You are on page 1of 30

Psychological Review Copyright 1991 by the Ameri can Psychological Association~ Inc.

1991, Vol. 98, No. 2, 224-253 0033-295X/91/$3.00


Culture and the Self." Implications for Cognition, Emotion, and Motivation
Hazel Rose Markus
University of Mi chi gan
Shinobu Kitayama
University o f Or egon
People in different cultures have strikingly different construals of the self, of others, and of the
interdependence of the 2. These construals can influence, and in many cases determine, the very
nature of individual experience, including cognition, emotion, and motivation. Many Asian cul-
tures have distinct conceptions of individuality that insist on the fundamental relatedness of indi-
viduals to each other. The emphasis is on attending to others, fitting in, and harmonious interde-
pendence with them. American cuRure neither assumes nor values such an overt connectedness
among individuals. In contrast, individuals seek to maintain their independence from others by
attending to the self and by discovering and expressing their unique inner attributes. As proposed
herein, these construals are even more powerful than previously imagined. Theories of the self
from both psychology and anthropology are integrated to define in detail the difference between a
construal of the self as independent and a construal of the self as interdependent. Each of these
divergent construals should have a set o fspecific consequences for cognition, emotion, and motiva-
tion; these consequences are proposed and relevant empirical literature is reviewed. Focusing on
differences in self-construals enables apparently inconsistent empirical findings to be reconciled,
and raises questions about what have been thought to be culture-free aspects of cognition, emotion,
and motivation.
In Amer i ca, "t he squeaky wheel gets t he gr e a s e : In Japan,
"t he nai l t hat st ands out gets pounded downy Amer i can par -
ent s who are t r yi ng to i nduce t hei r chi l dr en to eat t hei r supper s
ar e fond o f saying "t hi nk of t he st ar vi ng ki ds i n Et hi opi a, and
appr eci at e how l ucky you are t o be different f r om t h e m" Japa-
nese par ent s are likely t o say " Thi nk about t he f ar mer who
wor ked so har d t o pr oduce t hi s ri ce for you; i f you don' t eat it,
he wi l l feel bad, for his efforts wi l l have been in vain" (H. Ya-
mada, Febr uar y 16,1989). A smal l Texas cor por at i on seeki ng to
elevate pr oduct i vi t y t ol d its empl oyees t o l ook i n t he mi r r or and
say "I am beaut i f ul " 100 t i mes before comi ng t o wor k each day.
Empl oyees of a Japanese super mar ket t hat was recent l y opened
i n New Jer sey wer e i ns t r uc t e d t o begi n t he day by hol di ng
hands and t el l i ng each ot her t hat "he" or "she is beaut i f ul " (' 9,
Japanese Super mar ket , " 1989).
Such anecdot es suggest t hat peopl e in Japan and Amer i ca
may hol d st r i ki ngl y di vergent const r ual s of t he self, others, and
t he i nt er dependence of t he two. The Amer i can exampl es stress
at t endi ng to t he self, t he appr eci at i on of one' s difference from
others, and t he i mpor t ance of assert i ng t he self. The Japanese
exampl es emphasi ze at t endi ng t o and fitting i n wi t h ot hers and
Many thanks to Hiroko Akiyama, Nancy Cantor, Steve Cousins,
Susan Cross, Alan Fiske, Carol Gilligan, Tom Givon, Lawrence
Hirschfeld, Chic Kanagawa, John Kihlstrom, Joan Miller, Richard
Nisbett, Jeanne Oggins, Richard Shweder, Mark Snyder, Harry Trian-
dis, Hiroko Yamada, and Robert Zajonc for their extremely helpful
comments on earlier versions of this article, and thanks to Debbie
Apsley for preparing the manuscript.
Correspondence concerning this article should be addressed to Ha-
zel Rose Markus, Research Center for Group Dynamics--ISR, Univer-
sity of Michigan, Ann Arbor, Michigan 48106-1248, or to Shinobu
Kitayama, Department of Psychology, University of Oregon, Eugene,
Oregon 97403-1227.
t he i mpor t a nc e o f ha r moni ous i nt er dependence wi t h t hem.
These const r ual s of t he sel f and ot hers ar e t i ed t o t he i mpl i ci t ,
nor mat i ve t asks t hat var i ous cul t ur es hol d for what peopl e
shoul d be doi ng i n t hei r lives (cf. Cant or & Ki hl s t r om, 1987;
Eri kson, 1950; Veroff, 1983). Ant hr opol ogi st s and psychol o-
gi st s assume t hat such const r ual s can influence, and i n many
cases det er mi ne, t he ver y nat ur e o f i ndi vi dual experi ence (Cho-
dorow, 1978; Dumont , 1970; Geer t z, 1975; Gergen, 1968; Gi l l i -
gan, 1982; Hol l and & Qui nn, 1987; Lykes, 1985; Marsel l a, De
Vos, & Hsu, 1985; Sampson, 1985, 1988, 1989; Shweder & Le-
Vine, 1984; Smith, 1985; Tr i andi s, 1989; Weisz, Rot hbaum, &
Bl ackburn, 1984; Whi t e & Ki r kpat r i ck, 1985).
Despi t e t he growi ng body of psychol ogi cal and ant hr opol og-
i cal evi dence t hat peopl e hol d di vergent views about t he self,
most of what psychol ogi st s current l y know about human na-
t ure is based on one par t i cul ar vi e w- - t he so-cal l ed West er n
view of t he i ndi vi dual as an i ndependent , sel f-cont ai ned, au-
t onomous ent i t y who (a) compr i ses a uni que confi gurat i on of
i nt er nal at t r i but es (e.g., traits, abi l i t i es, motives, and values) and
(b) behaves pr i mar i l y as a consequence of t hese i nt er nal at t ri -
but es (Geertz, 1975; Sampson, 1988, 1989; Shweder & LeVine,
1984). As a result o f t hi s monocul t ur al appr oach t o t he self(see
Kennedy, Scheier, & Rogers, 1984), psychol ogi st s' under st and-
i ng of t hose phenomena t hat ar e l i nked i n one way or anot her t o
t he sel f may be unnecessar i l y r est r i ct ed (for some i mpor t ant
except i ons, see Bond, 1986,1988; Cousi ns, 1989; Fiske, i n press;
Maehr & Nicholls, 1980; St evenson, Azuma, & Hakut a, 1986;
Tr i andi s, 1989; Tr i andi s, Bont empo, Vi l l areal , Asai , & Lucca,
1988). In t hi s article, we suggest t hat const r ual s of t he self, of
others, and of t he rel at i onshi p bet ween t he sel f and ot hers may
be even mor e powerful t han previ ousl y suggest ed and t hat t hei r
influence is clearly reflected in di fferences among cultures. In
part i cul ar, we compar e an independent view o f t he sel f wi t h one
other, very different view, an interdependent view. The i ndepen-
224
CULTURE AND THE SELF 225
dent vi ew is most cl earl y exempl i fi ed in some si zabl e segment
of Amer i can culture, as well as i n many West er n Eur opean
cultures. The i nt er dependent vi ew is exempl i fi ed i n Japanese
cul t ure as well as i n ot her Asi an cultures. But it is al so charact er-
i st i c of Af r i can cultures, Lat i n- Amer i can cultures, and many
sout her n Eur opean cultures. We del i neat e how t hese di vergent
views of t he s e l f - - t he i ndependent and t he i nt e r de pe nde nt - -
can have a syst emat i c i nfl uence on var i ous aspect s o f cogni t i on,
emot i on, and mot i vat i on.
We suggest t hat for many cul t ures of t he worl d, t he West er n
not i on o f t he sel f as an ent i t y cont ai ni ng significant di sposi -
t i onal at t ri but es, and as det ached f r om cont ext , is si mpl y not an
adequat e descr i pt i on o f sel f hood. Rat her, i n many const rual s,
t he sel f is vi ewed as interdependent wi t h t he sur r oundi ng con-
t ext , and it is t he "ot her " or t he "sel f-i n-rel at i on-t o-ot her" t hat is
focal i n i ndi vi dual experi ence. One general consequence o f t hi s
di vergence i n sel f - const r ual is t hat when psychol ogi cal pr o-
cesses (e.g., cogni t i on, emot i on, and mot i vat i on) explicitly, or
even qui t e implicitly, i mpl i cat e t he sel f as a t arget or as a refer-
ent, t he nat ur e of t hese processes wi l l var y accor di ng t o t he
exact form or or gani zat i on of sel f i nher ent i n a given const r ual .
Wi t h respect t o cogni t i on, for exampl e, for t hose wi t h i nt erde-
pendent selves, i n cont r ast t o t hose wi t h i ndependent selves,
some aspect s of knowl edge r epr esent at i on and some of t he pr o-
cesses involved i n soci al and nonsoci al t hi nki ng al i ke are i n-
fl uenced by a pervasi ve at t ent i veness t o t he rel evant others i n
t he soci al cont ext . Thus, one' s act i ons are mor e l i kel y t o be seen
as si t uat i onal l y bound, and char act er i zat i ons of t he i ndi vi dual
wi l l i ncl ude t hi s cont ext . Fur t her mor e, for t hose wi t h i nt erde-
pendent const r ual s of t he self, bot h t he expressi on and t he expe-
r i ence of emot i ons and mot i ves may be significantly shaped and
governed by a consi der at i on o f t he r eact i ons of others. Specifi-
cally, for exampl e, some emot i ons, l i ke anger, t hat derive f r om
and pr omot e an i ndependent vi ew of t he sel f may be less preva-
l ent among t hose wi t h i nt er dependent selves, and sel f-servi ng
mot i ves may be r epl aced by what appear as ot her - ser vi ng mo-
tives. An exami nat i on of cul t ural var i at i on i n some aspect s of
cogni t i on, emot i on, and mot i vat i on wi l l al l ow psychol ogi st s t o
ask exact l y what is uni versal i n t hese processes, and it has t he
pot ent i al t o pr ovi de some new insights for t heori es of t hese
psychol ogi cal processes.
In t hi s analysis, we dr aw on r ecent research efforts devot ed t o
char act er i zi ng t he general differences bet ween Amer i can or
West er n views of per s onhood and East er n or Asi an perspec-
tives (e.g., Heel as & Lock, 1981; Hofst ede, 1980; Mar sel l a et al.,
1985; Rol and, 1988; Schwart z & Bilsky, 1990; Shweder, 1990;
Shweder & LeVine, 1984; St i gl er, Shweder, & Her dt , 1990;
Tr i andi s, 1989; Tr i andi s & Brislin, 1980; Wei sz et al., 1984). We
ext r act f r om t hese des cr i pt i ons many i mpor t a nt di fferences
t hat may exi st in t he specific cont ent , st ruct ure, and f unct i on-
i ng o f t he sel f - syst ems o f peopl e o f di f f er ent cul t ur al back-
grounds. The di st i nct i ons t hat we make bet ween i ndependent
and i nt er dependent const r ual s mus t be r egar ded as general t en-
denci es t hat may emer ge when t he member s o f t he cul t ure are
consi der ed as a whole. The pr ot ot ypi cal Amer i can vi ew of t he
self, for exampl e, may prove t o be most char act er i st i c o f Whi t e,
mi ddl e- cl as s me n wi t h a Wes t er n Eur ope a n et hni c back-
gr ound. It may be somewhat less descri pt i ve of women i n gen-
eral , or of men and women f r om ot her et hni c gr oups or soci al
classes.1 Moreover, we real i ze t hat t her e may well be i mpor t ant
di st i nct i ons among t hose views we di scuss as si mi l ar and t hat
t here may be views of t he sel f and ot hers t hat cannot easily be
classified as ei t her i ndependent or i nt er dependent .
Our i nt ent i on is not t o cat al og al l t ypes of sel f - const r ual s, but
rat her t o hi ghl i ght a vi ew of t he sel f t hat is oft en assumed t o be
uni versal but t hat may be qui t e specific t o some segment s of
West er n culture. We argue t hat sel f-const rual s pl ay a maj or role
in regul at i ng vari ous psychol ogi cal processes. Under st andi ng
t he nat ur e of di vergent sel f-const rual s has t wo i mpor t ant conse-
quences. On t he one hand, it allows us t o organi ze several appar -
ent l y i ncons i s t ent e mpi r i c a l f i ndi ngs a nd t o pose ques t i ons
about t he uni versal i t y assumed for many aspect s o f cogni t i on,
emot i on, a nd mot i vat i on (see Shweder, 1990). On t he ot her
hand, it per mi t s us to bet t er speci fy t he preci se rol e o f t he sel f i n
medi at i ng and regul at i ng behavior.
Th e Sel f: A De l i c a t e Ca t e g o r y
Universal Aspects of the Self
I n expl or i ng t he pos s i bi l i t y o f di f f er ent t ype s o f sel f - con-
st rual s, we begi n wi t h Hal l owel l ' s (1955) not i on t hat peopl e
everywhere are likely t o devel op an under s t andi ng of t hem-
selves as physi cal l y di st i nct and separ abl e f r om others. Head
(1920), for exampl e, c l a i me d t he exi s t ence o f a uni ver sal
schema of t he body t hat pr ovi ded one wi t h an anchor i n t i me
and space. Similarly, Al l por t (1937) suggest ed t hat t here must
exi st an aspect of per sonal i t y t hat allows one, when awakeni ng
each mor ni ng, t o be sure t hat he or she is t he same per son who
went t o sl eep t he ni ght before. Most recently, Nei sser (1988)
referred t o t hi s aspect of sel f as t he ecological self, whi ch he
defi ned as "t he sel f as percei ved wi t h respect t o t he physi cal
envi r onment : T am t he per son here i n t hi s place, engaged i n
t hi s par t i cul ar act i vi t y" (p. 3). Beyond a physi cal or ecol ogi cal
sense o f self, each per son pr obabl y has some awareness of inter-
nal activity, such as dr eams , a nd o f t he cont i nuous flow o f
t hought s and feelings, whi ch are pri vat e t o t he ext ent t hat t hey
cannot be di rect l y known by others. The awareness of t hi s un-
shar ed experi ence wi l l l ead t he per son t o some sense of an
inner, pri vat e self.
Divergent Aspects of the Self
Some under s t andi ng and some r epr esent at i on of t he private,
i nner aspect s of t he sel f may well be uni versal , but many ot her
aspect s of t he sel f may be qui t e specific t o par t i cul ar cultures.
Peopl e are capabl e o f bel i evi ng an ast oni shi ng vari et y of t hi ngs
about t hemsel ves (cf. Heel as & Lock, 1981; Mar sel l a et al., 1985;
Shweder & LeVine, 1984; Tr i andi s, 1989). The sel f can be con-
st r ued, f r amed, or concept ual l y r epr esent ed in mul t i pl e ways. A
cross-cul t ural survey o f t he sel f l ends s uppor t t o Dur khei m' s
(1912/1968) earl y not i on t hat t he cat egory of t he sel f is pr i mar -
The prototypical American view may also be further restricted to a
particular point in history. It may be primarily a product of late, indus-
trial capitalism (see Baumeister, 1987). For an analysis of the origins of
the independent view, see BeUah, Madsen, Sullivan, Swidler, & Tipton
(1985) and Weber 0958).
226 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
ily t he pr oduct o f soci al factors, and t o Mauss' s (1938/1985)
cl ai m t hat as a soci al category, t he sel f is a"del i cat e" one, subj ect
t o qui t e subst ant i al , i f not infinite, vari at i on.
The exact cont ent and st r uct ur e o f t he i nner sel f may differ
consi der abl y by culture. Fur t her mor e, t he nat ur e of t he out er or
publ i c sel f t hat deri ves f r om one' s rel at i ons wi t h ot her peopl e
and soci al i nst i t ut i ons may al so var y mar kedl y by culture. And,
as suggest ed by Tr i andi s (1989), t he si gni fi cance assi gned t o t he
private, i nner aspect s versus t he public, r el at i onal aspect s in
regul at i ng behavi or wi l l var y accordingly. I n fact, it may not be
unr easonabl e t o suppose, as di d numer ous ear l i er ant hr opol o-
gi st s (see Al l en, 1985), t hat i n some cultures, on cer t ai n occa-
sions, t he individual, i n t he sense of a set of si gni fi cant i nner
at t r i but es o f t he person, may cease t o be t he pr i ma r y uni t o f
consci ousness. I nst ead, t he sense of bel ongi ngness t o a soci al
rel at i on may become so st r ong t hat it makes bet t er sense t o
t hi nk o f t he relationship as t he f unct i onal uni t o f consci ous
reflection.
The cur r ent anal ysi s focuses on j us t one var i at i on in what
peopl e i n di fferent cul t ures can come t o bel i eve about t hem-
selves. Thi s one var i at i on concer ns what t hey bel i eve about t he
rel at i onshi p bet ween t he sel f and others and, especially, t he de-
gree t o whi ch t hey see t hemsel ves as separate f r om ot hers or as
connected wi t h others. We suggest t hat t he si gni fi cance and t he
exact f unct i onal rol e t hat t he per son assigns t o t he ot her when
defi ni ng t he sel f depend on t he cul t ural l y shar ed assumpt i ons
about t he separ at i on or connect edness bet ween t he sel f and
others.
Mother
Sibling
Xx:,x
Friend
A , Independent View of Self
Mother
Father
X X ~ Sibling
Friend~,N X X X x ( ~
~ ) . @ _ _ ( ~ C o - w o r k e r
Friend
B. Interdependent View of Self
Figure 1. Conceptual representations of the self. (A: Independent
construal. B: Interdependent construai.)
Two Co n s t r u a l s o f t h e Sel f: I n d e p e n d e n t
a n d I n t e r d e p e n d e n t
The Independent Construal
I n many West er n cul t ures, t here is a faith i n t he i nher ent
separ at eness o f di st i nct persons. The nor mat i ve i mperat i ve o f
t hi s cul t ure is t o become i ndependent f r om ot her s and to di s-
cover and express one' s uni que at t r i but es (Johnson, 1985; Mar -
sella et al., 1985; J. G. Miller, 1988; Shweder & Bourne, 1984).
Achi evi ng t he cul t ural goal of i ndependence requi res const r u-
i ng onesel f as an i ndi vi dual whose behavi or is or gani zed and
made meani ngf ul pr i mar i l y by reference t o one' s own i nt er nal
r eper t oi r e o f t hought s, feelings, and act i on, r at her t han by refer-
ence t o t he thoughts, feelings, and act i ons of others. Accor di ng
t o t hi s const r ual of self, to bor r ow Geer t z' s (1975) oft en quot ed
phrase, t he per son is vi ewed as "a bounded, uni que, mor e or
less i nt egr at ed mot i vat i onal and cogni t i ve universe, a dynami c
cent er o f awareness, emot i on, j udgment , and act i on or gani zed
i nt o a di st i nct i ve whol e and set cont rast i vel y bot h agai nst ot her
such whol es a nd a ga i ns t a soci al a nd na t ur a l ba c kgr ound"
(p. 48).
Thi s vi ew o f t he sel f deri ves f r om a bel i ef i n t he whol eness
and uni queness o f each person' s confi gurat i on o f i nt er nal at t r i b-
ut es (Johnson, 1985; Sampson, 1985, 1988, 1989; Wat er man,
1981). It gives ri se t o processes l i ke "sel f - act ual i zat i on; "real i z-
i ng onesel f " "expressi ng one' s uni que confi gurat i on o f needs,
rights, and capaci t i es, ' or "devel opi ng one' s di st i nct pot ent i al : '
The essent i al aspect o f t hi s vi ew involves a concept i on o f t he sel f
as an aut onomous, i ndependent per son; we t hus refer t o it as t he
independent construal of the self. Ot her si mi l ar labels i ncl ude
individualist, egocentric, separate, autonomous, idiocentric, and
self-contained. We assume t hat , on average, relatively mor e i ndi -
vi dual s i n West ern cul t ures wi l l hol d t hi s view t han will i ndi vi d-
ual s i n non-West ern cultures. Wi t hi n a given culture, however,
i ndi vi dual s wi l l var y i n t he ext ent to which t hey are good cul-
t ur al represent at i ves and const r ue t he sel f i n t he mandat ed way.
The i ndependent sel f must , of course, be responsive t o t he
soci al envi r onment (Fiske, i n press). Thi s responsiveness, how-
ever, is fost ered not so much for t he sake of t he responsi veness
itself. Rat her, soci al responsi veness often, i f not always, deri ves
from t he need t o st rat egi cal l y det er mi ne t he best way to express
or assert t he i nt er nal at t r i but es of t he self. Others, or t he social
si t uat i on in general , are i mpor t ant , but pr i mar i l y as st andar ds
o f reflected appr ai sal , or as sources t hat can verify and affirm
t he i nner core o f t he self.
The West ern, i ndependent view of t he sel f is i l l ust rat ed in
Fi gur e 1A. The l arge circle represent s t he self, and t he smal l er
circles represent specific others. The Xs are represent at i ons o f
t he vari ous aspect s of t he sel f or t he others. In some cases, t he
l arger circle and t he smal l circle i nt ersect , and t here is an X in
t he i nt ersect i on. Thi s refers to a r epr esent at i on of t he self-in-re-
l at i on- t o- ot her s or t o a par t i cul ar soci al rel at i on (e.g., "I am very
pol i t e in front of my professor"). An X wi t hi n t he sel f circle but
out si de o f t he i nt ersect i on represent s an aspect of t he sel f per-
ceived t o be relatively i ndependent of specific ot hers and, thus,
i nvar i ant over t i me and cont ext . Thes e sel f - r epr esent at i ons
usual l y have as t hei r referent some i ndi vi dual desire, preference,
at t ri but e, or abi l i t y (e.g., "I am creative"). For t hose with i ndepen-
dent const r ual s of t he self, it is t hese i nner at t ri but es t hat are
CULTURE AND THE SELF 227
mos t significant i n regul at i ng behavi or and t hat ar e assumed,
bot h by t he act or and by t he obser ver alike, t o be di agnost i c of
t he actor. Such r epr esent at i ons o f t he i nner sel f ar e t hus t he
mos t e l a bor a t e d i n me mo r y a nd t he mos t accessi bl e when
t hi nki ng of t he sel f (as i ndi cat ed by Xs i n Fi gur e IA). They can
be cal l ed core conceptions, salient identities, or self-schemata
(e.g., Ger gen, 1968; Mar kus, 1977; Stryker, 1986).
T h e I nt er dependent Co n s t r u a l
I n cont rast , many non-West ern cul t ures i nsi st , i n Kondo' s
(1982) t er ms, on t he f undament al connectedness o f huma n be-
i ngs t o each other. A nor mat i ve i mperat i ve o f t hese cul t ures is t o
ma i nt a i n t hi s i nt e r de pe nde nc e a mong i ndi vi dual s ( De Vos,
1985; Hsu, 1985; Mi l l er, 1988; Shweder & Bourne, 1984). Experi -
enci ng i nt er dependence ent ai l s seei ng onesel f as par t of an en-
compassi ng soci al rel at i onshi p and recogni zi ng t hat one' s be-
havi or is det er mi ned, cont i ngent on, and, t o a large ext ent orga-
ni zed by what t he act or percei ves t o be t he thoughts, feelings,
and act i ons of others i n t he rel at i onshi p. The Japanese experi -
ence of t he self, therefore, i ncl udes a sense of i nt er dependence
and of one' s st at us as a par t i ci pant i n a l arger soci al uni t (Samp-
son, 1988). Wi t hi n such a const r ual , t he sel f becomes mos t
meani ngf ul and compl et e when it is cast i n t he appr opr i at e
soci al rel at i onshi p. Accor di ng t o Lebr a (1976) t he Japanese are
most fully human i n t he cont ext of others.
Thi s vi ew of t he sel f and t he rel at i onshi p bet ween t he sel f and
ot hers features t he per son not as separat e f r om t he soci al con-
t ext but as mor e connect ed and less di fferent i at ed f r om others.
Peopl e are mot i vat ed t o fi nd a way t o fit i n wi t h rel evant others,
t o fulfill and creat e obl i gat i on, and i n general t o become par t of
var i ous i nt er per s onal r el at i onshi ps. Unl i ke t he i ndependent
self, t he si gni fi cant feat ures o f t he sel f accor di ng t o t hi s con-
st r ual are t o be f ound i n t he i nt er dependent and thus, i n t he
mor e publ i c component s o f t he self. We t herefore cal l t hi s view
t he interdependent construal o f the self. The same not i on has
been vari ousl y referred to, wi t h somewhat different connot a-
t i ons, as sociocentric, holistic, collective, allocentric, ensembled,
constitutive, contextualist, connected, and relational As wi t h t he
i ndependent self, ot her s are cri t i cal for soci al compar i s on and
self-validation, yet i n an i nt er dependent f or mul at i on o f t he self,
t hese ot hers become an i nt egral par t of t he set t i ng, si t uat i on, or
cont ext t o whi ch t he sel f is connect ed, fitted, and assi mi l at ed.
The exact manner i n whi ch one achi eves t he t ask of connect i on,
therefore, depends cruci al l y on t he nat ur e of t he cont ext , par t i c-
ul arl y t he ot her s pr esent i n t he cont ext . Ot her s t hus par t i ci pat e
actively and cont i nuousl y i n t he defi ni t i on of t he i nt er depen-
dent self.
The i nt er dependent sel f al so possesses and expresses a set of
i nt er nal at t ri but es, such as abi l i t i es, opi ni ons, j udgment s, and
per sonal i t y charact eri st i cs. However, t hese i nt er nal at t r i but es
are under s t ood as si t uat i on specific, and t hus as somet i mes elu-
sive and unrel i abl e. And, as such, t hey ar e unl i kel y t o assume a
powerful role in regul at i ng overt behavi or, especi al l y i f t hi s be-
havi or i mpl i cat es si gni fi cant others. I n many domai ns of soci al
life, one' s opi ni ons, abi l i t i es, and char act er i st i cs ar e assi gned
onl y s econdar y r ol e s - - t he y mus t i nst ead be const ant l y con-
t r ol l ed and regul at ed t o come t o t er ms wi t h t he pr i ma r y t ask of
i nt er dependence. Such vol unt ar y cont r ol of t he i nner at t r i but es
const i t ut es t he cor e of t he cul t ural i deal of becomi ng mat ure.
The under s t andi ng of one' s aut onomy as secondar y to, and con-
st r ai ned by, t he pr i ma r y t ask of i nt er dependence di st i ngui shes
i nt er dependent selves f r om i ndependent selves, for whom au-
t onomy and its expressi on is oft en afforded pr i ma r y signifi-
cance. An i ndependent behavi or (e.g., assert i ng an opi ni on) ex-
hi bi t ed by a per son i n an i nt er dependent cul t ure is likely t o be
based on t he pr emi se o f under l yi ng i nt er dependence and t hus
may have a somewhat different si gni fi cance t han it has for a
per son f r om an i ndependent culture.
The i nt er dependent sel f is i l l ust r at ed in Fi gur e lB. For t hose
wi t h i nt er dependent selves, t he si gni fi cant sel f-represent at i ons
(the Xs) are t hose i n rel at i onshi p t o specific others. I nt er depen-
dent selves cer t ai nl y i ncl ude r epr esent at i ons of i nvar i ant per-
sonal at t r i but es and abilities, and t hese represent at i ons can be-
come phenomenol ogi cal l y qui t e sal i ent , but in many ci r cum-
s t ances t hey ar e l ess i mp o r t a n t i n r egul at i ng obs er vabl e
behavi or and are not assumed t o be par t i cul ar l y di agnost i c o f
t he self. 2 I nst ead, t he self-knowledge t hat gui des behavi or is of
t he sel f-i n-rel at i on t o specific ot hers i n par t i cul ar contexts. The
f undament al uni t s of t he self-system, t he core concept i ons, or
sel f-schemat a are t hus pr edi cat ed on significant i nt er per sonal
rel at i onshi ps.
