Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

At least since Aristotle, language has been seen as distinctively human in its complexity. Ethologists have increased our appreciation of how other mammals—dolphins, chimpanzees, gorillas, and so on—employ sounds to signal one another in sophisticated ways, but humans, in conducting their everyday affairs, rely on spoken and gestural forms of intercourse to an unparalleled degree (Eibl-Eibesfeldt, 1989). Despite the centrality of language use in human society, social psychology textbooks often ignore the topic (Clark, 1985), and when they do pay attention it is to regard language as a mode of communication or a vehicle whereby humans transmit information, including ideas, thoughts, and feelings, from one to another.

A variety of philosophers and social scientists regard the view of language as primarily communicative in function as the conduit metaphor (Reddy, 1979). This metaphor is rooted in the commonsensical notion that, through speech, one person conveys information by inserting it into words and sending them along a communicative channel. People receive the words at the other end and extract the encoded thoughts and feelings from them. The conduit metaphor reinforces the idea that problems of meaning in human society are essentially referential—concerned with how concepts correspond to or represent reality—and that language operates to make propositions about the world (Pitkin, 1972).

Instead of using the conduit metaphor and referential approach to meaning, scholars recently have approached language as a medium of organized social activity, in which words are “performatives” (Austin, 1962) or “deeds” (Wittgenstein, 1958, par. 546). It is partly through language that humans “do” the social world, even as the world is confronted as the unquestioned background or condition for activity. Nonetheless, the conduit metaphor and “picture book” view of language, rather than the more dynamic or activist approach, still heavily influence social psychological theory and research. This chapter begins with a review of general statements in social psychology about language, and then examines language as action and the philosophical and social scientific background to this perspective. We review the so-called mapping problem—or the question of how utterances become linked to social actions. Sociolinguistics and discourse analysis provide rule-based answers to this question, arguing that actions are linked together through a combination of linguistic (grammatical) and social rules (such as those associated with politeness). After reviewing these approaches, we turn to perspectives in which rules play a less prominent role—Goffman’s frame analysis and discursive psychology. Finally, we discuss ethnomethodology and conversation analysis, in which rules are altogether abandoned as explanatory resources and investigators connect language to action through other means, such as the sequential organization of talk. We briefly discuss methodological approaches for each of these perspectives.

Language in Social Psychology

There are two main disciplinary “branches” to the field of social psychology—the psychological and the sociological (House, 1977). Along the psychological branch, it has been traditional to employ the conduit model of language. For example, a frequent topic along this branch is that of persuasion, and the well-known Yale communication model (Hovland, Harvey, & Sherif, 1953) poses a basic question about it: “Who says what to whom by what means?” The conduit model of language, which has been modified by more recent, cognitively-oriented approaches such as the elaboration likelihood and heuristic and systematic models (Chaiken, 1987; Petty & Cacioppo, 1986), includes four factors that are important to achieving persuasion—a communicator or source, a message, an audience, and a channel through which the message is conveyed. When, for example, audience members perceive a source as credible and trustworthy, they are more likely to be persuaded by what the source says. Over the years, such diverse public figures as (in the U.S.) Eleanor Roosevelt, Ronald Reagan, and Barack Obama have been seen as examples of persuasive source figures. Other “source” features—including likability, attractiveness, and expertise—also affect how audiences evaluate messages. Besides features of a source, researchers have studied characteristics of messages (capacity to arouse emotion or fear, quantity and timing of messages, discrepancy between message and target’s own position, etc.), targets (mood, motivation, etc.) and situations for their influence on persuasiveness.

In the sociological branch of social psychology, symbolic interactionists have been most concerned with language (Emirbayer & Maynard, 2011). This is no doubt due to the influence of Mead (1934), who originated the suggestion that humans employ significant symbols that, when emitted by one party, elicit the same response in that party as in the party to whom the symbol is directed. This suggestion assumes importance in a larger context than social psychology, however. Sociologists regard communication as achieving a solution to “the problem of meaning,” which Weber (1947) long ago identified as being at the core of social action. The defining criterion of such action is that it is a product of the interactive interpretations of society’s members. When Mead proposed the existence of significant symbols and the capacity for “taking the role of the other,” it seemed to represent a clear statement of how humans could form common understandings, produce mutual and complementary stances within what he called the “social act,” and also thereby provide for larger patterns of social life.

From ideas like Mead’s, and a more general concern with the problem of meaning, it is easy to see how social psychologists moved to the conduit metaphor when discussing human language, seeing it as a repository of significant symbols in which people package their ideas and feelings. Significant symbols include not only words but gestures as well, although there are two views of gestural communication. In one view, gestures are substituted for words. Thus, a hand wave stands for “hello,” a green light suggests “go,” a beckoning arm signifies “come on,” and so on (Hertzler, 1965, pp. 29–30). In the other view, gestures occupy a different “channel of communication” than words—a nonverbal one. In either view, because of the presumption that gestures encode referential meaning, the conduit metaphor is preserved. Although it is recognized that gestures and words are arbitrary and conventional and that they take on different senses according to the context in which they appear, individuals’ ability to encode their own experiences with words and gestures inexorably leads actors to share the same mental attitudes or states and to agree upon reference (Hewitt & Shulman, 2011). Shared agreement, in turn, makes collaborative activity possible.Footnote 1

Overall, then, language has been important to social psychology because it represents a vital medium whereby actors can communicate with one another and thereby set up joint projects according to preexisting social arrangements. In this view, the manipulation of significant symbols—i.e., thought—is a precursor to action. Behavior and action are the products or outcomes of pre-existing, common understandings achieved through language. A different view of language sees it as co-constitutive of social activity. That is, language and action are facets of a single process that participants collaboratively organize through their practices of speech and gesture.

Language Use and Action

The conduit metaphor implies that language is largely a vehicle for making propositions about the world. From this perspective, which is explicit or implicit in traditional social psychological research on language, problems of meaning involve how well linguistic concepts refer to, correspond with, or represent reality, including internal thoughts and feelings. For example, in the symbolic interactionist tradition, although it is recognized that words as symbols have an arbitrary relationship to what they represent, nevertheless speakers learn to associate a given word with the “same things or events, as do other speakers of the language” (Hewitt & Shulman, 2011, p. 34). To be used accurately, announcing a “fire” on one’s premises requires that this word designate, refer, or point to some actual conflagration and that others understand this and respond appropriately. A different idea, stemming from developments in what is called ordinary language philosophy, is that language is a site of social activity. A variety of scholars, including Austin, Ryle, Searle, and Wittgenstein, have recast problems of meaning and reference in traditional philosophy and, by extension, issues concerning how, and under what conditions, interactants communicate effectively with one another. Their approach avoids theorizing about the abstracting and generalizing process through which words refer or point to objects. Instead, it situates words in concrete, orderly contexts to appreciate how they achieve actions. Saying “there’s a fire in here” can perform a variety of social actions: announcing, teasing, scaring, joking, testing, or others, depending on the organization of conduct in which the saying occurs.

Speech Act Theory

The title of John Austin’s famous book, How to Do Things with Words, conveys the essence of speech act theory. Austin (1962, p. 12) questions “an old assumption in philosophy” that to say something is to state something in a propositional sense. Sentences that convey referential information, in Austin’s words, form “locutionary” acts, but many utterances do not describe, state, or report anything. That is, they do not state anything and cannot be evaluated for their truth, but rather are “illocutionary” performances. Examples, paraphrased from Austin (p. 5), are:

“I do” (take this woman to be my lawful wedded wife) (as uttered during a marriage ceremony)

“I name this ship the Queen Elizabeth” (as uttered when smashing the bottle against the stem)

“I give and bequeath my watch to my brother” (as occurring in a will)

“I bet you it will rain tomorrow”

Such utterances do not report or describe what a person is doing; as formulated, they achieve a designated activity, such as promising, naming, giving, or betting.

Reflecting on the characteristics of these performances or illocutionary acts, Austin came to view locutionary acts in a new way. He proposed that the “occasion of an utterance matters seriously” and that to understand how the utterance functions, the “context” in which it is spoken must be investigated together with the utterance itself (J.L. Austin, 1962, p. 98). That is, when we examine the occasion of locutionary or statement-like acts, we see that speakers are using them to ask or answer a question, give assurance or a warning, announce a verdict or intent, and so on. Accordingly, so-called statements or locutionary utterances also occur as some specific action; they too are performative rather than referential. In the end, Austin (1961, pp. 249–50) abandons the dichotomy between locutionary and illocutionary acts “in favor of more general families of related and overlapping speech acts.”

