“…But this is actually not the case in Hebrew: as shown in (1a) and (2a) above, a specific interpretation is also possible in gap relatives. 13 Conversely, if the specific D-layer is responsible 12 As an aside, note that such a weaker status of indirect object clitics is apparently at odds with some recent morphosyntactic analyses (Anagnostopoulou 2003, Adger & Harbour 2007, according to which third person indirect object clitics are more specified than direct object ones (in particular, they have a Person feature specification that direct object clitics lack.) 13 According to Boeckx, an alternative derivation satisfying the PUC involves an Agree relation between the two EPP-checking heads, an Agreeing C and AgrO.…”