Climbing up the structure tree in the acquisition of German

In contrast to the relative paucity of research on the acquisition of DGS, the main developmental milestones in the acquisition of German are well-documented for a variety of acquisition situations (section 4.2). However, whether the developmental sequence identified for the acquisition of German holds equally of (monolingual or bilingual) deaf learners’ development remains unexplored. The situation is not unique to the German context but holds equally of other oral languages even though deaf learners’ skills in some oral languages have been assessed in standardised tests (the case of English or French, for example). As statistical studies are not designed to provide a qualitative account of structure-building in the written language they contribute little to our understanding of written language development.

Apart from a lack of empirical data, research on bilingual deaf learners’ acquisition of the oral language is also marked by a lack of a consensus concerning the status of the written language vis-a-vis the spoken language. Indeed, the fragmented picture that emerges in the literature about (monolingual and bilingual) deaf children’s attainment of the written language reflects a lack of agreement about whether what is defined as written language can be conceived of independently from speech. The issue is crucial for an appropriate understanding of the acquisition of the written language by deaf children whose acquisition of the spoken language is bound to be delayed if not truncated owing to their hearing loss.

Based on the Interdependence hypothesis of the spoken language-written language relation (section 2.4.2.1) we have argued that deaf learners can attain the written language as a second language. By attributing an equal status to spoken language and written language, and assuming that the nature of their relation is reciprocal (rather than unidirectional), the Interdependence hypothesis allows for the conception of alternative routes in their acquisition, regarding not only the order of acquisition, but also preferences in the processing routes (phonemic, graphemic or both).

We argued further that phonological awareness, the metalinguistic skill regarded by most scholars in the field as the key skill for a successful literacy acquisition, might not only be a requisite for but also an outcome of written language development. Crucially, we assume that written language comprehension and production involve the orchestration of several skills (section 2.4.3) and that apart from the mastery of the writing system, learners who acquire the written language as a second language without or with only partial knowledge of the spoken language it relates to are confronted with the task of acquiring a grammar, its units and the principles that underlie their combination. Whether or not deaf learners attain the target structure in a way similar to their hearing peers is one of the key empirical questions we addressed in our analysis of the written narratives collected in our longitudinal investigation.

The developmental profiles we established for each participant on the basis of the diagnostic criteria elaborated reveals that sign bilingually educated deaf children, too, expand their initially elementary syntactic structure progressively, which is in line with the Structure-building hypothesis. However, we also acknowledged substantial variation among participants regarding their structural competence at the onset of the study (variation ranges from elementary to full structures) and the progress they make in the time covered by the study (for one participant, Simon, we found no evidence for a structural expansion).

By combining qualitative analyses of the data with quantitative measures of selected phenomena we have been able to assess the scope of variation and its potential role in the organisation of the participants’ multilingual knowledge. We remarked on the lack of variation in the data of some of the participants at the onset of the study. Their adherence to a rather rigid sentential pattern con?trasts with the behaviour of monolingual learners of the language, who tend to produce a variety of word orders, with some preference for verb-final structures. We assume that this difference in the linguistic behaviour results from a didactic focus on the attainment of a surface (main clause) SVX word order. We have argued that the benefit attributed to the mastery of such a fixed sentential format needs to be qualified from a developmental perspective as it might mislead rather than help learners in their attainment of the target word order. Learners who start out with the assumption that German is an SVO language will be confronted with the task of revising their assumptions about the order of elements in the left periphery (target V2 constraint) and the verb-complement structure (target verb bracket) much like L2 learners of German who initially adopt their L1 SVO structure (the case of L1 Romance learners). Several errors documented in the data corroborate this assumption.

The preceding observations also underline the relevance of an early provision of a rich and complex input (including main clauses with periphrastic verb forms, complex constructions with subordinated clauses) that makes it easier for the learner to discover the relationships between the different elements in the clause and the asymmetric structure of German main and embedded clauses. We remarked further that the participants’ production of target-deviant structures they do not encounter in their German input deserves special attention because these errors provide further insights into the underlying language learning processes and the nature of the learner grammars.

Striking similarities to productions of other L2 learners of German become apparent already at an early stage, at which only elementary (VP) structures are available. Participants produce hypotactic combinations of several sequences, using also functional elements, such as complementisers, despite their lack of the target grammatical features. By assumption, participants in this study, like other L2 learners of German with a more advanced level in their L1 narrative development, tend to use linguistic means available via their L1 to express complex meanings.

