# Getting others to do things

A pragmatic typology of recruitments

Edited by Simeon Floyd Giovanni Rossi N. J. Enfield

Studies in Diversity Linguistics 31

### Studies in Diversity Linguistics

### Editor: Martin Haspelmath

In this series:


# Getting others to do things

A pragmatic typology of recruitments

Edited by

Simeon Floyd Giovanni Rossi N. J. Enfield

Floyd, Simeon, Giovanni Rossi & N. J. Enfield (eds.). 2020. *Getting others to do things*: *A pragmatic typology of recruitments* (Studies in Diversity Linguistics 31). Berlin: Language Science Press.

This title can be downloaded at: http://langsci-press.org/catalog/book/263 © 2020, the authors Published under the Creative Commons Attribution 4.0 Licence (CC BY 4.0): http://creativecommons.org/licenses/by/4.0/ Indexed in EBSCO ISBN: 978-3-96110-278-5 (Digital) 978-3-96110-279-2 (Hardcover)

ISSN: 2363-5568 DOI:10.5281/zenodo.4017493 Source code available from www.github.com/langsci/263 Collaborative reading: paperhive.org/documents/remote?type=langsci&id=263

Cover and concept of design: Ulrike Harbort Typesetting: Giovanni Rossi, Sebastian Nordhoff, Naomie Nguyen, Elena Sheard Proofreading: Alexander Rice, Andreas Hölzl, Amir Ghorbanpour, Brett Reynolds, Eran Asoulin, Felix Kopecky, Giovanni Rossi, Jeroen van de Weijer, Jean Nitzke, Sean Stalley, Sebastian Nordhoff, Stefan Hartmann, Tom Bossuyt, Vadim Kimmelman, Viola Wiegand, Yvonne Treis Fonts: Linux Libertine, Libertinus Math, 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

# **Contents**


# **Chapter 1**

# **Recruitments and pragmatic typology**

# Giovanni Rossi

Department of Sociology, University of California, Los Angeles

# Simeon Floyd Department of Anthropology, Universidad San Francisco de Quito

# N. J. Enfield

Department of Linguistics, The University of Sydney

# **1 Introduction**

Getting others to do things is a central part of social interaction in any human society. Language is our main tool for this purpose. In this book, we show that sequences of interaction in which one person's behavior solicits or occasions another's assistance or collaboration share common structural properties that provide a basis for the systematic comparison of this domain across languages. The goal of this comparison is to uncover similarities and differences in how language and other conduct are used in carrying out social action around the world, including different kinds of requests, orders, suggestions, and other actions brought together under the rubric of *recruitment* (see §4 below). The project constitutes an exercise in pragmatic typology. We map out a possibility space for linguisticallymediated social actions and we use that possibility space as a grid for comparison between languages. This allows us to look for universals and cross-linguistic variation in this pragmatic domain. While other multi-authored publications present comparative findings from this project (e.g. Floyd et al. 2018), this book lays out the conceptual and methodological background for the project (Chapters 1–2)

Giovanni Rossi, Simeon Floyd & N. J. Enfield. Recruitments and pragmatic typology. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 1–23. Berlin: Language Science Press. DOI:10.5281/zenodo.4018370

Giovanni Rossi, Simeon Floyd & N. J. Enfield

and presents the findings language-by-language (Chapters 3–10). The book is intended to serve as a reference source for those interested in primary data on the phenomenon of recruitments in a diverse set of the world's languages.

# **2 Background on research on getting others to do things**

A landmark in research on requests and similar speech acts is Searle (1969; 1975), who built on Austin (1962). For Searle, speech acts have felicity conditions, which need to be met if the act is to succeed. For a request, the utterance should refer to a future act of the recipient and the speaker should believe that the recipient can do the requested act, among other conditions (Searle 1969: 66). Many of Searle's felicity conditions refer to mental states of participants. Building on this and the cognitively-grounded theory of Grice (1975), Brown & Levinson (1978; 1987) developed a theory of politeness in which requests featured prominently. The theory began with observations of similarities in pragmatic strategies in three unrelated languages and cultures (Tamil, Tzeltal and English). A theory of face – people's public self-image (Goffman 1967) – suggested universal pressures affecting social behavior, particularly in "face-threatening acts" such as requests.

Researchers in psychology engaged with the ideas of Searle, Grice, Brown & Levinson, seeking to test them with experimental methods. One puzzle concerned the literal meaning of an utterance (e.g. *Can you pass the salt?*) in the comprehension of the intended request, that is, whether or not the literal meaning must be computed first before inferring that a request is being made (Clark & Lucy 1975; Clark 1979; Gibbs 1979; Clark & Schunk 1980; Gibbs 1983; 1986a). Another puzzle concerned variation of request forms in terms of a single general principle: when making a request, a speaker first assesses what reason there might be for the recipient not complying, and then formulates an utterance to deal with the "greatest potential obstacle" they can anticipate (Gibbs 1985; Francik & Clark 1985;Gibbs 1986b;Gibbs & Müller 1988; Clark 1996).The obstacle mentioned may be generic, such as the recipient's inability to do what is requested (e.g. *Can you tell me what time it is?*), or more specific, such as the availability of a relevant object (e.g. *Do you happen to have a watch?*). This is closely related to the ideas of preconditions discussed by Searle (1969) and by Gordon & Lakoff (1971).

Linguists have studied the grammatical structures and pragmatic properties of the basic sentence types, all of which are used in requesting: imperatives, interrogatives, declaratives (Gordon & Lakoff 1971; Sadock & Zwicky 1985;Aikhenvald & Dixon 2017). They have also studied the connections between alternative lin-

### 1 Recruitments and pragmatic typology

guistic forms and social variables in events of requesting (Sinclair & Coulthard 1975; Ervin-Tripp 1976; 1981; Gordon & Ervin-Tripp 1984; Ervin-Tripp et al. 1990), including how these variables may affect the comprehension of the request.

A large body of research in the subfield of "cross-cultural pragmatics" (e.g. Blum-Kulka et al. 1989) has been devoted to the comparative study of linguistic patterns associated with requests across many languages. A unifying element of this tradition of research is a standardized methodology based on "discourse completion tasks" (Blum-Kulka 1982). In §5 below, we further discuss this tradition of research in relation to our pragmatic typological approach.

Research in the fields of conversation analysis, ethnomethodology, and interactional linguistics is perhaps closest to the approach taken here, for a few reasons. One is that the empirical source of data is recordings of informal interaction. Another is that the units of analysis are not clauses or sentences but moves in conversational sequences (Wootton 1981; 1997; Lindström 2005; Vinkhuyzen & Szymanski 2005; Heinemann 2006; Curl & Drew 2008; Craven & Potter 2010; Zinken & Ogiermann 2013; Drew & Couper-Kuhlen 2014a, among many others). These inductive approaches are grounded in the sequential development of interaction. Most conversation-analytic and interactional-linguistic studies of requesting to date have concentrated on a particular language. Comparison between languages has been only possible by drawing on results of distinct studies, each with their own particular focus and goals. Also, most conversation-analytic research tends not to be transparently quantitative. But structured quantitative analysis built on the back of a qualitative analysis has been shown to greatly enhance the analytic possibilities of comparative conversation analysis (e.g. Fox et al. 2009; Rossano et al. 2009; Stivers et al. 2009; Dingemanse et al. 2015).

The pragmatic domain of getting others to do things has been thought of in different ways. A first distinction is often made between getting someone to carry out a practical action and getting someone to provide information. Some work in the philosophy of language (e.g. Searle 1969) and in psycholinguistics (e.g. Clark 1979; Clark & Schunk 1980) tended to merge the two. But most work has distinguished between soliciting practical action and information, and has studied them as separate phenomena.

Another distinction has to do with the categorization of types or subtypes of social action. Two main approaches can be identified here. The first is to treat the domain of getting others to do things as a family of related but distinct speech acts or actions (e.g. directives, requests, hints) on the basis of distinct semantic/pragmatic features, for example those defining different degrees of forcefulness or coerciveness (e.g. Searle 1976; Wierzbicka 1991; Craven & Potter 2010). Giovanni Rossi, Simeon Floyd & N. J. Enfield

The second approach is to treat the domain as a single, generic type of social action, and to treat variations in the way this is implemented as pertaining to the level of linguistic practice (e.g. Ervin-Tripp et al. 1990; Wootton 1997; Rossi 2012).

Yet another distinction is drawn in terms of the temporality of the practical action being solicited: whether the action is carried out immediately, at the same place and time – such as passing a knife in a kitchen setting – or distally, at a later time and possibly different place – such as picking somebody up from work (see, e.g., Lindström 1999). Although much research encompasses both, studies of telephone calls obviously privilege the latter, whereas studies of face-to-face interaction privilege the former.

# **3 Background to the project and studies presented in this book**

The conceptual, empirical, and analytical work on recruitments presented in this volume was carried out by a team of researchers under the auspices of the *Human Sociality and Systems of Language Use* (HSSLU) project, a European Research Council Starting Grant awarded to Nick Enfield (2010–2014). The Recruitments Subproject was coordinated by Simeon Floyd and Giovanni Rossi, and was convened at the Max Planck Institute for Psycholinguistics in Nijmegen, the Netherlands, in collaboration with other members of the Interactional Foundations of Language project in Stephen Levinson's Language and Cognition Department (also encompassing Levinson's European Research Council Advanced Grant IN-TERACT, 2011–2015).

The authors of the chapters of this book each contributed to the comparative study in a number of ways. All contributed to the conceptual development of the project, including the content of the coding scheme, administering the coding scheme, and analyzing the results. Collection of video corpora used in the study was carried out by Julija Baranova (Russian), Joe Blythe (Murrinhpatha), Mark Dingemanse (Siwu), N. J. Enfield (Lao), Simeon Floyd (Cha'palaa), Giovanni Rossi (Italian, English) and Jörg Zinken (Polish).<sup>1</sup> Steve Levinson provided the context for this project to thrive, and was a key interlocutor at all points throughout the project. As an external collaborator, Paul Drew was present for many of the research meetings, and contributed much to the methodology and conception

<sup>1</sup>Part of the English data came from the Language and Social Interaction Archive created by Leah Wingard, available from San Francisco State University (http://www.sfsu.edu/~lsi/).

### 1 Recruitments and pragmatic typology

of the project. Séan G. Roberts provided crucial advice and assistance in developing the quantitative aspects of the comparative analysis (not reported in this volume). The coding and data analysis workflow built on and extended work in a closely related subproject of HSSLU that developed a pragmatic typology of other-initiated repair, coordinated by Mark Dingemanse and Nick Enfield. We also owe a debt to conceptual collaborators in discussion and data analysis over the life of the project: Lorena Pool Balam, Penelope Brown, Tyko Dirksmeyer, Paul Drew, Rósa S. Gísladóttir, Gertie Hoymann, Stephen C. Levinson, Lilla Magyari, Elizabeth Manrique, Ruth Parry, Séan G. Roberts, Jack Sidnell, Tanya Stivers and Francisco Torreira.

The development of the recruitments concept and the timeline of the project work and findings is as follows.<sup>2</sup>

The HSSLU project, which began in January 2010, featured three subprojects. One of these centered on actions of getting people to do things. On 7–9 October 2010, team members discussed requests and similar kinds of social actions in a UCLA workshop on "Action Ascription in Social Interaction". At a follow-up workshop on the same topic in Nijmegen on 18–19 March 2011, Enfield presented a first working definition of "recruitment" (Enfield 2011a). This was a reference point for a one-week intensive data workshop on recruitments held later that month (March 21–25, 2011), in which team members, together with Paul Drew as an external collaborator, delved into data and initial qualitative analysis of candidate recruitment sequences in the languages represented in this volume.

This collaborative work, along with a subsequent session on recruitments at a HSSLU project retreat on 20 April 2011 (Enfield 2011b), resulted in a first draft of the coding scheme for this volume, authored by Floyd and Rossi, and circulated within the project team on 25 October 2011. The project team met (on 27 March 2012) to discuss the first draft coding scheme. Notes by Floyd and Rossi were then circulated, followed by circulation of an updated coding draft on 16 April 2012.

In October 2012, a subgroup of team members – Enfield, Floyd, Rossi, and Dingemanse – carried out a first pilot study using the coding scheme. On November 2, 2012, Floyd presented the ongoing results of the recruitments project at a retreat at Schloss Ringberg, Germany. Later that month, the group met to discuss and plan a second pilot study (with a new coding scheme draft version 1.3), this time with all team members participating. During the week of November 20–27, 2012, the full team carried out a pilot of the coding scheme on all of the languages included in this volume. In December 2012, team members identified

<sup>2</sup> See http://recruitments.nickenfield.org/timeline/ for PDFs of the documents and memos mentioned in the rest of this section.

Giovanni Rossi, Simeon Floyd & N. J. Enfield

cases in their respective corpora for coding, and on January 20, 2013, the coding scheme was finalized (version 2.2). The full team then held an intensive coding week (January 21–25, 2013), following a shared set of further instructions circulated to the team by Floyd.

After the coding was completed, a coding consistency check was done, followed by a test for coder reliability across the team members. This process was overseen by Floyd and Rossi. Because each team member worked with data in languages that others had no access to, our coder reliability check was carried out using a reference set of English data. The reliability check established that some questions were coded in the same ways across the group, but it also revealed that some questions had not been coded consistently. The final step was for the team to carry out a re-coding of those questions in order to ensure coder reliability. The re-coding took place in December 2013.

Results of the comparative study were publicly presented in February 2014 by Floyd and Rossi at the UCLA workshop "About Face" (Floyd et al. 2014a), and then in June 2014 at the International Conference on Conversation Analysis at UCLA (Floyd et al. 2014b). The June 2014 presentation not only publicized the empirical findings of the comparative project, it also presented the key conceptual elements of our collective development of the concept of recruitment sequences. Other publications in which these ideas and findings have been discussed include Enfield (2014b), Drew & Couper-Kuhlen (2014b), Rossi (2015), Kendrick & Drew (2016), Zinken & Rossi (2016), and Floyd (2017).

While comparative findings from this project are presented in multi-authored publications (Floyd et al. 2014b; 2018), the present collection was convened as an opportunity for the individual researchers to lay out the project findings specific to their language of study.

# **4 Recruitment sequences defined**

Our use of the term *recruitment* reflects a shift from an approach centered around the speech act of requesting to one addressing the interactional process of getting others to act. The two main alternatives to our *recruitments* approach are: i) a definition of the phenomenon based on intentional states, such as someone's desire to have another do something, and ii) a definition of the phenomenon based on linguistic form, for example, focusing on imperatively-formatted utterances. While the former presents problems of evidence, the latter overly limits the scope, as we know that other types of strategies can be used in seeking assistance or collaboration. Our functional approach based on recruitment sequences in recorded

### 1 Recruitments and pragmatic typology

social interaction makes the identification of cases more objective and replicable, the analysis more falsifiable, and comparison across languages easier thanks to the natural control provided by sequential structure (Dingemanse & Floyd 2014), without having to restrict the scope of relevant linguistic patterns beforehand.

As we define it, a recruitment is a basic cooperative phenomenon in interaction consisting of a sequence of two moves with the following characteristics:<sup>3</sup>


Crucial to this phenomenon is the nature of the behavior instigated in Move B: a *practical* action involving physical work, typically the transfer of an object, the performance of a manual task, or the alteration of an ongoing bodily movement.

In this project, we restrict the target phenomenon by focusing on the recruitment of practical actions to be performed here and now. The recruitment of information and of future practical actions are excluded. At the same time, we are inclusive of any communicative behavior that causes someone to do something, independently of its verbal or nonverbal construction, and of whatever the speaker's exact intention may be. The identification of cases does not turn on the form of the instigating behavior but on the nature of the behavior instigated, and on the causal relation between the two.

Because of our focus on here-and-now cooperation, Move A and Move B must be temporally adjacent. This means that B must begin to deal with what A has said or done in the next few moments. In some cases, the provision of assistance or collaboration may be displaced because B initiates repair or defers fulfillment on some grounds (e.g. because they are momentarily busy). What is important is that the first response addresses the relevance of immediate cooperation. This obtains also when B refuses to fulfill the recruitment. Finally, there are cases in which B may ignore Move A and produce no response, or a response that does not address the relevance of their immediate cooperation. In these cases, additional measures are taken to preserve objectivity in the identification and inclusion of cases. When there is no uptake of the recruitment, we only consider cases in which: i) Move A involves an explicit, on-record practice, typically a

<sup>3</sup>By *move* we intend a unit of communicative behavior that may include language and/or other conduct (Enfield 2013: chap. 6; cf. Goffman 1981). This is a related but distinct concept to *turn*, which we understand as a move involving primarily language (see Schegloff 2007: 3–7).

### Giovanni Rossi, Simeon Floyd & N. J. Enfield

linguistic practice, that is known to regularly solicit compliance with recruitment (e.g. imperatives, explicit interrogatives such as *Can you x*, etc.); ii) Move A is repeated, either in the same or in another form, showing pursuit of response. These criteria mean that we exclude cases of implicit, off-record practices that could have potentially led to someone doing something but did not, and were not pursued. Finally, note that in cases in which the recruitment is fulfilled or granted in Move B, the sequence may be minimally expanded with a further move – Move C – registering appreciation or satisfaction, in other words, acknowledgment by A.

Recruitment sequences encompass a broad range of actions and interactional events that result in someone providing assistance or collaboration. Our definition does not restrict the focus to request-related events but includes offer-related events, where assistance is provided in response to someone's trouble (Curl 2006). The coding scheme, presented in Chapter 2, helps to distinguish offer-related events from others, while providing a framework for treating both as part of the broader phenomenon of recruitment.<sup>4</sup> The scheme also suggests criteria for discriminating between offers that are occasioned by another's trouble and offers that are not (p. 28).

Across the volume, we examine different kinds of recruitment events, from ones in which someone is told or asked to do something, to others in which someone responds to another's statement of need (individual or collective), to yet others in which someone responds to a wordless gesture requiring cooperation. Each chapter surveys a range of actions, sequential structures, and socialinteractional relations between recruiter and recruitee. The cooperative events we examine include both ones in which a participant is recruited to assist another in the realization of an individual goal and ones in which a participant is recruited to collaborate in the realization of a joint goal, involving shared commitments and responsibilities (Enfield 2014b; Zinken & Rossi 2016).

The components of a recruitment sequence and the terminology to describe them can be summarized as follows.

A recruitment seqence minimally involves:


<sup>4</sup>The coding scheme and the extensive collaborative work that contributed to its development (see §3 and fn. 2) predate a study by Kendrick & Drew (2016) that built on this collaborative work and used data collected within the project.

1 Recruitments and pragmatic typology

	- **–** fulfillment: a practical action involving physical work performed for or with A;
	- **–** rejection: the conveyance of inability or unwillingness to fulfill the recruitment;
	- **–** deferment of fulfillment;
	- **–** initiation of repair (often leading to fulfillment after repair).

If B ignores Move A, then the sequence must include an explicit, on-record practice of initiating recruitment and/or pursuit of response in order to be included.

The sequence can be expanded by a move c, doing acknowledgment.

The coding scheme (Chapter 2) provides detailed commentary and examples, elaborating on each element of recruitment sequences and on the criteria for their identification presented in summary form here.

# **5 Pragmatic typology**

Pragmatic typology is the comparative study of language use. It brings together conceptual and analytic tools from a range of disciplines including linguistics, conversation analysis, gesture studies, and anthropology. A key innovation of the approach in relation to tools for analysis in linguistics is the reference to features of a sequentially ordered exchange of actions in conversation, including the temporal unfolding of such exchange, and its social and normative context. Further, because of the reliance on video corpora, it incorporates both verbal and nonverbal conduct in the analysis of sequences of action. We are faced with the challenge that faces any comparative linguist, namely the need to distinguish between language-particular descriptive categories and language-independent comparative categories or "comparative concepts" (Haspelmath 2010). We submit that our appeal to features of conversational organization – outside the usual realm of "concepts" in the semantic sense – is an advance in the search for tools for linguistic comparison.

A landmark effort to carry out comparative pragmatics was the subfield of "cross-cultural pragmatics", launched in the 1980s. Building on speech act theory

### Giovanni Rossi, Simeon Floyd & N. J. Enfield

and politeness theory, cross-cultural pragmatics has studied the realization of requests across a large number of languages.<sup>5</sup> This work was motivated, on the one hand, by a search for similarities and differences in the use of language across cultures, and on the other hand, by an interest in the acquisition and development of pragmatic competence (see Woodfield 2008 for a review). Studies in this tradition have provided insights into culture-specific features of politeness and directness, and produced rich inventories of request realization patterns. However, these advances in systematic comparison of speech acts across languages have been limited by their methodology. By using written elicitation, in the form of a "discourse completion task", this work relied on speakers' metalinguistic beliefs about appropriate usage, rather than on direct observation of actual usage in situ. A first problem with this is that we cannot be sure if speakers' intuitions match with what they do in practice. A second problem is that such an approach is relatively low in ecological validity, and does not provide access to the kinds of empirical evidence that direct and repeated observation of behavior in recordings can provide.

Our approach to pragmatic typology has two fundamental elements: (i) the empirical analysis of verbal and nonverbal behavior in video recordings of naturally occurring interaction across languages, and (ii) a coding-based methodology for systematic comparison (see also Dingemanse & Enfield 2015; Rossi 2020). In some previous comparative work in conversation analysis and interactional linguistics, the comparison emerges from the cumulative results of distinct studies, each with their own particular focus and goals; this applies, for instance, to collections of studies of questioning (Steensig & Drew 2008), person reference (Enfield & Stivers 2007) and change-of-state tokens (Heinemann & Koivisto 2016). In other cases, the comparison is designed in advance and carried out jointly by reference to a common focus. Studies of this kind have examined, among other phenomena, the intersection of self-repair and turn-taking (Fox et al. 1996), otherinitiated repair (Egbert 1996; Egbert et al. 2009), epistemically authoritative second assessments (Sidnell & Enfield 2012), and requests (Zinken 2016). Finally, some structured comparisons involve a combination of qualitative and quantitative methods, including a coding scheme with a battery of standardized questions asked of hundreds of cases for each language; these include studies of gaze behavior (Rossano et al. 2009), turn-taking (Stivers et al. 2009), self-repair (Fox et al. 2009; 2010), question-answer sequences (Enfield et al. 2010), and other-initiated

<sup>5</sup> See House & Kasper (1981), Blum-Kulka et al. (1989), Sifianou (1992), Lee-Wong (1994), Le Pair (1996), Márquez-Reiter (2000), Tsuzuki et al. (2005), Rue & Zhang (2008), Félix-Brasdefer (2009), Ogiermann (2009), Peterson (2010), among many others.

### 1 Recruitments and pragmatic typology

repair (Dingemanse et al. 2015). The project reported on in this book falls within the third group. Ours is a mixed methodology in six defined steps, outlined in the next section.

# **6 A six-step method**

We now describe our six-step method for comparative team-based pragmatic typological research. We outline the process in idealized form, in part as a description of what we have done in this project and in part as a recipe for carrying out subsequent pragmatic typological team projects on other aspects of social interaction.


Giovanni Rossi, Simeon Floyd & N. J. Enfield

> as necessary background to the analysis of the corpora, along with the relevant biographical and ethnographic background. This means that the "Transcribe" step will likely require between 9 and 12 months of dedicated fieldwork. This is a valuable investment with broader payoff. An important outcome of this step is that these corpora will then be available for further research in the future.


1 Recruitments and pragmatic typology

researchers in Step 4 was done in the same way (as Dingemanse et al. 2015 do for other-initiated repair).

**Step 6. Model:** This final step involves statistical modeling of the coding results from Step 4. The data resulting from the large-scale coding scheme enables statistical modeling for quantitative assessment of patterns of association and interdependence between the phenomena coded for (as formulated in Step 3 and executed in Step 4). It is necessary to use multivariate statistics to control for interdependence among these variables (see Dingemanse et al. 2015 for an example of this). Steps 5 and 6 do not take a long time, but require special expertise.

This protocol requires a team science approach. Given the demanding combination of fieldwork (Steps 1, 2), expertise in comparative linguistics, interactional linguistics, and conversation analysis (Steps 3, 4), and quantitative approaches (Steps 5, 6), this could never have been done in any way other than by a team. Team science in linguistics is still rare and we were fortunate to have had the opportunity to do this here.

The findings reported in the language-specific chapters in this volume are primarily the product of qualitative analysis but also include quantitative findings particular to each data set.

# **7 Data**

This study is based on the analysis of corpora of audiovisual recordings of informal everyday language usage in social interaction in eight languages from five continents (see Figure 1). In building these corpora, we placed unattended cameras in household and community contexts to record social interactions as they were occurring naturally, using high standards for audio and video quality (see Step 1 discussed in §6 for more details).

The data were transcribed and translated by a language expert (see Table 1) with assistance from native speakers. The corpora range in size from about ten to over ninety hours of footage. In some cases, the corpus represents the largest available database for the language, especially in the case of unwritten minority languages like Cha'palaa (Chapter 3), Murrinhpatha (Chapter 7), and Siwu (Chapter 10). For larger-scale national languages like Italian, Lao, Polish, and Russian, other corpora may be available to some degree, but most of these are limited to written language, due to the intensive demands of transcription of spoken language.

Giovanni Rossi, Simeon Floyd & N. J. Enfield

Figure 1: World map showing locations of data collection for the eight languages involved in the study (Credit: satellite composition of Earth's surface by NASA).

Sampling procedures and criteria for inclusion/exclusion are detailed in Chapter 2.

# **8 This book**

The goal of this book is two-fold: to document the conceptual and methodological framework of our project (especially here and in Chapter 2) and to provide detailed qualitative/quantitative analyses of recruitment sequences in each of the eight languages: Cha'palaa, English, Italian, Lao, Murrinhpatha, Polish, Russian, and Siwu. Each language-specific study gives an overview of linguistic, gestural, sequential, and contextual features of recruitment sequences, following the categories defined in the coding scheme. While written to stand independently, the eight chapters adopt the coding scheme's common reference structure to facilitate navigation and comparison. At the same time, the chapters develop aspects and topics that are specific to each language and data.

By focusing on the phenomenon of recruitments, this large-scale collaborative study examines a domain of social action in interaction in which social relations are exploited, maintained, and potentially tested. We find that cross-linguistic diversity in this pragmatic domain is relatively low, considerably lower than the

### 1 Recruitments and pragmatic typology


Table 1: Languages covered in this volume, data sources, coding credits.

diversity observed in phonological, morphosyntactic, and semantic systems. This is in line with the idea that a species-wide infrastructure for interaction underpins the use of language, largely independent of the specific shape of that language (see Levinson 2000; 2006; Schegloff 2006; Enfield 2013; Enfield & Sidnell 2013; Stivers et al. 2009; Dingemanse et al. 2015). This is not to say that these pragmatic systems are identical. The chapters of this book show that there are differences. But we are struck by the commonalities that our approach reveals in a domain of language where many might expect to find radical variation.

Giovanni Rossi, Simeon Floyd & N. J. Enfield

# **Acknowledgments**

Our biggest thanks go to all contributors for their sustained work on the project, and for their patience with the editors. Thank you also to our collaborators for their input and intellectual engagement (see §3) and to our reviewers who helped to improve the volume enormously with their constructive commentary. We owe a particular debt of gratitude to Sebastian Nordhoff who directed the production of the volume with great generosity and patience. We gratefully acknowledge our funding bodies: the European Research Council (INTERACT and HSSLU projects), the Max Planck Society (Language and Cognition Department, directed by Stephen C. Levinson), the Dutch Research Council (NWO), the Australian Research Council, and the Academy of Finland.

# **References**


1 Recruitments and pragmatic typology


Giovanni Rossi, Simeon Floyd & N. J. Enfield


### 1 Recruitments and pragmatic typology


Giovanni Rossi, Simeon Floyd & N. J. Enfield


1 Recruitments and pragmatic typology


Giovanni Rossi, Simeon Floyd & N. J. Enfield


### 1 Recruitments and pragmatic typology


# **Chapter 2**

# **A coding scheme for recruitment sequences in interaction**

# Simeon Floyd

Department of Anthropology, Universidad San Francisco de Quito

# Giovanni Rossi

Department of Sociology, University of California, Los Angeles

# N. J. Enfield

Department of Linguistics, The University of Sydney

This chapter provides an annotated coding scheme for analyzing recruitment sequences in video-recorded social interaction. The scheme provided a basis for the research presented in the eight language-specific chapters of this book, and as such it gives necessary context for understanding the comparative project reported on here and in associated work. It is also intended to serve as a stepping stone for other researchers to use in the analysis of recruitment sequences in other languages. The scheme features guidelines for building collections and aggregating cases based on interactionally relevant similarities and differences among instances. The questions and categories featured in the scheme are motivated by inductive observations of interactional data, grounded in the framework outlined in the introduction to this volume.

# **1 Introduction**

The present coding scheme provides a way to systematically analyze a core set of formal and interactional features of recruitment sequences (defined in Chapter 1, §4). The scheme is the product of the collaborative qualitative study of such sequences in different languages, based on the analysis of video recordings

Simeon Floyd, Giovanni Rossi & N. J. Enfield. A coding scheme for recruitment sequences in interaction. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 25–50. Berlin: Language Science Press. DOI:10.5281/zenodo.4018372

Simeon Floyd, Giovanni Rossi & N. J. Enfield

of social interaction and focusing on the details of language and other conduct surrounding recruitments. Such analysis allowed us to identify recurrent socialinteractional dimensions and patterns of language usage, leading to the formulation of questions aimed at capturing these aspects in different languages. The coding scheme is therefore inductively derived from an iterative process of observation, analysis, and group discussion of naturally occurring data (see also Stivers & Enfield 2010; Dingemanse et al. 2016, among others).

The coding scheme (§6) is preceded by a definition of the phenomenon (§2), further specifications for inclusion/exclusion of a case from the data considered (§3), instructions for sampling and collecting cases (§4), and general guidelines for coding (§5). The questions and entries in the body of the scheme are extensively annotated to aid in the understanding of distinctions and replicability of coding.<sup>1</sup>

Besides documenting the analytical procedure of the project, the scheme is published here as a reference for future work, to foster comparable and cumulative research in the interactional domain of recruitment. The scheme can be applied to any type of face-to-face naturally occurring interaction featuring people getting others to do things for or with them.

# **2 Definition and terminology**

Recruitment sequences are defined as in Chapter 1, §4. A recruitment sequence is a sequence of two moves with the following characteristics:


For the purpose of this coding scheme, the following components of a recruitment seqence are identified:

• two participants: A (the recruiter) producing the instigating action, and B (the recruitee) responding to it;

<sup>1</sup>We include all of the original questions and entries. However, note that a few categories were not coded with sufficient reliability by the researchers involved in this project; such categories may require further training and calibration of coders, or else reformulation to make reliable coding possible.

2 A coding scheme for recruitment sequences in interaction

	- **–** fulfillment: a practical action involving physical work performed for or with A;
	- **–** rejection: the conveyance of inability or unwillingness to fulfill the recruitment;
	- **–** deferment of fulfillment;
	- **–** initiation of repair (often leading to fulfillment after repair).

If B ignores Move A, then the sequence must include an explicit, on-record practice of initiating recruitment and/or pursuit of response in order to be included.

The sequence can be expanded by a move c, doing acknowledgment.

# **3 Further specifications for the inclusion/exclusion of cases**

In including and excluding cases for coding, there are a number of further specifications that can help to delimit the phenomenon of recruitment from other kinds of related sequences of interaction.


Simeon Floyd, Giovanni Rossi & N. J. Enfield

	- i) After the table is set, A looks around in search of her plate, which is missing; B notices this and walks to the kitchen to get it.
	- ii) As A begins to eat her hamburger, B notices that the hamburger does not contain mayonnaise and passes the mayonnaise jar to A.

The first case should be included insofar as having a plate is necessary for A to be able to eat, as displayed by A's looking for it. The second case should be excluded insofar as the additional ingredient is not necessary for A to be able to have her hamburger, as shown by the fact that she is already eating it.

	- A: *Give me a knife*
	- B: ((gets the knife))
	- B: *Here you are* « not a separate seqence
	- A: ((reaches out to grab knife)) «

These should not be considered as separate recruitment sequences. At the same time, dependent or responsive recruitments should be distinguished from cases in which a practical action is independently recruited to deal with a contingency arising along the way toward the fulfillment of an earlier recruitment. For example:


These should be analyzed as two overlapping or nested recruitment sequences.

2 A coding scheme for recruitment sequences in interaction

5. **Repair initiation and solution.** When B responds to A's first attempt at recruitment (e.g. *I need a knife*) with repair initiation (e.g. *A knife?*) and A then provides a repair solution (e.g. *Yes*), this repair solution should be considered as Move A of a subsequent, separate recruitment case with its own unique identifier, even though the two cases belong to the same sequence (the sequential position of cases is coded in C01 below).


	- i) The vocative functions as a recruitment pursuit after one or more earlier attempts at recruitment have been made, e.g.



# **4 Guidelines for sampling interactions and collecting cases**

Coding must be based on a systematic and coherent collection of recruitment sequences from a broad sample of face-to-face interactions. The sample should include a range of different activities, settings, and speakers; it should contain both dyadic and multi-person interactions; and it should span both interactions that are task-focused (e.g. playing a game, preparing food, doing work together) Simeon Floyd, Giovanni Rossi & N. J. Enfield

and others that are talk-focused (e.g. just gossiping or chatting). Below are some guidelines for systematically sampling interactions and collecting cases.


The goal of these guidelines is to construct a sample that is representative of the diversity of the corpus at hand. How well this diversity represents social interaction in the target language will depend on how the corpus as a whole has been built.

# **5 General coding guidelines**

The coding sheet should contain a transcript of the core interactional moves of each recruitment sequence, including a basic transcription of verbal elements and a concise description of nonverbal elements (see §6 A. Basic data, below). This transcript is intended to make the coding data intelligible to other analysts and as a reference for coding; however, it is often not enough to be able to accurately code certain features of the sequence, for example, the strengthening of the recruiting move by means of prosody (question C11) or the visibility of a target object (question E05). For this reason, coding should be based whenever possible on direct inspection of audio and video streams, possibly supported by a more detailed transcript of the larger interaction.

### 2 A coding scheme for recruitment sequences in interaction

When in doubt, coders should choose the most conservative coding choice, or a choice that best reflects the potential equivocalness of the feature in question, or the "can't tell" option, if available. For any coding decision, coders should be able to provide a reasoned argument and evidence to support it. Comments about particular coding choices should be entered in a notes field at the end of the sheet so as to be available to other analysts examining the coding and as a bridge between the complexity of human behavior and its reduced representation in coding data (Stivers 2015).

In this project, we decided to compile a glossary of certain verbal practices that make up a language's repertoire of resources for initiating and responding to recruitment, including modal constructions (C07), mitigators and strengtheners (C11), and benefactive markers (C12). The coding process provides an opportunity for easily compiling such a glossary by creating an entry in a dedicated tab of the coding sheet every time a recurrent practice is identified in connection with a coding question. Questions suitable for glossary entries are marked with a superscript "gl" (e.g. C07gl).

# **6 The coding scheme**

### **A. Basic data**

This section records the basic data for every recruitment sequence, including Moves A and B. Each case has a unique identifier, which is used to locate it in a recording, refer to it in qualitative analysis, and for statistical purposes in quantitative analysis. Verbal elements are transcribed and translated and nonverbal elements are concisely described to make cases intelligible to other analysts.

### A01 **Language**

A02 **Unique identifier for the case.** Suggested format: recording\_timecode (e.g. Housemates\_3211246). In the rare event that two cases begin simultaneously, use an additional symbol to distinguish them (e.g. Housemates\_3211246a, Housemates\_3211246b).

A03 **Is a child involved as either recruiter or recruitee?**


Simeon Floyd, Giovanni Rossi & N. J. Enfield

> Make this choice according to your understanding of the duration of childhood in this particular culture.<sup>2</sup>


Descriptions of nonverbal behavior should be concise and pitched at an appropriate level of granularity. For example, ((gestures at the salt)) is too general and ((raises arm and extends index finger toward the salt)) is too elaborate. Different projects will have different requirements for the description of nonverbal behavior and cases differ in complexity, but for the example above, ((points at salt)) is the right level of granularity for most purposes. Moreover, descriptions of nonverbal behavior should stick to what is visible at that particular moment. For example, ((stands and walks toward spices shelf)) is more objective than ((goes to get salt)) because at that moment it is still not certain how B's compliance with the recruitment will develop.

<sup>2</sup> For different projects, different sociodemographic categories (based on age, ethnicity, socioeconomic status, etc.) may be flagged to allow for sorting or comparison (see, e.g., Stivers & Majid 2007). In this project, we decided to treat recruitments involving children separately to increase comparability among cultures and corpora.

### 2 A coding scheme for recruitment sequences in interaction

In this project, we decided not to record gaze behavior in the transcription of Move A's nonverbal component for reasons of economy, as there is a dedicated coding question about gaze (C16). However, we did record gaze in the transcript when it was used as a pointing gesture or "eye point" (Wilkins 2003) toward an object or location relevant to the recruitment, which made it easier to code this as a pointing gesture in question C02.

### **B. Recruitment categories**

This section contains questions designed to identify the main interactional "problem" or "point" of a recruitment sequence with reference to four categories, which were inductively derived from qualitative observation and analysis of recruitment episodes in our languages. This categorization serves multiple purposes, including grouping cases that share similar interactional contingencies, such as the visibility and accessibility of a target object (see questions E05, E06), and grouping cases in which Move A has a similar function, to test whether this has an effect on its form across languages.

In keeping with our recruitments approach, cases are assigned to categories primarily on the basis of B's response (see Chapter 1, §4). However, A's instigating behavior, along with any subsequent pursuits, should also be considered, especially in cases where B ignores it; here, the answer will reflect an understanding of what A's behavior is "going for" or working toward.

While it is possible to assign a case to more than one category, it is usually more useful to pick the most fitting or salient one. Also, while cases should be assigned to a category whenever possible, there will be cases that do not fit any of the categories but still fall within the broader definition of recruitment, for example, calling or motioning for a collective clinking of glasses, or initiating a joint recreational activity like playing chess.

Finally, a note on multiple cases belonging to the same recruitment sequence: when B responds to A's first attempt at recruitment (e.g. *I need a knife*) with repair initiation (e.g. *A knife?*) and A's subsequent action is constituted by a repair solution (e.g. *Yes*) followed by B's fulfillment (e.g. brings knife), questions B01- B04 should be answered in the same way across the two cases (the sequential position of cases is coded in C01). Example:


	- 1. yes
	- 2. no

"Giving an object" specifically refers to the physical transfer of a moveable object from the control of one person to another, usually released and grasped by the hands. This does not include cases in which B moves out of the way or otherwise facilitates A's taking possession of an object.

# B02 **Does B do a service for A in response to some behavior by A? Or, alternatively, was this the point of a sequence that went unfulfilled?**


A "service" is intended as a practical action involving some manipulation of the material environment (e.g. washing the dishes, feeding the chickens). While giving an object to someone can also be seen as a "service", it is a particular kind of service that is worth considering separately (see, e.g., questions E05, E06), so if you have answered "yes" to question B01, you should normally answer "no" to B02.

	- 1. yes
	- 2. no

"Altering" the trajectory of an ongoing behavior includes both adjusting or changing the behavior (doing the same thing in a different way) and ceasing the behavior altogether. These two kinds of alteration have different implications for how the recruitment sequence is fulfilled: doing something differently vs. not doing something anymore. Do not answer "yes" when the cessation of a behavior is incidental to performing a service or object transfer. Consider the following: if A says *Stop playing with your food and eat*, the cessation of eating is an integral part of the recruitment; however, if A says *Is there still some beer left?* and B stops watching the TV in order to go to the kitchen to get beer for A, the cessation of B's ongoing behavior is incidental to the recruitment.

2 A coding scheme for recruitment sequences in interaction

B04 **Does B do a practical action to address some current or anticipatable trouble for A?**


This question is aimed at capturing cases in which B provides assistance without this being solicited or expected by A, but rather instigated by A's display of current or anticipatable trouble (e.g. A arrives at a door with her hands full of heavy objects and B opens the door, or A grasps for the salt but is unable reach it so B pushes the salt closer).

### **C. Move A: The recruiting move**

### C01 **In the in-progress sequence, what is the position of Move A?**


Here we consider the sequential position of Move A, coding whether it is: the first and only attempt in a minimal sequence that is immediately completed ("one and only"), a first attempt in a longer sequence that is not completed in one go ("first of non-minimal"), a final attempt in a longer sequence ("last of non-minimal"), or a subsequent attempt that was neither the first nor the last ("*n*th"). When considering the sequential position of Move A, it is important to remember that certain preliminary moves, also referred to as "pre-requests" (see Levinson 1983: chap. 6; Rossi 2015), can function as recruiting moves on their own and lead to immediate completion.

```
Move A Are you using that pen? C01 = one and only
Move B ((passes A the pen))
```
However, preliminary moves may also be responded to with a go-ahead leading to a subsequent, more explicit Move A (or to multiple subsequent attempts).


Simeon Floyd, Giovanni Rossi & N. J. Enfield

> Other preliminary moves that cannot mobilize the relevant practical action in next position should not be considered as recruiting moves, though they are part of the recruitment episode. These typically involve generic presequences (Schegloff 2007: chap. 4):


*Only answer C02 if A06* ≠ *[not relevant] or [can't tell]*

# C02 **Concerning the nonverbal behavior, what does it consist of?**


These types of nonverbal behavior were derived from qualitative observation and analysis across languages, and were identified as recurrent relevant behaviors that either accompany verbal elements of the recruiting move or initiate recruitment on their own.

If multiple types of nonverbal behavior co-occur in the same recruiting move, answer this question based on the most salient behavior.

Pointing gestures include not only manual points, but also head points, lip points, and "eye points" (Wilkins 2003). As explained above in Section A, gaze should be considered here only when it is used to indicate an object or location relevant to the recruitment. Gaze used for recipient selection (Lerner 2003) is dealt with by question C16 and – in this project – is not be transcribed in A06.

*Only answer C03–C12 if A04* ≠ *[none] or [can't tell]*

### 2 A coding scheme for recruitment sequences in interaction

### C03 **Does the verbal behavior consist of a simple or complex construction?**


There are two main types of cases where verbal behavior should be coded as "complex":


### C04 **Does the verbal behavior include a directly-involved nominal referent?**


By "directly-involved nominal referent" we intend a referent that is the target object (e.g. *Water please*) or that is otherwise implicated in the recruited action (e.g. *Is the window open?* when the goal is to have B close the window). Such a referent may be encoded either with a full noun phrase (e.g. *Pass me the salt*) or with a pronominal element (e.g. *Pass me that*). Directly-involved nominal referents are easy to identify with most transitive predicates (e.g. *Clean the table*, *Light my cigarette*). With ditransitive (three-place) predicates, relevant referents will also include the "recipient" of the action (e.g. *Give it to Dad*, *Pass him* the lighter). Verbs such as *get* and *take* can also be treated as belonging to this group in that their semantics involves a location where an object is taken or gotten *from* (e.g. *Get it from the trolley*) (Fillmore 1977). Semantics aside, nominals can be directly involved in recruited actions in different ways and no single rule can capture all eventualities. But here are some examples that were collectively discussed during the project with an explanation of the rationale for the coding decision.

• *Sit on the chair*. Answer "yes" because A is telling B to sit specifically on the chair and not just anywhere (e.g. on the couch or floor); this referent is integral to the recruited action.

Simeon Floyd, Giovanni Rossi & N. J. Enfield


"Full noun phrases" typically involve open-class items referring to people, things, locations, whereas "pronominal" elements are reduced, closed-class pro-forms such as demonstratives and other deictics. For languages with the possibility of zero anaphora, if there is no overt pronominal form, stick to the surface and answer "no".

In the case of a complex verbal component (see question C03), consider it holistically; for example, a complex construction with multiple referents like *Get a pot from that cabinet and put it on the stove* should be coded as "yes, full noun phrase".

### C05 **What is the sentence type?**


In most languages, it is possible to distinguish different formal types of sentences that prototypically encode asserting or informing (declaratives), asking or questioning (interrogatives), and directing or ordering (imperatives) (Lyons 1977; Sadock & Zwicky 1985; König & Siemund 2007). These are formal, logico-semantic types that encode basic ways of dealing with propositional content. The criteria for assigning utterances to these three types may vary according to the internal organization of each language, but as a rule of thumb you can ask: how would this utterance be understood out of context? Some languages may have "other" major or minor sentence types (e.g. exclamatives, "insubordinated" *if* constructions, etc.). Since sentence type is based on there being a predicate, when Move A does not include a predicate (e.g. *Water please*), choose "there is no predicate".

*Only answer C06–C07 if C05* ≠ *"there is no predicate"*

### C06 **Is there a predicate that refers to the target action?**


This questions asks whether the action that is the projected outcome of the recruitment is explicitly referred to by a predicate in Move A. Examples:


*Only answer C07 if C05 = "interrogative", "declarative", or "other"*

C07gl **Does the predicate encode obligation, permission, ability, or volition to perform an act?**


This question is about specific modal categories: obligation/necessity (e.g. *You must finish your dinner*, *The door is to be shut*), permission/authorization (e.g. *May I have that last piece of cake?*), ability/possibility (e.g. *Can you pass me the salt?*, *You could start washing up*), volition/willingness (e.g. *I would like some water*, *Will you hand that to me please?*). These meanings must be semantically encoded. For example, a sentence like *You're standing on my foot*, although it may pragmatically oblige the recipient to step away, does not encode obligation. In English and other European languages, obligation, permission, ability, and volition are frequently encoded with modal

verbs such as *must*, *have to*, *need*, *may*, *can*, *will*. Other languages may use affixes or dedicated constructions. Cha'palaa, for example, encodes necessity with an infinitive verb followed by a finite 'be' verb.

*Only answer C08 if C05* ≠ *"imperative"*

C08 **Is the main subject overtly marked for person?**


"Overt grammatical person" refers to morphosyntactic and lexical categories in the language that encode the person of the subject-like argument of the verbal component, whether as a noun or noun phrase (e.g. *Grandma needs a blanket*), pronoun (e.g. *Can you pass the salt?*), clitic, verbal inflection (etc.), or a combination of these. This must be an overt, surface form. "Third person" refers to grammatical third person, regardless of whether the referent is a potential participant in the speech event (e.g. *Somebody should close the door*) or not (e.g. *The door should be closed*, *It's cold in here*).

Vocatives (e.g. *John, water please!*) do not constitute a form of person marking, but stand-alone pronouns do (e.g. *You, water!*). If you find overt grammatical subjects in your sample that bridge two or more of the categories listed above, code for the most specific person value that can be obtained from the construction, or choose "other" if you feel that none of the above choices apply.

	- 1. yes
	- 2. no
	- 3. can't tell

"Core constituents" refers to a predicate with its core arguments, which will normally be up to one argument with intransitive verbs (e.g. *You stop!*), up to two arguments with transitive verbs (e.g. *Can you close the door?*), and, in some languages, up to three arguments with ditransitive verbs (e.g. *You should give her a spoon*). Sometimes you will have to decide whether an element is a true ditransitive object (*Can you pass him the salad?*) or is marking a non-core beneficiary (as in the Italian: *Tienimi questo*, 'Hold **me** this'). In the former case, the answer to this question is "no"; in the latter, it is "yes".

Additional elements beyond core grammatical constituents typically belong to one of the following four categories:


*Only answer C10–C11 if C09 = "yes"*

# C10 **Does the move include a clausal explanation?**


A "clausal explanation" makes reference to a past, present, or future state of affairs or event that provides grounds for the recruitment or makes it more intelligible to B. This covers any kind of reason-giving, including accounts for untoward or imposing behavior (e.g. *Stop talking so loudly, I have a headache*) as well as more general explanations that make the recruitment more understandable or clear (e.g. *Keep stirring the sauce so it doesn't get lumpy*).

The clausal explanation must be built into Move A (single package). If the explanation is provided after a self-contained Move A has been produced (two packages) then there are two main possibilities.

i. Participant A gives a reason after no uptake comes from B or when it is not clear that B will comply. In this case the explanation effectively counts as a second attempt at recruitment and must be entered as a separate case. Example:


ii. Participant A gives a reason for the recruitment after it has been already fulfilled by B, or after B has clearly shown that they are on their way to comply. We can define this as a "post-hoc" explanation, justifying the launch of the recruitment sequence after it has been complied with. Such explanations are not part of Move A. Example:


# C11gl **Does the move include a mitigating or strengthening element?**


This question asks about elements that mitigate or soften the recruiting move (e.g. *Move the car if it's not too much trouble*, *Can I have a little water?*) or, alternatively, that strengthen or aggravate it (e.g. *I would really like some water*, *Get the key right now*). These elements may be clauses, phrases, adverbs, particles, affixes, or other forms. Do not consider clausal explanations (C10) when answering this question.

# C12gl **Is there formal benefactive marking?**


This question asks if the verbal component includes an explicit beneficiary of the recruited action, which may be A (e.g. *Could you move that a little bit for me please?*, *Read me a book!*) or, alternatively, another participant or

combination of participants: B (e.g. *Will you pass the cards so I can cut them for you*?, *Grab yourself a beer!*), both A and B (e.g. *Can you set the table for everyone*?), or a third party C (e.g. *Get him a fork!*). Formal benefactive marking includes datives (e.g. *Read me a book!*), prepositional phrases (e.g. *Could you move that a little bit for me please?*) and other resources such as specific constructions (e.g. *Do me a favor and*…). Constructions with verbs of need (e.g. *I need a lighter*) do not qualify as including formal benefactive marking.

# C16 **Does A gaze at B during Move A?**


The purpose of this question is to code gaze as a design feature of the recruiting move. What is relevant is whether A is looking at, or trying to establish eye contact with, B. Answer "yes" on the basis of A's behavior regardless of whether B perceives being gazed at or not. If you have reasons to believe that B does not perceive being gazed at by A, it is recommended to flag this in the general notes field.

# **D. Move B and Move C: Responding and acknowledging**

The questions in this section concern the responding move or Move B, and the potential expansion of the sequence with an acknowledgment or Move C.

# D01 **What is the response doing relative to the recruitment?**


*"Quickly fulfills" versus "Plausibly starts fulfilling"*. It is useful to distinguish between two ways of positively responding to recruitment: doing

the target action within a short time frame immediately after Move A and doing something that could plausibly be construed as the beginning of fulfillment (but still possibly equivocal), over a longer time span. To make this decision, put yourself in the position of participant A and ask what he or she would be aware of in the first few seconds after Move A. Note that some recruited activities inherently take more time than others (e.g. setting the table for lunch, getting an object that is far away) and should always be coded as "plausibly starts fulfilling". Cases in which B commits to later fulfillment (e.g. *Oh sorry I'm busy right now but I'll do that in half an hour*) should also be coded as "plausibly starts fulfilling".

*"Rejects"*. All clearly negative responses such as refusing (e.g. *No I won't do that*) and/or giving an account for non-compliance (e.g. *I'm too busy now*) should be coded as "rejects". Different types of rejections are distinguished by subsequent questions (D02 and D03).

*"Initiates repair", "ignores", and "other"*. Besides responding positively or negatively, B may respond to the recruiting move in other ways. One possibility is to initiate repair. Another is to ignore the recruiting move by not taking it up at all. This applies both to cases in which B would be in a position to hear/see the recruiting move but intentionally ignores it and to cases in which B might not have heard/seen the recruiting move (for example, because they are involved in a parallel activity, or too far away, etc.). Other cases in which the recruiting move is taken up but the response does not fit any of the above categories should be coded as "other". Examples of "other" responses are:


# D02 **Does the response include a positive or negative polar element?**


Positive polar elements include verbal/vocal elements such as *yes*, *okay*, *sure*, *mm hm* as well as nonverbal elements like a head nod. Negative polar elements include verbal/vocal elements such as *no*, *mh mh* as well as nonverbal elements like a head shake. Coding should take into account that linguistic systems differ. For example, in some languages like Mandarin and Cha'palaa, one way to do a polar response is by repeating the verb.

*Only answer D03 if A07* ≠ *[none] or [can't tell]*

# D03 **Does the response include a clausal explanation/account?**


See notes for question C10 on explanations and accounts.

# D04gl **Is there acknowledgment by A?**


"Acknowledgment" includes thanking (e.g. *Gracias!*), other expressions of gratitude (e.g. *Cheers, I appreciate it, Oh I'm so glad you can do this for me*), and more generally any positive conveyance of appreciation or satisfaction by the recruiter immediately after receiving a response indicating fulfillment. In some cases, fulfillment may be still ongoing or forthcoming at the time of the acknowledgment.

*Only answer D05–D06 if D04 = "yes"*

D05 **Transcribe and translate the acknowledgment** (*Muchas gracias* 'Thanks a lot'). If the acknowledgment includes nonverbal behavior, briefly describe it, e.g. ((nods repeatedly)).

# D06gl **Is there a subsequent move by B responding to the acknowledgment?**


Such a response may be conventionalized (e.g. *You're welcome*) or not (e.g. *Oh well I owed you this one*).

*Only answer D07 if D06 = "yes"*

D07 **Transcribe and translate the subsequent move by B responding to the acknowledgment.**

# **E. Other elements of the recruitment sequence**

The questions in this section code for other elements of the recruitment sequence beyond Moves A, B, C.

### E01 **Is there an evident local immediate beneficiary for the recruitment?**


The answer to this question is in principle independent of, and possibly incongruous with, the answer given to question C12 (which deals with *formal marking* of beneficiaries). E01 can be a tricky question, but try not to overthink the issue and choose the most straightforward answer. If in doubt, be conservative and answer "no".

# E02 **Is the interaction dyadic?**


For the answer to be "yes", there should normally be only two people in the video recording at the time at which the recruitment occurs; if there are three or more people, answer "no". In some cases, a stretch of interaction may be considered dyadic even though other people are present in the immediate vicinity but are clearly not part of the interaction.

### E03 **Is a vocative used?**


Vocatives normally involve a proper name, kin term, title, or similar, and provide a way of explicitly addressing the recruiting move to a specific recipient or set of recipients. The vocative may be built into Move A (e.g. *Can you pass me the knife John*?, *You two guys*,*come with me*) or be part of a summons-answer sequence that precedes the recruiting move (see also C01):

Summons *Hey Bob*! E03 = yes Answer *What?* Move A *Come here* Move B ((goes to A))

### E04 **Can A and B's relationship be characterized as socially asymmetrical?**


In this question we code for any salient social asymmetry between A and B, based on the researcher's knowledge of the society. The question refers to enduring asymmetries between A and B that hold across contexts. Social asymmetries can be based on age (e.g. older-younger siblings) as well as other kinds of social status (e.g. authority roles such as husband-wife, parent-child). The answer should be based on prescriptive norms and general cultural expectations of the community, not on the instantiation of the relationship in the recruitment sequence, so you should not use the recruitment sequence as a basis for your judgment: evidence for the asymmetry must be independent of it. Social asymmetry is gradient, so judge whether a dyad is *relatively* symmetrical or asymmetrical.

*Only answer E05–E06 if B01 = "yes"*

### E05 **Is the target object visible to A?**


### E06 **Does B have better access to the object in question than A?**


Typically, B is "using an object" when he or she is currently manipulating it. Cases in which B has been making use of the object all along and has only momentarily rested it somewhere when the recruitment is attempted should be coded as "yes, B is in possession of the object but is not using it". Possession does not require that B be the legal or socially recognized owner of the object; it applies to all cases where B has the object "on them" (e.g. in their pocket) as well as to cases where the object is enclosed into another possession of B's (e.g. their bag). For cases where relative closeness is relevant, try not to overthink the issue and answer "yes" only when there is a clear difference in distance (e.g. the object is within B's reach and visibly far from A).

*Only answer E07 if B02 = "yes"*

### E07 **Is B in charge of, or especially responsible for, the service in question?**


Only choose "yes" if the answer is clear. The kind of responsibility implied cannot be just a matter of proximity or availability, but must be linked to an individual and his or her social role, or derived from a previous agreement to do the action (preferably documented in the recording). As an example of the former, in Chachi society young girls are expected to bring water from the river to the house, and are more responsible for this task than males and people of other ages. As an example of the latter, in one case in the Italian corpus a woman agreed to add stock cubes to a soup but was then distracted and did not do it; fifteen minutes later, she was told to do the task. As with other questions, if you are unsure, be conservative and answer "no".

2 A coding scheme for recruitment sequences in interaction

# **Acknowledgments**

This coding scheme has benefited from the sustained input of all project contributors – Julija Baranova, Joe Blythe, Mark Dingemanse, N. J. Enfield, Simeon Floyd, Kobin H. Kendrick, Giovanni Rossi, Jörg Zinken – and of other participants and conceptual collaborators in data analysis and discussion – Tyko Dirksmeyer, Paul Drew, Rósa S. Gísladóttir, Steve Levinson, Elizabeth Manrique.

# **References**


Simeon Floyd, Giovanni Rossi & N. J. Enfield


# **Chapter 3**

# **Getting others to do things in the Cha'palaa language of Ecuador**

# Simeon Floyd

Department of Anthropology, Universidad San Francisco de Quito

This chapter describes the resources that speakers of Cha'palaa use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Cha'palaa, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). The present chapter extends our knowledge of the structure and usage of the Cha'palaa language with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

This paper offers a first description of one area of everyday interaction among speakers of the indigenous Cha'palaa language of Ecuador, sequences in which one party "recruits" the behavior of another for some practical action such as transferring an object or physically assisting with or collaborating in an activity. The analysis of these instances is based on a video corpus of informal conversation recorded in the Chachi communities where Cha'palaa is spoken. This area of Cha'palaa interaction is characterized by a tendency toward direct recruiting

Simeon Floyd. Getting others to do things in the Cha'palaa language of Ecuador. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 51–92. Berlin: Language Science Press. DOI:10.5281/zenodo.4018374

### Simeon Floyd

moves, employing an extensive set of imperative formats, within the contexts of the different rights and responsibilities of individuals in Chachi society.

# **1.1 The Cha'palaa Language**

The Cha'palaa language is spoken by the Chachi people in small communities and homesteads along the rivers of the Ecuadorian Province of Esmeraldas between the Andean foothills and the Pacific coast. It is one of the modern members of the Barbacoan family, which was once the dominant language family of the region corresponding to northern Ecuador and southern Colombia until it was displaced by Quechuan languages and, later, Spanish, in much of the Andean highlands. The Chachi people avoided the pressure of language shift by migrating to the coastal lowlands where they live today (Jijón y Caama 1914; DeBoer 1996; Floyd 2010). Estimates of the number of speakers vary between about 6,000 (Instituto Nacional de Estadística y Censos 2010) and 10,000 in Ethnologue (Lewis et al. 2014).

Until recently, Cha'palaa was a relatively unstudied language, with only a few descriptive sources by missionary linguists produced over the last few decades (Moore 1962; Lindskoog & Lindskoog 1964; Vittadello 1988); recent work by the author has begun to bring more aspects of the language and the interactive practices of its speakers to light (Floyd 2009; 2010; 2014a,b; 2015; 2016; 2018; Floyd & Bruil 2011; Dingemanse & Floyd 2014; Floyd & Norcliffe 2016; Dingemanse et al. 2017). Like many South American languages, Cha'palaa has a basic SOV word order with extensive agglutinating verbal morphology. Some of its grammatical features that are relevant for recruiting practices include its large imperative paradigm, its egophoric system (a distinctive type of epistemic marking), its complex predicate system, its morphological case markers, and other elements that will be described in the sections that follow.

# **1.2 Data collection and corpus**

The video corpus on which this work is based was constructed in accordance with a set of guidelines developed by and for the members of the comparative project being reported on in this volume (see Chapters 1–2). The corpora studied in this project feature informal conversation among friends and family, and participants received no special instructions other than to talk or go about their daily activities as they wished.

The Cha'palaa corpus was recorded by the author over a period between 2007 and 2015 in household and village settings in Chachi communities of north-west-

### 3 Getting others to do things in the Cha'palaa language of Ecuador

ern Ecuador. The majority of the recordings come from the Rio Cayapas area, particularly from its tributary the Rio Zapallo, and a few come from other areas. In most cases the camera was placed in a household or common area during regular daily activities and then retrieved after about an hour. All videos included adult participants (adolescents or older), including dyads and larger groups of family members and friends, sometimes changing configuration during filming, with children often coming and going. Participants were involved in cooking, eating, doing other household tasks, making handcrafts such as woven baskets, or simply relaxing and conversing.

The data considered for analysis consists of a sample of selections from the recordings (see Chapter 2, §4). The goal was to identify at least 200 recruitment sequences among adults. The nature of the Cha'palaa corpus, made up primarily of recordings in multi-generational households, meant that there were many cases in which one or both of the involved participants was a small child (not yet adolescent, below about 12 years). These were excluded from the present comparison in order not to introduce complicating issues of language development. Excluding these cases, which were twice as frequent as adult-only cases, necessitated reviewing a large sample of about 9.5 hours from 16 different recordings to reach a total of 205 cases (out of the initial 653, with 448 child-involved cases that were excluded). Excerpts from this sample of cases are presented below to illustrate the range of linguistic forms and practices that make up the Cha'palaa recruitment system.

# **2 Basics of recruitment sequences**

As defined in Chapter 1, §4, a recruitment is a basic cooperative phenomenon in social interaction consisting of a sequence of two moves with the following characteristics:


In general, "practical" actions in Move B (M-B) were considered to be goaldriven bodily movements or manipulations of the physical environment, and contrast with states of rest and inactivity. Such practical actions can often be thought of as "target" actions of participant A, when they are made explicit in

### Simeon Floyd

Move A (M-A). Also, participant A may often be a beneficiary of the recruited action, but in other cases both A and B together, and sometimes even mostly B, may benefit from the outcomes of the recruitment sequence. Further details relevant for this definition, including what happens when Move B rejects the recruitment, or when there is no response, will be discussed in §4. The next subsections give some basic examples of recruitment sequences. In the transcripts, ▶ and ▷ designate Move A and Move B, respectively.

# **2.1 Minimal recruitment sequence**

Extract 1 provides an example of a minimal recruitment sequence in Cha'palaa. In Move A, participant A turns to look at participant B, his wife, and uses an interrogative format to invoke a specific target action: 'did you sweep? did you sweep?'. As part of his recruiting move, he also adds a reason for doing this action: 'the child is in all that trash'. Participant B's reaction is to walk off camera and to return shortly with a broom, sweeping the spot indicated by participant A.

(1) CHSF2011\_01\_11S2\_1531121

```
▶ 1 a mankashyu mankashyu
          ma -n -kash -yu ma -n -kash -yu
          again-IPFV-sweep-EGO again-IPFV-sweep-EGO
          did you sweep? did you sweep?
  2 na tsamantsa ujtu'paatala
          na tsamantsa ujtu'-pala -tala
          small very.much trash-place-among
          the child is in all that trash
▷ 3 b ((leaves))
  4 (15.0)
▷ 5 b ((returns with broom, sweeps))
```
The format selected by A in line 1 illustrates how a distinctive feature of Cha'palaa's grammar, an "egophoric marker" (Floyd, Norcliffe & San Roque 2018; also referred to as "conjunct-disjunct" markers, see Hale 1980; DeLancey 1992; Bickel 2000; Creissels 2008; Dickinson 2000; Post 2013) is employed for the interactive function of instigating a behavior on the part of B. This type of knowledge-based morphology used in this context treats the addressee as the locus of knowledge (in statements the marker might associate with the speaker's perspective in a similar way). While this is a distinctive morphological resource of Cha'palaa, its usage for recruitment also fits a more general pattern of question-like formats. Looking at Move B, the uptake by B provides evidence that functionally this

3 Getting others to do things in the Cha'palaa language of Ecuador

question was taken as a request for the provision of a service, namely, sweeping up.

# **2.2 Extended recruitment sequence**

In the simplest sequences, B takes steps to accomplish the target action in Move B immediately after A produces Move A, but this does not always occur. Recruitment sequences sometimes feature more than one recruiting move; these cases may be "pursuits", in which A repeats a version of the recruiting move (Pomerantz 1984; Bolden et al. 2012) or other types of sequences in which the response to Move A comes later, such as in repair sequences like that seen in (2). Participant A, a woman who is washing her clothes on the shore, asks her friend participant B for a plastic tub, but before she passes it to A, B requests a clarifying confirmation of the target object, in line 2. After A provides this confirmation, B accomplishes the target action.

(2) CHSF\_2012\_08\_04S4\_1712020

```
▶ 1 a Daira ñaa inu tina ka' eede
          Daira ñu -ya i -nu tina ka -tu ere -de
          Daira 2SG-FOC 1SG-ACC tub grab-SR pass-IMP
          Daira you pass me the tub
▷ 2 b enstaa? ((pointing at tub))
          ensta-a
          this-Q
          this one?
▶ 3 a jee tsadekee
          jee tsa-de-ke-e
          yes SEM-PL-do-IMP
          yeah do that
▷ 4 b ((throws tub to A))
```
In these types of non-minimal sequences it is possible to observe "side sequences" (Jefferson 1972), "insert sequences" (Schegloff 2007: chap. 6), and other types of intervening interaction that may occur between the original Move A and the fulfillment of the recruitment. When the request is not fulfilled in the first Move B, this can generate further iterations of the M-A/M-B structure until the sequence is completed (or abandoned).

# **2.3 Subtypes of recruitment sequence**

Despite considerable overlap, the concept of recruitment is intended to capture a broader range of phenomena than terms like "request" or "directive" (see Chapter 1, §4). Four broad subtypes of sequence are further identified as a way to cat-

### Simeon Floyd

egorize and analyze cases. These categories distinguish cases in which the target action of M-A is best thought of as (i) the provision of a service, (ii) the transfer of an object, (iii) the alteration of some ongoing trajectory of behavior, or (iv) if there was no clear M-A and participant B stepped in to provide assistance in response to A's current or anticipatable trouble. This last category is not a request in that there is no on-record solicitation of a response, but is a recruitment in that practical assistance is instigated by A's visible trouble. The term "on-record" here refers to identifiable moves in social interaction that ask for or otherwise overtly signal the need for a target action; categories (i), (ii) and (iii) were required to be on-record in this sense, while (iv) was not.

Table 1: Relative frequencies of recruitment sequence subtypes in the Cha'palaa sample (*n*=205).


Extracts1 and 2 have already provided examples of the two most frequently occurring categories; in (1) participant A requests the provision of a service, sweeping the floor, and in (2) participant A asks for an object, a plastic tub. Extract 3 shows an example of an alteration of a trajectory of action: A notices that B is sitting in such a way that she appears uncomfortable, and tells her to alter the way she is currently sitting to a more restful position, giving the reason that otherwise her back will hurt. This example also helps us illustrate how benefit may be differently distributed in recruitment sequences as here the primary beneficiary is the recruitee herself.

(3) CHSF2011\_06\_25S2\_3916900

```
▶ 1 a leka leka leka beenbushu kiya
          rest rest rest back hurt
          rest rest rest, (your) back will hurt
```

```
▷ 2 b ((reclines))
```
As illustrated in Table 1, alterations of trajectory were the least frequent of the sequence types in the Cha'palaa sample. Preliminary analysis of the cases involving children – excluded from the comparative data set, as mentioned above

### 3 Getting others to do things in the Cha'palaa language of Ecuador

in §1.2 – show many more attempts to alter and correct behavior in those cases, suggesting that social status may play a role, and that adults may try to avoid such potentially face-threatening interactions among each other, while in similar interactions with children such recruitments may be the norm. Future work with child-involved cases stands to shed more light on these issues.

The last subtype of case that was included in the sample were sequences in which B steps in to assist A with some problem that, while usually evident from the context, has not been explicitly formulated by A. For example, in (4), B and several boys were sitting in front of the kitchen door, scraping and eating coconut shavings. When A begins to approach with a heavy load of bananas, B and the others first gaze at her and then proceed to move the bowls, stools, and other objects out of her way, and to lean to the side to allow her to pass into the kitchen more easily (Figure 1).


While in these types of cases there is no on-record M-A by A, the types of services and objects that B provides in such cases are the same types of local practical actions that are explicitly asked for in other instances (e.g. 'Move over so I can pass', etc.).

# **3 Move A: The recruiting move**

The formats used by participant A in M-A could be fully nonverbal, fully verbal, or a composite of verbal and nonverbal elements. This section describes the composition of M-A in both the visual and the spoken channels.

# **3.1 Fully nonverbal recruiting moves**

Most of the recruiting formats in the Cha'palaa sample included spoken elements; of the 205 cases sampled, only nine were fully nonverbal in M-A (excluding cases of trouble assistance, see §2.3). An example of a fully nonverbal case is shown in (5), in which A and B are taking care of an injured chick together. During a moment when no spoken conversation is ongoing, A holds out the chick for B to hold for a moment so A can free his hands to manipulate some thread. B responds to A holding the chick out by reaching up to take it (Figure 2).

### Simeon Floyd

(a) Participant B and accompanying children collecting coconut shaving in a bowl, while participant A is approaching with a load of bananas (in front of them, off camera) (line 1).

(b) Participant B and children see participant A approaching and move the bowl and stool out of her way (line 2).

(c) Participant B and children lean out of the way as participant A passes with her load of bananas (line 2).

Figure 1: Frames from Extract 4. Family members facilitate a woman's arrival with a large load of plantains that she needs to deposit in the kitchen.

### 3 Getting others to do things in the Cha'palaa language of Ecuador


Figure 2: Participant A holds out his hand and participant B hands the chick to him.

The main formats for nonverbal requests in the sample were holding out objects, as in (5), and reaching out to receive objects. These were also two of the major nonverbal formats seen accompanying verbal recruiting formats, the topic of the next section.

# **3.2 Nonverbal behavior in composite recruiting moves**

In most Cha'palaa recruitment cases M-A includes no nonverbal behavior that is salient or relevant for the sequence; instead, target actions or other elements are expressed verbally. However, in 55 of 205 cases some relevant nonverbal behavior occurred. As mentioned in the previous section, nine of these were independent nonverbal recruiting moves, but the other 46 were composites including verbal and nonverbal elements. While some nonverbal behavior was idiosyncratic and did not lend itself to categorization, several well-defined types of practices made up the majority of the nonverbal elements for Cha'palaa. In Cha'palaa three practices accounted for about 85% of all nonverbal behavior seen in recruiting moves (see Table 2). Pointing was the most common of these, accounting for 42% of nonverbal elements. Pointing gestures usually indicated an object, location, or person that was relevant for the recruitment in some way. The

### Simeon Floyd

next-most-common practice was holding out an object that is instrumentally involved in the recruitment, typically to be taken by B to do something with, as seen in (5), above. The other major practice in the sample was reaching out to receive an object. In addition, there was one instance of iconic-symbolic gesture (beckoning for B to approach A), and about 15% were heterogeneous practices that did not fit into any of the cross-linguistic coding categories applied in the comparative project.

Table 2: Types of nonverbal practices in recruiting moves (*n*=55).


Figure 3 shows a pointing gesture that accompanied the spoken recruiting turn 'give me the string there' (full sequence shown in Extract 10). Along with his indexical point, participant A also uses lip pointing, a practice observed commonly among Cha'palaa speakers (Dingemanse & Floyd 2014) and in many other languages (Sherzer 1973; Enfield 2001).

Figure 3: Screenshot from Extract 10; index finger and lip pointing as part of Move A.

In some cases the nonverbal behavior was relatively complex, as in (6) below, in which A first extends his arm and points at the menthol ointment he is re-

### 3 Getting others to do things in the Cha'palaa language of Ecuador

questing (Figure 4), saying 'give me that also' and, after a brief pause, providing a reason for the request ('I will smell a little') as he turns his palm upwards to receive the object. One interesting element of the nonverbal behavior in this case is that A has already extended his arm by the first part of M-A, suggesting that he has high expectations that the request will be fulfilled, possibly based on it being a relatively "low contingency" request (Curl & Drew 2008; Craven & Potter 2010).

### (6) CHSF2011\_01\_11S3\_2692960


Figure 4: A (center left) reaches out while requesting the menthol (in line 1).

In Cha'palaa, these three practices of pointing, holding out objects, and reaching to receive object made up more that 80% of the total nonverbal behavior seen in recruiting moves (see Table 2).<sup>1</sup> However, there is an asymmetry between M-A and M-B in recruitment sequences in that while M-B tends to involve nonverbal elements, especially the accomplishment of the target action, more than two thirds of M-As were in the verbal channel (71%, *n*=146/205), with only 29% of

<sup>1</sup>These are practices with deep roots, being among the first to appear developmentally (Masur 1983; Cameron-Faulkner et al. 2015).

### Simeon Floyd

M-As including nonverbal elements (*n*=59/205).<sup>2</sup> The next sections describe the verbal recruiting formats.

# **3.3 Verbal elements: construction types and subtypes**

The spoken elements of M-A are mainly made up of the morphosyntactic resources of the Cha'palaa language (sometimes also with elements of Spanish, as the primary local second language). Verbal elements were classified according to cross-linguistic syntactic categories of declarative, interrogative, imperative (see König & Siemund 2007), as well as cases of "no predicate", and "other" cases for predicates that do not fit well with any of the main categories (see Chapter 2, §6). Not all languages distinguish among sentence types in the same way, but in most cases Cha'palaa features very clear and unambiguous morphological distinctions on the verb associated with the three major sentence types (more on this below). As for frequency, imperatives outnumber the others considerably.


Table 3: Construction type of recruiting moves including spoken elements (*n*=192).

### **3.3.1 No predicate**

Reviewing all of the different morphosyntactic types and their functions, we can start with cases in which there was no predicate, which belong to no sentence type in a strict sense. These cases can be classified by a few simple categories: of the 22 cases without a predicate, 12 name an object to be transferred, 6 name places that were relevant for the target action, 3 are vocatives selecting the recipient, and one was an interjection. Because the last two categories do not specify any element of the recruited action, they generally occur as a second attempt

<sup>2</sup>Note, however, that of the 146 cases involving language, in 38 the presence or absence of nonverbal elements could not be ascertained due to the recruiter being off camera or with visual access impeded by another participant.

### 3 Getting others to do things in the Cha'palaa language of Ecuador

to a previous recruiting move that was not successful (see also Kendrick, Chapter 4, §4.2.4). The other formats can generally function as independent recruiting moves as well as subsequent attempts. For example, speakers can name destinations as a way to tell addresses to go to those places or take things to or from those places. Sometimes other grammatical resources come into play, like the locative case marker with the first-person pronoun in (7) that specifies that the addressee should do something to or for participant A. Object naming usually functions to request the object in question, sometimes with additional material like in M-A of (7), which also specifies a recipient, but which leaves the target action of giving or passing up to the recipient's inference. While unspecified, the requested action is usually obvious from the context, and so this type of recruiting practice can be considered relatively explicit or on-record (on "namings" see also Rossi 2015b: chap. 2).

Also worth noting here is that, in M-B, participant B acts towards the fulfillment of the target action, but she does so in a particular way: by delegating to a third party (see also Enfield, Chapter 6, §6; Blythe, Chapter 7, §4.2.2). This was a strategy sometimes observed when the target action was obviously easier for a third party, for example when they were closer to a target object, or of lower social status, both of which were the case for participant C in (7).<sup>3</sup>

(7) CHSF2012\_08\_04S4\_1524500

```
▶ 1 a inu jabon ((in water, points at soap on shore))
          i -nu jabon
          1SG-ACC soap
          to me, soap
▷ 2 b jabon tya'kide apa ñaa
          jabon tyatyu-ki-de apa ñu-ya
          soap throw -do-IMP father 2SG-FOC
          throw the soap, son, you
  3 ((points at soap))
  4 c ((child throws soap to A))
```
Aside from the 22 cases of M-A without a predicate, all other cases with verbal material in M-A included a predicate of some kind.

