Pragmatic Interpretation: Conversational Coherence

Pragmatic Interpretation: Conversational Coherence

Participants in conversation expect that utterances make sense at each moment in the interaction. This paper shows how meaning is constructed collaboratively and intersubjectively on the basis of two types of element: logical intelligible links between chunks of discourse provided by the interactants themselves and some contextual conversational resources, such as prior co-text, the concrete surrounding situation, or some background knowledge.

It follows that a definition of coherence will necessarily make reference to some recognisable property of relatedness or association among variable elements. In more concrete terms, it will be based on the assumption that each utterance is intended to mean something in the direction of achieving a goal in which the conversational actors mutually believe there is interest. In light of this assumption, Green regards coherence not only as a matter of textual properties but also as a matter of the ‘likelihood that the text’s audience will be able to make whatever inferences are necessary to relate the content of the individual sentences to each other in such a way that they support the inference of an orderly execution of an inferred plan to achieve an inferred objective’ (1996: 108). Conversationalists are thus mutually convinced that a ‘path to the goal will emerge from the interaction, given the cooperation of H, whom we expect to honour the spirit, if not the letter of our plans’ (McLaughlin 1984: 43, cf. Grice’s CP).

Edmonson (1981) explains the difference between cohesion and coherence. Cohesion describes the linguistic properties that are explicit in a text, such as repetition of key words, semantic associations, use of pronouns, ellipsis and other anaphoric expressions, the use of tense and aspect, that is, those devices whose role is to make it appear that the different utterances in a sequence are about the same referents or objects and which make ‘texture’ evident. Coherence, on the other hand, is a matter of the ‘interpretability’ of the semantic and interactional structure of a suprasentential stretch of language.

Consider the following example from Edmonson (1981: 6):

1. 
   A: John’s a good bloke. Is he coming tomorrow?
   

2. 
   A: John’s a good bloke. He is coming tomorrow.
   

appropriate because they do not perform any operation on the participants’ presumed knowledge. We can conclude with Edmonson that cohesion is ‘neither a necessary, nor a sufficient condition for coherence’ (1981: 5). In a coherent text we will find cohesion, i.e. identity of referents, but in a cohesive text we will not necessarily have coherence, i.e. the impression that a sequence of utterances constitute a thematic unit or whole.

As we can observe, the notion of coherence contains the ideas of ‘goal’/’objective’ and ‘plan’. A goal presupposes a plan, that is, a series of actions – the use of certain linguistic and conversational resources that will lead to its attainment. McLaughlin maintains that ‘much of the work in making a relevant contribution has to do with deducing the goals of the speaker, responding to those which one is intended to recognise, and integrating those responses with one’s own current plans and goals’ (1984: 41-2). An aspect that he finds critical in making a relevant contribution both to the immediately preceding utterance and to the ‘conversation-to-date’ is having a notion of the topic of that conversation.

In analysing the phenomenon of coherence in conversation, we can distinguish between functional organisation and topical organisation of discourse, which will be briefly described below.

2. Functional coherence

Functional organisation is concerned with the pragmatic aspects of language in use. McLaughlin (1984) sees the distinction between propositional or topical structure and pragmatic coherence in a way that is reminiscent of Austin’s distinction between what we are saying in an utterance (locutionary act) and what we are doing by it (illocutionary act), or between the organisation of ideas and the organisation of action.

To account for the linear or utterance-by-utterance level of functional organisation, McLaughlin uses Schegloff’s (1972) and Schegloff and Sacks’ (1973) notion of ‘conditional relevance’, according to which any utterance restricts the range of actions that subsequent turns may perform. Jefferson’s (1978) terms for this organisational aspect of discourse are ‘local occasioning’ and ‘sequential implicativeness’, and Goffman (1971) notes that the demand for conditional relevance seems to be taken for granted, since any utterance can be virtually interpretable as an appropriate reply if it occurs at the relevant point.

The smallest functional unit is the ‘adjacency pair’ (Schegloff and Sacks, 1973); other synonymous terms are the ‘couplet’, or ‘minimal dialogue unit’ (Goffman, 1976). Common examples of adjacency pairs are question-answer, accuse-deny/confess, request-grant/deny, etc. Goffman (1976) puts the generative mechanism for adjacency pairs down to the fundamental need in communication for the provision of feedback. This concept of action-reaction pair is important in discussing local functional organisation and hierarchical relations of pragmatic coherence.

