Pooling of resources in the organisation a multilingual competence

In a study dedicated to bilingual language acquisition scientific interest pertains not only to the development of either language, but also to potential developmental asynchronies in the attainment of the properties of the target languages, and the role of language contact phenomena in learners’ productions.

With respect to the first issue, the comparison of the participants’ developmental profiles in DGS and German reveals that variation in the attainment of the target structure is more pronounced in German than in DGS. As for the relative progress they make in either language, we could see that while participants are still dealing with the attainment of the target syntactic structure in German at the end of the recording time, they have already accomplished this task at the onset of the study in DGS. Individual variation in the command of this language, as we learned, pertains to properties involving the syntax-discourse interface. Taken on the whole, however, the signed narratives collected document a sophisticated command of DGS that is creatively used for narrative purposes. The apparent asynchrony in the development of the two languages comes as no surprise given the circumstances that determine this type of bilingual language acquisition. However, the theoretically founded assessment of the participants’ competences allowed for a precise evaluation of the developmental asynchrony between the two languages. This type of assessment is not only of interest from the perspective of developmental linguistics. It is also of relevance for the elaboration of didactic conceptions that aim at promoting deaf students’ bilingual development. By being knowledgeable about the competences available in either language, more specific measures can be devised to promote the students’ attainment of the target language skills.

Turning to language interaction in the course of the bilingual development, we have remarked on the consensus in the broader field of bilingualism research that exposure to two or more languages does not wind up in linguistic confusion. Bilingual language acquisition is rather characterised by a separate development of two independent systems (section 2.3). Language separation in the course of the bilingual development, however, does not exclude the possibility of a temporary interaction of the two languages, which would be reflected in developmentally constrained language contact phenomena.

The analysis of the participants’ development of DGS and written German provides further support for the assumption that his holds equally of bilingual language acquisition in deaf learners. Bilingual deaf signers, too, develop two separate systems early on. This finding deserves to be emphasised against the backdrop of the variety of languages and codes deaf learners are exposed to. Crucially, despite of exposure to and use of a sign language and a signed system, that is, two codes that use the visual-gestural modality of expression, we found no indication for a confusion. DGS is clearly distinguished from oral language, be it in its written, spoken or manual form.

Cross-modal language contact phenomena occur in the data collected in this study. However, the overall frequency of these phenomena was found to be low. As prejudices against language mixing as an instance of confusion continue to abound in the field of deaf education, it is important to emphasise that the phenomena observed are developmentally constrained. Crucially, we have seen that language contact phenomena involve specific grammatical areas in both languages, and that the type of constructions mixed changes as learners proceed in their development of the respective language. Once the target grammatical properties are established, mixing takes over other (pragmatic) functions. The bidirectional nature of the relation reflected in the evidence of language mixing in the signed and in the written productions is well in line with current assumptions about bilingual learners’ pooling of resources. Also, the low incidence of language mixing observed patterns well with the general picture obtained in bilingual language acquisition research. We have argued that evidence of crossmodal language mixing can be taken as an indication that sign bilingual learners, too, (tacitly) know that languages are alike in fundamental ways, irrespective of the modality of expression.

Turning to the candidates for language contact phenomena documented in the data collected, we have seen that they include borrowings at the levels of the lexicon and syntax. At the level of syntax, we paid particular attention to the relative order of complements and verbs as DGS and German differ with respect to the relative position of object complements (or other modifying constituents) and verbs in main clauses.

As for DGS, the analysis reveals that the target constraints on word order in DGS were mastered by all participants with the exception of Fuad. Because this participant produces target-deviant verb-complement sequences alongside target-like sequences at the onset of the study we assume that both values of the VP headedness parameter coexist in his learner grammar before he eventually fixes the parameter to the target value. Target-deviant SVX patterns produced occasionally by other participants typically involve elements that are also used in signed German (LBG) for the expression of grammatical relations. This is the case of constructions with the auxiliary pam, produced with a target-deviant word order across the board, with no indication of subsequent development toward the target. By assumption, the target-deviant word order is adopted from LBG, in which the sign pam, glossed auf, is used as a preposition to mark the relation between the verb and its complement in SVX sentential formats. Interestingly, the function served by this preposition in LBG is also reflected in written German productions of some of the participants, in which the preposition auf is used as an overt case marker at a time when the target case system is not yet mastered.