<sup>3</sup>A note on the translation of *apa* in (7), line 2: in Cha'palaa it is common to use affectionate vocatives that are the equivalent of "mommy" and "daddy" for small male and female children; the result is difficult to translate to English.

### Simeon Floyd

### **3.3.2 Imperatives**

In contrast with the more context-dependent cases without predicates, most of the time speakers gave more information about the target action by producing a predicate (89% of recruitments with spoken M-A). Of these, as noted in Table 3, imperative forms were by far the most frequent type of predicates seen in recruiting moves. The imperative sentence type in Cha'palaa does not consist of a single construction, but instead features several options (see also Enfield, Chapter 6, §4.3.1). Cross-linguistically, imperative verb forms tend to be relatively short, frequently consisting of just a verb root or a root with a minimal marker (Khrakovskij 2001; Aikhenvald 2010). Cha'palaa fits this pattern; its two main imperative constructions are a bare verb root or a suffix, *-de*, seen in line 2 of (7) above. Table 4 summarizes the different imperative formats observed in the sample.


Table 4: The Cha'palaa imperative paradigm, singular and plural forms (optionally marked for plural), and percentages of each format within the total of imperatives in the sample (*n*=137).

The bare root option is shown in (8). In a few limited contexts, declaratives can also occur as bare verb roots, so the comparable imperative format relies to a small degree on context for disambiguation. Cha'palaa has a system of complex predicates in which multiple roots combine in single predicates, where one of the roots, usually one of a set of verb classifiers, occurs farthest to the right, and takes the finite morphology (Floyd 2014a; see also Dickinson 2000 for a description of a similar system in a related language, Tsafiki). In most cases finite predicates take at least one verbal morpheme, but one of the options for forming imperatives is to use just the verb root. In (8) participant A takes this option, telling B to look at a magazine she is passing to her.

3 Getting others to do things in the Cha'palaa language of Ecuador

(8) CHSF2011\_06\_24S3\_1304600

```
▶ 1 a lenke' kerake ((hands B magazine))
          len -ke-tu kera-ke
          read-do-SR see -do
          read this
```
▷ 2 b ((takes magazine and reads))

In addition to the two most frequent imperative formats shown above in (7), line 2, and (8), Cha'palaa has three further imperative markers, plus distinct plural forms of each. Table 4 shows each of the formats' frequencies relative to the other imperative options in the sample. The bare verb form sometimes occurs with a lengthened vowel and a related shift in primary stress to the right, which might be considered a kind of strengthener (see §3.4). The next most-frequent format is the general imperative *-de*, which conveys an on-record wish that the addressee do the target action, and a speaker-directed imperative *-ka* that conveys that the speaker is the indirect object and beneficiary. Then there are two hortative forms for groups that the speaker includes him or herself in, one for stronger, "command" types of recruiting moves (*-da*), and one for weaker, "suggestion" types of recruiting moves (*sa*). This final polite option did not occur in the sample, perhaps in connection with the maximally informal nature of the recordings, which may lead more formal, delicate types of recruiting practices to be infrequent.

When using an imperative, it is possible to mark a beneficiary of three-place predicates like 'give' with a full noun phrase, as in (9).

(9) CHSF2011\_02\_14S3\_2673050

▶ 1 a inu jayu kude aamama shipijcha ((reaches towards B)) i -nu jayu ku -de aamama shipijcha 1SG-ACC a.little give-IMP grandma 'madroña' give me a little madroña grandma

```
▷ 2 b ((turns towards A, begins passing fruit))
```
There is also a special imperative marker that is only compatible with firstperson beneficiaries, *-ka*, illustrated in (10), line 3.

(10) CHSF2012\_01\_20S6\_2952739

1 a vieja old lady 2 b aa

huh?

### Simeon Floyd

▶ 3 a inu chuwa manka' kuka junu jee ((finger and lip point)) i -nu chuwa ma -n -ka -tu ku -ka junu jee 1SG-ACC vine again-IMFV-grab-SR give-IMP1 there yes give me the string there hey

```
▷ 4 b ((brings string))
```
When the speaker is included as a participant in the target action along with the interlocutor, one of two different hortatives may be used. The first, *-da*, was the only one of the two attested in the sample, indicating that it is probably used more frequently in general in informal contexts. Extract 11 shows a case of this hortative, when one teenager attempts to recruit another to go fishing. The sequence was unsuccessful and was abandoned when A did not respond to B after line 4.

### (11) CHSF2011\_02\_14S3\_3479997


Outside of the sample, looking into the video corpus more broadly, it was possible to find examples of the second hortative, *-sa*. This marker is identical to a dependent clause marker for modal complements, and it is likely that the hortative use developed through processes of "insubordination" (Evans 2007; Floyd 2016; Evans & Watanabe 2016) when the dependent clause developed a conventionalized main clause usage, and became incorporated into the imperative paradigm. It alternates with *-da* as a more "mitigating" option. At present it is so integrated into the imperative system that it takes the plural marker that only combines with imperative forms, *-i*. Extract 12 shows an example of the plural form of *-sa*.

(12) CHSF2012\_01\_07S1\_137560

```
▶ 1 a jisai lalaa
          ji-sai lala-ya
          go-HORT2.PL 1PL -FOC
          let's go
```
### 3 Getting others to do things in the Cha'palaa language of Ecuador

▷ 2 b ((leaves house with A))

Note that, at a finer level of categorization, the social actions in (11) and (12) may be also analyzed as "proposals" (Couper-Kuhlen 2014; Stivers & Sidnell 2016). These fall within the phenomenon of recruitment, which broadly encompasses sequences in which A obtains B's assistance or collaboration in doing something for or with them (see Chapter 1, §4).

The imperative system is flexible with respect to grammatical norms, plural marking is available but optional, and the motivations of speakers for choosing one of the three second-person imperatives or one of the two hortatives will require further research to be fully determined. This first analysis shows that such grammatical flexibility provides diverse options for different interactional contingencies.

### **3.3.3 Interrogatives**

While imperatives usually function as unambiguous on-record recruiting moves, the other two main sentence types, interrogatives and declaratives provide ways for instigating a fitted response to the recruiting move without going explicitly on-record, and have been discussed with respect to indirectness in speech acts (Searle 1969). Interrogatives often inquire about preconditions for the request (Levinson 1983), potentially launching a "pre-sequence" (Schegloff 1980; 2007; Rossi 2015a), but conventional use of interrogative formats for pre-requesting can result in the "collapse" of the pre-sequence to the point that interrogatives can act as independent request formats (Levinson 1983; Fox 2015). This is the case with the format seen in line 1 of (13), which inquires about the availability of an object ('is your saw not there?'), but which ends up being taken as a request for the object.

(13) CHSF2012\_01\_21S3\_2615530

is (your) saw not there?

```
▶ 1 a ñuchi serruchu tsutyuu ((off camera, outside of house))
          ñu -chi serruchu tsu-tyu-u
          2SG-POSS saw lie-NEG-Q
          is your saw not there?
  2 b aa
          huh?
  3 a serruchu tsutyuu
          serruchu tsu-tyu-u
          saw lie-NEG-Q
```
### Simeon Floyd

```
4 b enku (.) tanami ibain (.)
          en -ku ta -na -mi i -bain
          here-LOC have-be.in.POS-DECL 1SG-also
          also here, I have it (.)
  5 jayaa finberaya
          jayu -ya fi -n -bera -ya
          little.bit-FOC eat-PFV-still-FOC
          it 'eats' a little (it saws decently)
  6 seruchu tii ((to C))
          serruchu ti -i
          saw say-Q
          did (he) say 'saw'?
  7 c mm
          yeah
  8 b ((goes to get saw, returns))
▷ 9 Ebe jee ((holds out saw))
          Ebe hey
          here Ebe
```
Going through the interaction above line-by-line helps to illustrate how a question about the presence of an object is treated by the participants as a request for the object. In line 1 A inquires about the saw, using the verb 'lie', which is the appropriate positional verb for elongated objects on flat surfaces. Possibly because A is standing outside the house and did not have B's full attention, B displays some trouble hearing line 1 and initiates repair in line 2, occasioning a full repetition in line 3 (typical for an "open" repair initiator Drew 1997; see Floyd 2015 for a description of the Cha'palaa repair system). In line 4, B answers the question, confirming that the saw does in fact exist. But this is not all he does; in line 5 he also gives some information about the status of the object with respect to its function ('it saws decently'), giving evidence that he understands that lines 1 and 3 are geared towards getting the saw. Interestingly, B has chosen to respond to A even though he appears not to be fully certain of the target action, as in line 5 he requests further confirmation from his wife C. After this, B proceeds to fulfill the request in 8, but at no point has A overtly asked to be given the saw.

Interrogative formats like that seen above in (13) are frequent in cases of requests for transfers of objects: compare 7% of interrogative recruiting moves in the total sample (*n*=13/192) with 21% for object transfers (*n*=9/42). Additionally, if the object is not visible, interrogatives are used in 47% of cases (*n*=8/17); this is to be expected, because these are canonical contexts in which a participant might check the preconditions for a request before making it, thus avoiding rejection on the grounds of a faulty presupposition that the object was available (see also Rossi, Chapter 5, §3.3.3).

In other situations, other interrogative formats can be used. In cases of requests

### 3 Getting others to do things in the Cha'palaa language of Ecuador

for the provision of services, for example, a speaker might ask a question about a target action to convey that they would like an addressee to do this action. An example of this was seen in (1) with the question 'did/do you sweep?'. The common feature of all the different types of interrogative formats is that, each in different ways, they use what is on the surface a request for information as a way to request an activity.

### **3.3.4 Declaratives**

Declaratives are another format for less direct or off-record requests. These work by introducing a proposition about some state of affairs, but with an implicit understanding that some action should be taken by an addressee. Relative to imperatives, this format allows speakers to avoid overtly selecting an addressee for the recruitment. In some cases the addressee may be obvious from the context – if A is gazing directly at B, for instance (see Lerner 2003), or if the interaction is dyadic – while in other cases participants might self-select and construe themselves as the addressee. One cross-linguistically common format for declarative recruiting moves is deontic statements about how things should be, or what things need to be done (see also Rossi, Chapter 5, §3.3.4; Zinken, Chapter 8, §3.3.2; Baranova, Chapter 9, §3.3.3). Extract 14 provides one such example of a deontic construction in Cha'palaa, which is formed with the combination of an infinitive verb and a copula (a very common deontic construction type in South America, Müller 2013). In this case a husband A and wife B were working together to nail in some boards to repair a wall, and when A makes a statement about the task that should be done, B responds by altering the way in which she is performing the task.

(14) CHSF201

▶ 1 a tu- tu'pushujuntsaa kanu juaa tu- tu'pu-shujunsta-ya ka -nu ju-ya nail nail -REL.CL -FOC grab-INF be-FOC (one) must grab the part that was nailed

▷ 2 b ((grabs and moves board))

Research on some European languages has shown similar usages of deontic constructions (Zinken & Ogiermann 2011; Couper-Kuhlen & Etelämäki 2015; Rossi & Zinken 2016), suggesting that this particular strategy may be cross-linguistically recurrent. Apart from the specific deontic constructions seen above, a further wide range of declarative construction types can function as recruiting moves in the right contexts. For example, (15) gives us a case of a pursuit

### Simeon Floyd

of a recruitment that was not fulfilled after the first attempt, which was an imperative: 'look for lice on me'. While a bit taboo in Western cultures, picking parasites of each other is an important social interactive practice among peoples from different parts of the world, including the Chachis of Ecuador, for whom it is considered an affectionate form of behavior most common among family members. In this case, however, when A tells her husband B to groom her in this way, he displays no uptake, and continues a parallel line of conversation, leading to a second attempt by B in line 3, this time in a declarative format.

### (15) CHSF2012\_08\_04S4\_1524500

```
▶ 1 a inu mu keraa ((sits with back towards B))
          i -nu mu kera-a
          1SG-ACC lice look-IMP
          look for lice on me
  2 b ((no uptake, 88.0 unrelated conversation))
▶ 3 a ñaa inu mu keetyunkayu mm mm ((scratches head))
          ñu-ya i -nu mu kee-tyu-nkayu mm mm
          2 -FOC 1SG-ACC louse see-NEG-EV mm mm
```

```
you aren't looking for lice on me, hey
```
Participant A had been sitting with her back to her husband, giving him access to her hair for over a minute when she makes a second attempt at recruitment (line 3). This time she uses a declarative format, using a negation construction to call attention to a state of affairs that is not currently the case ('you aren't looking for lice'). Similar "negative observations" have been shown to be a format for complaining (Schegloff 1988; Rossi 2018). In light of the first, more overt recruiting move in line 1, this statement can be taken as a request that B do the relevant action.

### **3.3.5 Other construction types**

In addition to specific verbal morphemes, there are specialized verbal constructions that can be resources for initiating recruitment. One good example of this is a benefactive construction using the verb 'give' as an auxiliary to indicate beneficiaries, a construction which appears in several other unrelated local languages and may be a product of areal convergence (see Bruil 2008 on Ecuadorian Spanish and Quechua). While this construction literally asks one to 'give' the action, when the verb 'give' is used with a second verb it means 'do it for someone's benefit'. An example can be seen in (16).

<sup>4</sup> b ((no uptake, continues unrelated conversation))

3 Getting others to do things in the Cha'palaa language of Ecuador

(16) CHSF2011\_02\_14S3\_1828314


In line 1 A asks B to cook plantain, but uses the 'give' construction to mean that a third party, the other family members present, will benefit (the additional comment about filming is a bit of "camera behavior" in which participants in the recording make reference to the recording equipment). This benefactive 'give' construction can occur with any of the major sentence types, and is not dedicated solely to recruitment, but when it occurs in M-A of recruitment sequences it has the effect of introducing beneficiaries through a conventionalized use of a ditransitive predicate to modify the argument structure.

# **3.4 Additional verbal elements**

In addition to the predicate and its core arguments, there are other aspects of turn design that are relevant for the format of the recruiting move. This section reviews several of these non-core elements.

### **3.4.1 Strengtheners and mitigators**

Some non-core elements can be considered strengtheners or mitigators with respect to how they upgrade or downgrade the recruitment in terms of its claimed urgency, importance, appropriateness, ease of accomplishment, and so on. One common strengthener in Cha'palaa is the word *jee*, which shares several functions: it is the main positive polarity token ('yes'), a vocative often used to secure attention ('hey'), and a strengthening element in recruitments. Usually these different functions can be easily distinguished from their context of use, but *jee* generally needs to combine with other elements like a verb to be able to specify a target action in a recruitment sequence, as in (17) where it combines with the imperative verb 'look.'

### Simeon Floyd

(17) CHSF2011\_06\_25S2\_1428820


Here A asked B to help remove some broken pieces of fiber during basket weaving, finishing the spoken part of her turn with *jee*; B then takes the proffered fiber and proceeds to fulfill by doing the task. In this position, using *jee* to help secure the attention of B can be seen as a strengthener, although it may occur in other contexts doing different things (for example, in Extract 9, line 8, *jee* occurs in the fulfillment of an object transfer request: 'here take this').

Another quite different format for strengthening consists of modulating the volume of the spoken elements of the recruiting move. Extract 18 provides a good example of this strategy in the context of a pursuit sequence.The initial recruiting move in line 1 concerns A telling his wife B to hold onto a string so he can tie it. In line 3 he produces a second recruiting move giving more information about the position he wanted her to hold ('on the tip'), implying that her first attempt at fulfillment had not been totally acceptable. Then in line 5 he repeats the format from line 3, consisting of a noun *kapa*, meaning the 'side' or 'tip' of an object, and a locative suffix *-nu*, but now produces it at higher volume.

(18) CHSF2012\_01\_20S6\_2509823

```
▶ 1 a tadi
          ta -di
          have-come.into.POS
          hold (this)
▷ 2 b ((begins to hold string))
▶ 3 a mm kapanu
          mm kapa -nu
          yeah side/tip-LOC
          yeah, on the tip
▷ 4 b ((begins to hold more firmly))
▶ 5 a KAPANU
          kapa -nu
          side/tip-LOC
          ON THE TIP!
▷ 6 b ((holds firmly so that A can cut string))
```
3 Getting others to do things in the Cha'palaa language of Ecuador

```
7 a enu main kake
       e -nu main ka -ke
       here-LOC one grab-do
       grab one here
```
It is easy to see the difference in the two pronunciations of *kapanu* from the waveform of the audio recording (Figure 5, in which participant A increased the volume of a repeated M-A to upgrade its format in a pursuit sequence.

Figure 5: Waveform of lines 3–5 of Extract 18, comparing the different volume levels of two successive productions of the same word.

It is interesting to note that the two strengthening strategies discussed here, the particle *jee* and increased volume, both have connections to strategies for securing the attention of an addressee, either through using a vocative like 'hey!' or by making the words more perceptually salient by amplifying them. If other languages show a similar link between securing attention and strengthening recruiting moves, this may turn out to be a recurring strategy that combines securing attention to a recruitment and pursuing its fulfillment.

In addition to strengthening recruiting moves, Cha'palaa speakers also used different formats for mitigating them, or downplaying the contingency of the recruitment. One of these is a minimizing strategy that uses the word *jayu*, or 'a little bit', in order to frame a recruitment as something small, insignificant, or easily accomplished. Work on politeness strategies has noted that words or morphemes about smallness are a cross-linguistically common strategy for managing face-threatening acts like recruiting moves (Brown & Levinson 1987). In (19) A tells his wife to 'make rice', but then adds the word *jayu*. The word order is relevant because Cha'palaa is in most cases verb-final and modifiers like *jayu* generally precede their nouns, suggesting that here it was added on as a late-stage calibration of the contingency of the recruitment.

Simeon Floyd

(19) CHSF2011\_02\_15S4\_6949122

```
▶ 1 a arosya kee jayu
          aros-ya ke-e jayu
          rice-FOC do-IMP a.little
          make rice, a little.
```

```
▶ 2 b ((gets out of hammock))
```
Here it appears that the quantity of rice was not really the issue, and that *jayu* has more to do with minimization of the imposition of the recruitment.

### **3.4.2 Explanations**

Another type of strategy that may be used for mitigation of recruitments is the provision of explanations (including accounts and similar) in the same turn as the core recruiting component, providing background information that presents the target action as more necessary, justified, or reasonable (see Parry 2013; Baranova & Dingemanse 2016). For example, in (20) speaker A tells speaker B to 'clean the baby's face', which is very specific about the target action and its beneficiary, and could be a complete recruiting move on its own. However, A also adds the phrase 'it is dirty', which provides motivation for the target action.

(20) CHSF2011\_02\_15S4\_499970

```
▶ 1 a nanu kajuru mankijtikee kuchinuu
          na -nu kajuru ma -n -kijti-ke-e kuchinu-ju
          small-ACC face again-IPFV-clean-do-IMP dirty -be
          clean the baby's face, it's dirty
▷ 2 b ((gets up, takes baby and starts washing face))
```
Extra elements like mitigators and strengtheners were relatively uncommon. Explanations were present in 14% of cases that included a verbal element (*n*=27/ 192), while just 4% included mitigators (*n*=7/192), and strengtheners occurred in 10% of cases (*n*=20/192).

# **4 Move B: The responding move**

Like M-A, M-B, the responding move, can be fully nonverbal, fully verbal, or a composite. But in this sequential position, speech and nonverbal behavior are subject to different conditions than in the initial position of most recruitment sequences. Since the cases in the sample all involved practical actions that could be accomplished or begun during the interaction, most cases included some kind of

### 3 Getting others to do things in the Cha'palaa language of Ecuador

relevant nonverbal behavior in the response. More than half of cases (*n*=105/205) included a clearly identifiable nonverbal response, either as the only response (*n*=81/205) or as part of a composite move with verbal elements (*n*=24/2015). In a number of cases it is impossible to see whether there is a visual element of a response due to participant B moving off camera or behind another person or an object (*n*=79/205). In a smaller number of cases, responses only included verbal elements (*n*=21/205); since these cases include no practical action, they partly correlate with cases of rejection, while cases including nonverbal, practical actions tend to be cases of fulfillment.

Table 5: Fulfillment, rejection, and other response types in the Cha'palaa sample.


Table 5 shows the breakdown of different types of responses in M-B of the recruitment sequences of the sample. The data show an overall predominance of fulfillment, but also a number of other options for response, including rejection, which will be discussed below in §4.3. First, response formats will be addressed in §4.1 and §4.2.

# **4.1 Fully nonverbal responses**

Well over half (64%) of the responses in which B is visible in the video consisted of nonverbal elements only (*n*=81/126); in most cases this corresponded to the accomplishment or beginning of the target action (see also Rauniomaa & Keisanen 2012). Some target recruited actions could be accomplished quickly (e.g. A: 'to me, the soap' B: passes the soap), while for others B could only respond by beginning some activity that A can understand as projecting the completion of the target action (e.g. A: 'cook a little rice' B: begins cooking rice). Extract 21 gives an example of the former, a case of a request for the alteration of a trajectory of ongoing activity that was fulfilled immediately after M-A. Participant B had been holding a baby but was not devoting full attention and the baby was beginning

### Simeon Floyd

to slip out of his grasp. Participant A, noticing this, prompts B to hold the baby more firmly. B's change in behavior is immediate, includes no verbal elements, and is treated as a satisfactory fulfillment, in that it is no longer pursued by A beyond line 2.

```
(21) CHSF2011_01_11S3_4728040
```


▷ 2 b ((holds baby more carefully, baby stops slipping))

Also notable is the repetition of the recruitment predicate. Stivers (2004) observes that this type of repetition can be associated with urgency, and a similar connection can be made here: in the sample, several repetitions occur in recruitments dealing with alteration of trajectory in already-ongoing activities, which in this sense are more urgent than requests for services or objects, since the potential negative effects of not fulfilling the recruitment may be mounting while M-A is being produced (see also Extract 3, above, for a similar example of repetition in a case of urgency).

# **4.2 Verbal elements of responses**

In cases in which a spoken element was part of the response, some of these were rejections, especially when only a verbal element was present, as these cases included no practical activity fulfilling the request. However, verbal elements could accomplish other things as well in the sequential position of M-B. For example, in (9) above, participant B initiated repair with an interjection after the initial M-A (see Floyd 2015 on other-initiated repair in Cha'palaa). Another function that verbal elements of responses can accomplish is to manage the temporal contingencies of the sequence. For example, in (22), participant A asks participant B to take a basket, which A is holding out, but B is unable to immediately comply, so she makes it known that she intends to do the target action soon with the utterance 'wait a little'.

(22) CHSF2011\_06\_25S2\_3468149

```
▶ 1 a aanku ka' tsuude ((holding out basket))
          aanku ka -tu tsure-de
          there grab-SR lie -CAUS-IMP
          there get it and set it down
```
3 Getting others to do things in the Cha'palaa language of Ecuador

```
▷ 2 b kai keedi ((continues previous activities))
          kayi kera-di
          little see -come.into.POS
          wait a little
  3 (1.5)
```
▷ 4 b ((takes and moves basket))

In Figure 6 A is holding out the basket for B to take, but B's hands are busy (she appears to be rubbing saliva on a dry area of her arm). Often if there is no immediate nonverbal response, speakers pursue with further recruiting turns. However, because B conveys to A she will address the target action shortly, A simply waits with the basket outstretched for a couple of seconds; in Figure 6, B takes the basket and sets it down as A requested.

Figure 6: Participant A holds out basket for participant B and tells her to 'get it and set it down' (line 1).

# **4.3 Rejections and other non-fulfilling responses**

Most examples shown in previous sections have been fulfillments. Fulfillments by definition will fall into the nonverbal or composite categories, since they must include a practical nonverbal action or its beginning (although some nonverbal actions did not count as fulfillment). On the other hand, fully verbal responses tended to be rejections, as generally rejections require some on-record statement which, while possible to convey visually, tends to be spoken (*n*=7/8). Outright rejections were rare in Cha'palaa, and only 4% of total cases featured rejection as the response (*n*=8/205), compared to 47% which included fulfillment (*n*=97/205). The fulfillment rate is even higher (55%, *n*=69/125) when considering only sequence-final cases (Table 5), reflecting how unsuccessful first attempts

### Simeon Floyd

Figure 7: Participant B takes the basket and sets it down (line 4), after having delayed a moment, saying 'wait a little' (line 2).

can be pursued for eventual fulfillment. Another type of spoken response to unsuccessful first attempts is be repair initiation, which accounted for 9% (*n*=19/205) of total responses (predictably, this rate was much lower in sequence-final cases). Additional options included ignoring M-A, or "other" responses like delegating to a third party (see Extract 7), giving information that A can use to resolve the problem him/herself, making a counter-proposal, or pursuing some unrelated sequence. These additional types of responses were generally more frequent than overt rejection, and so it seems that Cha'palaa speakers tend to opt for less explicit ways of avoiding the uptake of recruitments besides overt rejection. In addition, the types of rejections that were seen were not on-record refusals (a flat-out 'no') but tended to take other forms. In rare cases rejections could be fully nonverbal, such as in (23), where a nonverbal recruiting move – A reaching out for a slingshot – is not responded to with a transfer of the object by B, who instead pulls the object away out of reach (Figure 8).


When there is spoken material in a rejection, most often it can be classified as an account or explanation for why B is unwilling or unable to comply (87% of rejections, *n*=7/8). Extract 24 is a good example of rejection through explanation. Participants A, B, and other friends are washing clothes together, but B is getting ready to leave while A still has more to wash. A suggests that B accompany her by

### 3 Getting others to do things in the Cha'palaa language of Ecuador

Figure 8: Participant A (in center in shirt with stripe) reaches for the slingshot (line 1) as participant B (right) pulls it back out of reach (line 2).

doing some more washing, using a declarative format ('you'll wash'). However, B has no more clothes to wash, so she offers this state of affairs as an explanation for why she cannot fulfill the request.

(24) CHSF\_2012\_08\_04S4\_1193345

```
▶ 1 a tsaaren manpipunaa manbije
          tsaa-ren man -pi -pu -nu -ya man -bije
          SEM -PRECIS again-water-put-INF-FOC again-time
          so (you) will wash one more time
▷ 2 b naaketaa manpipunu nejtaa deiñu
          naa-ke-tu-ya man -pi -pu -nu nejtu -ya de -i -ñu
          how-do-SR-FOC again-water-put-INF because-FOC CMPL-become-DR
          how can (I) wash since (it) is already finished?
▷ 3 nejtaa yumaa deiñu
          netju -ya yumaa de -i -ñu
          because-FOC now CMPL-become-DR
          since it is now finished
```
The rejection turn includes a main clause calling into question B's ability to perform the target action ('how can I wash?') as well as a clause providing an explanation, including the word *nejtu* which can be translated as 'because' or 'since' ('since it is now finished'). Requesters generally accepted such explanations amicably and did not insist, and such rejections do not seem particularly conflictive or strongly face-threatening. Here participant B is smiling as she rejects the request (Figure 9).

One final point about rejection formats is that there are relevant connections between the formats seen in the recruiting move and the formats seen in the

### Simeon Floyd

Figure 9: Participant A (center, foreground) asks participant B (right) if she will continue to accompany her washing clothes (line 1). Participant B rejects (line 2) and offers an explanation (line 3).

responding move. Particular first pair parts make relevant the provision of "typeconforming" responses (Raymond 2000; 2003). For example, the recruiting move in (25) is in the format of an interrogative clause inquiring about the existence of a target object, which is a common format for requesting objects that are not visible as discussed in §3.3.3. The response in line 2 both answers the question with the appropriate format and at the same time accomplishes rejection by appealing to an explanation citing the lack of the target object.

### (25) CHSF2011\_01\_11S2\_249991


Since B has done due diligence here by checking the basket to see if there are any limes, she does not end up being held accountable for non-compliance with the recruitment (see also Rossi 2015a and Chapter 5, §4.2.2). In many sequences which qualify as rejections speakers are able to maintain their affiliative stance, suggesting that people avoid the most fraught exchanges altogether when possible. For example, in (25) B rejects the recruitment with an affective, diminutive term 'little husband'. In general, the high fulfillment rate and low rejection rate indicate an orientation to affiliation in such sequences in Cha'palaa interaction.

3 Getting others to do things in the Cha'palaa language of Ecuador

# **5 Acknowledgment in third position**

After M-A and M-B, recruitment sequences may optionally include a move in third position by A that acknowledges the fulfillment of the recruitment. While in principle speakers of any language can make a positive assessment in this position, in some languages there are conventionalized resources that function as this type of "sequence-closing third" (Schegloff 2007) like the English *thank you*. Cha'palaa speakers are familiar with such linguistic resources through contact with Spanish, which has the format *gracias*, but when asked if there is a Cha'palaa equivalent, they end up puzzled and unable to think of anything. This illustrates how practices like saying 'thank you' can be highly variable across different populations (see Floyd, Rossi, et al. 2018). Other research on acknowledgments has reached similar conclusions, like Apte's (1974) observation that while thanking is relatively unmarked in American English in most contexts, in South Asia it is very marked except in a few specific contexts.<sup>4</sup> In Cha'palaa it appears that acknowledgment is not only marked, but that there is no conventionalized format for thanking in the language at all.

In Cha'palaa recruitment sequences, speakers tend to either close the sequence or continue some other conversational trajectory after M-B, where in other languages third-position acknowledgment practices are sometimes observed. The video corpus was collected in highly informal contexts, so acknowledgments might be expected to be infrequent for mundane requests among speakers who are highly familiar to each other (Floyd, Rossi, et al. 2018). However, even in these contexts speakers of other languages showed some evidence of orientation to this kind of "face work" (Goffman 1955; Brown & Levinson 1987), while speakers of Cha'palaa did not. Along with its preference for direct imperative formats over less direct interrogative and declarative forms, this suggests that some of the typical practices associated with politeness in English and many other languages are quite different among speakers of Cha'palaa.

# **6 Social asymmetries**

Differences in social status among people in interaction are highly significant for how recruitment sequences play out, but these are more difficult to charac-

<sup>4</sup>A number of other studies discuss norms of thanking in different languages and cultures (Eisenstein & Bodman 1986; Pedersen 2010; Ohashi 2013). Many studies are concerned with second language learning (Hinkel 1994; Intachakra 2004; Özdemir & Rezvani 2010; Cui 2012; Farashaiyan & Hua 2012, among others).

### Simeon Floyd

terize analytically than, for example, the morphosyntactic formats seen in those sequences. The most reliable method for assessing how social asymmetry may be relevant for a given society in the mundane, everyday contexts considered in this study is long-term ethnographic observation and participation in the community. Based on my experience in Chachi communities over a period of about 8 years, including one full year spending two weeks per month in the field (2008–2009), it is possible to generalize that the some of the most relevant types of social asymmetries are based on a combination of age, gender, and kinship relations. Grounded on this information, each dyad was classified as symmetrical or asymmetrical. While it is true that, at least to some extent, status is locally negotiated in every interaction, in practice rights and duties around recurrent household activities remain relatively stable from instance to instance (and it seems difficult for a society to function without a relatively stable distribution of rights and duties). The dyad classifications apply only for the comparable village and household settings of the corpus, involving recurrent activities like cooking and cleaning, but this relative stability is partially contingent on context and is not always stable for every dyad in every context. However, they are stable enough in these contexts to see some trends.

The Chachi people have a system of traditional law that governs questions of morality, based around strong gender roles and normative family structure, and punishing transgressions like adultery or marriage outside the ethnicity (Barrett 1925; Altschuler 1964; Floyd 2010). The male and female roles in the family are well-defined, and men and women are responsible for different tasks. Men usually participate in hunting and fishing, some agriculture, logging and canoemaking, while women are in charge of household work like cooking, cleaning and childcare, in addition to some agricultural tasks and handcrafts like basket weaving. Most of the mundane activities that made up the target actions in the sample of recruitment cases from the Cha'palaa corpus were the types of household activities that many Chachis consider to be women's responsibilities. For that reason, in most cases when men directed recruiting moves at women, typically men telling their wives to do things, such cases were classified as highstatus recruitments involving lower-status individuals (A>B). Additionally, children are accountable for a number of household responsibilities such as carrying buckets of water from the river and assisting adults in their tasks. While cases involving young children were excluded from the sample as described in §1.1, adolescents usually continue to be accountable for such tasks until marriage, so cases of adult family members like parents, grandparents, or aunts and uncles initiating recruitments with adolescents and young adults were also classified as

### 3 Getting others to do things in the Cha'palaa language of Ecuador

A>B. Cases with the inverse situation, when adolescents told their elders to do things or wives told their husbands to do things, were classified as low-status individuals initiating recruitment of higher-status individuals (A<B). All other cases among adults with no relevant family relationships were classified as symmetrical (A=B). This qualitative classification should be thought of as a flexible, pragmatic approach that takes into account both more stable aspects of social roles but is also attuned to situational factors for this data set.

Table 6: Relative frequencies of dyads by type of social (a)symmetry (*n*=201/205 cases). In four cases there was not enough information to classify the dyad.


The high rate of fulfillment versus rejection in the Cha'palaa recruitments sample may be in part accounted for because individuals who may not be socially entitled to make certain requests can simply avoid them as a way to avoid potential rejection. In their model of politeness, Brown & Levinson (1987: 69–74) proposed this option as preferable in cases in which the potential costs are too extreme to attempt the face-threatening act. While more than half of the recruitments in the sample were between individuals classified as socially symmetrical, there were also over four times more A>B recruitments than A<B recruitments. Based on this, it appears that Cha'palaa speakers are more likely to initiate recruitment of individuals with similar or lower social status than of individuals with higher social status (see also Enfield, Chapter 6, §6; Baranova, Chapter 9, §6; Dingemanse, Chapter 10, §5.2).

There is some evidence that the format of the recruiting move is also sensitive to social asymmetries in that the more direct formats like imperatives and no predicate are more frequent as social entitlement increases (see also Heinemann 2006, Curl & Drew 2008, and Craven & Potter 2010 on the concept of entitlement). Imperatives are "direct" in a straightforward sense, but no-predicate recruiting moves can also be considered very direct in that, like imperatives, they are usually on-record and understood as explicitly requesting a target action by way of naming objects, recipients, places and so on. If we compare direct formats with ostensibly off-record formats like interrogatives and declaratives, we can consider

### Simeon Floyd

this as a measure of directness. Table 7 compares percentages of imperative and no-predicate recruiting moves for the three dyad types (A>B, A<B, A=B).

Table 7: Percentage of direct (imperative + no predicate) formats in Move A by dyad type.


Across all cases, the rate of direct formats was approximately 78%. In cases of recruitments initiated by high-status individuals with low-status individuals, this rate rises to 84%, with a particularly high rate of no-predicate recruiting moves. However, in cases of recruitments initiated by lower-status individuals with higher-status individuals, the rate of direct formats falls to 67%. Among equal-status individuals, the rate is between these two values, at 75%. These results illustrate that the relative status of recruiter and recruitee can affect both the base rate of recruitments (Table 6) and the directness of the format selected (Table 7). Lower-status individuals are less likely to begin recruitment sequences, and more likely to use less direct strategies when they do. Higher-status individuals are more likely to begin recruitment sequences, and more likely to use more direct strategies. In this social context, this means that male heads of households initiate more recruitments, and women and young people are more often in the position to respond and comply, a finding that resonates with the observed social roles in the community.

# **7 Discussion**

This chapter has reviewed the particular ways that speakers of Cha'palaa address the common human problem of coordinating cooperative behaviors and joint actions in light of individual concerns about being imposed on by or imposing on others. Cha'palaa speakers draw on a wide range of spoken and nonverbal resources in order to accomplish this, and calibrate the formats they use in social interaction with respect to different contingencies. In about half (*n*=97/205) of Cha'palaa recruitment sequences the target action was accomplished, while in only a small percentage was there overt rejection. In another considerable portion of sequences the recruitment received no uptake but was abandoned and

### 3 Getting others to do things in the Cha'palaa language of Ecuador

not pursued. In some cases the risk of overt rejection may be too high a price to pay for pursuing the target action. The overall tendency in the sample appears to be to avoid rejection when possible.

The avoidance of rejection can be interpreted both as prosocial and as a reflection of social asymmetry. Chachi culture has been resilient over the centuries in part due to strict enforcement of traditional laws, but these laws are based on rigid norms concerning social roles, and there are strong expectations about the appropriate responsibilities for daily tasks linked to gender roles and age grades. On the one hand, the high rate of successful recruitments shows that Cha'palaa speakers are highly affiliative and cooperative. On the other hand, cases of disaffiliative rejections may be low in part because people "know their place" and do not initiate recruitment sequences at all when their social rights to do so are questionable (see Floyd 2017). Sequences in which lower-status individuals requested actions of higher-status individuals were indeed the least frequent in the sample, while higher-status individuals were not so restrained.

Many of the practices and tendencies described for Cha'palaa resemble those seen in other languages described in the literature cited above, and in this volume. However, in other ways Cha'palaa is distinct, including the grammatical forms employed (e.g. the large imperative paradigm), the types of target action requested (e.g. tasks involved in traditional basket weaving), and the cultural rationales behind the reasons and explanations offered as part of recruiting and responding moves. A lack of acknowledgment practices and a low frequency of indirect formats appears to place Cha'palaa on the low end of a cross-linguistic politeness scale. Even so, perhaps a better interpretation is that Cha'palaa speakers do their face-work by other means, as recruitments are mostly successful and face-threatening conflict is rare. Whether viewed as more prosocial or more hierarchical, the Cha'palaa recruitment system reflects deep social cohesion and interconnectedness that allows for individuals to instigate actions that go beyond their own lack of ability or willingness to act, and as such it plays an important role in Chachi society.

### Simeon Floyd

# **Abbreviations**


# **Acknowledgments**

This research was supported by European Research Council grant 240853 for the project *Human Sociality and Systems of Language Use* (HSSLU) directed by Nick Enfield, with additional support from the Language and Cognition Department directed by Steve Levinson at the Max Planck Institute for Psycholinguistics. Special thanks goes to members of the Chachi communities of Tsejpi, Zapallo Grande, Santa María, and all of the Chachi people who participated in the recordings and in the analysis, particularly Johnny Pianchiche who transcribed much of the corpus. Thanks also to my collaborators in the Recruitments Project: Julija Baranova, Joe Blythe, Mark Dingemanse, Paul Drew, Nick Enfield, Kobin Kendrick, Giovanni Rossi, and Jörg Zinken, and to the reviewers, for their comments and discussions that improved this work. Any errors are my own.

# **References**


3 Getting others to do things in the Cha'palaa language of Ecuador


### Simeon Floyd


3 Getting others to do things in the Cha'palaa language of Ecuador


### Simeon Floyd


3 Getting others to do things in the Cha'palaa language of Ecuador


Simeon Floyd

Vittadello, Alberto. 1988. *Cha'palaachi = El idioma cayapa*. Guayaquil: Pontificia Universidad Católica del Ecuador, Museos del Banco Central del Ecuador. Zinken, Jörg & Eva Ogiermann. 2011. How to propose an action as objectively necessary: The case of Polish *trzeba x* ("one needs to x"). *Research on Language & Social Interaction* 44. 263–287. DOI:10.1080/08351813.2011.591900

# **Chapter 4**

# **Recruitment in English: A quantitative study**

# Kobin H. Kendrick

Department of Language and Linguistic Science, University of York

This chapter describes the resources that speakers of English use when recruiting assistance from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in English, and reports languagespecific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of English with detailed attention to the properties of sequential structure in conversational interaction.

# **1 Introduction**

The recruitment of assistance is a basic social organizational problem for which participants in interaction have practiced solutions (Kendrick & Drew 2016; Chapter 1 of this volume). In our daily lives we carry out countless mundane courses of action: we may reach out and pick up a pen from a table, connect a power supply to a computer, turn the page of a book, put a dirty dish in the sink. For the most part, we execute these courses of action individually, whether alone or in the company of family, friends, or colleagues. If a trouble emerges – the pen is too far to reach, the power supply is nowhere to be found – we resolve the trouble

Kobin H. Kendrick. Recruitment in English: A quantitative study. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 93–146. Berlin: Language Science Press. DOI:10.5281/zenodo.4018376

### Kobin H. Kendrick

on our own as well (Kendrick 2017). But in the presence of others, a trouble in the realization of a course of action is a public event, and its resolution may become interactional achievement, not an individual one. Someone may see us visibly searching the environment (Drew & Kendrick 2018) or hear our imprecations as signs of trouble and therefore offer their assistance (Kendrick & Drew 2016). We need not, however, wait for those around us to take notice and volunteer to help. Using the resources of language and the body, we can agentively solicit solutions from others to practical problems that emerge in the course of our activities. We may use a gesture to point to a box of biscuits so that someone will hand it to us, ask someone to locate a bag that we cannot find, or direct someone to move over to make room for us on the couch. However someone comes to perform assisting actions such as these, whether occasioned by a trouble or solicited by a request, we will have in effect *recruited* them to give or offer assistance. This chapter presents a quantitative study of some such recruitment phenomena in English, focusing primarily on requests, as observed in a corpus of video recordings of everyday social interaction made in the US and UK.

# **1.1 Recruitment: initial specimens**

We will begin with a set of cases that outline, in broad strokes, the general domain of recruitment. The first is a case in which we observe an opportunity for recruitment, but in which no recruitment occurs. The extract comes from an interaction between a group of friends as they prepare a barbecue in a public park. Just prior to the extract, Alison, the woman in the white shirt in Figure 1, has been playing with a dog on the grass behind the picnic table. We then see her walk towards the table, stop, direct her gaze toward it, furrow her brow – a facial gesture obscured in Figure 1 to protect her identity – and place her hand on her chin, in an elaborated form of a "thinking face" (Goodwin & Goodwin 1986). She holds this complex of gestures, virtually motionless, for approximately 1.4 seconds as the other participants talk about a local concert series and prepare the meal. She then turns her head slowly to the right, a movement which takes approximately 0.6 seconds.

```
(1) BBQ 08:25
1 kim I used to work concerts in the park, in fuckin'
2 ( ).
3 (0.2)
4 kim beer garden. +[awesome.
5 don +[ah:
```
4 Recruitment in English: A quantitative study

Figure 1: Alison (white shirt) walks towards the table, stops, furrows her brow and places her hand on her chin, making a "thinking face". She then reaches out, picks up a lighter, walks to the front of the table, and sets it down.

Alison's visible bodily actions can be seen as a display of puzzlement, though the crux of the puzzle is initially obscure. As she turns her head, we come to see her actions as a visual search of the environment (Drew & Kendrick 2018) –

### Kobin H. Kendrick

she's evidently looking for something. The visible bodily resources that she uses to conduct the search, her facial and manual gestures, her head movement, allow it to be recognizable as such – she's not only looking for something; she's *doing* looking for something (Sacks 1984). After a search of approximately 2.0 seconds, she apparently spots the sought after object, a lighter, and announces the end of her search with *here it is* (line 11). This announcement, like her gestures, orients to the accountability of her actions (Garfinkel 1967), even though, as we can see in Figure 1*a*, only the eye of the camera is on her as she conducts her search.<sup>1</sup> She then reaches out and picks up the lighter (see Figure 1*b*), walks around to the front of the table, and sets it down (see Figure 1*c*) as she marks the completion of the course of action with *okay* (line 21). The lighter is later used to light the coals in the grill.

In this case we can see an opportunity for recruitment emerge, and then pass. Alison begins a course of action, encounters a trouble that disrupts its progressive realization, and makes this publicly available through her visible bodily actions. As we will see, explicit displays of trouble such as this can recruit the assistance of co-participants. But the opportunity here is twofold: just as Alison's visible bodily actions could have recruited a co-participant, so too could Alison have used resources of language to do so explicitly. Consider the following case, in which Kimmy searches for a paper bag and then asks her co-participants for its location. Paper torn from the bag had been used by Kimmy as kindling and will be used again to light the coals in the grill.

```
(2) BBQ 14:27
```

```
1 (3.7) ((Kimmy visibly searches as she walks
 2 around the table, see fig.2a-b))
▶ 3 kim where the fuck is my little #fire starting bag.
 4 ◩ #fig.2b
 5 ca? (°° °°) -----
 6 ali ( ) (4.3)
 7 ca? (°° + °°) -----
 8 ali +gazes down-->
 9 kim +fire starting# bag.=is that i*t?
▷ 10 ali +gestures with arm-->
 11 ◩ #fig.2c
 12 kim *reaches out-->>
```
<sup>1</sup>The individual frames within the figures in the chapter are designated as *a*, *b*, *c*, etc. from left to right.

### 4 Recruitment in English: A quantitative study

Figure 2: Kimmy (red pants) visibly searches around the table for 3.7 seconds (*a*) before she asks for assistance. After 4.3 seconds in which Kimmy continues her search, Alison gazes down into a bag to her right and gestures toward it with her arm. Kimmy reaches into the bag and retrieves the sought after object.

After she visibly searches near the grill and around the table for approximately 3.7 seconds (see Figure 2*a,b*), she asks her co-participants to locate the bag for her.<sup>2</sup> As Kimmy rounds the table, Alison, who has been involved in a quiet conversation with another participant, looks down at a paper bag to her right (line 8), gestures towards it with her hand (while holding a bunch of scallions, see Figure 2*c*). Kimmy then reaches out and takes the bag (see Figure 2*d*).

In contrast to (1), in which Alison encountered a trouble in the realization of a course of action and resolved the trouble on her own, in this case Kimmy encounters trouble, performs a remedial action that does not resolve it, and then recruits a co-participant to assist her, using linguistic resources to do so. The recruiting action, as we will call it, is an interrogative question about the location of an object. It explicitly and accountably asks the co-participants to locate the object and thereby to assist Kimmy in her search.

The diversity of linguistic and embodied practices that participants use to explicitly and accountably recruit one another to facilitate practical courses of action will be a major theme throughout this chapter (see §4). But the boundaries of recruitment are not so narrowly defined. Subtle visible bodily actions, through which a trouble becomes publicly recognizable, can recruit others to assist even when these actions are not, in the first instance, accountable as requests for assistance or other forms of solicitation. The following case, which comes from an

<sup>2</sup>The insertion of *the fuck* into the construction of the turn formulates this not merely as an inquiry, but also as a complaint; someone has moved *her* bag.

### Kobin H. Kendrick

interaction among a group of students in a common area of a university building, demonstrates this. Here Mark, the man in the patterned shirt in Figure 3, can be seen to encounter some difficulty as he looks across the table at a picture in a book held by Rachael. Rachael then holds the book up for him to see.

```
(3) RCE22a 23:15
 1 rac god that looks rude. ((about a picture in a book))
 2 (1.3)‡#(0.5)
 3 con ‡leans over and gazes at book-->
 4 ◩ #fig.3a
 5 con oh wow. .h heh
 6 +(0.8)
 7 mar +leans forward and gazes at book-->
 8 con that really do(h)es(hh)
 9 (0.4)+(0.6)*#
▶ 10 mar -->+tilts head to side-->
 11 rac *gazes at Ben-->
 12 ◩ #fig.3b
 13 (0.4)*(0.8)
▷ 14 rac -->*.....holds book up-->>
 15 mar what exactly is happening+# [in this.+
 16 mar +untilts head+
 17 ◩ #fig.3c
 18 rac [↑I don't know.
```
Figure 3: Mark leans forward and looks at Rachael's book (*a*). Rachael gazes up at Mark after he tilts his head to the side (*b*). She then holds the book up for him to see (*c*).

An assessment of the picture begins the sequence (line 1) and draws the attention of Connor, seated to Rachael's right (Figure 3*a*). Connor reacts to the picture

### 4 Recruitment in English: A quantitative study

with surprise (line 5), drawing the attention of Mark, who then leans forward and gazes at the picture from across the table (line 7). Mark holds this position for approximately 0.5 seconds and then torques his head back and to the side (cf. Schegloff 1998), such that the orientation of his head comes to approximate the orientation of the book (see Figure 3*b*). The torque of Mark's head makes public a minor trouble, namely that from his perspective, seated on the other side of the table, the picture appears upside-down and would therefore be difficult to see. Mark's head movement attracts Rachael's gaze (line 11), at which point she would be able to see his head in an unstable position and his gaze directed to the picture. Shortly thereafter she lifts the book and holds it up for Mark to see (line 14, see Figure 3*c*) and thereby resolves the trouble.

In contrast to (2), in which Kimmy employed a linguistic practice to recruit a co-participant to assist her explicitly and accountably, in this case Mark encounters a trouble in the realization of a course of action, performs a remedial action to resolve the trouble on his own, and thereby recruits Rachael's assistance. His visible bodily action exposes the trouble, making it public, and thereby provides an occasion for Rachael to assist him, voluntarily. The action in effect recruits Rachael, even though in the first instance it is recognizable and accountable as an action taken by Mark to resolve the trouble independently, without assistance.

# **1.2 The anatomy of recruitment**

With these cases in mind, we can now characterize recruitment and the interactional environment in which it occurs in more general terms. In each case, a course of action performed by an individual is impeded or disrupted, for example, by the lack of a necessary object (Extracts 1 and 2) or constraints on the interactional space (Extract 3). A set of methods exists with which participants can resolve such troubles, either individually, via self-remediation (Extract 1), or interactionally, via recruitment (Extracts 2 and 3). The nature of these methods and their organization is a central concern of research on recruitment. The methods are organized along a continuum and include requests for assistance; reports of troubles, difficulties, or needs; trouble alerts; embodied displays of trouble; and the projection and anticipation of troubles before they occur (Kendrick & Drew 2016).

A basic distinction can be made between methods that create a normative *obligation* for assistance by Other (e.g. the request in Extract 2) and those that create a systematic *opportunity* for such assistance to be given or offered voluntarily (e.g. the embodied display of trouble in Extract 3) (Kendrick & Drew 2014; 2016). This distinction also concerns who generates the possible solution to the trouble.

### Kobin H. Kendrick

With a request for assistance, Self generates a solution for Other to implement (e.g. Self identifies the object that will resolve her own trouble in Extract 2). In contrast, with forms of voluntary assistance, it is Other who generates the solution and either implements it directly or offers to do so (e.g. holding up the book in Extract 3). Recruitment thus encompasses the initiation of assistance by Self and Other as alternatives methods for the resolution of troubles (Kendrick & Drew 2016).

The methods for recruitment include not only those implemented through language (e.g. the verbal request in Extract 2) but also those implemented through visible bodily action (e.g. the visible searches of the environment in Extracts 1 and 2, and the torque of the head in Extract 3). Visible bodily actions that display difficulty, discomfort, or exertion, for example, create systematic opportunities for Others to give or offer assistance and thus constitute methods of recruitment (Kendrick & Drew 2016). Such visible bodily actions are commonly, though not exclusively, forms of self-remediation, that is, actions produced by Self to resolve troubles independently (e.g. the visible search in Extract 1). Remedial actions by Self commonly precede other methods of recruitment (e.g. a visible search precedes the request in Extract 2), which together with other evidence suggests that self-remediation is a preferred alternative in the organization of assistance in interaction (Kendrick 2017).

# **1.3 The present study**

This chapter reports on a quantitative study of some recruitment phenomena in English, as observed in a corpus of video recordings of everyday social interaction in the US and UK. As a contribution to a cross-linguistic comparison (Floyd et al. 2014), the study employs an operational definition of recruitment and examines cases along specific dimensions set out by a coding scheme (see Chapter 2). The study therefore focuses primarily, though not exclusively, on requests as "moves" – a term used in this volume, after Goffman (1969), for social actions – that recruit others to assist. It does not consider the full continuum of methods for recruitment identified by Kendrick & Drew (2016). The quantitative analyses presented in this chapter are descriptive in nature, reporting the relative frequencies and proportions of various coding categories. Inferential statistics are reported in the cross-linguistic comparative studies (e.g. Floyd et al. 2018).

The chapter is organized as follows. After a discussion of the corpus and collection (§2) and the basic structure of recruitment sequences (§3), the analysis considers the visible bodily actions and grammatical formats that participants use to construct recruiting moves (§4) and then turns to the ways in which par-

### 4 Recruitment in English: A quantitative study

ticipants respond (§5). The chapter concludes with a discussion of the operational definition of recruitment employed in the present study and the concept articulated by Kendrick & Drew (2016).

# **2 The corpus and collection**

The data for the study came from a corpus of 21 video recordings of social interactions between speakers of English in the US and UK with a total duration of 11 hours and 53 minutes. The interactions involved various activities, such as preparing a barbecue in a public park, eating a meal with friends, and playing a board game, as well as ordinary conversation. Interactions between children and caregivers were not included in the study. The video recordings came from a number of sources: (i) a set of recordings made by Giovanni Rossi in 2011; (ii) the Language and Social Interaction Archive (2014) by Leah Wingard; and (iii) a recording of a game of Monopoly by Heidi Kevoe-Feldman. Informed consent was obtained from all participants. In addition, two video recordings widely used in conversation-analytic research, "Chicken Dinner" and "Virginia", were also included. A total of 211 recruitment sequences were identified, using the criteria described in Chapter 2. The majority of recruiting moves in the resulting dataset were produced by speakers of a North American variety of English, whether recorded in the US or UK (*n*=149), with the remainder produced by speakers of British varieties (*n*=59) or non-native speakers (*n*=3). The transcripts employ conventions developed by Jefferson (2004) for the transcription of talk and those developed by Mondada (2014) for the transcription of visible bodily actions. A description of the multimodal transcriptions conventions can be found at the end of this chapter.

# **3 The structure of recruitment sequences**

# **3.1 Minimal sequences**

A minimal recruitment sequence includes two actions, referred to in the comparative study as *moves*, a recruiting move and a responding move. In the transcripts, ▶ and ▷ designate the recruiting and responding moves, respectively. In the following extract, for example, as Vivian and Shane sit together on a couch, Vivian tells Shane to move over.

### Kobin H. Kendrick

```
(4) Chicken Dinner 00:05 (simplified)
▶ 1 viv move over *a li:ttle,* can you?
▷ 2 sha *..........*moves over-->
  3 (.)*
  4 -->*
  5 sha yep.
  6 viv thanks.
```
Even before Vivian's turn has come to possible completion, Shane begins to comply with the request, as indicated by the preparation phase of his movement in the transcript. Upon the completion of his movement, Shane responds verbally with an answer to Vivian's tag question, at which point Vivian closes the sequence with a non-obligatory third move, a display of gratitude. The majority of recruitment sequences in the dataset were minimal, including only a single recruiting move (65.9%, *n*=139).

Most recruitment sequences in the dataset are organized as adjacency pairs (Schegloff & Sacks 1973; Sacks 1992; Schegloff 2007), in which the recruiting move creates a normative obligation for a response. This includes imperative requests like *move over a little*, which make embodied compliance the conditionally relevant response (Goodwin 2006; Kent 2012), as well as interrogative requests like *where the fuck is my little fire starting bag* in (2). But conditional relevance, understood as a normative obligation to produce a specifiable next action, was not a criterion for the identification of recruitment sequences. Indeed, visible bodily actions as subtle as the tilting of one's head – a move that does not accountably mandate a response – can effectively recruit another's assistance (see Kendrick & Drew 2016: 8).

# **3.2 Non-minimal sequences**

In a minority of cases, the sequence included more than two recruiting moves. One recurrent basis for this was the absence of a response to an initial move, as in the following extract. Here, after no one responds to her request for a fork, Donna pursues a response (Pomerantz 1984; Bolden et al. 2012).

```
(5) BBQ 52:19
1 ali >I'll've so:me.<
2 (0.3)
3 car heh
4 (0.4)
```
### 4 Recruitment in English: A quantitative study

```
5 jam it almost sounds like you're speaking an Asian
  6 language.
  7 ali hHA HAH HAH hah °hah°
▶ 8 don is [there a fork] over the[re k i : d s ? ]
  9 jam [I'll: so:me.]
  10 ali [>I'll've so:me.<]
  11 (0.2)+(0.8)+
  12 car +sets sausage on Ali's plate+
  13 ali yay:::
  14 (0.2)
  15 don no?
  16 jam what?
▶ 17 don fork.
  18 (0.5)
  19 jam oh.
  20 *(2.7) ((background talk omitted))
▷ 21 jam *picks up fork and hands it to Donna-->
  22 don thank *you.
  23 jam ----->*
```
Although the request occurs at the possible completion of a sequence by the other participants (lines 5–7), the sequence is contingently expanded (lines 9–10), resulting in overlap that obscures the request (lines 8–10). After no one responds or attends to the request, Donna produces a candidate answer to her question, itself designed as a question, and thereby pursues a response (line 15). This attracts the attention of James, who turns to look at Donna and initiates repair (see Kendrick 2015 for a review), providing an opportunity for Donna to reissue her request to a now available recipient. In this context, after a pursuit, the subsequent request takes a minimal form, simply the name of the requested object (line 17, see §4.2.4). This successfully initiates an object transfer and completes the sequence. Overall, 79.1 percent of recruiting moves (*n*=167) were sequence initial, whereas 20.9 percent (*n*=44) were subsequent attempts (e.g. pursuits or repair solutions).

# **3.3 Recruitment types**

Participants recruit each other to manage a variety of practical contingencies. To provide a general sense of the distribution for the comparative study recruitments were classified into four types. The most frequent type was the provision

### Kobin H. Kendrick

of a service, that is, the performance of a practical action for the recruiter (49.8%, *n*=105). Transfers of objects were also especially frequent (38.8%, *n*=82). Less frequent were sequences in which one participant stopped or altered the trajectory of another's actions (e.g. *leave it alone*, 7.6%, *n*=16) and those in which a visible trouble elicited a direct provision of assistance (3.8%, *n*=8).

# **4 Recruiting moves**

To recruit others to act on their behalf, participants in social interaction draw on an arsenal of resources, both linguistic and embodied. In this section we will review the most frequent forms of language and visible bodily action observed in the first move of recruitment sequences. We begin first with the body and examine the forms of visible bodily action that either constitute or accompany recruiting moves, and then turn our attention to language and consider the grammatical formats and linguistic components that participants use to recruit others through talk.

# **4.1 Visible bodily actions**

Language is not necessary for recruitment. Even a subtle movement of the body, as one maneuvers to inspect a picture from across a table or searches the local environment, can elicit a helpful action from a co-participant (see Kendrick & Drew 2016; Drew & Kendrick 2018). Such exclusively embodied recruiting moves are striking specimens, but they are rare (see also Extracts 8 and 9). Only 7.6 percent of recruiting moves in the dataset were exclusively visual (*n*=15). However, this number does not include visual recruiting moves that elicited offers of assistance which were not included in the operational definition of recruitment. More common were complex multimodal actions, in which the move to recruit had both linguistic and embodied components, such as asking for something and reaching for it concurrently (41.4%, *n*=82). But despite the abundance of visible bodily actions, a narrow majority of recruiting moves were exclusively linguistic, with no relevant visual components (51%, *n*=101).

When participants do use visible bodily actions, what do they look like? Table 1 presents the types, frequencies, and proportions of relevant visible bodily actions observed in the dataset.

The set of body behaviors identified as relevant is diverse, including those whose function is accountably communicative (e.g. pointing at an object) as well as those whose function may, in the first instance, be instrumental (e.g. visibly

### 4 Recruitment in English: A quantitative study


Table 1: Frequencies and proportions of visible bodily actions in recruiting moves (*n*=97).

searching for an object, on which see Drew & Kendrick 2018). But within this diversity, one body-behavioral resource emerged as dominant: the hands. Over two thirds of all visible bodily actions in recruiting moves involved manual gestures or manual actions (64.9%, *n*=63). Within the dataset as a whole, a third of all recruiting moves included relevant manual movements, in one form or another.

With the exception of visible displays of trouble, an example of which was given in (3), and visible searches, which can be seen in (2), the remainder of this section illustrates the forms of visible bodily actions observed in the first move of recruitment sequences. As the analysis of these cases will show, the different forms of visible bodily action differ in how and to what extent they specify what the recipient should do in response.

### **4.1.1 Pointing**

By far the most common form of visible bodily action used for recruitment was a pointing gesture. Points occurred not only in recruiting objects (*n*=11), where they index the object in demand, but also in recruiting services (*n*=15), where they designate a location for the action to be done, among other possibilities. In the following extract, a pointing gesture is used to recruit a co-participant to pass an object (see also Drew & Couper-Kuhlen 2014: 22–23). At the possible completion of a sequence, John raises his arm and points to a box of biscuits on the table (see Figure 4).

```
(6) RCE14_109822
```

```
1 ann I'm gonna have to actually A write the paper and
 2 then B get round to sorting it out.
 3 joh .hh ye(h)ah hh in the way these things do hhh[hh
 4 ann [phh hh
 5 +yeah.#
▶ 6 joh +points-->
 7 ◩ #fig.4
 8 (0.3)
 9 ann which is a *( )(0.1)+(0.2)*
 10 *sets down pen---*
 11 joh -->+,,,-->
 12
 13 joh *°°c[ookie°°+
 14 -->+
▷ 15 ann *reaches for box-->
 16 ann [biscuit? *biscui[t biscuit biscuit
 17 -->*sets box in front of John-->
 18 joh [°biscuit biscuit°
 19 ann °yeah°*
 20 -->*
 21 (0.3)
 22 ann shall I show you what I've-
 23 joh yea[h
 24 ann [pictures I've picked up
```
The pointing gesture by John is recognizable as a move to recruit Anne to act on the pointed-at object. But unlike linguistic requests, which formulate an action for the recipient to perform (e.g. *can you pass me the biscuits?*), a point does not specify a next action to be done. It instructs the recipient to redirect her attention to the object and invites her to search for its current relevance to the situation. In this case, the relevance of the biscuits is transparent. At the moment John's gestures reaches its apex, Anne's gaze is directed downward to a pen in her hands. John holds the gesture for approximately 700 ms until Anne quickly sets the pen down on the table, an action that displays her (late) recognition of the move to recruit her (lines 6–11). As he retracts his gesture and she reaches for the box, John softly names the object (line 13), a linguistic action that occurs after the recipient has begun to comply but before the recruitment has been fulfilled, a

### 4 Recruitment in English: A quantitative study

Figure 4: John points to a box of biscuits as Anne looks down at the pen in her hands.

position in which linguistic recruiting moves serve to "expedite" the completion of the transaction (Kent & Kendrick 2016).

### **4.1.2 Reaching out**

The shape and orientation of a manual action can not only index an object but also specify a relevant next action. Extending one's hand towards an object, with an open, vertical orientation, is recognizable as a move to recruit the recipient of the gesture to transfer the object. The recognizability of this as a recruiting move comes from the specific hand shape, which visibly anticipates taking the object (see Streeck 2009: 47 on prehensile postures). In the following extract, as Mark produces a request for candy from a bag held by Rachael, he simultaneously reaches out towards the bag (see Figure 5).

```
(7) RCE22a_690761
```

```
▶ 1 mar +ohh can I+ have# s*ome.
 2 +.........+reaches out-->
 3 ◩ #fig.5a
▷ 4 rac *holds bag out-->
 5 (0.8)+#(1.3)
```
### Kobin H. Kendrick


Figure 5: Mark reaches out as he asks Rachel for some candy. Rachel holds out the bag as Mark puts his fingers into it.

Even before Mark's verbal turn is complete, Rachael begins to fulfill the request, holding the bag out towards Mark (lines 1–4). Although the shape of Mark's hand anticipates receiving the bag, Rachael tilts the bag towards him, such that he can reach inside. This precipitates some difficulty as he inserts his fingers into the bag and fumbles as he tries to extract the candies (lines 6–11). Unlike pointing gestures, which occurred with recruitments of all types, reaching actions such as this occurred exclusively in recruiting objects.

### **4.1.3 Holding out**

As we have seen, transferring objects is a common contingency that participants recruit each other to manage. Just as an object can move from B to A, from the recruit to the recruiter, as it were, so too can it travel in the opposite direction, from A to B. Holding out an object towards a recipient initiates a transaction in which the recipient should take the object and act on it. Similar to a pointing gesture, which directs the visual attention of its recipient to an object but does not specify what he or she should do with it, holding out an object presents the recipient with a puzzle: what should be done? It is no surprise, then, that

### 4 Recruitment in English: A quantitative study

participants use this form of recruiting move in specific contexts and for specific objects that radically constrain the possibility space of relevant next actions. In the following extract, after Ellen finishes a bowl of cheesecake, she picks up the empty bowl and holds it out toward Daniel (see Figure 6).

```
(8) RCE26b_560620
  1 dan well but you could (.) have it with something savory.
  2 +like some beef or someth+ing.
  3 ell +........................+picks up bowl-->
  4 (0.7)+(0.8)#
▶ 5 ell -->+holds out to Dan-->
  6 ◩ #fig.6
  7 *(1.2)
▷ 8 dan *steps forward, reaches out, grasps bowl-->
  9 (0.2)+*
  10 ell -->+lets go of bowl, retracts-->
  11 dan -->*takes bowl, sets off camera-->
  12 (0.2)
  13 dan I'm not sure meringue+ beef would be the best=
  14 ell -->+
  15 dan =combination but*
  16 -->*-->>
```
This action recognizably recruits Daniel to take the bowl and perform some action with it. As Ellen's arm reaches maximum extension, Daniel steps forward from his position against the kitchen counter, reaches out to take the bowl, and then sets it in the sink off camera. But how does Daniel recognize that some action from him is due and select an appropriate response? The deictic aspect of Ellen's arm extension "points" to Daniel and thereby addresses the action specifically to him.<sup>3</sup> Holding out an object towards a recipient, with one's arm at maximum extension, not only addresses the action but also makes accountably relevant an embodied response in which the recipient takes the object. The visible form of the action does not, however, specify that the recipient should then put the object in the sink, as Daniel does. The solution to this puzzle, one which Daniel himself may have used, lies in the routine organization of the current activity (see Rossi 2014) and the local ecology of the room. Clearing dishes is a routine (and hence anticipatable) course of action after one has finished a meal,

<sup>3</sup>Note that this can also be done with gaze direction, but here Ellen averts her gaze as she holds out the bowl.

### Kobin H. Kendrick

Figure 6: Ellen holds her bowl out towards Daniel

and standing near the sink, Daniel is in a position to place the bowl in the sink on Ellen's behalf. Embodied actions such as this are analytically distinct from requests (e.g. *would you put this in the sink for me?*) on the grounds that, unlike such requests, they do not specify the action the recipient is to perform in next position.

# **4.1.4 Other gestures**

The most frequent forms of visible bodily action in recruiting moves, as we have seen, involve transferring objects from one participant to another, hand to hand. But manual gestures also accompany and constitute recruiting moves for practical actions, not only objects. In the following extract, after Rachael begins to turn the page of a book, revealing a picture on the next page, Mark leans forward and produces a verbal display of disgust (lines 1–3). Rachel, presumably unaware that Mark had seen the picture on the next page, abandons turning the page, allowing Mark to view the current one.<sup>4</sup> At this moment Mark produces two quick finger movements that iconically depict turning the page (see Figure 7).

<sup>4</sup>This, too, is a case of recruitment, but not the focus of the present analysis.

```
(9) RCE22a 35:46
 1 +(0.6)
 2 mar +leans forward-->
 3 mar euww:::
 4 (0.6)+#(0.8)+
▶ 5 mar -->+flicks finger twice+
 6 ◩ #fig.7a
 7 *(0.3)#(0.5)*
▷ 8 rac *turns page *
 9 ◩ #fig.7b
 8 (0.6)
 9 mar okay never mind.
 10 (0.3)
 11 mar I thought that was a mouth open.
```
Figure 7: Mark flicks his finger up and down and then Rachel turns the page of her book.

The iconic gesture visually depicts the action that Rachel should perform and thereby specifies the relevant response. After Rachel fulfills the recruitment, Mark accounts for his interest in the picture and brings the exchange to a close.

### Kobin H. Kendrick

### **4.1.5 Instrumental actions**

The forms of conduct we have seen thus far are used by participants to initiate transactions in which some practical action from a co-participant is accountably due. But not all forms of visible bodily actions observed in recruiting moves do this. In some cases, a participant who produces a recruiting move through language also performs an embodied action that facilitates the recruitment. In the following extract, for example, as Fabian directs Kate to put her coat on, he reaches out and pulls it up, facilitating the fulfillment of the sequence (see Figure 8).

```
(10) RCE02 04:00
```

```
1 fab uhm
 2 (1.8)
 3 fab but yeah.# are you co:ld.
 4 ◩ #fig.8a
 5 (0.5)
 6 kat mm hm.
 7 +(0.9)
 8 fab +.....-->
▶ 9 fab then cover yourself# up+ pro*perly.=
 10 -->+pulls coat-->
 11 ◩ #fig.8b
 12 kat *.....-->
 13 kat =well yeah but (0.5) oh
 14 (0.6)*(0.7)
▷ 15 kat -->*lifts herself off ground-->
 16 kat there's *dirt* all+ round the back of my
 17 -->*sits*
 18 fab -->+
 19 (0.6)
 20 kat [thing.
 21 fab [yeah.
 22 (0.6)
 23 fab and whose fault is that.
 24 (0.7)
 25 kat .tsk
```
### 4 Recruitment in English: A quantitative study

Figure 8: Fabian gazes at Kate before he asks whether she is cold. After she confirms that she is, he reaches out and begins to pull her coat up.

The embodied action in this case illustrates a distinct mechanism for recruitment. In the majority of cases in the dataset, a speaker produces a recruiting move through language, using a grammatical format, such as an interrogative or imperative, that normally encodes an obligation to respond. The recruiting move is thus recognizable and accountable as a social action that combines turnconstructional and sequence-organizational practices into a mechanism for recruitment. Visible bodily actions such as points and iconic gestures are analogous in that they recognizably initiate transactions in which a responding move is due (even if they do not fully specify what form it should take). A distinct mechanism, the one illustrated by this example, is for a participant to begin a course of action that *necessarily* involves co-participation. The recognition of the incipient course of action and the one's participation in its completion is a mechanism for recruitment in its own right. Here, as Fabian pulls Kate's coat up, he begins a course of action that requires participation from her and thereby recruits her to carry out the action with him. Kate evidently recognizes this and lifts herself off the ground to allow Fabian to pull her coat up.

# **4.2 Grammatical formats**

The grammar of a language includes a multitude of forms and formats that speakers use to construct turns at talk and produce recognizable social actions. For the recruitment of co-participants to act on one's behalf – surely one of the most basic of all social actions – the forms of grammar that speakers of English use most frequently come from only three basic types. Over 90 percent of linguistic recruiting moves in the dataset have an interrogative, imperative, or declarative grammatical format (see Table 2).

### Kobin H. Kendrick


Table 2: Frequency and proportion of grammatical formats in recruiting moves (*n*=196).

Referred to as sentence types by linguists (Sadock & Zwicky 1985; Palmer 2001; König & Siemund 2007), these grammatical formats institutionalize basic social relations (e.g. epistemic and deontic authority) and recurrent interactional contingencies (e.g. the redistribution of knowledge and the performance of practical actions) that all participants in social interaction must have ways to manage. The intricate relations between grammatical formats and social actions form a complex web, with no simple one-to-one correspondences (Schegloff 1984; Levinson 2013). Imperatives, for example, can and frequently do direct the actions of others (e.g. *drink that*), but so too can they offer (e.g. *have the last one*), admonish (e.g. *just watch it, okay?*), initiate repair (e.g. *pardon me*), or grant a request (e.g. *go for it*), among other possibilities (Kent & Kendrick 2016). But even within such a complex web of relations, order emerges, as particular forms are tied to general domains of action (Couper-Kuhlen 2014).

Given the large number of recruiting moves in the dataset that employ these basic formats, a complete enumeration of all types and subtypes is not possible within the confines of this chapter. Instead, for each format, we will examine a small set of cases in order to address a specific question or to bring a novel phenomenon into view. And for those recruitments without a predicate, which therefore do not belong to one of the three basic types, a simple discussion of their rather restricted context of use will suffice.

### **4.2.1 Interrogatives**

To recruit a co-participant to perform a practical action one can simply ask him or her to do so. In the dataset, the most frequent grammatical format for linguistic recruiting moves is the interrogative. Speakers generally use interrogatives to query the abilities or desires of recipients to perform an action (e.g. *can you pass*

### 4 Recruitment in English: A quantitative study

*me the butter, will you hand me that*) or to ask about the availability or location of objects (e.g. *do you have a cup, where's the bottle opener*). Such questions exploit an asymmetry between an unknowing speaker and a knowing recipient, indexed by the epistemic stance of interrogative grammar (Heritage & Raymond 2005; Heritage 2012), as a generic mechanism for recruitment. Traditionally referred to as indirect speech acts (Searle 1975; Brown & Levinson 1987) because they ostensibly concern the practical and personal contingencies of performing an action and not its performance per se, interrogatives such as these are better thought of as social action formats (Fox 2007) for recruitment, each with its own quirks and specifiable domains of use.

Table 3 presents the frequencies and proportions of interrogative recruiting moves identified in the dataset. Those with fewer than two attestations appear under other.


Table 3: Frequency and proportion of interrogative formats in recruiting moves (*n*=78).

Although in principle one could investigate each of these forms to arrive at a description of the specific socio-interactional conditions under which they occur (see Rossi 2015; Zinken 2015; Fox & Heinemann 2016; 2017), we will here restrict our discussion to a comparison of just two of these forms.

It has been suggested that the distinction between *can/could you* and *will/ would you* is "of relatively minor significance" (Curl & Drew 2008: 150, fn. 1). An examination of the distribution of these formats in the present dataset, however, suggests at least two possible differences. The first concerns the grantability of the request. The selection of *will/would you* over *can/could you* appears to orient

### Kobin H. Kendrick

to possible contingencies that may affect the grantability of the request, in line the observation by Curl and Drew that *can/could you* displays relatively little orientation to such matters. Consider, for example, the request in the following extract, which comes from the early moments of a family mealtime. After Britney hands her mother a plate, and as her mother begins to take her seat, she asks for the butter, using the form *can you*.

```
(11) SLF 24:53
  1 +(2.4)+*(0.2)*
  2 bri +picks up plate and holds out it to Mom+
  3 mom *takes plate*
  4 mom *tha:nk yo:u.
  5 *sets it on table-->
  6 (1.1)*
  7 -->*
▶ 8 bri c'n #you pass me the butter:.
  9 ◩ #fig.9
  10 (0.4)*(0.4)
▷ 11 mom *picks up butter-->
  12 mo? mm:
  13 (0.8)
  14 mom *d'we (.) go through enough butter* and bread last
  15 *holds it out to Britney----------*
  16 night?
  17 bri oh my go:sh.
```
There are few, if any, contingencies that could affect the grantability of this mundane request. It occurs in a setting where such requests (and their granting) are common. It is produced at a very precise moment – immediately after the mother has set her plate on the table, but before she has had an opportunity to begin a next course of action (lines 5–7) – thereby obviating one possible source of contingency. And it requests an object, the butter dish, that is directly next to the mother's plate, readily within her reach (cf. Keisanen & Rauniomaa 2012).

The request in the following extract, in contrast, occurs under less opportune circumstances. As Kimmy recounts a problem with a client at her work to Carrie (lines 1–6), Donna, who has not been involved in the telling, looks around the table, picks up a bottle of beer, inspects it, and sets it back down. She then points to a bottle of beer on the other side of the table and asks Carrie, the recipient of Kimmy's telling, to hand it to her, using the form *will you*.

