Models of Language Acquisition Inductive and Deductive Approaches

by Broeder, Peter; Murre, Jaap

• ISBN13: 9780198299899
• ISBN10: 0198299893
• Format: Hardcover
• Copyright: 2001-03-29
• Publisher: Oxford University Press

Summary

This book presents recent advances by leading researchers in computational modelling of language acquisition. The contributors have been drawn from departments of linguistics, cognitive science, psychology, and computer science. They show what light can be thrown on fundamental problems when powerful computational techniques are combined with real data. The book considers the extent to which linguistic structure is readily available in the environment, the degree to which language learning is inductive or deductive, and the power of different modelling formalisms for different problems and approaches. It will appeal to linguists, psychologists, cognitive scientists working in language acquisition,and to those involved in computational modelling in linguistic and behavioral science.

Author Biography

Peter Broeder is Assistant Professor at Tilburg University where his research involves unravelling processes of language change among ethnic minority groups.

Jaap Murre is a Research Fellow in the Department of Psychology, University of Amsterdam. Previously he worked as a scientist at the Applied Psychology Unit of the Medical Research Council in Cambridge. Dr. Murre also heads a research group sponsored by a PIONIER grant from the Netherlands Organization for Scientific Research (NWO).

Table of Contents

Introduction: The Computational Study of Language Acquisition	1	(8)
Peter Broeder
Jaap Murre
Introduction	1	(1)
Paradigms of Language Acquisition	1	(1)
Implementation: Theory and Experiment	2	(1)
Discovery of Existence Proof	3	(1)
Testing on Empirical Data	3	(1)
Conclusion	4	(5)
Part One: Words
Lexicalist Connectionism	9	(24)
Brian MacWhinney
Introduction	9	(3)
Grand Pretensions, Modest Reality	12	(5)
Lexical Items: an Achilles Heel?	17	(3)
A Solution to the Lexical Learning Problem	20	(2)
Using Lexical Representations	22	(6)
Discussion and Conclusions	28	(5)
Are SRNs Sufficient for Modelling Language Acquisition?	33	(22)
Noel Sharkey
Amanda Sharkey
Stuart Jackson
Introduction	33	(6)
Transfer, Forgetting, and Grammar Induction	39	(5)
Prediction Space Analysis	44	(2)
Extending Long-range Sequential Capabilities of SRNs	46	(5)
Discussion and Conclusions	51	(4)
A Distributed, yet Symbolic, Model for Text-to-speech Processing	55	(21)
Antal van den Bosch
Walter Daelemans
Introduction	55	(3)
SPC: a Symbolic, Single-route, Multiple-level Model	58	(5)
Analysing Orthographic Depth	63	(3)
Discussion and Conclusions	66	(10)
`Lazy Learning': Natural and Machine Learning of Word Stress	76	(27)
Steven Gillis
Walter Daelemans
Gert Durieux
Introduction	76	(3)
Metrical Markedness and Production Errors	79	(2)
The Learning Algorithm	81	(6)
Metrical Structure of Dutch	87	(2)
Experiment	89	(6)
Discussion and Conclusions	95	(8)
Part Two: Word Formation
Statistical and Connectionist Modelling of the Development of Speech Segmentation	103	(18)
Richard Shillock
Paul Cairns
Nick Chater
Joe Levy
Introduction	103	(1)
Corpus-based Research	104	(1)
Representing the Speech Signal	105	(2)
Statistical Modelling of the Corpus using a Connectionist Network	107	(10)
Discussion and Conclusions	117	(4)
Learning Word-to-meaning Mappings	121	(33)
Jeffrey Mark Siskind
Introduction	121	(1)
The Mapping Problem and its Motivation	122	(1)
Multi-word Utterances	123	(2)
Bootstrapping	125	(1)
Referential Uncertainty	126	(2)
Noise	128	(2)
Homonymy	130	(1)
The Noise-free Monosemous Case	131	(4)
Extensions to Handle Noise and Homonymy	135	(5)
Simulations	140	(4)
A Statistical Model of Noise	144	(5)
Experiments	149	(2)
Discussion and Conclusions	151	(3)
Children's Overregularization and its Implications for Cognition	154	(23)
Gary F. Marcus
Introduction	154	(1)
Models of Inflection	154	(13)
Implications for Cognitive Science	167	(4)
Discussion and Conclusions	171	(6)
A Recurrent Network with Short-term Memory Capacity Learning the German -s plural	177	(24)
Rainer Goebel
Peter Indefrey
Introduction	177	(3)
The German Plural System	180	(1)
The Model	181	(9)
Experimental Evidence on the German Plural	190	(2)
Discussion and Conclusions	192	(9)
Single- and Dual-route Models of Inflectional Morphology	201	(24)
Ramin Nakisa
Kim Plunkett
Ulrike Hahn
Introduction	201	(2)
Pattern Associators	203	(4)
When Are Two Routes Better Than One?	207	(3)
Description of the Inflection Systems	210	(3)
Comparing Single-route and Dual-route Models	213	(6)
Discussion and Conclusions	219	(6)
Part Three: Word Order
Formal Models for Learning in the Principles and Parameters Framework	225	(19)
Partha Nyogi
Robert C. Berwick
Introduction	225	(1)
The Triggering Model for Parameter Setting	226	(6)
Derivation of Transition Probabilities for the Markov TLA Structure	232	(2)
Convergence Times for the Markov Chain Model	234	(7)
Discussion and Conclusions	241	(3)
An Output-as-input Hypothesis in Language Acquisition	244	(29)
Loekie Elbers
Introduction	244	(4)
An Output-as-input Model	248	(4)
Three Phases	252	(4)
Some Evidence from Case Studies	256	(10)
Discussion and Conclusions	266	(7)
Notes on Contributors	273	(6)
Addresses	279	(4)
Index	283