Longitudinal Conversation Analysis - Introduction to the Special Issue

ABSTRACT

How do people’s interactional practices change over time? Can conversation analysis identify those changes, and if so, how? In this introductory article, we scrutinize the novel insights that can be gained from examining interactional practices over time and discuss the related methodological challenges for longitudinal CA. We first retrace CA’s interest in the temporality of social interaction and then review three lines of current CA work on change over time: developmental studies, studies of sociohistorical change, and studies of joint interactional histories. Existing work shows how the execution of locally coordinated actions and their meanings change over time; how prior actions inform future actions; and how resources, practices, and structures of joint action emerge over people’s repeated interactional encounters. We conclude by arguing that the empirical analysis of the microlevel organization of social interaction, which is the hallmark of CA, can elucidate the fine-grained situated interactional infrastructure that provides for the larger-scale social dynamics that have been of interest to other lines of research.

A society’s members encounter and know the moral order as perceivedly normal courses of action—familiar scenes of everyday affairs, the world of daily life known in common with others and with others taken for granted. (Garfinkel, 1964, p. 225)

Ethnomethodology (EM) and Conversation Analysis (CA) have set out to uncover the commonsense knowledge and procedures (or member’s methods) at work in everyday activities. But where do such procedures and knowledge—and their sharedness—come from? How do people adapt their methods for action and practical reasonings when entering new situations, interacting in new media, or when coming to know each other better over time? How do people come to build “the routine grounds of everyday actions” and to “encounter and know the moral order as perceivedly normal courses of action” (Garfinkel, 1964, p. 225)? How do practices change, and how might CA be used to track those changes?

This special issue scrutinizes how people’s practices for action and related normative expectations change over repeated interactions across time. Such change over time in interactional practices has only recently started to be explored systematically from an EMCA perspective (see the articles in Pekarek Doehler et al., 2018). Our aim is to showcase ongoing studies and synthesize prior research in longitudinal CA, to exemplify the novel insights to be gained from examining interactional practices over time, and to discuss methodological challenges related to this endeavor. The longitudinal approach to social interaction is motivated by an interest in change both on an individual and on a societal level. Longitudinal studies on the micro-organization of social interaction aim at understanding how the execution of actions and their meanings change over time; how prior actions inform actions further in the future than CA’s usual more immediate, local, timescale; and how resources, practices, and structures of joint action emerge and change over people’s repeated interactional encounters.

Longitudinal perspectives and temporality in prior EM and CA work

CA’s methodological focus tends to be on collections of instances of similar phenomena and hence a central concern with the stability of a given interactional mechanism across instances, interlocutors, and occasions. By contrast, the very purpose of longitudinal CA is to identify change, which entails a concern with both similarity and difference. It requires what Koschmann (2013, p. 1039) refers to as “same-but-different” analysis: To count as evidence for change over time, the phenomenon under scrutiny has to be different at time t2 from t1, yet similar enough to be interpretable as an occurrence of the same phenomenon—a token of the same type.

In conversation analytic work per se, the first strand of research addressing change over time in members’ practices has been concerned with what can broadly be labeled as “learning” or “development.” The first systematic study in CA was Wootton (1997) on the acquisition and the expanding situated adaptivity of a young child’s use of various request formats between the age of 10 and 37 months. Over the course of the 27 months, the child is shown to increasingly adapt her request formats to the circumstantial details of the interaction. For instance, the child increasingly chooses among different request formats as a function of whether there is a sequential basis for expecting parental compliance. Wootton’s study set a precedent for tracking an individual’s development over time by documenting change in social interactional practices for accomplishing a precise action (e.g., requests) within interactions with precise others (e.g., parents).

Wootton’s interest in change was not widely shared. Some other voices arguing for its study in those early days were Bilmes, Zimmerman, and Schegloff. Bilmes (1985) argued already that we sometimes have to turn to the larger conversational history beyond the sequence to understand why a certain action is produced and what action is actually performed. Zimmerman (1999) called more than a decade later for vertical (across time) comparative research in addition to horizontal investigations across settings, cultures, or languages. And Schegloff (2009, p. 400), while referring to the temporality of individuals’ lives, as well as social, cultural, and historical time, suggested that “The sort of data which has proven indispensable for CA work limits the possibility of comparative analysis at present, but not for long.” Today, the increasing number of longitudinal CA studies sets out to deepen our understanding of how participants’ practices, expectancies, understandings, and resources change over time, and how participants come to act the way they do at given moments in time.

Longitudinal CA: Three lines of research

Research that has explicitly used CA to chart change over time falls into three (overlapping) kinds:

1. identifying changes as a matter of speakers’ learning or development,

2. as a vector of sociohistorical/cultural change, or

3. as a community’s interactional history.

