Some notes on the ‘specificity effects’of optional resumptive pronouns

This paper discusses ‘specificity effects’ (Doron 1982) in the light of two recent approaches to resumption: Boeckx’s doubling analysis and Adger & Ramchand’s Agree-chain analysis. Boeckx analyses resumptive pronouns as functional heads encoding specificity; his approach cannot account for certain allowed ‘nonspecific’ functional readings (Sharvit 1999) nor for indirect object resumptive clitics, which lack specificity effects. Adger & Ramchand exploit unadorned individual variables and generally predict only specific/wide scope readings for movement relatives as well as for resumptive relatives. In general, any strict mapping between resumption and specificity fails to account for the fact that Hebrew gap relatives too allow for a specific interpretation; as a possible solution, I speculate that two different kinds of specificity may be relevant.