Tactile engagement of prospective next speakers in Indonesian multiparty conversations

Abstract

This article investigates the use of touch as a tool for engaging prospective next speakers within Indonesian multiparty conversation. We examine the lamination of touch onto questions directed towards specifically targeted recipients. First, we find that questions with touch are deployed when the physical environment complicates the attainment of mutual orientation. Second, when previously targeted recipients have failed to respond to a question, touch is added to follow-up questions that are deployed for pursuing a response. Third, touch is added to questions that are personal or that inquire about potentially delicate matters. This multimodal investigation of conversational turn-taking provides data from Colloquial Indonesian as basis for cross-linguistic comparison. In considering the volume of touches in these data we ask whether cultural and environmental factors might contribute to a haptic modification of ordinary turn-taking procedures. (Turn-taking, touch, multimodality, sociotopography)

INTRODUCTION

In managing participation in conversation, interactants may deploy a range of resources from three semiotic modalities: the audio-vocal, the visuo-corporal, and the manuo-tactile modalities. Human conversation can transpire in any of these modalities exclusively. For sighted, hearing participants, even though the cycling of turns-at-talk ultimately plays out within the audio-vocal modality, available resources for engaging prospective next-speakers may be drawn from all three modalities (audio-vocal, visuo-corporal, manuo-tactile). In this study, we explore the relationship between turn-taking organization—specifically, next speaker selection—and the manuo-tactile modality, as administered though interpersonal touch.

A small, but growing number of studies have explored the distribution of touch and its functions in interaction. These investigations have examined interpersonal touch in a range of interactional contexts such as adult-child interactions (Cekaite & Kvist Holm 2017; Goodwin & Cekaite 2018), medical examinations (Nishizaka 2007), caring interactions (Denman & Wilkinson 2011; Raia, Goodwin, & Deng 2020), and tactile sign language (Iwasaki, Bartlett, Manns, & Willoughby 2019; Iwasaki, Bartlett, Willoughby, & Manns 2022). Touch may be used for a range of purposes. For example, children in busy classrooms tap their teachers to summon their attention (Gardner 2015). People also employ touch to comfort each other or to build intimacy. According to Cekaite & Kvist Holm (2017), adults may use touch along with verbal resources as a form of ‘haptic soothing’ to calm children. Goodwin (2017) links touch within family interactions to functions including reconciliation, comforting, celebration and positive assessment, and grooming. These touch-based acts are done as independent responses to others’ actions. Goodwin & Cekaite (2013) highlight the role of touch in directives produced by parents to their children. They indicate that touch can be used to regulate mutual orientation, as well as to ‘recycle or upgrade the requested action’ (Goodwin & Cekaite 2013:136). It is the action import of interpersonal touch and its role in regulating mutual orientation that we investigate here, particularly with regard to urging specific co-participants to produce the next turn of talk.

As Goodwin & Goodwin (2004) argue, the management of participation is a complex multimodal activity, composed of verbal and embodied resources. They note that participants exploit ‘the semiotic resources provided by their bodies to construct a range of relevant displays about orientation toward others and the actions in progress’ (2004:239). That is, a speaker and a recipient can negotiate mutual orientation through a variety of bodily-visual actions (Ford & Stickle 2012), such as gesture (Streeck 1993), gaze, and body positioning (Kidwell 1997).

The first of Sacks, Schegloff, & Jefferson's (1974) ordered turn-taking rules (1a) relates to obligating co-participants into taking the floor at the next transition relevance place. Although the original turn-taking paper was principally based on audio recordings, Sacks and colleagues (1974:716–18) explain the basic allocation technique for ‘current speaker selecting next’ as involving ‘the affiliation of an address term (or some other device for achieving addressing, e.g., gaze direction) to a first pair-part’ (1974:717, emphasis added). Multimodal analyses of video recorded conversations have begun to elucidate these processes. For example, Lerner (2003) notes that a second-person pronoun (you) within a question will indicate that the person being addressed is expected to take the next turn. Specifically which person is being designated may be conveyed explicitly (via gaze or with address terms, etc.) or tacitly: the current speaker implicates another participant's greater knowledge of the events being discussed than their own, and that of the other present participants. Lerner notes that gaze on its own may be ineffective as a ‘recipient indicator’ if targeted recipients fail to notice the gaze of the speaker being directed toward them. More recently, a suite of ‘engagement tools’ (Blythe, Gardner, Mushin, & Stirling 2018) have been identified that increase the likelihood of securing a response from the targeted recipient. These include vocative expressions, recipient directed gaze, epistemic skewing toward the recipient, interrogative lexico-morphosyntax, and language-specific prosodic patterns associated with interrogativity (see also Stivers & Rossano 2010). Blythe and colleagues (2018) note that in multiparty conversations conducted in the Australian Aboriginal languages Gija, Jaru, Murrinhpatha, and Garrwa, uptake by prospective next speakers can be complicated by particular seating arrangements. When participants are not arranged within a semi-circular, L-shaped, or classic F-formation (Kendon 1990, 2010; see Figure 1), speakers deploy pointing gestures, raise their voices, turn their heads toward the targets, projecting their voices such that recipient-gaze is drawn toward the speaker, thus effecting the attainment of mutual gaze. When initial attempts at engaging prospective next speakers prove ineffective, speakers will augment their prior attempts by adding additional tools of engagement to a follow up question, such that transfer of speakership is effected.

