Both scholars and communicators operate on the premise that language is structured in an orderly fashion. An alternative view is that language is organized in a random fashion. Clearly, however, communicators treat language as tightly structured. A source of debate centers around whether the structure of language is innate in humans or is learned through socialization processes.
Noam Chomsky (1957, 1965) suggests that children are born with knowledge of a universal grammar (i.e., a set of principles that are common to all languages) that can be applied to any language. This fundamental knowledge of languages is an individual's linguistic competence. The ability to use a given language in a particular situation is an individual's linguistic performance. Scholars can use linguistic performance as a resource for inferring the character of linguistic competence.
Chomsky observed that while there are a fixed number of phonemes (i.e., meaningfully discriminated smallest units of sound) and morphemes (i.e., meaningfully discriminated blocks of phonemes), humans can construct an infinite number of sentences. This suggests that in learning a language, individuals are learning rules for producing sentences, rather than learning sentences themselves. Actual sentences are referred to as "surface structures." From them, linguists can infer deep structures (i.e., the basic rules of grammar that are part of a speaker's innate knowledge and are the same across languages). Chomsky, therefore, was concerned with inferring a theory of abstract sentence structure that could account for and generate grammatically correct sentences. In this sense, the grammar that Chomsky was seeking to describe was a generative one. Chomsky's grammar is also referred to as "transformational" because the few rules that operate to create an infinite number of sentences perform transformations on deep structure.
Some linguists and communication scholars examine the structure of suprasentential units. Their interest is to see how it is that a larger unit of discourse may be coherent and "hang together," or have cohesion. Karen Tracy (1985, p. 30) defines conversational coherence as "the fact that comments produced in conversation seem connected to each other in meaningful orderly ways." In this sense, conversational coherence is strongly related to topic organization. A "text" that is "coherent" is one in which topic is successfully managed. Different schools of thought take different positions with regard to how coherence works. For example, John Searle (1969, 1975) takes a speech act theory perspective. He sees conversation as being made up of series of speech acts that have rules for when and how a particular utterance is to be understood to perform a particular action. These rules provide for the recognizability of a particular turn as doing a particular action, and the have been called "felicity conditions." Michael Halliday and Ruqaiya Hasan (1976) describe different linguistic devices that provide for connection within discourse. Their focus is on written texts, but their work has been applied to oral communication as well. The cohesive devices they describe include, for example, references, conjunctions, ellipsis, substitution, and pronomial reference. Each of these requires connecting with something earlier in the text to be understandable and thus provides for the text to be seen as a coherent whole.
Conversation analysts have described many aspects of the structure of language as it is used to conduct the mundane business of everyday communication. Harvey Sacks (1984) observes that the structured character of everyday social life is a dominant feature. He suggests that one could examine almost any aspect of society to find its locally structured character. Conversation analysts examine tapes of naturally occurring conversations. At first, taped conversation was examined simply because it was easy to obtain. Subsequently, it was realized that because taped conversations can be replayed, they can be transcribed and examined in great detail. Thus, by examining audio and video recordings of naturally occurring interaction, conversation analysts have been able to discover a range of structural features of conversation that are part of the infrastructure through which a wide range of activities are conducted in and through conversation.
Some of the most important conversational structures are turn-taking, sequence organization, and repair.
Sacks, Emanuel Schegloff, and Gail Jefferson (1974) describe how turn-taking in conversation can be seen as a locally managed system. They point out that many different systematic processes for taking turns could be envisioned; for example, Starkey Duncan and Donald Fiske (1977) propose that turn-taking is based on the exchange of cues in talk. Sacks, Schegloff, and Jefferson describe the organization of turn-taking as communicators' organized solutions to the practical problem of who should talk when and for how long, and how speaker change should be arranged. They show how conversation proceeds in minimal units that could consist of a word, a clause, a phrase, or a sentence. These minimal units of talk (i.e., turn constructional units) are produced in such a way that it is hearable whether or not they are complete. One speaker produces such a unit. At its point of possible completion, a next speaker may take the conversational floor. If this does not happen, the current speaker could continue until the possible end of the next turn constructional unit, when there is another opportunity for speaker change to occur. These speaker change opportunities are called "transition relevance places." A speaker may bid for a longer turn at talk (i.e., one consisting of multiple turn constructional units) by starting his or her turn as a list (e.g., "First.… "). In this way, the interlocutor can project that the upcoming turn will consist of multiple items that are contained in multiple turn constructional units. This may also be done by projecting that there is a story to tell. Thus, turn-taking is locally managed, and it proceeds on a turn constructional unit by turn constructional unit basis.