McLaughlin (1984) draws on the considerable amount of literature at hand to speak of ‘pragmatic macro-structures’ in conversation, which correspond to the representation of a ‘global speech act’ or ‘macro-act’. He goes on to show that the ‘pragmatic topic’ refers to the perlocutionary effect of the global act, that is, to the speaker’s goal. He further specifies that one’s understanding of the actions (micro-acts) in the macro-structures and their importance will vary to the extent to which they are under ‘pragmatic focus’. This means that in terms of the role of the micro-acts in the speaker’s global plan, some will be main acts, others subordinate acts. It may happen that some pragmatic macro-structures do not have an explicit main act. In such cases, the micro-acts that constitute them will cumulatively amount to the value of a global act.

For illustration, we take the following dialogue, a familiar family scene:

2 B: What’s the time?

3 A: Dad’s leaving in a moment.

4 B: OK.

In other cases, main and subordinate actions are overtly manifest in the utterances, like in this dialogue, so typical of a teaching staff room:

2 B: How many copies?

3 A: Just two.

4 B: OK. There you go.

3. Topical coherence

The same idea that coherence is negotiated and jointly constructed in and through conversational interaction is formulated by other coherence theorists. For example, Bublitz (1988) defines topics as a combination of ‘topic subjects’, i.e. content, and the ‘topical actions’, which introduce, sustain, change or close them; Schegloff (1990) supports the idea that the fundamental building blocks in discourse are local sequential dependencies between actions and utterances; Gernsbacher and Givón (1995) maintain that coherence is not merely a matter of the conversational text itself, but a property which is gained through an emergent sense of directionality.

To strengthen this idea, the pragmatists’ view on topicality is different from that of the text grammarians in that the former consider not only the content but also the whole chain of related interactional events and actions, which Linell and Korolija (1997) term episodes. They take episodes to be ‘discursive events or action sequences, each delimited from prior and subsequent discourse, and internally bound together by something, e.g. a coherent topical trajectory and/or a common activity’ (1997: 167). In terms of minimal size, in their conception, episodes consist of ‘at least three turns, at least two of which must be substantial turns by different speakers’ (ibid.: 176, italics in original). By ‘substantial turns’ they mean turns expanded beyond being a minimal response. Episodes are structures that are situated above utterances and adjacent turns, but below whole speech events/activities.

Topicality (and, implicitly, coherence) is an incremental process developed on an utterance-to-utterance basis: ‘one “idea unit” makes another one possible and relevant, and once this new unit materializes, opportunities for further continuation, i.e. more idea units, are created’ (ibid.: 171). Thus, a topic emerges in the dialogue dynamics of responses and initiatives, being both the project and the product of coherence building (cf. Goffman 1983). Bergmann (1990) uses the terms ‘topic maintenance’ for the responsive actions and ‘topic progression’ for the initiatory steps. Topics can shade or drift into each other or they can be shifted or changed more or less abruptly.

Linell and Korolija (1997) make the point that although topic organisation is a property of discourse at the episode level, this is not an obligatory correlation. Their argument is that, generally, ‘topicality, coherence, and episode structure in discourse are activity-specific’ (1997: 171). For example, an activity type like gathering personal data about a client by a professional generates little local coherence beyond the question-answer pair but has global coherence derived from the ‘framing activity type, along with the macro-topical agenda’ (ibid.: 169). Quizzes and psychological tests are made up of topically unrelated questions, yet the activities are perceived as globally coherent by virtue of the activity-type context that compensates the lack of any relatedness between question-answer pairs.

Another aspect discussed by Linell and Korolija is topic maintenance as a dynamic process. This means that within episodes there are smoother or more radical topic transitions, which explains why the end of some episodes appears to be about a completely different topic from the initial one.

4. Contextual conversational resources

Linell and Korolija (1997) observe that the change of topic is rarely abrupt. In most of the cases, conversationalists initiate new episodes by resorting to ‘contextual resources, potential contexts that can be made into actual, relevant contexts through the activities of the interlocutors in dialogue’ (ibid.: 173). They insist on the potentiality of these contexts for being deployed as sources to the aforementioned effect.

According to Linell and Korolija, there are two main types of contextual resources: (i) immediate contextual sources and (ii) mediate (abstract) contextual sources. The former class comprises (ia) the prior (up to the simultaneously occurring) discourse in the encounter, i.e. co-text and (ib) talk-and-interaction situation, i.e. the immediate concrete environment (e.g. physical space, persons, objects, and extra-discursive events). The latter class includes (iia) discourse models (actors’ assumptions, beliefs, knowledge or understanding about the discourse in context, as well as models of their current and upcoming communicative projects); (iib) specific knowledge about people involved; (iic) the abstract situation definition, or the frame defining the encounter as an instance of a certain situated activity system or communicative genre and, related to this, the specific organisational context (in institutional interaction), and (iid) general background knowledge (about the world, of the language, communicative routines, and action types). (ibid.: 173-4, original emphasis).