Figure 1. The F-formation system (Kendon 1990, 2010). From left to right: semi-circular, L-shaped and classic F-formations with the o-space in the centre.

These factors attest to the organized relationships between the taking of turns at talk and semiotic resources administered via other modalities. There is, for instance, a developing body of evidence on the relationship between turn-taking organization and the visuo-spatial modality (Lerner 2003; Mondada 2007, 2013; Keisanen & Rauniomaa 2012; Kendrick & Holler 2017; Auer 2018). Less attention has been given to haptic behaviour within spoken interaction, particularly with regard to turn-taking and action sequencing, although Licoppe & Veyrier (2020) examine how court interpreters and asylum seekers use interjections and touch to manage turn transitions. Li's (2020) research on Mandarin conversations is another notable exception. Li (2020) demonstrates that touching can be employed alongside ‘jokings’ that are ‘sequentially disaligned’ from immediately prior talk, and ‘jokings’ that arise after a period of sustained disagreement. The additive functions of these touches include seeking appreciation, indicating the sequential misplacement of the ‘jokings’, and displaying intimacy with the other interactant. In this way touch manages the potential ‘relational import’ of the joking, particularly in the context of extended disagreement, ensuring recipients participate in the joking activity (Li 2020:20).

Our investigation centres on the use of touch as a haptic tool for engagement of prospective next speakers (Blythe et al. 2018) within Indonesian multiparty conversations. Touch has the potential to be as explicit (perhaps even more explicit) than any visual or talk-based resource for next speaker selection—at least for the party being selected. In some of the extracts we present, we find other (non-haptic) engagement tools such as address terms and second person pronouns used alongside touch. Touches laminated onto questions are not infrequent in our data (one every eight minutes, on average). Alongside their situated and systemic interactional motivations, we also consider both culture and the environment as possible factors motivating the use of tactile next-speaker engagement in these data.

DATA AND METHODS

A corpus of Indonesian multiparty conversations was collected in a variety of locations in Jakarta including homes, cafes, shopping centres, and public spaces. All data were collected with the informed consent of all participants, in accordance with ethical approvals granted by Macquarie University. All participants names presented below are pseudonyms. However, participants consented to the use of unaltered screenshots as part of the publication of the study's findings. Participants (forty-four females and twenty males) were recruited and recorded engaging in conversations with at least two other participants they knew well (e.g. friends, neighbours, family members), in locations where they regularly interact. Nine and a half hours of conversation were recorded using several video-cameras. Each participant wore a lapel microphone. Recordings were then transcribed with conversation analysis transcription conventions (Jefferson 2004; Hepburn & Bolden 2017). In this corpus, 1,533 questions were then identified using the Stivers & Enfield (2010) coding system that are implicated in selecting a next speaker, under rule (1a) of the turn-taking system (Sacks et al. 1974:704). We also found 336 tokens of interpersonal touching. The sheer frequency of touches in this collection (one token every 102 seconds, on average) dramatically exceeds those of all other corpora we have examined.¹ This raises the question of whether interpersonal touch might be a cultural feature that characterises Indonesian conversation. Here, we report on the analysis of seventy-one questions that included a touch from a question producer to a question recipient (or prospective question recipient). Of these questions, twenty-one also included an address term (29%), and nine included a second person pronoun (13%). The remaining forty-one questions (58%) included touch but no address terms nor second person pronouns. Regarding the touches themselves, they include single and multiple touches imparted by various parts of the body (e.g. palms, fingers, shoulders, thighs, feet, and elbows). When considering the propensity for touching in these data as a potentially cultural characteristic of Indonesian conversation, we are mindful that Jakarta has a very high population density, and that many of the recordings are conducted in locations that are cramped or crowded. We thus consider the spatial orientation of participants and whether their arrangement is conducive to the attainment of mutual orientation through gaze, and whether this arrangement is conducive to interpersonal touching.

In the sections to follow, we examine touches that accompany questions that attend to issues of participation management and action formation. First, we discuss examples where question producers’ use of touch address problems with mutual orientation by drawing recipient gaze. From there we examine cases in which touching accompanies follow-up questions that pursue responses to prior questions that have not thus far been satisfactorily answered. Third, we show touch being laid onto personal or potentially delicate questions to convey particular sorts of interpersonal meanings like concern or affection. These touches also draw recipient gaze but are not particularly geared towards resolving inadequate recipiency. In the final sections we discuss the use of touch as a tool for engaging prospective next speakers (Blythe et al. 2018), and whether the prevalence of questions with touch in these Indonesian conversations might be explained by environmental or cultural factors.

USING TOUCH TO ATTEND TO ISSUES OF MUTUAL ORIENTATION

When there are problems with mutual orientation, one function of touch is to engage a co-participant as the prospective next speaker. Gaze is an effective designator of recipiency when the speaker and targeted recipient look one another in the eye (Lerner 2003; Auer 2018, 2021; Blythe et al. 2018). Touches that are laminated onto questions are effective at ensuring that mutual gaze is achieved, such that the targeted recipient can see they are being addressed and thus obligated to respond. We see in the extracts to follow that speakers’ and recipients’ orientation toward each other is compromised by their seating arrangements, and that the touches are devoted to dealing with this issue. For example, in extract (1) the participants are seated in the corner of a room, leaning against the walls, such that they are not all facing each other directly.

