a* gather-prog *noˈkí-la]* do-mpass *ˈétʃ͡i* dem *tʃ͡aˈbôtʃ͡i* mestizo *aʔˈlì* and *ˈétʃ͡i* dem *riˈhò* man *ˈjûa* with *ba* cl

'When the man and the mestizo were gathered.'

'Cuando los juntaron al hombre y al mestizo.' <SFH 06 choma(36)/tx>

### 15 Complex clauses and complex predication

c. *taˈmò ˈkútʃ͡uala ˈtʃ͡ó beˈnèma ˈtʃ͡ó ˈnà ˈtʃ͡útimi riˈkám ˈnàwa noˈká ˈnà oˈtʃ͡êra noˈká tʃ͡aˈbè kiˈʔà ba taˈmò* 1pl.nom *ˈkútʃ͡ua-la* children-poss *ˈtʃ͡ó* also *beˈnè-ma* learn-fut.sg *ˈtʃ͡ó* also *ˈnà* then *ˈtʃ͡ú=timi* Q=2pl.nom *riˈká=mi* like=dem *[ˈnàwa* arrive.prs *noˈká]* do.prs *ˈnà* then *[oˈtʃêr-a* grow-prog *noˈká]* do.prs *tʃ͡aˈbè* before *kiˈʔà* before *ba* cl 'Our children will learn how you all grew up like before, how you used to live.' 'Nuestros hijos también van a aprender cómo crecieron ustedes antes, como vivían.' {SFH in61:00:46.7}, {SFH in61:00:49.2} d. *nabiˈsûra noˈkítʃ͡ino [nabiˈsûr-a* form.line-prog *noˈkí-tʃino]* move-ev 'It sounds like they are forming a line.' 'Se oye que se andan acomodando en fila.' < BFL 07 el 30 04] e. *haˈré ko=ti ke tʃ͡iˈkó=timi risoˈtʃ͡í má=m kaˈtêw-a noˈká-la ˈrú? haˈré=ko=ti* some=emph=1pl.nom *ke* neg *tʃ͡iˈkó=timi* tʃ͡iˈkó=2pl.nom *risoˈtʃ͡í* caves *má=mi* already=dem *[kaˈtêwi-a* keep-prog *noˈká-la]* do-rep *ˈrú* say.prs 'Don't some say that you all used to keep (corn) like that in caves?' '¿Que no dicen que unos guardaban así en cuevas?' {SFH

in61:03:43.4}

As shown in these examples, the light verb construction with *noˈká* is attested with heavy predicates that encode activities, e.g., *napaˈwía* 'get together' (64a), *ˈnàwa* 'arrive' (64c), *nabiˈsûra* 'get in a line' (64d) and *kaˈtêwa* 'store' (64e). In contrast, other light verbs may select for other types of predicates (e.g., the light verb *ni-* 'do' selects for stative and mental attitude predicates, as described in §15.6.1.2 below).

The verb *noˈká* has a morphologically related form *onoˈká*, a plural/pluractional form derived through a prefix which is aligned in color with the vowel quality of the first vowel of the root (see §8.2 for more details about pluractional marking in this language). While the semantically full version of this predicate may be used in contexts involving multiple participants, it is not clear that the light verb *onoˈká* marks plurality. Examples of this predicate in light verb constructions is shown in (65).

15.6 Complex predicates

(65) a. *suˈnù ʃuˈwâbuka ˈtʃ͡útimi riˈkám iˈtʃ͡à onoˈká tʃ͡aˈbè suˈnù* corn *suˈwâbuka* everything *ˈtʃ͡ú=timi* Q=2pl.nom *riˈká=mi* like=dem *[iˈtʃà* sow.prs *onoˈká]* do.prs *tʃ͡aˈbè* before 'Corn, everything, how did you use to plant the crops before.' 'El maíz, todo, cómo sembraban antes.' {SFH in61:01:02.7} b. *ˈtʃ͡ú ˈjêni ˈtʃ͡émi kaˈjèni onoˈká ˈétʃ͡i suˈnù itʃ͡iˈsûa ˈétʃ͡i biˈlé baˈmíbiri tʃ͡aˈbè ...?*

*ˈtʃ͡ú* how *ˈjêni* much *ˈtʃ͡é=mi* also=dem *[kaˈjèni* yield.harvest.prs *onoˈká]* do.prs *ˈétʃ͡i* dem *suˈnù* corn *itʃ͡i-ˈsûa* sow-cond.pass *ˈétʃ͡i* dem *biˈlé* one *baˈmíbiri* year *tʃ͡aˈbè* before 'And how much corn would the harvest yield in a year when you

used to sow, before?'

'¿Y cuánto se les daba de maíz cuando sembraban en un año, antes?' {SFH in61:02:42.5}

c. *ˈpé riˈpòpa ˈbítimi ˈá kawiˈká naˈʔá onoˈká ba? ˈpé* just *riˈpòpa* back *ˈbi=timi* just=2pl.nom *ˈá* aff *kawi-ˈká* bring.wood-ger *[naˈʔá* make.fire *onoˈká]* do.prs *ba* cl 'Just with your backs did you used to carry wood, that's how you used to make fire?'

'Nomás en el lomo ustedes traían leña, ¿así hacían lumbre?' < SFH 06 in61(233)/in >

There seem to be no restrictions as to the inflectional marking the light verb might bear. In most cases, the light predicate is inflected for present tense (as in, e.g., (65)), but in the following examples, the light verb is marked as imperfective (66a) or past tense (66b).

(66) a. *aʔˈlìtimi ˈtʃ͡útimi iˈjêni kotimi? ˈpîri ˈûatimi iˈtʃ͡à onoˈká-i tʃ͡aˈbè ko? aʔˈlì=timi* and=2pl.nom *ˈtʃ͡ú=timi* Q=2pl.nom *iˈjêni=ko=timi* go.pl=emph=2pl.nom *ˈpîri* Q *ˈjûa=timi* with=2pl.nom *[iˈtʃà* sow *onoˈká-i]* do-impf *tʃ͡aˈbè=ko?* before=emph 'And how did you go? With what did you use to sow before?' '¿Y ustedes cómo iban? ¿Con qué sembraban antes?' {SFH in61:02:15.5}

### 15 Complex clauses and complex predication

b. *ˈtʃ͡útimi riˈkám oˈmáwa onoˈkáli ba ... ˈtʃ͡útimi riˈká=m kaˈjèna ba ... ˈtʃ͡ú=timi* Q=2pl.nom *riˈká=mi* like=dem *[oˈmáwa* make.party.prs *onoˈká-li]* do-pst *ba* cl *ˈtʃ͡ú=timi* Q=2pl.nom *riˈká=mi* like=dem *kaˈjèna* yield.harvest.prs *ba* cl 'How did you make the parties? How good were the harvests?' '¿Cómo hacían ustedes las fiestas? ¿Cómo se les daban las cosechas?' {SFH in61:00:53.4}, {SFH in61:00:56.0}

Finally, the light verb *noˈká* is also frequently found in constructions where the more descriptive predicate is an adjective. This is shown in (67), with the adjectival predicate *ˈtʃ͡á*, 'ugly'.

(67) a. *ˈétʃ͡i ˈmí ʃiˈpá-am-ti ˈwé ˈtʃ͡á noko-ˈká siˈpá-ame ru-ˈwá riˈké tʃ͡aˈbèi m=aˈní-a biˈlá aˈní ˈétʃ͡i* dem *ˈmí* dem *siˈpá-ame-ti* use.peyote-ptcp-nmlz *ˈwé* int *[ˈtʃá* ugly *noko-ˈká]* do-ger *siˈpá-ame* use.peyote-ptcp *ru-ˈwá* say-mpass *riˈké* like.that *tʃ͡aˈbèi* before *mi=aˈní-a* dem=say-prog *biˈlá* really *aˈní* say.prs 'That the peyote shamans would do bad things before, they would say.' 'Que los raspadores hacían cosas malas raspaban antes, decían.' < SFH 06 in61(580)/in > b. *ˈwé ˈtʃ͡á nokoˈká ʃiˈpáami riˈhòli tʃ͡aˈbèe ko maˈnía biˈlá aˈní ˈwé* int *[ˈtʃá* ugly *noko-ˈká]* do-ger *siˈpá-ame* use.peyote-ptcp *riˈhò-li* man-vblz *tʃ͡aˈbè=ko* before=emph *m=aˈní-a* dem=say-prog *biˈlá* really *aˈní* say.prs 'They would do many bad things the peyote shamans before.' 'Hacían muchas cosas malas antes para raspar.' < SFH 06

in61(601)/in >

### **15.6.1.2 The** *ˈní-* **'do' construction**

Another light verb construction in Choguita Rarámuri involves the bound verbal root *ˈní-*. As a main predicate, this verb is a copular predicate, as exemplified in (68) (see §13.2.1 for more details of copular predicates).

15.6 Complex predicates

(68) a. *ˈnè uˈmûala ˈníli ˈétʃ͡i ba? ˈnè* 1sg.nom *uˈmûa-la* great.grandfather-poss *ˈní-li* cop-pst *ˈétʃ͡i* dem *ba* cl 'Was he my great grandfather?' '¿Era mi bisabuelo él?' < SFH 06 in61(117)/in > b. *"tamuˈhê biˈlá ˈlîna ˈníma riˈké pa ˈétʃ͡i ˈlîna ˈtòli ba" ˈhê biˈlá aˈníli ba tamuˈhê* 1pl.acc *biˈlá* really *ˈlîna* but *ˈní-ma* cop-fut *riˈké* perhaps *pa* cl *ˈétʃ͡i* dem *ˈlîna* but *ˈtò-li* take-pst *ba"* cl *ˈhê* dem *biˈlá aˈní-li ba*

> really say-pst cl "'It was going to be us, but it took him", that's what she said.' "'Ibamos a ser nosotros y fue él" así dijo.' {LEL tx5:04:26.3}

The following examples show the verb *ˈní-* used as a light verb (69). Like in the constructions with *noˈká*, these light verb constructions involve an order where the descriptive predicate precedes the auxiliary verb. In these constructions, the light verb may only appear with past tense marking, as exemplified below.<sup>10</sup> The descriptive (heavy) predicate in the construction is marked present or progressive.


<sup>10</sup>In contrast, the copular verb *ˈni-* may bear other TAM markers (see §13.2.1).

### 15 Complex clauses and complex predication

*[ˈnâta* remember.prs *ˈní-li]* do-pst 'They said it was finished down there, I can barely remember.' 'Dicen que se hizo allá abajo, yo apenas me acuerdo.' {SFH tx12:01:46.4} d. *ˈpé ˈbí ke me riˈʃí ˈníli ˈtʃ͡ó aˈlé ... ˈpé* just *ˈbí* just *ke* neg *me* almost *[riˈsí* be.tired.prs *ˈní-li]* do-pst *ˈtʃ͡ó* also *aˈlé* dub 'But one would not get tired.' 'Nomás que casi no se cansaba uno.' < SFH 06 in61(235)/in >

e. *aʔˈlì ˈnà ˈpéni maˈjêa ˈníli ˈwé aʔˈlá saʔˈmîami ˈníla ˈlé tʃ͡oˈnà ˈétʃ͡i kaˈwì aʔˈlì* and *ˈnà* then *ˈpé=ni* just=1sg.nom *[maˈjê-a* believe-prog *ˈní-li]* do-pst *ˈwé* int *aʔˈlá* well *saʔˈmî-ame* humid-ptcp *ˈní-la* cop-rep *aˈlé* dub *tʃ͡oˈnà* there *ˈétʃ͡i* dem *kaˈwì* land 'I believe that land was very humid.' 'Yo creo que esa tierra estaba muy bien mojado.' {LEL tx130:02:21.0}

The light verb construction with *ˈní* 'do' selects for stative or mental attitude heavy predicates: *ˈla* 'think' (69a), *kaˈlé* 'like' (69b), *ˈnâta* 'remember' (69c), *riˈsí* 'be tired' (69d) and *maˈjê* 'believe' (69e). This stands in contrast with the constructions with *noˈká* where the light verb selects for activity predicates.

### **15.6.1.3 The** *iˈsì* **'do' construction**

A third light verb construction involves the verb *iˈsì* 'do'. As the main predicate in a clause, this verb has the meaning 'do', as exemplified in (70).

