From Romance clitics to case Split accusativity and the person case constraint

We reconstruct the notion of dative in terms of a Q(⊆) category, denoting an ‘inclusion’ relation – to be understood roughly as ‘possession’. In the light of this, we reconsider interactions between dative and 1st/2nd person in Romance. In Italian 3rd person clitics have two separate lexicalizations for accusative and dative, 1st/ 2nd person clitics have a single dative-like lexicalization. We construe this phenomenon in terms of split accusativity (DOM), i.e. 1st /2nd person object clitics are embedded as datives, as opposed to 3rd person ones that alternate between dative and accusative. We also suggest a reworking of the Person Case Constraint (PCC) as a constraint on the interpretation of Q(⊆) (i.e. valuing of a Q(⊆) probe). Under it, a 1st/2nd person object clitic must be interpreted as the argument of Q(⊆) (i.e. value it) if present, leaving a dative clitic (i.e. bearing Q(⊆) morphology) without an interpretation.