### 4 Recruitment in English: A quantitative study

Figure 9: Immediately after the mother (in black) sets down her plate, Britney points to the butter and asks her to pass it.

### (12) BBQ 27:27


### Kobin H. Kendrick


Figure 10: Donna points to a bottle of beer as she asks Carrie to hand it to her.

Although Donna's request is produced at the possible completion of a turnconstructional unit (lines 6–7), suggesting that she has timed it so as not to interrupt, it occurs at a position in which the telling sequence is not yet possibly complete.<sup>5</sup> The request is thus interjected into an on-going activity and addressed to a participant whose status as the recipient of the telling renders her less than fully available to grant the request at that moment. To do so immediately would require that she suspend or postpone one action (i.e. responding to the telling) in order to produce the other. The selection of *will you* over *can you* in this case appears to orient to local contingencies such as these that influence the grantability of the request. Note that the request indeed runs into trouble as Carrie must initiate repair before she can fulfill it (lines 15–18). In the dataset, interrogative requests for which no discernable contingencies exist occur as *can/could you*,

<sup>5</sup>Note that Carrie has not yet responded to the telling, which she does at line 10 before she fulfills the request.

4 Recruitment in English: A quantitative study

whereas those that involve subtly more complex circumstances or actions appear as *will/would you*.

A second distinction between *can/could you* and *will/would you* concerns a specific type of request that occurs in one form but not the other. Requests that find fault in the actions – or more specifically the *inactions* – of the recipient occur as *can/could you* but not as *will/would you* (cf. Kent & Kendrick 2016). The following extracts illustrate such "fault-finding" requests.

(13) RCE06 15:48


In the first case, as Jessica and Sarah sit on a blanket on the lawn of a university campus, Jessica asks Sarah to move over to make room for her. The request includes an account that finds fault with Sarah's inaction, blaming the trouble (i.e. that Jessica does not have enough space on the blanket) on her. Note that after Sarah complies with the request, she immediately challenges the account, orienting to its fault-finding character. In the second case, after Ben evidently notices that Kerry, who has been eating a bowl of cereal, has milk on her chin, he asks her to remove it, and like the first case he also includes an account that (teasingly) finds fault with her inaction. In each of these, the speaker produces a multi-unit turn in a reqest + account format with no prosodic boundary between the two units, and the complex action that results both asks the recipient to perform an action and holds her to account for not having already done so.

### Kobin H. Kendrick

The motivation for the selection of *can/could you* over *will/would you* for such requests supports the conclusion that *will/would you* indexes greater contingency. For the speaker to find fault with the recipient's inaction, there should be no local contingencies that would have impeded the performance of the action and could therefore provide an account for the recipient's inaction.

In comparison to the differences that Curl & Drew (2008) observed between *can/could you* and *I wonder if* – a form of request that does not occur in the present dataset – the distinction between *can/could you* and *will/would you* is indeed subtle, and additional data must be brought to bear on this issue before a final verdict can be reached. But the data at hand suggest the two forms are not equivalent and the differences between them, while perhaps minor, are interactionally significant.

### **4.2.2 Imperatives**

Under what circumstances does one participant *tell* another to perform an action rather than *ask* him or her to do so? In general, speakers use imperatives in interactional contexts in which the sequential contingency of an interrogative request, which orients to a recipient's right to refuse, is unnecessary, such as after participants have agreed explicitly or tacitly to a collaborative activity (Wootton 1997; Rossi 2012; Zinken & Ogiermann 2013), or is otherwise inexpedient, such as when the situation calls for immediate action. In contrast to the epistemic stance of interrogatives, imperatives typically encode a deontic stance in which the speaker claims the authority (Stevanovic & Peräkylä 2012; 2014) or the entitlement (Craven & Potter 2010) to direct the actions of the recipient. Although imperatives implement a diversity of social actions (see §4.2 for examples), the recruitment of a co-participant to perform a practical action is among the most common. Referred to as directives by some (Goodwin 2006; Kent 2011; 2012) and requests by others (Rossi 2012; Couper-Kuhlen 2014), such imperatives name a practical action and thereby make relevant the performance of that action as a preferred response (see Kent 2011, on the preference organization of directive sequences).<sup>6</sup>

A basic distinction that runs through the set of imperatives in the dataset involves the complex relationship between (i) the imperative, (ii) the practical action it makes relevant, and (iii) the course of action in which they both occur

<sup>6</sup> In comparison to interrogatives, which have a number of highly frequent forms, imperatives are relatively homogeneous. The vast majority of imperatives in the dataset (83.6%, *n*=61) were simple verbal predicates; six (8.2%) were prohibitives formed with *don't* (e.g. *don't get it out* in Extract 3); and six (8.2%) included either let me or let's (e.g. *let me have that*).

4 Recruitment in English: A quantitative study

(see Kent & Kendrick 2016). As Wootton (1997) and Rossi (2012) have observed, a common home for imperatives is in the midst of a collaborative activity. While speakers can use imperatives to initiate courses of action (e.g. the passing of plates at the dinner table), it is more common that they use them to manage courses of action that have already been set in motion.

In (10) above, for example, before Fabian directs Kate to cover herself up properly with an imperative, he initiates the course of action with an interrogative (i.e. *are you cold?*). Similarly, in the following extract, after Hailey offers Britney pickles, an action that initiates a course of action, she uses an imperative to direct her to ask their father, an action that progresses the course of action.

(15) Sunday Lunch

```
1 hal did you want pickles Britney?
2 bri un-uh.
3 hal ask dad.
4 (0.5)
5 bri DA:::D.
```
The courses of action in which imperatives occur can have recognizable structures, such that the relevance of a specific next action is projectable on the basis of prior actions or events. In the example above, while the relevance of a subsequent offer to another member of the family may be projectable (e.g. on the basis of etiquette or other social norms), the delegation of this task to Hailey surely is not. It is a contingent and opportunistic development of the course of action (i.e. Britney's declination), not one that is anticipatable on the basis of the initial offer. As a point of contrast, consider the course of action in the following extract. Here an imperative directs a recipient to perform an action whose relevance *precedes* the imperative itself. During a break in a game of Monopoly, after Luke opens a can of beer, he notices that the beer is partially frozen (line 4). The beer then begins to overflow from the can (the result of a chemical reaction as the sudden decrease in pressure in the can lowers the freezing point of the beer, causing it to freeze and expand). As a solution to this emergent problem, Luke sips the frozen beer intermittently as it comes out.

```
(16) Monopoly Boys 37:41
```

```
1 *(0.2) *
2 luk *opens can*
3 (0.3)
4 luk o:h, [it's slushy.
```

```
5 ric [aw:.
 6 *(0.7) *(0.8) *(0.6) *(0.8)#
 7 luk *raises can*sips beer*lowers can*holds-->
 8 ◩ #fig.11
 9 ric what the hell.
 10 (0.5)
 11 luk do you have a *cup?
 12 -->*raises can-->
 13 (0.6)*
 14 -->*sips beer-->
 15 ric yeah but what's going on *with that. I've never seen
 16 luk -->*lowers can-->
 17 that before.
 18 (.)*
 19 luk -->*
▶ 20 ric hey. (.) drink that.=↑quick.↑=↑↑it's coming ba:ck.↑↑
 21 (0.2)
 22 luk I- (.) ahh
 23 ric COME ON.=IT'S COMING OUT.=IT'S GONNA GET ON MY
 24 my-* (.) ta:bl:e.*
 25 luk *raises can-->*sips-->>
```
The course of action that develops as Luke manages the problem has a projectable structure. At least two bases for this can be identified. Firstly, the very recognition of the problem allows for the projection of a set of possible solutions, such as sipping the frozen beer (lines 7 and 14) or pouring it into a cup (line 11). Thus once a problem has been recognized and publicly registered, as Rick does with *what the hell* in line 9, the provision of a possible solution becomes relevant. Secondly, after Luke has twice sipped the beer after it began to overflow, that he could or should do so again becomes anticipatable as a possible solution to the problem. In this way, the local structure of the sequence provides a basis for the projection of possible next actions.

With this in mind, we can now see that the imperative that Rick produces at line 20 (*drink that*) occurs in a position at which the action it directs Luke to perform is already relevant as a possible solution to the problem. Furthermore, before Rick directs him to do so, Luke has had an opportunity to perform the relevant action. Note that Rick produces the particle *hey*, a minimal form that alerts Luke of the reemergence of the trouble, and then pauses briefly before he issues the imperative (line 20). This prompts action from Luke and creates an

### 4 Recruitment in English: A quantitative study

Figure 11: Luke (white shirt) holds his gaze on the beer can for 0.8 seconds as frozen beer begins to emerge.

opportunity for him to act, one that he does not take. Thus both the relevance of the requested action and an opportunity to perform it *precede* the imperative. This stands in clear contrast to the imperative in (15) (*ask dad*), in which the relevance of the directed action and the opportunity to perform it both *follow* the imperative.

What is the consequence of this difference? The position in which an imperative request occurs "colors" its action, such that imperatives that follow the relevance of the requested action and an opportunity to perform it not only request a recipient to perform an action but also "admonish" him or her for not having already done so (Kent & Kendrick 2016). Within the present dataset, imperatives frequently occur in this po fsition and frequently do more than just recruit a co-participant to act (cf. Mandelbaum 2014). The data therefore suggest that speakers use imperatives not only for the management of practical courses of action – to recruit others to do things per se – but also for the *ex post facto* enforcement of social norms that regulate practical courses of action.

### Kobin H. Kendrick

### **4.2.3 Declaratives**

Just as one can ostensibly inquire into the abilities, desires, and future actions of co-participants to recruit them, so too can one inform co-participants of one's own desires, needs, or future actions to do so (see Stevanovic 2011; Childs 2012). The majority of declaratives in the dataset (*n*=19, 58%) include linguistic forms that index the obligations, volitions, or abilities of either the speaker or recipient. The most frequent forms (i.e. those with multiple attestations) are *you should* (*n*=4, 12%), *I want* (*n*=4, 12%), *I need* (*n*=2, 6%), *I'll have* (*n*=2, 6%), and *we need* (*n*=2, 6%).

Typically declarative grammar encodes an epistemic stance that is the inverse of interrogatives: the speaker has knowledge that the recipient lacks (Heritage & Raymond 2005; Raymond 2010; Heritage 2012). But as recruiting moves, declaratives nonetheless frequently exploit an epistemic asymmetry in which the recipient is in a K+ position. A speaker who informs a recipient of a need, as in the following extract, claims to know what should be done but not how to do it.

(17) BBQ 23:09

```
1 jam I was expecting like [a deposit in my accou(h)nt.
  2 ali [oh James.
  3 jam heh heh heh
  4 ali James.
  5 (0.2)
  6 ali really quick?=
  7 jam wha[:t.
  8 ali [can I get he:lp for something.
  9 jam what's up.
  10 ali can you put that down for just a minute.
  11 jam hold on lemme just get done with (this last piece)
  12 ((four lines omitted))
  13 jam wh[at's up babe.
▶ 14 ali [I need to check- I need to check th*e uhm (0.2) the
▷ 15 jam *stands up-->
  16 microphone quality?
  17 (0.5)
  18 jam do you have earphones?*
  19 -->*holds-->
  20 (0.3)
  21 ali yeah I do but I don't know *where the earphone plugin is.
```
### 4 Recruitment in English: A quantitative study

22 *jam* \*walks behind camera-->>

23 jam °okay.°

The mechanism of recruitment in such cases is analogous to that of a 'my side' telling (Pomerantz 1980), in which a speaker reports his or her limited access to an event and thereby elicits additional information from a recipient who has greater access (Childs 2012). In recruiting moves, however, a speaker reports his or her knowledge of a situation that requires action (i.e. a practical problem) and thereby elicits action from a recipient (i.e. a practical solution) who has a greater ability, availability, or obligation to perform the action. In the example above, the basis for James's ability to resolve the problem is made explicit earlier in the conversation when Alison reports that he used to own the same model of video camera. Recruitments such as this reveal a complex relationship between the epistemic status of the recipient (i.e. James knows how to operate the camera) and the recipient's obligation to provide assistance.

The selection of a declarative over other formats does not necessarily depend on the epistemic and deontic status of the speaker, however. The grammatical format of a recruiting move also affects the opportunity for response in important ways. Whereas interrogative requests constrain the response such that non-granting responses are dispreferred actions, declarative requests leave the response relatively open, allowing for a larger set of possible next actions (Vinkhuyzen & Szymanski 2005; Rossi & Zinken 2016). In the present dataset, the influence of the grammatical format can be seen indirectly in the quantitative distribution of responses. While over two thirds of interrogatives received a verbal response (69.2%, *n*=54), only half of the declaratives did (51.5%, *n*=17), with the remainder receiving an embodied response or no response at all. This shows that participants orient to the two formats in different ways. The preference for polar interrogatives to receive polar responses (Raymond 2003) may have contributed to this difference. Polar tokens, such as *yeah*, *okay*, *no*, and *nah*, occurred in response to 40.9 percent of polar interrogatives (*n*=27), in contrast to only 15.2 percent for declaratives (*n*=5). The quantitative distribution of responses in the dataset supports previous observations that declarative requests place fewer, or at least different, constraints on the response space than interrogatives do (cf. Rossi, Chapter 5, §4.2). The selection of a declarative format for recruitment thus affords greater agency to the recipient, who has an opportunity to select a response from a larger set of alternatives, including sequence initiating actions (cf. Kendrick & Drew 2014: 111).

Although the majority of declaratives in the dataset are modal statements that index the desires, abilities, or obligations of the participants, many are not. One

### Kobin H. Kendrick

type that calls co-participants to action without reference to such personal states is the announcement of the completion of a task (e.g. the familiar *dinner's ready*) (cf. Rossi 2018: 384). Announcements of task completion exploit the normative organization of complex activities, which can involve transformations of the participation framework, as a mechanism for recruitment. For this to work, the course of action that comes to completion (e.g. the preparation of food) must belong to a complex activity (e.g. the meal as a whole) that includes a subsequent course of action (e.g. serving and eating the food). The completion of one course of action makes the initiation of a next conditionally relevant (cf. Schegloff 2007: 213–215). For some activities, the participation framework also changes; a course of action that involves few participants (e.g. those who prepared the food) can transform into one that involves many (e.g. those who will eat the food). The relevance of the announcement thus derives from the need to solicit participation in the next phase of the activity. In the following extract, which again comes from the interaction between friends as they prepare a barbecue, an announcement by Kimmy that the coals are ready recruits Carrie to begin the relevant next course of action.

```
(18) BBQ 36:16
```

```
▶ 1 kim I think those coals# are ready for your sausages.
 2 ◩ #fig.12a
 3 ali yeah man.
 4 jam I think you *might wanna# pull up +the ra:ck? a little bit.
 5 ◩ #fig.12b
▷ 6 car *reaches for sausages-->
 7 kim +walks over to grill-->
 8 (0.2)
 9 kim *yep.
 10 car *picks up sausages-->
 11 (0.5)*(1.6)
 12 car *moves towards grill-->>
 13 jam (it's)+ pretty hot.
 14 kim -->+
```
The announcement also marks a transformation of the participation framework in the activity as a whole. Earlier in the interaction Kimmy had been recruited to light the coals (cf. Extract 2) and Carrie had revealed that she had brought sausages to cook on the grill. With the announcement, Kimmy informs Carrie that her task is complete and that the relevant next course of action can commence. Note that while Kimmy does return to the grill (see lines 7 and 14),

### 4 Recruitment in English: A quantitative study

Figure 12: Kimmy (red pants) announces that the grill is ready for Carrie's sausages. Carrie (black hat) then reaches for the sausages before she moves towards the grill.

she does not do so to assist Carrie with the preparation of the food, but rather to resolve a practical problem in the course of action that she had just announced as complete (see line 3, which is also a recruiting move). As this example shows, announcements of task completion can be used by participants to manage a transformation of participation across successive courses of action within a complex activity and thereby recruit others to assist.

### **4.2.4 Non-clausal**

As we have seen, the three most frequent grammatical formats for recruiting moves – interrogatives, imperatives, and declaratives – differ in important ways. But behind such differences lies a common linguistic structure: the clause. Defined as a predicate (e.g. a verb or a verb complex) and its associated arguments (e.g. noun phrases), a clause is a linguistic structure that, in the words of Thompson & Couper-Kuhlen (2005), "can be thought of as a crystallization of solutions to the interactional problem of signaling and recognizing social actions" (p. 484). The observation that 93.8 percent of verbal recruiting moves in the dataset have clausal formats (see Table 3) clearly supports the dominance of clausal formats in this domain of social action (or perhaps sequence-initiation more generally).

But what about the linguistic moves that do not have a clausal format? There are three recurrent types of phrasal recruiting moves in the dataset. First, a phrasal format can occur as a response to an other-initiation of repair that locates a clausal recruitment as a trouble source (e.g. A: *is there a fork over there ki:ds?* B: *what?* A: *fork* in Extract 4). Second, a phrasal format can occur as a pursuit of a response to a clausal format (e.g. A: *l'mme have that butter when yer through there.* A: *butter please.*). Third, a phrasal format can occur as an ad-

### Kobin H. Kendrick

dress term that either pursues a response (e.g. A: *Britney do you wanna help me set the table?* A: *Brit.*) or admonishes a recipient's actions (e.g. *shhh. Ow:en.* after an inappropriate remark). As this list indicates, aside from the use of address terms as admonishments, phrasal formats tend to occur in sequence-subsequent positions (see also Floyd, Chapter 3, §3.3.1) after an initiation of a sequence by a clausal format. An exception to this generalization is the phrasal °°*cookie*°° in (5), which occurs after an embodied recruiting move, a pointing gesture to a box of cookies, not a verbal one. This case nonetheless shows that phrasal formats tend to occur in sequence-subsequent positions after a more explicit move to recruit a co-participant, even if the initiation of the sequence is done without language.

# **4.3 Additional turn components**

In addition to the grammatical format of the recruiting move, participants also optionally select among a set of turn-constructional components that adjust or modify the action in various ways. In this section we review three such components that were frequently observed in the dataset: address terms, mitigations, and accounts.

### **4.3.1 Address terms**

Address terms, such as names (e.g. *Haley can you get the salt and peppy*), terms of endearment (e.g. *honey will you hand me that*), and person reference categories (e.g. *is there a fork over there ki:ds?* in Extract 5), frequently occurred with recruiting moves in the dataset. Nearly 20 percent (*n*=39) of linguistic recruiting moves included an address term, either as a turn-constructional component in turn-initial or turn-final positions or as a generic pre-expansion of the sequence (see, e.g., Extract 17, lines 2 and 4). In comparison, a study of 328 questions in English conversation – most of which were requests for information, confirmation, or repair – found that only 2.1 percent (*n*=7) occurred with an address term (Stivers 2010: 2777). This suggests that address terms occur more frequently with recruiting moves than in other domains of social action in English.

Two possible explanations for the conspicuously high proportion of address terms in recruitments present themselves. The first concerns a generic problem of social interaction, one that is especially acute when more than two participants are involved: the selection of who should act or speak next.<sup>7</sup> A participant can use an address term to designate a particular co-participant as the addressed recipient of the recruiting move and thereby select him or her as the one who should

<sup>7</sup>Note that 86.3% of sequences in the dataset, *n*=182, came from multiperson interactions.

### 4 Recruitment in English: A quantitative study

respond (cf. Sacks et al. 1974; Lerner 2003). Similarly, an address term can be used as a generic pre-expansion to secure the attention a particular co-participant before a base sequence (cf. Schegloff 1968; 2007). Because recruiting moves so frequently initiate courses of action – in contrast to requests for information, confirmation, or repair, which also frequently occur in non-initial positions (e.g. as insert expansions, post-expansions, and various "follow-up" questions) – the socio-interactional problem for which address terms are a solution arises with greater frequency. Some indirect evidence for this comes from the relatively low proportion of address terms with imperatives, which tend to occur in non-initial positions within a course of action (see §4.2.2): only 8.2 percent (*n*=6) of imperatives include an address term, whereas 30.8 percent (*n*=24) of interrogatives do (see also Zinken & Ogiermann 2013: 271, fn. 7; Rossi 2015: 96).

The second explanation for the high proportion of address terms concerns the nature of the action itself. A special relationship between requests, broadly understood, and address terms was first noted by Brown & Levinson (1978; 1987) who argued that such terms signal in-group membership between speaker and recipient and thereby mitigate the request's threat to the recipient's negative face (cf. Rendle-Short 2010: 1207). In an analysis of sequence-initiating actions in general, Lerner (2003) distinguishes between turn-initial and turn-final address terms and argues that the latter can "demonstrate a particular stance toward or relationship with a recipient under circumstances where that demonstration is particularly relevant" (Lerner 2003: 185; see also Clayman 2010; Butler et al. 2011). Although the present analysis does not distinguish between turn-initial and turn-final positions, the use of address terms to affirm one's relationship with a recipient as one solicits his or her assistance – a circumstance in which the relationship is especially relevant – is also a plausible explanation for the high proportion of address terms observed with recruitment (see also Zinken, Chapter 8, §6; Baranova, Chapter 9, §3.4).

### **4.3.2 Mitigators**

The linguistic construction of a recruiting move may also include a variety of practices that, in various ways, mitigate the action (e.g. *can I have a little bit*). The observation that speakers use linguistic devices to mitigate particular social actions can be traced back to linguistic research by Lakoff (1973) and Brown & Levinson (1987) on "hedges" (see Schneider 2010 for a review). From a conversation-analytic perspective, the use of practices to mitigate a recruiting move orients to the preference for agreement in conversation (Sacks 1987), as well as a general principle of social solidarity (Heritage 1984), in that such practices min-

### Kobin H. Kendrick

imize what the recipient must do to comply and thereby maximize the opportunity for an affiliative response. Consider, for example, the following extract. As James and his housemates prepare their respective dinners in a communal kitchen, he asks for an onion.

(19) RCE09 11:43 ▶ 1 jam has anyone got a spare onion I can borrow. 2 (.) 3 half an onion. ▷ 4 ben ye:ah, go in there.

The construction of the request includes at least three practices that mitigate the action. First, the selection of *a spare onion* over *an onion* anticipates a possible basis for rejection, namely that one of the housemates may have an onion but may need it for his or her own meal. This provides a recipient who wishes to reject the request with a means to do so, thereby minimizing the potential for discord. Second, the selection of the word *borrow* over *have* implies only a temporary transfer of possession. Even though one is not expected to return consumables such as onions, the selection of *borrow* nonetheless orients to the preference for agreement in that it ostensibly makes the request easier to grant. Third, after a short pause (just under 200 ms) in which no one responds, James issues a self-repair that minimizes the request in an even more transparent manner, cutting it in half. Transition space self-repairs such as this are common in requests and orient to the possibility of rejection that the absence of a response can indicate (Davidson 1984).

# **4.3.3 Accounts**

An account is an action, defined as a clausal turn-constructional unit in the present study, that provides a reason or explanation for a participant's action or inaction (see Robinson 2016 for a review). Early conversation-analytic research on accounts by Atkinson & Drew (1979) and Heritage (1984) showed, among others, that participants commonly provide accounts for responding actions that fail to align with normative expectations set by initiating actions (i.e. for dispreferred responses). Subsequent research has shown that accounts also accompany some initiating actions, most notably requests and directives (Goodwin 1990; Houtkoop-Steenstra 1990; Schegloff 2007; Raevaara 2011; Parry 2013; Baranova & Dingemanse 2016). A possible explanation for the occurrence of accounts with requests comes from Schegloff (2007), who argues that requests are dispreferred

### 4 Recruitment in English: A quantitative study

actions and that the construction of requests reveals a preference for offers (see Kendrick & Drew 2014 for counterarguments). One piece of evidence that Schegloff cites to support the dispreferred status of requests is the "regular" provision of accounts with them (p. 83). In light of the previous research on the relationship between requests and accounts, one would expect that accounts should be frequent within the present dataset. But in fact they are rare. Only 13.3 percent of linguistic recruiting moves (*n*=26) include an account (see Extracts 13, 14, and 16).

Why are accounts for recruiting moves so rare? To answer this question we must first take a step back and review the organization of recruitment (Kendrick & Drew 2016). In general terms, a recruiting move solicits or occasions a practical action from a co-participant in order to resolve a trouble encountered by its speaker in a practical course of action. The nature of such troubles varies: one may not have enough space on a couch (Extract 4); one may need a utensil in order to eat a meal (Extract 5); one may not be able to see a picture in a book from across a table (Extract 3); and so on. The recruiting move frequently, but not invariably, formulates a possible solution to the trouble for the co-participant to perform. An imperative such as *move over* (Extract 4) formulates an action that constitutes a possible solution to a practical problem, and an interrogative such as *is there a fork over there* (Extract 5) similarly refers to an object that could resolve the participant's problem. In contrast to the explicitness of the solutions, often embodied in the form of the recruiting move, the nature of the trouble is frequently left implicit, as participants treat the problems as transparently recognizable. In some cases, a problem has emerged in the interaction (e.g. the overflow of beer in Extract 16) and thus need not be stated. In others, the problem is recognizable by reference to the norms of an activity such as a meal (e.g. the lack of a fork). The explicit articulation of such troubles is the most common function that accounts for recruiting moves serve. Accounts are thus rare in the dataset because the "here and now" troubles that arise in face-to-face interaction are routine, often manifest explicitly in the situation, and are transparently recognizable to co-participants (see also Baranova & Dingemanse 2016: 647). Given this, the question becomes not why accounts are so rare, but why they should occur with requests at all. The fault-finding character of those requests that do include accounts (e.g. Extract 13, 14, and 16) points to one possible answer and an avenue for future research.

Kobin H. Kendrick

# **5 Responding moves**

The recipient of a recruiting move finds him- or herself in a position to respond. In this section we will consider this second move in a recruitment sequence. In contrast to the technical definition of a response as the second pair part of an adjacency pair (Schegloff & Sacks 1973; Schegloff 2007), the term is used more freely in the comparative study to refer to any relevant action in a next position to the recruiting move (see Chapter 2). This includes not only actions that complete the sequence, such as a move to fulfill a request, but also actions that, in one way or another, orient to the recruitment but leave the sequence open (e.g. a repair initiation). This includes cases in which participants recognize a trouble and volunteer assistance, thereby *initiating* a course of action from a second position (cf. Schegloff 2007 on "retro-sequences"). We will first consider the distribution of some general types of responding moves, so defined, that were observed in the dataset before we turn our attention to two specific types – deferring the recruitment and recruiting the recruiter – that require more detailed discussion.

# **5.1 Response types**

At the most general level, the recipient of a recruiting move can either fulfill the sequence, that is, carry out a relevant practical action, or opt not to do so. As Table 4 shows, in the majority of cases (61.6%, *n*=130) the sequence is fulfilled. Furthermore, if one considers only those recruiting moves that occur in a sequence-final position, including minimal two-move sequences and terminal moves in expanded sequences (see §2), the proportion of fulfillment increases to over 70 percent. This is clear evidence for the preference for agreement (Sacks 1987) and the principle of social solidarity (Heritage 1984), two specifications of the general bias towards cooperation in social interaction (see also Floyd et al. 2018).

Consistent with this, rejection was rare in the dataset. Less than one in ten moves to recruit co-participants were rejected. This included explicit rejections, such as *no* and *fuck you*, as well as various accounts that accomplish rejection (Drew 1984), such as *I don't know how to* and *I'll leave it to you*. More common were recruiting moves that received no response whatsoever. In such cases, the recipient produced neither a verbal response to the move nor a practical action that recognizably fulfilled (or began to fulfill) the sequence. This can be understood as an alternative to rejection, in that the recipient opts not to provide assistance but also not to reject the recruiting move explicitly. It can also be the result of a failure in addressing the move to a particular co-participant (see, e.g.,

### 4 Recruitment in English: A quantitative study

Table 4: Frequency and proportion of response types for all cases (*n*=211) and for sequence-final cases (*n*=166)


Extract 5). And in a small number of cases, the recipient of a recruiting move initiates repair (again, see Extract 5). Responding moves that did not correspond to the types described thus far were analyzed as "other" for the comparative study. Three of these will be examined in detail in subsequent sections.

As for the modality of the responding move, nearly three quarters included relevant visible bodily actions (72.6%, *n*=143). Given that recruitment involves the performance of a practical action (and not, for instance, the provision of information), this comes as little surprise. Fully nonverbal responses were not the norm, however, as these accounted for just over a third of all cases (35%, *n*=69), whereas multimodal responses with both verbal and nonverbal components were slightly more frequent (37.6%, *n*=74). Fully verbal responses without relevant nonverbal behavior were relatively infrequent (18.8%, *n*=37).

# **5.2 Deferring the recruitment**

Recruitment, by definition, involves the relevance of practical action in the here and now of the interaction and not proposals for some other time and place. The immediacy of recruitment poses a practical problem for participants who are already engaged in a course of action when a move to recruit them comes. The practical problem of multiple involvements (Toerien & Kitzinger 2007; Raymond & Lerner 2014) or multiactivity (Mondada 2011; Haddington et al. 2014) is one for which participants have practiced solutions. In this and the next section, we will review two practices that recipients of recruiting moves use to manage the emergence of multiple involvements.

The first is deferring the recruitment. Rather than abandon or suspend the course of action the recipient is engaged in, the recipient can continue the course of action, using resources of the body, and defer the recruitment verbally, as in the following cases.

(20) BBQ 23:09


(21) RCE22b 08:23

	- 2 [↑th:i:[ng:s:?
	- 3 mar [yeah?
	- 5 ◩ #fig.13

Figure 13: Lisa (far right) asks Megan, who is using a computer, whether she knows how to use the projector.

4 Recruitment in English: A quantitative study

In (20), repeated in part from (17), after Alison asks James to stop the course of action he is engaged in – peeling carrots – to help her, he responds with a request to defer his assistance until the completion of his task. In this way, James tacitly commits to assisting Alison, but he prioritizes the progressive realization of his current task over that of the recruitment. Likewise, in (21), while Megan is engaged with her computer, Lisa asks her whether she knows how to use a video projector, which can be understood in the situation as a request for Megan to assist her. Megan first responds to the format of the recruiting move, confirming that she does know how, and then responds to its action implication, deferring her assistance to a later point in time. In sequence-structural terms, deferments initiate pre-second insert sequences that displace the relevant next action (i.e. the assisting action) from next position in the sequence (Schegloff 2007) and thereby constitute one method that recipients have to manage the emergence of multiple involvements.

# **5.3 Recruiting the recruiter**

A second practice that participants have to manage the emergence of multiple involvements is for a recipient of the recruiting move to invert the recruitment sequence, that is, to recruit the recruiter to perform the action him- or herself. In the following extract, repeated with additional detail from (19), a recipient of a request, Ben, recruits its speaker, James, to fulfill the request himself.

(22) RCE09 11:43

```
1 jam has anyone got a spare onion I can borr#ow.
2 ◩ #fig.14a
3 *(.)
4 ben *turns to left-->
5 jam half an onion.
6 ben *ye:ah,* #go in there.
7 *......*points----->>
8 ◩ #fig.14b
```
The request comes as Ben waits for his food to be heated in the microwave in front of him (see Figure 14*a*). In response to the request, Ben rotates his body and gazes and points into an adjacent room (see Figure 14*b*). He confirms verbally that he has an onion for James and then directs him to go into the adjacent room to retrieve it. In principle, Ben could have walked into the adjacent room, retrieved the onion, and given it to James. That is, one solution to the problem of multiple

### Kobin H. Kendrick

Figure 14: Ben stands (white shirt) in front of the microwave as he waits for his food to be heated. In response to the recruiting move by James, Ben turns and points into the adjacent room as he recruits James to fulfill the recruitment on his own.

involvements is for a recipient of the recruiting move to abandon a course of action. By recruiting the recruiter, however, Ben neither abandons the course of action he is engaged in nor defers the recruitment until the course of action is complete; he suspends one course of action (the preparation of the meal) in order to recruit James to pursue the second course of action in his stead (cf. Blythe, Chapter 7, §4.2.2).

# **6 Discussion**

This chapter has presented a quantitative study of recruitment phenomena identified in a sample of video recordings of everyday interactions among speakers of English. Using an operational definition of recruitment designed for a crosslinguistic comparison (see Chapters 1–2), the present study has documented diverse forms of linguistic and embodied action that participants use to recruit the assistance of others. As we have seen, visible bodily actions were common; about half of all recruiting moves in the dataset included a relevant visual component. Recruiting assistance through visible bodily action alone, however, was relatively rare. The most frequent visible bodily actions observed in recruiting moves involved manual actions, such as pointing, reaching out, or holding out an object to be taken, which reflects the high proportion of hand-to-hand object transfers in the dataset. Of the grammatical formats observed in recruiting moves, interrogatives were the most frequent and exhibited the greatest variation, with multiple recurrent subformats (see Fox & Heinemann 2016; 2017). In contrast, imperatives, the second most frequent format, showed relatively little structural

### 4 Recruitment in English: A quantitative study

variation. Regardless of format, recruiting moves overwhelmingly elicited cooperative responses, a statistical trend that reflects the general preference for agreement in interaction (Sacks 1987) and testifies to the fundamentally prosocial nature of human behavior – one important finding to emerge from the comparative study (Floyd et al. 2018).

The operational definition of recruitment developed for the cross-linguistic comparison, to which the present study is but one contribution, differs in important respects from the articulation of the concept by Kendrick and Drew (2016; see also Drew & Kendrick 2018). Our articulation was rooted in the observation that offering and requesting, which had been understood as distinct forms of action, each with its own grammatical formats and sequential environments (see, e.g., Curl 2006; Curl & Drew 2008), in fact have an organizationally symbiotic relationship (Kendrick & Drew 2014). This observation, which itself has antecedents in previous research (Schegloff 1995; see also Heritage 2016), prompted us to collect and analyze not only requests in all forms, as indeed the present study has done, but also all *offers* and all actions that systematically occasion them, whether delivered through talk or embodied conduct. It became clear that the organizational symbiosis between offering and requesting centers on the recognition of troubles in the realization of practical courses of action and the alternative methods available to participants to resolve them (Kendrick & Drew 2016). Offers and requests, we observed, differ in two principal ways: who initiates the recruitment of assistance and who generates a possible solution to the trouble. With a request, the one who experiences the trouble, Self, formulates a possible solution for an Other to perform and thereby initiates the assistance (e.g. *move over a little, can you?* in Extract 4). With an offer, it is the Other who formulates a possible solution and initiates the assistance (e.g. *you want that*, Kendrick & Drew 2016: 7). The two actions thus involve an inversion of social relations (e.g. who initiates and who responds) and interactional contingencies (e.g. who generates the solution), yet both constitute methods for the recruitment of assistance. The cross-linguistic comparison, however, has *its* roots an investigation of requesting across cultures, one vestige of which is a somewhat equivocal treatment of offers (Chapter 2).

Another difference in the two articulations concerns the use of the term "recruitment" itself. Having conceptualized offers and requests as alternative methods by Self and Other for the resolution of troubles, we came to recognize them as parts in a complex social organization of action, which we termed *the organization of assistance* (Kendrick & Drew 2016). Recruitment – that is, one's having been recruited – is central to the organization of assistance. But it should not

### Kobin H. Kendrick

be understood as a category of action, one that somehow subsumes offering and requesting within it. Indeed, it was the very tendency towards the analytic conflation of interactionally distinct actions into categories that we meant to disrupt with the concept of recruitment. As we see it, recruitment does not refer to a type of social action, but rather to an outcome or effect that participants have alternative methods to achieve (Kendrick & Drew 2016: 2). In the language of speech act theory, recruitment is more akin to a perlocutionary effect (Austin 1962) than a class of illocutionary force (Searle 1976).

To illustrate the distinction and its ramifications, consider a recurrent source of trouble that routinely impedes practical courses of action and two alternative methods to resolve it. If a course of action requires an object for its completion (e.g. the lighter in Extract 1 or the bag in Extract 2), the absence of the object will necessarily disrupt the progressivity of the course of action. Among the set of methods that participants have to resolve such troubles interactionally are requests (e.g. Extract 2) and visible searches of the environment, which may occasion offers of assistance (see Drew & Kendrick 2018). In both cases, we would say that the Other provides assistance, whether solicited or volunteered, and has therefore been recruited. We would not say, however, that requests for objects and visible searches of the environment are themselves "recruitments". To do so would conflate analytically distinct forms of action into one conceptual category and thereby obscure the systematic, interactionally-relevant differences between them.

At risk of belaboring the point, consider the ways in which requests for objects differ from visible searches of the environment (cf. Kendrick & Drew 2016: 10– 11). First, the resources with which the participants construct the actions differ completely. A request specifies a solution to a trouble through linguistic forms or communicative gestures. In contrast, a visible search is in the first instance recognizable as an instrumental visible bodily action. Second, who initiates the sequence that comes to resolve the trouble differs. With a request Self initiates the sequence whereas with a visible search it is Other who does so. Third, the two actions differ in who generates the solution to the trouble. A request for an object formulates a solution for Other to perform whereas a visible search requires that Other recognize the trouble and generate a solution independently. Fourth, how the trouble manifests and hence becomes recognizable differs. With a visible search Self's actions embody the trouble whereas with a request the nature of the trouble is left implicit. Fifth and finally, the actions also differ in whether they establish a normative obligation for assistance as a response. A request for an object initiates an adjacency pair sequence in which assistance is

### 4 Recruitment in English: A quantitative study

a conditionally relevant response whereas a visible search for the environment does not.

On what basis, then, could one say that requests for objects and visible searches of the environment are instances of the same action or the same type of action? What unites them is not a similarity of *action*. A participant who searches for an object is not performing the same action, or even the same kind of action, as one who asks a co-participant for it. What unites them is the outcome they may achieve: the recruitment of assistance and the resolution of a trouble. If a category of "recruitments" exists, it is by virtue of this common interactional outcome, not a similarity in the methods participant use to arrive at it. The same argument holds equally for other methods of recruitment. Reports of troubles, for example, are analytically and interactionally distinct from requests (Kendrick & Drew 2016), and neither should be conflated into a single category of action, though each has its place in the organization of assistance.

Terminology aside, research on recruitment marks a shift of analytic focus away from singular actions (e.g. requests) and theoretical categories of action (e.g. directives). Rather than begin with an action and examine its implementation, research on recruitment begins with a social organizational problem – how do participants in interaction recognize and resolve troubles that emerge in practical, embodied courses of action? – and investigates its various solutions, the recruitment of assistance being one. This mode of analysis, which has its roots in classic conversation analytic research on the organization of interaction, is generic and widely applicable to the study of action.

# **7 Conventions for multimodal transcription**

Embodied actions are transcribed according to the following conventions developed by Mondada (2014).


### Kobin H. Kendrick


# **References**


4 Recruitment in English: A quantitative study


### Kobin H. Kendrick


4 Recruitment in English: A quantitative study


### Kobin H. Kendrick


4 Recruitment in English: A quantitative study


### Kobin H. Kendrick


# **Chapter 5**

# **The recruitment system in Italian**

# Giovanni Rossi

Department of Sociology, University of California, Los Angeles

This chapter describes the resources that speakers of Italian use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Italian, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Italian with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

Social life would not be called such if there were not a system for people to get one another's help. Whatever their language and culture, people need others to get by in the small and big practicalities of everyday life, be it passing food, moving a heavy object, or doing some other chore. This chapter documents the main practices that speakers of Italian use to recruit assistance and collaboration from others, as observed in video recordings of naturally occurring interaction, analyzed as part of the comparative project reported on in this volume. After a brief description of the Italian language (§1.1) and of the data used for the study (§1.2), I begin by illustrating the basic structure of recruitment sequences (§2). I then

Giovanni Rossi. The recruitment system in Italian. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 147–201. Berlin: Language Science Press. DOI:10.5281/zenodo.4018378

### Giovanni Rossi

survey the nonverbal and verbal practices used to design recruiting moves (§3), including pointing gestures, imperatives, different types of interrogatives and declaratives, and additional verbal elements. I then examine responding moves, focusing on how their design is fitted to that of the recruiting move (§4). I also discuss the occurrence and import of acknowledgment in third position (§5) and the role of social asymmetries (§6). The conclusion situates the findings in light of the cross-linguistic perspective adopted in the volume.

# **1.1 The Italian language**

Italian is a Romance language spoken by over 60 million people in Italy, Southern Switzerland, and by migrant communities in several other countries, the largest of which are found in the United States, France, and Canada (Lewis et al. 2014). While being characterized by a profusion of geographical variation, the Italian language has certain core features that are shared across regional varieties. Verbs inflect for person, number, tense, and mood. Nouns, pronouns, adjectives and articles inflect for gender and number. Subject pronouns are normally dropped, though they can be maintained for emphasis or contrast. Word order is flexible, but the basic order is Subject Verb Object (SVO).<sup>1</sup>

Of particular interest for the purposes of this chapter is the distinction among the three main sentence types identified by linguists cross-linguistically: imperatives, declaratives, and interrogatives (Sadock & Zwicky 1985; König & Siemund 2007; Aikhenvald 2010). In Italian, imperatives are distinguished from other sentence types by morphology and syntax, the rules of which are explained in §3.3.2. At the same time, there are generally speaking no morphosyntactic means for distinguishing declaratives from polar (yes/no) interrogatives. While it is commonly held that intonation compensates for this (e.g. Gili Fivela et al. 2015), interactional research urges caution in claiming straightforward mappings between intonation and polar questions (Rossano 2010). That said, recent work also provides evidence for the association of distinct intonation contours with specific types of polar questions, particularly questions involved in other-initiation of repair and related actions (Rossi 2015a; 2020). More evidence for the role of intonation in marking interrogative utterances in Italian comes from findings discussed later in this chapter (§3.3.3).

There is a growing body of studies on the Italian language in social interaction, including studies of family life and socialization (e.g. Sterponi 2003; Fatigante

<sup>1</sup> For more comprehensive descriptions of the grammar and sound patterns of Italian, see Lepschy and Lepschy (1988), Bertinetto and Loporcaro (2005), Maiden and Robustelli (2007).

### 5 The recruitment system in Italian

2007; Arcidiacono & Pontecorvo 2010; Pauletto 2017), storytelling (Monzoni & Drew 2009), medical interaction (e.g. Pino & Mortari 2012; Mortari & Pino 2014), and basic domains of social organization such as the question-answer system (Rossano 2010) and gaze behavior (Rossano 2012). Recent research has explored the linguistic design of social actions such as invitations (Margutti & Galatolo 2018). A study by Galeano & Fasulo (2009) has looked at request sequences between parents and children, including the use of address terms, preliminary questions, forms of requesting that are more or less coercive, the role of normative reasoning, and the structure of sequences of "concatenated" requests. These themes resonate with those explored in the present study. Informed by previous and ongoing work in this area (Rossi 2011; 2012; 2014; 2015a,c; 2017; 2018; Rossi & Zinken 2016), this chapter provides an overview of requesting behavior in informal interaction among adult speakers of Italian as part of the broader phenomenon of recruitment.

# **1.2 Data collection and corpus**

The video corpus on which this research is based was constructed in accordance with a set of guidelines developed by and for the members of the comparative project reported on in the volume (see Chapters 1–2). The video recordings were made between 2009 and 2013 in several locations within the province of Trento and the urban area of Bologna, in northern Italy. The interactions recorded were all informal, among family and friends, and involved not only casual conversation but also everyday activities such as cooking, having meals, and playing games. Participants received no instruction other than to go about whatever activity they were engaged in. From this corpus, I sampled 15 interactions for a total of 3.5 hours, yielding 221 recruitment sequences.

Conventions for transcription, glossing, and translation are explained in a dedicated section at the end of the chapter.

# **2 Basics of recruitment sequences**

As defined in Chapter 1, §4, a recruitment is a basic cooperative phenomenon in social interaction consisting of a sequence of two moves with the following characteristics:

**Move A:** participant A says or does something to participant B, or that B can see or hear;

### Giovanni Rossi

**Move B:** participant B does a practical action for or with participant A that is fitted to what A has said or done.

This is the basic structure and development of a recruitment sequence, an example of which is given in (1) below. Other details of what can happen, including what participant B can do in Move B to fulfill or reject the recruitment, are illustrated in later sections. In the transcripts, ▶ and ▷ designate Move A and Move B, respectively.

# **2.1 Minimal recruitment sequence**

Extract 1 exemplifies a typical recruitment sequence. Sergio and Plinio are washing the dishes. As Sergio finishes rinsing a baking pan, he turns to Plinio, who is wiping washed cutlery, and recruits his collaboration with an imperative request: *´PLInio; a´sciUga anche ˋQUEsta.* 'Plinio wipe this one too' (Move A). He then walks to Plinio and hands him the baking pan. In response, Plinio takes the baking pan and begins to wipe it (Move B).

```
(1) CampFamLava_1518767
  1 (33.0)
▶ 2 ser ´PLInio; a´sciUga anche ↘QUEsta. ((shakes baking pan over sink))
         NAME dry-IMP.2SG also this-F
         Plinio wipe this one too
  3 ((walks with baking pan to Plinio))
▷ 4 pli ((takes baking pan from Sergio))
  5 ((sets baking pan on counter and begins to wipe it))
```
In Move A, Sergio uses an imperative, a verbal form that is intimately connected to the process of recruitment by virtue of its semantics, which encodes the speaker's attempt to get another to do something (Lyons 1977: 746–748; Sadock & Zwicky 1985: 170–171). One of the properties of this verbal form is that it anticipates only the fulfillment of the recruitment, which is what Plinio does in Move B. In cases like (1), the recruitment sequence unfolds as an adjacency pair (Schegloff 1968; Schegloff & Sacks 1973), where the fulfillment of a practical action by participant B is normatively expected after what participant A says or does. In other cases, participant B's cooperation is not obliged by the recruiting move but rather occasioned by it, meaning that its absence may not be sanctionable or accountable in the same way.

5 The recruitment system in Italian

# **2.2 Non-minimal recruitment sequence**

People are often successful in recruiting others at the first go. But there are times when a first attempt fails, either because the recruiting move is not heard, seen, or understood, or because it is ignored. In yet other cases, the recruitment may be rejected. Extract 2 gives an example of this, illustrating how a first attempt at stopping someone from doing something is pursued with further attempts, generating a non-minimal sequence.

During a family dinner, Luca picks up a piece of the dessert before everyone has finished the main course (line 2). Olga, who is sitting across the table from him, notices this behavior and tries to rectify it by saying *<<h,f>in´tAnto il ˆdOlce si mangia mia aˋDESso>.* 'by the way the dessert is really not to be eaten now' (line 3), then adding *ˋDOpo.* 'later' (line 7). Instead of complying, Luca holds the pastry close to his mouth (line 8) and expresses his resistance with a "purse hand" gesture (line 10, Figure 1). This leads Olga to pursue the recruitment with further attempts.

(2) PranzoAlbertoni01\_1837927 1 +(2.1) +(0.9) 2 *luc* +picks up pastry +removes wrapper--> ▶ 3 olg <<h,f>in´tAnto il ˆdOlce si mangia +mia a↘DESso.> meanwhile the dessert IMPS eat-3SG PTC now by the way the dessert is really not to be eaten now 4 *luc* --------------------------------------->+raises pastry to mouth--> 5 (0.1)\*(0.1) 6 \*gazes up at Olga-->> ▶ 7 olg ↘DO+po. after later ▷ 8 *luc* -->+holds up pastry--> 9 (1.8)Δ(0.2)# ▷ 10 Δmakes "purse hand" gesture--> 11 ◩ #Figure 1 ▶ <sup>12</sup> olg puoi ^METterlo ↘lÀ Δ 'l dolce. ((points at tray)) can.2SG put=INF=3SG.ACC there the dessert can you put the dessert {back} there 13 *luc* --------------------------->Δ

### Giovanni Rossi

	- 18 (0.6)
	- 19 luc anche se mi piace però\_ even if 1SG.DAT please-3SG but though I do like it but
	- 20 (0.1)+(1.1)
	- 21 ---->+eats pastry-->>
	- 22 olg lo sai cos'è\_ ((points at large cake)) 3SG.ACC know-2SG what=be.3SG do you know what it is

Figure 1: Frame from Extract 2, line 11. Luca challenges Olga's first attempt to stop him from having dessert with a "purse hand" gesture (≈ 'what's the problem⁈')

Olga's first attempt to stop Luca from having dessert ahead of time is unsuccessful. Luca first shows non-compliance by bringing the pastry to his mouth

### 5 The recruitment system in Italian

(line 4), and then goes on to express overt resistance with a "purse hand" gesture (Poggi 1983; Kendon 1995). This emblematic gesture, where all the fingers are drawn together so as to be in contact with one another at the tips, may be roughly translated here as 'what's the problem⁈'. Olga pursues the recruitment by changing strategy, using an interrogative form instead: *puoi ˆMETterlo ˋlÀ 'l dolce.* 'can you put the dessert {back} there' (see §3.3.3 below). This second attempt is also unsuccessful. Luca continues his challenge by soliciting an account (*perché.* 'why', line 15). Olga then restates the norm of behavior invoked a moment earlier (*perché si mangia dopo;* 'because it is to be eaten later') and adds an enticement (*=te dao questo;* 'I'll give you this'), referring to a large cake (in the foreground of Figure 1) that will be the dessert's highlight. However, Luca continues to push back by saying that he is not going to eat from that large cake (line 17). A moment later, he goes ahead and eats the pastry he picked up a the beginning of the extract (line 21).

The development of the sequence shows the sustained relevance of compliance with the recruitment initiated by Olga, which she pursues with multiple attempts. This leads to an expansion of the basic two-part structure illustrated in §2.1 above.

# **2.3 Subtypes of recruitment sequence**

The phenomenon investigated in this project encompasses a range of social-interactional events that have in common the mobilization of someone's practical action. As discussed in Chapter 2, §6, most recruitment sequences fall into four broad subtypes. The two examples examined in the previous section illustrate two: the provision of a service (1), where someone is recruited to perform a manual task, and an (attempted) alteration of trajectory (2), where someone is recruited to stop or change an ongoing behavior. I now illustrate the two remaining subtypes, starting with object transfers.

Some time before (3), Furio has offered Sara a piece of the banana he is eating. In line 2, she asks him for one more piece.

(3) BiscottiPome01\_2168783

1 (1.3)

	- 3 (4.4)

### Giovanni Rossi

```
▶ 4 sar per fa↘VOre.
          for favor
          please
  5 (1.4)
▷ 6 fur ((gives Sara one more piece of banana))
          ((10 seconds not shown))
  7 sar grazie.
          thanks
```
Sara's initial recruiting move is followed by a long silence (line 3). She then pursues the request with the formulaic *per faˋVOre.* 'please'. After another, shorter silence, Furio eventually fulfills the recruitment by giving Sofia one more piece of banana.

The fourth subtype of recruitment sequence is trouble assistance, where participant B steps in to help in response to participant A's current trouble. In (4), Sergio is styling Greta's hair. During the process, a strand of dye-soaked hair rolls down on Greta's face (line 3, Figure 2*a*), causing her to gasp (line 5). As Sergio realizes what has happened (line 8), he promptly gathers the strand of hair and folds it back over Greta's head (lines 9–12, Figure 2*b*).

```
(4) Tinta_1445710
 1 +(2.3)Δ(0.4)#
 2 luc +kneads hair-->
 3 Δstrand of hair rolls down on Greta's face-->>
 4 ◩ #Figure 2a
▶ 5 gre °HH*HH†H ((gasps))
 6 *tilts head-->>
 7 †raises hand to face-->>
 8 ser <<f,h>↘Uu:+:.>
              oo::
▷ 9 +gathers strand of hair and folds it back up-->
 10 (0.3)#(0.8)+
 11 ◩ #Figure 2b
 12 ---------->+
 13 ser scusa_
        sorry
```
### 5 The recruitment system in Italian

(a) A strand of dye-soaked hair rolls down on Greta's face (line 4).

(b) Sergio helps Greta by gathering the strand of hair to fold it back up (line 11).

Figure 2: Frames from Extract 4, illustrating a case of trouble assistance.

Greta's gasp is seemingly produced as an instinctive reaction to the sudden discomfort of dye-soaked hair rolling down onto her face and possibly into her eye; it is arguably not intended or designed to elicit Sergio's help. What this shares with other recruiting behaviors, however, is that it makes apparent Greta's need for assistance, instigating Sergio to step in.

# **3 Formats in Move A: The recruiting move**

It has long been noted that people use a wide range of strategies to get others to do things (see Chapter 1, §2). In the framework of this project, this means looking at the resources that are available to people to design Move A, the recruiting move. Most of the literature on this topic focuses on verbal formats. But in face-to-face interaction, recruiting moves often involve a composite of verbal and nonverbal elements, and may also be fully nonverbal. This section surveys the range of options available to Italian speakers.

# **3.1 Fully nonverbal recruiting moves**

Fully nonverbal recruiting moves in Italian are much less frequent than ones involving language, making up only 10% of the cases (*n*=22/221). One reason for this

### Giovanni Rossi

is that fully nonverbal recruiting moves normally occur in relatively constrained contexts. Extract 5 provides an example.

Four friends have just finished playing a card game. Flavia announces how many points her team has to 'pay' (lines 1 and 4), that is, subtract from the previous score. A moment later, Bianca initiates a recruitment to retrieve the notepad on which the scores are kept.

```
(5) Circolo01_402024
 1 fla e ades te pago zinquantazin[que,
        and now 2SG.DAT pay-1SG fifty_five
        and now I'll pay you fifty-five
 2 bia [sì.
                                     yes
 3 (0.6)
 4 fla <<rall>cinquanta?> cinque,=
             fifty five
             fifty- five
 5 sil =<<p>sie+te.>
            seven
 6 bia +puts cards back on top of deck-->
 7 (0.1)+(0.3)*(0.8)+(0.5)+Δ#(0.2)Δ(0.6) +(0.6)+(0.2)Δ(0.7)
 8 bia *turns and gazes at notepad-------------------------------->
▶ 9 ---->+ +.....+points at notepad+,,,,,+
▷ 10 sil Δturns Δreaches for notepad Δpasses notepad-->
 11 ◩ #Figure 3
 12 +(0.1)Δ
 13 bia +puts glasses on-->>
 14 sil ----->Δ
```
Shortly after approving Flavia's count (line 2), Bianca turns to her right and gazes over at the notepad across the table (line 8), which is out of her reach but within Silvia's (Figure 3). Bianca then points at the notepad (line 9); Silvia turns toward the notepad, reaches for it, and passes it to Bianca (line 10). Silvia can be expected to comply with Bianca's request in that it is made in direct contribution to a shared activity that Silvia is participating in (see §3.3.2 below).

The action recruited here is embedded in the ordinary development of the ongoing activity (Rossi 2014). At the end of every game, the points for each team are counted and the scores updated in the game's record, which is kept on the

### 5 The recruitment system in Italian

Figure 3: Frame from Extract 5. Bianca points at the notepad; Silvia turns toward it (line 11).

notepad. For all the previous games, Bianca has been responsible for updating the record. So when Flavia marks the end of the count by repeating her team's score (line 4), the projectable next action is Bianca writing it down. This is an environment in which Bianca gazing and pointing at the notepad is all that is needed for Silvia to understand that she is being recruited to pass it.

For a fully nonverbal recruiting move to be successful, the action being recruited needs to be projectable. A common source of projection in informal interaction is the structure of an activity, which sets up expectations about people's actions within the activity (see Levinson 1979; Robinson 2013, among others). The structure of an activity is a form of common ground (Clark 1996: 93) that can and should be relied upon by participants when recruiting one another's collaboration. When the passing of an object is an expectable contribution to a joint activity, as in (5), participant A can minimize the recruiting move by simply making known the wanted object to participant B and preparing to receive it. Such minimization is motivated by fundamental principles of human communication (Grice 1975; Levinson 2000). These principles provide a common basis for the production of fully nonverbal recruiting moves across languages (see Kendrick Chapter 4, §4.1.3; Zinken, Chapter 8, §3.1; Baranova, Chapter 9, §3.1; Dingemanse Chapter 10, §3.4).

### Giovanni Rossi

# **3.2 Nonverbal behavior in composite recruiting moves**

Nearly half of all recruiting moves in the Italian data involve a combination of verbal and nonverbal elements (47%, *n*=96/206).<sup>2</sup> The types of nonverbal behavior that co-occur with language in these composite recruiting moves are given in Table 1.


Table 1: Nonverbal behaviors in composite recruiting moves (*n*=96).

The majority of nonverbal behaviors fall into three basic types identified in the comparative project (see Chapter 2, §6): pointing, holding out an object for someone to take and do something with, and reaching out to receive an object. But there are two other types that figure prominently in the Italian data. One is placing an object in a meaningful location for someone to do something with (Clark 2003: 249-50); the other is iconic gestures that depict the shape of the target object or action. In what follows, I focus on the use of pointing and iconic gestures, leveraging previous research on co-speech gesture to shed light on its role in recruitments.

Enfield et al. (2007) have shown that, when used by speakers to refer to localities, pointing gestures can take two main forms: "big" and "small". Big points are articulated with the whole arm, usually with head and gaze also oriented to the target. Small points are reduced in size and articulatory effort, with the head and gaze less frequently oriented to the target. Enfield et al. argue that the two pointing forms are functionally distinct. Big points are used when the information in the gesture is the primary, foregrounded component of the message, while small points are used when the speech is informationally foregrounded and the gesture adds to it in the background.

<sup>2</sup> Fifteen cases were excluded from the count as the recruiter was momentarily off camera or hidden by another participant at the time of the recruiting move.

### 5 The recruitment system in Italian

The argument is that big points occur when the location of a referent is focal (see Lambrecht 1994: chap. 5). In these cases, the speech typically contains a deictic element (such as 'here', 'there', 'this', 'that') but it is the gesture that supplies the key information. So it needs to be maximally accurate. Small points, on the other hand, occur in a variety of contexts where a referent seems "likely but not certain to be recognizable" (Enfield et al. 2007: 1730). In these cases, the speech should be sufficient for reference to be secured, but it might not. Speakers therefore strike a balance between the risks of "under-telling" and "over-telling" by adding a bit of extra information in their gesture.<sup>3</sup>

When we look at pointing gestures in recruiting moves, the Italian data suggest that their form is sensitive to the distinctions proposed by Enfield et al. (2007). Compare the following two cases, taken from the same interaction, where a group of friends are making cocktails.

When (6) begins, Silvio has just stopped pouring soda in a carafe and is proceeding to add gin. However, Bino and Fabio alert him to the fact that the quantity of soda is not yet sufficient (lines 1–2). By so doing, they recruit Silvio to add more soda (line 5). As Silvio begins to do this, Fabio produces another recruiting move – the one in focus here – aimed at further adjusting the trajectory of Silvio's actions. He tells Silvio to pour the soda *ˋQUA;* '{in} here', that is, into another container. The location of 'here' is supplied by a "big" pointing gesture (Figure 4*a*).

```
(6) MasoPome_2058866
```

```
1 bin non è mezza [bozza ancora.
         not be.3SG half bottle yet
         it's not yet half bottle
  2 fab [sì non è mezza bozza.
                        yes not be.3SG half bottle
                        right it's not half bottle
  3 (0.2)
  4 Δ(0.4)*(0.4) Δ(0.5) Δ(0.5)
  5 sil Δputs gin down Δpicks soda back up Δpours soda into carafe-->
  6 fab *gazes at other container----------------------------->
▶ 7 fab ↑´bU+ttalo gi+ù ↘QU#A;
          throw-IMP.2SG=3SG.ACC down here
          pour it {in} here
  8 +....................+points at other container-->
```
<sup>3</sup>On under-telling and overt-telling, cf. Grice (1975), Levinson (2000), Schegloff (2007a: 140).

### Giovanni Rossi

9 ◩ #Figure 4a 10 (0.2)+\*(0.5)+(0.4)+(0.1)Δ(0.1) 11 *fab* ---->+,,,,,,+ +....................--> 12 ----->\* 13 *sil* ----------------------->Δputs soda down--> ▶ <sup>14</sup> fab ´bUttalo \*giù +↘L#Ì; throw-IMP.2SG=3SG.ACC down there pour it {in} there 15 ..........................+points again at other container--> 16 \*gazes again at other container--------> 17 ◩ #Figure 4b ▷ 18 sil <<cresc>aspetta:Δ:+\*:.> wait-IMP.2SG wait 19 --------------->Δ 20 *fab* ----------------->+,,,--> 21 ------------------>\* 22 (0.1)+ 23 *fab* ,,,,,+

While saying *↑´bUttalo giù ˋQUA;* 'pour it {in} here', Fabio produces a big point, with the arm stretched out and the finger fully extended to pick out the other container with precision, his gaze fixed on the referent (lines 6–9, Figure 4*a*). When Silvio does not immediately comply, Fabio repeats the same composite utterance, changing only the deictic form ('here' → 'there') and using another big point (lines 11, 14–17, Figure 4*b*).

In Enfield et al.'s terms, Fabio's recruiting move here has a "location focus", that is, it is about where Silvio should pour the soda and designed to direct him to another container which he has apparently not considered using for the current purpose. A big point here is fitted to locating and identifying the target container.

The second case involves Fabio initiating a recruitment with an analogous verbal form, an imperative, which is again coupled with a pointing gesture. This time, however, the point is "small".

When (7) begins, the participants are debating over the qualities of vodka and gin, the two liquors they have on the table to make cocktails. Fabio and Silvio

### 5 The recruitment system in Italian

(a) Fabio uses a "big point" while saying *↑´bUttalo giù ˋQUA;* 'pour it {in} here' (lines 7–9).

(b) Fabio uses another "big point" while saying *´bUttalo giù ˋLÌ;* 'pour it {in} there' (lines 14–17).

Figure 4: Frames from Extract 6, illustrating a "big" pointing gesture accompanying an imperative request.

argue that gin is 'disgusting' and 'tastes like shit' (lines 3–4), which is reason for mixing it with a larger quantity of soft drink. Bino's subsequent repair initiation *come fa cagare.* 'what do you mean it tastes like shit' (line 5) projects his disagreement with the assessment (see Rossi 2015a: 279; Raymond & Sidnell 2019). To settle the issue, Fabio initiates a recruitment for Bino to taste the gin.

```
(7) MasoPome_1912588
```

```
1 bin ne fa due in più questa del gin.
       PTV make-3SG two in more this of-THE gin
       this contains two {percent} more {alcohol} than gin
2 (0.6)
3 fab sì vabè ma:[: il gin fa schifo.
       yes PTC but the gin make-3SG disgust
       yes well but:: gin is disgusting
4 sil [sì ma il gin fa cagare è quello magari che:.
                   yes but the gin make-3SG shit-INF be.3SG that maybe REL
                   yes but gin tastes like shit – that's maybe what:
5 bin come fa cagare.
       how make-3SG shit-INF
       what do you mean it tastes like shit
```
### Giovanni Rossi

```
6 (0.1)*(0.2)
 7 fab *turns and gazes at Bino-->
▶ 8 fab ↑ˇPRO*va a +tastarne: +en # go+↘ZA*T.+
         try-IMP.2SG to taste-INF=PTV one drop
         try taste some
 9 ---->*gazes at gin---------------------------->*gazes back at Bino-->
 10 +.............+points at gin+,,,,,,+
 11 #Figure 5a
 12 (0.2)
▶ 13 fab ↘TA#s[ta,
        taste-IMP.2SG
        taste
 14 #Figure 5b
▷ 15 bin [ma no così* liscio.
             but no like_this straight
             well not straight like this
 16 fab ---------------->*
```
The form of Fabio's pointing gesture here is quite different from what we have seen in (6). It is articulated with the lower arm only and with the finger not fully extended (Figure 5*a*). Also, instead of looking at the referent throughout, Fabio turns to Dino before speaking (line 7), then shifts his gaze to the gin, and then back to Bino before the end of the utterance (line 9). Fabio keeps looking at Bino also during the second imperative (*ˋTAsta,* 'taste', line 13), where notably he does *not* redo the pointing gesture (Figure 5*b*).

All this contributes to characterizing Fabio's gesture here as a "small point", conveying supplemental and possibly dispensable information. The recruitment here is not location focused; the goal is not to direct the recruitee to where he should put or do something, but rather to instigate action on a referent that has already been thematized (lines 3–5).

While there is only one bottle on the table containing straight gin (the glass bottle with the yellow label in Figure 5), a plausible alternative interpretation of Fabio's 'try taste some' is with reference to the gin mixed in the cocktail they have been making (contained in the green plastic bottles). In this context, a small point serves as an "informational safety net" (Enfield et al. 2007: 1734), available but inconspicuous, provided just in case the reference turned out to be ambiguous.

### 5 The recruitment system in Italian

(a) Fabio uses a "small point" while saying *↑ˇPROva a tastarne: en goˋZAT.* 'try taste some' (lines 8–11).

(b) Fabio does not repeat the gesture as he says again *ˋTAsta,* 'taste' (lines 13–14).

Figure 5: Frames from Extract 7, illustrating a "small" pointing gesture accompanying an imperative request.

I conclude this section on composite recruiting moves by illustrating the use of iconic gesture. Although less frequent compared to pointing, iconic gestures are approximately as frequent as the other main types of nonverbal behavior in the Italian data (see Table 1 above).<sup>4</sup> Extract 8 provides an example.

Before the extract begins, Rocco has unsuccessfully attempted to initiate a recruitment sequence with Loretta, who has been involved in a concurrent conversation. As the concurrent conversation ends with general laughter (line 1), Loretta finally answers Rocco's summons, clearing the way for Rocco's request.

(8) CampFamTavolo\_1803413


<sup>4</sup>This finding is consistent with experimental research on the relatively high frequency of iconic gestures by Italian speakers in other contexts (Campisi 2014).

### Giovanni Rossi


Figure 6: Frame from Extract 8, lines 3–5. Rocco makes an iconic gesture while saying *mi passe´rEsti un bicchier d'acqua norˆMAle;* '{would} you pass me a glass of plain water'.

Rocco's recruiting move includes an iconic gesture, with the thumb and index finger vertically aligned and kept at a distance (Figure 6). The gesture may repre5 The recruitment system in Italian

sent the size or height of a drinking glass, or metaphorically refer to the amount of water contained in it.

# **3.3 Verbal elements: construction types and subtypes**

In this section, I survey the range of options that Italian speakers have for designing the verbal component of recruiting moves. Italian speakers make use of all three main sentence types: imperatives, interrogatives, and declaratives. While imperatives are the most frequent, interrogatives and declaratives are also common (Table 2). The two cases of an "other" construction type feature the antecedent of a conditional sentence functioning as a main clause: (e.g. *questo se me lo mettete dentro* 'this one if you guys put it away for me').<sup>5</sup> Finally, Italian speakers also make use of utterances without a predicate, including noun phrases and single words. Since the use of such "minimal" utterances is sensitive to criteria related to those explained above for fully nonverbal forms (§3.1), I begin the analysis with these.


Table 2: Construction types in recruiting moves with a verbal component (*n*=199).

### **3.3.1 No predicate**

In §3.1, we saw that speakers do not use language when the action being recruited is projectable from the development of the ongoing activity (Rossi 2014). However, projectability is not an all-or-nothing dimension. Besides *fully* projectable actions, there are also *partially* projectable actions, some element of which cannot be anticipated by the recruitee and therefore needs to be verbally specified (Rossi 2015c: 54–57).

<sup>5</sup>This format is more frequent in other languages examined in this volume (see Dingemanse, Chapter 10, §3.2).

### Giovanni Rossi

In Extract 9, the card players we met in (5) are starting a new game. Bianca begins to deal the cards, giving out two at a time (lines 1–4). As the players have discussed previously in the interaction, dealing two cards at a time increases the chance that cards will cluster in combinations from the prior game. Here, however, Bianca has apparently forgotten about this.

```
(9) Circolo01_1948857
 1 Δ(0.8) Δ
 2 bia Δdeals two cardsΔ
 3 bia le robe se Δle fa senzaΔ: dove[rseΔle_]
       the things RFL 3PL.SCL do-3PL without must-INF-3PL.ACC
       things should be done without having to
 4 Δdeals two more cardsΔ Δdeals two more cards-->
▶ 5 fla [ˆUnΔa. ]
                                       one
 6 (0.4)
▶ 7 fla ˆUna.Δ
       one
 8 bia ---->Δ
 9 (0.1)Δ(1.3) Δ
▷ 10 bia Δtakes one card backΔ
```
Flavia initiates a recruitment for Bianca to alter the way she is dealing the cards by saying *ˆUna.* 'one' (line 5). This first iteration of Flavia's "naming" is simultaneous with Bianca dealing a third pair of cards (line 4). As Bianca is still in the process of doing this, Flavia repeats the recruiting turn (line 7). Bianca then complies by taking back one of the two cards she has just dealt.

In this environment, most elements of the recruited action are projectable: the target object (cards) and the action to be done with it (dealing). What is not projectable in light of Bianca's ongoing conduct is the object's quantity, which is what gets named.

Other no-predicate cases include nominal references to the object to be passed or manipulated (e.g. *coltello* 'knife') or to its location (e.g. *quell'altro* 'the other one') or destination. A no-predicate recruiting format allows the recruiter to verbally specify only what is necessary, leaving out what is not (cf. Mondada 2011; 2014; Sorjonen & Raevaara 2014).

### 5 The recruitment system in Italian

### **3.3.2 Imperatives**

Imperatives are the most frequent construction type used by Italian speakers to get others to do things (see Table 2). The Italian language has both morphological and syntactic means to distinguish imperatives from interrogative and declarative sentence types. Imperative endings are available for the second person singular of verbs in the main conjugation class (e.g. *parl-are* 'to speak', *parl-i* 'you speak' vs. *parl-a* 'speak!') and of certain irregular verbs. Another reliable cue, especially for morphologically ambiguous forms, is the position of clitic pronouns in the clause. In interrogatives and declaratives, clitic pronouns like *mi* 'to/for me' precede the main verb (*mi leggi un libro* 'you read a book for me'); in imperatives, they follow it (*leggimi un libro* 'read me a book').

Figure 7: Frame from Extract 1, line 2. Sergio says *´PLInio; a´sciUga anche ˋQUEsta.* 'Plinio wipe this one too' while Plinio is wiping washed cutlery.

We already encountered examples of imperatively formatted recruiting moves in the previous sections. Extract 1 exemplifies the typical environment for this construction type: before the recruitment sequence occurs, the participants have engaged in a joint activity or project (washing dishes) and the recruiting move is made made within this joint project to solicit an action that contributes to it (*a´sciUga anche ˋQUEsta.* 'wipe this one too') (Rossi 2012).

But previous examples also show that imperatives are not the only form occurring in such environments. Similar recruiting moves that further a joint project

### Giovanni Rossi

may also be formatted nonverbally (5) or without a predicate (9). In order to be understood, nonverbal and no-predicate recruiting moves require the full or partial projectability of the target action. The use of an imperative, on the other hand, is sensitive to the action not being projectable. Extract 1 again serves to illustrate this. When Sergio initiates a recruitment for Plinio to wipe the washed baking pan, Plinio is wiping cutlery (Figure 7). Wiping the baking pan is not a projectable next action at this point of the activity; it has to be "slotted into" what Plinio is currently doing (Rossi 2014: 318). This is grounds for using a clausal form that fully specifies the target action.

Consider another case, which can be directly compared against the nonverbal and no-predicate recruiting moves in (5) and (9). During the same card game, Flavia has just drawn a card that allows her to lay down a first combination (lines 1–2). Upon inspecting the cards played by Flavia, Bianca indicates a problem (line 4). She leans across the table and counts the cards while pointing at them (line 6). Then, after a brief pause, she tells Flavia *´mEti zo 'n altro ˋAMbo.* 'put down another double', which is needed to complete the combination. Moments later, Flavia fulfills the recruitment by laying down two sevens (line 11).

```
(10) Circolo01_677062
```

```
1 fla [una due tre quatro (che) te l'ho pescada? (.) to':?
           one two three four (CONN) 2SG.DAT 3SG.ACC=have-1SG draw-PCP INTJ
           one two three four – I drew it (.) here we go
  2 ((lays cards down in a new combination))
  3 cla ah [per-?
          oh because
          oh bec-
  4 bia [<<f,h>´NO:_> ((leans forward across table))
                   no:
  5 sil por[ca miseria.
          piggy misery
          holy cow
  6 bia [due quatro::_ ((points at cards))
             two four
             two four::
  7 (1.2)
▶ 8 bia ´mEti zo 'n altro ↘AMbo. ((keeps pointing at cards))
          put-NPST-2SG down one other double
          put down another double
  9 (2.5) ((Flavia looks at cards in her hand))
  10 fla de sete l' g'ho;
          of seven 3SG.ACC LOC=have-1SG
          I have one of sevens
```
### 5 The recruitment system in Italian

((10 seconds not shown))

▷ 11 fla ((lays down a double of sevens))

Bianca initiates the recruitment after Flavia has laid down an illegal combination of cards.The recruitment is aimed at solving a problem that has arisen during the game, but that was not projected by its structure. After Bianca first raises the problem (*<<f,h>´NO:\_>* 'no:', line 4), Flavia's silence indicates her uncertainty as to how to proceed. Also, the fact that Bianca needs to count the cards before she can instruct Flavia (line 6) shows that the next relevant action is hard to anticipate. Here, Bianca's pointing to the incriminated cards would not be enough for Flavia to understand what to do next (cf. Extract 5). The action being recruited needs to be fully articulated.

In sum, the imperative form is typically used to solicit actions that contribute to an already established joint project and that cannot be projected from its advancement (Rossi 2012; 2014). The imperative so used is usually bare and unmitigated (Rossi 2017). Other less frequent uses of the imperative are more likely to be mitigated with additional elements (see (§3.4).

### **3.3.3 Interrogatives**

Interrogatives are the second most frequent construction type after imperatives (see Table 2). As mentioned in §1.1, there are generally speaking no morphosyntactic means for distinguishing polar (yes/no) interrogatives from declaratives in Italian.<sup>6</sup> At the same time, recent research by the author has documented the association of distinct intonation contours with specific types of polar questions, particularly questions involved in other-initiation of repair and related actions (Rossi 2015a; 2020). This lends some support to the claim that intonation compensates, at least partly, for the lack of interrogative morphosyntax (e.g. Gili Fivela et al. 2015), though this should not be taken to imply a straightforward mapping between intonation and polar questions as a whole (Rossano 2010).

<sup>6</sup> In many areas of Italy, speakers may alternate or mix the national language with a local Romance vernacular (Maiden & Parry 1997). Regional Italian and vernacular are often inextricably interwoven in the speech of Italian speakers and both are integral parts of local Italian culture. In the province of Trento, where most my video recordings were made, the local Romance vernacular is the Trentino language. Unlike Italian, this language does have morphosyntactic means to distinguish between polar interrogatives and declaratives. This is due to the presence of subject clitics (Lusini 2013), which are positioned before the main verb in declaratives (e.g. *te gai* 'you have') and after the main verb in interrogatives (e.g. *ga-t* 'do you have'). See (18) and (22) for examples.

