4Priorities in discourse analysis
John Sinclair
From a linguistic perspective, the original discourse analysis work, revisited
in Chapter 1, was motivated by a wish to make a description of spoken
interaction, using the insights of the philosopher J.L.Austin (1962). Speech
act theory offered a functional theory of meaning. It also gave a partial
explanation of a class of descriptive problems in linguistics, namely those
which expose an inconsistency between the meaning given by a straightforward
description in terms of an established analytical framework, and a function
in discourse that requires an unconventional description. Austin’s notion of
‘illocutionary force’ was a powerful agent in reconceptualizing the way
language relates to the world.
It seemed, indeed, that the conventional meaning of an utterance was but
a stage in its interpretation; a preliminary statement of the organization of
the components drawn from general knowledge about language of this kind
to be found in grammars and dictionaries. When the utterance was viewed
in context, another set of criteria applied, building on the analysis-for-
meaning, and exhibiting the illocutionary force. So a statement like ‘It’s
getting late’ could acquire the status of a threat, a warning, a hint, a complaint,
etc., depending on how it was said and in what context. Its conventional
meaning was unaffected.
This argument suggested that there should be established a separate level
of language description, which used the output from the grammar and the
dictionary as input and which showed the relation between the utterances
and their function when deployed in discourse. This level was called the
level of discourse.
In suggesting a form of organization for the new level, the Hallidayan
model of a taxonomic hierarchy was adopted (Halliday 1961) and the level
of discourse was held to relate to the level of form as form did to the level
of phonology. The building blocks of discourse were the sentences and
clauses of the grammar, but they took on new values. In the same way that
the phoneme /s/ differs from the morpheme {s}, the sentence ‘I see’ differs
from the move ‘I see’.
The rank scale of act—move—exchange—(sequence)—transaction soon
concentrated on the exchange, much as grammar was concentrating on the
clause. Little was investigated above the exchange because it was recognized