Multi-cognition in L2 users:

More evidence from an object categorization task

by Japanese elementary school children*

MURAHATA, Goro Department of International Studies, Kochi University

Abstract

This study examines whether early limited exposure to a second language (L2) influences cognition of child L2 users. Research evidence, particularly from a ‘bilingual cognition’ perspective, has shown that knowledge of a L2 has effects on adult L2 users’ cognitive dispositions as well as their first language. However, there remains an issue of the timing of L2 learning influence on cognition. A picture categorization task was given to three groups of Japanese elementary school children with different amounts of English exposure. The purpose was to see how much exposure is necessary for L2 users to modify their object categorization preferences. It was found that though Japanese child L2 users fundamentally categorized objects based on the syntagmatic-thematic relation, the proportion of paradigmatic-taxonomic categorization steadily increased according to the cumulative amount of exposure. Response times also became faster. These findings accorded with those found in previous categorization studies with advanced adult L2 users. It was concluded that the findings in the present study provide further evidence of the early emergence of ‘multi-cognition’, a state of mind with more than one cognitive apparatus in L2 children. This reflects the very early effects of L2 exposure on cognitive dispositions.

1. Introduction

There is now an increasing body of research evidence that shows second language (L2)

users’ linguistic competence qualitatively differs in some way from that of both the first

language (L1) and the target language monolinguals (Tokumaru, 2002; Cook et. al., 2003;

Murphy & Pine, 2003). This unique competence, a compound state of mind with more than one

language (i.e., their L1 and interlanguage), has been called ‘multi-competence’ after Cook

�������������������������

* This study was supported in part by the Japan Ministry of Education, Culture, Sports, Science and Technology Grant-in-Aid for Scientific Research (#20520549).

(1991). Second language acqu

only L2 users’ linguistic com

categorizations (Athanasopou

distinctions (Cook et al, 2

thematic-taxonomic relations

either L1 or L2 monolinguals

for a comprehensive literature

L2 users’ unique cognition ‘m

mind with L1-established co

evidence for multi-competenc

the results of those studies wh

known about the influence of L

dispositions.

Fig

Studies on cognitive and

language (L1) on cognition (C

Choi et al., 1999). For exampl

of spatial terms in and kkita i

sensitivity to language-specifi

differ in concepts of space in t

relation regardless of containm

resulting in containment (put in

when they saw scenes with t

different and language-specif

uisition (SLA) and other related research has a

mpetence but also their cognitive dispositi

ulos et al., 2004; Athanasopoulos, 2009

2006; Kasai & Takahashi, 2005) and

(Murahata & Murahata, 2007) somehow d

(see Jarvis & Pavlenko (2008) and Cook &

e review). As Figure 1 below shows, Murahat

multi-cognition’, which referred to a compoun

ognition and ‘inter-cognition’ (Cook, 2008

ce and multi-cognition, however, has been ex

hich used advanced adult L2 users as particip

L2 learning on incipient child L2 users’ compe

gure 1. Multi-cognition (Murahata, 2010)

developmental psychology have shown stro

Choi and Bowerman, 1991; Choi, 1997; Ima

le, Choi et al. (1999) examined young childre

n two languages (English and Korean) to ex

c concepts with a cross-linguistic design. The

that the Korean verb kkita refers to actions re

ment or support while English makes a distinct

n) versus support or surface attachment (put o

those terms even 18-to-23 month years old

fic aspects of the scenes. Choi et al. (199

also shown that not

ions such as color

9), shape-substance

categorizations of

differ from those of

Bassetti (to appear)

ta (2010) called this

nd state of L2 users’

8, p. 21). Existing

xclusively based on

pants. Hence little is

etence and cognitive

ong early effects of

ai & Gentner, 1997;

en’s comprehension

xplore the timing of

ese terms strikingly

esulting in a tight-fit

ion between actions

on). They found that

d children gazed at

99) concluded that

�������� �� � ����������

children’s comprehension of those terms was guided by language-specific concepts realized by

their L1 (p. 263).

