Morphological Structure in Language Processing

532 Pages·2003·17.146 MB·

by  R. Harald Baayen, Robert Schreuder

Morphological Structure in Language Processing
W
Trends in Linguistics
Studies and Monographs 151
Editors
Walter Bisang
Hans Henrich Hock
(main editor for this volume)
Werner Winter
Mouton de Gruyter
Berlin · New York
Morphological Structure
in Language Processing
Edited by
R. Harald Baayen
Robert Schreuder
Mouton de Gruyter
Berlin · New York
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter GmbH & Co. KG, Berlin.
® Printed on acid-free paper which falls within the guidelines
of the ANSI to ensure permanence and durability.
Library of Congress Cataloging-in-Publication Data
Morphological structure in language processing / edited by R.
Harald Baayen, Robert Schreuder.
p. cm — (Trends in linguistics. Studies and monographs ;
151)
Includes bibliographical references and index.
ISBN 3-11-017892-3 (alk. paper)
1. Grammar, Comparative and general — Morphology —
Psychological aspects. 2. Psycholinguistics. I. Baayen,
R. Harald, 1958- II. Schreuder, Robert. III. Series.
P241.M5986 2003
415'9-dc22
2003044270
ISBN 3-11-017892-3
Bibliographic information published by Die Deutsche Bibliothek
Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data is available in the Internet at <http://dnb.ddb.de>.
© Copyright 2003 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this
book may be reproduced or transmitted in any form or by any means, electronic or mechan-
ical, including photocopy, recording or any information storage and retrieval system, with-
out permission in writing from the publisher.
Cover design: Christopher Schneider, Berlin
Printed in Germany.
Contents
Preface vii
Inflectional morphology and word meaning: Orthogonal or
co-implicative cognitive domains?
Aleksandar Kos tic, Tanja Markovic and Aleksandar Baucal 1
Visual processing of Italian verbs and adjectives: The role of
the inflectional family size
Daniela Traficante and Cristina Burani 45
Morphological resonance in the mental lexicon
Nivja H. de Jong, Robert Schreuder and R. Harald Baayen 65
Morphology and frequency: Contrasting methodologies
Michael A. Ford, William D. Marslen-Wilson and Matthew H. Davis 89
Derivational morphology in the German mental lexicon: A dual
mechanism account
Harald Clahsen, Ingrid Sonnenstuhl and James P. Blevins 125
The interplay of root, suffix and whole-word frequency in processing
derived words
Cristina Burani and Anna M. Thornton 157
On the role of derivational affixes in recognizing complex words:
Evidence from masked priming
Helene Giraudo and Jonathan Grainger 209
Morphological facilitation: The role of semantic transparency and
family size
Laurie Beth Feldman and Matthew John Pastizzo 233
Recognition of spoken prefixed words: The role of early conditional
root uniqueness points
Lee H. Wurm and Joanna Aycock 259
vi Contents
Lexical representation of morphologically complex words:
Evidence from Polish
Agnieszka Anna Reid and William DavidMarslen-Wilson 287
Identification of spoken prefixed words in French
Fanny Meunier and Juan Segui 337
Frequency effects in regular inflectional morphology:
Revisiting Dutch plurals
R. Harald Baayen, James M. McQueen, Ton Dijkstra and
Robert Schreuder 355
How does a child detect morphology? Evidence from production
Wolfgang U. Dressier, Marianne Kilani-Schoch and
Sabine Klampfer 3 91
Frequency effects in processing inflected Dutch nouns:
A distributed connectionist account
Matthew H. Davis, Maarten van Casteren and
WilliamD. Marslen-Wilson 427
When word frequencies do not regress towards the mean
R. Harald Baayen, Fermin Moscoso del Prado Martin,
Robert Schreuder and Lee Wurm 463
Spelling errors with a view on the mental lexicon: Frequency
and proximity effects in misspelling homophonous regular verb
forms in Dutch and French
Dominiek Sandra and Michel Fayol 485
List of contributors 515
Subject index 519
Preface
This volume brings together a series of studies of morphological processing
in Germanic (English, German, and Dutch), Romance (French and Italian),
and Slavic (Polish, Serbian) languages. The question of how morphologi-
cally complex words are organized and processed in the mental lexicon is
addressed from different theoretical perspectives (single and dual route
models), for different modalities (auditory and visual comprehension,
writing), and for language development. Experimental work is reported, as
well as computational and statistical modeling. The aim of this volume is to
provide an overview of the range of issues currently attracting research at
the intersection of morphology and psycholinguistics.
