Linguistic Colloquia Series
Friday, February 16, 2018
9:30 a.m.11:00 a.m.
Syntactic projection and distribution
In the modern period, there has been a tension between two ideas about the locus of syntactic relations. In the earliest syntactic models, syntactic relations were taken to hold between realized syntactic dependents. In subsequent elaborations of this tradition, a variety of relations were shifted onto abstract elements of the argument structures
associated with verbs and other predicates. Work over the past half century has shown the descriptive value of notions of participant roles and grammatical relations for the analysis of phenomena such as valence alternations. However, there have also been overextensions of these notions in analyses of patterns, such as those involving adjuncts, where the application of abstract models of argument structure and argument selection has been less fruitful. In part, the limitations of these models reflects the fact that they lack distributional information, which can be associated with dependents, and to which speakers are known to be sensitive. This talk outlines the basis for a contemporary synthesis, which aims to consolidate the insights of models based on dependents and argument structures, and suggests how some of these insights can be carried forward into the emerging big data era.