Current topics in CA
Conversation Analysis is an incredibly broad field, with scholars pursuing questions about fundamental questions such as how do people arrange who talks when, to very applied questions about how you get someone on board with a counseling service. Its breadth was evidenced at the recent International Conference for Conversation Analysis in 2014 and 2018 where over 500 scholars presented their work. Here I provide a quick glance at some of the broader topics that are currently hot in CA.
Recruitment
Since its inception researchers in CA have focused particularly on the sequences of actions that people use to structure their interactions. Indeed, the notion of conditional relevance that is crucial for our understanding of turn-taking relies on the idea that there are type-fitting pairs, where a specific type of first action asks for a specific type of response: greetings need return-greetings, questions need answers, and requests need assistance. This way of looking at social interaction thus assumes that actions are bounded categories: a turn-at-talk implements either a request, or a question, or an invitation. While people obviously can use their turns to pursue multiple goals—speakers accomplish turn-taking, can take up a stance, and so forth—they generally have one main action as their business, which a response subsequently has to address (see Levinson, 2013 for a nice discussion of this issue).
In recent years, however, various scholars have started to question this approach to social interaction. In anthropology people like Nick Enfield and Jack Sidnell have even questioned the entire notion of action itself as a foundation for social interaction (Enfield & Sidnell, 2017; Sidnell, 2017). The most prominent line of inquiry—as was evidenced by the best paper award at ICCA2018—comes from Kobin Kendrick and Paul Drew, who in a series of papers moved away from the division between offers of assistance and requests for assistance, and instead proposed a framework of Recruitment (Kendrick, & Drew, 2014, 2016; Drew & Kendrick, 2018).
Kendrick and Drew take as their starting point that "it is hardly possible to be copresent with others (...) without needing at some point to enlist someone's help or for someone to help." Assisting each other is a fundamental part of human interaction, and our lives would be a lot harder if we could not rely on other people for minor or major problems we encounter in our day-to-day business. Because helping each other is such a basic part of social interaction, simply displaying trouble may be a means of recruiting another's assistance. That assistance is then neither necessarily requested nor offered, it is simply provided. Indeed, others may even anticipate a potential trouble and take appropriate actions to prevent the trouble.
So instead of seeing requests and offers as different social actions, categorized based on who is the benefactor and who is the beneficiary, we can conceptualize social actions based on a recruitment scale, that reflects (i) whether the trouble is implicit or anticipated, or something in between; (ii) whether assistance by a co-participant is conditionally relevant or merely optional; and (iii) whether assistance is initiated by Self who encounters the trouble or Other. Research in this framework is only just starting, but it is already incorporated in discussions about accountability (Kent & Kendrick, 2016), cross-linguistic studies of language structure (Floyd et al., 2014), and the role of embodied behavior (Drew & Kendrick, 2018).
In recent years, however, various scholars have started to question this approach to social interaction. In anthropology people like Nick Enfield and Jack Sidnell have even questioned the entire notion of action itself as a foundation for social interaction (Enfield & Sidnell, 2017; Sidnell, 2017). The most prominent line of inquiry—as was evidenced by the best paper award at ICCA2018—comes from Kobin Kendrick and Paul Drew, who in a series of papers moved away from the division between offers of assistance and requests for assistance, and instead proposed a framework of Recruitment (Kendrick, & Drew, 2014, 2016; Drew & Kendrick, 2018).
Kendrick and Drew take as their starting point that "it is hardly possible to be copresent with others (...) without needing at some point to enlist someone's help or for someone to help." Assisting each other is a fundamental part of human interaction, and our lives would be a lot harder if we could not rely on other people for minor or major problems we encounter in our day-to-day business. Because helping each other is such a basic part of social interaction, simply displaying trouble may be a means of recruiting another's assistance. That assistance is then neither necessarily requested nor offered, it is simply provided. Indeed, others may even anticipate a potential trouble and take appropriate actions to prevent the trouble.
So instead of seeing requests and offers as different social actions, categorized based on who is the benefactor and who is the beneficiary, we can conceptualize social actions based on a recruitment scale, that reflects (i) whether the trouble is implicit or anticipated, or something in between; (ii) whether assistance by a co-participant is conditionally relevant or merely optional; and (iii) whether assistance is initiated by Self who encounters the trouble or Other. Research in this framework is only just starting, but it is already incorporated in discussions about accountability (Kent & Kendrick, 2016), cross-linguistic studies of language structure (Floyd et al., 2014), and the role of embodied behavior (Drew & Kendrick, 2018).
