On the sequential organization and genre-orientation of discourse units in interaction: An analytic framework

Introduction

Larger, coherent stretches of talk in conversation are constitutive of large parts of verbal interaction (Robinson, 2013). For instance, when people tell stories, provide explanations or resolve a dissent, they usually accomplish such larger stretches of talk. Different notions have been deployed to capture the above-the-sentence-level nature of such talk as opposed to single turns or moves, for instance ‘big packages’ (Sacks, 1995: II, 354), ‘large projects’ (Selting, 2000), ‘projects of activity’ (Robinson, 2013), ‘extended sequences’ (Mondada, 2012; Psathas, 1992), ‘coherent conversational units’ (Jefferson 1978: 219), ‘activity types’ (Levinson, 1979) or ‘discourse units’ (Houtkoop and Mazeland, 1985; Wald, 1978).

In this article, we outline an analytic framework within Interactional Discourse Analysis (IDA; cf. Morek et al., in press) that – taking up the notion of ‘discourse units’ (Wald, 1978) – synthesizes key work on such global practices in discourse analysis (DA), conversation analysis (CA) and genre analysis (GA). Taking a sequential approach to the participants’ orientations and their joint activities in terms of different practices, we analyze different types of discourse units. We tie together CA’s central insights into the mechanisms of sequential organization, the notion of discourse structures above the sentence level as discovered by DA, and the functional view on genres as pre-patterned solutions to communicative problems as put forward by GA.

Our ideas are thus indebted to a wide array of work in the aforementioned traditions, while at the same time they represent a particular research program. Our work has been largely concerned with children’s acquisition of the discursive skills to participate in discourse units, such as stories, explanations or arguments (Hausendorf and Quasthoff, 1996; Heller 2012, 2014; Morek 2012, 2015; Morek et al., in press; Quasthoff, 1997a). Our empirical, mostly comparative investigation into large corpora (see ‘Data and method’ section) of children’s interactions in adult–child settings, peer talk and classroom discourse called – and allowed – for the inductive development of an analytic framework fit to describe the different shares in the joint accomplishment of different global units.

In the following section we sketch the key ideas of previous work in DA, CA and GA that are pertinent to our notion of ‘discourse units’ and their analysis. Next we introduce the kind of data we drew on to develop our methodology and outline the analytic format. We then demonstrate how our GLOBE framework can be employed in the analysis of narratives, explanations and argumentations in conversation, focusing on one level of our framework, the organizational tasks or ‘jobs’, respectively. Following this we discuss conclusions as to the structural commonalities among discourse units of different genres as revealed by the analyses. Taking the task of ‘closing a discourse unit’ as an example, we finally demonstrate, how ‘devices’ and ‘forms’ as resources are used to accomplish a ‘job’ and how they differ in genre-specific ways.

Different insights into ‘coherent stretches of talk’

Among the different research traditions that have considered ‘coherent stretches of talk’, DA (e.g. Schiffrin et al., 2015), CA (e.g. Drew and Heritage, 2013; Sidnell and Stivers, 2013) and GA (e.g. Bauman, 2001; Bawarshi and Reiff, 2010; Luckmann, 1985) represent especially influential ones. While DA has contributed tremendously to our knowledge on how certain discourse units are structured (e.g. Labov and Waletzy, 1967; Sinclair and Coulthard, 1975) and marked linguistically (e.g. Schiffrin, 1987), CA has shed particular light on the local sequential accomplishment of ‘big packages’, such as stories (Jefferson, 1978; Mandelbaum, 2013; Sacks, 1974b), jokes (Sacks, 1974a) or trouble-telling (Jefferson, 1984). From a more anthropological and sociological perspective, approaches to genre as discursive practice have underlined the pre-patterned nature of such units and their goal-directedness with regard to recurrent communicative problems within a society (Bauman, 2001; Bergmann and Luckmann, 1995; Günthner and Knoblauch, 1995; Hanks, 1987; Luckmann, 1985, 1986).

In the following brief literature survey, we turn to the questions of how DA, CA and GA have systematically tackled the description of larger stretches of talk and how we take up some of these approaches’ basic theoretical and analytical tenets.