(70) a. *ˈhê ˈkwâ iˈsìli? ˈhê* who *ˈkwâ* who *iˈsì-li* do-pst 'Who did it?' '¿Quién lo hizo?' {MAF el1240:02:29.3} b. *ˈá ˈjén ˈá loˈmí tʃ͡í iˈsìa ba ˈá* aff *ˈjén* aff *ˈá* aff *loˈmí* know *tʃ͡í* how *iˈsì-a* do-prog *ba* cl 'I do know how to do that.' 'Sí se hacer eso.' {JLG co1236:03:43.0}

15.6 Complex predicates

As a light verb accompanying a heavy verb, the verb *iˈsì* bears tense/aspect inflection, while the descriptive predicate can only be marked present or progressive. As in the other light verb constructions described so far, the order is descriptive verb – light verb. This is shown in (71).

(71) a. *ˈpé biˈlá ˈpé kiˈlìm ˈtʃ͡óna iˈsìi ba ˈpé* little *biˈlá* really *ˈpé* little *kiˈlì=mi* slowly=dem *[ˈtʃón-a* smash-prog *iˈsì-i]* do-impf *ba* cl 'I would smash it slowly.' 'Le aplastaba despacito.' {BFL tx1:01:04.2} b. *ˈpé aʔˈlì biˈlá ˈpé kuˈrím kaˈjèna iˈsìa ruˈwá esˈkwêlatʃ͡i aʔˈlì ba ˈpé* just *aʔˈlì* later *biˈlá* really *ˈpé* just *kuˈrí=mi* recently=dem *[kaˈjèni-a* finish-prog *iˈsì-a]* do-prog *ru-ˈwá* say-mpass *esˈkwêlatʃ͡i* school *aʔˈlì* then *ba* cl

> 'That time it was said that they were just recently finishing the school.'

'Esa vez dicen que apenas estaba terminando la escuela.' {SFH tx12:02:35.8}

c. *aʔˈlì biˈlá ko ˈá riˈká ˈnà boˈtêja mokoˈʔôka biˈlá ko riˈpámi tʃ͡óm itʃ͡uˈkûba iˈsìli aˈlé ˈnà benˈtâantʃ͡i ba*

*aʔˈlì* and *biˈlá=ko* really=emph *ˈá* aff *riˈká* like *ˈnà* dem *boˈtêja* bottle *mokoˈʔô-ka* in.head.crown-ger *biˈlá=ko* really=emph *riˈpá-mi* up-supe *tʃ͡ó=mi* also=dem *[itʃuˈkûba* peek.prs *iˈsì-li]* do-pst *aˈlé* dub *ˈnà* dem *benˈtân-tʃ͡i* window-loc *ba* cl 'Like that, he was peeking through the window with the bottle in his head (like a crown).'

'Así como ya se asomaba por la ventana con la botella puesta.' {SFH tx152:02:14.4}

In these examples, the light verb is marked imperfective (71a), progressive (71b) and past (71c). In all instances, the heavy predicate is marked present tense or present progressive.

In the following example (in (72)), *iˈsì* is nominalized through an agentive participial marker and preceded by the adjectival predicate *ˈtʃ͡á* 'ugly', which contributes the main descriptive content of the construction. The light verb con-

### 15 Complex clauses and complex predication

struction is in turn the complement of a higher predicate (the motion verb *aˈjéna* 'go').

(72) *ˈtʃ͡á iˈsìkam ˈjénam koriˈmá ba [ˈtʃá* ugly *iˈsì-kame]* do-pst.ptcp *aˈjéna=mi* go.sg=dem *koriˈmá* korima *ba* cl "'The *korima* (fire bird) was bothering (lit. doing bad/ugly)".' "'Anduvo molestando (lit. haciendo feo/mal) el *korimá* (pájaro de fuego)".' {LEL tx5:03:43.0}

Whether the descriptive predicate is an adjective or a verb, the light verb in this construction appears to indicate that the event encoded by the construction is an activity.

### **15.6.2 Auxiliary verb constructions**

Auxiliary verbs are defined in this grammar as functional verbal elements that are combined with a semantically descriptive verbal predicate in a mono-clausal structure; the descriptive verbal predicate contributes lexical content to the construction, while auxiliary verbs may encode aspectual, temporal, modal or evidential information (Anderson 2006; Svenonius 2008). While auxiliary verb constructions and light verb constructions (and other complex predicate types) are proposed to be discrete classes of constructions (Butt & Geuder 2001) (cf. Anderson 2006), the boundary between them may not be sharp (Svenonius 2008). Two main criteria are assumed here in distinguishing auxiliary verb constructions from light verb constructions in Choguita Rarámuri: (i) while light verbs may impose selectional restrictions on their complements, auxiliary verbs do not impose such selectional restrictions nor do they contribute lexical-semantic (argument structural) content (Butt 1995); and (ii) while auxiliary verbs and light verbs both involve semantic bleaching, auxiliary verbs have less semantic content than light verbs or may lack descriptive content altogether (Butt 1995; see also Tubino Blanco et al. 2014).

Choguita Rarámuri features multi-verb constructions that may be analyzed as auxiliary verb constructions. These involve the paradigm of positional/posture verbs, which exhibit suppletion and affixation encoding number (singular vs. plural) and transitivity (stative, inchoative and causative) contrasts when functioning as main predicates in locative clauses. The morphological and syntactic properties of posture predicates are described in §13.2.3. The posture predicate paradigm is presented in Table 15.1.

### 15.6 Complex predicates


Table 15.1: Posture/positional predicates in Choguita Rarámuri

A subset of these predicates, namely stative and inchoative posture predicates, are deployed in auxiliary verb constructions encoding progressive aspect, indicating that an event takes place continuously over a given time frame. These constructions are analytic, with auxiliary verbs being separate prosodic words (the criteria for defining the prosodic word in Choguita Rarámuri are discussed in §11.1). The auxiliary verb bears tense marking (present, past or future), while the main lexical verb is inflected for present tense regardless of the tense marking on the auxiliary. Auxiliary verb constructions are exemplified in (73).

(73) a. *aʔˈlì tʃ͡i ko ˈétʃ͡i ˈnà ˈhônsa ko aʔˈlì biˈléna ˈtʃ͡ó bitiˈtʃ͡í siˈmíli ˈápu ko ˈnà baˈhîa moˈtʃ͡íli aʔˈlì* and *tʃ͡i=ko* tʃ͡i=emph *ˈétʃ͡i* dem *ˈnà* prox *ˈhônsa=ko* since=emph *aʔˈlì* and *biˈlé-na* one-incl *ˈtʃ͡ó* also *bitiˈtʃ͡í* house *siˈmí-li* gosg-pst *ˈnápi=ko* sub=emph *ˈnà* dem *[baˈhî-a* drink-prog *moˈtʃí-li]* sit.down.pl-pst 'And then from there he went to another house where people were drinking.'

### 15 Complex clauses and complex predication

'Y entonces de ahí se fue en otra casa donde estaban tomando.' {LEL tx5:01:43.4}

b. *ˈmá nataˈkêa buˈʔíli ˈnà biˈʔà roˈkò ˈmá* already *[nataˈkê-a* faint-prog *buˈʔí-li]* lie.down.sg-pst *ˈnà* then *biˈʔà* early *roˈkò* night 'He had already fainted before dawn.' 'Ya estaba desmayado en la madrugada.' {LEL tx5:04:03.7} c. *aʔˈlì ˈnè ko ˈmá bitiˈtʃ͡í ˈá buˈjèa aˈtí aʔˈlì* and *ˈnè=ko* 1sg.nom=emph *ˈmá* then *bitiˈtʃ͡í* house *ˈá* aff *[buˈjè-a* wait-prog *aˈtí]* sit.sg.prs 'And then I am waiting for her in the house.' 'Y entonces yo ya la estoy esperando en la casa.' {LEL tx19:01:17.9} d. *"tʃ͡ín oˈlá ko ˈétʃ͡i", ˈhê biˈlá ko ˈlàa aˈsáli ˈlé ruˈtûkuri ko ba tʃ͡í=ni* how=1sg.nom *oˈlá=ko* do.prs=emph *ˈétʃ͡i* dem *ˈhê* that *biˈlá=ko* indeed=emph *[ˈlà-a* think-prog *aˈsá-li]* sit.sg-pst *aˈlé* dub *ruˈtûkuri=ko* owl=emph *ba* cl "'That's how I did it to them" that's what he was thinking, the owl".' "'Así les hice a esos" eso estaba pensando, el tecolote".' {SFH tx152:07:01.9}

In contrast to light verbs, auxiliary verbs do not impose any selectional restrictions on the verb encoding the main lexical content of the clause, e.g., the descriptive verb of the construction may be an activity predicate (*baˈhî* 'drink' in (73a)), a telic (bounded) predicate (*nataˈkê* 'faint' in (73b)) or a stative predicate (*ˈlà* 'think' in (73d)). The progressive auxiliary verb construction encodes that an event is ongoing at some specified time frame internal to ongoing discourse: in (73a), an activity ('drinking') is carried out over a period of time (in this context, a drinking party). This construction may also denote that a state of affairs is transitory, rather than permanent: when combining with a stative predicate, this construction denotes a type of temporary activity, as in (73d), where a participant is in a state of thinking about an action carried out.

The verbs bearing inflection in these constructions are not completely devoid of the semantic contrasts they encode as main predicates in locative clauses. Specifically, number contrasts and some lexical distinctions are retained. In terms of their lexical meaning, auxiliaries may be selected in terms of the posture of referents, e.g., a man that has fainted lies down (73b), and the default collocation for human beings are the auxiliaries derived from the posture predicate 'sit'

15.6 Complex predicates

((73a) and (73c)) (for more discussion about default collocations and categorization involving posture predicates, see §13.2.3).

Number distinctions are also retained in auxiliary verbs. This is shown in (73a): the auxiliary employed *moˈtʃ͡í* encodes plural number as a positional verb and is used as an auxiliary in a context where the event described is a drinking party involving multiple participants; in contrast, in (73b) the auxiliary employed (*buˈʔí*) encodes singular number and is used in a context where the event involves a single participant (a man who is in the state of having fainted). Likewise, in (73c–d), the auxiliary verbs (*aˈtí* and *aˈsá*, respectively) encode singular number with singular subjects. Further examples of number contrasts in auxiliaries is exemplified in (74), where the auxiliaries in each case derive from the singular (*wiˈli*) and plural (*ˈhâwi*) forms of the positional predicate 'stand'.


In (74a), the auxiliary verb *wiˈli* encodes singular number in a construction denoting a transitory state of a location (the sandstone that is bloody), while in (74b) the auxiliary verb *ˈhâwi* encodes plural number in a context where an ongoing process (turning to face another side) is something that must be done by hearers of the advice provided in the narrative.

While number and other lexical contrasts are retained in auxiliation, not all forms in the posture/positional predicate paradigm are attested in auxiliary verb constructions. Specifically, there are no instances in the corpus where a causative positional predicate appears in an auxiliary verb construction. Furthermore, it appears the choice between a stative positional verb (*buˈʔí* 'lie down, sg', *aˈtí* 'sit, sg'

### 15 Complex clauses and complex predication

or *ˈhâwi* 'stand, pl') and an inchoative positional verb ( *moˈtʃ͡íwi* 'sit, pl' or *aˈsá* 'sit, sg') is arbitrary. That semantic bleaching in this construction is only partial is not surprising considering that posture/positional predicates are common lexical sources for auxiliary verb constructions encoding progressive aspect and that the multi-predicate constructions do not significantly differ semantically from their lexical sources in locative clauses (Bybee & Dahl 1989). But while the number contrasts are retained in auxiliary verb constructions, transitivity contrasts encoded by the posture predicates as main predicates are not retained in their auxiliary function.

### **15.6.3 Serial verb constructions**

Serialization forms complex predicates with specialized functions. Cross-linguistically, serial verb constructions can be characterized by the following properties (Foley & Olson 1985, Sebba 1987, Aikhenvald 2006):

	- a. Each serial verb construction contains at least two verbs without any overt marker of subordination or coordination.
	- b. Either verb of the construction could function as the predicate of its own clause.
	- c. The construction depicts what could be conceived as a single event.
	- d. The argument structure of the construction corresponds to that of a single clause, with a single internal and external argument.
	- e. The construction has a single set of tense, aspect, mood and polarity values.
	- f. Intonationally, serial verb constructions may behave like a single clause.

