We begin by reviewing data from Korean, Turkish, Italian and Spanish-speaking adults acquiring German without formal instruction. Our findings have shown that these learners transfer their L1 VPs: the Korean and Turkish speakers transfer a head-final VP and the Italian and Spanish speakers first transfer a head-initial VP and then switch its headedness to the correct, head-final value for German. Although functional projections in Korean and Turkish are head-final and in Italian and Spanish head-initial, all four groups of learners subsequently posit head-initial functional projections in German (which are not always target-like). We conclude that only lexical projections constitute the L2 learner's initial state; the development of functional projections is driven solely by the interaction of X'-Theory with the target-language input. We then discuss some studies on the acquisition of French by English speakers and of English by speakers of various L1s which purport to bring evidence to bear against our approach. Upon closer examination, the evidence turns out to offer further support for the position that the sole projections which the learner transfers from the L1 are lexical ones. Finally, we account for potentially problematic verb-raising data from French learners of English. Rather than taking the stance that French raising to Agr is transferred, we propose that L2 learners' identification of free morphemes as salient triggers leads to a misanalysis of verb raising in English. We also apply this idea to a reanalysis of the morpheme-order studies of the 1970s.

Vainikka and Young-Scholten (1994) propose an analysis of the acquisition of German by adult Korean and Turkish speakers based on the Weak Continuity account of L1 acquisition. They claim that L2 acquisition initially involves a bare VP whose (final) headedness is transferred from the learner's L1, with functional projections evolving entirely on the basis of the interaction of X'-Theory with the input. In this article, we extend this account to data from Italian and Spanish speakers learning German. Our analysis reveals that these learners initially posit a bare VP whose (initial) headedness is transferred from their native languages but, while still at the bare VP stage, they adopt the head-final VP of German. At this bare VP stage the morphological elements incompatible with the VP are not attested (e.g., auxiliary verbs, verbs marked for agreement and obligatory subjects). At the next stage of acquisition, similar to what Vainikka and Young-Scholten observed for the Korean and Turkish speakers, the Italian and Spanish speakers posit a head-initial functional projection. This projection further resembles the first functional projection observed in the acquisition of German by children (Clahsen, 1991) and involves optional verb-raising and the emergence of elements which typically appear in INFL (auxiliaries, modals) and in Spec (IP) (obligatory subjects).We conclude that child L1 learners and adult L2 learners build up syntactic structure in much the same manner and propose that the Weak Continuity approach accounts for all instances of syntactic acquisition.

In addition to adultlike nominative subjects, accusative and genitive subjects can be observed in the speech of young English-speaking children. In this article, two stages involving such oblique subjects are isolated; the two stages are separated by a stage characterized by adultlike nominative subjects. The proposed analysis relies on gradual, nonmaturational development of phrase structure. The early oblique subject stage is clearly attested in Nina's data, who until the age of 2 used oblique subjects without Infl elements; after her second birthday, nominative subjects and Infl elements co-occurred. The late oblique subject stage is attested in all four children studied. At a point when nominative subjects were used elsewhere, oblique subjects resurfaced in wh-questions.

No abstract

No abstract