(Universidade
Autónoma de Lisboa)
Bickerton (1981)
claims that the features of NPs in creoles crucially hinge on the distinction
between specific and non-specific. Other distinctions have been introduced by
Givón (1984), where referential and non-referential NPs are set apart, or
Mufwene (1986) who argues for a individuated vs. non-individuated distinction.
Over the last two decades, these views have been refined and the structure of
NPs/DPs has become the subject of a great deal of influential studies (Abney,
Longobardi, Kayne, Chierchia, etc.), mainly focusing on Romance and Germanic
languages. Along with this tendency, the interpretation of Bare NPs (BNP) has
gained a prominent position on the syntax-semantics interface.
A complex DP in Santome (ST), the
language under discussion, has basically the following structure:
(1) Dôsu
inen mina se mu glavi ku ...
NUM DET
child DET POS ADJ REL
‘My
two handsome kids that…’
Our talk will focus
on the following interrelated topics of its DP structure:
i) What is the exact
status of determiner se? Se attaches to a wide range of
categories, for instance pronouns, numerals, temporal adverbs or verbs, and
behaves like a clitic. It’s anaphoric nature covers a definite and referential
domain (in the sense of Lyons 1999). Especially interesting is the highly
specific interaction between se, animacy and pluralization (inen).
This interaction is also present in other modules of ST grammar, for instance
fronting strategies.
ii) How to derive a
DP structure in ST? The structure in (1) poses a potential problem for the
well-explored hypothesis that definite determiners (English the, French le,
etc.) are in Dº (cf. Longobardi) and derive the full-blown structure, since a
(non-partitive) number node is at least superficially in a higher position. In
a language lacking number/gender morphology it becomes questionable what would
trigger movement. This issue is, of course, paralleled by the lack of
inflexional morphology and the absence of verb movement, in ST and other
Creoles (e.g. Roberts 1999). Déprez (1999) argues, in fact, that languages with
poor morphology favour bare NPs (although bare NPs are not an indication of
poor morphology by itself).
iii) How do we embed
BNPs in DP structure? Due to the lack of true definite determiners in ST, BNPs
fill in a wide range of semantic interpretations (generic, kind, collective,
definite...) which on their turn hinge upon noun typology, argument structure,
etc. We will seek to subsume this variation under a set of features with +/-
(or unspecified) values that should on their turn be reflected in the syntactic
structure. Semantic un(der)specification has the advantage that syntax only
need to project the relevant features and therefore better complies to economy
principles.