### Giovanni Rossi

The analysis of recruitments provides further evidence for the role of intonation in marking interrogative utterances in Italian. Recruiting turns that make relevant acceptance or confirmation in the form of a polar answer (see Extracts 8,11,12) are normally produced with either a rise-fall or a rise from low intonation contour. In the main variety of Italian spoken in my corpus – Trentino Italian – these intonation contours are of the same type found on requests for confirmation and questioning repetitions (Rossi 2015a; 2020), and are both distinct from the hat-pattern and fall contours that are instead found on imperative and declarative recruiting turns (Rossi 2011). These intonation contours fulfill a criterion of formal distinguishability between construction types in that they "form a system of alternative choices that are mutually exclusive" (König & Siemund 2007: 278). On this account, I refer to recruiting turns systematically produced with intonation contours associated with polar questions as interrogatives.

There are three main subtypes of interrogative used in recruiting moves in Italian, and they are found in different interactional environments. The most frequent subtype is what I refer to as the simple interrogative (Rossi 2015c: chap. 3), which can be rendered in English with 'will you x'. Unlike its English translation, however, the construction does not contain any modal verb but only an action verb inflected for second person, simply asking if the recipient is going to do something (lit. 'you x?'). The action verb is typically preceded by a first person dative pronoun *mi* 'to/for me' expressing that the action is directed to, or for the benefit of, the speaker. The use of this interrogative subtype is illustrated in the following example, where a group of friends are playing cards.

Before the extract begins, Franco has gotten himself a piece of paper towel from a cabinet next to him in order to blow his nose. As he finishes wiping his nose (line 1), he turns back to the table (line 3), reengaging in the game. This is the context in which Beata recruits him to get a piece of paper towel for her too.

(11) CampUniTaboo01\_172458

```
1 fra ((finishes wiping nose, folds paper towel, puts it into pocket))
 2 san è veramente comunque per[verso [( )
        be.3SG really anyway perverse
        anyway {that thing} is really perverse ( )
 3 fra [((turns back to table))
▶ 4 bea [mi b- ˆDAi
                                      1SG.DAT b- give-2SG
                                      {will} you b- give
 5 anche a ↓↓me un pezzo di <<creaky>↘scOt[tex.>
        also to 1SG.ACC a piece of paper_towel
        a piece of paper towel to me too
```
5 The recruitment system in Italian


Unlike the imperative and no-predicate recruiting moves examined above, the request here is not part of a joint project. For one thing, it is unrelated to the ongoing game. Also, it is made at a point when Franco has just completed his own individual course of action with the paper towel and repositioned his body to reengage in the game with the other players. Turning back to get another piece of paper towel requires him to disengage from the game again. Such relation of discontinuity typically goes together with the fact that the action being recruited is in the interest of the requester as an individual. Rather than contributing to a shared goal, the action benefits the recruiter alone (Rossi 2012; 2015c: chap. 3).

Another subtype of interrogative is *puoi x* 'can you x', a modal construction asking about the ability of the recipient to do something. Much like in simple interrogative sequences, actions recruited using *puoi x* 'can you x' typically involve a departure from what the recruitee is currently doing. At the same time, what distinguishes the usage of *puoi x* 'can you x' is an anticipation of the recruitee's unwillingness to comply (Rossi 2015c: chap. 4; cf. Zinken, Chapter 8, §3.3.4). In (2), for example, Olga asks Luca if he 'can put the dessert' back on the tray (line 12) after a first attempt to get him to do so, which he has resisted (lines 8–11). By using a *puoi x* 'can you x' interrogative, Olga recognizes the problematic nature of the recruitment and attempts to overcome Luca's unwillingness by appealing to his cooperativeness, or put another way, by persuading him (cf. Zinken & Ogiermann 2011: 280–282). This orientation is reflected also in Olga's subsequent use of an enticement: 'I'll give you this' (line 16).

The third subtype of interrogative is *hai x* 'do you have x', a construction asking if the recipient is in possession of an object. Extract 12 gives us an example. A group of people are hanging out in the living room. Snacks and drinks are on the table, including beer and juice, but not milk.

Giovanni Rossi

```
(12) DopoProve09-2_293350
  1 mag Ada,
         NAME
         Ada
  2 ada ((looks up))
▶ 3 mag ^HAi un goccio di ^lAtte.
         have-2SG one drop of milk
         do you have a bit of milk
  4 (0.5)
▷ 5 ada mm hm[::? ((nods))
         mm hm::
  6 min [vuoi il succo? ((to Magda))
               want-2SG the juice
               do you want juice
  7 (0.5)
  8 mag [no grazie ( )
          no thanks ( )
▷ 9 ada [((stands up and walks to kitchen))
```
In line 1, Magda addresses Ada – the group's host – and asks if she has milk, which is not among the beverages available on the table. Ada responds with a positive polar token (*mm hm::?*, line 5), accompanied by nodding, and shortly after proceeds to fulfill the request (line 9).

The availability of an object is a precondition – a material and practical prerequisite – for the object to be passed or utilized by someone. In recruitment sequences, the function of a *hai x* 'do you have x' interrogative is to check an object's availability when this is uncertain, for example because the object is not visible (see also Floyd, Chapter 3, §3.3.3; Enfield, Chapter 6, §4.3.1). This subtype of interrogative, in other words, works as a pre-request (see Rossi 2015b and references therein). If the target object is available, the recruitee often responds by fulfilling the projected request immediately (see also Fox 2015), as in (12). Other response affordances of this form are illustrated in §4.2.2 below.

Regardless of subtype, interrogative recruiting moves make fulfillment of the recruitment contingent upon the recruitee's response. This distinguishes them from imperative recruiting moves, which instead assume compliance. One reason for a recruiter not to assume compliance is that the action being recruited is unrelated to what the recruitee is doing and, rather than contributing to a joint project, serves an individual goal of the recruiter. This is when a simple interrog5 The recruitment system in Italian

ative is normally used. If, in addition, the recruiter anticipates that the recruitee may be unwilling to comply, they can select a semantically and syntactically more complex interrogative – *puoi x* 'can you x' – to recognize the problematic or delicate nature of the recruitment. Yet another reason for not assuming compliance is when a precondition for recruitment is uncertain. When the object to be passed or utilized may not be available, recruiters can use a *hai x* 'do you have x' interrogative to check on this.

### **3.3.4 Declaratives**

Declarative recruiting moves are as frequent as interrogative ones in Italian (see Table 2) and, like interrogatives, fall into three main subtypes. The first subtype is personal modal declaratives, which include constructions expressing a person's obligation or necessity to do something, such as *devi x* 'you have to x' or 'you must x'.

In (13), Sofia and Furio are making cookies in Furio's kitchen. Before the extract begins, Sofia has left the table to weigh some of the ingredients on a scale. In line 1, she complains that she is having trouble turning the scale on.

```
(13) BiscottiMattina01_3000055
```

```
1 sof non si accende? non so, ((fiddles with scale))
          not RFL turn_on-3SG not know-1SG
          it doesn't turn on – I don't know
▶ 2 fur devi clic¯cAre:: [´plUrime ˇVOLte.
          must-2SG click-INF multiple times
          you have to press:: multiple times
▷ 3 sof [mhm. ((presses button again))
                              mhm
  4 fatto;
          done
  5 fur devi convincerla.
          must-2SG convince-INF=3SG.ACC
          you have to persuade it
```
Furio's recruiting move is responsive to Sofia's trouble. As she fiddles with the scale and signals a problem, Furio instructs her how to solve it. Sofia then complies and announces that she has succeeded. Note that, after the recruitment sequence is complete, Furio uses again the same *devi x* 'you have to x' form to reiterate how Sofia should handle the scale (line 5). While still connected to what Sofia has just done, the instruction in this position no longer refers to a here-and-now action and acquires broader temporal scope or applicability. This

### Giovanni Rossi

follow-up by Furio sheds light on the social-interactional import of *devi x* 'you have to x' relative to other recruiting formats. Similarly to the imperative, this form can be used to solicit a contribution to an undertaking that has already been committed to by the recruitee (see §3.3.2). However, while an imperative directs the recruitee to perform a here-and-now, one-off action, a *devi x* 'you have to x' declarative imparts an instruction that transcends the local circumstances and is applicable in the future (see Parry 2013; Raevaara 2017; cf. Zinken 2016: 117–130). Pressing the scale's button multiple times to turn it on is relevant not only for Sofia's current purpose but more generally every time she will have to operate the scale.

The second subtype of declarative is constituted by impersonal deontic constructions like *bisogna x* 'it is necessary to x', which express the obligation or necessity to do an action without tying it to a particular individual (see also Floyd, Chapter 3, §3.3.4; Zinken, Chapter 8, §3.3.2; Baranova, Chapter 9, §3.3.3).<sup>7</sup> Impersonal deontic declaratives have a complex pragmatics that extends beyond recruitment (see Zinken & Ogiermann 2011; Rossi & Zinken 2016). That said, an important affordance of this declarative subtype as a recruiting format is its potential to make participation in the necessary action negotiable. This means that different individuals may have to sort out who will take on the action.

Extract 14 is taken from the same interaction as (4). Sergio, Greta and Dino are chatting while Sergio styles Greta's hair. Before the extract begins, Greta has asked Sergio to remove a 'thingy' from her forehead, which turns out to be a wisp of hair (line 1). When Sergio realizes that the hair has glued up on Greta's forehead because some dye has run down on it, he initiates a recruitment sequence using an impersonal deontic declarative.

```
(14) Tinta_ 2051380
```

```
1 ser [questo_ ((holds wisp of hair))
          this
          this
  2 gre [(eh non lo so) c'ho un coso;>
          (PTC not 3SG.ACC know-1SG) LOC=have-1SG a thingy
          (well dunno) I have a thingy
▶ 3 ser scusa *↘SÌ +bisogna + * puΔ¯lI•re=
         sorry yes necessitate-3SG clean-INF
```

```
sorry yes it is necessary also to wipe
```
4 \*gazes at Dino--->\*gazes back to Greta's head-->

<sup>7</sup> In Italian, this can be grammatically achieved by using an impersonal verb (e.g. *bisogna tagliare il pane* 'it is necessary to cut the bread') or by intransitive constructions with a non-human subject (e.g. *c'è il pane da tagliare* 'the bread is to be cut').

5 The recruitment system in Italian


Wiping the dye away could in principle be taken on by any of the three participants, including the recruiter himself. Sergio is most immediately involved in the styling process and his apology *scusa* 'sorry' indicates that he is responsible for having let the dye drip on Greta's forehead. While saying the word *bisogna* 'it is necessary to', Sergio moves a hand (line 5), possibly in the direction of the paper towel, but then hesitates. At the same time, he gazes at Dino (line 4), inviting him to get involved (see Stivers & Rossano 2010; Rossano 2012: chap. 3).

Dino is arguably in a better position to do the wiping, one reason being that Sergio is wearing gloves that are stained with dye. Also, Dino has already assisted Sergio earlier in the styling process, seeing to similar side tasks such as cleaning. Here, too, Dino steps in to help, turning toward the paper towel on the table and reaching for it (line 6). As he begins to reach, however, Greta does the same (line 7). In the midst of this, Dino verbalizes his intention to take on the task (*faccio io.* 'I'll do it', line 10). It is not clear whether this verbal response is addressed primarily to Greta or Sergio; regardless, it reflects a negotiation over who should fulfill the recruitment.

This example shows that an impersonal deontic declarative such as *bisogna x* 'it is necessary to x' does not constrain participation in the action being recruited and can make a response relevant for multiple people. Although the responsibility for the action in question sometimes falls on a specific person (see Rossi & Zinken 2016), an impersonal deontic declarative can generate a negotiation of who the doer is ultimately going to be.

The third declarative subtype is constituted by factual declaratives: non-modal constructions that present a description of a state of affairs. Although the format cannot be defined by a single lexicosyntactic formula, they often refer to the lack of something (e.g. *manca sale* 'there isn't enough salt'), the reaching of a stage in a process (e.g. *bolle l'acqua* 'the water is boiling'), a property or quality of an object

### Giovanni Rossi

(e.g. *questo è un po' unticcio* 'this is a bit slimy'), or an untoward circumstance (e.g. *i piatti stanno bloccando lo scarico* 'the dishes are blocking the drain').

Like impersonal deontic declaratives, factual declaratives do not specify a recruitee. In addition, they also do not specify the action being recruited. When using a factual declarative, the recruiter relies on the recruitee's ability to infer the target action on the basis of a shared understanding of the practical circumstances (see also Kendrick, Chapter 4, §4.2.3; Enfield, Chapter 6, §4.3.1; Baranova, Chapter 9, §3.3.3; Dingemanse, Chapter 10, §3.2.2).

Utterances such as *it's cold in here* or *the matches are all gone* have been traditionally referred to as "indirect requests" or "hints" that allow the speaker not to commit to a request intention, leaving interpretation up to the recipient, and thus affording the option not to get involved (Ervin-Tripp 1976: 42; Brown & Levinson 1987: 69, 216; Weizman 1989). But using a factual declarative is not simply a matter of indirectness. More important, it allows the speaker to do more than just getting another person to do something. In everyday informal interaction, a recurrent function of factual declaratives alongside initiating recruitment is to inform the recipient of something they do not know (Rossi 2018).

In (15), Mirko is working with others in the kitchen. At the beginning of the extract, Emma walks in, addresses Mirko, and tells him that 'the feed drip has finished', referring to the intravenous drip being administered to a family member in another room.

```
(15) Camillo_ 2039498
```


5 The recruitment system in Italian

9 emm eh. INTJ right

The focal content of Emma's turn (*è finì la ↓ˋFLEbo.* 'the feed drip has finished') is prefaced by a formulation of the turn as an informing (*volevo ˋDIRte ↑¯chE* 'I wanted to tell you that'). This characterization of Emma's action is consonant with Mirko's first response in the form of a change-of-state token *a::h* 'o::h' (Heritage 1984a), which signals that his state of knowledge has changed and thus receipts the information reported by Emma as news (line 6). A moment later, Mirko expands his response with another unit, which includes an assessment of the news as 'good' and then a commitment to going and nursing Milena ('we can release Milena then'), showing his understanding of Emma's action not only as an informing but also as a request.

So factual declaratives are often used to inform the recruitee of something they do not know, which functions as a vehicle for recruiting their assistance or collaboration (Rossi 2018).

# **3.4 Additional verbal elements**

This section looks at verbal elements in the recruiting turn beyond the basic linguistic frame created by the construction type and subtype being used. As discussed in Chapter 2, §6, additional verbal elements tend to fall into four main categories: vocatives (e.g. *´PLInio; a´sciUga anche ˋQUEsta.* 'Plinio wipe this one too', Extract 1), benefactives (e.g. *tienimi questi un attimo* 'hold these for me one second'), explanations, and mitigators or strengtheners. The following subsections focus on the latter two categories and illustrate their usage in the context of imperative recruiting turns, to make comparison easier with cases examined in earlier sections.

### **3.4.1 Explanations**

Explanations, accounts, and more generally reason-giving occur at various places in interaction (see Goodwin 1987; Antaki 1994; Drew 1998; Waring 2007; Parry 2009; Bolden & Robinson 2011, among others). In recruitment sequences, explanations refer to circumstances that are grounds for the recruitment to be initiated or that make it more understandable or warranted (see Parry 2013; Baranova & Dingemanse 2016; Rossi 2017).<sup>8</sup>

<sup>8</sup>To count as an additional element rather than as a stand-alone recruiting turn, the explanation must be produced as an appendage to a construction type among those surveyed in §3.3.

### Giovanni Rossi

Extract 16 is taken from the same interaction as (1). Plinio and Rocco are in charge of drying the dishes that others are washing. As they wait for the next round of washed dishes to dry, Plino picks up a dishwasher tray and asks if it is going to be used again (line 1). After Agnese responds with 'no', Plinio puts the tray away (line 4). Shortly after this, Rocco tells Plinio to put away another tray that is lying on the floor.

```
(16) CampFamLava_591294
```

```
1 pli questo servirà ancora; ((holds up white tray))
          this serve-FUT-3SG again/still
          is this going to be used again
  2 (2.1)
  3 agn no.
          no
  4 (5.0) ((Plinio puts white tray away))
  5 (9.5) ((Plinio wanders between sink and dishwasher))
▶ 6 roc ˇMETti via anche quello lì ↘GIALlo ((points at yellow tray))
          put-NPST-2SG away also that there yellow
          put away that yellow one too
  7 che se no gli pestiam ↘SOpra;
          CONN if no 3SG.DAT step-1PL above
          otherwise we're going to step on it
▷ 8 pli ((picks yellow tray up and puts it away))
```
The recruitment is initiated within a joint project that recruiter and recruitee are involved in (see §3.3.2). The explanation appended to the imperative recruiting turn (*che se no gli pestiam ˋSOpra;* 'otherwise we're going to step on it') indicates that the recruitment is in the interest of both participants, with the goal of preventing an unwanted consequence, and thus articulates and specifies the contribution of the recruitment to their joint project. Of 17 explanations added to imperative recruiting turns in the Italian sample, 13 have an analogous function. For a more detailed account of the interactional processes involved in reasongiving for recruitments, see Baranova & Dingemanse (2016).

### **3.4.2 Mitigators**

Recruiting moves can include design features to mitigate or soften the imposition on the recruitee or, alternatively, to emphasize the urgency of the action being recruited (see Brown & Levinson 1987; Blum-Kulka et al. 1989). The following case gives us an example of mitigation.

5 The recruitment system in Italian

Extract17 is taken from the same card game as other examples examined above. Teammates Bianca and Flavia are consulting on their next move, while Clara and Silvia are waiting for their turn. During the wait, Silvia takes a piece of cake from a shared plate on the table (line 4). This occasions Clara's initiation of recruitment.

```
(17) Circolo01_1270484
```


The recruiting turn includes two mitigators: *per piazer* 'please' and *valà*, a northern Italian particle which in this context can be rendered with the English tag 'will you' – an appeal to the recipient's benevolence or goodwill. These two additional elements mark the imperative request as requiring some kind of redress (Brown & Levinson 1987). Such mitigators are normally not found in imperative requests of the kind illustrated in (1), (10), and (16). Take (10), for instance, which takes place during the same card game. In that sequence, the request contributes to the progress of the card game. In (17), by contrast, the request is for a good to be consumed by the requester alone, not unlike requests designed with a simple interrogative (11).

There is no space here to discuss the conditions that support the use of an imperative in (17) (see Rossi 2017 for an account). What is important to note is that

### Giovanni Rossi

the imperative request here differs functionally and interactionally from those seen above (see §3.3.2), and that this difference is associated with the use of mitigators.

To sum up this whole section on Move A (§3), I have surveyed a range of verbal and nonverbal resources that speakers of Italian have at their disposal for initiating recruitment. The use of fully nonverbal forms (e.g. simply pointing or reaching toward an object) is generally constrained to contexts that afford the projectability of the action being recruited. If language is needed to specify the action, Italian speakers calibrate the verbal component of the recruiting turn from phrasal or single-word formats to clausal ones. The use of alternative clausal types (imperative, interrogative, declarative) and subtypes is sensitive to a range of factors including the sequential and functional relation of the recruitment to what the recruitee is currently doing, the benefit brought by the action being recruited, the availability of objects, the anticipation of the recruitee's unwillingness, the negotiability of participation, and the performance of other actions (e.g. informing) as a vehicle for getting another to do something. I have also observed patterns in the use of pointing gestures and noted the frequency of iconic gestures in recruiting moves. Finally, the verbal component of a recruiting move can be enriched beyond the basic linguistic frame being used with additional elements. Focusing on imperative recruiting moves, we have seen that recruiters may add explanations to articulate the contribution of the recruitment to an ongoing joint project, or alternatively they may add mitigators to soften the use of an imperative format outside a joint project.

# **4 Formats in Move B: The responding move**

Like Move A, Move B can include nonverbal and/or verbal behavior. However, since the goal of a recruitment sequence is to mobilize practical action, fulfillment naturally requires nonverbal, physical work. This is often all the recruitee does in the responding move. When we look quantitatively at the modality of complying responses, over half are fully nonverbal (52.8%, *n*=75/142).<sup>9</sup>

In what follows, I consider the modality of the responding move with an eye to what it can tell us about the nature of the recruiting move. After examining particular kinds of verbal responses that may accompany nonverbal fulfillment, I

<sup>9</sup> Six cases were excluded from this count where it was not possible to ascertain whether the responding move did or did not include a verbal component.

### 5 The recruitment system in Italian

look at exclusively verbal responses that indicate a problem with the recruitment, including different ways of rejecting it.

# **4.1 Response modality**

As mentioned above, a fully nonverbal response is often all that is needed to fulfill a recruitment. However, fully nonverbal responses are not equally distributed across the dataset. Table 3 shows the modality of complying responses by recruiting format. Nonverbal, no-predicate, and imperative recruiting moves are more frequently responded to nonverbally; interrogative and declarative recruiting moves, by contrast, are more frequently responded to verbally.

Table 3: Modality of complying responses relative to the format of the recruiting move (*n*=142).


For imperative recruiting moves, the pattern is consistent with earlier research showing that the imperative format projects only the fulfillment of a request or directive (Wootton 1997; Goodwin 2006; Craven & Potter 2010; Kent 2011; 2012; Rossi 2012; 2015c: chap. 3).

In §3.1 and §3.3.1, we saw that nonverbal and no-predicate recruiting moves occur in similar environments as imperatives, namely within joint projects that support an expectation of compliance with recruitments serving the project's advancement (see Extracts 5 and 9). This suggests that, while nonverbal and nopredicate formats do not have the semantics of an imperative clause, they may be similarly understood as making relevant only the fulfillment of the recruitment (see also Dingemanse, Chapter 10, §4.1 on nonverbal recruiting formats receiving nonverbal responses).

The modality of responses to interrogative and declarative recruiting moves is also consistent with the findings of earlier research, discussed in the next section.

### Giovanni Rossi

As we will see, these recruiting formats make relevant more than one response option, with declaratives affording an open response space. We will also see that complying responses that include verbal elements involve more than the fulfillment of the recruitment.

# **4.2 Verbal elements of responses**

### **4.2.1 Accepting, confirming, and agreeing**

Earlier research has shown that interrogative recruiting formats, specifically polar interrogatives, are legitimately responded to with acceptance before fulfillment or with a negative answer (Wootton 1997; Raymond 2003; Craven & Potter 2010; Kent 2011; Rossi 2012; 2015c: chapp. 3–4). Unlike an imperative, a polar interrogative conveys that the recruitee's compliance is not being assumed (cf. Searle 1975: 74; Ervin-Tripp 1976: 60; Wierzbicka 1991: 159). In Italian, as in other languages, recruiting moves designed as polar interrogatives are accepted with a positive polar token. In (11), for example, Franco says *sì* 'yes' before fulfilling Beata's simple interrogative request; in (8), Loretta accepts a similar request from Rocco with a head nod.

Among the interrogative subtypes in Italian, the *hai x* 'do you have x' format exhibits special properties that have consequences for how the recruiting turn can be responded to. Like simple and *puoi x* 'can you x' interrogatives, a *hai x* 'do you have x' recruiting turn makes fulfillment contingent on the recruitee's response. But it does so in a different way. In §3.3.3, we saw that this format functions as a pre-request checking a precondition for recruitment. This affords two types of response that support the accomplishment of the sequence: one is immediate fulfillment, optionally accompanied by a positive polar answer (see Extract 12); the other is a go-ahead response (Schegloff 2007b: 30), confirming that the precondition obtains. An example of this is given in the extract below.

```
(18) Circolo01_2718316
```

```
▶ 1 sil ghe ´NAt? ((points at card combination))
           LOC PTV=have-2SG=2SG.SCL
           do you have any
▷ 2 cla una_
           one
  3 sil ↘DAmela?
           give-IMP.2SG=1SG.DAT=3SG.ACC
           give it to me
```

```
4 cla ((passes card))
```
### 5 The recruitment system in Italian

Clara responds to Silvia's *hai x* 'do you have x' interrogative by asserting that she has one unit of the target object. This is followed by Silvia producing another first pair part, this time in imperative form, which is responded to with fulfillment. For a more detailed account of expanded sequences like (18), see Rossi (2015b) and references therein.

While interrogatives make relevant at least two alternative types of verbal response, declaratives have been shown to afford an even wider range of options (Vinkhuyzen & Szymanski 2005; Rossi & Zinken 2016). In (15), for instance, Mirko responds to Emma's factual declarative with two distinct responses that address two different actions accomplished by her recruiting turn. The change-of-state token *a::h.* 'o::h' (line 6) treats what Emma has told Mirko as news, while his subsequent commitment to releasing Milena from the feed drip (line 8) orients to it as a request. In (14), Dino responds to Sergio's *bisogna x* 'it is necessary to x' declarative with *faccio io.* 'I'll do it', volunteering to do the necessary action. These examples already exhibit a wider range of response types than any of the recruiting formats we have considered so far.

The next example illustrates yet another type of response afforded by declarative recruiting turns: agreement. Fabio, Rino, and other friends are making a booklet of short readings, the printouts of which are scattered on the table. It is now time to type up the excerpts on the computer. When the extract begins, Fabio has just offered to dictate the excerpts to Rino. His question 'which one do we write up first' (line 1) implies an understanding that all the excerpts they have considered will eventually be included in the booklet. In response, Rino rejects this understanding and recruits everyone to make a selection of the readings for inclusion.

```
(19) Precamp01_831126
```


Before complying with Rino's recruiting move nonverbally (line 4), Fabio says *eh eˋSATto.* 'right exactly', by which he agrees with Rino's statement and the

### Giovanni Rossi

view of the world it presents. Fabio then strengthens his agreement with a nearverbatim repetition of the statement, a practice that is used to assert one's epistemic right over what someone else has just said (Stivers 2005; cf. Schegloff 1996). These types of responses are afforded only by declarative recruiting formats. In this particular case, the impersonal deontic construction used by Rino asserts the existence of a need or obligation, which may be agreed with or – as we see in the next section – disagreed with.

### **4.2.2 Rejecting, blocking, and disagreeing**

Another function of verbal elements in the responding move is to reject the recruitment. Rejection is a dispreferred response that thwarts the course of action initiated by the recruiter and poses a potential threat to social solidarity (see Heritage 1984b: 265–80; Brown & Levinson 1987; Schegloff 2007b: chap. 5). The dispreferred status of rejections is reflected in their design, as illustrated by the following examples.

```
(20) Capodanno02_655722
```

```
▶ 1 eva ma ˆmEteghe 'l ˆCO:so ↘prIma
         but put-IMP.2SG=3SG.DAT the thingy before
         but put the thi:ngy first
▷ 2 (0.7)
▷ 3 ada <<breathy>ma: pensavo ˆSOra.>
                   but think-IMPF-1SG above
                  but: I was thinking {to put it} on top
  4 eva <<pp>↑ah [vabem.>
               oh PTC
               oh okay
▷ 5 ada [↑↘SOra l'è pu gudu↘RIOso,
                   above SCL=be.3SG more pleasurable
                   on top is more delicious
(21) BiscottiPome01_1884369
▶ 1 azi Furio mi ´PREsti le chiavi del ga¯RAge
         NAME 1SG.DAT lend-2SG the keys of-the garage
         Furio {will} you lend me {your} garage keys
  2 che te le riporto alle ↘TRE;
         CONN 2SG.DAT 3PL.ACC return-1SG at-the three
         which I'm going to return to you at three
▷ 3 (4.7)
```
5 The recruitment system in Italian

```
▶ 4 azi ↑non ce le ↘HO.
           not LOC 3PL.ACC have-1SG
           I don't have {mine}
  5 (0.3)
▷ 6 fur eh öh eh sono mi- ¯Anche le mie chiavi di cat- di ˆCAsa.
          PTC uh PTC be.3PL mi- also the my keys of ho- of house
          well uh well they're m- also my c- house keys
```
These two examples illustrate some of the typical features of dispreferred responses that have been extensively documented in the literature: delays, prefatory particles (*ma* 'but', *eh* 'well'), hesitations (*öh* 'uh'), and the provision of reasons for not complying. These features are found in negative responses to variously formatted recruiting moves, including imperatives (20) and simple interrogatives (21).<sup>10</sup> But now consider another case where the recruitment is initiated with a *hai x* 'do you have x' interrogative (cf. Extracts 12 and 18).

```
(22) Circolo01_2718316
```

```
1 sil suo, ((points to Clara))
          hers
          {it's} hers
  2 (0.7)
▶ 3 fla öh ti ghe ´NAt? ((to Bianca))
          uh 2SG.NOM LOC PTV=have-2SG=2SG.SCL
          uh do you have any
▷ 4 bia no. ((shakes head))
          no
```
Like in (20) and (21), Bianca's 'no' is structurally dispreferred in that it does not support the accomplishment of the course of action initiated by the recruiting move. Yet it lacks all the features seen earlier. The explanation for this lies in the nature of the particular action performed by a *hai* x 'do you have x' interrogative. In §3.3.3, we saw that pre-requests check a precondition for a request to be made successfully. This means that a negative response to the pre-request is not a response to the projected request – that is, it is not a rejection. Rather, it is a blocking response (Schegloff 2007b: 30). A blocking response like Bianca's 'no' in (22) indicates a state of affairs – here, the unavailability of the target object – that prevents the further development of the activity and that is normally beyond the control of the recruitee, rather than a matter of disposition or uncooperative

<sup>10</sup>See also the responses in (2) above and in (23) below.

### Giovanni Rossi

behavior. For this reason, the negative response does not need to be mitigated in the same way a rejection does (Rossi 2015b).

The last example in this section illustrates a particular form of rejection that is afforded by declarative recruiting formats; here in particular by a declarative of the *bisogna x* 'it is necessary to x' subtype (cf. Extracts 14 and 19 above). Elena is sitting at the kitchen table, finishing her food. Across from her, Agata is loading the dishwasher and is doing so without pre-rinsing the dishes. When the extract begins, Elena points out the need to select a heavy wash cycle, on the grounds that the food they have eaten may otherwise not come off in the dishwasher.

```
(23) Capodanno02_21779
```

```
▶ 1 ele bisogna ´dArghe
         necessitate-3SG give-INF=3SG.DAT
         it is necessary to select
  2 en programma molto ↘ALto ↘Agata [eh, per]ché=
         a program very high NAME PTC because
         a very intense program Agata you know because
▷ 3 aga [↑↑mac↘CHÉ.]
                                      INTJ
                                      not at all
  4 ele =questo s' <<breathy>attacca en d' en ´M[Odo_>
         this RFL attach-3SG in of one manner
         this sticks so much
  5 aga [sì ma l'è l- la
                                              yes but SCL=be.3SG
                                              yes but it's l-
  6 g'avem giusto magnà.
         LOC=have-1PL just eat-PCP
        we've just eaten on it
  6 non è arivà neanche a secarse,
         not be.3SG arrive-PCP neither to dry_up-INF=RFL
         it hasn't even had the time to harden
```
Agata's response begins with the interjection *↑↑macˋCHÉ.* 'not at all' (or 'of course not'). With it, Agata confutes the veracity of the assertion expressed by Elena's declarative (*bisogna ´dArghe en programma molto ˋALto* 'it is necessary to select a very intense program'), in other words, she disagrees with it.

Disagreement is not found in rejections to imperative and interrogative recruiting moves as these are not treated as statements committing to the truth of a proposition. A statement of need, on the other hand, makes a claim about the material and social world, and exposes it to the evaluation of others against their own understanding of that world (see also Zinken & Ogiermann 2011). In (19), we saw that Furio agrees with Rino's *bisogna x* 'it is necessary to x' declarative with

### 5 The recruitment system in Italian

*eh eˋSATto.* 'right exactly' and then strengthens his agreement by repeating the statement. In (23), Agata does the opposite: after expressing her disagreement with *↑↑macˋCHÉ.* 'not at all', she goes on to dispute the grounds upon which the Eva's claim is based: although it may be true that the food sticks on plates, they have just finished eating so, according to Agata, the food has not yet had the time to cake on the plates. The implication is that, in her view, this makes the selection of a heavy wash cycle unnecessary.

In sum, this whole section (§4) has shown that formats in Move B are closely patterned relative to formats in Move A. The grammar and the particular actions accomplished by various recruiting formats place different constraints on, and provide different affordances for, how exactly the recruitee can comply with or reject the recruitment. For complying responses, imperatives project only nonverbal fulfillment, while interrogatives allow the recruitee to accept before fulfilling, and declaratives provide an open space of options, including receipting information, agreeing with what has been said, and volunteering assistance. For rejections, negative responses to most recruiting formats are normally marked as dispreferreds, with declaratives allowing for disagreement. Negative responses to *hai x* 'do you have x' interrogatives, however, are not designed as dispreferreds, as they do not constitute a rejection but a blocking response to a pre-request.

# **5 Acknowledgment in third position**

Across the languages examined in this volume, acknowledging fulfillment of a recruitment with a third-position turn like 'great' or 'thank you' is rare (Floyd et al. 2018). At the same time, Italian shows a relatively higher proportion of such turns than other languages (13.5%, *n*=20/148). This includes 8 cases of the dedicated expression *grazie* 'thanks', two examples of which are found in (3) and (8) above. Other cases involve positive assessments (e.g. *ottimo* 'excellent', *bravo* 'well done') and interjections such as *bom* 'alright' and *eh* 'right'; an example of the latter is found in (15) above.

If acknowledgment in third position is generally infrequent in recruitment sequences, what do we make of the cases where acknowledgment does occur? In a recent study (Zinken et al. 2020), we addressed this question with particular reference to thanking. What we found is that, in informal interaction, thanks are given to recognize another person's agency in providing assistance. In recruitment sequences that end with thanks, the recruiter treats fulfillment as not taken for granted and rather as the result of the recruitee's autonomous deci-

### Giovanni Rossi

sion to help. This happens most obviously in offer sequences, where assistance is provided without having been requested. We also observe thanking in delicate request sequences, where there is an anticipation of actual or potential unwillingness on the part of the recruitee; if unwillingness is successfully overcome, this is grounds for acknowledging the recruitee's compliance.

But thanking may also occur after compliance with unproblematic requests. Here, the recruiter treats compliance as not taken for granted even though, contextually, it is largely expectable. Thanking then functions reflexively to accentuate the recruitee's agency in providing assistance.

Even apparently reflexive practices, however, can be sensitive to the interactional environment in which they are used. Extract 24 serves to illustrate this.

During dinner at a family gathering, Plinio finds himself without a fork (*a me manca la forchetta. hah hah hah* 'I don't have a fork hah hah hah', line 1). One of the diners sitting across from him hears the comment and directs him to a container with forks located on a service table (*è lì:;* 'it's there', line 3). As Plinio looks over to where the forks are, Fabrizio walks in with a sponge cloth to wipe the service table, on which he accidentally spilled food moments earlier. Plinio calls out to Fabrizio and, after repeated attempts to get his attention (lines 6, 10), asks him to pass a fork (line 12). As Fabrizio turns to the forks container, he playfully rejects the request (*no:\_* 'no', line 16) and then quickly fulfills it by passing a fork, which Plino acknowledges with *grazie,* 'thanks' (line 19).

```
(24) NataleSala02_2007128
```

```
1 pli a me manca la forchetta. hah hah hah
       to 1SG.ACC lack-3SG the fork
       I don't have a fork hah hah hah
2 (0.6)
3 cle è *lìΔ:; ((points))
       be.3SG there
       it's there
4 pli *looks over-->>
5 fab Δapproaches table with sponge cloth-->
6 pli öeh:: <<all>FabriΔzio Fabrizio Fabrizio_>>
                  NAME NAME NAME
       uhm:: Fabrizio Fabrizio Fabrizio
7 fab ---------------->Δbegins to wipe table-->
8 (0.3)Δ(0.1)
9 fab ---->Δturns around with upper body-->
```
5 The recruitment system in Italian


Plinio's simple interrogative request initiates a course of action that is in his interest as an individual (see §3.3.3). As we saw in (11), a key aspect of these sequences is the lack of continuity between what is requested and what the recruitee is doing at the moment, which often requires departing from one's business in order to fulfill the request. In (11), Franco has to briefly disengage from the game to get Beata a paper towel; in (24), Fabrizio has to suspend his ongoing task of wiping the table to get Plinio a fork.

With this in mind, let us look more closely at some of the particulars of how Fabrizio's assistance is recruited and acknowledged in (24). A first notable feature is the work that Plino does to establish mutual attention. In his initial summons, Plinio repeats Fabrizio's name three times as he approaches the table and begins to wipe it (lines 5–7). Plinio then produces yet another, louder vocative (line 10), before Fabrizio finally turns around (line 11). Note that Fabrizio rotates only the upper part of his body, mainly his neck (Figure 8*a*); this body torque displays Fabrizio's commitment to his primary involvement in wiping the table

### Giovanni Rossi

(a) Plinio points to the forks container while saying *mi ´pAssi una forˆCHETta.* '{will} you pass me a fork'; Fabrizio has turned his neck and is looking at him while keeping a hand with the sponge cloth on the service table (line 13).

(b) Fabrizio passes a fork to Plinio while still keeping a hand on the service table (line 21).

Figure 8: Frames from Extract 24.

(Schegloff 1998). Note also that he keeps his hand with the sponge cloth on the table throughout the sequence (Figure 8*b*).

This configuration highlights Fabrizio's position as a participant with his own business to tend to, who is being recruited to assist in someone else's project. Fabrizio's agency in this episode is further underscored by his playful rejection of the request (*no:\_* 'no', line 16). By teasing Plinio with rejection, he draws attention to the fact that he has a choice, which helps to see his subsequent granting as an autonomous decision.

Treating compliance with a request as an autonomous decision is often a matter of construal, and a locus of cultural diversity (Zinken et al. 2020). However, some of the elements that seem particularly conducive to recognizing the recruitee's agency in (24) can be observed also in other cases where compliance is acknowledged. In (8), for instance, Rocco's request for a glass of water comes after an earlier attempt to get Loretta's attention; her agency as a recruitee is reflected in how she makes herself available only after she is done with a concur-

### 5 The recruitment system in Italian

rent conversation (*dimmi scusa.* 'tell me – sorry', line 2). In (3), Sara pursues her request for one more piece of banana with *per fa'VOre.* 'please' (line 4), which attributes Furio agency in deciding whether or not to share more of his food with her.

# **6 Social asymmetries**

This study is based on video recordings of informal interaction around the home, in the family, and in people's proximate community of friends and neighbors. My sample of recruitment sequences included only adult participants. I did not identify noticeable social asymmetries between the individuals participating in these sequences. Possible sources of asymmetry such as gender, age, or socioeconomic status did not emerge in the analysis.

Exploring the larger corpus, I identified one noticeably asymmetrical relation between a daughter-in-law and her mother-in-law. The social asymmetry in this dyad is reflected in the daughter-in-law's use of the second person formal pronoun *Lei* and corresponding verb inflections, which is not reciprocated by the mother-in-law. The sample did not include any recruitment sequences between these two individuals.

# **7 Discussion**

This chapter has provided an overview of how speakers of Italian recruit one another's assistance and collaboration in everyday informal interaction. Following the common structure adopted in the contributions to this volume, I have examined a range of interactional resources for initiating recruitment (Move A) and for responding to it (Move B), paying particular attention to the fit between the two moves.

For Move A, Italian speakers use all three main construction types found crosslinguistically: imperatives, interrogatives, and declaratives, as well as utterances without a predicate (e.g. *una* 'one'). While imperatives are the most frequent construction type, interrogatives and declaratives are also common. The verbal component of the recruiting move can be enriched with additional elements including explanations and mitigators, one function of which is to attune the basic format being used (e.g. an imperative) to the interactional circumstances.

The nonverbal component of the recruiting move often involves one of three basic behaviors found across languages: pointing, holding out an object for someone to take and do something with, and reaching out to receive an object. These

### Giovanni Rossi

behaviors can also function as recruiting formats on their own, without words. In speech-plus-gesture composites, functional distinctions can be observed between different forms of pointing ("big" and "small") that have been documented in other languages and interactional contexts (Enfield et al. 2007). At the same time, Italian speakers use a relatively high proportion of iconic gestures in Move A, which is consistent with previous research on the frequency of iconic gestures by Italian speakers in other settings (Campisi 2014).

In surveying the repertoire of strategies for initiating recruitment, I have also tried to account for the selection between alternative formats, examining a number of social-interactional factors that influence the design of the recruiting move. One is the projectability of the action being recruited, which is particularly important for the use of nonverbal and no-predicate formats. The criterion of projectability is grounded in fundamental principles of human communication, including informational calibration and the recognizability of action, which are likely to be shared across languages (Grice 1975; Clark 1996; Levinson 2000). Another factor that plays a role in the design of recruiting moves – particularly in the use of interrogative vs. imperative formats – is whether the action being recruited contributes to an already established joint project between recruiter and recruitee or else it initiates a new course of action that is in the interest of the recruiter as an individual. While these social-interactional concerns may be more prone to cultural variation, there is also evidence for their relevance across languages (see Kendrick, Chapter 4, §4.2.2; Zinken, Chapter 8, §3.3.1).

Yet another factor that influences the design of the recruiting move is the anticipation of the recruitee's actual or potential unwillingness to comply. In these cases, Italian recruiters use the interrogative format *puoi x* 'can you x' to recognize and attempt to overcome unwillingness through persuasion. While the basic concern for dealing with delicate recruitments may be universal (Brown & Levinson 1987), the particular formats used to do this are more likely to differ. In contrast to this, a form-function mapping that may be cross-linguistically valid is the one between the need to check a precondition for recruitment and the use of an interrogative format (see Floyd, Chapter 3, §3.3.3; Enfield, Chapter 6, §4.3.1). In the Italian data, this often involves querying the recruitee as to the availability of a target object with *hai x* 'do you have x'.

Factual declaratives that present a description of a state of affairs (e.g. *è finì la flebo* 'the feed drip has finished') are yet another format that appears to work similarly across languages (see Kendrick, Chapter 4, §4.2.3; Baranova, Chapter 9, §3.3.3; Dingemanse, Chapter 10, §3.2.2). An important affordance of this format emerging from the Italian data is its capacity to get others to do things by means

### 5 The recruitment system in Italian

of informing them of something they do not know. As for impersonal deontic declaratives (e.g. *bisogna x* 'it is necessary to x'), while these are not available in the same way across languages, they have counterparts in at least some of those examined in this volume (see Floyd, Chapter 3, §3.3.4; Zinken, Chapter 8, §3.3.2; Baranova, Chapter 9, §3.3.3).

Coming now to Move B, I have surveyed a range of options that Italian speakers have for complying with or rejecting a recruitment, paying special attention to how the use of alternative responding formats is sensitive to the nature of Move A.

Nonverbal fulfillment is the appropriate response to recruiting moves that project only compliance, such as those designed with an imperative. Verbal acceptance or confirmation (e.g. *sì* 'yes') is a relevant response to recruiting moves that formally anticipate the possibility of rejection or failure of the recruitment, first and foremost polar interrogatives. Similar principles of responding apply in other languages as well (see Kendrick, Chapter 4, §4.2.3).

With polar interrogatives that function as pre-requests (e.g. *hai x* 'do you have x'), a positive polar answer counts as a go-ahead, confirming that the precondition for recruitment obtains, whereas a negative answer counts as a blocking response which, unlike a rejection, may not need to be justified or mitigated. This pattern is based in generic properties of action and sequential structure (Schegloff 2007b; Kendrick et al. 2020).

Finally, declarative recruiting formats afford an open response space. Factual declaratives that convey new information to the recruitee, for instance, can be taken up with a news receipt (e.g. *ah* 'oh'). Other declaratives that make a claim about the material and social world (e.g. *bisogna x* 'it is necessary to x') can be agreed or disagreed with.

Looking beyond recruiting and responding moves, Italian speakers may acknowledge fulfillment of a recruitment by thanking (e.g. *grazie* 'thanks'), with positive assessments (e.g. *ottimo* 'excellent'), and with sequence-closing interjections (e.g. *eh* 'right'). Such acknowledgments in third position are uncommon in recruitment sequences around the world (Floyd et al. 2018). At the same time, Italian is one of two languages in this project, together with English, where acknowledgment is relatively more frequent. The occurrence of acknowledgment, particularly in the form of thanking, reflects a preoccupation with recognizing individual agency in the provision of assistance (Zinken et al. 2020).

In conclusion, the findings presented in this chapter show a tightly organized system of resources for recruiting assistance and collaboration. While the system is inflected according to the Italian language and culture, it shares many formal

### Giovanni Rossi

and functional elements with that of other languages, pointing to a common infrastructure for the management of cooperation in social life.

# **Transcription, glossing, and translation**

Transcripts follow basic conventions established in conversation analysis (Jefferson 2004; Hepburn & Bolden 2013). Prosodic features are represented according to GAT 2 conventions (Couper-Kuhlen & Barth-Weingarten 2011), which include symbols for indicating pitch movement on accented syllables and at the end of the utterance (e.g. *una riˆSTAMpa;*).<sup>11</sup> Elements of visible behavior are generally noted as comments in double parentheses ((nonverbal behavior)). For some extracts, visible behavior is represented in greater detail using Mondada's (2019) conventions. Interlinear glosses generally follow the Leipzig rules (Comrie et al. 2020). I have added a few abbreviations and shortened others for economy:


Free translations may include the following symbols:


<sup>11</sup>I have transcribed pitch movement on accented syllables only for focal turns. The grave accent character "ˋ" used in GAT 2 to represent falling pitch on an accented syllable was unavailable under the particular LaTeX setup used for transcripts in this volume, so I replaced it with the symbol "↘" (see, e.g., Extract 1, line 2).

5 The recruitment system in Italian

# **Acknowledgments**

Thank you to my fellow project members for their input and intellectual engagement over the years. I am also grateful to two reviewers for their helpful commentary. This research was supported by the European Research Council (grant no. 240853 to Nick Enfield) and by the Academy of Finland (grant no. 284595 to Marja-Leena Sorjonen).

# **References**


Clark, Herbert H. 1996. *Using language*. Cambridge: Cambridge University Press.

Clark, Herbert H. 2003. Pointing and placing. In Sotaro Kita (ed.), *Pointing: Where language, culture, and cognition meet*, 243–68. Mahwah: Lawrence Erlbaum.

### Giovanni Rossi


5 The recruitment system in Italian


### Giovanni Rossi


5 The recruitment system in Italian

*tween parents and children in Italian and Swedish families]*. Stockholm University. (Doctoral dissertation).


### Giovanni Rossi

(Studies in Language and Social Interaction 26), 303–334. Amsterdam: John Benjamins. DOI:10.1075/slsi.26.12ros


5 The recruitment system in Italian


# **Chapter 6**

# **Recruitments in Lao**

# N. J. Enfield

Department of Linguistics, The University of Sydney

This chapter describes the resources that speakers of Lao use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Lao, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Lao with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

# **1.1 Recruitments**

This chapter describes and analyzes aspects of the system of semiotic practices that speakers of Lao use when getting people to do things in the course of everyday life. As defined in this collaborative project (see Chapter 1, §4), a recruitment sequence involves one participant A doing or saying something to B, or such that B can see or hear it, and next, as a response, B doing something for or with A. The data are drawn from a corpus of video-recorded interaction collected in home and village settings in Vientiane, Laos. The approach taken here assumes

### N. J. Enfield

that the relevant unit of analysis is the *pair* of moves that constitutes the recruitment sequence, that is, both an initiating move by Person A – for example, a request or command, or a visible display of a need or difficulty – that would precipitate some assisting behavior, as well as the move by Person B that responds to it with a form of assistance or collaboration that benefits Person A or a larger social unit of which A is a part, whether it is an act of compliance, rejection, or something else. In this chapter, we examine properties of both moves in the sequence, and ask as to their forms, functions, and interrelations.

The observations offered here arise from research done in a major crosslinguistic project (see Chapter 1 and other chapters in this volume).<sup>1</sup> To maximize comparability, the data collected is tightly defined in scope. The studies rely solely on corpus materials from recordings of everyday home and village life. The interactions take place between relatives, neighbors, or people who otherwise know each other well. This implies that none of the interactions are formal or institutional in kind, which in turn means that the phenomena described in this chapter do not exhaust the resources that Lao speakers use in getting people to do things. For example, we shall see that in Lao village life, people seldom acknowledge the assistance given, for example by saying 'thank you', while in the more formal settings that are beyond the scope of this work, an idiom meaning 'thank you' is often used. A comprehensive account of the resources that Lao speakers rely on in recruitment sequences would require a broader collection of data.

# **1.2 The Lao language**

Lao is an isolating/analytic language of the Southwestern Tai branch of Tai-Kadai. It is spoken by about 20 million people mostly in Laos, Thailand, and Cambodia. It is a tone language, with five lexical tones. The tones are indicated in this chapter by a numeral at the end of each word (see Enfield 2007 for glossing conventions). Lao has open classes of ideophones, nouns, verbs and adjectives, and closed classes of tense/aspect/modality markers, modifier classifiers and noun class markers, and phrase-final and sentence-final particles. There is no inflectional morphology. Grammatical relations tend to be signaled via constituent order, though there is widespread zero anaphora and movement licensed by information-structure considerations. Several grammars of Lao are available (see Enfield 2007 and many references therein). For recent work on semantic, prag-

<sup>1</sup>With thanks to research collaborators in the *Human Sociality and Systems of Language Use* project (see Chapter 1, §3). Other publications presenting ideas and findings from this project include Enfield (2011a,b; 2014), Floyd et al. (2014; 2018), Rossi (2012; 2014; 2015), Drew & Couper-Kuhlen (2014), Kendrick & Drew (2016), Zinken & Rossi (2016), Floyd (2017).

6 Recruitments in Lao

matic, and conversational patterns in Lao, see Enfield (2009; 2010; 2013; 2015a; 2015b), Zuckerman (2017).

# **2 Data collection and corpus**

The corpus on which this work is based was constructed in accordance with guidelines developed by and for the members of the comparative project being reported on in this volume (see Chapter 1 for further information). Here are the key properties of the data:


The corpus from which the cases were drawn included video recordings collected by the author in Vientiane, Laos, between 2001 and 2011. Twelve interactions were sampled, with a combined duration of 2 hours 46 minutes, and a total of 222 cases of recruitments for this study. All interactions involved three or more participants. All recordings were made in family homes and village settings.

# **3 Basics of recruitment sequences**

This is a study of recruitment sequences, defined by members of the collaborative subproject on this topic in the following way:

### N. J. Enfield

"The subproject on recruitment (ways that cooperative action gets mobilized) focuses on sequences in which a move by one participant ("[Move A]", whether or not the move includes speech) leads immediately to a cooperative uptake behavior by another participant ("[Move B]"; this should be a practical bodily action, such as passing the salt, not simply giving information). We limit our scope to the here and now, thus precluding things like invitations where the uptake behavior would happen at a later place and time. […] On this definition, "recruitments" straightforwardly subsumes things like requests and proposals, but also includes cases in which it may be unclear or equivocal whether the initiating [Move A] was an overt "request" or similar, so long as it results in the cooperative behavior" (Enfield 2011b).<sup>2</sup>

# **3.1 Minimal sequence**

A basic or minimal recruitment sequence in the Lao data consists of these two moves, Move A and Move B, by Person A and B respectively. The two moves are indicated by ▶ and ▷ in the transcripts.

Here is a typical example. Person A says 'grind (it)', while holding some herbal medicine out for Person B, who is holding the relevant medicine-grinding paraphernalia.

```
(1) INTCN_020727a_326860
```
▶ 1 a fon3 vaj2 ((holding medicine for B to take)) grind IMP.RUSH grind (it)

▷ 2 b ((takes the medicine from A, prepares to grind it))

In another example, Person A is in an outdoor kitchen area, using a hose that delivers water pumped up from a well in the backyard. The pump is an electric one, and the switch that turns it on and off is located inside the house, several meters away from where Person A is standing. Two men are inside the house, close to the switch that turns the pump on and off. Person A calls out to them.

(2) INTCN\_030806e\_191591

▶ 1 a mòòt4 nam4 haj5 nèè1 extinguish water give IMP.SOFT turn off the water please

<sup>2</sup>The original formulation had "M1" (Move 1) and "M2" (Move 2). Cf. Enfield (2011a), Floyd et al. (2014), Rossi (2015), Kendrick & Drew (2016), Floyd (2017), and Chapter 1 of this volume.

6 Recruitments in Lao

▷ 2 b ((one of the men gets up and walks to the switch and turns the power for the water pump off))

A third example involves transfer of an object (as opposed to provision of a service as seen in the last two examples). Person A and Person B are in a household food preparation area. Person A asks Person B to pass a papaya.

```
(3) CONV_020723b_RCR_978740
```

```
▶ 1 a qaw3 maak5-hung1 qaw3 maa2 mèè4
          take CM.FRUIT-papaya take come IMP.UNIMP
          bring (me) a/the papaya
```

```
▷ 2 b qaw2 ((passing the papaya to A))
           take
           (here) take (it)
```
# **3.2 Non-minimal sequence**

Recruitment sequences sometimes feature more than one initiating move. Often this is because a first attempt does not get a response, and so the Move A part of the sequence is redone. This happens in the following case, in which Person A is asking her father to pass her a knife. Both attempts are done using an interrogative formulation, with the second attempt being done in more specific fashion than in the first attempt.

(4) CONV\_020723b\_RCR\_970010

```
▶ 1 a miit4 dêê3 phòq1
          knife Q daddy
          the knife Daddy?
  2 b ((no response))
▶ 3 a phòò1 miit4 thaang2 lang3 caw4 mii4 bòò3
          father knife way back 2SG.POL exist QPLR
          Dad a knife behind you, is there (one)?
▷ 4 b nii4 nii4
          here here
          here here
```
((reaches behind to look for the knife, finds it, passes it towards A))

Another reason a recruitment sequence can be extended beyond the minimal structure is that Person B may immediately delegate to another person, rather than carrying out the action herself (see also Floyd, Chapter 3, §3.3.1; Blythe, Chapter 7, §4.2.2). In the next example, when Person B is asked to go and get some trays in preparation to serve food, she does not carry out the action. Instead,

### N. J. Enfield

she turns to her younger sibling – Person C – and re-issues the initiating move, which Person C then immediately fulfills.

```
(5) INTCN_111202n_RCR_989020
```

```
▶ 1 a sòòng3 phaa2 nan4 song1 khaw5 maa2 haa3 kan3 ((to B))
          two tray.table DEM.EXT send enter come seek COLL
          those two tray tables, bring them in here together
```

```
▶ 2 b paj3 qaw3 maa2 ((to C, eye-pointing to trays))
          go take come
          go get (them)
```
▷ 3 c ((goes and gets trays))

# **3.3 Subtypes of recruitment sequence**

The recruitment sequences collected for this study were divided into four categories, distinguished by the kind of behavior they would elicit from Person B (see Chapter 2, §6 for discussion). The four categories are: (i) a service (such as turning off a switch), (ii) transfer of an object (such as a papaya), (iii) altering a current trajectory of action that Person B was on (such as telling someone to stop pouring), and (iv) assistance with some trouble that Person A was perceptibly experiencing (such as holding a door open for someone whose hands are full). Cases of service and object transfer are amply illustrated in above examples, and elsewhere throughout this chapter. We now illustrate the other two categories.

Following is an "alter trajectory" example. In this case, Person B is about to sit down on a rickety railing that appears unlikely to be able to bear his weight without breaking. Person A calls out repeatedly 'don't sit down!'. B responds by altering his trajectory of behavior, desisting from his path of sitting down, instead moving to sit elsewhere:

(6) INTCN\_030731b\_441300

```
1 ((B is going to sit on a weak railing))
▶ 2 a jaa1 paj3 nang1 dêj2 han5
         NEG.IMPV go sit FAC.NEWS DEM.DIST
         don't sit down
▶ 3 jaa1 paj3 nang1 dêj2 han5
         NEG.IMPV go sit FAC.NEWS DEM.DIST
         don't sit down
▷ 4 b ((desists from going to sit down on a rickety railing))
```
Next is a "trouble assist" example. Person A is preparing a salad-type dish, putting ingredients into a large pestle. She is holding the pestle in her hand. This

### 6 Recruitments in Lao

type of dish needs to be tossed prior to serving, and this is normally done using a spoon and a mortar-and-pestle in combination. At the moment in the interaction that we are focusing on in this example, Person B is looking directly at Person A, and can see that Person A does not have a spoon (Figure 1*a*). Rather than waiting for person A to ask, or letting them find a spoon themselves, Person B looks for a spoon (Figure 1*b*), locates one, picks it up, and places it in reach of Person A (Figure 1*c*), where Person A is subsequently able to pick it up and use it (Figure 1*d*).

### (7) INTCN\_030731b\_385660


Another example of the trouble assist type is discussed in Enfield (2014: 42). In the example described there in more detail, Person A is walking up a steep staircase with his arms full, holding a full basket of laundry. He approaches a nearly-closed safety gateway at the top of the stairs, which is blocking his way. He does not have a free hand to open the gate and pass. Seeing this, Person B – who is sitting at the top of the stairs with the gateway within arm's reach – does not wait for Person A to say anything, but reaches out to the gate and holds it open for Person A.

While these trouble assist cases are obviously not requests as such, they are recruitments as defined for the purposes of this study. For Person A to get Person B to do something, it is not necessary that their Move A is an on-record or intended signal for Person B. What is important is that Person B acts upon a sign, in the broadest sense, from Person A, and does so with an action that is, in some relevant sense, *for* Person A. Whether Person A means it or not, in these cases Person A's behavior results in Person B doing something for them.

This phenomenon relates to the kinds of action we would call "offers": often, when one person states a problem, another will offer to help. In that sense, offers are seldom truly initiating moves, but are occasioned by certain types of prior move (see also Curl 2006). In the case just described, Person B does not offer to help. Rather, they simply do the helping action in response to the prior move that revealed the need for assistance.

Table 1 shows the relative frequencies and proportions of the four types of recruitment sequences in the Lao corpus. The relative distribution of the types is heavily skewed. Moves that elicit a service account for over half of all cases,

### N. J. Enfield

(c) Person A (seated, toward back of frame), is involved in food preparation in which the next step requires a spoon; Person B (standing) is looking directly at Person A and can see that there is no spoon at hand.

(d) Person B turns and retrieves a spoon.

(f) Person A picks up the spoon to use.

Figure 1: Trouble assist and transfer of a spoon in (7).

### 6 Recruitments in Lao

and object transfers for over a third of all cases. By contrast, alter trajectory and trouble assist recruitments are infrequent, together accounting for fewer than one out of ten cases.



# **4 Formats in Move A: The initiating move**

Initiating moves in recruitment sequences may be formulated using verbal material alone (i.e. linguistic forms including words and grammatical constructions), nonverbal material alone (i.e. visible bodily behavior), or a combination of both verbal material and nonverbal material (referred to here as composite, cf. Enfield 2009).<sup>3</sup> As Figure 2 shows, 97% of all initiating moves in recruitment sequences have a verbal component, with a third of these featuring a nonverbal component in addition.

Table 2: Modality of initiating move (Move A).


<sup>3</sup> I use *verbal* to roughly denote the linguistic, symbolic, lexico-syntactic, vocal behavior in these Lao data, and *nonverbal* to roughly denote the visual, manual, gestural behavior. I use this distinction in the usual common sense way, despite known problems making the distinction definitive (see Enfield 2009).

### N. J. Enfield

# **4.1 Purely nonverbal initiating moves**

While fully nonverbal initiating moves are rare, they do occur. In an example, Person A points to a bag that had tamarind in it, which people present had been snacking on. Person B responds by stating that there is none left in the bag, thus orienting to the pointing gesture as something like a request to pass some of the tamarind.

```
(8) INTCN_111204x_RCR_495541
▶ 1 a ((points at item))
▷ 2 b bet2 lèèw4
          finished PRF
          (it's) finished (('there's none left'))
```
For another case, see Enfield (2013: 19–21, 46): Person A crawls forward in the direction of a basket that contains betel nut paraphernalia, and B responds by passing the basket to her and saying 'you'll chew?'. The behavior of crawling forward and reaching toward the basket was understood to be at least an attempt to obtain the contents of the basket to chew, and was perhaps even designed to elicit the other person's help.

It is notable that the initiating moves that were fully nonverbal include all of the trouble assist examples.

# **4.2 Types of nonverbal behavior in initiating moves**

As noted above, around a third of all initiating moves in recruitments in Lao have a component of visible bodily behavior. These forms of bodily conduct are of course quite varied, but there are some recurring types of visible behavior, as shown in this table.



A large number of examples involve pointing gestures, either by hand or some other vector-projecting body part (eyes, lips, etc.). These gestures often help to

6 Recruitments in Lao

locate something that is being asked for, or they may help to otherwise clarify what is intended. For example, in (5) above, a speaker eye-points to some tray tables as she asks her younger sibling to go and get them. For other examples in this chapter involving pointing, see (8) and (21).

Another common visible behavior accompanying initiating moves is for Person A to hold something out towards Person B. The following is a typical example, in which a man asks his son to cut some rattan shoots, while holding out the knife that he should use.

(9) INTCN\_111204x\_RCR\_48410


Figure 2: Person A (man in foreground, to right of frame) holds out a knife as he says 'cut ((that))' to Person B (man seated further back, to left of frame).

The third major category of visible behavior that accompanies initiating moves is reaching for something, usually an object that is being requested. See, for example, (18) below, in which Person A is asking for a piece of medicinal root as she holds out her hand, as if reaching to receive it.

### N. J. Enfield

# **4.3 Verbal elements**

### **4.3.1 Major sentence types**

In terms of linguistic form, a majority of the initiating moves in this Lao collection are full clauses marked as one or another of the three main sentence types: declarative, imperative, interrogative. As Table 4 shows, the relative frequency of these types is heavily skewed. Imperative forms account for around four fifths of all cases, with interrogatives and declaratives far less frequent.

Table 4: Sentence type.


In the following example of declarative formatting, Person A is sitting close to a large pot with live fish at the bottom of it. The water level in the pot is too low. Person B starts pouring water into the pot. As the water level rises, the fish start to thrash about, and water splashes onto Person A. He states '(that's) enough'. This assertion results immediately in Person B desisting and moving back from the pot.

(10) INTCN\_111203l\_618100

▶ 1 a qeej4 phòò2 lèèw4 – huaj5 yeah enough PRF – INTJ.ANNOYED hey, (that's) enough, gosh! ((moves body back away from pot that is splashing water from the fish))

▷ 2 b ((stops pouring water into the pot and moves back))

For another example of declarative formatting, see (20) below, in which Person A's statement 'you're blocking your brother' is an attempt to get Person B to move away. These examples of declarative formatting illustrate the indirect strategy by which people can get people to do things simply by describing a problem that needs solving. When Person A describes a problem, a cooperative Person B may respond by fixing that problem (see also Kendrick, Chapter 4, §4.2.3; Rossi, Chapter 5, §3.3.4; Baranova, Chapter 9, §3.3.3; Dingemanse, Chapter 10, §3.2.2).

6 Recruitments in Lao

Cases with interrogative formatting in the Lao data are mostly of two types. One type asks as to the existence or whereabouts of an object that Person A wants. For example, here Person A wants some betel nut to chew. She first asks 'permission' (a kind of ritual preliminary to issuing a request), and then asks 'is there anything to chew?'.

(11) INTCN\_020727a\_197007

```
▶ 1 a beng1 dee4 qanuñaat4 dèè1 ( )
          look FAC.ONRCD permission IMP.SOFT
          look, if I may ( )
▶ 2 khiaw4 maak5 mii2 ñang3 khiaw4 bòò3
          chew betel exist anything chew QPLR
          (I want to) chew betel nut, is there anything to chew?
          ((looking around for betel nut, grabbing hold of basket herself))
```

```
▷ 2 b qoo4 mii2 laø.bòò3 ((allows A to proceed))
          INTJ exist of.course
          oh, yes of course
```
See also (4) above. In that example, Person A wants a knife for food preparation. She first asks Person B (her father) a very general question, roughly 'the knife?', following it up with a more specific question 'Dad is the knife behind you?'. He then reaches back to retrieve the knife and pass it to her. Questions about where an object is, or whether it is available, are appropriate in precisely those situations in which the question is apposite – namely, when it is not known that the object can be provided or not (see also Floyd, Chapter 3, §3.3.3; Rossi, Chapter 5, §3.3.3).

In a second kind of question that is used for getting others to do things, a question may serve to somewhat indirectly draw attention to a problem that needs solving. In the next example, Person A is sitting at a neighbor's shop stall, watching as her neighbor threads pieces of meat onto skewers and piles them up, in preparation to grill meat for sale at her stall. Her question – 'why don't you ever grill any of these?' – can be interpreted as an oblique way of implying that Person A would like some to eat, and suggesting that Person B start grilling the skewered meat.

(12) INTCN\_111204q\_RCR\_15060

▶ 1 a khùù2 bòò1 piing4 cak2 thùa1 why NEG grill any time why don't you ever grill any of these

<sup>▷</sup> 2 b ((no response, continues threading meat onto skewers))

### N. J. Enfield

The imperative sentence type is the dominant one used in getting people to do things. In this sentence type, there is usually no subject (or if there is one, it is a form of person reference referring to the addressee, the intended agent of the requested action), and the verb has no marking for aspect or modality. In many cases, there is no further marking, while in others there is explicit marking by means of a sentence-final particle from a dedicated system of such particles.<sup>4</sup> The different particles allow speakers to denote a range of subtle or not-so-subtle distinctions in features such as expectation of compliance, minimization of imposition, and urgency (see Enfield 2007:63ff for detailed explication). Table 5 gives the figures for the forms of marking that occur more than five times in the Lao data (accounting for 124 cases).


Table 5: Forms of marking.

Finally, there is a dedicated negative imperative marker *jaa1*, meaning 'desist'. There are three cases in my corpus: see (6) above, and the following example. In this example, a group of people are seated in a village home, eating and talking. They are seated in a circle, without much space between them. On this occasion, one of the people, a middle-aged man, who is a son-in-law to the household, has been a guest in the house and is preparing to leave the village and not return for an extended period. Extended family members are gathering on this occasion. The man's niece wants to sit close to him, and she begins to push into the space next to him, requiring people to shift and make space. Her father (the man's brother-in-law) reacts by telling her not to go too close (this is an example of an "alter trajectory" type of recruitment). She ignores this, which is to say that she simply continues her trajectory of action.

<sup>4</sup>Cf. Floyd, Chapter 3, §3.3.2 for a similarly extensive imperative system based on morphological marking.

6 Recruitments in Lao

### (13) INTCN\_111204x\_RCR\_196441


### **4.3.2 Additional verbal types**

Here I note two further types of linguistic form that were used in initiating moves in the Lao data. First is the "no predicate" type, in which someone simply refers to the object being requested. Here is an example.

(14) INTCN\_111201k\_RCR\_343251

▶ 1 a phaa2 khaw5 lêk1 hanø naø luuk4 tray.table rice steel DEM.DIST TPC child the steel tray table, child

▷ 2 b ((outside the room, eventually returns with the tray table as requested))

It is worth noting the use here of the kin term *luuk4* 'child' as a vocative. This may help contribute to the understanding that the speaker is seeking to mobilize the child's assistance.

Second is the "bare vocative" type. In this type of utterance, a person is summoned by saying their name. That is, calling out 'John!' is functionally equivalent to saying 'John, come here!'. In (16), a foreman wants his tradesmen, who are working in a nearby building, to come and assemble for lunch. He calls out *saang1 qeej4* 'hey tradesmen!'. In another case (INTCN\_111202s\_980631), a girl is at her family rice fields, and wants her older brother, who is in a paddy field a hundred or so meters away, to come and help with a task. She calls *qaaj4-dong3* 'elder brother Dong!'. This would mobilize him to go and help her.

# **5 Formats in Move B: The responding move**

There is a range of things that Person B can do in the response move of a recruitment sequence. Table 6 gives a breakdown, from the 181 cases in the Lao data where it is possible to tell how initiating moves were responded to.

It is striking that nearly half of all cases are "no uptake" or "other". This may seem to imply that requests and similar actions are ignored half of the time (cf.

### N. J. Enfield


Table 6: Response action.

Blythe, Chapter 7, §4.2.4). But this is not what is going on here. Often it is because Person B does not hear or notice that the request is issued (recall that these are noisy village environments). Sometimes it is because people re-issue an initiating move before the other person has heard or had a chance to respond. When we restrict the count to "last of non-minimal sequence" plus "one and only" (see Chapter 2, §6), the proportions change a bit, specifically the proportion of "quickly fulfills" to "other" (Table 7).

Table 7: Response action when sequential position is "last of nonminimal" or "one and only".


A different breakdown of responses can be done using simple formal criteria. Table 8 shows the relative frequency and proportions of responses that are (i) nonverbal only, (ii) verbal only, and (iii) composite of both verbal and relevant nonverbal behavior.

The majority of responses in recruitment sequences (nearly two thirds) are fully nonverbal, and nearly three-quarters involve some form of relevant visible behavior.

### 6 Recruitments in Lao


Table 8: Response modality.

# **5.1 Fully nonverbal responses**

Fully nonverbal responses include behavior like the following:


These are common and straightforward kinds of scenario. Nothing more is done by Person B than simply complying with the desired behavior.

# **5.2 Verbal elements of responses**

The functional core of a response in a recruitment sequence is the bodily conduct that constitutes the assisting or collaborating behavior. As noted in the previous section, only a minority of the responses surveyed here have a verbal component. Few generalizations about these verbal aspects of responses are possible, but two points are worth mentioning.

First, there are cases in which the recruited action is itself a piece of verbal behavior, and not a bodily action like turning off a switch or passing something. In the following example, a mother-daughter pair (both adults) are sitting in a village house. The daughter's baby is asleep in a cradle in a nearby house. The daughter has sent a young girl to go and check on the baby, to see if it has woken up. As the young girl is walking over to the other house, the mother tells her daughter to call out to the young girl and instruct her to bring the baby over to them if it has awoken.

### N. J. Enfield

### (15) INTCN\_111204t\_769065


In another example, an elderly man is sitting in a village temple building where lunch has been prepared for a group of tradesmen who are working some distance away, in the temple grounds. He is with the tradesmen's foreman, who has just called out to the tradesmen to come and eat lunch. He wonders if the tradesmen did not hear him, and then asks if the elderly man – who he says has a suitably loud voice – could call out to them.

(16) INTCN\_111202n\_RCR\_892800


Second, rejecting a request or declining to comply is often done by verbally stating a reason for the rejection or declination. In 18 cases in the Lao data, there is a clausal statement of a reason. None of these are cases of fulfilling, or plausibly beginning to fulfill, a recruitment. Stated reasons for rejection include, for example, that the thing being asked for is not available, or that the addressee is not free to do the act being recruited. See examples of reasons given with rejections in the following section.

# **5.3 Types of rejections**

The Lao examples yield only seven cases of a response that rejects or explicitly signals that the person will not do what is asked. In most of the observed cases, the rejection or declination is done by stating a reason why Person B cannot do

6 Recruitments in Lao

what is asked of them. This aligns with the classical analysis of speech acts that refers to the felicity conditions of an action (Austin 1962). These conditions need to be presupposed if the intended speech act is to be consummated. For example, if a request is made, there are certain "preparatory conditions", including that Person B should be able to carry out the requested act (Searle 1969: 66). One way in which a person can reject or decline a request is to state or suggest that a preparatory condition does not hold (Labov & Fanshel 1977: 87–88).

In an example, Person A asks Person B to turn on the television. Person B's way of declining is to suggest that the television does not work.

### (17) INTCN\_111204t\_818990


Person B's line is formally a question, but the use of the question particle *tii4* is a way of conveying that you strongly suspect that the answer to your question is 'yes'. By suggesting in this way that the television does not work, Person B is directly attending to one of the preparatory conditions of the request or command being issued, namely that it is in fact possible to carry out the service requested (see Extract 8, in which the rejection move is done by stating that a requested food is finished up).

In a second example, Person A directly asks Person B to give them some of the herbal medicinal root that they are holding. Person B declines to do so, by stating that 'there is only one piece' of the root.

(18) INTCN\_111204x\_RCR\_296281

```
▶ 1 a qaw3 maa2 qaw3 maa2 ((reaching for requested object))
           take come take come
           give it here, give it here
```
▷ 2 b mii2 khòò5 diaw3 nùng1 exist joint single one there's only one piece

By saying that there is only one piece, Person B directly attends to one of the preparatory conditions of the request, namely that it is possible to fulfill it. Here it is not technically impossible to give the medicine, but the speaker is appealing not so much to what is possible, but to what is reasonable. When it comes to

### N. J. Enfield

goods such as medicines, Lao speakers tend to be willing to share, but in this case the addressee has only one piece of the medicine, and it is medicine that he is using to treat a current illness. His rejection appeals to the absence of a condition that would define a *reasonable* possibility to comply.

In a third example, Person A directly asks Person B to go and get a mortar and pound some papaya. Person B declines to do so, by conveying that 'there is no hurry' to do it, given the time frame of the food preparation that is going on.

(19) INTCN\_030731b\_695170

```
▶ 1 a paj3 qaw3 khok1 maø tam3 paj3
          go take mortar DIR.ALL pound go
          go and get a mortar to do the pounding
▷ 2 b qoo4 jaa1 faaw4 thòòq2
          INTJ NEG.IMPV rush INTJ
          oh, (let's) not rush.
```
Here, Person B is not disputing that the requested service is appropriate, nor that they are able to carry it out, but rather they are disputing that it needs to be done *now*.

A fourth example is from an "alter trajectory" sequence. A preparatory condition for this type of sequence is that Person B is currently engaged in a behavior that is somehow (potentially) problematic, such that it should be altered or halted. In (6) above, this condition was satisfied by the evident fact that the railing Person B was about to sit on was rickety. In the following case, Person A states that Person B is 'blocking her brother'. This kind of statement of a problem is a wellknown way of getting someone to do something (see Rossi 2018 and references therein), or at least, people may respond to such statements by helping, or at least offering to help. But in this case, Person B explicitly disputes the truth of the assertion made, thus denying that there is any problem in need of solving.

(20) INTCN\_111204x\_RCR\_153391


▷ 2 b bang3 qiñang3 kòq2 qaaj4-nik1 laaw2 hên3 dòòk5 block what Q.AGAIN elder.brother-Nick 3SG.FA see FAC.RESIST what am I blocking? Nick can see fine

In a final case, a man has been skinning catfish for some time and is now evidently tired of it, but he has not yet finished the job. He directs his wife, who is sitting nearby and also busy with laborious food preparation, to do this for him. She refuses, not by saying 'no', but by asking a question 'why don't you do it?'.

6 Recruitments in Lao

### (21) INTCN\_111203l\_682150

```
▶ 1 a qaw3 nang3 maa2 saj1 phii4
          take skin come put here
          put the skin ((of the fish)) in here
          ((pointing in direction of the fish skin, then to the bowl where it is to
          go))
▷ 2 b caw4 khùù2 bòò1 hêt1 san4
          2SG.POL why not do so
          why don't you do it
          ((pointing to fish that A already has in front of him, and could skin by
          himself))
```
Her question challenges a key presupposition of the initiating move by Person A, namely that her husband cannot (reasonably) do the action himself. This comes across in the context as a blunt refusal, yet it is still done using an indirect strategy.

The various forms of rejection observed in this section have the "indirect" quality that would be predicted by well-known social theories of language use. Brown and Levinson's (1987) theory of politeness predicts that "face-threatening acts" such as refusals will be more likely handled by off-record means. Instead of saying 'no' in the above cases, people instead give reasons, in the form of a reference to a problem with a preparatory condition for the speech act in question.

# **5.4 Acknowledgment in third position**

Lao speakers in informal family and village settings seldom say 'thank you' or anything resembling it. There are only two cases in the corpus in which there is arguably an acknowledgment by Person A that Person B has fulfilled a request or otherwise assisted. In both cases, this acknowledgment is a simple interjection of confirmation, meaning 'yes' or 'that's right'.

