Introduction

From the 1990s up to now, culture has incontestably become one of the most debated issues among political philosophers. The advocates of what is now usually called 'multiculturalism', such as James Tully, Charles Taylor, Bhikhu Parekh or Iris Marian Young, only to mention these names, have sought to make liberal political theory, as it has primarily developed since the publication of Rawls's A Theory of Justice, more sensitive to the 'cultural difference', that is, to the fate of minority groups like African American, women, religious, immigrant and ethnic groups, or indigenous and national minorities. But the question remains: Why does culture matter so much? And how can a liberal political theory take account of the importance culture has for individuals that is compatible with liberal values? In a nutshell: How can a liberal defence of the worth of cultures be provided?

Kymlicka's theory of minority rights provides probably the most important and currently discussed liberal defence of the worth of culture. Such a defence argues for the instrumental role culture plays in people's ability to make meaningful choices and lead a self-directed existence. However, one of the recurrent difficulties of this approach is making a convincing case for the protection of 'particular' cultures. After the publication of his first book Liberalism, Community and Culture (1989), Kymlicka was criticized by many for merely explaining why all individuals need membership in some culture, but not why they need to be members of a particular culture (see Buchanan, 1991, 54–55; Binder, 1993, 250–256; Margalit and Halbertal, 1994; Waldron, 1995, 105–110). If the importance of culture lies in its ability to enhance the autonomy of the group's members and to provide them with the tools for leading an autonomous existence, then most dominant cultures (mainly the liberal ones) can fulfil this function far better than many endangered minority cultures or, to put it differently, the particular culture in which some people are raised might provide fewer or less meaningful choices than other cultures, so that assimilation into another culture, or familiarity with more than one culture, could be more likely to extend and enrich our range of options. So it seems that by arguing merely for the instrumental role culture plays in an autonomous life, Kymlicka has not been able to make a convincing case for the protection of specific endangered cultures as his theory of minority rights is meant to do. In his second major book, Multicultural Citizenship (1995), Kymlicka sought to respond to these criticisms by insisting on the importance culture has for the identity of a group's members, for their secure and effortless belonging and their self-respect, and the difficulty of moving between different cultures. But this response, as it has no necessary logical connection with the autonomy argument first provided, does not seem to have appeased his critics either. Some maintain that Kymlicka's theory, in spite of the above explanations, still provides people with no more than the right to a culture and language, not the right to their culture and language.¹ In the same vein, this has not prevented others from maintaining that Kymlicka's theory provides plausible arguments to support the claim that access to a societal culture is a primary good, and that people have interests in accessing their own societal culture, but not the claim that access to one's own societal culture is a primary good (Patten, 2003, 183–187).

The above criticisms, whether sympathetic or not, are crucial to Kymlicka's endeavour as they are directed at the core of his theory of minority rights. In this paper, I would like to show that in order to provide a really convincing argument for the protection of minority cultures, we have no choice but to abandon Kymlicka's current instrumental account of cultures in favour of a conception of cultures viewed as intrinsically valuable. Does this shift of strategy necessarily lead to a kind of communitarian defence of the worth of cultures more or less at odds with liberal canons? I think not. I think that a liberal defence of the intrinsic worth of cultures can and must be provided and that such a defence is probably the one most likely to resist the criticisms mentioned above.

In the first part of the paper, I provide three reasons why Kymlicka's instrumental account of culture constitutes an obstacle to his argument for the protection of minority cultures and why it must be replaced. Next, turning to what I call the liberal defence of the intrinsic value of cultures I demonstrate that such a view can be found in the work of Joseph Raz, but that we must offer a slightly modified version of it if we really want to provide a 'liberal' defence in due form. Then I examine the rationale behind Denise Réaume's recent rejection of the kind of defence of the intrinsic value of culture developed in this paper, and I demonstrate that her critique is unconvincing because it is founded on dubious logical arguments. Afterwards, I explain why the liberal account of culture as intrinsically valuable can better meet the objections currently raised against Kymlicka's own instrumental account. Finally, in the last part, I briefly outline the kind of policy towards non-liberal minority cultures that would be in accordance with a liberal defence of the intrinsic value of cultures.

