This week’s syntax-semantics reading group meeting will take place Wednesday, November 2nd at 4pm in room 117. Bernhard will present joint work with Aron Hirsch, “Type disambiguation and logical strength”.
It has been widely argued that the grammar can allow for variability in the semantic type of certain elements. Partee (1987), for instance, proposed that nominals can be interpreted as entities (type e), predicates (type <e,t>), or quantifiers (type <et,t>), with general type-shifting mechanisms mapping between the different meanings. The possibility for type variability raises the question of how interpretation is constrained. One possibility, suggested by Partee and Rooth (1983), is that a given element is interpreted with the lowest type that can compose in the relevant syntactic environment. We, however, will observe cases where a higher type parse is not only available, but obligatory. The data will be based on modalized constructions involving alternatives, in particular focus association with only (Test 1) and questions (Test 2). We will suggest that, at least in these cases, a high type parse is preferred because it leads to a stronger meaning than would the competing low type parse.