Toward a Linguistic Anthropological Account of Deixis in Interaction: Ini and Itu in Indonesian Conversation Colorado Research in Linguistics. June 2009. Vol. 22. Boulder: University of Colorado. © 2009 by [Author full names as they appear in the author line]. Toward a Linguistic Anthropological Account of Deixis in Interaction: Ini and Itu in Indonesian Conversation Nicholas Williams University of Colorado at Boulder This paper presents an overview of research on deixis in linguistic anthropology. In line with other recent deixis theorists (e.g. Hanks), I suggest that deixis has not yet received sufficient theoretical nor empirical attention. I argue for the centrality of deixis, and demonstrative reference in particular, to an understanding of the fundamentally social and interactional nature of linguistic meaning. As an exercise in the analysis of deixis in interaction, I analyze the use of two nominal demonstratives (ini and itu) in colloquial Indonesian conversation. These demonstratives occur in what are known as “placeholder uses,” frequently in the context of a “word search.” Several instances of placeholder demonstrative use are analyzed, showing that differing types of “access” (perceptual, cognitive, social) (Hanks 2009) to the referent, as well as distinct indexical grounds, are what distinguish the meaning and use of these two demonstratives. These findings point to the importance of interactional data in the analysis of basic linguistic meaning. 1. Introduction: Indexicality, deixis and a theory of linguistic “meaning” What pervades all work in linguistic anthropology, from Boas to Silverstein and beyond, is the attempt to articulate the role of language with respect to culture and to understand what linguistic forms and/or practices mean (all problematic terms, as it turns out). What do linguistic forms/practices mean and how is this reflective of or reflected in the culture? This central problem has been approached in a number of ways over the years. Early linguists such as Boas and Greenberg worked from the “language is culture” approach, using recently developed conceptual tools (“phonemes,” “morphemes” and “syntax”) to describe the grammatical structure of “exotic” languages, particularly those native to North America. Based on the assumption that “language is culture”, these linguists masqueraded as anthropologists, “claim[ing] to be doing something anthropological by analyzing grammar” (Duranti 2003). However, their focus, and the focus of the long tradition of structural and generative linguists to follow, was not so much cultural meaning in the anthropological sense, but rather a far narrower conception of linguistic meaning or semantics as propositional content and truth-values. This “semantico-referential” approach to meaning takes the “symbol” as its prime object of interest. Symbols are characterized by their arbitrary association of the “signified” with the “signifier.” Symbols or pairings of signifier and signified (later “form-meaning pairs”) are pervasive in language and, in much of linguistics, typically assumed to be the unique characteristic of language that sets it apart from other semiotic systems. 1 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 2 In the groundbreaking 1976 paper “Shifters, Linguistic Categories and Cultural Description,” Michael Silverstein set an agenda for linguistic anthropologists by redefining “meaning” in anthropological inquiry. Building on the work of Peirce and others in semiotics, Silverstein provides a set of tools for understanding the role of language in society, with particular reference to how low-level linguistic forms and micro-practices (or “speech events”) are linked to “culture” or larger, macro-level processes of meaning-making. This perspective is presented in stark contrast to both the traditional semantico-referential approach and to the Austinian “speech acts” approach. Silverstein finds fault with the traditional semantic approach to meaning for its narrow focus on symbols, “traditionally spoken of as the fundamental kind of linguistic entity” (p. 27). If we are to understand language as a vehicle for the expression of cultural meanings, we will have to acknowledge additional sign types other than the symbol, including the icon and the index. While Austin’s and related approaches to “speech acts” and language use are helpful in refocusing our attention on the social life of language and away from a narrow view of propositional meaning, it is flawed in that it assumes a basic level of semantic-referential structure. Onto this basic layer of meaning, philosophers like Austin and Searle “tack on” the “performative use” of these basic linguistic categories. According to Silverstein, this entirely misses the point that reference is itself a performative act. There is nothing done in language that is not performative. If anything, reference is a relatively marginal type of action performed through the use of language (but with help from other semiotic systems such as gesture). This old critique bears reconsideration in light of much recent work in “interactional linguistics” which discusses the interactional “use” of various linguistic and/or grammatical “resources” to perform “actions” in conversation. We will leave this issue aside for the moment to examine the rest of Silverstein’s theory. However, the notion of reference as a performative speech act will be important later on and we will return to it. As an alternative to this traditional, symbol-obsessed, semantico- referential and proposition-based mode of analysis, Silverstein proposes an approach to meaning in language that focuses on the index, “those signs where the occurrence of a sign vehicle token bears a connection of understood spatio- temporal contiguity to the occurrence of the entity signaled” (p. 27), the meaning of which “always involves some aspect of the context in which the sign occurs” (p. 11). As we will see, this understanding of the index allows for a reinterpretation of all kinds of speech events (including referential speech events) as indexical of some element(s) of the “context in which they occur.” For Silverstein, this refocusing will ultimately lead us to the “most important aspect of the ‘meaning’ of speech” (p. 12). In addition to broadening our attention from the semantico-referential to the larger more encompassing field of indexical “meaning” in “culture,” what we gain from Silverstein’s approach is a far more nuanced understanding of the index itself, that most important tool for understanding links between language and 2 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 3 context. Using the cross-cutting dimensions of presupposing  creative and referential  non-referential, four distinct types of indexes are identified and discussed. The four types