A fundamental observation regarding turn-taking is that, ordinarily, one speaker speaks at a time, and speaker change recurs. While this is the canonical organization for turn-taking, variation can occur. For example, it is possible for overlap to occur if a speaker begins to take a turn when another person is at or near a point of possible turn completion. As Jefferson (1986) has shown, overlaps are often precisely placed, and they display one speaker's close monitoring of another's talk. It is also possible for a speaker to talk interruptively, coming in at a point that is not at or near a point of possible turn completion. Don Zimmerman and Candace West (1975) have suggested that interruption is practiced disproportionately by men interrupting women and therefore may be a way of enacting power. Schegloff (1987) calls this into question, pointing out that taking the floor via an interruption is tantamount to winning a battle but that the war is won only if the interruptive turn is actually taken up in subsequent talk. Nonetheless, other more subtle interpersonal activities—such as monopolizing the conversational floor or preventing the other speaker from taking an extended turn—can be accomplished by overlap and interruption.
Often, subsequent turns at talk are related to prior ones by a relationship that is stronger than just a serial relationship. That is, a next turn may be specifically "implicated" by a prior one, such that it can be heard to be officially "missing" if it does not occur. This strong relationship may be apparent in the way in which turns are structured. A strong relationship between turns is often referred to as "sequence organization." Schegloff and Sacks (1973) refer to pairs of turns of this kind as "adjacency pairs." Examples of adjacency pairs include summons/answer, question/answer, greeting/greeting, and compliment/response. According to Schegloff and Sacks (pp. 295-296), adjacency pairs have the following features:
- They are two utterances in length (i.e., a first "pair part" is followed by a second "pair part").
- The two utterances are adjacently placed (i.e., the second pair part must come right after the first pair part).
- Different speakers produce each utterance.
- First pair parts precede second pair parts.
- The pair type of which the first pair is a member is relevant to the kind of second pair part that is selected (e.g., if the first part is a question, the second pair part must be an answer and not a greeting).
If a second pair part is not produced after a first pair part is uttered, it will be heard as officially "missing." Delaying a second pair part, or postponing it through such items as "um" or "well," may be heard to foreshadow upcoming disagreement (Sacks, 1987). Thus, sequence organization is another environment in which subtle interpersonal dramas can be enacted.
Problems in talking together are generally resolved very soon after they happen. Schegloff, Jefferson, and Sacks (1977) describe how repairs can be made when problems occur. The first opportunity for a problem to be resolved is through "self-initiated" repair, where a speaker offers a correction or substitution to something that has been said (e.g., "She was giving me all the people that were gone this year, I mean this quarter, you know."). If the speaker does not correct the problem in or just after the turn in which the problem occurred, another speaker can initiate repair with a next turn repair initiator (i.e., a turn that can be heard to indicate a problem of some sort in the prior turn), thus putting the speaker of the problem in the position to remediate it. Schegloff, Jefferson, and Sacks (p. 368) provide the following example:
B: Well I'm working through the Amfat Corporation.
A: The who?
B: Amfat Corporation. It's a holding company.
Here, A produces a turn that calls into question, or initiates repair on, an item mentioned in B's first turn. Next B offers a redoing and explanation of the problematic item. If this type of repair does not occur, the second speaker can complete the repair, as the following example from Schegloff, Jefferson, and Sacks (p. 378) illustrates:
A: Listen to the pigeons.
B: Quail, I think.
Thus, the possibilities for repair include self-initiated self-repair, other-initiated self-repair, and other-initiated other-repair. The organization of repair is also a rich site for interpersonal activities, since it is a method through which intersubjectivity is negotiated. Repeated use of repairs may constitute an inability to understand one another, for instance (Ragan and Hopper, 1984).