(22) INTCN\_111203l\_636171

```
▶ 1 a qaw3 nii3 paj3 kaj3-kaj3
          take this go RDP -far
          take it far away
▷ 2 b ((picks up pot to move it))
  3 a qee5
          yeah
          yeah ((= yes, that's right))
```
### N. J. Enfield

```
(23) INTCN_020727a_559100
```

```
▶ 1 a ((crawls forward in direction of basket))
▷ 2 b caw4 khiaw4 vaa3 ((passes basket to A))
          2SG.POL chew QPLR.INFER
          you'll chew?
  3 a mm5
          yeah
          yeah ((= yes, that's right))
```
The data in this study are from highly informal settings. Acknowledgments of compliance or assistance are almost entirely non-existent in these settings, and when they do happen, as in these cases, they are not of the 'thank you' variety. Lao speakers do have a way of saying 'thank you' – the term is *khòòp5 caj3* – but it is restricted to more formal situations, or when speaking to strangers.

Other kinds of third position uptake practices following compliance moves in recruitment sequences are not frequent in this action context either. The following examples are typical of the Lao data in that they do not feature any acknowledgment following Move B.

(24) INTCN\_111203l\_644660


3 ((interaction continues))

6 Recruitments in Lao

# **6 Social asymmetries**

Social asymmetries in Lao social interaction can be defined in terms of a metaphor of height (see Enfield 2015b). In most dyads, one person is considered to be socially "above" the other person. Naturally it is not always a straightforward judgement as to who is above whom, given the sometimes fluid and contestable nature of social relations. But in the kinds of home and village settings focused on in this study, the social order is clear.<sup>5</sup> The core measure of social asymmetry in dyads is the relative birth order of siblings, and associated practices, many of which are linguistic in nature (Enfield 2015b). In the home and village, there is no ambiguity as to how most people relate to each other within this height-based conception of social difference. People are either related by kin or they are classified as such.

Where it was possible to determine the social asymmetries between dyads in the data described in this study – the three possibilities being that Person A is higher than Person B, the two are equal in status, or Person A is lower than Person B – here is what I found (Table 9).


Table 9: Social asymmetry.

Only one in five recruitment sequences features a lower-ranked person getting help from a higher-ranked one. Three in five are issued in a downward direction. This suggests support for Brown and Levinson's (1987: 69–74) flow-chart model by which people select from among various options when planning to carry out potentially face-threatening acts. At the first point of choice in their model, if a person judges that the potential threat to face is particularly high, they can choose not to carry out the act at all. This is arguably what accounts for the lower frequency of requests and similar actions directed toward higher-ranked people (see also Floyd, Chapter 3, §6; Baranova, Chapter 9, §6; Dingemanse, Chapter 10, §5.2).

When Lao speakers get lower-ranked people than themselves to do things, this is not just a preference, it reflects a strong asymmetry in entitlements (to expect

<sup>5</sup>This is not to say that people follow its associated linguistic norms to the letter; the norms can be flouted, negotiated, and contested in numerous ways.

### N. J. Enfield

assistance from lower-ranked people) and obligations (to provide assistance to higher-ranked people). This is especially apparent in cases of *delegation*: Person A asks lower-ranked Person B to do something, and Person B immediately delegates the task to Person C, who in turn is ranked lower than B (see also Blythe, Chapter 7, §6).

In a case from a family food preparation scene, when Person B is asked to go and scoop some jugged fish and bring it to use in cooking, she does not carry out the action. Instead, she turns to her younger sibling – Person C – and re-issues the command, which Person C then immediately fulfills.

### (27) CONV\_020723b\_RCR\_126590


Another example of delegation by lower-ranked Person B to a yet lower-ranked person is (5) above.

# **7 Conclusion**

This survey of semiotic resources for getting people to do things in Lao has concentrated on home and village interaction. The observations made here are not claimed to hold for the full range of contexts and domains in which Lao speakers operate, such as the formal and institutional settings that people sometimes find themselves in. That said, the informal home and village contexts discussed here are arguably the dominant ones in ordinary people's lives, and therefore require the core set of practices that any member of the Lao-speaking community should command. The overview presented here is therefore offered as a reference point for further work in this area.

Taken together, the above-described practices that Lao speakers use in getting each other to do things show two striking properties. First, they are varied and textured in kind: Lao speakers draw from a range of semiotic options (linguistic or otherwise) for formulating their moves in recruitment sequences. Second, when observed in operation in a corpus, these sets of options show characteristic properties of a functional system. The numerous statements of relative frequency

6 Recruitments in Lao

of options summarized in the many data tables provided above show precisely the skewed frequency distributions that are typical of functional systems across widely varying domains, from national economies to academic citation patterns to TV remote control handsets. Here we see the Pareto Principle – or the Law of the Vital Few and the Trivial Many – at work (Pareto 1971; see also Zipf 1949).<sup>6</sup> While many tools are available, a small number of them will carry the greatest functional load for those who use the system.

# **Abbreviations**


# **References**

Austin, J. L. 1962. *How to do things with words*. Oxford: Clarendon Press.

Brown, Penelope & Stephen C. Levinson. 1987. *Politeness: Some universals in language usage* (Studies in Interactional Sociolinguistics 4). Cambridge: Cambridge University Press.

Curl, Traci S. 2006. Offers of assistance: Constraints on syntactic design. *Journal of Pragmatics* 38(8). 1257–1280. DOI:10.1016/j.pragma.2005.09.004

<sup>6</sup>The phrase "Law of the Vital Few and the Trivial Many" is attributed to Joseph M. Juran in 1941.

### N. J. Enfield


### 6 Recruitments in Lao

diversity in the expression of gratitude. *Royal Society Open Science* 5(5). 180391. DOI:10.1017/S0047404516000385


# **Chapter 7**

# **Recruitments in Murrinhpatha and the preference organization of their possible responses**

# Joe Blythe

Department of Linguistics, Macquarie University

This chapter describes the resources that speakers of Murrinhpatha use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Murrinpatha, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Murrinhpatha with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

This chapter presents a first survey of recruiting moves and their responses in informal face-to-face conversation conducted in the Australian Aboriginal language Murrinhpatha. I begin by introducing the language and its speakers, and by discussing the corpus that informs this collection. In §2 I then illustrate some basic recruitment sequences and present the recruitment subtypes that we consider in the larger comparative project. In §3, I present the formats used as recruiting

Joe Blythe. Recruitments in Murrinhpatha and the preference organization of their possible responses. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 231– 280. Berlin: Language Science Press. DOI:10.5281/zenodo.4018382

moves, while in §4 I present the formats used as responses. The survey reveals a hierarchically governed array of responses, including structurally preferred compliant responses, as well as a range of dispreferred refusal formats, which either overtly or implicitly reject the recruitment. In §6 I discuss the possible effects of social asymmetry on recruitments in Murrinhpatha before concluding the chapter in §7.

# **1.1 The Murrinhpatha language**

Murrinhpatha is an indigenous regional lingua franca spoken by approximately 2700 people in Wadeye, Nganmarriyanga and in various smaller communities within the Fitzmaurice and Moyle Rivers region of Australia's Northern Territory (see Figure 1). It is spoken by people affiliated to the Murrinhpatha, Marri Ngarr, Marri Tjevin, Marri Amu, Magati Ke, Ngan'gityemerri and Jaminjung languages who, prior to the 1940s and 50s, would have been multilingual hunter-gatherers. Today all Aboriginal people in this region speak Murrinhpatha natively on a daily basis. It is one of only 18 traditional Australian languages still being acquired by children (AIATSIS 2005: 3). Until they encounter English at school, most children in Wadeye grow up as monolingual Murrinhpatha speakers (Kelly et al. 2010; Forshaw et al. 2017).

Murrinhpatha is a polysynthetic, headmarking language with grammaticalized kinship inflections. Its verbal morphology is templatic (Nordlinger 2010b). Complex predicates are comprised of bipartite stems, often consisting of discontinuous morphs. Nominal entities are classifiable in terms of ten semantically transparent noun classes, which do not form the basis for verbal agreement.

Previous research has described the language's genetic status (Green 2003), its complex polysynthetic verbal morphosyntax (Walsh 1976; 1996; 1987; Street 1980; 1987; Blythe 2009; 2010a; 2013; Nordlinger 2010a,b; Mansfield 2014b; Forshaw 2016; Forshaw et al. 2017), the system of nominal classification (Walsh 1993; 1997), syntax (Nordlinger 2011; Mujkic 2013), the marking of tense, aspect and mood categories (Nordlinger & Caudal 2012), and the kinship system (Blythe 2018). Interactional research has investigated person reference (Blythe 2009; 2010b; 2013), spatial reference (Blythe et al. 2016), teasing (Blythe 2012), and other-initiated repair (Blythe 2015).

# **1.2 Data collection and corpus**

Of the seventeen Murrinhpatha interactions sampled in this study, thirteen were collected by the author between 2007 and 2012 and four were collected in 2012 by

### 7 Recruitments in Murrinhpatha and the preference organization of responses

Figure 1: The Fitzmaurice and Moyle Rivers region of Australia's Northern Territory.

John Mansfield. The recordings were made either in the communities of Wadeye, Nganmarriyanga, or on the estates of one of the local clan groups. From 3.5 hours of transcribed Murrinhpatha conversation 145 recruitments were sampled.

Most of the recordings were made on picnics in the bush, away from the noisy community of Wadeye. For this reason many of the recruitments under examination relate to procurement of cigarettes or tobacco, or to the production of billy tea. They are generally low cost, low contingency requests for imminent action. In accordance with the guidelines of the project (see Chapter 1, §4) higher contingency requests for more distant future action were excluded from the collection.

### Joe Blythe

# **2 Basics of recruitment sequences**

As defined in Chapter 1, §4, a recruitment is a basic cooperative phenomenon in social interaction consisting of a sequence of two moves with the following characteristics:


Such sequences encompass requests for objects or other services as well as directives to move or modify behavior. They also include actions that occasion assistance or collaboration without necessarily having been produced with the intention to elicit that effect.

The basic minimal sequence will be illustrated below in §2.1 while non-minimal sequences will be discussed in §2.2. The subtypes will be elaborated in §2.3. In the transcripts, ▶ and ▷ designate Move A and Move B respectively.

# **2.1 Minimal recruitment sequence**

Extract 1 exemplifies a minimal recruitment sequence. The initial move by Mary is multimodally packaged as a composite utterance (Kendon 2004; Enfield 2009). The second person singular imperative verb *nangamutkathu* in line 2 is accompanied by eye-gaze toward Lily, directing her to 'give {something} here to me'. The vaguely expressed entity of the vegetable *mi*-class is minimally specified by the accompanying gesture. Mary's outstretched hand is open, ready to receive an item small enough to be passed by hand. This is inferable as either tobacco or a tobacco product. When ready, Lily passes Mary a *larrwa*, a conical tobacco pipe, packed with tobacco (*mi beka,* line 4).

(1) Da Ngarne 20091121JBvid03\_906530\_915256


### 7 Recruitments in Murrinhpatha and the preference organization of responses

This canonical minimal sequence consists of an initiating move (Move A, M-A) by participant A and a responsive move (Move B, M-B) by participant B. These canonical minimal sequences form the building blocks for non-minimal sequences.

# **2.2 Non-minimal recruitment sequence**

Interactionally, non-minimal sequences are less straightforward than the minimal sequences. Usually their non-minimality is brought about because the initiating move is problematic, because the expected responding move is not easily complied with, or because the recruitee is either unable to fulfill or is reticent about fulfilling the recruitment.

The non-minimal sequences are numerous and varied in type. In some sequences, the responsive move (M-B) becomes an initiating move for a subsequent sequence, as a counter or deflected sequence (see §4.2.2). This sequence might also be non-minimal. In other non-minimal sequences, the expected responsive move (M-B) does not eventuate and participant A pursues a response by reissuing, modifying, or elaborating upon the prior move (M-A<sup>2</sup> ). Alternatively (or additionally), there may be contingencies to be attended to by participant B before the responsive move can be produced. Thus, before committing to comply with a request, the recruitee might need convincing that s/he is capable of performing the requested action. This is exemplified in (2).

The three young men in (2) speak very little English and have few dealings with white people. Because they have no tobacco, one of them, Dave, tries to encourage Dom to procure some from a white man standing nearby.

### (2) Ngandimeli 20120715\_JB\_video\_GYHM100\_02\_51660\_68736

```
1 dav ngawu!
        hey!
 2 (1.2)
▶ 3 dav tjewirndurt thanadharrpu [mi beka ngarra ku karrim pangu warri ]
        tje -wirndurt tha -rna -dharrpu
        2SG.S.poke.RR(21).FUT-arise 2SG.S.poke(19).FUT-3SG.M.IO-ask
        mi beka ngarra ku karrim pangu warri
        NC:VEG tobacco LOC NC:ANM 3SG.S.stand(3).EXIST DIST Fa/So
        son get up and ask the white bloke standing there for tobacco
▶ 4 [((turns head, lip-points)) Figure 2 ]
 5 dom [((turns head to follow Dave's gaze)) ]
```
6 (0.4)

### Joe Blythe

```
▶ 7 dav narnawu:; (0.6) manitjpirr charge up ngamanu
         na -rna mani-dhatjpirr charge_up ngama -nu
         2SG.S.say(8).FUT-3SG.M.IO like-INTS recharge 1SG.S.say(34).FUT-FUT
         tell him something like "I'll become more lively
▶ 8 mi ngurduwinungi kardamatha (.) mangini pirditjme ngengerrennimenu.
         mi ngurdu -wi -nu =ngi kardamatha
         NC:VEG 1SG.S.shove(29).FUT-smoke-FUT=1SG.S.sit(1).FUT right_here
         mangini pirditjme nge -ngerren -neme -nu
         similar long_time 1DC.EX.S.sit(1).FUT-be_speaking-PC.M.NSIB-FUT
         as I sit here smoking (.) and thus we'll be able to sit and talk for ages"
  9 (0.4)
▶ 10 dav [kardu pa↑tha::; ]
          kardu patha-wa
          NC:HUMAN good –EMPH
         {he's a} good bloke!
▶ 11 [((points with thumb))]
  12 (0.2)
  13 dom I want- (0.2) give me smoke (.) I you:: (0.4) [(fix one and)] story.
  14 dav [( )]
  15 dav [yu.
          yeah.
▷ 16 dom [((gets up to go))
```
Having secured Dom's recipiency with a summons (line 1), Dave lip-points toward the white man nearby (line 4) and tells Dom to get up and ask him for tobacco (line 3, M-A<sup>1</sup> ). When Dom does not move after 0.4 seconds, he adds that he should provide the following rationale for providing tobacco: namely, that the boys will be revitalized and able to sit and talk for much longer (lines 7 and 8, M-A<sup>2</sup> ). When Dom (although smiling) still does not move after 0.4 seconds, Dave reassures him in line 10 that the white man is a good bloke (*kardu pathawa*, M-A3 ) 1 . Before complying with the request, at line 13 Dom rehearses what he will say to the white man in English. As Dave ratifies this rehearsal as adequate (line 15), Dom fulfills the request (line 16, M-B) by getting up to go and ask. Here the

<sup>1</sup>The nominal *kardu* class ordinarily pertains to Aboriginal people who can be related to as actual or classificatory kin. Non-Aboriginal people (social outsiders, effectively) are ordinarily grouped with animates in the nominal *ku*-class (Walsh 1997; Blythe 2015). Dave's initial reference to the white man at line 3 is with the *ku* classifier (*ku karrim pangu*, approximately 'the non-Aboriginal standing over there'). In the subsequent reference at line 10 Dave refers to him as *kardu patha*, literally 'good Aboriginal person'. The shift in classifier signals a pragmatic construal of the erstwhile alien as, for all intents and purposes, *kardu darrikardu* 'a fellow countryman', and thus as someone who can effectively be coerced into providing tobacco.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

Figure 2: Dom lip-points toward the white man off-screen who has tobacco (Extract 2, line 4).

contingencies – what exactly to say to the white man in a language he seldom uses – are dealt with before the responsive move is enacted.

A handful of sequences can be considered non-minimal because they consist of an initial move by participant A (M-A) followed by two responsive moves by participant B. The first of these responsive moves (M-B<sup>1</sup> ) expresses B's commitment to fulfill the recruitment, whereas the second (M-B<sup>2</sup> ) constitutes the actual fulfillment. We will encounter two of these three-move sequences below in (14).

# **2.3 Subtypes of recruitment sequence**

As outlined in Chapter 2, §6, recruitments mobilize a range of cooperative actions, which can be broadly categorized as being of four subtypes: i) service provision, i.e. performing a practical task for or with someone, ii) object transfer, i.e. giving someone an object, iii) alteration of trajectory, i.e. changing or stopping one's behavior, and iv) trouble assistance, i.e. stepping in to help someone in response to current or anticipatable trouble. Table 1 shows the relative proportions of the various recruitment subtypes within the Murrinhpatha corpus.

Service provision and object transfer sequences have already been exemplified in (2) and (1) respectively. The Murrinhpatha corpus contained no offers of assistance for evident trouble, possibly because all of the recordings were made

### Joe Blythe


Table 1: Relative proportions of recruitment subtypes (*n*=145).

outdoors in the open, rather than confined indoors where people may need, for example, to make way for each other (cf. Enfield, Chapter 6, §3.3). Extract 3 illustrates an alteration of trajectory recruitment. In this case the recruitee is exhorted to not cease an activity she was already engaged in.

(3) Dingalngu 20110730\_JB\_video\_GYHM100\_04\_253128

```
1 lau ((stares behind Maggie's ear))
  2 ((reaches into Maggie's hair)) 1.9 sec
  3 ((stops and scratches her own head)) 2.7 sec
▶ 4 mag awu kuka mere nawey- (.)
         awu kuka mere nawey
         no NC:ANM-TOP NEG STRI
         no! don't sto-
▶ 5 [nangiwewaytji [kuka tjirrangiwertirt weyida.
          na -ngi -weway =tji
          2SG.S.GRAB(9).FUT-1S.DO-examine_hair=2SG.S.SIT(1).FUT
          ku -ka tjirra -ngi -wertirt-weyida
          NC:ANM-TOP 2SG.S.WATCH(28).FUT-1S.DO-delouse-continue
          keep on looking in my hair for lice
  6 ali [hm hm hm hm [ha ha ha ha
▷ 7 lau [((resumes searching for lice))
```

```
8 kar yu
      yeah
9 (0.5)
```
In (3) Laura (at line 1) appears to notice something behind Maggie's ear (presumably, a louse) so reaches into her hair (line 2) to search for it. At line 3 she stops reaching and scratches her own head. At line 4 Maggie begins a negatively framed recruiting turn that is truncated midway through the verb. The negative morphosyntactic framing is replaced in self-repair by a positively framed recruit-

### 7 Recruitments in Murrinhpatha and the preference organization of responses

ing component which exhorts Laura to continue searching for the louse.<sup>2</sup> Laura resumes the search (line 7) before Maggie has even finished articulating her recruiting turn.

# **3 Formats in Move A: The recruiting move**

In multiparty interaction, two key dimensions of recruitments are the question of who is being recruited, and how that person comes to recognize what they are being recruited for. The successful recruiting move must address both the person-selection dimension (Lerner 2003) and the action ascription dimension (Levinson 2013). These dimensions can be separately handled through the visuocorporal modality, through the audio-vocal modality, or jointly handled through both as a composite, multimodal utterance. The *move* is the fundamental unit of social action within interaction (Enfield 2009; Goffman 1981). This semiotically rich unit is more often than not multimodal, that is, comprises verbal and kinesic components (see also Kendon 2004). In this paper, both kinesic behavior and spoken behavior are represented in the transcripts. I will be considering both person-selection and action ascription dimensions of the recruitment, as well as functional distinctions between the various forms of the recruiting moves.

# **3.1 Nonverbal behavior in recruiting moves**

Of the 145 recruiting moves in the collection, 92 (63%) have a seemingly relevant kinesic component. These nonverbal components include pointing, reaching out a hand to receive an object, holding out an object for a recipient to take, as well as iconic and conventionalized gestures.

Eye gaze and/or body torque toward the targeted recruitee can be critical in achieving the person-selection dimension of recruitment. Thus in (1) Mary manages the person selection issue by gazing toward Lily and reaching her arm out in her direction.<sup>3</sup> Other examples where person selection is successfully managed through eye gaze and physical embodiment include (4), (13), (14), (16) and (17).

<sup>2</sup>The repairable is not easy to translate. In all likelihood, the animate *ku* classifier is used to evoke the louse. The negatively framed repairable appears to have been shooting for something like 'don't stop searching for the critter'.

<sup>3</sup>The same arm also manages aspects of the action-ascription dimension. The open hand is ready to receive a small passable object.

Joe Blythe

# **3.2 Fully nonverbal recruiting moves**

Six of the 92 recruiting moves incorporating kinesic components were delivered entirely without speech. Extract 4 exemplifies this phenomenon.


Figure 3: Bruce, gazing at the spoon in Dave's cup, rubs his index- and middle-finger against his thumb (Extract 4, line 3).

In (4) Dave's left hand holds a cup of tea that is sitting on the ground. The cup has a spoon in it. At line 1 Dave is announcing his intention to convey a

### 7 Recruitments in Murrinhpatha and the preference organization of responses

message in the future to somebody who is not present. As he does this, Bruce fills his cup with tea (line 2). When he finishes this, he turns to face Dave and rubs his index- and middle-finger against his thumb (line 3, Figure 3). This abstract (and perhaps conventionalized) gesture is at least partly indexical in that it is oriented toward the spoon in Dave's cup – as is Bruce's eye gaze. Dave pauses as he passes the spoon to Bruce (line 4), and then resumes his announcement (line 5). The momentarily suspended lexico-syntactic channel belies no evidence for there even being a recruitment, as this sequence takes place entirely within the visuo-corporal modality.<sup>4</sup> Bruce manages the person-selection dimension of the recruitment by twisting his body and gaze toward Dave (and the spoon in his possession) and away from co-present Phil.

# **3.3 Verbal elements: construction types and subtypes**

In this section we consider the various grammatical structures that best characterize the verbal components of recruiting moves. In addition to the three basic sentence types, imperative, declarative and interrogative, we also consider those that lack a predicate altogether. The relative proportions of these construction types are given in Table 2.


Table 2: Proportions of construction types in the sample that include a verbal component (*n*=139).

### **3.3.1 Imperatives**

Imperative constructions are the most frequent of the verbal components of recruiting moves. Because they explicitly name the action to be performed and because the grammatical form of the predicate indexes the elicitation of action

<sup>4</sup>However, the recruitment sequence is at least partly evidenced by prosodic lengthening of the 'human' classifier *kardu*::: followed by the 0.9s of silence in line 5. This combination does suggest possible nonverbal activity.

### Joe Blythe

(Lyons 1977: 774–78; Sadock & Zwicky 1985: 170–71), in this collection they are the most overt, on-record method for recruiting action. Those that have second singular subjects (the majority) are used to single out the person being recruited. The imperative mood is morphologically distinguished from other moods. Both future and past indicative, as well as future and past irrealis moods are doublemarked within the template of the polysynthetic verb; firstly in the initial portmanteau classifier stems, and secondly in a morphological slot dedicated to marking TAM distinctions. This is not the case, however, with imperatives. In the imperative mood, the dedicated TAM slot remains unfilled (Nordlinger & Caudal 2012).

We have already encountered second person singular imperatives in (1) (*nangamutkathuya*, line 1) and (2) (*tjewirndurt* and *thanadharrpu*, line 3, and *narna*, line 7). Table 3 compares the imperatives *tjewirndurt* and *thanadharrpu* to their future indicative counterparts. The imperatives lack the future tense morpheme *-nu* that otherwise appears within future indicatives.


Table 3: Imperative forms compared with their future indicative counterparts; future indicative forms are doubly marked for future tense.

While the participation framework evoked by an imperative predicate with a second person singular subject will convey that specific addressing is being performed, recipients' identification of the intended target, within a multiparty setting, hinges on the particular person selection devices which accompany the predicate.<sup>5</sup> In (1) Mary's eye gaze and outstretched arm toward Lily serves to

<sup>5</sup>Lerner (2003: 182) suggests that the second person pronoun *you* is a "recipient indicator" but not a "recipient designator". "[S]peakers can indicate that they are addressing a specific participant in a manner that does not itself reveal who that individual is" (ibid: 183). In multiparty interaction, who specifically is being addressed through the use of the pronoun is managed through eye gaze or some other device, or inferentially when epistemic or deontic advantage

### 7 Recruitments in Murrinhpatha and the preference organization of responses

select Lily as the proper recipient for the 2SG.S inflected predicate *nangamutkathuya* 'give it here to me'. In (2) the kinterm *warri* 'father'/'son' (line 1)<sup>6</sup> serves to select Dom and not co-present Bruce (Dave's classificatory brother) as the intended target for the recruitment, and as the addressee for the 2SG.S inflected predicates *tjewirndurt* 'get up' and *thanadharrpu* 'ask him' in line 3. However, when recruiters are unconcerned about who specifically should fulfill the recruitment, the second person imperative predicate will have a non-singular subject.

(5) Thuykem20110901\_JB\_video\_GYHM100\_01\_381373


```
▶ 2 [((points at billycan))
```
3 phi [((removes his cap))

▷ 4 ((uses cap to take the hot billycan off the fire))

The three boys in (5) are classificatory brothers. In line 1 Dave exhorts his brothers to make some tea. The imperative verb *nangkarnuwardangu* is inflected as second person dual sibling: 'you two siblings go on and mix it up'. The nonspecific second person dual sibling subject is not accompanied by a vocative. As Dave initiates the recruitment he points at the billycan on the fire (line 1). He does not gaze at either of his two brothers. Thus, specifically which brother should concern himself with making the tea is left up to them to decide upon.<sup>7</sup> Actually,

is skewed toward a particular individual. However, as discussed in fn. 7, the inferences to be drawn from the English pronoun *you* differ from its Murrinhpatha counterparts.

<sup>6</sup>The "kinterm" *warri*, a recent innovation, is mainly used by young men. In Australian kinship systems it is very unusual for reciprocal kinterms (e.g. terms like *cousin* which apply equally in both directions, unlike *father* and *son*) to be used for persons separated by a single generation – although exceedingly common for two generations of separation. *Warri* may be a reanalyzed borrowing from the interjection *warriwarri*, which in the East Kimberley region of Western Australia and the Victoria River district of the Northern Territory (in the Jaru, Gija and Gurindji languages, amongst others), is produced as a sympathetic response by recipients who hear mention of a certain kinsman. In these languages, the term is used for fathers, sons, and other kin besides. Under similar circumstances, contrasting interjections are used for different classes of kin (McConvell 1982: 99; Blythe, Gija & Jaru fieldnotes 2016).

<sup>7</sup> In Murrinhpatha, second person predicates are marked for number (SG/DU/PC/PL), and (when DU or PC) gender (M/F), as well as siblinghood (siblings/non-siblings). In English however, the pronoun *you* is unmarked for number, or any other contrasts. This gives the languages different inferential affordances within in multiparty interaction. Upon hearing a Murrinhpatha inflected predicate with a 2SG subject, recipients can infer that the speaker *definitely* has as

### Joe Blythe

while Dave is speaking at line 1, Phil has apparently already taken it upon himself to make the tea. At line 3 he removes his cap, which at line 4 he uses to insulate his hand as he removes the hot billycan from the fire. He then goes on to make the tea.

### **3.3.2 No predicate**

Because, as the name suggests, the "no predicate" recruiting moves lack a predicate that expresses the action being elicited from the recruitee, they constitute a grab-bag mixture of structural possibilities. This category includes examples in which the sole lexical content is either an interjection or a vocative devoted to managing the person-selection dimension of the recruitment, leaving the actionascription dimension to be handled gesturally or through inference. More often however, with object-transfer requests, the object required is explicitly mentioned, as in (6).

```
(6) Nanthak2011 0828_JB_video_GYHM100_03_472600_479711
  1 kar ay kuraka djiwa karrimbuk[tharr.
         ay kura -ka djiwa karrim -buktharr
         oh NC:WATER-TOP that 3SG.S.stand(3)NFUT-be_red
         ay that tea is too strong ((too red))
▶ 2 ali [yawu munak [kura path- pathayu]=
                               yawu munak kura STRI patha=yu
                               hey sister NC:WATER STRI good =CL
                              hey sis, fresh water,
▶ 3 [ ((points to car))]
▶ 4 =murruwurlnyingka
         murruwurl-nyi -ngka
         beautiful-2SG.DO-eye/face
         beautiful face
  5 (0.7)
▷ 6 kar ma Rita ma nyinyirda tjewirndurttharra
         ma Rita ma nyinyirda tje -wirndurt-tharra
         but ♀name but ANAPH 2SG.S.POKE.RR(21).FUT-arise -ahead
         hey Rita, you get up for it
```
In (6) when Karen complains that the tea she is making is too strong (line 1), Alice, addressing her as *munak* 'sister', points to the car nearby and names the

specific addressee in mind; whereas the English pronoun *you* conveys that the speaker *perhaps* has a specific individual in mind – except when the participation frame is dyadic. The converse is also true for Murrinhpatha. When a second person predicate is *not* singular, then the inference to be drawn is that the speaker is *not* singling out any specific individual from the group of addressed recipients. Dave's gaze at the billycan, rather than at either of the two brothers, accords with the inference of *non-specificity* to be drawn from dual inflected subject.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

item required to solve the problem (*kura patha*, 'fresh water') and mitigating the illocutionary force of the directive with the compliment *murruwurlnyingka* 'you are beautiful' (line 4). Karen rejects the recruitment by deflecting it toward a somewhat younger woman (line 6).

Thirty percent of the no-predicate verbal recruiting moves (*n*=14/46) we can call "nominal-hither" constructions.These are exclusively used for object transfer recruitments. In these expressions an overt noun phrase is used to refer to the item being requested. The first element in the majority of Murrinhpatha noun phrases is the nominal classifier applicable to the relevant noun class. The nominal classifier may be followed by a noun, an adjective, a demonstrative and/or a numeral. However most Murrinhpatha noun phrases are under-elaborated: as bare nouns, as bare nominal classifiers, or as the nominal classifier plus a noun/adjective/demonstrative or numeral. If an item is being requested, eye gaze toward the desired item makes the targeted referent reasonably clear. Extract 7 exemplifies.

### (7) Nanthak2011 0828\_JB\_video\_GYHM100\_03\_879400

```
1 lil kapkathu [tepala;
        kap -gathu tepala
        receptacle-hither deaf
        the billycan here, deaf one
```
2 ali [((passes billycan to Lily))

In line 1 Lily leans toward her classificatory sister Alice and addresses her as *tepala* 'deaf one'.<sup>8</sup> The recruiting move consists of the noun *kap*, used to refer to the item being requested ('receptacle' < *cup*), here encliticized with the directional adverbial *-gathu* 'hither'. As she says this Lily gazes toward the billycan of tea which Alice then passes to Lily.

### (8) Ngantimeli20120715\_JB\_video\_GYHM100\_02\_389636

```
1 dom [warri (0.3) kurathu;
         warri kura -gathu
         Fa/So NC:WATER-hither
         Dad, a drink here!
2 [((points at billycan, Figure 4))
3 dav ((passes billycan to Dom))
```
<sup>8</sup> In face-to-face conversation, sisters and female cousins tend to address each as *tepala* 'deaf one', rather than address each other by name. This mild form of personal name avoidance does not extend to third person reference.

### Joe Blythe

Figure 4: Whilst holding an empty cup, Dom points to the billycan (Extract 8).

Extract 8 is almost identical to (7), except that rather than use a noun to specify the requested item, Dave, while pointing to the nearby billycan (line 2, Figure 4), uses the bare water-class classifier *kura* in conjunction with the 'hither' adverbial -*gathu* (line 1). In the absence of an explicit predicate, the deictic adverbial -*gathu* implies an object transfer recruitment by indicating the direction that the requested object ought to be transferred. The vast majority of these recruitments (92%, *n*= 13/14) are accompanied by eye gaze toward the object of desire.

# **3.3.3 Interrogatives**

In some languages like English and Italian interrogatives are a major sentence type utilized in recruiting moves (see Kendrick, Chapter 4, §4.2; Rossi, Chapter 5, §3.3), while in other languages like Cha'palaa, Lao, Polish, Russian, and Siwu interrogatives are much less frequent (see Floyd, Chapter 3, §3.3; Enfield, Chapter 6, §4.3.1; Zinken, Chapter 8, §3.3; Baranova, Chapter 9, §3.3; Dingemanse, Chapter 10, §3.2). In the Murrinhpatha dataset, there are only 4 recruiting moves that are built using interrogative structures. Three of these are built around the

### 7 Recruitments in Murrinhpatha and the preference organization of responses

interrogative word *ngarra* 'what'/'where', as the next examples illustrates. In (9) a multiparty conversation has undergone a schism. To facilitate legibility, extraneous overlapping talk has been removed from the transcript.

(9) Dingalngu20110730\_JB\_video\_GYHM100\_04\_231240 (transcript simplified)

```
1 ali bere memnginthawarrk (0.2) ng(h)arra
         bere mem -ngintha -warrk ngarra
         completion 3SG.S.10RR.NFUT-DU.F.NSIB-lose_oneself LOC
         the two of them got lost going
  2 (k(h)unungumng(h)intha) (0.4) ngarra Yilimu (1.0) ah ha
         kunungam -ngintha ngarra yilimu
         3SG.S.7go.EXIST-DU.F.NSIB LOC ♀name
         to where Yilimu is ((laughing))
  4 (.)
▶ 5 mag ngarra mi thawuy:.
         ngarra mi thawuy
         where/what NC:VEG chewing_tobacco
         where {is} some chewing tobacco?
  6 (1.0)
▷ 8 car mi thawuy:ka::: tjimngemardamardaka Yilimu damatha;=
         mi thawuy -ka tjim -nge -mardamarda-ka
         NC:VEG chewing_tobacco-TOP 2SG.S.1sit.NFUT-3SG.F.IO-wait_for -TOP
         yilimu damatha
         ♀name INTS
         as for chewing tobacco, you {should} really wait for Yilimu
  9 =mi wunku mi thawuy yulirn kandjinkadhukwurran.
          mi wanku mi thawuy yulirn
          NC:VEG also NC:VEG chewing_tobacco ashes
          kandjin -kadhuk=wurran
          3SG.S.22bring/take.NFUT-EXIST =3SG.S.6go.NFUT
          she has both chewing tobacco and ashes
  10 (0.7)
```
At line 5 Maggie requests chewing tobacco from anyone who might be able to provide it. She does so with the 'where/what' interrogative *ngarra*. At lines 8 and 9 Carol informs her that no-one present is able to fulfill her request and that, if she wants chewing tobacco, she will have to wait for another woman to return from fishing.

The remaining, solitary example of a polar interrogative recruiting move is not fulfilled, possibly because the polar question is produced in overlap. In Murrinhpatha, polar questions are not distinguished morphosyntactically from declaratives and, like declarative assertions, generally have falling intonation contours. Given that the linguistic cues to interrogativity are relatively thin, they may be poorly disposed toward recruiting assistance from others.

### Joe Blythe

### **3.3.4 Declaratives**

Declaratives are less direct recruiting moves than interrogatives. They do not make explicit the action being elicited. Furthermore, because they mostly have third singular subjects, they do not specify a particular target for the recruitment. As such, they generally highlight a problem. One of those present must take it upon themselves to remedy the issue, if they see fit to do so.

Extract 10 contains a non-minimal sequence. Initially Mary tries to get Edna to fill her cup with water (lines 1 and 2). At line 3 Edna implicitly rejects the recruitment, accounting for her non-compliance by both exclaiming the bottle to be empty (line 3) as well as demonstrating it to be empty by holding it up for Mary to see (Figure 5).

Figure 5: Edna holds up bottle and says *makura karrim* 'there's no water' (Extract 10, lines 3–4).

### (10) 20070728JBvid01c\_10378\_16778


### 7 Recruitments in Murrinhpatha and the preference organization of responses

▷ 5 mar ((gets up to get water))

As well as being an account for Edna's non-compliance, the syntactically declarative *makuraya karrim* also initiates a counter-recruitment. It does not specify what needs doing, nor who specifically should do it. Mary instantly gets up (line 5) and takes it upon herself to get some water.

As was the case in the previous extract, the third singular declarative predicate in (11) also does not specify a target for the recruitment. Feasibly, it might not even have been intentionally produced as a recruiting move.

```
(11) Thuykem 20110901_JB_video_GYHM100_01_810540
```

```
1 phi milkka ngarra¿
         milk-ka ngarra
         milk-TOP where
         where's the milk?
  2 (1.0)
  3 phi wurda damatha ma↑nandji marndarri.
         wurda damatha ma -nandji mam -rdarri
         NEG INTS NEG-NC:RES 3SG.S.do(8).NFUT-BACK
         there isn't any, it {must be} behind {in Wadeye}
  4 (0.3)
▶ 5 dav [awu milk karrimwa:. ]
          awu milk karrim -wa
          no milk 3SG.S.stand(3).EXIST-EMPH
          no, there is milk!
▶ 6 [((turns and gazes at camping box))]
▷ 7 bru ((stands up))
  8 dav [na:; manganart nawa:; ]
          na mangan -art na -wa
          TAG 3SG.S.grab(9).NFUT-get/take TAG-EMPH
          hey, he brought it, eh?
  9 bru [((goes to look for milk))]
```
When Phillip's inquiry about where the milk for the tea might be (line 1) yields no response after one second, he complains at line 3 that it must have been left behind in Wadeye. However, whilst turning to gaze toward the camping box where the milk ought be, Dave contradicts him, 'no', and reassures him that 'there *is* milk' (*awu karrimwa*, line 5), then further asserting that 'he' (the ethnographer) did in fact bring the milk (line 8). Upon hearing this, Bruce gets up (line 7) and takes it upon himself to retrieve it (line 9), fulfilling the recruitment at line 5 that may not have been intentionally initiated for him specifically to act upon. Feasibly, the recruitment is perhaps an incidental outcome of Dave's correcting

### Joe Blythe

Phillip's misunderstanding (and perhaps also incidental on Dave, like Phil, wanting milk in his tea).

# **3.4 Additional verbal elements**

In this section we examine additional elements within the recruiting move that are not core grammatical constituents. These might include vocative expressions like names and kinterms, interjections, benefactives, strengtheners and mitigators, and explanations.

### **3.4.1 Names, kinterms and interjections**

Personal names and kinterms used as vocatives address the person selection dimension of recruitments by picking out the intended recipient. We see personal names functioning as "recipient designators" (Lerner 2003: 182) in (20) and (22), and similarly functioning kinterms in (1), (6), (8), (15) and (21). In (7) and (23) we see similar use of *tepala* 'deaf one' as a characteristic form of address between women who are actual or classificatory sisters.

The interjection *yawu* 'hey', when used turn initially, can also be used as a recipient designator to elicit mutual eye gaze between recruiter and would-be recruitee. In line 5 of (18) the recruiter (Karen) does this before redirecting the recruitee's attention, with a point, to someone else.

### **3.4.2 Benefactives, strengtheners and mitigators**

Benefactive marking in recruitments makes explicit an alleged beneficiary. These are usually marked within the verbal template by bound indirect object pronouns; such as the first person singular -*nga* in *mi nangamardakutkathungadha* 'take a bit out **for me**' in (16), and the the first person non-singular inclusive -*nye* in *nanyengkarnu* 'mix in some fresh water **for us**' in (15). Recruiters can use first person non-singular inclusive pronouns strategically by including the addressee as a potential beneficiary of the solicited action, thus downplaying the perception of the benefit being for the recruiter alone.

Murrinhpatha deontic adverbials occur both as free-standing words or as bound morphs, some being incorporated into dedicated slots within the polysynthetic verbal template. Those that strengthen are more semantically transparent than those that mitigate. The strengtheners include the emphatic suffix *-wa* and the intensfiers *dhatjpirr* and *damatha* (as in *kura burrburr damatha* '{put in} cold water!'). The mitigating adverbials like -*ngadha*, often translated as 'for a while',

### 7 Recruitments in Murrinhpatha and the preference organization of responses

are difficult to gloss and are less well understood.<sup>9</sup> Other mitigators include adhoc compliments being paid to the recruitee (such as *murruwurlnyingka* 'you are beautiful', in line 4 of Extract 6).

### **3.4.3 Explanations**

Explanations or accounts for a recruitment may be added after the recruiting move, as in line 9 of (23), or they may precede it, as in (12).

```
(12) Nanthak 20110828_JB_video_GYHM100_03_537800_546223
```

```
1 kar [ngawu (1.0) thagilkilktharra
        ngawu tha -gilkilk-tharra
        hey! 2SG.S.POKE(19).FUT-hang -ahead
        hey! (1.0) poke {this} through the handle and carry it
2 [((picks up billycan with a stick, passing it toward Rita))
3 (0.5)
4 kar [karduka tjinengirdarribangnukun.]=panguwangu nabatjtharra.
        kardu -ka tjina -ngi -rdarri-bang -nukun
        NC:HUM-TOP 2SG.S.HEAT(27).FIRR-1S.IO-back -scald-FIRR
        pangu-wangu na -batj -tharra
        DIST-thither 2SG.S.GRAB(9).FUT-get/take-ahead
        you might scald me on the back, take it over that way
5 [ ((hands the stick to Rita)) ]
6 rit ((takes the hot billycan away to fill with cold water))
```
In (12) Rita is standing up on the beach ready to take a very hot billycan to where there is water with which to cool down the tea. At line 2 Karen picks up the billycan with a stick, placing the billy on the ground near Rita, meanwhile telling her to poke the stick through its handle in order to carry it (line 1). As she passes the stick to Rita, Karen explains in line 4 that Rita might scald her with the hot tea (*karduka tjinengirdarribangnukun*) then instructs her to 'take it over that way' (*panguwangu nabatjtharra*), through the gap where no one is sitting.

In the next section we consider the range of possible ways that would-be recruitees respond to a recruiting move, or not as the case may be.

# **4 Formats in Move B: The responding move**

A substantial body of research in conversation analysis investigates how the design of turns delivering initiating actions (Wootton 1997; Vinkhuyzen & Szymanski 2005; Curl 2006; Curl & Drew 2008; Craven & Potter 2010; Enfield et

<sup>9</sup> Some of these adverbials are translated, at least sometimes, with temporal semantics.

### Joe Blythe

al. 2010; Stivers & Rossano 2010; De Ruiter 2012; Rossi 2012; Kendrick & Drew 2016) impose constraints upon the sorts of responses they receive (Raymond 2003; Schegloff & Lerner 2009; Fox & Thompson 2010; Lee 2013; Thompson et al. 2015). In this Murrinhpatha dataset, only 46% of recruitments were either fulfilled promptly (24%, *n*=35) or indications were provided suggesting possible imminent fulfillment (22%, *n*=32). Counts on response types to particular recruiting formats do *not*, at this stage, suggest that any particular format (e.g. imperative, declarative, interrogative, etc.) is more or less likely to successfully elicit the desired response than any other format.

Just as the formats used in recruiting moves range between the overt, onrecord strategies, through to more covert, off-record strategies, so too do the range of possible responses. Overt on-record responses include both immediate compliance and relatively prompt rejection of the recruitment, while physical movements suggestive of possible compliance are more covert and less on-record. In this corpus overt on-record rejections are considerably less frequent than implicit rejections or non-fulfillments; such as counter-recruitments, deflected recruitments, and generally just ignoring the recruitment. Non-responses evade overt refusal or rejection of the recruitment. We will see evidence below that by neither complying nor committing to complying, ignoring a recruitment can usually be taken as an implicit refusal to comply.

# **4.1 Prompt or imminent compliance**

We have already encountered many recruitment sequences in which the response is physical compliance delivered relatively promptly without an accompanying verbal component (Extracts 1, 2, 3, 4, 5, 7 and 8). We have also seen in (5) how removing a hat and then removing a billycan from the fire suggest possible imminent compliance, which is ultimately followed by actual compliance. Possible imminent compliance can also be verbally hinted at without giving a commitment to actually comply, as (13) demonstrates.

(13) Nanthak 20110828\_JB\_video\_GYHM100\_03\_427300

```
▶ 1 lil ngarra kurayu.
         ngarra kura =yu
         what/where NC:WATER=CL
         where's the water/tea?
  2 ((gazes at Alice, Figure 6))
  3 (0.5)
```
### 7 Recruitments in Murrinhpatha and the preference organization of responses

Figure 6: Lily gazes at Alice (Extract 13, line 2).

Extract13 occurs near the beginning of a protracted episode of multiple recruitments, all of which deal, in some fashion, with the procurement of cold water for a very hot billy of tea.<sup>10</sup> Lily's question at line 1 (*ngarra kurayu*) is built around the 'where'/'what' interrogative *ngarra* and the bare *water*-classifier ('where is the tea/water?'). Whilst certainly a request, it can also be heard as a possible complaint. Although Lily's eye gaze is directed on Alice who is seated near the billycan, it is Karen, rather than Alice, who is preparing the tea. While Alice's reply *kuguk marrawangu* 'wait it's coming' does address the substance of the possible complaint (being slow in arriving), it does not commit to future compliance and is agnostic as to who will be responsible for ultimately fulfilling the request. The question of who will get the water remains unresolved for quite some time.

Extract 14 consists of two interlocking non-minimal sequences commencing with nonverbal recruiting moves. Each non-minimal sequence is of the threemove variety previously mentioned in §2.2, where participant B firstly commits to complying (M-B<sup>1</sup> ) with the recruitment, then actually complies with it soon after (M-B<sup>2</sup> ).

<sup>10</sup>The four, mostly elderly, women in this conversation are tired and feeling lethargic. The water required to cool the hot billycan is nearby on the beach, in a heavy 20-liter container. The women each display justifiable resistance to getting up and retrieving the water.

### Joe Blythe

At line 2 of (14) Dom (who has a cigarette in his mouth) leans forward. Whether leaning forward was intended as an offer is unclear, but either way it seems to occasion a recruiting move from Dave at line 3, where he holds his hand out to receive the cigarette. Dom's response to this recruiting move is semiotically and sequentially complex. The sweeping point from Dave to Bruce (see Figure 7) is an iconic depiction of the trajectory Dom intends the cigarette to travel along when Dave finishes with it. The drawing in the air conveys graphically that the recruitment is of the object transfer variety.

(14) Ngantimeli 20120715\_JB\_video\_GYHM100\_02\_196571

```
1 dom [mhm
```

```
2 [((leans toward Dave with cigarette in mouth))
```

```
▶ 3 dav ((holds out hand to receive cigarette))
```

```
▶▷ 4 dom ((sweeping point from Dave toward Bruce, Figure 7))
 ▷ 5 dav [nakurl ngaliwe nganamutnu. ]
           nakurl ngaliwe nga -rna -mut -nu
           later short 1SG.S.poke(19).FUT-3SG.M.IO-give-FUT
           I'll give a bit to him after
▶▷ 6 dom [((sweeping point from Dave toward Bruce))]
   7 ((takes a drag on the cigarette))
```
▷ 8 ((passes cigarette to Dave))

```
9 ((30 seconds of talk deleted, Dave smokes cigarette))
```
▷ 10 dav ((dave passes cigarette to Bruce))

This depictive point is repeated at line 6.<sup>11</sup> The gesture recruits Dave to pass the cigarette to Bruce when he is finished with it. In overlap with the repeat of the point, at line 5 Dave gives a verbal undertaking to comply with this recruitment (*nakurl ngaliwe nganamutnu* 'I'll give him the stub later'). This is the only vocal move in either of the two sequences.

Dom's sweeping points (lines 4 and 6) do more than merely recruit. By virtue of the fact that the cigarette is retained in Dom's mouth, they also can be seen as him giving an implicit undertaking to imminently comply with Dave's recruiting move at line 3. Dom's passing of the cigarette at line 8 is the eventual fulfillment implicitly promised at lines 4 and 6. Likewise, when Dave passes the cigarette to

<sup>11</sup>The repetition of the point is instantaneous and fluidly produced (and is hence more akin to reduplication than actual repetition), as if the invisible line in the air is being heavily bolded.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

Figure 7: Dom points from Dave to Bruce (Extract 14, line 4). This sweeping point is both an explicit-object transfer request and an implicit commitment to imminently comply with Dave's request for the cigarette.

Bruce at line 10, this can be seen as the fulfillment of the recruitment initiated by Dom that Dave had committed to fulfilling at lines 4 and 6.

Imminent possible compliance, or incipient compliance (Schegloff 1989; Kent 2012), can be projected visibly (as Dom does in lines 4 and 6 of Extract 14) or verbally (as Dave does in line 3 of Extract 14 and Alice does in line 4 of Extract 13). In the next section we will encounter a mixed-message example, where the physical responsive behavior contradicts the verbally delivered response.

# **4.2 Rejection and non-compliance**

The preferred response to a request, or any sort of recruiting move, is to comply with or fulfill the recruitment, or at least display that probable compliance is forthcoming. Anything less is dispreferred. The range of dispreferred alternatives is scalar. The most dispreferred alternatives are the overt refusals or rejections, which are vanishingly rare in this collection (*n*=3 from 145 recruitments, 2%). Only two refusals include the rejection token *awu* 'no'.

Of the 145 recruitment sequences in the Murrinhpatha collection, 54% (*n*=78) were not promptly complied with, nor was possible compliance projected as imminent. This may be because a request is problematic, unreasonable, or that other matters must be attended to before the recruitment can be fulfilled. The various possible alternatives to the preferred response include both explicit and implicit refusal. Delaying dispreferred responses can project that an imminent refusal is forthcoming (perhaps to be delivered with an overt rejection token), or that noncompliance is to be inferred from the silence that ensues. Other-initiated repair has the effect (whether intentional or otherwise) of delaying the expected com-

### Joe Blythe

pliance or refusal, such that potentially problematic requests become vulnerable to sequential deletion.

### **4.2.1 Overt rejections**

Overt refusals or rejections are socially dispreferred responses. As such, dispreferred responses tend to be delayed, mitigated, and accounted for (Heritage 1984: 265–80; Pomerantz 1984; Schegloff 2007: 58–96; Pomerantz & Heritage 2013).

Just prior to (15), the ethnographer poured himself a cup of hot tea from the billy and, before walking away from the scene, remarked that he likes hot unsweetened black tea. This is anathema to the four women in this extract, as they normally drink sweet white lukewarm tea from metal pannikins, which heat up when hot liquid is poured into them.

```
(15) Nanthak 20110828_JB_video_GYHM100_03_453900_460860
```

```
▶ 1 ali [munak kura pathadhatjpirr nanyengkarnu.
         munak kura patha-dhatjpirr na -nye -ngkarnu
         sister NC:WATER good -INTS 2SG.S.do(8).FUT-1NS.INC.IO-mix_into
         sister, mix in some fresh water for us
▶ 2 [((points to water-bottle/vehicle))
 3 (0.3)
▷ 4 kar [ya beremanangatha dendurr pigurdugurduk.
         ya beremanangatha dendurr pi -gurdu-gurduk
         HES never_mind.INTS hot 1NS.INC.sit(1).FUT-RDP-drink
         um, it really doesn't matter, we'll drink it hot
▷ 5 [((points into billycan))
 6 (.)
 7 ali [awu ku(h)rdu]nyidham(h)arrarrnukun[:;
         awu kurdu -nyi -dhamarrarr -nukun
         no 3SG.S.shove(29).FIRR-1NS.INC.DO-burn_throat-FIRR
         no! i(h)t might b(h)urn our throats!
 8 lil [ (h)a:wu;! ] [↓karraya;↓
            awu karraya
            no goodness!!
            n(h)o! good grief‼
```
At line 1 Alice, addressing Karen with the kinterm *munak* 'sister', tells her to mix cold water into the hot tea. Karen refuses the request at line 4. Her tonguein-cheek refusal echoes the ethnographer's earlier remark by insisting (sarcastically) that they will drink their tea hot. Despite the proposal being non-serious, the refusal is genuine. The dispreferred nature of the response is evident in the delay provided by the hesitation marker *ya*, approximately 'um'/'ah', and the

### 7 Recruitments in Murrinhpatha and the preference organization of responses

adverbial interjection *beremanangatha* 'it doesn't really matter'. The refusal to comply is implicit in the reason (albeit, a preposterous one) for not complying (we'll drink it hot!). The refusal elicits both disagreement and complaint from both Alice and Lily, whose responses at lines 7 and 8, respectively, are infused with laughter particles.

The overt refusal in (15) is verbally delivered. Furthermore Karen's physical behavior does not suggest any likelihood of her possibly complying in the future. Her physical behavior accords with her verbal behavior. However, in (16) the overt, vocally delivered dispreferred refusal is somewhat contradicted by the refuser's physical actions, which instead suggest possible imminent compliance.

```
(16) Nanthak 2011 0828_JB_video_GYHM100_03_760030_770043
```

```
1 kar (nga mi nanyemawathawarra.) ba berenguny berenguny
          nga mi na -nye -ma -watha-warra
          hey NC:VEG 2SG.S.hands(8).FUT-1NS.INC.IO-hand-make -ahead
          ba berenguny berenguny
          STRI OK OK
          (hey roll us a cigarette), oh it's alright, it's alright
▶ 2 ali [aa mi numigathungadha aa mi nangamardakutkathungadha.
          aa mi numi-gathu -ngadha aa mi
          ah NC:VEG one -hither-while Ah NC:VEG
          na -nga -mardakut-gathu -ngadha
          2SG.S.hands(8).FUT-1SG.IO-take_out-hither-while
          ah, give me one, take a bit out for me
▶ 3 [((holds out hand to receive)) -->
  4 (0.5)
▷ 5 ali [awu; mi nukunudha nginarr puleyu.
          awu mi nukunu-dha nginarr pule =yu
          no NC:VEG 3SG.M-PIMP poison_cousin esteemed=CL
          no it's from him (your) poison cousin ((FMBS))
  6 [(Karen gets out tobacco, Alice holds out hand, Figure 8))
▶ 7 ali mi mamawatha;
         mi ma -ma -watha
         NC:VEG 1SG.S.hand(8).FUT-hand-make
         I want to roll some
▶ 8 --> ((holds out hand to receive)) -->
  9 (0.7)
▷ 10 kar [thaninapartwardaya,
          thani -rna -part –warda=ya
          2SG.S.be(4).FUT-3SG.M.IO-leave-TEMP =CL
          leave it for him
  11 [((Karen looks into tobacco tin, Alice holds out hand))
```
The group of conversationalists in (16) have been sitting on the beach for a while, drinking tea and smoking. None of them have much tobacco left. At line 1

### Joe Blythe

Figure 8: While taking her tobacco out of her pocket, Karen says, *awu mi nukunudha nginarr puleyu* 'no it's from your poison cousin' (Extract 16, line 5).

Karen seems to request something, but then backs down, canceling the request.<sup>12</sup> At line 2 Alice combines a nominal-hither construction (*mi numigathungadha* 'one portion of/more tobacco over here') with an imperative verb (*nangamardakutkathungadha* 'take some out for me over here') to request tobacco from Karen's tin, meanwhile holding her hand out to receive it. At line 5 Karen refuses the request (*awu* 'no'), accounting for the refusal by claiming that it was provided by (or that it belongs to) her husband. However, rather than referring to him by name, or with a self-anchored kinterm as 'my husband' (*nangkun ngay*) (Blythe 2010b), she instead uses the alternative recognitional (Stivers 2007) *nginarr puleyu* '{your} poison cousin', implicitly anchored to her addressee, Alice. The kinterm *nginarr* – here, 'father's mother's brother's son' – connotes extreme avoidance; the implication being that the tobacco, like the kinsman, ought best be avoided. Despite this rationale being provided, Karen gets out the tobacco tin from her pocket, hinting that the provision of some tobacco is not out of

<sup>12</sup>The translation alleged for the utterance *nga mi nanyemawathawarra* is 'hey, roll us a cigarette'. Why Karen would say this is unclear, as she already has an unlit cigarette in her mouth! That said, her motives for canceling the request are perhaps clearer.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

the question (see Figure 8). Unfazed, Alice, still holding her hand out, pursues the request with *mi mamawathangadhaya* 'I'd like to roll some' (line 7).<sup>13</sup> At line 9 Karen again declines the request verbally (*thaninapartwardaya* 'leave it for him/on account of him') whilst inspecting the tobacco tin's contents (line 10), again hinting that possible compliance might be forthcoming. Despite the overt, verbally delivered refusals, Alice ultimately receives skerricks of tobacco from both Karen and co-present Lily, sufficient to roll herself a cigarette.

The dispreferred nature of the refusals are evident in the silence preceding the replies (0.5s at line 4 and 0.7s at line 9) and in the reason provided at line 5.<sup>14</sup> The hard line of the vocally delivered refusal is mitigated somewhat by the visual behavior that projects an alternative reality to that being projected verbally.

### **4.2.2 Implicit refusals: Counters, deflections and accounts.**

In the absence of an overt rejection token, with implicit refusals, rejection of the recruitment is inferable from the design of the responding move. Implicit refusal may be delivered solely as an account for not complying (as in line 6 of Extract 17). Two further varieties are counters and deflections. Both can have the effect of derailing recruitments. This is because the opportunities for compliance to be fitted sequentially, as responses to initiating actions, tend to rapidly evaporate. Extract 17 illustrates this with a counter-recruitment (cf. Kendrick, Chapter 4, §5.3).

### (17) Thuykem 2011 0824\_JB\_video\_GYHM100\_02\_1214705


<sup>13</sup>When Karen mentions *nginarr puleyu* '{your} poison cousin', Alice waggles the hand she is holding out (see Figure 8) and then continues to hold it there; thereby demonstrating that either, if the kinship relation is a genuine cause for concern, she is prepared to wear the consequences, or that Karen's excuse is fanciful and won't wash with her.

<sup>14</sup>While strictly speaking the gaps are not necessarily longer than various others which precede certain *preferred* second pair parts (cf. Gardner & Mushin 2015 for Garrwa conversation), they nevertheless reveal diminishing prospects for prompt compliance.

### Joe Blythe

```
5 (1.0)
▷ 6 gre ngay merengadha ngiku.
         ngay mere-ngadha ngi -ku
         1SG NEG –still 1SG.S.sit(1).FUT-get_going
         I can't move
```
At line 2 of (17) Greg holds out an unlit cigarette toward Mike who is seated in front of him. In the absence of a lighter, at line 1 he produces an imperatively formatted recruiting move *dadhawibuwathu* 'light the cigarette'. After 0.7s delay, Mike counters by firing back more-or-less the same recruiting move, *dadhawibu*, effectively 'light the cigarette {yourself}' (line 4). Greg refuses the counter recruitment by literally providing a "lame" excuse: *ngay merengadha ngiku* 'I can't move' (line 6); the account here serves as an implicit rejection. Greg's recruiting move remains unfulfilled.<sup>15</sup> In the next section below, we will see a further dramatic rejection delivered as a counter (at line 7 of Extract 23).

In (18) we see an implicit refusal via a deflected recruitment. Karen and Alice are both speaking to Maggie, a woman of about 90 years of age, who is hard of hearing. Just prior to this extract Maggie had been requesting chewing tobacco, but none was available (see Extract 9). Karen has just lit a cigarette, which she is holding in her hand. At line 3, Alice whispers to Karen that Maggie wants to smoke. Thus, she attempts to recruit Karen into passing Maggie her cigarette. Karen's dispreferred response is delayed initially by 0.7 seconds (line 4) and further delayed by the interjection *yawu* 'hey!' (line 5). The interjection initially draws Maggie's eye gaze toward her (Figure 9*a*), and then subsequently in the direction of her classificatory brother standing off-screen (Figure 9*b*). Karen then directs Maggie to ask the man off-screen (for permission to be granted the request).<sup>16</sup>

(18) Dingalngu 2011 0730\_JB\_video\_GYHM100\_04\_845780\_855106

1 kar nyiniwa kangkurl nyinyiyu kumban nyiniyu. nyini-wa kangkurl nyinyi=yu kumban nyini=yu ANAPH-EMPH wBSC 2SG =CL 3PL.S.6go.EXIST ANAPH=CL they're your grandsons, all of them

<sup>15</sup>After further unsuccessful attempts by Greg at recruiting someone to light it (see Extract 20), Mike eventually offers to light it. Offers, however are initiating moves rather than responsive moves.

<sup>16</sup>Karen's classificatory brother (*Kembutj*) has brought Maggie out bush, from the frail-aged hostel in Wadeye. By evoking him as a responsible person (given that he has taken responsibility for her wellbeing, at least for the day), she thereby abdicates any responsibility she might have, as provider of cigarettes, for the potentially detrimental effects smoking could have for an old woman.

7 Recruitments in Murrinhpatha and the preference organization of responses

2 (0.3) ▶ 3 ali °°purdiwinuwarda°° purdi -wi -nu –warda 3SG.S.30.FUT-swell-FUT-TEMP she wants to smoke 4 (0.7) ▷ 5 kar [yawu! (.) ] thadharrpu ngawu. (0.4) [kardu ngaynukun;] yawu tha -dharrpu ngawu kardu ngay-nukun hey! 2SG.S.19Poke.FUT-ask hey! NC:HUM 1SG -FIRR hey, you ask hey! the {brother} of mine ▷ 6 ◩ [ Figure 9a ] [ Figure 9b ] 7 (0.2) 8 ali kembutj [thadharrpu. kembutj tha -dharrpu man's\_name 2SG.S.19Poke.FUT-ask ask Kembutj ((for permission)) 9 kar [mama thadharrpu; (0.7) ngathan narna. mama tha -dharrpu ngathan mother 2SG.S.19Poke.FUT-ask brother na -rna 2SG.S.say(8).FUT-3SG.M.IO ask him mum. (0.7) ask {my} brother

Figure 9: (*a*) *yawu* 'hey!'; (*b*) *kardu ngaynukun* 'to my {brother}' (Extract 18, line 6).

The classificatory brother subsequently becomes drawn into the conversation (not shown in the extract). Maggie does not ask him for permission, and she does not receive a smoke. Her desire to to smoke remains unaddressed. Alice's recruitment initiation is derailed without the need for an overt refusal. Deflected recruitments reallocate responsibility for complying to a third party, such that the likelihood of the desired outcome arising is diminished.

### Joe Blythe

### **4.2.3 Other-initiations of repair**

As responsive moves that neither comply nor project compliance to recruitments, nor outrightly reject recruitments, other-initiations of repair (OIR) produced by the target of a recruitment are dispreferred responses. Not being of the category type projected by the recruiting turn (Raymond 2003; Heritage & Raymond 2012), other-initiations of repair results in delay of the expected category type response. This characteristic feature of dispreference can forecast imminent refusal of the recruitment (Schegloff et al. 1977: 380).

In (19) Karen, Alice, Lily and Maggie are conversing in a group as they sit on one side of a 4WD which has a trailer behind it. On the other side of the trailer, another group of women are also seated on the ground, and also being recorded on video as they converse. The car and the trailer creates a visual barrier between the groups that obscures their lines of sight. At line 1 of (19) Karen summons one of the women in the other group (Lily, apparently) to come and explain something to Alice. She does this with two interjections *yawu* 'hey!' and *kagawu* 'come here!' and with the second person singular imperative verb *thurduriyitjmani* 'try and explain it'. As she yells this summons she tries to look underneath the trailer to get a visual on her target. When this summons yields no result after 1.5s (line 2), Karen reissues the summons with another second person singular imperative verb *thurrumaniyethu* 'come here will you' (line 3). After further delay (0.6s, line 4), Lily initiates repair on the second person singular subjects of these verbs with the person-specific content question *nangkal* 'who'. At line 7 Karen specifies the previous speaker, Lily, as the target of the intended summons (*nyinyi nyinyi* 'you, you'), which is echoed by Alice at line 8. At line 10 Lily refuses the summons, invoking the video camera in accounting for the refusal.

(19) Dingalngu 2011 0730\_JB\_video\_GYHM100\_04\_341515\_350670

```
▶ 1 kar ↑YAWU kardu thurduriyitjmani kagawu!↑
          yawu kardu thurdu -riyitj –mani kagaw
          hey! NC:HUMAN 2SG.S.29.FUT-explain-try_to come_here
          HEY! try come here and explain {to her}
  2 (1.5)
▶ 3 kar thurrumaniyethu
          thurru -mani -gathu
          2SG.S.go(6).FUT-be_able-HITHER
```

```
come here will you
```

```
4 (0.6)
```

```
▷ 5 lil nangka:l;
          who
          who?
```
### 7 Recruitments in Murrinhpatha and the preference organization of responses

```
6 (0.2)
7 kar nyinyi [nyinyi.
       2SG 2SG
       you, you!
8 ali [nyinyi.
               2SG
               you!
9 (.)
10 lil ya nandji kanyinu nga ngay ngurdamyitjnganam.
       ya nandji kanyi-nu nga ngay
       HES NC:RES PROX -DAT Hey 1SG
       ngurdam -yitj=nganam
       1SG.SB.SHOVE.RR(30).NFUT-tell=1SG.SB.BE(4).NFUT
       I'm telling stories into this thing ((a video camera))
```
The delay induced by an other-initiation of repair can also have the effect that the necessity for the recruitee to comply, or account for not complying, disappears through the unrolling of interactional events (see also Dingemanse, Chapter 10, §4.2). Thus in (20) Greg continues attempting to enlist someone to light the cigarette. Turning to his right, he addresses Dom by name and instructs him with an imperatively formatted predicate (*dadhawibu*, line 1) to light the cigarette previously mentioned in (17). Dom does not have a clear view of Greg because Ray is sitting between them (see Figure 10). After two seconds delay, Dom initiates repair with the "open" interrogative *thangku* (Blythe 2015). Greg does not bother repairing the problematic recruiting move because by this stage, Mike (the target of the request in Extract 17), offers to light the cigarette by wiggling the fingers of the hand he is reaching out toward Greg (line 5). Greg then passes him the cigarette (line 6) and Mike lights it on a coal from the fire.

### (20) Thuykem 2011 0824\_JB\_video\_GYHM100\_02\_1222731\_1242143

```
▶ 1 gre Dom dadhawibu.
         Dom da -dhawibu
         ♂name 2SG.S.BASH(14).FUT-ignite_cigarette
         Dom light the cigarette
  2 (2.0)
▷ 3 dom [thang[ku.]
          what?
▷ 4 ◩ [Figure 10]
  5 mik [((wiggles fingers))
  6 gre ((passes cigarette to Mike))
  7 mik ((lights cigarette from a coal))
```
Joe Blythe

Figure 10: As Dom initiates repair (*thangku* 'what?'), his view of Greg is obscured by Ray (Extract 20, line 3).

### **4.2.4 Ignoring**

Of the 78 recruitments that were not promptly complied with, or for which possible compliance was projected as imminent, more than half were not noticeably responded to at all, and thus apparently ignored. That the lack of response should be taken as off-record implicit refusals cannot always be evidenced interactionally, as (21) demonstrates.

(21) Dingalngu 2011 0730\_JB\_video\_GYHM100\_04\_384070\_389631


### 7 Recruitments in Murrinhpatha and the preference organization of responses

In (21) Karen and Alice have been talking about some shellfish they have been eating. At line 2 Karen looks up to see Laura walking in front of her, rejoining the group. Gazing at Laura, Karen addresses her with the kinterm *nginarr* (MBDD, line 3) and requests that she make more tea. Laura continues walking slowly and then sits down where she had been previously been sitting. She does not make any tea, Karen does not pursue a response and tea is not mentioned again for some time. Although we cannot be entirely sure that Laura heard Karen's recruiting move, the recording reveals clear articulation from Karen and Laura was standing in front of her, right where her voice is being projected. There is no reason therefore to think Laura did not hear it. She appears instead to ignore the request completely.

For other examples, such as (22), we can be quite convinced that would-be recruitees refuse to acknowledge the recruiting move, by ignoring the recruiter altogether. At line 1 Dom picks up an empty billycan and peers into it. At line 2 he then targets co-present Mike (by name) and requests water from him with the nominal-hither construction (*kura pathathu kura patha*). Whether Mike actually hears Dom's request, or merely ignores him, is unclear.<sup>17</sup> Mike has been engaged in discussion with Bill, an ethnographer, about how much they will be paid for being recorded on camera, a discussion that Bill concludes at line 6, as he walks away from the scene.

(22) 20110824\_JB\_video\_GYHM100\_02\_1886635

```
1 dom ((picks up empty billycan, peers into it, replaces it))
2 Mike kura pathathu kura pa↓tha.
      Mike kura patha-gathu kura patha
      ♂name NC:WATER good –hither NC:WATER good
      Mike, some fresh water here
3 (0.4)
4 mik we'll get two hour Bill.
5 (0.2)
6 bill ok, (.) [(0.4) [puyya.
      OK puy =ya
      OK onward=CL
      alright carry on!
7 dom [kura pathath[u.
              kura patha-gathu
              NC:WATER good -hither
              water here
```
<sup>17</sup>Dom himself is unclear. His utterances at lines 2, 7 and 12 are all mumbled.

### Joe Blythe

8 [((gazes to his left)) 9 (0.3) 10 mik alright 11 (0.3) 12 dom [mi biskitkathu. mi biskit -gathu NC:VEG biscuit-hither give me a biscuit/the biscuits 13 [((touches Ray's leg twice, Figure 11)) 14 ray ((no response))

Figure 11: Dom touches Ray on the leg twice (Extract 22, line 13).

At line 7 Dom redoes the recruiting move with a repetition of the same construction *kura pathathu* 'water-hither'. He says this whilst gazing at various items located on the ground between Greg and Mike. Thus, in the absence of an explicit vocative, no particular recruitee is being targeted; and there is no uptake from the other young men. At line 13 Dom touches Ray twice on the leg (see Figure 11) and, with another nominal-hither construction (*mi biskitkathu*, line 12), requests biscuits from Ray. Ray does not respond and does not move. In order to avoid Dom's gaze, Ray turns his head slightly to his left, away from Dom who is seated slightly to Ray's right, but very much within Ray's "transactional segment" (Kendon 1990).<sup>18</sup> He is thus actively ignoring Dom. We know nothing of

<sup>18</sup>An individual's transactional segment is "the space into which he looks and speaks, into which he reaches to handle objects" (Kendon 1990: 211). It encompasses the arc projected 30° either side of the sagittal plane, as radiating out from individual's lower body (ibid: 212). When Ray tilts his head to the left, he torques his head to the left edge of his transactional segment, relegating Dom to the right periphery of his field of view. Thus, not looking at Dom requires active gaze avoidance on Ray's behalf.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

the reason for the non-fulfillment. Evidently, however, this is an utter, albeit implicit, refusal to comply with the request, and a refusal to even acknowledge the requestor's presence.

In (23) the rejection implicit in the silence that follows an ignored recruiting move is made explicit when recruitment is then pursued. The extract continues on from where (15) left off.

```
(23) Nanthak 2011 0828_JB_video_GYHM100_03_453900_460860
   1 ali [awu ku(h)rdu]nyidham(h)arrarrnukun[:;
          awu kurdu -nyi -dhamarrarr -nukun
          no 3SG.S.shove(29).FIRR-1NS.INC.DO-burn_throat-FIRR
          no! i(h)t might b(h)urn our throats!
   2 lil [ (h)a:wu;! ] [↓karraya;↓
              awu karraya
              no goodness!!
              n(h)o! good grief‼
   3 (0.7)
 ▶ 4 lil cupwangu nanyekut yawu. (.) haphapnu.
          kap -wangu na -nye -kut
          receptacle-thither 2SG.S.grab(9).FUT-1NS.INC.IO-gather
          yawu hap-hap -nu
          hey! RDP-half-DAT
          hey! put it evenly into our cups
   5 (1.3) ((Karen pours milk into billycan, Alice ignores Lily))
 ▶ 6 lil yawu tepala (0.4) kap!
          yawu tepala kap
          hey deaf receptacle
          hey deaf one! (0.4) cup!
▶▷ 7 ali >KURA PATHAWARRA NGAY YAWU!< (0.3)
          kura patha-warra ngay yawu
          NC:WATER good -ahead 1SG hey
          HEY! {BRING} ME / I {WANT} WATER FIRST
   8 (0.8)
   9 ali [PURDUNYIDHAMA]rrarr↓nu!
          purdu -nyi -dhamarrarr -nu
          3SG.S.shove(29).FUT-1NS.INC.DO-burn_throat-FUT
          it will burn your throat!
   10 kar [( )]
   11 (0.4)
   12 ali dendurr.
          dendurr
          hot
          it's hot!
```
At line 4 Lily instructs Alice (presumably, it is Alice she is gazing at) to 'put the tea half-and-half into the cups'. Alice does not return Lily's gaze, nor, while

### Joe Blythe

Karen pours milk into the billycan at line 5, does she concern herself with either tea or cups. When Lily at line 6 pursues a response with the interjection *yawu* 'hey!' and by addressing Alice as *tepala* 'deaf one', she elicits a fiery response from Alice in the form of a shouted counter-recruiting move: 'HEY! {BRING} ME / I {WANT} WATER FIRST', followed by a reason (line 9) for not serving out the tea prematurely ('it will burn your throat!'). The bald counter-recruiting move (which, incidentally, is also ignored) is neither delayed nor mitigated. In overlap with Alice, Karen at line 8 also shouts something that cannot be discerned. Karen, who at line 5 had been pouring milk into the billycan, like Alice, displays the irritation she had previously suppressed.

That such a substantial number of recruiting moves elicited no response, and are seemingly ignored, is alarming. Although this collection of sequences clearly deserves expanded investigation, it is already evident that "no-response" is to be considered a valid response. In some cases the initiating move is clearly problematic or perhaps difficult to comply with, but in other cases, we can evidently infer that the would-be recruitee considers the substance of the recruitment to not even merit an overt refusal.

# **5 Acknowledgment in third position**

Of the languages surveyed in this cross-linguistic project, only Italian and English showed at least some degree of acknowledgment of the recruitments' fulfillment; most languages had only a few if any (Floyd et al. 2018). There were only three in the Murrinhpatha collection, one being a simple nod, the others being seemingly ad-hoc acknowledgments which I will not elaborate on here.

# **6 Social asymmetries**

Most Australian Aboriginal societies are generally held to be egalitarian and nonhierarchical (e.g. Flanagan 1989; Boehm 1993; Peterson 1993). Social asymmetries are generally not reflected within grammatical contrasts, nor in the choice of lexical items used for address. In this dataset there are only a few occasions that we notice social asymmetry being born out within the interaction. One asymmetry that is brought into play is age, and the seniority that comes with greater experience. Elders are held in great esteem and may be referred to as *pule* 'respected'/'boss'. Age related seniority may lie behind Ray's refusal in (22) to even acknowledge his pesky younger brother's existence. Ray is the eldest of a group

### 7 Recruitments in Murrinhpatha and the preference organization of responses

of brothers and cousins who name themselves after a particular heavy metal band (Mansfield 2013; 2014a). Ray is said to be 'boss' for that group.

The by now familiar episode on the beach in which the four women resist fetching water for the hot tea is ultimately resolved when the three eldest women assert their seniority over Rita. In (24) particularly, Karen launches a sarcastic, melodramatic tirade at Rita.

