BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Towards empiricist models of language acquisition - Alexander Clar
 k (King's College London)
DTSTART:20140424T160000Z
DTEND:20140424T173000Z
UID:TALK50967@talks.cam.ac.uk
CONTACT:Alison Biggs
DESCRIPTION:Linguistic knowledge is on one view -- the empiricist view -- 
 based primarily on linguistic experience: language acquisition proceeds by
  general purpose inductive mechanisms of analogy and generalisation. From 
 this perspective the  innate biases of the learner have little or no lingu
 istically specific content.\nIn this talk we will describe a precise mathe
 matical theory of learning along these lines\; based on what is called dis
 tributional learning. These learning algorithms can now learn classes of g
 rammars that\, on our current understanding  are sufficiently powerful to 
 describe all aspects of natural language syntax.  We discuss three parts o
 f the theory: the theory of weak learning\, where we learn grammars that m
 erely generate the right set of strings\; strong learning\, where we recov
 er a grammar which generates an appropriate set of structural descriptions
 \, and probabilistic learning\, where we use indirect negative evidence to
  control overgeneralisation. \n\nTaken together we will argue that these p
 rovide a plausible and well articulated theory of language acquisition bas
 ed purely on distributional evidence\, and without any problematic appeals
  to semantic bootstrapping.
LOCATION:Faculty of Law\, Room LG19
END:VEVENT
END:VCALENDAR