Embodied action and embodied grammar
Although undergraduate students in communication studies are quickly taught that, contrary to popular belief, the far majority of communication is not non-verbal, it is obvious that communication is not purely linguistic. Language offers us a means of interaction that is more refined, and it is likely that thanks to our capacity for language, humans have built cultures, societies, and microwave ovens. But language is at the very least complemented by our bodily behavior, and plenty of what we do in social interaction indeed works better if we can rely on visual cues, such as facial expressions, gestures, and body torque.
It may thus be considered somewhat surprising that the discipline which places social interaction at its core, has so long paid relatively little attention to the "non-verbal". There were of course pioneers, such as Christian Heath and the late Charles Goodwin, who were analyzing body behavior long before CA was a respectable field. Goodwin demonstrated already in 1979 that linguistic structures are partly constructed in relation to how potential addressees react; people can show through for example facial expressions that they are going to disagree or disaffiliate, and speakers can shift their language in production to accommodate those projected problems (Goodwin, 1979). Since then Goodwin, and his partner Marjorie Goodwin, have done extensive studies of topics like gaze, gestures, and even vision.
Despite their foundational work, it has taken decades for the study of embodied action to become so institutionalized as to be indispensable, the way it is in now. This revolution has to a significant extent been the achievement of Lorenza Mondada who took a systematic approach to embodied behavior, even developing a transcription system to record the details that Conversation Analysts are so interested in (Mondada, 2014). Of course, this revolution would not have been possible without technological advancements; portable video camera's allow for easy and inconspicuous recording, and software nowadays makes it possible to do frame-by-frame analysis, which means that analysts can pinpoint the exact moment some bodily practice was started and can track and describe its movements in detail.
Studies in CA can now broadly focus on embodied behavior in two ways. The main analytical focus can be on some bodily behavior itself, on practices of embodied behavior. There is a tradition of studying pointing gestures that goes back to work by Adam Kendon (2004), and which was further explored by Mondada who examined how pointing is done in for example guided tours or shops. Similarly gaze has received significant attention, particularly as it relates to turn-taking, who talks when. Lerner (2003) started exploring how people can use gaze to select a next speaker, although it is clear from work by Rossano (2013) that the exact role of gaze, particularly in multiparty interaction, is still very poorly understood. But not all subjects need to be so broad: even something as small as an eye roll can be of interest (Clift, 2018).
Alternatively the focus can be how embodied practices are combined with verbal practices to design actions. As with CA in general, this line inquiry does not take a specific practice as its starting point, but investigates instead how embodied behavior is integrated into domains such as action formation. The goal - or assumption - may go so far as to show that the assumed boundary between language and the body does not exist. Keevallik (2018) for example investigates how depictions or demonstrations in dancing lessons can have grammatical functions in sentences (Keevallik, 2018). The grammatical structure of a clause cannot be adequately understood if we focus on what is traditionally called language proper.
It may thus be considered somewhat surprising that the discipline which places social interaction at its core, has so long paid relatively little attention to the "non-verbal". There were of course pioneers, such as Christian Heath and the late Charles Goodwin, who were analyzing body behavior long before CA was a respectable field. Goodwin demonstrated already in 1979 that linguistic structures are partly constructed in relation to how potential addressees react; people can show through for example facial expressions that they are going to disagree or disaffiliate, and speakers can shift their language in production to accommodate those projected problems (Goodwin, 1979). Since then Goodwin, and his partner Marjorie Goodwin, have done extensive studies of topics like gaze, gestures, and even vision.
Despite their foundational work, it has taken decades for the study of embodied action to become so institutionalized as to be indispensable, the way it is in now. This revolution has to a significant extent been the achievement of Lorenza Mondada who took a systematic approach to embodied behavior, even developing a transcription system to record the details that Conversation Analysts are so interested in (Mondada, 2014). Of course, this revolution would not have been possible without technological advancements; portable video camera's allow for easy and inconspicuous recording, and software nowadays makes it possible to do frame-by-frame analysis, which means that analysts can pinpoint the exact moment some bodily practice was started and can track and describe its movements in detail.
