Language Processing and Language Acquisition

393 Pages·1990·19.86 MB·English

by  Lyn Frazier

LANGUAGE PROCESSING AND LANGUAGE
ACQUISITION
STUDIES IN THEORETICAL PSYCHOLINGUISTICS
VOLUME 10
Managing Editors
Thomas Roeper, Dept. of Linguistics, University of Massachusetts at Amherst
Kenneth Wexler, Dept. of Brain and Cognitive Science, MI.T., Cambridge, Mass.
Editorial Board
Robert Berwick, Artifical Intelligence Laboratory, MI.T., Cambridge, Mass.
Manfred Bierwisch, Zentralinstitutfur Sprachwissenschaft, Akademie der
Wissenschaften der D.D.R.
Merrill Garrett, University of Arizona, Tucson
Lila Gleitman, School of Education, University of Pennsylvania
Mary-Louise Kean, University of California at Irvine
Howard Lasnik, University of Connecticut at Storrs
John Marshall, Neuropsychology Unit, Radcliffe Infirmary, Oxford
Daniel Osherson, M.l.T., Cambridge, Mass.
Yukio Otsu, Tokyo Gakugei University, Tokyo
Edwin Williams, University of Massachusetts at Amherst
The titles published in this series are listed at the end of this volume.
LANGUAGE
PROCESSING AND
LANGUAGE
ACQUISITION
Edited by
LYN FRAZIER
Department of Linguistics,
University of Massachusetts,
Amherst, U.S.A.
and
JILL DE VILLIERS
Department of Psychology, Smith College,
Northampton, U.S.A.
"~.
SPRINGER SCIENCE+BUSINESS MEDIA, B.V.
Library or Congress Cataloging-in-Publication Data
Language orocess'"g and language acquIsItIon! edIted by Lyn Frazler
and -.I,ll de Vllllers.
p. cm. -- (StudIes In theoretical psychcllngulStics)
Papers presented at a confe~ence held at the UnIversity of
Massachusetts, Amherst, In May 1988.
Includes bibliographical references.
ISBN 978-0-7923·0660-3 ISBN 978·94-011·3808·6 (eBook)
DOl 10.1007/978·94·011·3808·6
alk. pape""
1. Language acquIs~tlon--Cungresses. 2. Grammar. ComparatIve and
genera I--Congresses, 3, Psycho Ilngu 1st' cS--Congresses.
1. Fraz'er, Lyn, 1952- 11. Je V,II iers, -.li '] G .• 1948-
IIl. Ser les.
P118.L384 1990
401' .93--dc2C 90-11021
ISBN 978-0-7923-0660-3
Printed on acid free paper
All Rights Reserved
© 1990 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 1990
Softcover reprint of the hardcover 1s t edition 1990
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
This book is dedicated to
Charles Clifton
for his exceptional devotion to the field
and those who play in it
TABLE OF CONTENTS
ACKNOWLEDGMENTS IX
L YN FRAZIER and JILL DE VILLIERS / Introduction
DAVID LEBEAUX / The Grammatical Nature of the Acquisi
tion Sequence: Adjoin-a and the Formation of Relative
Clauses 13
ANNE V AINIKKA / The Status of Grammatical Default
Systems: Comments on Lebeaux 83
KEN WEXLER / On Unparsable Input in Language Acquisi-
tion 105
VIRGINIA VALIAN / Logical and Psychological Constraints
on the Acquisition of Syntax 119
THOMAS ROEPER and JURGEN WEISSENBORN / How to
Make Parameters Work: Comments on Valian 147
REIKO MAZUKA and BARBARA LUST / On Parameter
Setting and Parsing: Predictions for Cross-Linguistic Differ-
ences in Adult and Child Processing 163
NOBUKO HASEGAWA / Comments on Mazuka and Lust's
paper 207
JANET DEAN FODOR / Parameters and Parameter-Setting in
a Phrase Structure Grammar 225
JILL DE VILLIERS. THOMAS ROEPER and ANNE V AINIKKA /
The Acquisition of Long-distance Rules 257
VB
viii TABLE OF CONTENTS
AMY WEINBERG / Child Grammars - Radically Different, or
More of the Same?: Comments on de Villiers, Roeper and
Vainikka 299
DANA McDANIEL and HELEN SMITH CAIRNS / The
Processing and Acquisition of Control Structures by Young
Children 31 3
GREG N. CARLSON / Intuitions, Category and Structure:
Comments on McDaniel and Cairns 327
STEPHEN CRAIN, CECILE McKEE and MARIA EMILIANI /
Visiting Relatives in Italy 335
JANE GRIMSHAW and SARA THOMAS ROSEN / Obeying
the Binding Theory 357
HELEN GOODLUCK / Knowledge Integration in Processing
and Acquisition: Comments on Grimshaw and Rosen 369
LIST OF FIRST AUTHORS 383
INDEX 387
ACKNOWLEDGMENTS
This volume reports papers presented at the University of Massachusetts
Conference on 'Language Processing and Language Acquisition' held in
Amherst on May 7-8, 1989. The conference was supported by N.I.H.
Training Grant # HD-07327. We wish to thank the authors not only for
their papers and commentary, but also for reviewing each other's
papers. We are also grateful to Emmon Bach, Tova Rappaport, luli
Carter, Greg LaMontagne and especially Tom Roeper for help with
many stages of the conference and proceedings. Special thanks go to
Maggie Browning for the excellent commentary she provided at the
conference, and to Kathy Adamczyk for her extensive help and organi
zational wizardry.
IX
L YN FRAZIER AND JILL DE VILLIERS
INTRODUCTION
Studies of language acqUiSItion have largely ignored processing prin
ciples and mechanisms. Not surprisingly, questions concerning the
analysis of an informative linguistic input - the potential evidence for
