Professional Documents
Culture Documents
GUTMANN. I D E N T I T Y A N D DSMOCEACY
1 543
, , ,.
AMY GUTMANN
ries of justice ever do (Rawls 1971, 1999L; Carens 2000). In every existing
nonideal democracy, individuals identify themselves and are identified by
'
others,. especially
for public purposes,
with voluntary and involuntary
.
. .
groups defined by social markers such as nationality, ethnicity, gender,
'!'
race, age, disability, and political ideology. Even if none of these group
identities-nor all of them taken together-conipletely defines any individual, group identities both constrain and liberate individuals, depending
,
on the identitv.. the,individual. and the context f~ A ~ o i a h1996. 97-991.
Group
identities also play a central role in democratic politics, for better
" ,
'
and for worse, because most individuals can be politically influential only
in groups and some groups are far more justice friendlythan others.
..;) , ,,,
. The scholarship.on identity and democracy is internally divided over
: . :whether voluntary or involuntary group identities generally operate for the
better or the wane in democratic politics. The "school of culture" emplia,
'
sizes the essential contribution that cultural groups make to the lives of individuals in providing a sense of secure belonging and a set of scripts that
,,
'
give
meaning to individual lives. T h e school of culture therefore warns
,
against treating the ideal of the free and equal person as if actual individu,
als could be conceived independently of any and all cultural contexts. By
,; , , .
contrast, the "school of choice" emphasizes the value of individual freedoln
from involuntary groups, the freedom to criticize and revise culturally
;, .
given identities, and a correlative. right of free association. The school of
clloice therefore issues precisely the opposite warning from the school of
.,
'
cultu": Do not heat cu~turalgroups as if they were primordial, sovereign
,
authorities over individuals who should be accorded the civic standing of
free and equal Persons indemocraticsocieties.
,
,
: '
1" this essay, J propose a synthesis that avoids the temptation, on the
one hand, to elevate cultural groups over the individual, and, on the other
,
hand, to conceive of individuals free from socially given identities. I build
,,:
this synthesis on the insights offered by the schools of culture and choice at
.tl~eir
strongest, and I therefore criticize those claims that rest on a concep,
Lion of democratic societies as mere aggregations of either comprehensive
' ;
cultural groups or free and equal individuals. A synthetic penpecfive sib,' ,
,; ates the humanist ideals of freedom, opport~lnity,and civic equality for all
persons within nonideal democratic contexts and, by so doing, offen a per,.
spective on identity politics that is absent from ideal theories of justice. In
. the context of a society still suffering from a legacy of racial and gender dip
, , . crimination, for example, racial minorities and women often have no better political alternative than to engage in collective action with the aid of
,
identity groups and the aim of achieving greater civic equality and justice.
. . ,, For members of subordinated groups to eschew all identity associations
would be counterproductive to justice. Yet not all identity grot~psare jug
.
,
tice
friendly, and therefore a synthetic perspective needs to distinguish
.
,
: among different kinds of asciiptive identity politics. Without the National
Association for the Advancement of Colored People (NAACP), for exam!
..
,
Tile ever-growing attention to group identities and identity groups in normative political theory parallels their saliency in contemporav.democntic
politics. c~~~~~
of individuals l,ound together by a shared social identity
disability, or pb
such as nationality, ethnicity, gender, color, class,
litical ideology make political claims that range from exempting them.
selr,es from disproportionately burdensome laws to exercising sovereign^.
A large literature has developed over the past three decades centering
around the claims of identity groups in democratic politics, often begin;
ning with the premise that individual identities are socially constnlcted
(rather than essentialist) and then pursuing the political implications of srr
c.al construction. ~h~ idea ofthe social conshction ofidentity is so open
ended, however, that it has no political implications, leaving us with the
question of whether the use of identity in political analysis is at all meaningful (Bmbaker and Cooper 2000).
wllat is identity group politics7 is it reducible to interest group polis
tics? Are there any political implications of a shared group identity, for examp]e, for exemption from othewise valid laws or for group
criti&
of identity group politics often
it to tile free association of individ.
uals. Free association is thought to enable people to develop their identities
as they see lit, not as any group detemlines for them. Yet associational freedom permits group exclusivity, and exclusions based on ascriptive group
identity perpetuate a negative identity group politics, which leaves many
people-simply by virtue of their group identity-with less than equal free/
dam,
and civic standing. should democratic governnlenh
therefore constrain civic associations
to discriminate against
on
the basis of their group identities?
