The characteristics and outcome of Specific Language Impairment (SLI) in (successive) bilingual children have as yet very rarely been studied. The findings obtained up to now, though, clearly show that multi-/bilingualism alone does not amplify SLI (cf. Håkansson et al. 2003; Paradis et al. 2003; Restrepo & Guitiérrez-Clellen 2004). However, in order to clearly diagnose grammatical SLI, the outcome in both languages has to be identified. Language disorders that are observed only in the second language may be caused by external factors. Thus, a diagnostic dilemma appears, for only few studies characterising grammatical SLI in many of the first languages of bilingual children exist. We will present findings from a study referring to the early successive acquisition of German as second language (L2) in children with Turkish as their first language (L1). While most of the children in our study develop inconspicuously in both languages, some children show deviant forms and structures in both their Turkish and their German language production that may refer to grammatical SLI. The present analysis concentrates on Turkish case morphology. Successive-bilingual Turkish-German children who are suspected to be SLI-children show deviant profiles in L1 case morphology development. In this paper, the data of two normally developing children and two children who are suspected to be SLI will be presented and discussed.

The topic of this paper are the grammatical properties of spoken language. Do we need a totally new theory for modelling them? Are sentences still basic units? In which way can one handle the specific syntactic constructions? The answer given by some conversational analysts to these questions are insufficient in various aspects. These problems will be discussed in the following.