ArticlePDF Available

National Minorities and Ethnic Immigrants: Liberalism's Political Sociology



This article offers a sustained critique of Will Kymlicka's argument in MULTICULTURAL CITIZENSHIP that a liberal theory of minority rights ought to differentiate between two categories of ethnocultural groups, national minorities and ethnic immigrants, in terms of the rights that they should be accorded. National minorities are previously self-governing peoples who have been incorporated into states in which they do not constitute the majority group, either voluntarily (through treaties of cession or federation) or involuntarily (through conquest). Ethnic immigrants are cultural groups who have migrated to societies where they do not constitute a majority through voluntary immigration. According to Kymlicka, national minorities ought to possess "self-government rights" that enable them to erect separate institutions, and which may take the form of federal arrangements. Ethnic immigrants, by contrast, are entitled to "polyethnic rights" that facilitate their membership in common institutions, and include guarantees of non-discrimination and even exemptions from laws that disadvantage ethnic immigrants because of culturally specific practices. Kymlicka runs two different lines of argument to defend this taxonomy of ethnic groups and rights, which I term the argument from consent, and the argument from political sociology. In this article, I describe both arguments, and explain why neither can justify the hierarchical distinction that he draws between different ethnocultural groups. I also suggest that the shortcomings in both arguments suggest the limited work that cultural membership can do in a liberal theory of justice.
The Journal of Political Philosophy: Volume 10, Number 1, 2002, pp. 54±78
National Minorities and Ethnic Immigrants:
Liberalism's Political Sociology*
Law, University of Toronto
OVER the past decade, political philosophers have increasingly focussed
their gaze on the complex relationship between liberalism and the claims of
ethnocultural groups for the recognition by public institutions of their distinctive
ways of life. Conventionally, the liberal tradition has been viewed as being
indifferent or even hostile to these claims, in large part because of its commitment
to neutrality. In this context, neutrality is understood as the agnosticism of public
institutions in liberal democracies toward differing conceptions of the good. The
stance of liberal states toward religionÐone of both non-endorsement and non-
interferenceÐis often offered as a paradigmatic example. Inasmuch as
ethnocultural groups aim to enlist public institutions to preserve ways of life
based on a shared set of ends, the tension between these sorts of claims and the
central liberal tenet of neutrality is fairly clear. The proposal is to extend the
liberal treatment of religion to culture more generally. In Michael Walzer's
memorable phrase, ``the state is nationally, ethnically, racially, and religiously
Scholars sympathetic to the claims of ethnocultural groups have grappled with
the obstacle posed by liberal neutrality in two different ways. Some have argued
that as a matter of principle, liberal neutrality is an illegitimate standard for
political decision-making. They claim, for example, that neutrality is an
impossible ideal, since policies enacted for reasons that do not purposely seek
to advantage or disadvantage any conception of the good will inevitably be non-
neutral in their effects. Legislating with an awareness of the possibility of non-
neutral effects is therefore tantamount to legislating for non-neutral purposes,
albeit not explicitly. Neutrality becomes little more than a cruel facËade which
#Blackwell Publishers, 2002, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street,
Malden, MA 02148, USA.
*An earlier version of this paper was presented at the Harvard Center for Ethics and the
Professions. For helpful comments and discussions, I thank Arthur Applbaum, Joe Carens, Mary
Coleman, Kevin Davis, David Dyzenhaus, Bob Goodin, Pamela Hieronymi, Robert Howse, Richard
Katskee, Nancy Kokaz, Patrick Macklem, Nicholas Papaspyrou, Ira Parghi, David Schneiderman,
Arthur Ripstein, Ayelet Shachar, and three anonymous referees for the Journal of Political
Philosophy. I also thank Claire Hunter for editorial assistance and the Harvard Center for Ethics
and the Professions for ®nancial support. The usual disclaimer applies.
1Michael Walzer, What it Means to be an American (New York: Marsillo, 1996), p. 9.
conceals that it itself is ``highly discriminatory.''2The rejection of liberal
neutrality has led some to reject liberalism itself. For others, it has led to the
adoption of a perfectionist liberalism rooted in the value of autonomy.3
Other political philosophers have employed a different strategy. Instead of
arguing against neutrality as a matter of principle, they have turned to the
political practice of states that are widely accepted as liberal to determine
whether those liberal regimes are in fact neutral with respect to culture. These
critics argue that liberal democracies generally enshrine cultural rights, and
moreover, that they do so in two dramatically different ways. Liberal states draw
a fundamental distinction between two types of ethnocultural groups, ethnic
immigrants and national minorities. The former become citizens of liberal states
through individual and voluntary immigration, are geographically dispersed, and
lack the capacity to create the institutions of social, political and economic life.
Members of the latter are collectively incorporated, either involuntarily through
conquest, or voluntarily through treaties of cession or federation, are territorially
concentrated, and are more or less institutionally complete. These sociological
factors combine to create demands for different kinds of cultural rights within
each groupÐfor ethnic immigrants, rights that facilitate their inclusion in the
institutions of the broader society, and for national minorities, rights that enable
them to maintain a degree of institutional separateness. The policies of liberal
democracies, as a matter of fact, generally track these demands. The premise of
this enterprise is that a gap between the demands of liberal justice and the
practice of liberal democracies invites the questioning and recasting of the basic
principles of liberalism itself, including neutrality.4
In this article, I examine the theory of cultural rights that Will Kymlicka
develops in Multicultural Citizenship, because he seeks to straddle the divide
between these two strategies. On the one hand, Kymlicka offers an account
grounded in consent for the differential treatment accorded by liberal
democracies to ethnic immigrants and national minorities. On the other hand,
Kymlicka at times slips into what I term the argument from political sociology.
When Kymlicka argues in this mode, he claims that the practice of liberal
democracies with respect to ethnocultural groups tracks both the aspirations and
2Charles Taylor, ``The politics of recognition,'' Multiculturalism and the Politics of Recognition,
ed. A. Guttman (Princeton, N.J.: Princeton University Press, 1992), pp. 25±73 at p. 43. Taylor himself
does not seem to hold this position.
3Charles Taylor's work exempli®es the former result, whereas Joseph Raz's is an example of the
latter. Compare Charles Taylor, ``Atomism,'' Philosophy and the Human Sciences: Philosophical
Papers II (Cambridge University Press, 1985), pp. 187±210, and Joseph Raz, The Morality of Freedom
(Oxford: Clarendon Press, 1986).
4The leading ®gure here is Will Kymlicka, although important contributions have been made by
Charles Taylor. See Charles Taylor, ``Shared and divergent values,'' in Reconciling the Solitudes:
Essays on Canadian Federalism and Nationalism (Montreal & Kingston: McGill-Queen's University
Press, 1993) pp. 155±86; Will Kymlicka, Multicultural Citizenship (Oxford: Oxford University Press,
institutional capacities of those groups. The novel move in Kymlicka's account is
to argue that these facts of political sociology do important normative work.
There are two larger concerns driving my analysis. The ®rst is that accounts
like Kymlicka's ®nesse or ignore the tensions between the two different sets of
policies adopted by liberal democracies toward ethnocultural accommodation.
Policies which selectively promote the survival and ¯ourishing of the cultures of
national minorities, but not those of ethnic immigrants, often appear to be
discriminatory. Indeed, given that the presence of national minorities usually
predates that of ethnic immigrants, these preferences can easily be construed as
establishing different categories of citizenship on the basis of historical priority.
But in liberal democracies, differentiating among citizens simply on the basis of
prior membership, without additional justi®cation (for example, prior
sovereignty), appears to contradict the basic liberal commitment of giving
equal importance to the interests of every citizen. Moreover, there may be cases
where the demands of national minorities come into con¯ict with those of ethnic
immigrants.5In the face of these tensionsÐwhich will only grow as states
hitherto viewed as multinational become ethnically diverse through
immigrationÐa hierarchical distinction between different types of
ethnocultural groups will become at best a source of considerable
embarrassment, or at worst, a cause of moral outrage and social instability. If
liberal accounts like Kymlicka's cannot adequately justify these practices, they
may need to be reconsidered.6
The second is that Kymlicka's account of the correct stance of liberal
democracies toward ethnocultural difference furnishes an opportunity to re¯ect
5As a country with a large national minority, and the highest rate of immigration in the world,
Canada furnishes a number of vivid examples of these con¯icts. Efforts to de®ne Canada as a bilingual
and bicultural country in the 1960s, for example, provoked a strong backlash from ethnic immigrants
whose ancestry was neither English nor French. The response of the Canadian government was to
distinguish language from culture, describing Canada as both bilingual and multicultural.The
distinction between language and culture, though, is highly questionable, and has proved hard to
sustain. Another example of the con¯ict between the claims of national minorities and ethnic
immigrants can be found in Quebec. It is the policy of the provincial government to limit access to
publicly-funded primary and secondary education in English to persons whose parents were educated
in English in Canada. The distinction between old and new Canadians strikes many ethnic immigrants
as highly discriminatory. For an interesting discussion of some of these issues, see Joseph Carens,
``Immigration, political community, and the transformation of identity: Quebec's immigration politics
in critical perspective,'' Is Quebec Nationalism Just? Perspectives from Anglophone Canada, ed. J.
Carnes (Montreal & Kingston: McGill-Queen's University Press, 1995); Kenneth McRoberts,
Misconceiving Canada: The Struggle for National Unity (Toronto: Oxford University Press, 1997),
ch. 5.
