On the orchestration of linguistic devices in the acquisition of DGS

Sign language productions of bilingually educated deaf students have seldom been subjected to qualitative analyses that would provide a detailed picture of the structural knowledge attained. Commonly, general evaluations attribute bilingual students full competence in the language. The theoretically founded approach proposed in this work allowed for the scrutiny of empirical data with a view to assessing the learners’ command of the target grammar and to tracking down the origin of potentially remaining gaps. It is important in this context to acknowledge that caution is advised in research on the acquisition of a language whose grammatical description is ongoing, as it is the case of DGS. Nevertheless, although interpretations of the data might remain tentative at times, the findings obtained contribute to the overall objective of obtaining a better understanding of the language, its acquisition and use.

We remarked previously that only few studies are available on the acquisition of DGS. First insights into structure-building in infant native signers with a focus on their acquisition of verb agreement have been obtained in a longitudinal study conducted by Hanel (2005, cf. Plaza-Pust 2016 for a discussion). The DGS competence of bilingually educated deaf students in Hamburg was assessed in a rather global manner, as text production skills in written language were at the focus of the studies conducted by Schafke et al. (2004). The present study therefore represents a first attempt at assessing bilingual deaf learners’ competence in DGS and in written German.

Now, the participants in this study are all children of hearing parents who have been exposed to DGS or signed German early in their lives, but nevertheless with some years of delay when compared to native signers exposed to the language from birth. So, although participants were of an advanced age when we started to record data, whether or not they had a command of the full structure of DGS could not be presupposed with certainty. Also, because they used signed German (LBG) in their everyday communication within and without the school premises, a lack of differentiation between DGS and LBG could not be ruled out.

The developmental profiles established for each participant revealed that all participants had a command of the full sentential structure at the onset of the study and that grammatical processes associated with the functional IP and CP levels were operative, including verb inflection, signalling and marking of referential shift, subordination and question formation. Further development in the participants’ command of the language in the course of the time span covered by the study was observed in their use of the linguistic devices available for narrative purposes. For example, changes concerning the use of complex constructions (their variety, their frequency) were found to be reflected in the level of narrative complexity: the expression of temporal and causal relations as well as the use of complex clauses to recount the protagonists’ emotions and thoughts derived more structured and more detailed retellings of the picture story.

Further, our analysis revealed that not all dimensions that are pertinent to a linguistic use of the sign space were mastered by the participants at the onset of the study; however, they all made substantial progress in their ability to integrate the information from different levels of linguistic analysis in the time span covered by the present investigation.

For example, we remarked on the participants’ command of verb inflection at the onset of the study. However, full mastery of verb inflection, and several other grammatical phenomena in DGS (cf. section 3.1), involves not only the mor- phosyntactic but also the discourse level of linguistic analysis (apart from the lexical, involving the distinction of plain, agreement and spatial verbs). Recall that the choice of referential loci to mark agreement is modelled also by dis?course requirements for the purpose of creating cohesion. Indeed, a consistent use of loci in the sign space is crucial for an appropriate understanding of the narration produced. In this respect, we found participants to differ. Individual variation can be seen on a continuum, with learners who use referential loci in a contrastive and consistent manner early on at one end of the spectrum (the case of Muhammed, for example). At the other end of the continuum, we would see learners who mark verb agreement and referential shifts locally, at the sentential level, without paying too much attention to discourse requirements for the creation of cohesion (the case of Hamida, for example).

In a similar vein, we found participants to differ in the choice of linguistic forms for the introduction, reintroduction or maintenance of reference. The challenge here lies in the selection of overt reference forms vis-a-vis null elements, a choice that is bound not only to grammatical constraints, but also to discourse requirements as these elements fulfil different narrative functions depending on the discourse context they appear in. In this respect, our analysis reveals that null elements occur with a high frequency in the narratives of all participants. Further, we remarked on the referential ambiguity that arises through the choice of null elements in those narrative contexts where characters that have been out of narrative focus for some time are reintroduced as protagonists of a new narrative episode. As for the participants’ development in their choice of reference forms, the data document substantial progress in learners like Christa for whom we initially found a relatively high proportion of null subjects serving the function of reintroduction of referents. In more general terms, as variation in the choice of reference forms has also been observed in the narrative development of hearing learners acquiring spoken languages, we might conclude that irrespective of the modality of expression chosen, the disambiguation of reference forms is equally a task in signed and spoken narratives.

The progressive integration of information from different levels of linguistic analysis also becomes apparent in the participants’ expression of spatial relations as they narrate the protagonists’ locations and movements. In this case, too, we found that participants differ in the degree of detail they provide which reflects not only differences in narrative style, but also in the ability to integrate morphosyntactic, syntactic and discourse information. Recall that the target-like expression of spatial relations involves the appropriate selection of inflected spatial verb forms, word order, and h2-classifiers to background information.

All in all, the protracted development we observed in our data regarding the orchestration of linguistic devices for narrative purposes is well in line with the findings obtained in investigations on the acquisition of other sign languages, in which development of narrative skills is reported to continue at least to adolescence. It is important to note, however, that the available knowledge about nar?rative development in sign language learners continues to be fragmentary, with information available only on selected properties of individual sign languages. At the theoretical level, too, more research is needed that would provide further insights into the functions served by linguistic devices in diverse discourse contexts. Beyond the theoretical interest there is also an applied dimension, as information about the differential use of the language is also urgently needed in the training of the teaching personnel involved in sign bilingual education.

We know from studies on spoken language acquisition in hearing learners that grammatical properties are neither acquired instantaneously nor are they attained en bloc. This holds equally of the development of narrative skills. Indeed, full mastery of the orchestration of linguistic devices for narrative purposes has been documented to be the result of a protracted development also in hearing learners. As pointed out by Berman (2004: 265), “[i]n learning how to tell a story, as in other domains, acquisition is not an “all-or-nothing” leap from no knowledge to full knowledge. Rather, it involves partial knowledge and reorganization and integration of prior knowledge across different domains.” In addition, we also need to consider that the ability to construct a well-organised narrative text not only develops late, but has also been found to be manifested better in some narrative contexts than in others (Berman 2004: 265). In this respect, more research is needed to establish the potential scope of variation regarding the use of linguistic means in different discourse contexts (for example, by contrasting data from personal experience accounts and story retelling corpora).