One of Austin’s successors, Searle (1969, pp. 16–17), more forcefully states that the “unit of linguistic communication is not, as has generally been supposed, the symbol, word, or sentence … but rather the production of the symbol or word or sentence in the performance of a speech act,” and that a theory of language therefore needs a theory of action. For Searle, this theory is one in which a set of underlying, constitutive rules specifies how speech acts can be accomplished.

Both Austin (1962) and Searle (1969) attempt to come to grips with the well-known problem in the philosophy of language that a sentence with a given reference and predication can have an assortment of meanings. In terms of speech act theory, the “same” utterance can perform a variety of different speech acts. Searle’s (pp. 70–71) classic example is a wife reporting to her husband at a party, “It’s really quite late”:

That utterance may be at one level a statement of fact; to her interlocutor, who has just remarked on how early it was, it may be (and be intended as) an objection; to her husband it may be (and be intended as) a suggestion or even a request (“Let’s go home”) as well as a warning (“You’ll feel rotten in the morning if we don’t”).

Among speech act theorists, linking a given or “same” utterance to specific actions may involve what Austin (1962) called “felicity conditions,” or the set of circumstances that allow for the successful completion of a performative. Thus, for an act of promising to be effective, the promisor must intend to promise, have been heard by someone, and be understood as promising. Searle (1969, 1975), taking issue with Austin as well as others (H.P. Grice, 1957; Strawson, 1964) who base theories of meaning on speakers’ intentions, provides a sophisticated system of rules whereby the “direct” or “indirect” action a given sentence is intended to initiate can be consummated. For example, rules or conventions, according to Searle (1969) specify how an uttered promise is produced, what the preparatory conditions are (e.g., that the promise stipulates an act for someone that would not occur in the normal course of events), that the speaker intends to do the act as an obligation, and that the hearer recognizes the utterance as it was meant. These rules can be related to what Grice (1975) has called “conversational implicature,” a set of maxims that underlie and provide for the cooperative use of language.

Language Use as a Form of Life

Another important figure, and perhaps the most influential, in the ordinary language tradition is Ludwig Wittgenstein, who in his own early work was deeply committed to logical positivism and the idea that the function of language is to represent objects in the world. Subscribing to the referential approach to meaning, Wittgenstein thought that the fundamental question about language was the truth or falsity of its propositions. The philosopher’s main task was to translate complex sentences into their elementary units in order to assess its truth or falsity (Pitkin, 1972). Later, Wittgenstein disavowed this and any other rule-based approach to language, instead urging the examination of language practice—how individuals employ words and sentences in concrete situations.

In Philosophical Investigations (Wittgenstein, 1958) and other posthumous publications, he argued that language, rather than being a vehicle for naming things, conveying information, or enacting intentions according to rules, is an activity or form of life in its own right. For example, to analyze a single word in the language, and propose that there is a single definable class of phenomena to which it refers, is to neglect that words can be a wide variety of things depending on the various roles they occupy in a multiplicity of language games (Wittgenstein, 1958, para. 24). Consider the word “hello,” which we might define as a greeting. However, its status as a greeting depends on where, in a developing conversation, the item occurs (Schegloff, 1986). When a party uses the word after picking up a ringing telephone, the activity it performs is answering a summons rather than greeting the caller. Subsequently, there may be an exchange or sequence of salutations, and in that context “hello” does perform greeting. To discover the meaning of a word, then, it is not possible to rely on ostensive or referential or any other fixed definitions; one must examine the contexts of use. When contexts of use are similar, then words may be said to share what Wittgenstein (para. 67) called “family resemblances.” It is in the actual practice of placing words in particular contexts that such resemblances can be traced, and the lexical and other components of language appreciated as a form of life.

This emphasis on actual practice differs significantly from speech act theory, especially that of Searle. In Wittgenstein’s view, just as the word “hello” might appear in a variety of language games, so might the word “promise.” Rather than deriving meaning ostensively or from underlying constitutive rules, however, the word is always related to the force of the utterance in which it appears. Consider an example from an actual conversation at a family dinner table. Virginia, a teenager, has been asking her mother for a raise in her allowance, while her mother has been resisting the request. At one point, Virginia says, “I promise I never have enough money.”Footnote 2 Here, she is not making a promise in the conventional sense—assuring that she will do something in the future. Rather, in a context where she has made a request for an increased allowance and met with resistance, Virginia is complaining about her financial situation and justifying the request. Furthermore, by her use of “promise,” she intensifies her complaining/justifying actions. From a Wittgensteinian-informed perspective, an investigator would not try to derive meaning from definitions, from the rules of illocutionary force, or by inferring speaker intentions. Instead, the interest is in overt expressions, interactional contexts, and acts through which a word such as “promise” comes to life. Linguistic and interactional competence, in this view, consists in systematically relating given lexical items to other pieces of vocal (and bodily) conduct that signal how such items are produced and to be understood.

The “Mapping” Problem

According to speech act theory, the language that humans use can constitute an infinite variety of social actions (John R Searle, 1969). Austin (1962) suggests that there are on the order of a thousand or so actions, while Wittgenstein (1958, para. 23) proposes that there are “innumerable” activities in which language plays a part, including but by no means limited to “ordering, describing, reporting, speculating, presenting results, telling a story, being ironic, requesting, asking, criticizing, apologizing, censuring, approving, welcoming, objecting, guessing, joking, greeting.” This list can be indefinitely extended and shows that, as all the speech act theorists would argue, the communicative function of language, wherein people refer to objects and report their thoughts or feelings about them in a verifiable way, is only one among many modes of linguistic usage.

When social scientists regard language in this dynamic sense, as intimately tied to action, a seemingly simple problem still looms large for the investigator: How are we to know what the illocutionary force (action) of an utterance is? It is not tenable that the performative aspect of an utterance is somehow built into its form, for the reason stated above—the “same” utterance can perform a variety of acts. Put differently, the “form” of a sentence or utterance, including its syntactic structure, is often misleading about its status as an activity. For example, Levinson (1983, p. 275) mentions imperatives, which, despite their grammar as commands or requests, rarely appear as such in natural conversation. Rather, they occur “in recipes and instructions, offers (Have another drink), welcomings (Come in), wishes (Have a good time), curses and swearings (Shut up), and so on …”. As Levinson nicely formulates the problem of knowing the force of an utterance, it is one of mapping speech acts or social actions onto utterances as they occur in actual contexts. As we have seen, in ordinary language philosophy, there are two main solutions to this mapping problem, one being the rule-based approach of Austin, Searle, Grice and others, and the other being the practice-based approach of Wittgenstein. In contemporary social science, we also find these two approaches.

Sociolinguistics and Discourse Analysis

Language use as a topic is almost notable by its absence in social psychology, particularly of the psychological variety. Accordingly, we look elsewhere for sources of understanding the role of language in social interaction and social life. Prominent fields that relate linguistics, sociology, and anthropology include sociolinguistics and discourse analysis.

Sociolinguistics

Pioneers in sociolinguistics, such as Gumperz (1972), Hymes (1974), and Labov (1972b), were wrestling with a legacy of theorizing about language that posited its fundamental forms as being cognitive or minded phenomena. This legacy started with Ferdinand de Saussure’s (2011[1916]) famous distinction between langue, which comprises an underlying systematics across variations in social context, and parole, which consists of the actual speech that people produce. In de Saussure’s view, the proper focus of study was langue, the idea being that human cognition was the seat of linguistic structures and categories that guided people’s behavior. In more contemporary times, Noam Chomsky (1965) continued the cognitive legacy with his very influential notion of generative grammar, a set of psychologically based universal structures whose systematic transformations result in an infinite variety of human speech productions. With its emphasis on Cartesian mental properties, structural linguistics has always sought to decontextualize linguistic phenomena in favor of finding certain ideal properties of abstracted sentences. That is, the overwhelming tendency has been to view linguistic structure as extant outside of time and place, and hence not subject to social influence.

Sociolinguists, following scholars such as Firth (1935), Malinowski (1923), and others, were utterly dissatisfied with such a view. As Hymes (1974) has argued, the frame of reference of the social scientific investigation of language could not be linguistic forms in themselves, and must substitute the community context as a frame. Indeed, Labov (1972b) resisted the term sociolinguistics because he could not conceive of linguistic theory or method that did not incorporate a social component. The social component would include cultural values, social institutions, community history and ecology, and so on (Hymes, 1974). While sociolinguists agree that social influence is crucial to understanding linguistic structure, there are different perspectives on the relationship between society and language (Grimshaw, 1974) and varying strategies for investigating this relationship. The earliest sociolinguistic studies used dialect surveys to study speech variation among social networks and communities, finding that dialect variables were an excellent gauge of both social class and ethnic identity (Gumperz & Hymes, 1972).