This early sensitivity to the relationship between language and concepts strongly implies a

latent possibility of early L2 learning influence on L2 child users’ cognition. Thus, the present

study is an attempt to find further evidence for ‘multi-cognition’ in child L2 users of extremely

limited L2 exposure with special focus on a cognitive process ‘categorization’. ‘Categorization’

was chosen as focus of research because there is nothing more basic than ‘categorization’ to our

cognitive activities (Lakoff, 1987, pp. 5-6). Then, as Markman (1989, p. 11) says, it is a

fundamental cognitive process, involved in one way or another in almost any intellectual

endeavor. For instance, in identifying objects, in perceiving two things as similar, in acquiring

and using language, in all of these cognitive processes, categorization plays a major role.

2. Review of relevant studies

2.1 Terminology

Studies related to ‘categorization’, according to Smiley and Brown (1979), there are two

distinctive types of categorization. One type is based on how things are used or related to the

real world. Things in this category have a linear character. For example, ‘monkey’ and ‘banana’

are linearly related because of their ‘thematic’ relation: for example, ‘Monkeys love bananas.’

The other type is based on how things are associated or related with each other. Thus, things in

this category have the common character: for example, ‘monkey’ and ‘panda’ are related

categorically because they are both animals. Smiley and Brown (1979) called this the

‘taxonomic’ relation.

However, the pairs of syntagmatic-paradigmatic and thematic-taxonomic have been used

rather interchangeably in difference contexts of linguistics and SLA research literature. For

example, according to linguists such as de Saussure (1916), in categorizing or relating objects

around us as part of our mental activity, we use two different kinds of linguistic relations. One

has been called the ‘syntagmatic’ relation, which has the linear character of linguistic items

such as Dieu est bon (‘God is good’) (de Saussure, 1916, p. 170). The other ‘associative’

relation does not address such a linear character but rather a common character of linguistic

items that form groups of variously related members. These terms have been used in object

categorization or word association research (Ervin, 1961; Yoneoka, 1987; Wolter, 2001).

Other different terms have also been used to distinguish between pairs of dichotomized

relations referred to in different research contexts such as:

- heterogeneous-by-part-of-speech/homogeneous-by-part-of-speech (Brown & Berko, 1960)

- relational-contextual/descriptive-analytic, descriptive-whole/inferential-category (Chiu, 1972)

- functional/conceptual (Moran, Mefford & Kimble, 1964)

������������ �� �� ��������������� ���

- iconic, enactive/functional, logical (Moran, 1973)

- thematic/taxonomic (Ji, Zhang & Nisbett, 2002; Smiley & Brown, 1979).

All of these pairs, however, share a general idea that the former denotes the organization of

objects based on how they are used or their relationships with the real world (functional,

thematic, complementary, syntagmatic relations) while the latter denotes the organization of

objects based on how they are associated or their relationships with each other (associative,

taxonomic, paradigmatic relations) (Smiley & Brown, 1979, p. 250). Thus, in this study the

compound terms ‘SynThem’ and ‘ParaTax’ respectively are used to subsume diverse

terminology in the literature (Murahata, to appear).

2.2 Background to concept research: categorizing objects and word association

There have been two different lines of concept research in the field of developmental

psychology: categorizing objects and word association. Though research studies along these

lines have been conducted in different arenas with different research methods, they keep track

of the same concern, that is, how words and concepts are organized in the mind or in the mental

lexicon (Smiley & Brown, 1979, p. 250). For example, cognitive psychologists from the word

association tradition hold the view that associative responses to stimulus words reflect the

functioning of concept and thought processes of an individual (Zareva, 2007, p. 124). In the

sections that follow we will briefly review some of the main findings in those concept research

studies which seem relevant to the present study.

2.2.1 Categorizing objects by monolinguals and L2 users

As for categorizing objects by monolinguals, it has been shown that young children prefer

categorizations based on the SynThem relation while mature monolinguals prefer

categorizations more based on the ParaTax relation (Ervin, 1961; Smiley & Brown, 1979). This

change of categorization preference has been called the thematic-taxonomic

(SynThem-ParaTax) shift (Entwisle et al., 1964). For example, Smiley and Brown (1979)

explored the shift by using a picture matching task where participants were given a standard

picture and two other pictures, either thematically related or taxonomically related, and were

asked which of the two pictures went best with the standard word. They found that the higher

the participants’ ages, the clearer the shift from SynThem to ParaTax categorization preferences

they showed.

