On the Nature of Dialogue
One of my continuing interests is the technical understanding of human dialogue - how it is structured, how people engage in dialogue, how dialogue is interpreted by people, how people choose and create their contribution to dialogue and the like. I am interested in structure, but more as well.
Dialogue communicates. After the details of an interaction have been forgotten, perhaps after two bilingual individuals have forgotten which language they used, communicated residues remain. Attitudes, obligations, sale prices, promises and far more are retained, wholly dependent on the words and other interaction used, but distinct and not simply structural.
So, What is the nature of dialogue?
One might hope that some direct extension of speech act theory would work, but the phenomena seem to be different, calling for a qualitative shift (perhaps in addition to changes in views of speech acts.) However, such hopes are not easily satisfied.
In the book (On) Searle on Conversation (Benjamins, 1992) Searle gives an essay on conversation, focused on " Could we get an account of conversations parallel to our account of speech acts?"(p.7) Searle makes two points:
1. If speech acts are the 'units' of conversational structure, then one can find few 'interesting' interactions between speech acts. (p.2)
2. Conversations seem to not be subject to constitutive rules.
He also "suggests that for a minimal understanding of conversation as a theoretical entity, it must be seen as an expression of shared intentionality " -- of a different sort than just the sum of the individual intentions of the participants.(p.4)
The book includes essays of a number of noted commentators, commenting on Searle's position, and ends with Searle's comments on their comments.
In his concluding comments he notes that "... the form of intentionality that is characteristic of the individual illocutionary act is quite different in structure from the form that is characteristic of the collective intentionality of entire conversations, or even portions of conversations." and later "... the conversation as a whole is characteristically not representational, (and so) it does not represent an additional state of affairs. So there need be no additional representational propositional content in the entire conversation beyond that of the individual speech acts of which it is composed. There is no additional level of meaning that goes with the conversation as opposed to the meaning of the individual speech acts." (p. 147.)
This pessimistic conclusion is perhaps understandable given the choice of study topics that led to it, but it seems mistaken. I am tempted to cite for contrast the possibly equally partial view associated with McCluhan, the "the medium is the message."
Other opinions in the literature are similarly pessimistic. Often conversation is simply labeled as a species of chaos.
In contrast, there are several approaches that are yielding their own sorts of results.
- The Discourse Representation Theory, originating in a team led by Hans Kamp, has yielded both its own line of progress and some spinoff alternatives.
- The Conversational Analysis (CA) movement in sociology represents a high sensitivity to the realities of human interaction. It is thriving both in making its own kind of progress and in influencing related work.
- Cognitive Linguistics is growing in activity and insight. While it is mostly focused on the sentence and below, there is published discourse work as well.
- Activity in Computer Science, especially Artificial Intellegence, is actively building models of communicating agents (including non human agents), and also studying very carefully how human interaction proceeds.
Recognizing all of this, the topic of dialogue is still poorly understood and poorly represented in theory, especially from a linguistic point of view. Unlike the situation for grammar, there are not many competing views of the nature of dialogue. Overall there is a shortage of specific ideas.
This idea itself is not widely accepted. For example, there is still widespread willingness to take either grammar or speech act theory as a model of dialogue, to assume that the key to understanding dialogue is finding its structure (rather, say, than how or what it communicates), and to believe that with enough study, the "meaning" of a dialogue will be found somewhere in the sequences of symbols and associated sensations.
Certainly there is room for fresh conceptions of human interaction. Ideas from a wide range, not just common culture, need to be examined. There are many potential sources, each with a mature tradition of practice and training. they include courtroom dialogue, diplomacy, labor negotiations, literature and literary criticism, and textual interpretation in religious traditions. Many non-western sources of ideas could be profitably brought to bear. Understanding of interaction using language is in a very preliminary stage, in need of inspiration.
Send email to Bill Mann: firstname.lastname@example.org