Séminaire de Recherche en Linguistique

Inflectional morphology in Semitic and, more generally, in Afroasiatic, is encoded by both prefixes and suffixes. In Semitic, the perfect/past verbal forms have only suffixes while the imperfect/nonpast forms have both prefixes and suffixes. We try to explain this correlation, by building on the cartrographic hierarchies in the clause. We take PastP to be merged higher than FutureP (Cinque 1999), PersonP > NumberP > GenderP (Shlonsky 1989, Nevins 2002) and first person > second person > third person (Shlonsky 2000 a.o).

Our analysis of the apparent allomorphy in this system is embedded in a syntactic perspective on morphology (Koopman, Kayne, Collins & Kayne, a.o), assuming ‘one feature per head’ (Kayne ), early, as opposed to late insertion of terminals and syntactic constraints such as a generalized ‘multiply-filled Comp filter’, and an extension of Cinque’s analysis of movement within DP - (N(P) alone or two variants of pied-piping - to V(P) movement in the inflectional zone of the clause.

By grounding our analysis exclusively on independently-motivated syntactic principles and constraints, our analysis explores what happens if one eschews morphology-specific mechanisms such as fission and impoverishment.  More generally, our analysis shows what one would need in the syntax in the absence of a specific morphological component governing the order of insertion of morphemes and their form. To illustrate our approach, we provide an analysis of the the inflectional paradigms of Hebrew and Tamazight Berber.