(I)

I see three major reasons why Kymlicka's instrumental account of culture poses a problem for his defence of the rights of minority cultures. First, such an account cannot satisfactorily be distinguished from what we may call, according to Waldron's (1995) use of the term, the 'cosmopolitan view' of culture. According to such a view, endorsed not only by Waldron but also by Buchanan (1991, 52–64; 1998) and Habermas (1996; 2005, 17–18),² cultures have no inherent or intrinsic value. Their value is only instrumental and resides in their capacity to provide their members with the cultural goods necessary to make meaningful choices. But it matters little what particular form those cultural goods may take and this is why particular cultures need no protection. If there is anything such as 'cultural rights', those rights should not be restricted to access to cultures of origin — too often mistakenly conceived of as homogeneous wholes, and thus reified into closed totalities — but rather grant access to a variety of cultural goods and traditions. To claim the contrary would be tantamount to saying that certain collective goods, or certain cultures, have a value in themselves, and must be protected out of consideration for their individual members. But such a view is incompatible with the usual liberal standards that grant individual persons only a moral value.

Kymlicka rejects the above conception, not because it provides an inaccurate appraisal of the worth of cultures, but because it does not give due weight to the distinction between choice and circumstances. In effect, since the publication of his first book (1989, chapters 8 and 9; 1995, chapter 6), the standard argument Kymlicka advances for the protection of minority cultures has relied on the choice/circumstance distinction made by most liberal egalitarians since Rawls (the so-called 'luck egalitarians'), according to which egalitarian politics must, to be fair, strive to compensate as far as possible for inequality of circumstance rather than choice. If belonging to a culture were simply a matter of choice, it would be hard to make a case for a politics of protection of cultures since this could be interpreted as a form of public support for expensive tastes over less expensive ones. However, it turns out that, in the same way as the brute luck of people's belonging to certain economic, social, racial or sexual groups puts them at a disadvantage in comparison with other groups, the brute luck of people's belonging to certain minority cultures, such as non-sovereign minority nations, puts them at a disadvantage in comparison with majority cultures, disadvantage arising, not from choice, but from unchosen circumstances. A consistent egalitarian politics should then consider the compensation for the inequalities of circumstance related to cultural membership in the same way as all the other inequalities that are due to the brute luck of people's belonging to certain social groups. The mistake made by most cultural cosmopolitans would be to place cultural claims only (or mainly) on the choice side of the choice/circumstance distinction, thereby downplaying or ignoring the important inequalities deriving from the brute luck of cultural membership.

What is at issue, however, is not whether the rationale behind Kymlicka's theory of minority rights is legitimate. It probably is.³ What is at issue is whether his instrumental account of culture is appropriate for supporting such a theory. And one may suspect it is not. Such an account does not explain why people need access to their own societal culture and why cultural minority rights, especially under the guise of self-government rights, are the kind of remedial measures necessary or most appropriate for rectifying the inequalities of circumstance related to cultural membership. It does not explain why justice requires that access to cultures of origin be protected rather than access to another culture, or why access to a variety of cultural goods and traditions should not be facilitated. In sum, contrary to the cosmopolitan view, Kymlicka's egalitarian approach to culture rightly stresses the inequalities of circumstance deriving from cultural membership, but his functional account of culture does not support, any more than the cosmopolitan view, the necessity of appealing to a politics of protection of threatened cultures as the only or the best remedy.

One might argue nevertheless that it is unjust to claim that Kymlicka's work provides no explanation as to why people need access to their own societal culture and why the protection of their particular culture by means of cultural rights is the appropriate remedy to inequalities of cultural membership. In fact, Kymlicka provides two distinct explanations. First, integrating into another culture would be too costly a process and too great a sacrifice that people should not be required to bear. It would be analogous to choosing to take a vow of perpetual poverty (1995, 85–86). Secondly, culture would be crucial to personal identity and self-respect, and consequently, the bonds of language and culture would be generally too strong to give up (1989, 175–176; 1995, 89–90). I will examine each of these responses in turn. As we will see, they provide two further reasons for considering Kymlicka's instrumental account of culture as a problem in his defence of the rights of minority cultures.