What these lines of research share is a basic rationale behind the ordering of single instances—their chronological organization—as well as a range of methodological challenges pertaining to issues of comparability over time.

Developmental studies

Developmental studies track aspects of people’s pathways from novices or “not yet competent members” (Schegloff, 1989) to “competent members” (Goodwin, 2018, chapter 1.2.3) or experts. Prime examples of this type of longitudinal CA are Wootton’s (1997) pioneering work and subsequent studies on children’s or second-language speakers’ development of interactional competences. Developmental CA studies scrutinize time spans ranging from several months up to several years. They document how participants’ ways of dealing with a certain interactional problem change over time. Accordingly, their analytic focus is either on practices for accomplishing a specific action (e.g., ways of requesting, Wootton, 1997) or on the use of a linguistic resource for interaction. Longitudinal developmental studies typically follow individuals (such as a given child) in their interactional encounters with precise others (such as the child’s parents) over time; often, they are case studies. There are a few cross-sectional studies in the field in which change in interactional practices and resources is investigated by comparing members of different age groups, speakers of different levels of proficiency, or novices versus more “competent” members (see, e.g., Stivers et al., 2018, who compare children’s and adults’ responses to questions).

Following Wootton’s (1997) seminal work, a range of longitudinal CA studies has been concerned with children’s development (for earlier studies in developmental pragmatics, see, e.g., Ochs & Schieffelin, 1979). Forrester (2008) investigated the emergence of self-repair; Filippi (e.g., 2009, 2015) examined the development of various aspects of interactional competence, such as gazing and the display of recipiency, as well as the development of recipient design. Hausendorf and Quasthoff (1996) analyzed the development of narrative competencies in the context of storytelling within the family. Pfeiffer and Anna (2021/this issue) study children’s acquisition of interactional competence concerning the use of one linguistic resource, the interjection oh in German, showing how this verbal resource is increasingly used, in combination with gaze, to recruit adults’ assistance in problem solving sequences.

A more recent line of longitudinal research has been conducted in CA-SLA—CA research on second-language acquisition—regarding the development of interactional competence by speakers of a second language (for an overview see Skogmyr Marian & Balaman, 2018). This work has investigated change over time in speakers’ practices for turn taking (Cekaite, 2007), disagreeing (Pekarek Doehler & Pochon-Berger, 2011), opening storytellings (Hellermann, 2008; Pekarek Doehler & Berger, 2018), repairing conversational trouble (Hellermann, 2011), and more generally managing participation (Nguyen, 2011). Other studies have documented how linguistic resources become used for new interactional purposes (e.g., Kim, 2009; Pekarek Doehler, 2018) or are variously coupled with embodied conduct over time (Eskildsen & Wagner, 2013, 2015). In this Special Issue, Pekarek Doehler and Balaman (2021/this issue), tracking one participant’s Skype interactions over four years, document the routinization of a precise grammatical construction as a social action format for dealing with social coordination in video-mediated interactions. Skogmyr Marian (2021/this issue), in turn, shows how participants’ practices for launching a complaint develop over time, ensuing in the emergence of increased prefatory work in complaint initiation. In her study, longitudinal tracking of individuals is combined with a cross-sectional study of different proficiency levels.

Finally, there are a couple of studies that tracked novices in professional interactions. Brouwer and Wagner (2004) analyze second-language speakers doing business calls. Nguyen (2011) documents how a pharmacy student develops turn-design practices in advice-giving sequences that are increasingly recipient designed. Nguyen (2019) studies the trajectory of a hotel employee’s acquisition of competence in English as a lingua franca via interaction with customers. Martin and Sahlström (2010), in turn, address learning through the lens of repair practices in physiotherapy treatment. A case apart is the investigation of change in conversational practices of participants with aphasia by Wilkinson et al. (2007).

Developmental studies have their specific methodological challenges. Developmental trajectories of individuals mostly cover large time spans. Therefore, recorded data are necessarily discontinuous and do not document the whole developmental history. Temporal gaps between recording dates and possible related developmental gaps between novices and more expert participants may be considerable. Important steps of trajectories and patterns of emergence may be missing, and relevant factors affecting change and relevant turning points over the course of the change may not be available in the data. Therefore, while these studies provide conclusive evidence for developmental change over time, an explanation of developmental patterns often cannot be given because nonrecorded events impinging on the target practices or resources cannot be assessed.