The role of morphological complexity for lexical processing has been
studied most extensively for the visual domain. In this volume, three lines
of current research are represented. The first line of research addresses the
role of paradigmatic relations between complex words in the mental lexi-
con. The chapter by Kostic, Markovic, and Baucal introduces an informa-
tion-theoretic approach to the inflectional paradigms in Serbian. The con-
sequences of the size of inflectional paradigms in Italian for lexical proc-
essing is studied in the chapter by Traficante and Burani.
Paradigmatic relations are also present outside the domain of inflection.
While paradigmatic relations in inflection are often highly structured by
features such as case, number, person, tense, and aspect, the paradigmatic
relations in a morphological family brought about by derivation and com-
pounding are only loosely structured. Nevertheless, the size of the mor-
phological family has been observed to be a relevant predictor for lexical
processing. The chapter by De Jong, Schreuder, and Baayen explores the
way in which the family size effect is modulated by linguistic context. The
chapter by Ford, Marslen-Wilson, and Davis studies the family size effect
in English in its relation to a range of other frequency measures. Feldman
and Pastizzo report an investigation of the relation between family size and
semantic transparency.
A second line of research in visual studies represented in this volume
addresses the role of morphological decomposition in lexical processing.
The chapter by Clahsen, Sonnenstuhl, and Blevins addresses the role of
decomposition for German productive derivational suffixes, and reports
frequency effects for the derived forms combined with full cross-modal
priming for those forms. The chapter by Burani and Thornton studies the
contribution of root and affix to morphological decomposition for high and
low frequency complex words.
viii Preface
A third line of research addresses the question of whether morphological
regularities affect lexical processing during the early stages of visual identi-
fication. The chapter by Giraudo and Grainger reports experimental re-
search suggesting that morphological effects arise only after visual identifi-
cation has been completed.
Four studies in this volume address aspects of the processing of com-
plex words in the auditory modality. The chapers by Wurm and Aycock
and Meunier and Segui investigate the processing of prefixed words, the
chapter by Baayen, McQueen, Dijkstra, and Schreuder studies suffixed
words. Wurm and Aycock provide further evidence for the relevance of the
conditional root uniqueness point in English. Meunier and Sequi show for
French that the stem of prefixed words is not processed as if there were no
preceding prefix. Baayen et al. document frequency effects for regular in-
flected words in Dutch. The chapter by Reid and Marslen-Wilson reports a
series of experiments in Polish documenting stem and affix priming, a suf-
fix-suffix interference effect similar to that observed in earlier work for
English, as well as the importance of semantic transparency.
The final four chapters complement the preceding studies in visual and
auditory lexical processing. The chapter by Dressier, Kilani-Schock, and
Klampfer addresses the question how children acquire morphology. The
chapter by Davis, Van Casteren, and Marslen-Wilson shows by means of
distributed connectionist modeling that data previously argued to support a
dual-route model can also be accounted for within a single-route frame-
work. The issue of the reliability of corpus-derived frequency data for ex-
perimental designs using frequency measures as diagnostics for decompo-
sition is discussed in the chapter by Baayen, Moscoso del Prado Martin,
Wurm, and Schreuder, in reply to some of the issues raised in the chapter
by Ford, Marslen-Wilson, and Davis. In the final chapter, Sandra and Fayol
address the role of morphological structure in written language production,
documenting for Dutch and French how frequency and proximity give rise
to spelling errors for homophonous regular verb forms.
The editors would like to thank Lanneke van Dreumel for editorial help.
Inflectional morphology and word meaning:
Orthogonal or co-implicative cognitive domains?
