In this chapter, we review work that explores communication accommodation through talk, using a variety of qualitative methods (see Gallois & Giles, Reference 121Gallois, Giles, Tracy, Ilie and Sandel2015, for a recent review; see also http://catqualstudies.weebly.com/ for a list of many papers using qualitative methods to study communication accommodation). In this way, we engage a substantive body of research which has been conducted somewhat independently of the quantitative work for which Communication Accommodation theory (CAT) is, arguably, best known (see Chapters 3 & 4).
There are many forms of qualitative research, ranging from thematic analyses of interviews through to analyses grounded entirely in naturally occurring conversational behavior (e.g., ethnography of communication; discourse analysis, DA; conversation analysis, CA). Studies framed by CAT have often extended sociolinguistic research on language variation that examines the association between phonological, lexical, and syntactic variation to (usually pre-existing) social categories. Other studies have used one or another form of ethnography or DA, and have linked interpretations of talk to existing or emerging social categories and to the dynamics of positioning in intergroup conversation; we discuss a number of such studies below.
In addition, we afford special emphasis to CA, a methodology with its own long research tradition (see Sidnell & Stivers, Reference Sidnell and Stivers2012), and we include a new CA analysis of data on encounters between American police officers and drivers. This analysis highlights two issues: (1) the ways in which CA can be compared to CAT (see Chapter 2) and (2) the potential benefits to researchers of accommodation in using this well-validated method. We also point to some of the differences between CA and other qualitative research framed by CAT. In addition, we review DA studies, along with other qualitative methods.
As many of the chapters in this volume show, CAT has long been known for its subtle experimental studies of the communicative moves people make in interpersonal and intergroup contexts; it can be regarded as a robust intergroup theory of interpersonal communication (see Chapter 7). It would be wrong to assume, however, that CAT research does not explore the richness of talk. On the contrary and across the world, there is a long and strong tradition of qualitative research framed by CAT. This work dates at least from the early 1980s (e.g., Coupland, Reference Coupland1984; Shockey, Reference Shockey1984). Both these studies linked detailed phonological analysis to intergroup convergence. In Coupland’s research, a travel agent changed phonological features toward those of her customers. This was likely to be in the interests of better customer relations and sales, although Coupland only examined behavior and not attitudes or motivation. In Shockey’s research, an American converged to the British exploded dental stop /t/, presumably in the interests of better comprehension in the United Kingdom (again, only behavior, not motivation, was analyzed). Indeed, the special issue of the International Journal of the Sociology of Language (Giles, Reference Giles1984) on communication accommodation, a seminal publication in the history of the theory, contained several articles analyzing discourse and talk. The emphasis put by CAT on actual behavior in interaction made the theory very attractive for scholars in sociolinguistics, as well as in psychology and communication.
Discourse Research and CAT
In contrast to the experimental and questionnaire-based research that characterize the quantitative work on CAT, qualitative studies have always emphasized the details of linguistic behavior in situ, and the impact of identity dynamics on interaction and communication. This work began with the detailed observation of phonological changes in intergroup contexts, as illustrated by the papers above. Soon, there was DA work on ingratiation and over-convergence, exemplified by Platt and Weber’s (Reference 122Platt and Weber1984) study of Anglo-Saxon employers in Singapore and their (unsuccessful or misplaced) attempts to converge to local versions of English. They found a similar situation with the Singaporean employees, whose attempts to converge to Australian English were perceived negatively, as ingratiating. Studies of underaccommodation also appeared, such as the detailed exploration by Hamilton (Reference Hamilton, Giles, Coupland and Coupland1991) of communication between a daughter and her mother, who had dementia. In an interpretive discourse analysis, Hamilton invoked the non-approximation strategies (see below, and Coupland, Coupland, Giles, & Henwood, Reference Coupland, Coupland, Giles and Henwood1988) by analyzing the daughter’s response strategies to ambiguous pronoun references, off-topic comments, or misunderstandings by the mother. Thus, her work was particularly concerned with interpretability. Hamilton, indeed, was one of the first (if not the first) to invoke the concept of unintentional underaccommodation, with specific reference to communication-disabled people like those with dementia. This research made clear the benefits of close analysis of ongoing conversation in situ to explore the dynamics of accommodation.
