Skip to content
2005 LSA Institute Linguistic Society of America









LSA.221 | Storage and Computation in the Mental Lexicon: An exemplar-based approach

Harald Baayen
TR 10:10-11:50
location: 32-124
course web site:

In linguistic theories of the lexicon, it is often assumed that storage has to be minimized and computation maximized. This makes sense from a descriptive point of view. Rules highlight the structure of the language; the mere listing of forms in a lexical list would be uninformative and uninteresting. By minimizing storage (listing) and maximizing computation (rule-based analysis), we may expect to obtain better insight into the regularities and subregularities in the lexicon.

The question addressed in this course is whether a theory of the lexicon that is as parsimonious as possible with respect to lexical storage is appropriate for understanding the mental lexicon. I will argue that parsimony of linguistic description does not map straightforwardly onto processing theories of the lexicon, and that we need a theory of the mental lexicon that incorporates a memory for morphologically complex words, irrespective of whether these complex words are regular or irregular. Such a memory should not be thought of as an unstructured list of unstructured words, but rather as a paradigmatically organized database of often highly structured exemplars, jointly driving probabilistic analogical generalization in comprehension and production.

Prerequisites: Participants in this course might consider reading in advance the short papers listed below, which illustrate the ongoing debate between dual route and single route theories of the lexicon. The exemplar-based approach proposed in my course holds a middle ground between these two more extreme positions.

Seidenberg, M. S. and Gonnerman, L. M. (2000), “Explaining derivational
morphology as the convergence of codes,” Trends in Cognitive Sciences 4, 353-361.

Pinker, Steven and Ullman, Michael (2002), “The past and future of the past
tense,” Trends in Cognitive Sciences 6, 456-462.

Pinker, Steven and Ullman, Michael (2002), “Combination and structure, not
gradedness, is the issue: Reply to McClelland and Patterson,” Trends in Cognitive Sciences , 6, 472-474.

McClelland, James L. and Patterson, Karalyn, (2002), “'Words or Rules' cannot exploit the regularity in exceptions: Reply to Pinker and Ullman,” Trends in Cognitive Sciences 2002 , 6, 464-465.