Dear all,
like David Keeble, I am interested in spreading activation models. The ones
I am familiar with (such as Stemberger, J. (1985) 'An interactive activation
model of language production' in Ellis, A. (ed) Progress in the Psychology
of Language, Lawrence Erlbaum) are concerned with lexical access and
retrieval, and offer a very nice explanation of certain types of speech
error. The models I know only attempt to account for lexical production and
reception (i.e. decoding) and assume context-independent processing at this
level - hence no inferencing. This assumption is supported by
psycholinguistic priming experiments on the context-independent processing
of, for example, ambiguous words - there are lots of references to work in
this area in Jean Aitchison's Words in the Mind.
I haven't seen these models applied to text interpretation, and it seems to
me a bit of a strange thing to try; although I don't know exactly what kind
of connections David is talking about, it sounds very much as if you would
need a model of inferencing in addition to a model of decoding processes
here.
Barbara MacMahon
This archive was generated by hypermail 2b29 : Wed Apr 19 2000 - 12:07:29 GMT