The biclausal structure of Haitian clefts

Haitian Creole Se: A Copula, a Pronoun, Both or Neither? On the Double Life of a Functional Head

“…As described in DeGraff (1992) and following, se appears, in fact, to be essentially incompatible with these predicates. (7) Jan bèl John goodlooking John is good looking (8) Jan nan lekòl la John at school Det John is at school Furthermore, while it is standard to assume that one of the central role of the copula in many languages is to serve as a support for tense morphemes, the Haitian Creole element se has been repeatedly observed to be in complementary distribution with the TMA markers and the negative particle of the language (Lumsden 1990, DeGraff 1992, Déprez & Vinet 1991,1992, 1997 Tu n'étais pas un bon professeur Ou (*se) pa te (*se) yon bon pwofesè 1 Clearly then, positing that se in (1) is some sort of copula requires minimally an account of its absence in these and other contexts. Pre-theoretical support for the view that the Haitian Creole element se is a pronominal form rather than a copula can also be found in commonly occurring sentences.…”

“…Déprez and Vinet 1997 analysis: 1.2 Basic assumptions of the analysis: Déprez and Vinet (1997) analysis is an attempt to link distinct uses of the element se, in predicative constructions such as (1), on the one hand, and in clefting constructions such as (13) on the other hand: (13) Se Jan Mari renmen Se John Mary likes It is John Mary likes In this respect, D&V's analysis is in the same vein as Lumsden's (1990) original proposal but it further attempts to provide an additional unification between se and the element ye that appears in sentence final position when certain predicates are fronted: (14) Se frè mwen, Jan ye Se brother my , John ye 4 A comparable debate is entertained in a interesting paper by M. Baptista (199 ) on the nature of the morpheme e in Capverdean Creole. She argues against the analysis of e as a pronoun and proposes to consider it as a functional projection generated in Agr o .…”

“…Under D& V's analysis, predicate restrictions on the distribution of se are taken to follow from general constraints (the HMC) on the movement of the predicate to the functional projection Pred 6 . Following Lumsden (1990), D&V assume that lexical predicates can move to occupy the head of the Pred projection as in (16) In such cases, predication is direct and se does not surface. Within the Minimalist framework, this movement could be seen as enforced by a strong feature in the Pred head.…”

“…(17) This structural approach to the distribution of se with distinct predicates makes an interesting prediction. In contrast to Lumsden's (1990) who takes the distributional restrictions of se to be a consequence of the categorial nature of the predicate, the presence/absence of se is here taken to result from structural constraints. The latter approach predicts that in principle, all types of predicates should be able to move to the Pred head position, provided they are not separated from this head by an intervening functional projection.…”

“…Recent research on the element se in Haitian Creole predicative constructions such as (1) seem to have reached opposite conclusions: (1) Jan se yon pwofesè Jan se one teacher John is a teacher Lumsden (1990) and Déprez and Vinet (1997) on the one hand, have independently argued that se is essentially a functional head in the verbal functional structure (IP and related projections) of Haitian Creole. The common intuition at the basis of their somewhat diverging views is that se is some sort of copula that serves as a link for predication in specific cases.…”