The Japanese Psychonomic Society
NII-Electronic Library Service
The JapanesePsychonomic Society
'Ine
JapaneseJeurnato.fRsychonomicScience
13007,VoE,25,No.2,279-280
Summary
ofAwardedPresentationIP44
The
attribute
of
in
Chinese-Japanese
bi
language-switching
linguals'
word
recognition
HuiShuJiN
andKazuhiko
YoKosAwA
The
U'niversit.v
of
Tokyo*Two
monolingual lexicaldecisiontasks and a cued language-switching task were uscd toexplore thecognitive processingof two-Chinese-character compound words inproficient
Chinese-Japanese
bilinguals,The
results showed one-way lacilitationandinhibition
from
Ll
toL2 whenbillngualsperformed lexicaldecision tasksinboth languages. respectively, Interactive
interfer-ence and facilitationwere observed
in
when sporadic language-switching isrequired.Key
words:Chinese-Japanese
bilinguals,language-switching, language setBilingualscan switch from one language to the
other indailyconversation without
hesitation.
How dothey control theirboth
languages? According toGreen
(1998),
it
is
the language schema{or
language
set) which regulates the output from theword
recog-nition system by altering the activation levelsof
representations. Ifa language schema isincharge of
languagc activation, then will manipulating the
lev-els of language schema activation influence
bilin-guals'language processing?
This
is
one of theques-tionswe explored inthepresent study.
Another question that we examined
in
thisstudyi$
whetherbilinguaL
lexical
access islanguage-selective. There might be an inLeractionbetween
lexical
processing inboth languages,especially whenthe two languages share most parts of a common
writing system, such as
Chinese
andJapanese.
Al-though
Chinese
and Japanese belong to differentlinguistic
families,
and differgreatly in grarnmar,both
languages useChinese
characters(Kanji)
in
morphological representations.
There
are manywords that are identical
in
theChincse
andJapanese
]anguage
atboth
graphemic and scmantic Ievels(cognates).
How wM thesimilarity inmorphological repre$entations affect cognitive processingby
Chi-nese-Japanese bilinguals? In the present study weexamined
bilinguals'
language switching inChinesc
* Department of Psychology, Graduate
School
ofHumanities and
Seciology.
The
University
ofTokyo, 7-3Ll Hongo, Bunkyo-ku, Tokyo
O033
and
Japanese,
focusing onthe
comprchension of visually presented words,Method
Stimuli and Tasks
Four
types oftwo-Chinese-character
(Kanji)
compound words were used asstimuli:
(1)
words that are idcnticalin
Chinese
andJapanese
at both graphemic and sernantic levels (cog-natcs);(2)
words specific toJapanese
which do nothave semantic meaning inChinese
(Japanese
words);(3)
words specific toChinese,
whichdo
nothave
meaning inJapanese
(Chinese
words);(4)
non-wordsineither language
(pseudo-word$),
There were 25stimuli of each type.Allof thesingle Chinese
charac-ters
(Kanji)
that appeared ineach compound wordhad
the same morpheme inboth Chinese andJapa-nese. The frequency and
familiarity
of stimuli were strictly contro]led.
Two
monoLingual, lexicaldccision tasks(one
in
Japanese
and the otherin
Chinese),
and a cuedlan-guage-swiLching task,were conductcd, In
the
mone-ltngual
lexical
decision tasks,a total of 100 stimulLwere presented in white at the center of a
black
computer screen background.
The
cued language-switching task had 200 trials,with 100 stimuli pre-$ented twice, one timein
agreen coler, the other ina red celor. The color ofthe
stimuli toldthepartici-pant inwhich Ianguage the
lexical
decisionshouldbemade, depending on the
instructions,
Participants
The
participants were 18 nativespeakers of Chinese with an advanced
level
ofThe Japanese Psychonomic Society
NII-Electronic Library Service
The JapanesePsychonomic Society
280
The
Japanese
Journal
ofPsychonomic
Science
Vol.
25,
No.