A growing and wide-ranging scholarship on group identity and associ.
ational freedom in nonideal democracies addresses this question. It suggests that ' W h o should decide what and how in democratic politics?"
cannot, be answered witllout taking more
of the political role of
group identities and civic associations in nonideal contexts than ideal the@
h:
, ,
..
pie, Nrican Americans would be worse off than they are today, whereas
without the Klu Kluv Klan (KKK) they would be better off, yet both are ascriptively based associations. To tell African America~lsthat they should
not identify as a grollp for the purposes of democratic politics is to tell
them to ignore both history and contemporary reality; to tell them that they
should only identify as an ascriptive group would be similarly misleadin8
Many commentators on identity politics condemn it without qualifiwhile many others uncritically applaud i t To understand and evaluate different kinds of identity groups and how they function in nonideal
delnocratic contexts, 1 begin by recognizing the irreducible tension'between individual freedom and. the shaping of i~~dividual
identities and
democratic politics by groups. When we recognize this irreducible tension,
we are in a better position to.understand the re1ationship:among group
identities, democratic politics, and the ideals of individual freedom, oppor!
tunity, and civic equality The first section asks what identity groups are;
and what is the relationship between identity and interest group politics,
Is interest
Can identity group politics be reduced to interest group
group politics, as commentators suggest, "an inherent part of the governing
fiocess ?f ademocracy" while identity group ~olitics"is antithetical to the
basic principle of one indivisible nation" (Connerly 1997)7 The next two
sections examine two opposing schools of thought on the role of group
identity in democratic politics: the schools of culture and of choice. The f i ~
nal section develops the implicatipns of a more synthetic perspectiveby r b
visiting the divisive issue of group rights and by raising the question of liow
missing identities might alter democratic landscapes.
I
An identity group, as the name suggests, is bound up with who people are1
not merely with what they want, The distinctive and definingfeature oflan
identity group is the identification of its members as a certain kind of petson, and therefore its members' muti~alidentification with people of that
kind, where the kind is a consequential social category (Fearon 1999).This
mutual identification around a social category is independent of-yet alsa
clearly compatible with-the pursoih of instrumental ends by the group1
When a group of people who mutually identify around a social c
act in politics on the basis of a group identity-whether for the sak
gaining recognition for the group or furthering the interests of the group
they are part of identity group politics.
Identities and therefore identity groups are socially constructed.
this, however, is not to say much more than that genes and physiognolnie
do not determine our social identities. "People are limited by, but they~arb
not prisoners of, their genes, their physiognomies, and their histories in settling on their own identities. And ifpowerful social forces motivate idelltity
exploration-as they seem to do in our age-it is the constructivist face of
identity that seems the lhore real" (Laitin 1998a, 21). Almost everytl~ing
that informs our social identities-such as race, gender (as distinct from biological sex), etlinicity, and nationality-can be called social c o n s h c tions. To say that racial, gender, ethnic, and national identities are social
constmctions is not to say that they are any easier to change than our genetic inheritance or physiognomy. Most African Americans, women, and
deafpeople cannot '!passn for white, men, or hearing individuals; they can
reinterpret. these ascriptive identities but they cannot give them up. For
Some people, ethnic identities that are connected to a native language may
he somewhat easier to change, by a decisionto speak a second language
and give up certain customary ways of acting, but even an ethnic identity
can be difficultto alter except by generational change.