Chiu (1972) conducted a cross-linguistic study to see how monolingual children of

different linguistic backgrounds categorize objects. They were given an object selection task

where Chinese- and English-speaking participants were asked to select any two out of the three

objects in a set which were alike or went together. It was found that Chinese children preferred

to select objects on the basis of the SynThem relation, while American children on the basis of

�������� �� � ����������

the ParaTax relation. This finding was of importance in that the ParaTax preference for

categorizing objects cannot be universally manifested in all groups of older ages as often

depicted as the SynThem-ParaTax shift, and also that the language of children could possibly

influence their categorization preference patterns.

From a socio-psychological perspective, Ji et al. (2002) examined how adults from

different linguistic (cultural) backgrounds categorize objects in order to see a linguistic

(cultural) influence on categorizing objects. They presented monolingual English and Chinese

college students with triads of SynThem and ParaTax relations (‘Cow’ [the standard], ‘Milk’

[the SynThem relation to the standard], ‘Panda’ [the ParaTax relation to the standard]) and

asked them to decide which went best with the standard. The results revealed that English

monolinguals showed a marked preference for categorization on the basis of ParaTax, while

Chinese monolinguals showed a preference for categorization on the basis of SynThem. This

finding was well in accordance with the finding by Chiu (1972) for a possible linguistic

influence on object categorization.

Murahata and Murahata (2007) conducted a cross-linguistic experiment by replicating the

data collecting method used in Ji et al. (2002). Participants were Japanese adult users of English

with different levels of English proficiency and English monolinguals. They were asked to

judge how strong the relations between ParaTax objects and SynThem objects were. They

found that English monolinguals judged ParaTax relations to be more strongly than Japanese

users of English, and that advanced Japanese L2 users judged ParaTax relations more strongly

than less advanced Japanese L2 users.

2.2.2 Word association by monolinguals and L2 users

The other line of concept research concerns what word come to mind when people are

given a word as stimulus. Wakabayashi (1973) examined word associations by Japanese

monolinguals and English monolinguals. He found a substantial difference between the two

groups. While the English monolinguals overwhelmingly produced ParaTax associations, the

Japanese counterparts produced much more SynThem associations than English monolinguals.

Though Wakabayashi himself did not suggest a linguistic influence on word association

preferences, his result coincided with the findings in the case of categorizing objects.

Reviewing a series of his own and associates’ research into word associations by Japanese

speakers of different ages, Kashu (1973) found no straight SynThem-ParaTax shift in Japanese

speakers as often been observed among English speakers such as Smiley and Brown (1979).

Japanese speakers were more likely to make SynThem associations across all age groups.

Moreover, the proportion of ParaTax associations further decreased as their ages went higher.

This finding was replicated in several word association experiments by Araki (1995).

Moran (1973) longitudinally examined cross-linguistic differences in the growth of

cognitive dictionaries between young Japanese and American children by using a free word

������������ �� �� ��������������� ���

association task. He found a marked shift from SynThem to ParaTax associations among

American children as they grew older. This clearly supported Ervin’s (1961) finding that

English-speaking children of increasing ages showed marked increases in ParaTax responses.

On the other hand, a reverse pattern of association preferences was found in Japanese children.

They showed an overwhelming preference for SynThem associations with increasing

frequency as they grew older. Moran’s (1973) findings show that the preference changes the

SynThem-ParaTax shift for American children and no preference shift (i.e., the

SynThem-SynThem shift) for Japanese children. This suggests a theoretical consequence for

the growth of monolingual children’s cognitive dictionaries that there may be a universal

preference modification pattern of word associations in young children. In other words,

children first organize words in their mental lexicon based on the SynThem relation and later

modify or stay with its organization as a consequence of acquiring their L1.