DA in a certain sense derives its raison d’etre from studying global units in different kinds of verbal encounters. Often following a sociolinguistic or ethnographic interest in understanding naturally occurring verbal exchanges in different fields, DA exceeded sound, word and sentence as the traditional linguistic units of analysis as well as the preoccupation with the single speech acts in early pragmatics. In terms of global linguistic structures, DA established an early notion of discourse units and their substructures (Sinclair and Coulthard, 1975; Van Dijk, 1980) and yielded insights into the internal structuring and the marking of discourse functions (Schiffrin, 1987).

Among the latter, narratives received by far the most analytical attention. Labov and Waletzky (1967) in particular have been pioneers in describing how these units are organized along the structural elements abstract, orientation, complicating action, resolution, evaluation and coda. These elements, however, do not capture how narratives are interactionally brought about. Subsequently, the discourse analytic description of narratives has been refined not only to grasp different realizations of narratives (Ochs and Capps, 2001), but also to acknowledge the interactional basis of storytelling (e.g. Bamberg, 1997; De Fina and Georgakopoulou, 2012). Among the rich stock of conceptions and findings of DA, its early insight – especially in contrast to the majority of early research topics in CA (see later) – is that there indeed is a level of global structuring prominent in spoken interaction that can be observed in different discourse types and activities. The term ‘global’ here does not refer to the overall structural organization (Schegloff, 2010; see later) of whole interactions (e.g. a lesson in school or a doctor–patient consultation), but to coherent units in such interactions that are ‘larger’ than adjacency pairs and exhibit a specific sequential organization.

This insight into globality as an orderly mechanism of interaction is most prominently explicated by Wald (1978). While most discourse analytic studies examine particular types of global discourse structures, Wald focuses on global discourse structures as such by examining how participants interactively and sequentially organize the production of ‘discourse units such as narratives, jokes, explanations etc’. According to Wald, discourse units are produced and recognized as coherent building blocks in conversation because participants themselves mark off their beginning and ending from the surrounding turn-by-turn talk. Their internal sequential realization follows a particular structural pattern for each type of discourse unit. Furthermore, when producing a discourse unit, participants establish special rules for turn-taking (Sacks et al., 1974): one participant is assigned the role of a ‘principal speaker’ and holds the right to the floor until the closing of the multi-unit turn is interactively established (for diverging management of speaker roles in argumentative discourse units, see ‘Dealing with divergent validity claims – argumentations’ later). In demonstrating how participants orient themselves to the production of these global discourse units, Wald took a perspective that is very similar to that of CA (cf. Houtkoop and Mazeland, 1985). Following Wald, we henceforth use the term ‘discourse unit’ to denote the entire conversational unit in which a narrative (or joke, explanation, etc.) is interactively prepared, produced and closed. With ‘multi-unit-turn’, we refer to the larger turn that is produced by the principal speaker within a discourse unit.

While the vast majority of conversation analytic studies have been focusing on ‘very small, crystalline bits of the conversational machinery’ (Jefferson, 1988: 418), recent work in CA more and more directs attention to activities in action formation and ascription (Levinson, 2013) and to ‘larger’ phenomena such as ‘interactional projects’ (Schegloff, 2007), ‘extended sequences’ (Mondada, 2012; Psathas, 1992) or ‘overall structural organization’ (Robinson, 2013; Schegloff, 2010). Early on, it was Sacks (1995: 3–31, 222–281, 354–359) who already pointed to the relevance of ‘big packages’, namely ‘more-than-a-sentence-long-coherent-bit-of-talk’ (p. 225).

As in most studies on larger units below entire conversations, stories have gained by far the most attention also in CA (Jefferson, 1978; Mandelbaum, 2013; Sacks, 1974b). Sacks (1995: II, 3–31, 222–228) demonstrated that telling a story in conversation requires the teller to employ methods to keep possession of the floor for an extended turn-at-talk (or ‘multi-unit-turn’) and to display to the recipients what kind of activity they are to expect and react to (Sacks, 1995: II, 21). The most typical method to project stories is a story preface (e.g. I heard something wonderful today, Sacks, 1995: II, 18). Houtkoop and Mazeland (1985), as well as Psathas (1992), take into account that there are other activities (e.g. direction-giving) that require participants to interactively tackle an ‘entry’ and a ‘closing problem’. Altogether, CA’s insights into conditional relevance as a basic organizational mechanism of interactions are of central importance for understanding the interactive and sequential realization of discourse units. With respect to multi-unit turns as such, CA has also pointed to the fact that extended sequences have their ‘parts interconnected by virtue of the problem to be solved’ (Psathas, 1992: 119).