Other language contact phenomena involving the use of LBG elements include constructions with ist, a sign created to express the German copula form. This LBG sign occurs in constructions that appear to be used as unanalysed formulae. Christa’s file 1 interrogative sequences, for example, include this element in combination with the interrogative marker where, a phenomenon that disappears by file 3. Other hybrid patterns that are neither compatible with DGS or German involve the use of the sign have in combinations with predicative adjectives or the use of before to mark past tense.

Turning to candidates for borrowing in written German productions, we remarked upon the phenomenon of verb drop. At the early stage of elementary grammars, verb drop was found to occur in sequences that would require the copula in target German, such as existential “da+X” patterns. We have argued that copula drop in the early productions of the participants is an ambiguous phenomenon. Because DGS knows no copula and “da x” (det_exist x) patterns are target-like in that language, copula drop represents a candidate for language mixing from that language. However, the optional realisation of elements at the VP stage has been found to occur also in other acquisition situations, which suggests that we are dealing with a developmentally constrained phenomenon. A different situation obtains in the analysis of copula drop at later stages of development, that is, at a time when functional elements are available to the learners. As sequences with copula drop alternate with target-like constructions at a more advanced stage, we assume that the variation observed reflects the availability of various structural options. Copula drop as an option, though target-deviant, might be reinforced by DGS. Such a reinforcing effect of the other language in bilingual learners has also been observed in the development of bilingual learners in other acquisition situations.

At the level of word order, we found occasional candidates for language mixing. Elements in some constructions appear to be arranged in accordance with the figure-ground principle. Some sequences appear to follow the verb final order of DGS. In general, however, we found no reinforcing effect of DGS on the attainment of the underlying SOV order in German, which we assume to be also an effect of the early focus on the surface SVX pattern remarked upon previously. Lexical borrowing from DGS becomes apparent in some verbless clauses that involve expressions which would require the use of periphrastic noun-verb combinations in German. The drop of the verb in these constructions reflects the partial overlap with equivalent DGS expressions at the lexical level.

A more subtle type of language mixing concerns the use of da (‘there’) serving the function det_loc would fulfil in DGS (that is, to assign a location to a referent) or the translation of simultaneously expressed complex DGS meanings (as is the case of complex classifier constructions) into sequentially expressed German propositions. We remarked on the relevance of paying attention to the German elements used in such translations as they reflect remaining lexical and structural gaps in the host language (the use of full NPs in the expression of spatial relations following the figure-ground principle, for example, reflects the lack of the pronominal system at the time). We also noted that some target-deviant sequences that represent candidates for borrowing deserve further scrutiny because of their pioneering potential in the development of the target structure. This potential could be exploited in the teaching/learning of the language pro

vided it is acknowledged. Combinations of propositions including the adjunction of prepositional phrases to the right of the noun they modify, for example, appear to involve borrowing from complex structures of DGS. Such borrowings could serve such a pioneering function when interpreted as precursors in the development of relative clauses. The observation that participants in this study tend to produce series of main clauses, rather than complex sentential constructions at later stage might serve as an indication that they did not exploit the potential implicit in these constructions any further.

Summarising, the sophisticated nature of structural and lexical borrowing provides intriguing insights into the multilingual knowledge developed by learners acquiring two languages of different modality of expression. Clearly, this difference does not prevent them from creatively using their linguistic resources. Only the type of mixing we have attributed to the use of LBG raises the question about the alleged benefit attributed to the use of a hybrid system. While the hybrid type of contact phenomena we observed might serve the function of a relief strategy in the absence of target grammatical properties, caution is advised in the interpretation of phenomena that might indicate that learners are misled in their attainment of some grammatical properties. Clearly, more research is needed about the characteristics of the communication in LBG to determine the potential influence of this system on the learners’ development in either language.