Meeting the Systematicity Challenge Challenge

Abstract

From Fodor and Pylyshyn's celebrated 1988 systematicity argument in favour of a language of thought (LOT ), a challenge to connectionist models arises in the form of a dilemma: either these models do not explain systematicity or they are implementations of LOT . From consideration of this challenge and of systematicity in domains other than language, defenders of connectionism have mounted a parallel systematicity argument against LOT which results in a new self-defeating dilemma, what I call here the systematicity challenge challenge (SCC): either LOT does not explain nonlinguistic systematicity, or it is in no better position than its rivals to explain any systematicity, even linguistic systematicity. In this paper, first, I critically examine the SCC and some considerations that seem to support it. Second, I offer a response to the SCC by: (1) showing that LOT was never meant to be a cognitive model restricted only to linguistic systematicity, and (2) formulating a new argument in favour of LOT from nonlinguistic systematicity. Third, I argue that there is a central assumption underlying the SCC and maintain that it is mistaken. I conclude that the classical systematicity challenge continues to be fully valid for linguistic and nonlinguistic domains. © 2012 Philosophy Documentation Center.