Debate: Language and Big Data


Friday 24 March
16:30 – 18:20 (4:30 PM – 6:20 PM) Hall N1

Can Big Data provide a different perspective on Language Competence?

The question of how far one can go in explaining language using only the statistics embedded in the language data has been debated for many years within Linguistics, Psychology and Neuroscience. Recent developments in computational modelling (e.g., deep learning) and the availability of very large corpora of language, behavioral and imaging data bring a new perspective to this question. Two world-leading researchers at the cutting edge of Machine Learning/Computational Linguistics Marco Baroni and Edward Grefenstette will describe and provide a critical evaluation of current models that do not rely on linguistic structure. Their presentations will be followed by presentations by two other world-renowned scholars, W. Tecumseh Fitch and Ina Bornkessel-Schlesewsky who will discuss implications and alternative viewpoints.

Moderator: Nick Chater, Behavioural Science Group, Warwick Business School, United Kingdom

Edward Grefenstette, DeepMind, United Kingdom

Marco Baroni, Department of Information Engineering and Computer Science, University of Trento, Italy

W. Tecumseh Fitch, Department of Cognitive Biology, University of Vienna, Austria

Ina Bornkessel-Schlesewsky, School of Psychology, Social Work and Social Policy University of South Australia