Document Preview:
The Influence of Bilingualism on Cognitive Strategy and Cognitive Development Author(s): Sandra Ben-Zeev Reviewed work(s): Source: Child Development, Vol. 48, No. 3 (Sep., 1977), pp. 1009-1018 Published by: Blackwell Publishing on behalf of the Society for Research in Child Development Stable URL: http://www.jstor.org/stable/1128353 . Accessed: 11/11/2011 11:04 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Blackwell Publishing and Society for Research in Child Development are collaborating with JSTOR to digitize, preserve and extend access to Child Development. http://www.jstor.org The Influence of Bilingualism on Cognitive Strategy and Cognitive Development Sandra Ben-Zeev Bilingual Education Service Center BEN-ZEEV, SANDRA. The Influence of Bilingualism on Cognitive Strategy and Cognitive Development. CHILDD EVELOPMEN1T97, 7, 48, 1009-1018. The hypothesis is that mutual interference between the bilingual child’s 2 languages forces the child to develop particular coping strategies which in some ways accelerate cognitive development. The sample consisted of 2 groups of Hebrew-English bilingual children, 1 group tested in the United States and the other group tested in Israel, and 2 groups of monolingual Jewish children, with those tested in the United States speaking only English and those tested in Israel speaking only Hebrew. In all groups parent occupation and education level were similarly high. WISC IQ was estimated from 4 subtests-Similarities, Digit Span, Picture Completion, and Picture Arrangement-and…The Influence of Bilingualism on Cognitive Strategy and Cognitive Development
Author(s): Sandra Ben-Zeev
Reviewed work(s):
Source: Child Development, Vol. 48, No. 3 (Sep., 1977), pp. 1009-1018
Published by: Blackwell Publishing on behalf of the Society for Research in Child Development
Stable URL: http://www.jstor.org/stable/1128353 .
Accessed: 11/11/2011 11:04
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Blackwell Publishing and Society for Research in Child Development are collaborating with JSTOR to digitize,
preserve and extend access to Child Development.
http://www.jstor.orgThe Influence of Bilingualism on Cognitive
Strategy and Cognitive Development
Sandra Ben-Zeev
Bilingual Education Service Center
BEN-ZEEV, SANDRA. The Influence of Bilingualism on Cognitive Strategy and Cognitive Development.
CHILD DEVELOPMENT, 1977, 48, 1009-1018. The hypothesis is that mutual interference
between the bilingual child’s 2 languages forces the child to develop particular coping
strategies which in some ways accelerate cognitive development. The sample consisted of 2
groups of Hebrew-English bilingual children, 1 group tested in the United States and the
other group tested in Israel, and 2 groups of monolingual Jewish children, with those tested in
the United States speaking only English and those tested in Israel speaking only Hebrew. In all
groups parent occupation and education level were similarly high. WISC IQ was estimated
from 4 subtests-Similarities, Digit Span, Picture Completion, and Picture Arrangement-and
statistically controlled. In spite of lower vocabulary level, bilinguals showed more advanced
processing of verbal material, more discriminating perceptual distinctions, more propensity to
search for structure in perceptual situations, and more capacity to reorganize their perceptions
in response to feedback.
A child growing up in a bilingual environment
in which the two domains of language
usage and the two sets of interlocutors overlap
is continually faced with interference between
his two languages (Weinreich 1953). If there
are different strategies of language learning
(Bloom 1970), one might expect bilinguals to
adopt characteristic strategies due to the difficulties
they encounter. In the present paper
it is suggested that the bilingual child develops
a language strategy which helps him to resolve
interlingual interference. The strategy requires
increased scanning of verbal input-including
feedback from his own output-and vigilance
for inconsistency of rule use or failure of the
input to adhere to an interpretable rule system.
Even monolinguals need to keep different
sets of language rules separated from each
other, since a single language takes somewhat
different forms as it applies to different speech
registers and different domains of discourse.