are summarized in the following chart, from Silverstein (1976): Of particular interest to Silverstein (1976, 2003) and much later scholarship that draws on the notion of indexicality (e.g. Ochs 1990, 1992, Inoue 2004, Bucholtz 2011, many others) are the “non-referential” indexes, especially those non-referential “creative (performative)” ones. Ochs’ (1990) article further developed this concept of the non-referential performative index, adding an additional dimension, direct/indirect. Sharing some similarity to Silverstein’s (2003) indexical order, this notion of direct/indirect indexes helped further our understanding of the link between performed social identities (e.g. gender) and linguistic practices (e.g. Japanese sentence-final particles). As Ochs’ pointed out, it is not the case that sentence-final particles in Japanese directly index being-a- woman or her femininity, but rather that such particles index an “affective stance,” which then indexes a type of “female voice” in Japanese speech. In this way Japanese sentence-final particles indirectly index gender (conceived of as a performed and socially constructed category of identity). This work is foundational to our contemporary understanding of the links between macro-level social categories and micro-level linguistic practices. However, this is just one direction to take Silverstein’s ideas. In the linguistic anthropological literature of the “third paradigm” (Duranti 2003), less attention has been given to the referential side of Silverstein’s diagram, and even less to the referential presupposing top left corner. What we find in this area are issues typically left to linguists, taken up by semanticists and sometimes pragmaticists, or those scholars probably categorizable in Duranti’s (2003) second paradigm. The most notable and well known example of “referential presupposing” indexes is “deixis.” Variously defined by different authors, “deixis” for Silverstein seems to mean “spatio-temporal deixis,” mainly demonstratives and tense. As he describes it, deixis is “maximally presupposing, in that the contextual conditions are required in some appropriate configuration for proper indexical reference,” and “some aspect of the context … is fixed and presupposed” (p. 34). When using a “deictic” expression (e.g. English this or that, 3 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 4 here or there), speakers make reference to cognitively or perceptually “accessible” (Hanks 1990) objects (real or abstract). For Silverstein, such accessible objects “exist” for both speaker and addressee and are thus “presupposed,” “otherwise the use of the deictic token is inappropriate” (p. 33). Since Silverstein’s landmark account of “shifters” and indexicality as crucial parts of the theory of cultural meaning making, a fair amount of progress has been made on that neglected upper left corner of the index diagram. In particular, William Hanks (1990, 1992, passim) has argued for a more complex understanding of deictic reference that takes into account not only the object of reference (Hanks’ “denotatum”), but also the “indexical ground” and the relation between the two. For Hanks and others (e.g. Agha 1996), use of deixis is not necessarily presupposing, but rather constitutive of the interactional context of the utterance. Agha, for instance, argues, “Deictic spatialization effects are not the outcome of ‘coding’ relationships between deictic categories and preexisting spatial realities” (p. 679). That is, deictic expressions do not necessarily (en)code presupposed notions of spatial reality, but rather might be considered more “creative” (in the sense of Silverstein 1976). As Agha puts it, “deictic usage indexically situates spatial representations in relation to contextual variables whose values are only specified during the course of discursive interaction” (my emphasis). That is, what is presupposed in an act of deictic reference is itself only specified within an instance of situated (talk-in-)interaction, typically through “co(n)textual superposition.” In the remainder of this paper I will further explore the category of deixis as it fits into Silverstein’s broader theory of indexicality. Rather than relegate it to “traditional linguists” who are likely to give it a semantico-referential treatment, I argue that (spatial) deixis is a crucial feature of situated language use in interaction that requires serious consideration by (linguistic) anthropologists. Deictics play a crucial role in connecting instances of language use to the immediate and broader context. Deictic reference is accomplished through co- production with other speakers and is accompanied by use of multi-modal signs (gestures, eye gaze), which are often crucial for interpretation. Through the use of data from video recorded conversation in Indonesian, I will argue that deictic practice works to constitute (rather than simply reflect or encode spatial aspects of) the immediate context of interaction. I will also show that, in line with Hanks (1990), what is fundamental to deictic forms is “access” rather than any spatial notion. Finally, I will suggest additional pragmatic effects of deictic use in interaction (in Indonesian) that might be interpretable with reference to notions of indexicality and stance (Du Bois 2007). In this way I aim to situate deictic “referential practices” (Hanks 1990) firmly within a linguistic anthropological approach to language and meaning. 2. Approaches to deixis 4 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 5 Deixis has suffered a history of marginalization in the study of language, alternately being handed off from philosophy to semantics, taken up by pragmatics and occasionally linguistic anthropologists. As a phenomenon it has been largely ignored by most generative and descriptive-typological linguists alike, getting only a cursory treatment in the majority of reference grammars. This is not to say that deixis is lacking a significant literature, which it is not. The tradition of studies of deixis dates back to Buhler (1934), at least, and has been followed up by Lyons (1977, 1982), Fillmore (1982), Weissenborn and Klein (1982), Levinson (1983, 1994, passim), and Anderson and Keenan (1985), among others. These studies represent mostly philosophical as well as semantic and/or pragmatic approaches to deixis and deictic use. Even by 1983, as Levinson states in his Pragmatics textbook, theoretical approaches to deixis were underdeveloped. Since the publication of that textbook, and particularly in the past two decades (1990 – 2010), there has been somewhat of a surge in studies of deixis. This new literature has come from a number of fields, in particular linguistic anthropology (Hanks 1990, 1992, 2005, 2009, etc., Agha 1996), conversational analysis and interactional