Overall Structural Organization
Two general features of the overall structure of a conversation are how it is begun and how it is ended. Schegloff (1968, 1986) describes how conversations on the telephone are opened. He shows that although there is a canonical order of adjacency pairs enacted by interactants (e.g., summons/answer, identification/recognition, greeting/greeting, "how are you"/"how are you," reason for the call), these should not be thought of as the enactment of a script. The series of sequences can be deviated from at any point in the conversation. For example, the answer to the summons may be a "Hello" instead of an expected corporate identification. A caller may take this up. Problems in recognizing one another may occur. The response to "How are you?" may be "Oh, okay, I guess" and thus may indicate that there might be some trouble that could become elaborated. Both following and deviating from the canonical format may have implications for the relationship between the interactants.
Schegloff and Sacks (1973) describe a canonical way that telephone calls may be brought to a close. They show how the orderliness of methods for closing telephone calls (and, by implication, the orderliness of other features of conversation) represents the interactants' solutions to the technical problem of how to suspend the relevance of speaker exchange. That is, once a conversation has begun, speaker exchange recurs. Speakers are thus faced with the problem of working together to agree to suspend the relevance of exchanging turns. This is accomplished by producing a preterminal sequence in which it is possible for interlocutors to recognize that the conversation is possibly complete, but that previously unmentioned mentionables could be inserted. Thus, for example, after the apparent completion of some previous spate of talk, an "Okay" shows that the current speaker is not advancing the talk at this point and yields the floor to the other speaker. If the other speaker passes up this opportunity by producing a reciprocal "Okay," then the conversation may be treated as closed. This will be indicated by an exchange of "Goodbye." Thus, sequence organization can provide a structure for the closing of a conversation.
Because talk is organized on the basis of speakers taking minimal turn constructional units, with the potential for speaker exchange recurring after each next turn constructional unit, special work must be done if, for example, a speaker is to take the floor for an extended turn at talk to tell a story. In this case, the prospective storyteller produces a story preface in which he or she indicates that he or she may have something to tell. These turns vary with regard to how strongly they project an upcoming story. First pair part turns such as "You wanna hear a story my sister told me last night" (Sacks, 1974) or "You'll never guess what happened to me today" strongly project a story. They also make a forwarding response strongly relevant, putting the prospective recipient(s) in the position of aligning as recipients of an extended turn by saying "Okay" or "What?" Alternatively, a turn may simply announce some news, such as "Shawn ate lobster this afternoon." Recipients could treat this as the news and offer some assessment such as "Wow," or they could treat it as projecting a story, in which case they could solicit further talk about it with a turn such as "He did?" Thus, the temporary suspension of turn-by-turn talk for the taking of an extended turn to tell a story is interactively worked out at the possible beginning by interactants.
In the body of the story, recipients can become passive recipients, producing continuers such as "mm hm" or "uh huh" (which show them to be attending but do not shape the story by showing what they make of it). Alternatively, recipients may produce assessments, such as "Wow" or "How terrible" that show what they are making of the story-so-far. Turns of this kind are considered to be more "active" recipiency because they may influence the course of the storytelling, particularly if they display an understanding of the storytelling that is not in line with that of the teller. The most active kind of recipiency (i.e., the kind of turns that may have the most effect on how the telling unfolds) are first pair part turns that actively require the teller to take a particular kind of turn next. According to Jenny Mandelbaum (1989), a first pair part recipient turn, placed just at the point where the teller is apparently about to make fun of one of the recipients, can (by actively influencing what the teller says next) divert the storytelling from making fun of a recipient and convert it into a storytelling in which something good happened. Recipients of storytellings can show different kinds of understanding of the point that the storyteller is trying to make. When a recipient goes along with the point the teller appears to be making, he or she is aligning with the teller. The recipient is affiliating when he or she shows support for the action or perspective apparently espoused by the teller. While alignment and affiliation are not limited to storytelling, storytelling is an environment in which alignment and affiliation often become observable issues for interactants. The recipients' alignment and/or affiliation, or lack of them, are primary ways in which interpersonal work can be undertaken in the storytelling environment.
