Further development

Expansion of the VP structure: coexistence of VP and IP structures

The first periphrastic verb constructions with a correct placement of objects inside the verb bracket appear in file 3. Before, Muhammed produces some constructions with modal verbs in file 2. However, as the object is placed after the verbal complex, these sequences remain ambiguous regarding the expansion of the VP by an additional structural projection, the IP. While the appearance of the complement after the periphrastic verb form in example (439) could result from a failure to correctly set the headedness of the VP, the word order in (440) (i.e. the placement of the negator after the modal verb, the object appearing after the lexical verb) suggests that “kann nicht + X” may be used as a formula at the time. In addition, we might speculate that the modal verb expression is borrowed from DGS and used as an unanalysed or idiomatic expression at a time when the necessary structure is not yet available in L2 German.

In this context, a note is due on the use of determiners. Although this is not a domain we will deal with in depth in this study, the determiners appearing examples (439) and (440) illustrate that Muhammed uses a combination of a definite and an indefinite article. This unusual combination is likely to result from vocabulary training measures, involving the learning of the determiner-noun pairs by rote. That both determiners are used at a time makes apparent, perhaps more so

than if the participant had learned to combine only one at a time, that the grammatical constraints determining their combination are not yet mastered.

The sentence-final placement of the finite verb in Muhammed’s first weil (‘because’)-introduced embedded clause (441) is target-like, but represents an exception as all other weil-clauses in subsequent files appear with main clause order.

Constructions with periphrastic verb forms. As of file 3, periphrastic verb constructions appear with a correct placement of objects inside the verb bracket (compare example (442), a construction with an auxiliary, and example (443), a sequence with a modal verb), providing evidence of a structural layer above the VP, the IP, and the target-like fixation of the VP-headedness parameter.

It must be noted, however, that the lexical verb in example (442) lacks the prefix ge-, which shows that participle formation is not fully mastered at the time (notice that some participles appear in their correct form, such as the participle of the verb schlafen (‘to sleep’) in example (444)). Subsequent recordings show that the task remains to be tackled by the end of the recording time. The sequence in (444) is also illustrative of another phenomenon, namely the target-deviant use of the auxiliary haben (‘to have’) in the place of the copula sein (‘to be’) in combination with an adjective (mUde ‘tired’, in this case). Note that the auxiliary would be target-like if it appeared only in combination with the participle geschlafen.

Example (445), in which the auxiliary appears with the adjective traurig (‘sad’)), documents a similar error in file 4.

Coexistence of VP and IP structures. We remarked previously that verb drop remains a constant and a frequent phenomenon in Muhammed’s narratives. As we can see in examples (446)-(449), which make up the passage describing the boy and the dog’s falling into the water, verb drop occurs in constructions that provide information about the locations of the main characters. To describe such spatial locations in target German would require the use of the copula or the existential main verb sich befinden (‘to be situated’). Notice that the concatenation of propositions does not involve connecting elements such as adverbials (e.g. dann, ‘then’) or conjunctions (e.g. weil, ‘because’, or und ‘and’), which Muhammed uses in other parts of the text. Consequently, the text passage serves more the purpose of picture description than of narrating connected events.

Main verbs vs. non-thematic verbs. From a structural perspective, it seems, however, Muhammed does not yet fully exploit the IP structure as main verb raising is not productive until the end of the recording time. Consider in this

respect the preverbal placement of the sentence-internal adverb in example (450) or the preverbal position the negator in file 4 in example (451).

Note, additionally, the use of da (‘there’) in sequences like (450). The use of this element in combination with the subject, as well as its placement after the referent and prior to the main verb derives a sequence that is reminiscent of referential establishment in DGS constructions (recall that the determiner det in DGS is often annotated as da). By assumption, this usage of da (serving the function of a referential marker) differs from the predicative function this element fulfils in other contexts, as illustrated in example (452). This sequence is also an example of how Muhammed expresses a complex meaning through a concatenation of propositions. Verb drop in these constructions reflects the continuing lack of structural and lexical means in his written German. Attributive constructions with the copula are not productive. Neither is the complementiser dass (‘that’) (note that dass would be used to introduce the embedded clause subcategorised by the psychological verb denken (‘to think’) of the main clause; alternatively, an unintroduced embedded clause with main clause verb placement could be used).

Complex sentential constructions. Muhammed occasionally produces complex constructions with embedded clauses introduced by the complementiser weil (‘because’) (see examples (453) and (454)). The verb appears in second position after the complementiser with the exception of example (454) in which hinaus (‘out’) appears preverbally. Notice that in target German hinaus functions as a separable prefix of the verb hinausklettern (‘climb out’), which implies that it appears preverbally in embedded clauses. However, because the verb appears in the infinitive form (we disregard the spelling error in this context, as the target form should correctly be klettern, ‘to climb’, which he produces correctly later in the narrative), and all other weil-clauses appear with V2 it seems likely that the

word order in (454) results from the use of an unanalysed form rather than from the correct raising of the verb to a head-final INFL.