An i nt er dependent sel f cannot be pr oper l y char act er i zed as a
bounded whole, for it changes st r uct ur e wi t h t he nat ur e of t he
par t i cul ar soci al cont ext . Wi t hi n each par t i cul ar soci al situa-
t i on, t he sel f can be differently i nst ant i at ed. The uni queness of
such a sel f deri ves f r om t he specific confi gurat i on o f rel at i on-
shi ps t hat each per son has devel oped. Wha t is focal and obj ect i -
fied i n an i nt er dependent self, t hen, is not t he i nner self, but t he
relationships of t he per son t o ot her act ors ( Hamaguchi , 1985).
The not i on of an i nt er dependent sel f is l i nked wi t h a moni st i c
phi l osophi cal t r adi t i on i n whi ch t he per son is t hought t o be of
t he same subst ance as t he rest of nat ur e (see Bond, 1986; Phi l -
lips, 1976; Rol and, 1988; Sass, 1988). As a consequence, t he
rel at i onshi p bet ween t he sel f and other, or bet ween subj ect and
obj ect , is assumed t o be much closer. Thus, many non-West ern
cul t ures i nsi st on t he i nsepar abi l i t y o f basi c el ement s (Gal t ung,
1981), i ncl udi ng sel f and other, and per son and si t uat i on. I n
Chi nese culture, for i nst ance, t here is an emphasi s on synt hesi z-
i ng t he const i t uent par t s of any pr obl em or si t uat i on i nt o an
i nt egr at ed or h a r mo n i o u s whol e ( Moor e, 1967; Nor t hr op,
1946). Thus, per sons are onl y par t s t hat when separ at ed f r om
t he l arger soci al whol e cannot be fully under s t ood (Phi l l i ps,
1976; Shweder, 1984). Such a hol i st i c vi ew is i n opposi t i on t o
t he Car t esi an, dual i st i c t r adi t i on t hat char act er i zes West er n
t hi nki ng and in whi ch t he sel f is separ at ed f r om t he obj ect and
f r om t he nat ur al wodd.
Exampl es o f the interdependent self. An i nt e r de pe nde nt
vi ew of t he sel f is c ommon t o many of t he ot herwi se highly
diverse cul t ures of t he worl d. St udi es of t he ma i nl a nd Chi nese,
for exampl e, s ummar i zed i n a r ecent book by Bond (1986),
show t hat even among t he most r api dl y moder ni zi ng segment s
o f t he Chi nese popul at i on, t her e is a t endency for peopl e t o act
2 For a discussion of how interdependent selves strive to maintain a
balance between internal (private) and extensive (public) representa-
tions, see T. Doi (1986).
228 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
pr i mar i l y i n accor dance wi t h t he ant i ci pat ed expect at i ons of
ot hers and soci al nor ms r at her t han wi t h i nt er nal wishes or
per sonal at t r i but es (Yang, 198 lb). A pr e mi um is pl aced on em-
phasi zi ng collective welfare and on showi ng a sympat het i c con-
cer n for others. Thr oughout t he st udi es of t he Chi nese r epor t ed
by Bond, one can see t he cl ear i mpr i nt of t he Conf uci an empha-
sis on i nt er r el at edness and ki ndness. Accor di ng t o Hsu (1985),
t he supr eme Chi nese vi rt ue, jen, i mpl i es t he person' s capabi l i t y
to i nt eract wi t h fellow huma n bei ngs in a sincere, polite, and
decent fashion (see al so Elvin, 1985).
Numer ous ot her exampl es o f cul t ures i n which peopl e are
likely t o have some versi on of an i nt er dependent sel f can al so be
i dent i fi ed. For exampl e, Tr i andi s, Mar i n, Lisansky, and Bet an-
c our t 0984) have de s c r i be d t he i mp o r t a n c e o f s i mpat i co
among Hi spani cs. Thi s qual i t y refers t o t he abi l i t y to bot h re-
spect and share ot hers' feelings. In char act er i zi ng t he psychol -
ogy of Fi l i pi nos, Chur ch (1987) descr i bed t he i mpor t ance t hat
peopl e at t r i but e t o smoot h i nt er per sonal rel at i ons and to bei ng
"agreeabl e even under difficult ci r cumst ances, sensitive to what
ot hers are feeling and wi l l i ng t o adj ust one' s behavi or accor d-
ingly." Similarly, Weisz (in press) r epor t ed t hat Thai s pl ace a
pr emi um on self-effacement, humility, deference, and on t r yi ng
to avoi d di st ur bi ng others. Among t he Japanese, it is si mi l arl y
cr uci al not t o di st ur b t he wa, or t he har moni ous ebb and flow o f
i nt er per sonal rel at i ons (see al so Geer t z, 1974, for char act er i za-
t i ons o f si mi l ar i mper at i ves among t he Bal i nese and Mor oc-
cans).
Beat t i e (1980) cl ai med t hat Af r i cans are al so ext r emel y sensi-
tive t o t he i nt er dependenci es among peopl e and view t he worl d
and ot hers i n it as ext ensi ons of one anot her. The sel f is vi ewed
not as a hedged cl osure but as an open field. Similarly, Mar r i ot t
(1976) ar gued t hat Hi ndu concept i ons assume t hat t he sel f is an
open ent i t y t hat is given shape by t he soci al cont ext . In his
insightful book, Ka ka r (1978) descr i bed t he Hi ndu' s i deal of
i nt er per sonal fusion and how it is accompani ed by a per sonal ,
cul t ural sense o f hell, whi ch is separ at i on f r om others. In fact,
Mi l l er, Bersoff, and Har wood (1990), i n a recent , careful l y con-
t r ol l ed st udy on mor al reasoni ng, found t hat I ndi ans r egar d
responsi veness t o t he needs o f ot hers as an obj ect i ve mor al obl i -
gat i on t o a far great er ext ent t han do Amer i cans. Al t hough t he
sel f-syst ems of peopl e f r om t hese cul t ures ar e mar kedl y differ-
ent in many ot her i mpor t ant respect s, t hey appear t o be al i ke in
t he great er val ue (when c ompa r e d wi t h Amer i cans) t hat is at-
t ached to pr oper rel at i ons wi t h others, and i n t he r equi r ement
t o flexibly change one' s own behavi or i n accor dance wi t h t he
nat ur e o f t he rel at i onshi p.
Even i n Amer i can culture, t here is a st r ong t heme o f i nt erde-
pendence t hat is refl ect ed in t he values and act i vi t i es o f many of
its subcul t ures. Rel i gi ous groups, such as t he Quakers, expl i ci t l y
val ue and pr omot e i nt er dependence, as do many smal l t owns
and r ur al communi t i es (e.g., Bellah, Madsen, Sullivan, Swidler,
& Ti pt on, 1985). Some not i on o f a mor e connect ed, ensembl ed,
i nt er dependent self, as oppos ed to a sel f-cont ai ned, i ndepen-
dent self. is al so bei ng devel oped by several o f what Sampson
(1989) cal l s " pos t moder n" t heori st s. These t heor i st s are ques-
t i oni ng t he soverei gnt y o f t he Amer i can vi ew of t he mat ur e
per son as aut onomous, sel f - det er mi ned, and unencumber ed.
They argue t hat psychol ogy is cur r ent l y domi nat ed by a view of
t he per son t hat does not adequat el y reflect t he ext ent t o which
peopl e everywhere are cr eat ed by, const r ai ned by, and respon-
sive to t hei r var i ous i nt er per sonal cont ext s (see Ger gen & Ger -
gen, 1988; Gi l l i gan, 1982; Miller, 1986; Tajfel, 1984).
Furtherdefinitionoftheinterdependentself. Theor i st s of Jap-
anese cul t ure are begi nni ng t o char act er i ze t he i nt er dependent
sel f much mor e speci f i cal l y t han was pr evi ousl y at t empt ed.
These descr i pt i ons offer some mor e refi ned i deas of how an
i nt er dependent view of sel f can depar t mar kedl y from an i nde-
pendent view o f self(see Nakane, 1970; Plath, 1980; R. J. Smith,
1983). For exampl e, bui l di ng on a st udy of L. T. Doi 0973) ,
Bachni k (1986) wrot e
(in Japanese society) rather than there being a single social reality,
a number of possible perspectives of both self and social life are
acknowledged. Interaction in Japanese society then focuses on the
definition of the appropriate choice, out of all the various possibili-
ties. This means that what one says and does will be different in
different situations, depending on how one defines one's particu-
lar perspective versus the social other. (p. 69)
I n Japan, t he word for self, fibun, refers to "one' s share of t he
shar ed life space" ( Hamaguchi , 1985). The self, Ki mur a (cited
i n Hamaguchi , 1985) cl ai med, is "nei t her a subst ance nor an
at t r i but e having a const ant oneness" (p. 302). Accordi ng t o Ha-
maguchi (1985), for t he Japanese, "a sense of i dent i fi cat i on wi t h
ot hers (somet i mes i ncl udi ng conflict) pre-exi st s and selfness is
conf i r med onl y t hr ough i nt er per s onal r el at i onshi ps . . . . Self-
ness is not a const ant l i ke t he ego but denot es a fluid concept
whi ch changes t hr ough t i me and si t uat i ons accor di ng to inter-
per sonal rel at i onshi ps" (p. 302).
The Japanese ant hr opol ogi st Lebr a (1976) defi ned t he es-
sence of Japanese cul t ure as an "ethos of soci al r el at i vi sm" Thi s
t ransl at es i nt o a const ant concer n for bel ongi ngness, reliance,
dependency, empathy, occupyi ng one' s pr oper place, and reci-
proci t y. She cl ai med t he J apanes e ni ght mar e is excl usi on,
meani ng t hat one is fai l i ng at t he nor mat i ve goal of connect i ng
t o others. Thi s is i n shar p cont r ast to t he Amer i can ni ght mare,
which is t o fail at separat i ng from others, as can occur when one
is undul y i nfl uenced by others, or does not st and up for what
one believes, or when one goes unnot i ced or undi st i ngui shed.
An i nt er dependent view of sel f does not result in a mergi ng of
sel f and other, nor does it i mpl y t hat one must always be i n t he
company of ot hers to funct i on effectively, or t hat peopl e do not
have a sense of t hemsel ves as agents who are t he ori gi ns of t hei r
own actions. On t he contrary, it t akes a hi gh degree of self-con-
t r ol and agency t o effectively adj ust onesel f t o vari ous i nt er per -
sonal cont i ngenci es. Agent i c exercise o f cont rol , however, is
di r ect ed pr i mar i l y to t he i nsi de and t o t hose i nner at t ri but es,
such as desires, per sonal goals, and private emot i ons, t hat can
di st ur b t he har moni ous equi l i br i um of i nt er per sonal t ransac-
t i on. Thi s can be cont r ast ed wi t h t he West ern not i on o f cont rol ,
whi ch pr i mar i l y i mpl i es an asser t i on of t he i nner at t ri but es and
a consequent at t empt t o change t he out er aspects, such as one' s
publ i c behavi ors and t he soci al si t uat i on (see al so Weisz et al.,
1984).
Gi ven t he Japanese not i on o f cont r ol t hat is inwardly di -
rect ed, t he abi l i t y t o effectively adj ust in t he i nt er per sonal do-
mai n may f or m an i mpor t ant basi s o f self-esteem, and i ndi vi du-
al i zed styles o f such adj ust ment t o soci al cont i ngenci es may
cont r i but e to t he sense of self-uniqueness. Thus, Hamaguchi
CULTURE AND THE SELF 229
(1985), for exampl e, r e por t e d t hat for t he Japanese, " t he
st rai ght forward cl ai m of t he naked ego" (p. 303) is exper i enced
as chi l di sh. Sel f-assert i on is not vi ewed as bei ng aut hent i c, but
i nst ead as bei ng i mmat ur e. Thi s poi nt is echoed in M. Whi t e
and LeVine' s (1986) descr i pt i on of t he meani ng of sunao, a t er m
used by Japanese par ent s t o char act er i ze what t hey val ue i n
t hei r chi l dren:
A child that is sunao has not yielded his or her personal autonomy
for the sake of cooperation; cooperation does not suggest giving
up the self, as it may in the West; it implies that working with
others is the appropriate way of expressing and enhancing the self.
Engagement and harmony with others is, then, a positively valued
goal and the bri dge--t o open-hearted cooperation, as in sunao--
is through sensitivity, reiterated by the mother's example and en-
couragement. (p. 58)
Kumagai (198 l ) sai d sunao "assumes cooper at i on t o be an act
of affi rmat i on of t he sel f " (p. 261). Gi vi ng in is not a sign of
weakness; rather, it reflects t ol erance, self-control, flexibility,
and maturity.
The role o f the other in t he interdependent self. In an i nt erde-
pendent view, i n cont r ast t o an i ndependent view, ot her s wi l l be
assi gned much mor e i mpor t ance, wi l l car r y mor e weight, and
wi l l be relatively focal i n one' s own behavior. Ther e are several
di r ect consequences of an i nt er dependent const r ual o f t he self.
Fi r s t , r el at i ons hi ps , r at her t ha n bei ng me a ns for r eal i zi ng
var i ous i ndi vi dual goals, wi l l oft en be ends in and of t hem-
selves. Al t hough peopl e everywhere mus t mai nt ai n some relat-
edness wi t h others, an appr eci at i on and a need for peopl e wi l l
be mor e i mpor t ant for t hose wi t h an i nt er dependent sel f t han
for t hose wi t h an i ndependent self. Second, mai nt ai ni ng a con-
nect i on t o ot hers wi l l mean bei ng const ant l y aware of ot hers
and focusing on t hei r needs, desires, and goals. I n some cases,
t he goal s of ot hers may become so focal i n consci ousness t hat
t he goal s of ot hers may be exper i enced as per sonal goals. I n
ot her cases, fulfilling one' s own goal s may be qui t e di st i nct f r om
t hose of others, but meet i ng anot her ' s goals, needs, and desi res
wi l l be a necessar y r equi r ement for sat i sfyi ng one' s own goals,
needs, and desires. The assumpt i on is t hat whi l e pr omot i ng t he
goal s of others, one' s own goal s wi l l be at t ended t o by t he per son
wi t h whom one is i nt er dependent . Hence, peopl e may actively
wor k t o fulfill t he ot hers' goal s whi l e passively moni t or i ng t he
r eci pr ocal cont r i but i ons f r om t hese ot hers for one' s own goal -
ful fi l l ment . Yamagi shi (1988), i n fact, suggest ed t hat t he Japa-
nese feel ext remel y uncomf or t abl e, much mor e so t han Amer i -
cans, when t he o p p o r t u n i t y for such passi ve moni t or i ng o f
ot hers' act i ons is deni ed.
Fr om t he s t andpoi nt of an i ndependent , "self-ish" self, one
mi ght be l ed t o r omant i ci ze t he i nt er dependent self, who is ever
at t uned t o t he concer ns o f others. Yet i n many cases, responsi ve
and cooperat i ve act i ons are exerci sed onl y when t her e is a rea-
sonabl e assurance o f t he "good-i nt ent i ons" of others, namel y
t hei r commi t ment t o cont i nue t o engage i n r eci pr ocal i nt erac-
t i on and mut ual suppor t . Clearly, i nt er dependent selves do not
at t end t o t he needs, desires, and goal s o f all others. At t ent i on t o
ot hers is not i ndi scr i mi nat e; it is highly selective and wi l l be
most char act er i st i c of r el at i onshi ps wi t h "i n- gr oup" member s.
These are ot hers wi t h whom one shares a c o mmo n fate, such as
f ami l y member s or member s o f t he same l ast i ng soci al group,
such as t he wor k gr oup. Out - gr oup me mbe r s ar e t ypi c a l l y
t r eat ed qui t e differently and are unl i kel y t o experi ence ei t her
t he advant ages or di sadvant ages o f i nt erdependence. I ndepen-
dent selves are al so selective i n t hei r associ at i on wi t h ot hers but
not t o t he ext ent of i nt er dependent selves because much less of
t hei r behavi or is di rect l y cont i ngent on t he act i ons of others.
Gi ven t he i mpor t ance o f ot hers i n const r uct i ng real i t y and regu-
l at i ng behavior, t he i n- gr oup- out - gr oup di st i nct i on is a vi t al
one for i nt er dependent selves, and t he subjective bounda r y of
one' s "i n- gr oup" may t end t o be nar r ower for t he i nt er depen-
dent selves t han for t he i ndependent selves (Tri andi s, 1989).
To i l l ust rat e t he r eci pr ocal nat ur e o f i nt er act i on among t hose
wi t h i nt er dependent views, i magi ne t hat one has a f r i end over
for l unch and has deci ded t o make a sandwi ch for hi m. The
conversat i on mi ght be: "Hey, Tom, what do you want i n your
sandwich? I have turkey, sal ami , and chees e: Tom responds,
"Oh, I l i ke turkey." Not e t hat t he f r i end is given a choi ce because
t he host assumes t hat f r i end has a ri ght , i f not a duty, t o make a
choi ce reflecting his i nner at t ri but es, such as preferences or
desires. And t he f r i end makes his choi ce exact l y because o f t he
bel i ef i n t he same assumpt i on. Thi s scr i pt i s " n a t u r a l ; however,
onl y wi t hi n t he i ndependent view of self. Wha t woul d happen i f
t he f r i end were a vi si t or f r om Japan? A likely response t o t he
quest i on "Hey, Tomi o, what do you want?" woul d be a little
moment of bewi l der ment and t hen a noncommi t al ut t er ance
l i ke "I don' t know." Thi s happens because under t he assump-
t i ons of an i nt er dependent self, it is t he r esponsi bi l i t y of t he host
t o be abl e t o "r ead" t he mi nd of t he f r i end and offer what t he
host percei ves t o be t he best for t he fri end. And t he dut y o f t he
guest , on t he ot her hand, is t o receive t he favor wi t h grace and
t o be pr epar ed t o r et ur n t he favor i n t he near future, i f not ri ght
at t he next moment . A likely, i nt er dependent scr i pt for t he same
si t uat i on woul d be: "Hey, Tomio, I made you a t ur key sandwi ch
because I r emember t hat l ast week you sai d you l i ke t ur key
mor e t han beef y And Tomi o wi l l r espond, "Oh, t hank you, I
really l i ke t urkey"
The r eci pr ocal i nt er dependence wi t h ot her s t hat is t he sign o f
t he i nt er dependent sel f seems t o requi re const ant engagement
o f what Mead (1934) meant by t aki ng t he rol e of t he other. It
involves t he wi l l i ngness and abi l i t y t o feel and t hi nk what ot her s
are feeling and t hi nki ng, t o absor b t hi s i nf or mat i on wi t hout
bei ng t ol d, and t hen t o hel p ot hers satisfy t hei r wishes and
real i ze t hei r goals. Mai nt ai ni ng connect i on requi res i nhi bi t i ng
t he "I" perspect i ve and processi ng i nst ead f r om t he "t hou" per-
spective (Hsu, 1981). The r equi r ement is t o "r ead" t he ot her' s
mi nd and t hus t o know what t he ot her is t hi nki ng or feeling. In
cont rast , wi t h an i ndependent self, it is t he i ndi vi dual ' s responsi -
bi l i t y t o "say what' s on one' s mi nd" i f one expect s t o be at t ended
t o or under st ood.
Co n s e q u e n c e s o f a n I n d e p e n d e n t o r a n I n t e r d e p e n d e n t
Vi e w o f t h e Se l f
Tabl e 1 present s a brief, highly si mpl i fi ed s umma r y o f some
o f t he hypot hesi zed differences bet ween i ndependent and i n-
t er dependent const r ual s of t he self. These const r ual s of sel f and
ot her are concept ual i zed as par t o f a r eper t oi r e of self-relevant
schemat a used t o evaluate, organi ze, and regul at e one' s experi -
ence and act i on. As schemat a, t hey ar e pat t er ns of one' s past
behavi ors as well as pat t er ns for one' s cur r ent and fut ure behav-
230 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
Table 1
Summary of Key Differences Between an Independent and an
Interdependent Construal of Self
Feature compared Independent Interdependent
Definition
Structure
Important features
Tasks
Role of others
Basis of self-esteem a
Separate from social context
Bounded, unitary, stable
Internal, private (abilities, thoughts,
feelings)
Be unique
Express self
Realize internal attributes
Promote own goals
Be direct; "say what's on your mind"
Self-evaluation: others important for
social comparison, reflected
appraisal
Ability to express self, validate
internal attributes
Connected with social context
Flexible, variable
External, public (statuses, roles,
relationships)
Belong, fit-in
Occupy one's proper place
Engage in appropriate action
Promote others' goals
Be indirect; "read other's mind"
Self-definition: relationships
with others in specific
contexts define the self
Ability to adjust, restrain self,
maintain harmony with
social context
a Esteeming the self may be primarily a Western phenomenon, and the concept of self-esteem should
perhaps be replaced by self-satisfaction, or by a term that reflects the realization that one is fulfilling the
culturally mandated task.
iors (Neisser, 1976). Markus and Wurf (1987) called this assort-
ment of self-regulatory schemata the self-system. Whenever a
task, an event, or a situation is self-relevant, the ensuing pro-
cesses and consequences are likely to be influenced by the na-
ture of the self-system. The self-system has been shown to be
instrumental in the regulation of intrapersonal processes such
as self-relevant information processing, affect regulation, and
motivation and in the regulation of interpersonal processes
such as person perception, social comparison, and the seeking
and shaping of social interaction (see Cant or & Ki hl st rom,
1987; Greenwald & Pratkanis, 1984; Markus & Wurf, 1987, for
reviews). The goal of this article is to further specify the role of
the self-system in behavior by examining how these divergent
cultural self-schemata influence individual experience.
In the current analysis, we hypothesize that the independent
versus interdependent construals of self are among the most
general and overarching schemata of the individual' s self-sys-
tem. These construals recruit and organize the more specific
self-regulatory schemata. 3 We are suggesting here, therefore,
that the exact organization of many self-relevant processes and
their outcomes depends crucially on whether these processes
are rooted in an independent construal of the self or whether
they are based primarily on an interdependent construal of the
self. For example, in the process of lending meaning and coher-
ence to the social world, we know that people will show a
heightened sensitivity to self-relevant stimuli. For those with an
independent view of self, this includes information relevant to
one's self-defining attributes. For one with an interdependent
view of self, such stimuli would include information about sig-
nificant others with whom the person has a relationship or in-
formation about the self in relation to another person.
Affect regulation involves seeking positive states and avoid-
ing negative ones. Positive states are those that enhance or pro-
mote one's view of the self, and negative states are those that
challenge this view. For a person with an independent view of
self, this involves seeking i nformat i on that confirms or en-
hances one's internal, private attributes. The most desirable situ-
ations are those that allow one to verify and express those im-
portant internal attributes and that convey the sense that one is
appropriately autonomous. In contrast, for a person with an
interdependent view of self, one might expect the most desir-
able states to be those that allow one to be responsive to one's
immediate context or that convey the sense that one is succeed-
ing in his or her interdependent relationships or statuses.
A third important function of the self-concept suggested by
Markus and Wurf (1987) is that of motivating persons, of mov-
ing them to action. The person with an independent view of self
should be motivated to those actions that allow expression of
one's important self-defining, i nner attributes (e.g., hardwork-
ing, caring, independent, and powerful), whereas the person
with an interdependent view of self should be motivated to
3 What these very general cultural self-schemata of independence or
interdependence mean for a given individual's articulated view of self
cannot be specified, however. The self-concept derives not only from
the cultural self-schema that is the focus herein but from the complete
configuration of self-schemata, including those that are a product of
gender, race, religion, social class, and one's particular social and devel-
opmental history. Not all people who are part of an independent cul-
ture will thus characterize themselves as independent, nor will all
those who live as part of an interdependent culture claim to be interde-
pendent. Within independent and interdependent cultures, there is
great diversity in individual self-definition, and there can also be
strong similarities across cultures. For example, many artists, whether
Japanese or American, may describe themselves as nonconformist,
innovative, and breaking with tradition. And many aspects of their
behavior are indeed very similar. Yet, nonconformity Japanese-style
and nonconformity American-style, although similar in some respects,
will not, because of the differences in their supporting cultural con-
texts, be identical. For Japanese, nonconformity is a privilege afforded
only to selected, talented individuals whose deviance from the norm of
interdependence is implicitly sanctioned by the rest of society. For
Americans, nonconformity is regarded as every individual's birthright.
CULTURE AND THE SELF 231
t hose act i ons t hat enhance or foster one' s rel at edness or connec-
t i on t o others. On t he surface, such act i ons coul d l ook r emar k-
abl y si mi l ar (e.g., wor ki ng i ncr edi bl y har d t o gai n admi ssi on t o a
desi rabl e college), but t he exact source, or etiology, o f t he ener-
gi zi ng mot i vat i on may be powerful l y different (De Vos, 1973;
Maehr & Ni chol l s, 1980).
In t he following sect i ons, we di scuss t hese i deas i n f ur t her
det ai l and review t he empi r i cal l i t erat ure, whi ch suggests t hat
t her e ar e si gni f i cant cogni t i ve, e mot i ona l , a nd mot i vat i onal
consequences of hol di ng an i ndependent or an i nt er dependent
vi ew of t he self.
Consequences f or Cognition
I f a cognitive act i vi t y i mpl i cat es t he self, t he out come of t hi s
act i vi t y wi l l depend on t he nat ur e of t he self-system. Specifi-
cally, t here are t hr ee i mpor t ant consequences o f t hese di vergent
sel f-syst ems for cogni t i on. Fi r st , we may expect t hose wi t h i n-
t er dependent selves t o be mor e attentive and sensitive t o ot hers
t han t hose wi t h i ndependent selves. The at t ent i veness and sensi-
tivity t o others, char act er i zi ng t he i nt er dependent selves, wi l l
result i n a relatively great er cognitive el abor at i on o f t he ot her or
of t he sel f-i n-rel at i on-t o-ot her. Second, among t hose wi t h i nt er-
dependent selves, t he uni t of r epr esent at i on o f bot h t he sel f and
t he ot her wi l l i ncl ude a r el at i vel y speci fi c soci al cont ext i n
which t he sel f and t he ot her ar e embedded. Thi s means t hat
knowl edge about persons, ei t her t he sel f or others, wi l l not be
abst r act and gener al i zed across cont ext s, but i nst ead wi l l re-
mai n specific t o t he focal cont ext . Thi r d, a consi der at i on of t he
soci al cont ext and t he r eact i ons o f ot hers may al so shape some
basic, nonsoci al cognitive activities such as cat egor i zi ng and
count erfact ual t hi nki ng.
In expl or i ng t he i mpact o f di vergent cul t ural const r ual s on
t hi nki ng, we assume t hat how peopl e t hi nk (the process) i n a
soci al si t uat i on cannot be easi l y separ at ed f r om what t hey t hi nk
about (the cont ent ; Shweder, 1990; Shweder & Bourne, 1984).