1 . Yet many people do change their ascriptive identities over time, somelimes quite deliberately and strategically. Tlle social conshuction of identity is most evident when identities are "constructed and reconstructed
as'social opportunities change" (Laitin 1998a, :20). To account for such
uhange, especially among ethnic minorities, political scientists propose a
raltibnal-choice theory about the social constmction of identity. A rationalhoice theory posits that people retain or change their social identities acording to what best satisfies their interests,~andwhat best satisfies their
interests depends in turn on the payoffs people predict from the alternate
identities that are available to them. According torational-choice theory, in
any given social context, people choose the available identity that they predict will best serve their interests. When identity change is possible, on this
theory, it is interest driven. Even ascriptive identity, wlien it is not simply a
social given, is one among many tools in the arsenal of interest-based politics. But this does not mean that identity politics is reducible to interest
pdlitics; what it meansis that identities and interests interact, and the twb
kihds of politics work together. Both need to be understood for what each
qontributesto understanding politics.
i l l ' , A rationalchoice theory that admits all kinds of actions as interest
based (by positing interests that conform to all the revealed preferences of
beople) is a hollow shell into which all actions can be fit and said to conford to its premise: people act in ways that reveal (and therefore reflect)
'?interests. Actions reveal people's preferences, and their interests are
to be their revealed preferences. All political action can be placed
such a rational-cl~oiceframework before it is analyzed. Used as an
lsifiable framework of analysis, a rationalchoice theory cannot claim
,identity.politics is reducible to interest group politics in any substan.meaningful sense. If we stipulate that all human behavior is interest
t h e n everything is redircible to interest group politics, but not in a
hy that informs our understanding of how democratic politics works. To
say that human behavior is interest driven is to say nothing, for example!,
understand actions that reflect group identification or moral commihn
on the part of people and those that reflect notliing more than a .
welfare goal. T h e most plausible alternative to reducing all identity pa
to interest politics does not deny that changes in identity areinteres
driven. Nor does it claim that identity politics conflicts with a politibs
based on interests. Rather it suggests that identities inform people's int
ests. Because identity often informs interest, identity cannot be reduced,
interest. Interests do not always, or even generally, precede identily inlb
way that would pennit an insightful observer to explain people's beha
by their prior interests, without knowing their identities.
How then is an identity group analytically.different from an interes
group? T h e ideal typeof interest group organizes around a shared instrul
mental interest of individuals (Olson 1971). T h e ideal type of iden '
group organizes around mutual identification among its members. In
litical practice, most organized groups are both identity and interest
groups. Members are drawn to the group because of their mutual ident
cation and because they share an instrumental interest pursued by
group. David Truman subsumes identity and interest groups, and eve
thing Madison called a "faction," into his definition ofinterest groups: "ad
group that. .:. makes certain claims upon other groups . . . for the est
lishment, maintenance or enhancement offorrns of behavior that are I
plied by the[ir] shared attitudes" (1971, 33). More analytical definitions or
identity a n d interest groups that are not so broad as to verge o
meaningless are preferable.
T h e defining features of identity and interest groups-the mutual
identification of people with one another, on the one hand, and the see
ing together of a shared inshumental goal, on the other-are often mu
allv reinforcing. Most identiw. nrouvs
- . .pursue instrumental interests of the1
members and thereby encourage more people to see their identities ,a$ '
bound up with a group. Many interest groups encourage people to join'bx :'
orienting themselves around some mutual identification that is broadei.
than the specific interests that they are pursuing at any given time. Some
interest group theorists call these solidary.incentives (Cigler.and Loomis
1995.9). T h e greater the role that identity plays illattracting and retainin
members, the more a group is a n interest group. T h e greater the role tha
the pl~rstlitof shared instrumental interests .ulavs
, in atiractine: and retain
ing members, the more a group is an interest group. Most groups that.adt
in democratic politics are a combination of both, and for two good reasons
First, people's identification with one another influences their sense
what they .want, and people's sense of what they want influences
whom they identify. Second, individuals who identify with others are b
able to organize politically, and organized groups can be far more p
cally effective than an equal number of ~~nilllied
citizens (Moe.1980). Th
observation provides the single most salient answer to the question olwl~y
identity matters so much, in democratic politics. Group identificawhether it be around gender, race, religion, sexual orientation, eth,nationality, age, disability, or i d e o I ~ ~ ~ - ~ r o v ino
d e less
s "real" a set
ns for people to organize politically than purely material incena.As important, the two kinds of reasons-identity and interest-are ofen mutually reinforcing.