First, as Denise Réaume has convincingly demonstrated, the claim that the protection of particular minority cultures or languages remains the best remedial policy as regards inequality of cultural membership depends on the assumption that the obstacles to mastering another language or to integrating into another culture are too great and cannot be overcome. But this assumption is based only on contextual and practical considerations. It is not conceptually impossible to think that such constraints could be circumvented. It is not impossible, for instance, to put in place 'a very careful process of integration designed to support the self-esteem of members of the minority as they learn their new culture', or 'to make good the economic losses of the adult members of the minority, knowing that their children, raised within the new culture, will not suffer this cost' (Réaume, 2000, 249). Nor is it impossible to contemplate implementing programmes that facilitate and support the gradual transferral of people from one culture to another by enabling 'minorities to raise their children biculturally and then expecting the bicultural children of these families to complete the transfer to the majority culture and language with their own children' (p. 250). In brief, what is to be proscribed, and what has proved to be prejudicial to minority cultures, is forced assimilation.⁴ But gentle, gradual, supportive and facilitating integration into another culture could be considered, if not a better, at least as good a remedial solution to inequality of cultural membership as the protection of particular cultures and languages. It could even be less costly for the society as a whole and provide members of minority cultures with more opportunities to lead a flourishing existence. As we can see, the main weakness of the cost-of-transfer argument is that if we could find a way to eliminate or significantly reduce these costs, however realistic, 'this would seriously weaken the claim to accommodation designed to allow a minority to maintain its linguistic practices' (p. 249), or its own culture. The latter claim can stand up only if culture is envisioned as intrinsically valuable. But the moment culture is viewed as only instrumentally valuable, the cost-of-transfer argument can no longer by itself suffice as a ground for a politics of protection or accommodation of particular languages and cultures.⁵

Let's now turn to the identity and self-respect argument. As we saw earlier, the standard objection against Kymlicka's instrumental account of culture is that it bases the case for the protection of minority cultures on a view of culture as a 'context of choice', whereas most members of minority cultures usually base their demands on a perception of their culture as a 'context of identity', whatever the range of options it provides.⁶ However, as some critics have rightly insisted (Forst, 1997, 66), it is not in the end an argument for culture as a 'context of choice', but for culture as a 'context of identity' on which Kymlicka, in Multicultural Citizenship, bases the differential treatment enjoyed by national minorities in relation to immigrant groups. In other words, only the view of culture as a specific societal context of socialization in which people are raised, and form their identity can explain why national minorities are allowed to retain their culture rather than assimilate into majority nations as immigrants do. The view of culture as a context of choice has nothing to do with this, otherwise integration into majority nations, which usually offer greater opportunities, would prove to be as attractive for immigrants as for national minorities. In order to clarify his position, Kymlicka gives the following answer: I admit that my argument here was unclear, but what I meant to argue was that considerations of identity provide a way of concretizing our autonomy-based interest in culture. In principle, either the minority's own culture or the dominant culture could satisfy people's autonomy-interest in culture, but considerations of identity provide powerful reasons for tying people's autonomy-interest to their own culture. Identity does not displace autonomy as a defence of cultural rights, but rather provides a basis for specifying which culture will provide the context for autonomy. (Kymlicka, 2001, 55)

The only problem with this response is that, the moment culture is described instrumentally, as a mere tool for making meaningful choices, the connection between the two views of culture, as a context of choice and as a context of identity, remains largely contingent. It is not necessary for individuals to grow up in their own culture in order to fully exercise their capacity for autonomy. This capacity could develop as well in 'a freewheeling cosmopolitan life, lived in a kaleidoscope of cultures' (Waldron, 1995, 99). The fact that we need our own culture in order to exercise autonomy is based only on empirical observations and practical considerations about the role cultural membership plays generally in the life of individuals and the harm its absence or distortion causes to individuals. But, as Kymlicka himself notes, these are only 'general trends'. There might be other people, at other times, who might have no such bonds, for instance, stateless people or people who 'seem most at home leading a truly cosmopolitan life, moving freely between different societal cultures' (Kymlicka, 1995, 90). And this proves that people's own culture as the context in which to exercise their autonomy is nothing more than a means among others, perhaps an important one, but one which nevertheless is to be considered as a matter of choice and preference, not of need and necessity. This conclusion becomes inevitable as soon as culture is depicted merely as an instrumental context of choice deprived of any intrinsic value. So long as culture as a context of choice and culture as a context of identity are not internally linked, and the latter does not become an intrinsic, non-instrumental condition of the former, people's bond to their own culture will look like a contingent fact whose importance varies according to circumstances and individual preferences rather than a basic need, a 'primary good', which should be taken into consideration, and that governments would have to respond to.

