LI0057 Computational Modelling of Linguistic Processes

(Also known as COMPUTATIONAL LINGUISTICS 2 Half course)

Prerequisites

Content and teaching

• Major theme: Connectionist (neural net) modelling of language processes (weeks 1-8)
• Minor theme: Computational models of the evolution of recursive language (weeks 7-10).

The course will involve extensive essential reading. The usual weekly pattern will be:

• Lecture introducing topic and reading.
• Tutorial discussion of reading.

In addition, there will be an introduction to practical exploration of a package for running neural net simulations (XERION).

Apart from the practical work with a neural net package, the course will have no taught practical component. But students will be encouraged to use their own computational expertise to write small programs exploring various aspects of the issues raised by the reading.

Instructors

The course will be taught by Prof. J Hurford, with occasional participation by research associates currently working in this field. These associates are also willing to supervise coursework projects for this course (see below).

Assessment

EITHER ONE

report of about 4000 words (or involving equivalent work in programming or experiment),

OR TWO

reports of about 2000 words each (or involving equivalent work in programming or experiment).

A ``report'' may be

EITHER

an essay critically surveying and comparing at least two (for a 2000-word report) or three (for a 4000-word report) of the works studied,

OR

a description of a small computational experiment designed to test, (in)validate, or otherwise throw light on, one or more of the works read.

Students have a free choice within these options, but are advised to negotiate their choices early with me (Jim Hurford).

The deadline for the first report of two will be the end of Week 6, and the deadline for the second report of two, or the single `big' report will be the first day of the following term).

Weekly Readings

1. Bechtel, William, and Adele Abrahamsen, 1991 Connectionism and the Mind: an Introduction to Parallel Processing in Networks, Blackwells, Oxford. (Hereafter `B&A') Chapters 1 and 2, pp.1-65.

2. B&A Chapters 3 and 5, pp.66-105, 147-175.

3. B&A Chapter 6, pp.175-210. and Rumelhart, David E., and James L.McClelland, 1986 ``On Learning the Past Tenses of English Verbs'', in Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Vol.2, Psychological and Biological Models, ed by James L.McClelland and David E.Rumelhart, MIT Press, 216-271.

4. Pinker, S and Prince, A. 1988 ``On language and connectionism: Analysis of a parallel distributed processing model of language acquisition'', Cognition, 28:73-193.

5. Fodor, J.A. and Pylyshyn, Z.W. 1988 ``Connectionism and cognitive architecture: a critical analysis'', Cognition, 28:3-71.

6. Elman, Jeffrey L., 1993 ``Learning and development in neural networks: the importance of starting small'' Cognition, 48:71-99.

7. Batali, John, 1994 ``Innate Biases and Critical Periods: Combining Evolution and Learning in the Acquisition of Syntax''. In Artificial Life VI, ed. by Rodney Brooks and Pattie Maes, MIT Press, 160-171. and Christiansen, Morten, and Devlin, Joseph, 1997 ``Recursive inconsistencies are hard to learn: a connectionist perspective on universal word order correlations.'' In Proceedings of the 19th Annual Cognitive Science Society Conference, 113-118, Lawrence Erlbaum Associates, Mahwah, NJ.

8. Batali, John, 1998 ``Computational Simulations of the Emergence of Grammar'', In Approaches to the Evolution of language: Social and cognitive bases, edited by James R Hurford, Chris Knight and Michael Studdert-Kennedy, Cambridge University Press, 405-426.

9. Kirby, Simon, forthcoming, ``Learning, bottlenecks and the evolution of recursive syntax'', and Hurford, James R, forthcoming ``Social transmission favours linguistic generalization'', both to appear in Approaches to the evolution of language: the emergence of phonology and syntax, edited by Chris Knight, Michael Studdert-Kennedy and James R Hurford, Cambridge University Press.

10. Batali, John, forthcoming, ``The negotiation and acquisition of recursive grammars as a result of competition among exemplars'', to appear in Linguistic evolution through language acquisition: formal and computational models, edited by Ted Briscoe, Cambridge University Press.