This is where the notion of genre comes in. Among the various frameworks which make central use of a concept of genre (for reviews, see e.g. Günthner and Knoblauch, 1995), only those that explicitly base ‘genre’ on interactive practices are relevant for our argument. Bakhtin and Emerson (1986) were among the first to point out that language manifests itself in the form of speech genres, which are available as routinized resources for interlocutors to coordinate and interpret their joint activities. Their rationale is reflected in more recent approaches as developed in linguistic anthropology (e.g. Bauman, 2001; Hanks, 1987), the sociology of knowledge (Bergmann and Luckmann, 1995; Luckmann, 1985, 1986) including its linguistic extensions (Günthner and Knoblauch, 1995; Kotthoff, 2007; Linell, 2009), and in the New Rhetoric (Miller, 1984). In this sense, genres can be seen as routinized communicative solutions to recurrent communicative problems of a speech community (Bergmann and Luckmann, 1995: 291). As such, they are resources for mutual sense-making in discursive practices, drawing on the fact that the type of these problems is functionally reflected in the pattern-like generic structure of each corresponding actual verbal exchange. For example, narratives present a generic solution to the recurrent communicative problem of sharing experiences and aligning moral evaluations (Bergmann, 1987). The particular generic features of narrating are derived from this basic function (see ‘Sharing personal experiences – narratives’ later). In this view, each patterned multi-unit turn is the communicative solution to a specific recurrent task in societies. Thus, globality and genre-orientation are different tasks that have to be dealt with at the same time.

The approach of communicative genres has acknowledged not only the internal and external structure of genres, but also their situated realization (Günthner and Knoblauch, 1995). In terms of these structural differentiations, the term ‘discourse unit’ (Wald, 1978) can be considered as encompassing the internal structure and the situated realization: while genre has been used to refer to conversational activities of a very different scope (e.g. sequences of reproaches and justifications as well as job interviews), ‘discourse units’ clearly specify a specific structural range marked-off by the participants in situating them. It thus underlines the sequential orderliness such units require, so that explicating the association between discourse units and genres adds insights into the interplay of sequential structures and communicative functions.

The notion of genre needs to be distinguished from another neighboring concept, that of ‘activity types’ (Levinson, 1992). With the latter notion Levinson (1992) refers to ‘goal-defined, socially constituted, bounded, events with constraints on participants, setting […] and the kind of allowable contributions’ (p. 69). Like genres, activity types are more or less ‘prepackaged’ and ‘culturally recognized activities’ (Levinson, 1992). Yet in contrast to genres, activity types do not necessarily involve talk; paradigm examples are lessons and job interviews on the one hand, and football games and tasks in a workshop on the other. Another difference between the two notions concerns functional aspects. Although both approaches assume that genres and activities are built on structural elements that serve particular functions, the functions are conceptualized differently. While from Luckmann’s perspective genres have evolved to solve recurrent societal problems, Levinson (1992) rather emphasizes the individual character of goals (p. 71). Since we are especially interested in verbal interaction, we use the term ‘genre’. In binding it to discourse units, we also address a smaller scope in comparison to the examples given by Levinson.

To sum up, the insights contributed by DA, CA and GA have informed the development of an analytical framework that allows describing the (1) interactional organization, (2) pragmatic function of participants’ genre-specific moves and (3) linguistic and embodied resources they use when producing a discourse unit.

Accordingly, the questions we would like to ask are as follows: How do participants jointly organize the accomplishment of discourse units? How do they display to each other that a multi-unit turn is to be produced and what kind of unit is underway? How do discourse units with different genres differ from each other, and what organizational features do they share?

The analytic framework – GLOBE ¹ – is used to tackle analytically the global sequential organization and genre orientation in an integrated methodological procedure.