(II)

Thus far I have tried to show that the instrumental account of culture is non-viable for any consistent defence of the rights of minority cultures such as the account Kymlicka puts forth in his work. Now the question arises as to whether it is possible to provide an intrinsically valuable account of culture compatible with liberal canons and which, as far as possible, avoids the defects of the communitarian account of culture. I think such an account of culture is expressed by Joseph Raz. According to Raz (1986, 200), something has an intrinsic value when it has a value in itself, independently of the value of its actual or possible consequences. By contrast, something has an instrumental value when it derives its value from its actual or possible consequences. Raz makes a distinction between different categories of intrinsically valuable goods, for example, between ultimate goods (those whose value need not be explained or justified by reference to other goods or values) and constituent goods (those that render the existence of the former possible and whose value need be explained or justified by reference to these other goods). For Raz (1986, 204–205), having an autonomous life would be an ultimate good. Collective goods and, among them, those that provide a sufficient range of acceptable and meaningful options, would be constituent goods, that is, the kind of intrinsically valuable goods the value of which cannot be explained or justified without reference to other goods, in the present case, individual autonomy. For instance, the cultural context provided by a form of life would be a constituent good as it allows individuals to make valuable choices.

Two things are worth noting in this very shortened account of Raz's view of culture. First it has to be distinguished from most variants of the communitarian account of culture for which a collective good such as culture can have an inherent value that is non-derivative, in other words, a value totally independent of its contribution to the well-being of individual human beings, a position that may be called, following Michael Hartney's use of the term (Hartney, 1995, 206–207), 'value-collectivism'. By contrast, according to Raz's view of culture, a collective good such as culture may have an intrinsic value, but this value can be only derivative : only the lives, well-being and autonomy of individual human beings have ultimate intrinsic value, and collective goods such as culture derive their value from their contribution to the lives, well-being and autonomy of individual human beings. Hartney (1995, 206) calls this position 'value-individualism', and he distinguishes it from another doctrine that is officially rejected by Raz (1986, 198), one called 'moral individualism'.

Is there a distinction to be made between 'value-individualism' (presumably endorsed by Raz) and 'moral individualism' (explicitly rejected by him)? I think there is, and this is the second aspect I want to emphasize in Raz's position. There is a distinction to be made between two kinds of individualism: one ascribes an intrinsic value whereas the other only an instrumental value to culture. What Raz rejects is not the humanistic principle, probably shared by all liberal theories, that the well-being of individual human persons, or the quality of human life, has ultimate value, that the goodness or the badness of anything derives ultimately from its contribution to the latter. What he rejects is the idea that some collective goods such as culture would have only an instrumental role to play in the well-being of persons and the quality of human life. As Raz puts it: 'A moral theory will be said to be individualistic if it is a humanistic morality which does not recognize any intrinsic value in any collective good. In other words, individualistic moralities are humanistic moralities which hold that collective goods have instrumental value only.' (Raz, 1986, 198). As we saw earlier, this is precisely the position held by Kymlicka. But how exactly does Raz's view of culture differ from other, more instrumental, ones? What distinguishes instrumental and intrinsic humanistic accounts of culture?

I think the best analysis of the difference has been done by Denise Réaume (2000, 246–247).⁷ The moment individual autonomy, the self-directed life, the ability to plan one's life, is recognized as an ultimate intrinsically valuable good (a central part of human well-being and of what makes human life worthwhile), what role does culture have to play in this? According to one response (e.g. that of Kymlicka), that role is only instrumental. In other words, culture is only causally related to the ability to make choices, and to human flourishing generally. Other goods (wealth, health, education, financial power, political influence, etc.) could contribute as well to the ability to make choices and to human flourishing. Another response, that of Raz emphasized by Réaume, considers that the role of culture is intrinsic. Culture would be, not causally and externally, but logically and internally connected with a self-directed life. According to the instrumental reading, a cultural context of choice is sufficient for acquiring the capacity to plan one's life but is not a necessary condition for it. According to the intrinsic reading, a cultural context of choice represents a necessary condition of that capacity, a constitutive element of it, one without which a self-directed life would be impossible. The moment one recognizes that A has intrinsic value and that B is logically connected to A, it follows that derivatively B has intrinsic value. As Réaume puts it: 'To say that the use of culture is a logical condition of something else that has intrinsic value is to make the cultural context for choice itself intrinsically valuable. The pursuit of a life plan cannot be regarded as an objective independent of a context giving it meaning.' (Réaume, 2000, 247)

