L3 Workshop: Multilingual Language Acquisition, Processing and Use
Donnerstag, 10. Januar 2019
15.15 – 16.45 Uhr
Maria Teresa Guasti (Universität Milano Bicocca)
My claim is that language acquisition should be the starting point for building a theory about language complexity and language architecture. I will start from the observation that third person direct object clitics are omitted for a shorter period in Spanish, Romanian and Greek than in Catalan, French, Italian (Gavarró, Torrens and Wexler 2010). This cross-linguistic discrepancy revealed by child language comes about because clitics are different objects in the two sets of languages: affixes in the former set of languages and heads of DPs in the second set. Consequently, sentences including clitics undergo different derivations in the two sets of languages. In the former set, clitics are inserted in the clausal spine in a clitic phrase (CLP) (Sportiche 1996) and a null object is inserted in the base generated position. In the second set, clitics are inserted as DPs in the object position and undergo movement as DPs and as heads. I will claim that evidence for movement is the presence in these languages of the partitive clitic, which is absent in the other set of languages. Partitive clitics feature a split configuration, whereby the clitic and its remnant are in different positions (Io ne ho prese molte = I of+them have taken-fem-pl many). As a corollary, I will derive the fact that since clitics move in these languages, they will trigger past participle agreement. I will provide evidence from a survey of dialects spoken in Italy (Manzini & Savoia 2005) that languages with past participle agreement generally have the partitive clitic and I will provide evidence that this split accounts for some differences in the acquisition pattern of clitics (reflexives and dative clitics) within a language. In conclusion, clitics are more difficult to acquire in Catalan, French, Italian, because they are more complex. This case provides an example of how language acquisition data can shed light on the nature of linguistic objects and on the mechanisms involved in their combination in a syntactic structure.