Variation in linguistic patterns is a prominent theme in sociolinguistics. Besides dialect usage, another example of variation is code switching (Ervin-Tripp, 1972), or the manner in which members of a single community juxtapose, in the same situation, speech belonging to different grammatical systems (Auer, 1999; Breitborde, 1983). The uses of code-switching include but are not limited to quoting others, selecting a particular addressee (by using his or her native language), marking something as an interjection, reiterating a remark in one language by using another, and qualifying or specifying a generalization. Here are two examples from Gumperz (1982, pp. 77–78), where the code switching goes from Spanish to English or vice versa:

[Interjecting:]

A: Well I’m glad I met you.

B: Andale pues (O.K. swell). And do come again. Mm?

[Reiterating:]

A: I was … I got to thinking vacilando el punto ese (mulling over that point) you know? I got to thinking this and that reason.

In terms of a classic topic in social psychology—that of identity—sociolinguists suggest that code switching reflects speakers’ ability to categorize situations, interlocutors, and social relationships and thereby to make inferences and judgments about the appropriate and relevant speech forms to produce. Accordingly, sociolinguists examine the relation of diverse languages to self-concept, personality, and status attitudes. Other core topics in sociolinguistics are language conflict, loyalty, and maintenance, as well as the structure and organization of pidgin and creole languages.

Methodology in Sociolinguistics

Methodologically, the field of sociolinguistics relies on sampling a particular speech community (Gumperz, 1972, p. 16) or group whose speakers “share knowledge of the communicative constraints and options governing a significant number of social situations” to interview subjects or informants and record how they talk. Investigators use a variety of interview-based elicitation techniques (Chambers, 2008). In order from most formal to most casual style, these techniques include “word list” (subjects read a prepared inventory); “minimal pairs” (words with one phoneme that is different, as with cat and bat); “reading passage” (a prepared text); and “interview style” (subjects recall a car accident, or a fire in the toaster or other experience); and “casual style” as when a subject talks to someone else in the household or takes a telephone call.

The best data from a sociolinguistic standpoint is that which minimizes the speaker’s self-awareness, but there is what Labov (1972b, p. 209) famously termed the observer’s paradox: “the aim of linguistic research in the community must be to find out how people talk when they are not being systematically observed; yet we can only obtain these data by systematic observation.” Among the most successful of attempts to resolve this paradox—because it stimulates spontaneous talk—involves asking subjects to talk about a situation in which their life was in danger. Labov also points to the importance of recording and studying language use in natural groups, which anticipates the techniques in conversation analysis and discursive psychology.Footnote 3

Discourse Analysis

Related to sociolinguistics, and representing an effort to become more theoretically sophisticatedFootnote 4 about the relationship between language and society, is discourse analysis. The term “discourse analysis” can be used to refer to a number of quite different research traditions. Along with the linguistic discourse analysis discussed here, there is historical discourse analysis that usually focuses on written texts (Armstrong, 1983; Foucault, 1979), “critical discourse analysis” which combines social criticism with the analysis of textual material (Fairclough, 1992), and the social psychological discourse analysis that has come to be called “discursive psychology,” which will be discussed later in this chapter.

“Discourse” broadly includes both textual and spoken forms of language and refers to language production as it is organized external to the unitary sentence or clause (Stubbs, 1983), although, as van Dijk (1997b, p. 6) suggests, the field could include studies of what ordinarily is called prosody, including “pronunciation, emphasis, intonation, volume and other properties” contributing to the “sound structures of discourse.” Usually, discourse analysis is concerned with the orderly connections between clauses and sentences, rather than with the structuring of those units alone. That is, even when concerned with small units or characteristics of speech, discourse analysts go beyond these boundaries to discover how contexts of various kinds enter into the constitution of such units. Thus, as Coulthard (1977) notes, discourse analysis overlaps partially with pragmatics, a subfield in linguistics that is distinguished from traditional concerns with syntax and semantics by the interest in how language users take the social environment into account when producing and understanding speech forms. Discourse analysis is multitopical and multidisciplinary, with scholars from anthropology, artificial intelligence, communications, philosophy, psychology, and sociology contributing to the enterprise (Stubbs, 1983; van Dijk, 1985).

Some discourse analysts are interested in formalizing the relationship between language and other sociological variables. For example, Grimshaw (1989) models the discourse process as involving a “source,” or originator of some manipulative speech move, a “goal,” or target of the move, an “instrumentality,” which is the speech act itself, and a “result” or outcome that the source pursues. The particular speech act a source employs is constrained according to the three variables of power, affect, and utility. Grimshaw’s approach complements Labov and Fanshel’s (1977) concern with rules of discourse by emphasizing rules deriving from essentially social considerations of appropriateness as based on participants’ cultural and social knowledge. A less formalistic approach to describing discourse and its social parameters—how discourse as action involves topic selection, overall or schematic organization, local meanings, choice of words, style, and rhetorical devices—can be found in van Dijk (1997a). Viewing discourse as action, van Dijk also stresses the importance of context and power in the analysis of text and talk.

Given the multitopical and multidisciplinary character of discourse analysis, it is difficult to define any unitary methods. As Wood and Kroger (2000, p. 28) put it, “In discourse analysis, the units of analysis are variable and may range from words, phrases, and sentences to paragraphs or even larger units.” Starting small, one could consider a progressive approach to discourse analysis, where linguistic methods would be appropriate for studying the order of words, phrases, or clauses in sentences—their syntax. Semantic and cognitive psychological approaches might take on the next level, having to with the assignment of meaning to whole sentences or clauses. Ethnographic methods would be appropriate for analysis of style and variation in speech and text, which discourse analysts consider particularly important because of their relation to the accomplishment of social identity. As Gee (2010, p. 28) states:

People build identities and activities not just through language, but by using language together with other “stuff” that isn’t language. If you want to get recognized as a street-gang member of a certain sort you have to speak in the “right” way, but you also have to act and dress in the “right” way, as well. You also have to engage (or at least, behave as if you are engaging) in characteristic ways of thinking, acting, interacting, valuing, feeling, and believing. You also have to use or be able to use various sorts of symbols (e.g., graffiti), tools (e.g., a weapon), and objects (e.g., street corners) in the “right” places and at the “right” times.

Gee (2010) proposes that the “same is true of doing/being a corporate lawyer,” and it is clear that getting at the various components to style and identity may best be accomplished through field research. Beyond stylistic matters, critical communication methods may handle rhetorical aspects of discourse and other schematic themes or global meanings that discourse may constitute.

As van Dijk (1997b, p. 13) notes, each step along this methodological pathway involves “structures that are further removed from the traditional scope of linguistics” until the social sciences become especially relevant for the study of action and interaction. For these topics, many discourse analysts such as van Dijk consider ethnomethodology and conversation analysis to provide the relevant methodological tools (see below), although these subfields tend not to deal with the formal and written aspects of language that discourse analysis includes as part of its concerns (Stubbs, 1983). It is clear that discourse analysis is not only multitopical and multidisciplinary, but also multi-methodological.

Goffman and Frame Analysis

Sociolinguistics and discourse analysis emphasize the importance of microanalysis of minute particles of speech and single interactional events as a means for understanding the social and social psychological dimensions of language use. Akin to the speech act tradition, both areas invoke rule-like mechanisms for connecting social conditions, environments or structures to these particles and events. In Goffman (1983)’s work—which, in the latter part of his career, involved an increased interest in language use—we see less emphasis on the connective or even causal approach to rules and more concern with social actors’ agency and rule usage. As background to discussing Goffman’s (1981) later focus on talk, we consider the theoretical context in which that focus resides.

Goffman (1983) argues that the corporeal and interactional “face to face” or “body to body” situation—whether in urban or rural areas, business or family, and independent of socioeconomic class, gender or ethnic categories—should be the primary focus for understanding social interaction. That is, the same rules and conventions, applying to turn-taking, physical distance between speakers, and other matters, prevail in social interaction regardless its broader context. Or to take a more specific example: Goffman refers to a “contact” ritual, such as any service encounter where customers may form a queue as they await their turn at being helped. Although the queue could be organized according to externally structured attributes of involved parties (e.g., age, race, gender, or class), normal queuing “blocks” or filters out the effects of such variables in favor of an egalitarian, first-come, first-serve ordering principle.