They specify that, in general, ‘givenness’ is discussed in terms of the triplet (ia), (ib) and (iid) – the most relevant types of contextual sources, a classic triplet in the literature: Clark and Marshall’s (1982) ‘prior conversation’, ‘joint perceptual experience’, and ‘joint membership’, Givón’s (1989) parallel concepts of textual, deictic and generic contexts, and Givón’s (1995) ‘episodic model of the current text’, ‘model of the current speech-situation’, and ‘model of permanent generic-lexical knowledge’.

5. Conclusion

Coherence is manifested at the level of text (cohesion), in the structure of content (topical/propositional coherence), and in the organisation of actions (pragmatic/functional coherence). Cohesive devices are indicative of a topic space inhabited by the same ‘thematic participants’, but that is not always a guarantee of coherence as the latter emerges in the ‘collaborating minds’ of conversationalists (cf. Gernsbacher and Givón 1995: viii). Some coherence theorists treat propositional and pragmatic coherence separately (McLaughlin 1984), while others (Gernsbacher and Givón 1995; Linell and Korolija 1997) have a holistic view on coherence and regard it as a combination of content and actions.

Coherence is comprehended in terms of the communicative goals and plans of the conversational actors, who join interpretative efforts to help each other in realising them. It follows that actors actively organise the world through the processes of ‘collaboratively and intersubjectively constructed cognition’, for ‘if we were not able to do such framing, we would be lost in a murk of chaotic experience and probably would not have survived as a species in any case’ (Bruner 1990: 56, original emphasis).

Bibliography:

Austin J. L. 1962. How to Do Things with Words. Oxford: Clarendon.

Bergmann, J. 1990. “On the local sensitivity in conversation.” In I. Markova and K. Foppa (eds.) The Dynamics of Dialogue. Hemel Hemstead: Harvester Wheatsheaf, pp. 201-26.

Bruner, J. 1990. Acts of Meaning. Cambridge, Mass.: Harvard University Press.

Bublitz, W. 1988. Supportive Fellow-Speakers and Cooperative Conversations. Amsterdam: John Benjamins.

Clark, H. H., C. R. Marshall 1982. “Definite reference and mutual knowledge.” In A. Joshi, B. Webber, and I. Sag (eds.) Elements of Discourse Understanding. Cambridge: Cambridge University Press, pp. 10-63.

Edmonson, W. 1981. Spoken Discourse: A Model for Analysis. London and New York: Longman.

Gernsbacher, M. A., T. Givón (eds.) 1995. Coherence in Spontaneous Text. Amsterdam: John Benjamins.

Givón, T. 1989. Mind, Code and Context. Hillsdale, NJ: Lawrence Erlbaum.

Givón, T. 1995. Functionalism and Grammar. Amsterdam: John Benjamins.

Goffman, E. 1971 Relations in Public. New York: Basic Books.

Goffman, E. 1976. “Replies and responses.” Language in Society, 5: 257-313.

Goffman, E. 1983. “Felicity’s Condition.” American Journal of Sociology, 89: 1-53.

Green, G. 1996. Pragmatics and Natural Language Understanding. New Jersey: Lawrence Erlbaum.

Grice, H. P. 1975. “Logic and Conversation.” In P. Cole and J. L. Morgan (eds.) Syntax and Semantics, vol. III. New York: Academic Press, pp. 41-58.

Jefferson, G. 1978. “Sequential aspects of story-telling in conversation.” In J. N. Schenkein (ed.) Studies in the Organisation of Conversational Interaction. New York: Academic Press, pp.219-48.

Linell, P. and N. Korolija 1997. “Coherence in multi-party conversation.” In T. Givón (ed.) Conversation: Cognitive, Communicative and Social Perspectives. Amsterdam/Philadelphia: John Benjamins Publishing Company, pp. 167-205.

McLaughlin, M. L. 1984. Conversation – How Talk Is Organised. Beverly Hills, London, New Delhi: Sage Publications.

Schegloff E. A. 1972. “Notes on a conversational practice: formulating place.” In D. Sudnow (ed.) Studies in Social Interaction. New York: Free Press, pp. 75-119.

Schegloff, E. A. 1990. “On the organisation of sequences as a source of “coherence” in talk-in-interaction.” In B. Dourval (ed.) Conversation Organisation and Its Development. Norwood NJ: Ablex, pp. 51-77.

Schegloff, E. A., H. Sacks 1973. “Opening up closings.” Semiotica, 8 (4): 289-327.

Van Dijk, T. A. 1998. “The pragmatics of discourse.” In A. Kasher (ed.) Pragmatics – Critical Concepts. London: Routledge.