Expansion of the VP structure

As mentioned previously, there is no evidence of the raising of main verbs to INFL until file 5. Before, phrasal verbs appear in their unanalysed form and adverbs occur preverbally as in examples (628)-(629), produced in file 3. However, periphrastic verb constructions with objects and adverbials inside the verb bracket appear 5 months earlier, that is, in file 4 (compare example (630)), providing evidence of a structural position above the VP. The assumption of the availability of the IP at this stage is further corroborated by the target-like placement of the negator after the copula in sequences like (631). Finally, main verb raising to INFL in file 5 is reflected in the target-like sentence final placement of separable prefixes of phrasal verbs, such as an (‘on’) and unter (‘down’) in examples (632) and (633) (note, though, that the verb does not correctly agree with the subject argument in example (632)).

A task to be tackled: V2 constraint. While we may safely conclude that Christa’s initial L2 German VP grammar has been expanded by an additional IP layer, the analysis of the data does not allow for a clear-cut conclusion concerning her acquisition of the target V2 constraint. Crucially, there is no evidence of target-like non-subject V2 constructions until the end of the recording time considered here. We noted previously on the predominance of V2 in Christa’s narratives. A more detailed analysis of the V2 sequences identified reveals that these follow the SVX pattern (cf. Figure 4.21 which provides an overview of verb placement in main clauses, with a more differentiated account of the different verb second patterns observed, cf. also Table D-6 in Appendix D). In the whole corpus we find no instance of an XVS sequence, and only one instance of a verb subject pattern in the last file. Incidentally, the general adherence to the SVX pattern and the rare production of V3 constructions also reflect the general absence of narrative specifications about temporal relations in Christa’s written productions. In the narratives of other participants, the adverb dann (‘then’), used to indicate the succession of narrative events, occurs fairly frequently and it often appears in V3 constructions. Christa, however, only produces two sequences with dann in the whole corpus, namely one in file 2 and one in file 5, both of them V3 constructions. Hence, we can only conclude that while she produces only few errors regarding verb placement, her written productions do not provide sufficient information about her attainment of the V2 constraint.

Figure 4.21: Main clause verb placement in Christa’s narratives.

Question formation. Regarding question formation, the interrogative sequence in example (634), produced in file 5 is target-like, but patterns with the title of the picture story elicited (which she also uses as a title at the beginning of the file 5 narrative). Interestingly, Christa already used the same question twice earlier, in file 2 (see examples (635) and (636)), once with the verb in the target-like position and another time with a target-deviant verb placement. These observations lead us to conclude that there is no sufficient evidence in Christa’s narratives to unambiguously establish whether the mechanisms for question formation are in place.

Subordination. Finally, regarding complex clauses, (637) is a remarkable example in that it involves target-like verb final placement in an embedded clause and the use of the complementiser bis (‘until’). As this is the only instance of an embedded clause with this word order, we can only speculate on the implementation of the head-final value of the IP and the availability of the CP layer at this stage.