As developmental studies typically seek to keep the target participant’s partners stable over the observed time spans by focusing, for instance, on his or her interaction within a family or a classroom context, a further tricky issue pertains to the intertwinedness of language and communicative development and larger processes of socializations. Rather than seeking to tease apart what in the observed change in participants’ practices pertains, for example, to language learning, and what pertains to change in social relations between participants over time, some researchers have insisted on the reflexive relationship between change in speakers’ interactional competence and their evolving social relations with others (Kasper & Wagner, 2014). Change in practices and resources may constitute and simultaneously reflect the participants’ evolving epistemic entitlements (e.g., to ask for help or to correct others) and their progressively establishing closer interpersonal ties (Brouwer & Wagner, 2004; Hellermann, 2008; Nguyen, 2011).

The cumulative evidence from developmental CA studies shows that, when interacting in novel situations or languages, participants build on interactional abilities they have developed since infancy to deal with generic features of social interaction. Yet they continuously recalibrate these in the course of becoming more competent members in the second language in professional or other communities of practice. For instance, beginner second-language speakers have been shown to employ only “basic” methods for disagreeing, such as the use of turn-initial polarity markers (instead of more complex dispreferred formats); they then refine and diversify resources in ways that show increased sensitivity for preference organization in the process of becoming more efficient second-language speakers (Pekarek Doehler & Pochon-Berger, 2011). Many of the cited studies show that increasing interactional competence does not simply involve an enriched repertoire of linguistic forms but also a diversification of practices and the capacity to use linguistic resources in more locally adaptive ways. These findings evidence that moving from “not yet competent” to “competent” members implies the ability to deploy increasingly context-sensitive and recipient-designed conduct and ultimately more active and more agentive participation in interaction. Becoming a member thus means both to learn how to perform one’s own agentive and creative participant’s work and to coproduce and understand the interactional order in accountable ways.

Importantly, longitudinal CA concerned with developmental issues has empirically corroborated a conceptualization of learning and of competence as situated and embodied: Learning is defined as a social practice and competence as being shaped through participants’ joint actions within the multisemiotic local ecologies of social interactions (Firth & Wagner, 1997; Markee & Kasper, 2004).

Studies of sociohistorical change

CA studies of sociohistorical change are interested in the sociocultural historicity of interactional patterns. Compared to studies on interactional competence, they cover much longer time spans, ranging from years to possibly centuries—yet the latter will only become possible in the future, when interactional data from an increasingly remote past will be available. Of course, studies of sociohistorical change usually do not include the same individual participants at different points in time. Still, it is equally important for studies of sociocultural change to attend to the same type of participants in the same activity types and sequences to evidence that the observed change is due to change in the sociocultural environment and not, e.g., to change in participant type, activity, or sequential environment (Clayman & Heritage, 2021/this issue).

CA work on sociohistorical change has mainly materialized through a series of studies conducted by Clayman and Heritage (e.g., Clayman & Heritage, 2002, 2021/this issue; Heritage & Clayman, 2013) on journalists’ question design in U.S. presidential press conferences between the 1950s and the beginning of this millennium. They provide a unique model of interactional studies of sociohistorical change in interactional practices. These studies document the changing tenor of journalistic questioning over several decades, thereby tracking change in the practices not of individuals but of members of a professional group, i.e., journalists, in a precise organizational domain of activity, i.e., presidential press conferences. The set of studies stands out for its combination of qualitative CA with quantitative methods (for a discussion see Clayman & Heritage, 2021/this issue). They show how, over decades, journalists’ questioning increases in adversarialness. Having picked a central action type from public political discourse, the authors have been able not only to document historical change in action formats, they have provided groundbreaking evidence of how the relationship between journalists and leading politicians, journalistic habitus, and ethos have changed over time.

The aforementioned methodological requirements sometimes prove to be hardly possible to fulfill. The whole sociocultural context may have changed in a way that it is not possible to keep important contextual features constant. For instance, Clayman and Heritage (2021/this issue) report that the type of press conference they studied was initially a closed, confidential meeting with the president and his staff, deemed to provide journalists with background information, whereas after WWII, it developed into a public, broadcasted event. Therefore, a thorough comparison of journalists’ questioning under these different conditions has to take into account the change of the whole interactional event, the properties of mediation, and the shift in participation frameworks, rather than offering an analysis simply in terms of how the formatting of questions has changed.

A different type of study of sociohistorical change in interactional practices has recently been undertaken by Couper-Kuhlen (2021) in her comparative analysis of participants’ use of okay in American English in the 1960s and the 1990s/early 2000s. This study stands out for its pioneering status as an analysis of linguistic change in terms of interactional practices rather than in terms of forms or more or less stable form-function mappings. Together with recent work on the routinization of grammatical patterns for specialized purposes (Pekarek Doehler & Balaman, this issue) and on how patterns of language use may have grown out of interaction (Couper-Kuhlen, 2011; Pekarek Doehler, in press), it spearheads advances in our understanding of grammar-in-interaction over time. Another kind of sociohistorical change that remains to be studied in more detail concerns interactional practices that develop due to the (evolution of) mediation of interactional exchanges. The shift from landline phones to cell phones or from face-to-face meetings to video-mediated interaction (see Luff et al., 2003) are cases in point.