As for word association by bilinguals, Yoneoka (1987) tried to replicate Moran’s (1973)

finding of the SynThem-ParaTax shift among English and Japanese speakers of different ages.

She found a clear shift in the association patterns among English speakers, from SynThem in

young children to ParaTax in adults. On the other hand, Japanese speakers’ responses did not

show a general SynThem-ParaTax shift. This supported Moran’s (1973) finding.

The findings in the studies above related to concept development or change, either by

monolinguals or bilinguals, hold two implications. Firstly, young children overwhelmingly

categorize lexical items or concepts in the mind based on the SynThem relation. Then later as

they grow older they seem to modify their categorical preference patterns to either the ParaTax

or SynThem relation depending on the language they are exposed to as their L1. In the case of

Japanese monolinguals, they first categorize objects based on the SynThem relation and even as

they grow older they are still likely to categorize objects in fundamentally the same way based

on the SynThem relation. One the other hand, English monolinguals come to categorize objects

based more on the ParaTax relation as they grow older. Secondly, this L1-influenced

categorization preference could again be modified by exposure to another language. For

example, Japanese adult users of English came to categorize objects based more on the ParaTax

relation as their L2 proficiency went higher.

3. The study

3.1 Research questions

There are three research questions in the present study. The first two are related to cognitive

processes, the object categorization either based on the SynThem or ParaTa relation, and the

third is related to the period of time taken for each response and cognitive processing:

�������� �� � ����������

1) Do Japanese child L2 users basically categorize objects based on the SynThem

relation as often found in previous studies?

2) Does L2 exposure influence the way incipient child L2 users categorize objects?

3) Are there any differences in response times between groups of different amounts of

L2 exposure?

The first question is to see if the SynThem categorization preference can also be found in child

L2 users. The second question concerns whether or not as English exposure increases such

children shift their categorical preferences, say, categorize objects based more on the ParaTax

relation as English monolinguals do. The third research question asks about any differences in

response times between groups of different amounts of L2 exposure. Hunt and Agnoli (1991)

hold that examining ‘current practice’ is one of the best ways to determine cross-linguistic

effects on cognition as psychologists and psycholinguists often study differences in the time

required to perform a certain task. Moreover, as Cook (1997) concluded after reviewing several

empirical studies, “L2 users are less effective in speed of processing” (p. 289) languages, either

L1 or L2. Thus we can predict that children with more L2 exposure respond slower than those

with little L2 exposure.

3.2 Participants

Participants of the present study were 76 children, 5th and 6th graders (10 to 11 year-olds),

from two elementary schools, K-ES and S-ES. Both schools are located in central cities of a

Western prefecture in Japan. Thus, little difference exists in socio-economic backgrounds

between the children from the two elementary schools. However, at the time of the experiment,

the children at K-ES received no formal English teaching. On the other hand, the children at

S-ES had been receiving two hours a week English teaching from the 3rd grade. In English

class children engage in a variety of language activities by Japanese teachers of English (JTEs),

who received formal teacher training at a national university and are qualified as English

teachers. They try to use English most of the time during the class. S-ES also has a full-time

ALT who helps those JTEs by team teaching. Some children at S-ES had also learned English

outside of school for more than one year.

The participants were divided into three groups according to the amount of English

exposure. The Non-EX group (K-ES children) consists of 26 children with no formal English

learning at school and outside of school; the EX group (S-ES children), 32 with two hours a

week English learning at school but no English learning or less-than-one-year English learning

outside of school; the EX-EX group (S-ES children), 18 with two hours a week English

learning at school and more-than-a-year English learning outside of school.

������������ �� �� ��������������� ���

3.3 The task

A picture categorization

(2002) was given to the partic

goods and Vehicles, and four tr

trial, three pictures were given

ParaTax stimulus. Figure 2 b

standard stimulus, the banana

the ParaTax stimulus.

Table 1 shows all the stimuli f

Table 1Entity Types

Animals

Daily Goods

Vehicles

(or matching) task after Smiley & Brown (

cipants. There were three different entity type

rials for each entity type, which make up 12 tr

n as stimuli: the standard stimulus, the SynThe

below shows an Animal entity trial. The mo

bottom left the SynThem stimulus, and the

Figure 2. An animal entity trial

for the 12 trials used in the present study.

