Inter-modal go-betweens: language borrowing and the unclear role of LBG
Candidates for language mixing at the time when learners are dealing with the implementation of the IP reflect learners’ pooling of resources.
Agreement: overgeneralisation of “auf”. Verb raising to INFL is tied to the feature checking (agreement, case-marking). As mentioned previously, the participants’ overt marking of subject-verb agreement varies throughout the recording time. With respect to the relation of the verb and its complement arguments, the data reveal that during the phase in which grammatical processes relating to the IP become available there is a remarkable increase of constructions with the preposition auf (‘on’). The diversity of constructions involving this preposition is illustrated in the sequences produced by Fuad in file 3, repeated here for convenience (685)-(687).
As we can see, auf is correctly used to case mark the object with verbs that subcategorise for this preposition (685). However, examples (686) and (687) suggest that auf serves a more general function, namely, the one of marking the relation between transitive verbs and their objects. It is interesting to note in this context that the use of auf in the sense outlined previously is also remarked upon in the study on written German skills of bilingually educated deaf students in Hamburg (cf. Schafke 2005: 273, Gunther et al. 2004: 241f.), which provides additional support for the assumption that patterns of mixing relate to the language systems bilingual children learn (Genesee 2002: 187). In other words, language mixing is not a random, but a systematic phenomenon.
By assumption, as we advanced in Plaza-Pust (2008b) three phenomena conspire in the use of auf as a free morpheme to express this grammatical relation, namely, (a) the borrowing of DGS PAM which is commonly translated as AUF (686), (b) the analysis of the morphological components of agreement verbs in DGS and subsequent translation into German through the use of the German case-marking preposition auf (687), and (c) the remaining gaps regarding the German case-marking and determiner system.
While a detailed discussion of the acquisition of the case and determiner system is beyond the scope of this study, it is worth mentioning that the data gathered show that this area, like the domain of inflectional morphology, remains to be mastered by the end of the recording time. Participants use articles, but errors in case and number indicate that the choice occurs randomly. It seems plausible to assume therefore that the use of auf to overtly express the relation of the verb with its complement is used to fill the gap regarding the target morphology. Having said this, however, we must also take up our previous observations in chapter 3, dedicated to the participants’ DGS competence, in which we remarked upon a generalised erroneous use of pam with a target-deviant word order (basically, pam seems to prompt the choice of an SVX format) (cf. section 188.8.131.52).
Hence, it seems, borrowing occurs in both directions as in either language the respective element is associated with properties of the other language. Given the advanced level of DGS of the participants in this study this is a somewhat surprising finding. While we cannot ultimately determine the origins of this phenomenon we might speculate on the influence of third factor, that is, the third code these bilingual learners are exposed to, namely, LBG. Could it be the case that auf (with the corresponding mouthed element) is used in the communication via this signed system as a hybrid element that eventually influences the status of both pam and auf in DGS and written German respectively? Unfortunately, we have to leave this question unanswered because we have no reliable data on the LGB input and output in the communication of the participants with their interlocutors. It is interesting to note, though, that the feedback obtained from some members of the teaching personnel basically confirms the use of this element as a case marker by teachers and students.
Determiners: “da”. In a similar vein, though less consistently, the adverb da is used with the function the determiner detexist (often notated as da) would fulfil in DGS (that is, the establishment and maintenance of reference). Consider, for example (688), produced by Muhammed in file 4, in which da appearing to the right of the subject mimics referential establishment as it would occur in DGS. In this case, too, it seems, an element of the host language is used to serve a function it would fulfil in the source language.
Copula drop. The range of variation produced during the reorganisation phase tied to the implementation of the IP includes verbless clauses which would require the use of the copula verb sein (‘to be’) in German, a phenomenon that was already remarked upon in our discussion of the learners L2 grammar at the VP stage. Recall our call for caution in that context owing to the observation that copula drop also occurs in the early productions of other learners of German. Presently, as copula drop continues to occur at a time when more advanced structures are available we are confronted once again with the task of determining the origin of this persistent phenomenon.
Summarising, our analysis of copula drop at this stage reveals that this phenomenon typically occurs (a) in clauses with the adverbial da (‘there’) or prepositional phrases and (b) in predicative constructions. At closer inspection the following scenarios become apparent:
Copula drop in predicative constructions. One participant, Christa, provides no evidence of a productive use of the copula in predicative constructions. Christa uses the suppletive form ist (‘is’) as of file 2 in combinations with da (‘there’), das (‘that’), wer (‘who’), and, in file 4, with the expletive es (‘it’) (see example (689)). However, she consistently drops the copula with predicative adjectives (compare (690)).