Interactional histories

Interactional histories (Deppermann, 2018) concern change in communities of practice (Lave & Wenger, 1991). They focus on histories of interactions that include—at least in part—the same participants, studying participants’ mutual adaptations over time. Examples include families, couples, peer groups, professional teams, and classrooms. Communities can be lifelong and enduring or rather transient (Hazel, 2017; Mortensen, 2017). In contrast to social categories, such as parents, policemen, second-language speakers, or journalists, communities of practice are social groups that develop a we-relationship based on spending time together in interaction and often also in co-presence (Schütz, 1932/1960). Their members share experiences and produce joint action by which shared knowledge and reciprocal expectations of action and understanding develop (Clark, 1996). Studies of interactional histories set out to evidence the cumulative nature of joint experiences. A development of and within a community of practice is examined in Beach’s (2009) “natural history of a family cancer,” which retraces the ways in which a family navigates through diagnosis, treatment, and prognosis for cancer based on 61 calls between the family members recorded over a period of 13 months. Brouwer and Wagner (2004) provide an earlier example of an analysis of a short interactional history, documenting how a newcomer to a professional interaction progressively adapts the way he engages in telephone openings over a series of business calls with the same interlocutor. In this Special Issue, Deppermann and Schmidt (2021/this issue) track the interactional history of the appropriation of a novel expression within a transient community of practice. They analyze how an artistic core concept from traditional Japanese culture (wabi sabi) is appropriated over a series of theater rehearsals. Initially being introduced by the director, the actors first take wabi sabi up in a quotative manner; they explore its meaning and finally begin to use it as part of their own repertoire of esthetic assessment.

In sum, if an action leads to a successful accomplishment in interaction, it serves as a precedent to be reused at a next occasion of the same type; actions that cause non- and misunderstandings or unwanted effects instead may be avoided and replaced by other options. Co-membership in a community of practice thus enables systematic recipient design and increasing adaptivity to types of contexts because action formation can be built on prior experiences with the addressee.

The longitudinal perspective on interactional histories is important for understanding the formation and change of human culture. Methodologically, this is done by tracking the emergence and change of shared practices, routines, and understandings in and via social interaction within a community. Unlike in developmental CA studies or in existing CA research on sociocultural change, the analytic focus is often not narrowed down to a given practice or resource in precise sequential environments, but may include larger stretches of interaction, such as specific sequences in therapeutic encounters (Voutilainen et al., 2018) or scenes in a series of theater rehearsals (Hazel, 2018; Norrthon, 2019; Schmidt & Deppermann, 2020). Ideally, a study on interactional histories starts with the very first occurrence of an instance of the use of a resource, a practice, or an interactional problem to be solved and tracks consecutive occurrences without a gap. A research design of this kind allows for representing the whole trajectory of emergence and change, thus enabling not only the identification of interactional patterns but also the discovery of how one instance of interaction may lead to a next. It allows identifying how interactional practices at point t1 inform interactional practices at t2 and thereby to discover and to test if and in which ways members use interactional experiences to build their future actions and understandings.

This has been demonstrated by Deppermann (2018) for instructions in driving lessons. Based on recordings of the entire interactional history between a driving instructor and his driving student, the study documents how the instructor over time progressively simplifies the action formats of his instructions to the student, both grammatically and regarding their informational contents: Across their shared interactional history, the instructor adapts to the student’s increasing expertise, understandings, and shared expectations. Similarly to Broth (2017), the study shows that over a learning history, instructional sequences become more condensed, and instructors use increasingly economic linguistic means to index the basis of knowledge that students have to mobilize to execute the required action. Intelligibility of most economical communicative means thus presupposes a shared interactional history of the participants.

This kind of “from the beginning without a gap” design requires a sampling procedure that can only be executed if the researcher is able to identify the point in time when a community of practice comes into being and when a certain practice is implemented for the first time. This is possible, e.g., with new classes at school, a newly formed team of workers or an acting ensemble, or a new therapeutic relationship. Studies of interactional histories mostly cover stretches of time between one day (Harjunpää et al., 2021), a couple of weeks (Deppermann, 2018; Hazel, 2018), and several months (Deppermann & Schmidt, this issue; Voutilainen et al., 2018). Voutilainen et al. (2011) have investigated how psychotherapy is treated as a longitudinal development by both therapists and clients through their referring back to their prior interactions (see also Bercelli et al., 2013; Voutilainen et al., 2018). Mondada (2018) shows how participants to an urban participatory project, over the course of six years, draw on shared past encounters within the project and mobilize these as resources for shaping their current actions.