Prior to extract (1), Ami and Sari had been talking about health issues that elderly people experience. Over lines 1 to 7, Ami and Sari mention various age-related diseases (e.g. heart attack, stomach ulcers, liver problems, and diabetes), culminating in a collaborative list over lines 6, 7 and 9. The question in focus is produced by Lela at line 13 and is addressed to Ami. The arrowed lines prior to the question show the bodily movements that occur before and within Lela's turn.

After a silence, Sari lists the final symptom, kuning ‘jaundice’ in line 9, but closes the list in line 11 with sekarang mah ‘nowadays’. At this point Lela starts to change her body posture and glances at Ami. During the intra-turn pause at line 10, Lela gazes at Ami (Images 1 and 2) then touches Ami's arm with an open palm (Image 3). Lela then raises and lowers her hand to implement a second touch; at which point, they establish mutual gaze (Image 4) and Lela asks Mak Titi sakit? ‘Is Mrs Titi unwell?’ (at line 13). There is a gap in line 14 and Sari produces a continuer (‘Uh huh’). Lela then produces a possible complaint in line 16. After a long gap, Ami deflects the possible complaint by saying, ‘Well, she's just like that, mother, {she's} a homebody’ (Iye sakit begitu aja die mah mak jarang keluar, line 18), which effectively states that she is unsure about Mrs. Titi's condition.

In extract (1), Ami and Sari had been exchanging turns, with Lela not bidding for the floor. In addition, Lela's head, trunk, and legs are positioned away from Ami and Sari. Before touching and questioning Ami, Lela reorients herself toward Ami and Sari's interactional space. By deploying touch just prior to the question, Lela accomplishes a shift in the local participation structure such that Ami's gaze is drawn toward her and away from Sari. Note too, that Lela had asked this question earlier in the conversation, but it is not shown in the transcripts. This may be part of the basis for her account/complaint at line 16 (and perhaps her touch).

Extract (2) offers another example of a problematic seating arrangement where touch is used to elicit a specific recipient's orientation. The conversation involves Tubi, Ucu, and Ucu's family and takes place at Ucu's house. Tubi is visiting Ucu, who were best friends during childhood. The other participants include Ucu's sister (Asih), Ucu's daughters (Nada and Rina), Ucu's son-in-law (Eman), and Ucu's niece (Nana). Nana is a late addition to the conversation, joining in after dropping by Ucu's home. In the arrowed lines 15 to 17, Asih reaches over to touch Nana who she is addressing (line 16), drawing her gaze toward her (line 17). Prior to extract (2), Tubi and Nana had been talking about their friends who had passed away. Tubi and Nana had not seen each other for many years after having attended school together and they did not recognize each other.

Addressing Tubi in line 1 Nana asserts Nana ga ngenalin ‘I wouldn't have recognized you’. Here, Nana uses her own name for self-reference (Djenar 2007). Nana and Tubi then laugh. After a brief silence, Tubi takes the floor at line 6 and tells Nana that he met his old friend (Oim) and could not recognize him. However, he struggles to remember where he last met him. At line 8, Nana responds by saying that Udah meninggal juga ‘{Oim} has already passed away as well’ and indicating Oim's neighborhood (Benteng Dalem). After a silence, Tubi provides a receipt (‘Yeah, Benteng Dalem’) and then tells Nana that he went there. After Nana's go-ahead response in line 12, Tubi continues to speak about the day he met Oim, enacting how he could not recognize him. After the brief silence at line 15 where Asih shifts her gaze toward Nana, she asks her Emang cing Oim masih ada Na? “Is Uncle Oim still alive, Na?”. Having established mutual gaze (Image 7), Nana confirms that Oim has indeed passed away (line 18), to which Asih provides an oh-prefaced response at line 20.

We now focus on Asih's question at line 16 emang cing oim masih ada na? ‘Is uncle Oim still alive, Na?’. Asih touches Nana's arm as she asks this question which also includes a post-positioned address term (Na) (Lerner 2003; Hamdani, Barnes, & Blythe 2022). Asih shifts her gaze toward Nana at line 15 (Image 5), but Nana is gazing at Tubi. Asih then produces a polar question in line 16, enquiring as to whether or not uncle Oim—the person being referred by Nana and Tubi—is still alive. Asih is sitting a little behind Nana and, as her turn approaches possible completion, she leans forward and tries to secure Nana's orientation by touching her elbow (Image 6b). The combination of touch and address term appears to be dealing with the fact that Nana keeps gazing toward Tubi, despite having progressed a substantial way through her turn. Because Asih is seated behind Nana, Nana cannot see who Asih is addressing without turning her body backward to meet Asih's gaze. She does this at line 17 (see Image 7), replying that Oim has already passed away (line 18). Note too that Asih's question revives a topic that Nana and Tubi had already moved past. Thus in addition to dealing with an issue of embodiment, the touching question may also deal with the fact that the topic of the talk has advanced.

The mutual orientation issue in extract (3) relates to how the four neighbours are seated on a bench. Before this extract, Mali, a fifty-eight-year-old street food vendor, had been talking about his age. Over lines 1 to 2, Mali continues his story about younger people in his village looking older than him. After a long silence in line 3, Muiz selects Ifan as the next speaker by touching him on the shoulder and asking him about his age. After a gap, Ifan jokes in line 6 that he is twenty-five. Muiz acknowledges this at line 8 with a change-of-state token before Mali informs Muiz that Ifan is lying to him. Ifan smiles and changes his body position so that he and Muiz can see each other (Image 9). Ifan then provides a serious answer over lines 13 to 16, explaining that he is thirty-two.

