Oesterreichisches Forschungsinstitut fuer Artificial Intelligence (OeFAI)
Schottengasse 3, A-1010 Wien
Tel.: +43-1-53361120, Fax: +43-1-5336112-77, Email: sec@ai.univie.ac.at
-------------------------------------------------------------------------
VORTRAG
*******
Martin CORLEY, Ph.D.
University of Edinburgh
TOWARDS A SATISFACTORY MODEL OF LANGUAGE COMPREHENSION
Until recently, the dominant view in cognitive psychology has been that
the processing system endeavours to form the most complete and detailed
representation of external stimuli possible (Pylyshyn, 1984). However,
a growing body of evidence from work in visual cognition suggests that
processing may satisfy a viewer's current needs without resulting in a
full analysis.
This talk outlines an attempt to discover whether a similar principle
governs human language comprehension: our working hypothesis is that
humans do not necessarily fully process language, but simply understand
enough of the input to satisfy their current communicative needs. We
present evidence from current work showing the operation of such a
satisfaction principle, and discuss future attempts to work out how
it is parameterised.
Zeit: Montag, 22.Mai 2000, 18:30 Uhr pktl.
Ort: OeFAI, Schottengasse 3, 1010 Wien 1.
OESTERREICHISCHES FORSCHUNGSINSTITUT
FUER ARTIFICIAL INTELLIGENCE
o.Univ.-Prof.Dr.Robert Trappl