linguistics (M. Goodwin 1990, C. Goodwin 1999a, 1999b) and cognitive psycholinguistic approaches (Levinson 1996). Additionally, more traditional descriptive semantico-referential and typological approaches have produced accounts of deictics, particularly demonstratives (Anderson and Keenan 1985, Himmelmann 1996, Diessel 1999, Dixon 2003). A recent overview of the this literature is found in Sidnell (1998). Schegloff’s (1972) paper on “formulating place” should also be mentioned here, as it informed some of the work on deixis in interaction, although it does not address deixis directly. While in the past few years a number of linguists have published studies of demonstratives and/or deixis in particular languages that draw on these interactionally and ethnographically informed approaches (e.g. Bickel 1997, Enfield 2003a, 2003b, Hayashi and Yoon 2006), other work is continuing to be produced which takes the old, simply semantic and problematic view of deixis criticized by Hanks (1990, 2009) among others. It will be useful now to briefly summarize past approaches and previous understandings of deixis, as well as the anthropological and interactional critique. This will set us up to look closely at some examples of Indonesian deictic demonstrative use in interaction. Traditional approaches to deixis are mainly descriptive and typological in nature. Fillmore and Lyons first discussed deixis from a semantico-referential perspective and contributed to our basic understanding of deixis as a phenomenon distinct from context-referring phenomena such as anaphora and indexicality. More recently, the more “traditional” descriptive-typological approach has been interested in cataloguing the cross-linguistic variation in deictic forms and functions, delineating parameters for universal and language-specific features of deictic systems. This work is exemplified in Weissenborn and Klein (1982), Anderson and Keenan (1985), Himmelmann (1996), Diessel (1999) and Dixon (2003), among others. This work has increased our understanding of what can be encoded in deictic systems in a diverse set of languages, thus providing a base 5 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 6 from which to further investigate the everyday use of such forms. Himmelmann (1996) is a good example of a study of the functions of diverse deictic (demonstrative) systems across a number of languages. However, there is reason to believe that this well-developed typology of deixis, as well as the “functional” approach to the “use” of these “forms,” is problematic. If we recall Silverstein’s (1976) warning, it is not the case that certain semantico-referential aspects of language “structure” are already there, to be left to the formal semanticists and descriptive linguists to discuss, and that these “resources” are then put to “use” in interaction. Instead, deictic reference is just one type of language use, one type of performative speech act. While it is true that deictic practices involve certain pragmatic effects, it is misunderstanding the phenomenon to separate the basic “meaning” of the forms from their “use in interaction.” Enfield (2003a) has shown very clearly that video data of situated talk-in-interaction is crucial for an accurate analysis of the semantically encoded aspects of demonstratives’ meanings in Lao. In the case of Lao, it turns out that what is “encoded” semantically is not the expected “proximal” vs. “distal” distinction, but rather a more basic notion of NOT HERE for the “distal” form, while the “proximal” form has no semantically encoded meaning. Agha (1996) also cautions us against relying on seemingly intuitive and straightforward notions of spatial meanings “encoded” in deictic forms, arguing instead that deictics project “spatial schemas” which are only interpreted and achieve “spatialization effects” through usage in which higher-order superpositions contextualize deictic usage. That is, “aspects of context routinely superimpose spatial construals on deictic usage” (p. 644). This perspective calls to mind, and in fact builds directly upon, the work of Hanks (1990, 1992, 2009 and passim), which provides us with a new way of understanding deixis. Situated in a phenomenological approach to interaction and grammatical practices (including deictic referential practice), and incorporating rich, ethnographic details relating to the broadly defined context of use, Hanks (1990) brings together a new approach to the study of deictic reference, one which presents even more fundamental problems with the traditional, descriptive- typological approach. Of particular relevance here is the notion of context and the features of that context “encoded” in or relevant to the interpretation of deictic use. In his 2009 article Hanks presents the following simplified diagram to summarize his approach to (spatial) deixis: While previous approaches have focused largely on the object of reference, the denotatum, defined as locatable at some relative distance (“proximal,” “medial” 6 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 7 or “distal”) from the speaker (or “deictic center”), Hanks explodes the possibilities by explicitly formalizing the three pieces of any occasion of deictic reference. Thus, any occasion of deictic reference points to a denotatum or object of reference (the figure), while also specifying the type of “indexical ground” of reference and the relationship between the two. This is reminiscent of Du Bois’ (2007) “stance triangle,” which is itself characterized by three reference points, the “stance object” and two separate but dialogically related subjects, connected in one act of stance-taking by relations of alignment, evaluation and positioning. The similarities between these conceptual frameworks will be left here, but a comparison and potential unification of these frameworks deserves much attention. What is crucial to understand about Hanks’ approach to deictic reference is the increased number of variables and features relevant for an interpretation of deictic use. Hanks (2009) is strongly critical of the “spatialist, egocentric” bias in studies of deixis. Rather than understanding deixis as defined in relation to the here-now-speaker deictic center, including the reference to an indexical ground as a relevant parameter allows for different types of deictic reference, including not only “speaker oriented,” but other more “socio-centric” orientations such as “addressee oriented” or “speaker and addressee oriented.” Relatively underexplored is the cross-linguistic variation in categorization of the object of deictic reference as, for example, in terms of animacy, gender, number, etc. Finally, and most important for the Indonesian data we are about to see, the relation between the indexical ground/origo and the object of reference “may be spatial, distinguishing for instance relative proximity, inclusion or orientation. But space is just one sphere of context. Other spheres attested in deictic systems include time, perception, memory versus anticipation,” etc. What is fundamental in interpreting instances of deictic reference is the type of “access” indexed by the form. This “access” may be cognitive, perceptual or socially defined. With this background in mind, let us shift to an analysis of some actual examples of deictic reference. The data to be considered come from video recordings of three speakers of Indonesian living in Boulder, CO. The speakers are all bilingual in English and Indonesian. 