Let us now consider how Muiz engages Ifan as the next speaker. Muiz looks toward Ifan from the beginning of line 3. At line 4 Muiz indicates a single addressee by producing the second person singular pronoun kamu. Muiz, however, is sitting behind Ifan who is gazing forward; so they cannot possibly meet each other's gaze. Just after kamu, Muiz adds touch to the wh-question word berapa ‘how many/how much’ (Image 8). By doing this, he ensures that the second person singular pronoun kamu ‘you’ explicitly targets Ifan. One might query, though, how the unaddressed recipients—Mali and Roni—might come to understand that kamu is directed to Ifan. Mali has just revealed his age before this extract, which makes kamu a relevant reference form for Muiz to address him with. As Hassall (2013) has noted, kamu is mainly used in an age-oriented way, for people who are ‘equal or lower status, or to address children’. It could potentially be applicable to Roni who is younger than Muiz, but Roni is likely to be able to see Muiz reaching out to Ifan (Image 9). By contrast, Ifan is less able to see Roni, who he might take as being addressed by Muiz's kamu question. That is to say, Ifan is really the only party for whom this question might possibly be ambiguous, and this is exactly what the touch disambiguates.

In each extract in this section, the question producers were not actually speaking before producing their questions. As such they have interceded into the talk, taking the floor (legitimately, under rule 1b of the turn-taking system) and then reallocating it to their targeted recipients (under rule 1a; Sacks et al. 1974:704). We can see in each case that there are obstacles to establishing mutual gaze. Question producers must deal with seating arrangements that restrict visual access of their targeted recipients. This is summarized in Figure 2.

Figure 2. A birds’ eye view of the seating arrangements in extracts (1) (left), (2) (middle), and (3) (right).

In extract (1) (Figure 2, left), Sari, Lela, and Ami sit with their backs against brick walls. Although L-shaped, this is not an F-formation as the participants’ transactional segments do not overlap. When Lela experiences rheumatic pains, she stretches her legs out, leaving a space between herself and Ami. She also disengages from the conversation for a while, gazing away from the other two participants. In both extracts (2) and (3), the question producers are seated behind the question recipients, which complicates the recipients’ return of gaze. These problematic seating arrangements seem to encourage Asih (Figure 2, middle) and Muiz (Figure 2, right) to deploy touch as they address their questions, which secures the recipiency of their targets who turn around to look them in the eye. Consider also how the question producers progress towards touching their targets. Images 1 and 2 show Lela (in extract (1)) moving her head before gazing toward Ami. Lela's first attempt at securing Ami's recipiency by touching her arm (Image 3, line 12) is followed by a second touch in Image 4 (line 13), which overlays her question. Similarly, the question producers in extracts (2) and (3) monitor the targeted recipients before launching their questions. Asih raises her left arm in the middle of her turn (Image 6a, line 16) before touching Nana's right arm (Image 6b) just prior to the completion of the turn. Muiz raises his arm before producing his turn (line 3 of extract (3)) so that his hand meets Ifan's shoulder (Image 8) before his question is complete.

In this section we have seen touch being used to manage problems with mutual orientation between question producers and question recipients. Touch is a useful addition when people are oriented in formations that complicate the attainment of mutual gaze, although touch is only deployable to recipients within reach of the speaker. A key outcome of these touches is that the speaker elicits the gaze of the targeted recipient, so they can plainly see that they are being addressed and selected to speak next, under rule 1a of the turn-taking system. Another point of consideration is that in extracts (1) and (2) touch is laminated onto questions that inquire about delicate matters (Mrs. Titi being unwell and Oim having possibly passed away). We consider the caring dimension of tender touches further below, after first examining extracts in which touch is implicated in pursuing responses to an earlier question.

TOUCH USED IN PURSUING RESPONSES AND IN REDIRECTING QUESTIONS TO NEW RECIPIENTS

This section examines how speakers employ touch when pursuing a response. We also see touch being used to redirect follow-up questions toward new—more proximal—participants, in the hope of securing an answer. That is, question producers use touch to address issues related to recipiency, sequence organization, or both. Their use of touch points towards issues of conditional relevance and the preference for selected next speakers to take the floor (Stivers & Robinson 2006; Schegloff 2007). Being a haptic tool of engagement, touch can augment the semiotic resources that were previously deployed in unsuccessful attempts at securing a response (Stivers & Rossano 2010; Blythe et al. 2018).

Extract (4) is taken from a conversation amongst four male relatives—Anis, Ical, Iyan, and Aldi—who are sitting at a table nearby a shopping centre. Prior to this extract, Aldi, Anis, and Iyan had been talking about a computer Aldi and Anis had recently bought. Ical begins asking Aldi and Anis about the specifications of the computer. The turns in focus are at lines 20 to 22, in which Ical asks Anis a question before re-directing the question to Aldi.