Such an ordering principle belongs to what Goffman (1983: 5) calls the interaction order, which consists of “systems of enabling conventions, in the sense of ground rules for a game, the provisions of a traffic code, or the syntax of a language.” The interaction order is relatively autonomous order of organization, both in relation to the broader social organization and to the psychological properties of the actors. Hence, Goffman wanted to promote it as a target of social scientific study in its own right. Although the interaction order consists largely of rules or conventions, violations do not threaten the game or the language as much as they serve as resources for accomplishing the very projects that adherence itself involves, including the definition of self and the creation or maintenance of social meaning (Goffman, 1971, p. 61):

Given that a rule exists against seeking out a stranger’s eyes, seeking can then be done as a means of making a pickup or as a means of making oneself known to someone one expects to meet but is unacquainted with. Similarly, given that staring is an invasion of information preserve, a stare can then be used as a warranted negative sanction against someone who has misbehaved—the misbehavior providing and ensuring a special significance to overlong examination.

Actors, in this view, do not range between naive conformity and blatant rule breaking. Rules, says Goffman (1971, p. 61) make possible a set of “nonadherences” which, according to how we classify the interactional work they do, have a variety of meanings. In social psychological terms, actors’ orientation to the interaction order rests on commitments that in one way or another (through adherence or violation) enable the self to emerge and be preserved (Goffman, 1971; Rawls, 1987). The interactional rules do not tightly constrain actions; they are more like rough guidelines that permit actors to accomplish a variety of social projects, depending on how they align themselves with respect to those rules or guidelines.

This point about actors’ capacity for flexible alignment to rules is most fully developed in Frame Analysis, Goffman’s (1974) major treatise on the “organizational premises” of ordinary activity and the “reality” of everyday experience. This work, and particularly a chapter entitled “Frame Analysis of Talk” brings full attention to the use of language in interaction. Much of everyday experience goes beyond literal activity and has numerous figurative aspects, which are especially visible in talk (1974). In particular, Goffman argues that rather than using terms such as speaking and hearing to characterize the production and understanding of utterances, analysts must see how participants display a stance with respect to those utterances. A speaker, for instance, may employ a variety of production formats when talking, so that he/she says something as principal (one whose position is represented in the talk) or as animator (who simply speaks the words representing another’s position). As principal or animator, one can also project a particular identity or figure (ranging from that of the speaker to identities of fictitious and actual others). Finally, a speaker can be a strategist who acts to promote the interests of an individual on whose behalf he/she is acting. In a way complementary to speakers, hearers also take up different alignments or participation statuses—ratified recipient, overhearer, eavesdropper, and so on. Eventually, Goffman (1979) referred to the frame analysis of talk as an investigation of the “footing” or stances that participants constantly change over the course of an utterance’s production.

Methodologically, Goffman based his research on a combination of ethnography and observation. Some contemporary scholars have used these methods to develop and critique his work. For example, in her studies of gender in public places, Gardner (1989, 1995) argues that Goffman’s claim about public order is that it is gender-neutral. This caused him to overlook the inherently gendered quality of social interaction. In analyzing public behavior from a feminist perspective, Brooks-Gardner foregrounds how women, who (along with minorities) are vulnerable to various forms of harassment and discrimination, experience public places differently from men.

Goffman’s method of frame analysis, and the corollary concept of footing or stance, has been taken up in a variety of ways. This method provides tools for distinguishing among and structural bases for the multiple identities people enact in a given situation. Where Goffman often utilized fictitious examples, more recent studies of actual interactions demonstrate the utility of his methodological orientations. Maynard (1984) analyzes how in plea bargaining, district attorneys and public defenders strategically shift footings to align with, and distance themselves from, the structural roles in which they are embedded. Thus, a public defender may animate his client’s wishes while also distancing himself from them. Similarly, Clayman (1988) shows how news interviewers achieve neutrality by shifting to the footing of animator, rather than principal, of challenges to an interviewee. That is, interviewers attribute challenging questions to a third party. For example, an interviewer on a public television station in the U.S. once asked the South African ambassador to the United States about a state of emergency that had been imposed in his country in this way: “Finally Mister Ambassador, as you know, the critics say that the purpose of the state of emergency … is to suppress political dissent, those who are opposed to the apartheid government of South Africa. Is that so?” (Clayman, 1988, p. 482 with simplified transcript). By referring to “critics,” the interviewer tacitly claims that the challenge originates with others and not himself.

Other research further extends Goffman’s concepts by deploying them for the analysis of multi-modal activities, including gesture and environment as well as talk (Mondada, 2012). C. Goodwin (2007a) shows how “participation frameworks” are enacted in the physical alignments of actors’ bodies toward one another. When people engage in teaching activities, for example, they configure their bodies to establish joint attention to an object. Disruptions of this organized embodied expression lead to shifts in participants’ affective stance in the interaction. Similarly, M. Goodwin (2006) shows how embodied alignments (“facing formations”) can affect whether children comply with their parents’ directives, and also how bodily stances can accomplish inclusion and exclusion within young girls’ peer groups (M. H. Goodwin, 2007b).

Discursive Psychology

Discursive psychology is a European (mostly British) social psychological approach that takes an “action oriented” understanding of language as its point of departure. Scholars such as Billig (1987), Edwards (1997), Edwards and Potter (1992), Potter (1996), Potter and Wetherell (1987) and Antaki (1994) have questioned the “cognitivist” presuppositions predominant in current social psychology. The cognitivism that discursive psychologists critique coincides with the conduit metaphor. In cognitivism, “we start with a given, external world, which is then perceived and processed, and then put into words” (Edwards, 1997, p. 19), and language is understood as a transparent medium used for transfer of ideas concerning the external reality and inner worlds of humans. In contrast with this view, discursive psychologists study accounts and accounting—how everyday descriptions of people, their behavior, and their mental states are in themselves actions (Antaki, 1994). Descriptions are produced in particular occasions to do particular things, such as blaming, justifying, explaining, and so on (Buttny, 1993). Descriptive themes as accounts involve courses of action, mental and emotional states, and identities. We discuss these themes in order.

Accounts of Courses of Action

Following Schegloff (1989), and Potter and Wetherell (1987), Edwards (1997, p. 8) remarks, “accounts of actions are invariably, and at the same time, accounts for actions.” Two distinct aspects of these accounts involve scripts and dispositions (Edwards, 1997). In describing events in terms of scripts, the speakers often implicitly (and sometimes explicitly) propose that what happened followed a routine pattern in the given circumstances. The course of action is then presented as expected, as ordinary, and as “natural:” in short, as one that follows a script. On the other hand, events can also be described as breaches of the script, as something unusual and unexpected. When events are described as breaches of the script, dispositions often come into play. Dispositions are “pictures” of the actor implied by the description of the course of action; two such relevant dispositions are the personality and the moral character of the actor. Speakers use scripts and dispositions as explanatory resources in pursuing their local interactional goals (Edwards & Potter, 1992). For example, a speaker complaining about another party’s conduct may propose that it is part of a recurrent—or scripted—pattern of violating social norms (Edwards, 1995).

Accounts of Mental and Emotional States

Discursive psychologists are interested specifically in the ways in which the participants’ states of knowledge figure in talk (Edwards, 1997). They examine how emotional and cognitive states are practically accomplished, and how local interactional goals are pursued in and through them. Cognitive states are achieved, for example, through the ways in which statements, stories and descriptions are designed and received in conversation. As conversation analysts have shown (see below), speakers produce their talk carefully to show their understanding of the recipients’ knowledge or “epistemic” states. By the same token, recipients may show, through their own action, whether the things that were told were new information or already known by them (Sorjonen, 2001).

Discursive psychology also investigates descriptions of affect, or the ways in which speakers avow their own emotions and ascribe them to others. In line with other social constructionist approaches (Harré, 1986), this research centers on the use of emotion words (rather than non-lexical expressions of emotion) and their role in actions such as “assigning causes and motives of action, in blamings, excuses, and accounts” (Edwards, 1997, p. 170). Thus, during an argument, one spouse may blame the other’s jealousy as the source of his anger and their fights (Edwards, 1995). Emotion descriptions are an essential resource in accounting for action. Moreover, as Edwards (1997) points out, emotion descriptions can be embedded in routine scripts—as when, for instance, a particular event, such as having a child, evokes a particular emotion, such as happiness. Emotion descriptions also can be part of dispositions—e.g. when a specific emotion, such as inclination towards jealousy, is used to explain non-routine courses of action.Footnote 5

Accounts of Identity

Identity is a third theme in discursive psychology. In and through their talk, speakers present themselves, those to whom they talk, and those about whom they talk, as having particular identities, and being particular kinds of persons. Like mental states, identity is, as Antaki and Widdicombe (1998) put it, both an achievement and a tool in performing particular actions in talk (Edwards & Potter, 1992). Thus, in blaming another, or defending one’s own (or the other’s) actions, speakers ascribe and avow particular motives and personality features, and thereby construct identities (Potter & Wetherell, 1987). For example, a speaker criticizing protesters may divide them into two groups—those who have genuine motives and those who do not. This, in turn, allows the speaker to support the protest in general while objecting to certain aspects of it, such as violence. The speaker’s construction of the protesters’ identities has reflexive implications for her own: she is someone who supports the protesters’ cause, but not the protesters per se.