The first systematic theoretical statement of CAT to take full account of research on discourse and talk was probably Coupland et al.’s (Reference Coupland, Coupland, Giles and Henwood1988) framework. This version of the theory explored the non-approximation strategies of interpretability, discourse management, and interpersonal control, adding them to the original approximation strategy (as convergence, divergence and maintenance were re-labelled). This work was significantly based on DA of talk between cognitively and socially active elderly people attending senior citizen day centers and their younger recruited interlocutors. Partly for this reason, qualitative work on CAT became associated with communication in aging and health. For example, it produced important work on the dilemmas of accommodative strategies and intergenerational communication – simplifying language for interpretability could also be demeaning or result in reduced opportunities for social engagement (Ryan, Giles, Bartolucci, & Henwood, Reference Ryan, Giles, Bartolucci and Henwood1986). This work not only added to the complexity of CAT, but also to its comprehensiveness.
In the 1980s and 1990s, qualitative studies using CAT were carried out mainly by sociolinguists, and emphasized the type of DA that is typical there. These scholars examined the ways in which language changes were linked to movements toward or away from social groups. This work continues to the present day. For example, Chakrani (Reference Chakrani2015) found that speakers of various dialects of Arabic with higher or lower social prestige converged to or diverged from each other’s dialects in conversation as a function of their language attitudes. In an interesting example of triangulation, Chakrani used a very detailed interpretive DA around lexical choice, supplemented by analysis of nonverbal behavior and elicitation of attitudes through interviews and questionnaires. In one example, speakers of high-prestige and low-prestige dialects co-constructed a joke about a word with different meanings across dialects. The high-prestige speaker showed a positive attitude to the other speaker’s convergence, and also to her dialect. A third speaker (of another high-prestige dialect), however, diverged from the low-prestige speaker by refusing to understand the joke, and indicated a negative attitude to the low-prestige dialect.
In a similar vein, and using both CAT and CA, Nilsson (Reference Nilsson2015) provided both a broad examination of language variation from nine areas in western Sweden, and a detailed study of convergence of language variants in particular interactions. Initially, Nilsson used sociolinguistic interviews to establish the phonological variants that speakers typically used. Then, conducting a detailed CA of recordings of informal contexts like family dinners, she found that people who typically used different variants converged to those of their interlocutors when they repeated words or parts of phrases used by interlocutors in the immediately preceding turn; that is, they format tied. This convergence occurred when interlocutors were oriented to the same immediate conversational goal. It sometimes occurred when there was disagreement, possibly as a signal of solidarity, given that convergence did not occur in all format tying. Given the intergenerational nature of the conversations, Nilsson pointed to the potential of this kind of convergence in explaining language shift, and she highlighted the usefulness of CAT for this type of analysis. This work points to the need for future research in order to understand accommodation in linguistic variants as an interactional resource in conversation.
Once again in the intergenerational context, Sandel, Chao, and Liang (Reference Sandel, Chao and Liang2006) looked at code-switching among multilinguals across generations in Taiwan, noting the convergence from local languages to the more prestigious Mandarin, which was greater in the younger generations. In this case, younger people showed their identification with the larger society by speaking Mandarin, which meant that in some cases they diverged from their own (older) family members by not code-switching back to the local language.
What emerges is that behavioral convergence in naturally occurring conversation, which may not be deliberate or even within awareness, is associated with identity and fluctuations in it. Convergence may also be a practice to accomplish actions such as agreements, or to manage social solidarity in the face of disagreement. Other sociolinguistic studies have examined other features of language than the sound system. For example, in a critical sociolinguistic study, Rajadurai (Reference Rajadurai2007) examined shifts in syntax and semantics, as well as phonology, to outgroup variants of Malaysian English in classroom conversations. Rajadurai noted that teachers converge downward to the variants of students for strategic reasons, such as added respect and control. It would be difficult if not impossible to achieve this kind of understanding by other means.
The different approaches of experimental/survey and qualitative studies on CAT in focus and emphasis meant that the two strands of research have been conducted largely independently of each other. There were relatively few mixed-methods studies, although quantitative and qualitative studies often appeared alongside each other in the same edited books (e.g., Giles, Coupland, & Coupland, Reference Giles, Coupland and Coupland1991). In this century, on the other hand, there has been an increase in the number of qualitative studies, a greater diversity of methods and contexts, and a coming-together of experimental and survey research combined with detailed studies of discourse (e.g., Gasiorek, Reference Gasiorek2013; Hewett, Watson, & Gallois, Reference Hewett, Watson and Gallois2015). Furthermore, there are now software tools (e.g., Wordsmith, Word Cloud, Leximancer, Discursis) that facilitate the analysis of language and talk, and these have also contributed to the increase in the analysis of qualitative data.