I : , . Once we recognize the reality of identity group politics and its inter'ction with interest group politics, along with its social constniction, we
oid reification, which is the single most common charge of academic
ltics who suggest that analyzing identity politics as a real phenomenon
hntails approving o f " 'groupness' itself" (Brubaker and Cooper 2000, 31).
uite the contrary, we are now in a better position to consider two serious
alleliges that identity politics presents in democratic politics. First, illib
al i d e n t i t y ' g i ~ uerect
~ s obstacles to the civic equality ofsome members,
oftin women and disadvantaged minorities, who do not want to choose between their cultural group identity and their recognition as equal.citizens.
Must members of illiberal cultural groups be forced to choose? Second,
Juluntary groups claim the right to associate as their members see fit and
therefore the right to exclude those they deem unfit for membership.
hould civic associations that preiudicially exclude people associated with
~istorically disadvantaged identity groups be permitted to discriminate
eden at the cost of civic equality, equal liberty or opportunity for those who
t excluded? In examining how the schools of culture and choice address
ese challenges, I develop a more synthetic perspective on identity and
emocracy, which is consistent with furthering the ideals of civic equality,
'c liberty, and opportunib for all persons, regardless of their group iden-
,,
W'
Jj,.,.
I.
The School o f C u l t u r e : Q u e s t i o n i n g C l a i m s a b o u t
Comprehensiveness
,.
i?'See Kymlicka 1995a, 18. What Kymlicka calls a societal cblture, Joseph Raz and
bUhii Margalit call a pervasive culture. A pe~vasiveculture "defines or marks a variety of forms or styles oflife, types ofactivity, occupation, pursuit, and relationship;
\J(ithnational groups we expect to find national cuisines, distinctive architectural
siyles, a common language, distinctive literary and artistic traditions, national music, customs, clress, ceremonies and holidays, etc. None of these is necessary:' Peoplecwlla share-or come close to sharing-a peniasive culture and whose "identiiy
lodetermined a t least in part by their culture" are "~erionscandidates for the right
to self-determination" in the account offered by Margalit and Raz (1994, 114).
.. , , 1!.,;
$,',
"
548
Set. 3
CITIZENSHIP,
IDENTITI,AND POLITICAL
PARTICIPATID~
CUTMANN
I D E N T ~ T YA N D DSMOCRICI
549
Group rights that exempt members of some groups from pllblic pol
that impose unfairburdens and special aid to overco
vantages illustrate this prospect. The problem is.that other group
may violate basic individual rights. Group rights that
of individuals in order to protect an endangered cult
fend absolute sovereignty for the cultural group il
~h~ cllallenge is therefore to distinguish behveen
mate group rights (Levy 1997; Lukes 1997; William
The cotinby that is the theoretical home ofgr
example of the defensibility of special exemptions
unfair burdens on members of an identity group (because of their.
identity). Such special exemptions are meaningfully considered a k
group right, since they apply toindividuals only by virtue o f t
identity.' In 1990, the Royal Canadian Mounted Police decided to ex
a group from $ long-standing rule. They exempted
wide-brimmed hat that has long been part of their required ~m
exemption made it Possible for Sikhs to join the Mounties wi
up their identity as Sikhs. One might consider this decision i
the rule is far from earthshaking in its significance-but it is quite
of many rules that have raised public controv
Mounties' decision met with six years of protests,
to parliament signed by 210,000 Canadians and de
antiturban badges and antiSikh slogans and'ending with an ap
to the Canadian Supreme Court,which refused to hear tlie chal
(Winsor 1996).
