Last Tuesday, June 2, Jad Wehbe superbly and successfully defended his dissertation, entitled Plural Predication and Scope. Here is the abstract:
A central question in formal semantics concerns the division of labor between lexical semantics and the compositional system. Given a particular inference pattern, we can ask whether this inference is encoded in the lexicon or whether it arises from a particular syntactic interaction between different compositional ingredients. This question is especially important in the domain of plural predication, where prominent accounts attribute different types of phenomena to generalizations over the lexical meanings of predicates. This thesis addresses this question through three case-studies: (i) cumulative inferences, (ii) reciprocal alternations, and (iii) the contribution of together. I argue for a framework for plural predication in which these different phenomena result from scope interactions between a limited set of syntactic operators, including pluralization operators and other scope-taking elements. Thus, I argue for a more impoverished lexicon and a richer syntax in the domain of plural predication.
The scope-based account I argue for faces an immediate explanatory challenge: while lexical analyses can appeal to lexical stipulations to capture the distribution of readings in each construction, the scope-based account appears to overgenerate unattested readings. I argue that these lexical stipulations can be replaced by independently motivated grammatical constraints that cut across different constructions. More specifically, I propose that two such constraints play a key role in plural predication: (i) a generalized economy constraint, inspired by Fox’s (2000) scope economy, which rules out LFs when they can be proven to be equivalent to simpler LFs and (ii) a constraint that requires implicit arguments to take lowest scope (Fodor and Fodor, 1980). The result is a framework where the different ingredients of plural predication are in principle free to take scope in different positions, as long as they don’t violate these independently needed constraints.
Finally, the scope-based account I argue for has some broader architectural consequences. The first consequence concerns the nature of economy constraints in the grammar. I show that in order to predict the desired distribution of readings, the generalized economy constraint has to be evaluated relative to an encapsulated system which only has access to some of the information that is needed to compute truth-conditions. My account therefore provides a new domain where we see evidence for a modular deductive system (following Fox, 2000; Gajewski, 2002; Fox and Hackl, 2006, a.o.). The second consequence concerns universal constraints that determine what kinds of meanings can be lexicalized in a single lexical item. I explore the idea that there is an economy constraint on lexicalization, which optimizes the expressive power of the lexicon, relative to its size. This constraint essentially predicts that inferences which can be generated independently in the syntax can’t be encoded in the lexicon.
Congratulations, Jad!!