Approximation
Perhaps unsurprisingly, much of the qualitative work on accommodation is about approximation. This kind of accommodation was studied first, and for many scholars it is still the part of the theory they know best. In a novel approach, Azuma (Reference Azuma1997) studied in detail the speeches of Emperor Hirohito of Japan. He found that, contrary to expectations, the Emperor converged to lower-status versions of Japanese – as Hirohito spoke a version reserved for the Emperor, this behavior amounted to downward convergence to everyone. It is likely that Hirohito, in the light of changes after the defeat in World War II, endeavored in this way to present a more accessible image to his people. He seems to have been successful, as his popularity increased with his use of these versions of Japanese.
In a different context, Yaeger-Dror (Reference Yaeger-Dror1988) found, in a study of approximation in Israel, that immigrant speakers in the mass media (singers) from Arabic linguistic backgrounds first converged to speakers of the dominant form of Hebrew, the koine, in their use of particular sounds that differed in Hebrew and their first language. Then, as the confidence and perceived ethnolinguistic vitality (see Giles, Bourhis, & Taylor, Reference Giles, Bourhis, Taylor and Giles1977) of the immigrant groups in the larger society increased, the singers became more likely to maintain their dialect versions of the sounds.
There has been much research like this, often in classroom situations and involving detailed observation of convergence. In an interesting non-classroom case, Bonnin (Reference Bonnin2014) looked at convergence to the other’s voice, mainly in terms of word choice and style, between psychiatrists and patients in consultations in Argentina. Findings indicated that both interlocutors attempted to adopt the terms and the level of complexity used by the other (perhaps resulting in some overaccommodation, although the study did not look at this). Interestingly, Bonnin describes the study as not about CAT, as (in his view) CAT studies are experimental. This misunderstanding is not unusual among qualitative researchers, and an important activity for scholars interested in CAT is to correct this misperception.
Overaccommodation
Almost as soon as CAT was initially put forward by Giles (Reference Giles1973), it became clear that the straightforward idea of convergence to people one liked or groups one wished to join, and divergence from those who were disliked, was – however elegant a concept – too limited to capture the panoply of interpersonal and intergroup interactions. First, people differ in the extent to which they have the skills to accommodate; in addition, they may have complicated motives for accommodating or not. Beyond this, interlocutors differ in their reactions to a given behavior in talk, as a function of their own attitudes and motivations but also of the conversational dynamics and what actions are being oriented to as relevant by people in the interaction. The complexity, evident in research on conversational behavior, began to be reflected in the theory. One of the early additions to the theory was overaccommodation (see Chapter 5), in which speakers converge (or accommodate in other ways) to stereotypes about their conversational partners, rather than to the partners’ actual behavior (Ryan et al., Reference Ryan, Giles, Bartolucci and Henwood1986). Platt and Weber (Reference 122Platt and Weber1984) had earlier observed this in Singapore, but Ryan and her colleagues explored it in detail through conversations between older people and health professionals, leading to a CAT-satellite model called the Communicative Predicament of Ageing (CPA), in which older people are subjected to stereotyped communication from others, no matter what they do themselves (see Chapter 7).
Patronizing communication – the use of intonation patterns that resemble baby talk (“elder-speak”) and words and phrases that infantilize the care receiver, have been explored in depth as indications of overaccommodation (Giles & Gasiorek, Reference Giles, Gasiorek, Schaie and Willis2011). Early work indicated that how words are perceived was crucial. For example, Edwards and Noller (Reference Edwards and Noller1993, in a quantitative study of impressions) found that elderly people perceived elder-speak as less patronizing than did younger nursing or psychology students. Such perceptions are central to conversational dynamics, and there was a great opportunity to explore overaccommodation and reactions to it in ongoing talk. Unfortunately, this path was not taken, and research on this complex form of behavioral accommodation has been relatively uncommon since the 1990s. A recent DA study in Quebec by Lagace, Tanguay, Lavallée, Lapante, and Robichaud (Reference Lagace, Tanguay, Lavallée, Laplante and Robichaud2012), however, found that overaccommodation in the form of patronizing communication is alive and well in aged care homes, and that residents resent it, but accommodate to it, perpetuating the CPA. In a theoretical development, Ryan and her colleagues (Savundranayagam, Ryan, & Hummert, Reference Savundranayagam, Ryan, Hummert, Weatherall, Watson and Gallois2007) extended the model to the predicament of people with disabilities (Duggan, Robinson, & Thompson, Reference Duggan, Robinson, Thompson and Giles2012; see Chapter 7), and there have been a few studies looking directly at discourse from this frame.