Studies in CA can now broadly focus on embodied behavior in two ways. The main analytical focus can be on some bodily behavior itself, on practices of embodied behavior. There is a tradition of studying pointing gestures that goes back to work by Adam Kendon (2004), and which was further explored by Mondada who examined how pointing is done in for example guided tours or shops. Similarly gaze has received significant attention, particularly as it relates to turn-taking, who talks when. Lerner (2003) started exploring how people can use gaze to select a next speaker, although it is clear from work by Rossano (2013) that the exact role of gaze, particularly in multiparty interaction, is still very poorly understood. But not all subjects need to be so broad: even something as small as an eye roll can be of interest (Clift, 2018).
Alternatively the focus can be how embodied practices are combined with verbal practices to design actions. As with CA in general, this line inquiry does not take a specific practice as its starting point, but investigates instead how embodied behavior is integrated into domains such as action formation. The goal - or assumption - may go so far as to show that the assumed boundary between language and the body does not exist. Keevallik (2018) for example investigates how depictions or demonstrations in dancing lessons can have grammatical functions in sentences (Keevallik, 2018). The grammatical structure of a clause cannot be adequately understood if we focus on what is traditionally called language proper.
Quantification and experimental CA
CA is fundamentally an inductive method. Indeed, scholars regularly start by studying one piece of interaction, asking at each point why the participants perform that particularly, in that particular way, and that particular moment. This means that quantitative methods are close to anathema to CA researchers. The more traditional approaches in the social sciences, such as experiments, require that researchers generalize across broad patterns and ignore the fine details. Moreover, the strict methodological standards that are required for generalization mean that experiments don't deal with natural data. That is of course not to say that experiments have no scientific value, but they clash with CA in a very fundamental way. Or at least, they used to.
In recent years even scholars in CA have started to embrace more quantitative and experimental methodology. This may be in a relatively simple form of counting practices. For example, Stivers and Sidnell (2016) studied how children design proposals when playing. They found that there are two major forms 'Let's X' and 'How about X,' and that they are used to propose new activities or modify an activity in progress respectively. To support this claim, they counted the frequency of each combination and used a statistical test to show that this association was statistically significant. Such an analysis can make findings more palatable for quantitative scholars.
Alternatively, CA can make use of proper experimental methods. The collaboration of CA with medical sciences requires that conclusion cannot be based merely on inductive analyses: they ideally need to be supported by randomized control trials. For example, a study done by Heritage et al. (2007) had two groups of general practitioners ask their patients in different ways whether they had additional concerns. By only varying this one condition conclusions could be drawn about whether using some instead of any might work better.
Finally, the questions researchers have sometimes can only be properly investigated by using experimental methods. Gaze, as pointed out earlier, has become of significant interest to CA scholars in recent years. But measuring gaze is tricky; we can only estimate what people are looking at. To get a detailed picture, we need to use eye-tracking technology, and that inherently means people have to wear goggles. Obviously, when people are sitting with these massive goggles in a stimulus-free environment as they did in a series of studies by Holler and Kendrick (2015), that does not create the most natural interaction. But gaze is mostly not done consciously, and so this may not be a problem. Similarly, Stevanovic and colleagues have been putting participants in motion capture suits to get a more accurate reading of how they move their body when interacting. These are cases where new technologies allow us to address a whole new array of questions.
The fact that quantitative and experimental methods are only now being applied in CA has some advantages and disadvantages. It can be seen as evidence that CA is becoming a more respected field and that there is a need and desire to communicate with fields that are more resistant to qualitative methods as the only source of evidence. It also means that, unlike for example psychology, CA does not have to go through decades of scientifically dubious methods with the aim of getting publishable results: we already now how research should be done. However, the start is not promising. CA scholars are not always well-versed in these methods and so we see that not only are the same old faulty statistical methods are applied, but they can even be wrongly applied. Some scholars use inferential statistical methods, when the conditions for those methods, such as randomized sampling, have not been met. While this is a problem that is rampant in the social sciences and humanities, it'd be a shame if CA loses its gold standard of using deviant cases as evidence for social norms and instead comes to rely on statistics that are way too often wrong as to be useless.