A relatively infrequent construction in Choguita Rarámuri involving motion verbs meets the criteria for serial verb constructions: (i) the verbal complex lacks any overt marker of subordination or coordination; (ii) each verb of the construction may also function in non-serial constructions as a main predicate; (iii) the two verbs form depict a single event; (iv) the overall argument structure corresponds to that of a single clause; (v) the construction features a single value for tense, mood, aspect and polarity; and (vi) the construction may have a prosodic contour that corresponds to a single Intonational Phrase (IP). These properties are illustrated in (76).

15.6 Complex predicates

(76) a. *ˈétʃ͡i ko ˈwé naˈlàʃia naˈwàli ˈétʃ͡i=ko* dem=emph *ˈwé* int *[naˈlà-si-a* cry-mot-prog *naˈwà-li]* arrive-pst 'She arrived crying.' 'Llegó llorando.' {LEL tx5:03:09.6} b. *ˈhê aˈnè aˈníʃia naˈwàli ˈétʃ͡i naˈmú niˈrà ʃuˈwá ba ˈá ruˈwè-li ˈhê* dem *aˈn-è* say-appl *[aˈní-si-a* say-mot-prog *naˈwà-li]* arrive-pst *ˈétʃ͡i* dem *naˈmú* something *niˈrà* relatives *suˈwá* everybody *ba* cl *ˈá* aff *ruˈw-è-li* say-pst 'A relative arrived saying, telling everybody.' 'Llegó diciendo un familiar, diciéndoles a todos.' {LEL tx5:04:07.0} c. *aʔˈlì ˈétʃ͡i ˈápu roˈwéa ˈúmi ko ˈá biˈlá ritiˈwá ˈtʃ͡ú tʃ͡uˈrú aˈtí aʔˈlì* and *ˈétʃ͡i* dem *ˈnápi* sub *[roˈwé-a* women.race-prog *ˈhúmi]=ko* run.pl.prs=emph *ˈá* aff *biˈlá* indeed *riti-ˈwá* see.-mpass *ˈtʃ͡ú* how *tʃ͡uˈrú* much *aˈtí* sit.sg.prs 'And then the ones running the ariweta race see how many things there are (lit. sit) (things people bet).' 'Y entonces las que andan corriendo carrera de ariweta ven que tanto va.' {LEL tx19:03:10.4} d. *apaˈlì ˈmá biˈlé ˈtòoru ne paˈtʃ͡ûsia iˈnâliro riˈpáti raˈbô ˈnápi* when *aʔˈlì* then *ˈmá* already *biˈlé* one *ˈtò-ru* take-pst.pass *ne* int *[paˈtʃû-si-a* drip-mot-prog *iˈnâli-li-ro]* go-pst-mov *riˈpá-ti* up-all *raˈbô* hill 'When one was already taken it goes dripping something by the top of the hill.' 'Cuando ya lleva uno va goteando algo por arriba del cerro.' {LEL tx5:06:17.1}

The meaning of this construction is that of motion with associated manner. In Choguita Rarámuri serial verb constructions, the open class verb is marked with progressive aspect, while the motion verb, the defining member of the construction, may be inflected with a variety of tense values (e.g., past tense (76a–b, d) or present tense (76c)).<sup>11</sup> Thus, the open class predicate bears argument marking for

<sup>11</sup>In example (76d) the final stem syllable and past tense suffix syllable have identical onsets, and undergo stem-suffix haplology (see §9.4.3.1).

### 15 Complex clauses and complex predication

the entire clause, while the motion predicate bears tense marking for the entire clause. The open class verb may additionally bear associated motion marking, as in (76a–b,d). The motion predicate may also be marked with an unproductive motion suffix, *-ro*, as in (76d).<sup>12</sup> Each verb is a separate prosodic word (in contrast to V-V incorporation constructions, where the two verbs are combined in a stem, equivalent to a single prosodic word, as described in §15.6.4 below).

A second construction In Choguita Rarámuri has properties of a serial verb construction and involves predicates of speaking. This construction is exemplified in (77).

(77) a. *aʔˈlì biˈlá tʃ͡oˈná tʃ͡uˈkúli ˈtʃ͡ó ˈlé bo baˈjèa aˈnía esˈta toˈwí tʃ͡oˈná paˈtʃ͡á atiˈkó maˈjêli ... ˈétʃ͡i riˈmò ba aʔˈlì* and *biˈlá* indeed *tʃ͡oˈná* there *tʃ͡uˈkú-li* be.bent-pst *ˈtʃ͡ó* also *aˈlé* dub *bo* emph *[baˈjè-a* call-prog *aˈní-a]* say-prog *ˈétʃ͡i* dem *ˈtá* def *toˈwí* boy *tʃ͡oˈná* there *paˈtʃ͡á* inside *ati=ˈkó* sit.sg=emph *maˈjê-li* believe-pst *ˈétʃ͡i* dem *riˈmò* frog *ba* cl 'And then the boy was there calling it (lit. calling saying), he thought that the frog was in there' 'Y entonces allí estuvo llamándole (lit. llamando diciendo) el niño, pensó que allí adentro estaba el sapo.' {SFH tx152:05:25.2} b. *aʔˈlì biˈlá ko ti ˈétʃ͡i ruˈká biˈlámi kom ˈnâra atʃ͡aˈní ba aʔˈlì* and *biˈlá=ko=ti* indeed=emph=1pl.nom *ˈétʃ͡i* dem *ru-ˈká* say-ger *biˈlá=mi=ko=ni* indeed=dem=emph=1sg.nom *[ˈnâri-a* ask-prog *atʃaˈní]* make.noise.prs *ba* cl 'Then saying I'd ask (lit. ask make noise) her.' 'Entonces diciendo yo le preguntaba (lit. pregunto hago ruido).' {SFH tx43:01:45.3}

In (77a), the second verb of the complex verb construction is the verb that defines the construction (*aˈní* 'say'), while the first verb describes the manner of speaking (*baˈjè* 'call'). Both verbs are marked with progressive aspect and describe a single event with a single internal and external argument. In (77b), the construction encodes that the act of speech (encoded with the verb *atʃ͡aˈní* 'make noise') involves asking.

<sup>12</sup>The cognate suffix in Mountain Guarijío (*-tó/ro*) is described with a distributive sense, e.g., 'to go do X in more than one place' or 'do X more than once' (Miller 1996: 165). It is not possible to determine if the Choguita Rarámuri *-ro* suffix has the same meaning given how infrequently it is attested in the corpus.

### 15.6 Complex predicates

Serial verb constructions are not generally documented in Uto-Aztecan languages, but some complex verb constructions are analyzed as involving serialization in Southeastern Tepehuan (Tepiman) (García Salido 2007b) and Northern Paiute (Numic) (Thornes 2011), where an open class verb combines with a verb of a restricted class (primarily motion predicates in Southeastern Tepehuan, and motion and posture predicates in Northern Paiute). The constructions analyzed as serial verb constructions in Northern Paiute involve a single phonological word, which can be characterized as verb-verb compounding or incorporation. A different set of constructions in Choguita Rarámuri have this formal property and are addressed next.

### **15.6.4 V-V incorporation (secondary verb constructions)**

Choguita Rarámuri features V-V incorporation construction. As discussed in §9.4, the Choguita Rarámuri verbal morphological structure includes a domain, namely the "aspectual stem", where suffixes encoding desiderative, associated motion and auditory evidential meanings are transparently related to independent verbs in the language. Table 15.2 lists these suffixes, their grammaticalized meanings and their independent lexical verb sources.

Table 15.2: Choguita Rarámuri aspectual suffixes and their lexical counterparts


Verbal roots attaching these suffixes have properties that are characteristic of complex predicates, with the derivational suffixes in Table 15.2 attaching to a verbal root of an open class and forming a verb cluster. In the resulting complex, the suffix may introduce its own arguments to the construction and take the embedded phrase as its semantic complement. The verb complex has a single value for tense, aspect, mood and polarity.

These structures resemble a multi-verb construction documented across Uto-Aztecan languages and known in the Uto-Aztecanist literature as the "secondary verb" construction (Crapo 1970; Thornes 2011). These constructions involve a verb which may be used as a main verb in a clause or may be appear phonologically bound to another verb. In their suffixed form, secondary verbs encode

### 15 Complex clauses and complex predication

aspectual or adverbial meanings. In Choguita Rarámuri, V-V incorporation is productive and involves two verbal predicates in sequence, the first of which is an open class verb followed by the defining member of the construction. V-V incorporation is exemplified in (78). Similar multi-verb constructions in other Uto-Aztecan languages have been alternatively analyzed as involving serialization (e.g., in Northern Paiute (Numic) (Thornes 2011)) or light verb constructions (e.g., in Hiaki (Taracahitan) (Tubino Blanco et al. 2014)).


In addition to the desiderative, associated motion and auditory evidential, indirect causative constructions also involve V-V incorporation: as detailed in §15.1.5 and exemplified in (79), the complement of the jussive predicate *nuˈlé* 'order, command' is a clause where the lower verb is additional marked with the jussive verbal affix *nula* 'order, command' deriving a co-lexicalized structure within the complement. This is another instance of V-V incorporation.

15.6 Complex predicates

	- aff some=emph wood indeed gather.with.hands-order *nulu-ˈrîwi* order-mpass *ˈru* say.prs *ba* cl *ˈkútʃ͡i* small *ˈkûruwi=ko* children=emph 'Some send them to bring wood, the children.' 'Unos los mandan a traer leña, a los niños.' {LEL tx73:00:31.2}

# **Appendix A: Verbal suffixes**


### A Verbal suffixes


### **A.1 The Derived Stem: inchoative and transitivity markers**

The first identifiable layer in the suffixation domain of morphologically complex verbs is the Derived Stem. This verbal domain includes semantically restricted, unproductive derivational suffixes (an inchoative suffix and three transitive suffixes). These suffixes are restricted to attach to a semantically defined class of verbs, change-of-state verbs.

### **A.1.1 Inchoative** *-bá*

The inchoative suffix is productively used with positional or stative predicates to indicate a dynamic change-of-state; a state is turned into a process, meaning 'to become X'. This stress-shifting suffix is exemplified in (1b, d).

	- b. *ˈmá aʰkaˈbátʃ͡anali*
		- *ˈmá* already *aʰka-ˈbá-tʃ͡a-na-li* be.sweet-**inch**-tr.plc-tr-pst 'S/he has already sweetened it.' 'Ya lo endulzó.' < BFL 05 2:56/el >

c. *ˈwé raˈtâame ˈhú ˈwé* int *raˈtâ-ame* be.hot-ptcp *ˈhú* cop.prs 'It is very hot.' 'Está muy caliente.'

A.1 The Derived Stem: inchoative and transitivity markers

d. *rataˈbámam pa rata-ˈbá-ma=mi* be.hot-**inch**-fut.sg=dem *pa* cl '(so) it will become hot' '(para que) se caliente' {LEL tx68:0:46.2}

### **A.1.2 Transitive** *-nâ*

The transitive *-nâ* suffix is a stress-shifting suffix that attaches to change-of-state predicates, increasing the valency of the verb stem. Stems derived with this transitive suffix are "alternating stems", where surface tonal patterns are conditioned by the stress properties of inflection markers: a HL tone if the inflection morpheme is stress-shifting and a L tone if the inflection morpheme is stress-neutral (see §11.3.7.3 for more details).

The following examples show transitive derivations with suffix *-nâ*: the intransitive (inchoative) forms is not suffixed (e.g., (2a)), and the corresponding transitive counterpart is marked with the transitive *-nâ* suffix (e.g., (2b)).