Data and method

We draw on a variety of corpora that contain video and audio recordings of naturally occurring interactions in diverse settings. These recordings have been collected as part of different research projects, all of which investigated how differently aged children participate in the interactive accomplishment of discourse units. ² These data include 168 hours of family talk, peer talk among friends and classroom discourse (see Quasthoff and Kern, 2007; Heller, 2012; Morek, 2012). Table 1 gives an overview of the recordings used here.

OPEN IN VIEWER

Table 1.

Data.

Project	Year	Setting	Hours	Participants
DisKo	2011–2014	Family talk	31	12 German speaking families with children aged 10–12 years
DisKo	2011–2014	Peer talk	22	12 German speaking cliques of children aged 10–12 years
DisKo	2011–2014	Classroom talk	38	5 German language classes (5th/6th grade)
Heller (2012)	2009–2011	Family talk	25	11 families with German, Turkish and Vietnamese as their first language, children aged 6 years
Heller (2012)	2009–2011	Classroom talk	19	4 German language classes (1st grade)
DASS	2001–2003	Family talk	16	5 German speaking families with children aged 6–8 years
DASS	2001–2003	Classroom talk	17	2 German classes (1st/2nd grade)

Transcription followed the conventions GAT 2 proposed by Selting et al. (2011). Sequences were selected for transcription according to our interest in narrative, explanatory and argumentative talk. Starting from a collection of similar sequences in which participants deal with comparable communicative problems, we inductively uncovered from the data how different participants sequentially organize the solution of the respective generic problem. Comparing in detail a huge number of cases finally allowed for the extraction of constitutive series of organizational tasks that participants deal with in narrative, explanatory or argumentative discourse units. Drawing on a first version of GLOBE (Hausendorf and Quasthoff, 1996), we call these organizational tasks ‘jobs’ (Psathas, 1992; Quasthoff, 1997a; Sacks, 1995: II). The analytic level of ‘devices’ describes the kind of conversational activities by which participants fulfill these jobs, while the level of ‘forms’ refers to the linguistic and multimodal surface realizations of the semantic-pragmatic devices (see Table 3).

Jobs

Jobs represent organizational tasks that have to be interactively fulfilled by the participants in the joint achievement of a discourse unit (cf. Psathas, 1991, 1992). In accordance with Wald (1978; see the previous section), discourse units are defined in terms of their unity, brought about – among other features – by their being delimited from the surrounding turn-by-turn talk (see also Quasthoff, 1997a, 1997b). Since the organizing jobs explicate the steps that are derived from the particular genre’s communicative function, the jobs vary according to the respective genre. Due to their genre-orientation, the jobs are assumed to be applicable for all instances of the particular genre and are thus context-free: if a discourse unit was successfully performed, these tasks must have been fulfilled, no matter how or by whom.

Table 2 sums up which jobs could be inductively identified on the basis of a multitude of narrative, explanatory and argumentative global sequences within the aforementioned corpora. The horizontal level indicates the sequential order of jobs.

OPEN IN VIEWER

Table 2.

Organizational jobs in different genres.

Narration	Establishing topical relevance	Topicalizing an event	Elaborating a course of events	Closing	Transition
Explanation	Establishing topical relevance	Constituting an explanandum	Explicating procedural, conceptual and/or causal relations	Closing	Transition
Argumentation	Constituting dissent	Establishing an obligation to provide reasons	Providing and challenging reasons	Closing	Transition

In each of the genres, the series of jobs represents an orienting framework of globality and genre-orientation that participants in our data observably mutually address when treating the respective communicative problems. They always handle these genre-specific jobs; however, they do this in a locally occasioned and context-sensitive way. For example, we find instances where participants seemingly ‘skip’ a job and/or fulfill two distinct jobs conjointly at one go, that is, by means of one utterance (see our examples in ‘Sharing personal experiences – narratives’ and ‘Explicating knowledge – explanations’ later). Such seemingly deviant cases do not undermine our finding that narrative, explanatory and argumentative discourse units are constituted by the aforementioned series of five jobs. Rather, they demonstrate the mechanism of context-specific realization of context-free resources (Sacks et al. 1974): the ways the jobs of a particular genre are interactively accomplished by specific participants in a particular conversational context, that is, the devices and forms, serve as a contextualizing resource for the ongoing interaction. ³