6I assume, for the purposes of the paper, the validity of the cross-cutting distinction that Kymlicka
draws between two different kinds of claims that ethnocultural groups can makeÐclaims to
protection against economic and political decisions of the larger society (external protections), and
claims by groups to restrict the rights of their own members (internal restrictions). Kymlicka argues
that a liberal theory of cultural rights demands the former and rejects the latter. Although I largely
agree, my sense is that the distinction is much more complex than he suggests. Be that as it may, the
distinction is not relevant to my discussion because it applies equally to the claims of ethnic
immigrants and national minorities. As a consequence, it does not ®gure into the hierarchical
distinction that Kymlicka draws between the two types of groups.
on the place of political sociology in political philosophy. One of the notable
features of the argument in Multicultural Citizenship is the normative work done
by the political practices of liberal democracies toward ethnocultural minorities.
In this respect, Kymlicka's work is part of a methodological trend in
contemporary political thought. For example, Joseph Carens' recent work on
culture and citizenship, Stephen Macedo's discussions of civic education in
culturally diverse societies, and David Miller's new book on distributive justice
all weave detailed examinations of current social practices into their normative
accounts.7To a considerable extent, the contextual turn in contemporary
political philosophy is a reaction to the perceived abstractness of the hypothetical
choice positions employed by John Rawls, Ronald Dworkin and others. My
worry, though, is that in the shift to context, inadequate attention has been paid
to questions of methodologyÐthat is, what are the right and the wrong ways for
political sociology to ®gure into normative argument? Kymlicka's theory of
minority rights serves as a useful vehicle for exploring these issues.
Kymlicka's argument in Multicultural Citizenship begins with a sociological
account of ethnocultural difference in liberal states. Although many states are
multicultural, for Kymlicka, that general term encompasses two patterns of
cultural diversity, tied to two different sorts of ethnocultural groups. The ®rst
type of group are national minorities, previously self-governing peoples who have
somehow been incorporated into states in which they do not constitute the
majority group. Incorporation may have occurred involuntarily, for example,
through conquest, or voluntarily, through treaties of cession or federation.
Kymlicka gives a number of examples of national minoritiesÐthe Que
ÂcËois, the
Scots and the Catalans. The second type of group are ethnic immigrants, who
typically come to live in liberal democracies through individual or familial
immigration. Kymlicka focuses on instances of voluntary immigration. As a
consequence, he puts persons who have immigrated involuntarilyÐeither
because they were expelled (refugees) or abducted (former slaves) from their
homelandsÐto one side. Kymlicka notes that, until relatively recently,
immigration was largely a New World phenomenon. He accordingly has in
mind many of the ethnic groups that ®gure prominently in North American social
lifeÐJews, Chinese, Indians, and so on.
National minorities and ethnic immigrants differ along a number of
dimensions. The ®rst, as I have mentioned, is their manner of incorporationÐ
collectively and involuntarily or voluntarily; as opposed to, individually and
7Joseph Carens, Culture, Citizenship, and Community: A Contextual Exploration of Justice as
Evenhandedness (Oxford: Oxford University Press, 2000); Stephen Macedo, Diversity and Distrust:
Civic Education in a Multicultural Democracy (Cambridge, Mass.: Harvard University Press, 1999);
David Miller, Principles of Social Justice (Cambridge, Mass.: Harvard University Press, 1999).
voluntarily. Kymlicka also considers the two types of groups to differ in their
relationship to what he calls a societal culture, a central sociological category in
his theory. Kymlicka de®nes a societal culture as:
a culture which provides its members with meaningful ways of life across a full
range of human activities, including social, educational, religious, recreational, and
economic life, encompassing both the public and private spheres. These cultures
tend to be territorially concentrated and based on a shared language . . . and they
involve not just shared memories or values, but also common institutions and
Societal cultures, then, possess a number of features. The ®rst is their scope.
Societal culturesÐwhich Kymlicka calls a ``shared vocabulary of tradition and
convention''9Ðcover most areas of human activity, encompassing both the
public and the private spheres. This serves to distinguish societal cultures from
various subcultures of the innumerable social groups that exist in liberal societies.
A culture that is con®ned to just one or some types of activityÐfor example,
religious, social and/or educational lifeÐis in this respect incomplete, because it
does not offer options extending across all aspects of a person's life.
Another feature of societal cultures is their territorial concentration. At the
very least, this means that societal cultures exist in a de®ned geographic space;
human activity within that space is governed by the body of tradition and
convention that constitutes that culture. Moreover, Kymlicka would probably
claim that societal cultures are tied to place in a stronger senseÐthe shared
vocabulary of tradition and convention is framed around social narratives that
are geographically speci®c, for example, through references to concrete historical
events. A third feature of societal cultures is a common language. Language
would appear to be the medium through which traditions and conventions are
created and transmitted. Since cultures are goods that are created and
experienced by more than one person, a common basis for communication
would appear to be essential. Finally, societal cultures consist of practices that are
institutionally embodied. Traditions and conventions claim to be normative
standards for conduct, and thus demand that people lead their lives in accordance
with them. Moreover, those practices occur in institutions that re¯ect those
practices, and whose existence enables those practices to occur.
Kymlicka's principal point is that national minorities possess and are capable
of possessing societal cultures, whereas ethnic immigrants do not and cannot.
National minorities, for example, possess cultures which speak to most areas of
public and private life. They typically occupied a speci®c territory or homeland at
the time of incorporation, and despite the existence of rights of internal mobility
within nation-states, continue to predominate in certain regions. National
minority groups also share a common language which is usually distinct from
8Kymlicka, Multicultural Citizenship, p. 76.
that spoken in the rest of the country. As well, they are institutionally complete,
in that it is possible for them to engage in culturally meaningful practices in their
own institutions across a broad swath of social, economic and political life.
Ethnic immigrants, by contrast, presumably possessed their own societal cultures
in their countries of origin. However, the brute fact of immigration has severed
their links with those sources of tradition and convention. Crucially, it is
sociologically impossible or infeasible for them to recreate societal cultures based
on their traditions and languages in their new home. As Kymlicka states, ethnic
bring their language and historical narratives with them. But they have left behind
the set of institutionalized practices, conducted in their mother tongue, which
actually provided culturally signi®cant ways of life to people in their original
homeland. They bring with them `a shared vocabulary of tradition and convention',
but they have uprooted themselves from the social practices which this vocabulary
originally referred to and made sense of.10
The inability of ethnic immigrants to sustain a complete set of separate
institutions is of crucial importance. But ethnic immigrants are also incapable of
maintaining distinct societal cultures because they lack a distinct homeland
within their new state.
Why does this distinction matter? Kymlicka's theory of cultural rights is a
liberal theory, and liberalism, on his account, is de®ned by its commitment to
individual autonomy. In particular, individuals possess two fundamental
interestsÐthe freedom to lead their lives from the inside according to values
that they endorse, without fear of punishment or discrimination, and the freedom
to question those beliefs and examine them in light of whatever information,
arguments and examples that our culture provides. These two interests justify the
traditional liberal freedomsÐthe ®rst, rights to privacy and security of the
person, and the second, rights to expression and freedom of association. But
Kymlicka's distinctive contribution is the proposition that societal cultures are
another crucial precondition of individual autonomy. Societal cultures are
important because they present the options from which we choose or construct
our conceptions of the good. Moreover, societal cultures attach value to different
options, so that, ``in the ®rst instance,''11 our beliefs about the value of ways of
life are culturally conditioned.
The importance of societal cultures to individual freedom is a normative
account, suggesting that membership in societal cultures is a fundamental
interestÐa primary good in the Rawlsian sense. However, although Kymlicka is
not clear on this point, he seems to suggest that people in fact realize the
importance of societal cultures, and that access to those cultures will therefore be
the subject of political demands. But the dramatically different capacities of
10Ibid. at p. 77.
11Ibid. at p. 83.
national minorities and ethnic immigrants to develop and sustain their own
societal cultures lead them to assert their need for societal cultures in two very
different ways. Unable to recreate their own societal cultures, ethnic immigrants
demand access to the societal culture of the dominant society, through their
inclusion in common institutions. However, institutional integration is not
synonymous with assimilation. Ethnic immigrants demand the right to express
their cultural particularity in family life and in voluntary associations. Moreover,
they demand that public institutions not privilege the ways of life of some ethnic
groups over others, and this may at times involve group-speci®c accommodations.
National minorities, by contrast, are capable of maintaining distinct societal
cultures and demand the means to do so. In practical terms, they desire a degree of
institutional separateness. Indeed, these institutions are comprehensive in scope,
so that national minorities are in effect ``asking to set up a parallel society.''12
Political practice in multicultural states generally tracks these demands. Ethnic
immigrants hold ``polyethnic rights'' which facilitate their membership in
common institutions. Although traditionally con®ned to guarantees of non-
discrimination, polyethnic rights now extend to anti-racism policies, the reform
of educational curricula to re¯ect the histories and contributions of all immigrant
groups, public funding for cultural practices, and even exemptions from laws that
disadvantage ethnic immigrants because of culturally speci®c practices. National
minorities, by contrast, hold ``self-government rights'' which enable them to erect
separate institutions. Self-government rights might take the form of federal
arrangements that provide that members of a national minority constitute the
majority in a federal sub-unit, or the establishment of reserved lands in the case of
small indigenous populations. Moreover, these self-governing political entities
possess jurisdiction over subject matters relevant to cultural survival.