However, when studying, e.g., families or peer groups, complete interactional histories will not be available. Since their starting point is not available, histories cannot be monitored and recorded continuously, and their temporal extension is too large. Although in such data, causal attribution of change will prove to be more difficult because decisive experiences may have been missed, it still allows for the identification of patterns and dynamics of change and opens a window onto processes of mutual adaptation. Hazel (2018), for instance, documents the methodical procedures through which actors subtly change the ways they jointly play a precise part of a scene across a series of rehearsals.

Research in this field shows how the appropriation of new ways of acting, new understandings, and ways of coacting happens over repeated occasions of interaction. Studies document the emergence of routine practices and how these may involve structural reduction of practices adapted to joint experiences. Recipient design fosters not only change in practices but also the routinization of practices for a given action and hence change in action formation. Furthermore, some of the cited research provides evidence for how participants themselves talk into being past encounters and hence time. By this, they offer key examples for an emic analysis of change across time.

This kind of longitudinal approach encounters some principled methodological challenges as well. Interactional histories shade into large time spans ranging over several years. Long-term processes of socialization and biographical change cannot be addressed because they do not occur within the time span of recordable interactional histories. In addition, there are processes of change in societal practices that cross-cut communities of practice because members are acting in diverse communities and contexts; members of social groups also may orient to each other without interaction (e.g., via mediated experience).

The three strands of longitudinal CA complete each other in furthering our understanding of the historicity of social interactional practices and resources as implemented in the hic et nunc accomplishment of actions in interaction. They shed light on how individuals build and continuously adapt their methods for action in interactions with others in locally relevant, experience-based ways; how members come to a sharedness of such methods; and how these methods may evolve over generations as features of larger processes of sociocultural change.

Methodological challenges

Longitudinal studies in CA face at least three methodological challenges: (a) how to warrant comparability over time, (b) how to provide evidence for the robustness of change, and (c) how to maintain an emic perspective on change (for an earlier discussion see Wagner et al., 2018).

Comparability

Comparability is a key issue for any longitudinal and cross-sectional research (just as for cross-cultural or -situational comparison, Deppermann & Mondada, 2021; Sidnell, 2009). Most existing longitudinal research outside of CA seeks to warrant comparability through strict control of variables based on experimental or semi-experimental data generation techniques. By contrast, and in line with CA’s commitment to naturally occurring data, longitudinal CA follows people in their naturally occurring environments. This in turn asks for data collection techniques that are apt to warrant a “natural control” of variables (Stivers et al., 2009). This is different from some classical studies in CA, which can be agnostic as to who produces a given practice, in what situation (but see Sidnell, 2009 for horizontal comparative studies), and at what moment in time.

When it comes to analyzing how practices or resources for accomplishing a given action (repairs, requests, instructions, questions, etc.) change over time, comparability needs to be maintained by keeping participants (or the category of participants, such as teachers or journalists), speech-exchange systems, activity, participation framework, and sequential environment of a target phenomenon all constant. Such comparability is key for understanding the observed change over time as being explainable in terms of repeated interactional experiences over time or sociohistorical change rather than as being due to differences in situational contexts, coparticipants, sociocultural conditions, and so forth.

Depending on the object studied (resources, practices, participants’ own references to past experiences), there may be some variation in the range of variables calling for a “natural control.” Some studies on interactional histories track strictly the same participants in the same activity type, but they focus on more extended activities instead of precise sequential environments (e.g., Bercelli et al., 2013; Hazel, 2018; Voutilainen et al., 2018). Others show how participants use the same resource on different interactional occasions in various sequential environments (e.g., Deppermann & Schmidt, this issue). Developmental studies, by contrast, typically proceed by collection-based analysis of recurrent practices occurring in precise sequential environments at different moments in time.

Robustness of the empirical evidence of change

Robustness of the empirical evidence of change can be reached through collection-based analysis at different moments in time, sometimes complemented by quantification. Studies usually proceed by dividing the timeline into “earlier” and “later” stages, with possible “in-between” stages, so as to document how systematic the change is between moment t, t + 1, t + n. This is a key requirement, given that learning, development, and sociocultural change are not necessarily linear processes; that different practices for accomplishing the same action may coexist at a given moment in time; and that a given practice or resource may be used, then recede, and then reappear over time. Furthermore, existing evidence suggests that change in members’ practices over time is not absolute: A resource or practice does not disappear but may be complemented by other resources or practices and change in its relative frequency (Clayman & Heritage, 2002; Wootton, 1997). Exactly this point requires a critical discussion regarding the need and the limits of quantification in longitudinal CA (Clayman & Heritage, 2021/this issue). In studies of interactional histories, which track the development of single cases (social groups in the lifeworld, communities of practice), in contrast, it is most important that sampling follows the principle of “from the beginning without gap” as far as possible (see the aforementioned).