2
Tab]e
Mean
Response
Times
(in
milliseconds) of 3 tasks.cognates; "JVkr":
Japanese
words)1{"PSE":
pseudo-words; "CW": Chinese words; "C'':
Monolingual Lexical
Decision
Tasks
Cued
Language-Switching
TaskJapanese
Task
Chinese
TaskJapanese
trialsChinese
trials BilingualsChineseJapanese
PSE CWCJWPSE
JWccwPSE
CWcJWPSE
JWCCW
746 722 624 662 735 762 607 622666
682
587
901
839
669596719755
869 755 825808
895
723
761745
906
730
781
851
923
731 782nese Emd I4 native speakers of
Japanesc
with anintermediate levelof
Chinese.
Results
andDiscussion
Thc
reaction times of correct responsesfor
the
word conditions and the non-word conditions were
analyzed separately using t-tests. The mean
re-sponse times are shown intable1.
The
responses to word conditions inmonolingualtasks are
discussed
first,
The
cognates wereproc-essed significantly
faster
than those words specific toL2
[Chinese
bilinguals,t(17}=3.08,P<.Ol;
Japanese
bilinguals,
t(,11)=2.60,P<.05].The
resu!tsindicated
a one-wayfacilitation
effectfrom
Ll
toL2.
An
oppositefacilitation
orinhibition
effectfrom
L2
to Ll was not revcaled, This result indicatesthatbi]in-guals can effectively activate the conceptual
repre-sentation via
both
Chinese
andJapanese
ertho-graphic rcpresentations, facilitatingtheIexical proc-essing inL2, However, theirprocessing inLl was not affected by the L2 lexicalstatus. Regarding the responses to non-words, pseuclo-words were muchharder
to reject as non-werds whenbilinguals
per-forrned
thelexical
deeision
taskin
L2
compared toLl
[Japanese
bilinguals,t(11)=3.12,P<.Ol].Moreo-vcr,
Chinesc
bilingualswere affected by L2 lexicalstatus as
indicated
bv
thecorrectness of their lexical'
decisions. Itsuggests that
lexical
candidatesfrom
the task-irrelevantlanguage are activated, which
in-dicatesthat lexicalacce$s isnot language-sclective,
With
respect tothecuedlanguage-swit.ching
task,interactive
interierence
and facilitationwerere-vealed
between
thebillnguals'
two languages, not enlyfrorn
L1
toL2,
but
alsoin
theoppositedirection,
Cognates wcre processed significantly fasterthan
words specific toone
language
[Chinese
bilinguals,
t{]7)=4.55,
P<.Ol;
Japanese
bilinguals,t(10)=2.42,P<.05].
Slower
responses and more errors weremade to non-words specific to the current
task-relevant language than to
pseudo-words
[Chinese
bilinguals,t(17)==5.06,
P<.Ol;
Japanese
bilinguals,t{10}=4.66,
p<,Ol].
Since
cognateshave
identical
orthographic and semantic representations
in
both
languages, they were simultaneously activated via
both Chinese and
Japanesc,
resulting infasterre-sponses.
However, slower responses to language-specific
word$ could also be interpretedas thcresult of
inhi-bition
from
the task-irrelevantlanguage. Forexam-ple,
Chinese
words are non-words inJapanese,
and a"No"
response
based
uponJapanese
is
correct.This
"No"
response from
Japanese
isaccompanied even ifin
Chinese
lexical
decisions
whereJapanese
is
notthe
task-relevant
language.
In
contrast, responses tocognate$ are always "Yes"
in
both
languages,
result-inginfasterprocessing. Furthermore, the
simultane-ous activation of two languages also strongly
inter-feredwith responses tonon-words. When a bilingual
was making a
lexical
decision
in
onelanguage,
words specific totheotherlanguage
were also activated atthe same time. Consequently, an inhibitorycontrol
of the non-response language would be needed,
which resulted
in
a slewcr rcsponse when the non-words were words specific totheotherlanguage,
Reference
Green, D.W. 1998 Mental control of the bilingual
lexico-semantic system. Bitingualism] Language
and