1. Twelve trials used as stimuli in the taskStandard SynThem ParaTax

Monkey Banana Panda

Cow Milk Sheep

Camel Horse Desert

Fish Crab Fishing-ro

Pencil Notebook Pen

Fork Salad Chopstick

Ring Wedding Necklace

Piano Notes Violin

Train Railroad Bus

Truck Logs Taxi

Car Road Bicycle

Airplane Clouds Helicopte

(1979) and Ji et al.

es: Animals, Dailey

rials in total. In each

em stimulus, and the

onkey on top is the

panda bottom right

x

od

ks

e

e

er

�������� �� � ����������

These 12 trials were random

participant was given the stan

pictures (the SynThem stimu

stimulus. Then they were to d

deciding, the participant was r

pointing to either of the alterna

Figure

The time-recorder beside the

responses, that is, from the tim

when he or she pointed to one o

Still Green) was used to measu

To find the answer to the

basically categorize objects ba

the means of the two respons

were calculated. Figure 4 show

the two response types. The m

trials]) while the mean of the

(73.3% [2.93/4.00]) of the Ja

SynThem relation. A t-test w

response types was statistically

the answer to the first research

prefer to categorize objects ba

mly given to the participants. As shown in F

ndard picture (stimulus) alone and then wa

ulus and the ParaTax stimulus) in circles b

decide which stimulus best went with the sta

required to say in Japanese, ‘Kocchi (This one

ative stimuli.

3. The ‘object categorization’ experiment

experimenter measured the length of time ta

me when the participant saw the two pictures i

of the pictures. A time-keeping software called

ure the length of time.

4. Results and discussion

first research question, that is, whether Japa

ased on the SynThem relation as often found

e types (SynThem and ParaTax) for all the p

ws its results. There was a large difference be

ean of the SynThem response was 2.93 (SD=

ParaTax response was 1.07 (SD=1.231). N

apanese children’s categorization responses

was performed on the data and the differenc

y significant (t (227) = 11.429, p < .000 (two-

h question was positive in that Japanese childr

ased on the SynThem relation.

Figure 3, first each

as shown two more

below the standard

andard stimulus. On

e)’, with their finger

aken for each of the

n circles to the time

d SGwatch (© 2005

anese child L2 users

in previous studies,

participants (N=72)

etween the means of

1.233; Max=4.00 [4

early three quarters

were based on the

ce between the two

-tailed)). Therefore,

ren overwhelmingly

������������ �� �� ��������������� ���

Figure 4. Means of SynThem and ParaTax responses for all the participants (N=76)

Secondly, the means of the two response types for each of the three groups were calculated

to examine if there were any differences between groups of different amounts of L2 exposure.

That is, is it possible to find any L2 exposure effects on categorization preferences in child L2

users? Figure 5 shows group means of SnyThem and ParaTax responses. As you can see, the

proportions of SynThem responses (grey bars) constantly declined while those of ParaTax

responses (black bars) rose. Particularly important was the large gap between the non-EX group

and the EX and EX-EX groups. The means of the ParaTax relation for the latter groups (1.15

and 1.61) was nearly twice or three times as large as that for the non-EX group (0.59). Statistical

analyses (ANOVA) also showed significant differences between the two categorical types (F (1,

73) = 59.532, p < .000) and the interaction between group and categorical type (F (2, 73) =

2.93

1.07

SymThem

ParaTax

3.41

2.84

2.39

0.59

1.15

1.61

0

0.5

1

1.5

2

2.5

3

3.5

4

Non-EX EX EX-EX

SynThemParaTax

Figure 5. Group means of SynThem and ParaTax responses�

�������� �� � ����������

6.103, p < .004). These results are evidently an indication of L2 learning influence on child L2

users’ objects categorization. That is, as a consequence of learning a L2, Japanese children

gradually shifted their object categorization preference patterns from the SynThem relation to

the ParaTax relation.