3. Demonstrative use in interaction A recent addition to the literature on deixis in interaction is Hayashi and Yoon’s (2006, reprinted as Hayashi and Yoon 2010) work on demonstratives and “word formulation trouble” cross linguistically. These authors show that one overlooked and under-appreciated interactional use of deictics (here demonstratives, specifically) is as placeholders in the course of the interactional activity known as “word search.” They delineate three types of such use: placeholder demonstratives (which fill a syntactic slot and maintain the “need for progressivity” and allow speakers to complete their turn without completing the act of reference), avoidance use (placeholders used to avoid producing an 7 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 8 impolite or unacceptable referent) and interjective hesitators (like english “uhm” and Spanish “este”, used to indicate occurrence of a word search but not filling any syntactic or semantic role in the utterance). Hayashi and Yoon show that placeholder demonstratives are pervasive across a range of languages (Chinese, Korean, Japanese, Indonesian, Finnish, Maliseet-Passamaquoddy, among others). The prevalence of demonstratives, in particular, used in this type of interactional trouble is attributed to their “pointing function and the invocation of participant access” (Hayashi and Yoon 2006). Their account draws heavily on Hanks’s (1990, 1992) analysis of deictic usage, which posits “access” (perceptual, cognitive, social) as the fundamental feature of demonstratives and (spatial) deixis more generally. They show that different demonstrative forms index different types of participant access to knowledge of the intended referent. In Korean, for example, medial and distal forms are both used as placeholders during word search. However, while the medial demonstrative indexes “shared access” (indexical ground = speaker + addressee), the distal demonstrative indexes “remote access for speaker” (indexical ground = speaker [only]). These differences in “participant access” follow directly to the Hanks-ian notions of relational type and origo or indexical ground, and, in turn, have important consequences for the nature of the following interaction. As Hayashi and Yoon show, the “shared access” medial forms lead to more overt other-participant involvement in the word search (through offering a possible completion and/or maintaining sustained attention through eye-gaze or other bodily and linguistic practices), while the “remote (from speaker) access” distal forms result in less other-participant involvement and minimal “uh-huh” types of responses. While Hayashi and Yoon have drawn our attention to the crucial role played by demonstratives and deictic expressions in talk-in-interaction, one potential drawback of this approach is that we are still starting with form, some kind of basic underlying structural distinctions, and then investigating the “use” of these forms in interaction to discover something about “language use in interaction.” As many have shown and Enfield (2003a) has so articulately demonstrated, the home of language is face-to-face interaction and it is in interactional data that we will discover the basic encoded semantic distinctions, not only how those distinctions are “put to use” as “interactional resources.” With this in mind, let us turn to some examples of Indonesian demonstrative use in interaction. 4. Indonesian placeholder demonstratives Like English, Lao and many other languages, Indonesian has a two-term adnominal demonstrative system (this and that), traditionally described as “proximal” vs. “distal” (e.g. Sneddon 2006). The forms in Indonesian are ini and itu. However, this is just one part of a more elaborate system of (spatial) deixis in Indonesian which includes a three-way locative demonstrative system (sini, situ, 8 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 9 sana, roughly: ‘here’, ‘there’ (medial?), ‘over there’ (distal?)), two adverbial demonstratives (begini and begitu, roughly: ‘like this’ and ‘like that’), as well as several other miscellaneous forms including (at least) anu (from Javanese, roughly: ‘uhm’/‘whatchamacallit’), nih and tuh (“discourse particles,” grammaticalized forms of ini and itu, with elusive meanings and functions). For the time being we will have to limit ourselves to ini and itu, and in particular, their use as placeholders in the context of word searches. It should be noted that the precise characterization of ini and itu semantics, as well as their function in various forms of discourse and conversational interaction, requires further study. Previous work has initiated this (Himmelmann 1996, Sneddon 2006), but these studies have made limited use of conversational interactional data and do not distinguish the conventional semantics of these forms as from their pragmatic force. The following examples, taken from Wouk (2005), demonstrate the placeholder use of ini and itu in relatively straightforward contexts. (1) terus mengenai hadiah-hadiah-nya itu, apa then about REDUP-gift:GEN DEM what dari e: e itu, e Karang Taruna Nana sendiri from uh uh DEM uh Karang Taruna Nana self "Then as for the presents, (were they) what from uh uh that, uh your own Karang Taruna (name of an organization)." In line 2 the distal demonstrative itu serves as a placeholder for the noun Karang Taruna Nana sendiri. In this example several markers of repair occur, including apa and e: (3x). Quite frequently instances of placeholder repair are encountered in which there is no other indication of repair or difficulty recalling the word. For example, (2) o: kalo gitu udah ini dong, lancar oh if like:this already DEM EMPH fluent bahasa inggris-nya language English-GEN "Oh, in that case (he’s) already this, fluent in English" Here the “proximal” demonstrative ini is used as a placeholder for the adjective lancar. We see, then, that placeholder demonstratives in Indonesian may stand in for nouns and adjectives (or even verbs or parts of words, not shown here), and 9 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 10 that these placeholder “repairs” may be produced fluently (as in example (2)) without any indication of word formulation trouble aside from the use of the demonstrative. An analysis of the differing functions of placeholders in fluent non-word search production is outside the scope of this paper, but ultimately will have to be accounted for. In fact, a broader approach to demonstratives in all occasions of use will likely shed light on both what is semantically encoded and what pragmatic implicatures underlie the diverse observable occasions of use. Following below is an example of prototypical placeholder demonstrative use in a word search from the data collected for this study. Note here that in line (1) we see a fluently produced speaker-completed placeholder demonstrative. Of more interest here is the use of ini in line (3), where the speaker’s involvement in a word search is clearly indicated by the repetition of ini, the initially cut off production of in- and the lengthened vowel [i::] on the second production, the micro-pause in line (4) and the self-addressed ‘whatchamacallit’question, apa nama-nya? at the end of line (3). (3) 1 A: ya udah ini aja, tari-saman kita yeah already this just k.o. dance we latiha:n. [.hh practice “yeah ok, ((let's)) just ((do)) this, tari saman, we(('ll)) practice," 2 L: [m'm= mhm "mhm," 3 A: =sama buka in- ini:: >apa nama-nya(˚)?< also open thi- this what name-GEN "and open in-, ini::, what's it called?" { [A directs gaze at V, far left] } 4 (.) 5 V: booth= booth "booth," 6 A: =booth .hh ntar makanannya:: (.3) kita booth later food-NYA we { ^[V nods] } 10 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 11 masak ra- rame-ra↑me juga cook ra- all.together also bisa ama [yokke can with Y. "booth, then we can also cook the food all together, ((along)) with Yokke." In this excerpt the participants are discussing what they would like to do at an upcoming cultural event. They have been discussing types of dance. Apparently they have settled on the tari saman. In line (1), A indicates that this topic of conversation is settled and that they should move (ya udah might be glossed as "yeah enough already," but without the negative connotation associated with that English phrase). In line (3) A continues with another turn, suggesting another thing that they might do at the event. While a demonstrative like ini alone might not typically indicate a word search, in this instance the nature of A's production of ini indicates that she is engaged in word search. Her first attempt is cut short (in-). In the second production of this proximal demonstrative the final vowel is lengthened quite extensively. This type of "sound stretch" is typical of word searches (Hayashi 2003, Goodwin and Goodwin 1986). Immediately following this use of the demonstrative, A utters the common phrase apa namanya, somewhat equivalent to English whatchamacallit, though here it is functioning more like an "interjective hesitator," rather than a placeholder. At the same time, A directs her gaze directly at V, indicating that she is inviting assistance from V for the completion of her word search (see figure 21). The micro-pause following her turn serves to open up the floor and allow V to enter into the word search activity in progress. V's suggestion for a completion is agreed upon by A. A's recycling of the previously searched-for referent, booth, is immediately followed by a quick, but perceptible, nod by V directed at A. This series of gestures and vocal practices points to the careful attention that participants pay to the ongoing activity. In a series of alternating turns, A and V manage to co-construct and complete this word search activity. Most importantly here is the observation that the speaker, A, maintains directed eye-gaze with the recipient, V, throughout the activity. We can conclude that word search in Indonesian at least has the possibility to begin as a multi-participant activity. Word search is not necessarily initiated through the "characteristic" diversion of eye-gaze and production of a "thinking face," as suggested by Goodwin and Goodwin (1986). 1 In the video stills, speaker V is on the far left, L is in the middle, and A is on the far right. 11 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 12 Figure 2. A: [sama buka in- ini:: >apa namanya(˚)?< Of crucial importance here is the use of ini and how we are to account for its use in the context of the word search. While it is clear from the discussion so far that this (and probably any) use of a placeholder demonstrative in Indonesia is uninterpretable without consideration of accompanying multi-modal practices (particularly eye gaze), the question remains, what is the function of ini and what does its “meaning” contribute to the unfolding interaction? What can this instance of ini used as a placeholder tell us about the indexical ground, the object of reference and the relation between the two that are “encoded” or “schematized” (Agha 1996) in the deictic form in general? The occurrence of an associated gesture and directed eye gaze between participants here is crucial. These multi-modal aspects of the interaction superimpose (in the sense of Agha 1996) an interpretation of the deictic reference as accessible to both speaker and addressee. However, it is not clear that the 12 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 13 demonstrative form itself includes addressee in its indexical ground. In the next example we will see how ini can be used in the context of a word search, accompanied by different gaze and gesture practices, to index speaker-only access. We therefore conclude that only speaker is included in the indexical ground for ini. As is typical for an adnominal demonstrative and is clear from the nominal element that replaces ini, the denotatum type or object of reference is a ‘thing’ (as opposed to ‘region’, for example, as for the deictic term “here”). The relational type here appears to be “inclusive,” meaning that the speaker has access to knowledge of the intended referent. Following Hanks and Hayashi and Yoon, we could represent this schematically as: Table 1. Form Denotatum type Relational type Indexical ground ini ‘the one’ Inclusive speaker(+addressee?) From this characterization of ini as making reference to a thing in an inclusive relationship to the speaker, the implication follows that this demonstrative, ini, indexes immediate access (of the speaker) to knowledge of the intended referent. As seen in the example above, this indication of speaker-access can be adjusted to include speaker and hearer access to knowledge of the referent through accompanying use of gesture and mutually directed eye gaze. Example (4) shows another use of ini as a placeholder, in this case used by the speaker to indicate speaker-only access and avoid overt other-participant involvement in the word search. (4) 1 A: o- isn't it crazy? dan orang-orang di swiss { English } and RED-person in Switzerland gitu-gitu ya? (.5) like.that-RED DP 2 mereka ↑tuh- e::: ini lho, apa nama-nya? e: setuju they DEM uh:: this DP what name-GEN uh agree 3 (.3) untuk bayar tax lebih mahal. (.5) karena mereka tahu bahwa de[ngan bayar tax they know that with(by) pay tax "oh, isn't it crazy? and people in Switzerland ((and places)) like that, you know? they, uh::, ini lho, what's it called? uh:, agree to pay more expensive [higher] taxes, because they know that with paying taxes..." 