grammatical parameter setting - have also been ignored. Especially in
linguistic approaches to language acquisition, the role of language
processing has not been prominent. With few exceptions (e.g. Goodluck
and Tavakolian, 1982; Pinker, 1984) discussions of language perform
ance tend to arise only when experimental debris, the artifact of some
experiment, needs to be cleared away. Consequently, language pro
cessing has been viewed as a collection of rather uninteresting perform
ance factors obscuring the true object of interest, namely, grammar
acquisition. On those occasions when parsing "strategies" have been
incorporated into accounts of language development, they have often
been discussed as vague preferences, not open to rigorous analysis. In
principle, however, theories of language comprehension can and should
be subjected to the same criteria of explicitness and explanatoriness as
other theories, e.g., theories of grammar. Thus their peripheral role in
accounts of language development may reflect accidental factors, rather
than any inherent fuzziness or irrelevance to the language acquisition
problem.
It seems probable that an explicit model of the way(s) processing
routines are applied in acquisition would help solve some central
problems of grammar acquisition, since these routines regulate the
application of grammatical knowledge to novel inputs. Several examples
are given below, intended to illustrate the kinds of puzzles to which a
processing solution might contribute. For a start, it opens the possibility
of linguistic input influencing acquisition, and thus playing a role in
acquisition theory, without complete handwaving in the account of how
the input was analyzed correctly in the first place. In the absence of a
processing theory, the only way input data can influence grammar
acquisition is by unanalyzed input directly fixing a grammatical parame
ter or by the acquisition system using "compounds" of linguistic
relations already present in the child's current grammar.
1
Lyn Frazier and Jill de Villiers (eds.), Language Processing and Language Acquisition,
1-11.
© 1990 Kluwer Academic Publishers.
2 L YN FRAZIER AND JILL DE VILLlERS
To begin, we may consider what drives the acquisition process. Why
does the child abandon some intermediate grammar (G)? Most
j
theories appeal somehow to new input. But this is problematic. If a new
input is consistent with G then there is no need for G to be changed.
j, j
However, if the novel input is not consistent with G then the child
j,
presumably must assign the input some analysis which is incompatible
with G,. The child must then be willing to alter the grammar on the
basis of an ungrammatical input. This is a potentially risky process
especially if grammar acquisition involves the setting of parameterized
principles where setting the value of a single parameter may have wide
consequences permeating the entire syntactic system (cf. Roeper and
Williams. 1987).
Several papers in this volume make explicit or implicit reference to
the view that language acquisition is a parameter setting process, though
there are some unresolved disagreements among the positions about
the desirability of "linked" versus "unlinked" parameters. On the view
that parameters may be linked, acquisition is seen as an economical
process whereby rapid progress can be made on the basis of limited
input trggers. However, there is a concomitant cost attached for lin
guistic theory if parameters are not independent, as Fodor (this volume)
points out. On the "linked" parameters view, little input will be required.
but it will be essential for the crucial input to be analyzed correctly.
Given unlinked parameters, accuracy in data analysis will be less
central, but more input will be needed. On either view, it becomes quite
important to investigate precisely how a description is assigned to an
"informative input". i.e. one serving as evidence for grammar alteration.
The papers in this volume address this question, outlining several
different types of solutions. Lebeaux (this volume) emphasizes that both
parsing and acquisition reflect core grammar. He provides an account
of the development of relative clauses that introduces the notion that
the grammar/parser is "geologically layered", such that the child's
errors always represent a return to certain default positions. In contrast
to most existing proposals, he give a specific view of how inputs are
treated. and the connnection between acquisition and parsing. He
suggests UG equips the child with a set of default operations which
apply to analyze inputs extending beyond the current grammar. These
defaults are east in a form which allows them to apply directly to a
novel input, determining the default grammatical analysis assigned to
aspects of the phrase marker which are underspecified given the child's