Analysis of Written German data and outline of the empirical chapters

In the investigation of the participants’ command of written German we have used the diagnostic criteria we established in section 4.3 for the assessment of the main structural properties of German associated with the VP, IP, and CP levels respectively. We conducted qualitative and quantitative analyses of the data summarised in the following.^[1]

Developmental profiles. Based on the results obtained concerning their development of written German in the time covered by files 1-5, we established a developmental profile for each participant. Individual profiles are summarised in a schematic manner following the template provided in Table 4.8. Examples illus?trate the participants’ command of the respective property of German. Potential candidates for language mixing are provided in a separate line, shaded in grey.

Table 4.8: Template used for the sketch of participants’ German profiles.

Participants’ profiles are followed by a more in-depth discussion of their competence at the onset of the study and the progress they make in the course of the time span covered by the first five samples of the longitudinal study.

Word order (verb placement). Regarding word order, we were interested to establish whether participants correctly set the VP and IP headedness values, and whether they adhered to the target V2-constraint. Target-deviant patterns were scrutinised for a potential impact of borrowing from DGS. In addition to qualitative measures assessing verb placement in main and embedded clauses, we used a quantitative measure to determine the absolute and relative frequencies of the different word order patterns identified for main clauses. The results are summarised following the template in Table 4.9. V2 sequences were distinguished into non-subject-verb (XVS), subject-verb (SV), and non-subject-verb sequences without an overt subject (XVX) because this was deemed to contribute to a better assessment of word order variation that might be associated with the attainment of the V2 constraint. Note that SV(X) sequences are ambiguous concerning (a) V2, because they could be based on a grammar without the V2-constraint (English, Italian) and (b) VP headedness, unless the verb is followed by another constituent. XVX structures indicate that sentence-initial non-subjects are integrated into the sentence structure, while subject-drop indicates that the correct setting of the pro-drop parameter remains to be tackled. Because V3 structures typically result from the adjuntion of adverbials to the left periphery of the sentence, indicating that the V2 constraint is not yet acquired, we counted these separately.

Furthermore, we also counted V1, Vend and verbless sequences separately. As for V1 sequences, indicating that the topicalisation of non-subject or subject XPs to a position left of the verb is not yet mastered, we further differentiated these into verb-subject (VS) and verb-non-subject sequences (VX), because the former have been found to occur during reorganisation phases associated with the implementation of V2. Verb-final clauses are of particular interest in this study for two reasons. First, child L1 learners of German have been found to initially produce verb final clauses. Because German is an SOV language this is a fundamental step in their development of the target language. Verb final clauses could also be an indication of borrowing given that in DGS verbs always appear in sentence final position.

Finally, we also considered sequences with verb drop by counting instances of verb omission in obligatory contexts. We were interested to establish the frequency of verb drop to the extent that this phenomenon might be developmentally motivated (initial stages, optional constituents) or an effect of borrowing, bound to certain expressions in the other language and lexical deficits in the host language.

Table 4.9: Template used for the summary of results on main clause verb placement.

Because the overall frequency of embedded clauses was found to be extremely low, we decided to include the results of quantitative measures in the discussion section on an exceptional basis only.

Morphosyntax. At the level of morphosyntax we focused on verb inflection not only with a view to determining the participants’ command of the target inflection paradigm but also to establish whether processes related to the IP level were operative (notably, finiteness distinction, verb raising, subject-verb agreement). In a first step we distinguished target-like forms, errors and omissions. Target-like forms were broken down in accordance with the person/number forms identified in the data. Verb inflection errors were further distinguished into the erroneous use of infinitives (-fin) and other erroneous forms (x), including forms with a wrong inflection ending of the main verb and those errors that pertained to the non-finite part of periphrastic verb constructions. Omissions were broken down into verb drop and copula drop. This differentiation, although not always easy to establish, was done to obtain further insights into the frequency of copula drop, which is known to represent a common phenomenon during the initial stages of language acquisition. In addition, because in DGS knows no copula, a protracted copula drop could be interpreted as an effect of cross-linguistic influence. Table 4.10 illustrates the template used for the summary of the results obtained.

Table 4.10: Template used for the summary of results on verb inflection.*

* F=file, n=total clauses, V=verbs produced (total), I=AGR/TNS (verb inflection) errors, -fin=infinitives, x=other inflection errors, 0V=total verbless, 0 cop=copula drop

• [1] A preliminary presentation and discussion of the main results of the qualitative measures waspublished in Plaza-Pust (2008b). To put these results into perspective, we decided to carry outadditional quantitative measures for the present study.