The third research question was whether or not there are any differences in response times

between groups of different amounts of L2 exposure. Behind this question is a psychological

premise that one of the best ways to determine cross-linguistic effects on cognition is to

measure differences in the time required to perform a certain task (Hunt & Agnoli, 1991). Table

2 and Figure 6 show group means of response times (seconds) for each of the three groups.

Table 2. Group means (SDs) of response times (seconds)

Non-EX EX EX-EX

(N=26) (N=32) (N=18)

Means 2.331 1.982 1.813

SD 1.816 1.023 0.957

Figure 6. Group means of response times (seconds)

As you can see in Figure 6, there are clear differences in response times between groups. To our

great surprise, the response times became faster as L2 exposure increased (2.331 seconds for

the Non-EX group, 1.982 seconds for the EX group, 1.813 seconds for the EX-EX group)

contrary to our prediction. One-way ANOVA was performed on the data and the result was

statistically significant (F (2, 73) = 10.760, p < .000). Furthermore, t-tests were performed on

the data to see differences between groups. As the t-test (two-tailed) values in Table 3 show, all

of the pairs between the three groups were statistically significant, indicating clear differences

2.331

1.982

1.813

1.6

1.7

1.8

1.9

2

2.1

2.2

2.3

2.4

Non-EX EX EX-EX

������������ �� �� ��������������� ���

in response times between groups. This indicates that L2 learning not only influences child L2

users’ cognitive processes, but also influences their current practice, that is, their on-line

cognitive processing.

Table 3. T-test (two-tailed) values of response times between groups

EX EX-EX

Non-EX t (56) = 3.027, p < .003 t (42) = 4.257, p < .000

EX t (48) = 2.026, p < .043

The main findings of the present study can be summarized as follows. Firstly, Japanese

child incipient L2 users largely categorized objects on the SynThem relation. This finding is in

support of those found in the previous concept and word association studies such as

Wakabayashi (1973), Kashu (1973), Yoneoka (1987), and Murahata and Murahata (2007).

Secondly, as L2 exposure increased, they modified, though slightly, their categorical

preferences and came to categorize objects based more on the ParaTax relation as adult English

monolinguals do (Smiley & Brown, 1979).

Thirdly, there were clear differences in response times between groups of different

amounts of L2 exposure. Existing research evidence has shown, as Cook (1997, pp. 281-283)

summarized, that bilinguals are less effective than monolinguals in the speed of linguistic

processing such as anaphora comprehension tasks, acceptability judgments, decoding tasks

such as ‘Mark the third letter from the left’ and list recognition and lexical decision tasks.

Contrary to these findings in linguistic processing tasks, in the present study the child L2 users

with more L2 exposure responded faster than those with less L2 exposure in an object

categorization, that is, non-linguistic, task. Though what was attributable to this ‘acceleration

effect’ of cognitive processing in a non-linguistic task was not clear, this finding implies an

additive aspect of bilingualism which maintains that knowing another language other than the

L1 extends our cognitive capabilities in some way. In other words, learning “a second language

increases the normal capacity of the individual and, so, confers a benefit rather than creates a

problem” (Cook, 1997, p.289).

5. Conclusion

This study was an attempt to explore whether L2 learning influences cognitive processes

and processing of child L2 users who are learning English in a very limited L2 exposure context.

The findings depicted in detail above evidently showed it does. As the amount of L2 exposure

increased, child L2 users not so dramatically but definitely shifted their categorical preferences

toward those of adult English monolinguals and their processing speed also became faster.

Accordingly, these findings certainly provide further evidence for a L2 users’ compound state

�������� �� � ����������

of mind with more than one cognitive apparatus, that is, multi-cognition.

What is the most striking finding was that even two-hour-a-week L2 exposure made a

significant difference in children’s cognitive dispositions. If this very early influence of L2

learning on children’s cognition is to be recapitulated in future studies, this could possibly form

a presumption for SLA research particularly from a bilingual cognition perspective. That is, the

threshold of cognitive modification as a consequence of L2 learning can be manifested far

earlier than has been expected.