In recent years, researchers have begun to combine sociolinguistic and social-psychological methods in the detailed study of overaccommodation in talk. For example, Rosemary Baker and her colleagues (Baker et al., Reference Baker, Angus, Smith-Conway, Baker, Gallois, Smith, Wiles and Chenery2015) conducted a DA of conversations between people with dementia and their professional carers in aged care homes. They found that overaccommodation – in this case, inappropriate use of questions apparently based on stereotyped expectations – resulted in less engagement in the conversation by care receivers than did accommodation by searching for topics and words that care receivers engaged with. Accommodation by carers resulted in more engagement, even by people with severe dementia. The study made use of one of the newer software tools for analysis of discourse, Discursis, which allows visualization of conversations turn-by-turn (see Angus, Smith, & Wiles, Reference Angus, Smith and Wiles2012, for an explanation of this software). Angus and his colleagues (Angus, Watson, Smith, Gallois, & Wiles, Reference Angus, Watson, Smith, Gallois and Wiles2012) also used Discursis to visualize accommodation in conversations between health professionals and patients, and found that convergence in word use was associated with conversations perceived as more effective.
Underaccommodation
As research has continued, it has become clear that underaccommodation – the failure to move toward the communication of a conversational partner or to take account of the partner’s conversational needs – is probably the most ubiquitous form of nonaccommodation, and indeed of miscommunication in general. Underaccommodation in approximation was originally called maintenance, and tended (in studies of perceptions of accommodation in language and accent) to have similar (albeit less extreme) consequences as divergence (i.e., negative perceptions). In fact, this behavior is far more complicated than a simple refusal to change language or style. In recent years, Gasiorek and her colleagues (see Chapter 5) have found that the perceived motive for underaccommodation can be the most important influence on evaluations of it. For example, Gasiorek (Reference Gasiorek2013, study 1), in a thematic analysis of open-ended questionnaire responses to underaccommodation, identified six main responses to this behavior, based on perceived motive. Responses ranged from requests for clarification to open confrontation.
Underaccommodation has emerged as a key feature of discourse in health and disability. The study by Baker and colleagues described above shows that responses to underaccommodation by people with dementia is central to their engagement in a conversation. This work is similar to the work by Hamilton (Reference Hamilton, Giles, Coupland and Coupland1991), although the more recent study is more explicitly framed as underaccommodation. Likewise, Cretchley, Gallois, Chenery, and Smith (Reference Cretchley, Gallois, Chenery and Smith2010) examined conversations between people with schizophrenia and their carers, using the data-mining software Leximancer to map the parts of the conversations most associated with the carer and the person with schizophrenia. They found that accommodative responses – for example, questions around topics the care receiver showed interest in – were successful in producing conversational engagement by these underaccommodating people.
Others have studied underaccommodation as miscommunication. Susan Baker and her colleagues (Baker, Gallois, Driedger, & Santesso, Reference Baker, Gallois, Driedger and Santesso2011) conducted a DA of interviews with people with arthritis and similar diseases and their doctors (also using Leximancer). They found numerous cases of subtle miscommunication. For example, they noted that both patients and doctors maintained their different uses of the word “normal.” For patients, “normal” meant back to normal, or back to where they were before the diagnosis. Doctors, on the other hand, used “normal” to mean a new normal after the diagnosis. The miscommunication resulting from this underaccommodation by both sides passed largely unnoticed. On the other hand, patients did feel that they needed to accommodate (“reach out”) to doctors in their language.
In a different context, Hewett and colleagues (Reference Hewett, Watson and Gallois2015, study 2) conducted a DA of underaccommodation and reactions to it, in medical record entries written by doctors of different specialties. They found, first, that doctors who were not from the specialty of the doctor who made an entry had difficulty in understanding the ingroup terms, abbreviations, and “hidden agendas.” Generally, however, they excused the other doctors’ behavior by saying that it was appropriate to that specialty, or that they did not really need to understand the entry. In this way, they could maintain solidarity with their (unseen) colleagues; that said, the impact on patient care is potentially more problematic.