Extensive research on soci al cogni t i on i n t he past decade has
suggest ed t he power of cont ent i n soci al i nference (e.g., see Fi ske
& Taylor, 1984; Mar kus & Zaj onc, 1985, for reviews). It is t he
nat ur e of t he r epr esent at i on (e.g., self, anot her person, a weed,
or cl am chowder) t hat gui des at t ent i on, and t hat det er mi nes
what ot her rel evant i nf or mat i on wi l l be ret ri eved t o fill i n t he
gap o f avai l abl e sense dat a. For exampl e, i nves t i gat i ons by
DAndr ade (1981) and Johnson- Lai r d (1983) i ndi cat e t hat t he
great er t he f ami l i ar i t y wi t h t he st i mul us mat eri al s, t he mor e
el abor at e t he schemat a for f r ami ng t he pr obl em, and t he bet t er
t he pr obl em solving. I n general , t hen, how a given obj ect is
cul t ural l y cons t r ued and r epr esent ed i n me mor y shoul d i mpor -
t ant l y i nfl uence and even det er mi ne how one t hi nks about t he
object. Accordingly, t he di vergent r epr esent at i ons o f t he sel f we
descr i be shoul d be expect ed t o have var i ous consequences for
all cogni t i on rel evant t o self, others, or soci al rel at i onshi ps.
More interpersonal knowledge. I f t he mos t significant ele-
ment s of t he i nt e r de pe nde nt sel f ar e t he s el f - i n- r el at i on- t o-
ot hers el ement s, t her e wi l l be a need, as well as a st r ong nor ma-
t i ve d e ma n d , for knowi ng a nd unde r s t a ndi ng t he soci al
sur r oundi ng, par t i cul ar l y ot hers i n di r ect i nt er act i on wi t h t he
self. That is, i f peopl e conceive o f t hemsel ves as i nt er dependent
par t s of l arger soci al wholes, it is i mpor t ant for t hem t o be
4 . 0
. > ,
= a.9
3. 8
3. 7
"0
3.6
m~
8 3. s
L_
a. 3. 4
3. 3
E
Sel f to Ot her
Ot her t o S - "
Eastern Western
Cul t ur a l B a c k g r o u n d
Figure 2. Mean perceived similarity of self to other and other to self by
subjects with Eastern and Western cultural backgrounds.
sensitive t o and knowl edgeabl e about t he ot hers who are t he
copar t i ci pant s i n var i ous rel at i onshi ps, and about t he soci al
si t uat i ons t hat enabl e t hese rel at i onshi ps. Mai nt ai ni ng one' s re-
l at i onshi ps and ensuri ng a har moni ous soci al i nt er act i on re-
qui res a full under s t andi ng of t hese others, t hat is, knowi ng
how t hey ar e feeling, t hi nki ng, and likely t o act i n t he cont ext of
one' s rel at i onshi ps t o t hem. It follows t hat t hose wi t h i nt erde-
pendent selves may devel op a dense and ri chl y el abor at ed st ore
of i nf or mat i on about ot her s or of t he sel f i n rel at i on.
Ki t ayama, Mar kus, Tummal a, Kur okawa, and Kat o (1990)
exami ned t hi s i dea i n a st udy r equi r i ng si mi l ar i t y j udgment s
bet ween sel f and other. A t ypi cal Amer i can fi ndi ng is t hat t he
sel f is j udged t o be mor e di ssi mi l ar t o ot her t han ot her is t o t he
sel f (Hol yoak & Gor don, 1983; Srul l & Gael i ck, 1983). Thi s
fi ndi ng has been i nt er pr et ed t o i ndi cat e t hat for t he t ypi cal
Amer i can subject, t he r epr esent at i on o f t he sel f is mor e el abo-
r at ed and di st i nct i ve i n me mor y t han t he r epr esent at i on o f an-
ot her person. As a result, t he si mi l ar i t y bet ween sel f and ot her is
j udged to be less when t he quest i on is posed about a mor e
di st i nct i ve obj ect (Is se/ f si mi l ar t o other?) t han when t he ques-
t i on is posed about a less di st i nct i ve obj ect (Is other si mi l ar t o
self?.). If, however, t hose wi t h i nt er dependent selves have at l east
as much knowl edge about some ot hers as t hey have about t hem-
selves, t hi s Amer i can pat t er n of fi ndi ngs may not be found.
To t est t hese predi ct i ons, Ki t ayama et al. (1990) c ompa r e d
st udent s f r om East er n cul t ural backgr ounds (students f r om In-
dia) wi t h t hose f r om West er n cul t ural backgr ounds ( Amer i can
st udent s). As shown i n Fi gur e 2, for t he Wes t er n subj ect s,
Ki t ayama et al. r epl i cat ed t he pr i or fi ndi ngs i n whi ch t he sel f is
percei ved as significantly mor e di s s i mi l ar t o t he ot her t han is
t he ot her t o t he self. Such a fi ndi ng is consi st ent wi t h a br oad
range of st udi es showi ng t hat for i ndi vi dual s wi t h a West ern
backgr ound, supposedl y t hose wi t h i ndependent selves, self-
knowl edge is mor e di s t i nct i ve a nd densel y el abor at ed t han
knowl edge about ot her peopl e (e.g., Gr eenwal d & Prat kani s,
1984). Thi s pat t er n, however, was nonsi gni fi cant l y reversed for
t he I ndi an subjects, who j udged t he sel f t o be somewhat mor e
si mi l ar t o t he ot her t han is t he ot her t o t he self. It appear s, t hen,
232
HAZEL ROSE MARKUS
that for the latter, more interdependent subjects, knowledge
about others is relatively more elaborated and distinctive than
knowledge about the self. Asymmetry in similarity judgments
is an indirect way to evaluate knowledge accessibility, but a
more direct measure of cross-cultural differences in knowledge
of the other should reveal that those with interdependent selves
have more readily accessible knowledge of the other.
Context-specific knowledge of self and other. A second con-
sequence of having an interdependent self as opposed to an
independent self concerns the ways in which knowledge about
self and other is processed, organized, and retrieved from mem-
ory. For example, given an i nt erdependent self, knowledge
about the self may not be organized into a hierarchical structure
with the person's characteristic attributes (e.g., intelligent, com-
petent, and athletic) as the superordinate nodes, as is often as-
sumed in characterizations of the independent self. In other
words, those with interdependent selves are less likely to orga-
nize knowledge about the "self in general" or about the "other in
general?' Specific social situations are more likely to serve as the
unit of representation than are attributes of separate persons.
One learns about the self with respect to a specific other in a
particular context and, conversely, about the other with respect
to the self in a particular context.
In exploring variations in the nature of person knowledge,
Shweder and Bourne (1984) asked respondents in India and
America to describe several close acquaintances. The descrip-
tions provided by the Indians were more situationally specific
and more relational than those of Americans. Indi an descrip-
tions focused on behavior; they described what was done,
where it was done, and to whom or with whom it was done. The
Indi an respondents said, "He has no land to cultivate but likes
to cultivate the land of others," or "When a quarrel arises, he
cannot resist the temptation of saying a word" or "He behaves
properly with guests but feels sorry if money is spent on them?'
It is the behavior itself that is focal and significant rather than
the i nner attribute that supposedly underlies it. Notably this
tendency to provide the specific situational or interpersonal
context when providing a description was reported to character-
ize the free descriptions of Indians regardless of social class,
education, or literacy level. It appears, then, that the concrete-
ness in person description is not due to a lack of skill in ab-
stracting concrete instances to form a general proposition, but
rather a consequence of the fact that global inferences about
persons are typically regarded as not meaningful or informa-
tive.
Americans also describe other people in terms of the spe-
cifics of their behavior, but typically this occurs only at the
begi nni ng of relationships when the other is relatively un-
known, or if the behavior is somehow distinctive and does not
readily lend itself to a trait characterization. Rather than saying
"He does not disclose secrets" Americans are more likely to say
"He is discreet or principled?' Rather than "He is hesitant to
give his money away," Americans say "He is tight or selfish?'
Shweder and Bourne (1984) found that 46% of American de-
scriptions were of the context-free variety, whereas this was true
of only 20% from the Indian sample.
A study by J. G. Miller (1984) on patterns of explanation
among Indi an Hindus and Americans revealed the same ten-
dency for contextual and relational descriptions of behavior
AND SHINOBU KITAYAMA
among Indi an respondents. In the first phase of her study, re-
spondents generated two prosocial behaviors and two deviant
behaviors and then explained why each behavior was under-
taken. For example, in the prosocial case, respondents were
asked to "describe something a person you know well did re-
cently that you considered good for someone else:' Miller coded
the explanations for reference to dispositional explanations; for
reference to social, spatial, temporal location; and for reference
to specific acts or occurrences. Like Shweder and Bourne
(1984), she found that on average, 40% of the reasons given by
American respondents referred to the general dispositions of
the actor. For the Hi ndu respondents, dispositional explana-
tions constituted less than 20% of their responses.
In a second phase of the study, Miller (1984) asked both Amer-
ican and Indian respondents to explain several accounts of the
deviant behaviors generated by the Indi an respondents. For ex-
ample, a Hi ndu subject narrated the following incident:
This concerns a motorcycle accident. The back wheel burst on the
motorcycle. The passenger sitting in the rear jumped. The mo-
ment the passenger fell, he struck his head on the pavement. The
driver of the motorcycle--who is an attorney--as he was on his
way to court for some work, just took the passenger to a local
hospital and went on and attended to his court work. I personally
feel the motorcycle driver did a wrong thing. The driver left the
passenger there without consulting the doctor concerning the
seriousness of the injury--the gravity of the situation--whether
the passenger should be shifted immediately--and he went on to
the court. So ultimately the passenger died. (p. 972)
Respondents were asked why the driver left the passenger at
the hospital without staying to consult about the seriousness of
the passenger's injury. On average, Americans made 36% of
their attributions to dispositions of the actors (e.g., irresponsi-
ble, pursuing success) and 17% of their attributions to contex-
tual factors (driver's duty to be in court). In comparison, only
15% of the attributions of the Indians referred to dispositions,
whereas 32% referred to contextual reasons. Both the American
and the Indi an subjects focused on the state of the driver at the
time of the accident, but in the Indian accounts, the social role
of the driver appears to be very important to understanding the
events. He is obligated to his role, he has a job to perform.
Actions are viewed as arising from relations or interactions with
others; they are a product of obligations, responsibilities, or
commitments to others and are thus best understood with re-
spect to these interpersonal relations. This preference for con-
textual explanations has also been document ed by Dalai,
Sharma, and Bisht (1983).
These results call into question the exact nature of the funda-
mental attribution error (Ross, 1977). In this error, people, in
their efforts to understand the causes of behavior, suffer from
an inescapable tendency to perceive behavior as a consequence
of the internal, personal attributes of the person. Miller's (1984)
Indian respondents also explained events in terms of properties
or features of the person, yet these properties were their role
relationships--their socially determined relations to specific
others or groups. Because role relationships necessarily impli-
cate the social situation that embeds the actor, it is unclear
whether the explanations of the Indi an respondents can be
viewed as instances of the fundamental attribution error. It may
be that the fundamental attribution error is only characteristic
of those with an independent view of the self.
CULTURE AND THE SELF 233
0 . 6
t-
~
,_ 0 . 5
. 0 . 2
~. ~g. 0 . 4
<~ 0 . 3
" 9
o . . - 0 . 2
~ . m 0. 1
. o
o. 0
Amer i can st udent s
Japanese st udent s
Twent y Cont ext ual i zed
St at ement s Format
Test
Figure 3. Mean proportion of psychological attributes endorsed by
American and Japanese students in two self-description tasks.
The t endency t o descr i be a per son i n t er ms o f his or her
specific behavi or and t o speci fy t he cont ext for a given behavi or
is al so evi denced when t hose wi t h i nt er dependent selves pr o-
vi de sel f - descr i pt i ons. Cousi ns 0989) c ompa r e d t he sel f-de-
scr i pt i ons of Amer i can hi gh school and college st udent s wi t h
t he sel f-descri pt i ons o f Japanese hi gh school and college st u-
dents. He used t wo t ypes of free-response format s, t he or i gi nal
Twent y St at ement s Test (TST; Kuhn & Mc Pa r t l a nd, 1954),
whi ch si mpl y asks " Who Am I?" 20 consecut i ve t i mes, and a
modi f i ed TST, whi ch asks subj ect s t o descr i be t hemsel ves in
several specific si t uat i ons (me at home, me wi t h friends, and me
at school). Whe n r espondi ng t o t he or i gi nal TST, t he Japanese
sel f-descri pt i ons were l i ke t hose of t he I ndi ans in t he Shweder
and Bour ne (1984) study. They were mor e concret e and role
specific ("I pl ay t enni s on t he weekend"). In cont rast , t he Amer i -
can descr i pt i ons i ncl uded mor e psychol ogi cal t rai t or at t r i but e
char act er i zat i ons ("I a m opt i mi st i c, " a n d " I a m friendly"). How-
ever, i n t he modi f i ed TST, where a specific i nt er per sonal con-
t ext was pr ovi ded so t hat r espondent s coul d envi si on t he situa-
t i on (e.g., me at home) and pr esumabl y who was t her e and what
was bei ng done t o whom or by whom, t hi s pat t er n of results was
reversed. As shown i n Fi gur e 3, t he Japanese showed a st r onger
t endency t o char act er i ze t hemsel ves i n psychol ogi cal t rai t or
at t r i but e t er ms t han di d Amer i cans. I n cont r ast , Amer i cans
t ended t o qual i fy t hei r sel f-descri pt i ons, cl ai mi ng, for exampl e,
"I a m somet i mes l azy at homey
Cousi ns (1989) ar gued t hat t he or i gi nal TST essent i al l y iso-
l at es or di sembeds t he ' T' f r om t he r el at i onal or si t uat i onal
cont ext , and t hus sel f - descr i pt i on becomes art i fi ci al for t he Jap-
anese r es pondent s , who ar e mor e a c c us t ome d t o t hi nki ng
about t hemsel ves wi t hi n specific soci al si t uat i ons. For t hese re-
spondent s, t he cont ext ual i zed f or mat "Descr i be your sel f as you
are wi t h your fami l y" was mor e "nat ur al " because it l ocat es t he
sel f i n a habi t ual uni t of represent at i on, namel y i n a par t i cul ar
i nt er per sonal si t uat i on. Once a defi ni ng cont ext was speci fi ed,
t he Japanese r espondent s were deci dedl y mor e wi l l i ng t o make
gener al i zat i ons about t hei r behavi or and t o descr i be t hemsel ves
abst ract l y usi ng t rai t or at t r i but e charact eri zat i ons.
Amer i can st udent s, in cont r ast t o t hei r Japanese count er-
part s, were mor e at home wi t h t he or i gi nal TST because t hi s
t est elicits t he t ype of abst r act , si t uat i on-free sel f-descri pt i ons
t hat f or m t he core of t he Amer i can, i ndependent self-concept.
Such abst r act or gl obal charact eri zat i ons, accor di ng t o Cousi ns
(1989), reflect a cl ai m of bei ng a separat e i ndi vi dual whose
nat ur e is not bound by a specific si t uat i on. Whe n r espondi ng t o
t he cont ext ual i zed sel f - descr i pt i on quest i ons, t he Ame r i c a n
st udent s qual i fi ed t hei r descr i pt i ons as i f t o say "Thi s is how I
am at home, but don' t assume t hi s is t he way I a m everywhere?'
For Amer i can respondent s, selfness, pur e and simple, seems t o
t r anscend any par t i cul ar i nt er per sonal rel at i onshi ps.
Basi c cognition in an interpersonal context. One' s view o f sel f
c a n have an i mpact even on some evi dent l y nonsoci al cognitive
activities. I. Li u (1986) descr i bed t he emphasi s t hat t he Chi nese
pl ace on bei ng l oyal and pi ous t o t hei r super i or s and obedi ence
t o t hem, whet her t hey are parent s, empl oyers, or gover nment
officials. He cl ai med t hat most Chi nese adher e t o a specific rul e
t hat st at es " I f your super i or s ar e present , or i ndi rect l y involved,
in any si t uat i on, t hen you are t o r espect and obey t hem" (I. Li u,
1986, p. 78). The power and t he i nfl uence of t hi s rul e appear t o
go consi der abl y beyond t hat pr ovi ded by t he Amer i can admoni -
t i on t o "r espect one' s e l de r s : I. Li u (1986) ar gued t hat t he st an-
dar d of sel f-regul at i on t hat involves t he at t ent i on and consi der -
at i on o f ot hers is so pervasi ve t hat it may act ual l y const r ai n
verbal and i deat i onal fluency. He r easoned t hat t aki ng account
of ot hers i n every si t uat i on is oft en at odds wi t h i ndi vi dual as-
ser t i on or wi t h at t empt s at i nnovat i on or uni que expressi on.
Thi s means, for exampl e, t hat i n an unst r uct ur ed creat i vi t y t ask
i n which t he goal is t o generat e as many i deas as possible, Chi -
nese subj ect s may be at a relative di sadvant age. In a si mi l ar vein,
T. Y. Li u and Hsu (1974) suggest ed t hat consi der at i on of t he
rul e "r espect and obey ot hers" uses up cognitive capaci t y t hat
mi ght ot herwi se be devot ed t o a t ask, and t hi s may be t he rea-
son t hat Chi nese nor ms for some creat i vi t y t asks fall bel ow
Amer i can nor ms.
Char t i ng t he differences bet ween an i ndependent sel f and
i nt e r de pe nde nt sel f may al so i l l umi na t e t he cont r over s y
sur r oundi ng t he debat e bet ween Bl oom (1981, 1984) and Au
0983, 1984) over whet her t he Chi nese can r eason count erfac-
t ual l y (for a t hor ough review o f t hi s debat e, see Moser, 1989).
Bloom' s st udi es (1981) on t he count erfact ual began when he
asked Chi nese- speaki ng subj ect s quest i ons l i ke " I f t he Hong
Kong gover nment were t o pass a law r equi r i ng t hat all ci t i zens
bor n out si de of Hong Kong make weekl y r epor t s of t hei r activi-
t i es t o t he police, how woul d you react?" Bl oom not ed t hat his
r es pondent s cons i s t ent l y ans wer ed "But t he gover nment
hasn' t," "It c a n ' t ; or "It won' t?' Pr essed t o t hi nk about it anyway,
t he r espondent s became f r ust r at ed, cl ai mi ng t hat it was unnat u-
ral or un- Chi nese t o t hi nk i n t hi s way. Amer i can and Fr ench
r espondent s answered si mi l ar quest i ons r eadi l y and wi t hout
compl ai nt . Fr om t hi s and subsequent st udi es, Bl oom (198 l ,
1984) concl uded t hat Chi nese speaker s"mi ght be expect ed t ypi -
cal l y t o encount er difficulty i n mai nt ai ni ng a count erfact ual
perspect i ve as an active poi nt o f or i ent at i on for gui di ng t hei r
cognitive activities" (1984, p. 21).
Au (1983) chal l enged Bloom' s concl usi ons. Usi ng different
234 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
stimulus materials and also different translations of the same
stimulus materials, she reported that Chinese subjects per-
formed no differently from their Western counterparts. The
controversy continues, however, and many investigators remain
unconvinced that the differences Bloom and others have ob-
served in a large number of studies on counterfactual reasoning
are solely a function of awkward or improper translations of
stimulus materials.
Moser (1989), for example, discussed several of Bloom' s
(1981, 1984) findings that are not easily explained away. He
described the following question that Bloom (1981, pp. 53-54)
gave to Taiwanese, Hong Kong, and American subjects in their
native language.
Everyone has his or her own method for teaching children to re-
spect morality. Some people punish the child for immoral behav-
ior, thereby leading him to fear the consequences of such behavior.
Others reward the child for moral behavior, thereby leading him
to want to behave morally. Even though both of these methods
lead the child to respect morality, the first method can lead to
some negative psychological consequences--it may lower the
child's self-esteem.
According to the above paragraph, what do the two methods have
in common? Please select only one answer.
A. Both methods are useless.
B. They have nothing in common, because the first leads to nega-
tive psychological consequences.
C. Both can reach the goal of leading the child to respect mor-
ality.
D. It is better to use the second.
E. None of the above answers makes sense. (If you choose this
answer, please explain.)
Bloom (1984) reported that 97% of American subjects re-
sponded C, but that only 55% of the Taiwanese and 65% of the
Hong Kong respondents answered C. In explaining his results,
he wrote:
Most of the remaining Chinese-speaking subjects chose D or E
and then went on to explain, based on their own experience and
often at great length and evidently after much reflection, why, for
instance, the second method might be better, or why neither
method works, or why both methods have to be used in conjunc-
tion with each other, or perhaps, why some other specified means
is preferable. For the majority of these subjects, as was evident
from later interviewing, it was not that they did not see the para-
graph as stating that both methods lead the child to respect moral-
ity, but they felt that choosing that alternative and leaving it at that
would be misleading since in their experience that response was
untrue. As they saw it, what was expected, desired, must be at a
minimum an answer reflecting their personal considered opin-
ion, if not a more elaborated explanation of their own experiences
relevant to the matter at hand. Why else would anyone ask the
question? American subjects, by contrast, readily accepted the
question as a purely "theoretical" exercise to be responded to ac-
cording to the assumptions of the world it creates rather than in
terms of their own experiences with the actual world. (Bloom,
1981, p. 54)
It is our view that the differences in response between the
Americans and the Chinese may be related to whether the re-
spondent has an independent or interdependent construal of
the self. If one's actions are contingent on, determined by, or
made meaningful by one's relationships and social situations, it
is reasonable to expect that respondents with interdependent
selves might focus on the motivation of the person administer-
ing the question and on the nature of their current relationship
with this person. Consequently, i n the process of responding,
they might ask themselves, "What is being asked of me here?
What does this question expect of me or require from me?
What are potential ramifications of answering in one way or
another in respect to my relationship with this person?" In Le-
bra's (1976) terms, what is "my proper place?" in this social
interaction [i.e., me and the interviewer], and what are the "obli-
gations attached to [it?]" (p. 67). To immediately respond to the
question as a purely abstract or theoretical exercise would re-
quire ignoring the currently constituted social situation and the
nature of one's relationship with the other. This, of course, can
be done, but it does not mean that it will be easily, effortlessly, or
automatically done. And this is especially true when the prag-
matics of a given context appears to require just the opposite. It
requires ignoring the other's perspective and a lack of attention
to what the other must be thinking or feeling to ask such a
question. One's actions are made meaningful by reference to a
part i cul ar set of contextual factors. If these are ignored or
changed, then the self that is determined by them changes also.
Those with relatively unencumbered, self-contained, indepen-
dent selves can readily, and without hesitation, entertain any of
a thousand fanciful possible worlds because there are fewer
personal consequences--t he bounded, aut onomous self re-
mains essentially inviolate.
One important implication of this analysis is that people with
interdependent selves should have no disadvantage in counter-
factual reasoning if the intent of the questioner and the demand
of the situation is simply to test the theoretical reasoning capaci-
ties of the person. One such situation would involve an aptitude
test such as the Scholastic Aptitude Test (SAT). Indeed, on the
quantitative portion of the SAT that requires substantial hypo-
thetical and counterfactual reasoning (e.g., "If Tom walked 2
miles per hour, then how far will he have walked in 4 hours?"),
both Taiwanese and Japanese children perform considerably
better than their American peers (Stevenson et al., 1986).
It would appear important, therefore, to distinguish between
competence and performance or between the presence of par-
ticular inference skills and the application of these skills in a
particular pragmatic context (see also Laboratory of Compara-
tive Human Cognition, 1982). The discussion thus far implies
that regardless of the nature of the self-system, most people
with an adequate level of education possess the skills of hypo-
thetical reasoning and the ability to think in a counterfactual
fashion. Yet, the application of these skills in a particular situa-
tion varies considerably with the nature of the self-system.
Some people may invoke these skills much more selectively. For
those with interdependent selves, in contrast to those with inde-
pendent selves, a relatively greater proportion of all inferences
will be contingent on the pragmatic implications of a given
situation, such as the perceived demands of the interviewer, the
convention of the situation, and the rules of conversation.
Do styles of thinking and inference vary above and beyond
those that derive from the pragmatic considerations of particu-
lar social situations? This question has yet to be more carefully
addressed. However, given the tendency to see people, events,
and objects as embedded within particular situations and rela-
tionships, the possibility seems genuine. Chiu (1972), for exam-
ple, claimed that the reasoning of American children is charac-
CULTURE AND THE SELF 235
t er i zed by an i nf er ent i al - cat egor i cal style, whereas t he
reasoning of Taiwanese Chi nese subjects displays a rel at i onal -
contextual style. When Amer i can chi l dren descri bed why two
objects of a set o f three objects went together, t hey were likely t o
say "because t hey bot h live on a f ar m" In contrast, Chinese
chi l dren were mor e likely t o display a rel at i onal -cont ext ual
style, putting two human figures together and claiming t he two
go together "because the mot her takes care of the baby." I n the
latter case, t he emphasis is on synthesizing features into an orga-
nized whole. Bruner (1986) referred t o such differences as aris-
ing from a paradi gmat i c versus a narrative mode of thought. In
the former, the goal is abstraction and analyzing c ommon fea-
tures, in the latter, establishing a connect i on or an interdepen-
dence among t he elements.
Consequences f or Emotion
I n psychology, emot i on is often viewed as a universal set of
largely prewi red i nt ernal processes o f sel f-mai nt enance and
self-regulation (Buck, 1988; Darwi n, 1896; Ekman, 1972; Le-
Doux, 1987). Thi s does not mean, t hough, t hat emot i onal expe-
rience is also universal. On t he contrary, as suggested by ant hro-
pologists Rosaldo (1984), Lut z (1988), and Sol omon (1984), cul-
ture can play a central role in shapi ng emot i onal experience. As
with cognition, i f an emot i onal activity or reaction implicates
the self, the out come o f this activity will depend on the nature of
t he self-system. And apar t f r om t he fear i nduced by bri ght
lights and l oud sounds, or the pleasure produced by a sweet
taste, there are likely t o be few emot i ons t hat do not directly
implicate one' s view of t he self. Thus, Rosaldo (1984) cont ended
"feelings are not substances t o be discovered in our bl ood but
social practices organi zed by stories t hat we bot h enact and tell.
They are st ruct ured by our forms o f underst andi ng" (p. 143),
and we would add, specifically, by one' s const rual of t he self. I n
an extension of these ideas, Lut z (1988) argued t hat although
most emot i ons are viewed as universally experienced "nat ural "
human phenomena, emot i ons are anyt hi ng but natural. Emo-
tion, she cont ended, "can be viewed as cultural and interper-
sonal product s of nami ng, justifying, and persuading by people
in relationship t o each other. Emot i onal meani ng is t hen a so-
cial rather t han an individual achi evement - - an emergent prod-
uct o f social life" (Lutz, 1988, p. 5).
Among psychologists, several cognitively oriented theorists
of emot i on have suggested t hat emot i on is i mport ant l y impli-
cated and embedded in an actual social situation as const rued
by t he person (e.g., De Riviera, 1984; Roseman, 1984; Scherer,
1984). Accordingly, not onl y does the experience o f an emot i on
depend on the current const rual of the social situation (e.g.,
Frijda, Kuipers, & ter Schure, 1989; Shaver, Schwartz, Ki rson,
& O' Connor, 1987; C. Smith & EUsworth, 1987), but the experi-
enced emot i on in t ur n plays a pivotal role in changi ng and
t ransformi ng the very nature of the social situation by allowing
a new const rual of the situation t o emerge and, furthermore, by
instigating the person t o engage in cert ai n actions. Fr om the
current perspective, construals o f the social situation are con-
strained by, and largely derived from, const rual s of the self,
others, and t he relationship between t he two. Thus, emot i onal
experience should vary systematically with t he const rual of t he
self.