In line 1 Ical asks Anis Tapi ramnya? ‘How about the RAM?’ (Random Access Memory). Anis and Aldi both answer at lines 3 and 4 by providing the size of the RAM. After a silence, Ical carries out a third position repair (Schegloff 1992) at line 6. Maintaining his gaze toward Anis, he instead asks about the hard drive. Anis answers at line 8, treating the question as inquiring about the size of the drive (‘500{GB}’). Aldi also answers in overlap (line 9) by attempting to explain which type of drive they selected, then, elaborating at line 11, by specifying the brand (‘WD Blue’), which he pronounces [wede].³ After a silence, Iyan corrects Aldi's pronunciation of the brand name (‘Wi Di’ [widi]) at line 13 which Aldi accepts at line 14 by repeating twice. Ical then tries to link the brand name to a person, Si itu ‘That person’, but he fails, leaving a 1.4-second silence in line 16. At lines 17 and 18, Aldi and Anis reconfirm their choice of hard drive. After a brief silence at line 19, Ical shifts his gaze from Iyan to Anis, establishes mutual gaze, and then asks at line 20 what WD is (Image 10). Ical then shifts his gaze to Aldi and revises his question in line 22 (‘What is the acronym?’, i.e. ‘What does the acronym stand for?’), touching Aldi three times as he does so. Aldi responds twice in lines 24 and 26 with an account stating that he does not know what the acronym means.

The initial version of Ical's question at line 20 does not receive an immediate response. His gaze is directed toward Anis as he delivers this question, but Anis, who is returning his gaze, shows no signs of responding. His second attempt at line 22 asks specifically what the acronym stands for. Alongside the revisions to his talk, Ical shifts his gaze and touches Aldi's right thigh three times (Images 11a,b,c). Ical deals with the lack of uptake from the recipient he originally selected through gaze by using touch and gaze, in combination, to redirect the revised question toward Aldi. Aldi and Anis purchased the computer together. If Anis doesn't know what WD means then Aldi might be able to answer. These touches are thus central in redirecting the question toward a new target with knowledge about the computer. Aldi does take the floor by providing two dispreferred responses, namely accounts for his inability to explain the acronym.

The next example is from a conversation between neighbours: Ima, Yaya, Titi, and Yuda. Ima, Titi, and Yaya are sitting together in the local shop that Ima runs. Before this extract, the focus of the talk had been on issues surrounding subsidized liquefied petroleum gas (LPG), and Titi had asked Ima how to source an LPG container. There are two questions with accompanying touch in this extract, and both are produced by Titi. We focus on the second one at line 18.

At line 1, Titi gazes toward Ima and asks ‘So, what are we going to use?’. Ima gazes away from Titi, shifts her gaze toward Yaya, and answers at line 3 that they are going back to using kerosene again. In line 6, Yaya gazes at Ima and then laughs. At line 7, Titi—who is still gazing at Ima—raises her voice, and responds Kalo minyak tanahnya ada juga ‘If the kerosene is available’.

At line 11, Titi turns her gaze toward Yaya (who is gazing at Ima), touches her on the arm, and asks her a question, Kalo ga ade gimana Ya? ‘What if it's not {available}, Ya?’. Yaya responds whilst turning toward Titi, with whom she establishes mutual gaze (Image 13), noting that kerosene is expensive. At line 13 Titi receipts this (non-answer) response with a minimal post-expansion nah ‘See! (I told you)’. At lines 15 and 16, Titi gazes toward Ima and asks her about the price of kerosene. Ima, however, is not gazing at Titi, nor at Yaya, so she does not respond. At line 18, Titi pursues a response by changing her target. She shifts her gaze toward Yaya, touches Yaya on the elbow with her finger (Image 14), and then asks her Yang duluan aja berapa duit? ‘How much money was it?’. After 0.5-seconds of silence Yaya turns toward Titi and they establish mutual gaze at line 19 (Image 15). Again, Titi pursues a response at line 20 with the turn increment Seliternya itu? ‘Per litre?’. Yaya answers in overlap at line 21 (‘thirteen {thousand rupiah}’).

As noted above, there are two questions produced by Titi that are accompanied with touch. Ima's shop is quite cramped so the participants are oriented such that Yaya cannot gaze at both Titi and Ima at the same time. Her first question at line 11 is thus similar to those discussed in the previous section in that the touch accompanying this question deals with this problem of mutual orientation. Titi uses touch (as well as a post-positioned address term) to solicit the gaze of Yaya in order to engage her as next speaker.

At line 18, however, Titi uses touch to pursue a response to her question from lines 15 and 16. As she produces this second question, Titi is gazing at Ima but Ima is not returning her gaze. (Note that Yaya aborts her turn at line 14 because Titi's question to Ima overlaps her). Like Ical in extract (4), Titi revises her question such that it now asks about the former price of kerosene. With the addition of touch, she redirects her revised question toward Yaya who is closer to her, and who previously displayed knowledge about the price. Note too that as Titi produces the revised question, Yaya, who had been gazing toward Ima, shifts her gaze to Titi before responding, which she does after mutual gaze has been established. Titi's touch is thus effective in soliciting Yaya's orientation and in eliciting an answer to her question.

The final extract in this section includes touches that do not solicit the gaze of the target and the eventual, dispreferred response is delayed. The conversation is between four students, Turi, Dini, Dira, and Caca, who are all classmates. Before this extract, they had been talking about rain and floods in South Jakarta. A drop of water then hit Turi, which she thought was a raindrop. When Dini, Dira, and Caca question whether it was rain or a droplet from the tree above, the four girls begin looking up into the tree. Dini asks questions accompanied by touch at lines 6, 8, and 11. The initial touch deals with a problem of mutual orientation while her subsequent touches pursue a response. As the extract begins, Turi jokes that fruit might fall on them because they are sitting under the tree.

