View Our Catalog

Join Our E-Mail List

What's New

Sign Language Studies

American Annals of the Deaf

Press Home

Toward a Deaf Translation Norm

Previous Page

Next Page


Code-Based and Inference-Based Comprehension Models

There has been a gradual shift in understanding how people comprehend language. The message model (de Saussure, 1922, [1916] 1974) is an early model used to explain how interlocutors understand language. It was based on the following idea: thoughts are generated by a central thought process; these thoughts are encoded into a transmitted linguistic signal; this signal is perceived by a hearer, decoded, and the thought is received.

Authors have argued the linguistic code does not contain enough information for this simple message model to work (Sperber and Wilson 1995, 1986; Blakemore 1992, 1987; Grice 1975). The linguistic code underspecifies the information needed by the audience in order to fully understand what is being said. Hence Grice (1981, 1978, 1975) developed the concept of the cooperative principle; the idea that in conversation we aim to be a useful interlocutor because we have expectations about the rules of conversational communication. From this idea he developed nine maxims of conversation deemed to rule conversational interaction and first introduced the idea of relevance (Grice 1975).

His idea of relevance focused on the inference system and inference-based communication, relying not only on the surface form of a message but what underlying meaning there might be to that message. The surface form of the meaning is often called literal, and message meanings not following this surface-form meaning are taken to be nonliteral. A classic example of nonliteralness is irony, as when a speaker says, “Why don’t you take all day?” Within the context of criticizing someone for taking a long time, we know that to understand this utterance we need to derive more than its surface-form meaning. There could be a variety of clues, including the expectation that the task being undertaken by the addressee should not, literally, “take all day.” In this case the addressee is supposed to infer from the utterance something other than its literal meaning.

Although many theorists (Bach and Harnish 1979) used this inference-based model to augment the code-based model, Sperber and Wilson (1995, 1986) use Grice as a starting point to focus on inferential communication, examining relevance in “psychologically realistic terms.” Grice started examining identifying “what is said,” and RT started to address how “what is said” is identified (Carston 2004, 13).

Relevance Theory

This theory of pragmatics is grounded in information processing and cognitive theories of linguistic communication with its central tenet: “the aim of information processing is to recover as many contextual effects as possible for the least cost of processing” (Blakemore 1992, 34).

Human beings use certain “behaviour which makes manifest an intention to make something manifest–ostensive behaviour or simply ostension” (Sperber and Wilson 1995, 49). This is a general property of human interaction; the desire to point out information and to communicate this information has been intentionally pointed out. The hearer uses inference to understand there was an ostension; a coded communication such as language can be used to strengthen this ostensive-inferential communication, where ostensive-inferential communication can be defined as,

Ostensive-inferential communication: the communicator produces a stimulus which makes it mutually manifest to communicator and audience that the communicator intends, by means of this stimulus, to make manifest or more manifest to the audience a set of assumptions. (Sperber and Wilson 1995, 63)
Within linguistic communication, the processing effort the hearer makes to understand an utterance needs to be considered to be worth making. This effort is seen as worthwhile when overt communication is occurring; ideally in this situation, the speaker is deemed by the hearer to be optimally relevant. The speaker intends the hearer to believe she (the speaker) is being optimally relevant when she speaks, and two principles bind RT.[1]
Cognitive Principle of Relevance

Human cognition tends to be geared to the maximization of relevance.

(Wilson and Sperber 2002, 254)

Communicative Principle of Relevance

Every ostensive stimulus conveys a presumption of its own optimal relevance. (Wilson and Sperber 2002, 256)


1. This follows the convention adopted by Sperber and Wilson (1986) where the communicator is assumed to be female and the audience male.
Previous Page

Next Page