(2) a. *ˈmá tʃ͡iˈwáli siˈpútʃ͡a ˈmá* already *tʃ͡iˈwá-li* tear-pst *siˈpútʃ͡a* skirt 'The skirt already tore.' 'Ya se rompió la falda.' < SFH 07 1:17-21/el > b. *ˈá riˈwè! tʃ͡iwaˈnâra! ˈá* aff *riˈwè* leave.imp.sg *tʃ͡iwa-ˈnâ-ra* tear-tr-pot 'Leave it, you are going to tear it!' '¡Déjalo! ¡Lo vas a trozar!' < SFH 07 1:17-21/el >

### **A.1.3 Pluractional transitive** *-tʃa*

The pluractional transitive *-tʃ͡a* suffix is a stress-neutral transitive suffix that can attach to the same change-of-state predicates that may be marked with the transitive *-nâ* suffix described in Appendix A.1.2. The suffix, historically reconstructed to Proto-Uto-Aztecan (Heath 1978), may encode (i) an event that is performed repeatedly (the succession or discernible, discrete events) or (ii) that more than one entity is affected by an event. Of limited productivity in Choguita Rarámuri,

### A Verbal suffixes

this stress-shifting suffix is found with a pluractional sense. Compare the transitive form with suffix *-na* in (3b) and the pluractional transitive form with suffix *-tʃ͡á* in (3c) of the same base predicate *kuʔˈrí* 'to turn'.


Non-pluractional uses of transitive *-tʃ͡a* can also be found. An example of this is provided in (4b):

	- b. *niˈhê rataˈbátʃ͡ama koˈʔámi niˈhê* 1sg.nom *rata-ˈbá-tʃ͡a-ma* heat-inch-tr.pl-fut.sg *koˈʔ-ámi* eat-ptcp 'I'm going to heat up the food.' 'Voy a calentar la comida.' < LEL 06 4:151/el >

A.2 The Syntactic Stem: causative and applicative markers

### **A.1.4 Transitive** *-bû*

There is a third transitivizing suffix, *-bû*, which is also unproductive and lexically restricted. This stress-shifting suffix is exemplified in (5b):<sup>1</sup>

(5) a. *toˈwí ˈmá ˈmóli toˈwí* boy *ˈmá* already *ˈmó-li* go.up-pst 'The boy already went up.' 'Ya se subió el niño.' < BFL 06 4:189/el > b. *ˈmáni moˈbûuro ˈmá=ni* already=1sg.nom *mo-ˈbû-ru* go.up-tr-pst.pass

'I was already taken up.'

'Ya me subieron.' < BFL 06 4:189/el >

### **A.2 The Syntactic Stem: causative and applicative markers**

After the derived stem domain, the next verbal zone in the Choguita Rarámuri verbal complex is the "syntactic stem" which includes suffixes in suffix positions S3 to S5 that encode valence-increasing operations, namely applicative and causative operations.

### **A.2.1 Applicatives**

### **A.2.1.1 Applicative** *-ni*

The Applicative *-ni* suffix increases the valency of the verb, adding a benefactive argument ('to do X for Y'). This suffix is unproductive and lexically conditioned by the roots to which it attaches. The contrast between a basic, two-place base predicate and its applicative derivation is exemplified in (6b).

(6) a. *ˈnè ˈmá ˈʔwîma suˈnù ˈnè* 1sg.nom *ˈmá* now *ˈʔwî-ma* harvest-fut.sg *suˈnù* corn 'I'll harvest corn now.' 'Ya voy a pizcar maíz.' < LEL 06 4:151/el >

<sup>1</sup>Lengthening of the transitive suffix vowel in example (5b) is triggered by the past passive suffix. This effect is discussed below (§9.4.3.3).

### A Verbal suffixes

b. *ˈʔwînimon oˈlá ˈnè ˈjéra suˈnù ˈʔwî-ni-ma=ni* harvest-appl-fut.sg=1sg.nom *oˈlá* cer *ˈnè* 1sg.nom *ˈjé-ra* mom-poss *suˈnù* corn 'I will harvest the corn for my mom.' 'Le voy a pizcar el maiz a mi mamá.' < BFL 06 5:146/el >

### **A.2.1.2 Applicative** *-si*

The suffix *-si* is another unproductive, lexically conditioned stress-neutral applicative marker that increases the valency of the verb by adding a benefactive argument. This applicativesuffix is exemplified in (7b):

(7) a. *ˈpáka ˈpá-ka* throw-imp.sg 'Throw it!' '¡Tira!' < BFL 06 5:147/el > b. *taˈmí ku ˈpáʃiri peˈlôta taˈmí* 1sg.acc *ku* rev *ˈpá-si-ri* thow-appl-imp.sg *peˈlôta* ball 'Throw the ball back at me!' '¡Tírame la pelota de vuelta!' < BFL 06 5:147/el >

### **A.2.1.3 Applicative** *-wi*

A third Applicative suffix in position S3 is *-wi*, another stress-neutral, unproductive suffix that adds a benefactive argument to a transitive predicate. This suffix is exemplified in (8b):

	- *ˈmí=ni* 2sg.acc=1sg.nom *naˈpítʃ͡i* fire *aˈtʃ͡wi-wi-ma* sit.appl-appl-fut.sg *oˈlá* cer *toˈwí* boy 'I will sit down your boy next to the fire.' 'Te voy a sentar al niño cerca de la lumbre.' < BFL 06 6:146/el >

A.2 The Syntactic Stem: causative and applicative markers

### **A.2.2 Causative** *-ti*

The Causative *-ti* suffix is a stress neutral suffix that introduces an agent (causer) argument to the argument structure of a predicate. Causativization applies to both intransitive and transitive verbs. In the causative construction exemplified in (9b), the object corresponds to the subject argument of its basic, non-causative counterpart. The introduced agent argument causes the undergoer to perform the activity described by the verbal root.

	- *ˈnè=mi* 1sg.nom=2sg.acc *riˈmê-ni-ra* make.tortillas-appl-pot 'I can make you tortillas.' 'Yo te hago tortillas.' < BFL 08 1:161/el > b. *ˈmín ˈnè onoˈlá riˈmêentima*
		- *ˈmí=ni* 2sg.acc=1sg.nom *ˈnè* 1sg.nom *ono-ˈlá* father-poss *riˈmê-ni-ti-ma* make.tortillas-appl-caus-fut.sg 'I will make you make tortillas for my dad.' 'Te voy a hacer que le hagas tortillas a mi papá.' < BFL 08 1:161/el >

The Causative suffix has two lexically determined allomorphs, *-ti* and *-ri*. The allomorphy is also partially phonologically determined, since there is a phonological process that devoices voiced/lenis consonants after another consonant (a derived environment stemming from stress-conditioned syncope) (cf. §3.4.4). Examples of the distribution of allomorph *-ti* are provided in (10).

(10) a. *ˈlàantiki ˈlàna-ti-ki* bleed-caus-pst.ego 'I made them bleed.' 'Lo hice sangrar.' < SFH 05 1:102/el > *\*làan-ri-ki* b. *sikiˈréptiki sikiˈrép-ti-ki* cut-caus-pst.ego 'I made them cut.'

A Verbal suffixes

> 'Lo hice cortar.' < BFL 05 1:113/el > *\*sikiˈrép-ri-ki* c. *oʔˈpéʃtia oʔˈpés-ti-a* vomit-caus-prog 'S/he is making them vomit.' 'Los está haciendo vomitar.' < BFL 05 1:136/el > *\*oʔˈpés-ri-a*

The Causative *-ti* suffix is highly productive, displaying no restrictions as to the bases to which it can attach.

### **A.2.3 Applicative** *-ki*

The Applicative suffix *-ki* (S5) is another productive, stress-neutral suffix. This suffix introduces an additional argument to one-place or two-place predicates. The argument introduced is a benefactive or malefactive argument,<sup>2</sup> i.e., the object can be favorably or adversely affected. This suffix is exemplified in (11).


In (11b), the Applicative introduces a benefactive argument as an unmarked object (*ˈmí* '2sg.acc') to a basic transitive predicate (an argument which would be expressed through a postpositional phrase in a non-applicative construction).

<sup>2</sup>The cognates of Choguita Rarámuri Applicative suffixes in the closely related Mountain Guarijío (Miller 1996) introduce other semantic roles in addition to benefactive/malefactive (e.g., instrumental). There is however no evidence that the suffix -ki or any of the other Applicatives in Choguita Rarámuri introduce semantic roles other than the benefactive/malefactive.

A.3 The Aspectual Stem

The distribution of this suffix does not exhibit restrictions on its distribution, unlike the Applicative suffixes in suffix possition S3 (described in Appendix A.2.1s above).

### **A.3 The Aspectual Stem: desiderative, associated motion and evidential markers**

Another layer of the verbal stem is the Aspectual Stem, composed of suffixes in positions S6 to S9, marking desiderative, associated motion, and auditory evidential. The prosodic and morphosyntactic properties of these suffixes are discussed in §11.5.3 and §15.6.4, respectively. As discussed in §15.6.4, these suffixes may be analyzed as involving V-V incorporation.

### **A.3.1 Desiderative** *-nále*

The disyllabic Desiderative suffix *-ˈnále* is a stress-shifting suffix of agent-oriented modality. Derived from the verb *naˈkí* 'want', it has the meaning 'X wants to/feels like doing X', where the 'wanter' and the subject of the desideratum predication are correferent (when these two arguments are not correferent, a periphrastic construction must be used). Examples from context are shown in (12).

(12) a. *ˈnè biˈlá niˈjúrka seˈbánili ba ˈnè* int *biˈlá* indeed *niˈjúri-ka* fight-ger *seˈbá-nale* reach-desid *ba* cl 'He really wanted to reach it (against all odds).' 'Lo quería alcanzar realmente (a fuerzas).' {BFL tx191:1:31.6} b. *aʔˈlì ˈnà moʔoˈtʃ͡íki tʃ͡uˈkúrili tʃ͡apiˈnála aʔˈlì* and *ˈnà* dem *moʔoˈtʃíki* headboard *tʃ͡uˈkúri-li* go.around-pst *tʃ͡api-ˈnále-a* grab-desid-prog 'And then he was going around near the headboard wanting to get him.' 'Y entonces andaba por la cabecera queriéndolo agarrar.' {LEL tx5:1:11.3}

As discussed and exemplified in §11.5.3, the desiderative and other affixes belonging to the aspectual stem undergo truncation when attaching CV suffixes encoding TAM categories. In the case of the desiderative suffix, it's truncated form is *-na*.

### A Verbal suffixes

### **A.3.2 Associated motion** *-simi*

The Associated Motion suffix *-simi* is a stress-neutral suffix derived from the freestanding motion verb *simi* 'go (sg.)' (in it's truncated form, this suffix surfaces as *-si*). Verbs marked with this suffix encode an event that is carried out while in motion (e.g., 'X goes along doing V'). This is exemplified in (13).

	- b. *toˈwí ˈwé ˈnârisimi buʔuˈtʃ͡ími toˈwí* boy *ˈwé* int *ˈnâri-simi* ask-mot *buʔuˈtʃ͡ími* road 'The boy is going along the road asking a lot.' 'El niño va preguntando muchas cosas por el camino.' < SFH 08 1:148/el >

### **A.3.3 Auditory evidential** *-tʃane*

The auditory evidential *-tʃ͡ane* suffix is a productive epistemic modality marker that indicates that the evidence of the proposition encoded by the predicate has an auditory (i.e. non-visual) source ('it sounds like X is taking place'). This stressneutral suffix is exemplified in (14).

(14) a. *ˈtʃ͡éti toˈrí ˈmá toˈrétʃ͡ane ˈtʃ͡éti* def.bad *toˈrí* chicken *ˈmá* already *toˈré-tʃ͡ane* cackle-ev 'It sounds like the chicken are already cackling' 'Ya se oyen cacarear las gallinas' < SFH 08 1:160/el > b. *tʃ͡oˈnítʃ͡ane maˈtʃ͡í tʃ͡oˈní-tʃ͡ane* fight-ev *maˈtʃ͡í* outside 'It sounds/it seems like fighting is happening outside.' 'Se oye/parece que pelean afuera.' < BFL 08 1:17/el >

In these constructions, the source for the evidence is the noise generated by the event itself that the predicate describes (e.g., 'cackling' or 'fighting' in (14a) and

### A.4 The Finite Verb: voice, tense, aspect and mood markers

(14b), respectively). The evidence can also come indirectly from another event, as in example (15), where the speaker infers that dancing will take place because of other non-visual cues (i.e., sound of the rattles used in the dance, people talking about starting to dance, etc.).

