A Connectionist Investigation of Linguistic Arguments from the Poverty of the Stimulus: Learning the Unlearnable

The principal goal of Chomskyan linguistics is to account for language acquisition; the core postulate is that the child has available linguistically detailed innate knowledge, known as Universal Grammar (UG). Critical to this postulate is the poverty of the stimulus argument that the child’s target grammar is “hopelessly underdetermined by the fragmentary evidence available” (Chomsky, 1968). UG is assumed to account for those “properties of language that can reasonably be supposed not to have been learned” (Chomsky, 1975). Our research demonstrates, however, that important statistical properties of the input have been overlooked, resulting in UG being credited for properties which are demonstrably learnable; in contradiction to the most celebrated argument from the poverty of the stimulus Chomsky’s structuredependence argument (e.g. Chomsky, 1975) a simple recurrent network (Elman, 1990), given input modelled on child-directed speech, is shown to learn the structure of relative clauses, and to generalize that structure to subject position in aux-questions. The result demonstrates that before a property of language can reasonably be supposed not to have been learned, it is necessary to give greater consideration to the indirect positive evidence in the data and that connectionism can be invaluable to linguists in that respect.