A structural account of Conservativity

This paper explores an account of conservativity based on copy theory of movement, based on suggestions in Chierchia (1995), Fox (2002), Ludlow (2002) and Sportiche (2005). In this approach, conservativity is not a constraint on the lexicon, but rather just a by‐product of the syntax‐semantics interface. The reason for the (apparent) absence of non‐conservative determiners from natural languages is that in entering chain relations in the syntax, they would lead to quantificational clauses truth‐ conditionally equivalent to ones created by regular conservative determiners. When these clauses have non‐trivial meanings (i.e., non‐contradictory and non‐tautological), then the relevant non‐ conservative determiner’s meaning might very well exist, but we cannot be sure as we could obtain the same sentence meaning with some conservative function instead. In other cases, the clauses created would have trivial meanings, and this would be the reason for their absence (Gajewski 2002). I then discuss three problems for this idea: (i) DPs in subject position, (ii) late merge and (iii) raising constructions (Hallman 2012). In response, I propose that (1) DPs always move and (2) every movement operation triggers a triviality check. I will then show that with these assumptions one can account for all of the three challenges above.Keywords: syntax, semantics, generalized quantifier theory, conservativity, conservativity universal, copy theory of movement