PSYCHO-COMPUTATIONAL MODELS OF SUBSET PRINCIPLE COMPLIANCE IN SIMULATED LANGUAGE LEARNING

Previous research has proposed that any model of language learning should use the Subset Principle to guide hypothesis selection when the language domain contains at least two languages such that one is a subset of the other (Gold, 1967; Berwick, 1985; Manzini & Wexler, 1987; Wexler & Manzini, 1987). Informally, the Subset Principle states that the learner should select a language that is: a) compatible with the input data, and, b) does not properly contain any other language that is compatible with the input data. We will investigate, from both an empirical and theoretical perspective, psychocomputational models of language learning that abide by the Subset Principle. We intend “psychocomputational models” to include computational models that are in line with research in psycholinguistics, developmental psychology and linguistics (Sakas, 2004). Empirical research will focus on the creation and analysis of simulated language learners equipped with memory for past grammars. A comparison study will also be done between traditional total ordering learners and our partial ordering learners. Theoretical research will be concerned with investigating how the shape of the language domain, in terms of both cross-language ambiguity and the partial ordering of subset-superset relationships, would affect learner performance. Although there has been some computational research that attempts to address the problems that are introduced when learning in domains that contain superset languages, our research will make its contributions by directly modeling the Subset Principle within a psychologically compelling framework.