Drawing on Sacks’ work in conversation analysis, Antaki and Widdicombe (1998) emphasize the centrality of categorization in the construction of identity: “to have an identity” entails being “cast into a category with associated characteristics or features.” Categories can, of course, be numerous, the most general ones including age, ethnic, gender and professional categories. A key challenge in investigating categorization, as Antaki and Widdicombe point out, is to show how participants orient to a particular categorization, and how this orientation is consequential for their joint courses of action.

Methodology in Discursive Psychology

In this section, we have reviewed three broad and interrelated areas of description as action: accounts of courses of action, accounts of mind and affect, and accounts of identities. In all these fields, discursive psychologists seek to show how the design and reception of descriptions contributes to particular social actions. This research program raises once again the mapping problem: on which basis can we say that a type of description contributes to a particular social action? Discursive psychology tends to blend the methods of both discourse analysis and conversation analysis. More specifically, Potter (2012) traces three strands of methodological influence that developed over the years. First, with influences from Billig’s rhetorical psychology (Billig, 1987), are open-ended interviews and group discussions from which investigators identify “interpretative repertoires,” or categories and idioms such as those surrounding identities. Second, in a discourse analytic way, discursive psychology began dealing with naturalistic data including talk (legal arguments, parliamentary debates, news interviews) as well as texts (newspaper reports). This strand has been important for the studies of course-of-action accounts as well as socially constructed mental and emotional states. Third, in recent years, the research methodology of discursive psychology has come very close to that in conversation analysis: “Indeed, at times these two fields blur together” (Potter, 2012, p. 122). Accordingly, we postpone further discussion of methodology in discursive psychology to explore the interrelated traditions of ethnomethodology and conversation analysis and their associated methodological practices.

Ethnomethodology and Conversation Analysis

Ethnomethodology proposes that there is a self-generating order in everyday activities (Garfinkel, 1967) and takes a unique approach to the problem of mapping utterances onto actions in at least two ways. First, where Goffman’s frame analysis relaxed the theoretical hold that rules could have in explaining linguistic conduct, ethnomethodology argues that rules should instead be treated as topics and features of the activities they are said to organize. In so doing, it extricates rules from their traditional conceptual status in social theory. That is, in ethnomethodology there is no attempt to explain linguistic or other behavior by reference to rules. Instead, the analytic tactic is to examine rules empirically as resources for actors, who use them for various situated projects and ends of their own. It is not that behavior is unconstrained, disorderly, or arbitrary, but that rules, if they are operative at all, figure as part of actors’ own practices of reasoning and ways of organizing a social setting. People are artful users of rules, often invoking them in an ex post facto, rhetorical manner to describe the morality of some way of life. For example, jurors retrospectively invoke legal standards to depict how they arrived at a verdict, even when the route involved substantial commonsense, non-standardized reasoning (Garfinkel, 1967). Ethnomethodologists also have shown how residents at a halfway house use the “convict code” to account for disregard of the official ways of doing things (Wieder, 1974). In another study, Zimmerman (1970) demonstrates how staff members at a social welfare agency get their “people processing” job done, in part, through departing from routine policies while still providing an accountable (defensible) sense of having conformed to them. Rules, to repeat, are features of actions rather than explanations for them.

The Transition to Conversation Analysis

Another unique aspect of ethnomethodological research is its concern with indexical expressions (Garfinkel, 1967; Garfinkel & Sacks, 1970), or utterances whose meaning and understandability depend on the context or circumstances in which they appear. While it is generally recognized that “deictic” utterances, such as “this,” “that,” “here,” “there,” and so on, assume particular meaning according to their speech environment, Garfinkel developed this insight further (1967) arguing that all talk is fundamentally indexical and context- dependent. One major, orderly aspect of “context” is an utterance’s sequential placement. Conversation analysis theorizes that an utterance’s force as an action of a particular type derives from such placement (Heritage, 1984; Maynard & Clayman, 1991). Thus, rather than linguistic or social rules, it is sequential organization and the interaction between speakers and hearers that have primary analytic utility in describing utterances as action (Schegloff, 1991). Overall, ethnomethodology and conversation analysis have affinities with the Wittgensteinian “form of life” approach to the mapping problem, in which actual, orderly linguistic practice (rule usage and sequence organization) is brought to the fore of analytic inquiry.

With its commitment to the study of naturally occurring talk, conversation analysis in particular aims to rebuild sociology as a natural observational science (Sacks, 1984, 1992) in three senses: (1) it is possible to formally describe social actions and activities, because (2) these actions and activities are methodical occurrences, and (3) the methods by which a single action or activity is composed are generalizable to and reproducible in other situations. In pursuing the goal of studying talk and building a science of interaction, conversation analysts have generated a sizable research literature over the past 35 years (Clayman & Gill, 2012; Heritage, 1984; Sidnell & Stivers, 2012). Furthermore, conversation analysts have published on a wide variety of social psychological issues related to interaction. Many of these publications appear in other journals, but if we confine ourselves to Social Psychological Quarterly, these issues include relationship and ritual in relation to topical talk (Manning & Ray, 1993; Maynard & Zimmerman, 1984), doctor-patient communication (Gill, 1998; Heath, 1989; Lutfey & Maynard, 1998; Peräkylä, 1998; Stivers, 2007), epistemological orientations in ordinary as well as institutional settings of talk (Heritage & Raymond, 2005; Speer, 2012), and emotion displays (J. Whalen & Zimmerman, 1998; Wilkinson & Kitzinger, 2006), among others (Hepburn & Potter, 2011; G. H. Lerner, 1996; Whitehead, 2009). In 1987, there was a special issue of SPQ on “Language and Social Interaction” (Maynard, 1987) that included topics such as single episode analysis (Schegloff, 1987), forgetfulness as a resource (C. Goodwin, 1987), the job interview as an interactional event (Button, 1987), and the organization of 911 calls for help (Whalen & Zimmerman, 1987).

In maintaining a commitment to examining naturally occurring social action, conversation analysis avoids treating language as a variable to be manipulated, tested, or related to other variables. We explore the implications of this stance in the next section. Here, the point is that conversation analysts’ major social scientific concern has been with endogenous (internally orderly) features of “talk-in-interaction” (Schegloff, 1991). The primary focus in conversation analysis is sequence organization. In the next section, we illustrate one form of this organization, while also identifying others.

Organization of Sequences: Adjacency Pairs

It is well established that conversational interaction occurs in a tightly ordered serial or (as mentioned) sequential fashion. Sequential structure is exemplified in the adjacency pair, a ubiquitous type of unit that includes such conversational objects as questions + answers, requests + grantings or refusals, invitations + acceptances or declinations, and many other such pairs. Characteristically, adjacency pairs are (1) two-utterances in length, (2) adjacent to one another, (3) produced by different speakers, (4) ordered as a first part and a second part, and (5) typed, so that a first part requires a particular kind of second part (Schegloff & Sacks, 1973).

Moreover, adjacency pairs are characterized by “conditional relevance”—conditional on the occurrence of an item in the first slot, or first pair-part (e.g., the question), the occurrence of an item in the second slot, or second pair-part (e.g., the answer to the question), is expected and required. When second pair parts do not occur, their absence is noticeable and treated as accountable by first speakers, who may then interpret the recipient as “ignoring” them, “snubbing” them, not hearing them, or otherwise resisting the their initial action. Importantly, it is the participants themselves that make inferences concerning these kinds of actions. Conversation analysts work secondarily with these participant-based inferences, insofar as they are displayed in the ongoing interaction.

Adjacency pairs can be expanded in three ways. One way is through an insertion sequence between first and second pair-parts of a basic sequence. With invitation sequences, for example, a recipient may need pertinent details before providing a reply (Schegloff, 1972, p. 78):

1. A: Are you coming tonight? [First pair part of base adjacency pair]

2. B: Can I bring a guest? [Insertion first part]

3. A: Sure. [Insertion second part]

4. B: I’ll be there. [Second pair part of base adjacency pair]

Here, the “base” sequence—an invitation (line 1) and its reply (line 4)—is separated by the insertion sequence at lines 2 and 3.

Two other ways of enlarging the adjacency pair involve “pre-” and “post-”expansions (Schegloff, 2007, p. 27), both of which are illustrated below. The base sequence includes the invitation and its acceptance at lines 3 and 4. A pre-invitation sequence occurs at lines 1 and 2, where A checks out B’s circumstances and B indicates a possible availability with a “go-ahead” signal. Other responses to a “pre” are a “blocking move” that forestalls the production of the base sequence, and a “hedging” move that is a kind of wait-and-see response.

