Lecturer: Klaus Abels (UCL)
Title: On the Poverty of the Stimulus: Evidence from Artificial Language Learning and from Ellipsis
Abstract: The classic poverty of the stimulus argument holds that competent speakers of a language have knowledge that a general purpose inferencing mechanism could not have arrived at on the basis of an unstructured hypothesis space together with the primary linguistic data. I will summarise recent work on artificial language learning experiments with the poverty of the stimulus paradigm which suggest substantial learner biases structuring the hypothesis space. Current work I am involved in probes biases that have been invoked in an explanation of the asymmetries in word-order patterns in the noun phrase (Universal 20) and the Germanic verb cluster. I then turn to ellipsis. Given the absence of clearcut overt evidence for structure in the ellipsis site, one can view elliptical structures as a real-life poverty of the stimulus experiment. Constraints on structures that are rare in the wild provide the best argument for an underlying structure of the hypothesis space. I will suggest that multiple sluicing, sluicing with case mismatches, and sluicing with non-canonical pied-piping behaviour are cases in point. The results favour a view with very abstract (as opposed to construction specific) biases that end up structuring the grammar.
Date and time: September 9-11 2019, 13.00-16.00
Participation: free
Location: KU Leuven, Faculty of Arts, Brussels Campus, Warmoesberg 26, 1000 Brussels
Room number: B07-16 (Mon-Tue) & 6108 (Wed)