The main advantage of the intrinsic, as opposed to the instrumental, reading of the value of culture is that it makes the protection of culture something necessary and indispensable to individuals. As Réaume puts it, '...if individuals are entitled to the protection of their ability to choose a life plan, the context that makes choice possible must deserve protection' (Réaume, 2000, 247). According to the intrinsic reading, cultural membership is no longer a mere preference among others, it becomes a primary good, a basic need that any government has to recognize and be responsive to.

Nevertheless, a measure of caution needs to be exercised with regard to the intrinsic reading I have thus far examined based on Raz's. A variant of this reading is possible, one I consider misguided, and which could perhaps voice in some way the position espoused by Raz and/or Réaume's interpretation of it: 1. A self-directed life (as a central part of human well-being and a worthwhile life in general) is intrinsically good and valuable;
2. Cultural membership is a logical condition of a self-directed life;
3. Therefore, cultural membership is intrinsically good and valuable.

The weakness of this reasoning is its logical inconsistency. If A has a certain property and B is a logical condition of A, it does not follow that B also has this property. If being altruistic is a virtue, and having certain psychological traits (intelligence, extroversion, sensitivity, imagination) is a necessary condition for being altruistic, it does not follow that the mere fact of having these traits is per se a virtue.⁸ In the same way, if a self-directed life is intrinsically valuable and if cultural membership is a logical condition of a self-directed life, it does not follow that cultural membership is also intrinsically valuable. The only consistent conclusion we can derive from these premises is that cultural membership in conjunction with a self-directed life (in other words, a culture that contributes to individual autonomy) is something intrinsically valuable. This conclusion certainly reflects the meaning of Raz's distinction between 'ultimate' and 'constituent' or 'derivative' intrinsically valuable goods. However, such a distinction must be voiced in our reasoning clearly and without confusion. The simple fact that someone has access to a culture or uses a cultural context does not tell us anything about the value of that culture or cultural context, even if one assumes that cultural membership represents a logical condition of an autonomous life and that autonomous life is intrinsically valuable. The reason for that is that cultural membership forms, as a logical, not a causal antecedent, a necessary, not a sufficient condition of a self-directed life. It could perfectly turn out — as is very often the case — that belonging to a culture cannot allow the full exercise of one's ability to plan one's life, for instance, when such a culture is oppressive, closed or authority-based. The only thing that cultural membership tells us as a logical condition of a self-directed life is that the latter could not occur without the former, not that cultural membership will always give rise to a self-directed life. I think that, in order to properly reconstruct the argument that has to be made for a suitable defence of the politics of protection of minority cultures, and to make it both logically consistent and theoretically compatible with the usual liberal standards, one should revise the previous reasoning slightly as follows: 1. A self-directed life (as a central part of human well-being and a worthwhile life in general) is intrinsically good and valuable;
2.Cultural membership is a logical condition of a self-directed life;
3'. Therefore, any cultural membership that contributes to a self-directed life is intrinsically good and valuable.

In my opinion, this conclusion corrects the defects of Kymlicka's account of culture while preserving the central core of his theory. In effect, contrary to Kymlicka's instrumental account of culture, it makes the protection of culture something indispensable to individuals, it makes cultural membership a real primary good, no longer a mere preference, desire or interest people would have in belonging to a culture. But it also preserves the thrust of the liberal argument Kymlicka makes for the protection of minority cultures: not any culture, but only the liberal ones, only the cultures that foster or, at least, do not jeopardize the possibility of a self-directed existence, deserve to be protected and are to be envisioned as authentic primary goods, as something necessary and indispensable to individuals, in a nutshell: as intrinsically valuable goods. I call this the liberal defence of the intrinsic value of cultures.