Explaining change

Although identifying patterns of change (tendencies, directions, end points) in itself has the potential of offering key insights into human sociality, most work in longitudinal CA is additionally interested in explaining change. This means relating empirically observed change to factors such as learning or development, shared interactional experience, or sociocultural context as explanatory templates. Far from pursuing unmotivated looking, longitudinal CA studies start with a data collection design that is purposefully configured in ways to document change. There is, admittedly, a risk of circularity involved in such an enterprise, and therefore causal factors for change must be identifiable independently from the focal phenomenon under analysis. For example, if changes in turn design in instructions are to be linked to students’ learning, then students’ acquisition of skills must be evidenced by other data than the turn design; if changes in practices are to be explained by changes in sociocultural values, the relevance of the latter for the participants must be proven independently.

A further challenge—and arguably the most central one—is how to maintain an emic, i.e., participant-relevant, perspective on change. The issue here lies in the mutual accountability of conduct. Participants do display orientation to local accountability of one another’s conduct to various degrees; sometimes they explicitly show an orientation to their interactional histories (Bercelli et al., 2013; Mondada, 2018; Voutilainen et al., 2018). We also know from research on interactions between “competent” and non- or not-yet-competent members that people index the accountability and problematicity of conduct, e.g., through repair initiation, by talking in installments, speaking in an overarticulate manner, and seeking understanding displays more frequently (Svennevig, 2018). Yet as Garfinkel and Sacks (1970, p. 344) put it, conduct is competent when it “specifically not matters for competent remarks,” i.e., when it is perceived as such by members and provides no grounds for comment or repair. People usually do not display that they or others are doing something “differently” at a given moment than they did before. Although change often becomes relevant and is assessed in pedagogical contexts (institutional but also within the family or in settings involving novices), in many everyday situations less competent ways of doing things interactionally may be received by co-participants with a “let it pass” (Firth, 1996; but see Schegloff, 2009), rather than being treated as departures from normative conduct. Such heightened permissibility or acceptability of interactional conduct and change in normative expectations is exactly what often characterizes interactions with “less competent” members such as children, second-language speakers, or people with language disorders (e.g., Goodwin et al., 2002; Kurhila, 2006; Wilkinson et al., 2007). The consequence is that analytic accounts of longitudinal change cannot always be systematically grounded in observable procedural consequences (Schegloff, 1992) of local conduct, as they are not always accountably treated as different or as more or less competent. In either case, the variation in local practices that is documented through sequential analysis is post-analytically interpreted in terms of such issues as increased competence, development, learning, socialization, and sociocultural change.

The contribution of longitudinal studies of practices to the enterprise of CA and beyond

The cited strands of longitudinal CA complement each other in offering unique insights into human sociality. Longitudinal studies in CA open a window onto the processes that shape, over time and experience, how members become members, how they come to coproduce and understand the interactional order in accountable ways on every single occasion. They not only offer unprecedented insights into how the social order, institutions, and communities are shaped and reshaped through processes of social interaction over time. They also have the potential to contribute to empirically corroborate our understanding of social and psychological notions such as “socialization,” “social integration,” “learning,”or “competence” as grounded in the microdetails of social interaction.

Longitudinal studies of interactional histories, the development of interaction competence, and sociohistorical change shed novel light on the core CA concern with the nature of interactional practices and structures by offering insights into how practices and structures evolve and how participants adapt them over time. Here are a few examples:

• Recipient design: Prior experience with and knowledge about the co-participant’s knowledge, status, preferences, etc., is a prerequisite for neatly tailored recipient design of actions (Schegloff, 1972). Studies of interactional histories can show how prior experience is used for producing future actions that adapt to specific interactional partners (Broth, 2017; Deppermann, 2018). They can also help us to understand better how indexicality, implicitness, and innuendo use and index common ground in interaction (Deppermann & Schmidt, this issue). Similarly, studies of interactional competence (in first or second language) document how becoming a more competent member implies a diversification of practices that allows for increasingly recipient-designed (Wootton, 1997) and context-sensitive conduct (Pekarek Doehler & Berger, 2018), and so do studies of novices in workplace settings (Brouwer & Wagner, 2004; Nguyen, 2011).

• Understanding in interaction: Participants clearly understand more than they overtly display. Reliance on background knowledge of various sorts is indispensable for conversational action and understanding. Longitudinal studies can show how knowledge and common ground accumulate by virtue of interactional experience (Deppermann & Schmidt, this issue). This allows us to comprehend how meanings of conversational action are understood without the need to turn to speculative cognitive ascriptions based on intuition or theoretical models the situated relevance of which is not warranted. In particular, inferences that are constitutive of action-ascription and emotive-evaluative connotations of conversational action can become transparent through longitudinal analysis to a degree that cannot be achieved by recourse to sequential analysis alone.