(15) *ˈnápi ˈlé ˈmá awiˈmêt͡ʃani ˈnápi* sub *aˈlé* dub *ˈmá* already *awi-ˈmê-t͡ʃani* dance-fut.sg-ev 'It sounds like they are about to dance.' 'Se oye como que van a bailar.' < SFH 07 1:140/el >

As the desiderative and associated motion suffixes, the auditory evidential suffix undergoes truncation when attaching outer CV suffixes encoding TAM distinctions. The truncated form of the auditory evidential suffix is *t͡ʃa*.

### **A.4 The Finite Verb: voice, tense, aspect and mood markers**

The final layer of morphology in inflected verbs in Choguita Rarámuri involves suffixes in positions S9–S11 , the Finite Verb level suffixes. The inflectional categories encoded by these suffixes involve mood distinctions (including imperative and reportative), voice, tense, and aspect, with number and person marginally conflated with tense/aspect in portmanteaux suffixes. This final layer of morphology is required in all inflected verbs.

### **A.4.1 Passive**

### **A.4.1.1 Past passive** *-ru*

This suffix is a productive marker that encodes the object argument of the active transitive base has been promoted to subjecthood, while also encoding past tense. The subject of the active construction is not overtly expressed in the corresponding passive construction. As discussed in §9.4.3.3, the past passive suffix triggers lengthening of a preceding stressed syllable. Example (16b) illustrates the past passive sense and concomitant lengthening in the stressed vowel of the base.

(16) a. *tòsnaleni tò-si-nale=ni* take-mot-desid=1sg.nom 'I want to go along taking them.' 'Quiero írmelas llevando.' < BFL 06 5:149/el >

### A Verbal suffixes

b. *ˈtòoru grabaˈdôra ˈtò-ru* take-pst.pass *grabaˈdôra* recorder 'The recorder was taken.' 'Se llevaron la grabadora.' < SFH 08 1:45/el >

### **A.4.1.2 Future passive** *-pa*

This is a productive, stress-shifting suffix that concomitantly marks a passive derivation and future tense. The following example illustrates the contrast between a basic active construction (17a) and a future passive construction (17b).


The onset of this suffix undergoes lenition (for more details on this process, see §3.5.1 above).

### **A.4.1.3 Medio-passive** *-rîwa, -wá*

This suffix is used in constructions where the actor participant is backgorunded or left unspecified, and the undergoer participant is emphasized. The mediopassive suffix has two allomorphs, *-riwa* and *-wa*. Both allomorphs are stressshifting. Examples (18) illustrate the medio-passive morphological construction:

(18) a. *aˈnáwka biˈlá rupuˈnáruwa ba aˈnáwi-ka* measure-ger *biˈlá* really *ripu-ˈná-riwa* tear-tr-mpass *ba* cl 'While measuring it, it is tore.' 'Midiendo se troza.' {BFL tx1:1:21.4}

A.4 The Finite Verb: voice, tense, aspect and mood markers

b. *ˈnápi ˈlé riˈmênuwa ˈtʃ͡ém fedeˈrîko reˈmê ˈnápi* like *aˈlé* dub *riˈmê-ni-wa* tortillas-appl-mpass *ˈtʃ͡ém* Mr. *fedeˈrîko* Federico *reˈmê* tortillas 'It seems like tortillas are being made for Mr. Federico.' 'Como que le hacen tortillas a Don Federico.' < SFH 07 2:69-72/el >

### **A.4.1.4 Conditional passive** *-sûwa*

The conditional passive suffix *-sûwa* is a stress-shifting suffix that is productively used in complex clauses cumulatively expressing a conditional relationship and passive voice. The predicate marked with the conditional passive is the predicate of the protasis clause (describing the condition), not the apodosis (describing the potential result). An active-conditional construction is contrasted with a passive conditional construction in (19).


### **A.4.2 Future**

### **A.4.2.1 Future singular** *-ˈmêa, -ma*

There are two future tense suffixes in Choguita Rarámuri: *-mêa* ~*-ma*, for future, singular subject, and *-pô* for future, plural subject. Historically, these suffixes developed from Proto-Sonoran *\*mi(l)a* 'go, run, sg', and *\*po* 'go, run, pl' (Miller 1996: 133). The future singular suffix has an unstressed allomorph (*-ma*) and a stressed allomorph (*-mêa*). Both the unstressed and stressed allomorphs of the future singular suffix are exemplified in (20).

### A Verbal suffixes

(20) a. *ˈhê ˈná=ni siˈpútʃ͡a sipuˈ-tá-mo ˈlá ˈhê* it *ˈná=ni* prox=1sg.nom *siˈpútʃ͡a* skirt *sipuˈ-tá-ma* skirt-vblz-fut.sg *oˈlá* cer 'I will wear this skirt.' 'Me voy a poner esta falda.' < BFL 07 Sept 6/el > b. *ˈmá mukuˈmêa raˈjénali ˈmá* already *muku-ˈmêa* die-fut.sg *raˈjénali* sun 'There will be an eclipse.' (lit. 'The sun will die.') 'Va a haber un eclipse.' (lit. 'se va a morir el sol.') < SFH 05 2:63/el >

As described in §10.8.4, Choguita Rarámuri has epistemic modality markers that indicate the degree of certainty speakers have towards the actuality of an event. These modal particles are frequently found in future tense constructions, as exemplified in (20a). This example also illustrates the phonological effect that these particles have on the inflected verb's final vowel, namely vowel deletion. Forms lacking such particles have a neutral interpretation with respect to the speaker's commitment to the expectation that the event encoded by the predicate will take place or not in the future.

### **A.4.2.2 Future plural** *-pô*

The future plural suffix *-pô* is a stress-shifting suffix used when the subject is either first or second person plural. Clauses with a third person plural subject may optionally be marked with the future plural suffix or the future singular suffix. The future plural suffix is exemplified in (21).

(21) a. *ke naˈkíu-po ruˈwá taˈmí ˈwísia ruˈwá ˈétʃ͡i arˈiwála ke* neg *naˈkíwi-po* allow-fut.pl *ru-ˈwá* say-mpass *taˈmí* 1pl.acc *ˈwí-si-a* take-mot-prog *ru-ˈwá* say-mpass *ˈétʃ͡i* dem *arˈiwá-la* soul-poss 'It's said that we can't let (the *korimáka*) get to us, because they say it goes along taking our souls.' 'Dicen que no hay que dejarnos de él (del *korimáka*) porque dicen que nos roba el alma.' {LEL tx5:5:00.6}

A.4 The Finite Verb: voice, tense, aspect and mood markers

b. *ruˈbô ru-ˈbô* say-fut.pl 'They will say something.' 'Van a decir.' < SFH 04 1:27/el >

When used with the first person plural, the construction has a hortative reading ('let us do X'). The hortative use of the future plural suffix is illustrated in (22).


### **A.4.3 Motion imperative** *-mê*

The Motion Imperative suffix *-mê* is a stress shifting suffix. It is a productive suffix that often occurs in conjunction with the imperative suffix *-sa* (in position S8). Motion Imperative constructions with the suffix *-mê* have the meaning 'go and do X!', used for a single addressee. When unstressed, the suffix vowel reduces to *i* or undergoes complete deletion, following the general unstressed vowel reduction and deletion processes operating in the language (§5.2). This suffix is exemplified in (23).

(23) a. *ˈàamsa ˈà-me-sa* give-mot.imp-imp.sg 'Go give it to her!' '¡Ve y dáselo!' < SHF 04 1:112/el >

### A Verbal suffixes

b. *iˈʃîmi iˈsî-mi* urinate-mot.imp 'Go and urinate!' '¡Ve a orinar!' < BFL 08 1:13/el > c. *ˈjúrka osiˈmêra ˈlé ˈjúri-ka* take-imp.sg *osi-ˈmê-ra* write-mot.imp-pot *aˈlé* dub 'Go take him to see if he writes.' 'Ve y llévalo a ver si escribe.' < BFL 08 1:94/el >

When there are multiple adressees, the motion imperative construction involves the stress-shifting suffix *-pi* (with stress-shifting allomorph *-bô*), followed by the imperative plural suffix *-si*. This is exemplified in (24).

(24) a. *osiˈbôsi*

*osi-ˈbô-si* write-mot.imp.pl-imp.pl 'You all go and write' '¡Vayan a escribir!' < BFL 05 2:94/el > b. *taˈmí ku ˈàkipisi*

*taˈmí* 1sg.acc *ku* rev *ˈà-ki-po-si* look.for-appl-mot.imp.pl-imp.pl 'You all go and look for it for me!' '¡Vayan a buscármelo!' < BFL 08 1:164/el >

### **A.4.4 Conditional** *-sâ*

This is a productive, stress-shifting suffix used in constructions that express a conditional relationship in the active voice (contrast with the conditional passive suffix described in Appendix A.4.1.4 above). The verbal predicate marked with the conditional suffix is the predicate of the protasis clause. Steele (1975) reconstructs the cognate form of this suffix for Proto-Uto-Aztecan as meaning "must/speaker wish" (1975: 216). This stress-shifting suffix is exemplified in (25).

(25) a. *ˈwé waˈrínami ˈnísako, ˈá mahaˈwá ˈwé* int *waˈrín-ame* light-ptcp *ˈní-sa=ko* cop-cond=emph *ˈá* aff *mahaˈwá* be.affraid 'If she (the other runner) is really fast, she gets affraid' A.4 The Finite Verb: voice, tense, aspect and mood markers

'Si es muy ligera (la otra corredora), sí le tiene miedo' {LEL tx19:0:45.1}

b. *riˈhé ukuˈsâ ro, tʃ͡ú t͡ʃé=timi riˈkám méra? riˈhé* hail *uku-ˈsâ* rain-cond *ro,* Q *tʃ͡ú* how *tʃ͡é=timi* how=1pl.nom *riˈká=mi* like=dem *mé-ra?* scare.away-pot 'And when it would hail? How did you guys scare it away?' '¿Y cuando llovía granizo? ¿Cómo lo espantaban?' < SFH 07 in 243/in >

### **A.4.5 Irrealis**

### **A.4.5.1 Irrealis singular** *-mê*

The irrealis singular suffix is used in constructions where the speaker has no certainty that a particular event will take place in the future, or if a particular event holds true in a hypothetical or contingent world. This stress-shifting suffix is highly productive (I have not documented any restrictions on its occurrence), and is used when the subject argument is singular. Examples of its use are presented in (26).


### A Verbal suffixes

### **A.4.5.2 Irrealis plural** *-pi*

Irrealis constructions with a plural subject argument are marked with the suffix -pi. This suffix is stress-neutral and, like the irrealis singular suffix described above, is highly productive. This suffix has two allomorphs, with a voiced and a voiceless stop onset (*-pi* and *-bi*). Examples are shown in (27).

(27) a. *ˈmá ˈtôbi ˈlé ma ˈmá* already *ˈtô-bi* bury-irr.pl *aˈlé* dub *ma* perhaps 'Maybe they will bury it already.' 'A la mejor ya lo van a enterrar.' < SFH 08 1:3/el > b. *koʔˈnálpi ˈléti ˈmáo koʔ-ˈnále-pi* eat-desid-irr.pl *aˈlé=ti* dub=1pl.nom *ˈmáo* perhaps 'Perhaps we might want to eat.' 'A lo mejor vamos a querer comer.' < BFL 06 5:140/el >

### **A.4.6 Potential** *-râ*

This suffix is used in constructions expressing the possibility of occurrence of an event, ability or wishes (with an optative reading). This is a stress-shifting suffix and is exemplified in (28).


A.4 The Finite Verb: voice, tense, aspect and mood markers

### **A.4.7 Imperative**

### **A.4.7.1 Imperative singular** *-kâ*

Imperatives may be marked through the bare stem, but there are also affixal exponents of imperative mood. One of such markers is suffix *-kâ*, a productive, stress-shifting suffix. This suffix is exemplified in (29).

(29) a. *ˈkíti naˈlàka! ˈkíti* neg *naˈlà-ka* cry-imp.sg 'Don't cry!' '¡No llores!' < BFL 05 2:89/el > b. *ˈwé simiˈkâ! ˈwé* int *simi-ˈkâ* go.sg-imp.sg 'Go!'

'¡Ve!'

### **A.4.7.2 Imperative singular** *-sâ*

Another imperative suffix used in constructions where the addressee is singular is *-sâ*. This stress-shifting suffix is exemplified in (30).