Kymlicka provides a rich sociological account of the politics of ethnocultural
difference in liberal states, which in no small part accounts for Multicultural
Citizenship's in¯uence among social scientists and legal scholars interested in
questions of cultural difference. However, Kymlicka's goal is to go beyond mere
description, and to provide a normative argument that seeks to justify this
pattern of political practice. Upon closer examination, though, Kymlicka offers
not one argument, but two. The ®rst is an argument for the distinction between
ethnic immigrants and national minorities that turns on the manner in which
different ethnocultural groups were incorporated into the political communities
of which they are a part. Call this the argument from consent. According to
Kymlicka, ethnic immigrants have waived the right to live in accordance with
12Ibid. at p. 15.
their own cultures through the decision to immigrate to a society in which they
knew they would constitute a minority, whereas national minorities, again
because of the manner of their incorporation, have not waived that right.
The second argument is what I will call the argument from political sociology.
Faced with dif®culties with the argument from consent, Kymlicka at times
emphasizes the fact that the practice of liberal democracies with respect to
ethnocultural groups tracks both the aspirations and institutional capacities of
those groups. When arguing in this mode, he takes the claims of ethnocultural
groups to be exhaustive of the normative arguments worth considering for
cultural rights, and as a consequence does not give much weight to claims that
are absent from political discourse. As well, he regards the viability of societal
cultures as a crucial variable in an account of just institutions, in that justice does
not demand that we create institutions that cannot survive. The accuracy of this
account matters, because of the structure of this normative argument. More
fundamentally, the normative work done by political sociology in this strand of
Kymlicka's argument is highly questionable.
A. The Argument from Consent
Kymlicka begins the argument from consent by examining the importance of
cultural membership to the liberal self, through explaining the value of cultural
membership to individual autonomy.13 As I mentioned earlier, Kymlicka's
principal point is fairly uncontroversialÐthat cultures can support individual
autonomy because they present options from which we construct our conceptions
of the good, and render those options meaningful by assigning them some
value.14 The controversial move in Kymlicka's argument is to claim that persons
have a fundamental interest not only in cultural membership per se, but in
membership in their own cultures. That is, membership in aculture is not
suf®cient to secure individual autonomy. The obvious objection to the
importance given by Kymlicka to membership in one's own culture is that it
does not yield the distinction that he draws between different categories of
13This is an argument that he ®rst presented in Liberalism, Community, and Culture (Oxford:
Clarendon Press, 1989).
14However, as Andrew Kernohan has argued in Liberalism, Equality and Cultural Oppression
(Cambridge: Cambridge University Press, 1998), a culture that devalues or debases certain identities
can undermine individual autonomy.
Kymlicka is unclear at times on whether it is open to persons to reject those options, or even to
reject the value that has been assigned to them by their culture. Thus, at one point, he states that
``freedom involves making choices amongst various options, and our societal culture ...provides
these options'' (83). However, an account of our fundamental interests that precluded the re-
evaluation, or even the rejection, of culturally endorsed ways of life would not be recognizably liberal.
As John Tomasi has noted (in ``Kymlicka, liberalism, and respect for cultural minorities,'' Ethics, 105
(1996), 580±603 at p. 591), liberalism has historically been identi®ed with a rejection of the ancien
regime, and the endorsement of the possibility of personal experimentation free from the strictures of
social convention. Kymlicka's point must be that cultural narratives are the inescapable starting points
for our deliberations about conceptions of the good, although they do not de®nitively de®ne the set of
alternatives from which we must choose.
ethnocultural groups. Assuming such a fundamental interest to exist, by its own
terms it should apply equally to all persons, regardless of the particular type of
ethnocultural group to which they belong. But Kymlicka argues that only
national minorities, and not ethnic immigrants, are entitled to access to their own
cultures. Even worse, Kymlicka's distinction appears to embody two mutually
inconsistent views on the value of cultural membership. The treatment Kymlicka
accords to national minorities certainly re¯ects the importance that he ascribes to
membership in one's own culture. But the policy he endorses for ethnic
immigrantsÐaccess to the dominant culture of the society to which they
emigrate, albeit on fair termsÐsuggests that what matters is not membership in
any particular culture, including one's own culture, but membership in aculture.
Note that one need not be a cosmopolitan, and adopt Jeremy Waldron's
skepticism of the existence of distinct cultures, or of the importance of
membership in one culture, to hold the latter view.15 Rather, even if we believe
cultures to be distinct, and cultural membership to be a singular phenomenon,
the special importance of membership in one's own culture has not been
established. Indeed, the fact that ethnic immigrants often thrive in their new
environments (witness America prior to the 1950s), albeit after initial dif®culties
of transition, is reason to doubt that an ongoing membership in one's original
culture is a fundamental interest.
This contradiction over the value of one's own culture infects Kymlicka's
scheme of cultural rights. If one's own culture is unimportant, it is not clear how
Kymlicka can justify polyethnic rights for ethnic immigrants.16 Recall that these
rights include, inter alia, group-speci®c accommodations that facilitate
culturally-speci®c practices. Exemptions from Sunday-closing laws and the
alteration of dress codes are two oft-cited examples. But if justice only demands
that persons be offered membership in the dominant culture, it is not clear why
these accommodations are required at all. Kymlicka might defend these
accommodations on the ground that they contribute to the internal
heterogeneity of the dominant culture. But this argument fails to make sense of
polyethnic rights, because those rights only apply to culturally mandated
practices that are group-speci®c, not to any and all ways of life. Thus, the
promotion of preference diversity is not a suf®cient rationale for polyethnic
rights. Kymlicka's support of polyethnic rights, then, can only be explained if
one's own culture does matter, even to ethnic immigrants. But if it does, it is
unclear how he can stop at polyethnic rights, and not proceed to self-government.
In order to make his distinction workable, Kymlicka has two options. He
could conceivably argue that the interest of ethnic immigrants in cultural
membership is somehow different from that of national minorities. But this is
15See generally Jeremy Waldron, ``Minority cultures and the cosmopolitan alternative,'' University
of Michigan Journal of Law Reform, 25 (1992), 751±93.
16Joseph Carens has made a similar point. Joseph Carens, ``Liberalism and culture,''
Constellations, 4 (1997), 35±47 at p. 44.
tantamount to saying that one's own culture matters less for ethnic immigrants
than for national minorities. Kymlicka rightly does not take this position,
because it is conceptually impossible. Fundamental interestsÐtheories of the
goodÐare by hypothesis interests which everybody has. To begin to draw
distinctions between persons with respect to fundamental interests violates a
cardinal premise of modern political philosophyÐnamely, as Kymlicka has
written elsewhere, ``that the interests of each member of the community matter,
and matter equally.''17 Kymlicka therefore argues not that ethnic immigrants
possess a lesser interest in cultural membership, but rather that they have waived
the right to live in accordance with their own cultures through the decision to
immigrate to a society in which they knew they would constitute a minority. As
he says in a crucial passage:
nor is it unjust that the American government (and other Western democracies) have
decided not to give immigrants the legal status and resources needed to become
national minorities. After all, most immigrants (as distinct from refugees) have
chosen to leave their own culture. They have uprooted themselves, and they know
when they have come that their success, and that of their children, depends on
integrating into the institutions of English-speaking society.18
In short, immigrants have consented to a less extensive scheme of cultural rights.
Although super®cially attractiveÐand re¯ective of political discourse of liberal
democraciesÐthis argument does not withstand scrutiny. Consider ®rst the
normative signi®cance of the decision to immigrate. Kymlicka argues that
immigration amounts to a form of consentÐa choice for which ethnic
immigrants can be held responsible. But in the liberal tradition, persons are
only held responsible for states of affairs that arise from choices that are made
freely. What is the liberal conception of freedom? At the very least, freedom
entails freedom from physical coercion. For this reason, Kymlicka puts former
slaves and refugees to one side because their immigration was involuntary, either
because they were physically abducted or ¯ed in the face of physical harm. These
are genuinely hard cases for any liberal theory of minority rights. However, for
liberal egalitarians, like Rawls and Dworkin, choices made against the
background of material inequality also lack moral force because they are not
freely taken. This creates two problems for Kymlicka. First, Kymlicka has
elsewhere con®rmed his commitment to the egalitarian version of liberalism. We
can assume that he wishes his theory of minority rights to cohere with his views
on distributive justice. Second, most immigrants, or at least a very signi®cant
proportion, migrate in search of economic opportunity. Although immigration is
often prompted by other reasonsÐfamily reuni®cation, personal relationshipsÐ
economic motivations predominate. The key point here is that economic
opportunities are unevenly distributed on a global scale, owing to factors that
17Will Kymlicka, Contemporary Political Philosophy (Oxford: Oxford University Press, 1990) at
p. 4.
18Kymlicka, Multicultural Citizenship, pp. 95±6.
egalitarians would regard as morally arbitrary.19 This means that in the majority
of cases the decision to immigrate is not free, such that voluntary immigration
cannot validate distinctions drawn between different categories of ethnocultural
Kymlicka is obviously bothered by this problem. As he acknowledges, ``the
line between involuntary refugees and voluntary immigrants is dif®cult to draw,
especially in a world with massive injustice in the international distribution of
resources.''20 His response, though, is not to re-examine his scheme of minority
rights, but to argue that the correct solution is to remedy the unjust international
distribution of resources. But it is inconsistent for Kymlicka to not address this
problem, because he refuses to apply the consent argument with equal vigour to
national minorities who joined the larger political community against a
background of material inequality.