Copula drop after overgeneralisation. Another participant, Simon, initially uses the copula in a range of target-like contexts, including predicative constructions, but also in combination with main verb infinitives. The alternation of sequences with and without a copula involving the same items as illustrated in (691) and (692) occurs as of file 3, in which the rigid SVX sentential pattern is given up and the incidence of verbless clauses increases.
Alternation of copula drop and target-like copula sequences (copula drop after target-like use). The drop of the copula alternates with the target-like use and is restricted to certain contexts, in particular, constructions involving (nicht) da or weg (‘(not) there’; ‘gone’), as is the case in the narratives of Fuad (693) and Hamida (694).
Clearly, what these scenarios indicate is that the drop of the copula at the VP stage, which is ambiguous regarding a potential influence from DGS given the lack of functional elements in the learner grammars at that stage, needs to be distinguished from the persistent drop of the copula at later stages. Participants produce these verbless clauses and target-like constructions in which they demonstrate their knowledge of a variety of contexts the copula is used in German. Thus, copula drop at this stage is indicative of a coexistence of diverse grammatical options that might be reinforced by the grammatical properties of DGS which lacks copula verbs. As pointed out by Tracy (2000: 25), for those errors that are also attested in monolingual acquisition of German the question arises as to whether bilingual children might take more time in “correcting misanalyses”, especially in the case where the other language reinforces the erroneous hypothesis (cf. also Muller 1998).
Lexical borrowing. Participants produce a series of verbless clauses containing expressions like Angst (‘fear’) (695) or bescheid (‘information’) (677) which are indicative of language mixing at the lexical level: both languages include lexical elements to express ‘to be frightened’ or ‘to let sb. know’, but the lexical overlap is only partial as German, unlike DGS, does not have a verb to express the meanings, but uses periphrastic verb-noun combinations instead (i.e. “Angst haben”, “Bescheid geben”). The use of “Angst” or “bescheid” as predicates in clauses like (695)-(677) is thus indicative of the borrowing of these expressions from DGS and the lack of the target idiomatic expressions.
It is interesting to note that this type of lexical borrowing is also observed in the narratives analysed by Schafke (2005: 271) and Gunther et al. (2004: 240f.); compare the following example (697) of a participant in their study, Thomas, who also draws on DGS. The example is remarkable in that “Bescheid” appears with the infinitive marker -en and is combined with the preposition “auf” (example from Gunther et al. 2004: 240).
Code-switching. By assumption, learners also resort to a pragmatically driven type of mixing which would be reflected in the use of DGS-like constructions for narrative purposes. (698), produced by Fuad, seems to involve the type of role shift characteristic of storytelling in a sign language like DGS in that it mimics the thoughts of the story character. However, the non-manual components used to signal the change of perspective in DGS (e.g. eye gaze, body shift) are not “translated”.
Individual variation in the implementation of the IP
Summarising, the variation observed in the transition from the VP to the IP grammar involves the coexistence of alternative structural patterns that are indicative of a reorganisation of the learner grammars. A similar variation was not observed in the case of Maria: Instead, the analysis of the data suggests that the IP is already established in her learner grammar at the onset of the recording. Whether or not her previous development involved a similar transition stage cannot be decided here. At the other end of the spectrum of individual variation, we are confronted with the written productions of Simon in which we find no evidence of variation along the lines described previously: this learner does not produce inflected verb forms, neither does he use periphrastic verb constructions, and adverbs and the negator appear in the preverbal position in the narratives produced toward the end of the recording time.
-  As we remarked upon in Plaza-Pust (2008), auf serving the function of an overt case markeris reminiscent of the function served by the preposition of in English (compare "Poirot is enviousof Miss Marple” in which the preposition assigns accusative case to Miss Marple, cf. Haegeman1994: 173). Moreover, as auf is available in German, learners are easily tempted to overtly markgrammatical relations at this stage which is in line with the insights gathered in other acquisitionsituations in which learners temporarily make these relations transparent (A. Hohenberger, pers.communication).
-  Note that this phenomenon needs to be distinguished from the expression of direct speechwhich the participants in this study almost always correctly signal via quotation marks and correspondent introductory expressions like “the boy says...”.