In recent years even scholars in CA have started to embrace more quantitative and experimental methodology. This may be in a relatively simple form of counting practices. For example, Stivers and Sidnell (2016) studied how children design proposals when playing. They found that there are two major forms 'Let's X' and 'How about X,' and that they are used to propose new activities or modify an activity in progress respectively. To support this claim, they counted the frequency of each combination and used a statistical test to show that this association was statistically significant. Such an analysis can make findings more palatable for quantitative scholars.
Alternatively, CA can make use of proper experimental methods. The collaboration of CA with medical sciences requires that conclusion cannot be based merely on inductive analyses: they ideally need to be supported by randomized control trials. For example, a study done by Heritage et al. (2007) had two groups of general practitioners ask their patients in different ways whether they had additional concerns. By only varying this one condition conclusions could be drawn about whether using some instead of any might work better.
Finally, the questions researchers have sometimes can only be properly investigated by using experimental methods. Gaze, as pointed out earlier, has become of significant interest to CA scholars in recent years. But measuring gaze is tricky; we can only estimate what people are looking at. To get a detailed picture, we need to use eye-tracking technology, and that inherently means people have to wear goggles. Obviously, when people are sitting with these massive goggles in a stimulus-free environment as they did in a series of studies by Holler and Kendrick (2015), that does not create the most natural interaction. But gaze is mostly not done consciously, and so this may not be a problem. Similarly, Stevanovic and colleagues have been putting participants in motion capture suits to get a more accurate reading of how they move their body when interacting. These are cases where new technologies allow us to address a whole new array of questions.
The fact that quantitative and experimental methods are only now being applied in CA has some advantages and disadvantages. It can be seen as evidence that CA is becoming a more respected field and that there is a need and desire to communicate with fields that are more resistant to qualitative methods as the only source of evidence. It also means that, unlike for example psychology, CA does not have to go through decades of scientifically dubious methods with the aim of getting publishable results: we already now how research should be done. However, the start is not promising. CA scholars are not always well-versed in these methods and so we see that not only are the same old faulty statistical methods are applied, but they can even be wrongly applied. Some scholars use inferential statistical methods, when the conditions for those methods, such as randomized sampling, have not been met. While this is a problem that is rampant in the social sciences and humanities, it'd be a shame if CA loses its gold standard of using deviant cases as evidence for social norms and instead comes to rely on statistics that are way too often wrong as to be useless.
References
Clift, R. (2018). Signs of Trouble: Embodiment in Dissent. Lecture presented at the 5th International Conference on Conversation Analysis, Loughborough University, UK.
Drew, P. & Kendrick, K.H. (2018). Searching for Trouble: Recruiting Assistance through Embodied Action. Social Interaction: Video-Based Studies of Human Sociality, 1(1). doi: 10.7146/si.v1i1.105496
Enfield, N. & Sidnell, J. (2017). On the concept of action in the study of interaction. Discourse Studies, 19(5), 515-535.
Goodwin, C. (1979). The Interactional Construction of a Sentence in Natural Conversation. In: G. Psathas (ed.), Everyday language: Studies in natural ethnomethodology (pp. 97-121). New York: Irvington.
Floyd, S. et al. (2014). Recruitments across languages: A systematic comparison. Paper presented at the 4th International Conference on Conversation Analysis, University of California at Los Angeles, CA.
Heritage, J., Robinson, J.D.,Elliott, M.N., Beckett, M., & Wilkes, M. (2007). Reducing patients' unmet concerns in primary care: the difference one word can make. Journal of General Internal Medicine, 22(10), 1429-1433.
Holler, J. & Kendrick, K.H. (2015). Unaddressed participants' gaze in multi-person interaction: Optimizing recipiency. Frontiers in Psychology, 6(98). doi: 10.3389/fpsyg.2015.00098
Kendon, A. (2004). Gesture: Visible action as utterance. Cambridge: Cambridge University Press.
Kendrick, K.H. & Drew, P. (2014). The putative preference for offers over requests. In: P. Drew & Couper-Kuhlen, E. (eds.), Requesting in Social Interaction (pp. 87-113). Amsterdam: John Benjamins.
Kendrick, K.H. & Drew, P. (2016). Recruitment: offers, requests, and the organization of assistance in interaction. Research on Language and Social Interaction, 49(1),1-19.