For the consent argument to do the work Kymlicka demands of it, it should
explain not only why ethnic immigrants do not have rights to self-government,
but also why national minorities do have those rights. National minorities who
have been incorporated involuntarily clearly can invoke this right because they
have not consented to join the larger political community. But what about
national minorities who have joined voluntarily? As Kymlicka suggests, the terms
of federation or cession, either contained in constitutions or treaties, are of
central importance in these sorts of cases. However, Kymlicka is quick to note
the dif®culties with this argument.21 The decisive problem is that these treaties or
the terms of federation vary enormously in their speci®city and content, ranging
from extensive rights of self-government to almost none at all. As a consequence,
for Kymlicka, they produce arbitrary distinctions among national minorities, a
result that he does not accept. But how can Kymlicka make this argument? His
only option is to acknowledge the source of these differences: the varying degree
of relative inequality of bargaining power between minority and majority groups.
Inequality in bargaining power, in turn, is a function in part of material
inequality (historically, this was particularly true for indigenous peoples in North
America). But if Kymlicka is willing to accept that choices made against a
background of material inequality do not carry moral force here, then how can
he insist that they should count in the case of ethnic immigrants?
The argument from consent is also problematic because it does not produce a
justi®cation for polyethnic rights. As Kymlicka frames it, an argument from
consent does not in itself specify the terms that are consented to. The relevant
terms are those that the parties agree to, either through a process of negotiation,
or more likely, a process of unilateral offer and acceptance. Those
19See the essays in Global Justice, ed. I. Shapiro and L. Brilmayer (New York: New York University
Press, 1999).
20Kymlicka, Multicultural Citizenship, p. 99.
21In this respect, he follows the work of Patrick Macklem in ``Distributing sovereignty: Indian
nations and equality of peoples,'' Stanford Law Review, 45 (1993), 1311±67.
understandings may be explicit or implicit. Only if polyethnic rights are part of
those understandings can ethnic immigrants then assert them. Kymlicka, though,
does not justify polyethnic rights on this basis, with good reason. Historically,
immigration policies have been assimilationist, demanding that immigrants
accept not only principles of political justice, but also the lifestyle preferences of
the majority society. By de®nition, this approach rules out polyethnic rights.
What other arguments are available to him? Kymlicka argues that those rights
further the project of integrating ethnic immigrants into the dominant society.
Alternatively, he also grounds those rights in some notion of fairness. But note
that these justi®cations have little to do with consent. Rather, they appeal to
some other value.
If the consent argument fails, then what are we left with? We are pushed back
to the claim that the interests of ethnic immigrants and national minorities in
their own cultures are somehow different. But Kymlicka cannot make that
argument because it offends the basic notion of moral equality that lies at the
foundation of any theory of justice. In the end, then, Kymlicka has not offered a
good justi®cation for the hierarchical distinction between ethnic immigrants and
national minorities drawn by most liberal democracies. Although I cannot
develop the point here, Kymlicka's dif®culty in justifying that distinction suggests
that culture cannot do all of the work he wants it to do in a liberal theory of
B. The Argument from Political Sociology
The argument from consent does not justify the distinction between the treatment
accorded to ethnic immigrants and national minorities. Faced with this dif®culty,
Kymlicka at times retreats into what I call the argument from political sociology.
Perhaps the most interesting methodological feature of Multicultural Citizenship
is the normative work done by political sociology. In this section, I argue that
Kymlicka's use of political sociology in normative argument is conceptually
¯awed. But what does political sociology mean? I take that term to encompass at
least three different types of empirical or descriptive claims. First, there are the
aspirations or expectations of members of ethnocultural groups. The demand of
national minorities for institutional separateness is an illustrative example.
Second, there are questions of viability or possibility. For my purposes, the
relevant issue is the capacity of ethnocultural groups to embody their cultural
practices institutionally. Finally, there are current political practices. Into this
category I place the institutional arrangements that currently exist in liberal
democracies to accommodate ethnocultural difference.
There are four different ways in which these varieties of political sociological
facts can ®gure into normative political theory, three of which are correct, but
one of which represents a serious mistake.
First, the expectations or aspirations of ethnocultural groups can ®gure into an
analysis of political discourse. The idea is to articulate the claims of justice
implicit in the political claims of social groups. The value of this exercise is
empirical, because it reveals what kinds of normative arguments members of
these groups make. In this respect, Kymlicka is not alone. Applied to problems of
cultural difference, this approach was pioneered by Iris Marion Young, and has
been used with great effectiveness most recently by Melissa Williams.22 But there
are two cautions to keep in mind when analyzing political discourse in search of
normative arguments. The ®rst is that these arguments merely represent claims of
justice. Even after these claims have been carefully described, from a normative
perspective, the important question remains unansweredÐthat is, whether those
claims of justice should count. Used in this way, political sociology sets the stage
for normative analysis, but is not a substitute for it (as both Young and Williams
acknowledge). The second caution is that political discourse must not be taken as
an exclusive source of normative arguments, because this would unduly narrow
the scope of normative inquiry. There may be good arguments that have not been
advanced in political discourse. As well, the claims of social groups are often
shaped by the very institutions and practices that those claims are used to
scrutinize. The preferences of members of social groups often adapt to existing
social arrangements, so that their expectations and aspirations approximate or
rationalize the status quo. If this is true, then it is question-begging to use those
preferences as a benchmark against which to measure the justice of institutions,
because they are the products of those very institutions themselves. Indeed, it may
be perverse, if those institutions are designed to generate certain sorts of political
preferences. Stephen Macedo argues, for example, that one of the central aims of
a liberal constitutional order is to transform religious views that are incompatible
with the principles of liberal democracy.23
Second, political sociology, in the form of current political practices, can play
an important role in the process of re¯ective equilibrium.24 Current political
practices may represent our considered judgments or the ®xed points in our
search for abstract principles of justice. The goal of re¯ective equilibrium is to
ensure that those principles accord with and explain our considered judgments.
We arrive at this equilibrium through the following kind of process. We begin by
proposing a set of abstract principles of justice, which we adopt provisionally,
and compare them against our considered judgments. If there is a disjunction
between political practices that seem right and those principles that we have
provisionally adopted, then we have a choice. Either we reconsider and reframe
22Iris Marion Young, Justice and the Politics of Difference (Princeton: Princeton University Press,
1990); Melissa S. Williams, Voice, Trust, and Memory: Marginalized Groups and the Failings of
Liberal Representation (Princeton, N.J.: Princeton University Press, 1998).
23Stephen Macedo, ``Transformative constitutionalism and the case of religion: defending the
moderate hegemony of liberalism,'' Political Theory, 26 (1998), 56±80.
24John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), pp. 19±20
and 47±50.
the principles of justice, and again measure them against those ®xed points, or we
revise our considered judgments to accord with fundamental principle. The
important point to note is that the process of revision extends both ways, so that
we must be willing to revise political practices that we initially take as givens.
Third, empirical claims about viability or possibility can serve as the empirical
premises of normative arguments that have the following structure. Imagine
some social goal or principle, whose importance has been established
independently through moral reasoning. Our objective is to secure that goal
here and now, through whatever concrete measures are appropriate. An
argument for those measures is a normative argument of what is to be done.
But that argument will also have an important empirical component, relevant to
the question of means. For example, we must consider what is practical and
feasible given our understanding of human behavior, such as how people respond
to certain incentives. As well, questions of political culture and comparative
institutional advantage are important. These sorts of arguments, combining
normative goals with empirical insights with respect to means, ®gure prominently
in the design of just institutions. The discussions in the Federalist Papers of the
obstacles that factions pose to the effective functioning of democracies is a
famous example. Into this category also fall arguments relating to the
implementation of moral norms through positive law. The caution here is that
we must be skeptical of empirical premises that appear to be ®xed givens. This is
especially so when claims of viability or possibility are in fact a function of the
very institutions that we want to reform. This is another manifestation of the
problem of recursiveness that presented itself with respect to aspirations or
expectations. Here, existing institutions unduly limit our sense of the possible.
Finally, political sociology could ®gure into normative political theory in a
crude and direct manner. The argument would be that facts of political sociology
of all sorts carry normative force simply because they exist. Call this the claim of
per se normativity. It typically applies to current political institutions and
expectations or aspirations. Whatever the particular fact, this claim comes in a
weak and a strong version. The weak claim is that the mere existence of a fact
counts, but that it is not the only consideration that matters. Other factors may
outweigh it. The strong claim is that a sociological fact is so weighty that it is
right merely because it is exists. The effect of the strong claim on normative
reasoning can be understood as the replacement of ®xed points with immovable
ones in the process of re¯ective equilibrium. If this is the case, then the claim of
per se normativity transforms the process of re¯ective equilibrium, such that the
principles of justice merely rationalize existing facts, instead of striving to justify
them. The fundamental problem of the claim of per se normativity, in either its
weak or its strong form, is that it does not provide an argument for why
sociological facts should themselves carry normative force. It derives an ought
from an is. As Jeremy Waldron has reminded us, the liberal tradition has been
characterized from the outset by a refusal to accept tradition and established
social practice as the basis of political order; rather, liberalism is synonymous
with a demand for justi®cation of the social world.25 For this reason, claims of
per se normativity typically rely on some other value, which then replaces those
facts as the primary source of the normative claim.
