A program for experimental syntax: Finding the relationship between acceptability and grammatical knowlege

Abstract

There has always been interest in the methodology of acceptability judgment collection, as well as the reliability of the results. It seems, though, that the past several years have seen an increase in the number of studies employing formal experimental techniques for the collection of acceptability judgments, so much so that the term experimental syntax has come to be applied to the use of those techniques. The question this dissertation asks is whether the extent of the utility of experimental syntax is to find areas in which informal judgment collection was insufficient, or whether there is a complementary research program for experimental syntax that is more than just a methodological footnote to the informal judgment collection of theoretical syn- tax. This dissertation is a first attempt at a tentative <i>yes<\i>: the tools of experimental syntax can be used to explore the relationship between acceptability judgments and the form or nature of grammatical knowledge, not just the content of grammatical knowledge. This dissertation begins by identifying several recent claims about the nature of grammatical knowledge that have been made based upon hypotheses about the nature of acceptability judgments. Each chapter applies the tools of experimental syntax to those hypotheses in an attempt to refine our understanding of the relationship between acceptability and grammatical knowledge. The claims investigated include: that grammatical knowledge is gradient, that grammatical knowledge is sensitive to context effects, that the stability or instability of acceptability reflects underlying differences in grammatical knowledge, that processing effects affect acceptability, and that acceptability judgments have nothing further to contribute to debates over the number and nature of dependency forming operations. Using wh-movement and Island effects as the empirical basis of the research, the results of these studies suggest that the relationship between acceptability and grammatical knowledge is much more complicated than previously thought. The overarching conclusion is that there is a program for experimental syntax that is independent of simple data collection: only through the tools of experimental syntax can we achieve a better understanding of the nature of acceptability, and how it relates to the nature of grammatical knowledge.