(30) a. *koʔˈsâ! koʔ-ˈsâ* eat-imp.sg 'Eat!' '¡Come!' b. *ˈmà-sa ˈmà-sa* run-imp.sg 'Run!' '¡Corre!' < BFL 04/11/06/el >

### **A.4.7.3 Imperative plural** *-sì*

Imperative constructions where the are multiple addressees are distinguished from imperatives with a single addressee with a productive, stress-shifting suffix, *-sì*. Examples of this suffix are provided in (31).

### A Verbal suffixes

	- b. *taˈmí ku riˈwíisi taˈmí* 1sg.acc *ku* rev *riˈwí-i-si* find-appl-imp.pl 'You all find it for me!' '¡Vayan a encontrármelo!' < BFL 08 1:16/el >

### **A.4.8 Reportative**

The reportative suffix is an evidential suffix that indicates that the speaker's source of information is hearsay. This productive marker, also used in direct quotation constructions, is a stress-neutral suffix which is added to the dependent verb of the complex sentence. This switch-reference system is restricted, as it is not generalized to all constructions involving dependent clauses in Choguita Rarámuri.

### **A.4.8.1 Reportative different subject** *-la*

When the notional subjects are not correferential, the dependent verb suffixes the different referent reportative *-la* suffix (32).

(32) a. *ˈá biˈlá oˈkám tʃ͡aˈnía ne ka ˈhémi isiˈmâtala ruˈá tʃ͡aˈbè ˈá* aff *biˈlá* really *oˈká=mi* many=dem *tʃ͡aˈní-a* sound-prog *ne* int *ka* ka *ˈhémi* here *isiˈmâta-la* pass.pl-rep *ru-ˈwá* say-mpass *tʃ͡aˈbè* before 'Many people<sup>i</sup> say that they<sup>j</sup> used to pass through here long time ago.' 'Muchas personas<sup>i</sup> dicen que por aquí pasaban<sup>j</sup> mucho antes.' {LEL tx223:4:17.6} b. *tʃ͡iˈnà ba ˈétʃ͡i biˈlá ˈtòola ruˈá aliˈwâla ba ˈétʃ͡i* dem *ˈnà* dem *ba* cl *ˈétʃ͡i* dem *biˈlá* really *ˈtò-la* take.pst.pass-rep *ru-ˈwá* say-mpass *aliˈwâ-la* soul-poss *ba* cl

> 'They<sup>i</sup> say that that one<sup>j</sup> got his soul stolen there'

'Cuentan<sup>i</sup> que a ese<sup>j</sup> ahí le llevó el alma' < BFL rihói mukúri 6/tx >

A.4 The Finite Verb: voice, tense, aspect and mood markers

c. *ˈnápu riˈká ˈlátimi ˈwé uˈkúla ruˈá ˈníam tʃ͡aˈbèe=ko ba ní ˈnápi* sub *riˈká* like *oˈlá=timi* cer=2pl.nom *ˈwé* int *uˈkú-la* rain-rep *ru-ˈwá* say-mpass *ˈní-ame* cop-ptcp *tʃ͡aˈbè=ko* before=emph *ba* cl *ní* int 'So you all say that it used to rain a lot ling time ago.' 'Pues así como dicen ustedes que llovía mucho antes.' {SFH in243:0:21.0}

### **A.4.8.2 Reportative same subject** *-lo*

When the notional subjects are correferential, the dependent verb is marked with the same referent reportative *-lo* suffix (33).

(33) a. *ˈá biˈláko aˈní ˈmâgre neˈhê amaˈtʃ͡îkolo ruˈá ˈá* aff *biˈlá=ko* really=emph *aˈní* say.prs *ˈmâgre* nuns *neˈhê* 1sg.nom *amaˈtʃ͡î-ko-lo* pray-appl-rep *ru-ˈwá* say-mpass 'The nuns<sup>i</sup> say that they<sup>i</sup> prayed for me.' 'Las monjas<sup>i</sup> dicen que (ellas<sup>i</sup> ) me rezaron.' b. *maˈnuêliko ˈwé biˈlá riˈkúlo ˈrú maˈnuêli=ko ˈwé biˈlá riˈkú-lo ˈrú*

Manuel=emph int really get.drunk.sg-rep.ss say.prs 'Manuel<sup>i</sup> says he<sup>i</sup> got drunk.' 'Manuel<sup>i</sup> dice que (él<sup>i</sup> ) se emborrachó.'

### **A.4.9 Past** *-li*

Past tense is marked by the suffix *-li*, a stress-neutral suffix. The past perfective both situates the event in a point prior to the time of the speech act and indicates that the event has been completed. Examples of this construction are given in (34).

(34) a. *aʔˈlìko ˈmá biˈlá ʃiˈnêami wiʰˈkâ ˈsíli tʃ͡oˈnà ˈétʃ͡i ˈná ˈétʃ͡i riˈhòi biˈtêritʃ͡i aʔˈlì=ko* and=emph *ˈmá* already *biˈlá* really *siˈnê-ame* every-ptcp *wiʰˈkâ* many *ˈsí-li* arrive-pst *tʃ͡oˈnà* there *ˈétʃ͡i* dem *ˈná* prox *ˈétʃ͡i* dem *riˈhòi* man *biˈtêritʃ͡i* house 'And then everybody arrived there, to that man's house.' 'Y ya llegaron todos ahí a la casa de ese señor.' {LEL tx32:11:43.4}

### A Verbal suffixes

b. *ˈhê aˈnè aˈníʃia naˈwàli ˈétʃ͡i naˈmú niˈrá ʃuˈwá ba ˈá ruˈwèli ˈhê* it *aˈn-è* say-appl *aˈní-si-a* say-mot-prog *naˈwà-li* arrive-pst *ˈétʃ͡i* dem *naˈmú* thing *niˈrá* relative *suˈwá* everybody *ba* cl *ˈá* aff *ruˈw-è-li* say-appl-pst 'He arrived saying that, that relative, he told everybody.' 'Llegó diciendo un familiar, les dijo a todos.' {LEL tx5:4:07.0}

### **A.4.10 Past egophoric** *-ki*

Verbs marked with the past egophoric *-ki* suffix encode an event carried out in the past by a first-person subject in statements and by a second-person subjects in questions, have optional nominative-marked pronominal marking. This stressneutral suffix is exemplified in (35).

```
(35) a. pen soˈmáki bi
```
*pe=ni soˈmá-ki=bi*

just=1sg.nom wash.head-pst.ego=just

'I just washed my head.'

'Nomás me lavé la cabeza.' {JLG co1235:6:45.8}

b. *ˈmí biˈtʃ͡éni kaˈlí piˈt͡ʃínula nuˈlèki ˈró ˈmí* 2sg.acc *biˈtʃ͡é=ni* turn=1sg.nom *kaˈlí* house *piˈt͡ʃí-nula* sweep-order *nuˈl-è-ki* order-appl-pst.ego *ˈrú* say.prs 'I told you to sweep the house!' '¡Te dije que barrieras la casa!' < BFL 06 4:145/el >

For some speakers, this suffix is used in the past tense when the object argument is first person. This is shown in (36).

(36) a. *ˈétʃ͡i taˈmí ˈúrki riʔˈréti ˈétʃ͡i* dem *taˈmí* 1sg.acc *ˈúri-ki* take-pst.ego *reʔˈré-ti* down-all 'He took me down (the river).' 'Él me llevó para abajo.' < FLP 06 in61/in > A.4 The Finite Verb: voice, tense, aspect and mood markers

b. *"ˈpé ke taˈmí ˈàki" ˈhê aniˈká ˈpé* just *ke* neg *taˈmí* 1sg.acc *ˈà-ki* give-pst.ego *ˈhê* like.that *ani-ˈká* say-ger "'They didn't give me" that's how you tell them.' "'A mi no me dieron" así se dice.' {JLG co1237:5:45.5}

As shown in (37a–b), the past egophoric suffix is also used with verbs encoding past tense and a second person subject in interrogative constructions. In contrast, interrogative clauses in the past tense with third person argument subjects are marked with the past tense *-li* suffix, as in (37c).


### **A.4.11 Imperfective** *-e*

In contrast to the past tense suffix, the imperfective emphasizes the internal duration of the event depicted by the predicate. The Choguita Rarámuri imperfective encodes an incomplete or habitual event that takes place over a period of time. The imperfective is marked with the stress-neutral suffix *-e*, a marker which does not display any allomorphy or occurence restrictions. Due to the general process of post-tonic vowel reduction that operates in the language (see §5.2), this suffix is realized with the high, front vowel *-i* for most speakers. Examples are provided in (38).

### A Verbal suffixes

	- b. *aˈwísinili*

*aˈwí-si-nale-i* dance-mot-desid-impf 'She wanted to go along dancing' 'Quería irse bailando' < SFH 07 2:72-73/el >

c. *ˈhúmtʃ͡ani ˈhúmi-tʃ͡ane-i* take.off.pl-ev-impf 'It sounded like they were taking off' 'Se oía como que se arrancaban' < SFH 07 1:7/el >

### **A.4.12 Progressive** *-a*

The progressive is encoded by the stress-neutral suffix *-a*, and it indicates that the event described by the predicate is an ongoing process which is independent of time reference. Uses of this marker are exemplified in (39).


A.5 The Subordinate Verb: deverbal morphology

'We like it a lot indeed, to make women, girls and also young children run a race.'

'Nos gusta mucho hacer correr a las mujeres, a las niñas y a los niños chiquitos.' {LEL tx19:00:23.6}, {LEL tx19:0:26.5}

### **A.4.13 Indirect causative** *nula*

In indirect causative constructions, a causer exerts indirect manipulation on a causee which retains certain degree of autonomy. Indirect causative constructions in Choguita Rarámuri involve a periphrastic construction in which a main jussive predicate takes the caused event as a complement. The lower predicate is marked with the suffix *-nula*. This stress-neutral suffix is derived from the independent verb *nula* 'to order, to command'. The main predicate can be inflected with any tense or aspect, but the lower predicate marked with *-nula* is closed to further suffixation. For more etails about the indirect causative construction, see Appendix A.4.13. Examples of the indirect causative are provided in (40).

	- already=1sg.nom send-pst.ego buy-order store 'I already sent him to the store to buy' 'Ya lo mandé comprar a la tienda.' < BFL 06 2:48/el >

### **A.5 The Subordinate Verb: deverbal morphology**

### **A.5.1 Temporal** *-tʃi*

The temporal *-tʃ͡i* suffix is a stress-neutral morpheme added to predicates of adverbial clauses which encode a temporal relation between two events (translated

### A Verbal suffixes

into English as 'when' clauses). The base for affixation of this suffix is a verb inflected for progressive aspect. The following examples (in (41)) illustrate the use of this suffix.

	- *ˈnè* 1sg.nom *ˈkrîlitʃ͡i* Creel *si-ˈmêa* go.sg-fut.sg *ˈmá* already *suˈwíb-a-tʃ͡i* finish-prog-temp *ˈhê* it *ná* this *taˈrári* week 'I'm going to Creel when this week is finished.' 'Voy a ir a Creel cuando acabe esta semanana.'

### **A.5.2 Epistemic** *-o*

The epistemic modality suffix marks lower predicates of complement clauses of main predicates that express a psychological or mental state, like 'think', 'dream', 'sing' or 'say'. The use of this suffix is exemplified in the examples in (42).

(42) a. *riˈmùini ˈnáptim noˈkáo riˈmù-i=ni* dream-impf=1sg.nom *ˈnápi* sub=2pl.acc *=timi* move-ep *noˈká-o* 'I used to dream that you all were moving.' 'Yo soñaba que ustedes se movían.' b. *aʔˈlì ˈnà kotʃ͡iˈká buˈʔílo maˈjêli*

*aʔˈlì* and *ˈnà* then *kotʃ͡i-ˈká* sleep-ger *buˈʔí-li-o* lay.down.sg-pst-ep *maˈjê-li* think-pst 'And then he thought he was asleep (laid down sleeping).' 'Nomás que pensó que estaba dormido' {LEL tx5:0:35.0}

For some speakers, the epistemic suffix *-o* is also attested in V-V incorporation constructions, as shown in (43).

