Contact a Humanities Office or Academic unit.
Find your course outlines.

Marr’s razor and syntactic movement in grammar, parsing, and acquisition

Date: Tuesday November 26, 2019

Time: 11:30 am to 1:00 pm

Location: TSH 201

Guest Speaker: Dr. Jon Sprouse

Bio: Dr. Sprouse’s research focuses on experimental approaches to syntactic theory, and the relationship between syntactic theory and sentence processing. His ultimate goal is to construct a comprehensive theory of the mental representations of sentences and the cognitive processes that construct those representations. Given the complexity of this problem, he would also like to explore the power of various methodological tools for uncovering the nature of the language faculty. Dr. Sprouse’s primary focus has been acceptability judgment experiments, with recent expansions into corpus analyses, EEG, and fMRI.

Abstract: Syntactic theories are often evaluated using something like Occam’s razor – we syntacticians typically prefer the theory with the smallest set of grammatical operations. As such, the syntactic operation “movement” has been invoked to account for at least two classes of dependencies: A’-dependencies, which involve clause-edge positions; and A-constructions, which involve argument positions. In this talk, I’d like to explore a different metric of success that we can call Marr’s razor: syntactic theories should have consequences for other aspects of the theory of language, such as sentence processing and language acquisition. I will discuss three sets of studies, each using a different methodology, designed to ask whether we have evidence for syntactic movement in A’ and A-dependencies according to Marr’s razor. The first is a set of working memory studies designed to test the claim that constraints on syntactic movement can be reduced to limited working memory capacity. The second is a set of EEG studies designed to look for ERP indices of the parsing processes required by movement in both A’ and A-dependencies. The third is a set of Bayesian modeling studies using real child-directed input designed to test whether the existence of an innate movement operation (as in some syntactic theories) leads to more or less child-like behavior during acquisition. Taken together, the results of these studies suggest that there is good evidence for movement in A’-dependencies, but there is almost no evidence for movement in A-dependencies (according to these specific tests).