(III)

Denise Réaume rejects the liberal defence of the intrinsic value of cultures. She contends that this sort of defence works only when it remains at the abstract level. Access to a culture would be a logical condition of a self-directed life in the sense that all human life, to be worthwhile and autonomous, requires some cultural context. But, according to her, this would be insufficient to establish the intrinsic value of a particular cultural context. As she puts it: 'The English linguistic context is not a logical condition of planning a life, any more than German, Cantonese, or Tagalog is. A life can, in principle, be equally satisfactorily planned in any language' (Réaume, 2000, 247). In other words, the intrinsic account would be confronted with the same kind of difficulties as the instrumental one: it would provide people with no more than the right to a culture and language, not the right to their culture and language. The intrinsic account would explain why all individuals need membership in some culture, but it could not explain, any more than the instrumental account, why they need to be members of a particular culture. Finally, the only difference between the intrinsic and the instrumental accounts of culture would be that, for the intrinsic view, only culture as an abstract idea and a logical condition of a self-planned life would have intrinsic value. But all particular cultural contexts would have instrumental value only, in the sense that they would provide their members with the concrete linguistic and cultural tools that enable them to lead a worthwhile and autonomous life, none of them having intrinsic worth.

This point of view seems to be misguided. As previously mentioned, the only valid conclusion we can derive from the assumption that an autonomous life is intrinsically valuable and that a cultural context is a logical condition of such an autonomous life is not that all cultural contexts are intrinsically valuable (or that culture as an abstract idea is something intrinsically valuable), but that a cultural context contributing to an autonomous life is intrinsically valuable. Starting with this premise, consider now the following reasoning: 4. A cultural context that contributes to a self-directed life is intrinsically good and valuable;
5. C is a cultural context that contributes to a self-directed life;
6. Therefore, C is intrinsically valuable.

What this reasoning says is that C is intrinsically valuable the moment it contributes to a self-directed life, the moment it possesses all the features of a society that fosters or does not jeopardize the well-being and autonomy of individuals, thus contributing to the flourishing of human life. The English, German, Cantonese or Tagalog cultural contexts may be other specific instantiations of the idea we have of an intrinsically valuable culture if they contribute to the well-being and autonomy of individuals. If they do so, they deserve to be protected as specific cultures because membership in those cultures proves to be intrinsically valuable. Members of those cultures can demand protection of their specific culture, not on the grounds that it is unique or that it forms a context of identity, but on the grounds that, as an exemplar of a liberal culture, a culture that promotes autonomy and contributes to human flourishing, it is to be considered as intrinsically valuable.⁹ The mistake made by Réaume, a logical one, lies in believing that the intrinsic account of culture cannot establish the intrinsic value of a particular cultural context simply because no particular context can be a logical condition of anything. But this is a fallacy. A specific cultural context can have intrinsic value while at the same time not be a logical condition of anything else. Let us consider the following reasoning: 7. A cultural context is a logical condition of a self-directed life;
8. A cultural context that contributes to a self-directed life is intrinsically good and valuable;
9. C is a cultural context that contributes to a self-directed life;
10. Therefore, C is intrinsically valuable.

What this reasoning shows is that the first premise is perfectly compatible with the conclusion, that it is not necessary for a particular cultural context to be a logical condition of a self-directed life in order to contribute to such a self-directed life and to be therefore intrinsically valuable. The first premise expresses only the general statement that a self-directed life cannot occur without a cultural context. The moment C instantiates such a cultural context, it naturally does not follow that C becomes a logical condition of a self-directed life in general. It follows, however, that C forms a particular sample of a cultural context without which a self-directed life could not occur. C makes possible individual autonomous existences without being a logical condition of individual autonomy. The English, German, Cantonese, or Tagalog cultural contexts can also make autonomous lives possible without being a logical condition of an autonomous life generally speaking.

(IV)

Once we have analysed and rejected the objections raised by Réaume against the liberal defence of the intrinsic value of cultures, it remains to be demonstrated that the liberal intrinsic account can better meet the usual objections raised to the instrumental account of culture espoused by Kymlicka.