There remain some important issues that we should discuss on the relationship between L2

learning and L2 users’ cognition. For example, why do L2 users come to respond faster in a

non-linguistic task as their L2 exposure increases? Is it also true for on-line processing of other

cognitive tasks? Secondly, how do an individual’s cognitive processes and processing change

over time? All of the studies related to this kind of research to date have been conducted

cross-sectionally. Conducting longitudinal studies is of particular importance to find out about

real shifts or changes of cognition in L2 users.

References

Araki, N. (1995). Shogakusei no Gengorensou nikansuru Shinrigakuteki Kenkyu (A

psychologial study of word association by elementary school students). Tokyo: Kazama

Shobo.

Athanasopoulos, P. (2006). Effects of the grammatical representation of number on cognition

in bilinguals. Bilingualism: Language and cognition, 9, 89-96.

Athanasopoulos, P. (2009). Cognitive representation of color in bilinguals: The case of Greek

blues. Bilingualism: Language and Cognition, 12, 83-95.

Athanasopoulos, P., Sasaki, M., & Cook, V. J. (2004). Do bilinguals think differently from

monolinguals? Evidence from colour categorization by speakers of Japanese. Paper

presented at the 14th European Second Language Association Conference. San Sebastian,

Spain.

Brown, R. & Berko, J. (1960). Word association and the acquisition of grammar. Child

Development, 31, 1-14.

Chiu, L-H. (1972). A cross-cultural comparison of cognitive styles in Chinese and American

children. International Journal of Psychology, 7, 235-242.

Choi, S. (1997). Language-specific input and early semantic development: evidence from

children learning Korean. In D. I. Slobin (ed.) The Cross-linguistic study of language

������������ �� �� ��������������� ���

acquisition: vol. 5. Expanding the contexts (pp. 41-133). Hillsdale, NJ: Lawrence Erlbaum

Associates.

Choi, S. & Bowerman, M. (1991). Learning to express motion events in English and Korean:

the influence of language-specific lexicalization patterns. Cognition, 41, 83-121.

Choi, S., McDonough, L., Bowerman, L., & Mandler, J. M. (1999). Early sensitivity to

language-specific spatial categories in English and Korean. Cognitive Development, 14,

241-268.

Cook, V. J. (1991). The poverty-of-stimulus argument and multi-competence. Second

Language Research, 7, 103-117.

Cook, V. J. (1997). The consequences of bilingualism for cognitive processing. In A. M. B. de

Groot & J. F. Kroll (eds.) Tutorials in bilingualism: psychological perspectives (pp.

279-299). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Cook, V. J., Iarossi, E., Stellakis, N. & Tokumaru, Y. (2003). Effects of the L2 on the syntactic

processing of the L1. In Vivian Cook (ed.) Effects of the second language on the first (pp.

193-213). Clevedon, UK: Multilingual Matters.

Cook, V. J. (2008). Multi-competence: black hole or wormhole for second language acquisition

research? In Z. H. Han (ed.) Understanding second language process (pp. 16-26).

Clevedon, UK: Multilingual Matters.

Cook, V. J., Bassetti, B., Kasai, C., Sasaki, M., & Takahashi, J. A. (2006). Do bilinguals have

different concepts? The case of shape and material in Japanese L2 users of English.

International Journal of Bilingualism, 10, 137-152.

Cook, V. J., & Bassetti, B. (to appear). Language and Bilingual Cognition. London:

Psychology Press.

de Saussure, F. (1916). Course in general linguistics. C. Bally & A. Sechehaye (eds.) R. Harris

(Translated and Annotated) (1983). Chicago and La Salle, IL: Open Court.

Entwisle, D. R., Forsyth, D. F., & Muuss, R. (1964). The syntactic-paradigmatic shift in

children's word associations. Journal of Verbal Learning and Verbal Behavior, 3, 19-29.

Ervin, S M. (1961). Changes with age in the verbal determinants of word-association. American

Journal of Psychology, 74, 361-372.

Hunt, E. & Agnoli, F. (1991). The Whorfian hypothesis: A cognitive psychology perspective.

Psychological Review, 98, 377-389.

