The Representational Deficit for Dependent Relations hypothesis
In order to be able to develop appropriate grammatical assessments and remedial programmes, we need a precise account of the range of grammatical deficits found in children with SLI. Thus, van der Lely has advanced the Representational Deficit for Dependent Relations (RDDR) hypothesis (van der Lely, 1994, 1998) to explain the grammatical (syntactic) deficit in G-SLI subjects. This hypothesis accounts for the range of deficits found in G-SLI and, moreover, the optionality of grammatical operations (rules) which characterises these children's grammar. The hypothesis has stimulated and directed research exploring areas not previously investigated in SLI children and has proved useful in characterising the strengths (e.g., negative particles, Davies, PhD research) and deficits (e.g., question formation) of English and Greek-speaking SLI children (van der Lely & Battell, submitted; Savrakaki, 2001, in press). The potential clinical utility of the RDDR hypothesis has encouraged the development and evaluation of an assessment and therapeutic programme for severe syntactic deficits (Ebbels, PhD research).
The cognitive function and structure of language processing in children with SLI: A neuropsycholinguistic investigation
Questions concerning the validity of SLI subgroups and whether domain-specific or domain-general mechanisms underlie language processing are the focus of work funding by the Wellcome Trust, commencing in 2002. Recent work using event related potentials (ERPs) to investigate the neural correlates of sentence processing in children (Friederici & colleagues) has led us to plan work on the neural instantiation of syntactic and semantic processing in SLI and normally developing children. This work, to be carried out in collaboration with Professors Angela Friederici (director of the Max-Planck Institute of Cognitive Neuroscience, Leipzig), Johnson & Csibra (Birkbeck) examines the syntactic and semantic processing of spoken sentences and the effects of general cognitive load on sentence processing in the G-SLI and WF-SLI subgroups. The research will test competing hypotheses of SLI (the RDDR vs. Input-processing deficit hypotheses). By applying cognitive-neuroscience methods to the study of SLI, we aim to examine basic research questions through the study of language pathologies. Several projects being carried out by postdoctoral fellow and as part of doctoral studies are directly related to this work which is central to the centre.
