UMass Amherst
UMass Amherst | Library | Umail | Spire | People Finder 

Search

Match case Regex

Categories

Creative Commons License
This weblog is licensed under a Creative Commons License.

« UMass Amherst Linguists at NELS 38 | Main | Call: Syntax of the World's Languages III»

Charles Yang Colloquium

Charles Yang
Penn

Tipping Point

Friday, October 12, 3:30 pm, Machmer W-26

  • Why do we say "redness" but not "redth"? (We do say "wide-width".)
  • Why do we "donate money to charity" but not "donate charity money"? (We do "give charity money" as well as "give money to charity".)
  • Why did three classes of weak verbs in Old English collapse into only one, the regular class, by the time of Modern English?

Underneath these three classic puzzles (Aranoff 1976, Baker 1979, Kiparsky 1976) is the problem of linguistic productivity. Cast in terms of language acquisition, we ask, How does the child make just the right level of linguistic generalizations, which are almost always laden with exceptions? We approach the problem by developing a calculus of cost-benefit analysis, under which the cost of a potentially productive process and the exceptions to it can be measured and optimized. What emerges is the pattern of the Tipping Point, a place where the increasing cost of storing and processing more and more exceptions outweighs the benefits from the automatic application of a productive processes. Empirical case studies include the structure of the German lexicon, the U-shape learning in morphology, the acquisition of dative constructions, and the demise of the weak verb classes in Old English.