13 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 14 5 V: [tax-nya balik lagi ke [tax-NYA return again to mere[ka: "the tax ((will)) go back to them again." 6 A: [uh'uh mereka dapat servis: gitu [uh-huh 3P.PL get service like.that "uh-huh, uh-huh, they get service, you know." The excerpt in (4) presents an example of word search, again initiated by A, in which eye-gaze is directed away from the co-participants, resulting in A's own eventual production of the searched-for referent. Note that in fragment (6), V engages in co-participant completion to help the talk move forward. Here no such assistance is invited, and A ends up completing her own word search. The other two participants display understanding that A is involved in a word search through their uninterrupted eye-gaze directed at A. So, while V and L do not participate vocally in the activity, their eye-gaze serves as an acceptable and appropriate response to the ongoing word search. In this case an interruption or attempted completion by one of these co-participants would indicate a lack of attention to A's current activity expressed through both talk and diverted eye- gaze. Their silence and directed eye-gaze is thus a salient form of participation in the interaction, while simultaneously they acknowledge A’s indication of speaker- only access. In line (2) A begins the vocal component of her word search with the "filler" e:::, immediately followed by the proximal demonstrative ini. However, an examination of the video data indicates that A has already diverted her gaze by the end of the tuh-. Prior to this, A's gaze is directed at the two participants (V in particular - see figure 3). The (-) at the end of tuh here indicates an abrupt, probably glottal stop, closure, cutting short the production of this form tuh. This abrupt stop, along with a raised intonation, might itself be the first vocal indication of a word search. During the rest of the word search A maintains eye- gaze away from the other participants (see figure 4). During the (.3) second pause following setuju, the first word in her completion of the previous word search, A shifts her gaze back to the other two participants. This excerpt clearly demonstrates the role of gaze in the life of a word search. Eye-gaze can be used strategically by the speaker to either invite (example 3) or discourage (example 4) co-participant involvment in the completion of the search. It is not the case (contra Goodwin and Goodwin 1986), that word searches are characteristically defined by diverted eye-gaze. The diversion or maintenance of mutual eye-gaze between speaker and hearers during word search difficulty are resources used by the speaker for the production of different types of interaction. Such multi-modal practices work to superimpose particular interpretations of deictic reference. In this case, direction of eye gaze reinforces the indexing of speaker-only access, 14 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 15 while in example (3) directed eye gaze + a pointing gesture worked to include the addressee(s) in the indexical ground and thus index shared-access. Figure 3. A:o- isn't it crazy? dan orang-orang di swiss gitu-gitu ya? (.5) { English } and person-RED in Switzerlandlike.that-RED DP 15 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 16 Figure 4. 2 mereka ↑tuh- e::: ini lho, apa nama-nya? e: setuju (.3) 3P.PLDEM uh this DP what name-GEN uh agree 3 untuk bayar tax lebih mahal. (.5) karena mereka tahu bahwa to/for pay tax more expensive because 3P.PL know that 4 de[ngan bayar tax wi[th pay tax With these two examples, we have tried to make three claims: (1) demonstratives in Indonesian are not simply reflective of the immediate spatial context of the utterance, but rather they contribute to the work of constituting the context by indexing particular types of participant access to knowledge of the intended referent; (2) the particular type of participant access indexed follows directly from the indexical ground, denotatum type and relation between these 16 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 17 two features encoded by demonstrative; and (3) these characteristics of the schema (Agha 1996) of the deictic form are recoverable from the interaction and should not (and cannot) be based only on exophoric, supposedly “basic” situational use of demonstratives to refer to perceptually accessible objects. It is NOT the case that “basic” spatial meanings map metaphorically onto non-spatial “endophoric” contexts. Instead, the “meaning” of deictic reference forms is constructed in multi-modal interaction. A final example of the “distal” demonstrative itu used as a placeholder will help to reinforce these claims. (5) 1 V: mbak Yeny mbak Ully malah(an) nari (.) miss Y. miss U. in.fact dance buat, (.2) for "Y ((and)) U actually dance for ..." 2 A: [i:ya yeah "yeah" {A directs gaze to V} 3 L: [ehh [tunggu mbak= eh 2 wait miss 3 "hey! hold on," {A's gaze is directed to L} 4 V: [itu. that "itu" 5 A: =buat [iya spanyol. for yeah spain/spanish "... for, yeah, spanish ((dance))." {A briefly directs eye gaze to V again} 2 ehh is a frequent attention grabber or marker of interruption. That is, ehh is a resource used to "take the floor" during a conversation. 3 mbak and other address terms are used for second person reference. Her the reference is to A, to whom L's utterance is addressed. 17 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 18 In this case V experiences trouble formulating the word spanyol, eventually supplied by A in line 5. To indicate her involvement in a word search, V first produces two brief pauses, a micro-silence followed by a slightly longer (.2) second silence in line 1. This word search differs substantially from the types of word searches involving ini that we have examined previously. In the previous cases, both recipients indicated acknowledgement of the word search through different modes of involvement. In example (3) this involved co-participant completion in the production of the searched-for referent. In example (4) this acknowledgement was indicated by the maintenance of eye-gaze on the part of the recipients. Co-participant completion was not a relevant or appropriate form of interaction in (4) because of A's eye-gaze diversion. By diverting her eye-gaze away from the recipients, A indicated that she did not desire assistance in the completion of her word search. On the other hand, V's word search in (5) is interrupted by L's turn at line 3. This indicates that L is not attending to V's experience of word formulation trouble. While A does eventually complete the search through a form of co-participant completion in line 5, this is uttered at a much lower volume than the previous discourse. A also has begun shifting her gaze away from V toward L in response to L's abrupt interruption and attempt to take the floor. A’s offering of a candidate for co-participant completion here might follow from some kind of pressure to complete the reference. While placeholders in Indonesian are used for “vague reference” and “avoidance use” (e.g. the frequent use of itu-nya (DEM-3.POSS) to refer to a male sexual organ, the actual term obviously to be avoided in polite speech), this would not seem to work in this case because the speakers are discussing what kind of dance they will perform in an upcoming event. To refer to the type of dance vaguely with a placeholder like itu is dispreferred in this context since explicit reference to the dance-to-be-performed is needed. If we try to recover the three elements of this deictic form from this example, it seems that in contrast to ini, itu is defined by its non-immediate relational type between a thing (denotatum type) and the speaker (indexical ground). We could schematize this as follows: Table 2. Form Denotatum type Relational type Indexical ground itu ‘the one’ non-immediate speaker In this case the relation of non-immediacy with reference to the speaker indexes “remoteness of access” for the speaker. Since the hearer/addressee is not included in the indexical ground (as with ini), the deictic reference might be interpreted as remote or not to the addressee. In terms of involvement in the ongoing word search and attempt at reference, this means that the addressee may or may not directly participate. Thus, as we see in example (5), the addressee (A) becomes involved through directed eye gaze and eventually offering a possible 18 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 19 completion (spanyol). However, this leaves the potential for itu-placeholders to be “filled in” by the speaker alone, without any involvement from the addressee(s). Clearly more examples are needed to draw firm conclusions regarding the characteristic features of these two adnominal demonstratives. However, this paper has shown that the “meaning” of demonstratives is not necessarily based on exophoric spatial uses, but can instead be shown to follow from their use in situated interaction. Future work will have to show whether these conclusions are valid if we consider the wider range of demonstrative uses across situations, activities and types of discourse. We tentatively hypothesize that these findings regarding the meaning of Indonesian demonstratives when used as placeholders will extend to exophoric spatial uses as well, thereby undermining the supposed spatial basis of demonstrative meaning and use (cf. Hanks passim). 5. Toward a (linguistic anthropological) account of deixis in interaction In this paper I have shown that an anthropological approach (broadly conceived) to demonstrative use, and deixis in general, is needed to account for the “meaning” and interpretation of this important part of language. Deixis represents a core example of the contextualized and contextualizing nature of linguistic practice and meaning. Detailed micro-analysis of demonstrative use in naturally-occurring interaction can shed light on their meaning(s) and lead to insights unavailable based on the analysis of hypothetical examples. The analysis presented here aims to promote the claim that language is socially constituted and an “emergent” product use in real-time interaction. Further evidence will come from more detailed analyses of language use and social interaction. In addition to the semantic and pragmatic “meanings” encoded in these demonstrative forms, it seems quite likely that this practice of placeholder use does additional work for participants in interaction. What I would like to suggest here is that use of a placeholder demonstrative as a type of repair shares something in common with the other-initiated repair discussed by Besnier (2010) in his book on the production of gossip on Nukulaelae atoll. In this work Besnier suggests that the use of non-referential forms (like “he is such a …”) in initial reference position, which evokes other-initiated repair (“who?”, i.e. “who is he?”), works to invite co-participant production of gossip, which ultimately removes culpability and blame from the gossip-initiator. Similarly, in the case of placeholder demonstrative use, the use of such a vague reference form (akin to a “recognitional” form, cf. Himmelmann 1996) in what looks like “initial” position, occasionally evokes other-participant repair and/or involvement in the word search. This allows speakers to invite other-participant involvement in the production of reference, a fundamental feature of language use and everyday talk and interaction. This avoidance of speaker-only production of reference might relate to the relative rights of speakers and addressees to do reference. This might have to do with shared knowledge among the participants about others’ epistemic 19 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 20 rights. For example, in example (3) V is invited to make reference to the booth because she is much more involved in the preparations for the cultural event being discussed than either A or L is. V, then, holds greater epistemic rights to do the reference in this case, and A’s use of a placeholder demonstrative might be interpreted as an attempt to “downgrade” her own epistemic rights inherent to first position in the adjacency pair (Heritage and Raymond 2005). The immediately preceding analysis represents the results of a pilot study into the meaning and use of demonstratives in Indonesian conversation. Future research will need to provide a more comprehensive account of demonstrative meaning, including the relationship between the interactional functions discussed here and the apparent “exophoric” functions commonly proposed as the “basic” function of demonstratives. Future work will require a greater amount of data as well as a more ethnographic approach. Hanks has drawn linguistic anthropology’s attention to deixis and referential practice as a phenomenon of importance to the field. However, we are still lacking descriptions of deixis and referential practice in a wide sample of languages. Echoing Hanks (2009), this paper makes the call for greater attention to deixis in linguistic anthropology and other approaches to the study of language (use) and social interaction. 