Imai, M. & Gentner, D. (1997). A cross-linguistic study of early word meaning: universal

ontology and linguistic influence. Cognition, 62, 169-200.

Jarvis, S. & Pavlenko, A. (2008). Crosslinguistic influence in language and cognition. New

York: Routledge.

�������� �� � ����������

Ji, L., Zhang, Z., & Nisbett, R. E. (2002). Culture, language and categorization. Unpublished

manuscript. Queens University, Kingston, Ontario. Quoted in R. E. Nisbett (2003). The

geography of thought: How Asians and Westerns think differently ... and Why. New York:

Free Press.

Kasai, C., & Takahashi, J. A. (2005). Effects of L2 on the cognitive decision of shapes and

mass: A report of the cases of Japanese University students. Bulletin of the Faculty of

Regional Studies, Gifu University, 17, 21-33.

Kashu, K. (1973). Nihongo no rensou mekanizumu (Mechanism of word association among

Japanese speakers). Jinbunronkyu (Humanities Review: Journal of the Literary

Association of Kwansei Gaguin University), 23, 1-22.

Lakoff, G. (1987). Women, fire, and dangerous things. Chicago: University of Chicago Press.

Markman, E. M. (1989). Categorization and naming in children. Cambridge, MA: The MIT

Press.

Moran, L. J. (1973). Comparative growth of Japanese and north American cognitive

dictionaries. Child Development, 44, 862-865.

Moran, L. J., Mefford, R. B., & Kimble, J. P. (1964). Idiodynamic sets in word associations.

Psychological Monographs, 78, (2, Whole No. 579).

Murahata, G. (2010). Shogakko Eigo ga Jidou no Ninchi ni Ataeru Eikyo (Influence of English

learning on elementary school children’s cognition). Paper presented at the 36th Annual

Conference of The Japan Society of English Language Education (Kansai University,

Osaka).

Murahata, G. (to appear). Conceptual preference modification in incipient Japanese-English

bilingual children: the case of categorizing objects on the basis of 'SynThem' and 'ParaTax'

relations. SELES (Shikoku English Language Education Society) Journal, 30.

Murahata, G. & Murahata, Y. (2007). Does L2 learning influence cognition?: evidence from

categorical and thematic organization of objects. Research Reports of the Department of

International Studies (The Department of International Studies, Faculty of Humanities and

Economics, Kochi University), 8, 17-27.

Murahata, G. & Murahata, Y. (2008). V. Cook’s multi-competence and its consequences for

SLA research and L2 pedagogy. Research Reports of the Department of International

Studies (The Department of International Studies, Faculty of Humanities and Economics,

Kochi University), 9, 109-128.

Murphy, V. A., & Pine, K. J. (2003). L2 influence on L1 linguistic representations. In V. J.

Cook (ed.) Effects of the second language on the first (pp. 142-167). Clevedon, UK:

Multilingual Matters.

������������ �� �� ��������������� ���

Smiley, S. S., & Brown, A. L. (1979). Conceptual preferences for thematic or taxonomic

relations: a nonmonotonic age trend from preschool to old age. Journal of Experimental

Child Psychology, 28, 249-257.

Tokumaru, Y. (2002). Cross-linguistic influence of L2 English on L1 Japanese in

Japanese-English bilinguals. Proceedings II Simposio International Bilinguismo.

Wakabayashi, N. (1973). Nihongo no imirensou no zure to eigokyoiku. Eigokyoiku (The

English Teachers’ Magazine), Special Issue, January, 68-82.

Wolter, B. (2001). Comparing the L1 and L2 mental lexicon: a depth of individual word

knowledge model. Studies in Second Language Acquisition, 23, 41-69.

Yoneoka, J. (1987). Word associations of Japanese and Americans compared with Moran and

the S-P shift. The Kumamoto Shodai Ronshu (The Review of Kumamoto University of

Commerce), 33, 117-133.�

Zareva, A. (2007). Structure of the second language mental lexicon: how does it compare to

native speakers’ lexical organization? Second Language Research, 23, 123-153.

�������� ���� ��� �������� ��� ����

��������� �������� ��� ����

�������� �� � ����������