(24) Nanthak 2011 0828\_JB\_video\_GYHM100\_03\_509480 (simplified transcript)


Karen contrasts Rita, as young (*kardu mardinybuy* 'a young girl', line 1) and radiant (literally, a 'rainbow', *kemnyekekngime*, line 3) with the other white-

### Joe Blythe

haired women (*damnyiwebawaywardangime*, line 8) with one foot in the grave (*ku wakay warda manda warda* 'for whom death is near', line 11). This fanciful comparison breaks the deadlock because Rita gets up (line 9) in order to take the billycan to get some cool water (see Extract 12). She has drawn the short straw here as she herself is a grandmother and is Karen's junior by merely two years!

# **7 Discussion**

In most conversation-analytic research on preference structure, dispreferred second pair parts are analyzed in terms of their dispreference features as delayed, hedged, accounted for, etc. An implicit criterion for this approach is detection of the dispreferred second pair part for analysis of these features. An empirical question then is: when an expected response is absent, can its notable absence be legitimately read as a dispreferred response?

When conversation analysis was in its infancy, telephone recording technologies were adopted more widely by conversation analysts than was video. Most of the seminal works on preference organization were conducted on phone call data. Because participants speaking on the phone are not co-located in space, when requests are made, seldom can the substance of the request be fulfilled immediately. Thus phone call requests are normally higher contingency, future actions, for which arrangements need to be made in advance. The substance of the request may well be the actual reason for the call (Sacks 1992; Schegloff & Sacks 1973; Couper-Kuhlen 2001). Usually, the possible imposition on the requestee is foregrounded, becoming the substance of deferential behavior and politeness considerations. Preliminaries need to be dealt with through backgrounding and pre-sequences (Schegloff 1980; 2002; 2007). However, like each dataset in our comparative project, the Murrinhpatha corpus consists entirely of casual faceto-face conversation amongst friends and family. All of the recruitments call for similarly immediate action to be performed within the general vicinity. A likely outcome of this is that, at least in the Murrinhpatha corpus, there are no prerecruitment sequences (but see Floyd, Chapter 3, §3.3.3; Rossi, Chapter 5, §3.3.3).

This chapter has presented the Murrinhpatha system of language use pertaining to recruitments. As per the other chapters, it has surveyed the range of possible recruiting formats followed by the array of possible actions and formats in the responding move.

Here I concentrate the discussion on response types and their relative proportions. The payoff in considering response options paradigmatically, as a set of alternatives, is immediately evident (see also Thompson et al. 2015). From among

### 7 Recruitments in Murrinhpatha and the preference organization of responses

the range of possible responses, "no-response" (ignoring) substantially emerges as a legitimate option existing intermediately between overt compliance and overt rejection (see Figure 12).<sup>19</sup> Extracts 21–23 show that, at least for Murrinhpatha speakers, silence plus a lack of physical action following a recruiting move can be understood not as a harbinger of imminent refusal, but as actual implicit refusal. There is reason, however, to think that this state of affairs is not culturally specific to Murrinhpatha speakers.

Figure 12: Relative proportions of response types. Projected possible compliance includes visible behavior that hints at fulfilling the recruitment, as well as explicit commitments to future compliance. Alongside prompt compliance these are the preferred responses. Implicit rejections included counters and deflections, as well as rejections delivered as accounts for non-compliance (see §4.2.2). Pragmatically, ignoring is a "morphologically unrealized" subtype of implicit rejection.

Discussing an example reproduced below as (23), Levinson (1983: 320-321) demonstrates how a two-second silence following a pre-request is taken to be a negative response to the pre-request. The pre-request deals with the call-taker's availability, a prerequisite condition for arranging a future meeting.<sup>20</sup> The caller's reading of the silence as conveying unavailability ultimately proved to be un-

<sup>19</sup>The denominator has been reduced here from 145 to 139 due to the untypable responses: those where the vocal component of the move is insufficiently audible to be adequately categorized, and/or when the respondent is obscured from view or off-screen.

<sup>20</sup>Levinson suggests that the two-second silence at line 3, following the caller's pre-request, is

### Joe Blythe

founded (presumably, the call-taker was actually checking his/her schedule during the silence). Irrespective of the caller drawing the wrong conclusion, the extract illustrates how silence following a specifically allocated first-pair part mobilizes the inferential machinery such that a sub-optimal outcome is imagined.

```
(25) (Levinson 1983: 320-21)
```

```
1 cal So I was wondering would you be in your office on
2 Monday (.) by any chance?
3 (2.0)
4 cal Probably not
5 tak Hmm yes=
6 cal =You would?
7 tak Ya
8 cal So if we came by could you give us ten minutes of your time?
```
In the absence of pre-sequences, a no-response following a conditionally relevant first-pair part is hearable *not* as projecting an impending block of a yet-toemerge base-sequence, but rather as non-fulfillment of, or non-compliance with, the first-pair part of the base sequence. Ignoring is the "zero-morph" of responses to recruitment. No-response is a meaningful declining response that stands in paradigmatic opposition to fulfillment, as one "format" within a range of dispreferred alternative formats that explicitly reject the substance of the recruitment (overt refusals), implicitly reject it (ignoring, counters, deflections, accounts as rejections), or defer the expected base second pair part (OIR).<sup>21</sup> The utility of no-response lies in conveying rejection without leaving any on-record token of rejection.

While the rate of non-compliance in Murrinhpatha is high, the rate of noresponse is strikingly high. However, we should be careful to interpret these

allocated by the turn-taking system to the call-taker, as the next-selected speaker. As such, the call-taker owns the silence. The caller hears the silence as a projecting a dispreferred negative response to the pre-request, which would effectively block the caller's projected request. Preempting the blocking response, the caller answers his/her own question, wrongly as it seems. Having then established the call-taker's availability, the request eventuates at line 8. "Note here the remarkable power of the turn-taking system to assign the absence of any verbal activity to some particular participant as his turn: such a mechanism can then quite literally make something out of nothing, assigning to a silence or pause, itself devoid of interesting properties, the property of being A's, or B's, or neither A's nor B's" (Levinson 1983: 321).

<sup>21</sup>In the protracted episode with the hot billycan on the beach, all participants but especially Rita use the full range of these refusal formats to doggedly resist recruitment after recruitment. In this battle of wits, twenty-seven (!) recruiting moves were produced before possible imminent compliance was projected.

### 7 Recruitments in Murrinhpatha and the preference organization of responses

high rates as reflecting a cultural difference, as they might at least partly influenced by the nature of the interactions and people represented in the sample used for this study. Many cases come from interactions among old and relatively infirm participants who are recruited to do things such as lifting heavy water bottles, which requires a high level of physical exertion. Other cases involve demands that are silly or unreasonable, such as Karen's instructions to Maggie in (18) that she ask her brother for permission to smoke (Blythe 2017). Nevertheless, the high no-response rate still raises interesting questions, especially for politeness theorists and intercultural communication researchers. If making requests is inherently face-threatening for the requestor, why would Murrinhpatha speaking recruiters risk threats to their positive face when the likelihood of refusal is so substantial? Do cultural expectations based on demand sharing (Peterson 1993) diminish potential threats to the recruiter's positive face such that the chance of refusal merits the risk? Might ignoring recruitments actually be the politest method for declining them? Is ignoring a mechanism for coping with *humbug*? 22 Is the reason many Europeans working in Aboriginal communities feel excessively overburdened by *humbug* (Gerrard 1989) because they do not imagine ignoring to be an acceptable option for refusing requests? I will not attempt to answer any of these questions here. However, the fact that they emerge from these results underscores the immense value in taking an emic perspective on social interaction: taking video recordings of informal conversation conducted within a single social group as baseline interactional data; allowing researchers to ground their understanding of cultural expectations upon members' normative responses to recurrent social actions. Having then compared practices from other social groups, using analogous datasets (as per the approach of pragmatic typology), intercultural communication researchers can draw on these data to better understand communication between participants from different cultural and linguistic backgrounds.

<sup>22</sup>*Humbug* is a colloquial Aboriginal English term for the annoying pressure placed on an individual with the intention of eliciting material goods or future deeds. When a person *humbugs* someone, they make persistent demands and requests for such things as food, money, tobacco, and lifts in vehicles; perhaps even performed with a degree of with menace or intimidation (Gerrard 1989; Blythe 2001: 40–42).

### Joe Blythe

# **Abbreviations**


# **Acknowledgments**

The corpus was prepared with the consultative assistance of Kinngirri Carmelita Mardigan, Mawurt Ernest Perdjert, Lucy Tcherna, Phyllis Bunduck, Gertrude Nemarlak, Desmond Pupuli and Jeremiah Tunmuck. Thanks also to Mark Crocombe at the Kanamkek Yile Ngala Museum and Languages Centre in Wadeye for logistical support. I am grateful to John Mansfield for making the conversations he recorded available for this collection. The paper has benefited from the comments and advice of Giovanni Rossi, Simeon Floyd, Scott Barnes and two anonymous reviewers. I'm grateful for the lively discussions with fellow collaborators in the HSSLU project – Julija Baranova, Paul Drew, Mark Dingmanse, Tyko Dirksmeyer, Nick Enfield, Simeon Floyd, Rósa Gísladóttir, Kobin Kendrick, Steve Levinson, Ely Manrique and Giovanni Rossi. This research was funded by the European Research Council (StG 240853) and the Australian Research council (DP110100961, DE130100399).

7 Recruitments in Murrinhpatha and the preference organization of responses

# **References**


### Joe Blythe


### 7 Recruitments in Murrinhpatha and the preference organization of responses


### Joe Blythe

*in Aboriginal Australia* (Oceanic Linguistic Monographs 24), 86–106. Sydney: University of Sydney.


### 7 Recruitments in Murrinhpatha and the preference organization of responses

Bruner (eds.), *Interaction in human development*, 139–153. New York: Lawrence Erlbaum.


# **Chapter 8**

# **Recruiting assistance and collaboration in Polish**

# Jörg Zinken

Department of Pragmatics, Leibniz-Institute for the German Language

This chapter describes the resources that speakers of Polish use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Polish, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Polish with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

This chapter provides an overview of many of the practices for recruiting another person's assistance or collaboration in Polish. The data for this overview come from a corpus of video recordings of informal everyday interactions in the homes of families living in urban areas of Poland. As this chapter will show, recruitment practices in Polish follow many of the regularities that we have observed for other languages in the larger project reported in the present volume. Some distinctive aspects of Polish recruitments, such as the diverse imperative, impersonal, and infinitive formats of recruiting moves, are also discussed.

Jörg Zinken. Recruiting assistance and collaboration in Polish. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 281–324. Berlin: Language Science Press. DOI:10.5281/zenodo.4018384

### Jörg Zinken

# **1.1 The Polish Language**

Polish is an Indo-European language that belongs to the West-Slavic branch of the Slavic language family. Polish is spoken by about 40 million people worldwide, of whom about 37 million live in the Republic of Poland in Central Europe.

Polish has a long tradition of grammatical description (comprehensive grammars are Bąk 2010 and Strutyński 2006). Although it is characterized by relatively free word order, its basic word order is SVO. There is a rich tradition of pragmatic work in Polish linguistics, but work on the basis of recorded interaction has been virtually absent until recently (though see Labocha 1985; 1986). Grammatical features relevant to recruitment practices include verbal aspect, the absence of interrogative syntax, a relatively elaborate imperative paradigm, and impersonal modal constructions with the verbs *trzeba* 'it is necessary to' and *można* 'it is possible to'.

# **1.2 Data collection and corpus**

The Polish corpus of video recordings was built outside the comparative project this volume reports on. Most recordings were made in 2009 as part of a comparative project on *Sharing responsibilities in English and Polish families*, funded by the British Economic and Social Research Council (ESRC). For that project, matched corpora of video recordings of everyday life in British and Polish families were collected. Participants were asked to record everyday activities, such as mealtimes, cooking, or playing with their children. The Polish corpus from that project includes 10 hours of recordings made by six families. Other recordings have been made during field visits since then. These further data amount to 3.5 hours of recordings made by three families. The restriction to family interaction distinguishes the Polish data from other languages examined in this volume, which include recordings of informal interaction beyond family contexts. The reader might want to keep this caveat in mind when comparing the results across languages.

The recordings were made in the capital city Warsaw and in Lublin, a university city in the southeast of Poland. This means that all recordings come from Eastern regions in Poland. The data considered for comparison consisted of coded samples from the recordings, with the goal of coding at least 200 recruitments. Most of the families had young children, and many recruitment sequences found in the data included a child, either as "recruiter" or as "recruitee". To maximize comparability of the data across languages, only recruitments in which both participants are adults were considered for the study reported on here. Six hours and

8 Recruiting assistance and collaboration in Polish

thirty minutes of recordings were sampled to identify 215 recruitment sequences.

Transcripts may include up to three tiers for each line. The first tier represents the original talk and/or other conduct following the conventions of conversation analysis (Jefferson 2004); the second tier gives a word-by-word or morpheme-bymorpheme gloss of the talk following the Leipzig glossing rules and abbreviations (Comrie et al. 2020);<sup>1</sup> the third tier gives a more idiomatic English translation.

# **2 Basics of recruitment sequences**

As defined in Chapter 1, §4, a recruitment is a basic cooperative phenomenon in social interaction consisting of a sequence of two moves with the following characteristics:


A recruitment sequence can have a minimal shape, consisting only of two moves, or it can have a more complex shape. I begin with examples illustrating this difference. In the transcripts, ▶ and ▷ designate Move A and Move B, respectively.

# **2.1 Minimal recruitment sequence**

Extract 1 provides an example of a minimal recruitment sequence. The participants are seated at the table for supper. At lines 1–2, Ilona asks Jacek to pass her the salad bowl (Move A); at line 3, Jacek passes her the salad bowl (Move B).

```
(1) PP2-1_2224980
```
here you are

```
▶ 1 ilo wiesz co podaj mi kochanie jeszcze
          know.2SG what pass.IMP me dear still
          you know what, pass me some more
  2 sałatki
          salad.GEN
          salad, dear
▷ 3 jac bardzo proszę ((passes salad bowl))
          very plead.1SG
```
<sup>1</sup> In addition to the standard abbreviations, I also use hrt for hortative and ptc for particle.

### Jörg Zinken

# **2.2 Non-minimal recruitment sequence**

Recruitment sequences are non-minimal when the recruiting move is done again (e.g. to clarify it, to make it more forceful, or because the recruiting speaker is not certain whether it has been heard) before it is complied with or rejected. In (2), Piotr is sitting at the dinner table with a baby on his lap. He tells Aga at line 1 that cheese has dropped to the floor (Move A). Aga arrives at the table and puts her coffee cup down. This might, but need not be, a move preliminary to doing the target action of picking up the cheese (line 2 is not marked with ▷ to allow for this uncertainty). Piotr then redoes the recruiting move in truncated form (line 3), and Aga picks up the cheese from the floor immediately thereafter.

(2) PP5-4\_0154810a

```
▶ 1 pio tutaj ser u- (0.4) tu ser upadł jeszcze
         here cheese dr- here cheese drop.down.3SG.PST still
         here cheese has- here cheese has dropped down also
  2 aga ((arrives at table, places coffee cup on the table))
▶ 3 pio ser upadł
         cheese drop.down.3SG.PST
         cheese has dropped down
▷ 4 aga ((picks up cheese))
```
# **2.3 Subtypes of recruitment sequence**

In the comparative project, we distinguish between four types of recruitments: i) sequences in which B provides a service, ii) sequences in which B gives an object to A, iii) sequences in which B alters the trajectory of his or her current behavior, and iv) sequences in which B does something to address A's current or anticipatable trouble.

We have already seen examples for two of these: (2), where A recruits B to pick up something that has dropped to the floor, was an example of a recruitment the point of which was that B provide a service; (1), where A recruits B to pass the salad bowl, was an example of a recruitment the point of which was that B give an object to A. Passing an object can be a particular kind of service, but we consider object requests separately, because such requests are numerous and they constitute a distinct domain (see also Zinken 2015).

Extract 3 is an example of a recruitment the point of which is to alter some current conduct by B. Ilona is putting sugar into Jacek's tea. At line 4, Jacek's recruiting turn *już=już* 'already already' (or, more idiomatically, 'enough enough')

### 8 Recruiting assistance and collaboration in Polish

gets Ilona to stop sweetening the tea further (cf. Stivers 2004 on multiple sayings as a practice for indicating that some course of action should be halted).

```
(3) PP2-5_949800
  1 ilo proszę:: ((spoons sugar into A's tea))
         plead.1SG
         here you are
  2 jac dzię[kuję bardzo
         thank.1SG very
         thank you very much
  3 ilo [słodzę:: [ci mężu
              sweeten.1SG you.DAT husband.VOC
             I sweeten it for you, my husband
▶ 4 jac [już= już
                         already already
                         enough enough
```

```
▷ 5 ilo ((stops putting sugar into tea))
```
Finally, (4) is a case in which Move B addresses some current or anticipatable problem of A's. Piotr is trying to cut pizza, but he is also holding a baby on his lap. The baby has started to pull the table mat with the pizza plate on it towards himself, and Piotr is in the difficult position of having to juggle trying to cut pizza, holding the baby, and controlling the baby's hands, all at the same time. Piotr's trouble is both visible and audible (*kurczę*, loosely translatable as 'damn', line 3). The two children, Łukasz and Przemek, laugh at Piotr's predicament (lines 4 and 5), but his wife Aga announces help and shortly thereafter comes to the table and takes the baby from Piotr.

```
(4) PP5-4_0134460
```

```
1 pio ((cuts pizza on his plate))
```

```
▶ 2 bab ((pulls the table mat))
```

```
▶ 3 pio ku::rcze no,
         EXPL PTC
         damn no
```

```
4 łuk Hh::
```

```
5 prz A: hhahaha .H
```

```
▷ 6 aga już go ci biorę stamtąd
         already him.ACC you.DAT take.IPFV.1SG from.there
         already I'm taking him for you from there
```

```
▷ 7 ((comes to the table and takes the baby))
```
### Jörg Zinken

Cases of assistance with current or anticipatable trouble can also often be analyzed as eliciting a service of some sort. What separates them as a category, however, is that A might not have designed their conduct to recruit assistance. Nonetheless, B's practical action is occasioned by some conduct in what then becomes Move A.

# **3 Formats in Move A: The recruiting move**

# **3.1 Fully nonverbal recruiting moves**

Sometimes recruiting moves are fully nonverbal. For example, a person can simply point to an object that they want to be given, or they can reach out to receive an object, or – as in the following case – a person can hold out an object and thereby recruit another person to take it and do something with it. In (5), Marta and Karol are searching for a particular medication in the fridge. Marta takes a package out of the fridge, inspects it, and then holds it out for Karol to take it from her. Karol then takes the package from Marta (Figure 1).

```
(5) (PP6-3_1920720)
```

```
▶ 1 mat ((holds out package))
```

```
▷ 2 kar ((takes package))
```
Figure 1: Karol takes the package from Marta (Extract 5, line 2).

### 8 Recruiting assistance and collaboration in Polish

Fully nonverbal recruiting moves can be successful when the context provides a rich scaffold that secures the other person's attention and makes the point of the recruitment transparent (Rossi 2014 and Chapter 5, §3.1; see also Kendrick, Chapter 4, §4.1.3; Baranova, Chapter 9, §3.1; Dingemanse, Chapter 10, §3.4). The recruitment in (5) occurs in a context in which Marta and Karol are already engaged in the activity of inspecting various packages of medicine they have in the fridge. Fully nonverbal practices make up 6.5% (*n*=14) of all recruiting moves.

# **3.2 Nonverbal behavior in composite recruiting moves**

Moves initiating recruitment in face-to-face interaction often involve a combination of verbal and nonverbal conduct. For example, when A asks B to pass the butter, A might also gaze towards B, a practice that can serve to address the relevant person (Lerner 2003). Here, however, we coded only conduct that aids the recipient in identifying the target object and/or action. Table 1 provides an overview of the types of nonverbal behavior found in composite recruiting moves in the Polish data.


Table 1: Types of nonverbal behavior in composite recruiting moves (*n*=77).

Aga's turn in (6) includes two recruiting components, both of which are accompanied by relevant nonverbal conduct. In line 2, when Aga formulates a request to be given the baby, she also stretches out her arm to receive him (Figure 2*a*). This is an example of a "reach to receive" gesture, although it also has an iconic element, because Aga would not actually grasp the baby with her outstretched hand (when Piotr hands Aga the baby, she takes him with both hands). After the completion of this recruiting move, Aga immediately launches the next element in her turn: another request for Piotr to sit down (line 4). As she formulates this request, Aga also slightly pushes back the chair (Figure 2*b*). Such manipulation of objects involved in the target event was coded as an "other" form of nonverbal

Jörg Zinken

conduct.


▷ 7 pio ((walks to Aga, hands over baby))

(a) Aga stretches out arm (line 3). (b) Aga pushes back chair (line 5).

Figure 2: Frames from Extract 6.

By pushing back the chair, Aga indicates where Piotr should sit down, but also makes sitting down more straightforward for Piotr. Like pointing gestures, "other" forms of nonverbal conduct often indicate a relevant object. However, they also commonly make the object more useable for the intended purpose, and thereby increase the transparency of the recruiting move. In another example, (15) below, Bogusia places a salad bowl on a small counter between the kitchen and the living room, saying *jeszcze proszę sałatę* 'also please the salad'. Placing the salad bowl there makes it more easily accessible to the others, and, in conjunction with the verbal turn, constitutes a transparent request for somebody to bring the salad to the living room table in preparation for the meal.

8 Recruiting assistance and collaboration in Polish

# **3.3 Verbal elements: construction types and subtypes**

Verbal elements in recruiting moves were classified with reference to three crosslinguistically common sentence types: declarative, interrogative, and imperative. As there is no interrogative syntax in Polish, questions are recognizable through intonation, the use of question particles, and what one speaker knows that the other knows (for a conversation-analytic discussion, see Weidner 2013a). Further construction types are recruiting turns without a predicate and others with a verb in the infinitive. Imperative recruiting turns are the most common. At the same time, we find other formats in more than half of recruiting moves with a verbal component (see Table 2).<sup>2</sup>

Table 2: Construction type of recruiting moves including spoken elements (*n*=199).


### **3.3.1 Imperatives**

Polish has a relatively complex imperative paradigm. Morphological imperatives exist for the second person singular and plural, and the first person plural. Periphrastic hortative constructions exist for the third persons and the first person singular, as well as for formal (V-form) second person reference (on imperatives and hortatives, see Van der Auwera et al. 2013).

In the examined corpus, nearly all imperatives are in the second person. The only exception is found in (7) where Henio uses the third person hortative construction *niech* + verb (roughly, 'may it x').<sup>3</sup> Henio's move at lines 4–5 recruits Bogusia to leave the camera on by saying *niech to jeszcze ten* 'may this still that one', where *ten* 'that one' is a demonstrative pronoun that in spoken Polish often functions as a dummy term. Here, it stands in for an otherwise expectable third

<sup>2</sup>The missing 0.5% is due to one case in which the verbal component of Move A is inaudible.

<sup>3</sup>Another instance of this format can be found in (30), where the recruitment is initiated but not completed.

### Jörg Zinken

person singular verb, such as *filmuje* or *nagrywa* 'it records'. Bogusia complies at lines 8–9, reformulating the hortative utterance, this time with the "missing" third person predicate.

(7) PP3-1\_2348380a 1 bog ((moving to turn off camera)) 2 hen ((getting up from the table)) 3 dobrze to jeszcze jeszcze to póki Magda good.ADV then still still then while Magda alright then while Magda is still ▶ 4 je to niech to jeszcze ten eat.3SG then HRT this still that.one eating then may this still that one 5 (0.4) ▷ 6 bog tak¿> osta- [Magda ostatnia od[chodzi od yes las- Magda last leave.3SG from yes? Las- Magda leaves the table last? 7 hen [(no) [tak PTC yes *no* yes ▷ 8 bog stołu tak,= niech będzie table.GEN yes HRT be.3SG right, may it be ▷ 9 sfilmowane no filmed.PASS PTC filmed *no* 10 hen niech będzie sfilmowane HRT be.3SG filmed.PASS

may it be filmed

In the remainder of this section, I discuss only second person morphological imperatives. Among these, I distinguish three turn formats: imperatives with perfective aspect marking, imperatives with imperfective marking, and the (perfective) double imperative, *weź zrób x* ('take do x'). Perfective imperatives are by far the most common in the corpus (*n*=68), followed by imperfective imperatives (*n*=13),<sup>4</sup> and by the double imperative (*n*=12).

**3.3.1.1 Perfective imperatives.** Perfective imperatives are the most common subtype of imperative recruiting format in the data. Work on the selection of imperatives for requesting action demonstrates that such recruiting moves convey an expectation of compliance (Wootton 1997; Goodwin 2006; Craven & Potter

<sup>4</sup>One of these is the monoaspectual *siedź* 'be/remain sitting'.

### 8 Recruiting assistance and collaboration in Polish

2010; Kent 2011; 2012; Rossi 2012). This expectation is, in informal interaction among friends and family, typically grounded in the fact that the requested action is integral to a wider activity to which the recipient is already committed (Wootton 1997; Rossi 2012). The two examples of recruiting moves with perfective imperatives that we have seen so far, (1) and (6), illustrate this. In (1), Jacek is available for jobs such as passing the salad bowl on the basis of his participation in the mealtime event, and the imperative orients to this availability. In (6), Piotr is already engaged in finding arrangements for feeding his son (see his question in line 1) and the imperative recruiting move is designed as a step in this wider activity (see also Zinken & Deppermann 2017). Recruiting moves formatted as perfective imperatives will be discussed repeatedly in later sections (Extracts 8, 19, 22, 26, 32, 34, 35, 37) and I therefore do not provide further examples here.

**3.3.1.2 Imperfective imperatives.** Imperfective imperatives have repeatedly concerned linguists working on Slavic languages (see Forsyth 1970; Lehmann 1989; Benacchio 2010).<sup>5</sup> From the perspective of sequential analysis, it is striking that imperfective imperatives are used in positions where the relevant action has already been brought into play by the other person, or is the direct consequence of what has occupied the interaction in the just prior turns. In other words, although we might think of requests and directives as good examples of sequentially first actions (Sorjonen et al. 2017), imperfective imperative turns in Polish are never textbook examples of first pair-parts (Schegloff 2007). In fact, imperfective imperatives are often used in second position to accomplish actions such as giving a go-ahead (Lehmann 1989; Zinken & Deppermann 2017). In the domain of recruitments, imperfective imperatives treat an action as already "authored" by the other (Zinken 2016: chap. 8).

In (8), Ania is urging her mom Ela to start dinner, because she has to leave in ten minutes (lines 1–2). The turn-initial *no* in Ela's agreeing response (line 3) conveys that it is obvious to Ela that the meal is to start now (on turn-initial *no*, see Weidner 2013b), while her *już* 'already' conveys that in fact everything is on track. She then extends her turn to address a directive to Ania, namely to serve the food for the younger children (line 5). This recruiting turn is designed with an imperfective imperative. It begins, again, with the particle *no*, which here expresses Ela's stance that serving the food is the obvious consequence of Ania's wish to speed things up. In response, Ania begins serving the food.

<sup>5</sup>This literature is mostly concerned with Russian data but is relevant also to Polish.

### Jörg Zinken

```
(8) PP1-1_0145540
```

```
1 ani do:bra: mo- czy możemy zjeść¿ =bo ja muszę
          good.ADV ca- Q can.1PL eat.INF because I must.1SG
          okay, can we eat, because I'll have to
  2 za dziesięć minut wyjść.
          after ten minutes go.out.INF
          leave in ten minutes
  3 ela no już:=
          PTC already
          no already
  4 tad =◦jest już◦
           is already
           it's already done
▶ 5 ela ↑no to już nakładaj im.
           PTC then already serve.IMP.IPFV them
           no then serve them (their food) already
  6 (0.5)
▶ 7 ela mięsko weź na stół¿ (.) surówka jedna druga¿
          meat take.IMP.PFV on table slaw one other
          put the meat on the table, the one salad and the other
```
We can think of recruiting moves as having a deontic side (telling the other that they should do something) and an information side (telling them what to do). Imperfective imperatives add little to the information side of a recruiting move. Insofar as the recruitment concerns a new action at all, that action, as in line 5 of (8), is framed as a direct consequence of what has come before. Imperfective imperatives mainly deal with the deontic side of the prospective action, that is, they give the go-ahead to, insist on, or prohibit an action that already concerns the other or is inferably relevant (see also Extract 37) (Zinken 2016: chap. 8). Note that Ela extends her turn with another directive detailing what exactly Ania is supposed to serve the younger children (line 7), reformulating the previous recruiting component *nakładaj* 'serve them (the food)' (line 5) in a more informative manner – this reformulation is done with a perfective imperative.

**3.3.1.3 Double imperative.** A double imperative construction that is recurrent in spoken Polish takes the form of the perfective verb 'take' (*wziąć*, imperative: *weź*) plus the relevant action verb, also with perfective aspect. In (9), Jacek is talking to his children about observations the children have made at the local swimming pool. The turn in line 1 belongs to that conversation. His wife, Ilona, is in the process of clearing the table, and she is scraping the last bits of a vegetable salad onto Jacek's plate. Some of the salad is sticking to the spoon and Ilona

<sup>▷</sup> 8 ani ((puts salads on the table))

### 8 Recruiting assistance and collaboration in Polish

recruits Jacek to scrape it off (lines 3–5) while holding out the spoon for him to take.

```
(9) PP2-5_2002280
  1 jac trudno powiedzieć
        difficult say.INF
        difficult to say
  2 dlaczego akurat [( )
        why exactly
        why exactly
▶ 3 ilo [weź kochanie
                       take.IMP dear
                      take dear
▶ 4 [((holds out spoon for Jacek to take))
▶ 5 zgarnij z łyżki
        collect.IMP from spoon.GEN
        gather (it) from the spoon
```

```
▷ 6 jac ((takes salad spoon, scrapes salad onto his plate))
```
Ilona's recruiting move allows for a serial interpretation ('take the spoon and scrape off the salad'). However, *weź* is grammaticalized to the extent that it is unproblematically used to solicit actions that do not involve taking anything. In other words, it functions as a particle rather than as the imperative of the verb 'take' (see Zinken 2016: chap. 7). Simple perfective imperatives are often selected to recruit actions as part of an established joint project, as outlined in §3.3.1.1. However, the 'take-V2' format is selected in situations in which B's commitment to the relevant project is not evident, although it often is expectable at a more general level. Here, Jacek is not involved in clearing the table when Ilona makes her request, but he is one of the adults responsible for organizing the mealtime event at a more general level (see also Extracts 17, 30, 31) (Zinken 2016: chap. 7; Zinken & Deppermann 2017).

### **3.3.2 Declaratives**

Declarative recruiting turns make up nearly a quarter of all the attempts that contain a verbal element (*n*=46). These can be further divided into subtypes. One group are non-modal descriptions in the third person (*n*=12) as in (2) (*ser upadł* 'cheese has dropped down'). Sometimes, third person descriptions can also recruit B by telling another, third person what B will do. In the present corpus, these are cases where one parent says to a child what the other parent will do, thereby recruiting the adult for that action (such cases are not part of the group

### Jörg Zinken

of non-modal descriptions, which are restricted to objects requiring action). In (10), for example, Ilona first proposes to the child that 'we' can put a special ointment on a scratch that the child has. Both parents seek and receive approval from the child for this course of action. At that point, the question remains as to which of the parents will go to get the ointment. It would seem that Ilona is in a better position to do so, as Jacek has the child on his lap and is feeding him. However, at line 1, Ilona formulates the target turn, addressed to the child: *tatuś posmaruje takim kremem* 'dad will put on this ointment' and in so doing mobilizes Jacek to put the child on an adjacent chair and leave to get the cream.

(10) PP2-1\_3410860

```
▶ 1 ilo tatuś po- posmaruje takim kremem
         dad smear.PFV.3SG such.INS cream.INS
         daddy wi- will apply this ointment
```

```
▷ 2 jac ((puts child on adjacent chair, leaves))
```
Another group of recruiting turns in declarative format involve impersonal constructions with a verb expressing deontic modality (*n*=9) (see also Floyd, Chapter 3, §3.3.4; Rossi, Chapter 5, §3.3.4; Baranova, Chapter 9, §3.3.3). In Polish, turns with the impersonal modal verb *trzeba* 'it is necessary to / one has to' are a practice for recruiting another person's collaboration (Zinken & Ogiermann 2011). In (11), the family have been crafting together; some glue remains on a piece of paper in the center of the table. When Marta tells her daughter not to play with the glue, this becomes an occasion for Karol to formulate what 'is necessary' to do, namely to throw the glue away (line 5). In overlap with Karol's turn, Marta begins extending her arm and then picks up the paper sheet with the glue on it and throws it away (see Zinken 2016: chap. 6).

```
(11) PP6-1_4228840
```

```
1 mar to już tym klejem się Gabrysiu
         this already this.INS glue.INS REFL Gabrysia
         don't play with this glue
  2 nie baw wiesz:¿
         not play.IPFV.IMP know.2SG
         already Gabrysia
  3 (.)
  4 mar [(on już troszkę) ]
           it already a.bit
           it has already somewhat
▶ 5 kar [to już trzeba wyrzucić]
          this already necessary throw.away.INF
          it is necessary to throw this away already
```
8 Recruiting assistance and collaboration in Polish

6 mar (.) [zasechł dry.3SG.PST dried out ▷ 7 [((picks up paper sheet with glue))

The modal verb *móc* 'can' is sometimes used with person marking in turns that recruit another person. In (12), Olek addresses his daughter Kasia with a proposal to give her toddler son (Olek's grandson) something to eat. He uses an infinitive recruiting form (see §3.3.4 below). Kasia responds by asking her son whether he would like to eat something (line 2) but then turns back to her father and recruits him to mount a contraption designed to hold small children, a kind of cloth child chair, on an ordinary chair (line 3). This recruiting turn is in the form of a declarative with second person singular marking.<sup>6</sup>

### (12) MiBrApr2012\_0459322


Recruiting moves in this format build on the other's displayed or assumed readiness to contribute to the relevant matter (see also Extract 27), in this case on Olek's suggestion or proposal that the child should or could eat something (line 1).

Turns with a performative verb in the first person are also used to recruit another person's action (*n*=7). A turn format that is specialized for recruiting another person to provide an object is built with the verb *prosić/poprosić* 'plead, ask' (*poprosić* is the perfective form) in the first person plus the item as direct object

<sup>6</sup>Note that Kasia extends her recruiting turn at line 6 with a predicate-less unit (see §3.3.3 below).

### Jörg Zinken

in the accusative (see Ogiermann 2015; for a usage of this form outside object requests, see Weidner 2015). In (13), Kasia asks Dorota to pass the horseradish. Kasia's talk in line 1 closes a prior, unrelated interaction.

```
(13) MiBrApr2012_0643192
```

```
1 kas dobrze na razie Józio nie chce siedzieć
          good.ADV on time NAME not want.3SG sit.INF
          okay, for now Józio doesn't want to sit
▶ 2 po- poproszę m- chrzan¿
          pl- plead.PFV.1SG m- horseradish.ACC
          I ask (for) m- horseradish
▷ 3 dor ((passes horseradish))
```
Announcements in the first person (singular or plural, *n*=4) can also recruit another person's collaboration. A type that occurs a few times in the corpus involves a family member announcing that 'we will say grace' (*pomodlimy się*), which recruits the others to move into the appropriate posture. Other declarative formats are attested as single cases, such as that of the second person non-modal declarative turn in (14). Paweł and Klaudia are preparing a salad and in line 1 Klaudia brings a peeled cucumber to where Paweł is standing, for him to slice. In line 2, Paweł recruits Klaudia to give him a bowl; in response, Klaudia turns to the cupboard and gets a bowl out.

```
(14) PP4-1_0812980
  1 (4.0) ((Klaudia walks towards Paweł))
▶ 2 paw dałabyś mi m- (.) miskę,
         give.2SG.F.PST.COND me.DAT m- bowl.AKK
         you'd give me a b- (.) bowl
▷ 3 kla (turns to cupboard, gets bowl))
```
At first glance, this turn looks just like a request formulated as a second person yes/no question (see §3.3.4 below). However, in this sequential position, the prosody of the turn – with stress on the first syllable of *dałabyś* 'you'd give' and level turn-final intonation – clearly marks it as a statement.

# **3.3.3 No predicate**

Recruiting turns without a predicate are common in the Polish corpus (see also Extract 28). These are most often names of objects (*n*=16) requiring some action. In (15), the family are busy laying the table for supper. Talk is about a nearaccident that the family dog has had with a car (lines 1–5). Bogusia is taking

8 Recruiting assistance and collaboration in Polish

things out of the fridge. At line 7, she puts a bowl of salad onto the worktop and says *jeszcze proszę sałatę* 'also, please, salad', recruiting an unspecified family member to take the salad and put it on the table. *Proszę* 'I plead, please' is the imperfective form of the same verb that we have encountered in the object request in (13). This imperfective form is commonly used in actions of passing or offering an object to another person. In this function, it is best translated as 'please' or 'here you are'. *Sałatę* 'salad' here is not an argument of *proszę* 'I plead, here you are' but a stand-alone item naming the object that has been made available for taking by somebody.

(15) PP3-2\_0338665a


Naming an object does not select a particular person for the job at hand. A generic danger of such an "untargeted" recruiting move is that others can choose not to feel addressed (unless addressing is done in other ways, e.g., through gaze). This is what happens here: all the remaining family members have turned to the dog, and the recruiting move remains unanswered (and is pursued by Bogusia a few moments later).

Some recruiting turns without a predicate only "activate" another person with a vocative, leaving the required action to be inferred.<sup>7</sup> In (16), the family are

<sup>7</sup> In this project, we use the term "vocative" to refer to proper names addressing the recruiting move to a person, and not just in relation to vocative case.

preparing for a craft activity with their children, making an earthworm. This preparation involves getting the children to come to sit down at the table (lines 1–3) and making space on it. Marta is in the process of stowing things away in a cupboard; Karol is on the other side of the table. He picks up a piece of crockery that is on the table and places it closer to Marta, saying *mamuśka* 'mommy' (line 6). This initiates a recruitment for Marta to stow away the crockery as well and thereby make more space on the table. It might be that Marta first misunderstands Karol's recruiting turn as summoning her to the table (in extension to his directives towards his daughter a bit earlier, lines 1 and 3). Her initial response (line 7) is fitted to either recruitment – to sit down or to clear away the piece of crockery – but her subsequent, redone response (line 9) is specifically fitted to a recruitment to clear away the crockery.

```
(16) PP6-1_8650
```


### **3.3.4 Question formats**

Since Polish does not have interrogative morphosyntax, I speak more generally of "question formats" in this section. Question formats are morphosyntactically

### 8 Recruiting assistance and collaboration in Polish

equivalent to declaratives in Polish. But they become recognizable as questions through prosody and the distribution of epistemic rights among participants (Weidner 2013a) or the use of question words in the case of content questions. The particle *czy* can be used turn-initially to mark a polar question but is rarely used in spoken Polish and not at all in the data at hand. About 7% of all recruiting moves in the Polish corpus (*n*=15) have a question format.<sup>8</sup>

Nine recruiting moves in question format project a polar response. Polar questions are sometimes used to indirectly recruit B for some action. In (17), the family have sat down for supper, and at line 1 one of the sons implores the parents (both of them, using a second person plural double imperative) to turn on the TV. His mom, Aga, turns to dad, Piotr, with a question: *włączymy¿* 'do we turn it on?'. A yes-response to this question would imply that somebody should now turn on the TV. In the current situation, seating arrangements are such that Piotr is best placed to do such a job, while Aga is also feeding the baby. Immediately upon completion of the question, Piotr turns his gaze towards the TV and quickly moves up his torso, presumably getting up from the table. Ultimately, though, he interrupts that movement and rejects the proposal to turn on the TV, because this would make it more difficult for the researchers to listen to the recording (line 7).

```
(17) PP5-5_47880b
```


Recruiting turns that, at least at first glance, merely ask for a decision or for information are in danger of being treated as just that. In (18), Karol and his two

<sup>8</sup> In addition, a few of the cases described below as infinitives could also be counted as "questions".

### Jörg Zinken

daughters are sitting at a table, preparing for a crafts activity. Marta is not sitting yet but standing behind one of the daughters, doing the girl's hair. At line 1, she asks her husband, who is sitting at the other side of the table, *masz tam wolne jedno krzesełko* 'do you have one free stool there?'. It might be evident to Karol that Marta is asking that question because she still needs a stool to sit on at her side of the table. However, he does not take the opportunity to hand a stool to her – that is, he does not take up her question as a recruiting move – but instead treats 'do you have one free stool there?' merely as a request for information.

```
(18) PP6-1_0520400
```

```
1 (3.4)
▶ 2 mar masz tam wolne jedno krzesełko?
          have.2SG there free one stool
          do you have one free stool there?
▷ 3 kar mam
          have.1SG
          I do
  4 mar ((walks around table, carries stool to her place))
```
Questions that (ostensibly) ask whether B can or will do some action are used conventionally in various languages to recruit another person (see, e.g., Kendrick, Chapter 4, §4.2.1 on English). These question formats are understood as recruiting moves also in Polish. In the present Polish corpus, however, such recruiting moves are rare (*n*=4). What is more, they are resisted in three out of four cases, either by ignoring the recruiting move altogether (see below, §4) or by overtly displaying annoyance while complying. This indicates that, in Polish, this format might be restricted to attempts at recruitment that are judged by A to be particularly sensitive (see also Rossi 2015: chap. 4 and Chapter 5 of this volume, §3.3.3 on 'can you x' requests in Italian). On the other hand, in the one case where the recruitment is not resisted, shown below as (19), the question format is treated as overly cautious (see also Zinken & Ogiermann 2013 on a similar case).

Dorota wants Wiesia to take a plate out of the cupboard for Dorota's granddaughter.<sup>9</sup> She first formulates a perfective imperative turn (line 4), but seeing that Wiesia has already started moving toward the table to sit down (line 3) as just suggested by Kasia (line 1), Dorota immediately changes the formulation of her recruiting turn. This is in polar question format (line 5), selected here possibly because complying will now require Wiesia to depart from her current trajectory of sitting down (see Wootton 1997; Rossi 2012). The verbal response accompany-

<sup>9</sup>Zinken (2016: chap. 4) provides a more detailed discussion of this case.

### 8 Recruiting assistance and collaboration in Polish

ing the nonverbal compliance (line 6) begins with the particle *no*, which in turninitial position can indicate that the previous turn communicated something that is obvious (Weidner 2013b), followed by *tak* 'yes' with marked prosody involving high pitch onset and lengthening (see also Bolden 2017 for Russian). As a whole, this verbal response seems to indicate that Wiesia's compliance with the request is obvious and need not have been questioned.

(19) Pa02Apr2012\_0823880

```
1 kas siadasz?
          sit.IPFV.2SG
          are you sitting down?
  2 (0.6)
  3 wie [((starts moving towards table, then stops))
▶ 4 dor [↑daj jej talerz mamo (przepraszam)
          give.PFV.IMP her.DAT plate mom.VOC (apologize.1SG)
          give her a plate mom (I'm sorry)
▶ 5 dasz jej¿
          give.PFV.2SG her.DAT
         (will) you give her?
▷ 6 wie no t↑a::k ((turns to cupboard for plates))
          PTC yes
          no yes
▷ 7 ((walks towards cupboard))
```
Other recruiting turns in question format are used even more rarely, and are attested only as single cases in the corpus. For example, a speaker can try to get another person to stop doing something by (ostensively) demanding an account (*po co robisz x* 'why are you doing x').<sup>10</sup> Or they might ask 'who will do x' to get somebody to volunteer (*kto wyjmuje naczynia ze zmywarki* 'who is taking the dishes out of the dishwasher').

### **3.3.5 Infinitive**

Infinitive constructions are functionally versatile in a way that is particularly relevant to the domain of recruitments. Depending on context, prosody, and lexical turn construction, they can embody various "directive-commissive" actions from requests to offers to suggestions to proposals (see Couper-Kuhlen 2014). Similar to turns in the no-predicate category, infinitive turns cannot be categorized for sentence type. One way of thinking about this construction is to treat it as an

<sup>10</sup>See (35) for a use of this format as a way of rejecting recruitment.

### Jörg Zinken

elliptical construction that has developed out of a modal (declarative or interrogative) sentence, e.g. '[you must] tie your laces', '[shall we] make a salad?' etc. (see Deppermann 2006 on such "deontic infinitives" in German). Striking features of this construction as a recruiting move are its modal vagueness (it is not always clear whether the relevant action is something that must or could or should be done) and its impersonality: it does not formally specify who should or must do the relevant action. Consider (20), presented earlier as (12). In line 1, Olek suggests to his daughter that her son (his grandson) should or could be given some sausage to eat, using an infinitive turn.

### (20) MiBrApr2012\_0456292

```
▶ 1 ole jakiejś wędlinki może mu dać
         some.GEN sausage.GEN maybe him.DAT give.INF
         maybe to give him some sausage / maybe he should be given some sausage
▷ 2 kas (Józienko) chcesz coś zjeść
          NAME.VOC want.2SG something eat.INF
         Joseph you want to eat something?
  3 wiesz co możesz mu (.) zmontować
         know.2SG what can.2SG him.DAT mount.INF
         you know what, you can mount for him
  4 to siedzenie (z y::)
         this seat with INTJ
         this seat ( )
```
Olek's turn in line 1 could be a strong suggestion that Kasia could give the child some sausage, or it could equally be a tentative proposal that she give the child some sausage. In response, Kasia asks her son whether he would like to eat, but then moves to a counter-request for Olek to mount the child seat in preparation for the child's meal. All cases of (deontic) infinitives in the corpus are mitigated with *może* 'maybe', which gives them the quality of a suggestion or proposal, rather than of a blunt order (see also Wierzbicka 1991; Królak & Rudnicka 2006).

# **3.4 Additional verbal elements**

Like recruitment sequences, individual recruiting moves can be more or less complex. In this section, I consider verbal elements beyond those required by the argument structure of the predicate. These include mitigators or strengtheners, vocatives, the provision of reasons in a turn with multiple turn-constructional units, benefactives, or adverbs that suggest a connection of the recruited action to ongoing activities.

8 Recruiting assistance and collaboration in Polish

### **3.4.1 Mitigators and strengtheners**

The previous case (20) already provided an example of an additional verbal element, namely the mitigator *może* 'maybe', which softened the deontic force of the infinitive. The following recruiting move has the form of an impersonal declarative (see also §3.3.2 above). After Ilona and Jacek have agreed to swap childcare duties so that Jacek could finish his meal (lines 1–7), Ilona formulates the target turn, initiating recruitment for Jacek to turn off the camera (lines 8–9). This turn contains elements that mitigate the recruiting move and seem to acquiesce to an earlier suggestion (*chyba* 'probably' and *faktycznie* 'really').

(21) PP2-2\_2315590

```
1 ilo może ja się nim zajmę skończysz co¿
          maybe I REFL he.INS occupy.1SG finish.2SG what
          maybe I take care of him, you finish, okay?
  2 (1.0)
  3 ilo skończysz y:: z:jeść.
          finish.2SG INTJ eat.INF
          you finish eh eating
  4 (.)
  5 jac dobrz[e
          good.ADV
          okay
  6 ilo [skończysz¿
               finish.2SG
               you finish?
  7 (.)
▶ 8 ilo dobrze ↑to chyba już można
          good.ADV then probably already possible
          okay, then it is probably really already possible
▶ 9 wyłączyć faktycznie
          turn.off really
          to turn (it) off
▷ 10 jac ((turns camera off))
```
### **3.4.2 Vocatives**

Vocatives are present in roughly 12% of all recruiting moves (*n*=27). Vocatives can be inserted at the beginning of the recruiting move to single out the addressed party and mobilize the addressee's attention. In (22), Jacek is involved in a conversation with his children, Asia and Bolek, with his body facing them. Ilona's

### Jörg Zinken

recruiting move in line 6 begins with a substantial portion of talk that is preliminary to the request, and that can serve to ascertain that Jacek will be attending to Ilona's talk by the time the request is formulated.

```
(22) PP2-5_1423040
```


Sometimes, vocatives are inserted at the end (*n*=7) or in the middle (*n*=6) of a recruiting turn-constructional unit (TCU). Extract 1, reproduced here as (23), illustrates such a case, where a vocative is inserted after a move has become recognizable as a recruiting one but before the TCU's possible completion. Jacek is gazing at Ilona, and when she turns her gaze to him, she formulates a request for the salad bowl. The request turn begins with a turn-initial element, *wiesz co* 'you know what' (line 2), like (22) did. The vocative *kochanie* 'dear' (line 2) comes after Ilona has told Jacek to give her something, but before telling him

### 8 Recruiting assistance and collaboration in Polish

what to give her. Such a vocative can do work to disambiguate between potential addressees, although this does not seem to be the case here: Jacek is already being addressed through gaze, and the only other people present are two young children who are engaged in a separate conversation. Turn-final and turn-medial vocatives might rather be doing some affiliational work in recruitments, as we will see in §6 below (cf. Lerner 2003). Possible functional differences between these two positions will require further research to be elucidated.

```
(23) PP2-1_2224980
```

```
1 jac ((gaze to Ilona))
```

```
▶ 2 ilo wiesz co podaj mi kochanie jeszcze
          know.2SG what pass.IMP me dear still
          you know what, pass me some more
  3 kas sałatki
          salad.GEN
          salad, dear
```

```
▷ 4 jac bardzo proszę ((passes salad bowl))
          very plead.1SG
          here you are
```
### **3.4.3 Reasons**

Sometimes, speakers give a reason for recruitment (*n*=21). Reasons can be given to make a request easier to understand and comply with (Baranova & Dingemanse 2016). In (24), the recruiting move might be barely intelligible without the appended reason. Aga is holding her baby Feliks in her arms, and the baby has fallen asleep. Piotr, the family father, is admonishing the two sons, Przemek and Łukasz, to stop mucking about. At line 3, Aga admonished the others to be quiet – a recruitment that might be difficult to make sense of, and be hardly acceptable to the others without the subsequent reason.

```
(24) PP5-1_301160
  1 prz hehehe
  2 pio je::dz (że) Łukasz n[o:
         eat.IPFV.IMP (that) Lukasz PTC
         eat now Lukasz no
▶ 3 aga [sz::
                                 sh
▶ 4 bo Feliks mi zasnął
         because NAME me.DAT fall.asleep.3SG.PST
         because Feliks has fallen (me) asleep
▷ 5 ((Piotr, Przemek, Łukasz gaze at Aga))
```
### Jörg Zinken

But reasons can also have other interactional motivations. In (25), the provision of a reason seems to be mainly a vehicle for doing affiliational work between partners. The pair's toddler, Staś, has been pleading to get a dummy for some time (also in line 1). In line 3, his mom Ilona gives in. She recruits her partner Jacek to bring the dummy, and she expands this recruiting move with a reason that expresses her exasperation in a humorous way.

```
(25) PP2-2_1616090
```

```
1 sta khykhy Hha .Hh::=monia?
                      dummy?
  2 (0.8)
▶ 3 ilo monia. monia=tatusiu przy[nieś tego
        dummy dummy daddy bring.IMP this
        dummy, dummy, daddy get that
▷ 4 jac [już.
                              already
▶ 5 ilo monia bo ja dostanę: [choroby nerwowej
        dummy because I get.1SG illness nervous
        dummy because I am having a nervous breakdown
▷ 6 jac [((puts down cutlery))
```

```
▷ 7 ((gets up))
```
Reasons are not always introduced with a *bo* 'because' and appended to the recruiting component. In (26), Kasia starts her turn with an observation: the toddler fed by Wiesia has a runny nose. This observation then becomes the grounds on which Kasia incrementally builds an extended recruiting turn.<sup>11</sup>

```
(26) Pa02Apr2012_1127560
```

```
1 wie czekaj mniejszy kawa[łek
        wait.IPFV.IMP smaller piece
        wait, a smaller piece
▶ 2 kas [katar
                               cold/runny nose
▶ 3 ↑podaj [husteczkę ((point towards tissues))
        pass.PFV.IMP tissue.ACC
        pass a tissue
▷ 4 dor [((gets up))
```
<sup>11</sup>In fact, the observation might have been sufficient to mobilize Dorota to get a tissue: Dorota starts getting up after the first word of the recruiting TCU, before Kasia has formulated the object she wants to be passed.

8 Recruiting assistance and collaboration in Polish

### **3.4.4 Benefactives**

Speakers sometimes formulate the beneficiary of the recruitment, which may be the recruiter (*n*=14) or another participant (*n*=28), usually a child. However, as (26) illustrated, formulating the beneficiary is not obligatory in spoken Polish even with recruitments that involve 'giving' or 'passing' something.The question therefore arises as to what function benefactives serve. One context in which benefactives are used is contrastive, as shown in the next example, where Bogusia is getting Magda her promised dessert, biscuits, and Henio recruits Bogusia to get something else for him (line 4).

```
(27) PP3-1_1236810
```

```
1 bog dobrze dobrze już wyjmę te pieguski=
         good.ADV. good.ADV already take.out.PFV.1SG these cookies
         okay okay, I'm already taking the cookies out
  2 mag =pieguski marki:zy >pieguski mark[izy
          cookies biscuits cookies biscuits
  3 hen [to-
                                      then-
▶ 4 a mi możecie dać tego piernika
         and me.DAT can.2PL give.INF this.GEN gingerbread.GEN
         and to me you can give that gingerbread
  5 (0.8)
▷ 6 bog dobrze
         good.ADV
         okay
▷ 7 ((brings gingerbread to the table))
```
### **3.4.5 Adverbs embedding the recruitment in a larger activity**

Adverbs such as *jeszcze* 'still, also', *też* 'also', and *już* 'already' can connect the recruitment to a larger activity (*n*=8). In (28), Klaudia and Paweł are preparing a meal. On his way to the fridge, Paweł stops and turns around, looking at the oven, apparently unsure about what to do next. A moment later, the *jeszcze* 'still, also' in Klaudia's recruiting move (line 3) marks the recruited action as part of the larger activity of gathering ingredients for the meal they are preparing (see also Extracts 1, 2, 15, and 32).

```
(28) PP4-1_620160
```

```
1 paw a- °czekaj°
            wait.IPFV.IMP
        a- wait
```
Jörg Zinken

```
2 (1.0)((Paweł stops, turns towards oven))
▶ 3 kla sera jeszcze
          cheese.GEN also
          (we need) cheese still
▷ 4 paw prosz: ((opens fridge, passes cheese))
          plead.1SG
          here you are
  5 kla dzię:ki¿
          thanks
          thanks
```
The temporal adverb *już* 'already' can connect the recruitment to a larger course of action by marking the requested action out as a temporal milestone (e.g. the endpoint) within that activity. In (29), Ela has been offering her daughter Gabi various items of food. Tadek requests at line 4 that she stop distracting the daughter from eating what she has on her plate (a recruiting move that Ela disregards at line 6).

(29) PP1-1\_1230310

```
1 ela Gabi może chcesz ka- tego brokułka?
          NAME maybe want.2SG po- this.GEN broccoli.GEN
          Gabi maybe you want some of this broccoli
  2 (0.8)
  3 gab nie::
          no::
▶ 4 tad nie mieszaj już jej
          not confuse.IPFV.IMP already her.DAT
          don't confuse her now / stop confusing her now
▶ 5 [niech ona je to co
          may she eats.3SG this what
          she should eat (let her eat) what-
▷ 6 ela [a może dać ci marchewkę.
          and maybe give.INF you.DAT carrot.ACC
          or maybe you want a carrot
```
# **4 Formats in Move B: The responding move**

The space of possible next actions by participant B after a recruiting move by participant A can be partitioned into two nested sets. At one level, B can either produce some response to the recruiting move or not respond to it at all; if B responds to the recruiting move, the response can either work towards complying with the recruitment or embody non-compliance.

### 8 Recruiting assistance and collaboration in Polish

Let us consider the first of these distinctions: responding in some way vs. not responding at all. A lack of response to the recruiting move is not uncommon in the Polish data: there are 23 such cases in the corpus (11%) in which B "ignores" the recruiting move as it were (cf. Blythe, Chapter 7, §4.2.4). A closer look, however, reveals that many of these cases are more benign.<sup>12</sup> A potential recruitee who is already involved in some work – especially if it is work related to a wider activity within which the recruitment emerges – might claim some allowance for not attending to the recruitment "just now". The clearest cases of this come from nonverbal requests. For example, think back to the activity of two people checking the medicines they have in the fridge (Extract 5). Marta repeatedly takes a package of medicines from the fridge, inspects it, and then holds it out for Karol to take. In that interaction, there are two instances where Marta holds the package out for Karol to take, but Karol is still inspecting the package he was given previously. Noticing that Karol is not attending to her gesture, Marta puts the new package on the table, from where Karol eventually takes it. It seems plausible that potential recruitees can also use their being occupied strategically as a way to avoid responding to a recruiting move. Extract 30 might be an example of this (also discussed in the context of deontic infinitives, see Extract 20 above). Olek is beginning to mount a kind of child seat for his toddler grandson, following a request by Kasia to do so (lines 3–4). At line 6, Kasia incrementally extends her request with another one, namely that Olek should also put a cushion onto the chair (so that the toddler would sit high enough to be securely held in place by the child seat, and to be able to reach the table). At this time, Olek has already begun mounting the child seat, and there is no response to this subsequent recruiting move. At lines 8–9, Wiesia reformulates the request made by Kasia at line 6, but again, Olek does not respond to this but carries on trying to unravel parts of the child seat. At lines 11–12, Wiesia incrementally extends the request and thus provides another occasion for Olek to provide a response, which he does not do (line 13). At line 14, Wiesia announces that she will bring this cushion herself. Shortly after this, Olek puts the part of the child seat that he has been wrestling with down on the table and starts walking towards the next room, at the same moment as Wiesia. Seeing that Olek is now (presumably) on his way to get the cushion, Wiesia stops and walks back to where she was working in the kitchen, and formulates another increment to the request, specifying the kind of cushion (lines 17–18).

<sup>12</sup>These cases are more difficult to quantify in that it is not always clear whether a person has genuinely not heard a request or simply does not want to hear it (more on this below).

### (30) MiBrApr2012\_0456292

	- 7 ole ((begins mounting child seat on chair)) (1.8)
	- 10 (0.2)
	- 13 (3.2)
	- 14 wie zaraz przyniosę in.a.moment bring.1SG I'll bring it in a moment
	- 15 (1.6) ((Olek puts child seat down on table))
	- 20 ((Olek comes back with cushion after some time))

### 8 Recruiting assistance and collaboration in Polish

In short, there is a series of recruiting moves here, and Olek does not produce an on-record response to any of them. Instead, he starts a move that is conceivably the complying response (lines 15–16) in a position where it is contiguous to a prior turn that was *not* a recruiting move (line 14). "Ignoring" another person's requests would seem to be a socially sensitive matter. However, this might be mitigated here by the fact that Olek is already involved in work on the child seat. The cushion might only be required once the child seat itself is fixed to the chair. In other words, Olek's non-responsiveness may not be treated as "ignoring" if it can be accounted for as him being busy with step 1 of the project of preparing a seat for the toddler (securing the child seat to the chair) before moving on to step 2 (providing a cushion).

Another context in which B sometimes does not respond, but may not be fully held accountable for "ignoring" the recruiting move, is when this is formatted as an impersonal declarative (see §3.3.2 above; see also Rossi & Zinken 2016). Consider (31), where the family are at the dinner table and Jacek is feeding his toddler son on his lap (line 1 is part of that interaction). At lines 2–3, Ilona formulates an impersonal declarative: *Stasiowi by się przydał widelczyk*, roughly: 'a fork for Staś would be useful'. This turn is prefaced with *wiesz co* 'you know what', which marks it as being addressed to some individual (Lerner 2003). However, Ilona does not use any formal resources that would convey who is to get the required fork (she is cutting food on her plate and gazes down throughout her turn). It is the fact that the recruitment attempt is concerned with a childcare matter that makes the turn relevant for her partner. However, Jacek does not provide any response. Ilona fills the emerging silence with another short turn thinking out loud (line 5). When Jacek still does not begin any response to the recruiting move, engaging instead in a short exchange with his son (lines 6–7), Ilona begins a new turn, which explicitly addresses the recruiting move, in different form, to her daughter, Iza (lines 8, 9, 11).

(31) PP2-2\_241620a


### Jörg Zinken


We now turn to recruiting moves that receive a response.

# **4.1 Fully nonverbal responses**

One way – arguably the basic way – of responding to a recruiting move is to do the relevant action. There are many such cases among the examples discussed so far (Extracts 2–3, 5–6, 8–14, 21–22, 24, 26). Fully nonverbal compliance is common when a recruited action can be performed quickly and easily (Rauniomaa & Keisanen 2012): passing a knife across the table, picking up something that has dropped to the floor etc. Out of 69 cases of such quick compliance in the data, 50 (72%) come without any verbal element. What is maybe more surprising is that fully nonverbal compliance is also common in cases where doing the relevant action takes more time, where it is necessary to create certain conditions for the requested action first: going to the kitchen in order to fetch a spoon, for example. There are 72 cases in the data where B's next move after a recruiting one is the first step of a compliant response, but where that compliance takes a bit longer (or might become stalled after that first move). Of these recruitments, 45 cases (63%) do not involve any verbal response. Extract 32 illustrates such a case. Jacek and Ilona are talking to their son about possible places where he could search for his lost ball (line 1 is a contribution to this conversation). At lines 2 and 5, Ilona recruits Jacek to also look 'here', that is, in a corner of the room. In response, Jacek takes steps in that direction and begins moving back some furniture to look for the ball. His response is not accompanied by any verbal turn.

8 Recruiting assistance and collaboration in Polish

```
(32) PP2-1_3936480
  1 jac może być też
         can.3SG be.INF also
         it can also be (there)
▶ 2 ilo i może jeszcze Jace::k¿
         and can.3SG still NAME
         and maybe also, Jacek?
  3 (.)
  4 jac °hm°
▶ 5 ilo [y:: rzuć okiem [o tutaj w:: (°° °°)
          INTJ throw.PFV.IMP eye.INS PTC here in
          eh, have a look o here, in ( )
  6 [((head nod)) [((head nod))
▷ 7 jac ((turns and searches for ball))
```
# **4.2 Verbal elements of responses**

Verbal elements accompanying complying responses to recruitments can be ordered according to their grammatical complexity (cf. Thompson et al. 2015). The simplest verbal responses are polar responses that indicate (upcoming) compliance or reject the recruitment. As mentioned above, such responses might be more relevant for recruited actions that are not quick and easy. However, the relevance of a particular type of verbal response might also depend upon the form of the recruiting move. Recruiting moves in polar question format grammatically project a polar response that accepts the recruitment (Raymond 2003).<sup>13</sup> Quick and easy compliance can diminish the usefulness of accepting – after all, acceptance should occur before the actual compliance. Out of the four conventional request moves in polar question format in the Polish data, one receives a polar response (*no tak* 'PTC yes', Extract 19), while the other three are "problematic" recruitments (see §3.3.4 above). Out of 93 imperative recruiting moves, only one receives a polar response: a flat-out rejection with *nie* 'no' (Extract 34, see below). The action of accepting conveyed by a positive polar response does not seem to be relevant in response to imperative recruiting moves in Polish (see also Craven & Potter 2010; Rossi 2012 for English and Italian). This does not mean that there are no verbal responses to imperatives. However, these verbal responses emphasize compliance rather than accepting the recruitment, e.g. *proszę bardzo* 'here you are', *masz* 'here you are', literally 'you have', *już przyniosę* 'already I bring it' (see Zinken 2016: chap. 5, for a discussion).

<sup>13</sup>But cf. Thompson et al. (2015) for an argument against this view.

### Jörg Zinken

Another response token is *dobra* or *dobrze* 'okay', which appears to indicate compliance "in principle", in a situation where maybe immediate compliance is not possible, or the recruitee does not know how to go about the recruited action (see Extract 27). A practice found repeatedly in the corpus is to begin a verbal response with a temporal adverb, for example, *już* 'already'. In response to an onrecord request, *już* can be produced as a response not just to indicate compliance but also to treat the request as urgent (see Extract 25). Clausal responses with a turn-initial *już* enact stronger agency and initiative on the part of the recruitee (see Extract 4).

In general, clausal responses do more than simply indicating compliance. In (33), Ania notices that the sauce she requested earlier has not been poured on her food, and she asks for it again (line 2), designing her turn as "having to ask again". In other words, her turn is formatted not just (and maybe not primarily) as a request, but as a complaint. The clausal response in line 4 is fitted to this "double-barreled" first action (Schegloff 2007: 76; see also Kitzinger et al. 2013; Rossi 2018). The response indicates not only or not so much compliance with the request, but manages the disaffiliational undercurrent through a relatively elaborate verbal offering of the sauce.

(33) PP1-1\_0509630b

```
1 (1.6)
▶ 2 ani ale ja jeszcze poprosiłam (.) y °sosiku° =moment
         but I still ask.PFV.PST.1SG INTJ sauce.DIM.GEN moment
         but I still asked for some sauce wait a moment
  3 (1.0) ((Ania picks up her plate, walks toward cooker))
▷ 4 ela ↑no to ma:sz.
         PTC then have.2SG
         no then here you have (some)
▷ 5 ((serves Ania sauce))
```
# **4.3 Types of rejections**

Recruitments are rejected in 23 cases in the corpus (nearly 11%). Overt rejection with just the response particle *nie* 'no', however, occurs in only one case (Extract 34). Klaudia and Paweł are having supper, and when Paweł moves up his fork with melted cheese sticking to it, Klaudia pleads with him to give her the cheese. Paweł responds curtly with a 'no', and then turns to the family dog squealing at his legs.

8 Recruiting assistance and collaboration in Polish

(34) PP4-1\_2301200

1 paw ((moves up fork with melted cheese sticking to it)) ▶ 2 kla da::j mi tego żółtego sera give.PFV.IMP me.DAT this.GEN yellow.GEN cheese.GEN (do) give me some of that yellow cheese ▷ 3 paw nie¿ no no

4 ((turns to dog))

It is questionable whether Klaudia's request in (34) was serious. It is more likely that she did not really expect Paweł to scrape the cheese from his own food and pass it to her. Her plea for the cheese might more plausibly be part of some kind of tease between the two, and this also puts Paweł's seemingly blunt rejection in a different light. In any case, rejections are overwhelmingly done in ways that avoid being blunt in one way or another.

One way of rejecting a recruitment is to question the need for the requested action (see Zinken & Ogiermann 2011). A format for rejection in Polish that at least ostensibly does this is *po co* 'what for'. However, this format does not really seem to question the need for the requested action – a reason is never provided in response, and is never pursued. Instead, questioning the need in this format works as a practice for rejecting a recruitment (cf. Bolden & Robinson 2011 on account solicitations with *why*). In (35), Wiesia is walking around the flat with her toddler granddaughter. At line 2, Dorota, who is sitting at the kitchen table, recruits Wiesia to turn on the light in the corridor where she and the toddler are (in fact, Wiesia had just switched the light off, but Dorota might not have noticed). Wiesia does not respond to this recruiting move, continuing instead a turn addressed to her granddaughter (lines 1 and 3). Dorota repeats her recruiting move in line 5 and, after some silence, Wiesia rejects the recruitment with *a po co (światło)* 'but why (light)' (line 8).

(35) Pa02Apr2012\_0725770b


	- 3 wie może coś zjesz maybe what eat.PFV.2SG maybe you'll eat something

### Jörg Zinken

```
4 (.)
▶ 5 dor zapal tam ↑światło.=°mamusiu°
         turn.on.PFV.IMP there light mom.DIM.VOC
         (do) turn the light on there mommy
  6 (0.4) ((Wiesia walking towards kitchen with toddler))
▶ 7 dor Pol[uniu:¿
         NAME.DIM.VOC
         Polly
▷ 8 wie [a po co [(światło)
            and.but for what (light)
            but why (light)
  9 dor [Poluniu zjesz jeszcze salami?
                           Pola.DIM.VOC eat.PFV.2SG still salami
                           Polly will you eat some (more) salami?
```
A common element in turns rejecting a recruitment is an informing TCU that can be taken as providing an explanation for not complying. Sometimes such turns begin with a rejection token (*nie* 'no', *n*=2) but more commonly they do not.

Extract 36 is a case where a recruitment is rejected with a *nie* 'no' plus explanation. This case comes from the same setting as the previous Extract 35. Dorota is asking Wiesia and the toddler, who are walking around the flat, to come to the table to eat something as the rest of the family are having breakfast. Wiesia initiates repair at line 3, and Dorota redoes the recruiting move, addressing it now only to the toddler (line 4). However, Wiesia apparently does not notice this and responds with a rejection token (*nie* 'no') and an appended explanation (line 5).

```
(36) Pa02Apr2012_0714730a
```

```
▶ 1 dor cho::dźcie zjeść z nami
          come.IPFV.IMP.2PL eat.INF with us.INS
          come (you two) eat with us
  2 (0.2)
▷ 3 wie proszę¿
          plead.1SG
          excuse me?
▶ 4 dor chodź córuś może zjesz coś.
          come.IPFV.IMP daughter.DIM maybe eat.2SG something
          come (my) daughter maybe you'll eat something
▷ 5 wie nie:: ja jestem po śniadaniu.
          no I am.1SG after breakfast.LOC
          no, I have had breakfast
  6 dor ale nie do ciebie mówię(h) ((laughter))
          but not to you talk.1SG
          but I am not talking to you
```
### 8 Recruiting assistance and collaboration in Polish

More commonly, a rejection is accomplished with just an explanation for not doing as requested. In (37), Ania has sat down with her back straight to the camera, and the participants have just commented on this. At line 1, Ela directs Ania to sit 'here', on the chair next to the one she is sitting on. Ania does not respond to the initial recruiting move, with her gaze directed at the free chair Ela is indicating. When Ela redoes the recruiting move in amended format, Ania rejects this with a turn composed of two units, each of which formulates a reason for not taking the "better" chair: *ale ja nie zostanę* 'but I am not staying' (line 4) and *ja już jestem po śniadaniu* 'I have already had breakfast' (lines 4–5).

(37) PP1-1\_0615520b

```
▶ 1 ela usiądź tu ((points to vacant chair))
          sit.down.PFV.IMP here
          sit down here
  2 (0.8)
▶ 3 ela siada:j
          sit.IPFV.IMP
          (do) sit (down)
▷ 4 ani ale ja nie zostanę ja już jestem
          but I not stay.PFV.1SG I already be.1SG
          but I am not staying, I have already
▷ 5 po śniad(h)a(niu)(h)
          after breakfast
          had breakfast
```
# **5 Acknowledgment in third position**

As in the other languages examined in the comparative project (Floyd et al. 2018), acknowledgment of compliance is rare in the Polish data: only 3 cases were found. We have seen two of these in (22) and (28), where the recruiter thanks after receiving a requested object. The third case, shown below, is also an object request. Kasia asks Georg to pass the horseradish across the table. Georg does this, accompanying the action with a verbal turn, *prosz*:: 'here you are'. Kasia takes the horseradish and quietly says *dziękuję* 'thanks'.

(38) MiBrApr2012\_0552334


### Jörg Zinken

3 kas °dziękuję° thank.1SG thanks

Thanking is a way of recognizing another's agency in providing assistance (Zinken et al. 2020). It is also worth noting that in two out of three cases, the recruitee points to his compliance with *proszę* 'please, here you are', which might make the provision of an acknowledgment more likely.

# **6 Social asymmetries**

The videos in the Polish corpus were recorded by families in their homes. Social asymmetries enter the picture in so far as interactions are sometimes between parents and their adult children.The interactions mostly take place in the parents' homes, and both the setting and the social relationship might contribute to some deference on the part of the adult children. No strong influence was noticed in terms of the ratio of fulfillments to rejections. However, one striking aspect in the formulation of recruiting moves is the common use of vocatives by adult children when attempting to recruit their parents. Out of 24 recruitment sequences in which the recruiter was analyzed as occupying a higher social position than the recruitee, only 3 (12%) contained a recruiting move with a vocative. But out of 23 recruitment sequences in which the recruiter was analyzed as occupying a lower social position than the recruitee, 10 (43%) contained a recruiting move with a vocative (e.g. Extract 35 above).

Extract 39 is one of those rare cases where a father uses a vocative in addressing a recruiting move to his adult daughter, Dorota. Olek has his toddler grandson on his lap, and the toddler wants to get off to walk around. This has been problematic before, because the toddler has a sausage in his hand, and Dorota – whose home is this is – does not want the little ones to run around with food in their hands. Olek addresses a turn to Dorota, in which he raises this problem and thereby recruits her to do something about it (a recruited action that is about to be made more specific in line 3). Dorota responds in two ways: she rejects the plan to 'go', attributed to the toddler (lines 2 and 5), while walking towards him, and taking the sausage from him and putting it on a plate (line 7), thus creating the circumstances in which the toddler can have his wish to walk around granted.

(39) Pa02Apr2012\_1227960

▶ 1 ole nie mo- ciocia on chce iść patrz <sup>o</sup> not pos- aunt he want.3SG go.INF look.IPFV.IMP PTC you can't- aunt, he wants to go look *o*

8 Recruiting assistance and collaboration in Polish

```
▷ 2 dor n[ie ((gaze at toddler, eyebrows raised))
          no
▶ 3 ole [trzymaj go na ( )
            hold.IPFV.IMP him.ACC on
           hold him ( )
  4 kas ale to
          but this
▷ 5 dor nie ((gaze at toddler, eyebrows raised))
          no
  6 kas poprostu go
          simply him.ACC
▷ 7 dor tutaj to hopsa i można iść ((takes sausage from toddler))
          here this hop and possible go.INF
          here (we put) this, hop, and you can go
```
Of particular interest is Olek's use of the category term *ciocia* 'aunt' (line 1). Dorota is in fact Olek's daughter and the toddler's aunt. Olek addresses Dorota in her family relationship role to the toddler, who is the target of the recruitment (see also Extracts 10, 16, and 25 above). For one thing, a vocative addresses the recruiting move to a particular person in a multiparty setting; at the same time, it provides a slot in which the choice of vocative item can be used to mobilize or acknowledge particular social relationships (see also Kendrick, Chapter 4, §4.3.1; Baranova, Chapter 9, §3.4).

# **7 Discussion**

This chapter provided an overview of practices that speakers of Polish use for the organization of collaboration and assistance in informal family settings. In many respects, the Polish data are consistent with findings from other languages in the cross-linguistic project, and with expectations based on the extant literature.

For example, the findings show that there seem to be hardly any verbal turn formats that could *not* become part of a recruiting move: imperative, declarative, and interrogative turn shapes are all attested, as are turns without a predicate and interactional moves without any talk. This supports the contention that drawing on others' cooperation is a fundamental facet of human sociality that does not make any specific demands on grammatical structures (Tomasello 2008). Imperatives are the most common sentence type in recruiting moves, as we would expect given that imperatives are dedicated to the delivery of directive actions (e.g. Aikhenvald 2010). Also, the findings support arguments for a bias towards prosocial orientations at work in human interaction (e.g. Heritage 1984). Rejections are much less frequent than compliant responses, and are mostly done by

### Jörg Zinken

providing explanations for non-compliance, rather than by bluntly rejecting the recruitment. Even cases in which a person does not respond to a recruiting move at all show traces of such a prosocial orientation: recruitees skillfully orient to aspects of the situation that could make their lack of response accountable in terms other than "ignoring".