But here the rule differences are relatively
superficial and a matter of usage, whereas the
bilingual child must learn to distinguish two
different language rule systems. The most direct
strategy to keep the two languages from
mutual interference is to pay particular attention
to the systematic aspects of both languages.
A major hypothesis of this study is that
highly bilingual children process syntactic rules
with special flexibility. The language learning
process requires any child to venture inferences,
some of which are incorrect because of
overgeneralization or idiosyncratic interpretation
(Ervin-Tripp 1973). The bilingual situation
is likely to encourage the child’s initiative
in rule inference. Having two referent symbols
for most referents, the bilingual child learns
early that words are not intrinsic but arbitrary.
Forced to have alternate means for the expression
of a given idea, he becomes aware early
that the relation between an idea and its
means of expression is problematic. The bilingual
child has experienced himself making
many mistakes concerning which language is
required in a given situation. He has probably
also made more mistakes in inferring syntactic
rules than have other children, given his more
limited exposure time to each language. As
a result of these factors, the bilingual child is
more than ordinarily aware that language
learning is hypothesis testing and that it reThis
research was submitted in partial fulfillment of the Ph.D. degree requirements at the
University of Chicago, Committee on Human Development. The author wishes to thank advisors
Carol Feldman, Wilbur Hass, Robert LeVine, and Richard Hodges. Thanks are also due to Aliza
Fine, Shoshanna Kaniel, and Esther Mandelker, and to the principals, teachers, and students of
the various Hebrew day schools and suburban schools in which testing was done. Address reprint
requests to Sandra Ben-Zeev, Bilingual Education Service Center, 500 South Dwyer
Avenue, Arlington Heights, Illinois 60005.
[Child Development, 1977, 48, 1009-1018. @ 1977 by the Society for Research in Child Development, Inc. All rights
reserved.] 1010 Child Development
quires effort. He is more prepared to reject
syntactic hypotheses when feedback signals
that this is necessary and to switch quickly to
alternative interpretations or to a search for
new interpretations.
A second hypothesis concerns possible
effects of bilingualism on semantic processing.
In association tasks with monolingual children
only the older children begin to give many
associations which place the stimulus word
into a sort of system, that is, associations based
on similarity, contrast, or superordination. We
do not yet know the means by which such
categorical associations develop. If they develop
strictly through induction from experience
with particular words in varied contexts,
then bilinguals cannot be expected to develop
categorical associations sooner. On the contrary,
bilinguals have relatively limited experience
with particular words. The words they
hear are divided between two languages, so
there is less opportunity for them to hear particular
words frequently. This vocabulary inexperience
might even retard the appearance
of categorical associations, if this is the mechanism
by which they come to occur.
On the other hand, we can expect pre- cocious appearance of categorical associations
from bilingual children if the ability to associate
to words categorically is enhanced by the frequent experience of conflict in the interpretation
of word meaning and sentence
structure.
The latter prediction, that is, precocious
development of categorical associations in the
bilinguals, must be tempered by the expec- tation that the bilinguals will also have a
vocabulary deficit. Limited word knowledge could hinder production of categorical associ- ations in spite of a strategy oriented in that
direction on the part of the child.
A third hypothesis is that the bilingual child’s awareness of the systematic basis of
language structure generalizes to increased
ability to analyze nonverbal structures.
There is empirical evidence of the facilitating
influence of bilingualism on cognition.
Peal and Lambert (1962) concluded that their
10-year-old French-English bilinguals per- formed better on tests of intellectual reasoning
than did comparable monolinguals. They
also found that the abilities of the bilinguals
were more differentiated. This study, however,
failed to control for native ability and there
were indications that the bilinguals in the study
may have been more intelligent at the outset,
since they were more advanced in grade level
over their comparison group of the same age.
Another problem was that the criterion for
whether a child was bilingual or monolingual
did not involve a sharp bifurcation and also
depended heavily on vocabulary score. If one
effect of bilingualism is to retard vocabulary,
as is proposed here, then it is biasing to use
vocabulary in the less well known language
as a criterion for bilingualism.
