Retour vers liste

Détail de la contribution


The big DP hypothesis: Reconciling the internal and external syntax of cliticization

Abstract/Résumé: Clitic pronouns raise a number of questions for syntactic theory, among others: 1) the inventory of syntactic categories: do they realise the same category as determiners or a different category?, 2) the format of syntactic categories: do they realize a reduced structure with respect to the phrases they stand for?, 3) the types of movement available: do they involve an instance of X-movement?, 4) their peculiar behaviour in language acquisition and language disorders: why are they often omitted in these populations? To answer some of these questions, we focalize on the following Generalisations: a) clitics can realize VP-internal complements, but cannot realize VP-external complements (like temporal, causal, manner adjuncts). Clitics only exist in the case of complements that display case morphology. While NOM, ACC, DAT, GEN, LOC and INSTR case are attested in languages with a rich case morphology, no “temporal” or “causal” case are attested; b) the dependence built by clitics is local. Clitics undergo movement from VP-internal merge positions to their spell-out positions; c) clitics merge with functional heads belonging to the extended projections of verbs. They are not found inside DPs, APs, PPs at Spell-out. These generalisations can be captured by one single hypothesis, previously suggested to account for clitic doubling: the clitic and the associate DP are merged as a 'big' DP (Torrego 1995, Belletti 1999, 2005, Cecchetto 2000, Uriagereka 2005). The big-DP hypothesis can capture the parallelism between the availability of cliticization (Generalis. a) and the extraction possibilities (Generalis. b). Complements that are cliticized are also those that allow extraction of clitics. Big DPs occur in dedicated positions of the clause, not available in the extended projections of categories other than verbs (Generalis. c). This is related to the fact that clitic doubling is possible in clauses but not in DPs, APs, and PPs. The above mentioned proposals of clitic doubling however have the disadvantage of considering the clitic as the D head of the big DP. In spite of prima facie evidence (e.g. identical paradigms of clitics and determiners in languages like French), clitics are not determiners. Robust crosslinguistic evidence suggest that they realise different categories. While definite articles spell out copies of the phi-features of the noun (Giusti 2008), as confirmed by their realisations with e.g. prenominal adjective be-l, be-i, be-gli “nice + phi-features” (Cardinaletti, Giusti 2011), clitics have more structure: they realise (reduced) projections lacking the highest layer of nominal expressions (Cardinaletti, Starke 1999). If clitics are not Ds, a slight reformulation of the big DP hypothesis is therefore called for.