The present analysis suggests several ways in which emo-
tional processes may differ with t he nature of the self-system.
First, t he pr edomi nant eliciting conditions of many emot i ons
may differ markedly accordi ng t o one' s const rual of t he self.
Second, and mor e i mpor t ant , whi ch emot i ons will be ex-
pressed or experienced, and with what intensity and frequency,
may also vary dramatically.
Ego-focused versus other-focused emotions. The emot i ons
systematically vary accordi ng t o the extent t o which t hey follow
from, and also foster and reinforce, an i ndependent or an inter-
dependent const rual of the self. This is a di mensi on t hat has
largely been ignored in t he literature. Some emotions, such as
anger, frustration, and pride, have t he individual' s internal at-
tributes (his or her own needs, goals, desires, or abilities) as the
pr i mar y referent. Such emot i ons may be called ego focused.
They result most t ypi cal l y from t he bl ocki ng (e.g., "I was treated
unfairly"), t he satisfaction, or t he confi rmat i on (e.g., "I per-
formed better t han others") of one' s internal attributes. Experi-
encing and expressing these emot i ons furt her highlights these
self-defining, i nt ernal at t ri but es and leads t o addi t i onal at-
t empt s t o assert t hem in public and confi rm t hem in private. As
a consequence, for those with i ndependent selves t o operate
effectively, t hey have t o be "experts" i n the expression and expe-
rience of these emotions. They will manage the expression, and
even the experience, of these emot i ons so t hat t hey mai nt ai n,
affirm, and bolster the const rual o f t he self as an aut onomous
entity. The public display of one' s own internal attributes can be
at odds with t he mai nt enance of interdependent, cooperative
social interaction, and when unchecked can result in interper-
sonal confront at i on, conflict, and possibly even overt aggres-
sion. These negative consequences, however, are not as severe as
t hey might be for interdependent selves because t he expression
of one' s internal attributes is the culturally sanct i oned task o f
t he i ndependent self. I n short , t he cur r ent analysis suggests
that, in cont rast t o those with more interdependent selves, the
ego-focused emot i ons will be more frequently expressed, and
perhaps experienced, by those with i ndependent selves.
I n cont rast t o the ego-focused emotions, some other emo-
tions, such as sympathy, feelings of interpersonal communi on,
and shame, have anot her person, rather t han one' s internal at-
tributes, as the pr i mar y referent. Such emot i ons may be called
other focused. They typically result f r om being sensitive t o the
other, t aki ng the perspective of the other, and at t empt i ng t o
pr omot e interdependence. Experiencing these emot i ons high-
lights one' s i nt er dependence, facilitates t he reci procal ex-
changes of well-intended actions, leads t o further cooperative
social behavior, and t hus provides a significant form o f self-val-
idation for interdependent selves. As a consequence, for those
with interdependent selves t o operate effectively, t hey will have
t o be "experts" in the expression and experience of these emo-
tions. They will manage t he expression, and even the experi-
ence, o f these emot i ons so t hat t hey mai nt ai n, affirm, and rein-
force t he const rual of t he self as an interdependent entity. The
other-focused emot i ons often discourage the aut onomous ex-
pression of one' s internal attributes and may lead t o inhibition
and ambivalence. Al t hough among i ndependent selves these
consequences are experienced negatively (e.g., as timidity) and
can, in fact, have a negative impact, t hey are tolerated, among
interdependent selves, as t he "business o f living" (Kakar, 1978,
236 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
p. 34). Creat i ng and mai nt ai ni ng a connect i on t o ot her s is t he
pr i ma r y t ask of t he i nt er dependent self. I n short , t hi s anal ysi s
suggest s t hat , i n c ont r a s t t o t hos e wi t h mor e i nde pe nde nt
selves, t hese ot her-focused emot i ons wi l l be mor e frequently
expr essed and per haps even exper i enced among t hose wi t h i n-
t er dependent selves.
Ego-focused emotions--emotions that foster and create inde-
pendence. In a compar i s on o f Amer i can and Japanese under -
graduat es, Mat sumot o, Kudoh, Scherer, and Wal l bot t (1988)
f ound t hat Amer i can subj ect s r epor t ed exper i enci ng t hei r emo-
t i ons longer t han di d Japanese subjects, even t hough t he two
gr oups agr eed i n t hei r or der i ng of whi ch emot i ons were experi -
enced l ongest (i.e., j oy = sad > anger = gui l t > fear = shame =
disgust). Amer i cans al so r epor t ed feeling t hese emot i ons mor e
i nt ensel y t han t he Japanese and r epor t ed mor e bodi l y symp-
t oms (e.g., l ump i n t hr oat , change i n breat hi ng, mor e expressive
react i ons, and mor e verbal reactions) t han di d t he Japanese.
Finally, when asked what t hey woul d do t o cope wi t h t he conse-
quences of var i ous emot i onal events, significantly mor e of t he
Japanese st udent s r epor t ed t hat no act i on was necessary.
One i nt er pr et at i on of t hi s pat t er n o f fi ndi ngs may assume
t hat most of t he emot i ons exami ned, wi t h t he except i on of
shame and possi bl y gui l t , are what we have cal l ed ego-focused
emot i ons. Thus, peopl e wi t h i nde pe nde nt selves wi l l at t end
mor e t o t hese feelings and act on t he basi s of t hem, because
t hese feelings ar e r egar ded as di agnost i c of t he i ndependent
self. Not t o at t end t o one' s i nner feelings is oft en vi ewed as bei ng
i naut hent i c or even as denyi ng t he " r eal " self. I n cont r as t ,
among t hose wi t h mor e i nt er dependent selves, one' s i nner feel-
i ngs may be less i mpor t ant i n det er mi ni ng one' s consequent
actions. Ego-focused feelings may be r egar ded as by- pr oduct s of
i nt er per sonal rel at i onshi ps, but t hey may not be accor ded pri vi -
l eged st at us as regul at ors o f behavior. For t hose wi t h i nt erde-
pendent selves, it is t he i nt er per sonal cont ext t hat assumes pr i or -
ity over t he i nner at t ri but es, such as pri vat e feelings. The l at t er
may need t o be cont r ol l ed or de- emphasi zed so as t o effectively
fit i nt o t he i nt er per sonal cont ext .
Gi ven t hese di fferences in emot i onal processes, peopl e wi t h
di vergent selves may devel op very di fferent assumpt i ons about
t he et i ol ogy of emot i onal expressi ons for ego-focused emot i ons.
For t hose wi t h i ndependent selves, emot i onal expressi ons may
literally "express" or reveal t he i nner feelings such as anger,
sadness, and fear. For t hose wi t h i nt er dependent selves, how-
ever, an emot i onal expressi on may be mor e of t en r egar ded as a
publ i c i ns t r ument al act i on t hat may or may not be rel at ed di -
rectly t o t he i nner feelings. Consi st ent wi t h t hi s analysis, Mat su-
mot o (1989), usi ng dat a f r om 15 cul t ures, r epor t ed t hat i ndi vi d-
ual s f r om hi er ar chi cal cul t ures (that we woul d classify as bei ng
general l y i nt er dependent ; see Hofst ede, 1980), when asked t o
rat e t he i nt ensi t y of an angry, sad, or fearful emot i on di spl ayed
by an i ndi vi dual i n a phot ogr aph, gave l ower i nt ensi t y rat i ngs
t han t hose f r om less hi er ar chi cal cul t ures. Notably, al t hough t he
degr ee o f hi erarchy i nher ent i n one' s cul t ures was st rongl y re-
l at ed t o t he i nt ensi t y rat i ngs given t o t hose emot i ons, it was not
rel at ed t o t he cor r ect i dent i f i cat i on of t hese emot i ons. The one
except i on t o t hi s fi ndi ng was t hat peopl e f r om mor e hi erarchi -
cal cul t ures (those wi t h mor e i nt er dependent selves) were less
likely t o correct l y i dent i fy emot i onal expressi ons of happi ness.
Among t hose wi t h i nt er dependent selves (often t hose f r om hi er-
archi cal cultures), positive emot i onal expressi ons are most fre-
quent l y used as publ i c act i ons in t he servi ce of mai nt ai ni ng
i nt er per sonal har mony and, thus, are not regarded as par t i cu-
l arl y di agnost i c of t he act or' s i nner feelings or happi ness.
For t hose wi t h i nt er dependent selves (composed pr i mar i l y of
rel at i onshi ps wi t h ot hers i nst ead o f i nner attributes), it may be
very i mpor t ant not t o have i nt ense experi ences of ego-focused
emot i ons, and t hi s may be par t i cul ar l y t r ue for negative emo-
t i ons l i ke anger. Anger may seriously t hreat en an i nt er depen-
dent sel f and t hus may be highly dysfunct i onal . In fact, some
ant hr opol ogi st s expl i ci t l y chal l enge t he uni versal i st view t hat
al l peopl e experi ence t he same negative emot i ons. Thus, in Ta-
hi t i , anger is hi ghl y f ear ed, a nd var i ous ant hr opol ogi cal ac-
count s cl ai m t hat t here is no expressi on of anger in t hi s cul t ure
(see Levy, 1973; Sol omon, 1984). It is not t hat t hese peopl e have
l ear ned t o i nhi bi t or suppress t hei r "real " anger but t hat t hey
have l ear ned t he i mpor t ance o f at t endi ng t o others, consi der i ng
others, and bei ng gentle i n al l situations, and as a consequence
very little anger is elicited. I n ot her words, t he social real i t y is
const r ued and act ual l y const r uct ed in such a way t hat it does
not l end i t sel f t o t he st r ong experi ence, l et al one t he out bur st , of
negat i ve ego- f ocus ed e mot i ons such as anger. The s ame is
cl ai med for Ukt a Eski mos (Briggs, 1970). They are sai d not t o
feel anger, not to express anger, and not even t o t al k about
anger. The cl ai m is t hat t hey do not show anger even in t hose
ci r cumst ances t hat woul d cer t ai nl y pr oduce compl et e out rage
in Amer i cans. These Eski mos use a wor d t hat means"chi l di sh"
t o l abel angr y behavi or when it is observed in foreigners.
Among t he Japanese, t here is a si mi l ar concer n wi t h avert i ng
anger and avoi di ng a di sr upt i on of t he har mony of t he soci al
si t uat i on. As a consequence, exper i enci ng anger or receiving
anger signals may be relatively rare events. A st udy by Miyake,
Campos, Kagan, and Bradshaw (1986), which compar ed Japa-
nese and Amer i can infants of 11 mont hs o f age, provi des sug-
gestive evi dence for t hi s cl ai m. These investigators showed each
i nfant an i nt erest i ng t oy and pai r ed it wi t h a mot her ' s vocal
expressi on of joy, anger, or fear. Then t hey measur ed t he chi l d' s
l at ency to r esume l ocomot i on t owar d t he t oy aft er t he mot her ' s
ut t erance. The two gr oups of i nfant s di d not differ i n t hei r
react i ons t o expressi ons of j oy or fear. But, aft er an angr y vocal
expressi on o f t he mother, t here was a st r i ki ng difference be-
t ween t he two groups. The Japanese chi l dr en r esumed l ocomo- .
t i on t owar d t he t oy aft er 48 s, Amer i can chi l dr en aft er onl y 18 s.
It may be t hat t he Japanese chi l dr en are relatively mor e t r auma-
t i zed by t hei r mot her ' s anger expressi ons because t hese are such
rare events.
Notably, i n t he West, a cont roversy exists about t he need, t he
desirability, and t he i mpor t ance o f expressi ng one' s anger. As-
sumi ng a hydraul i c model o f anger, some argue t hat it is neces-
sar y t o express anger so as t o avoi d boi l i ng over or bl owi ng up at
a l at er poi nt (Pennebaker, 1982). Ot her s argue for t he i mpor -
t ance of cont rol l i ng one' s anger so as not t o ri sk losing control.
No such cont roversy appear s t o exi st among t hose in pr edomi -
nant l y i nt e r de pe nde nt cul t ur es, wher e a s eemi ngl y unchal -
l enged nor m di rect s i ndi vi dual s to r est r ai n t hei r i nner feelings
and par t i cul ar l y t he overt expressi on o f t hese feelings. I ndeed,
many i nt er dependent cul t ures have wel l -devel oped st rat egi es
t hat r ender t hem exper t at avoi di ng t he expressi on of negative
emot i ons. For exampl e, Bond (1986) r epor t ed t hat i n Chi na
CULTURE AND THE SELF 237
di scussi ons have a cl ear st r uct ur e t hat is expl i ci t l y desi gned t o
prevent conflict f r om er upt i ng. To begi n wi t h, di scussant s pres-
ent t hei r c ommon pr obl ems and i dent i fy al l t he const r ai nt s t hat
al l t he par t i ci pant s mus t meet . Onl y t hen do t hey st at e t hei r
own views. To West erners, such a pat t er n appear s as vague,
beat i ng ar ound t he bush, and not get t i ng t o t he hear t of t he
mat t er, but it is par t of a careful l y execut ed st rat egy of avoi di ng
confl i ct , and t hus per haps t he exper i ence of negative emot i ons.
Bond, i n fact, not ed t hat among school chi l dr en i n Hong Kong
and Taiwan, t her e is a t endency t o cooper at e wi t h opponent s
even in a compet i t i ve r ewar d st r uct ur e and t o rat e fut ure oppo-
nent s mor e positively t han ot hers who wi l l not be opponent s
(Li, Cheung, & Kau, 1979, 1982).
In a recent cross-cul t ural compar i s on of t he el i ci t i ng condi -
t i ons of several emot i ons, Mat s umot o et al. (1988) al so f ound
t hat Japanese r espondent s appear t o be avoi di ng anger in cl ose
rel at i ons. Specifically, for t he Japanese, closely rel at ed ot hers
were rarel y i mpl i cat ed i n t he exper i ence of anger. The Japanese
r epor t ed feeling anger pr i mar i l y i n t he presence o f strangers. It
t hus appear s t hat not onl y t he expressi on but al so t he experi -
ence o f such an ego- f ocus ed e mot i on as anger is effectively
avert ed wi t hi n an i nt er dependent st r uct ur e of rel at i on. Whe n
anger arises, it happens out si de o f t he exi st i ng i nt er dependence,
as in conf r ont at i on wi t h out -groups (e.g., Samur ai warfare i n
feudal Japan). I n cont r ast , Amer i cans and West ern Eur opeans
r epor t exper i enci ng anger pr i mar i l y in t he presence of closely
rel at ed others. Thi s is not surpri si ng, given t hat expressi ng and
exper i enci ng ego-focused, even negative emot i ons, is one vi abl e
way t o assert and affirm t he st at us of t he sel f as an i ndependent
entity. Cons i s t ent wi t h t hi s anal ysi s, St i pek, Wei ner, a nd Li
(1989) f ound t hat when de s c r i bi ng s i t uat i ons t hat pr oduc e
anger, Chi nese subj ect s were much mor e likely t han Amer i can
subj ect s t o descr i be a si t uat i on t hat happened t o someone else
("a guy on a bus di d not give up a seat t o an ol d woman"). For
Amer i cans, t he maj or st i mul us t o anger was t he si t uat i on where
t he i ndi vi dual was t he vi ct i m ("a f r i end broke a pr omi se t o
me").
Ot her emot i ons, such as pr i de or guilt, may al so differ ac-
cor di ng t o t he nat ur e of t he medi at i ng self-system. As wi t h
anger, t hese expressi ons may be avoi ded, or t hey wi l l assume a
s omewhat di f f er ent f or m. For exampl e, i f def i ned as bei ng
pr oud of one' s own i ndi vi dual at t ri but es, pride may mean hu-
bri s, and its expressi on may need t o be avoi ded for t hose wi t h
i nt er dependent selves. 4 Consi st ent wi t h t he i dea t hat pr i de i n
one' s own per f or mance may be i nhi bi t ed among t hose wi t h
i nt er dependent selves, St i pek et al. (1989) found t hat t he Chi -
nese were deci dedl y less likely t o cl ai m t hei r own successful
efforts as a source o f pr i de t han were Amer i cans. These investi-
gat ors al so r epor t ed t hat t he emot i on of gui l t t akes on some-
what different connot at i ons as well. Among t hose wi t h i ndepen-
dent selves, who are mor e likely t o hol d stable, cross-si t uat i onal
beliefs and t o consi der t hem sel f-defi ni t i onal , "vi ol at i ng a law
or a mor al pr i nci pl e" was t he mos t frequently ment i oned cause
o f gui l t . Amo n g Chi nese, however, t he mos t c o mmo n l y re-
por t ed source of gui l t was "hur t i ng ot hers psychologically."
Other-focused emotions--emotions that create and foster in-
terdependence. Those wi t h i nt er dependent selves may i nhi bi t
t he experi ence, or at l east t he expressi on, o f some ego-focused
emot i ons, but t hey may have a hei ght ened capaci t y for t he expe-
r i ence and expressi on o f t hose emot i ons t hat deri ve pr i mar i l y
from focusing on t he other. I n Japan and Chi na, for exampl e,
t here is a much great er i nci dence of cosl eepi ng, cobat hi ng, and
physi cal cont act bet ween mot her and chi l d t han is t ypi cal l y
t r ue i n mos t Wes t er n count r i es. The t r a di t i ona l J apanes e
mot her car r i es t he chi l d on her back for a l arge par t o f t he first 2
years. Lebr a (1976) cl ai med t hat Japanese mot her s t each t hei r
c hi l dr e n t o f ear t he pa i n o f l onel i ness, wher eas Wes t er ner s
t each chi l dr en how t o be al one. Japanese and Chi nese soci al i -
zat i on pract i ces may hel p t he chi l d devel op an i nt er dependent
sel f i n t he first place, and at t he same t i me, t he capaci t y for t he
exper i ence of a relatively great er var i et y of ot her-focused emo-
tions.
The gr eat er i nt er dependence t hat results bet ween mot her s
and t hei r chi l dr en i n Japan is reflected i n t he fi ndi ng t hat t he
cl assi fi cat i on of i nfant s accor di ng t o t he nat ur e of t hei r at t ach-
ment s t o t hei r mot her s (i.e., secure, ambi val ent , and avoidant)
depar t s mar kedl y f r om t he pat t er n t ypi cal l y obser ved i n West-
er n dat a. Specifically, many mor e Japanese i nfant s ar e classi-
fied as "ambi val ent l y at t ached" because t hey seem t o experi -
ence deci dedl y mor e stress following a br i ef separ at i on f r om t he
mot her t han do Amer i can i nfant s (Ai nswort h, Bell, & St ayt on,
1974; Mi yake, Chen, & Campos, i n press). Thi s fi ndi ng al so
i ndi cat es t hat a par adi gm l i ke t he t ypi cal st r anger si t uat i on is
i nherent l y l i nked t o an i ndependent vi ew o f sel f and, thus, may
not be appr opr i at e for gaugi ng at t achment i n non-West ern cul -
tures.
In Japan, soci al i zat i on pract i ces t hat fost er an i nt ense close-
ness bet ween mot her and chi l d give ri se t o t he feeling ofamae.
Amae is t ypi cal l y defi ned as t he sense of, or t he accompanyi ng
hope for, bei ng lovingly car ed for and involves dependi ng on
and pr esumi ng anot her ' s i ndul gence. Al t hough, as det ai l ed by
Kumagai and Kumagai (1985), t he exact meani ng of amae is
open t o some debat e, it is cl ear t hat "t he ot her " is essent i al .
Whe n a per son exper i ences amae, she or he "feels t he f r eedom
t o do what ever he or she wills" whi l e bei ng accept ed and car ed
for by ot hers wi t h few st ri ngs at t ached. Some say amae is a t ype
of compl et e accept ance, a phenomenal r epl i cat i on o f t he i deal
mot her - i nf ant bond (L. T. Doi , 1973). Fr om our poi nt o f view,
exper i enci ng amae wi t h respect t o anot her per son may be i nher -
ent in t he f or mat i on and mai nt enance o f a mut ual l y r eci pr ocal ,
4 In interdependent cultures, i f pride is overtly expressed, it may
often be directed to a collective, of which the self is a part. For example,
the Chinese anthropologist Hsu (1975) described an event in which a
Japanese company official showed a "gesture of devotion to his office
superior which I had never experienced in the Western world" (p. 215).
After talking to Hsu in his own small, plain office, the employee said,
"Let me show you the office of my section chief." He then took Hsu to a
large, elaborately furnished office, pointed to a large desk, and said
proudly, "This is the desk of my section chief." Hsu's account makes
clear that this was not veiled cynicism from the employee, just com-
plete, unabashed pride in the accomplishments of his boss. Americans
with independent self-systems can perhaps understand this type of
pride in another's accomplishment i f the other involved is one's relative,
but it is typically unfathomable in the case ofone' s immediate supervi-
sor. Without an understanding of the close alignment and interdepen-
dence that occurs between employees and supervisors, the emotion
experienced by the employee that prompted him to show offhis super-
visor's office would be incomprehensible.
238 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
i nt er dependent r el at i onshi p wi t h anot her per son. I f t he ot her
per son accept s one' s amae, t he r eci pr ocal r el at i onshi p is symbol -
ically compl et ed, l eadi ng t o a si gni fi cant f or m of self-valida-
t i on. If, however, t he ot her per son rejects one' s amae, t he rel a-
t i onshi p wi l l be i n j eopardy,
For t he pur pos e of compar i ng i ndi genous feelings, such as
amae, wi t h t he mor e uni versal ones, such as anger and happi -
ness, Ki t ayama and Mar kus (1990) used a mul t i di mens i onal
scal i ng t echni que, whi ch allows t he i dent i fi cat i on of t he di men-
si ons t hat i ndi vi dual s habi t ual l y or spont aneousl y use when
t hey make j udgment s about si mi l ar i t i es among var i ous emo-
t i ons. Recent st udi es have demons t r at ed t hat peopl e are capa-
bl e of di st i ngui shi ng among var i ous emot i ons on as many as
seven or ei ght cogni t i ve di mensi ons (Mauro, Sato, & Tucker,
1989; C. Smi t h & El l swort h, 1987). I n t hese st udi es, however,
t he di me ns i ons have been speci f i ed a pr i or i by t he exper i -
ment er and given expl i ci t l y t o t he r espondent s t o use i n descri b-
i ng t he emot i ons. Whe n t he di mens i ons are not pr ovi ded but
al l owed t o emer ge in mul t i di mens i onal scal i ng st udi es, onl y
t wo di mensi ons ar e t ypi cal l y i dent i fi ed: act i vat i on (or excite-
ment) and pl easant ness (e.g., Russell, 1980). And it appear s t hat
most West er n emot i ons can be r eadi l y l ocat ed on a ci r cumpl ex
pl ane defi ned by t hese t wo di mensi ons. Thus, al t hough peopl e
ar e capabl e of di s cr i mi nat i ng among emot i ons on a subst ant i al
number of di mensi ons, t hey habi t ual l y cat egori ze t he emot i ons
onl y on t he di mens i ons of act i vat i on and pl easant ness.
Mor e recently, Russel l (1983; Russell, Lewi cka, & Ni i t , 1989)
a ppl i e d t he s ame t e c hni que t o sever al non- Wes t er n cul t ur al
gr oups and r epl i cat ed t he Amer i can findings. He t hus ar gued
t hat t he lay under s t andi ng of emot i onal exper i ence may i ndeed
be universal. Russel l used, however, onl y t hose t er ms t hat have
cl ear count er par t s i n t he non-West ern gr oups he st udi ed. He
di d not i ncl ude any emot i on t er ms i ndi genous t o t he non-Wes-
t er n gr oups such as amae. It is possi bl e t hat once t er ms for such
i ndi genous feeling st at es are i ncl uded i n t he analysis, a new
di mensi on, or di mensi ons, may emerge. To expl ore t hi s possi bi l -
ity, Ki t ayama and Mar kus (1990) s ampl ed 20 emot i ons f r om t he
Japanese language. Ha l f of t hese t er ms were al so f ound in En-
glish and were sampl ed so t hat t hey evenly covered t he ci r cum-
pl ex space i dent i fi ed by Russell. The r emai ni ng t er ms were
t hose i ndi genous t o Japanese cul t ur e and t hose t hat pr esuppose
t he presence o f others. Some (e.g., fureai[feeling a close connec-
t i on wi t h someone el se] ) refer pr i mar i l y t o a positive associ a-
t i on wi t h ot hers (rather t han events t hat happen wi t hi n t he i ndi -
vi dual , such as success), whereas ot hers refer t o i nt er per sonal
i sol at i on and confl i ct (e.g., oi me [the feeling o f i ndebt edness] ) .
Japanese college st udent s r at ed t he si mi l ar i t y bet ween 2 emo-
t i ons for each o f t he 190 pai r s t hat coul d be made f r om t he 20
emot i ons. The mean percei ved si mi l ar i t y rat i ngs for t hese pai r s
were t hen submi t t ed t o a mul t i di mensi onal scaling.
Repl i cat i ng pa s t r esear ch, Ki t a y a ma a nd Ma r kus (1990)
i dent i fi ed t wo di mens i ons t hat closely cor r es pond t o t he activa-
t i on and t he pl easant ness di mensi ons. I n addi t i on, however, a
new di mens i on emerged. Thi s t hi r d di mens i on r epr esent ed t he
ext ent t o whi ch t he per son is engaged i n or di sengaged f r om an
i nt er per s onal r el at i onshi p. At t he i nt er per s onal engagement
end were what we have cal l ed ot her-focused emot i ons, such as
shame, f ureai [feeling a cl ose connect i on wi t h s omebody else],
and shi t ashi mi [feeling f ami l i ar ] , whereas at t he di sengagement
end were f ound some ego-cent ered emot i ons, such as pr i de and
tukeagari [feeling puffed up wi t h t he sense of sel f - i mpor t ance] ,
al ong wi t h sl eepi ness and bor edom. Thi s i nt er per sonal engage-
me nt - di s e nga ge me nt di mens i on al so di f f er ent i at ed bet ween
ot herwi se ver y si mi l ar emot i ons. Thus, pr i de and el at i on were
equal l y positive and hi gh i n act i vat i on, yet pr i de was percei ved
as consi derabl y less i nt er per sonal l y engaged t han elation. Fur-
t her mor e, anger and shame were very si mi l ar in t er ms o f activa-
t i on and pl easant ness, but shame was much hi gher t han anger
i n t he ext ent of i nt er per sonal engagement .
Mor e i mpor t ant , t hi s st udy l ocat ed t he i ndi genous emot i ons
wi t hi n t he t hr ee- di mensi onal st ruct ure, per mi t t i ng us t o under -
st and t he nat ur e of t hese emot i ons in reference t o mor e uni ver-
sal emot i ons. For i nst ance, amae was low in act i vat i on, and
nei t her positive nor negative, fairly aki n t o sleepiness, except
t hat t he f or mer was much mor e i nt er per sonal l y engaged t han
t he latter. Thi s may i ndi cat e t he passive nat ur e of amae, involv-
i ng t he hopef ul expect at i on of anot her person' s favor and i ndul -
gence wi t hout any active, agent i c sol i ci t at i on of t hem. Compl e-
t i on of amae depends ent i rel y on t he ot her person, and, there-
fore, amae is uni quel y ambi val ent i n its connot at i on on t he
pl easant ness di mensi on. Anot her i ndi genous emot i on, oime,
involves t he feeling of bei ng psychol ogi cal l y i ndebt ed t o some-
body else. Oi me was l ocat ed at t he very negative end of t he
pl easant ness di mensi on, percei ved even mor e negatively t han
such uni versal negative emot i ons as anger and sadness. The
ext r eme unpl easant ness of oi me suggests t he aversive nat ur e of
unmet obl i gat i ons and t he press of t he need t o fulfill one' s obl i -
gat i ons t o ot hers and t o r et ur n favors. It al so under scor es t he
si gni fi cance of bal anced and har moni ous rel at i onshi ps in t he
emot i onal life of t hose wi t h i nt er dependent selves.