i,
The Mounties'example illustrates hvo important but o
points: ( I ) no controversial claim about identity is needed to def
empting members of a group even from some rules that reflect no in
discriminate; and (2) somegroup rights can provideeffect
specting individual rights, which classical~liberal defender
rights have neglected but need not oppose. On a Lockea
nlent, thegovemment is not required to exempt minoriti
neutral laws that unequally burdentheir religious beliefs, but neithe
prohibited.from doing so as long as the exemption does not violate s
one else's rights.! A. crucial feature of this exemption is that it li
burdens (borne by members oTa group through no faultof their o
:
idua) rights. By contrast, other group rights-violate basic individual rights. No democratic
pective that takes the interests of individuals seriously can defend cul1 practices that violate basic individual rights, whether of members or
onmetnberi of a cultural group.5 Because the moral value of the group
es from the way it supports the interests of individuals over time,
taining a customary practice of the group at the expense of the basic
ies and opporlunities of individuals is morally indefensible (Blake
ands the right of self-governance over its customary
se practices include extensive discrimination. in
sic rights of some of its members (often women
(children), its demands are suspect to the extent that the larger society
IdothenCise protect those rights. The group may still have eaensive
ts' of self-govemment, but self-governing rights-like all rights-can
mpeting rights. Even when groups cannot have
emment, they may be justifiably guaranteed r e p
tion within a legislature that would otllenuise routinely pass dist the group, which the courts would uphold
teed representation f0r.a group is inore probdemocratic the group is and the more its. memedoms for their votes to be counted as part of the
emerge that call for a far more nuanced view of
lly offered: '(1) group rights of self-government
all or,nothing, and (2) group rights should be
ssedon the basis of how well they protect the.basic interests-includividuals compared to the available institutional
I
.,
e aesthetic value of cultural practices is a differentissue, which I cannot adnmthis essay. Charles Taylor attributes to individualsan interest in their culurviving over time, since people want to be assured that their descendants will
the same culture as they do. This raises the difficult question ofwhat the barests (and therefore the basic rights) of individuals are and how they should
rmined (Taylor 1994, 51-7j); and folio sense of how heated the debate over
re can be, see David Biomwich, Michael Walzer, and Taylor in Dissent
. .
I'
as oppression. Many of .the group rights that are prominently dein contemporary politics-and also in political theory-threaten to
ate the subordination of half the world's population, namely
en. The cultural relativist view is that the cultural identity of women
s attached to these practices, which are not identified as oppressive
cause this interpretation is itself identified as foreign to the indigenous
tice. To try to obliterate the practices is viewed as tantamount to trying
rate these women, to do away with their cultural identities, with
are. T h e relativizing move here allows a cultural identity to go
1 and lets it block any vindication of the basic rights of
n. O n this analysis,, cultural identity precludes the vindication of
is as much of an imn's equal righti with men, but this
tlieoretical construct as any. A cc~lhtral identity is assumed-alever actually shown-to comprehend the entirety of a person's
nd a cultural identity is allowed to become morally imperial and
to deny basic rights to women. Group rights to enforce child
iriage,, forced marriage, polygamy (or, more accurately, polygyny), clictomy, discriminato~ydivorce, unequal schooling of girls and boys,
nequal property rights threaten the basic rights of women in a way
ument from cultural identity alone has ever been able to bvef
?ere any reason that the dominant members of cultural groups
ranted the right to resbict the basic rights of women?
I within the school of culture explicitly defends a broad culthin liberal democracies to restrict the basic rights of women.
alit and Moshe Halbertal argue that, "Protecting cultures out
right toculture may'take the form o f a n obligation [ofliberal
vernments] to support cultures that flout.the rights of the iniberal society" (1994,491). This defense of the right to culgoes s i far as to claim that liberal democracies have an
ibligation to support cultures in their midst that Rout the rights of women,
ndividnals. Ultra-Orthodox Jewish culture in 1srael;accoidment, must be suppoited by the state in its totality, which
g the rights ofwomen in schooling, divorce law, and inheiecting Ulha-Orthodox men as the guardians of the authora publicly supported cultural community (493). One
is treated as independent ofany particular culture and that
it a culh~ralc o m m u n i t y ? ~ ~ iiss the right that undermines
ral relativism. If individrlal identities are so comprehen'
fi,,.
,,
'
,~,For,~an
eltended discussion, see Okin, Moller, et al. 1999, Nussbaum and
Clover 1994, and Nussbaum 1997b.
e right of exit to a free market society is also the only right of individuals reced by Chandran Kukathas who clefends the right of minority groups-no mas
ow~illiheral-to be left aldne (1992, ,105-39). Wr a critiqueof Kukathas on the
gl~t~ofexit;see.Kymlicka,
199Sa; 190,.234-35, n. 18.