Aleksandar Kostic, Tanja Markovic and Aleksandar
Baucal
Typically, linguistics and psycholinguistics treat word meaning and inflectional
morphology as two domains that do not interact. The aim of the present study is to
evaluate this assumption empirically. The evaluation is based on a contrast between
two types of information probabilities for inflected noun forms in Serbian. One,
derived from all nouns of a particular gender (e.g., accusative singular in feminine
gender) is generic across words and thus free of meaning. This is a "type" prob-
ability. The other is related to the probability of a particular word in a particular
inflected form, and is a "token" probability (e.g., the accusative for the noun
meaning "chair"). Because it is specific to a particular noun, "token" probability is
related to its meaning. A pattern whereby the cognitive system is sensitive only to
information of type probability of inflected forms, would be consistent with the
claim that inflectional morphology and word meaning are cognitively independent
domains. By contrast, an outcome whereby the cognitive system is sensitive to
information of token probabilities would suggest that the two domains are not truly
independent.
In a lexical decision experiment, six inflected forms of feminine nouns were
presented. Nouns were divided into two groups that differed significantly in the
patterning of the amount of information carried by six inflected forms. This differ-
ence derived from differences in token probability only. Variability in the pattern-
ing of decision latencies to inflected forms was predicted by the average amount of
information for both type and token probabilities. This finding indicates that the
cognitive system is sensitive to "token" as well as type probabilities. Because the
"token" probability distribution is word specific, this outcome suggests that inflec-
tional morphology constitutes an intrinsic part of a word's meaning. Accordingly
the domains of word meaning and inflectional morphology should be treated as co-
implicative rather than independent cognitive domains.
1. Introduction
Linguists and psycholinguists traditionally treat word meaning and inflec-
tional morphology and syntax as two orthogonal aspects of language. Thus,
for example, a particular syntactic structure can encompass an infinite
number of grammatical sentences that differ in meaning. Likewise, the
2 Kostic, Markovic and Baucal
taxonomy and classification of grammatical forms as offered by descriptive
linguistics derive from a set of morphological alternations within a par-
ticular word class, irrespective of the meaning of the individual words that
belong to that class. The declension of a noun of a particular gender, for
example, is specified in terms of the morphological properties of cases
singular and plural, but not in terms of the meaning of the noun. Likewise,
conjugation specifies morphological properties of a verb including person
and grammatical number, irrespective of the meaning of the particular
verb. In general, the taxonomy and classification of grammatical forms of
different word types is based on generic morphological properties of a
particular word class or subclass within a particular word type (e.g., femi-
nine nouns).
With the above considerations in mind, inflectional morphology and
word meaning satisfy two formal requirements of separability. The first
requirement is the absence of an interaction between the two domains.
Inflectional alternations within a defined paradigm (say declension) do not
alter word meaning. The second requirement is that a set of inflectional
alternations should apply to all instances of a defined class. In the case of
nouns this means that all nouns can appear in all seven cases and two
grammatical numbers. In that respect, a formal description of inflectional
morphology is, in principle, distinct from the meaning of individual words.
Psycholinguists adopt a similar position. In addition to accepting a for-
mal distinction between these two aspects of language, psycholinguists
assume that inflectional morphology and word meaning are also distinct
cognitive domains that do not interact. Morphologically complex words are
described in terms of their base form and affixes. While the base form is
idiosyncratic to a particular word and relates to its meaning, the repertoire
of affixes is a generic property that pertains to a class of words. Stated
generally, psycholinguistic models that describe how we process morpho-
logically complex words embrace, with no exception, the distinction be-
tween inflectional morphology and word meaning (cf. Taft and Forster,
1975; 1976; Taft 1979,1981, 1994).
The aim of the present study is to evaluate the orthogonal status of the
two domains and reexamine the cognitive status of inflectional morphology
as independent from word meaning. The basis for such an evaluation dis-
tinguishes between two kinds of probabilities for inflected noun forms: the
probabilities of inflected forms per se, irrespective of word meaning (e.g. if
word is a feminine noun, what is the probability that it will end with a suf-
fix e), and the probabilities of inflected forms for an individual word (e.g.