The fi ndi ng t hat t he Japanese r espondent s clearly and reli-
abl y di s cr i mi nat ed bet ween ego-focused emot i ons and ot her-
focused emot i ons on t he di mens i on of i nt er per sonal engage-
ment versus di sengagement st rongl y suggests t he val i di t y of t hi s
di st i nct i on as an essent i al component of emot i onal experi ence
at l east a mong J apanes e and, per haps , a mong peopl e f r om
ot her cul t ures as well. In a mor e recent study, Ki t ayama and
Mar kus (1990) f ur t her t est ed whet her t hi s t heor et i cal di men-
si on of emot i on al so underl i es and even det er mi nes how fre-
quent l y peopl e may exper i ence vari ous emot i ons and whet her
t he frequency of emot i onal exper i ence vari es wi t h t hei r domi -
nant const r ual of sel f as i ndependent or i nt er dependent .
Ki t ayama and Mar kus (1990) first sampl ed t hree emot i ons
c ommon i n Japanese cul t ure t hat were expect ed t o fall under
one of t he five t ypes t heor et i cal l y deri ved f r om t he cur r ent anal -
ysis. These t ypes are l i st ed in Tabl e 2. Ego-focused positive
emot i ons (yuet ukan [feeling superi or], pri de, and tukeagari[feel-
i ng puffed up ] ar e t hose t hat are most t ypi cal l y associ at ed wi t h
t he conf i r mat i on or ful fi l l ment of one' s i nt er nal at t r i but es, such
as abilities, desires, and needs. Ego-focused, negative emot i ons
(anger, f ut ekusare [sulky feel i ng], and yokyuf uman [frust ra-
t i on] ) occur p r i ma r i l y when such i nt er nal at t r i but es ar e
bl ocked or t hr eat ened. Al so i ncl uded were t hose cor r espond-
ingly positive or negative emot i ons associ at ed with t he mai nt e-
nance or enhancement of i nt er dependence. Thus, t hree emo-
t i ons are commonl y associ at ed wi t h t he affi rmat i on or t he com-
pl et i on o f i nt e r de pe nde nt r el at i ons hi ps ( f ureai [feeling o f
connect i on wi t h someone] , shitashirni [feeling of f ami l i ar i t y t o
CULTURE AND THE SELF 239
someone] , sonkei [feeling of r espect for s omeone] ) and t hus
were desi gnat ed as positive and ot her focused. I n cont rast , some
negative emot i ons are t ypi cal l y deri ved f r om one' s fai l ure t o
offer or r eci pr ocat e favors t o rel evant ot her s and t hus t o fully
par t i ci pat e in t he rel at i onshi p. They are t hus closely l i nked t o
di st ur bance t o i nt er dependence and a consequent desi re t o re-
pai r t he di st ur bance. They i ncl ude oime [feeling of i ndebt ed-
ness ], shame, and guilt. Finally, as not ed before, i nt er dependent
selves are likely t o t ol erat e ambi val ence r egar di ng one' s i nt erde-
pendent st at us wi t h some rel evant others. Interestingly, some
emot i ons are uni quel y l i nked t o t hi s i nt er per sonal ambi val ence.
Thr ee such emot i ons (amae [hopeful expect at i on of ot hers' i n-
dul gence or favor], tanomi [feeling l i ke rel yi ng on someone] ,
and sugari [feeling l i ke l eani ng on s omeone] ) were exami ned.
Japanese r espondent s r epor t ed how frequently t hey experi -
enced each of t he 15 emot i ons l i st ed i n Table 2. The five-factor
st r uct ur e i mpl i ed by t he t heor et i cal desi gnat i on of t he 15 emo-
t i ons t o one of t he five t ypes was veri fi ed in a conf i r mat or y
fact or anal ysi s (J/)reskog, 1969). A cor r el at i on mat r i x for t he five
t ypes is given in Tabl e 3. Ther e was a st r ong cor r el at i on be-
t ween positive and negative ego-focused emot i ons, as may be
expect ed i f bot h of t hem are der i ved f r om and al so foster and
rei nforce an i ndependent const r ual o f self. Fur t her mor e, t hese
ego- f ocused emot i ons ar e cl earl y di s t i nct f r om t he ot her-fo-
cused emot i ons. Thus, nei t her positive nor negative ego-focused
emot i ons had any significant rel at i onshi p wi t h ot her-focused,
posi t i ve emot i ons. Int erest i ngl y, however, t hese ego- f ocused
emot i ons were significantly associ at ed wi t h t he ambi val ent and,
t o a l arger extent, wi t h t he negative ot her-focused emot i ons,
suggest i ng t hat t he exper i ence o f ego-focused emot i ons, ei t her
positive or negative, is r eadi l y accompani ed, at l east i n Japanese
culture, by t he felt di st ur bance of a rel at i onshi p and, thus, by a
st r ong need t o rest ore harmony. Alternatively, bei ng embedded
Tabl e 2
The 15 Emotions and Their Meaning
Emotion type
(factor) Emotion Meaning
Ego focused
Positive Yuetukan
Tukeagari
Pride
Negative Futekusare
Yokyufuman
Anger
Other focused
Positive Fureai
Ambivalent
Negative
Shitashimi
Sonkei
Amae
Tanomi
Sugari
Oime
Shame
Guilt
Feeling superior
Feeling puffed up with the
sense of self-importance
Sulky feeling
Frustration
Feeling of connection
with someone
Feeling of familiarity to
someone
Feeling of respect for
someone
Hopeful expectation of
someone' s indulgence
and favor
Feeling like relying on
someone
Feeling like leaning on
someone
Feeling of indebtedness
Tabl e 3
Correlations Among the Five Types of Emotions
Emotion 1 2 3 4 5
Ego focused
1. Positive
2. Negative .70 - -
Other focused
3. Positive - . 05 - . 18 - -
4. Ambivalent .35 .63 .40
5. Negative .49 .69 .18 .43
i n a highly r eci pr ocal rel at i on and feeling obl i ged t o cont r i but e
t o t he rel at i onshi p may somet i mes be percei ved as a bur den or
pressure, hence r ender i ng sal i ent some o f t he ego-focused emo-
tions. 5 Finally, t he t hr ee t ypes o f ot her-focused emot i ons (posi-
tive, ambi val ent , and negative) ar e al l positively cor r el at ed (see
Tabl e 3).
Can t he frequency o f exper i enci ng t he five t ypes of emot i ons
be pr edi ct ed by one' s pr edomi nant const r ual of sel f as i ndepen-
dent or i nt er dependent ? To addr ess t hi s issue, Ki t ayama and
Mar kus (1990) al so asked t he same r espondent s ei ght quest i ons
desi gned t o measur e t he ext ent t o whi ch t hey endor se an i nde-
pendent const r ual o f sel f (e.g., ' Ar e you a ki nd o f per son who
hol ds on t o one' s own view?"; "How i mpor t ant is it t o hol d on t o
one' s own view?") and ei ght cor r espondi ng quest i ons desi gned
t o measur e t he ext ent t o whi ch t hey endor se an i nt er dependent
const r ual of sel f (e.g., "Are you t he ki nd o f per son who never
forgets a favor pr ovi ded by others?"; "How i mpor t ant is it t o
never forget a favor pr ovi ded by others?"). Consi st ent wi t h t he
cur r ent analysis, t he frequency o f exper i enci ng bot h positive
and negative ego-focused emot i ons significantly i ncr eased wi t h
t he i ndependent const r ual of self. They were, however, ei t her
negatively rel at ed (for positive emot i ons) or unr el at ed (for nega-
t i ve emot i ons) t o t he i nt e r de pe nde nt c ons t r ua l o f self. I n
mar ked cont r ast t o t hi s pat t er n for t he ego-focused emot i ons,
al l t hr ee t ype s o f ot her - f ocus ed e mot i ons wer e si gni f i cant l y
mor e frequent l y exper i enced by t hose wi t h mor e i nt er depen-
dent const r ual s of self. These emot i ons, however, were ei t her
unr el at ed (for positive and negative ot her-focused emot i ons) or
negatively rel at ed (for t he ambi val ent emot i ons) t o t he i ndepen-
dent const r ual o f self.
Consequences for Motivation
The st udy of mot i vat i on cent ers on t he quest i on of why peo-
pl e initiate, t er mi nat e, and per si st i n specific act i ons in par t i cu-
l ar ci r cumst ances (e.g., At ki nson, 1958; Mook, 1986). The an-
swer given t o t hi s quest i on i n t he West usual l y involves some
t ype of i nt ernal , i ndi vi dual l y r oot ed need or mot i ve - - t he mo-
tive t o enhance one' s sel f-est eem, t he mot i ve t o achieve, t he
motive t o affiliate, t he mot i ve t o avoi d cognitive conflict, or t he
mot i ve t o self-actualize. These mot i ves are assumed t o be par t
5 On these occasions, perhaps interdependent selves are most clearly
aware of their internal attributes. Such awareness (the honne in Japa-
nese) may be typically accompanied by a situational demand (the tate-
mae in Japanese).
240 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
of t he uni que, i nt er nal core of a person' s self-system. But what is
t he nat ur e of mot i vat i on for t hose wi t h i nt er dependent self-sys-
tems? Wha t f or m does it take? How does t he ever-present need
t o at t end t o ot her s and t o gai n t hei r accept ance i nfl uence t he
f or m o f t hese i nt er nal , i ndi vi dua l mot i ves? Ar e t he mot i ves
i dent i fi ed i n West er n psychol ogy t he uni versal i nst i gat ors of
behavi or?
As wi t h cogni t i on and emot i on, t hose mot i vat i onal processes
t hat i mpl i cat e t he sel f depend on t he nat ur e o f t he self-system.
I f we assume t hat others wi l l be relatively mor e focal i n t he
mot i vat i on of t hose wi t h i nt er dependent selves, var i ous i mpl i -
cat i ons follow. Fi r st , t hose wi t h i nt er dependent selves shoul d
express, and per haps experi ence, mor e o f t hose mot i ves t hat are
soci al or t hat have t he ot her as referent. Second, as we have
not ed previously, for t hose wi t h i ndependent selves, agency will
be e xpe r i e nc e d as an effort t o expr ess one' s i nt e r na l needs,
rights, and capaci t i es and t o wi t hst and undue soci al pressure,
whereas among t hose wi t h i nt er dependent selves, agency wi l l
be exper i enced as an effort t o be recept i ve t o others, to adj ust t o
t hei r needs and demands, and t o r est r ai n one' s own i nner needs
or desires. Mot i ves rel at ed t o t he need t o express one' s agency or
c ompe t e nc y (e.g., t he achi evement mot i ve) ar e t ypi c a l l y as-
sumed t o be c o mmo n t o all i ndi vi dual s. Yet among t hose wi t h
i nt er dependent selves, st ri vi ng t o excel or accompl i sh chal l eng-
i ng t asks may not be i n t he servi ce o f achi evi ng separat eness
and autonomy, as is usual l y assumed for t hose wi t h i ndependent
selves, but i nst ead in t he servi ce of mor e fully real i zi ng one' s
c onne c t e dne s s or i nt er dependence. Thi r d, mot i ves t hat ar e
l i nked t o t he self, such as sel f - enhancement , self-consistency,
sel f-veri fi cat i on, sel f-affi rmat i on, and sel f - act ual i zat i on, may
assume a ver y different f or m dependi ng on t he nat ur e of t he sel f
t hat is bei ng enhanced, veri fi ed, or act ual i zed.
More interdependent motives? Mur r ay (1938) as s embl ed
what he bel i eved t o be a compr ehensi ve l i st o f human mot i va-
t i ons (see al so Hi l ga r d, 1953, 1987). Many o f t hese mot i ves
seem most rel evant for t hose wi t h i ndependent selves, but t he
l i st al so i ncl udes some mot i ves t hat shoul d have par t i cul ar sa-
lience for t hose wi t h i nt er dependent selves. These i ncl ude defer-
ence, t he need t o admi r e and wi l l i ngl y follow a superi or, t o serve
gladly; similance, t he need t o i mi t at e or emul at e others, to agree
and believe; affiliation, t he need t o f or m f r i endshi ps and associ -
ations; nurturance, t he need t o nouri sh, ai d, or pr ot ect anot her;
succorance, t he need t o seek ai d, pr oj ect i on, or sympat hy and t o
be dependent ; avoidance of blame, t he need t o avoi d bl ame,
ost r aci sm, or puni s hme nt by i nhi bi t i ng unconvent i onal i m-
pul ses and t o be well behaved and obey t he law; and abasement,
t he need t o compl y and accept puni s hment or sel f-deprecat i on.
Many of t he soci al mot i ves suggest ed by Mur r ay seem t o cap-
t ur e t he t ypes of st ri vi ngs t hat shoul d char act er i ze t hose wi t h
i nt er dependent selves. Whe n t he cul t ural i mperat i ve is t o seek
connect edness, soci al i nt egrat i on, and i nt er per sonal harmony,
most of t hese mot i ves shoul d be t ypi cal l y exper i enced by t he
i ndi vi dual as positive and desi rabl e. In cont r ast , when t he cul-
t ur al t ask cent ers on mai nt ai ni ng i ndependence and separat e-
ness, hol di ng any of t hese mot i ves t oo st rongl y (e.g., si mi l ance
and succorance) of t en i ndi cat es a weak or t r oubl ed personality.
Thus, Murray, for exampl e, gave t he need t o compl y t he pej ora-
tive l abel o f need for abasement.
The l i mi t ed evi dence for t he i dea t hat t hose wi t h i nt er depen-
dent selves wi l l experi ence mor e of t he social or i nt er dependent
mot i ves comes f r om Bond (1986), who s u mma r i z e d sever al
st udi es expl or i ng t he mot i ve pat t er ns of t he Chi nese (see al so
McCl el l and, 196 l ). He f ound t hat t he level of vari ous motives
are a fairly di r ect refl ect i on of t he col l ect i vi st or gr oup- or i ent ed
t r adi t i on of t he Chinese. Thus, Chi nese r espondent s show rel a-
t i vel y hi gh levels o f need for abas ement , soci al l y or i ent ed
achi evement , change, endurance, i nt r acept i on, nurt urance, and
order; moder at e levels of autonomy, deference, and domi nance,
and succorance; and low levels o f i ndi vi dual l y or i ent ed achieve-
ment , affiliation, aggressi on, exhi bi t i on, heterosexuality, and
power. The soci al l y or i ent ed achi evement motive has, as its ulti-
mat e goal, a desire t o meet expect at i ons o f significant others,
whereas t he i ndi vi dual l y or i ent ed achi evement motive i mpl i es a
st r i vi ng for achi evement for i t s own sake ( di scussed later).
Hwang (1976) found, however, t hat wi t h cont i nui ng r api d social
change i n Chi na, t here is an i ncrease i n levels of exhi bi t i on,
autonomy, i nt r acept i on, and het erosexual i t y and a decrease in
levels of deference, order, nur t ur ance, and endurance. Int erest -
ingly, it appear s t hat t hose wi t h i nt er dependent selves do not
show a great er need for affiliation, as mi ght at first be t hought ,
but i nst ead t hey exhi bi t hi gher levels of t hose motives t hat re-
flect a concer n wi t h adj ust i ng onesel f so as t o occupy a pr oper
pl ace wi t h r espect t o others.
The motive for cognitive consistency. Anot her powerful mo-
tive assumed t o fuel t he behavi or of West erners is t he need to
avoi d or r educe cognitive conflict or di ssonance. Classic di sso-
nance occurs when one says one t hi ng publ i cl y and feels an-
other, qui t e cont r ast i ng t hi ng privately (Fest i nger & Carl smi t h,
1959). And such a confi gurat i on pr oduces par t i cul ar difficulty
when t he pri vat e at t i t ude is a sel f-defi ni ng one ( Gr eenwal d,
1980). One mi ght argue, however, t hat t he st at e of cognitive
di s s ona nc e ar i s i ng f r om c ount e r a t t i t udi na l behavi or is not
likely t o be exper i enced by t hose wi t h i nt er dependent selves.
Fi r st it is t he i ndi vi dual s' roles, statuses, or positions, and t he
c ommi t me nt s , obl i gat i ons, a nd r es pons i bi l i t i es t hey confer,
t hat are t he const i t uent s o f t he self, and in t hat sense t hey are
self-defining. As out l i ned in Fi gur e l , one' s i nt er nal at t ri but es
(e.g., private at t i t udes or opi ni ons) are not regarded as t he signifi-
cant at t r i but es of t he self. Fur t her mor e, one' s private feelings
are to be regul at ed i n accor dance wi t h t he requi rement s of t he
si t uat i on. Rest r ai nt over t he i nner sel f is assi gned a much hi gher
val ue t han is expressi on of t he i nner self. Thus, Ki ef er (1976)
wrote:
Although Japanese are often acutely aware of discrepancies be-
tween inner feelings and outward role demands, they think of the
l a t t e r . . , as the really important center of the self. Regarding
feelings as highly idiosyncratic and hard to control, and therefore
less reliable as sources of self-respect than statuses and roles, the
Japanese tends to include within the boundaries of the concept of
self much of the quality of the intimate social group of which he is
a member. (R. J. Smith 1985, p. 28)
Mor e recently, T. Doi (1986) has ar gued t hat Amer i cans are
deci dedl y mor e concer ned wi t h consi st ency bet ween feelings
and act i ons t han are t he Japanese. I n Japan t here is a vi r t ue in
cont rol l i ng t he expressi on o f one' s i nner mos t feelings; no vi r t ue
accr ues f r om expressi ng t hem. Tr i andi s (1989), for exampl e,
r epor t ed a st udy by Iwao (1988), who gave r espondent s a series
o f scenari os and asked t hem t o j udge which responses woul d be
CULTURE AND THE SELF 241
appr opr i at e for t he per son descr i bed i n t he scenari o. I n one
scenari o, t he daught er br i ngs home a per son f r om anot her race.
One of t he possi bl e r esponses given was "t hought t hat he woul d
never al l ow t hem t o ma r r y but t ol d t hem he was i n favor of t hei r
ma r r i a ge " Thi s answer was r at ed as best by onl y 2% of Amer i -
cans. I n shar p cont r ast , however, it was r at ed as best by 44% of
t he Japanese. Among t he Amer i cans, 48% t hought it was t he
wor st response, whereas onl y 7% of t he Japanese r at ed it as t he
worst.
Common motives in an interdependent context. Of t hose mo-
tives assumed by Mur r ay (1938) and Hi l gar d (1987) t o be univer-
sally significant, t he achi evement mot i ve is t he most wel l -docu-
ment ed exampl e. Variously defi ned as t he desi re t o over come
obst acl es, t o exer t power, t o do somet hi ng as well as possible, or
t o mast er, mani pul at e, or organi ze physi cal objects, huma n be-
ings, or i deas ( Hal l & Lindzey, 1957; Hi l gar d, 1987), t he achieve-
ment mot i ve is t hought t o be a f undament al huma n charact eri s-
tic. However, t he dri ve for achi evement i n an i nt er dependent
cont ext may have some very different aspect s f r om t he motive
for achi evement i n an i ndependent cul t ural cont ext . In a recent
anal ysi s of t he cont ent and st r uct ur e of val ues in seven cul t ures
(i.e., Aust ral i a, Uni t ed States, Spai n, Fi nl and, Germany, Israel,
and Hong Kong), S. H. Schwart z and Bi l sky (1990) f ound a
confl i ct bet ween values t hat emphasi ze i ndependent t hought
and act i on and t hose t hat emphasi ze r est r ai ni ng of one' s own
i mpul ses i n all sampl es except Hong Kong. In t he Hong Kong
sampl e, sel f-rest rai nt appear ed t o be qui t e compat i bl e wi t h i nde-
pendent t hought and act i on.
Al t hough al l i ndi vi dual s may have some desi re for agency or
cont r ol over t hei r own act i ons, t hi s agency can be accompl i shed
i n vari ous ways (Maehr, 1974). Pushi ng onesel f ahead of ot hers
and actively seeki ng success does not appear t o be uni versal l y
val ued. An i l l umi nat i ng anal ysi s o f cont r ol mot i vat i on by Weisz
et al. (1984) suggests t hat act i ng on t he worl d and al t eri ng t he
worl d may not be t he cont r ol st rat egy of choi ce for al l peopl e.
I nst ead, peopl e i n many Asi an cul t ures appear t o use what is
t er med secondary control. Thi s involves accommodat i ng t o ex-
i st i ng real i t i es "somet i mes vi a acts t hat l i mi t i ndi vi dual i sm and
per sonal aut onomy but t hat enhance per cei ved al i gnment or
goodness of fit wi t h peopl e, objects, or ci r cumst ances" (Weisz
et al., 1984, p. 956).
The Ame r i c a n not i on o f achi evement i nvol ves br e a ki ng
away, pushi ng ahead, and gai ni ng cont r ol over sur r oundi ngs.
How do selves concer ned wi t h fitting i n and accommodat i ng t o
exi st i ng real i t i es achieve? The quest i on of achi evement mot i ve
i n an i nt er dependent cont ext is al l t he mor e compel l i ng be-
cause ma ny o f t he mos t col l ect i ve soci et i es o f t he wor l d
cur r ent l y appear ext r emel y pr eoccupi ed wi t h achi evement . In
an anal ysi s o f Chi nes e chi l dr en' s st or i es, for exampl e, Bl u-
me nt ha l (1977) f ound t hat t he mos t c o mmo n behavi or was
achi evement - or i ent ed i n nat ure, t he second mos t frequent was
al t r ui sm, and t he t hi r d was soci al and per sonal responsibility.
Among j uni or hi gh school st udent s i n Japan, t he mot t o "pass
wi t h four, fail wi t h five" is now common. Thi s refers t o t he fact
t hat i f one is sl eepi ng 5 hr a night, he or she is pr obabl y not
st udyi ng har d enough t o pass exams. It appear s, however, t hat
t hi s st r ong emphasi s on achi evement mot i vat i on is, i n par t ,
ot her mot i vat ed. It is mot i vat ed by a desi re t o fit i nt o t he gr oup
and t o meet t he expect at i ons of t he group. I n t he chi l d' s case,
t he gr oup is t he family, and t he chi l d' s mi ssi on is t o enhance t he
soci al st andi ng o f t he f ami l y by gai ni ng admi ssi on t o one of t he
t op universities. The mot i ve t o achi eve need not necessar i l y
reflect a mot i ve t o achieve for "me" per sonal l y ( Maehr & Ni -
cholls, 1980). It can have soci al or collective origins. Chi l dr en
are st ri vi ng t o achi eve t he goal s o f others, such as f ami l y and
t eacher s, wi t h whom t hey ar e r eci pr ocal l y i nt er dependent .
Cons i s t ent wi t h t hi s not i on, Yu (1974) r e por t e d t hat t he
st rengt h of achi evement mot i vat i on was cor r el at ed positively
wi t h f ami l i sm and filial pi et y St ri vi ng for excellence necessar-
ily involves some di st anci ng or separ at i ng f r om some others,
but t he separ at i on allows t he chi l d t o pr oper l y accompl i sh t he
t ask of t he st udent and t hus t o fulfill his or her role wi t hi n t he
f ami l y
Several st udi es by Yang (Yang, 1982/1985; Yang & Liang,
1973) have sought t o di st i ngui sh bet ween t wo t ypes of achieve-
meri t mot i vat i on: i ndi vi dual l y or i ent ed and soci al l y ori ent ed.
I ndi vi dual l y or i ent ed achi evement mot i vat i on is vi ewed as a
funct i onal l y aut onomous desi re i n whi ch t he i ndi vi dual strives
t o achieve some i nt er nal i zed s t andar ds of excellence. I n con-
t r ast , soci al l y or i ent ed achi evement mot i vat i on is not funct i on-
ally aut onomous; rather, i ndi vi dual s persevere t o fulfill t he ex-
pect at i ons o f si gni fi cant ot hers, t ypi cal l y t he f ami l y ( Bond,
1986). Wi t h soci al l y or i ent ed achi evement , when t he specific
achi evement goal is met , t he i nt ense achi evement mot i vat i on
formerl y evi dent may appear t o vanish. Thi s anal ysi s i ndeed fits
many anecdot al r epor t s i ndi cat i ng t hat once admi t t ed i nt o t he
college of t hei r choice, or hi r ed by t hei r pr ef er r ed company,
Japanese high school and college st udent s ar e no l onger par t i cu-
l arl y i nt erest ed i n achi evement .
Once a new goal is est abl i shed, of course, t he soci al l y ori -
ent ed achi evement motive may be easily reengaged by any fig-
ure who can serve as a symbol i c subst i t ut e for f ami l y members.
A l ongi t udi nal survey conduct ed i n Japan over t he l ast 30 years
(Hayashi , 1988) has repeat edl y shown t hat appr oxi mat el y 80%
o f t he Japanese, regardl ess of sex, age, educat i on, and soci al
class, pr ef er a ma na ge r wi t h a f at her l i ke char act er (who de-
mands a l ot mor e t han officially r equi r ed i n t he work, yet ex-
t ends his car e for t he person' s per sonal mat t er s even out si de of
work) over a mor e West er n- t ype, t ask- or i ent ed manager (who
separat es per sonal mat t er s f r om wor k and demands as much as,
yet no mor e t han, officially required). I n a l arge number of
surveys and exper i ment s, Mi s umi and his col l eagues ( summa-
r i zed i n Mi s umi , 1985) have de mons t r a t e d t hat i n J apan a
l eader who is bot h demandi ng and per sonal l y car i ng is most
effective regardl ess of t he t ask or t he popul at i on exami ned (e.g.,
college st udent s, whi t e-col l ar workers, and bl ue-col l ar workers).
Thi s is i n mar ked cont r ast t o t he maj or concl usi on r eached in
t he l eadershi p l i t erat ure i n t he Uni t ed States, which suggests
t hat l eadershi p effectiveness depends on a compl ex i nt er act i on
bet ween char act er i st i cs of leaders, char act er i st i cs of followers,
and, mos t i mpor t ant , on t he nat ur e of t he t ask (Fiedler, 1978;
Hol l ander, 1985). Accor di ng t o our analysis, i n Japan as well as
in ot her i nt er dependent cultures, it is t he per sonal at t achment
t o t he l eader and t he ensui ng obl i gat i on t o hi m or her t hat most
st rongl y mot i vat e peopl e t o do t hei r work. Mot i vat i on medi at ed
by a st r ong per sonal rel at i onshi p, t hen, is unl i kel y t o be cont i n-
gent on factors associ at ed wi t h t he specific t ask or envi ron-
ment .