Kymlicka fares quite poorly with respect to the manner in which he
incorporates political sociology into his normative argument. Consider ®rst his
treatment of the expectations or aspirations of members of ethnocultural
groups. One of the striking features of Kymlicka's analysis is his focus on what
ethnic immigrants and national minorities demand of liberal states. His
principal error is to regard these demands as largely exhaustive of the
arguments worth considering. Thus, when presented with arguments that vary
from the ones he claims are made in actual political discourse, he downplays
them. This arises particularly with respect to the question of institutional
separateness for ethnic immigrants. He does not take this claim seriously
because, in his view, it is unrepresentative of political discourse. The following
passage is characteristic:
there is little evidence that immigrants are seeking national rights, rather than
polyethnic rights. Some commentators interpreted the `ethnic revival' in the United
States in the 1970s as a repudiation of integration into the mainstream society. But
as I noted [earlier] . . . this is dubious. . . . The ethnic revival, in other words, involved
a revision in the terms of integration, not a rejection of integration.26
I do not deny that there is value to articulating the claims of justice implicit in
the demands of social groups. Indeed, as Kymlicka rightly notes, there is a special
need to characterize those arguments fairly and accurately, because they are so
often misunderstood or misrepresented in debates over cultural difference. But
the fact that ethnic immigrants do not make certain arguments does not mean
that they should not be making them, or that more ambitious claims are not
Narrowing the scope of normative analysis to existing claims is also question-
begging, because those aspirations or expectations are to a considerable extent
the function of background institutional conditions. This is the lesson of Benedict
Anderson's famous treatment of nationalism, which detailed the deliberate way
in which a national consciousness was created in many modern nation-states
with no historical antecedents.27 Kymlicka himself acknowledges this fact when
he states that ``differences in the mode of incorporation affect the nature of
minority groups, and the sort of relationship they desire with the larger
society.''28 But if expectations or aspirations are a function of the very
25Jeremy Waldron, ``Theoretical foundations of liberalism,'' Philosophical Quarterly, 37 (1987)
26Kymlicka, Multicultural Citizenship, pp. 97±8; also see p. 67, para. 2.
27Benedict Anderson, Imagined Communities: Re¯ections on the Origin and Spread of
Nationalism (London: Verso, 1991).
28Kymlicka, Multicultural Citizenship, p. 10; also see p. 11, para. 2.
institutions we wish to scrutinize, then relying on them is problematic because
they provide no external standard for evaluation. Even worse, relying on
expectations creates perverse incentives for states to mold preferences so as to
insulate institutions from political criticism.
Indeed, Kymlicka goes even further at certain points, and ascribes normative
signi®cance to the demands of ethnocultural groups simply because those
demands exist. This is clearest in a later essay in which Kymlicka examines the
implications of his theory for Hispanics in the United States.29 The sociological
status of Hispanics is ambiguous, because this group encompasses both national
minorities (such as Puerto Ricans) and ethnic immigrants (such as Central
Americans), each with the aspirations that one would expectÐthe former
demanding institutional separateness, the latter demanding integration on fair
terms. What is striking is the normative implication that Kymlicka draws from
these facts of political sociology. He argues that the ``aspirations of Hispanic
groups are simply too divergent to create a single form of multiculturalism
appropriate for all Hispanic groups.''30 This is an argument that multiculturalism
should be tailored to suit the aspirations and expectations of ethnocultural
groups. But this is no more than the derivation of an ought from an is.
Related problems occur in Kymlicka's treatment of institutional completeness.
Kymlicka differentiates the legitimate claims of ethnic immigrants and national
minorities on the basis of their capacities to sustain societal cultures. Thus,
national minorities have a right to self-government (rights to institutional
separateness) because they possess the requisite institutional capacity, whereas
ethnic immigrants only possess polyethnic rights (rights to institutional
integration) because they lack this capacity. This is an account of the design of
just institutions of the style that I described earlier. Kymlicka's mistake is to take
the institutional capacities of ethnocultural groups as factual givens, and to base
his account of just institutions around them. In reality, these institutional facts
are highly contingent. They are a function of existing distributions of resources
and political power. Moreover, there is a recursive relationship between
institutional capacities and the rights that groups currently possess, making it
unfair to regard them as decisive considerations in arguments for why or why not
those current arrangements should be altered. Institutions are a function of
rights, not the other way around. Additionally, as with expectations or
aspirations, incorporating institutional capacities incorrectly into normative
argument creates perverse incentives for states.31 Carens argues in this vein that a
29Will Kymlicka, ``Ethnic associations and democratic citizenship,'' Freedom of Association, ed. A.
Guttman (Princeton, N.J.: Princeton University Press, 1998) pp. 177±213 at pp. 208±10.
30Ibid. at p. 210 (emphasis mine).
31Ironically, Kymlicka employs this argument himself, in rejecting the lack of the capacity to create
societal cultures as a reason not to grant rights to self-government of national minorities. He draws a
distinction between the ``potentiality of societal cultures'' (100) and current or present capacity. But if
Kymlicka is willing to relax his empirical assumptions for national minorities, why does he refuse to
do so for ethnic immigrants?
danger of a ``contextual approach'' to political philosophy ``is that it will fail to
recognize injustices that are pervasive and deeply embedded in practice.''32
Kymlicka's tendency to take the institutional capacities of ethnocultural
groups as ®xed givens is worth re¯ecting on because it is premised on a mistaken
view of the relationship between sovereignty and territory. His assumption
throughout is that sovereignty requires a de®ned territory over which it may be
exercised. Kymlicka is in distinguished company here; this assumption underlies
traditional conceptions of sovereignty and political community. It follows that
only ethnocultural groups that are territorially concentrated can realistically
exercise rights to self-government and, in Kymlicka's view, that only those
groups possess those rights. Accordingly, national minorities, which by de®nition
are territorially concentrated, possess rights to self-government, whereas ethnic
immigrants, who are territorially dispersed, do not.
Kymlicka's mistake is to give short shrift to institutional arrangements that
would allow communities to govern themselves without a de®ned territory, and
hence challenge the claim that sovereignty is inextricably linked to territory.33
Here, the facts of political sociology, particularly those drawn from comparative
experience, far from supporting Kymlicka's arguments, can instead serve as a
useful fund of examples that both expose and destabilize his institutional
assumptions.34 I have already alluded to immigrant communities whose
institutions serve important quasi-public functions, without the need for a
de®ned geographic base of the kind that public governments (nations, provinces,
municipalities) possess. Moreover, other institutional arrangements that
explicitly detach the capacity for self-government from territory are possible.
For example, in Canada, it has been suggested that aboriginal peoples who live in
urban areas, and who are interspersed with the general population and hence
lack the territorial basis for traditional forms of self-government, could
nevertheless govern themselves with respect to social services, child welfare,
housing, and primary and secondary education through a ``community of
interest'' model.35 Under this model, a local aboriginal council would be vested
with legal powers over these matters and the power to tax; jurisdiction over
aboriginal peoples would be voluntary.
32Carens, Culture, Citizenship, and Community,p.4.
33To be fair, Kymlicka does discuss the Millet system of the Ottoman empire (at pp. 156±8),
although in connection with the distinction between external protections and internal restrictions. As
well, in his most recent work, Kymlicka seems to have recognized the variety of institutional
mechanisms available to accommodate ethnocultural difference within liberal democracies (W.
Kymlicka and W. Norman, ``Citizenship in culturally diverse societies: issues, contexts, concepts,''
Citizenship in Diverse Societies, ed. W. Kymlicka & W. Norman (Oxford: Oxford University Press,
2000) pp. 1±41 at pp. 24±30.
34Carens, Culture, Citizenship, and Community, pp. 4±5. Elsewhere, I have argued that
comparative case-law can serve the same function in constitutional adjudication, in ``Globalization
in search of justi®cation: toward a theory of comparative constitutional interpretation,'' Indiana Law
Journal, 74 (1999), 819±92.
35Royal Commission on Aboriginal Peoples, Report of the Royal Commission on Aboriginal
Peoples (Ottawa: Minister of Supply and Services Canada, 1996), vol. 2, pp. 274±8.
More radical are proposals for non-territorial forms of federalism, whereby a
community (for example, a religious minority, or an indigenous people) would
possess institutions with the power to promulgate laws that bind its members
throughout a nation-state. This approach to federalism imagines political
communities in non-territorial terms, and, depending on how membership is
determined, could in principle operate independently of the consent of those
governed by a community's laws. The kind of legal pluralism contemplated by
non-territorial federalism was and continues to be in operation, both prior and
subsequent to independence, in the former colonies of British Africa.36 Some
contemporary proposals suggest that the jurisdiction of non-territorial political
communities should be equivalent to the jurisdiction of territorially based federal
sub-units;37 others suggest that the jurisdiction of non-territorial political
communities should be rather limited. A leading example of the latter is the
system of Muslim personal law in India, which is limited in scope to the
regulation of marriage, divorce, child custody, and inheritance.38
I hasten to add that by listing these institutional options, I do not endorse them.