Collaboration between teller and recipients is also necessary at the possible end of the storytelling. When the teller produces turns that suggest that the storytelling is possibly complete, the recipient must show a realization of the possible completion in order for turn-by-turn talk to resume. Thus, conversational storytelling is interactively constructed throughout its course. This demonstrates that the normative structures for conversation that have been described above are just that—normative structures that are constructed and enacted by interactants. That is to say, they are not fixed rules, but rather, they are the communicators' organized solutions to the structural problems that are embodied in the activity of interacting.
Chomsky, Noam. (1957). Syntactic Structures. The Hague: Mouton.
Chomsky, Noam. (1965). Aspects of a Theory of Syntax. Cambridge, MA: MIT Press.
Duncan, Starkey, and Fiske, Donald W. (1977). Face to Face Interaction: Research, Methods, and Theory. Hillsdale, NJ: Lawrence Erlbaum.
Ellis, Donald. (1999). From Language to Communication, 2nd edition. Mahwah, NJ: Lawrence Erlbaum.
Fromkin, Victoria, and Rodman, Robert. (1993). An Introduction to Language, 5th edition. Fort Worth, TX: Harcourt Brace Jovanovich.
Goodwin, Charles. (1984). "Notes on Story Structure and the Organization of Participation." In Structures of Social Action: Studies in Conversation Analysis, eds. J. Maxwell Atkinson and John C. Heritage. Cambridge, Eng.: Cambridge University Press.
Halliday, Michael A. K., and Hasan, Ruqaiya. (1976). Cohesion in English. London: Longman.
Jefferson, Gail. (1986). "Notes on 'Latency' in Overlap Onset." Human Studies 9:153-184.
Mandelbaum, Jenny. (1989). "Interpersonal Activities in Conversational Storytelling." Western Journal of Speech Communication 53:114-126.
Ragan, Sandra, and Hopper, Robert. (1984). "Ways ToLeave Your Lover." Communication Quarterly 32:310-317.
Sacks, Harvey. (1974). "An Analysis of the Course of aJoke's Telling." In Explorations in the Ethnography of Speaking, eds. Richard Bauman and Joel Sherzer. Cambridge, Eng.: Cambridge University Press.
Sacks, Harvey. (1984). "Notes on Methodology." InStructures in Social Action: Studies in Conversation Analysis, eds. J. Maxwell Atkinson and John C. Heritage. Cambridge, Eng.: Cambridge University Press.
Sacks, Harvey. (1987). "On the Preference for Agreement and Contiguity in Sequences in Conversation." In Talk and Social Organization, eds. Graham Button and John Lee. Clevedon, Eng.: Multilingual Matters.
Sacks, Harvey; Schegloff, Emanuel A.; and Jefferson, Gail. (1974). "A Simplest Systematics for the Organization of Turn-Taking for Conversation." Language 50:696-735.
Schegloff, Emanuel A. (1968). "Sequencing in Conversational Openings." American Anthropologist 70:1075-1095.
Schegloff, Emanuel A. (1986). "The Routine as Achievement." Human Studies 9:111-152.
Schegloff, Emanuel A. (1987). "From Micro to Macro:Contexts and Other Connections." In The Micro-Macro Link, eds. Jeffrey Alexander, Bernhardt Giesen, Richard Münch, and Neil Smelser. Berkeley: University of California Press.
Schegloff, Emanuel A.; Jefferson, Gail; and Sacks, Harvey. (1977). "The Preference for Self-Correction in the Organization of Repair in Conversation." Language 53:361-382.
Schegloff, Emanuel A., and Sacks, Harvey. (1973). "Opening Up Closings." Semiotica 7:289-327.
Searle, John. (1969). Speech Acts. New York: Academic Press.
Searle, John. (1975). "Indirect Speech Acts." In Syntax and Semantics, Vol. 3: Speech Acts, eds. Peter Cole and Jerry L. Morgan. New York: Academic Press.
Tracy, Karen. (1985). "Conversational Coherence: ACognitively Grounded Rules Approach." In Sequence and Pattern in Communicative Behavior, eds. Richard Street and Joseph Capella. Baltimore, MD: Edward Arnold.
Zimmerman, Don, and West, Candace. (1975). "Sex Roles, Interruptions and Silences in Conversation." In Language and Sex: Difference and Dominance, eds. Barrie Thorne and Nancy Henley. Rowley, MA: Newbury House.