'
10. One might revise the perspective to say thata person has
only in her personaliv identity asshe defines it, not as defined by other m
her cultural group. Tliis starting point would svpport many more col
pendent rights of individuals than the right h exit. In'dividtrals wbuld
fight to resist having a cultural group impose laws and rules
their personality iden6Hes:Buf this is not the view of personality identi
--I;+
...A
U..IL+-~ A-F~nrl T h ~ view
r
<In+$
not n ~ m i nnv:cleantdis
t
s,and so co~nmonlyexcused or even iustiiy isn't the formal right of exit an adequate antidote to granting cultuihl groups the authority to.heat their members in illiberal ways? Once
t to exit from cultural groups, they are imof individual consent and respect for perntly justify subordinating individuals in
tura1,groups that deprive them of the necessary conditions for makal right of exit can he an adequate
ich truly voluntary associations treat
present societal cultures-by the very
re not voluntary associations. When
~iasi-com~rehensivecultural groups violate rights withthe support of
tic governments, they greatly reduce the basicliberties and o p p o ~
hat their members would otherwise have to live a life oftheir own
sing either inside or outside t h e communily. An Ultra-Orthodox
an in Israel is not educated to support herself or her children outride
e'ultra-Orthodox community. She is not taught the legitimacy of
tioning her tightly circumscribed place in the Ulha-Orthodox com;
claim that she has an overriding interest in preserving a culturim that her overriding interest is to
ive set of laws over which she has no
reting or altering."This is not a claim thatanyond can make
.
claiming the subordination of women is jusinterest lies in being subordinated. For anyse, women, like men, must be given equal
want to live within therange of
om in the first place cannot validate
want or need this freedom. No theoretical
nation ofwomen within~culturesthdt deny .
al right of exit. For women to,Ilave
ler they want to remain within a culhra]
to exit thatgroup, a democraticsociety
ht to exit, an education for equal citizen.
r basic rights as well. An effective right to
ns at minimum being able to exit and still support oneself and
ildren. Securing women these rights means not ceding cultural
s within a democracy the authority to control schooling and family
as to effectively deny women the same basic liberties and opportuni,.
..
..!
'
556
sac. 3
CITIZENSHIP,
IDENTITI,
AND
POL~T~CAL
PAI~.IICIPATION,
:
ties as men. Democratic states may need to tolerate some violations of1
sic rig\lts when political intervention would make matters worse. bd
pragmatic argument of this kind denies any obligation on
democratic state to support cultural groups at the expense of vi
basic rights of individuals.
Taken at its strongest, the school of culture recognizes that ther
group right to subordinate individuals to a societal culture by via
their basic rig&. At its weakest, it erects indefensible obstacles to vin
jng the basic rights of members and nonmembers of cultural groups, E
if the societal culture of the Israeli Ultra-Orthodox Jewish communi
self-contained for members of that community, its effects on nonm
would still be morally and politically relevant. T h e ~oliticsof cultura
norities rarely can be isolated from a broader politics, which to extendt
~sraeliexample, profoundly affects the liberty and opportunity of Palesti
ans, Israeli Arabs, and non-Orthodox Israeli Jews.
T h e politics of cultural identity groups also shapes the identities
ture generations in ways that the school of culture assumes to be le
extensions of a group right to self-determination. Cultural groups w
democratic societies are a n intricately interdependent part of the
ety. They must be able to interact politically and often economicall
with people outside their groups, and people outside their groups wi
This interdependence, moreover, affects the liberty, opporh~ni
standing of both outsiders and insiders to these groups since each-dep
on the other to ensure the equal protection of rights. If some membe
cultural groups identify only with a single group, it is therefore not be
the comprehensive conditions of their lives are independent of the lar
ciety. (Ultra-Orthodox Jewsin Israel depend heavily on support fro
of society for their schools, police protection, and national securi
ben of culh~ralgroups also influence the larger politics of democra
eties as much as many other citizens, sometimes even more s o (
example also illustrates.) Members of distinctive cultural grou
ten more dependent on the larger society tosupport their distin
This is not a n argument against the dependency of cultural groups o
influence on the larger society Quite the contrary, it is an argument ?,
a politics or political theory that posits the comprehensiveness of societ
tures and therefore depends on denying the interdependence of cultur
cultural communities within democratic societies.
Wlll
~.
..
,I
564
set., C
E N S ~ ~IpD ,E N T ~ T I ,A N D
POLITICAI.PA
..I
1302 (8).
565
i
i
!
I
iI
;i
1
:j
:'i
!
!I
1
!
. .
ne