Anisfeld (Note 1) reanalyzed the Peal
and Lambert data to control for intelligence
by eliminating subjects until groups could be
matched for Kuhlman-Anderson IQ. She found
that bilinguals still differed in that they.excelled
on the Ravens Progressive Matrices Test,
a test considered high on “g.” Anisfeld’s procedures
do not correct for the questionable
selection criteria in the original study.
lanco-Worrall’s results (Ianco-Worrall
1972; Worrall, Note 2) comparing AfrikaansEnglish
bilinguals of ages 4-9 with comparable
monolinguals were different. She did not find
that her bilinguals were better on tests of
ability to deal flexibly with structures or on
tests of operational thinking. She did find,
however, that her bilinguals were more oriented
toward meaningful stimuli and that they
were better able to conceive of the word as
an arbitrary symbol. Her failure to find linguistic
differences in structural flexibility could
reflect her criterion for bilingualism. She did
not include a test of translation ability, although translation ability is the most direct indication
that the child has been able to deal with the
interlingual conflict. Her South African IQ
test is not described in detail, so its effectiveness
as a control is not clear. In addition,
for most of her bilinguals each parent was associated
exclusively and differentiatingly with
just one of the two languages. The complete
predictability of interlocutor may possibly diminish
the essential interlingual conflict which
is presumed to be the basic motivating factor
in the cognitive situation of the bilingual child.
For the bilinguals used in the study to be
described below, the language in which any
particular interlocutor might speak to the child
was less predictable than in the Ianco-Worrall
study. Intelligence was controlled by standardized
tests. Vocabulary level was not used as
a criterion for determination of bilingualism.
The cognitive dependent variables were selected
in accordance with the hypotheses described
above. Sandra Ben-Zeev 1011
Method
SUBJECTS
There were 96 children divided into four
sample groups. One group of Hebrew-Englishspeaking
bilingu|al children was tested in the
United States, and a second group was tested
in Israel. There were two corresponding monolingual
groups, one tested in the United States
speaking only English and one tested in Israel
speaking only Hebrew. The age range was 5-4
to 8-6 years (mean = 7-0 years). There was
an approximately equal sex distribution.
All subjects were middle-class Jews. Occupation
of fathers was academic or professional.
According to the parents, average age
of onset of bilingualism for the United States
sample was 22, years and, for the Israeli bilingual
sample, 3,2 years.Most of the United States bilinguals were
children of Israeli parents living temporarily in the United States. Most attended Hebrew
day school in which both Hebrew and English
are spoken and taught. Both languages are
heard by them in other environments as well.
Testing was done at three Hebrew day schools
in the Chicago area and one in Brooklyn, and
also in the children’s homes. Testing of United
States monolinguals was in a suburban Chicago
public school or in their homes.
The bilinguals tested in Israel were immigrants
from the United States or South
Africa. They were tested in their homes, which
were in suburbs of Jerusalem and Tel Aviv.
Among peers in these immigrant neighbor- hoods both languages are common, although school classes are exclusively in Hebrew. The
Hebrew monolinguals were tested in a Tel Aviv
school selected after consultation with the
Ministry of Education because of its location
in an established neighborhood of families of
European background with high occupational attainment.
PROCEDURE
All children were tested individually and
with tests in their appropriate language. Except
for the Israeli bilinguals, who were
tested by an Israeli psychologist, the investigator
administered the tests. The order of administration
of the tests was fixed.
Screening was by means of a 15-item
translation test in which sentences to be translated
alternated from one language to the
other. Vocabulary was kept very simple, but
sentence translation required changes in word
order, construction, and sometimes idiomatic
adaptation. If a potential monolingual subject
showed understanding of any part of the sentences
in the second language, or if he spoke
another foreign language, he was eliminated
from the sample. Potential bilinguals who
failed to translate more than two sentences or
whose translations were forced and literal were
also eliminated, but this was rare.
Two scores determined the degree of
bilingualism among subjects selected for inclusion
in the bilingual sample, one derived
from the translation test and another from an
18-item bilingual word association test. These
two were found to be the most representative