Abstract Deprez, Guella & Sleeman

Definiteness-based L1/L2 and specificity effects

In recent work on L2 acquisition of English determiners, Ionin, Ko & Wexler (2004a) (henceforth I,K,&W) argue that L2-English article choice by L1-article-less-language-speakers (Korean, Russian) fluctuates between the definiteness setting and the specificity setting of the Article Choice Parameter. Since the specificity setting is instantiated in neither the L1 nor the L2, I,K&W claim that the adult learners still have access to UG.

This paper presents the results of studies of Arabic learners of French L2, Dutch learners of Arabic L2, and American learners of Dutch L2 that focus on the pragmatic factors contributing to article choice, and more specifically on the roles of specificity and definiteness. We show that despite the presence of determiner systems based on a definiteness opposition in their L1 and L2, these learners’ choice of articles manifests a clear specificity bias in early L2 acquisition: overuse of the definite article is significantly tied to [+specific] contexts and overuse of the indefinite article, to [-specific] contexts. A possible interpretation of these surprising and interesting results is that they provide even stronger support for UG access than I,K&W’s own previous results: UG is accessed here not in the absence of L1 input, but preferably to L1 settings.

However, that learners with a definiteness-based L1 learning a definiteness-based L2 have recourse to the specificity setting in UG seems unexpected and the primarily pragmatic rather than syntactic nature of the relevant data suggest that there could be another interpretation for these results. Ko, Ionin & Wexler (2004b) (henceforth K,I&W) show that Korean learners of L2 English systematically overuse the with indefinites in [+partitive] contexts. Following Wexler (2003), K,I&W attribute this overuse to the lack of the Maximality (or uniqueness) presupposition associated with the definite article. If, following much current semantic literature, presuppositions are taken to be lexically triggered, then such results could be assumed to merely reflect an incomplete L2 lexical entry for the. In the same spirit, a possible interpretation of our (and, a fortiori, I,K&W’s) results is that the specificity bias observed is not due to recourse to UG, but to an incomplete acquisition of the lexical entries of the L2 articles. On this view, resort to specificity-tied-choices of determiners suggests that even for adult L2 learners, pragmatically based distinctions must be acquired. Following Sleeman (2004), we suggest that for adults, in contrast to children, pragmatic based principles are already in place, but the Pragmatics-Lexicon/Syntax interface needs to be acquired in L2. Our data suggest that the encoding of familiar vs unfamiliar information (i.e. one aspect of definiteness) by a(n) (in)definite article is acquired earlier by I,K&W’s, K,I&W’s, and our L2 learners than maximality or the encoding of specificity. This hypothesis gets support from other constructions.

References

Ionin, T., H. Ko & K. Wexler (2004a). “Article semantics in L2 acquisition: the role of specificity”. Language Acquisition 12.1, 3-69.

Ko, H., T. Ionin & K. Wexler (2004b). “Parallels between L1- and L2-acquisition of determiners: the role of partitivity”. Poster presented at the 29th Boston University Conference on Language Development.

Sleeman, P. (2004). “The acquisition of definiteness distinctions by L2 learners of French”. In Linguistics in the Netherlands 2004, ed. by L. Cornips & J. Doetjes, 158-168.

Wexler, K. (2003). “Maximal trouble”. Paper presented at CUNY Sentence Processing Conference, MIT.