магистрант Сатыбалдиева И.П.

Региональный социально-инновационный университет

Discourse Markers in Relevance Theory (RT)

The framework of Relevance Theory has produced quite a large amount of literature on discourse markers since it provides an interesting pragmatic model for their analysis.

Relevance Theory (RT) builds from Grice’s ideas that speakers make use of principles or maxims in conversations. That is a theory of pragmatics based on cognitive principles. In this cognitive model, the effects of new information are worked out by the hearers against the background of existing assumptions. Whilst a cognitive environment is shared by all members of a speech community, the work left for the hearers is to choose a context for an utterance so as to make the correct inferences about the speaker’s meaning.

The choice of context is constrained by the principle of optimal relevance: ‘Every act of ostensive communication communicates the presumption of its own optimal relevance’. (1, p.158) The conversational aim to be relevant is assessed with the principle that dispensable cognitive effort should be avoided. Less cognitive effort means greater relevance.

Discourse connectives (RT terminology) do not have a unitary semantic account. According to Blakemore (2), some discourse connectives encode concepts, and others encode procedures. These two types of meaning echo Austin’s distinction between describing and indicating, and/or Grice’s distinction between saying and conventionally implicating. This distinction between conceptual and procedural meaning reflects a cognitive distinction between representation and computation.

A word with conceptual meaning contributes to the content of assertions. This type of word encodes elements of conceptual representations. The information on how these representations are to be used in inference is encoded by words with procedural meaning.

Most words encode concepts; for example teacher encodes the concept TEACHER, or rich encodes the concept RICH. An example of a discourse connective that encodes conceptual meaning would be the following:

What I think we need, you see is a room with a table, that is to say, a table which students could sit around. Blakemore (3, p. 328)

The encoded meaning of table could be recovered as a concept of something large for people to sit at, or the meaning could be recovered as the concept of something smaller. The function of the reformulation indicated by that is to say in the example helps the hearer to recover the first concept: a large item for students to sit at. The proposition explicitly communicated by the reformulation is a proposition about the conceptual representation that the hearer has to derive from the host utterance.

Procedural meaning cannot form part of the communicated message. The function of procedural meaning is to guide the hearer towards the interpretation of a message.

John is handsome, but he is always moaning.

But expresses that some aspect of the interpretation of the second utterance contradicts an implication derivable from the first. In the example, John has a good quality to be the speaker’s love interest, that is, he is handsome. The fact that he is always moaning contradicts the implication that he is a good candidate to be the speaker’s love interest. But is a discourse connective encoding procedures activating some kind of inference. The encoded meaning of discourse connectives of this type is not part of the communicated message. The communicated message is the result of the inferences triggered by these discourse connectives.

The function of discourse connectives is to guarantee that the correct context is selected at minimal processing cost. Discourse connectives are valuable means of constraining the interpretation of utterances whilst complying with the principle of relevance.

According to Blakemore (4), there are three ways in which the information of an utterance can be relevant: it may help to derive a contextual implication; it may reinforce an existing assumption; or it may introduce conflict to an existing assumption.

a. Max fell. David pushed him.

b. Pushing someone can cause the pushed item/person to fall.

(b) is a contextual implication derived from (a). Discourse connectives can introduce contextual implications as well:

A: You drink too much.

B: So (what)?

In the following examples the answer of speaker B is supposed to question the intended relevance of what speaker A has just uttered. In fact, B is asking a rhetorical question since speaker B has already drawn a conclusion regarding the intentions of speaker A's utterance. Thus, a proposition introduced by the discourse connective so has to be interpreted as a conclusion.

The information of an utterance can be relevant because it may strengthen an existing assumption as in the following example:

a. David isn’t here.

b. We shall have to cancel the meeting.

c. If David isn’t here, we shall have to cancel the meeting. (4, p.135)

The information in (a) is relevant to provide further evidence for the assumptions in (b) and (c).

Some discourse connectives are concerned with strengthening. For example, after all indicates that the proposition introduced by it is evidence for an assumption that has just been made accessible, as in the following example:

You need to buy those shoes. After all, they are gorgeous.

Other connectives that are concerned with strengthening according to Blakemore are besides, moreover, furthermore, also (when utterance-initial), and indeed.

The information in an utterance can be relevant because it may contradict an existing assumption as is the case in the following set of utterances:

a. If Barbara is in town, then David will be here.

b. Barbara is in town.

c. David isn’t here. (4, p.136)

The information in (c) is relevant because it contradicts and eliminates the existing assumption in (a) when (b) is spoken by the hearer.

However is an example of a discourse connective which introduce denials:

I am not busy. However, I cannot speak to you.

However indicates that the proposition it introduces, i.e., I cannot speak to you, is inconsistent with a proposition the speaker assumes the hearer has derived as a contextual implication from the first utterance. From the utterance I am not busy, the hearer would derive that the speaker can speak to him/her.

The study of discourse connectives in RT is extensive. However, most of the analyses so far have not been formalized in a syntactic or semantic manner. Moreover, this type of analysis is rather difficult to formalize. RT makes implicatures reliant both on individual memory organization and on indefinite principles of processing effort that possibly differ across speakers and hearers. Achieving general rules and predictions is problematic for a Relevance theoretical account.

If RT were formalized (there have been a few attempts so far), it is uncertain whether the outcome would be insightful in a general theory concerning a variety of issues such as how context affects the interpretation of presuppositions, or explicate why there is wide agreement on their interpretations.

The first clash with this framework and our approach is that Blakemore (2) argues that discourse markers do not have a function of markers of relationships or connections between units of discourse. In our analysis, we will establish that it is one of the crucial elements for the meaning of anyway, it relates two utterances or even a large group of them.

Bibliography:

1. Dan Sperber and Deidre Wilson. Relevance: Communication and Cognition. Blackwell, Oxford, 1986.

2. Diane Blakemore. Relevance and Linguistic Meaning. Cambridge University Press, Cambridge, 2002.

3. Diane Blakemore. Or- parenthicals, that is - parenthicals and the pragmatics of reformulation. Journal of Linguistics, 43:311-339, 2007.

4. Diane Blakemore. Understanding Utterances. An Introduction to Pragmatics. Blackwell, Oxford, 1992.