Dini's first touch (Image 16b, line 6) is laminated onto a question which inquires as to what type of tree they are under. This transpires as she and Turi gaze up into the tree. Despite the girls’ arrangement in semi-circular F-formation, their preoccupation with the tree impinges on their mutual orientation. Thus Dini's fist on Turi's knee attempts to target her as the recipient of the question and select her as the next speaker. Dira also asks a similar question at line 7, but it is not addressed to anyone in particular. As Turi continues to inspect the tree, Dini repeatedly touches her several times (Images 17, 18a, and 18b). Thus, despite Dira offering a candidate type of tree at line 12, these touches display Dini's expectation that Turi should provide an answer. Her persistent pursuit of an answer may derive from Turi's earlier claim (at line 1) that the tree has fruit—a claim that Dini immediately calls into question. So perhaps Dini's questions provide Turi with opportunities to support (or abandon) her earlier claim. Nonetheless, Dini's repeated touches from line 8 onwards (and her gaze at line 13, see Image 19) display that she is pursuing a response from Turi, pressuring her to take the floor and produce a turn. Eventually at line 14 Turi produces an account for not having provided an answer, which is a type of dispreferred response. The delayed response is pipped at the post by Dira's answer at line 12. Dira, who is gazing up into the tree, does not observe Dini tapping on Turi's knee. Dira's self-selection means that Turi's eventual reply is not ultimately produced as the ‘next’ turn, even though the response is elicited under turn-taking rule 1a. In summary, Dini's first touch at line 6 principally deals with the absence of mutual orientation, as do the subsequent touches, but these subsequent touches pursue the absent response by ramping up the pressure to respond.

In this section, we have shown that touch may be used by question producers to pursue responses from question recipients. As in the previous section, some of these pursuits (e.g. in extracts (5) and (6))—but not all of them—relate to problems with mutual orientation. Furthermore, the seating arrangements in extracts (4) and (6) are F-formations, so a complicated seating arrangement is clearly not the only reason participants deploy touch when attempting to engage would-be next speakers. Touch appears to be a highly effective strategy when resources like ‘interrogative lexico-morphosyntax, interrogative prosody, recipient-focused epistemicity, and speaker gaze’ (Stivers & Rossano 2010:4) have not been sufficient to gain a prompt response. It also seems that touch offers a subtle strategy for readdressing questions, shifting the responsibility for responding in a way that is clear to the new recipient, but without necessarily making this explicit to the previous one. Importantly, touched recipients usually return their gaze to the speakers before providing responses to their questions.

All examples thus far reveal that although touching generally takes place within eyeshot of at least some non-selected recipients, it generally transpires within their peripheral vision. This is because the gaze of both addressed and non-addressed recipients tends to be drawn toward the speakers’ eyes and mouth. Consequently, one affordance of the manuo-tactile modality is that a public-yet-private connection is established between the speaker and the target. The intimate nature of this tactile connection is available only to the person being selected to speak, and it is this intimate connection we examine in the next section.

PERSONAL AND POTENTIALLY DELICATE QUESTIONS

In this section we revisit the potentially delicate nature of certain questions. Recall from extracts (1) and (2) that touch was employed with questions which enquired about whether particular persons were unwell or actually alive. These are sensitive matters that were broached with touches which would appear to convey concern about the referents’ wellbeing. Like those examples, certain ‘touching-questions’ in the following extracts belie an intimate quality that signals respect, affection, or admiration to the addressee, or displays empathy or concern for the wellbeing of the referent or the addressee. Some of these touches are deployed when new topics are raised for discussion. As with the earlier extracts, these touches still tend to elicit recipient gaze, but the lead-up to the questions are not precipitated by obvious problems that might impinge upon the attainment of mutual engagement. Instead, these interpersonal touches seem principally devoted to conveying tenderness or affability, particularly through the additional attainment of mutual gaze.

In extract (7), we return to the neighbors Ami, Sari, and Lela. Their prior discussion about fasting (for Mawlid, the birth of Prophet Muhammad) has run its course (at line 1). At line 3 Ami addresses a question to Sari that raises a new topic which is possibly sensitive. As she does this, she touches Sari on the arm.

Once the earlier discussion has run its course (line 1) Ami turns her gaze to Sari—who is gazing at Lela—and asks a question at line 3 about Sari's child, Hendra, touching her as she does so (Image 20). Ami extends her turn in line 5, Baek? ‘Good?’. Sari gazes toward Ami (Image 21) and replies saying that she assumes he is better. She also accounts for not being very sure by saying that she has not called him (line 9). From there she adds Ga tau ‘{I} don't know’ (line 11) but says nothing further on the matter.

In extract (7), Ami and Sari have been exchanging turns with one another while Lela is positioned off to the side. Although Sari is gazing towards Lela as the question begins, her body remains broadly oriented towards Ami. In addition, the topic Ami is raising—Sari's son—has the potential to tacitly select Sari as next speaker, and his name is in fact the first item in the turn. Despite all that, she chooses to touch Sari while asking her topic-initiating question. Why might this be? Evidently, Sari's response (mendingan kali ‘better perhaps’) indicates that Hendra may have been less than good in the past and that his improvement cannot be ascertained. Put simply, this appears to be a delicate topic. Ami's touch might therefore address this delicacy (and the possibility of disalignment and disaffiliation), softening the question and showing Ami's genuine concern about the matter. Sari's lack of expansion on the topic and her claim not to know keep open the possibility of trouble.