Impact of Qualitative Research on CAT
Qualitative studies that examine actual language behavior in situ have importantly shaped the theoretical development of CAT. This research has highlighted the subtlety of intergroup communication in many contexts. For example, Watson and her colleagues (see Chapter 8) elucidated the importance of emotional expression and attention to emotional needs in conversations between health professionals and patients. In recent mixed-methods work with Discursis, Watson, Angus, Gore, and Farmer (Reference Watson, Angus, Gore and Farmer2015) found that convergence to an optimal level was most effective in dealing with open disclosure explanations provided by health professionals to patients or their families about medical errors and adverse events. In the next section, we take a different approach, and invoke CA to elucidate accommodation in natural interactions.
Conversation Analysis and Accommodation
The studies reviewed so far have, to a greater or lesser extent, taken account of features of CAT outside the immediate conversation, even though their main emphasis has been on the dynamics of the talk in situ. Indeed, recent versions of CAT (Gallois, Ogay, & Giles, Reference Gallois, Ogay, Giles and Gudykunst2005; Giles & Soliz, Reference Giles, Soliz, Braithewaite and Schrodt2014) have theorized conversational tactics and dynamics as important constraints of accommodation, that may modify or even negate initial orientation to the conversation. Nevertheless, many studies have taken group memberships like ethnicity, gender, age, or occupation as starting points from which to consider language in conversation (see Chapter 7). CA eschews this process, concentrating on participants’ orientations to relevant actions in talk (see Weatherall, Reference Weatherall and Giles2012). We explore this method empirically in the next sections, in order to explore the potential of CA to add value to research with CAT; one of us (AW), an experienced conversation analyst, conducted the CA, using data collected by Giles, Linz, Bonilla, and Gomez (Reference Giles, Linz, Bonilla and Gomez2012). The aim of what follows was to use CA to ground subjective attributions of accommodation in what people are doing in talk. The interactions we examined were two American traffic stops. The officer in one interaction received the highest accommodation score of all the traffic stops (ratings were provided by naïve judges; see Giles et al., Reference Giles, Linz, Bonilla and Gomez2012). For the second case, the driver received the highest nonaccommodation score.
Traffic stop interactions are a form of institutional talk (see Choi & Giles, Reference Choi, Giles and Giles2012). The study of institutional talk is important in conversation analysis, and this method is used to examine how institutions are produced by participants as they orient to the joint accomplishment of its business. For example, Weatherall (Reference Weatherall and Moore2015) showed how call-takers for a dispute resolution service accomplished a neutral stance toward callers’ complaints. In the analysis presented below, we propose that participants’ (mis)alignment to early actions initiated by traffic officers are likely causes of raters’ strong impressions of accommodation and nonaccommodation.
Category Membership, Social Identities, and the Accomplishment of Action
A CA approach to the study of social interaction involves limiting the significance of identity matters to the social or category memberships observably relevant to the participants themselves in the mutual accomplishment of action. In this way, CA is different from (and arguably antithetical to) some other forms of qualitative analysis, where talk is interpreted in terms of existing or theoretically determined categories and concepts. This issue has been particularly controversial in gender and language research, which largely rests on assumptions of the omni-relevance of gender to social interaction (Weatherall & Gallois, Reference Weatherall, Gallois, Holmes and Meyerhoff2003).
The practices recruited in the mutual accomplishment of actions in talk, which CA aims to explicate, involve social and psychological matters that are handled interactionally. Doing things together requires establishing and maintaining “properly aligned participants” (Jefferson & Lee, Reference Jefferson and Lee1981). Being aligned refers to interactants who are jointly engaged in the same activity. For example, Jefferson (Reference Jefferson1988) documented some of the ways troubles were told about in interaction. This work showed that a properly aligned recipient in a troubles-telling displays a similar stance toward the problem as the teller. A properly aligned troubles-telling recipient shows agreement and intersubjective understanding of the teller’s problem and stance. In the analysis that follows, the alignment between police officer and driver is central to the course of the interactions.