20 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 21 References Agha, Asif. 1996. “Schema and superimposition in spatial deixis.” Anthropological Linguistics 38: 643-682. Anderson, Stephen R. and Keenan, Edward L. 1985. “Deixis.” In: Shopen, Timothy (ed.), Language typology and syntactic description, Vol. 3, 259-308. Cambridge: Cambridge University Press. Besnier, Niko. 2009. Gossip and the Everyday Production of Politics. Honolulu: University of Hawai’i Press. Bickel, B., 1997. “Spatial operations in deixis, cognition, and culture: where to orient oneself in Belhare.” In Nuyts, J., Pederson, E. (Eds.), Language and Conceptualization. 46–83. Cambridge University Press, Cambridge. Bucholtz, Mary. 2011. White kids: language, race and styles of youth identity. Cambridge: Cambridge University Press. B hler, Karl. 1934. Sprachtheorie: Die Darstellungsfunktion der Sprache. Jena: Fischer. [1990. Theory of language: The representational function of language, translated by Donald F. Goodwin. Amsterdam/Philadelphia: John Benjamins.] Diessel, Holger. 1999. Demonstratives: Form, function, and grammaticalization. Amsterdam/Philadelphia: John Benjamins. Dixon, R. M. W. 2003. “Demonstratives: A cross-linguistic typology.” Studies in Language 27: 61–112. DuBois, John, 2007. “The stance triangle.” In Englebretson, R. (Ed.), Stancetakinig in Discourse: Subjectivity, evaluation, interaction. Benjamins, Amsterdam, pp. 139–182. Duranti, Alessandro. 2003. “Language as Culture in U.S. Anthropology: Three Paradigms.” Current Anthropology 44(3):323-348 Enfield, Nicholas J., 2003a. “The definition of WHAT-d’you-call-it: semantics and pragmatics of ‘recognitional deixis’.” Journal of Pragmatics 35, 101–117. Enfield, Nicholas J., 2003b. “Demonstratives in space and interaction: data from Lao speakers and implications for semantic analysis.” Language 79, 82–117. Fillmore, Charles J. 1982. “Toward a descriptive framework for spatial deixis.” In: Jarvella, Robert J. and Klein, Wolfgang, Speech, place, and action: Studies in deixis and related topics, 31-59. Chichester: John Wiley & Sons Ltd. Goodwin, Charles and Goodwin, Marjorie. 1986 "Gesture and Coparticipation in the Activity of Searching for a Word.” Semiotica 62(1-2):51-75. 21 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics, Volume 22 (2009) 22 Goodwin, M. 1990. He-Said-She-Said: Talk as Social Organization among Black Children. Indiana Univ. Press. Goodwin, C. 1999a. “Pointing as Situated Social Practice.” Ms, UCLA. Goodwin, C. 1999b. “Action and Embodiment within Situated Human Interaction.” Ms, UCLA. Hanks, William F., 1990. Referential Practice: Language and Lived Space among the Maya. University of Chicago Press, Chicago. Hanks,William F., 1992. “The indexical ground of deictic reference.” In: Duranti, A., Goodwin, C. (Eds.), Rethinking Context. Cambridge University Press, Cambridge. Hanks, William F., 1996. “Language form and communicative practices.” In: Gumperz, J., Levinson, S. (Eds.), Rethinking Linguistic Relativity. Cambridge University Press, Cambridge, pp. 232–270. Hanks, William F., 2005. “Explorations in the deictic field.” Current Anthropology 46, 191–220. Hanks, William F. 2009. “Fieldwork on Deixis,” in Journal of Pragmatics. Hayashi, Makoto. 2003. “Language and the body as resources for collaborative action: A study of word searches in Japanese conversation.” Research on Language and Social Interaction 36: 109-141. Hayashi, Makoto and Yoon, Kyung-eun. 2006. "A cross-linguistic exploration of demonstratives in interaction: With particular reference to the context of word-formulation trouble," in Studies in Language 30:3, 485-540. Heritage, John and Geoffrey Raymond. 2005. “The Terms of Agreement: Indexing Authority and Subordination in Talk-in-Interaction.” Social Psychology Quarterly, 68:1, 15-38. Himmelmann, Nikolaus. 1996. “Demonstratives in narrative discourse: A taxonomy of universal uses.” In: Fox, Barbara (ed.), Studies in Anaphora, 205–54. Amsterdam/Philadelphia: John Benjamins. Inoue, Miyako. 2004. “What Does Language Remember?: Indexical Inversion and the Naturalized History of Japanese Women.” Journal of Linguistic Anthropology 14(1):39–56. Levinson, Stephen C. 1983. Pragmatics. Cambridge University Press. Levinson, Stephen C. 1994. Deixis. In R. E. Asher (ed.) Encyclopedia of Language and Linguistics, 2:853-57. Levinson, Stephen C. 1996. “Frames of reference and Molyneux’s question: cross-linguistic evidence.” In P. Bloom, M. Peterson, L. Nadel and M. Garrett (eds.) Language and Space: 109-169. MIT. Lyons, John. 1977. Semantics. Vols. 1 & 2. Cambridge: Cambridge University Press. Lyons, J. 1982. “Deixis and subjectivity: Loquor, ergo sum?” in R. J. Jarvella & W. Klein (eds.) Speech, Place and Action. John Wiley & Sons. Sidnell, Jack. 1998. “Deixis.” In: Verschueren, Jef; Östman, Jan-Ola; Blommaert; Jan, and Bulcaen Chris (eds.), Handbook of Pragmatics 1998, 1-28. Amsterdam/Philadelphia: John Benjamins. 22 Colorado Research in Linguistics, Vol. 22 [2010] https://scholar.colorado.edu/cril/vol22/iss1/6 DOI: https://doi.org/10.25810/0ghq-wk77 Toward a linguistic anthropological account of deixis in interaction: ini and itu in Indonesian conversation 23 Silverstein, Michael. 1976. “Shifters, verbal categories and cultural description.” In Basso, K., Selby, H. (Eds.), Meaning in Anthropology. School of American Research, Albuquerque, pp. 11–57.Silverstein, Michael. 2003. Silverstein, Michael. 2003. “Indexical order and the dialectics of sociolinguistic life.” Language and Communication 23:193-229. Sneddon, James. 2006. Colloquial Jakartan Indonesian. Canberra: Pacific Linguistics. Ochs, E. 1990. “Indexicality and socialization" in Cultural psychology: The Chicago symposia, ed. by J. Stigler, G. Herdt, & R. Shweder. Cambridge: Cambridge University Press. Ochs, Elinor, 1992. “Indexing gender.” In: Duranti, A., Goodwin, C. (Eds.), Rethinking Context. Cambridge University Press, Cambridge, pp. 335–358. Schegloff, EA. 1972. “Notes on a conversational practice: formulating place.” In Studies in Social Interaction, ed. DN Sudnow, pp. 75–119. New York: Free Press. Weissenborn, J rgen and lein, Wolfgang (eds.). 1982. Here and there: Cross- linguistic studies on deixis and demonstration. Amsterdam/Philadelphia: John Benjamins. Wouk, Fay. 2003. “The syntax of repair in Indonesian.” Unpublished manuscript. University of Auckland. 23 Williams: Toward a Linguistic Anthropological Account of Deixis in Interaction Published by CU Scholar, 2010 Colorado Research in Linguistics 6-2009 Toward a Linguistic Anthropological Account of Deixis in Interaction: Ini and Itu in Indonesian Conversation Nicholas Williams Recommended Citation CRIL Style sheet