In the final extract, the question producer, Ifan, uses touch to convey his interest in the recipient's (Muiz's) personal history. Before this extract, Muiz had been telling Roni about his age, and how he has been getting weaker as he gets older. Roni jokingly suggests the reason for Muiz's weakness is that he quit smoking too early. Mali then tells Ifan that he has never seen Muiz smoking.

At line 1, Mali lists Muiz's children (Keli, Isit, and Badrun) who do not smoke. Muiz confirms this in line 2, saying that none of his children smoke cigarettes. At line 3, Ifan turns his body toward Muiz and asks him whether he smokes. Most of Ifan's question is produced in overlap with Muiz who, initially facing Roni, asserts in line 4 that he used to smoke. Midway through his question Ifan touches Muiz on the knee (Image 22), which has the effect of drawing Muiz's gaze towards him (Image 23). Having secured his gaze, Ifan prolongs his touch (Image 24) as he reframes the question in the past tense (‘Did you used to smoke?’, line 5), this time adding a post-positioned address term bapak ‘father’. At line 6, Muiz raises his eyebrows and answers, saying that he quit smoking in 1992. It is during this response that Ifan retracts his hand (see Image 25). He then receipts Muiz's claim at line 7, and again at line 10. At line 8 Roni adds that Muiz used to be a heavy smoker. Over lines 11 to 13, Roni again says jokingly to Muiz that quitting smoking has caused his weakness, which Muiz mildly resists at line 15, sehat? ‘healthy?’, before laughing.

In this extract, Ifan has contorted his body position (Schegloff 1998) such that he is tilted towards Muiz. As Ifan commences his question at line 3, Muiz is gazing at Roni. However as he feels Ifan's hand, his gaze is drawn toward Ifan as he produces talk at line 4 that is perhaps responding to Ifan's question. Ifan's second question, now framed in the past tense, solicits an answer that is very specific about when Muiz stopped smoking. Furthermore, the post-positioned address term bapak ‘father’ that Ifan uses at line 5 is a respectful form of address used by younger men when addressing older men (Ewing & Djenar 2019).⁴ Despite the second question being perhaps a little redundant, his prolonged touch, the address term, and the reworked question help elicit additional personal information. Having established mutual gaze at line 3, Ifan sustains his gaze toward Muiz throughout his reply (see Image 25). The mutual gaze followed by sustained unilateral gaze, the respectful post-positioned address term, in addition to Ifan's prolonged touch, all evidence a respectful, personalized interest in Muiz, which—given the potentially detrimental effects of smoking—perhaps inquires as to his wellbeing.

Questions with touch establish a double conduit of connectivity between the question producer and the question recipient; that is, the interpersonal connection established through mutual eyegaze and touch (especially through sustained touching) affords a more personal and profound connection than is requisite for merely securing recipiency. The touched recipients in these extracts answer their questions (appropriately, as next selected speakers) but without any observable reaction to the physicality of these intercorporeal enquiries. Because interpersonal touches actually (or at least, potentially) impinge upon one's fellow interlocutors’ personal space, in the following sections we consider cultural and environmental dimensions of the phenomenon we're exploring, and how they might impact on the more general or universal organization of conversational turn-taking.

TOUCH AS A TOOL OF ENGAGEMENT

The production of turns-at-talk with few gaps and minimal overlap, by different speakers, is a miraculously precise undertaking that yields its output predominantly within the audio-vocal modality. Yet the mechanisms required to bring this off within face-to-face multiparty interaction are inherently multimodal. The first of Sacks et al's (1974) turn-taking rules is a case-in-point. Unless the methods of recipient designation are tacit (Lerner 2003), or deployed using address terms like names or kinterms (Sacks et al. 1974; Hamdani et al. 2022), sighted recipients are generally reliant on gaze to determine specifically who is being engaged to speak next. Thus speaker-gaze toward the target, pointing gestures, elevated prosody, and vocative expressions that draw recipient gaze toward the target are engagement tools that involve the visuo-spatial modality (Blythe et al 2018; Auer 2018), as does touch when layered onto questions directed towards specific targets. In this respect, touch as a tool of engagement merely adds the manuo-tactile modality to an enterprise that is already inherently multimodal. Like the other gaze-drawing tools, touch principally serves to indicate which recipient is being selected to respond while the design of the question indicates how they should respond.

We have demonstrated that the Indonesian question producers in our corpus use touch in a variety of ways. Touch is employed to deal with problems of mutual orientation between question producers and recipients, to pursue responses to questions, and to inquire about delicate or personal matters. Touch appears to be especially useful for overcoming challenging seating arrangements that disfavour mutual gaze, and for changing the recipiency of a question. In addition, it may be useful for designing actions that may risk disalignment and/or disaffiliation (cf. Li 2020).

Like Li (2020), the present study has demonstrated that touch can be employed for special, additive tasks in action formation. Our analyses have shown that certain touches accompanying questions are less driven by next speaker recruitment than by the relational import of the question. By including a touch (or touches), the question producer can guide a recipient to inspect a question for special import; not just questioning ‘simpliciter’ (see Schegloff 1996). Like the process of action formation in general (Enfield & Sidnell 2017), the nature of this import (e.g. delicacy or affection) is realized in and through the ‘thick particulars’ (Lerner 2003:190) of the conversation.

