BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Towards a Stochastic Model of Linguistic Competence - Shalom Lappi
 n - King's College\, London
DTSTART:20101126T120000Z
DTEND:20101126T130000Z
UID:TALK27983@talks.cam.ac.uk
CONTACT:Thomas Lippincott
DESCRIPTION:In recent years computational linguists\, psycholinguists\, an
 d even some theoretical linguists have adoped a probabilistic view of ling
 uistic knowledge. The primary motivation for this approach is a concern to
  incorporate the gradient effects and soft\, defeasible constraints eviden
 t in speakers' variable judgements on acceptability into the theory of lin
 guistic competence. On this view knowledge of a language is identified dir
 ectly with a language model and the probability distribution over the stri
 ngs of a language that it specifies. I will take up some of the problems i
 nvolved in developing a viable stochastic representation of competence and
  suggest possible solutions to these problems. I will also look at the con
 nections between probabilistic theories of learning and a stochastic model
  of grammar. Finally\, I will consider several consequences that such a mo
 del has for the competence-performance distinction.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR