When Swedes begin to learn German: from V2 to V2

Ute Bohnacker (1)

1 Skane University Hospital [Lund]

This article investigates verb placement, especially Verb second (V2), in post-puberty second language (L2) learners of two closely related Germanic V2 languages: Swedish and German. Håkansson, . (2002) have adduced data from first language (L1) Swedish-speaking learners of German in support of the claim that the syntactic property of V2 never transfers from the L1 to L2 interlanguage grammars. Regardless of L1, learners are said to follow a hypothe-sized universal developmental path of L2 German verb placement, where V2 is mastered very late (only after Object–Verb, OV, has been acquired), if ever. Explanations include the notion of Subject–Verb–Object (SVO) being a more basic, ‘canonical' word order (e.g. Clahsen and Muysken, 1986), so-called ‘vulnerability' of the C-domain (Platzack, 2001), and ‘processability', according to which SVX and Adv–SVX (i.e. V3) are easier to process (i.e. produce) than XVS (i.e. V2) (e.g. Pienemann, 1998). However, the empirical data comes exclusively from Swedes learning German as a third language, after substantial exposure to English. When these learners violate V2, syntactic transfer from English, a non-V2 language, cannot be ruled out. In order to control for this potential confound, I compare new oral production data from six adult Swedish learners of German, three with prior knowledge of English and three without. With an appropriate elicitation method, the informants can be shown to productively use non-subject-initial V2 in their German after four months of exposure, at a point when their interlanguage syntax elsewhere is non-targetlike (VO instead of OV). Informants who do not know English never violate V2 (0%), indicating transfer of V2-L1 syntax. Those with prior knowledge of English are less targetlike in their L3-German productions (45% V2 violations), indicating interference from non-V2 English. These results suggest that, contra Håkansson . (2002), learners do transfer the property of V2 from their L1, and that L2 knowledge of a non-V2 language (English) may obscure this V2 transfer. The findings also suggest that V2 is not difficult to acquire , and that V2 is not developmentally dependent on target headedness of the VP (German OV) having been acquired first.

Mots clés: Social Sciences & Humanities

Ute Bohnacker. When Swedes begin to learn German: from V2 to V2. Second Language Research, 2006, 22 (4), pp.443-486. ⟨10.1191/0267658306sr275oa⟩. ⟨hal-00572106⟩