• Sequence organization: Routine sequences in long-standing relationships like families and professional teams involve sequential shortcuts, highly specialized sequence types, participant-specific kinds of contributions, and locally specific turn-taking organizations. Longitudinal studies can elucidate participants’ knowledge and competences that are required for the participation in routines of this kind and that allow, e.g., for anticipating a partner’s needs or next steps in a joint project. Developmental studies demonstrate the increasing complexification of the sequential organization of activities such as story openings (Hellermann, 2011; Pekarek Doehler & Berger, 2018) or the initiation of complaints (Skogmyr Marian, this issue), showing how participants over time begin to use prefaces and other prefatory means to project these activities so as to assure coparticipants’ alignment. There is also initial evidence that the ways participants handle preference organization vary across levels of interactional competence (Pekarek Doehler & Pochon-Berger, 2011). A longitudinal approach may also uncover motivations for responses and participants’ interpretations that may seem unwarranted or unintelligible (“neurotic“) from an outsider’s perspective, but which become understandable by their roots in shared interactional histories (see Goodwin et al., 2002).

• Action formation: Interactional experience is a key to the local choice of resources. The study of the history of success and failure of action formats and the use of specific resources and their interactional uptake helps us to understand why certain forms are shaped in specific ways and end up being preferred over others, and why reduced or “elliptical” forms of action may develop over interactional histories and be perfectly mutually intelligible to participants (Deppermann, 2018). This contributes to current concerns with how grammar routinizes as practical solution for recurrent socio-interactional purposes (Pekarek Doehler & Balaman, this issue) and, ultimately, with the interactional motivations of linguistic change (Couper-Kuhlen, 2021; Pekarek Doehler, in press). Studies of socio-interactional practices over longer periods of time can in turn reveal how normative expectations and related accountabilities regarding types of actions (say, negative interrogative questioning) in certain types of situations (say, presidential press conferences) change over historical time (Clayman & Heritage, 2021/this issue inter alia).

Taken together, longitudinal CA studies confirm the omnirelevance of the generic organizational issues identified in classic CA work, yet they show that time (in the shape of individuals’ or communities’ iterative engagement in social interaction or in terms of larger sociohistorical processes) is crucial for understanding the becoming and the variation of the practices people deploy to deal with these issues.

The nature of change in interactional practices and the contributions to this issue

The longitudinal dynamics of change in practices are multifaceted. The studies gathered in this Special Issue attest to different directions of such change.

Stabilization/routinization

Pekarek Doehler and Balaman show how a participant initially uses different resources to deal with a recurrent interactional task; over time, a stable, uniform social action format emerges that can be shown to be the result of “interactional streamlining”—that is, participants select and routinize the solution that is most effective and least liable to engender problematic responses on the part of coparticipants.

Diversification

Skogmyr Marian documents how participants acquire a growing repertoire of different practices that can be used to implement a specific action (complaining). Diversification allows for a more context-sensitive choice of practices. Deppermann and Schmidt show how a lexical form is used in a variety of contexts over time. This leads to context-specific meanings, which add to the meaning potential (Norén & Linell, 2007) of the resource. And Pfeiffer and Anna evidence how a young child over time expands the action types accomplished by means of the expression “oh X” in concert with their bodily conduct.

Increasing adaptivity

Pfeiffer and Anna, Pekarek Doehler, and Balaman, as well as Skogmyr Marian document how interactional resources are used in ways that are increasingly sensitive to addressees and local contexts of action. This can mean that children learn to coordinate resources in more systematic ways (Pfeiffer & Anna), that a novice to video-mediated interactions develops a format that suits a variety of medium-specific interactional contingencies (Pekarek Doehler & Balaman), or that second-language speakers over time increasingly use resources that allow implementing an action in ways that are more sensitive to recipients’ prior knowledge and their ongoing responses (Skogmyr Marian).

Increasing complexity

Skogmyr Marian shows how participants deploy increasingly structurally complex practices to implement a sensitive action, i.e., complaining. More complex practices are tailored to specific contingencies of the interactional context.

Reduced complexity

Repeated-task performance and frequent use of a resource can entail increasingly lean forms of executing an action or activity. Reduction can concern both formats of individual action and interactional sequences.

Qualitative shift in practices

Clayman and Heritage show how longitudinal processes can lead to diachronic change in the kind of practices that are used to implement a certain action, i.e., questioning presidents. Change can be both categorical (emergence of new practices, while older practices disappear) or gradual (changes in frequency and preference of some practices over others). This kind of change can—but need not—involve a reinterpretation of practices.