Furthermore, we have seen that the apparent coexistence of alternative grammatical options provides important insights into the dynamics that characterise the organisation of language knowledge. Progress in language acquisition, as becomes apparent in different acquisition situations, is bound to variation and the resolution of conflict situations resulting thereof. It must be pointed out, however, that caution is advised in the interpretation of variation in learner data because it can only be established a posteriori whether it represents a temporary phenomenon preceding the eventual implementation of the target option.

At the level of word order, variation in the written productions of the participants in this study initially pertains to the relative order of the verb and the com?plement in constructions with periphrastic verb forms and phrasal verbs, before the target complement-verb order is eventually implemented. The situation is more complex regarding the attainment of the V2 constraint which involves various grammatical phenomena (verb raising, topicalisation, case marking). In this respect it became apparent that the inclusion of new target-like grammatical features associated with the V2 constraint does not occur to the immediate exclusion of the previously available target-deviant ones. This is reflected in the production of V3 patterns that result from the adjunction of non-subject constituents in the left periphery. Interestingly, the variation observed in the left periphery patterns well with the alternate use of V2 and V3 patterns documented in studies on other learners of L2 German.

In order to determine whether grammatical processes associated with an expanded sentential structure (including functional layers) were operative we used distributional (verb placement) and morphological criteria (verb inflection). At first sight, the analysis of the written narratives regarding verb inflection reveals a picture that is reminiscent of the findings obtained in previous descriptive studies in that error frequency rates are high. However, our analysis also makes apparent that there is variation at the individual level. While all of our participants produce errors at the onset of the study and continue to do so by the end of the study, they differ in the frequency of the errors produced. Whereas the overall proportion of errors raises in the narratives of Simon, it remains at about the same rate in the final narratives of Muhammed and Christa. For three participants, namely, Maria, Fuad and Hamida we acknowledge a decrease of the overall error rate by the end of the study.

With respect to the nature of the deficits observed, we remarked that the range of errors observed is limited and that it is developmentally constrained. We have also seen that variation in the use of finite and non-finite forms in sentential contexts that would require overt marking of subject-verb agreement does not represent a phenomenon that is unique to this type of acquisition situation, as substantial variation is also typically encountered in the productions of L2 learners of German. In line with current assumptions in L2 acquisition research, we have argued that the apparent optionality in the marking of subject-verb agreement reflects difficulties in the morphological realisation of abstract features, rather than the lack of functional categories.

We have expanded on this hypothesis by suggesting that the different types of errors identified reflect remaining deficits regarding the information encoded, whereby target-like inflection is assumed to involve the interaction of several grammatical modules (the lexicon, morphology, syntax). Errors produced at a time when the distributional evidence indicates that verb raising is operative suggest that deficits remain at the interface between morphology and syntax as person-number encoding remains problematic. Other errors indicate remaining gaps in the knowledge about the selective properties of the verbs (e.g. auxiliary selection) and morphological rules (participle formation). A rule-based verb inflection with irregular verbs indicates a deficit at the interface between morphology and lexicon.

While the eventual integration of information from different levels of linguistic analysis was not accomplished by most participants by the end of the recording time covered in the study, the productions of one learner, Maria, reveal that she has a command of verb inflection at the time. We are left therefore with a picture of deaf learners’ attainment of the target inflectional morphology that is definitely less idiosyncratic than traditional analysis made us believe. In addition, striking similarities become apparent between deaf and other L2 learners of the language.

In summary, by the end of the recording time, not all learners have gone all the way in the development of the target German grammar. With the exception of one participant, the developmental profiles established provide evidence for the structural expansion of elementary grammars and for grammatical processes associated with an expanded sentential structure. In Simon’s written productions, by contrast, we found no conclusive evidence for an evolution along these lines. Individual variation between the other participants pertains to the progress they make in the course of the study as some participants, for whom we found evidence for the implementation of the IP, do not produce evidence for the availability of the CP (that is, target-like complex clauses or interrogative clauses). Also, only some of the participants adhere to the V2 constraint by the end of the recording time. Nevertheless, the analysis of the data allows for the conclusion that participants in this study “climb up” the structure tree much like other L2 learners of German. This is an important conclusion given the myths that surround the acquisition of written language by deaf students.