A.5 The Subordinate Verb: deverbal morphology


### **A.5.3 Gerund** *-ká*

The gerund suffix *-ká* occurs in subordinate clauses and marks a non-finite verbal construction denoting an ongoing event which occurs simultaneous to another event. This stress- neutral suffix is exemplified in (44):

(44) a. *púlako biˈlá niwaˈrîa ˈnàri biʔˈrìnka baˈtʃ͡á ba biˈlé ˈtá kuˈʃìti ba ˈpúla=ko* belt=emph *biˈlá* really *niwa-ˈrîwa* make-mpass *ˈnàri* then *biʔˈrì-na-ka* roll.up-tr-ger *baˈtʃ͡á* first *ba* cl *biˈlé* one *ˈtá* det *kuˈsì-ti* stick-instr *ba* cl 'The belt is made by rolling it up first with a stick.' 'La faja se hace enrollándolo primero con un palo.' {BFL tx1:0:24.9} b. *aʔˈlì ˈnà kotʃ͡iˈká buˈʔílo maˈjêli aʔˈlì* and *ˈnà* then *kotʃ͡i-ˈká* sleep-ger *buˈʔí-l-o* lay.down.sg-pst-ep *maˈjê-li* think-pst 'And then he thought he was asleep (laid down sleeping).' 'Nomás que pensó que estaba dormido.' {LEL tx5:0:35.0}

As discussed in §15.2.1, verbal predicates in apodosis clauses of conditional constructions are often marked with the gerundive *-ká* suffix. Very frequently, verbs marked with the *-ká* suffix appear in verb chaining structures conveying a temporal relation of chronological overlap or chronological sequence (with some extended semantic meanings in some cases). For more details about this, see §15.5.

### A Verbal suffixes

### **A.5.4 Purposive** *-ra*

The purposive suffix *-ra* is a stress-neutral suffix which derives a noun from a finite verb inflected for progressive aspect. The purposive indicates that the derived noun is an instrument or means involved in carrying out the event described by the predicate. This suffix is not limited to a few lexical items, and may be productively added to any finite verb inflected for progressive aspect. The forms in (45) exemplify this nominalization process.

(45) a. *ˈpòara ˈpò-a-ra* cover-prs-purp 'lid' (lit. 'for covering') 'tapadera' (lit. 'para tapar') < SFH 07 in242/in > b. *oˈsìara oˈsì-a-ra* write-prog-purp 'pen' (lit. 'for writing') 'pluma' (lit. 'para escribir')

### **A.5.5 Participial** *-ame*

One final layer of morphologically complex verbs involves one optional layer of morphology (in suffix position S12), where a finite verb may attach morphemes that encode subordination. As discussed in §8.5.1, affixation of the participial suffix *-ame* to transitive bases or intransitive bases with an unergative argument, derive nominalizations with an agentive meaning ('the one who performs V') (46a). Patientive nominalizations, on the other hand, are formed through attachment of the participial suffix *-ame* to passivized verbal bases (46b) or to intransitive verbs with a theme as subject argument (46c). Finally, theme nominalizations are derived though attachment of the participial *-ame* suffix to a medio-passive base (46d).

(46) a. *waʔˈlûm ˈwe ˈnà raˈhâame buˈsêame tʃuˈkú naˈʔî waʔˈlû=mi* big=dem *ˈwe* int *ˈnà* then *raˈhâ-ame* lit.up-ptcp *buˈs-ê-ame* eye--have-ptcp *tʃuˈkú* sit.sg.prs *naˈʔî* here 'A big one with lighten up eyes is (lit. sits) here.' 'Muy grande, con unos ojotes está (se sienta) aquí.' {LEL tx5:0:59.2}

### A.5 The Subordinate Verb: deverbal morphology

b. *ˈétʃi ˈmi ˈnà ˈtòruame tʃiˈhônsa ko ba ˈnà ˈmá wiʰˈkâ riˈhòrame ba ˈétʃi* dem *ˈmi* dist *ˈnà* dem *ˈtò-ru-ame* take-pst.pass-ptcp *tʃiˈhônsa=ko* then=emph *ba* cl *ˈnà* dem *ˈmá* already *wiʰˈkâ* many *riˈhò-r-ame* live.people-r-ptcp *ba* cl 'It was then taken there because there were a lot of people already.' 'Para allá fue llevada entonces porque ya había mucha gente.' {SFH tx12:0:48.2} c. *ˈwé aʔˈláa ˈnísa ka baʰˈtâri ba ˈwé aʔˈlá waˈsiame ˈnísa ka ba ˈwé* int *aʔˈlá* well *ˈní-sa* cop-cond *ka* irr *baʰˈtâri* corn.beer *ba* cl *ˈwé* int *aʔˈlá* well *waˈsiame* cook-ptcp *ˈnísa* cop-cond *ka* irr *ba* cl 'It would be good the corn beer if it is well cooked.' 'Sí será bueno el tesgüino si está bien cosido.' {LEL tx68:2:30.0} d. *ke biˈlé tʃaˈpí ˈétʃi liˈmôsna aniˈrîami ko, muˈkuî ko ke* neg *biˈlé* uno *tʃaˈpí* grab *ˈétʃi* dem *liˈmôsna* charity *ani-ˈrî-ame=ko* say-mpass-ptcp=emph *muˈkuî=ko* women=emph

'She doesn't take what it is called charity, the women. '

'No agarra lo que se dice 'limosna', las señoras.' {LEL tx19:5:20.7}








guage and Linguistic Theory 82), 267–289. Dordrecht: Springer. DOI: 10.1007/ 978-94-007-0137-3.






Aguilar, Andrés, 5, 170, 186, 188, 429, 434, 498 Aikhenvald, Alexandra Y., 596 Alderete, John D., 3, 150 Alegre, Francisco Javier, 15, 207 Alexander, Jennifer Alexandra, 171 Álvarez González, Albert, 563 Ameka, Felix K., 21, 480, 492, 493 Anderson, Gregory D. S., 592 Anderson, Stephen R., 337 Andrews, Avery D., 563 Anttila, Arto, 22 Arellanes Arellanes, Francisco, 66 Aronoff, Mark, 272 Ayoker, Otto Gwado, 23 Babel, Molly, 71 Baker, Mark C., 163, 301, 583 Banerji, Robindra, 418 Barnes, Jonathan, 138, 334 Barreras, Isabel, 235, 242 Bascom, Burt, 169 Batista, Dolores, 13 Beckman, Mary E., 22, 134, 177, 405 Bennett, Ryan T., 408 Bennett, Wendell Clark, 12 Berez-Kroeker, Andrea L., 23 Bickel, Balthasar, 2, 263, 273, 348, 350, 358 Bickmore, Lee, 403 Blevins, James P., 272 Blevins, Juliette, 406

Booij, Geert, 160, 161 Bowern, Claire, 584 Brambila, David, 12, 65, 68, 161, 162, 165, 233, 235, 273, 292, 342, 354, 364, 423, 459, 462 Brugman, Claudia, 584 Bugaeva, Anna, 464 Burgess, Don H., 1, 6, 8, 9, 12, 15, 16, 89, 178, 207, 349, 395, 552 Butt, Miriam, 583, 584, 592 Bybee, Joan L., 218, 596 Bye, Robert Arthur, 12 Caballero, Gabriela, 3–5, 20, 24, 32, 64, 66, 133, 136, 137, 150, 151, 167–175, 177, 180, 181, 183– 189, 203, 207, 265, 273, 292, 293, 322, 326, 328, 404, 405, 408, 410, 411, 413, 415, 418, 423, 427–429, 434–437, 439, 445, 498 Campbell, Lyle, 9, 11, 150 Caponigro, Ivano, 563 Carrillo Carrillo, Araceli, 170 Carroll, Lucien, 4, 5, 64, 66, 133, 136, 137, 167, 170, 171, 173, 175, 181, 207, 292, 405, 408, 410, 411, 415, 423, 435 Casad, Eugene H., 222, 578, 583 Casaus, Michael, 14, 17, 207 Chaparro Gardea, Rosa Isela, 24 Charney, Jean Ormsbee, 550

Chávez-Peón, Mario E., 208, 213, 214 Clayton, Ian D., 70 Cohen, Antonie, 179 Comrie, Bernard, 32, 473 Crapo, Richley H., 317, 599 Cristofaro, Sonia, 21, 471, 553 Dahl, Östen, 596 Dakin, Karen, 231, 235, 242 Davidson, Lisa, 127 Dayley, Jon Philip, 150, 225, 300, 535 De Chene, Brent, 337 Dedrick, John M., 222, 578, 583 Demers, Richard, 4, 64, 161, 169 Dench, Alan, 22 Di Napoli, Jessica, 188 DiCanio, Christian T., 66 Dixon, RMW, 161 Downing, Laura J., 188, 403, 434 Dresher, B. Elan, 406 Dryer, Matthew S., 451, 472, 473, 475, 518 Elenbaas, Nine, 3 Embriz Osorio, Arnulfo, 1 Enfield, N. J., 362, 366 Erker, Daniel, 127 Escalante, Fernando, 222, 309 Estrada-Fernández, Zarina, 178 Evans, Nicholas, 22 Everdell, Michael, 367 Everett, Daniel L., 161 Felix Armendáriz, Rolando G., 473, 554, 578 Fitzgerald, Colleen M., 404 Foley, William A., 596 Fowler, Catherine S., 9 Gaby, Alice Rose, 493

García Salido, Gabriela, 2, 473, 563, 578, 583, 599 Garellek, Marc, 5, 170, 172, 177, 179– 183, 408, 429–431, 434, 498, 499 German, Austin, 5, 174, 175, 273, 293, 413, 415, 418, 427, 428 Geuder, Wilhelm, 583, 584, 592 Gil Burgoin, Carlos Ivanhoe, 170 Godjevac, Svetlana, 434 González Rodríguez, Luis, 15 Gordon, Matthew K., 4, 133, 136, 169, 188, 405 Gouskova, Maria, 447 Graf, Dafna, 404 Gralow, Frances L., 434 Green, Thomas, 161 Grimes, Joseph E., 4, 64, 169 Guerrero Valenzuela, Lilián, 2, 472, 473, 475, 487, 563, 578, 583 Guillaume, Antoine, 531 Guion, Susan G., 4, 5, 64, 136, 169– 171, 435 Gurevich, Naomi, 73 Hagberg, Larry, 4, 64, 161, 169 Hagberg, Lawrence Raymond, 169 Hale, Ken, 550 Hanks, William F., 366 Harrison, Sheldon P., 406 Haspelmath, Martin, 531, 537 Haugen, Jason D., 2, 65, 203, 237, 238, 306, 309, 367 Hayes, Bruce, 22, 134, 178, 401, 404, 405 Heath, Jeffrey, 278, 282, 411, 480, 605 Hill, Jane H., 3, 9,11,150, 353, 367, 550 Hill, Kenneth C., 3, 150, 367 Hilton, K Simón, 12

Himmelmann, Nikolaus P., 362 Hualde, José Ignacio, 168, 204, 208 Hyman, Larry M., 22, 133, 134, 169, 188, 189, 334, 403, 405, 418, 419, 434 Inkelas, Sharon, 22, 134, 263, 270, 333, 338, 403, 419, 434 Islas Flores, Bianca Paola, 12, 366, 457 Itô, Junko, 406 Jacobsen Jr, William H., 550 Jacobsen, Steven E., 550 Jara, Yolanda Valdez, 12 Jeanne, Laverne Masayesva, 550 Jelinek, Eloise, 309 Jun, Sun-Ah, 5 Kager, René, 3, 64, 161, 168, 408, 447 Kapatsinski, Vsevolod, 3, 5 Kataoka, Reiko, 71 Kavitskaya, Darya, 334 Kawahara, Shigeto, 113 Kenstowicz, Michael, 161 Kingston, John, 188 Kiparsky, Paul, 338, 445 König, Ekkehard, 523 Kuang, Jianjing, 188, 208 Kubozono, Haruo, 208 Ladd, D Robert, 179 Ladd, D. Robert, 22, 177 Ladefoged, Peter, 66, 90 Lahiri, Aditi, 406 Langacker, Ronald W., 2, 9,11, 65,150, 237, 272, 301, 354, 425, 497 Lastra, Yolanda, 18 Lehmann, Christian, 565 Levi, Jerome M., 15, 16