Forms

The forms refer to the linguistic and embodied resources for realizing the functional role of the particular device. They constitute the unit’s audible and visible production and can be seen analytically as the level on which participants mark structures and functions and thus help the co-participant to identify the sort of discourse unit being underway. Achieving recognizability (Levinson, 2013) is done by forms. As devices and forms are different for different genres as well as for the different interactional roles, specific forms for establishing topical relevance or dissent are also resources for projecting the particular type of discourse unit: ‘This morning I had a row’ is heard as initiating a story (see excerpt 1) due to the reference to a tellable event; ‘Why are you making the dumpling wet?’ (see excerpt 2) states the lack of relevant knowledge and thus asks for an explanation; and ‘But they also need fully qualified grooms’ (see excerpt 3) contradicts an assumed claim and thus opens argumentation (see ‘Interactional jobs across genres and across contexts’ later for a necessarily short sample analysis of closing-devices and forms).

It is on the levels of devices and forms, then, where participants contextualize the generic pattern, relying on the recognizability of the genre but shaping it linguistically according to the immediate communicative and social demands.

Table 3 sums up the three descriptive levels. Due to space reasons, in the analyses in the following section we focus analytically on explicating the jobs. The devices and forms, however, being the material basis of how the jobs are accomplished, are actually addressed to, albeit not systematically in a technical sense (but see ‘Interactional jobs across genres and across contexts’ later). We present one excerpt each for narrative (excerpt 1), explanatory (excerpt 2) and argumentative (excerpt 3) discourse activities and show how the participants orient to and sequentially deal with the jobs constituting genre-specific globality. For reasons of comparison, all examples have been taken from family dinner talk between children and their parents.

OPEN IN VIEWER

Table 3.

Descriptive levels of GLOBE.

Descriptive levels	Analytical focus
Jobs	Dyad/interactional team: global communicative tasks to be accomplished jointly be the participants
Devices	Participants in the discourse unit (e.g. principal speaker – recipient): moves designed to the fulfillment of the organizational jobs and the requirements of a particular genre
Forms	Linguistic, prosodic and nonverbal resources for realizing the devices

Discourse units of different genres: An analysis with GLOBE

Interactional jobs across genres and across contexts

Our analyses in the previous sections have treated discourse units as they come in the real communicative world, namely in the given manifestation of a particular genre. ⁸ However, in our theoretical considerations in ‘Different insights into “coherent stretches of talk”’ earlier, we have argued, in line with Wald (1978), that globality, organizing a multi-unit structure as such, poses its own sequential problems and thus calls for its own conversational resources. Consequently, an analytically transparent access to the reconstruction of members’ practices also has to provide a descriptive format, allowing for this more general global sequential organization beyond the particular requirements of each genre, but still in light of the genre-orientation. Taking a comparative look at the three genres (cf. Table 2) makes it obvious that these jobs solve similar organizational problems across genres, thus shedding light on their quality as resources to deal with globality in general.

The first job participants have to tackle in order to embed a discourse unit of any kind into the ongoing talk is to direct each other’s attention to a particular topic or problem. This is done by establishing topical relevance or establishing a dissent.

As was pointed out in the analyses, all genres then include a job which establishes a global sequential implication triggering the multi-unit turn. Due to the concurrence of globality and genre-orientation in actual discourse units, this projecting task is accomplished by genre-specific variations: ‘topicalizing’ a tellable event, ‘establishing an explanandum’ or ‘establishing an obligation to give reasons’. This comes as no surprise since it is always a particular communicative problem and as such a specific activity that has to be projected.