1. A: Whatcha doin’? [Pre-Expansion]

2. B: Not much [Go-ahead signal]

3. A: Ya wanna drink? [First Pair Part of base adjacency pair]

4. B: Yeah [Second Pair Part of base adjacency pair]

5. A: Okay [Post-Expansion]

In this example, subsequent to the base sequence can be a post-expansion, which in this case is a “sequence closing third” (line 5). By doing closing, this move minimizes the post-expansion (which can be more extensive) and allows for movement to a next topic.

Other Kinds of Sequence Organization

Turn-Taking

Conversations may consist of a series of adjacency pairs and their expansions that also involve recurring transfer of speakership. The ordering of speaker change, as well as the size and content of a speaker’s turn, is not predetermined in ordinary conversation but instead is free to vary (Sacks, Schegloff, & Jefferson, 1974). Moreover, change of speakership is so tightly articulated that both gap and overlap are minimized (Lerner, 1989; Sacks et al., 1974; Emanuel A. Schegloff, 2000). Participants methodically allocate turns of talk through a set of ordered options, including current speaker selecting the next speaker, the next speaker self-selecting, or current speaker continuing to speak (Sacks et al. 1974). In coordinating exchange of speakership and tightly articulating sequences, participants who take a turn of talk are required to display their understandings of a previous speaker’s turn. Methodologically, this is considered a proof procedure: the second speaker’s turn serves as a resource by which the first speaker may check whether a turn was heard correctly. Moreover, the second speaker’s turn aids the analyst in characterizing the action of the first turn.

Repair Sequences

Given the elaborate and systematic organization of adjacency pair sequences and turn-taking, how are interactional troubles managed? That is, how do participants handle errors, mishearings, glitches in turn transition, problems of meaning, and the like? The answer is that the there are ways of both initiating and accomplishing repair, including devices by which the turn-taking system itself is deployed to fix problems. For example, “if two parties find themselves talking at the same time, one of them will stop prematurely” (Sacks et al., 1974, p. 701), thereby permitting the other “current speaker” to talk in the clear. Dropping one’s own turn effectively enacts the practice of a current speaker selecting a next speaker, and ensures the “one speaker at a time” feature of conversation. The devices and sequences involved in repair organization are many and complex and are important resource for the preservation of mutual understanding and the achievement of intersubjectivity (Hayashi, Raymond, & Sidnell, 2013; Schegloff, 1992).

Overall Structural Organization

Single conversations can be said to have discrete aspects to their organization, such as openings, topical structure, and closings. One import of this is that an utterance may obtain its status as an action through its relation to this overall organization. When a caller, just after introducing herself to a call recipient, produces “Was Bryan home from school ill today?” its placement just there (after self-identification) helps inform the call recipient that this inquiry is the “reason for the call.” In general, overall structural organization has to do with how participants form an entire occasion of interaction from beginning to end. Analysis may involve how the placement of utterances and other embodied devices in this overall organization informs the construction and understanding of these devices as turns, adjacency pairs, and social actions (Schegloff, 2007, p. xiv).

Conversational Epistemics

Domains of sequencing (adjacency pairs, turn taking, repair, overall structural organization, and others) provide the basis for much of the vigorous research agenda in conversation analysis. A recent addition to this literature, although reaching back to discussions of “practical epistemology” (Whalen & Zimmerman, 1987) and related phenomena, has involved what is called “epistemics in action” (Heritage, 2012a) and the “epistemic engine” (Heritage, 2012b) of conversational interaction. In an early exploration of these topics, Heritage and Raymond (2005) show how, in doing “assessments” or evaluating social objects and experiences, speakers exhibit their epistemic stance or knowledgeable position regarding the object or experience. Consider this example (Heritage & Raymond, 2005, p. 30, simplified transcript):

Emma: How was your trip?

Lottie: Oh god wonderful Emma.

Emma: Oh isn’t it beautiful down there?

As Heritage and Raymond (2005, p. 30) observe, Lottie is the most knowledgeable about her trip and her assessment (“Oh god wonderful”) reflects that direct knowledge, which Emma lacks. Emma, however, follows with an oh-prefaced negative interrogative (“Oh isn’t it beautiful down there”), whose referent is the location rather than the trip and asserts Emma’s own “generalized experience of Palm Springs rather than Lottie’s more immediate experiences there.”

More than exhibiting epistemic positions, participants manage their rights to assess such objects and experiences according to their own states of knowledge. They do this management by way of what they say and how they say it in their turns as these occupy first or second position in assessment-type adjacency pair sequences. With regard to other kinds of action besides assessments, it is the case that access to knowledge or who has “primary epistemic status” may take precedence over syntax and intonation in forming such actions. A good example is the conveyance or requesting of information. Using declarative syntax (“Things have arrived from Barker and Stone House”) when one has primary access to the state of affairs suggests the conveyance of information. However, using the same kind of declarative syntax (“You’re divorced currently”) when a recipient is the knowledgeable one, asks for confirmation (Heritage, 2012a, p. 8). An upshot of this line of research is that participants in conversation must monitor the distribution of knowledge between themselves and others as a condition of being competent interactants. It follows that professional analysis of conversational interaction also needs to pay attention to epistemic work going on in the construction of turns and actions.

Methodology in Conversation Analysis

Conversation analysis (CA) involves several methodological orientations: (1) analyzing utterances as actions, (2) engaging in sequential analysis, (3) analyzing participant orientations, (4) regarding interactional detail as a site of social organization, and (5) using both single and multiple episodes along with deviant cases for analyzing phenomena.

We have already discussed (1) how utterances perform actions, which is a tenet that cross-cuts the various perspectives on language use and social interaction. In everyday conduct, participants perform an immense variety of social actions that include informing, criticizing, insulting, complaining, giving advice, requesting, apologizing, joking, and so on. For CA, the crucial element in identifying actions and their interactional force is (2) sequential analysis, which includes features of turn-taking and adjacency pairs along with (3) the above-mentioned “proof procedure” whereby investigators discipline their characterization of actions by attention to a recipient’s displayed understanding of a speaker’s talk. (4) Transcription conventions developed by Jefferson (1983, 2004) depict silences and their duration, overlapping talk, sound stretches, emphasis and other such matters. These details are important to CA because they enter into and help constitute social actions. Once unique to CA, the transcription process and its capture of interactional detail have been adopted by discourse analysts and other investigators as well.

The CA perspective aims to develop claims about systematic structural organization in interaction. Such claims are supported by substantial accumulations of instances of a practice, each instance of which the investigator examines as an individual case. In fact, there can be analyses in which the analyst uses resources developed from past work to explicate a single episode of talk (Schegloff, 1987). A prominent methodological device is examining departures from an interactional regularity, or what is known as deviant case analysis, which allows researchers to validate empirical findings and discern larger patterns in which a practice helps achieve particular social actions. For example, in a study of diagnostic news about HIV infection, Maynard (2003) found a practice contrary to patterns documented in a variety of health care settings where clinicians, in delivering diagnostic findings, overwhelmingly work to shroud bad news and expose good news. In the HIV clinic, counselors often delivered the bad news of being HIV-positive as forthrightly as they presented the good news of HIV-negative status. In other words, rather than shrouding the bad news, they exposed it. Examining these deviant cases revealed that the counselors were attempting to “crack the emotional nut”—the often stoic way in which clients would receive bad news about HIV infection. The tactic was meant to prompt the discussion of what Peräkylä (1995) calls dreaded issues that are associated with HIV and AIDS by facilitating the flow of interaction between counselor and client.

Language Use, Action, and Social Structure

Thus far, we have concentrated on interaction, suggesting that social psychology benefits from understanding how parties use language in an immediate sense to perform joint endeavors of all sorts. Of course, as parties talk and gesture to one another, more than completely local interests and social organization may be at stake, and this means that questions regarding “social structure” come to the fore. Roughly following Zimmerman and Boden’s (1991) reflections on talk and social structure, we consider two main approaches to probing the interrelation of language, action, and social structure. First, we consider a macrodirectional approach in which facets of social structure are seen to affect patterns of language in use. Second, we consider a more dialectical approach in which there are reflexive relations between language use and social structure.

Macrodirectional Approach: Social Categories and Language Use

Investigators often see social structure as consisting of such forms as age, gender, class, and other socio-demographic categories—as well as culture, institutions, and complex organizations—which condition the use of language in specifiable ways. “In such a framework,” Zimmerman and Boden (1991, p. 5) remark, “talk and, indeed, all interaction of actual actors in social situations is seen as a product of those social forces.” This is the strategy in experimental and survey-based social psychology that examines how social structural arrangements condition language and social interaction, and emphasizes the relationship between social statuses or categories (e.g., race, gender, class, and age) and language.