Accommodative and Nonaccommodative Traffic-Stop Interactions
For this chapter, two traffic stop interactions were identified from a larger sample (n = 313) included in a study of the influence of ethnicity in police–civilian interactions (Giles et al., Reference Giles, Linz, Bonilla and Gomez2012). In the original study, data had been content-analyzed for pre-determined information, behavior had been rated by judges for communication quality, and comparisons had been made based on whether the interactions involved white or Latino interlocutors (see Chapter 9). The analysis here, however, is entirely grounded in the details and sequencing of the talk.
The recordings of the traffic stop interactions – involving two different white male officers and two different white male drivers – rated as most and least accommodative were transcribed using the Jeffersonian system (Jefferson, Reference Jefferson and Lerner2004), which is conventional in conversation analysis. The two extracts below show the two traffic stop beginnings, which include the opening, the reason for the stop, and getting the driver’s documents.
Extract 1 presents the early moments of the interaction from the accommodating officer case. The driver has complied to the request to pull over with no problems. The recording begins with the officer asking the driver how he is.
Extract 1: Opening of traffic stop interaction rated as the most accommodating (officer)
01 O: How’s it goin? 02 D: How you doin? 03 O: Pretty good. 04 D: (All [right.) 05 O: [Hey >the reason< I stopped you is your, (1.0) (t-) uh: 06 rear tail, (.) license plate has got a- (0.5) hitch in the 07 front and it’s blocking >a part of the< numbers, so it’s 08 hard to see. 09 (0.8) 10 D: [ Really? ] 11 O: [By law- Y]eah. By law you have to have that totally 12 unobstructed. 13 (0.7) 14 O: and: you have a light that usually lights it up.=It looks like 15 it’s not working at the (t-) at the time being. 16 (.) 17 D: (Can I step [out and) 18 O: [Uh: in a second you can.=Do you have your 19 driver’s licens[e, r]egistration and insurance^ 20 D: [Yep.]
A finite set of activities can occur in the openings of social interactions. They are summons-answer sequences, greetings, identification sequences, and personal state inquiries (Schegloff, Reference Schegloff1968). Typically, the openings of institutional interactions are more truncated than mundane conversational openings. For traffic stops, the officer summons the driver by requesting him or her to pull over. In the earlier case, the only other activity typical of conversational openings is a personal state inquiry, which the officer initiates with How’s it going? (line 1). Instead of providing an answer to the question, the driver counters it by asking how are you doing? (line 2). The counter expedites the how-are-you sequence and progresses the interaction by passing the conversation floor back to the officer. The officer responds to the driver’s inquiry with pretty good.
By countering the how-are-you sequence, the driver shortens the conversational opening and immediately gives the conversational floor back to the officer. In doing so, the driver curtails an activity that would otherwise postpone – by just a turn of talk – finding out why he has been pulled over. After responding to the driver’s personal state inquiry the officer continues talking, providing the reason for the stop, which is that the license plate on the driver’s vehicle is not clearly visible.
After a short delay, the driver responds to the reason for the stop with really, a ritualized display of disbelief that regularly occurs in response to news (Wilkinson & Kitzinger, Reference Wilkinson and Kitzinger2006). The officer orients to the delay as the driver not understanding the significance of the infringement, because he starts explaining the problem in terms of the law. He cuts off the beginning of that explanation that is in overlap with the driver’s turn, to respond to the driver’s display of disbelief with yeah. The officer then continues with a second aspect of the problem – the tail light is not working.
In response to the second aspect of the problem, the driver asks if he can get out of his car and have a look. The request is an interrogative designed for a yes answer. It demonstrates the driver accepting the right of the officer to control what the driver can do. The officer projects the nature of the request before it has been fully articulated, interrupting the driver and refusing his request, or at least indicating there will be a delay before it is granted by saying in a second you can. The time formulation second minimizes the time lag which, in reality, turns out to be a lot longer.
Another activity visible in the earlier extract is the officer’s request for the driver’s documentation – do you have your … (lines 18–19). The driver indicates his compliance with the request with a yep early in the officer’s turn. What is meant by early is the driver claiming next speakership before the prior turn is brought to possible completion. The driver is showing recognition of the action being done by the officer. That recognition is because the request has a somewhat routine form that has three expected elements. On the video, it can be seen that the driver is also complying with the request by starting to retrieve the relevant documentation.
