Prior Structure in Language Acquisition: A Consensus Workshop

Date: August 31 - September 1, 2017

Organizers: Giosuè Baggio & Terje Lohndal

Venue: NTNU Norwegian University of Science and Technology, Trondheim

In recent years, research in the language sciences has been revitalized by the emergence of novel research programs and by the radical reconfiguration of some existing ones. More compact characterizations of Universal Grammar have been produced in parallel with alternative frameworks in the theory of grammar and language learning. Modelling techniques such as Bayesian probability and game theory have been applied in computational and experimental work on language learning, processing and transmission, generating new problems and results. Cognitive neuroscience, genetics and comparative ethology are changing our understanding of the biological preconditions of language. Finally, technological advances such as deep learning are bringing neural networks back to the fore. None of these projects views language as the effect of the environment on a blank slate, but they differ as to the nature and scope of the prior structure that enables language learning in humans.

The aim of the Workshop is to explore the possibility that a common core of learning principles, constraints or mechanisms exists, shared by all or most of these different approaches to language. Specifically, we will bring together scholars from different sub-disciplines of the language sciences and from different research traditions in search for consensus on what kind of prior structure enables first language acquisition, and on what may be its evolutionary origins.

The invited speakers at the Workshop are:

David Adger Queen Mary University, London

Jennifer Culbertson University of Edinburgh

Adele Goldberg Princeton University

Simon Kirby University of Edinburgh

Ianthi Tsimpli University of Cambridge

Charles Yang University of Pennsylvania

Call for papers

We invite additional contributions to the Workshop in the form of concise (word limit is 5000), tightly argued papers on the topic. Papers must address at least one of these issues:

  • What is the nature of prior structure in first language acquisition: constraints on learning, inductive biases, specific computational mechanisms, principles of grammatical or conceptual structure etc.?
  • What is the scope of prior structure in first language acquisition: does it include properties of sound only, or does it affect grammar acquisition too?
  • Where is critical evidence for the options above more likely to come from: typology, experimental and computational research on language acquisition and transmission, cognitive neuroscience, genetics, comparative biology etc.?
  • What is the most effective methodological approach at this stage of research: to posit minimal or no prior structure, pursue an empiricist approach and increase prior structure at each failed attempt to model or explain the data; or, conversely, to posit as much structure as seems reasonable, and then reduce it by falsification or correction vis-à-vis empirical data; or a mixture of the two?

Papers must be submitted by March 31, 2017 to the following email address: . All papers will be peer reviewed. Accepted papers will be presented at the Workshop. We aim to publish the proceedings of the Workshop either in a special issue of an international linguistics or cognitive science journal or as a book with one of the major international academic publishers.