Other findings might be cross-linguistically more restricted. For example, imperative recruiting moves with imperfective verbal aspect in Polish display that the recruiting move does not convey new information. Speakers can use this resource to indicate that the other person should have acted already – that they already knew what to do (cf. Kent & Kendrick 2016). Conventionally indirect (Brown & Levinson 1987) recruiting practices, such as questions about the ability or willingness to do something, are very rare in the examined corpus. Instead, declarative turns and turns without a predicate make up nearly half of recruiting moves with a verbal element that are not imperatives. These turn formats have received little attention in the literature relative to their prominence in (Polish) informal everyday interaction.

# **Acknowledgments**

I am grateful to two anonymous reviewers for their detailed and helpful comments on an earlier version of this chapter.

# **References**


8 Recruiting assistance and collaboration in Polish


### Jörg Zinken


8 Recruiting assistance and collaboration in Polish


### Jörg Zinken

*in action* (Studies in Language and Social Interaction 30), 27–63. Amsterdam: John Benjamins.


# **Chapter 9**

# **Recruiting assistance and collaboration in Russian**

# Julija Baranova

Department of Language and Communication, Radboud University

This chapter describes the resources that speakers of Russian use when recruiting assistance and collaboration from others in everyday social interaction. The chapter draws on data from video recordings of informal conversation in Russian, and reports language-specific findings generated within a large-scale comparative project involving eight languages from five continents (see other chapters of this volume). The resources for recruitment described in this chapter include linguistic structures from across the levels of grammatical organization, as well as gestural and other visible and contextual resources of relevance to the interpretation of action in interaction. The presentation of categories of recruitment, and elements of recruitment sequences, follows the coding scheme used in the comparative project (see Chapter 2 of the volume). This chapter extends our knowledge of the structure and usage of Russian with detailed attention to the properties of sequential structure in conversational interaction. The chapter is a contribution to an emerging field of pragmatic typology.

# **1 Introduction**

The work in this chapter was carried out as part of the comparative project on recruitment systems in eight languages presented in this volume. Chapter 1 defines recruitment as an interactional phenomenon; Chapter 2 outlines the coding scheme and explains the comparative categories used in the analysis. The present chapter offers an overview of the main practices used by speakers of Russian to recruit assistance and collaboration from their peers and family members in everyday activities such as talking, having dinner, or cooking together. The data

Julija Baranova. Recruiting assistance and collaboration in Russian. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 325–367. Berlin: Language Science Press. DOI:10.5281/zenodo.4018386

### Julija Baranova

come from a set of video recordings made by the author in Russia. The chapter begins with a brief introduction to the Russian language and a description of the corpus. I then present the basic structure of recruitment sequences. The recruiting move is analyzed in a separate section that discusses nonverbal elements of its composition as well as its linguistic formats (imperative, declarative, and interrogative). Attention is also given to additional elements such as reasons and diminutives that mitigate the recruiting move. Then the chapter discusses responding moves, broadly divided into complying and non-complying. Towards the end, I discuss the expression of gratitude and the role of social (a)symmetries in recruitment sequences. Finally, I summarize the findings and present some ideas for future research.

# **1.1 The Russian language**

Russian is an East-Slavic language of the Indo-European language family. About 150 million people speak Russian as their first language. Russian is an official language of the Russian Federation, Belarus, Kazakhstan, Kyrgyzstan, and Tajikistan. The basic word order is SVO (Hawkins 1983; Tomlin 1986). Interrogatives are formed mainly through intonation, question words, and interrogative particles. Russian verbs come in aspectual pairs: perfective and imperfective. They inflect for tense, person, number, and on certain occasions for gender. Russian nouns are marked for gender (feminine, masculine, and neuter), number (singular and plural), and case (six cases). The grammar of Russian has been extensively studied, but we are only beginning to understand how Russian is used in everyday conversational interaction (e.g. Bolden 2003; 2004; 2008; Robinson & Bolden 2010; Baranova 2015; Baranova & Dingemanse 2016). This paper offers a contribution to this line of work by focusing on the recruitment system in informal Russian.

# **1.2 Data collection and corpus**

The corpus on which this work is based was constructed in accordance with guidelines developed by and for the members of the comparative project reported on in this volume (see Chapters 1–2). Russian data come from nineteen recordings made by the author during three field trips to Russia in 2011 and 2012. The recordings took place in several locations in the region of Chelyabinsk, at participants' homes, and on two occasions at their work places. The interactions were all informal involving friends and family. The total sampled recording time was 3 hours and 20 minutes, resulting in 200 recruitment cases. The length of the

### 9 Recruiting assistance and collaboration in Russian

sample per recording varied from 10 to 25 minutes.

Interlinear glosses in the data extracts generally follow the Leipzig glossing rules (Comrie et al. 2020), albeit with some simplification in terms grammatical categories that are less relevant for the purposes of the chapter. The focus in the glosses is mainly on grammatical tense, aspect, case, and gender.

# **2 Basics of recruitment sequences**

As defined in Chapter 1, §4, a recruitment is a basic cooperative phenomenon in social interaction consisting of a sequence of two moves with the following characteristics:


This is the basic and canonical sequence, an example of which is given in the following section. Other details of what can happen, including what participant B can say or do in Move B to fulfill or reject the recruitment, are illustrated in later sections. In the transcripts, ▶ and ▷ designate Move A and Move B, respectively.

# **2.1 Minimal recruitment sequence**

When a recipient responds to Move A with immediate fulfillment of the recruitment, the result is a minimal recruitment sequence. This is illustrated in (1) where several family members are gathered for dinner at Lida's place. The extract starts with an offer sequence between Tanya and her young child.

(1) 20120114\_family\_visit\_2\_164605

```
1 tan mozhet malaka?
          maybe milk.GEN
          maybe some milk?
  2 chi ((nods with his head))
  3 (0.7) ((Tanya turns away from child and towards Lida))
▶ 4 tan malaka ((nods))
          milk.GEN
          some milk
```
### Julija Baranova

▷ 5 lid ((takes milk from refrigerator, pours it into a cup, and places the cup on the table in front of the child))

Figure 1: Tanya initiates a recruitment for milk (Extract 1, line 4).

Figure 2: Lida gets milk from the refrigerator (Extract 1, line 5).

Tanya offers her child some milk (line 1) and he accepts the offer (line 2). However, Tanya is unable to get out from the table easily. She recruits Lida's assistance using a no-predicate construction: *malaka* 'some milk' (line 4, Figure 1). Lida starts complying immediately (line 5, Figure 2). This recruitment sequence is minimal as consists of a recruiting turn followed by fulfillment with no other actions in between, such as repair initiations or redoings of the recruiting move. Fulfillment is entirely nonverbal: Lida takes the milk out of the refrigerator, pours it in a cup, and puts it on the table in front of the child.

9 Recruiting assistance and collaboration in Russian

# **2.2 Non-minimal recruitment sequence**

Non-minimal recruitment sequences are sequences where fulfillment does not immediately take place. Instead, the initial response is something other than fulfillment or compliance, such as a question or a rejection. Sometimes there is no immediate relevant response at all and the recruiting move is effectively ignored. In these cases, recruiters may pursue compliance, for example by offering a repair solution, answering a clarification question, offering a reason for the recruitment, or by simply redoing the recruiting move.

An example of a non-minimal recruitment sequence can be found in (2). The scene features Maria and her adult daughters Katya and Olga. Maria is standing at the kitchen counter talking to Olga, who is in an adjacent room. At one point, Maria places a cup with hot water on the table for Katya who is about to make herself some instant coffee.

(2) 20110827\_Family\_2\_820127


it looks like this one has been {dipped} in the milk

### Julija Baranova

```
13 mar ta da:. v malake:
        DEM.F yes in milk.LOC
        that one, yes, {it's been dipped} in the milk
```

```
▷ 14 ((gives a teaspoon to Katya))
```
15 kat spasiba ((putting coffee into her cup with the given spoon)) thanks thanks

Figure 3: Katya is about to put instant coffee into the cup of hot water (Extract 2).

At line 8, Katya initiates recruitment of Maria by using an imperative construction with rising-falling intonation. That is, she starts with a high pitch and end with a low one: *↑daj lo:shku druguju pazhalusta* '↑give {me} another spoon please'. Instead of immediately complying, Maria initiates repair: 'a spoon? (.) another one?'. With this repair initiation, she claims to have trouble hearing or understanding Katya's recruiting turn. Katya responds with the confirmation *uhu:m* 'uhu:m' (line 10). It appears, however, that Maria's 'a spoon? (.) another one?' is not a simple repair initiation; it also embodies a kind of challenge (see Baranova 2015). Maria's turn may be understood to be using a claim of trouble of hearing or understanding as a way to question the need for recruitment (Schegloff 2007: 102–106). When Katya issues the request, she is holding a teaspoon (Figure 3), raising the obvious question as to why she cannot use the one she already has in her hand. At line 12, Katya expands on her initial recruiting move by orienting to just this question and supplying the reason: 'it looks like this one has been {dipped} in the milk'. So the first part of Katya's response (line 10) targets the potential problem of hearing, while the second part offers a reason that defends the relevance and purpose of recruitment here (line 12).

### 9 Recruiting assistance and collaboration in Russian

Maria's repair initiation ('a spoon? (.) another one?') delays compliance and expands the recruitment sequence into a non-minimal one, in which the recruiter supplies a repair solution and a reason to back up her original recruiting turn. Also, both lines 10 and 12 serve here as renewals of the original recruiting turn, making a response relevant (Davidson 1984; Pomerantz 1984). Maria complies at line 14 by giving a clean teaspoon to Katya. The recruitment sequence is closed off with an acknowledgment *spasiba* 'thanks' in line 15.

# **2.3 Subtypes of recruitment sequence**

In the larger comparative project, we distinguish four main recruitment types based on the nature of the response by the recruitee (see Chapter 2, §6). As Table 1 shows, the provision of a service is the most frequently encountered response type in the Russian sample. We saw an example of this in (1), where the recipient Lida fulfilled the recruitment by pouring milk into a cup for the recruiter. Recruitments resulting in the passing or moving of an object fall into the category of object transfers as exemplified by (2), in which a spoon was handed to the recruiter. The two remaining recruitment types involve alterations of trajectory of behavior (e.g. getting someone to desist from doing something) and assistance in response to visible or anticipatable trouble (e.g. open the door for someone when their hands are occupied). I discuss these in the following two extracts.


Table 1: Relative frequencies of recruitment sequence subtypes (*n*=200).

Alterations of trajectory form the second largest group of recruitment sequence types in the current sample. In (3), Marina is visiting her mother-in-law Anna. Both women are sitting at the kitchen table. Marina is holding her small dog on her lap and playing with it, while Anna is sitting next to her having dinner.

### Julija Baranova


Figure 4: Anna tells Marina to remove her dog from the table (Extract 3, line 3)

Marina is playing with her dog at the table (lines 1–2). Assuming that there is a special relationship between dogs and their owners, Marina's play with her dog might be seen as a private activity that does not include Anna. Nonetheless, Anna intervenes, which might be seen as a delicate matter. This may be why Anna's recruiting move is accompanied by a reason: '{so that} I finish eating in peace' (line 3). The request-reason combination implies that finishing eating the meal in peace is incompatible with the presence of the dog at the table. Marina

### 9 Recruiting assistance and collaboration in Russian

orients to this negative implication by offering a counter-reason: 'but we aren't doing anything to you' (line 7).

The examples discussed so far involve on-record verbal recruiting moves that make explicit the type of practical action being recruited. By contrast, recruitments of the trouble-assist type feature visible trouble but no on-record request to solve the trouble and no instruction as to how to do so. While there is no explicit initiation of recruitment, recruitees nevertheless provide assistance. This assistance may involve altering behavior, transferring an object, or performing a service. Why are such cases recruitments at all? They certainly share features with requests and other on-record recruitments. First, there is the issue of accountability. While a participant who merely sees that someone is in need is presumably less accountable for failing to assist than someone who is the addressee of an on-record recruitment, it can be argued that their failure will still be noticeable. Second, trouble-assist recruitments are hardly distinguishable from verbal recruiting moves that verbalize a trouble using a declarative statement. For instance, in (15) discussed later in this chapter, the recruiting turn consists of a declarative statement that makes the speaker's trouble clear: her toddler is chewing on a paper napkin. In response, the toddler's grandmother takes the napkin away from her.

Extract 4 illustrates how a participant can assist another person after observing the trouble that they are experiencing. This fragment is taken from a conversation between Inna and her adult niece Sasha. The women are in Inna's narrow kitchen when Sasha's mobile phone starts ringing in the corridor. Sasha visibly struggles to stand up from the kitchen bench as the table blocks her movements.

```
(4) Niece_1_1517800
  1 ((mobile phone rings))
  2 sas eta minia [kto-ta patirial
         DEM I.GEN somebody lost
         that's me somebody is looking for
▶ 3 [((struggles standing up))
  4 inn (tak) [(0.2) ( )
          so
          so (0.2) ( )
▷ 5 [((pulls the table for Sasha to pass through))
  6 sas ((passes through the opening between the table and the kitchen cabinet))
```
Sasha stands up from the kitchen bench with visible difficulty. She is squeezed between the table and the kitchen cabinets, unable to pass through. Inna is sitting

### Julija Baranova

Figure 5: Inna pulls the table so that Sasha can leave (Extract 4, line 5)

just in front of Sasha. Inna pulls the table to make more space for Anna to leave (line 5, Figure 5), which Anna is then able to do.

In this example, Sasha does not explicitly recruit Inna's assistance, but Inna gets it all the same. Important here is that Sasha might not be free to push the table forward as this would put Inna in an uncomfortable position. Also, Inna is the host here and bears some responsibility for the comfort of her guests. These features make it more likely that Inna will offer assistance without an on-record recruiting move being made.

To summarize, I have introduced four main recruitment types: performing services, transferring objects, altering behavior, and trouble assistance. While the first three types are straightforward and refer to the nature of the recruited action, the last type is different, but it should still be seen as belonging to the domain of recruitment in its broad definition (see Chapters 1–2). Trouble-assist recruitments do not involve an on-record initiating move. One of the participants assists another when a trouble manifests itself. This assistance can involve performing a service, transferring an object, or altering a behavior.

# **3 Formats in Move A: The recruiting move**

While the previous section was mainly concerned with what kind of assistance or collaboration is being recruited, in this section the focus is on the format or formulation of the recruiting move. Numerous strategies are observed, the use of which is influenced by both the immediate situational context (Rossi 2015) and cultural preferences for (in)directness (Ogiermann 2009; Bolden 2017). Initi9 Recruiting assistance and collaboration in Russian

ating moves in recruitment sequences might be fully nonverbal, fully verbal, or a combination.

# **3.1 Fully nonverbal recruiting moves**

In some situations, verbalizing a recruiting move appears unnecessary and a mere gesture might be clear enough to indicate what kind of assistance or collaboration is being called for. In (5), Pavel is one of Anna's guests at a dinner gathering. The extract begins when Anna offers Pavel a drink.

```
(5) 20120602_family_friends_2_1085520
  1 ann Pavel ↑chaj kofe
         name tea coffee
         Pavel, tea, coffee?
  2 (0.7)
  3 pav .hhhh chijku esli tol'ka luchshe
              tea.DIM.GEN if only better
         .hhhh if {possible} better some tea
  4 ann ((takes a tea bag from [the box))
  5 pav [((lifts his cup and looks into it))
  6 o:pa
         INTJ
         oh
  7 u minia eshio jest' An'
         with I.GEN still is name.VOC
         I still have some, Anna
  8 ann [((turns to different speaker)) ↑Ir
                                     name.VOC
                                      Ira?
  9 pav [((finishes his tea))
  10 (0.9)
  11 ira (ni budu [spasiba)
          NEG be.FUT.1SG thanks
          I won't, thank you
  12 ann [((puts tea bag on the [table ))
▶ 13 pav [((holds out his cup for Anna))
▷ 14 ann ((puts tea bag into Pavel's cup))
▷ 15 ((takes the cup, pours hot water into it, and gives it back to Pavel))
```
Pavel accepts Anna's offer by specifying that he would like tea (line 3). Then he notices that there is still some tea left in his cup and this is what he tells Anna

### Julija Baranova

Figure 6: Pavel holds out his cup and Anna puts a tea bag into it (Extract 5, lines 13–14).

at line 7: 'I still have some, Anna'. Anna treats this as a rejection of her offer because she immediately turns to Ira to offer tea to her. At line 13, Pavel holds out his cup towards Anna and she takes this gesture as a request for tea. With no questions asked she puts a teabag in Pavel's cup and fills it with hot water.

Such nonverbal recruiting moves can only be successful in environments that maximally disambiguate them (Rossi 2014 and Chapter 5, §3.1; see also Kendrick Chapter 4, §4.1.3; Zinken, Chapter 8, §3.1; Dingemanse, Chapter 10, §3.4). In our case, the meaning of Pavel's gesture is clear in the context of the preceding offer sequence.

This class of nonverbal recruiting moves is different from those involved in trouble-assist type recruitments like (4). Rather than simply making a problem visible in an off-record way, fully nonverbal recruiting moves like (5) involve on-record practices for soliciting a practical action by the recipient.

In my Russian recruitments corpus there are 31 fully nonverbal recruiting moves. This number is high compared to other languages examined in the comparative project (see other chapters in this volume). One reason for this is the relatively high number of cases in which speakers initiate clinking glasses with one another, thus getting the other to drink (see below).

# **3.2 Nonverbal behavior in composite recruiting moves**

Recruiting moves are often composite utterances consisting of both verbal and nonverbal elements. In 87 cases in my corpus, recruiters combine nonverbal and verbal elements in Move A. Nonverbal elements observed in initiating moves are

### 9 Recruiting assistance and collaboration in Russian

of four main types, as found across languages in this volume: pointing, holding out an object, reaching for an object, and iconic gestures (see Chapter 2, §6). In my Russian sample, two more specific subcategories can be identified: holding a glass/cup out for clinking and holding a glass/cup out to receive a drink (see Table 2). The emergence of these categories can be explained by a prevalence of celebratory gatherings in my sample of informal interactions. The category "other" in the table includes recruitments where a speaker places an object on the table for a recipient to take it or refill it.



In this section I illustrate some of the attested nonverbal practices. Extract 6 gives us an example of reaching to receive an object. As Inna asks her husband Fyodor to pass her magnifying glass, she reaches out to receive it.

```
(6) 20110816_Sisters_A_1_332247
```


### Julija Baranova


10 fyo ((retrieves the magnifying glass from his pocket and hands it over to Inna))

Figure 7: Inna reaches with her hand towards Fyodor (Extract 6, line 2).

Figure 8: Fyodor gives the magnifying glass to Inna (Extract 6, line 10).

Inna uses the imperfective imperative verb *davaj*. She complements the verbal component of her recruiting move with a gestural component: stretching out her hand in Fyodor's direction with her palm turned upwards (line 2, Figure 7). Inna

### 9 Recruiting assistance and collaboration in Russian

holds this gesture until she receives her magnifying glass (line 10, Figure 8). Fyodor and Inna coordinate their moves: they bring their hands close to each other. An advantage of gestures such as reaching out is that they can persist through time in a way that a verbal utterance cannot. By holding the gesture after the verbal component of a recruiting move has been spoken, Inna may, for instance, emphasize the urgency of the recruitment and encourage prompt compliance. Another possible function of this gesture is to minimize Fyodor's efforts, as he does not need to bring the magnifying glass all the way to Inna but only meet her hand halfway.

Extract 7 illustrates the use of a pointing gesture in a recruiting move. Maria has just taken a seat on the kitchen bench with her back blocking the view of the video camera. Her daughter Katya alerts Maria to this problematic state of affairs. After Maria fails to respond, Katya makes an explicit request for Maria to change her position at the table and sit on the chair that she is pointing to with her index finger.

```
(7) 20110827_Family_2_437830
  1 mar [Kir padvin'sia ((to the cat))
          name:cat.VOC move.over.IMP.PFV.SG
          Kira, move over
  2 [((sits down on the kitchen bench next to the cat))
  3 kat [( )
  4 (0.4)
  5 kat ja patom k kantsu u nivo zabrala,
         I later towards end from him took.away
         later, towards the end, I took {it} away from him
  6 (1.0)
▶ 7 kat e:ta
         PTC
         well
  8 (0.3)
▶ 9 kat ty naverna [sela v't kak ras
         you.SG probably sat.F DEM just right
         you've probably sat down exactly
  10 [((points at camera))
  11 (0.9)
▶ 12 kat zakrylasia [na stul tuda sadis'
         covered.REFL.F on chair there sit
```
### Julija Baranova

	- 14 (0.6)

Figure 9: Katya points with her index finger (Extract 7, line 13).

In the initiating move, Katya produces a statement: 'well (0.3) you've probably sat down exactly' (lines 7–9). Intonationally and informationally, this statement sounds unfinished. The accompanying point to the camera (line 10), however, completes the trouble statement and clarifies its import. When Maria fails to respond, Katya adds more information about the problem together with an explicit recruiting component: '{it} got obscured, sit on the chair there' (line 12). This turn further explicates the problem and offers a solution for it. At the same time, Katya now points to the chair with her index finger (line 13, Figure 9). In response, Maria partially complies. Instead of taking a seat on the pointed-to chair, she shifts on the bench, partially uncovering the view of the camera.

In the next example, the recruiting move contains two relevant nonverbal elements: holding out an object and pointing (in this case, a head point). The example is from an interaction between Sasha, the host, and Ksenia, a friend. Earlier in the interaction, Sasha presented Ksenia with one of her own photographs as a gift. Here, Ksenia asks Sasha to put the photograph on top of her jacket.


### 9 Recruiting assistance and collaboration in Russian

```
2 ((holds out photograph))
3 [tam ((head point))
        there
        there
```
### ▷ 4 sas ((takes the photograph and leaves))

Figure 10: Ksenia holds out the photograph; Sasha reaches out to take it (Extract 8).

Ksenia instructs Sasha to put the photograph on top of Ksenia's jacket, which is in the corridor. During the production of the request, she holds the photograph and stretches her arm in Sasha's direction, gazing at her. Subsequently, she verbally refers to the place where the photograph should be put: *tam* 'there'. At the same time, she head-points in the direction of the corridor. Sasha takes the picture and leaves the room.

Ksenia and Sasha's hand gestures in (8) are similar to those we saw in (6). In that example, Inna's reaching for the magnifying glass facilitated Fyodor's action of handing it to her. Likewise, in (8), Ksenia's holding out the photograph makes it easier for Sasha to take it. By easier I mean that the distance Sasha's hand has to travel to take the object is shorter. Another example of holding-out gesture can be found in (5), where the speaker holds out his cup for the recipient to take it and fill it with tea.

In this section I have described some nonverbal elements observed in recruiting moves, including pointing gestures, holding out objects, and reaching for objects. Nonverbal elements not only complement recruiting turns with relevant information, they also pursue and facilitate compliance.

### Julija Baranova

# **3.3 Verbal elements: construction types and subtypes**

The Russian language provides its speakers with a multitude of linguistic formats to initiate recruitments. The relative frequency of imperatives, declaratives, interrogatives, and no-predicate constructions are shown in Table 3. Chapter 2, §6 explains in more detail how recruiting moves were coded for verbal elements. In this section, I discuss the Russian imperative, declarative, and interrogative forms. For an example of a no-predicate construction I refer the reader to (1) above: 'some milk' is a minimally designed recruiting turn only containing a reference to the object being requested.

Table 3: Sentence type of recruiting moves including spoken elements (*n*=159). For non-minimal recruitment sequences, only the first attempt is counted in this table.


I have analyzed and coded the linguistic structure of recruiting turns based on their sentence type only, without reference to their intonational contour. Note that, in Russian, imperative utterances can have rising-falling intonation. I discuss this issue in the following section on imperatives.

### **3.3.1 Imperatives**

Imperatives are the most frequent format of recruiting moves in my sample of informal Russian interactions. Imperatively formatted moves have recently been identified as a default strategy in another corpus of informal Russian interaction (Bolden 2017). While research on Italian and English has shown that imperatives are typically used in the context of ongoing joint projects in which the recruitee's commitment has been secured (e.g. Rossi 2012; 2015; Zinken 2016), Russian imperatives appear to be used in a broader range of recruitment contexts. Bolden (2017) shows that Russian imperatives are used in contexts where Italian and English speakers would normally opt for interrogatives. These are contexts where the i) recipient is not yet committed to the project of which the recruitment is part, ii) contexts where compliance requires the recipient to abandon or alter his/her

### 9 Recruiting assistance and collaboration in Russian

own ongoing activity, and iii) contexts where compliance involves a relatively elaborate physical activity.

Aspect is important in the Russian verb system. As mentioned in the introduction to this chapter, Russian verbs can be perfective or imperfective. It is not yet entirely clear which recruitment contexts prefer which aspectual type, but it appears that the imperfective is more often used when compliance can be expected (Benacchio 2002; cf. Zinken, Chapter 8, §3.3.1 on the usage of perfective and imperfective imperatives in Polish).

Imperatives can be singular or plural. Singular imperatives are used when there is just one addressee. Additionally, imperatives may be of the perfective aspectual type (see Extracts 2, 3, and 8) or the imperfective aspectual type (see Extract 6). Perfective and imperfective imperatives have a plural form when there are multiple recipients or when the second person polite plural pronoun *vy* is used. The latter situation is illustrated in the next example. The recruitment sequence in (9) comes from a conversation between school custodians who are having lunch in the staff room. Alifa is about to join her colleagues at the table who are already having soup.

(9) 201220120\_colleagues\_casual\_2\_498040

```
1 ali ((takes a bowl from the cabinet and puts it on the table next to Anna))
  2 ann [(losh)ku ( )
           spoon-ACC
           a spoon ( )
▶ 3 ali [Anna- Anna Batkiyevna,
           name name patronymic
           Anna- Anna the daughter of a father
  4 ann aye:
          INTJ
          hey
  5 (0.3)
▶ 6 ali pazhalsta nakla:dyvajte mneh
          please put.IMP.IMPFV.PL I.DAT
          you may do {some} serving for me please
  7 (0.4)
  8 ver khahahm[hmhm ((laughter))
  9 (mar) [(Ret')kiyevna
                   patronymic
                  daughter of the (radish)
▷ 10 ann [((puts her loaf of bread on the table))
  11 ? [h.hehehe
```
Julija Baranova

12 (0.5)

▷ 13 ann ((takes Alifa's bowl from the table))


16 [eta to:chna it exactly exactly

Figure 11: Alifa makes a request for Anna to serve her some soup (Extract 9).

This recruitment sequence is rather complex. Here I focus only on the format of the imperative Alifa uses to recruit Anna's assistance. Alifa asks Anna to serve her some soup starting in line 3, where she draws Anna's attention with a non-serious patronymic: 'Anna- Anna the daughter of a father'. A patronymic is formed by adding a suffix to the father's first name. Ivan's daughter, for instance, would have *Ivanovna* as her patronymic. The use of a patronymic goes hand in hand with the use of plural 'you' as a polite form of address. The formation of a patronymic using the generic 'father', as Alifa does here, is occasionally used as a joking patronymic when the real one is not known. Alifa continues with 'you may do {some} serving for me please'. Her use of the imperfective imperative with the plural ending *-te* makes it clear that she is addressing the recipient with the plural 'you'.

Another example of an imperative with a plural ending is given in (10). This time, the imperative is of the perfective type and is directed to multiple recipients. 9 Recruiting assistance and collaboration in Russian

Pavel is about to go to the village brook with a family guest. The host of the gathering, Inna, is trying to persuade her son Dennis and grandson Kostia to join them.

(10) Family\_dinner\_Country\_A\_2\_1038680

```
1 lid na::: e:: ↑na ruchej pajdiote?
         to to brook.ACC go.FUT.2PL
         to::: uhm:: are you guys going to the brook?
  2 pav ((head nod))
▶ 3 inn aha, eh [Dennis ajda-te [sxadite Kostia,
         uhuh name PTC.IMP-PL go.IMP.PFV.2PL name
         yeah, eh, Dennis come on you guys, go ((plural)) Kostia
  4 [((touches Dennis)) [((reaches for Kostia's arm))
  5 kos [((turns his head and looks at Inna))
▷ 6 pajdiom ((looks at Dennis))
         go-FUT-1PL
         shall we go?
```
<sup>▷</sup> 7 den ((bows his head to the side))

Figure 12: Inna touches Dennis' arm as she recruits him and Kostia to go to the brook (Extract 10, lines 3–4).

In line 1, Lida is talking to her husband Pavel, who is standing outside the window. She is asking whether Pavel is indeed going to take their guest to the village brook. After Pavel's confirmation at line 2, Inna initiates a recruitment involving Dennis and Kostia: *aha, eh Dennis ajda-te sxadite Kostia* 'yeah, eh, Dennis come on you guys, go ((plural)) Kostia'. As in (9), the imperative *sxadite* 'go' has

### Julija Baranova

Figure 13: Inna reaches for Kostia's arm as she recruits him and Dennis to go to the brook (Extract 10, lines 3–4).

the plural ending *-te*. While Kostia indirectly agrees to go to the brook by inviting Dennis at line 6 ('shall we go?'), Dennis seems reluctant (line 7). His lack of response is noticeable. In the interaction subsequent to the extract, Dennis provides several reasons why he cannot go, while Inna insists on his participation. Dennis and Kostia do end up joining Pavel and the family guest on their walk to the brook.

Occasionally, Russian imperatives are combined with particles. For instance, the imperative in (3) contains the sentence-initial particle *nu* (*nu Marish, pusti: ejo, ja pa- pae:m spako:jna* 'sweet Marina, let her go {so that} I finish eating in peace'). This particle has multiple functions, one of which is to convey insistence on carrying out the requested action (Bolden 2016; 2017). Imperatives can also sometimes be preceded by the particle *na*. This particle is also found as a stand-alone directive. In spoken Russian it conveys the meaning of 'take' or 'here you are', as illustrated in (11). In this example, Fyodor and Inna are visiting their daughter Nadya and Nadya's young daughter. Inna is entertaining the toddler.

```
(11) Granddaughter_605308
```

```
▶ 1 nad na daj yej von vazachku ana budit sidet' s nej ((to Inna))
           PTC give her PTC vas.DIM she will be sit.INF with her
           take {this}, give her this little basket, she'll be sitting with it
           ((puts the basket with sweets in toddler's hands))
```

```
2 (0.3)
```
▶ 3 nad na. ((to toddler)) PTC take / here you are

9 Recruiting assistance and collaboration in Russian

	- 5 inn ((returns the basket to the table))

Figure 14: Nadya lifts the basket from the table to give it to the child (Extract 11, line 1).

Nadya's first recruiting turn (line 1) is directed at Inna.The turn begins with the particle *na*, which has the meaning of 'take' in this case. The particle is followed by the verb *daj* 'give'. This results in a complex recruiting expression that combines the actions of 'taking' and 'giving'. Nadya's second recruiting turn (lines 3–4) is directed at the toddler and again begins with the particle *na* and combines multiple actions: 'take', 'go', 'treat'. She says: 'take, go, for grandmother (.) treat grandmother with sweets'.

Another imperative type makes use of a double-verb construction, where the first verb has a frozen imperative form and the second denotes the required action (see Zinken 2013 and Chapter 8, §3.3.2 on a comparable construction in Polish). The construction in Russian combines the verb 'give' with a relevant action verb. Its use is shown in (12). Participants in this interaction are friends who have gathered at Ksenia's apartment for dinner and drinks. Ksenia's elderly mother enters the room, where people are seated, and addresses a request to Ksenia.

(12) 20110813\_School\_Friends\_2\_618255a

```
1 mom Ksiush [pirashki-
        name.VOC pastries
        Ksenia, pastry
```
Julija Baranova

```
2 kse [Vo:f
             name.VOC
             Vova
▶ 3 mom ty davaj vyzyvaj etaj (.) gazafshi:tsu [nu-,
       you give.IMP call.IMP.IMPFV DEM.F gas.worker PTC
       go ahead call the (.) gas worker
▷ 4 kse [nu xva:tit
                                          PTC enough
                                         enough {already}
▷ 5 [((waves))
 6 mom kak su:xa ta
       how dry PTC
       but how dry
```
Figure 15: Mom requests that Ksenia make a call (Extract 12, line 3).

Mom's request (*ty davaj vyzyvaj etaj (.) gazafshi:tsu* 'go ahead call the (.) gas worker') consists of the frozen imperative *davaj* combined with the imperative verb expressing the requested action, here *vyzyvaj* 'call'. In this context, where no object transfer is involved, *davaj* loses its independent meaning of 'giving' and takes on the meaning of 'come on' or 'go ahead'. Mom's request to call gas services is met with resistance by her daughter Ksenia, who says 'enough {already}' and literally waves the recruitment away (lines 4–5). After this response, Mom supports her recruitment with a reason ('but how dry'). As it becomes clear from the unfolding conversation, she is referring to the pastry that was made earlier in the day which turned out dry due to presumed problems with the gas (see also line 1).

For interrogatives, Russian mainly relies on intonation along with the use of in-situ question words and particles. An imperative construction can be given a

### 9 Recruiting assistance and collaboration in Russian

question-like quality by means of intonation, which can be applied to any relevant turn-constructional unit. This leads to a hybrid recruiting format containing both imperative and interrogative features. Such imperatives with interrogative features can be found in (2) and (8). The recruiting turn in (2) has rising-falling intonation on the word 'give': *↑daj lo:shku druguju pazhalusta* '↑give me another spoon please'. In (8), the interrogation is done with the final particle *a?* uttered with rising intonation: *Sash, palazhi mne na ku:rtku a?* 'Sasha, put {it} on top of my jacket eh?'. These cases can be contrasted with imperatives containing no interrogative features in (3), (6), (7), (11), (12).

So, even though imperatives in Russian are used in a wider range of contexts than, for instance, in Italian and English, the Russian system of imperatives shows greater diversity, involving aspectual pairs (imperfective and perfective), distinction in number (singular and plural), the use of interrogative features, and diminutives particles on the verb (see §3.4.1).

### **3.3.2 Interrogatives**

Although imperatives can be considered a default way of recruiting assistance and collaboration in Russian, we also find recruiting turns that are interrogatively formatted. In the next example, Ksenia is visiting her friend Sasha. She asks whether Sasha will let out a guest who is already in the corridor and about to leave her place.

(13) 20110826\_Old\_friends\_A\_2\_66555

```
1 kse ((points towards the corridor))
2 (0.3)
3 kse pra↑vodish=
       let_out.FUT.2SG
       will you let {her} out?
4 sas =uhum,
        uhum
5 ((leaves the room to let the guest out))
```
First, Ksenia points to the corridor (line 1). When there is no response, she asks 'will you let {her} out?'. This recruiting turn has an interrogative format. Similarly formatted recruiting turns in English ('would/will you x?') tend to occur when where are perceived contingencies or obstacles to compliance, and where the recruiter has a low degree of entitlement to make the request (Curl & Drew 2008). In this case, Ksenia's entitlement is an issue. By asking whether Sasha will let her

### Julija Baranova

Figure 16: Ksenia points towards the corridor (Extract 13, line 1).

guest out implies that Sasha has failed in her responsibilities as a host. Ksenia, too, is a guest here and her entitlement to initiate such a recruitment is arguably low.

In another type of interrogative strategy, *wh*-questions can also be used to recruit assistance. In (14), several girlfriends are looking at Sasha's photographs. Ksenia is curious about the photographs that Sasha and Lida are talking about.

### (14) Old\_friends\_B\_1\_302784


Sasha and Lida are talking about one of Sasha's photographs, which Lida is holding in her hands. Already at line 1, Ksenia expresses her interest in the pictures by asking: 'what's there?', which can be seen as the initiation of the recruitment sequence. Her interest becomes even clearer when Ksenia uses the interrogative 'which one' at line 4, together with the imperative 'show {me}'. Ksenia is entering into someone else's ongoing conversation. This justifies the use of the

### 9 Recruiting assistance and collaboration in Russian

Figure 17: Lida shows the photograph she is holding to Ksenia (Extract 14, line 5).

interrogative construction. The recruitment is immediately fulfilled as Lida turns the photograph around so Ksenia can see it.

### **3.3.3 Declaratives**

We have seen that sometimes participants in interaction notice that someone is in need of help and they offer practical assistance without being explicitly asked to do so (see Extract 4). In other cases, the need for help might not be that apparent and the nature of the trouble needs to be verbalized. This is demonstrated by the following extract from an interaction between family members who gathered at Nadya's place. Inna is holding Nadya's toddler on her lap when Nadya initiates a recruitment.

```
(15) Family_dinner_B_2_649099
```

```
▶ 1 nad ma:m ana salfetku von zhujot
          mama.VOC she napkin.ACC there chew.3SG
          Mom, she's chewing on the napkin
  2 inn ((leans her head towards the toddler))
  3 e: ((to the toddler))
          INTJ
          hey
  4 (0.4)
  5 inn e:
          INTJ
          hey
  6 ((removes the napkin from toddler's hands))
```
### Julija Baranova

Figure 18: Nadya tells Inna than her toddler is chewing on a napkin (Extract 15, line 1).

Instead of instructing Inna to remove the napkin from her toddler's mouth, Nadya simply describes the problem that needs to be addressed (see also Kendrick, Chapter 4, §4.2.3; Rossi, Chapter 5, §3.3.4; Enfield, Chapter 6, §4.3.1; Dingemanse, Chapter 10, §3.2.2). Inna responds by leaning towards the child and removing the napkin. Nadya's request is indirect as it does not explicitly ask for any assistance and does not specify the practical action required from the recipient. Inna, however, acts quickly and removes the napkin from the toddler.

Recruitments of the trouble-assist kind (§2.3) and statements of the kind shown in (15) are similar in certain respects. Seeing or being alerted to a source of trouble and being able to act upon it seems sufficient for the recruitee to step in and solve the problem. Note that Nadya is facing the child and has better visual access to the toddler's behavior than Inna does (see also Rossi 2018). On the other hand, Nina is in a better position to solve the problem because she is the one closest to the toddler.

Another way of conveying that some action is required is to state that it 'needs' to happen (see also Floyd, Chapter 3, §3.3.4; Rossi, Chapter 5, §3.3.4; Zinken, Chapter 8, §3.3.2). At a memorial dinner, with the entire family present, Pavel's daughter Lena asks whether it is necessary to eat the rice porridge. In what follows, Pavel tries to convince Lena to eat the porridge that is traditionally consumed at memorials.

9 Recruiting assistance and collaboration in Russian

```
(16) memorial_1_424599
  1 len a chio mnoga kashi nada sjest' -ta?
         PTC what a.lot porridge need.MOD eat.INF-PTC
         does one need to eat a lot of porridge?
  2 pav ne:t
         n:o
▶ 3 Le:n. (.) hm (.) nada abiza:til'na=
         name.VOC need.MOD necessarily
         Lena (.) hm (.) one necessarily needs
▶ 4 =etu vot kashu sjest'
          DEM PTC porrige eat.INF.PFV
          to eat this porridge
  5 (0.7)
▶ 6 pav [lozhichku
          spoon.DIM.ACC
          a little spoon
  7 [((scoops some rice with a spoon))
  8 ((brings the spoon to Lena's [plate))
  9 [lo:zhichku fsio ravno nada
```
spoon.DIM.ACC anyway need.MOD a little spoon is still necessary

Figure 19: Pavel serves his daughter Lena some porridge (Extract 16).

Rice porridge is traditionally served at Russian memorial dinners. When Pavel's daughter expresses her reluctance to eat it, he tries to persuade her to do so. He uses an impersonal declarative construction: *nada* 'one needs' (line 3). This construction is similar to the Polish *trzeba* (Zinken & Ogiermann 2011). Pavel combines the impersonal declarative with the address term *Lena*. The use of this

### Julija Baranova

address term is pragmatically marked because it is already clear who the addressee is: Pavel is responding to Lena's question. At the same time, the impersonal declarative expresses a general requirement to eat the porridge. Pavel even uses the adverb *nada abizatil'na* 'necessarily', which strengthens the statement. He also takes the liberty of serving his daughter some porridge without securing her acceptance (lines 7–8). This is another piece of evidence that Pavel considers eating rice porridge to be an obligation in this context, regardless of a person's own wishes. Pavel does, however, orient to the girl's reluctance to eat the porridge by using the diminutive *lozhichku* 'a little spoon' (lines 6 and 9). This makes it clear that, although not eating the porridge is out of the question, it would be sufficient to eat only a little bit.

To summarize this whole section (§3.3), we have seen that Russian recruiting turns come in four main linguistic formats. Imperatives are the most widely used format, followed by no-predicate constructions, declaratives, and interrogatives. Russian imperatives come in aspectual pairs: perfective and imperfective. They also have singular and plural forms. Declarative recruiting turns can be compared to recruitments of the trouble-assist type: declaratives often verbalize a trouble; trouble-assist recruitments do not involve language but feature trouble that is visible in the context.

# **3.4 Additional verbal elements**

The core elements of a linguistically-formulated recruiting move (i.e. a predicate with its core arguments) can be complemented by additional verbal elements, among which are vocatives (see Extracts 3, 8, 9, 10, 15, 16 for examples), benefactive markers (e.g. 'you may do {some} serving for me please', Extract 9), reasons (see Extract 3 above and §3.4.2 below), and mitigators. In the next subsections, I focus on additional verbal elements that mitigate the recruiting move and elements that explain it.

# **3.4.1 Mitigators**

Recruitments always involve some degree of imposition on recipients. Because of the potential threat to "face" and to the social relationships at hand, recruiters sometimes use strategies to mitigate the potential imposition of a recruitment (Brown & Levinson 1987). Here I illustrate several strategies.

The following extract features two types of mitigation: one is marked on the noun and the other on the imperative form of the verb. Vladimir and his wife 9 Recruiting assistance and collaboration in Russian

Julia have their family over for food and drinks. At line 2 Vladimir produces a recruitment directed at Julia: *daj-ka riumki nam* 'give us {some} glasses'.

(17) cooking\_3\_226998 1 (1.9) ▶ 2 vla daj -ka [riumki nam give-PTC glasses.ACC we.DAT give us {some} glasses 3 jul [((looks at Vladimir)) 4 (0.3) ▶ 5 vla [s pamidorchikam with tomato with a little tomato ▷ 6 jul [((opens the kitchen cabinet and gets several glasses))

Figure 20: Vladimir recruits Julia (Extract 17).

The imperative *daj* 'give' is accompanied by the diminutive particle *-ka* which makes the action sound more casual and low-cost. Possibly in response to the absence of an immediate response by Julia in line 4 – the next transition-relevance place (Clayman 2013) – Vladimir increments his recruiting turn by adding *s pamidorchikam* 'with a little tomato'. The diminutive *-chik-*, added to the basic form *pamidoram*, formally attenuates Julia's effort in serving the vegetable (Ogiermann 2009).

We saw this strategy in (16) as well, where Pavel is persuading his daughter to eat rice porridge: *lo:zhichku fsio ravno nada* 'a little spoon is still necessary'.

### Julija Baranova

The diminutive *-ich-*, added to the basic form *lozhku*, minimizes the effort that his daughter would have to make in order to comply.

Recruiters can also use a diminutive address term as an expression of their affection for the recipient. This may serve as a way of downplaying the recruiter's desire to impose. We saw this in (3), where Marina is visiting her mother-in-law Anna. In Anna's recruiting turn *nu Marish, pusti: ejo, ja pa- pae:m spako:jna* 'sweet Marina, let her go {so that} I finish eating in peace', the name Marina is rendered in the diminutive form *Marisha*. In addition to the affectionate vocative, Anna provides a reason for the recruitment.This is another mitigating device, discussed in the next section.

### **3.4.2 Reasons**

Complementing a recruiting turn component with a clause that offers a reason for the recruitment goes beyond mere mitigation (e.g. Waring 2007; Parry 2013). The current sample includes 21 recruitment sequences in which the recruiter gives a reason in support of the recruitment. Reasons for informal recruitments in the Russian sample deal with i) informationally underspecified recruitments, ii) delicate recruitments, and iii) recruitments involving actions beyond requesting, for instance joking and complaining (Baranova & Dingemanse 2016). I now give examples of reasons supplied in the contexts of an underspecified recruitment and a delicate recruitment.

Extract 18 shows a recruitment sequence where the reason provided by the recruiter adds information that is crucial for compliance. Several family members are having dinner together on the porch of a country house. One of them, Julija, has gone outside to take some photos. She is a guest visiting from abroad. Julija's uncle, Pavel, was sleeping when Julija left the table. So, at the beginning of this extract, Pavel is unlikely to be aware of Julija's whereabouts.

```
(18) Family_dinner_Country_A2_876874
```

```
1 ((Pavel joins the others at the table after being outside))
  2 pav dozhdik zamarasil [u vas
         rain.DIM drizzle.PST.PFV with you.PL
         it has started drizzling in your area
▶ 3 lid [pasmatri,
                                look.IMP.PFV.2SG
                                take a look
  4 vyjdi iz-za: ako:li-
         go_out.IMP from fen-
         go out behind the fen-
```
### 9 Recruiting assistance and collaboration in Russian

Figure 21: Pavel has just put his jacket on to go outside (Extract 18).

As Pavel returns (line 1) and while he is commenting on the weather (line 2), Lida addresses a recruiting turn to him: 'take a look go out behind the fen-', instructing Pavel to go outside, which is in conflict with his observation that it is raining. Up until this point, Lida's recruiting turn is also lacking information on what Pavel should do once he is outside. However, Lida goes on to provide a reason that deals with these issues: 'uhm Julija went to take pictures I think'. This reason refers to Julija who is out in the rain. She is a family member who is visiting from abroad. The family does not see her often in person, which makes her an honored guest. Pavel agrees to comply by saying *shias (pajdu)* 'in a bit (I'll go)' (line 7). After getting his jacket, he leaves the porch to find Julija outside.

### Julija Baranova

The previously discussed Extract 3 also contains a reason: *'sweet Marina, let her go {so that} I finish eating in peace'*, which deals with the delicacy of the requested action. The reason implies that finishing dinner in peace is incompatible with the presence of the dog at the table. Marina orients to this negative implication with a counter-reason 'but we aren't doing anything to you'.

This section (§3.4) has discussed additional verbal elements used to mitigate and explain recruitments in the Russian sample. Russian vocatives, imperatives, and nouns can be formatted with diminutives and minimizers. Recruiters can also increase the chances of compliance by providing a reason. Reasons supply information necessary for fulfillment, or explain a recruitment that is otherwise unclear, delicate, or imposing.

# **4 Formats in Move B: The responding move**

After having discussed recruiting moves, I now analyze the types of response they receive. Table 4 shows the distribution of the main types of response relative to the format of the recruiting turn.<sup>1</sup>


Table 4: Types of response by format of the recruiting turn (*n*=159).

Most recruitments in the Russian sample are fulfilled. Since fulfillment involves some practical action, most responses are nonverbal: 145 recruiting moves received a fully nonverbal response.<sup>2</sup> In three cases, the recipient's response was not visible or hearable. In the remaining 52 recruitments, the response involved

<sup>1</sup>Nonverbal recruiting moves and recruitments of the trouble-assist type are not included in this table. Response categorized as "other" do not fit any of the main types or are unclear.

<sup>2</sup> In 39 cases in which the response was analyzed as fully nonverbal, the recruitee does say something while fulfilling the recruitment nonverbally but what they say is unrelated to the recruitment sequence.

### 9 Recruiting assistance and collaboration in Russian

a relevant verbal element. Such verbal responses can co-occur with fulfillment; at other times, they are indicative of rejection or delay in compliance. In what follows, I illustrate several response types in recruitment sequences.

# **4.1 Compliance**

Compliance with a recruitment is usually evident from recipients' nonverbal behavior as they proceed to give an object to the recruiter, perform a service, or cease/alter their ongoing behavior (see Extracts 1, 4, 6, 8, and 9, and 13). Occasionally, verbal elements complement such nonverbal responses.

Sometimes the verbal element expresses the recipient's commitment to fulfill the recruitment prior to the actual fulfillment. Consider again (13), where Ksenia and Sasha were in the kitchen and Lena asked Sasha to let a guest out of her place. Sasha first responded to the recruiting move with the confirming expression *uhum* and then displayed behavior consistent with this: she left the kitchen and went to the corridor to let the guest out. In this case, leaving the kitchen may not in itself be a clear indication of compliance since her behavior is not visible once she leaves the kitchen. So Sasha's confirmation helps to convey to the recruiter Ksenia that she will indeed comply.

We saw a similar response to Lida's request 'take a look go out behind the fen-' in (18). Here the recruitee Pavel displays his commitment to fulfilling the recruitment with *shias (pajdu)*, literally 'now (I'll go)' (line 7). The Russian word *shias* is used to indicate unstable and still changeable time in the present and, in combination with a future/imperfective inflection of the verb, it can also convey the meaning of 'in a bit'. Basically, Pavel indicates that he will comply in the very near future. Instead of immediately going outside, he goes into the house to get his jacket. The verbal element *sejchas (pajdu)* was necessary to make it clear that he would carry out the recruitment when his actual behavior could have been interpreted as non-compliance.

# **4.2 Non-compliance**

In the Russian sample, rejecting a recruitment was never done with a simple 'no' response. Rejection was usually achieved by counter-proposals and the giving of reasons why compliance would not take place.

Consider again (3), where Anna asks Marina to take the dog away from the table so that Anna can 'finish eating in peace'. Marina does not comply but instead gives a counter-reason, i.e. a reason not to comply: 'but we're not doing anything to you'. Marina keeps holding the dog on her lap for another five minutes.

### Julija Baranova

Another, even more indirect non-compliance strategy is to ignore the recruiting move, to give no response to it either way (see also Blythe, Chapter 7, §4.2.4). This is illustrated in the next example, where several school custodians gathered for lunch in their staff room. Vera makes a request of Anna, while Lena and Marina are involved in an unrelated conversation.

```
(19) 20120120_colleagues_casual_2_339070
```

```
1 ((Anna is standing in front of the open cabinet))
▶ 2 ver [grenki tam eshio [dastan'
          breadsticks there also take_out.IMP.PFV.SG
          also get the bread sticks out
  3 len [a?
                                 INTJ
                                 huh?
  4 mar ni uexala eshio?
         NEG go_away.PST.PFV.3SG yet
         hasn't she left yet?
  5 len kto?
         who
         who?
  6 (0.7)
  7 ann ((opens up the cabinet))
  8 mar ((finger [points towards the wall behind her))
▶ 9 ver [Sasha padi tozhe pajest=
                  name probably also eat.FUT.3SG
                  Sasha will probably also have {some}
  10 len =dir↑ektar
          director
          the director?
  11 (0.4)
  12 len [.hhhh ana- ushla v- ushli ani s Lugaevaj
               she went_away.F in/to went_away.PL they with surname
          .hhh she left to-, she left with Lugaeva
```

```
13 ann [((looks into the cabinet and takes out two bowls))
```
At line 2 Vera makes a request of Anna: 'also get the bread sticks out'. Anna is standing in front of the cabinet, where the breadsticks are, ready to take something out of it. So she is the most suitable person to get the breadsticks. However, she does not respond to Vera's request and seems to be looking in Marina's direction instead. Vera pursues her request at line 9 by offering a reason for it: 'Sasha will probably also have {some}'. After this, Anna still shows no signs of compliance, but Vera does not pursue the request further. It may be that Anna has

### 9 Recruiting assistance and collaboration in Russian

Figure 22: Vera makes a request for Anna to get breadsticks out of the cabinet (Extract 19).

simply failed to notice or register the ongoing recruitment. But in this case, the recruiter does not treat the absence of a response as a problem of hearing or registering the recruitment, as she would do by repeating the request, for instance. Instead, Vera offers a reason. There is still no response from Anna and the recruitment remains unfulfilled. Instead of getting the breadsticks, Anna takes two bowls out of the cabinet (line 13).

A recipient may initiate repair in response to a recruiting move, as in (2), where Katya asked Maria to give her another spoon. Maria's initial response was a repair initiation: 'a spoon? (.) another one?'. It seems that the repair initiation is used here not only to indicate a problem of hearing, but also as a challenge to Katya's request, questioning whether the recruitment is necessary. Katya's response – explaining that the spoon she has is dirty – targets both potential readings of Maria's repair initiation.

In sum, most recruitments in the Russian sample were fulfilled. Often, fulfillment involved nothing more than performing the relevant practical action. Sometimes the responding move featured a verbal component, for example conveying the recipient's commitment to comply, before fulfillment actually occurred. In non-compliance, verbal responses included giving reasons not to comply and initiating repair.

### Julija Baranova

# **5 Acknowledgment in third position**

A recruitment sequence is a paired sequence consisting of a recruiting move and a responding move. This sequence can potentially be extended with a "sequenceclosing third" (Schegloff 2007) in the form of an acknowledgment by the recruiter of the assistance provided, such as 'thank you'. In only three out of 176 fulfilled recruitments (2%) did the recruiter produce some kind of acknowledgment of the recruitee's efforts. On one occasion, the acknowledgment was the confirmative *uhum* and the other two involved the conventional expression of gratitude *spasibo* 'thanks' (see Extract 2). In informal interaction, it seems that expressing gratitude for another's assistance or collaboration is a universally rare practice, with a minority of languages (including English and Italian) showing slightly greater frequency of occurrence (Floyd et al. 2018).

# **6 Social asymmetries**

The phenomenon of recruitment is sensitive to social asymmetries between recruiters and recruitees (Brown & Levinson 1987). Such asymmetries may be referenced through the format of the recruiting move. So recruiting assistance or collaboration in the workplace may be done differently than in an informal setting (Garvey 1975; Corsaro 1977; Dixon 2015; Takada & Endo 2015). Also, recruitments involving children are known to have different features from the ones that only involve adults (Drew & Heritage 1992). Child-directed recruitments were not included in this study.

Relative social status is difficult to operationalize. In this study, I took indicators of relative social status to be participants' ages, the kind of relationship between them, and the way they address each other in the recording. In Russian culture, older people tend to be accorded a higher status. This is often expressed in the way they are addressed. Normally, a combination of their first name and their patronymic is used. Also, the use of the plural 'you' *vy* is preferred over the singular *ty*. However, if the older person is a close family relation, he or she can be addressed with the singular *ty* and the corresponding kin term, such as 'grandmother', sometimes in combination with the person's first name, e.g. 'aunt Olga'. When there was no apparent age difference between the recruiter and recruitee, I looked at the relationship between them. Friends, spouses, siblings, and in-laws were considered to be of equal status when the participants were of an approximately same age. Such relationships are characterized by the use of the singular 'you' *ty* and only first names when referring to or addressing each other.

### 9 Recruiting assistance and collaboration in Russian

Based on these criteria, the relationship in each recruiter-recruitee dyad was categorized according to the three types defined in the project's coding scheme (see Chapter 2, §6), namely, i) participants of equal status (A=B); ii) recruiter has higher status than recruitee (A>B); and iii) recruiter has lower status (A<B). As Table 5 shows, the majority of recruitment sequences in the Russian sample (61%) involved participants that were in a socially symmetrical relationship. In 35% of cases the relationship between the participants was considered to be asymmetrical based on the above criteria.

Table 5: Relative frequencies of dyads by type of social (a)symmetry (*n*=200).


Social status did not have a straightforward effect on the type of response that recipients produced. Fulfillment, rejection, and non-response rates were relatively equally distributed across the (a)symmetry types. The main finding here is that, if there are asymmetries, recruitment is more likely to be initiated in a downward direction. Recruitments initiated from a lower-status position were relatively rare. This is in line with Brown and Levinson's (1987) theory of politeness: if the imposition is in an upward direction (A<B), then there is a greater "threat to face" than in other kinds of case (A>B, A=B). When the threat to face is higher, potential recruiters would be more likely not to carry out the facethreatening act at all and instead to perform the desired action themselves, or to mobilize someone of a lower or equal status for the task (see also Floyd, Chapter 3, §6; Enfield, Chapter 6, §6; Dingemanse, Chapter 10, §5.2).

# **7 Conclusion**

This chapter has presented an overview of recruitment practices in conversational Russian. Imperatives are the most frequent linguistic format of recruiting moves. This is in line with Bolden's (2017) conclusion that Russian imperatives are used in a wider range of recruitment contexts compared to languages such as English and Italian. Moreover, Russian imperatives form a diverse set. First,

### Julija Baranova

they feature an aspectual distinction between perfective and imperfective. Second, imperatively formatted recruiting turns can be produced with interrogative features and diminutive particles.

Similar to Italian and Polish, declarative recruiting turns in the Russian sample may be designed with an impersonal predicate *nada*, which can be translated into English as 'one needs to' or 'it is necessary to'. In this manner, the speaker can frame the recruitment in terms of shared responsibilities that hold for the recipient, but also for the speaker and perhaps the entire community (see also Zinken & Ogiermann 2011; Zinken 2016: chap. 6).

Russian has a rich diminutive morphology. Diminutive vocatives were observed in recruiting turns, expressing speakers' affection for recipients, and may orient to the relationship between recruiter and recruitee. Diminutive nouns and particles can also be used to minimize the perceived imposition on recipients (see also Bolden 2017).

In responding moves, overt rejections are dispreferred and nonverbal fulfillment is the most frequent response. Rejection is usually done by means of reasons and counter-proposals. Overt refusals to comply were not found in the sample. It is not clear from our sample whether social asymmetry affects the way a recruiting move is formatted. Future research on this question may offer further insights into how recruitees are selected and how recruiting turns are designed.

# **References**


9 Recruiting assistance and collaboration in Russian

opening Russian conversations. *Human Communication Research* 34(1). 99–136. DOI:10.1111/j..2007.00315.x


### Julija Baranova


9 Recruiting assistance and collaboration in Russian


# **Chapter 10**

# **Recruiting assistance and collaboration: A West-African corpus study**

# Mark Dingemanse

Centre for Language Studies, Radboud University

Doing things for and with others is one of the foundations of human social life. This chapter studies a systematic collection of 207 recruitments of assistance and collaboration from a video corpus of everyday conversations in Siwu, a Kwa language of Ghana. A range of social action formats and semiotic resources reveals how language is adapted to the interactional challenges posed by recruitment. While many of the formats bear a language-specific signature, their sequential and interactional properties show important commonalities across languages. Two tentative findings are put forward for further cross-linguistic examination: a "rule of three" that may play a role in the organization of successive response pursuits, and a striking commonality in animal-oriented recruitments across languages that may be explained by convergent cultural evolution. The Siwu recruitment system emerges as one instance of a sophisticated machinery for organizing collaborative action that transcends language and culture.

# **1 Introduction**

Doing things for and with others is one of the foundations of human social life. The question of how we recruit assistance and collaboration has venerable roots in ethnography (Malinowski 1923; Frake 1964) and in the philosophical study of speech acts (Austin 1962; Searle 1969). Yet it has only recently become possible to address it more systematically using records of actual behavior in conversation (Drew & Couper-Kuhlen 2014). Here I study one of the most concrete forms of prosociality in everyday social interaction: recruitments, when someone gets another to carry out a practical action for or with them. Examining the interactional

Mark Dingemanse. Recruiting assistance and collaboration: A West-African corpus study. In Simeon Floyd, Giovanni Rossi & N. J. Enfield (eds.), *Getting others to do things: A pragmatic typology of recruitments*, 369–421. Berlin: Language Science Press. DOI:10.5281/zenodo.4018388

### Mark Dingemanse

practices by which people come to do things for and with each other contributes to our understanding of the role of language in human sociality.

Much prior work on requesting in social interaction has focused on how requests are shaped by participants' claims of entitlement (Heinemann 2006; Curl & Drew 2008) or how formats are selected depending on the degree of imposition on a recipient (Brown & Levinson 1978; Fukushima 1996). To bring out differences clearly, such analyses often contrast a small number of formats under broad social or situational asymmetries. Complementing such approaches, this study presents a survey of the recruitment system of one language based on a systematic collection of 207 recruiting and responding moves from a corpus of informal conversation. By focusing on the recruitment of practical actions, we can observe a range of factors that shape how people get others to do things in everyday interaction.

One way of understanding the organization of verbal and nonverbal resources in recruitment sequences is as addressed to a set of interactional challenges. People have to reach a joint understanding of who will carry out the practical action and why; what exactly needs to be done and when; how to coordinate bodily behavior and manipulate the physical environment; how to relate the desired action to preceding, ongoing, and projected activities; and other contingencies that require some degree of implicit or explicit calibration (Clark 2006; Goodwin & Cekaite 2013; Enfield 2014). The elements of recruitment sequences appear to be adapted to these challenges, which provides us with a roadmap to the interactional practices surveyed in this chapter (Table 1).

Not all resources make their appearance in every recruitment sequence. When people are already in a dyadic interaction, close to each other, and involved in an activity with a projectable structure, a recruiting move and its response can be minimal, even nonverbal (Rossi 2014). In other situations, interactional contingencies may need to be negotiated more explicitly, bringing a wider range of practices into play. This way, the recruitment system provides for a flexible organization of verbal and nonverbal resources adapted to the task of organizing assistance and collaboration.

# **1.1 The Siwu language**

Siwu is a Kwa language spoken north of Hohoe in Ghana's Volta Region. It has somewhere between 15.000 and 25.000 speakers depending on how the diaspora community is counted. This paper is based on Siwu as spoken in the village of Akpafu-Mempeasem. Siwu is a language in which grammatical relations are established primarily by word order (which canonically is SVO) along with an ex-

### 10 Recruiting assistance and collaboration: A West-African corpus study

Table 1: Interactional challenges to be negotiated in recruitment sequences, along with some of the interactional practices mobilized to address them.


tensive system of noun classification and agreement. The earliest lexical records for the language date back to the late 19th century, and there are recent sketches of phonology, morphosyntax, and the repair system (Kropp Dakubu & Ford 1988; Dingemanse 2015).

Studies of informal social interaction in West African languages are rare, as linguists have traditionally privileged phonetics, phonology and morphosyntax over semantics, pragmatics and language use (but see Ameka 1991; Obeng 1999; Meyer 2010 for prior work on interactional routines in some West African languages). By describing practices for getting another's assistance or collaboration in Siwu, this paper contributes not only to the documentation of this language, but also to a larger program of understanding how language is shaped by and

### Mark Dingemanse

for social interaction. As we shall see, interactional practices in a basic domain such as getting assistance and collaboration combine universal structural properties with language-specific resources. So the practices and principles described here are of broad relevance to the cross-linguistic study of recruitments and of talk-in-interaction.

# **1.2 Data collection and corpus**

This work is based on a video corpus of naturally occurring conversations in Siwu, collected from consenting participants over the period 2007–2013. The target behavior was maximally informal social interaction: the primary ecology of language in use and the most promising baseline for cross-cultural comparison (Dingemanse & Floyd 2014). All of the recordings were made outdoors, where most social interaction between family and friends happens. The recordings cover dyadic as well as multiparty conversations between family and friends. To achieve a diverse and representative collection of recruitment sequences, multiple 10-minute stretches from a total of 11 different interactions were exhaustively sampled, amounting to a total of almost 3 hours of conversation in everyday settings.

A first sweep through this corpus identified a total of 389 candidate recruitments, which amounts to over two recruitments for every minute sampled. This includes 173 cases involving small children as recruiter or recruitee, reflecting the fact that children engage in interactive prosocial behavior from a young age (Warneken & Tomasello 2013). Such recruitment sequences stand out from other cases in a number of ways, most striking among them a higher number of noticeably absent responses and concomitant response pursuits (see §5).

To avoid skewing the sample and to maintain comparability with other languages, recruitment sequences involving small children were not included in the core collection for Siwu, leaving only sequences involving adults and children roughly from age eight onward (when they are clearly treated as having their own deontic authority, along with typical domestic rights and duties).<sup>1</sup> This leaves a core collection of 207 recruiting moves initiating 146 independent recruitment sequences. Even this conservative count finds roughly one recruiting move for every minute of conversation sampled, showing the fundamental importance of these interactional practices for social life.

<sup>1</sup>Any boundary drawn in order to achieve comparability is debatable. In §5 I discuss excluded cases and offer some observations on notable differences.

### 10 Recruiting assistance and collaboration: A West-African corpus study

# **2 Basics of recruitment sequences**

There are many ways of conceptualizing assistance and collaboration in interaction, giving rise to a variety of terms and definitions in prior work. To achieve cross-linguistic comparability, the focus of this study is on sequences of interaction where one participant *recruits* another to do something practical. The phenomenon of recruitment is defined as a sequence of two moves with the following characteristics (see Chapter 1, §4):


This definition characterizes the phenomenon as a conversational sequence, implying that a variety of semiotic resources may be used to implement it. The sequential nature of the definition means that we can use the "natural" or "sequential control method" (Dingemanse & Floyd 2014; cf. Zimmerman 1999) to locate comparable cases across settings and societies. The main focus is on practical actions in the here-and-now. Of course, people also recruit assistance or collaboration for matters that cannot be fulfilled immediately (e.g. building a house or borrowing a car). These cases are beyond our scope here, though they are likely to involve substantially similar resources.

# **2.1 Minimal recruitment sequence**

Many recruiting moves are minimally formatted and straightforwardly complied with. In (1), participants are checking some batches of rice (Figure 1). Eku asks Yawa to give her 'the deep calabash one' (line 1), referring to some rice in a deep calabash resting at Yawa's feet. She reaches out to receive it (line 2) in anticipation of Yawa handing it over (line 4). In the transcripts, ▶ and ▷ are used to mark the moves in focus, distinguishing initiating and responding moves where relevant. The individual frames within the figures are designated as *a*, *b*, *c*, etc. from left to right.


Mark Dingemanse

▶ 2 [((reaches out for calabash, Fig. 1a))

```
3 yaw àrĩ abùà agbagba[rà ló
       rice it:exceed it:IDPH.be.large FP
       this rice is really large-grained
```
▷ 4 [((takes calabash and hands it to Eku, Fig. 1b))

5 eku àba ɔrãrã ànaà. it:have weight too it's heavy, too

Figure 1: (*a*) recruiting move by Eku (sitting right, lines 1–2); (*b*) responding move by Yawa (line 4).

This recruitment is minimal in the sense that it consists of an initiating move – Eku's 'take the deep calabash one and let me see' (Figure 1*a*) – and a single response – Yawa taking the calabash and handing it to Eku (Figure 1*b*). About two thirds of all independent recruitment sequences in the corpus (102 out of 146) have this kind of simple two-part structure of initiating move and response.

# **2.2 Non-minimal recruitment sequence**

The complex interactional challenges at play in everyday recruitments are easy to overlook in minimal sequences, where a pre-existing shared focus of attention, physical co-presence, and activity structure conspire to enable a simple request that is immediately fulfilled. About one third of independent recruitment sequences (44 out of 146) take more than one attempt to reach completion. In such non-minimal sequences, the levels of coordination are pulled apart a bit, similar to the way in which an exploded-view diagram can show the elements and order of assembly of a complex piece of machinery.

Two common ways in which non-minimal sequences happen are (i) when a response to the recruiting move is noticeably absent or delayed, which often results in the recruiter pursuing a response, and (ii) when a recruitee claims a

### 10 Recruiting assistance and collaboration: A West-African corpus study

problem of hearing or understanding and initiates repair. Extract 2 illustrates the first type (for the second type, see §4.2 below). Beatrice is cleaning some pots and pans while Afua, her mother, is holding Beatrice's infant. When the infant becomes increasingly restless, Afua asks Beatrice to wash her hands and take him over (line 1). Beatrice immediately provides an affirmative verbal response (line 2), but in the next 10 seconds she appears to continue her current involvement, even taking up another pot to clean. This leads to multiple response pursuits by Afua (lines 5–6, 8) until Beatrice carries out the requested action.

(2) Kitchen1\_1052883

```
▶ 1 afu Beatrice fore nrɔ̃ si àba àakɔ ũ=
          PSN wash hands LNK 2SG:come 2SG:FUT:take him
          Beatrice wash your hands, so you can come and take him
▷ 2 bea =ao
          yes
          yeah
  3 afu nɛ ɔ̃ũ bùa ɔsɛ
          TP he.TP exceed 3SG:sit
          cause he's done sitting
  4 (10.0) ((Beatrice takes up another pot and starts cleaning it))
▶ 5 afu Beatrice mɛ sɔ fore nrɔ̃
          PSN me says wash hands
          Beatrice I said wash your hands
▶ 6 si àba àa kɔ ũ si ɔnyũa kàku ɔɔbiɛ ló.
          LNK 2SG:come 2SG:FUT take him LNK 3SG:stop cry crying FP
          so you can come and take him, so he'll stop crying
▷ 7 bea aoo: ((speeds up and finishes cleaning, starts washing her hands))
          yes
          ye:s
  8 afu nɛ ɔ̃ũ bùa ose
          TP he.TP exceed 3SG:sit
          because he's done sitting
  9 (37.0) ((Beatrice finds a towel, dries her hands, and walks towards Afua;
          baby cries))
▷ 10 bea ooo! ((picks up baby))
          EXCL
          ooh!
  11 (3.0) ((baby calms down))
```
This case illustrates a range of practices commonly used to manage the interactional challenges posed by recruitment sequences. For Afua, this includes using a proper name to secure joint attention (line 1), providing a reason that orients to Beatrice's current involvement, pursuing response by marking the follow-up as a resaying (lines 5–6), and invoking Beatrice's responsibilities for the task at hand (lines 3, 6). For Beatrice, this includes using affirmative responses to signal

### Mark Dingemanse

willingness to comply (lines 2, 7), visibly speeding up and shifting tasks to signal imminent availability (line 7), and finally carrying out the requested action (line 10). All of the practices noted here are discussed in more detail below.

That there are non-minimal sequences means that not all 207 recruiting moves in the core collection are independent events: some are pursuits of response following problems in compliance or other-initiations of repair.<sup>2</sup> In what follows, where relevant, I make a distinction between *initial* (or *independent*) versus *subsequent* recruiting moves, and I reserve the term *recruitment sequence* for the full sequence – minimal or non-minimal – an initial recruiting move gives rise to.

# **2.3 Subtypes of recruitment sequences**

The practical actions instigated by recruiting moves can be classified into types. Three common ones are (i) the transfer of an object from B to A, (ii) the provision of a service by B for A, and (iii) the alteration of a trajectory of action. We have seen an object transfer in (1), where a calabash changes hands, and the provision of a service in (2), where a mother is recruited to take care of her child. The notion of "service" is the broadest of the three and it is no surprise that this turns out to be the most frequent category in the corpus (Table 2).



Extract 3 below illustrates the third type of recruitment, where one person asks another to alter an ongoing trajectory of behavior. Yao and Afua are producing palm oil when Lucy stops by their compound to ask something (line 1, 4). She happens to position herself right before the camera. Yao draws attention to this and asks her to move aside.

<sup>2</sup> In §5.1, I discuss an apparent limit to the number of pursuits observed.

### 10 Recruiting assistance and collaboration: A West-African corpus study

```
(3) Palmoil1_1118517
  1 luc ǹdɔrɛ̃ kasorekɔ̃ misee? ((moves in front of camera))
          firewood LOC:gather:place 2PL:go:Q
          are y'all going to the firewood place?
  2 yao m[m
  3 afu [mm
  4 luc mikɛlɛgu ilɛ?
          2PL:go.with place
          where are you going to bring {it}?
▶ 5 yao nyɔ àta àbɔrɛ gu fɔ ɛh
          look 2SG:PROG 2SG:move with your HES
          look move away with your uh
▶ 6 ɔpò m̀mɔ nɛ həhəhəh ndza marɔ̃
          tub there TP how 3PL:call
          tub there həhəhəh what-d'you-call-it
▷ 7 luc ((steps aside and takes a look at the camera))
  8 ↑hɛ↑ (.)↑ah↑
```
Yao's request to 'move away with your uh tub there' (lines 5–6) is not a response to Lucy's question. Instead it launches a new course of action, with the turn preface 'look' marking a departure from the current course of action (Sidnell 2007) and helping to redirect Lucy's attention to the camera, which is behind her. She turns around and takes a look at the camera (line 7), producing two high-pitched exclamations of surprise (line 8) which also claim unawareness of the situation and therefore serve to account for her prior actions.

The sequential definition of recruitments used here relies on the recognition of Move B as a practical action for or with participant A. This opens the door to a further possible distinction with regard to how Move B arises. Often, it is prompted by an explicit request in Move A, as we have seen in the examples so far. But it can also arise in anticipation of a current or imminent need. This is illustrated in (4).

Emma, a blind woman, is inside a room while some others are chatting and preparing food outside. One of them, Aku, is sitting in the doorway. When it becomes clear that Emma is going to go outside (line 1), Aku stands up from the doorway to make way for her (line 4).

```
(4) Compound4_2054269
```

```
▶ 1 emm [((audibly takes some shuffling footsteps toward doorway, Fig. 2a))
```

```
2 kof [mmakosò
         kin.F.junior
         aunty
```
Mark Dingemanse


10 emm ((leans against portal and takes a careful step down))

Figure 2: (*a*) Aku sits in the doorway as Emma approaches from inside (line 1); (*b*) Aku stands up and frees doorway (line 4). Kofi is not visible in the frame.

Cases like (4), in which someone responds to anticipated trouble, can be challenging to identify because the recruiting move itself is not on-record: Emma

### 10 Recruiting assistance and collaboration: A West-African corpus study

does not ask Aku to get up. However, in this case, another participant happens to provide a running commentary that supports an analysis of this event as a recruitment. Kofi, a distant relative hanging around and engaging in occasional chats with the others, first cautions Emma to be careful stepping out the door, then describes what happened in causal and sequential terms, stating how one behavior occassioned another: 'so because you're coming, she stood up' (lines 7, 9). This comment glosses Aku's assistance as relevant and potentially expected given the context.

Fully nonverbal recruitments like (4) are in the minority and straddle the boundary between offers of help and responses to requests (Curl 2006; Couper-Kuhlen 2014). One reason they are interesting is that off-record cues may, over time, develop into conventionalized signals, and may come to be seen as part of an ordered paradigm of interactional practices (Manrique & Enfield 2015). For instance, on urban sidewalks, an audible footstep is often sufficient to "ask" others to make space, and appears to be preferred over an explicit request, a format that tends to be reserved for subsequent attempts. In the following sections, we will explore a range of formats that are more directly on-record as requests for assistance or collaboration.

# **3 Formats in Move A: The recruiting move**

# **3.1 Nonverbal behavior in recruiting moves**

Most recruiting moves are multimodal utterances composed of speech and bodily behavior. The semiotic resources work in concert to produce the recruiting move, with a division of labor appropriate to the affordances of each modality (Goodwin 2000; Clark 2012). Three common forms of nonverbal behavior found in recruiting moves are: (i) reaching to receive an object, illustrated in (1) above; (ii) holding out an object; and (iii) pointing, illustrated in the following case.

Eku is preparing food. Her teenage daughter Kpɛi has just come back from school and is standing next to the water tank. Eku starts with an imperative *su* 'take', then self-repairs to ask Kpɛi to check whether there is water in the tank. After receiving confirmation, she produces a complex request that involves taking a container, filling it with water, pouring that water somewhere, then putting it on the fire (lines 3–6). The under-specification of the verbal content is made up for by a series of pointing gestures, three of which are illustrated in Figure 3.

### Mark Dingemanse

	- 5 (0.4)

Figure 3: Pointing gestures accompanying (*a*) 'take this gallon' (line 3), (*b*) 'pour some water' (line 4), (*c*) 'put it on the fire' (line 6).