Indeed, as Ayelet Shachar has demonstrated, systems of religious personal law have
systemically worked to the disadvantage of women, and for that reason should be
accommodated within the family law regimes of liberal democracies only with
extreme caution.39 But these examples do refute the institutional assumptions
underlying Kymlicka's claims of viability or possibility, which are key empirical
premises in his argument for why ethnic immigrants lack rights to self-government.
The availability of institutional options that divorce sovereignty from territory
demands that Kymlicka squarely face the normative justi®cations for the distinction
between ethnic immigrants and national minorities. It also illustrates the
contingency of the predominant institutional options in liberal democracies.
What unites Kymlicka's mistakes is a tendency at times to give normative
signi®cance to contingent facts of political sociology.40 What is driving Kymlicka's
analysis? His concern, I think, is less with justice than with the political stability of
liberal democracies that are ethnically diverse. This becomes clear in the latter part
of his book, where Kymlicka explicitly addresses the question of whether
polyethnic rights and self-government rights undermine or support social unity.
His conclusion is rather revealing. On the one hand, polyethnic rights support social
36T. W. Bennett, ``Con¯ict of lawsÐthe application of customary law and the common law in
Zimbabwe,'' International and Comparative Law Quarterly, 30 (1981), 59±103.
37For a proposal that a non-territorial ``Aboriginal Peoples Province'' be created in Canada, see
David J. Elkins, Beyond Sovereignty: Territorial and Political Economy in the Twenty-First Century
(Toronto: University of Toronto Press, 1995) ch. 5.
38David Pearl and Werner Menski, Muslim Family Law, 3rd edn (London: Sweet & Maxwell,
1998), ch. 2.
39Ayelet Shachar, ``Group identity and women's rights in family law: the perils of multicultural
accommodation,'' Journal of Political Philosophy, 5 (1998), 285±305, and ``On citizenship and
multicultural vulnerability,'' Political Theory, 38 (2000), 64±89.
40Yael Tamir has made a similar point, in ``Theoretical dif®culties in the study of nationalism,''
Canadian Journal of Philosophy, 22 Supp. (1996), 65±92 at p. 79.
unity because they facilitate the participation of ethnic immigrants in societal
cultures. Self-government rights, however, have exactly the opposite effect because
they acknowledge the existence of multiple political communities within a nation-
state, thereby raising the question of which community is the object of citizens'
primary allegiance. In this light, limiting the range of groups that can legitimately
claim rights to institutional separateness is of pressing practical concern.
Fortunately, the existing practice of liberal states offers a sharp categorical
distinction between ethnic immigrants and national minorities.41 Kymlicka's goal is
to justify that practice. One would have expected Kymlicka to approach thistask by
con®ning himself to the argument from consent. Instead, he slips into the argument
from political sociology, which is methodologically suspect.
In this context, Kymlicka's emphasis on the normative claims implicit in
political discourse is extremely understandable. His goal is to address the concern
that immigrant multiculturalism is inherently separatist, and is the ®rst step down
the road toward minority nationalism. Kymlicka's response is to examine the
actual implications of public policies associated with immigrant
multiculturalism, as well as the demands of ethnocultural groups. Neither
these policies nor the demands of ethnic immigrants seek to transform ethnic
immigrants into national minorities, with their attendant demands for
institutional separateness and rights of self-governance. Accordingly, Kymlicka
dismisses this fear as ``a red herring without any basis in reality.''42 Kymlicka
here is a participant in a political debate over the logic and consequences of
existing multiculturalism policies. His emphasis on the concrete facts of political
sociology is valuable, because debates on the accommodation of cultural
difference in liberal democracies are often characterized by more heat than light
over what multiculturalism currently means in practice. However, Kymlicka's
sociological groundedness is purchased at a normative price, because it limits his
access to a critical stance from which to scrutinize existing political practices.
But even if Kymlicka is correct to use political sociology in the manner that he
does, his descriptive sociology leaves a great deal to be desired and creates
41In this connection, the following admission is remarkably candid: ``there is a more realpolitik
reason for emphasizing the sharp distinction between immigrants and national minorities, rather than
viewing them as simply two poles on a ¯uid and amorphous continuum. One of the most common and
in¯uential objections to minority rights for any group is that it would lead us down a `slippery slope'
in which more and more groups would demand more and more rights, leading to the eventual
disintegration of society. . . . I believe that progress on the rights of minorities will only come about if
we effectively tackle this `slippery slope' view. And to do so, we need to show that ethnocultural
groups do not form a ¯uid continuum, in which each group has in®nitely ¯exible needs and
aspirations, but rather that there are deep and relatively stable differences between various kinds of
ethnocultural groups.'' Will Kymlicka, ``Do we need a liberal theory of minority rights? Reply to
Carens, Young, Parekh and Forst,'' Constellations, 4 (1997), 72±87 at p. 80.
42Kymlicka, ``Ethnic associations and democratic citizenship,'' p. 199.
signi®cant problems for him. Consider ®rst the issue of viability. Kymlicka's
argument is as follows:
P1Individuals have a fundamental interest in cultural membership, because
cultures provide the context of choice that is a necessary condition for the
exercise of individual autonomy.
P2Only a societal culture can serve as a suitable context of choice.
C1Individuals have a fundamental interest in membership in societal cultures.
P3Ethnocultural groups differ in their capacities to build and sustain societal
cultures in states in which they do not constitute the majority. In general,
national minorities possess this capacity, whereas ethnic immigrants do not.
C2National minorities can meet their fundamental need for cultural membership
by maintaining their own societal cultures, while ethnic immigrants cannot.
There are a number of dif®culties with this argument. Let us assume P1to be true,
and focus on P2and P3. What is Kymlicka's argument for P2? Kymlicka never
fully explains why only societal cultures can provide a meaningful context of
choice. To discover what his argument might be, one place to turn to is the
distinction he draws between societal cultures and the other species of culture
that are somehow de®cient or lacking. Most relevant here are the various
subcultures prevalent in liberal societies characterized by the fact of reasonable
pluralism. In discussing subcultures, Kymlicka refers to ``the distinct customs,
perspectives, or ethos of a group or association, as when we talk about a `gay
culture'.''43 Into this category, he places ``the various lifestyle enclaves, social
movements, and voluntary associations which can be found in any modern
society.''44 Kymlicka appears to emphasize two signi®cant differences between
societal cultures and subcultures. The ®rst is one of scopeÐsocietal cultures
provide meaningful options across a broad range of social, political and
economic life, encompassing both the public and private spheres. Subcultures, by
contrast, speak to a much narrower range of activities. The second difference is
one of institutional embodiment. Again, the distinctive feature of societal
cultures, for Kymlicka, is that practices occur in institutions. Presumably,
subcultures lack institutional speci®cation. Thus, when referring to ethnic
immigrants, Kymlicka claims that ``they have left behind the set of
institutionalized practices, conducted in their mother tongue, which actually
provided culturally signi®cant ways of life to people in their original
43Kymlicka, Multicultural Citizenship, p. 18.
45Ibid. at p. 77 (emphasis mine).
It is readily apparent that the second of these distinctions is empirically false. In
liberal societies, the cultural practices of many social groups, including ethnic
immigrants, are institutionally embodied. Note here that Kymlicka appears to
have a rather expansive de®nition of institutions, broad enough to encompass
``schools, media, economy, [and] government.''46 The de®ning feature of
institutions would appear to be that they regularize or concretize social
interactions. Assuming this to be Kymlicka's de®nition of institutions, then
many social groups possess them, including ethnic immigrants. These
institutions range from small social clubs to nationwide cultural and religious
organizations. Nevertheless, despite disparities in size, there are institutions
which facilitate practices which are culturally meaningful, albeit only to
Kymlicka might then respond that the institutions of subcultures are
somehow less signi®cant in comparison to the institutions of societal cultures.
But this claim is false as well. Consider the institutions of ethnic immigrants.47
In North America, there is a long tradition of immigrant communities building
institutions that serve important functions. Many ethnic immigrant
communities operate educational institutions of varying sophistication,
ranging from after-class or weekend schools to secondary schools where
instruction is received both in English and the language of their country of
origin. Immigrant communities have built hospitals and nursing homes. They
operate social service agencies that address a wide range of social problems,
from spousal and elder abuse to alcoholism to dif®culties with cultural
integration. They possess media outlets, offering mother tongue news coverage
in print, radio and television. Kymlicka's frequent references to the ``home and
voluntary...associations,''48 and to the tendency of immigrants to ``maintain
some of their old customs regarding food, dress, [and] religion''49 misdescribes
the complex and rich patterns of social life mediated through the institutions of
ethnic immigrant communities. Indeed, as my examples illustrate, many of
these institutions serve quasi-public functions, or operate in the public sphere.
Granted, these institutions are not governmental, in that they lack sovereignty
and jurisdiction, including the powers to tax and coerce. But even here, the
reality is more complicated. In some ethnocultural communities, for example,
membership fees imposed by religious institutions (for example, synagogues),
coupled with the high costs of exit ¯owing from the control exerted by those
institutions over access to important goods (for example, burial services,
religious marriages, or even social acceptance) combine to create incentives to
contribute ®nancially to those institutions as strong as the legal obligation to
46Ibid. at p. 76.
47For a similar argument, see Chandran Kukathas, ``Multiculturalism as fairness: Will Kymlicka's
Multicultural Citizenship'',Journal of Political Philosophy, 5 (1997), 406±27 at p. 415.