In sum, Extract 1 shows a smooth and expedited progression through the activities that occur at the beginning of the traffic stop. The driver is compliant with the officer’s requests. The driver also displays his unawareness of the problems with his vehicle by displaying surprise – really – and asking to be able to confirm it with his own eyes. That request displays the driver’s orientation to the interaction as a traffic stop and his interlocutor’s membership category as an officer of the law. The officer denies that request, but indicates that permission of the request will be granted later. The sensitive handling of denials is a regular feature of dispreferred response types (Schegloff, Reference Schegloff2007), which may contribute to the impression of an accommodating officer. In CAT terms, the two interlocutors have accommodated each other, especially in interpersonal control; they have treated each other as individuals within the constraints of the context. Even though it was the police officer who received the highest score for accommodation, the CA also points to the interactional accomplishment of institutional identities and action. Furthermore, the CA elucidates the ways in which this was accomplished through alignment from the start through the whole opening.
A very different opening is evident in Extract 2, which is taken from the interaction rated as the most nonaccommodating; in this case, the nonaccommodating interlocutor is the driver. Trouble is evident from the start of the recording, where the officer is not satisfied with the way the driver has complied with the request to pull over. When the driver has pulled over satisfactorily, the officer launches straight into the reason for the stop – the lack of a light over the license plate.
Extract 2: Opening of traffic interaction rated as the most nonaccommodating (driver)
01 ((Dispatcher talking on radio)) 02 O: PULL AHEAD! 03 21.0) 04 O: Pull off to the si:de=of the road up he:re. 05 (.) 06 D: ((driver turns to the left to look at officer)) 07 D: (Huh?/Sorry? 08 O: >Phull off to the side of the road up here.< 09 D: <Off to the side of the [road?>] 10 O: [YEAH^ ]>You’re in the >middle of the< 11 LAne^< 12 (0.2) 13 D: Well, (.) (it’s because [you-)° 14 O: [Pull o:ff >°to° the< side of the 15 road over here. 16 (26.0)((Driver pulls over. Officer gets back in police 17 vehicle and follow driver.)) 18 (26.0) ((Officer talks to dispatch before leaving police 19 vehicle again to approach driver)) 20 (17.0)((Officer approaches driver’s car)) 21 O: >I just.< pulled you over >‘cause you don’t have a light 22 over your °license plate°.=It’s not a big< dea:l. (.) but, 23 >do you have< you:r (.) registration, insurance and > 24 (license with you)<? 25 D: °Yup°
At the beginning of the above extract, it is clear that the driver complied with the request to pull over, but did so in a way that was unsatisfactory for the officer. The officer instructs the driver to move the car forward (Pull ahead, line 2), which the driver fails to do. The officer orients to that failure as a problem in understanding, because he expands the instruction, providing more detail on what is required (pull off to the side of the road up here, line 4). The driver claims a trouble in understanding, which the officer addresses by repeating where he wants the driver to move to – off to the side and further up the road. The driver repeats the instruction, which shows that his trouble is not with hearing, but with understanding. The officer addresses the driver’s understanding by explaining the problem with the position of the driver’s vehicle – you’re in the middle of the lane. The driver begins to respond to that explanation with well, a turn beginning that regularly marks some kind of disalignment with the prior turn. It seems likely that a justification was in the offing, but the officer interrupts the driver to reissue the direction about where to move to, which the driver then complies with.
With the pulling-over sequence completed, the officer could engage with other opening activities, such as a greeting or a how are you sequence. However, he curtails the opening and launches directly into the reason for the stop, which is that there is no light over the license plate. The officer assesses the problem as not a big deal. The short silence after that assessment provides a conversation slot where a response to the news about the lack of a light or to the assessment of its severity could be made – this is what happened in Extract 1. In this case, after the beat of silence where a next speaker can talk, the officer continues speaking with a next relevant activity, which is the request for the driver’s documentation. The driver indicates his compliance with that request verbally when the turn has been brought to completion.
The extended pulling-over sequence in Extract 2 is in stark contrast with the same activity in Extract 1. Following the pull-over, the officer launched directly into the reason for the stop – none of the other mundane activities of conversational openings were in the beginning of this interaction. The driver did not take a turn-taking opportunity to respond to the reason for being pulled over; nor did he indicate an early willingness to comply for documentation, which the driver in Extract 1 showed was possible. In CAT terms, the analysis shows that both interlocutors are nonaccommodating (even though it was the driver who received the highest score on this variable). The CA reveals their misalignment with respect to the action.