Besides the three consecutive pointing gestures, this sequence reveals a range of verbal elements that enter into the design of recruiting moves, to which we now turn.

# **3.2 Verbal elements: constructions for formulating recruiting moves**

Recruiting moves come in different *formats*, conventionalized linguistic practices that deliver social actions (Thompson & Couper-Kuhlen 2005; Fox & Heinemann 2016). For recruiting turns that include a predicate, it is possible to distinguish between a number of constructions and grammatical moods (Table 3). There is a small number of recruiting turns that do not feature a predicate (for instance, combining 'hey' with a pointing gesture to draw someone's attention to an actionable matter). Also, in 11 mixed cases, formats are combined. The basic con-

### 10 Recruiting assistance and collaboration: A West-African corpus study

struction types reviewed here can be further enriched with a range of final particles and other elements, described in the next section.


Table 3: Verbal formats of 172 recruiting moves (excluding 35 fully nonverbal cases).

As Table 3 shows, all construction types occur in initial as well as subsequent position. However, there are some patterns that suggest an ordering of resources. For instance, 7 out of 8 interrogatives are found in initial position (the only subsequent case is a response pursuit that repeats an initial interrogative). So an interrogative is never selected as an upgrade of another format. But the opposite does occur, as when an initial interrogative is reformulated as a proposal in (10) below. Conversely, some non-predicative formats like *anɔ:* 'y'hear?' in (20) occur only in subsequent position, as a result of the fact that one can pursue a response to a recruiting move by repeating only part of it – in this case, the final tag.

Linguistic labels such as those in Table 3 are employed here for ease of reference. However, the analysis of these formats below is focused more on understanding the interactional work done with these formats, each of which has its own affordances for social action. To briefly preview the interactional work done with the main constructions: imperatives allow people to direct each other's actions; *si*-prefaced recruiting moves present a requested action as a logical consequence; declaratives are noticings that present reasons for action; negative interrogatives mark deviations from expected courses of action; and jussives frame recruitments as suggestions for courses of action.

### **3.2.1 Imperative**

The basic imperative in Siwu consists simply of the bare verb, usually morphologically unmarked for person and number, though occasionally the second per-

### Mark Dingemanse

son plural prefix *mi-* can be found. Some imperatives feature just a verb phrase (e.g. *sa mà* 'chase them away', Extract 6), others add a beneficiary (e.g. *su tã mɛ* 'gimme back', lit. 'take give me', Extract 22) or a more elaborate specification of the desired action (e.g. *ba fore mɛ ndu* 'come pour me water', Extract 21). Serial verb constructions, as in the latter two examples, are common.

Although a plural form of the imperative does exist, most recruiting moves are unmarked for person or number, even when the recruited action is taken up by multiple people. An example of this is given in (6), where one participant notices some goats getting too close to the food and issues a directive to 'chase them away'. Her recruiting move is unaddressed and unmarked for person or number, and is taken up by two people who are closer to the goats than the recruiter is (lines 5, 6).

(6) Cooking1\_1545188

```
1 ((goats approach food))
▶ 2 afu sà ma
         chase them
         chase them {away}
  3 (0.5)
  4 afu sà [ma
         chase them
         chase them {away}
▷ 5 taw [kai (0.4) [↑kai ((waves arm))
            INTJ INTJ
            skai kai
▷ 6 adz [hî hî, hî, hî ↑híì↑ ((waves arm))
                      INTJ INTJ INTJ INTJ INTJ
                      hî hî, hî, hî ↑híì↑
  7 ((goats flee the scene))
```
Imperatives are by far the most common construction type in the Siwu data, accounting for 59% of all recruiting moves and over 70% of moves featuring speech. As we will see below, there are several ways of designing imperative recruiting moves to specify consecutive actions (*si*-prefacing, §3.3.2) or to mark fine differences in stance or illocutionary force (final particles, §3.3.3).

### **3.2.2 Declaratives and interrogatives**

Some recruiting moves in the collection come in the form of declaratives. All of them are noticings of some actionable event or matter that requires attention. In (7), two women are chatting while preparing food. Vicky is in the process of

telling a story when she sees a chicken coming up behind Tawiya. She interrupts her telling mid-turn (line 3) to tell Tawiya of the chicken, marking it as a piece of advice with the final particle *ló* (line 4), which results in Tawiya shooing away the chicken using the animal-oriented interjection *shuɛ* (line 5). Without missing a beat, Vicky then resumes the story by recycling material from the turn she abandoned (line 7).

(7) Compound4\_1600030


A similar case happens later in the same interaction, when Tawiya has put a pan on the fire next to her and Vicky sees it sliding from one of the firestones, at risk of toppling. Vicky notifies Tawiya by pointing out the trouble and Tawiya responds by righting the pan.

(8) Compound4\_1655650

1 ((pan slides off one of the firestones))

```
▶ 2 vic kãrã te ìturu. ((points to trouble))
          pan it:PROG it:tilt
          the pan is tilting
```
3 taw ↑mm↑ ((turns to look, repositions the pan))

In both cases, the declarative formatting is well suited to delivering a verbal "noticing" of some actionable trouble which the other may not have noticed yet

### Mark Dingemanse

and is in a good position to resolve (see also Kendrick, Chapter 4, §4.2.3; Rossi, Chapter 5, §3.3.4; Baranova, Chapter 9, §3.3.3).<sup>3</sup>

Question-formatted recruiting moves are rare in the Siwu data, and the most common type is a negative interrogative format. In (9), Dora spots somebody walking off and asks 'hey, aren't you bringing me water?' (line 1). The negative interrogative design gives the recruiting move a complaining quality (Heinemann 2006) and appears to orient to a decreased likelihood of immediate fulfillment. Indeed, Efi indicates that she will be going someplace else before coming back. Dora's response provides further evidence of the complaint-like quality of the initial formulation: 'it's because of you this woman has not bathed yet' (line 5).

```
(9) Maize1_6136999
```

```
▶ 1 dor HÀƐ: AITÀ BƆ MƐ NDU:
         INTJ 2SG:NEG:PROG bring me water:Q
         HE:Y AREN'T YOU BRINGING ME WATER?
  2 (1.0)
  3 efi losɛ kàto ngbe loba.
         1SG:go top here 1SG:come
         I'm going up, I'll be back
  4 (1.3)
  5 dor ƆƆNYA FƆ ƆSO ƆRƆ̃GO GƆǸGBE ŨIPIE NDU
         2SG:PFV:see 2SG reason woman REL:here 3SG:NEG:bathe water
         YOU SEE IT'S BECAUSE OF YOU THIS WOMAN HAS NOT BATHED YET
```
In (10), mealtime is approaching and Afua calls out to her fellow clan member Eku asking 'won't you eat food?', the plural *mi*- signaling that Eku is in the company of others. When no response follows, she upgrades the recruiting move, shifting from an interrogative to a jussive format, discussed in the next section. The recruitment attempts are ignored and then abandoned as the conversation lapses.

```
(10) Cooking1_1266243
```

```
▶ 1 afu Daa Ɛku (.) mìite mìde ara:
         sister PSN 2PL-NEG-PROG 2PL-eat thing
         Sister Eku (.) won't you eat food?
```

```
2 (0.4)
```
<sup>3</sup>A reviewer points out that the beneficiary of the target action here is not clearly the person producing the recruiting turn, making the recruitment akin to what Couper-Kuhlen (2014) describes as "suggestions". However, such suggestions in Couper-Kuhlen's English data are "likely to be resisted in everyday conversation" (p. 635) and often have the recipient as the primary beneficiary; here, no such resistance is in evidence and the beneficiary is neither self nor other alone, but both together.

10 Recruiting assistance and collaboration: A West-African corpus study

```
▶ 3 afu mìba mìade adera.
         2PL:come 2PL:FUT:eat food
         you should come and eat food
  4 ((interaction lapses))
```
### **3.2.3 Jussives**

Recruiting moves can be formulated as proposals using the verb forms *ba* 'come' and *tã* 'give', which can be structurally characterized as jussives. The first is often heard in the formulaic proposal 'come let's eat' that is routinely addressed to passers-by when people are sharing a meal. In (11), Ruben invites Kodzo to share a meal, though Kodzo declines. This first person plural formulation is the most commonly encountered version of the *ba* 'let's' format. An instance of the second person plural version is found in (10) above.


Another jussive format frames the recruitment as a proposal with a beneficiary, owing to the semantics of the *tã* auxiliary, derived from *tã* 'give'. We have seen one example in (1), where the beneficiary is the recruiter herself ('let me look'); in (12), it is a third person ('let him sit by your side'). Evidence of the auxiliary status of *tã* comes from the occurrence of negative forms like 'don't let me get sore' in (13). If *tã* were a bona fide verb here, it would require the benefactive to follow immediately after it (*tã mɛ* 'give me'); instead, it conveys a jussive meaning 'let {it}' and the main predicate is *bɛbɛrɛ* 'burn, feel sore'.

(12) Maize3\_673020 (see Extract 20 for full sequence)

```
4 aku tã ũ ɔsɛ i fɔ kɔrɛ.
       let him NOM-sit LOC 2SG side
       let him sit by your side
```
(13) Compound5\_366774

```
1 aku daa tã bɛbɛrɛ mɛ
        NEG let burn me
        don't let me get sore
```
### Mark Dingemanse

Interrogatives and jussives are typically classified as more indirect than imperatives, and prior work in cross-cultural pragmatics suggests "a strong preference for conventional indirectness" in languages like English and German (Ogiermann 2009). In Siwu, by contrast, imperative constructions are the main workhorse for recruiting moves, and interrogatives and jussives play relatively minor roles. With imperatives, declaratives, interrogatives and jussives, we have exhausted the basic grammatical distinctions found in recruitment predicates in Siwu.

# **3.3 Additional verbal elements**

### **3.3.1 Vocatives**

One prerequisite for fulfilling a recruitment is establishing who will do it. In multiparty interaction, vocatives – linguistic resources such as proper names and interjections used for addressing people – provide one way to address recruiting moves to specific participants and to get their attention. We saw this in earlier examples where recruiting moves are prefaced by proper names: 'Beatrice wash your hands […]' (2) and 'Sister Eku, won't you eat?' (10). In both cases, the recruitment happens in multiparty interaction, and the vocative helps to cut across established participation frameworks and activities to address a specific recipient.

Proper names and other terms of address can also show up in summons-answer sequences preceding the recruiting move. Though not an "additional element" in such cases (see Chapter 2, §6), I discuss them here because of their connection to vocatives. An example is given in (14) below. Bella calls her mother with the vocative *mama* and, after getting an answer, asks her to get up and sit elsewhere while preparing the food. A summons-answer sequence serves the role of establishing an open channel for interaction (Schegloff 1968). Other examples can be found in (15), (31), and (33).

### (14) Cooking1\_1188540

```
1 bel mama.
```
2 mom ↑m


### 10 Recruiting assistance and collaboration: A West-African corpus study

Vocative interjections like 'hey' can be used in the same two sequential environments: as a summons separate from the recruiting turn or as an element within the turn. We saw an example of the latter in (9), where Dora addresses someone in the distance with 'hey, aren't you bringing me water?'.

### **3.3.2 Marking consecutive actions and giving reasons**

Many recruiting moves in the collection consist of an imperative followed by the specification of a consecutive action that is introduced using the morpheme *si*. An example of this can be found in (2), 'wash your hands *si* you come take him'. For this morpheme, I adopt the term "linker" from Ameka's (2008) analysis of Ewe *né*, a form with a similar range of uses. In (15), Eku asks her daughter Afua to take a broom and sweep the compound, introducing the second element of the action with *si*.

```
(15) Neighbours_4593390
```

```
1 eku Afua
         PSN
         Afua
  2 (0.8)
▶ 3 eku su ibubù si kà afifiɛ ngbe.
         take broom LNK IMM you:PLUR~sweep here
         take a broom and sweep here
  4 afu ((gets up to take broom))
```
In these and other examples, there is a complex recruiting turn specifying more than one action, where the first action (usually formatted as an imperative) appears to be a first step for later actions, and the later actions are introduced in a *si*-prefaced subordinate clause. In this context, *si* can often be translated as 'then', 'so that', or 'in order to' (Table 4). Sometimes the *si* clause refers to a component of the recruited activity (e.g. 'come take the child', 'sweep here'), while in other cases it need not be done by the recipient (e.g. 'so that I can wash my hands', 'so that {he} be dressed'). What unites all cases is that *si* marks a consecutive relation in which one action follows another, the first often addressing a precondition for the one introduced by *si*.

A case discussed earlier, reproduced in part as (16) below, provides a closer look at the relation between different stages of recruitment and the design of *si*-prefaced recruiting formats. Eku first launches a bare imperative format, then self-repairs and turns it into a question about a necessary precondition: 'take uh: is there water there?'. The self-repair reveals an orientation to the conditions

### Mark Dingemanse

Table 4: First steps and consecutive actions in multi-part recruiting turns.


necessary for fulfilling the recruitment. Once it is clear that this condition is met, she goes on to formulate a recruiting turn combining an imperative and a *si*-prefaced target action (line 3). That turns out to be only the first in a series of actions requested of Kpɛi, all introduced by *si*-prefaced clauses: '*si* you take this gallon', '*si* you pour some water', '*si* you put it on the fire'. This supports the analysis of *si* as encoding consecutive actions. The consecutive action verbs all have irrealis mood and so can be described collectively as linked by a form of co-subordination, a situation similar to the linker *né* in Ewe (Ameka 2008).<sup>4</sup>

```
(16) Maize3_276559a (excerpted from Extract 5)
```

```
▶ 1 eku su ɛ:. ndu pia mmɔ: ((points in direction of water tank))
          take HES water be there:Q
          take uh:. is there water there?
  2 kpɛ mm.
          INTJ
          mm.
  3 eku su fore si àsu ɛh gálɔn gangbe ((points to gallon))
          take pour LNK 2SG:take HES gallon AGR:this
          pour it {and} take uh this gallon
```
The consequential or consecutive reading of *si* opens up the possibility for *si*prefaced clauses to be used in providing reasons for recruitment. An example is given in (17). Mom calls on Sesi, her teenage son, to bring her a 'knife and uh tub' (line 1). When, moments later, Sesi arrives with only a knife, she repeats the request for a tub, now adding a *si*-prefaced reason: 'so I {can} peel the cassava'

<sup>4</sup>Homophonous with *si* 'so that' as a marker of consecutive action is a *si* 'if' form that introduces conditional antecedents. It is possible that the two are related, which would render *si* heterosemous and would make the *si*-prefaced format akin to independent if-clauses (Ford & Thompson 1986), which have been found in many languages to develop into a dedicated request format (Evans 2007; Lindström et al. 2016). However, many of the *si*-prefaced recruiting turns do not lend themselves to a conditional reading; indeed, they tend to be closer to the consequent ('then') than to the antecedent of a conditional.

### 10 Recruiting assistance and collaboration: A West-African corpus study

(line 4). Peeling the cassava is an activity for which one needs a knife and a container. By mentioning this activity and marking it as a consecutive action, Mom renews the relevance of getting the tub and adds weight to her repeated request.

```
(17) Neighbours_662742
  1 mom Sesi bɔ mɛ ipɛmi ku ɛɛ kàpoi anɔ:?
          PSN bring me knife and HES tub:DIM you:hear:Q
          Sesi bring me a knife and uh tub y'hear?
  2 (14.0)
  3 ses ((arrives with knife))
▶ 4 mom hɛ bɔ mɛ kàpoi, si lòyɛrɛ igbedi. ((receives knife))
          INTJ bring me tub:DIM LNK 1sg:peel cassava
          hey bring me a tub so I {can} peel the cassava
  5 (0.8)
  6 mom bɔ mɛ kàpoi
          bring me tub:DIM
          bring me a tub
  7 ses ((goes off to get tub))
  8 (23.0)
  9 ses ((arrives with tub))
```
Because of their consecutive meaning, *si*-prefaces can be used to present "inorder-to motives" (Schutz 1962) in interaction. In such cases, the *si*-prefaced clause is the motive for which the recruitment is a means, as here for Mom's request to be brought a tub so she can peel the cassava.

Another type of reason that people may use in recruitment sequences refers to "because-motives" (Schutz 1962). These are not marked with *si* but presented as declarative statements. We saw both types together in (2), where Beatrice was told 'wash your hands *si* you come take him' (an *in-order-to* motive), 'cause he's done sitting' (a *because*-motive).

Reasons are provided in 26 out of 207 initial and subsequent recruiting moves. Most commonly, they occur in pursuits of response when there was a problem in uptake, as we saw in (2) and (17). In the relatively rarer cases when they occur in first position, they may be designed to help disambiguate a request (21) or to anticipate a question about rights and duties that might otherwise come up. These functions of reason-giving, which can be summarized as rendering recruitments more intelligible and making fulfillment more likely, correspond closely to those found in a dedicated study of a collection of 56 recruitment sequences featuring

### Mark Dingemanse

reasons in Russian (Baranova & Dingemanse 2016; see also Baranova, Chapter 9, §3.4.2).

### **3.3.3 Mitigating and strengthening recruiting moves**

In their seminal work on the structure of therapeutic interaction, Labov & Fanshel (1977) noted that some linguistic devices appeared to soften requests ("mitigators") while others may serve to strengthen them ("aggravators"). Conversation-analytic work since then has showed that such devices can be understood with reference to the sequential structure of interaction (Heritage 1984; Schegloff 2007). We have already seen some of the strategies for upgrading the strength of subsequent versions of recruiting moves, for instance by adding a marker of resaying or by providing a reason.

Like many West-African languages, Siwu has a system of final particles, two of which are of particular interest with regard to the question of how people can modulate the force of recruiting moves. The final particle *ló* conveys 'I advise you', implying no claim about prior knowledge. The form *ní* conveys 'you should have already understood', implying a claim about prior knowledge and a complaint that this has not been acted upon. The two forms are never found together in the same utterance in the corpus, and seem to occur in complementary sequential positions.

We saw a case of *ló* in (7), where Vicky noticed a chicken behind Ella and told her about it so she could take action. One affordance of *ló* is its "no fault" quality (Heritage 1984: 271). Its usage does not imply prior knowledge and so it does not blame the other for failing to know or notice something. This is why it can also serve as a gentle nudge that makes a recruiting move sound more affiliative. In terms of sequential position, it tends to occur in initial but not in subsequent versions of recruiting moves, as seen in the following case.

Emilia is preparing porridge in the kitchen while Aku is sitting outside, a few meters away, with her back to Emilia. After a lapse in the conversation (see Hoey 2015), Emilia calls on Aku to bring her bowl, with the implication that she can get some food. The recruiting move contains *ló*, marking it as advice and perhaps orienting to the possibility that Aku, sitting outside, may not be aware that food is ready to be served. When Aku does not respond immediately, Emilia pursues a response by first calling her, then repeating the recruiting move, this time without *ló* (line 6).

(18) Cooking1\_521410

1 (7.0)

10 Recruiting assistance and collaboration: A West-African corpus study

```
▶ 2 emi Aku bɔ mɛ fɔ irɔi ló
          PSN bring me your bowl FP.advice
          Aku bring your bowl ló
  3 (1.4)
  4 emi Aku
          PSN
          Aku
  5 aku mm.
          CONT
          mm
▶ 6 emi bɔ mɛ fɔ irɔi
          bring me your bowl
          bring me your bowl
▷ 7 aku ((gets up))
```
The final particle *ní* is almost a mirror image of *ló*. It rarely occurs in the initiating turn of a recruitment sequence and instead appears in subsequent versions that pursue a response. In (19), Emma is shuffling across the compound heading towards an overturned bench which she cannot see (this happens moments after Extract 4, where Aku stood up from the doorway to let her through). Aku instructs Emma to 'pass here'. When Emma does not appear to be listening and instead places her cane on the overturned bench, Aku pursues response by saying 'pass here *ní*', the *ní* particle marking the recruitment as something that should have been understood and acted on already.<sup>5</sup>

```
(19) Compound4_2076833
```
1 emm ((blind, walking with cane, is about to stumble over overturned bench))

```
▶ 2 aku ki ngbe. ((pulls Emma's arm))
          pass here.
          pass here
```

```
3 emm ((places cane on turned-over bench))
```

```
▶ 4 aku ki ngbe ní ((pulls Emma's and leads her around the bench))
          pass here FP
          pass here ní
```

```
5 emm ((lets herself be led by Aku))
```
In sum, the final particles *ló* and *ní* help to manage accountability by making claims about the recipient's knowledge (or lack of knowledge) about what they should be doing. *Ló* can be seen as a general dispensation, conveying 'I advise

<sup>5</sup> Similar strengthening uses of *ní* are found in a sequence analyzed in (28) and (29) below, where a mother attempts to get her teenage son to run an errand.

### Mark Dingemanse

you' without implying a complaint; *ní* conveys the reverse: 'you should have known this and acted on it already', and therefore holds the recruitee accountable for the failure to respond. These usages are in line with the use of the particles in non-recruitment contexts, where they have similar implications.

Another device that can be used to strengthen recruiting moves is *anɔ:* 'you hear', illustrated in (20), where a little boy is making tottering steps around three women: Aku, Charlotte, and Emma. Aku produces a request: 'let him sit by your side'. Then Charlotte adds 'his mother is winnowing rice', accounting for the unavailability of the primary caregiver. Although neither the request nor the reason for it are clearly addressed, the fact that two of three participants present have jointly formulated a request plus reason makes a response relevant by the third participant, Emma. When no response follows, Aku upgrades her request by specifying the action and adding a strengthening particle *anɔ:* 'y'hear?' (line 8).<sup>6</sup>

```
(20) Maize3_673020
  4 aku tã ũ ɔsɛ i fɔ kɔrɛ.
         BEN him NOM:sit LOC 2SG side
         let him sit by your side
  5 (0.5)
  6 cha ɔ̃ ɔnyĩ tó ɔ fɛ kàmɔ.
         3SG.POSS mother PROG 3SG winnow rice
         his mother is winnowing rice
  7 (0.8)
▶ 8 aku puta ũ (.) anɔ:?
         lift him 2SG:hear:Q
         pick him up (.) you hear?
  9 (0.8)
         ((continues in Extract 30 below))
```
*Anɔ:* 'y'hear?' is a tag question with affirmation as the preferred response.<sup>7</sup> Adding it to a recruiting move has the effect of soliciting a commitment to fulfill the recruitment: after all, admitting to hearing a request makes it harder to escape the normative requirement to comply with it.

<sup>6</sup> In (30), which continues this extract, the particle is repeated on its own in a further pursuit of response.

<sup>7</sup>Another instance can be found in (17), line 1, above.

### 10 Recruiting assistance and collaboration: A West-African corpus study

# **3.4 Fully nonverbal recruiting moves**

So far we have reviewed a range of linguistic, verbal resources for building recruiting moves. Only 23 independent recruiting moves in the corpus are fully nonverbal. These can be arranged according to the degree to which they are presented and treated as on-record. An off-record nonverbal recruiting move was illustrated in (4) above, where some imminent trouble on the part of one participant provides a reason for another participant to help out. In such cases, the trouble does not make a response conditionally relevant (Schegloff 1968): participant A cannot be said to have asked anything, and participant B cannot be held accountable for inaction. On-record nonverbal recruiting moves are rare in the Siwu data (3 independent sequences, 9 moves in total), and only seem to happen when the recruitment occurs as part of an already established activity sequence which can provide the context for their interpretation (Rossi 2014 and Chapter 5, §3.1; see also Kendrick, Chapter 4, §4.1.3; Zinken, Chapter 8, §3.1 Baranova, Chapter 9, §3.1).

One situation where we find such nonverbal recruiting moves is when a prior request has made relevant the execution of a related subtask. In (21), an extended recruitment sequence is initiated when Atasi tells Eku to 'get some water so I can wash my hands'. The *si*-prefaced reason here (see §3.3.2) helps to disambiguate and specify the request: one might need water for any of a number of purposes, with consequences for the quantity desired and the container to be used – in (5) above, for example, a gallon of water is needed for cooking, and in (9) an even larger quantity is needed for taking a bath. With the request and its reason made clear, Eku's standing up (line 2) marks a commitment to provide this service, and her return with a calabash with water (some 20 seconds later) marks the start of compliance. Now a series of nonverbal actions ensues in which Atasi holds out her hands and Eku pours some water in response (lines 18–22), a process that is repeated five more times until the task is completed.

(21) Compound5\_846793


2 eku ((stands up to fetch water))

((20 seconds pass, during which an unrelated story is told by a third party, after which Eku returns with a calabash of water and Eku and Atasi stand together))

▶ 18 ata ((holds out hands and assumes 'washing hands' position))

### Mark Dingemanse

```
▷ 19 eku ((pours water over A's hands))
  20 (2.3)
▶ 21 ata ((opens hands palms up for more water))
▷ 22 eku ((pours more water))
          ((actions in 21-22 repeated five times))
  33 ata ((shakes water off her hands, walks back to seat))
```
Cases like (21) show that recruitments can assume a fractal nature, where an initiation and its response can set up a context for a number of subsidiary sequences. To the extent that such subsidiary sequences occur in the context provided by the base sequence and are part of a default script associated with the base activity, they are often implemented nonverbally.

A recruitment episode with subsidiary sequences as in (21) raises the question of how we can distinguish between a series of recruitments versus a sequence of behaviors done in the service of one instance of recruitment (see also Chapter 2, §3). The most reductive approach would be to stipulate that only base sequences count as recruitments. So, in (21), there would be a single Move A ('come pour me water so I can wash my hands', line 1) and its fulfillment would be the full sequence of moves implementing that complex action, starting when Eku stands up to get the water (line 2) and ending when Atasi shakes the water off her hands (line 33). However, this analysis would fail to capture the contingent nature of Atasi's repeated nonverbal requests for more water (lines 21–22ff). The number of times water has to be poured is not preestablished and is under Atasi's control, while for the pouring of the water she fully depends on Eku. Therefore, Atasi's opening up her hands palm up is analyzed here as a Move A in its own right and Eku's pouring more water as a corresponding Move B, and a series of such moves in quick succession expands the base adjacency pair.

Another example of a fully nonverbal recruiting move is given in (22). Bella is holding Aku's phone and taking a call Aku asked her to pick up. Speaking into the phone, she notes she is 'not sister Aku'. When it becomes clear the caller wants Aku, Aku asks Bella to give the phone back (line 2). After a place in which a response would have been relevant (line 3), she asks again, now with an added gesture of reaching out to receive the phone (line 4). When Bella continues to speak on the phone, Aku produces one more response pursuit, this time fully nonverbal (line 6), after which she is handed back the phone.

10 Recruiting assistance and collaboration: A West-African corpus study

### (22) Neighbours\_818304


hello person:CQ FOC hello, who is this?

Figure 4: Aku reaches to receive the phone in an upgraded response pursuit (line 6).

Like the subsidiary recruitments in (21), the response pursuit in (22) occurs in an environment where it is already abundantly clear what needs to be done and by whom. So both cases fit the generalization that fully nonverbal requests tend to occur only when the activity structure, participation framework and prior context render verbal specification unnecessary (Rossi 2014).

Mark Dingemanse

# **3.5 Animal-oriented recruitments**

Recruitments are defined in this study as interactional sequences with human participants, in line with a focus of the larger research project on human sociality. However, people also have interactional practices oriented towards animals (Bynon 1976; Spottiswoode et al. 2016). Indeed, humans are not alone in producing communicative signals aimed at other species (Krebs & Dawkins 1984). Animal-oriented recruitments provide an interesting limiting case of how semiotic resources adapt to situations in which there are radical asymmetries in agency and linguistic capability between interactants.

In Siwu, as in many other languages, animal-oriented recruitments often involve a set of dedicated interjections (Ameka 1992). Two examples occurred in extracts discussed earlier, relevant portions of which are reproduced below. In (23), Tawiya's interjection *kai* can be said to effectively recruit the goats to go away, and in (24), the interjection *shuɛ* has a similar effect on the chicken.<sup>8</sup>

(23) Cooking1\_1545188 (excerpted from Extract 6 above)

```
5 taw kai, [↑kai ((waves arm))
6 adz [hm, hm, ↑hm, hm↑ ((waves arm))
7 ((goats flee the scene))
```
(24) Compound4\_1600030 (excerpted from Extract 7 above)


The shape of at least some of these animal-oriented interjections appears not to be arbitrary but motivated. Take *shuɛ* 'shoo', the interjection for chasing away domestic fowls. A survey of functional equivalents reported for other languages from around the world shows that shooing words seem to converge on sibilant sounds, variously transcribed as *s*, *ʃ*, *š*, *ç* (Table 5).<sup>9</sup>

<sup>8</sup>Conversation analysis shies away from attributing intentions to participants in interaction, aiming instead to base analyses on publicly observable sequences of behavior (Heritage 1990). This methodological stance renders CA suitable for analyzing at least some forms of nonhuman animal communication (Rossano 2013).

<sup>9</sup>Most of the sources cited do not give phonetic renditions, so forms are presented here without adjustments. The table presents a sample of typologically diverse languages selected by searching grammars and dictionaries for forms translated as 'shooing/chasing away chicken/fowl'.

### 10 Recruiting assistance and collaboration: A West-African corpus study

Table 5: Interjections for 'shoo' and words for 'chicken' in 17 languages from 11 phyla around the world, showing strong convergence towards sibilant sounds in the interjections but not in the words for 'chicken'.


Sibilant sounds show up in shooing words in a diverse sample of languages, many of which are not historically related. Some of the commonalities may be due to language contact. After all, the domestic fowl (*Gallus g. domesticus*) has itself been culturally dispersed (Liu et al. 2006) and some words may have traveled along. However, it is unlikely that the global similarities can be explained solely by cultural diffusion, as this would predict words for 'chicken' to show similar global commonalities, which they do not (Table 5). Nor can the global similarities be explained solely by inheritance from a common ancestor, as this would require a temporal stability that even basic vocabulary is not known for; and again, words for 'chicken' do not show such global similarities. A more parsimonious explanation is that some sounds are more effective than others for the goal of shooing birds, and come to function as cultural attractors biasing the transmission of shooing words – a form of convergent cultural evolution.

Convergent cultural evolution has been put forward as an explanation for a range of cross-linguistic similarities (Caldwell 2008; Dingemanse et al. 2013; Blythe 2018). Animal-oriented interjections present a particularly illuminating

### Mark Dingemanse

view of the phenomenon, as the evolutionary landscape to which such words must adapt is strongly constrained by the perceptual and behavioral systems of the animals in question. The effectiveness of prolonged sibilants in shooing words for domestic fowls can be connected to the fact that continuous high frequency sounds are among the sound stimuli domestic fowls are most aversive to (MacKenzie et al. 1993).

Owing to the narrow ranges of behavior they seek to elicit, animal-oriented signals may present one of the few areas of language that can be truthfully said to bring behavior under the control of some stimulus, as Skinner (1957) envisioned. The principle of semiotic adaption to perceptual systems is likely to hold across a wide range of animal-oriented communicative signals across languages.<sup>10</sup>

# **4 Formats in Move B: The responding move**

So far, we have considered the design of Move A, the move by which a recruitment is initiated or pursued. But a recruitment sequence is not complete without a Move B. In what follows, we consider the design of Move B and the further development of the sequence, from simple closure in the case of fulfillment to sequence expansion in the wake of resistance and rejection.

# **4.1 Nonverbal and verbal elements of responses**

Since recruitments by definition involve getting another to perform a practical action, many relevant responses are nonverbal and simply consist of the doing of the target action. Examples of this are shown in Figures 1*b*, 2*b*, and 5*b*, and further examples are transcribed in Extracts 3, 4, 15, 17, 18, 19, 21, and 22. About two thirds of responses to initial recruiting moves are fully nonverbal, and the great majority of these fulfill the target action or plausibly start doing so.

Although the focus here is on the composition of Move B, an important factor in its design is the format used in Move A, which initiates the recruitment sequence. Consider the relative frequency of fully nonverbal responses. Table 6 gives the proportion of fully nonverbal Moves B relative to the format of Move A. It shows that nonverbal Moves A are followed by a fully nonverbal Move B in 77% of cases; the remaining 23% is either composite or verbal only. On the other

<sup>10</sup>In an ethnological study of domestic fowls, Fischer (1972) shows that sound stimuli featuring repeated low-frequency sounds are most likely to induce following. This generates the prediction that, across languages, words for calling domestic fowls will feature more repetition and lower-frequency sounds than words for shooing them.

### 10 Recruiting assistance and collaboration: A West-African corpus study

hand, responses to interrogative recruiting moves are fully nonverbal in only 17% of cases.<sup>11</sup>

Table 6: Proportion of nonverbal Moves B relative to the format of Move A in Siwu recruitment sequences


Recruiting formats in Move A can be ranked on a cline from more to less coercive (Brown & Levinson 1978). One way to explain this cline is in terms of the "response space" created by the formats (see Vinkhuyzen & Szymanski 2005; Rossi & Zinken 2016). As we saw above, nonverbal recruiting moves occur only in situations where the context makes abundantly clear what is requested, which places considerable constraints on the response space and makes relevant immediate (and nonverbal) fulfillment. Imperatives similarly push fairly directly for fulfillment and leave little room for other types of responses (see, e.g., Kent 2012; Rossi 2012). At the other end of the spectrum, interrogative recruiting turns in Siwu tend to be negative interrogatives like 'why don't you', which formulate things either as complaints or proposals, both of which allow verbal or composite responses and push less directly for fulfillment.

One of the main uses of verbal material in the responding move is to signal commitment to fulfilling the recruitment. We see this in (25). Becca, seated on a low bench, is winnowing rice; Ama, who is trying on a new dress, comes standing with her back to Becca and says 'fix me' (Figure 5*a*). Becca immediately responds 'now, I'm coming', takes a second to put down the rice winnower and stands up. Then she carries out the requested action, zipping up Ama's dress (Figure 5*b*).

(25) Tailor\_995460

1 (3.0) ((Ama walks towards Becca))

▶ 2 ama di mɛ ((comes standing with back to Becca, Fig. 6a)) fix me fix me

<sup>11</sup>See Rossi, Chapter 5, §4.1 for comparable distributions of fully nonverbal responses in Italian.

### Mark Dingemanse


Figure 5: (*a*) Ama stands with her back to Becca (line 2); (*b*) Becca zips up Ama's dress (line 5).

So verbal responses can claim a commitment to fulfilling a recruitment when something stands in the way of immediate fulfillment. At the moment the are produced, these are, of course, claims rather than demonstrations. We saw this in (2), where Beatrice said 'yes' to a request while finishing another activity. She was subsequently held accountable for not stopping the other activity soon enough. So recruiters may hold the recruitee accountable when verbal claims become incongruent with visible actions.

Sometimes verbal elements of responses can address aspects of the design of a recruiting turn. For instance, in (8), Vicky notified Ella that a pan was sliding off a firestone. Ella responded by righting the pan and by uttering a high-pitched response token *↑mm↑*, marking Vicky's noticing as something counter to expectation. Another example where the nonverbal element of the response fulfills the recruitment while a verbal element responds to the formulation of the recruiting move is given in (26) below. Odo, carrying a small metal pan holding some food that is possibly hot, walks towards a bench to sit down but finds Bella standing in his way. He issues a crude request to Bella to get out of the way, which she does,

### 10 Recruiting assistance and collaboration: A West-African corpus study

but not without voicing her disapproval at his formulation with the response token *woo:*.

(26) Neighbours\_880320

```
▶ 1 odo rùi bie kakɔiɔ sɛ wãrã.
          uproot find place:INDEF sit rest
          get out of the way and find somewhere {else} to relax
▷ 2 bel woo: ((steps aside to make way))
          INTJ
          woo:
```

```
3 odo ((sits down on bench))
```
A number of features of turn design conspire to make Odo's recruiting move akin to an extreme case formulation and give it complaint-like qualities (Pomerantz 1986). The verb *rùi* literally means 'uproot'; the indefinite marker *ɔ* attached to *kakɔi* 'place' works to suggest that Bella should be anywhere but where she is; and the construal of her current action as 'relaxing' implies that Bella, perhaps unlike Odo, has nothing to do. Bella's interjection of disapproval *woo:* appears to be addressed to these features.

In sum, we have seen that the bulk of complying responses to recruitments are nonverbal. Verbal elements of responses may vary in relation to the format of the recruiting move and may be used to (i) claim commitment when something stands in the way of immediate fulfillment and (ii) respond to action affordances of the design of the recruiting move. But a further, major role for verbal elements of responses to recruitments is in the domain of resistance and rejection, to which we now turn.

# **4.2 Repair, resistance, and rejection**

Sometimes recruitments are not immediately fulfilled, but questioned, resisted, or even rejected. Resistance and rejection rarely come in the form of explicit claims of unwillingness. Rather, participants have a variety of ways to avoid immediate compliance (Kent 2012), though none of them comes for free: as we will see, resistance and rejection (and more generally, dispreferred responses) tend to lead to interactional turbulence.

"Repair" refers to the practices people use to deal with problems in speaking, hearing, and understanding (Schegloff et al. 1977). In (27), Mom and Dad are preparing food with Sesi and some other family members close by. Following a joke, Dad produces extended laughter, in overlap with which Mom asks Sesi to get something, the request infused with a laughter particle. In response, Sesi

### Mark Dingemanse

initiates repair using 'what?' and Mom redoes the recruiting move, providing a more explicit formulation, after which Sesi complies.

(27) Neighbours\_4875900 (Dingemanse 2015: 234)

1 dad həh hɛ hɛ hɛ HA [HA HA HA HA HA HA

▶ 2 mom [Sesi su ɛ(h)ɛh iraɔ̀ tã mɛ PSN take HES thing:INDEF DAT me Sesi take uhuh: the thingy for me 3 ses be: what:Q what 4 mom su kadadìsɛ̃ĩbi bɔ mɛ. take small.pot.DIM bring me take the small pot and bring it to me

▷ 5 ses ((complies by bringing small pot))

An other-initiation of repair starts a side sequence (Jefferson 1972), signaling some trouble that first needs to be resolved before the base sequence can be resumed. A side effect is that the position where a response would be relevant is pushed back at least until the embedded side sequence is closed (in Extract 27, until after line 4). This makes repair initiation a powerful tool that can also be used for secondary purposes (Sacks 1992; Schegloff 1979). Earlier we saw how affirmative verbal responses may claim alignment with the goal of a recruitment, but may also hold off actual fulfillment. In a similar way, repair initiations claim communicative trouble but at the same time can be a device for protracting a sequence and delaying fulfillment (see also Blythe, Chapter 7, §4.2.3).

Consider (28), where Sesi is asked to fetch a bag to go get a load of plantain from a household in a neighboring hamlet. Although Mom's formulation is sufficiently vague to allow Sesi to choose a fitting bag himself ('from inside this thing here', line 2, a reference to a shed nearby), he initiates repair by asking 'what d'you mean bag?'(line 3).<sup>12</sup> The other people present are quick to respond: Aunty taunts 'you'll just go with your bare hands?' and Dad suggests 'your school bag', a suggestion which, after laughs all around, is elaborated by Aunty to reveal the absurdity of Sesi's question (line 8). After this barrage of non-serious responses, Mom's seemingly serious follow-up question (line 9) remains unanswered by Sesi.

<sup>12</sup>The dismissive connotation of the indefinite marker *ɔ* in *bagɔ* is hard to capture in translation. Possible alternatives are 'whatever bag?', 'which bag?', 'what bag?'.

### 10 Recruiting assistance and collaboration: A West-African corpus study

(28) Neighbours\_1131171

```
▶ 1 mom ba su ira ní, ba- ba fe àdi ɛɛ-
         come take thing FP come come pass 2SG:take HES
         come get {the} thing, come come pass {so} you take uh
  2 ɛ bagì i iraɔ amɛ mmɔ ní.
         HES bag LOC thing:INDEF inside there FP
         uh a bag from inside this thing here
▷ 3 ses mmɛ bágɔ:
         which bag:INDEF:Q
         what d'you mean bag?
  4 (0.9)
  5 aun nɛ nrɔ̃-nrɔ̃ aàsɛ[:
         CONJ hand~DIST 2SG:FUT:go:Q
         so you'll just go with your bare hands?
  6 dad [fɔ skúl bagì.
                            3SGPOSS school bag
                            your school bag
  7 ((all laugh together))
  8 aun kɛlɛ adi sìko sɛ si àsu.
         go 2SG:remove books ? LNK 2SG:take
         throw your books out and take it
  9 mom bagì na i ɛɛ ngbe gɔ fɔ ɔse sia áwu sa mmɔ:
         bag lack LOC HES here REL your father put clothes farm there:Q
         there's no bag uh where your dad puts his farming clothes?
  10 bel shuɛ: (.) màkɔkɔ maũ ta madaa kutsùɛ ní.
         INTJ chicken they.TP PROG they:disturb ear FP
         shoo: the chickens are disturbing
  11 (3.0)
```
So here we have a recruiting turn followed by a repair initiation that not all parties to the conversation take entirely seriously as an indication of trouble. What the repair initiation *is* taken as becomes clear later in the interaction, when half a minute has passed and there is still no sign of Sesi fulfilling the request. As (29) shows, Mom pursues a response, upgraded with *mlàmlà* 'quickly' and a final particle *ní* (line 38), implying, as we have seen in §3.3.3, that the recruiting move should have been attended to before. In the continued absence of a response, Aunty observes that 'kids are difficult' and Mom adds 'kids are extremely difficult' in a second-position upgrade that allows her to agree yet also assert her own epistemic access to the matter (Heritage & Raymond 2005). The extract starts 27 turns or 35 seconds after line 10 in (28).

(29) Neighbours\_1131171 (continues from Extract 28)

37 (2.4)

Mark Dingemanse

```
▶ 38 mom bɔ: mlàmlà ní.
          bring IDPH.quickly FP
          bring it quickly now!
  39 (1.0)
  40 aun màbi bɔle.
          children have:force.
          kids are difficult
  41 mom màbi ba ɔle pápápápápápa
          children have force IDPH.extremely
          kids are extremely difficult
```
Mom and Aunty's statements that kids are 'difficult' treat Sesi's troubles in this sequence as related to his teenager status rather than as a true problem in hearing or understanding. In fact, they seem to take Sesi to be exploiting repair in order to delay or even avoid fulfilling a recruitment – a possibility that also puts his behavior in (17) and (27) in a new light.

Repair is not the only way to resist recruitment. Several other ways are illustrated in (30), which continues from (20) above. Three women are chatting together. Aku and Charlotte have asked Emma to watch over a little boy for a moment while his mother is occupied with a task in a neighboring compound. At line 7, Emma ignores the initial recruiting move. Following a response pursuit by Aku, Emma then objects 'I don't know who's picking him up' (line 10), a crafty formulation that enables her to imply that she is unwilling to fulfill the recruitment without going on record as saying so. Aku formulates a high-pitched response pursuit '↑you hear?↑', reasserting the relevance of a response to the request. Following this second pursuit, Emma produces a well-positioned yawn, hearable as a claim of tiredness and by implication inability (line 13). In a final bid to secure compliance, Aku repeats the recruiting move, now adding 'I myself {will do it} when I'm back', thereby trying to overcome Emma's unwillingness by proposing to share the task but also accounting for her own inability to do it immediately.

(30) Maize3\_673020 (continues from Extract 20 above)

```
7 (0.8)
```

```
▶ 8 aku puta ũ (.) anɔ:
        lift him 2SG:hear:Q
        pick him up (.) you hear?
 9 (0.8)
 10 emm lèiye ngɔ toòputa ũ ní
```
1SG:NEG:know REL:who PROG:SCR:lift him FP I don't know who's picking him up *ní*

### 10 Recruiting assistance and collaboration: A West-African corpus study

```
11 (1.0)
▶ 12 aku ↑anɔ:↑
         2SG:hear:Q
         ↑you hear?↑
  12 (0.7)
  13 emm mmmhhh ((yawn))
  14 (1.1)
  15 aku la ũ si lò ba (.) mmɛ nìtɔ si lò ba.
         hold him LNK 1SG come I self LNK 1SG come
         hold him until I'm back (.) I myself {will} when I'm back
```
So we see here that a recruiting move can be resisted by simply ignoring it (line 7), claiming a lack of knowledge as to who should fulfill the recruitment (line 10), or producing a yawn where a response would have been relevant (line 12). Of note is that, throughout, Emma avoids going on record as being unwilling, revealing the lengths to which participants will go to avoid directly rejecting a recruitment (see also Blythe, Chapter 7, §4.2 and §7).

The yawn, a physical display functioning as a claim of unavailability, brings us into the territory of accounts (Heritage 1988), that is, the explanations that often accompany dispreferred responses. Embodied accounts such as Emma's yawn are relatively rare, and special in being off-record. More commonly, accounts are verbal and on-record, as in (13), where Dora asked 'aren't you bringing me water?' and Efi answered 'I'm just going up here, I'll be back', accounting for her failure to fulfill the recruitment by noting a competing commitment.

Yet another way to resist recruitment is to propose another course of action, and by far the rarest way to reject a recruitment is to actually say 'no'. Both of these occur in (31), below. Odo is asked to hold Aku's child for a moment. Other participants include Mercy, a 3-year-old child, Hope, Odo's 9-year-old son, a hairdresser, and her client, both visibly occupied. Even though Aku has already walked up to Odo and is holding up the child before him, Odo declines. He does so using a complex turn format featuring a declination, a reason, and an alternative course of action: 'no, I didn't give birth to the child (.) I'm like ( ), give it to uh' (line 4). The features of this turn are all consistent with what we know about the design of dispreferred responses (Levinson 1983: 334–35; Heritage 1984: 265–66).

(31) Compound5\_737320

1 aku ee, Odoi! voc PSN:DIM hey, little Odo

### Mark Dingemanse

Figure 6: (*a*) Aku (rightmost) approaches Odo (with hand on water drum) holding out her infant (line 3); (*b*) after Odo's refusal, Hope (foreground) is recruited to hold the infant (line 13).


### 10 Recruiting assistance and collaboration: A West-African corpus study

In response to Odo's rejection, Aku starts to formulate a name 'Me-', then self-repairs to Odo's son 'Hope', walking away from Odo and asking Hope to hold the child. Odo meanwhile finishes his word search and says 'Mercy' (line 7), likely the name that Aku abandoned. Odo then begrudgingly volunteers to take the child after all, since he had other plans for his son Hope (line 9), but Hope already comes running towards Aku and Odo. Aku takes issue with Odo's suggestion (lines 11–12) while Hope takes over the child (line 13). The expansion of the sequence after Odo's rejection is typical of what happens after dispreferred responses (Schegloff 2007).

Summing up, how do people resist recruitment? Not without collateral damage to the conversational sequence. They may initiate repair, which has the effect of buying some extra time, but as the side sequence closes a response is still relevant and they are likely to provide it (27), or be held accountable for failing to do so (28). They may try to ignore the recruiting move, but are likely to be held accountable for failing to respond, as in (29) and (30). They can provide a reason (12), propose another course of action, say no outright, or any combination of these things (31), but all of these tend to lead to post-expansion of the sequence (Schegloff 2007: chap. 7).

In short, it seems the deck is firmly stacked against resistance and rejection, and the organization of interactional resources point to fulfillment as the most expedient way to reach sequence closure. This reflects an observation made in some of the earliest work on the organization of preferred/dispreferred actions: such actions 'are both inherently structured and actively used so as to maximize cooperation and affiliation and to minimize conflict in conversational activities' (Atkinson & Heritage 1984: 55).

# **4.3 Acknowledgment in third position**

Sometimes, a two-part recruitment sequence is followed by an expression that has the interactional function of closing the sequence: a "sequence closing third" (Schegloff 2007). An example is given in (32), where Awusi tells Yawa to pour water in a pan with plantain to be put on the fire. While Yawa is pouring, Awusi says *milɛɛ* 'that's good' to indicate that there is now enough water in the pan.This expression is also used when one is poured a drink, to indicate 'this is enough'.

(32) Maize3\_286780

1 awu fore ndu- fore ndu i bɛrɛdzo amɛ. ((points to pan with plantain)) pour water- pour water in plantain pan pour water- pour water in the plaintain {pan}

### Mark Dingemanse

2 yaw ((takes jerrycan, pours water)) 3 awu milɛ:

AGR.N:be.good tha:t's good

One type of sequence closing third that is not attested in the Siwu collection is an acknowledgment like 'thank you'. The simple and immediate practical actions studied here never receive verbal expressions of gratitude in Siwu. Instead, such expressions appear to be reserved for more momentous occasions, for instance when people have spent a day assisting each other with manual labor on the farm or in town. The importance of expressing gratitude in such cases is enshrined in a Siwu greeting routine often heard in the morning: *gu fɔ kɔmakade karabra* 'for your work yesterday', which is answered with *(gu) fɔ kpɛ:* 'and yours'.

The absence of acknowledgments like 'thank you' in everyday recruitments in Siwu stands in contrast with accounts of frequent thanking practices in some other societies (Aston 1995; Becker & Smenner 1986). However, these studies tend to focus on service encounters, which are quite different from the kinds of recruitments studied here (Apte 1974). One crucial difference is that everyday recruitments are almost always*repayable in kind*. The comparative results of the project reported on in this volume suggest that thanking and other ways of verbalizing gratitude are less necessary because of an implicit norm that, where possible, we hold ourselves available and are willing to help others in turn (Floyd et al. 2018), a norm that underlies the web of interdependence and reciprocity in resourcesharing that is typical of human societies (Melis et al. 2016). In contrast, service encounters present an asymmetry: we obtain services or goods that we do not control or produce ourselves, so paying back in kind is harder, which makes it more important to verbally express gratitude.<sup>13</sup>

# **5 Sequential structure and social asymmetries**

# **5.1 A "rule of three" in social interaction?**

Non-minimal sequences amount to a little less than a third of initial recruitments in the core collection (44 out of 146). Most of them are resolved after one pursuit (33 cases); the remaining ones take two pursuits (10 cases) except for one case

<sup>13</sup>Children, like adults in service encounters, are also frequently in the position of not being able to pay back in kind. So perhaps the fact that children are socialized (in some societies) to say 'thank you' and indeed to use more prolix forms in general is a reflection of this asymmetry in agency.

### 10 Recruiting assistance and collaboration: A West-African corpus study

with three pursuits.<sup>14</sup> We see the same in other-initiated repair in Siwu, where non-minimal sequences amount to about a quarter of 153 independent sequences and resolving a single troublesome bit of talk tends to take just one, sometimes two, and rarely more than three other-initiations of repair (Dingemanse 2015).<sup>15</sup> So recruitment and repair usually take only one attempt (as in a minimal sequence), sometimes two, and seldom three or more attempts (Table 7).

Table 7: Distribution of independent sequences of recruitment and other-initiated repair and number of attempts (adult interaction only).


If this pattern proves representative and robust, it may point to a "rule of three" (or a "three strikes" principle) in social interaction: a disruption of progressivity in pursuit of a fitting response rarely takes more than three successive attempts, with a preference for fewer attempts. Research is needed here, starting with the identification of deviant cases, which may reveal to what degree it is a consequence of the structure of complex social action, and to what extent participants orient to it as a socially normative phenomenon.<sup>16</sup> Perhaps the needs addressed in recruitment and repair can overwhelmingly be solved in one go, and the increasingly lower frequency of cases with more than one attempt is in line with an expected probabilistic distribution. Perhaps participants balance intersubjectivity and progressivity (Heritage 2007), and three attempts mark a tipping point where pursuits become too disruptive to overall progressivity. This may also be

<sup>14</sup>The only cases involving more than three attempts are those involving small children. As we will see below, these cases are dissimilar in other ways as well, a key difference being that small children are not held accountable for misunderstandings and failures to response in the same way as other participants.

<sup>15</sup>I am indebted to Nick Enfield for our discussion of this pattern in sequences of other-initiated repair. The general pattern seems to be confirmed even in conversations involving people with Parkinson's disease, where one might expect more protracted sequences of other-initiated repair (Griffiths et al. 2015). In his discussion of self-repair, Schegloff notes that "[a]lthough not common, two successive repairs on a same repairable, yielding (together with the repairable) three tries at that bit of talk, are not rare" (Schegloff 1979: 277).

<sup>16</sup>An indication that a "rule of three" may relate specifically to disruptions of progressivity (as opposed to being a general limit on repeated behavior) is that multiples of successful recruitment sequences in close succession do occur, as in (21), which features at least six nonverbal requests and responses.

### Mark Dingemanse

a fruitful area for cross-species comparison (cf. Wilkinson et al. 2012 on repeated requests for meat sharing among chimpanzees), linking to a more general theme of communicative persistence.

# **5.2 Social asymmetries**

An interest in social asymmetries has long been a prominent feature of crosslinguistic studies of requests (Brown & Levinson 1978; Blum-Kulka et al. 1989). On the basis of this literature, one might expect the organization of assistance in interaction to be influenced by social asymmetries, such that, for instance, the selection of one format over another, or the nature of responsive actions, would differ depending on the relative social status of participants.

There is one large set of recruitments where social asymmetries clearly play a role: those involving small children as recruitees (recall that these were collected separately from the 207 cases that make up the core collection of Siwu recruitments, §1.2). The following extract is from a multiparty conversation in which a mother asks her toddler, less than 2 years of age, to come to her. The sequence involves six pursuits until compliance in line 11.

```
(33) Cooking1_93710
▶ 1 mom Sise (.) ba.
         Sise come.
         Sise (.) come
  2 (0.8)
  3 mom ba.
         come.
         come
  4 (0.4)
  5 esi mama sɔ ba.
         mother QT come
         Mom says come
  6 mom ↑ba:↑
         come
         ↑co:me↑
  7 (0.6)
  8 esi ↑ma↑ma sɔ ↑ba↑
         mother QT come
         ↑mom↑ says ↑come↑
  9 (1.2)
  10 esi ↑MA↑MA SƆ BA
         mother QT come
         MOM SAYS COME
```
### 10 Recruiting assistance and collaboration: A West-African corpus study

```
▷ 11 chi ((turns and walks towards mother))
  12 ama ɔ nyɔ nɛ yaa.
          3SG watch it IDPH.absently
          he was just staring yaa ((absently))
  13 mom ((holds up underpants))
  14 chi ((steps into underpants))
```
This sequence differs in several ways from those considered so far. The number of pursuits appears to flout the "rule of three" (though none of the participants individually puts in more than three attempts). The pursuits are all simple repetitions with few changes except in prosody, in stark contrast with other pursuits we saw earlier which involve reformulations and reason-giving. Despite many pursuits, the child does not provide any form of response until the nonverbal action in line 11, and there is no evidence here that the child has mastery of strategies like repair initiation or other practices that others use in non-minimal sequences. Whereas recruiting and responding moves usually tend to be taken as a matter between recruiter and recruitee, here two other participants join in pursuing a response (lines 5, 8, 10), and a third provides an account for the lack of response of the child (line 12), showing that its absence is seen as accountable while at the same time implying that the child cannot (yet) speak for itself.

Combined, these observations suggest that child recruitees are treated differently. They are treated as still having to learn how to respond to recruiting moves, and they are not held accountable for their interactional conduct and for possible troubles in understanding in the same way that other participants typically are. While it may be tempting to say the child is treated this way *because* of a social asymmetry, it is at least as plausible to say that cases like this show how social asymmetries are socially constructed and reinforced. The sequence is a socialization routine as much as an attempt to get the child to do something.

Social asymmetries also surface in sequences other than those involving very young children. Particularly telling of the social construction of asymmetries are moments when participants orient to them. Recall some of the turbulent sequences involving Sesi – a teenager – and his parents and alloparents. When, in (29), Sesi's aunt and mom note that 'kids are difficult', they invoke the category of *kids*, which forms a contrast set with *adults*, to make a complaint about Sesi's unwillingness. It may be a universal feature of teenage behavior to try and find ways to escape household chores. Likewise, it may be a universal feature of caregiver talk to complain about this. That is one way in which social asymmetries can become tangible in interaction.

### Mark Dingemanse

Although I have focused so far on evidence for social asymmetries in the moment-by-moment unfolding of the interaction, such social asymmetries do not emerge out of nothing. Knowledge about social membership categories and kinship relations is usually available, or at least assumed to be available, to participants in interaction (Terkourafi 2005; Enfield 2013), and so these categories and relations may also influence social interactions without being explicitly oriented to in talk. The most relevant durable social asymmetries for Siwu speakers are grounded in a combination of age and kinship relations. Older age generally comes with higher social status, and kinship structure provides a framework for allocating rights and duties (such that parents and alloparents can exercise deontic authority over younger kin). Based on this, most recruitment sequences in the corpus can be classified as involving a dyad that is either: (i) symmetrical with A and B having approximately the same social status, or (ii) asymmetrical with A higher in status than B, or (iii) asymmetrical with A lower in status than B (Table 8).<sup>17</sup>


Table 8: Social asymmetry of participants in 146 independent recruitment sequences.

For a large majority of participants in recruitment sequences, there is no evidence of a social asymmetry between them, reflecting the fact that a lot of everyday social interaction in the corpus is between peers. In about one fifth of cases, participant A can be considered higher in social status than participant B; most commonly, these are cases where parents or alloparents address younger people in the household. In contrast, there are only 6 cases where participant A is clearly lower in social status than participant B. The relative paucity of such cases suggests that people may be somewhat less likely to recruit the assistance or collaboration of others who are higher in social status – possibly as a way to avoid resistance, rejection, or other types of interactional turbulence (Brown & Levinson 1978; see also Floyd, Chapter 3, §6; Enfield, Chapter 6, §6; Baranova,

<sup>17</sup>For 18 cases, it was not possible to assess this with sufficient confidence.

### 10 Recruiting assistance and collaboration: A West-African corpus study

Chapter 9, §6). So social asymmetries may influence how likely people are to recruit assistance or collaboration from others.

Do social asymmetries also influence matters of formulation or format selection? An analysis of the core collection of recruitment sequences provides little evidence that social asymmetry (as operationalized here) is a decisive factor in format selection or in the design of responsive actions.<sup>18</sup> Instead, as we have seen throughout this study, many matters of formulation and selection appear to be more directly affected by local factors such as establishment of joint attention, relation to ongoing activities, and sequential position as initial or subsequent. This fits a recurring theme in systematic comparative work on informal conversation: micro-scale local factors like attention, participation framework, and sequential position seem more directly consequential than macro-sociological factors like social status, power, or politeness.

# **6 Conclusions**

The domain of recruitments provides a microcosm of how linguistic resources combine with bodily conduct and adapt to social interaction. Malinowski, observing everyday social interaction on the Trobiand Islands a century ago, noted that "the structure of all this linguistic material is inextricably mixed up with, and dependent upon, the course of the activity in which the utterances are embedded" (Malinowski 1923: 311). Recruitments provide a privileged locus for observing this intertwining of speech and action.

Some of the resources used in recruitment sequences bear a language-specific signature. For instance, Siwu makes available a *si*-prefaced format to mark consecutive actions in larger projects, and final particles like *ló* and *ní* for mitigating and strengthening recruiting moves. But beneath the language-specific resources, the recruitment system appears to be fundamentally cut from the same cloth across languages and cultures. Recruiting and responding moves are adapted to

<sup>18</sup>The following elements of format design and selection did not seem to be affected by the absence, presence, or direction of social asymmetry: type of recruitment (object transfer versus service); verbal or nonverbal means for recruitment; construction types (imperative, interrogative, declarative, *si*-prefaced); presence or absence of an account in the recruiting turn; use of mitigating or strengthening devices; relative frequency of fulfillment versus resistance or repair; presence or absence of an account in the response. For three variables, there are not enough cases in the collection to draw firm conclusions about a possible role for social asymmetries: the relative frequency of recruitments to alter an ongoing trajectory of behavior; the relative frequency of assistance prompted by current or anticipatable trouble; and the relative frequency of resistance and rejection.

### Mark Dingemanse

recurrent interactional challenges, from calibrating joint commitments to specifying practical actions and managing activity structure. The Siwu recruitment system appears to be one instantiation of a sophisticated machinery for organizing collaborative action that transcends language and culture.

# **Transcription conventions and abbreviations**

Conversational transcripts follow the conventions developed by Jefferson (2004). In addition, words in free translations with no direct equivalent in the original material are {marked so}. Interlinear glosses follow the Leipzig glossing rules (Comrie et al. 2020) with the following additions: cont continuer • fp final particle • hes hesitation marker • ing ingressive • lnk linker • o object marker • plur pluractional reduplication • psn person name • scr subject cross-reference marker. Conflicts between conversation analytic conventions and Leipzig glossing rules (e.g. marking of self-repair vs. morpheme breaks using dashes) are resolved in favor of the former.

# **Acknowledgments**

Thanks to Rev. Wurapa and Ruben and Ella Owiafe for welcoming me into the Siwu-speaking community, and to Ɔdimɛ Kanairo for assistance in transcription and translation. *Mi ndo karabra ló!* Thanks to the other members of the Recruitments project, especially Simeon Floyd, Giovanni Rossi, and Nick Enfield for feedback and guidance; to Felix Ameka for helpful comments; to Steve Levinson for fostering a department at MPI Nijmegen where the project could blossom; and to Herb and Eve Clark for hosting me at Stanford and providing the best writing environment. Sharon Rose helped me clear up shooing words in Gurage. The fieldwork and data collection underlying the study was funded by ERC Starting grant 240853 to N. J. Enfield. The author is supported by NWO Talent grant 016.vidi.185.205.

# **References**


10 Recruiting assistance and collaboration: A West-African corpus study


Breugel, Seino. 2014. *A grammar of Atong*. Leiden: Brill.


### Mark Dingemanse


Enfield, N. J. 2007. *A grammar of Lao*. Berlin: Mouton de Gruyter.


10 Recruiting assistance and collaboration: A West-African corpus study


### Mark Dingemanse


10 Recruiting assistance and collaboration: A West-African corpus study

trajectories and sequences of actions. *Language Sciences* 58. 8–21. DOI:10.1016/j.langsci.2016.02.008


### Mark Dingemanse


10 Recruiting assistance and collaboration: A West-African corpus study

Valdman, Albert & Kevin J. Rottet. 2009. *Dictionary of Louisiana French: As spoken in Cajun, Creole, and American Indian communities*. Jackson: University Press of Mississippi.

van den Berg, René. 1989. *A grammar of the Muna language*. Dordrecht: Foris.


Aikhenvald, Alexandra Y., 2, 64, 148, 319 Altschuler, Milton, 82 Ameka, Felix K., 371, 388, 396, 397 Amha, Azeb, 397 Antaki, Charles, 177 Apte, Mahadev L., 81, 408 Arcidiacono, Francesco, 149 Aston, Guy, 408 Atkinson, J. Maxwell, 130, 407 Austin, J. L., 2, 138, 221, 369 Bąk, Piotr, 282 Baranova, Julija, 74, 130, 131, 177, 178, 305, 326, 330, 356, 390 Barrett, Samuel Alfred, 82 Barth-Weingarten, Dagmar, 194 Becker, Judith A., 408 Benacchio, Rosanna, 291, 343 Bickel, Balthasar, 54 Blum-Kulka, Shoshana, 3, 10, 178, 410 Blythe, Joe, 232, 236, 258, 263, 273, 397 Bodman, Jean W., 81 Boehm, Christopher, 268 Bolden, Galina B., 55, 102, 177, 194, 301, 315, 326, 334, 342, 346, 364 Bolton, H. Carrington, 397 Breugel, Seino, 397 Brown, Penelope, 2, 73, 81, 83, 115,

129, 176, 178, 179, 184, 192, 320, 354, 362, 370, 399, 410, 412 Bruil, Martine, 52, 70 Butler, Carly W., 129 Bynon, James, 396, 397 Caldwell, Christine A., 397 Cameron-Faulkner, Thea, 61 Campisi, Emanuela, 163, 192 Caudal, Patrick, 232, 242 Cekaite, Asta, 370 Childs, Carrie, 124, 125 Clark, Herbert H., 2, 3, 157, 158, 192, 370, 379 Clayman, Steven E., 129, 355 Comrie, Bernard, 194, 283, 327, 414 Corsaro, William A., 362 Coulthard, Malcolm, 3 Couper-Kuhlen, Elizabeth, 3, 6, 67, 69, 105, 114, 120, 127, 194, 204, 270, 301, 369, 379, 380, 384 Craven, Alexandra, 3, 61, 83, 120, 181, 182, 251, 290, 313 Creissels, Denis, 54 Cui, Xuebo, 81 Curl, Traci S., 3, 8, 28, 61, 83, 115, 120, 137, 209, 251, 349, 370, 379 Davidson, Judy, 130, 331 Dawkins, Richard, 396

De Ruiter, Jan Peter, 252 DeBoer, Warren, 52 DeLancey, Scott, 54 Deppermann, Arnulf, 291, 293, 302 Dickinson, Connie, 54, 64 Dingemanse, Mark, 3, 7,10–13,15, 26, 52, 60, 74, 130, 131, 177, 178, 305, 326, 356, 371–373, 390, 397, 402, 409 Dixon, R. M. W., 2 Dixon, Sally, 362 Drew, Paul, 3, 6, 8, 10, 61, 68, 83, 93– 95, 99–102,104,105,115,120, 125, 130–132, 137–139, 149, 177, 204, 251, 252, 349, 362, 369, 370 Egbert, Maria M., 10 Eisenstein, Miriam, 81 Endo, Tomoko, 362 Enfield, N. J., 5–8, 10–12, 15, 26, 60, 158–160, 162, 192, 204, 206, 209, 211, 216, 225, 234, 239, 251, 370, 379, 397, 412 Ervin-Tripp, Susan M., 3, 4, 176, 182 Etelämäki, Marja, 69 Evans, Nicholas, 66, 388 Fanshel, David, 221, 390 Farashaiyan, Atieh, 81 Fasulo, Alessandra, 149 Fatigante, Marilena, 148 Félix-Brasdefer, Julio César, 10 Fillmore, Charles, 37 Fischer, Gloria J., 398 Flanagan, James G., 268 Floyd, Simeon, 1, 6, 7, 52, 54, 60, 64, 66, 68, 76, 81, 82, 85,100,132,

137, 187, 193, 204, 268, 317, 362, 372, 373, 408 Ford, Cecilia E., 388 Ford, Kevin, 371 Forshaw, William, 232 Forsyth, James, 291 Fox, Barbara A., 3, 10, 67, 115, 136, 172, 252, 380 Frake, Charles O., 369 Francik, Ellen P., 2 Fukushima, Saeko, 370 Galatolo, Renata, 149 Galeano, Giorgia, 149 Gardner, Rod, 259 Garfinkel, Harold, 96 Garvey, Catherine, 362 Gerrard, Grayson, 273 Gibbs, Raymond W., 2 Gili Fivela, Barbara, 148, 169 Goffman, Erving, 2, 7, 81, 100, 239 Goodwin, Charles, 94, 177, 379 Goodwin, Marjorie Harness, 94, 102, 120, 130, 181, 290, 370 Gordon, David, 2, 3 Green, Ian, 232 Grice, H. Paul, 2, 157, 159, 192 Griffiths, Sarah, 409 Haddington, Pentti, 133 Hale, Austin, 54 Haspelmath, Martin, 9 Hawkins, John A., 326 Heinemann, Trine, 3, 10, 83, 115, 136, 370, 380, 384 Hepburn, Alexa, 194 Heritage, John, 115, 124, 129, 130, 132, 137, 177, 184, 256, 262, 319,

362, 390, 396, 403, 405, 407, 409 Hinkel, Eli, 81 Hoey, Elliott M., 390 House, Juliane, 10 Houtkoop-Steenstra, Hanneke, 130 Hua, Tan Kim, 81 Intachakra, Songthama, 81 Jefferson, Gail, 55, 101, 194, 283, 402 Jijón y Caama, Jacinto, 52 Kasper, Gabriele, 10 Keisanen, Tiina, 75, 116, 312 Kelly, Barbara, 232 Kendon, Adam, 153, 234, 239, 266 Kendrick, Kobin H., 6, 8, 93–95, 99– 105, 107, 114, 119, 121, 123, 125, 131, 137–139, 193, 204, 252, 320 Kent, Alexandra, 102, 107, 114, 119– 121, 123, 181, 182, 255, 291, 320, 399, 401 Khrakovskij, Viktor S., 64 Kitzinger, Celia, 133, 314 Klamer, Margaretha Anna Flora, 397 Koivisto, Aino, 10 König, Ekkehard, 38, 62, 114, 148, 170 Krebs, John R., 396 Królak, Emilia, 302 Kropp Dakubu, Mary Esther, 371 Kruspe, Nicole, 397 Labocha, Janina, 282 Labov, William, 221, 390 Lakoff, George, 2, 129 Lambrecht, Knud, 159 Le Pair, Rob, 10

Lee, Seung-Hee, 252

Lee-Wong, Song Mei, 10 Lehmann, Volkmar, 291 Lerner, Gene H., 36, 69, 129, 133, 239, 242, 250, 252, 287, 305, 311 Leslau, Wolf, 397 Levinson, Stephen C., 2, 15, 35, 67, 73, 81, 83, 114, 115, 129, 157, 159, 176, 178, 179, 184, 192, 239, 272, 320, 354, 362, 370, 399, 405, 410, 412 Lewis, M. Paul, 52, 148 Lindskoog, Carrie A., 52 Lindskoog, John N., 52 Lindström, Anna, 3, 4 Lindström, Jan, 388 Liston, Jerry L., 397 Liu, Yi-Ping, 397 Lucy, Peter, 2 Lusini, Sara, 169 Lyons, John, 38, 150, 242 MacKenzie, J. G., 398 Maiden, Martin, 169 Majid, A, 32 Malinowski, Bronislaw, 369, 413 Mandelbaum, Jenny, 123 Manrique, Elizabeth E. M., 379 Mansfield, John, 232, 269 Margutti, Piera, 149 Márquez-Reiter, Rosina, 10 Masur, Elise Frank, 61 Matisoff, James A., 397 McConvell, Patrick, 243 Melis, Alicia P., 408 Meyer, Christian, 371 Miller, Mark T., 397 Mondada, Lorenza, 101, 133, 139, 166 Monzoni, Chiara M., 149 Moore, Bruce R., 52

Mortari, Luigina, 149 Mujkic, Elena, 232 Müller, Neele Janna, 69 Müller, Rachel A. G., 2 Mushin, Ilana, 259 Norcliffe, Elisabeth, 52, 54 Nordlinger, Rachel, 232, 242 Obeng, Samuel Gyasi, 371 Ogiermann, Eva, 3, 10, 69, 120, 129, 171, 174, 186, 294, 296, 300, 315, 334, 353, 355, 364, 386 Ohashi, Jun, 81 Oswalt, Robert L., 397 Özdemir, Çiğdem, 81 Palmer, F. R., 114 Pareto, Vilfredo, 227 Parry, Mair, 169 Parry, Ruth, 74, 130, 174, 177, 356 Pauletto, Franco, 149 Pedersen, Jan, 81 Peräkylä, Anssi, 120 Peterson, Elizabeth, 10 Peterson, Nick, 268, 273 Pino, Marco, 149 Poggi, Isabella, 153 Pomerantz, Anita, 55, 102, 125, 256, 331, 401 Pontecorvo, Clotilde, 149 Post, Mark W., 54 Potter, Jonathan, 3, 61, 83, 120, 181, 182, 251, 290, 313 Raevaara, Liisa, 130, 166, 174 Rauniomaa, Mirka, 75, 116, 312 Raymond, Geoffrey, 80, 115, 124, 125, 133, 161, 182, 252, 262, 313, 403

Rendle-Short, Johanna, 129 Rezvani, Seyed Ali, 81 Robinson, Jeffrey D., 130, 157, 177, 315, 326 Rossano, Federico, 3, 10, 148, 149, 169, 175, 252, 396 Rossi, Giovanni, 4, 6, 8, 10, 35, 63, 67, 69, 70, 80, 81, 109, 115, 120, 121, 125, 126, 129, 148, 149, 156, 161, 165, 167–172, 174– 177, 179, 181–183, 186, 204, 222, 252, 287, 291, 300, 311, 313, 314, 334, 336, 342, 352, 370, 393, 395, 399 Rottet, Kevin J., 397 Rudnicka, Kinga, 302 Rue, Yong-Ju, 10 Sacks, Harvey, 96, 102, 129, 132, 137, 150, 270, 402 Sadock, Jerrold M., 2, 38,114,148,150, 242 San Roque, Lila, 54 Schegloff, Emanuel A., 7, 15, 29, 36, 55, 67, 70, 81, 99, 102, 114, 126, 129, 130, 132, 135, 137, 150, 159, 182, 184, 185, 190, 193, 252, 255, 256, 262, 270, 291, 314, 330, 362, 386, 390, 393, 401, 402, 407, 409 Schneider, Stefan, 129 Schunk, D. H., 2, 3 Schutz, Alfred, 389 Searle, John R., 2, 3, 67, 115, 138, 182, 221, 369 Sherzer, Joel, 60 Sidnell, Jack, 10, 15, 27, 67, 161, 377 Siemund, Peter, 38, 62, 114, 148, 170 Sifianou, Maria, 10

Sinclair, John McHardy, 3 Skinner, B. F., 398 Smenner, Patricia C., 408 Sorjonen, Marja-Leena, 166, 291 Spottiswoode, Claire N., 396 Steensig, Jakob, 10 Sterponi, Laura, 148 Stevanovic, Melisa, 120, 124 Stivers, Tanya, 3, 10, 12, 15, 26, 31, 32, 67, 76, 128, 175, 184, 252, 258, 285 Streeck, Jürgen, 107 Street, Chester S., 232 Strutyński, Janusz, 282 Szymanski, Margaret H., 3, 125, 183, 251, 399 Takada, Akira, 362 Terkourafi, Marina, 412 Thompson, Sandra A., 127, 252, 270, 313, 380, 388 Toerien, Merran, 133 Tomasello, Michael, 319, 372 Tomlin, Russell S., 326 Tsuzuki, Masako, 10 Valdman, Albert, 397 van den Berg, René, 397 Van der Auwera, Johan, 289 Vinkhuyzen, Erik, 3, 125, 183, 251, 399 Vittadello, Alberto, 52 Walsh, Michael J., 232, 236 Waring, Hansun Zhang, 177, 356 Warneken, Felix, 372 Watanabe, Honoré, 66 Weidner, Matylda, 289, 291, 296, 299, 301

Weizman, Elda, 176 Wierzbicka, Anna, 3, 182, 302 Wilkins, David P., 33, 36 Wilkinson, Ray, 410 Woodfield, Helen, 10 Wootton, Anthony J., 3, 4, 120, 121, 181, 182, 251, 290, 291, 300 Zhang, Grace Qiao, 10 Zimmerman, Don H., 373 Zinken, Jörg, 3, 6, 8, 10, 69, 115, 120, 125, 129, 149, 171, 174, 175, 183, 186, 187, 190, 193, 204, 284, 291–294, 300, 311, 313, 315, 318, 342, 347, 353, 364, 399 Zipf, George Kingsley, 227 Zwicky, Arnold, 2, 38, 114, 148, 150, 242

# Getting others to do things

Getting others to do things is a central part of social interaction in any human society. Language is our main tool for this purpose. In this book, we show that sequences of interaction in which one person's behavior solicits or occasions another's assistance or collaboration share common structural properties that provide a basis for the systematic comparison of this domain across languages. The goal of this comparison is to uncover similarities and differences in how language and other conduct are used in carrying out social action around the world, including different kinds of requests, orders, suggestions, and other actions brought together under the rubric of recruitment.