242 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
The self-related motives. The motive t o mai nt ai n a positive
view o f the self is one motive t hat psychologists since James
(1890) t hrough Greenwal d (1980), Hart er (1983), Steele (1988),
and Tesser (1986) have assumed t o be universally true. What
constitutes a positive view o f self depends, however, on one' s
const rual of t he self. 6 For those with i ndependent selves, feeling
good about onesel f t ypi cal l y requires fulfilling t he tasks asso-
ciated with being an i ndependent self; t hat is, being unique,
expressing one' s i nner attributes, and asserting onesel f (see Ta-
ble 1). Al t hough not uncont est ed, a reasonable empi ri cal gener-
alization f r om the research on self-related motives is t hat West-
erners, particularly those with high self-esteem, t ry to enhance
themselves whenever possible, and this t endency results in a
pervasive self-serving bias. Studies with Amer i can subjects dem-
onstrate t hat t hey take credit for their successes, explain away
their failures, and in various ways t r y t o aggrandi ze themselves
(e.g., Gilovich, 1983; Lau, 1984; J. B. Miller, 1986; Whitley &
Frieze, 1985; Zuckerman, 1979). Mai nt ai ni ng self-esteem re-
quires separating onesel f from others and seeing onesel f as dif-
ferent from and better t han others. At 4 years old, children
already show a clear self-favorability bias (Harter, 1989). When
asked t o compar e themselves with others with respect t o intelli-
gence, friendliness, or any skill, most children t hi nk t hey are
better t han most others. Wylie (1979) report ed t hat Amer i can
adults also consider themselves t o be mor e intelligent and more
attractive t han average, and Myers (1987), in a national survey
o f Amer i can students, found t hat 70% o f students believe t hey
are above average in leadership ability, and with respect t o the
"ability t o get along with ot hers, ' 0% t hought t hey were below
average, 60% t hought t hey were in the t op 10%, and 25% t hought
t hey were in the t op 1%. Moreover, as document ed by Taylor
and Brown (1988), among Ameri cans, most people feel t hat
t hey are more in cont rol and have more positive expectations
for themselves and their future t han t hey have for ot her people.
Thi s t endency t owar d false uni queness pr esumabl y derives
from efforts of those with i ndependent selves t o mai nt ai n a
positive view of themselves.
The motive t o mai nt ai n a positive view o f t he self may as-
sume a somewhat different form, however, for those with inter-
dependent selves. Feeling good about one' s interdependent self
may not be achieved t hr ough enhancement of the value at-
t ached t o one' s internal attributes and the at t endant self-serving
bias. Inst ead, positive feelings about t he self should derive from
fulfilling t he tasks associated with being interdependent with
relevant others: belonging, fitting in, occupyi ng one' s proper
place, engaging in appropri at e action, pr omot i ng others' goals,
and mai nt ai ni ng har mony (see Table 1). This follows for at least
two reasons. First, peopl e with i nt erdependent selves are likely
t o be motivated by other-focused emotions, such as empat hy
and oi me (i.e., the feeling of psychological indebtedness) and t o
act in accordance with t he perceived needs and desires of their
part ners in social relations, and this may produce a social dy-
nami c where individuals strive t o enhance each other' s self-es-
t eem. I n such r eci pr ocal rel at i onshi ps, ot her enhancement
coul d be more i nst rument al t o self-enhancement t han direct
attempts at self-enhancement because the latter are likely t o
isolate t he individual f r om t he net work of reciprocal relation-
ships. Second, self-esteem among those with interdependent
selves may be based in some large measure on their capacity to
exert control over their own desires and needs so that they can
i ndeed belong and fit in. As not ed earlier (see also Weisz et al.,
1984), such self-control and self-restraint are i nst rument al to
the ability t o flexibly adjust t o social contingencies and thus are
highly valued in interdependent cultures. Indeed, self-restraint
together with flexible adjustment is often regarded as an impor-
t ant sign of the moral mat uri t y of the person.
A developmental study by Yoshida, Kojo, and Kaku (1982,
Study 1) has document ed t hat self-enhancement or self-promo-
t i on are perceived quite negatively in Japanese culture. Second
(7-8 years old), third (8-9 years old), and fifth graders (10-11
years old) at a Japanese elementary school were asked how their
classmates (including themselves) would evaluate a hypotheti-
cal peer who comment ed on his own superb athletic perfor-
mance either in a modest , self-restrained way or in a self-en-
hanci ng way. The evaluation was solicited on the di mensi on of
personality ("Is he a good person?") and on the di mensi on o f
ability ("Is he good at [the relevant athletic domai n] ?") . As
shown in Figure 4A, the personality o f the modest peer was
perceived much more positively t han was t hat of the self-en-
hanci ng peer. Furthermore, this difference became more pro-
nounced as the age (grade) o f the respondents increased. A simi-
lar finding also has been reported for Chinese college students
in Hong Kong by Bond, Leung, and Wan (1982), who found
t hat individuals giving humble or self-effacing attributions fol-
lowing success were liked better t han those giving self-enhanc-
ing attribution. The most intriguing aspect of the Yoshida et al.
(1982) study, however, is their finding for the ability evaluation,
which showed a compl et e crossover interaction (see Figure 4B).
Whereas the second graders t ook the comment of the peer at
face value, perceiving the self-enhancing peer to be more com-
petent t han t he modest peer, this trend disappeared for the
t hi r d graders, and t hen compl et el y reversed for t he fifth
graders. Thus, the fifth graders perceived t hat the modest peer
was more compet ent t han the self-enhancing peer. These find-
ings indicate t hat as children are socialized in an interdepen-
dent cul t ural cont ext , t hey begi n t o appreci at e t he cul t ural
value of self-restraint and, furthermore, to believe in a positive
association between self-restraint and ot her favorable attributes
of the person not only in the social, emot i onal domai ns but also
in the domai ns of ability and competence. Although it is cer-
tainly possible for those with i ndependent selves to overdo their
self-enhancement (see Schlenker & Leary, 1982), for the most
part, the Amer i can prescription is to confidently display and
express one' s strengths, and those who do so are evaluated posi-
tively (e.g., Greenwald, 1980; Mullen & Ri ordan, 1988).
Self- or other-serving bias. Given the appreciation that those
with interdependent selves have for self-restraint and self-con-
trol, the various self-enhancing biases t hat are c ommon in West-
ern culture may not be prevalent in many Asian cultures. In an
initial exami nat i on o f potential cultural variation in the ten-
dency t o see onesel f as different f r om others, Mar kus and
Ki t ayama (in press) admi ni st ered questionnaires containing a
series o f false-uniqueness items t o large classes of Japanese col-
lege students in Japan and t o large classes of Ameri can college
6 For a compelling analysis of how self-esteem is related to culture,
see Solomon, Greenberg, and Pyszczynski (in press).
CULTURE AND THE SELF 243
l -
. _ o
( g
m
( g
>
LIJ
1 4 - -
1 2 -
1 0 - -
8
A. Per sonal i t y B. Abi l i t y
1 4 - -
i -
. O 1 2
- I
i
m
> l O
LI,,I
- - - 0 - Modest
X Sel f - enhanci ng
I s I I I I
2 3 5 2 3 5
G r a d e G r a d e
1 5 = m o a t p o s i t i v e
1 = m o s t n e g a t i v e
Figure 4. Mean evaluations by second, third, and fifth graders. (A: Personality of target person. B: Ability
of target person. Drawn from results reported by Yoshida, Kojo, and Kaku, 1982.)
st udent s i n t he Uni t ed States. I n bot h cases, t he classes were
chosen t o be represent at i ve of uni versi t y st udent s as a whole.
They asked a series o f quest i ons of t he f or m " Wha t pr opor t i on
of st udent s i n t hi s uni versi t y have hi gher i nt el l ect ual abi l i t i es
t han yoursel f?" Ther e were mar ked di fferences bet ween t he
Japanese and t he Amer i can st udent s i n t hei r est i mat i ons o f
t hei r own uni queness; t he Amer i cans di spl ayed significantly
mor e false uni queness t han t he Japanese. Amer i can st udent s
assumed t hat onl y 30% o f peopl e on average woul d be bet t er
t han t hemsel ves on var i ous t rai t s and abi l i t i es (e.g., memory,
at hl et i c ability, i ndependence, and sympat hy), whereas t he Japa-
nese st udent s showed al mos t no evi dence of t hi s false uni que-
ness. I n most cases, t he Japanese es t i mat ed t hat about 50% of
st udent s woul d be bet t er t han t hey were or have mor e of a given
t r ai t or ability. Thi s is, of course, t he expect ed fi ndi ng i f a repre-
sentative sampl e of college st udent s were eval uat i ng t hemsel ves
i n a relatively nonbi ased manner .
I n a recent series of st udi es conduct ed i n Japan wi t h Japanese
college st udent s, Takat a (1987) showed t hat t her e is no self-en-
hanci ng bi as in soci al compar i son. I n fact, he f ound j ust t he
o p p o s i t e - - a st r ong bi as i n t he self-effacing di r ect i on. Par t i ci -
pant s per f or med several anagr am pr obl ems t hat were al l eged t o
measur e me mor y abi l i t y Af t er compl et i on of t he t ask, t he par -
t i ci pant s were pr esent ed wi t h t hei r act ual per f or mance on some
of t he t ri al s and al so t he per f or mance o f anot her per son pi cked
at r a ndom f r om t he pool of subj ect s who had al l egedl y com-
pl et ed t he st udy The di r ect i on o f t he sel f - ot her difference was
mani pul at ed t o be ei t her favorable or unfavorabl e t o t he subject.
The dependent measur es were col l ect ed i n a pri vat e si t uat i on t o
mi ni mi ze sel f-present at i onal concerns. Fur t her mor e, because
it was consi der ed possi bl e t hat t he subj ect s mi ght st i l l bel i eve
t hey had a chance of seei ng t he ot her per son af t er war d, in a
fol l owup st udy t he "ot her person" was r epl aced wi t h a comput er
pr ogr am t hat al l egedl y si mul at ed t he t ask per f or mance of t he
average college st udent .
Sever al s t udi es (e.g., Goet hal s , 1989; Mar ks , 1984; Wyl i e,
1979) reveal t hat wi t h r espect t o abi l i t i es, Amer i cans t ypi cal l y
give t hemsel ves hi gher rat i ngs t han t hey give t o others. Thus,
when a compar i s on wi t h anot her is unfavorabl e t o t he self, t he
sel f - enhancement hypot hesi s pr edi ct s t hat Amer i cans shoul d
show little confi dence i n t hi s est i mat e of t hei r abi l i t y and seek
f ur t her i nf or mat i on. Thi s, i n fact, was t he case in an Amer i can
st udy by J. M. Schwart z and Smi t h (1976), whi ch used a pr oce-
dur e very si mi l ar t o Takata' s (1987). Whe n subj ect s per f or med
poor l y relative t o anot her per son, t hey had very little confi-
dence i n t hei r own score. These Amer i can dat a cont r ast sharpl y
wi t h t he Japanese dat a. Takata' s st udy shows a t endency exact l y
t he opposi t e of sel f-enhancement . Fur t her mor e, t he pat t er n di d
n o t depend on whet her t he compar i s on was made wi t h anot her
per son or wi t h t he comput er pr ogr am. The Japanese subj ect s
felt great er confi dence i n t hei r sel f-eval uat i on and were less
i nt er est ed i n seeki ng f ur t her i nf or mat i on when t hey had unfa-
vorabl e sel f-eval uat i ons t han when t hey had favorable ones. Si m-
ilarly, Wada (1988) al so r epor t ed t hat Japanese college st udent s
were convi nced of t hei r level of abi l i t y on a novel, i nf or mat i on-
i nt egr at i on t ask af t er fai l ure f eedback, but not af t er success
feedback. These dat a suggest what mi ght be cal l ed a modes t y
bi as or an ot her - enhancement bi as in soci al compar i son.
A si mi l ar modes t y bi as among t hose wi t h i nt er dependent
selves has al so been suggest ed by Shi kanai (1978), who st udi ed
t he causal at t r i but i on for one' s own success or failure i n an
abi l i t y t ask. Typically, Amer i can subj ect s bel i eve t hat t hei r i n-
t er nal at t r i but es such as abi l i t y or compet ence are ext remel y
i mpor t ant t o t hei r per f or mance, and t hi s is par t i cul ar l y t he case
when t hey have succeeded (e.g., Davi s & St ephan, 1980; Gi l mo r
& Rei d, 1979; Gr eenber g, Pys zczyns ki , & Sol omon, 1982;
Weiner, 1986). I n t he Shi kanai study, Japanese college st udent s
per f or med an anagr am t ask. Hal f of t hem were subsequent l y
l ed t o bel i eve t hat t hey scored bet t er t han t he average and t hus
" s ucceeded" whereas t he ot her hal f were l ed t o bel i eve t hat t hey
scor ed worse t han t he average and t hus "failedY Subj ect s were
t hen asked t o choose t he most i mpor t ant factor i n expl ai ni ng
t he success or t he fai l ure for each o f l 0 pai r s made f r om t he 5
possi bl e causes for per f or mance (i.e., ability, effort, t ask diffi-
cul t y [or ease], luck, and ment al - phys i cal "shape" of t he day).
Shi kanai anal yzed t he average number of t i mes each cause was
pi cked as mos t i mpor t ant (possi bl e mi n i mu m of 0 and maxi -
mum of 4). As shown i n Fi gur e 5, a modes t y bi as was agai n
obt ai ned, especi al l y aft er success. Wher eas fai l ure was at t r i b-
ut ed mai nl y t o t he l ack of effort, success was at t r i but ed pr i mar -
244 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
4 r--
3 ~ -
0
- 2
0
E
1
0
7
Success
Fai l ure
Abi l i t y Effort Task Luck Shape
Di f f i cul t y/ Ease
Causes
Figure 5. Mean importance rating given to each of five causes following
success and failure. (Drawn from results reported by Shikanai, 1978.)
ily to the ease of the task. Furthermore, the potential role of
ability in explaining success was very much downplayed. In-
deed, ability was perceived to be more important after a failure
than after a success, whereas task difficulty (or its ease) was
regarded to be more important after a success than after a fail-
ure. Subsequent studies by Shikanai that examined attribution
of success and failure of others did not find this pattern (Shi-
kanai, 1983, 1984). Thus, the pattern of "modest" appraisal
seems to be specific to the perception and the presentation of
the self and does not derive from a more general causal schema
applicable to both self and others. For others, ability is impor-
tant in explaining success. Yoshida et al. (1982, Studies 2 and 3),
who studied explanations of performance in a Japanese elemen-
tary school, found the tendency to de-emphasize the role of
ability in explaining success as early as the second grade.
Observations of a tendency to self-efface, and not to reveal
the typical American pattern of blaming others or the situation
when explaining failure, have been made outside of the experi-
mental laboratory as well. In a study by Hess et al. (1986), Japa-
nese mothers explained poor performance among their fifth
graders by claiming a lack of effort. In marked contrast, Ameri-
can mothers implicated effort in their explanations but viewed
ability and the quality of the training in the school as equally
i mport ant . This study also required the chi l dren to explain
their own poor performance by assigning 10 points to each of
five alternatives (ability, effort, training at school, bad luck, and
difficulty of math). Japanese children gave 5.6 points to lack of
effort, but American children gave 1.98 points. H. Stevenson
(personal communication, September 19, 1989) noted that in
observations of el ement ary school classrooms, Japanese
teachers, in contrast to American teachers, rarely refer to differ-
ences in ability among their students as an explanation for per-
formance differences, even though the range of ability as as-
sessed by standardized tests is approximately the same. Those
with interdependent selves thus seem more likely to view intel-
lectual achievement not as a fixed attribute that one has a cer-
tain amount of, but instead as a product that can be produced
by individual effort in a given social context. 7
The nature of modesty The exact nature of these modesty,
self-effacing, or other-enhancing biases has yet to be specified.
Perhaps those from i nt erdependent cultures have simply
learned that humility is the desired response, or the culturally
appropriate response, and that it is wise not to gloat over their
performance or to express confidence in their ability This inter-
pretation implies that the modesty biases observed in the stud-
ies described herein are primarily the result of impression man-
agement and that the subjects involved actually could have held
different, perhaps opposite, beliefs about themselves and their
ability However, it is also possible that these other-enhance-
ment biases reflect, or are accompanied by, psychologically au-
thentic self-perceptions. There are two related possibilities con-
sistent with this suggestion.
First, given the press not to stand out and to fit in, people in
interdependent cultures may acquire through socialization a
habitual modest-response tendency In large part, it may be a
function of the need to pay more attention to the other than to
the self, just as the self-serving bias is believed to result from a
predominant focus on the self(see Ross & Fletcher, 1985). Con-
sequently, for those with interdependent selves, whenever cer-
tain aspects of self need to be appraised in public, a modest,
self-effacing pattern of responses may occur spontaneously Fur-
thermore, this modesty can be motivated by many other-fo-
cused emotions that are central to the construal of self as an
interdependent entity From an independent viewpoint, such
modesty seems false and the result of suppressing a "natural"
pride in one's attributes. Yet, such pride is only natural within a
view of the self as an independent entity From an interdepen-
dent view, modest responses may be experienced quite posi-
tively and engender the pleasant, other-focused feelings that are
associated with connecting and maintaining interdependence.
Such positive, other-focused feelings also may be responsible
for the finding that Japanese students are more convinced of
and more confident in their ability after failure than success.
The satisfaction of doing well that can accompany good perfor-
mance on a novel, decontextualized task may be mitigated by
the threat of potential uniqueness and uncertainty over how to
respond to it. Moreover, i fa predominant basis of self-esteem is
how well one fits in and preserves relationships and interper-
sonal harmony, then failing to distinguish oneself with a highly
successful performance may not be particularly devastating, s
7 Of course, because those in Asian cultures believe high ability to be
a result of effort does not mean that they do not differentiate between
ability and effort. In all likelihood, they believe that effort and ability
are related in a multiplicative fashion to determine performance. Thus,
for instance, in a recent study by Stipek, Weiner, and Li (1989), Chinese
respondents reasoned, just as their American counterparts did, that if
a person shows the same level of performance with much less effort
expended on the task, the person must have a high level of the relevant
ability. Our point is simply that those in Asian cultures believe that
abilities are relatively more changeable over a long span of time
through the effort the person expends.
8 As noted, achievement may sometimes be construed as a means to
complete one's interdependence, as may well be the case for a Japanese
high school student who studies hard to gain admission to a prestigious
college. In this case, failure may well be extremely troubling for those
with interdependent selves.
CULTURE AND THE SELF 245
Cer t ai nl y it wi l l not be as devast at i ng as it is t o t he per son whose
sel f-est eem rest s pr i mar i l y on doi ng well i ndi vi dual l y and on
separ at i ng onesel f f r om others.
Second, among t hose wi t h i nt er dependent selves, t her e may
not be an awareness of one' s own abi l i t y i n general or i n t he
abst ract . I nst ead, one' s own abi l i t y i n a given t ask under a given
condi t i on may be i nferred f r om what ever cues ar e avai l abl e i n
t he specific si t uat i on i n whi ch t he t ask is per f or med. And what-
ever is i nferred i n t hi s way may be exper i enced as aut hent i c and
genuine. For exampl e, upon recei pt of f eedback about t hei r
ability, i nt er dependent selves may first at t end and t hi nk not so
much about t hei r abi l i t y as about t he appr oval or di sappr oval of
t he per son who gives t he feedback. I f appr oval or di sappr oval
can be st rongl y and unambi guousl y i nf er r ed, t hen t he per cep-
t i on of appr oval or di sappr oval may pr ovi de a st r ong heur i st i c
cl ue about abi l i t y; i f one receives appr oval , one mus t have hi gh
abi l i t y in t hi s si t uat i on, whereas i f one receives di sappr oval ,
t hen one mus t have l ow abi l i t y i n t hi s si t uat i on. In t he absence
of a st rong, endur i ng bel i ef about one' s abi l i t y in t he abst r act ,
such a heur i st i c may pr ovi de a subjectively genui ne sel f-apprai -
sal. Thi s anal ysi s al so suggest s why t hose wi t h i nt er dependent
selves may be convi nced o f t hei r l ow abi l i t y aft er a fai l ure feed-
back t o a much great er ext ent t han t hey ar e convi nced of t hei r
hi gh abi l i t y aft er a success feedback. Because of t he preval ent
soci al nor ms for pol i t e behavi or i n i nt er dependent cultures, di s-
appr oval can be mor e unequi vocal l y i nf er r ed f r om negat i ve
feedback t han appr oval can be i nf er r ed f r om positive feedback.
These suggest i ons about t he source of a modes t sel f-apprai sal
have yet t o be empi r i cal l y t est ed, but t hey are wor t hy o f careful
i nqui r y becaus e t hese f or ms o f s el f - appr ai s al ma y be qui t e
uni que t o i nt er dependent cultures. On t he basi s of empi r i cal
evi dence, however, t hi s much seems clear: Those wi t h i nt erde-
pendent selves wi l l t ypi cal l y not cl ai m t hat t hey are bet t er t han
others, wi l l not express pl easure i n t he st at e o f feeling super i or
to others, and i ndeed may not enj oy it. A st rong, pervasi ve
mot i ve for sel f - enhancement t hr ough t aki ng per sonal cr edi t for
success, denyi ng per sonal r esponsi bi l i t y for failure, and believ-
i ng onesel f t o be bet t er t han average may be pr i mar i l y a West-
er n phenomenon. It is aki n t o bei ng t he nai l t hat st ands out.
So far, t he empi r i cal evi dence on cul t ural var i at i on i n self-re-
l at ed mot i ves is l i mi t ed largely t o di fferences i n sel f-enhance-
ment versus ot her enhancement . However, ot her sel f-rel at ed
motives, such as sel f-affi rmat i on (Steele, 1988), self-verification
(Swann & Read, 198 l ), and sel f-act ual i zat i on (Maslow, 1954),
may al so differ across cul t ures i n si mi l ar ways. A series o f st ud-
ies by Steele has shown t hat t he negative psychol ogi cal i mpact
of one' s own mi sdeed, bl under, or publ i c embar r as s ment can be
r educed once anot her, si gni fi cant aspect of t he sel f is act i vat ed
and affi rmed. Thus, one' s t hr eat ened self-worth can be r est or ed
by a r emi nder of anot her, unt hr eat ened aspect o f t he self(e.g., "I
may not be athletic, but at l east I ' m creative"). To t he ext ent t hat
very different aspect s of sel f ar e highly val ued among t hose wi t h
i nt er dependent selves, t hi s process of sel f-affi rmat i on may al so
differ. For t hose wi t h i ndependent selves it wi l l be t he i nt er nal
at t r i but es of sel f t hat may most effectively offset each ot her and
reest abl i sh t hr eat ened sel f-est eem, whereas for t hose wi t h i nt er-
dependent selves it may be t he mor e publ i c aspect s of t he self,
l i ke one' s significant soci al roles, st at uses, and i mpor t ant i nt er-
per sonal rel at i ons, t hat mus t be focal i n sel f-est eem mai nt e-
nance. Thus, sel f-affi rmat i on for an i nt er dependent sel f wi l l
requi re an oppor t uni t y t o ensure t hat one is fitting i n and en-
gaging i n pr oper act i on i n a given si t uat i on.
I n a si mi l ar vein, exact l y what is veri fi ed i n self-verification
and what is act ual i zed i n sel f-act ual i zat i on may al so differ con-
si derabl y acr oss cultures. Currently, it is c ommon t o assume
t hat i ndi vi dual s are mot i vat ed t o veri fy and act ual i ze an i nt er-
nal l y coherent set of at t r i but es t hat t hey r egar d as significant.
Our pr esent analysis woul d imply, however, t hat peopl e wi t h
i nt e r de pe nde nt selves ma y st r i ve t o ver i f y a nd act ual i ze t he
mor e publ i c qual i t i es of t he s e l f - - t he ones t hat al l ow t hem t o
concei ve of t hemsel ves as r espect abl e and decent par t i ci pant s
i n si gni fi cant i nt er per sonal rel at i onshi ps.
Fur t her mor e, among t hose wi t h i nt er dependent selves, self-
ver i f i cat i on a nd s el f - act ual i zat i on may even be achi eved
t hr ough t he r eal i zat i on of some mor e general , abst r act forms of
rel at i on, t hat is, one' s r el at i onshi p t o or one' s rol e i n soci et y or
even i n t he nat ur al or cosmi c syst em. The sel f - descr i pt i on st ud-
ies revi ewed earl i er suggest t hi s possibility. In general , t he self-
de s c r i pt i ons o f t hose wi t h i nt e r de pe nde nt sel ves have been
f ound t o be qui t e concret e and si t uat i on specific (see Cousins,
1989). Ther e is, however, one i nt erest i ng, rel i abl e except i on t o
this. Subj ect s f r om Asi an cul t ural backgr ounds ( pr esumabl y
t hose wi t h pr edomi nant l y i nt er dependent selves) of t en pr ovi de
ext r emel y gl obal sel f-descri pt i ons, such as "I a m a uni que cre-
ation," "I a m a huma n bei ng" "I a m an organi c f or m" and "I
a m a pr oduct of my envi r onment : ' It coul d appear t hat t hese
st at ement s ar e t oo abst r act t o be i nformat i ve i n any pr agmat i c
sense (Rosch, 1978). The l ack of i nf or mat i on cont ai ned i n t hese
descri pt i ons, however, may be mor e appar ent t han real. Not e
t hat t hese gl obal st at ement s pr esuppose a view o f t he worl d as
an encompassi ng whol e in whi ch t hese subj ect s percei ve t hem-
selves t o be a par t or a par t i ci pant . And for t hese subjects, it may
be t hese rel at i onshi ps t hat must be veri fi ed and act ual i zed.
We have suggest ed t he different forms t hat some sel f-rel at ed
mot i ves mi ght assume i f t hey ar e based in an i nt er dependent
r at her t han an i ndependent const r ual of self. Fur t her empi r i cal
wor k is r equi r ed t o det er mi ne whet her t he t ypes o f sel f-rel at ed
mot i ves descr i bed her ei n are i ndeed as preval ent i n East er n
i nt er dependent cul t ures as t hey have been f ound t o be i n West-
ern, par t i cul ar l y Amer i can, cultures. It coul d be t hat t hese self-
rel evant mot i ves are not par t of t he set o f uni versal i ndi vi dual
st ri vi ngs, 9 but i nst ead an out gr owt h of an i ndependent self-sys-
t em r oot ed i n t he press for separ at i on and i ndi vi duat i on.
Co n c l u s i o n s
We have descr i bed t wo di vergent const r ual s o f t he s e l f - - a n
independent vi ew and an i nt er dependent view. The mos t signifi-
cant di fferences bet ween t hese t wo const r ual s is in t he rol e t hat
is assi gned t o t he ot her i n sel f - def i ni t i on. Ot her s a nd t he
sur r oundi ng soci al cont ext ar e i mpor t ant i n bot h const r ual s,
but for t he i nt er dependent self, ot her s are i ncl uded within t he
boundar i es of t he sel f because rel at i ons wi t h ot hers i n specific
9I t is intriguing that Murray's (1938) original study of motives, as
well as Hilgard's (1953, 1987) update of it, did not include any of the
self-focused motives that are so central to current research on the self.
246 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
cont ext s are t he defi ni ng feat ures o f t he self. I n t he words of
Lebr a (1976), t he i ndi vi dual is i n some respect s "a fract i on" and
becomes whole when fitting i nt o or occupyi ng one' s pr oper
pl ace i n a soci al uni t . The sense of i ndi vi dual i t y t hat accompa-
ni es an i nt er dependent sel f i ncl udes an at t ent i veness and re-
sponsi veness t o ot her s t hat one ei t her expl i ci t l y or i mpl i ci t l y
assumes wi l l be r eci pr ocat ed by t hese others, as well as t he
wi l l ful management of one' s ot her-focused feelings and desi res
so as t o mai nt ai n and f ur t her t he r eci pr ocal i nt er per sonal rel a-
t i onshi p. One is consci ous o f where one bel ongs wi t h r espect t o
ot hers and assumes a recept i ve st ance t owar d t hese others, con-
t i nual l y adj ust i ng and accommodat i ng t o t hese ot hers i n many
aspect s o f behavi or (Azuma, 1984; Weisz et al., 1984). Such acts
of fitting i n and accommodat i ng are oft en i nt r i nsi cal l y reward-
ing, because t hey give ri se t o pl easant , ot her-focused emot i ons
(e.g., feeling of connect i on) whi l e di mi ni s hi ng unpl easant ones
(e.g., shame) and, f ur t her mor e, becaus e t he sel f - r est r ai nt re-
qui r ed i n doi ng so forms an i mpor t ant basi s of sel f-est eem.