48Kymlicka, Multicultural Citizenship, p. 78.
49Ibid. at p. 14.
pay taxes in political communities.50 In others, the doctrines of private law (for
example, trust, contract) have been successfully used to enforce coercively the
norms of conventional morality, through the threat of expulsion and/or the
denial of a share of communal property.51 To be sure, in the wake of the rise of
the welfare state, these institutions have receded in importance. But they have
always been present, and as the state retreats their importance will likely
Kymlicka's point may be that these institutions, although signi®cant, are not as
expansive in scope as the institutions of societal cultures. That is certainly true,
especially since the institutions of ethnic immigrants rarely extend into the
economic and political spheres.53 However, this does not mean that subcultures,
including the cultures of ethnic immigrants, lack institutional embodiment.
Rather, it only means that those cultures are institutionally incomplete. What this
means is that considerations of scope, both of institutions and of the norms of
tradition and convention themselves, appear to be the central feature that
distinguishes societal cultures and subcultures. But why does this difference
matter? Presumably, Kymlicka wants to argue that the narrow scope of
subcultures somehow renders them incapable of serving as contexts of choice.
On purely functional grounds, this is clearly false. Recall what Kymlicka says
about the value of cultural membership:
Freedom involves making choices amongst various options, and our societal culture
not only provides these options, but also makes them meaningful to us. People make
choices about the social practices around them, based on their beliefs about the
value of these practices . . . And to have a belief about the value of a practice is, in
the ®rst instance, a matter of understanding the meanings attached to it by our
Again, assume P1to be correct. Why then are subcultures any less capable of
serving as the context of choice than societal cultures? It cannot be that
subcultures are de®cient either because they do not present options or because
50Chandran Kukathas is thus mistaken when he ignores the high costs of exit from ethnocultural
minorities in his theory of cultural rights, in ``Are there any cultural rights?'' Political Theory,20
(1992), 105±39.
51Denise G. Re
Âaume, ``Common law constructions of group autonomy: a case study,'' Ethnicity
and Group Rights, ed. I. Shapiro and W. Kymlicka (New York: New York University Press, 1997)
pp. 257±89, and ``The legal enforcement of social norms: techniques and principles,'' Citizenship,
Diversity and Pluralism: Canadian and Comparative Perspectives, ed. A. Cairns, J. Courtney, D.
Smith, P. MacKinnon, and H. Michelman (Montreal: McGill-Queen's University Press, 1999)
pp. 177±201.
52It is also worth noting that these institutions developed in the absence of policies encouraging
ethnic immigrants to view themselves as colonists with the goal of establishing institutionally
complete societies in the mirror image of the mother country. In urban areas, communities of ethnic
immigrants are suf®ciently large and geographically concentrated to enable a similar process to occur.
53Even here, the reality is more complex. Suzanne Model has described the role of networks among
ethnic immigrants that serve as sources of employment opportunities in ``The ethnic niche and the
structure of opportunity: immigrants and minorities in New York City,'' The ``Underclass'' Debate:
Views from History, ed. M. Katz (Princeton, NH.: Princeton University Press) pp. 161±93.
54Kymlicka, Multicultural Citizenship, p. 83.
they do not assign them value. Subcultures certainly do both. Indeed, this is the
de®nition of a culture. Perhaps the problem, then, is that subcultures do not
present the right kinds of options, or more precisely, do not present relevant
options. Maybe subcultures speak to such a narrow range of our existence that
they cannot assist in our attempts to frame, question and revise a conception of
the good. This is de®nitely true for some of the subcultures that Kymlicka refers
to. For example, it hard to imagine an intelligible conception of the good
emerging from ``a `bureaucratic culture'.''55 But, as David Miller has noted, the
subcultures of ethnic immigrants are qualitatively different.56 Recall that, on
Kymlicka's own account, they consist of a shared vocabulary of tradition and
convention that were formerly part of societal cultures in immigrants' countries
of origin. The alleged obstacle posed by the fact of immigration is that those
cultures cannot be actualized because of a lack of institutions. But this
characterization assumes that the cultures of ethnic immigrants speak to
important aspects of their lives as much as the societal culture of the dominant
society; the only problem is one of institutionalization.
Kymlicka might respond that the fact of immigration does more than sever the
traditions and conventions of ethnic immigrants from their institutional means of
support. It also narrows the range of activities the cultures of ethnic immigrants
contemplate; it converts them from societal cultures into subcultures. Instead of
speaking to both the public and private realms, ethnic immigrant subcultures
merely speak to our intimate and associational lives. However, even if this claim
were true, it is not clear why this transformation would render the cultures of
ethnic immigrants less able to serve as contexts of choice. In liberal societies, our
conceptions of the good are often most closely tied to the private sphere. Personal
choices regarding family life, friendship and religious faith often lie at the core of
our most important projects. By contrast, our involvement in the political and
economic sphere is rarely of this character. Indeed, in liberal democracies, our
economic and political commitments are often viewed instrumentally, in terms of
their ability to enable us to pursue our most cherished projects. Inasmuch as the
cultures of ethnic immigrants are limited in scope because of their failure to speak
to economic and political life, this does not undermine their relevance to
individual autonomy.
The institutional viability and relevance of ethnic immigrant subcultures that
are partial in scope demonstrates that P2is false. As a consequence, P3loses its
force, since it is not necessary for ethnic immigrants to possess their own societal
cultures to secure their fundamental interest in cultural membership. In addition,
this reality also suggests that the political claims of ethnic immigrants for
recognition of their particularity in public institutions is more complex than
Kymlicka suggests. Instead of uniformly demanding integration on fair terms into
55Ibid. at p. 18.
56David Miller, On Nationality (Oxford: Oxford University Press, 1995) at pp. 121±2.
common institutions, at least some ethnic immigrants aspire to a mix of inclusion
and institutional separateness. The liberal distinction between public and private
helps to illuminate the topography of these claims. In matters economic and
political, the demand of ethnic immigrants is for inclusion in common
institutions on fair terms. The concern is that institutional separateness will
lead to unfair distributions of important social goods, and to economic and
political marginalization. In matters familial and associational, though, at least
some ethnic immigrants are sympathetic to the idea of institutional separateness.
Integration is perceived as a threat to the maintenance of cultural distinctiveness.
The diverse set of institutions of civil society organized on the basis of
ethnocultural difference in multicultural states testi®es to the strength of this
The distinction between public and private, however, only takes us so far. As I
mentioned earlier, many institutions of ethnic immigrants serve quasi-public
functions. That is, they serve functions that are or could plausibly be provided by
governments. What is striking is that, in Canada, many ethnic immigrant groups
now receive and demand government support for those institutions. This
tendency is most marked with respect to social services, particularly in large
urban areas with substantial immigrant populations. The best way to serve the
needs of these populations is through services that accommodate cultural
difference. Public institutions are capable of rising to this challenge, but what has
evolved instead in many cases is a combination of public support and provision
by organizations tied to particular ethnocultural groups.
What implications does this have for Kymlicka's argument? Inasmuch as the
argument from political sociology turns on the aspirations and institutional
capacities of ethnocultural groups, the sharp dichotomy that Kymlicka draws
between ethnic immigrants and national minorities cannot withstand an
encounter with reality. At best, these groups lie on a sociological continuum
that Kymlicka attempts to shoehorn into his rigid categories. Thus, the argument
from political sociology cannot compensate for the failure of the argument from
consent. Even worse, the argument from political sociology fails on its own
Kymlicka might respond to many of my criticisms by branding them utopian, a
term that is open to at least two interpretations. A utopian political philosophy
may be one that does not speak to urgent problems in the here and now, and
which for that reason is somehow de®cient. Utopian political philosophy fails in
its mission because it does not aim to alter or elucidate the terms of reference for
political debates in which we are now engaged. Conversely, a political philosophy
may be utopian not because of its aim, but because of an ignorance of or
indifference toward the empirical realities surrounding its implementationÐthe
question of means. This charge has been leveled recently against the work of
egalitarian liberals.57 The danger here is not only irrelevance, but also that such a
philosophy could be seized upon by political radicals set on realizing it no matter
what the cost. Indeed, some of the most disastrous social experiments of the last
century were founded on an unyielding and rigid adherence to political
philosophical ideals. To avoid this danger, political philosophy must pay heed
to existing expectations, political practices and institutional capacities, in the
manner that Kymlicka does.
How could one respond to these charges? To be sure, some of the best political
philosophy is written against the background of contemporaneous political
events or debates. Indeed, context may provide both the spark and the materials
for re¯ection on the most dif®cult problems of justice. Without a doubt, the most
admirable feature of Multicultural Citizenship is that it speaks to pressing
political questions confronting liberal democracies today. And Kymlicka's
impressive attempt to provide workable and practical policies that grow out
of, and build upon, current political practices, expectations and institutional
capacities clearly represents a heartfelt desire that political philosophers
contribute to political discourse in a way that is both useful and responsible,
by steering clear of the dangers of irrelevance and intellectual imperialism.
My quarrel with Kymlicka is not with the idea that context and facts should
count in political philosophy. Indeed, my own criticisms of KymlickaÐwith
respect to both the argument from consent, and the link he draws between
territory and sovereigntyÐrely on political sociology in order to defeat his
normative claims. Rather, my concern is that, in the argument from political
sociology, he does not count context and facts in the right way. As we strive for
relevance, we must avoid the temptation to bend our theories around political
realities, for if we do, political philosophy surrenders its critical stance. Without
our ideals, we lack the ability to appreciate what is lost when public policies fall
short of principle.