In both cases, after the officer has secured the driver’s documents, dispatch is contacted to check that everything is in order. The interactions look very different, however, and seem likely to produce different results.
As can be seen from this analysis of the openings, the grounded approach of CA sheds a different light on these interactions. The analysis proceeds from grounded observations of what participants are oriented to as relevant local courses of action, accomplished over sequences of turns of talk. It does this rather than interpreting the talk in terms of its meaning in the interaction and with reference to larger institutional and societal issues, as discourse analysts would do. It also fleshes out the behavior in terms of conversational practices – the drivers’ and officers’ turn-by-turn responses to each other’s behavior – instead of pre-existing strategies based on social identities. In particular, CA does not consider the pre-existing category of ethnicity, which is often assumed to impinge on identity (whether it does so or not). In this way, the CA foregrounds the locally relevant nature of social categories.
Conclusion
The work reviewed in this chapter, and the CA presented here, highlight the contribution that qualitative research has made to CAT, as well as pointing to some directions for the future. For one thing, it is clear that accommodation is partly in the eye of the beholder (both speaker and receiver), but is likely also to be grounded in the accomplishment of actions as they unfold in interaction. There may be moment-to-moment changes in the relevant identities, or in CA terms category memberships, and in orientation to other speakers based on language and communication behavior. There is no substitute for examining conversation in natural settings to capture this aspect of accommodation.
A challenge for researchers using CAT is the complexity of the theory. The complexity arises from the attempt to capture interactional dynamics as well as the antecedents and consequences of an interaction. There is room for much more work on conversational tactics and immediate responses in CAT, and the kind of detailed analysis provided by CA and some forms of sociolinguistic and DA is probably the only way to examine it fully. Some of the new language-analysis software may help researchers to analyze larger bodies of data more efficiently; programs like Discursis, which allow visualization of a conversation turn-by-turn, seem especially promising here. Furthermore, in addition to the ways in which conversational participants make their behavior more similar or different, there are a myriad of ways in which they manage the discourse, help each other to understand the interaction (or not), and allow each other to interact as individuals rather than in roles (or not). For example, how is underaccommodation responded to in succeeding turns in conversational talk? What difference does this make to the overall flow of the conversation? Methods like CA and DA are well-suited to this task.
Much of CAT’s theoretical development has resulted from observations of actual language behaviors. A conversation-analytic approach also insists on grounded observations of actual interactional behavior. Much qualitative research in CAT is based on observations of language, focusing on accommodation at the phonological, lexical, syntactic, and pragmatic levels. CA adds a focus on action – what participants themselves are doing in talk. It also shows that actions are a joint accomplishment; each turn of talk can stall or further an action launched in a previous turn of talk. CA examines what is happening in practice for the participants, turn-by-turn, in talk. How the theoretical differences between CA and the analysis of behavior within CAT can be reconciled is a future challenge, but in any case, there is huge value in this grounded analysis of action in talk.
The Principles of CAT described in Chapter 3 synthesize many years of research in this area, most of it experimental and quantitative in approach. The CA we present here, as well as the emphasis on talk in action in much of the research reviewed in this chapter, indicate the need for more attention to behavior in interaction. Principle 3 comes the closest to interactional dynamics, positing that “the degree and quality of individuals’ accommodation in interaction is a function of both their motivation to adjust and their ability to adjust.” We would add to this another principle (Principle 8), along the following lines: “The degree and quality of individuals’ accommodation or nonaccommodation in interaction is a function of the interactional dynamics, including turn-by-turn actions, interactional accomplishment of immediate conversational goals, and the (mis)alignment of speakers in terms of the personal and/or social identities negotiated.” This means that, as has always been the case, actual talk in interaction is at the very core of CAT.
In sum, qualitative research has been part of CAT from the start and has increased in volume and impact in recent years. This research has been impactful, as the analysis of behavior in interaction has immediate resonance for speakers. The increase in mixed-methods research has also helped to strengthen the links between behavior in situ and studies of its antecedents and consequences. We expect these trends to continue in the future, so that CAT will continue to present a robust analysis of both (non)accommodative behavior and attitudes.
Acknowledgment
We are very grateful for Doug Bonilla’s help in the collection of data on traffic stops analyzed in this chapter.
