Linguistic encoding is seen as playing a necessary but not solely sufficien
t role in speaker meaning by philosophers of Ordinary Language, such as Gri
ce and Strawson. Despite well-rehearsed problems with some of Grice's and S
trawson's specific theories, this general model has much to recommend it to
present day linguistics. Recent accounts have tended either overtly to den
y the existence of a code, such as those offered within the framework of in
tegrationism, or radically to limit its contribution to speaker meaning. Ac
counts of this latter type tend to dwell on the fact that the linguistic co
de cannot explain all aspects of the meaning of an utterance in context, an
d therefore to deny that encoded meaning can be propositional. Defining a p
roposition as a set of conditions for truth, however, it is possible to mai
ntain that encoded meaning determines a proposition, expressed by a sentenc
e, which is complete in itself, but radically underspecified with respect t
o the proposition expressed by any utterance of that sentence in context. S
uch an approach can offer a way of addressing the rather ambiguous place af
forded to the linguistic code in Levinson's account of pragmatic intrusion,
and Sperber and Wilson's relevance theory. (C) 2001 Elsevier Science B.V.
All rights reserved.