Epistemic Boundedness and the Argument from the Poverty of the Stimulus in Phonology
Seminar Speaker, Charles Reiss, Concordia
Linguistics Frontiers Series

Much recent work either denies the relevance of Poverty of the Stimulus arguments in phonology, or else applies them to toy language learning experiments in a manner that is very different from the classic application to AUX inversion in syntax. I discuss the relationship between internalism and nativism, and shows that recent empiricist/anti-nativist reasoning in phonology is partly due to an implicit rejection of internalism.  Expanding on Halle's (1978) `Knowledge  Unlearned and Untaught' paper, I argue that simple, familiar phenomena like the phonology of the  English plural (*cats, dogs, bushes, etc.) *actually  support the classic argument from the Poverty of the Stimulus,  since the learned rule is not extractable directly from observation. The rule can only be learned  by a LAD that requires that rules refer to natural classes that are defined *intensionally.  *Because of rule ordering, the assimilation rule's triggers do not form a natural class *extensionally. * The talk also relates the argument from the PoS to Fodor's notion of* epistemic boundedness*.