Spread

Starting from innovators, resources, and practices (Deppermann & Schmidt) but also action sequence and activity types can spread throughout a community of practice. Spread is linked to the emergence of communal common ground (Clark, 1996), which allows innovative resources to be used more widely, with more interactional partners. Enhanced common ground allows for more efficient interaction through more reduced forms and a higher degree of implicitness and indexicality (e.g., use of pronouns instead of full lexical forms).

Thoughts on an integrative model of change: Interconnecting the strands of longitudinal CA and their relation to other studies of change

As we stated earlier, different types of longitudinal studies of social-interactional practices can shade into each other. This is due to the fact that the different time-scales of interactional conduct (sequential, cross-sequential, cross-event, historical) are related to each other. Yet these relations have not yet been explored in detail. Existing studies in longitudinal CA have discovered a rich phenomenology of processes of change. Still, it has not yet been attempted to formulate an integrative model that can accommodate and interrelate the different findings so as to provide a more comprehensive picture of change in interactional practices over time.

Here, we can only tentatively try to link empirical findings from longitudinal CA to other theoretical approaches to linguistic and sociocultural change that have identified similar processes with different methods and that have proposed overarching dynamics and models of change. What we’d like to argue is basically this: The detailed empirical analysis of the microlevel organization of social interaction, which is the hallmark of CA, can elucidate the fine-grained, situated interactional infrastructure that provides for dynamics that have been identified in large-scale corpus-based studies; and CA can also shed light on larger-scale social orders. For that purpose, and in a somewhat speculative manner, we refer to models of linguistic and sociocultural change of language emanating from usage-based linguistics (e.g., Bybee, 2010) and the sociology of knowledge (Berger & Luckmann, 1966) to sketch the possible relationship between interactional, individual, and sociocultural processes in the change of interactional practices.

An integrative picture could look like this. The sequential and simultaneous organization of action in interaction is the primary site in which resources are employed, actions are produced and tried out, interactional experiences are made, and intersubjectivity is accomplished. It is the site where interactional competences are exercised, elaborated, and rehearsed, and where members are treated in locally accountable ways as more or less competent members. This happens within specific sociocultural contexts with agents endowed with specific kinds and levels of competences. Cumulative encounters with interactional partners, across a plethora of locally accomplished interactions within a community, create interactional histories (Deppermann, 2018). Over interactional experiences and histories, practices become habitualized by repetition (Berger & Luckmann, 1966) if their use is effective and validated by consistent responses by interactional partners. In that very process, practices stabilize, repertoires to deal with varying interactional tasks diversify, and coordination routines develop (Clark & Wilkes-Gibbs, 1986). Routinization implies reduction and simplification of practices, both on the level of turns and sequences and of multimodal resources: Forms become more economical, they fuse together and get abridged (Köhler, 1986); taken for granted elements become presupposed and elided; sequences become compacted (Wakin & Zimmerman, 1999). Frequency plays a major, yet not the only role, for the increasing entrenchment of routine practices (Bybee, 2010; Divjak, 2019; Schmid, 2020, ch. 2). Frequency itself depends on social factors: The recurrence of interactional environments and, most importantly, interactional partners’ uptake, their interpretation, response, and reinforcement of emerging ways of talking and acting affect the frequency of actions and the resources used to implement these actions (Tomasello, 2003). Entrenchment of individuals’ practices includes faster production, chunking, unit formation (from smaller units that were originally distinct), and automatization of execution (Bybee, 2010; Divjak, 2019). Recurrent interactional experiences hence ground the longitudinal development of resources and procedures put to work by individuals, who, as they reach and flourish in adulthood, act in more and more varied contexts and communities.

Sociohistorical change equally starts with and materializes with changes in individuals’ practices that depart from or add to practices that are normative or conventional at the time. To become diachronically relevant, innovative individual conduct must be taken up by others so as to spread and to become part of shared and relevant options on a community level, finally leading to sociocultural change—a change that involves “conventionalization” (Schmid, 2020, ch. 3). Conventionalization of linguistic and pragmatic means is a way of institutionalization (Bergmann & Luckmann, 1995), i.e., the sanctioned sedimentation of routine solutions for recurrent interactional problems (see also Berger & Luckmann, 1966).

In a nutshell, then, the sedimentation of practices—and the related normative expectations—grows out of member’s cumulative experience with others, and this experience across co-participants, situations, and time is what in turn warrants the very sharedness of the commonsense knowledge and repertoires of practical procedures that Garfinkel (1967) was concerned with.

With this background in mind, the core contribution of longitudinal CA for understanding change in human conduct lies in the fact that it allows a sociological respecification of psychological notions such as learning and competence and shared knowledge or identity/membership (belonging to a community) as well as a grounding of our understanding of cultural change in the minute observation of locally situated, mutually coordinated social practices.