CP structures

The production of embedded clauses introduced by a complementer (compare examples (420) and (421) of an L1 and an L2 learner respectively) and target-like question formation (compare examples (422) and (423) of an L1 and an L2 learner respectively) reflect the expansion of the IP structure through the projection of the complementer phrase (CP). Following the asymmetry hypothesis of German sentence structure discussed in section 4.1.4, we assume that the headedness of the IP is head final in CP structures.

Variation in L1 German acquisition. Contrary to the generalised assumption that the acquisition of verb final word order for embedded clauses in children acquiring L1 German would be flawless, there is evidence of word order variation in productions of L1 learners. Some children, as for example ‘Benny’, have been found to produce a range of different constructions including V2, V1, and Vend structures (see examples (424)-(402) from Fritzenschaft et al. 1991). Four weeks after the production of examples (424)-(402), verbs appear in the target final position in all complementer introduced clauses produced by this learner (Fritzenschaft et al. 1991).

In some monolingual L1 German learners variation has also been observed regarding question formation. Gawlitzek-Maiwald and Tracy (2005: 288), for example,

remark on the production of interrogative clauses with finite verbs in final position at a time when the child adheres to the V2 constraint for main clauses (for further illustration compare the main clause V2 and interrogative verb final examples in (427) and (428) from Gawlitzek-Maiwald and Tracy 2005: 288). According to these authors, some learners appear to reserve specific structural formats for specific functions (in this case V2 for main clauses and Vend for wh-questions) (cf. Gawlitzek-Maiwald & Tracy 2005: 288). Interestingly, a few weeks later (at the ages 2;2-2;4) the authors observe the coexistence of both VE and V2 question formats (cf. (428)). Two months later, only V2 formats are produced.

Variation in L2 German acquisition. L2 German learners have been found to produce embedded clauses relatively early in their development. Typically, word order in early embedded clauses “reflects” main clause word order. In Plaza-Pust (2000: 249) we remarked that word order in embedded clauses produced by an L2 German Italian learner was target-deviant during the first two thirds of the recording time (cf. example (429)).

Similar to the variation observed in the implementation of V2, target-like verb placement in embedded clauses (430), emerging toward the end of the recording time considered in Plaza-Pust (2000), coexists with target-deviant word orders until the end of the recording time so that it remains unclear whether the target-like word order is ultimately implemented (Plaza-Pust 2000: 257). Interestingly, Plaza-Pust remarks that the introduction of target-like verb-final embedded clauses correlates with an increased productivity of syntactic complementisers such as dass (‘that’) and ob (‘if’), the availability of new adverbial conjunctions such as solange (‘as long’) and bevor (‘before’) and the differentiated use of relative pronouns. This evidence is argued (Plaza-Pust 2000: 274) to provide further support for the assumption of a dynamic relation between structural development and the expansion of the lexicon (notice that the implementation of FCs in child learner grammars has also been found to be followed by lexical spurts, Radford 1990).