Written German competence at the onset of the study

Muhammed’s L2 German learner grammar at the onset of the recording time can be described as a VP grammar. Grammatical processes such as subject-verb agreement or verb raising run vacuous because the relevant functional projections are not yet available.

Word order. Muhammed uses a diversity of word order patterns in file 1. Apart from SVX constructions (cf. (431)), he produces a range of V3 patterns that result from (a) the non-application of verb raising (to INFL) with the effect that sentence- internal adverbs appear between the subject and the verb (compare example (432)), (b) the adujuntion of an adverbial phrase in sentence-initial position as in example (433), or, (c) the application of the figure-ground principle which would hold in the equivalent DGS construction as in example (434). Further, there is one verb final sequence in this file (example (435)) which occurs with a main verb infinitive.

Word order and language contact. Some sequences in Muhammed’s file 1 represent candidates for language mixing. For example, the sequence in (434) could be categorised as a V3 sequence with a postverbal prepositional phrase. Yet there are two elements in this sequence that deserve further attention. First, there is the preposition mit (‘with’), which is erroneously chosen in the place of on (the boy is lying on the deer’s back). Secondly, there is the repetition of the reference to the deer. Why would the deer appear once at the beginning of the sequence, as if preposed to the clause, and then occur another time at the end of the clause, in a prepositional phrase indicating the location of the boy? At closer inspection, and considering DGS as potential source of what looks like the ground-figure order characteristic of that language, the sequence might be reinterpreted as a translation from an equivalent DGS sequence. Indeed, apart from the ground-figure order, the expression “liegen mit Hirsch” could be reinterpreted as a sequential translation of a meaning that would be simultaneously expressed in DGS by means of a complex classifier construction. Example (436) represents another candidate for borrowing from DGS. Not only do the constructions that would require the copula appear without a verb. The question answer pair is also reminiscent of question answer pairs used in DGS for narrative purposes.

Verb inflection. Examples (431)-(433) above show that Muhammed already produces some verb forms that are correctly inflected for person and number in file 1 (cf. also (437) below). However, apart from the forms geht (‘goes’), sagt (‘says’) and schaut (‘looks-at’), all other verbs produced in this file appear in their infinitive form (cf. (434)-(435) above, and (437) below), which suggests that inflection is not rule-based. Further, and unlike other participants, Muhammed does not produce constructions with the copula sein (‘be’) at the onset of the study.

Figure 4.5: Verb inflection errors and verb drop in Muhammed’s file 1.

Figure 4.5 provides an overview of the errors produced by Muhammed in this file in the area of verb inflection, including the relative frequency of verb drop (copula drop making up 70% of the instances of the percentage of 33.3 of constructions with verb drop). As we can see, the erroneous use of infinitive forms outnumbers the relative frequency of verb drop. Erroneous verb endings make up only a small percentage of the overall error rate (5% of 60% verb inflection errors).

Finally, note that Muhammed already produces sequences with verbs that take clausal complements, such as wunschen (‘wish’) (437), but does not yet master the target selective properties of these verbs (in the case of (437) target German would require an infinitive clause). Examples such as (437) show that Muhammed expresses complex meanings by combining different propositions paratactically, at a time when he lacks the necessary structural means that would allow him to integrate the related propositions syntactically. We might speculate further that Muhammed also draws on his knowledge of complex syntax in DGS (recall that he produces complex clauses in that language such as the one provided in (438), repeated here for convenience).