As observed, Kymlicka's instrumental account differs from the cosmopolitan view of culture only in that it takes more seriously into account the distinction between choice and circumstances. But it shares with the cosmopolitan view the idea that cultures have no inherent or intrinsic worth. Their value is only instrumental and resides in their capacity to provide their members with the cultural goods necessary to make meaningful choices. The problem with such a conception, however, is that it does not explain why justice requires the protection of cultures of origin rather than a facilitated access to a variety of cultural goods and traditions. The liberal defence of the intrinsic value of culture can respond to this problem by emphasizing what I said earlier: members of endangered minority cultures can demand protection of their specific culture on the grounds that it is an exemplar of a culture that contributes to individual well-being and autonomy, to human flourishing and, as it were, one which is to be considered as intrinsically valuable. Such a demand is what is required by justice because such a culture forms a sample of a cultural context with intrinsic worth, and that access to another culture, an even more secure and prosperous one with a larger range of opportunities is, contrary to the instrumental account, no longer attractive. It is no longer attractive because the former already offers its members the cultural environment suitable to lead a meaningful, flourishing and autonomous existence. If its only defect is that it is less secure than another culture, the appropriate measure is surely not to convince its members that it would be preferable for them to leave their culture and enter another one, but to make it more secure with suitable rights. If it offers a lesser range of opportunities than another, then once again, the appropriate measure is not for the members to leave their culture but to consolidate its basic structure, making it more secure, so that in the short or the long run it will give them access to a wider range of opportunities.

The thrust of the above response to the usual objections raised with regard to Kymlicka's instrumental account of culture is that endangered minority cultures deserve to be protected with appropriate rights, not primarily because such cultures form a context of identity for their members, not even because it would be too costly for them to move to another culture, but only because they are, according to liberal standards, cultures with intrinsic worth. The fact that those cultures form a context of identity or that it would be too costly for the members to leave their culture is naturally further reason to protect them. But the first and fundamental reason for providing them with due protection is that they form intrinsically valuable cultures. And this has important consequences for the two other problems that Kymlicka's instrumental view of culture poses.

First, Réaume's objections to the cost-of-transfer argument no longer have weight. Gradual and facilitating integration into another culture, were it possible, can no longer be considered as good a remedial solution to inequality of cultural membership as the protection of particular cultures and languages, even if it were less costly for the society as a whole or if it provided members of minority cultures with more opportunities to lead a flourishing existence. Such a possibility appears attractive only when culture is viewed as merely instrumentally valuable, where the cost-of-transfer argument cannot by itself suffice as a ground for a politics of protection of cultures because it is based only on considerations given to practical constraints, which can be logically overcome. But the above possibility is no longer attractive when intrinsic worth is ascribed to a culture. In such a case, it seems reasonable to believe that justice should require that such a culture be protected through appropriate rights and resources, and that the transferral to another culture be envisioned as an alternative only in extreme cases of urgent need, when the decay of such a culture can no longer be avoided without excessive cost to the whole society, but not in other circumstances.¹⁰

Second, the liberal defence of the intrinsic worth of culture can demonstrate that the two aspects of culture discussed earlier, as a context of identity and as a context of choice, are, contrary to the instrumental account, in no way contingent, but internally linked. What does this mean? This means that a shared context of identity is considered to be a necessary condition of culture as a context of choice. If a self-directed existence is to be exercised somewhere, it is to be exercised in a concrete, particular cultural context. And if we assume that every concrete, particular culture forms a context of identity for its members, then it follows that a self-directed life cannot occur without a context of identity. As we have seen, it is far from sure that, according to the instrumental account, a context of identity is indispensable for people to exercise their capacity for autonomy. Any cultural context of choice might prove to be as fruitful as any other one for people in order to exercise their capacity for autonomy and such a capacity could develop as well in 'a freewheeling cosmopolitan life, lived in a kaleidoscope of cultures'. But then, it becomes difficult to understand why governments should satisfy the preferences of those deeply attached to their culture with corresponding rights and resources. It must be demonstrated that culture as a context of identity is, not a contingent and instrumental, but a necessary, condition of culture as a context of choice. How can this be demonstrated? The supporters of the instrumental account of culture, and many cosmopolitans among them, would surely reject this claim as implausible. The latter simply cannot take into account numerous cases where culture as a context of identity cannot be considered by people as a necessary condition for exercising their capacity for autonomy. What about people who, like Salman Rushdie, have grown up in different cultural contexts and formed what we might call a hybrid identity? What about complex, non-homogenous, multicultural contexts where people become completely bilingual, are perfectly able to move within two or more different cultural environments, and develop allegiances to more than one national community? I will only make a few remarks to clarify this point.