Once a discursive activity has been organizationally prepared, performing the actual activity can begin. Depending on the kind of sequential implication, the accomplishment of the core job also differs. It depends on the type of communicative problem what interactive roles are established and what procedures are employed (see Table 3). The analyses demonstrated that a principal speaker was only regularly installed in narrative and explanatory activities. Thus, establishing a principal speaker is not a resource that organizes globality in general, but rather depends on the specific roles of participants determined by the particular genre drawn upon. A narrator and an explainer appear as principal speakers due to their epistemic tasks in fulfilling the genre’s purpose: the narrator normally is the only participant who has access to the event (‘entitlement to experience’, cf. Sacks, 1995: 242–248) or the story to be told as well as to its particular significance. The explainer claims to possess the relevant knowledge to be transferred in the process of explaining. It is this particular epistemic status that gives them the special right to the floor. In contrast, in argumentation the communicative problem solved by argumentative patterns does not allocate such a specific role to one of the participants in that it – in clear distinction from explaining – implies more or less epistemically equal roles of proponent and opponent.

Closing refers to the task of interactively displaying that the genre’s communicative purpose has been accomplished. Organizing the transition into the succeeding turn-by-turn-talk (or the next discourse unit) mirrors the global sequential tasks of the first job ‘establishing topical relevance’ in that it organizes the return to the locally organized turn-by-turn talk.

A schematic representation of the organizational jobs in establishing globality, overarching specific genres, is given in Figure 1.

OPEN IN VIEWER

Figure 1.

Organizational jobs establishing globality.

In our analytical approach, the jobs account for the integration of the sequential and genre-oriented aspects of discourse units. The same organizational jobs in their genre-specific shape have proven to be applicable for all instances of the genres discussed here (Heller, 2012; Morek, 2012; Ohlhus, 2014). They pertain to different types of stories (e.g. stories of personal experience, fantasy stories, big stories and small stories; Bamberg and Georgakopoulou, 2008), explanations (e.g. why, what, how) and argumentations (e.g. persuasive vs explorative; Ehlich, 2014). Thus, they are ‘context-free’ resources (Sacks et al., 1974: 699). At the same time, these jobs are ‘capable of extraordinary context-sensitivity’ (Sacks et al., 1974): participants realize the jobs in context-specific ways and thus constitute particular contexts.

These genre-specific and context-specific realizations of the global interactional jobs become visible on the levels of devices and forms (see the ‘Data and method’ section). While our analyses in ‘Discourse units of different genres: An analysis with GLOBE’ foregrounded the level of jobs, they already included descriptions of the ways the respective participants dealt with the jobs, namely the devices and the linguistic forms each of the interlocutors made use of. In order to fully demonstrate the analytic power of GLOBE, we will now turn to showing systematically how genre-specific devices and forms make up the difference among narrative, explanatory and argumentative discourse units. Due to space restrictions we cannot do so for each of the five jobs, but will choose ‘closing’ a discourse unit to illustrate our point.

For all of the three genres discussed in this article, closing refers to the task of bringing the globally structured sequence to a jointly ratified ending by interactively displaying that the genre’s communicative purpose has been accomplished. The shares that the involved participants have in this job as well as the devices and forms they use necessarily differ:

In this way, it should become clear that the job-related devices and forms borrow essentially from the pre-patterned solutions for certain problem types. They provide the orientation framework (Günthner and Knoblauch, 1995) for the participants in view of all possible contextualized realizations of a genre and at the same time allow for the particular contextualization at hand.

Conclusion

We hope to have shown that our analytic approach explicating globality, genre-orientation and sequential orderliness of single moves presents a suitable tool for describing how participants co-construct discourse units and display their particular kind in interaction. The global character of discursive activities is revealed in two respects: (1) by virtue of their being a complex interactive format fulfilling particular communicative functions with the aid of patterns available to the speech community (genres), and (2) by virtue of their fulfilling special sequential tasks in organizing multi-turn units as opposed to sequences of single turns.

The fact that a prominent way of accomplishing more complex communicative tasks implies the format of discourse units makes a strong case for integrating central insights into sequence- and genre analyses. This is why Interactional Discourse Analysis applies the three-level descriptive format GLOBE that has been demonstrated in its descriptive potential.

There is a surplus of this integrated heuristics in its context-free and context-sensitive dimensions with respect to applied research questions: due to its defining functional equivalences on the level of jobs, it allows for comparing how participants of, for example, different competence levels accomplish narrative, explanatory and argumentative activities. Likewise, in its full-fledged form, it has been shown to provide a suitable format to grasp different contextual realizations of global discursive activities (e.g. classroom discourse, peer talk).