Typically, t hen, it is ot hers r at her t han t he sel f t hat serve as t he
referent for or gani zi ng one' s experi ences.
Wi t h an i ndependent const r ual o f t he self, ot hers are less
cent ral l y i mpl i cat ed i n one' s cur r ent sel f-defi ni t i on or identity.
Certainly, ot hers are i mpor t ant for soci al compar i s on, for re-
flected appr ai sal , and i n t hei r rol e as t he t arget s of one' s act i ons,
yet at any given moment , t he sel f is assumed t o be a compl et e,
whole, aut onomous entity, wi t hout t he others. The defi ni ng fea-
t ures of an i ndependent sel f ar e at t r i but es, abi l i t i es, traits, de-
sires, and mot i ves t hat may have been soci al pr oduct s but t hat
have become t he " pr oper t y" o f t he sel f - cont ai ned i ndi vi dual
(see Sampson, 1989) and t hat ar e as s umed t o be t he source of
t he i ndi vi dual ' s behavior. The sense o f i ndi vi dual i t y t hat accom-
pani es t hi s const r ual of t he sel f i ncl udes a sense o f onesel f as an
agent, as a pr oducer of one' s act i ons. One is consci ous of bei ng
i n cont r ol over t he s ur r oundi ng si t uat i on, and of t he need t o
express one' s own t hought s, feelings, and act i ons t o others, and
is relatively less consci ous of t he need t o receive t he t hought s,
feelings, and act i ons of others. Such act s of st andi ng out are
oft en i nt r i nsi cal l y r ewar di ng because t hey elicit pl easant , ego-
focused emot i ons (e.g., pride) and al so r educe unpl easant ones
(e.g., frustration). Fur t her mor e, t he act s of st andi ng out , t hem-
selves, f or m an i mpor t ant basi s of sel f-est eem.
The Role of the Self
The relative i mpor t ance t hat is accor ded t o ot hers i n t hese
two const r ual s has a wi de range of psychol ogi cal i mpl i cat i ons.
In t hi s article, we have out l i ned some of t he cognitive, emo-
t i onal , and mot i vat i onal consequences of hol di ng a vi ew o f t he
sel f t hat i ncl udes ot her s and t hat requi res ot hers t o defi ne t he
self. Al t hough a r api dl y expandi ng vol ume of st udi es suggest
t hat some aspect s o f cogni t i ve f unct i oni ng are relatively har d-
wi red, many feat ures of t he way peopl e perceive, cat egori ze, or
assign causal i t y are pr obabl y not basi c processes t hat deri ve i n
any st r ai ght f or war d way f r om t he f unct i oni ng o f t he human
machi ner y or "hardware. " Rat her, t hese processes are t o a large
ext ent per sonal , refl ect i ng t he nat ur e o f t he sel f t hat anchors
t hem. Thus, t hey reflect al l o f t hose factors, i ncl udi ng cul t ural
aspects, t hat j oi nt l y det er mi ne t he self. I f one percei ves onesel f
as embedded wi t hi n a l arger cont ext of whi ch one is an i nt erde-
pendent par t , it is likely t hat ot her obj ect s or events will be
percei ved i n a si mi l ar way. For exampl e, a given event involving
a par t i cul ar act or wi l l be percei ved as ari si ng f r om t he situa-
t i onal cont ext o f whi ch t hi s act or is an i nt er dependent par t ,
r at her t han as s t emmi ng solely f r om t he at t ri but es of t he actor.
Or, i n answeri ng any quest i on, one' s first t endency may be t o
consi der t he par t i cul ar soci al si t uat i on t hat is defi ned by t he
cur r ent i nt er act i on (e.g., t eacher - s t udent , wor ker - co- wor ker ,
and younger-el der) and t hen t o gauge t he range o f responses
t hat are most appr opr i at e t o t hi s si t uat i on. These const r ual s of
sel f ar e pr obabl y abst r act ed t hr ough early pat t er ns of di r ect
i nt eract i ons wi t h par ent s and peers. The way peopl e initially,
and t hus thereafter, most nat ur al l y or effortlessly percei ve and
under s t and t he worl d is r oot ed i n t hei r sel f-percept i ons and
sel f-underst andi ngs, under st andi ngs t hat are t hemsel ves con-
st r ai ned by t he pat t er ns of soci al i nt eract i ons char act er i st i c of
t he given culture.
Consequences for Self-Processes
Our di scussi on of t he cognitive, emot i onal , or mot i vat i onal
consequences has by no means exhaust ed t he range of pot ent i al
consequences of hol di ng an i ndependent or i nt er dependent
const r ual o f t he self. Consi der first t he set of processes con-
nect ed by a hyphen to t he self. It is r easonabl e t o ~ssume t hat all
of t hese phenomena (e.g., sel f-at ~rmat i on [Steele, 1988], self-
ver i f i cat i on [Swann, 1983], sel f - consci ousness [ Feni gst ei n,
Scheier, & Buss, 1975], sel f-cont rol [Carver & Scheier, 1981],
sel f-act ual i zat i on [Maslow, 1954], or sel f - handi cappi ng [Jones
& Berglas, 1978]) coul d assume a somewhat different form de-
pendi ng on how i nt er dependent t he sel f is wi t h others.
Sel f-est eem for t hose wi t h an i ndependent const r ual of t he
sel f depends on one' s abilities, at t ri but es, and achi evement s.
The mos t wi del y used measur e of sel f-est eem, t he Rosenberg
Sel f-Est eem Scale, requi res t he endor sement o f i t ems l i ke "I am
a per son of wort h" or "I a m pr oud of my abi l i t i es: ' Sel f-est eem
associ at ed wi t h an i nt er dependent sel f coul d i ncl ude endorse-
ment of si mi l ar items, al t hough what it means t o be, for exam-
ple, a per son of wor t h coul d well have a different meani ng. Or
hi gh sel f-est eem may be mor e st rongl y associ at ed wi t h an en-
dor sement o f i t ems t hat gauge one' s abi l i t y t o r ead t he si t uat i on
and t o r espond as requi red. I f t hi s is t he case, a t hr eat or a
chal l enge t o t he sel f may not come in t he f or m of feedback t hat
one is unl i ke a cher i shed concept i on of t he i nner or di sposi -
t i onal sel f ( dumb i nst ead of smar t ; submissive rat her t han domi -
nant) but i nst ead i n t er ms of a t hreat o f a di sr upt i on of, or a
di sconnect i on from, t he rel at i on or set o f rel at i ons with which
one forms an i nt er dependent whole.
The focus on t he di st i nct i on bet ween i ndependent versus
i nt er dependent selves has t he pot ent i al to pr ovi de a means of
i nt egrat i ng research on a large number of separat e per sonal i t y
const ruct s. One of t he significant di st i nct i ons t hat appear s re-
peat edl y t hr oughout West ern psychol ogy reflects a var i at i on
among i ndi vi dual s in how t uned in, sensitive to, or i ent ed to-
ward, focused on, or concer ned t hey are wi t h others. The i nt r o-
ver si on- ext r aver si on di mensi on reflects t hi s difference, as does
t he i nner - di r ect ed- out er - di r ect ed di st i nct i on (Rei sman, Den-
ne y, & Gl azer, 1950). Ot her rel at ed di st i nct i ons i ncl ude high
versus low sel f-moni t ori ng (Snyder, 1979), per sonal i dent i t y
CULTURE AND THE SELF 247
versus social identity (Cheek, 1989; Hogan, 1975), public versus
private self-consciousness (Fenigstein, 1984), social orientation
versus individual orientation (Greenwald, 1980), collectivism-
individualism (Hui, 1988; Triandis, 1989), and field i ndepen-
dence- f i el d dependence (Wi t ki n & Goodenough, 1977). I n
fact, Wi t ki n and his colleagues descri bed a fi el d-dependent
person as one who includes others within t he boundari es of t he
self and who does not make a sharp distinction between the self
and others. Many of the empi ri cal findings (described in Wit-
ki n & Goodenough, 1977; Witkin, Goodenough, & Ol t man,
1979) about t he i nt er per sonal expert i se and sensitivities o f
field-dependent people are similar t o those descri bed herein
for people with interdependent selves.
unbelievable. People conform, obey, diffuse responsibility i n a
group, allow themselves t o be easily persuaded about all man-
ner o f things, and become hopelessly commi t t ed t o others on
the basis of mi ni mal action (e.g., see Myers, 1989). Even within
highly individualist Western culture, most people are still much
less self-reliant, self-contained, or self-sufficient t han t he pre-
vailing cultural ideology suggests t hat t hey should be. Perhaps
Western models of t he self are quite at odds with act ual i ndi vi d-
ual social behavior and should be reformulated t o reflect the
substantial i nt erdependence t hat characterizes even Western
individualists. Sampson (1989) has recently argued t hat t he real-
ity of globalization and a shri nki ng world will force j ust such a
rethinking of the nature of the individual.
Consequences for Social Psychological Phenomena
Ot her social behaviors may also depend on one' s medi at i ng
model of the self (see Triandis, 1989, for a recent analysis o f
some of these effects). Thus, for one with an interdependent self,
conformi t y may not reflect an inability t o resist social pressure
and t o stick by one' s own perceptions, attitudes, or beliefs (the
defining features of the self). Inst ead, conformi t y t o part i cul ar
others with whom t he ot her is interdependent can be a highly
valued end state. It can signify a willingness t o be responsive t o
others and t o adjust one' s own demands and desires so as t o
mai nt ai n t he ever-important relation. The conformi t y observed
for these subjects with interdependent selves when surrounded
with others who form part of an i mpor t ant social unit, could
well be much higher t han typically observed. However, confor-
mi t y t o t he desires and demands o f those outside the i mpor t ant
social uni t or the self-defining i n-group may not be required at
all. Thus, for those with interdependent selves, a t ypi cal Asch-
t ype conformi t y par adi gm involving subjects and strangers as
confederates may result in less conformi t y t han typically ob-
served in Amer i can studies.
Studies of ot her phenomena such as social facilitation or so-
cial loafing could also produce differential effects, dependi ng
on t he self-systems of the subjects. Should those with interde-
pendent construals of the self show pr onounced social facilita-
t i on compar ed with those with individual selves? Or should
those with interdependent selves be less susceptible t o social
loafing (decrements in performance when one' s individual con-
t ri but i on t o t he gr oup pr oduct cannot be identified; see Har-
kins, Lat anr, & Williams, 1980)? Our analysis is also relevant t o
t wo of the central probl ems i n Western psychol ogy- - t he incon-
sistency between attitudes and behavior and t he inconsistency
between personality and behavior. As we have noted, interde-
pendent selves do not prescribe or require such a consistency
bet ween one' s i nt er nal at t ri but es and one' s act i ons. Conse-
quently, the press for consi st ency should be much less i mpor-
t ant and much less bemoaned when not observed. I n fact, con-
sistency from an interdependent perspective may reflect a lack
o f flexibility, insensitivity t o t he cont ext , rigidity, or i mma-
turity.
Furt her analysis of the consequences o f different construals
of t he self may also prove fruitful in underst andi ng some basic
social psychological questions. Social psychologists report t hat
people are enormousl y influenced by others, oft en t o an extent
t hat the investigators and certainly individuals themselves, find
Construals of the Self and Gender
Many i mpor t ant gender differences may also be linked t o
divergent construals o f t he self. Recent feminist t heory on em-
pathy suggests t hat relations have a power and a significance in
women' s lives t hat have gone largely unrecogni zed (e.g., Belenky,
Clinchy, Goldberger, & Tarule, 1986; Jordan & Surrey, 1986; J.
B. Miller, 1986; Stewart & Lykes, 1985). An awareness of and
sensitivity t o others is described as one o f rnost significant fea-
tures of the psychology o f women. I f this is t he case, t hen self-
esteem and self-validation should depend not only on being
able t o do a j ob well, but on fostering and sustaining relation-
ships. As Gilligan (1986) claimed, a willingness and an ability
t o care are st andards o f self-evaluation for many women. This
theoretical work is forging a new vision of dependence, one t hat
is similar in many ways t o some East ern views. Being depen-
dent does not invariably mean being helpless, powerless, or
without control. It often means being interdependent. It thus
signifies a conviction t hat one is able t o have an effect on others
and is willing t o be responsive t o others and t o become engaged
with t hem. I n other words, there is an alternative t o selfishness
(which implies t he excl usi on of others) besides selflessness
(which is t o imply the exclusion of t he self or self-sacrifice):
There is a self defined in relationship t o others (see Chodorow,
1978; Gilligan, 1982; Markus & Oyserman, 1988).
Difficult Questions
Car r yi ng out the research necessary t o systematically investi-
gate t he range o f basic consequences of having one or anot her
const rual of the self raises several compl ex questions. Some of
these we have onl y t ouched on. For example, a persistent issue
is how deep or pervasive are these cultural differences? Are t he
observed differences pri mari l y a reflection of differences in
styles of behavioral expression, or do t hey also reflect differ-
ences in the phenomenol ogy accompanyi ng t he behavior? I f
there are nor ms against the display or expression of anger, what
happens t o t he nature o f the felt anger? I n ot her words, is it the
case, as we suggest here, t hat these nor ms can somet i mes be
internalized t o t he extent that t hey det ermi ne the nature of
one' s experience? For example, a recent study by Bontempo,
Lobel, and Triandis (1989) compar ed the public and private
responses of individuals from a collectivist culture with those o f
i ndi vi dual s f r om an i ndi vi dual i st culture. The researchers
asked respondents t o indicate how enjoyable it would be t o
248 HAZEL ROSE MARKUS AND SHINOBU KITAYAMA
engage in a time-consuming, individually costly behavior such
as visiting a friend in the hospital. Only i n the public condition
did individualists claim that the behavior would be enjoyable.
The collectivists, in contrast, claimed that the behavior would
be enjoyable even when their responses were private.
The view that altruistic behaviors are only seemingly altru-
istic and that they are public actions without any subjective,
private foundation can perhaps be traced to the insistence of
Western psychologists on the i nt ernal attributes (feeling,
thought, and traits) as the universal referents for behavior. They
have thus understandably failed to attend to the possibility of
the other as a referent for behavior, and thus to the possibility of
other-focused emotions. There is, however, the possibility that
such emotions can motivate genuine, other-oriented, altruistic
behaviors, without any conscious, or even unconscious, calcu-
lation of individual payoff, and as such serve as the important
glue of interdependent relationships.
Another thorny issue centers on the assessment of cultural
differences. The use of introspective reports, for example,
which are typically quite useful in the study of cognition, emo-
tion, and motivation, may be problematic in cross-cultural re-
search because within a given cultural context, people have
little access to the absolute extent of their attention or respon-
siveness to others. This may explain, for example, why Triandis
et al. (1988) found that those with collective selves do not report
a greater than average awareness of or concern for the demands
of others. Another persistent issue is that of translation and
equating stimuli and questionnaires. Can psychologists readily
assume that when an American and a Japanese use the word
embarrass it indicates a similar emotional experience? Can they
hypothesize, for example, that those with interdependent selves
should show more high self-monitoring (i.e., attention to the
behavior of others) than those with independent selves, and
then assume that a translation of Snyder's (1979) scale into Japa-
nese or Chinese will be sufficient to reflect these differences?
One may even ask to what extent a construct such as self-moni-
toring can be unequivocally defined across different cultures
with remarkably different construals of self.
In sum, we have argued that the view one holds of the self is
critical in understanding individual behavior and also in un-
derstanding the full nature of those phenomena that implicate
the self. A failure to replicate certain findings in different cul-
tural contexts should not lead to immediate despair over the
lack of generality of various psychological principles or to the
conclusion of some anthropologists that culturally divergent in-
dividuals inhabit incomparably different worlds. Instead, it is
necessary to identify the theoretical elements or processes that
explain these differences. We suggest that how the self is con-
strued may be one such powerful theoretical element.
References
Ainsworth, M. D. S., Bell, S. M., & Stayton, D. (1974). Infant-mother
attachment and social development. In M. R Richards (Eds.), The
introduction of the child into a social world (pp. 95-135). London:
Cambridge University Press.
Allen, N. J. (1985). The category of the person: A reading of Mauss's
last essay. In M. Carrithers, S. Collins, & S. Lukes (Eds.), The category
of the person: Anthropology, philosophy, history (pp. 26-35). Cam-
bridge, England: Cambridge University Press.
Allport, (3. W. (1937). Personality: A psychological interpretation. New
York: Holt.
Atkinson, J. (Ed.). 0958). Motives in fantasy, action and society New
York: Van Nostrand.
Au, T. K. (1983). Chinese and English counterfactuals: The Sapir-
Whorf hypothesis revisited. Cognition, 15, 162-163.
Au, T. K. (1984). Counterfactuals: In reply to Alfred Bloom. Cognition,
17, 289-302.
Azuma, H. 0984). Secondary control as a heterogeneous category.
American Psychologist, 39, 970-971.
Bachnik, J. M. (1986). Time, space and person in Japanese relation-
ships. In J. Hendry & J. Webber (Eds.), Interpreting Japanese society:
Anthropological approaches (pp. 49-75). New York: Oxford Univer-
sity Press.
Baumeister, R. E (1987). How the self became a problem: A psychologi-
cal review of historical research. Journal of Personality and Social
Psychology, 52, 163-176.
Beattie, J. (1980). Representations of the self in traditional Africa.
Africa, 50, 313-320.
Belenky, M. E, Clinchy, B. M., Goldberger, N. R., & Tarule, J. M.
(1986). Women& ways of knowing: The development of self, voice, and
mind. New York: Basic Books.
Bellah, R. N., Madsen, R., Sullivan, W. M., Swidler, A., & Tipton, S. M.
(I 985). Habits of the heart: Individualism and commitment in Ameri-
can life. Berkeley, CA: University of California Press.
Bloom, A. (1981 ). The linguistic shaping of thought. Hillsdale, N J: Erl-
baum.
Bloom, A. (1984). Caution--the words you use may effect what you
say: A response to Au. Cognition, 17, 281.
Blumenthal, E. P (1977). Models in Chinese moral education: Perspec-
tives from children's books. Dissertation Abstracts International,
37(10-A), 6357-6358.
Bond, M. H. (1986). The psychology of the Chinese people. New York:
Oxford University Press.
Bond, M. H. (Ed.). (1988). Thecross-culturalchallengetosocialpsychol-
ogy. Beverly Hills, CA: Sage.
Bond, M., Leung, K., & Wan, K.-C. (1982). The social impact of self-ef-
facing attributions: The Chinese case. Journal of Social Psychology,
118, 157-166.
Bontempo, R., Lobel, S. A., & Triandis, H. C. (1989). Compliance and
value internalization among Brazilian and US. students. Unpublished
manuscript.
Briggs, J. (1970). Never in anger. Cambridge, MA: Harvard University
Press.
Bruner, J. (1986). Actual minds, possible worlds. New York: Plenum
Press.
Buck, R. (1988). Human motivation and emotion (2nd ed.). New York:
Wiley.
Cantor, N., & Kihlstrom, J. (1987). Personality and social intelligence.
Englewood Cliffs, NJ: Prentice-Hall.
Carver, C. S., & Scheier, M. E (1981). Attention and self-regulation: A
control theory approach to human behavior. New York: Springer-Ver-
lag.
Cheek, J. M. (1989). Identity orientations and self-interpretation. In
D. M. Buss & N. Cantor (Eds.), Personality psychology: Recent trends
and emerging directions (pp. 275-285). New York: Springer-Verlag.
Chiu, L. H. (1972). A cross-cultural comparison of cognitive styles in
Chinese and American children. International Journal of Psychol-
ogy, 7, 235-242.
Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis
and the sociology of gender. Berkeley, CA: University of California
Press.
Church, A. T. (1987). Personality research in a non-Western culture:
The Philippines. Psychological Bulletin, 102, 272-292.
CULTURE AND THE SELF 249
Cousins, S. (1989). Culture and self hood in Japan and the U.S. Journal
of Personality and Social Psychology 56, 124-131.
Dalal, A. K., Sharma, R., & Bisht, S. (1983). Causal attributions of
ex-criminal tribal and urban children in India. Journal of Social
Psychology, 119, 163-171,
DAndrade, R. (1981). The cultural part of cognition. Cognitive Science,
5, 179-185.
Darwin, C. R. (1896). The expression of emotions in man and animals.
New York: Philosophical Library.
Davis, M. H., & Stephan, W G. (1980). Attributions for exam perfor-
mance. Journal of Applied Social Psychology 10, 235-248.
De Riviera, J. (1984). The structure of emotional relationships. In R
Shaver (Ed.), Review of personality and social psychology: Emotions,
relationships, and health (pp. 116-145). Beverly Hills, CA: Sage.
De Vos, G. A. (1973). Socialization for achievement: Essays on the cul-
tural psychology of the Japanese. Berkeley: University of California
Press.
De Vos, G. (1985). Dimensions of the self in Japanese culture. In A.
Marsella, G. De Vos, & E L. K. Hsu (Eds.), Culture and se/f(pp.
149-184). London: Tavistock.
Doi, L. T. (1973). The anatomy of dependence. Tokyo: Kodansha.
Doi, T. (1986). The anatomy of self" The individual versus society Tokyo:
Kodansha.
Dumont , L. (1970). Homo hierarchicus. Chicago: University of Chicago
Press.
Durkheim, E. (1968). Les formes elementaires de la vie religieuse [Basic
forms of religious belief] (6th ed.). Paris: Presses Universitarires de
France. (Original work published 1912)
Ekman, P. (1972). Universals and cultural differences in facial expres-
sion of emotion. In J. K. Cole (Ed.), Nebraska symposium on motiva-
tion (pp. 207-283). Lincoln: University of Nebraska Press.
Elvin, M. (1985). Between the earth and heaven: Conceptions of the
self in China. In M. Carrithers, S. Collins, & S. Lukes (Eds.), The
category of the person: Anthropology, philosophy, history (pp. 156-
189). New York: Cambridge University Press.
Erikson, E. (1950). Identification as the basis for a theory of motiva-
tion. American Psychological Review, 26, 14-21.
Fenigstein, A. (1984). Self-consciousness and the overperception of self
as a target. Journal of Personality and Social Psychology, 47, 860-
870.
Fenigstein, A., Scheier, M. E, & Buss, A. H. (1975). Public and private
self-consciousness: Assessment and theory. Journal of Consulting
and Clinical Psychology 43, 522-527.
Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of
forced compliance. Journal of Abnormal and Social Psychology 58,
203-210.
Fiedler, E E. (1978). Recent development in research on the contin-
gency model. In L. Berkowitz (Ed.), Group processes (pp. 209-225).
New York: Academic Press.
Fiske, A. P. (in press). Making up society: The four elementary relational
structures. New York: Free Press.
Fiske, S. T., & Taylor, S. E. (1984). Social cognition. Reading, MA: Ad-
dison-Wesley.
Frijda, N. H., Kuipers, P., & ter Schure, E. (1989). Relations among
emotion appraisal and emotional action readiness. JournalofPerson-
ality and Social Psychology 57, 212-228.
Galtung, J. (1981). Structure, culture, and intellectual style: An essay
compari ng Saxonic, Teutonic, Gallic and Nipponic approaches. So-
cial Science Information, 20, 817-856.
Geertz, C. (1974). From the native's point of view: On the nature of
ant hropol ogi cal underst andi ng. In K. Basso & H. Selby (Eds.),
Meaning in anthropology (pp. 221-237). Albuquerque: University of
New Mexico Press.
Geertz, C. (1975). On the nature of anthropological understanding.
American Scientist, 63, 47-53.
Gergen, K. J. (1968). Personal consistency and the presentation of self.
In C. Gordon & K. J. Gergen (Eds.), The self in social interaction:
Classic and contemporary perspectives (Vol. 1, pp. 299-308). New
York: Wiley.
Gergen, K. J., & Gergen, M. M. (1988). Narrative and the self as rela-
tionship. In L. Berkowitz (Ed.), Advances in experimental social psy-
chology (Vol. 21, pp. 17-56). New York: Academic Press.
Gi l l i gan, C. (1982). In a different voice: Psychological theory and
women's development. Cambridge, MA: Harvard University Press.
Gilligan, C. (1986). Remappi ng the moral domain: New images of the
self in relationship. In T. C. Heller, M. Sosna, & D. E. Wellbery (Eds.),
Reconstructing individualism: Autonomy individuality and the self in
Western thought (pp. 237-252). Stanford, CA: Stanford University
Press.
Gilmor, T. M., & Reid, D. W. (1979). Locus of control and causal attri-
bution for positive and negative outcomes on university examina-
tions. Journal of Research in Personality 13, 154-160.
Gilovich, T. (1983). Biased evaluation and persistence in gambling.
Journal of Personality and Social Psychology 40, 797-808.
Goethals, A. (1989, April). Studies of false uniqueness. Paper presented
at the Research Center for Group Dynamics Seminar, Institute for
Social Research, University of Michigan, Ann Arbor, MI.
Greenberg, J., Pyszczynski, T., & Solomon, S. (1982). The self-serving
attributional bias: Beyond self-presentation. Journal of Experimen-
tal Social Psychology 18, 56-67.
Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision
of personal history. American Psychologist, 35, 603-618.
Greenwald, A. G., & Pratkanis, A. R. (1984). The self. In R. S. Wyer & T.
K. Srull (Eds.), Handbook of social cognition (Vol. 3, pp. 129-178).
Hillsdale, NJ: Erlbaum.
Hall, C. S., & Lindzey, G. (1957). Theories of personality New York:
Wiley.
Hallowell, A. I. (1955). Culture and experience. Philadelphia: Univer-
sity of Pennsylvania Press.
Hamaguchi, E. (1985). A contextual model of the Japanese: Toward a
methodological innovation in Japan studies. Journal of Japanese
Studies, 11, 289-321.
Harkins, S. G., Latan6, B., & Williams, K. (1980). Social loafing: Allo-
cating effort or taking it easy? Journal of Experimental Social Psy-
chology 16, 457-465.
Harter, S. (1983). The development of the self-system. In E. M. Hether-
ington (Ed.), Handbook of chiM psychology: Vol. 4. Socialization, per-
sonality and social development (4th ed.). New York: Wiley.
Harter, S. (1990). Causes, correlates and the functional role of global
self-worth: A life span perspective. In R. J. Sternberg & J. Kolligian,
Jr. (Eds.), Competence considered (pp. 67-97). New Haven, CT: Yale
University Press.
Hayashi, C. (1988). National character of the Japanese. Tokyo: Statisti-
cal Bureau, Japan.
Head, H. (1920). Studies in neurology. London: Oxford Uni versi t y
Press.
Heelas, P. L. E, & Lock, A. J. (Eds.). (1981). Indigenous psychologies:
The anthropology of the self. London: Academic Press.
Hess, R., Azuma, H., Kashiwagi, K., Dickson, W. P., Nagano, S., Hollo-
way, S., Miyake, K., Price, G., Hatano, G., & McDevitt, T. (1986).
Family influences on school readiness and achievement in Japan
and the United States: An overview of a longitudinal study. In H.