Levinson, Stephen C., 379, 480, 485, 492, 493 Lindenfeld, Jacqueline, 222 Lionnet, Andrés, 12, 68, 90, 220, 235, 273, 295, 296, 301 Lipski, John M., 208 Longacre, Robert E., 578, 579, 582 MacWhinney, Brian, 333 Maddieson, Ian, 63, 66, 90, 113, 171 Makáwi, Martín, 13 Manaster-Ramer, Alexis, 4, 64, 169 Martínez Fabián, Constantino, 583 Martínez-Paricio, Violeta, 168, 408 Matthews, Peter H., 5 McCarthy, John J., 334 McDonough, Joyce, 198, 403, 434, 500 McKenzie, Andrew, 550 McMahon, Ambrose, 64, 169 Menn, Lise, 333 Merino Mares, Reynaldo, 12 Merrill, William L., 1, 6–9, 11, 12, 15– 17, 89, 207 Mester, Armin, 406 Miller, Wick R., 6, 9, 11, 65, 109, 152, 154, 161, 162, 165, 198, 220, 230, 235, 238, 242, 246, 282, 292, 295, 296, 309, 320, 334, 342, 353, 354, 356, 362, 367, 369, 380, 395, 396, 399, 409, 423, 450–452, 455, 458, 459, 488, 501, 518, 525, 527, 549, 550, 554, 563, 565, 578, 583, 598, 610, 615 Mithun, Marianne, 9, 11 Monaka, Kemmonye C., 189, 434 Morales Moreno, José Isidro, 2, 12, 64, 65, 68, 178, 354, 358,

395–397, 399, 451, 461, 545, 568, 578 Morales Muñoz, Marco Vinicio, 17 Mosel, Ulrike, 34 Munro, Pamela, 3, 64, 136, 150, 169, 404 Myers, Scott, 403 Neely, Kelsey C., 375 Nespor, Marina, 22, 177, 178, 401 Neumann, Joseph, 15 Newman, Paul, 220, 295 Nichols, Johanna, 2, 263, 273, 348, 350, 358 Noonan, Michael, 471, 538, 553 Nordell, Norman, 208 O'Meara, Carolyn, 379, 487 Ochs, Elinor, 375 Odden, David Arnold, 403 Olson, Mike, 596 Orgun, Cemil Orhan, 22 Paciotto, Carla, 17, 18 Paster, Mary, 445 Patiño Velázquez, EP, 5, 170 Pennington, Campbell W., 15 Pérez Báez, Gabriela, 379 Pierrehumbert, Janet B., 22, 177 Pintado Cortina, Ana Paula, 12, 13, 15–18, 207 Poser, William John, 134 Prieto, Pilar, 176, 416 Prince, Alan, 334 Pullum, Geoffrey K., 349 Remijsen, Bert, 23, 133, 136, 434 Revithiadou, Anthi, 348 Reyes Taboada, Verónica, 2, 169 Riad, Tomas, 434

Rice, Curtis Calvin, 406 Rice, Keren, 23, 136, 263 Roettger, Timo, 136 Rolle, Nicholas R., 287, 418, 427 Sapir, Edward, 2, 150, 166, 550 Schieffelin, Bambi, 375 Sebba, Mark, 596 Selkirk, Elisabeth, 22, 177, 178, 401, 406 Servín, Enrique, 12, 13, 68 Shaul, David Leedom, 169 Shaw, Jason A., 113 Shue, Yen-Liang, 186 Sicoli, Mark, 208 Siemund, Peter, 523 Silva, David J., 171 Silver, Shirley, 9 Simpson, Jane, 263 Smalley, William A., 161, 535 Smith, Jennifer L., 261 Stassen, Leon, 225, 531, 535 Steele, Susan M., 358, 618 Stolz, Thomas, 225 Stump, Gregory T., 5, 350 Svenonius, Peter, 583, 584, 592 Thompson, Sandra A., 553, 558, 562 Thornes, Tim, 317, 599, 600 Timberlake, Alan, 334 Tona, Ana E., 178 Toosarvandani, Maziar, 579, 582 Tubino Blanco, Mercedes, 317, 584, 592, 600 Ussishkin, Adam, 404 Valdez-Jara, Yolanda, 178, 395, 396 Valiñas, Leopoldo, 6

Van der Hulst, Harry, 133, 136, 169, 405 Van Heuven, Vincent J., 133, 136, 434 Van Valin Jr, Robert D., 472, 473, 475, 578, 583 Vázquez Soto, Verónica, 473 Villalpando-Quiñonez, Jesús, 2, 12, 354, 364, 546, 578 Voegelin, Charles F., 65 Vogel, Irene, 22, 177, 178, 401 Welmers, Wm E., 418 Whorf, Benjamin L., 2, 418 Wilkins, David P., 485 Willard, E. Rainbow, 338 Wilson, Stephen A., 134 Withgott, Meg, 263 Woo, Nancy, 4, 64, 169 Wood, Esther, 220, 295 Yip, Moira, 169 Zamora Alarcón, Oscar, 1 Zec, Draga, 134 Zepeda, Ofelia, 535 Zigmond, Maurice L., 166 Zingg, Robert Mowry, 12 Zoll, Cheryl, 22, 270, 406, 407

Zwicky, Arnold M., 349

# **Language index**

Ainu, 464 Arawakan, 168 Athabaskan, 198, 434, 500 Aztecan, 4, 11, 64, 170, 207, 238, 282, 480 Bantu, 188, 434 Barasana, 90 Basque, 204 Azkoitia Basque, 168 Cahitan, 4, 5, 11, 64, 161, 169, 480 Cahuilla, 282 Chadic, 171, 434 Chickasaw, 188 Chinook Wishram Chinook, 168 Chinookan, 168 Chitumbuka, 434 Comanche, 161, 535, 550 Cora, 64, 169, 231, 235, 473 Corachol, 4, 11, 64, 169, 231, 418, 473 Coreguaje, 434 Cupan, 150, 480 Cupeño, 3, 150, 282, 353, 550 Dagbani, 188 Edo, 171 Eudeve, 11

French, 243, 514

German, 514 Greek, 168 Guarijío, 6, 7, 9, 11, 13, 65, 109, 152, 154, 161, 163, 220, 282, 295, 307, 395, 396, 399, 409, 487 Mountain Guarijío, 65, 161, 162, 198, 199, 203, 220, 230, 235, 238, 242, 246, 292, 295, 309, 320, 334, 342, 353, 354, 356, 362, 367, 369, 380, 395, 395, 399, 423, 423, 450, 451, 452, 455, 458, 459, 488, 489, 501, 518, 525, 527, 549, 550, 554, 563, 565, 578, 583, 598, 610 River Guarijío, 235, 242, 399, 554, 578 Gur, 188 Hausa, 171, 434 Hebrew, 168 Hiaki, 13, 317, 583, 584, 600 Hopi, 4, 11, 64, 169, 169, 282, 447, 550 Huichol, 4, 64, 169, 418, 473 Italian, 168, 188 Japanese Tokyo Japanese, 113 Kawaiisu, 166 Luiseño, 282

### Language index

Macedonian, 168 Mam Maya, 338 Mayo, 4, 5, 11, 13, 64, 161, 169, 487 Mono, 282 Muskogean, 188 Naasioi, 90 Nahuatl, 5, 136, 136, 170, 207, 208 Balsas Nahuatl, 4, 64, 169, 169, 171, 435 Milpa Alta Nahuatl, 418 Navajo, 198, 434, 500 Northern Paiute, 317, 579, 599, 600 Mono Lake Northern Paiute, 71 Southern Nevada Northern Paiute , 71, 71 Numic, 9, 11, 71, 71, 150, 161, 225, 238, 301, 317, 535, 550, 579, 599, 600 Opata, 11 Opatan, 11 Papiamentu, 434 Papuan, 90 Pima, 12 Lower Pima, 487 O'ob Nook Pima, 13 Pirahã, 161 Polish, 168 Quechua Imbabura Quechua, 168 Rarámuri Norogachi Rarámuri, 12, 29, 65, 90, 161–163, 220, 235, 292, 342, 354, 354, 364, 423, 459, 462, 546, 578 Ojachichi Rarámuri, 12, 68

Pahuírachic Rarámuri, 178 Rochéachi Rarámuri, 12, 64, 65, 178, 354, 395, 396, 397, 399, 451, 461, 545, 568, 578 Samachique Rarámuri, 12 Urique Tarahumara (Rarómari), 178, 395 Western Rarámuri (Tarahumara), 12, 178, 395, 552 Serbo-Croatian, 434 Serrano, 282 Shekgalagari, 434 Shoshone Tümpisa Shoshone, 150, 535 Sonoran, 11, 409 Proto-Sonoran, 235, 409, 615 Southern Paiute, 150, 166, 282, 480, 550 Spanish, 9, 18, 25, 30, 89, 90, 91, 134, 167, 168, 168, 175, 176, 176, 176, 178, 207, 207, 208, 209, 209, 210–215, 217, 221, 224, 230, 242, 256, 257, 287, 332, 334, 341, 370, 387, 415, 416, 436, 477, 483, 520, 571, 572 Swedish Stockholm Swedish, 434 Takic, 3, 11, 238, 353, 550 Tara-Guarijío, 65, 154, 550 Taracahitan, 11, 152, 161, 165, 198, 222, 238, 309, 317, 353, 472, 473, 487, 501, 583, 584, 600 Tarahumara, *see* Rarámuri Tarahumara-Guarijío, 9 Tarahumaran, 9, 11 Tepehuan Northern Tepehuan, 4, 12, 13, 64, 169, 169, 170

Language index

Santa María de Ocotán Tepehuan, 169 Southeastern Tepehuan, 447, 473, 599 Southern Tepehuan, 169 Tepiman, 4, 11, 64, 169, 169, 170, 225, 238, 282, 301, 404, 409, 447, 473, 487, 535, 550, 599 Terena, 168 Tohono O'odham, 404, 535, 550 Trique Ituñoso Trique, 66 Tubar, 11, 235 Tucano, 90 Tucanoan, 90, 434 Tupi, 168 Icua Tupi, 168 Tübatulabal, 11 Uto-Aztecan, 2, 3, 9, 11, 13, 150, 150, 161, 169, 170, 198, 204, 207, 225, 231, 237, 238, 272, 282, 300, 301, 305, 306, 315, 353, 358, 367, 401, 404, 409, 418, 435, 473, 480, 487, 497, 501, 535, 550, 563, 582, 584, 599, 600 Northern Uto-Aztecan, 11 Proto Northern Uto-Aztecan, 282 Proto Southern Uto-Aztecan, 282 Proto-Uto-Aztecan, 64, 65, 136, 150, 169, 235, 237, 278, 282, 300, 354, 404, 605, 618 Southern Uto-Aztecan, 11, 133 Volta-Niger, 171

Warihio, *see* Guarijío

Wixárika, 418 Yaqui, 4, 5, 11, 13, 64, 161, 169, 222, 309, 472, 473, 475, 475, 487, 563, 578, 583 Zapotec San Pablo Güilá Zapotec, 66 Zoque, 168

# A grammar of Choguita Rarámuri

This book provides the first comprehensive grammatical description of Choguita Rarámuri, a Uto-Aztecan language spoken in the Sierra Tarahumara, a mountainous range in the northern Mexican state of Chihuahua belonging to the Sierra Madre Occidental. A documentary corpus developed between 2003 and 2018 with Choguita Rarámuri language experts informs the analysis and is the source of the examples presented in this grammar. The documentary corpus, which consists of over 200 hours of recordings of elicited data, narratives, conversations, interviews, and other speech genres, is available in two archival collections housed at the Endangered Languages Archive and at UC Berkeley's Survey of California and Other Indian Languages.

Choguita Rarámuri is a highly synthetic, agglutinating language with a complex morphological system. It displays many of the recurrent structural features documented across Uto-Aztecan, including a predominance of suffixation, head-marking, and patterns of noun-incorporation and compounding (Sapir 1921; Whorf 1935; Haugen 2008b). Other features of typological and theoretical interest include a complex word prosodic system, a wide range of morphologically conditioned phonological processes, and patterns of variable affix order and multiple exponence. Choguita Rarámuri is also of great comparative/historical importance: while several analytical works of Uto-Aztecan languages of Northern Mexico have been produced in the last years (Guerrero Valenzuela 2006, García Salido 2014, Reyes Taboada 2014, Morales Moreno 2016, Villalpando Quiñonez 2019, inter alia), many varieties still lack comprehensive linguistic description and documentation.