Utterance processing

Work on how users understand meaning in pragmatics has traditionally focused on the inferences through which hearers attribute an intention to mean something on the part of the speaker. As we have already noted, Grice himself was fairly ambivalent about the status of speaker meaning (i.e. meaning intended by the speaker to be recognized by the hearer as intended) in communication. It was only in subsequent work on how users understand or process meaning that speaker meaning was assumed to be synonymous with communicated meaning. This is part of the general move away from a philosophical, analytical framework to a cognitive, explanatory framework in approaching the issue of how users understand pragmatic meanings. An explanatory framework for instance, aims to ground the processing of pragmatic meaning in a more general model of cognitive processes. It thus approaches utterance processing primarily from the perspective of cognitive representations and how various cognitive processes give rise to them.

Relevance theory, which we briefly introduced, is the most widely utilized explanatory framework to have emerged to date, as it focuses for the most part on how hearers reconstruct speaker meanings. It is generally assumed in this approach that the utterance is the basic unit of analysis. The cognitive and communicative aspects of utterance processing are linked through the notion of relevance, which is defined in terms of information (Sperber and Wilson 1995). Information itself is more specifically represented through particular beliefs, thoughts, desires, attitudes, intentions and so on. As we previously indicated, information has greater value for users when it is new as opposed to given, which in turn depends on, amongst other things, the common ground between the users in question. Furthermore, it is generally assumed that users will prefer, other things considered equal, information that is more easily processed, which in turn relates to issues of clarity, economy and processability. Relevance theory formalizes these as two forces that jointly determine the relevance of information from a cognitive perspective. This new distinction is defined in terms of contextual effects, while clarity, economy and processability all come under the scope of processing effort. Contextual effects are divided into three types in relevance theory: (a) those that produce new information, (b) those that confirm or strengthen an existing assumption, and (c) those that contradict or weaken an existing assumption. When discussing contextual effects that arise in an individual user’s mind, they are termed positive cognitive effects, which are contrasted with cognitive effects that do not provide any new information or do not strengthen/weaken existing assumptions for that user.

For example, if I take my car into a garage because it has developed a steering problem, a variety of positive cognitive effects might arise depending on what is said by the mechanic:

[a] The garage said the problem was serious. Thus, I realized that I’d be taking the bus home.

[b] I had assumed that the problem was with the steering rack. The garage said the problem was with the steering rack.

[c] I had assumed that the problem was with the steering rack. The garage said the problem was with the tires.

The cognitive effect described in [a] involves new information for the user, namely, the thought that I will be taking the bus home (as well as contradicting the prior assumption that I would be driving home in my car). The cognitive effect in [b], on the other hand, represents the strengthening of my prior belief that the steering problem is due to a problem with the steering rack. Finally, the cognitive effect outlined in [c] involves contradicting my prior belief that the steering problem is due to a problem with the steering rack, and new information, namely, the assumption that the problem with the steering is due to the tires. These are thus all instances of positive cognitive effects.

In relevance theory it is claimed that relevance is critical for both communication and cognitive processing more generally. The interplay of cognitive effects and processing effort is what determines relevance of information for an individual, which Wilson and Sperber (2004: 609) define in the following way:

The general idea is that the greater the cognitive effects, and the smaller the processing effort, the greater the relevance of information. In the case of cognitive processing, it is claimed that “Human cognition tends to be geared to the maximization of relevance”, or what is termed the Cognitive Principle of Relevance (Sperber and Wilson [1986]1995: 260). This amounts to the claim that users will pay attention to information that has greater relevance for them. In the case of communication, it is claimed that “Every act of ostensive communication conveys a presumption of its own optimal relevance”, or what is termed the Communicative Principle of Relevance (ibid.). Ostensive refers to the assumption on the part of users that an utterance (or behavior more generally) is intended by the speaker to mean something, while optimal relevance refers to the assumption that the utterance is “relevant enough for it to be worth the audience’s effort to process it” and is “the most relevant one compatible with communicator’s abilities and preferences” (ibid.: 270). This amounts to the claim that when speakers make it clear to the hearer they are communicating something through an utterance (or other kind of behavior), what they are communicating can be assumed by the hearer to be relevant to them. These assumptions underpin the mechanisms by which pragmatic meanings are processed by users.

Let us consider the following example where B’s response makes available a (particularized) implicature.

The question is, how does A work out what B means by his response here? From a Relevance theoretic perspective, A proceeds on the assumption that B’s utterance is optimally relevant to her. In this situation, B’s utterance provides ready access to both an explicit and implicit premise. The explicit premise arises from a pragmatically enriched understanding of B’s response (what is termed an explicature), namely, that the person they mutually know as Graeme has allocated a specified time to do things in that mutually known workplace in five minutes from the time of the speaker’s utterance. The implicit premise made available is the contextual assumption that “Graeme helps with computing problems” (which is triggered in part by the existential presupposition that Graeme has an office hour). By combining this explicit premise with this implicit premise, A can arrive at the implicit conclusion that B thinks Graeme can help A with her computing problem, and furthermore that he thinks A should ask Graeme for help rather than him (i.e. two related implicatures). This overall interpretation presumably satisfies A’s expectations of relevance.