Stevenson, H. Azuma, & K. Hakuta (Eds.), Child development and
education in Japan (pp. 147-166). New York: Freeman.
Hilgard, E. R. (1953). Introduction to psychology New York: Harcourt,
Brace.
250 HAZEL ROSE MARKUS AND SHI NOBU KITAYAMA
Hilgard, E. R. (1987). Psychology in American: A historical survey, New
York: Harcourt Brace Jovanovich.
Hofstede, G. (1980). Culture's consequences: International differences in
work-related values. Beverly Hills, CA: Sage.
Hogan, R. (1975). Theoretical egocentrism and the problem of compli-
ance. American Psychologist, 30, 533-540.
Hol l and, D., & Qui nn, N. (1987). Cultural models in language and
thought. Cambridge, England: Cambridge University Press.
Hollander, E. P. (1985). Leadership and power. In G. Lindzey & E.
Aronson (Eds.), Handbook of social psychology (Vol. 2, pp. 485-
537). New York: Random House.
Holyoak, K. J., & Gordon, P. C, (1983). Social reference points. Journal
of Personality and Social Psychology, 44, 881-887.
Hsu, E L. K. (1975). Iemoto: The heart of Japan. New York: Wiley.
Hsu, E L. K. (1981). American and Chinese. Passage to differences.
Honolulu: University of Hawaii Press.
Hsu, E L. K. (1985). The self in cross-cultural perspective. In A. J.
Marsella, G. De Vos, & E L. K. Hsu (Eds.), Culture and se/f(pp.
24-55). London: Tavistock.
Hui, C. H. (1988). Measurement of individualism-collectivism. Jour-
nal of Research in Personality, 22, 17-36.
Hwang, C. H. (1976). Change of psychological needs over t hi rt een
years. Bulletin of Educational Psychology (Taipei), 9, 85-94.
Iwao, S. (1988, August). Social psychology's models of man: Isn't it time
for East to meet West? Invited address to the International Congress
of Scientific Psychology, Sydney, Australia.
James, W. (1890). Principles of psychology, New York: Holt.
A Japanese super mar ket in New Jersey. (1989, Apri l 6). New York
Times, p. 4.
Johnson, E (1985). The Western concept of self. In A. Marsella, G. De
Vos, & F. L. K. Hsu (Eds.), Culture and self. London: Tavistock.
Johnson-Laird, P. N. (1983). Mental models: Towards a cognitive science
of language, inference, and consciousness. Cambridge, MA: Harvard
University Press.
Jones, E. E., & Berglas, S. (1978). Control of attributions about the self
through self-handicapping strategies: The appeal of alcohol and the
role of underachievement. Personality and Social Psychology Bulle-
tin, 4. 200-206.
Jordan, J. V., & Surrey, J. L. (1986). The self-in-relation: Empathy and
the mot her - daught er relationship. In T. Bernay & D. W. Cant or
(Eds.), The psychology of today's women (pp. 81-104). Cambridge,
MA: Harvard University Press.
J6reskog, K. G. (1969). A general approach to confirmatory maxi mum
likelihood factor analysis. Psychometrika, 34, 183-202.
Kakar, S. (1978). The inner world." A psychoanalytic study of childhood
and society in India. Delhi, India: Oxford University Press.
Kennedy, S., Scheier, J., & Rogers, A. (1984). The price of success: Our
monocultural science. American Psychologist, 39, 996-997.
Kiefer, C. W. (1976). The danchi zoku and the evolution of metropolitan
mind. In L. Austin (Ed.), The paradox of progress (pp. 279-300).
New Haven, CT: Yale University Press.
Kitayama, S., & Markus, H. (1990, August). Culture and emotion: The
role of other-focused emotions. Paper presented at the 98th Annual
Convention of the Ameri can Psychological Association, Boston.
Kitayama, S., Markus, H., Tummala, P., Kurokawa, M., & Kato, K.
(1990). Culture and self-cognition. Unpublished manuscript.
Kondo, D. (1982). Work, f ami l y and the self: A cultural analysis of Japa-
nese f ami l y enterprise. Unpublished doctoral dissertation, Harvard
University.
Kuhn, M. H., & McPartland, T. S. (1954). An empi ri cal investigation of
self-attitudes. American Sociological Review, 19, 68-76.
Kumagai, H. A. (1981). A dissection of intimacy: A study of "bi pol ar
posturing" in Japanese social interaction--amaeru and amayakasu,
indulgence and deference. Culture, Medicine, and Psychiatry, 5, 249-
272.
Kumagai, H. A., & Kumagai, A. K. (1985). The hidden ' T' in amae:
"Passive love" and Japanese social perception. Ethos, 14, 305-321.
Laboratory of Comparative Human Cognition. (1982). Culture and
intelligence. In R. J. Sternberg (Ed.), Handbook of human intelh'gence
(pp. 642-719). London: Cambridge University Press.
Lau, R. R. (1984). Dynamics of the attribution process. Journal of
Personality and Social Psychology, 46, 1017-1028.
Lebra, T. S. (1976). Japanese patterns of behavior. Honolulu: University
of Hawaii Press.
LeDoux, J. E. (1987). Emotion. In V. Mount Castle (Ed.), Handbook of
physiology: Vol. 1. The nervous system (pp. 419--459). Bethesda, MD:
Ameri can Physiological Society.
Levy, R. (1973). The Tahitians. Chicago: University of Chicago Press.
Li, M.-C., Cheung, S.-E, & Kau, S.-M. (1979). Competitive and coopera-
tive behavior of Chinese children in Taiwan and Hong Kong. Acta
Psychologica Taiwanica, 21, 27-33. (From Psychological Abstracts,
1982, 67, Abstract No. 11922)
Liu, I. (1986). Chinese cognition. In M. H. Bond (Ed.), The psychology
of the Chinese people (pp. 73-105). New York: Oxford University
Press.
Liu, T. Y., & Hsu, M. (1974). Measuring creative thinking in Taiwan by
the Torrance test. Testing and Guidance. 2, 108-109.
Lutz, C. (1988). Unnatural emotions: Everyday sentiments on a Microne-
sian atoll and their challenge to Western theory Chicago: University of
Chicago Press.
Lykes, M. B. (1985). Gender and individualistic vs. collectivist bases for
notions about the self. In A. J. Stewart & M. B. Lykes (Eds.), Gender
and personality: Current perspectives on theory and research (pp. 268-
295). Durham, NC: Duke University Press.
Maehr, M. (1974). Culture and achievement motivation. American Psy-
chologist, 29, 887-896.
Maehr, M., & Nicholls, J. (1980). Culture and achievement motivation:
A second look. In N. Warren (Ed.), Studies in cross-cultural psychol-
ogy (Vol. 2, pp. 221-267). New York: Academic Press.
Marks, G. (1984). Thinking one's abilities are unique and one's opin-
ions are common. Personality and Social Psychology Bulletin, 10,
203-208.
Markus, H. (1977). Self-schemas and processing information about the
self. Journal of Personality and Social Psychology, 35, 63-78.
Markus, H., & Kitayama, S. (in press). Cultural variation in the self-
concept. In G. R. Goethals & J. Strauss (Eds.), Multidisciplinary per-
spectives on the self. New York: Springer-Verlag.
Markus, H., & Oyserman, D. (1988). Gender and thought: The role of
the self-concept. In M. Crawford & M. Hamilton (Eds.), Genderand
thought (pp. 100-127). New York: Springer-Verlag.
Markus, H., & Wurf, E. (1987). The dynamic self-concept: A social
psychological perspective. Annual Review of Psychology, 38, 299-
337.
Markus, H., & Zajonc, R. B. (1985). The cognitive perspective in social
psychology. In G. Lindzey & E. Aronson (Eds.), Handbook of social
psychology (3rd ed., pp. 137-230). New York: Random House.
Marriott, M. (1976). Hi ndu transactions: Diversity without dualism. In
B. Kapferer (Ed.), Transaction and meaning (pp. 109-142). Philadel-
phia: Institute for Study of Human Issues.
Marsella, A., De Vos, G., & Hsu, E L. K. (1985). Culture and self. Lon-
don: Tavistock.
Maslow, A. H. (1954). Motivation and personality New York: Harper.
Matsumoto, D. (1989). Cultural influences on the perception of emo-
tion. Journal of Cross-Cultural Psychology, 20, 92-105.
Matsumoto, D., Kudoh, T., Scherer, K., & Wallbott, H. (1988). Anteced-
CULTURE AND THE SELF 251
ents of and reactions to emotions in the United States and Japan.
Journal of Cross-Cultural Psychology, 19, 267-286.
Mauro, R., Sato, K., & Tucker, J. (1989). A cross-cultural analysis of the
cognitive dimensions of human emotion. Unpublished manuscript,
University of Oregon, Eugene, OR.
Mauss, M. (1985). A category of the human mind: The notion of per-
son; the notion of self [W D. Halls, Trans.]. In M. Carrithers, S.
Collins, & S. Lukes (Eds.), The category of the person: Anthropology,
philosophy history(pp. 1-25). Cambridge, England: Cambridge Uni-
versity Press. (Original work published 1938)
McClelland, D. C. (1961). The achieving society New York: Free Press.
Mead, G. H. (1934). Mind, self and society Chicago: University of Chi-
cago Press.
Miller, J. B. (1986). Toward a new psychology of women (2nd ed.). Bos-
ton: Beacon Press.
Miller, J. G. (1984). Culture and the development of everyday social
explanation. Journal of Personality and Social Psychology, 46, 961-
978.
Miller, J. G. (1988). Bridging the content-structure dichotomy: Culture
and the self. In M. H. Bond (Ed.), The cross-cultural challenge to
social psychology (pp. 266-281). Beverly Hills, CA: Sage.
Miller, J. G., Bersoff, D. M., & Harwood, R. L. (1990). Perceptions of
social responsibilities in India and in the United States: Moral imper-
atives or personal decisions? Journal of Personality and Social Psy-
chology, 58, 33-47.
Misumi, J. (1985). The behavioral science of leadership: An interdisciplin-
ary Japanese research program. Ann Arbor, MI: University of Michi-
gan Press.
Miyake, K., Campos, J., Kagan, J., & Bradshaw, D. L. (1986). Issues in
socioemotional development. In H. Stevenson, H. Azuma, & K.
Hakuta (Eds.), ChiM development and education in Japan (pp. 239-
261). New York: Freeman.
Miyake, K., Chen, S., & Campos, J. J. (in press). Infant temperament,
mother' s mode of interaction, and attachment in Japan: An interim
report. In I. Bretherton & E. Waters (Eds.), Growing points of attach-
ment theory and research. Monographs of the Society for Research in
ChiM Development. Chicago: University of Chicago Press.
Mook, D. G. (1986). Motivation: The organization of action. New York:
Norton.
Moore, C. A. (Ed.). (1967). Int roduct i on: The humani st i c Chi nese
mind. In The Chinese mind: Essentials of Chinese philosophy and
culture (pp. 1-10). Honolulu: University of Hawaii Press.
Moser, D. (1989). I f this paper were in Chinese, would Chinese people
understand the title? Unpublished manuscript, Indiana University.
Mullen, B., & Riordan, C. A. (1988). Self-serving attributions in natural-
istic settings: A meta-analytic review. Journal of Applied Social Psy-
chology, 18, 3-22.
Murray, H. A. (1938). Explorations in personality New York: Oxford
University Press.
Myers, D. (1987). Social psychology (2rid ed.). New York: McGraw-HiU.
Myers, D. (1989). Social psychology (3rd ed.). New York: McGraw-HiU.
Nakane, C. (1970). Japanese society Berkeley: University of California
Press.
Neisser, U. (1976). Cognition and reality: Principles and implications of
cognitive psychology San Francisco: Freeman.
Neisser, U. (1988). Five kinds of self-knowledge. Philosophical Psychol-
ogy, 1, 35-59.
Northrop, E S. C. (1946). The meeting of East and West. New York:
Macmillan.
Pennebaker, J. W. (1982). The psychology of physical symptoms. New
York: Springer-Verlag.
Phillips, D. C. (1976). Holistic thought in social science. Stanford, CA:
Stanford University Press.
Plath, D. W (1980). Long engagements: Maturity in modern Japan.
Stanford, CA: Stanford University Press.
Reisman, D., Denney, R., &Glazer, N. (1950). The lonely crowd." A study
of the changing American culture. New Haven, CT: Yale University
Press.
Roland, A. (1988). In search of self in India and Japan: Toward a cross-
cultural psychology Princeton, N J: Princeton University Press.
Rosaldo, M. Z. (1984). Toward an anthropology of self and feeling. In
R. A. Shweder & R. A. LeVine (Eds.), Culture theory." Essays on mind,
self, and emotion (pp. 137-157). Cambridge, England: Cambridge
University Press.
Rosch, E. (1978). Principles of categorization. In E. Rosch & B. B.
Lloyd (Eds.), Cognition and categorization. Hillsdale, N J: Erlbaum.
Roseman, I. J. (1984). Cognitive determinants of emotion: A structural
theory. In P Shaver (Ed.), Review of personality in social psychology
(Vol. 5, pp. 11-36). Beverly Hills, CA: Sage.
Ross, L. D. (1977). The intuitive psychologist and his shortcomings:
Distortions in the attribution process. In L. Berkowitz (Ed.), Ad-
vances in experimental social psychology (Vol. 10, pp. 173-220). New
York: Academic Press.
Ross, M., & Fletcher, G. J. O. (1985). Attribution and social perception.
In G. Lindzey & E. Aronson (Eds.), The handbook of social psychol-
ogy (3rd ed., Vol. 2, pp. 73-122). New York: Random House.
Russell, J. A. (1980). A circumplex model of affect. Journal of Personal-
ity and Social Psychology, 39, 1161-1178.
Russell, J. A. (1983). Pancultural aspects of the human conceptual
organization of emotions. Journal of Personality and Social Psychol-
ogy, 36, 1152-1168.
Russell, J. A., Lewicka, M., & Niit, T. (1989). A cross-cultural study of a
circumplex model of affect. JournalofPersonalityandSocialPsychol-
ogy, 57, 848-856.
Sampson, E. E. (1985). The decentralization of identity: Toward a re-
vised concept of personal and social order. American Psychologist,
40, 1203-1211.
Sampson, E. E. (1988). The debate on individualism: Indigenous psy-
chologies of the individual and their role in personal and societal
functioning. American Psychologist, 43, 15-22.
Sampson, E. E. (1989). The challenge of social change for psychology:
Globalization and psychology's theory of the person. American Psy-
chologist, 44, 914-921.
Sass, L. A. (1988). The self and its vicissitudes: An "archaeological"
study of the psychoanalytic avant-garde. Social Research, 55, 551-
607.
Scherer, K. R. (1984). Emot i ons as a mul t i - component process: A
model and some cross-cultural data. In P. Shaver (Ed.), Review of
personality and social psychology: Emotions, relationships, and
health (pp. 37-63). Beverly Hills, CA: Sage.
Schlenker, B. R., & Leary, M. R. (1982). Social anxiety and self-presen-
tation: A conceptualization and model. Psychological Bulletin, 92,
641-669.
Schwartz, J. M., & Smith, W. P (1976). Social comparison and the
inference of ability difference. Journal of Personality and Social Psy-
chology, 34, 1268-1275.
Schwartz, S. H., & Bilsky, W (1990). Toward a theory of the universal
content and structure of values: Extensions and cross-cultural repli-
cations. Journal of Personality and Social Psychology, 58, 878-891.
Shaver, E, Schwartz, J., Kirson, D., & O' Connor, C. (1987). Emotion
knowledge: Further exploration of a prototype approach. Journal of
Personality and Social Psychology, 52, 1061-1086.
Shikanai, K. (1978). Effects of self-esteem on attribution of success-
failure. Japanese Journal of Experimental Social Psychology, 18, 47-
55.
Shikanai, K. (1983). Effects of self-esteem on attributions of others'
252 HAZEL ROSE MARKUS AND SHI NOBU KITAYAMA
success or failure. Japanese Journal of Experimental Social Psychol-
ogy, 23, 27-37.
Shikanai, K. (1984). Effects of self-esteem and one's own performance
on attribution of others' success and failure. Japanese Journal of
Experimental Social Psychology, 24, 37-46.
Shweder, R. A. (1984). Preview: A colloquy of culture theorists. In
R. A. Shweder & R. A. LeVine (Eds.), Culture theory: Essays on mind,
self, and emotion (pp. 1-24). Cambridge, England: Cambridge Uni-
versity Press.
Shweder, R. A. (1990). Cultural psychology: What is it? In J. W. Stigler,
R. A. Shweder, & G. Herdt (Eds.), Cultural psychology: Essays on
comparative human development (pp. 1-46). Cambridge, England:
Cambridge University Press.
Shweder, R. A., & Bourne, E. J. (1984). Does the concept of the person
vary cross-culturally? In R. A. Shweder & R. A. LeVine (Eds.), Cul-
ture theory: Essays on mind, self, and emotion (pp. 158-199). Cam-
bridge, England: Cambridge University Press.
Shweder, R. A., & LeVine, R. A. (Eds.). (1984). Culture theory: Essays on
mind, self, and emotion. Cambridge, England: Cambridge University
Press.
Smith, C., & Ellsworth, P. C. (1987). Patterns of appraisal and emotion
related to taking an exam. Journal of Personality and Social Psychol-
ogy, 52, 475-488.
Smith, R. J. (1983). Japanese society: Tradition, self, and the social order.
Cambridge, England: Cambridge University Press.
Smith, R. J. (1985). A pattern of Japanese society: In society or know-
ledgement of interdependence? Journal of Japanese Studies, 11, 29-
45.
Snyder, M. (1979). Self-monitoring process. Advances in Experimental
Social Psychology 12, 85-128.
Solomon, R. C. (1984). Getting angry: The Jamesian theory of emotion
in anthropology In R. A. Shweder & R. A. LeVine (Eds.), Culture
theory." Essays on mind, self, and emotion (pp. 238-254). Cambridge,
England: Cambridge University Press.
Solomon, S., Greenberg, J., & Pyszczynski, T. (in press). A terror man-
agement theory of social behavior: The psychological functions of
self-esteem and cultural worldviews. Advances in Experimental So-
cial Psychology
Srull, T. K., & Gaelick, L. (1983). General principles and individual
differences in the self as a habitual reference point: An examination
of self-other j udgment s of similarity Social Cognition, 2, 108-121.
Steele, C. (1988). The psychology of self-affirmation: Sustaining the
integrity of the self. In L. Berkowitz (Ed.), Advances in experimental
social psychology (Vol. 2 I, pp. 181-227). San Diego, CA: Academic
Press.
Stevenson, H., Azuma, H., & Hakuta, K. (1986). Child development
and education in Japan. New York: Freeman.
Stewart, A. J., & Lykes, M. B. (Eds.). (1985). Conceptualizing gender in
personality theory and research. In Gender and personality: Current
perspectives on theory and research (pp. 2-13). Durham, NC: Duke
University Press.
Stigler, J. W, Shweder, R. A., & Herdt, G. (Eds.). (1990). Culturalpsy-
chology: Essays on comparative human development. Cambridge, En-
gland: Cambridge University Press.
Stipek, D., Weiner, B., & Li, K, (1989). Testing some at t ri but i on-emo-
tion relations in the People's Republic of China. Journal of Personal-
ity and Social Psychology, 56, 109-116.
Stryker, S. (1986). Identity theory: Developments and extensions. In K.
Yardley & T. Honess (Eds.), Sel f and identity (pp. 89-104). New York:
Wiley.
Swann, W B., Jr. (1983). Self-verification: Bringing social reality into
harmony with the self. In L Suls & A. G. Greenwald (Eds.), Psycholog-
ical perspectives on the se/f(Vol. 2, pp. 33-66). Hillsdale, NJ: Erl-
baum.
Swann, W. B., Jr., & Read, S. J. (1981). Self-verification processes: How
we sustain our self-conceptions. Journal of Experimental Social Psy-
chology, 17, 351-372.
Tajfel, H. (I 984). The social dimension: European developments in so-
cial psychology Cambridge, England: Cambridge University Press.
Takata, T. (1987). Sel f-deprecat i ve t endenci es in self-evaluation
through social comparison. Japanese Journal of Experimental Social
Psychology, 27, 27-36.
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social
psychological perspective on mental health. Psychological Bulletin,
103, 193-210.
Tesser, A. (1986). Some effects of self-evaluation maintenance on cog-
nition and action. In R. M. Sorrentino & E. T. Higgins (Eds.), Hand-
book of motivation and cognition: Foundations of social behavior (pp.
435-464). New York: Guilford Press.
Triandis, H. C. (1989). The self and social behavior in differing cultural
contexts. Psychological Review, 96, 506-520.
Triandis, H. C., Bontempo, R., Villareal, M. J., Asai, M., & Lucca, N.
(1988). Individualism and collectivism: Cross-cultural perspectives
on self-ingroup relationships. Journal of Personality and Social Psy-
chology 54, 323-338.
Triandis, H. C., & Brislin, R. W. (Eds.). (1980). Handbook of cross-cul-
tural social psychology (Vol. 5). Boston: Allyn & Bacon.
Triandis, H. C., Marin~ G., Lisansky, J., & Betancourt, H. (1984). Sim-
patla as a cultural script of Hispanics. Journal of Personality and
Social Psychology, 47, 1363-1375.
Veroff, J. (1983). Contextual determinants of personality Personality
and Social Psychology Bulletin, 9, 331-344.
Wada, M. (1988). Information seeking in self-evaluation of ability [Ab-
stract]. In Proceedings of Japanese Psychological Association Meet-
ing, 52, 222.
Waterman, A. S. (1981). Individualism and interdependence. Ameri-
can Psychologist, 36, 762-773.
Weber, M. (1958). The Protestant ethic and the spirit of capitalism (T
Parsons, Trans.). New York: Scribner.
Weiner, B. (1986). An attributional theory of emotion and motivation.
New York: Springer-Verlag.
Weisz, J. R. (in press). Culture and the development of child psychopa-
thology: Lessons from Thailand. In D. Cicchetti (Ed.), Rochester
Symposium on Developmental Psychopathology (Vol. 1). New York:
Cambridge University Press.
Weisz, J. R., Rothbaum, E M., & Blackburn, T. C. (1984). Standing out
and standing in: The psychology of control in America and Japan.
American Psychologist, 39, 955-969.
White, (3. M., & Kirkpatrick, J. (Eds.). (1985). Person, self, andexperi-
ence: Exploring Pacific ethnopsychologies. Los Angeles: University of
California Press.
White, M., & LeVine, R. A. (1986). What i san liko(good child)? In H.
Stevenson, H. Azuma, & K. Hakuta (Eds.), Child development and
education in Japan (pp. 55-62). New York: Freeman.
Whitley, B. E., Jr., & Frieze, I. H. (1985). Children' s causal attributions
for success and failure in achievement settings: A meta-analysis.
Journal of Educational Psychology, 77, 608-616.
Witkin, H. A., & Goodenough, D. R. (1977). Field dependence and
interpersonal behavior. Psychological Bulletin, 84, 661-689.
Witkin, H. A., Goodenough, D. R., & Oltman, P. K. (1979). Psychologi-
cal differentiation: Current status. Journal of Personality and Social
Psychology, 37, 1127-1145.
CULTURE AND THE SELF 253
Wylie, R. C. (1979). The self-concept: Vol. 2. Theory and research on
selected topics. Lincoln: University of Nebraska Press.
Yamagishi, T. (1988). Exit from the group as an individualistic solution
to the free-rider problem in the United States and Japan. Journal of
Experimental Social Psychology, 24, 530-542.
Yang, K. S. (198 la). The formation of change of Chinese personality: A
cultural-ecological perspective [In Chinese]. Acta Psychologica Tai-
wanica, 23, 39-56.
Yang, K. S. (198 lb). Social ori ent at i on and i ndi vi dual moder ni t y
among Chinese students in Taiwan. Journal of Social Psychology,
113, 159-170.
Yang, K. S. (1982). Causal attributions of academic success and failure
and their affective consequences. Acta Psychologica Taiwanica, 24,
65-83. (From Psychological Abstracts, 1985, 72, Abstract No. 13126)
Yang, K. S. (1986). Chinese personality and its change. In M. H. Bond
(Ed.), The psychology of the Chinese people (pp. 106-170). New York:
Oxford University Press.
Yang, K. S., & Liang, W. H. (1973). Some correlates of achievement
motivation among Chinese high school boys [In Chinese]. Acta Psy-
chologica Taiwanica, 15, 59-67.
Yoshida, T., Kojo, K., & Kaku, H. (1982). A study on the development
of self-presentation in children. Japanese JournalofEducational Psy-
chology, 30, 30-37.
Yu, E. S. H. (1974). Achievement motive, familism, and hsiao: A replica-
tion of McClelland-Winterbottorn studies. Dissertation Abstracts In-
ternationaL 35, 593A. (University Microfilms No. 74-14, 942)
Zuckerman, M. (1979). Attribution of success and failure revisited, or:
The motivational bias is alive and well in attribution theory. Journal
of Personality, 47, 245-287.
Recei ved Fe br ua r y 1, 1990
Revi si on recei ved J une 28, 1990
Accept ed July 11, 1990
Ame r i c a n P s y c h o l o g i c a l As s o c i a t i o n Today' s Dat e
S u b s c r i p t i o n Cl a i ms I n f o r ma t i o n
This form is provided to assist members, institutions, and nonmember individuals with any subscription problems. With the appropri-
ate information provided, a resolution can begin. If you use the services of an agent , please do NOT duplicate claims through them and
directly to us. PLEASE P R I N T C L E A R L Y A N D I N I N K I F P O S S I B L E .
P I ~ P UL L NAMB OR ~ N A ~ O F E ~ l f f l T l ~ N
~DRESS
C r r Y S T A T ~ C O U N T R Y Z I P
YoUR N^MB AND rHON~ NUMS~
Trn~s) VoL/YR.
ICa~mt oR c u s - r o ~ ~'UMS~ ( UA~ S~ FOt ~ ON ~ pAS'r tSSUa t ~ t . )
DATE YOUR ORD~ WAS MAE.I~) ( OR mOl ~ )
P. O. N U M B E R
I ~ U ~ ^ ~ C H E C ~ C ~ t A R G n . Cm~. / C, um c u ~ u ~ D ^ 1 " ~ _ _
(If possible, s c a d a c o p y , f r o n t a n d b a c k , o f y o u r c a n c e l l e d c h e c k t o h e l p u s in
o u r r ~ . ~ h o f y o ~ clalm.)
I s s u e : M ~ o .. l ~ w o ~
I s s o ~ s ) No. / MoN~
Thank you. Once a claim is received and resolved, delivery of replacement issues routinely takes 4-6 weeks.
a mmm i i i a m i i a m i m m i ~ m i i i i m i ( T o BB F H' J - J ~ OUT B Y A ~ A S T A P F ) m i i i i m i i I B m i m" m i m ~ ~ i ~ i
DATE 1 ~ . DATE O F ACTION
ACTION T A~ INV. No. & DATE
S T ~ NA~ LABEL #, DATE
l i b i i l l I f H a l l a i D U l l i I f ~ i I 1 ~ i I n i i m i i I l i a i l ~ l i b i l i a ~ i ~ I ~ i i i i i i i i i I ~ i ~ i
SgND TInS FORM XO: APA Subscription Claims, 1400 N. Uhle Street, Arlington, VA 22201
P L E A S E D O N O T R I ~ O V E . A P H O T O C O P Y M A Y B E U S E D .