Social Class

Perhaps the best known work in this area is that of Bernstein (1961, 1972), who proposed that middle and working-class children learn two very different linguistic “codes”—an “elaborated” and “restricted” code, respectively, with the features of each determined by the forms of social relations in different communities. Middle-class subcultures assert the primacy of the individual “I” over collective “we,” which results in an elaborated code characterized by flexible organization and a range of syntactic options. In contrast, in working-class communities the collective “we” is used over the “I,” and the result is a restricted, more rigid code with low levels of syntactic and vocabulary selection, and implicit rather than explicit meanings (Bernstein, 1972). These two class-based codes, Bernstein argues, help account for middle-class children’s success and working-class children’s lack of success in school.

Bernstein’s argument generated a vigorous response. The argument was related to the notion that, in the U.S., low-income African American children upon entering school were “culturally deprived” and capable of engaging only in “emotional cries” and a “non-logical mode of expressive behavior” (Bereiter, Engelman, Osborn, & Reidford, 1966, pp. 112–113). Portraying Bernstein’s as well as Bereiter and Engelmann’s (1966) analyses as deficit models, Labov (1972a) demonstrates that the “nonstandard English” spoken in U.S. African-American communities is not “restricted” in its flexibility or range of options for syntax or vocabulary and, in certain ways, exhibits impressive linguistic, social, and cultural complexity and competence on the part of the speakers. More recently, Goodwin (1990) shows how skilled urban African-American youth are in various linguistic activities (especially disputing) whereby they display and generate “character” and achieve localized social organization. Thus, Labov has argued that there is no relationship between language use or the “codes” employed in poor and working-class African-American communities and failure in school. Instead, “failure” may lay within the school as a social institution that does not adapt to the cultures of the diverse communities it serves. Controversy about whether linguistic repertoires represent “differences” or “deficits” continues (Baugh, 1999; J. Edwards, 1979; Giles & Robinson, 1990).

Gender

Studies of the relationship between language and social stratification are related to numerous comparisons of speech practice—based on cross-cultural, gender, and ethnic differences. Perhaps most prominent are investigations of linguistic divergences between women and men. Differences between men’s and women’s speech appear to be enough for Tannen (1990) to propose that males and females speak different “genderlects.” Early research suggested that women are more expressive in intonation; that they use more adjectives and intensifiers, including “so,” “such,” “quite,” “vastly,” and “more”; that they make more precise determinations of color (Key, 1972); that they employ more fillers, such as “um” and “you know;” and that they more often use affectionate address terms, such as “dear,” “honey,” and “sweetie” (West & Zimmerman, 1985). As it turns out, when researchers examine these items as simple markers or indicators of female speech, only a few show any consistent patterning. Compared to men, women produce speech in phonetically more correct forms (Thorne & Henley, 1975) and vary their pitch and intonation more (West & Zimmerman). Also, there is evidence that females are more likely to interpret remarks indirectly rather than directly (Holtgraves, 1991), and that men may initiate more “unilateral” (as compared “collaborative”) topic changes in interaction (Ainsworth-Vaughn, 1992; West & Garcia, 1988).

Whether there are distinct “genderlects” is controversial, however. Conversation analysts have “de-gendered” certain actions that are commonly thought to be distinctive to women or men (Speer & Stokoe, 2011a). The tradition of research concerned with asymmetries between men and women that was initiated by West and Zimmerman (1983), which found that men interrupt women more than the reverse in cross-sex conversations, has shown few consistent results (Aries, 1996; Kitzinger, 2008). Other status and power differences (Kollock, Blumstein, & Schwartz, 1985) as well as processes intrinsic to the interaction (Okamoto & Smith-Lovin, 2001), including participation rates and manner (topic-changing behavior) may overshadow a characteristic such as gender. In her detailed analysis of interruptions in mixed and same-sex conversations, Kitzinger (2008) found no evidence that males interrupt more often than females. Following Sacks et al. (1974), she reports that actions commonly coded as “interruptions” are simply instances where two incipient speakers begin to talk at the same time because they projected the grammatical and pragmatic completion of a turn. Moreover, where an action is in fact an interruption, it is often not done in a competitive way to prevent someone from completing a turn at talk, but rather is cooperative. For example, a listener may offer help in a word search in which a current speaker is engaged.

Similarly, investigators have contradicted a common claim that women use tag questions—where the speaker appends (“tags”) a question to the end of a declarative statement—more often than men (Lakoff, 1975). However, Potter and Hepburn (2011) show that, contrary to the belief that such questions make statements less assertive and more polite they can in fact be used in ways that are both coercive and invasive. For example, speakers can use tag-questions to propose that recipients know, or ought to already know something, and thereby predispose those recipients into a position that aligns with the speaker. Finally, in her analysis of mixed-sex meetings, Ford (2008) shows that contrary to the belief that men and women have different participation styles, both use the same strategies to assume and retain speakership.

Talk and Social Structure: Dialectics and Reflexivity

Dialectics

A dialectical approach to talk and social structure involves social structure as both the cause and outcome of spoken interaction. Language is the site of the production and reproduction of socio-demographic, cultural, institutional, and organizational forms characteristic of the overall society. It is therefore important to know both the local and broad context in which utterances occur, making it incumbent on the investigator to engage in ethnographic inquiry to complement the analysis of recorded speech. Indeed, there is considerable writing about the role of ethnography in studying talk, and Duneier and Molotch (1999) provide an excellent example as well as methodological discussion.Footnote 6

The dialectical premise is central to cognitive sociology (Cicourel, 1981), informing the work of students of talk in such institutional settings as preschools (Corsaro, 1979, 1996), schools (McDermott, Gospodinoff, & Aron, 1978; Mehan, 1979; Phillips, 1982), universities (A. Grimshaw, 1989), doctor’s offices or hospitals (Cicourel; Fisher, 1983; Silverman, 1987; Strong, 1979; Waitzkin, 1991) and courts (Danet, 1980; Molotch & Boden, 1985). As an example of this approach, Mehan (1991) argues that the “social facts” of school systems—including designations of disability and special needs—derive from the “practical work” of educators engaged in interaction with students, parents, and other professionals in a series of “microevents” that occur in the classroom, testing sessions, and meetings.

As another example of the dialectic approach, Corsaro (1992) develops an “interpretive” approach to childhood socialization, which challenges the view of socialization as a linear progression from the tabula rasa of infancy and childhood to full fledged adulthood, as if the individual only gradually and in an individualistic stage-like fashion becomes more competent and social over the course of years. Drawing on the classic works of Piaget and Vygotsky and adding more contemporary views of Bourdieu (1991) and Giddens (1991), Corsaro observes that children are from the outset embedded in social relations and networks (including those of peers) enabling them to discover and construct a meaningful existence. Thus, the social structural context in which children are embedded is important because it provides these relations and networks. However, children are not acted upon so much as they shape in their use of language and in social interaction the contours and structures of their everyday lives. The study of socialization, accordingly, demands close attention to children’s lifeworlds as well as social structural contexts (Eder, 1995). The dialectical approach is compatible with the work of European theorists including not only Bourdieu and Giddens (1984), but also Habermas (1979) and others and their concerns with language, ideology, and social reproduction.

Reflexivity

A reflexive analysis of language use, action, and social structure sees the interaction order and the institutional order of formal organizations as having complex interrelationships not adequately described in causal—or even reciprocally causal—terms. The interaction order is comprised of mechanisms of turn taking and other sequential organizations, which provide the resources for producing and understanding what is being said and done in concert (Zimmerman & Boden, 1991). As Goffman (1983) pointed out, the interaction order and its constituent devices are basic or primordial in the sense of underlying, preceding, being organized independently of any social structural context in which talk occurs. Further, it is invariant with respect to historical and cultural variation, while nonetheless being sensitive to it.