Keevallik, L. (2018). What Does Embodied Interaction Tell Us About Grammar. Research on Language and Social Interaction, 51(1), 1-21.
Kent, A. & Kendrick, K.H. (2016). Imperative Directives: Orientations to Accountability. Research on Language and Social Interaction, 49(3), 272-288.
Lerner, G. (2003). Selecting next speaker: The context-sensitive operation of a context-free organization. Language in Society, 32(02), 177-201.
Levinson, S. (2013). Action formation and ascription. In J. Sidnell & T. Stivers (eds.), The handbook of conversation analysis (pp. 101-130). Chichester, UK: Wiley-Blackwell.
Mondada, L. (2014). Conventions for multimodal transcription. Available from https://franz.unibas.ch/fileadmin/franz/user_upload/redaktion/Mondada_conv_multimodality.pdf
Rossano, F. (2013). Gaze in Conversation. In: J. Sidnell & T. Stivers (eds.), The handbook of conversation analysis (pp. 308-329). Chichester, UK: Wiley-Blackwell.
Sidnell, J. (2017). Action in interaction is conduct under a description. Language in Society, 46(03), 313-337.
Stivers, T. & Sidnell, J (2016). Proposals for activity collaboration.Research on Language and Social Interaction, 49(2),148-166.
Drew, P. & Kendrick, K.H. (2018). Searching for Trouble: Recruiting Assistance through Embodied Action. Social Interaction: Video-Based Studies of Human Sociality, 1(1). doi: 10.7146/si.v1i1.105496
Enfield, N. & Sidnell, J. (2017). On the concept of action in the study of interaction. Discourse Studies, 19(5), 515-535.
Goodwin, C. (1979). The Interactional Construction of a Sentence in Natural Conversation. In: G. Psathas (ed.), Everyday language: Studies in natural ethnomethodology (pp. 97-121). New York: Irvington.
Floyd, S. et al. (2014). Recruitments across languages: A systematic comparison. Paper presented at the 4th International Conference on Conversation Analysis, University of California at Los Angeles, CA.
Heritage, J., Robinson, J.D.,Elliott, M.N., Beckett, M., & Wilkes, M. (2007). Reducing patients' unmet concerns in primary care: the difference one word can make. Journal of General Internal Medicine, 22(10), 1429-1433.
Holler, J. & Kendrick, K.H. (2015). Unaddressed participants' gaze in multi-person interaction: Optimizing recipiency. Frontiers in Psychology, 6(98). doi: 10.3389/fpsyg.2015.00098
Kendon, A. (2004). Gesture: Visible action as utterance. Cambridge: Cambridge University Press.
Kendrick, K.H. & Drew, P. (2014). The putative preference for offers over requests. In: P. Drew & Couper-Kuhlen, E. (eds.), Requesting in Social Interaction (pp. 87-113). Amsterdam: John Benjamins.
Kendrick, K.H. & Drew, P. (2016). Recruitment: offers, requests, and the organization of assistance in interaction. Research on Language and Social Interaction, 49(1),1-19.
Keevallik, L. (2018). What Does Embodied Interaction Tell Us About Grammar. Research on Language and Social Interaction, 51(1), 1-21.
Kent, A. & Kendrick, K.H. (2016). Imperative Directives: Orientations to Accountability. Research on Language and Social Interaction, 49(3), 272-288.
Lerner, G. (2003). Selecting next speaker: The context-sensitive operation of a context-free organization. Language in Society, 32(02), 177-201.
Levinson, S. (2013). Action formation and ascription. In J. Sidnell & T. Stivers (eds.), The handbook of conversation analysis (pp. 101-130). Chichester, UK: Wiley-Blackwell.
Mondada, L. (2014). Conventions for multimodal transcription. Available from https://franz.unibas.ch/fileadmin/franz/user_upload/redaktion/Mondada_conv_multimodality.pdf
Rossano, F. (2013). Gaze in Conversation. In: J. Sidnell & T. Stivers (eds.), The handbook of conversation analysis (pp. 308-329). Chichester, UK: Wiley-Blackwell.
Sidnell, J. (2017). Action in interaction is conduct under a description. Language in Society, 46(03), 313-337.
Stivers, T. & Sidnell, J (2016). Proposals for activity collaboration.Research on Language and Social Interaction, 49(2),148-166.