57Elizabeth Anderson, ``What is the point of equality?'' Ethics, 109 (1999), 287±307.
... Methodologically, following a problem-driven approach and adopting a context-sensitive ethical enquiry, I assume that the examination of linguistic justice requires immersion in individuals' particular contexts rather than abstraction from them (Carens 2000). My discussion adopts the contextual shift undergone by contemporary political thought according to which political practices have been woven into normative arguments (Choudhry 2002). This fact, along with my intention to integrate a normative perspective into a real-life setting, have led the focus of my analysis. ...
Full-text available
In this article, I aim to analyse language rights in relation to groups of immigrant origin. Liberal democracies are reluctant to consider immigrant groups as subjects entitled to the same set of language and cultural rights enjoyed by national minorities. However, the trend towards increasing levels of immigration is configuring new cultural and language correlations within territorial boundaries that provoke responses that problematise a fixed conception of language rights. Drawing on theories of liberal multiculturalism, I examine the case of claims for language recognition in the Spanish autonomous cities of Ceuta and Melilla and its normative implications. In these territories, factors such as size, concentration, and the historical ties of Arabic- and Berber-speaking communities challenge conventional approaches to minority groups’ rights based on a national versus immigrant minority distinction. I argue that these approaches are not satisfactory for language claims in these two cities and that a contextual approach is better suited to conceptualising the recognition of language rights.
Should speakers of immigrant languages have fewer language rights than members of long-settled nations? In this article, I argue that the interests of immigrant and native groups are normatively equally recognition-worthy. Any legitimate differential recognition must follow from the different interests immigrants have in language. I distinguish three categories of immigrants: strictly temporary immigrants, immigrants that have become or are to be citizens and linguistically very wealthy immigrants. The purpose of distinguishing these types is to get a better grasp on the needs and interests of immigrants, by creating sub-profiles. I argue that the first group is entitled to minimal language recognition. The second group, I argue, should be granted equal consideration of their language interests, but that does not necessarily mean equal language recognition. The third group—linguistically very wealthy immigrants such as English-speaking immigrants—typically outrivals the local population in terms of the fulfilment of their language identity interests. Their language interests warrant the need of integration into the national language(s), not the need for equal language recognition.
The European Union is bound to respect the national identities of member states. States might, on occasion, define their national identities in ways that breach inclusive constitutional values (e.g. equality) protected under TEU Article 2. The assumption behind the recognition of diverse constitutional identities is the peaceful coexistence of both, which is challenged by illiberal national developments. We see the relationship between the constitutional recognition of exclusive values (e.g. dominant ethnicity or religion) and inclusive constitutional values as a zero-sum game; every gain by the proponents of emergent authoritarianism translates to a loss for constitutional democracy. While exclusive norms appear in virtually every constitutional system, a critical mass of exclusive values can lead to the hollowing out of a democratic order, both at the national and supranational levels. To try to identify the line where this shift happens, we rely on the limits of toleration and recognition of exclusive norms and identity elements of minority communities in liberal theories of multiculturalism (e.g. Raz, Taylor, Kymlicka). We think that the case of illiberal minorities raises structurally similar theoretical questions, insights and experiences as the dilemma defined above, namely the challenge of illiberal states undermining fundamental EU values.
Full-text available
This chapter first reviews former Soviet approaches to language policy and then offers a comparative analysis of evolving practices in the independent post-Soviet countries from 1991 onwards. In presenting changes in the management of the language regimes that the successor states have deliberately brought about to overturn major aspects of Soviet language policy and to re-institute the national languages, the focus is on language and citizenship legislation, education and language acquisition, titular language assessment, efforts of language institutions to modernize, standardize, or purify the newly (re-) instituted languages, bilateral and multilateral agreements and minority protection, and (the management of) bilingual and multilingual language practices. Examples are drawn from the Baltic states (Estonia, Latvia, Lithuania), the Slavic states (Belarus, Ukraine, the Russian Federation), the Transcaucasus states (Armenia, Azerbaijan, Georgia) and the Central Asian states (Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan). In conclusion we briefly examine the scholarly treatment of policy issues in a selection of the successor states.
The central notion in this article is 'pluriform accommodation,' a term that we have coined to defend two lines of thought. The first is a plea for inclusive and consequential neutrality; the second is a closely linked plea for reasonable accommodation. With 'pluriform accommodation' we emphasize that the multicultural recognition scope should be expanded. The need for inclusive and accommodative rules, laws, and practices is a matter of principle and as such cannot be reduced to the inclusion of people with an immigration background who bring with them all kinds of ethnocultural and religious practices, convictions, and traditions. Furthermore, the enshrined freedom of religion does not provide the needed protection for the multiplicity of conscientious identifications, convictions, and strong allegiances that might be central to one's sense of self. We argue that we should not engage in (top down) debates about the rights of individuals and groups of different types and think in terms of identity hierarchies, but instead should consider the various claims being made (bottom up), aiming for common standards and criteria to assess the validity of these claims and the reasonableness of the associated accommodations.
It is widely considered that minorities’ rights and dignity are protected in liberal societies particularly in western liberal countries up to a level humanly possible. However, in this article I show that it’s not as much heroic a theory as we conceive specifically when the liberal society is committed to accommodate plurality of ideas and lifestyles. After briefly explaining principal tenets of classical liberalism I claim that some fundamental weakness of the theory makes it untenable in liberal plural society let alone non-western society of diversity.
Migration creates linguistic pluralism within immigrant-receiving states, and migrants frequently encounter linguistic barriers upon entry and resettlement into new societies. Immigrant-receiving states therefore confront the need for a language policy, to address utilitarian communication challenges, and to help ensure social cohesion in the face of demographic change.
Any conceptual or empirical analysis of the relations between ‘nationalism’ and ‘multiculturalism’ must begin by acknowledging the ambiguity of such notions—and setting aside those of their possible meanings which would make that analysis fruitless or impracticable from the outset. Thus, in the discussion that follows, the word ‘multiculturalism’ will refer exclusively, not to the fact of cultural diversity—which is characteristic of most contemporary liberal democracies—but to a specific kind of political response to that fact, so as to avoid the confusions deriving from ‘the [widespread] tendency to slide from descriptive to normative uses’1 of that most equivocal term. Similarly, and in contrast with the assumption that a nation can be defined as an ethnically homogeneous community—an assumption seemingly embraced by some of the leading scholars in the field, who tend to equate nationhood with cultural distinctiveness,2 thereby leaving it to others to account for the historical process by which nationalist movements actually invented the distinctive ‘culture’ of their nation-to-be3—I will adopt a more consensual and, at any rate, less unduly restrictive definition of that second, equally capacious notion. For the purpose of this article, the word ‘nation’ will refer to a community of people characterised by some common cultural features, mutual recognition, ‘the anonymity of membership’4and an aspiration to collective political self-determination that distinguishes it from an ethnic group (although an ethnic group whose identity is being threatened is likely to begin to think of itself as a nation). For while ‘ethnic groups can transform themselves into national ones and national communities may define their identity in terms of common ethnic origins […], this broad area where ethnicity and nationhood overlap does not make the two phenomena identical. National unity need not refer to common descent and ethnic groups need not understand themselves as separate political communities within the wider society.’5
Elsewhere, I have argued that comparative case-law can serve the same function in constitutional adjudication, in``Globalization in search of justi®cation: toward a theory of comparative constitutional interpretation
  • Culture Carens
Carens, Culture, Citizenship, and Community, pp. 4±5. Elsewhere, I have argued that comparative case-law can serve the same function in constitutional adjudication, in``Globalization in search of justi®cation: toward a theory of comparative constitutional interpretation,'' Indiana Law Journal, 74 (1999), 819±92.
  • David Pearl
  • Werner Menski
David Pearl and Werner Menski, Muslim Family Law, 3rd edn (London: Sweet & Maxwell, 1998), ch. 2.
Yael Tamir has made a similar point, in``Theoretical dif®culties in the study of nationalism
Yael Tamir has made a similar point, in``Theoretical dif®culties in the study of nationalism,'' Canadian Journal of Philosophy, 22 Supp. (1996), 65±92 at p. 79.
Chandran Kukathas is thus mistaken when he ignores the high costs of exit from ethnocultural minorities in his theory of cultural rights, in``Are there any cultural rights?
Chandran Kukathas is thus mistaken when he ignores the high costs of exit from ethnocultural minorities in his theory of cultural rights, in``Are there any cultural rights?'' Political Theory, 20 (1992), 105±39.
The legal enforcement of social norms: techniques and principles
  • Denise G Re Âaume
Denise G. Re Âaume,``Common law constructions of group autonomy: a case study,'' Ethnicity and Group Rights, ed. I. Shapiro and W. Kymlicka (New York: New York University Press, 1997) pp. 257±89, and``The legal enforcement of social norms: techniques and principles,'' Citizenship, Diversity and Pluralism: Canadian and Comparative Perspectives, ed. A. Cairns, J. Courtney, D. Smith, P. MacKinnon, and H. Michelman (Montreal: McGill-Queen's University Press, 1999) pp. 177±201.