If the interaction order is primordial in this sense, conversation analysts have shown the implications in various ways. One implication is that the fundamental organization of conversational turn taking may be different in institutional as compared with ordinary settings. Thus, where in ordinary conversation turn size, turn content, and turn order are free to vary and are subject to local management, in settings such as courtrooms (Atkinson & Drew, 1979), the jury deliberation (Manzo, 1996), classrooms (McHoul, 1978; Mehan, 1979), psychological testing (Marlaire & Maynard, 1990; Maynard & Marlaire, 1992), news interviews (Clayman & Heritage, 2002), clinical settings (Heritage & Maynard, 2006), and the survey interview (Maynard, Houtkoop-Steenstra, Schaeffer, & Zouwen, 2002), this is not the case. Attorneys, teachers, newscasters, clinicians, or survey interviewers ask questions, and witnesses, students, interviewees, patients, or respondents must answer. From these elemental observations, a wide range of consequences follow in regard to how professionals, in collaboration with lay and other participants, organize such actions as accusing and denying in the courtroom, teaching, testing, and showing learning ability in classrooms and diagnostic clinics, being “neutral” and expertly informative in the news interview, eliciting talk about delicate and sensitive personal matters in the medicine, or achieving the “standardization” of social measurement in the survey interview.

Race and Gender as Categories-in-Interaction

A reflexive approach to language, action, and social structure means understanding how sequential organization and other aspects of the interaction order can be deployed in ways that are sensitive to the contingencies and relevancies of a society’s larger structures. In recent years, conversation analysts and discursive psychologists have devoted much attention to the interactional organization of social structural categories in everyday and institutional settings. Rather than look at correlations among category membership and social outcomes, or at the subject positions created by abstract discourses, this line of research focuses on how social categories are realized in concrete interactions, and with what consequences. Here, we review recent interactional research on two social categories that occupy a significant place in contemporary society: race and gender.

With respect to race, conversation analytic research (Whitehead, 2009; Whitehead & Lerner, 2009) has shown how racial categories are managed in everyday interaction. Because the invocation of race categories of which one is not a member could invite negative inferences about oneself—for example, that one is racist—participants use various strategies to deflect them when talking about race. In particular, they use formulations that refer to race in general, rather than a specific group; qualify racial references by implying or claiming that they are only relevant to the instance being discussed; and mention race as an afterthought about someone who has already been adequately described. Moreover, when using race to explain an event, speakers allude to it instead of mentioning it explicitly, leaving listeners to fill in the gaps through what is known in common about racial groups.

Other research, by Stokoe and Edwards (2007) and Buttny (1997) has examined how race categories are used in reports about absent third parties. For example, Stokoe and Edwards show how race is incorporated into complaints to mediators about troublesome neighbors, and how police and suspects deploy these racial categories in interrogations. Suspects use such categories to make counter-claims against alleged victims, while police use them to question suspects about the specifics of an offense. In these cases, participants use race and allegations of racism to amplify the egregiousness of a reported transgression. Such usage strengthens a speaker’s focal action, such as complaining or accusing. Finally, Stivers and Majid (2007) examined the relationship between race and talk in medical interactions involving physicians, children, and parents. After coding physicians’ practices for speaker selection—such as gaze and terms of address—they correlated these with socio-demographic variables, including race and educational attainment. With respect to race, they found that when a parent is black, or less educated and Latino, physicians are more likely to direct their questions to parents than children. In contrast with other correlation-based studies reviewed above, Stivers and Majid combine statistical analysis with a careful, systematic coding schema grounded in a conversation analytic examination of concrete, interactional practices.Footnote 7

As with race, recent research on gender-in-interaction has examined how participants orient to, enact, and reproduce gender categories and norms in everyday interaction (Speer & Stokoe, 2011b). In line with path-breaking studies by Garfinkel (1967) and West and Zimmerman (1987), this line of research understands gender as an ongoing accomplishment, rather than a socio-structural attribute that people simply possess. People actively bring their actions into line with prevailing gender norms and expectations. In so doing, they reproduce taken-for-granted, commonsense beliefs and assumptions about what is “natural” for males and females, and hold one another accountable for violating them. For example, Land and Kitzinger (2011) show how participants produce gender as an interactional phenomenon by positioning themselves as members of one particular category—e.g., woman, as opposed any of the others from among which they could have selected, such as mother, daughter, friend, employee/employer, colleague, patient, etc. In this way, gender can be made relevant to, and procedurally consequential for, the interaction underway (Kitzinger, 2005). Similarly, Speer and Parsons (2006) and Speer (2011) examine how transgender individuals who seek access to cross-sex hormones and ratification from their psychiatrists for sex reassignment surgery, provide evidence that other people perceive them as having attributes of the opposite sex. This evidence is inserted into the conversation by means of reported third-party compliments about their gender-relevant attributes (Speer, 2011, p. 157–158).

Other research has focused on the reproduction of commonsense gender norms and expectations. Stokoe (2008, 2011) shows how repair practices associated with word selection in everyday conversation—e.g., replacing the word “girl” with “woman” mid-sentence—is a pervasive way in which participants abide by and enforce gender norms in local interactions. In her analysis of police interrogations, Stokoe (2010) shows how suspects use gender to perform a particular action—denying an accusation—as well as how this trades on and reproduces normative beliefs about the proper way for males to treat females. Thus, male suspects construct themselves as particular kinds of men to refute allegations of criminal conduct. For example, in response to officers’ allegations, suspects will produce “category-based denials” in which they claim that they are “not the kind of men” who would hit women. Such denials partition men into two groups, those who hit women and those who do not, and locate suspects in the latter category.

Conclusion

Language is a primary medium of social behavior and, as such, deserves center stage in the panoply of social psychological topics. Indeed, other topics in social psychology, including exchange, bargaining, justice, socialization, deviance, health, ethnic relations, and collective behavior (to name a few) necessarily involve interactive speech processes, which makes language use perhaps the most basic of all social psychological phenomena. This is, we have argued, not so much because language is a vehicle of communication; rather, it is a resource for action and activity. One action humans sometimes perform is “communicating” information of various kinds, but this is one among many other activities, such as arguing, promising, requesting, apologizing, joking, and greeting.

Influenced by ordinary language philosophy, recognizing that words do not have stable dictionary or ostensive meanings, and that the “same” utterance has different interpretations according to its context of use, researchers oriented to language use wrestle with the basic question of how utterances perform or are mapped onto specifiable actions. Sociolinguistics and discourse analysis answer this question in one way by suggesting that some combination of linguistic and social rules link words and activities together. This answer comes close to the theoretical model provided by the speech act theory of Austin and Searle. Frame analysis also presumes some normative connection between utterances and actions, while giving freer rein to actors’ strategic calculations and decision making in regard to rule adherence as participants exhibit different stances in relation to the talk being produced. Finally, ethnomethodologists, conversation analysts, and discursive psychologists propose that in their ongoing conduct, participants themselves use rules in the service of performing various activities. Rules, therefore, are only one possible facet of the practices whereby actors order speech productions to accomplish and understand the active force of these utterances. This way of solving the “mapping problem” by attending to user practices is closer to Wittgenstein’s idea of language games.

Moreover, in the conversation analytic view, importance is attached to how actors combine their utterances in a sequenced fashion. That is, the sequential organization of talk-in-interaction is a “primordial site of social action,” which implies that this organization needs investigation and explication before the orderliness of conduct and action in institutional and other social structural arenas can be analyzed fully. This assertion implies a point of contact between conversation analysts and Goffman’s concern with the interaction order. Among sociolinguists, discourse analysts, and cognitive sociologists, however, the argument is that participants’ actions are not completely local in terms of either genesis or effect. It behooves the analyst to import the context or setting of talk ethnographically to analyze speech patterning and interactive order properly.

Overall, the understanding of spoken language has moved from the conduit metaphor to an “action” orientation (Heritage & Clayman, 2010), and ever more realms of language use related to social psychology are coming under the microscope and set an agenda for further study. Scholars (Hayashi et al., 2013; Kitzinger, 2012) are renewing the investigation of repair—or how participants correct errors in hearing, speaking, and understanding in achieving mutual understanding. Emotions and affect, which did not garner sociological attention until late in the twentieth century (Turner & Stets, 2006), are also now getting conversation analytic scrutiny (Peräkylä & Sorjonen, 2012; Ruusuvuori, 2012). Studies in clinics are deepening our understanding of doctor-patient communication not only in primary care medicine (Gill & Roberts, 2012; Heritage & Maynard, 2006; Stivers, 2007), but also in psychotherapy (Peräkylä, 2008) and other clinics, and in circumstances involving aphasia (C. Goodwin, 2003), autism (Maynard, 2005), and other disabilities (Antaki & Wilkinson, 2012). Morality as expressed in and through talk-in-interaction is a burgeoning area of research (Rawls, 2010; Stivers, Mondada, & Steensig, 2011; Turowetz & Maynard, 2010). As such studies attest, and as we noted at the outset of this chapter, language and its use are distinctively human and also complex. Nevertheless, recent and forthcoming developments demonstrate strongly that talk, text, and social interaction are eminently susceptible of scholarly investigation. Such investigation documents the orderliness and organization that inhere in language as participants use it in their everyday social and social psychological contexts.