CULTURE, THE ENVIRONMENT, AND TOUCHING BEHAVIOUR

Finally, we consider whether the use of touch in next speaker selection might have a cultural dimension, or whether other factors like the environment might play a role. The central question is whether Indonesian people touch each other in conversation more often than people from other cultures. And if so, what would be driving this? In the Indonesian data presented here, it is quite clear that laminating touch onto questions has a robust effect of drawing the gaze of the touched recipient toward the speaker. However, we do find substantial variation in gaze behaviour across different cultural groups. For instance, Meyer (2017:153) characterises the Wolof of Senegal as having a ‘low-gaze’ culture where ‘[i]t is not uncommon to briefly or constantly touch one's interlocutor’. He suggests that in Wolof conversations touch is preferred over gaze for next-speaker allocation, although the application of touch to questions is not discussed. In comparative research specifically focusing on gaze's relationship to questions, Rossano, Brown, & Levinson (2009) compare conversations in Yélî Dnye, Italian and Tzeltal. The Yélî Dnye and Italian data contained high volumes of mutual gaze as the participants often faced each other, while participants in the Tzeltal conversations prefer side-by-side seating arrangements and tend to avoid mutual gaze. A range of factors were identified that increase the likelihood of securing mutual gaze across question-answer pairs but touch (in these dyadic conversations) was not reported upon. The relationship between touch and questions has seldom been investigated previously (however see Chen's (2022) dissertation on Mandarin conversation).

Halls’ (1966) theory of proxemics was the basis for large-scale cross-cultural studies by Sorokowska and colleagues (2017) on preferred interpersonal distances, and by Sorokowska and colleagues (2021)⁵ on the prevalence of different types of affective touch (hugs, embraces, kisses, and stroking) vs. different types of social relationships (one's children, partners, male and female friends). These studies reveal that people in warmer climates exhibit closer contact behaviour—with more touch—than people from colder climates, and that kisses and stroking are more prevalent between partners and between parents and their children. Across the board there is evidence that women touch women more than men touch men (Sorokowska et al. 2021) and there is strong evidence for cross-cultural differences in preferred interpersonal distances, although older people who are acquaintances or intimates generally prefer greater distancing than younger people (Sorokowska et al. 2017, see also Remland, Jones, & Brinkman 1995). Sorokowska et al. (2021) find that conservative values and religiosity are corellated with reduced touching between the sexes. These surveys (and other earlier studies)⁶ are reliant on participants’ self-reports and are not designed to deal with frequency of touches. They do however strongly support environmental factors as influencing touching behaviour. As far as we are aware, there are no corpus-based studies that explicitly compare touching behaviours across languages and cultures within multiparty interactions. Be that as it may, if compared to the CIARA project's Australian Aboriginal language and ‘remote’ Australian English corpora (Blythe, Stirling, Mushin, & Gardner n.d.), in which questions accompanied by touch are either vanishingly rare or absent, the volume of touches in this Indonesian corpus is particularly striking.

The CIARA English and Australian language conversations were collected in the remote outback where population densities are extremely low.⁷ By contrast, these Indonesians conversations were conducted in Jakarta, a built-up urban environment with a population density of 13,000 people per square kilometre (Martinez & Masron 2020). Conversationalists who are familiar with one another are often seated within close proximity, in cramped locations that are not necessarily conducive to the return of mutual gaze (Kendon 1990; Blythe et al. 2018), but are very conducive to physical engagement. These environmental factors may be consequential in accounting for the high incidences of touching, none of which preclude against cultural norms and values also contributing to the prevalence of touch. As potentially causal factors, the environment vis-à-vis culture need not be thought of as in conflict.

Culture is enacted within social interaction (Schegloff 2006) by participants who occupy the places and spaces within the environments they live in—thus their relationship with the landscape is socially constructed. This notion lies behind Palmer and colleagues’ sociotopographic hypothesis (Palmer, Lum, Schlossberg, & Gaby 2017): namely, that the ways humans interact socially with their physical environment influences how they conceptualise space and the ways they talk about places and locations, particularly in their choices of spatial frames (see also Lum, Palmer, Schlossberg, & Gaby 2022). Relatedly, de Dear and colleagues (2021) suggest that ease with which ‘remote’ (non-Aboriginal) speakers of Australian English and ‘remote’ speakers of the Aboriginal languages Gija and Murrinhpatha point accurately to distant locations within conversation does not seem to be determined by language, nor culture, but does reflect participants’ long-standing social engagement with their environments. Perhaps the social relationships people have with their physical environment also influences certain fundamentals of social interaction, such as how participants conduct themselves when taking turns at talk.

Our report contributes toward the study of turn-taking practices across languages and cultures. The cross-cultural/cross-linguistic enterprise seeks to understand which dimensions of language use are universal and which exhibit variation. When we encounter apparent variation, the next step is to determine whether such differences are culturally determined, or whether alternative explanations are more appropriate. Perhaps these data show ordinary practices that are otherwise predominantly audio-vocal and visuo-spatial, being adapted to the compact urban environment; which is not to suggest that these practices should be unique to participants occupying crowded compact environments, rather that the ways people inhabit the environments they socialise in enables certain practices to flourish, and others to become disfavoured. Our prediction is that other urbanised societies with high population densities may also favour the use of touch in next-speaker engagement. We hope this multimodal investigation into the nexus between questions and touch provides fertile ground for future cross-cultural comparisons.