The contention that culture as a context of identity is a condition of culture as a context of choice is not incompatible with the existence of hybrid cultural identities, or identities formed within non-homogenous, multicultural contexts. Many individuals have built complex identities, or have become perfectly bilingual or multilingual, due to numerous migrations, or their involvement in a rich cultural environment with many influences from different traditions. The case of Salman Rushdie is only one example. But when one speaks of a culture as a context of identity, one does not claim that such a culture should be perfectly pure. Nothing is said about the character of that culture. Some cultures are more homogenous, as is the case with the German cultural context. Some others are penetrated by the influence of multiple traditions. India is a good example. Individuals will come to develop more uniform identities in the former case, more complex and hybrid identities in the second case, or often when they have moved from one country to another. In each case, culture as a context of identity continues to form a condition of culture as a context of choice. What is excluded by this statement is not the possibility of 'mongrelized' identities. It is not even the possibility of debating, revising and transforming a primary acquired identity in light of other cultural traditions or constructing other aggregated identities. In a nutshell: The core idea that our selves are normatively prior to our ends, in the sense that no end is exempt from possible re-examination, remains untouched. What is excluded, however, is the empirical possibility that such an examination can be executed, not only in a vacuum — all cosmopolitans agree on this point — but also by someone who would not have been raised in a specific cultural or linguistic environment in which he/she has built a primary identity, in other words, a 'self' who would exist prior to or independently of any process of socialization in a particular culture and regard externally all traditions, languages and cultural heritages as a mere source of cultural supplies at his/her disposal. There is no doubt that Kymlicka also excludes such a possibility. But in considering cultures as mere tools for people's autonomy, the necessity for individuals to grow up in a specific culture in order to make valuable choices or develop further identities, whatever the character of such a culture, simply disappears and becomes a mere contingent fact.

However, if culture as a context of identity is a necessary condition of culture as a context of choice, the intrinsic account considers it as an insufficient condition. In other words, and still contrary to the instrumental view advocated by Kymlicka, the primary reason for protecting a culture as a context of identity is not that its members have deep bonds to their cultural heritage and cherish their identity, but that it forms a valuable context of choice that enables them to lead autonomous lives and, as such, represents a cultural context with intrinsic worth. It is not sufficient for a culture to form a lively context of identity cherished by its members for it to deserve protection. People may be said to 'need' a culture as a context of identity only so long as it represents a condition of their autonomy. Otherwise, there is no basic need or primary good that governments would have to be sensitive to. Cultures must be protected as contexts of identity only so long as these contexts also form genuine contexts of choice, but not as contexts of identity per se because, if one assumes that a context of choice supposes a context of identity, a context of identity does not necessarily represent a genuine context of choice having intrinsic value.

(V)

To end this discussion, let me examine the last point. What does the liberal defence of the intrinsic worth of cultures have to say about non-liberal (i.e. authoritarian, hierarchical, patriarchal and inegalitarian), indigenous or national minorities not deemed appropriate to provide their members (mostly the women and children) with the kind of cultural environment suitable to lead autonomous lives? Until now, two prominent approaches have been developed within the liberal tradition to meet this challenge. The first is an autonomy-inspired policy of intervention in illiberal minority cultures, sometimes not unfavourable to measures close to assimilation (Raz, 1986, 424; 1994, 82; Okin, 1999, 22–23). The second is a tolerance-based policy of abstention, seeing in pluralism (Galston, 1995, 2002; Spinner-Halev, 2005) or in the freedom of association (Kukathas, 1992) the core liberal value and leading most often to a defence of exit rights. The main weakness of the first approach is that it undervalues all cultures deemed non-liberal, separates too drastically liberal and illiberal cultures, and as a consequence, idealizes the conditions for autonomy in our own consumer, materialistic societies (Spinner-Halev, 2005, 162). But the second approach is not spared from difficulties. Liberal pluralists and associationists underestimate the sociological and psychological constraints on leaving one's culture and freely exercising exit rights without state action (Reitman, 2005). Given the pitfalls faced by the two approaches, some liberals have recently attempted at overcoming the traditional intervention/abstention policy options (Weinstock, 2005). As we can see, the literature on non-liberal minority cultures and their internal minorities is vast and complex, and I do not intend to make a detailed argument in what follows. My aim will be only to sketch the general outline of the avenue most compatible with the liberal defence of the intrinsic worth of cultures advocated in this paper. In my view, this avenue is an autonomy-based defence of exit rights.

Rather than avoiding the intervention/abstention options, an autonomy-based defence of exit rights should seek to specify where and when a state should intervene or refrain from intervention in a manner that gives due weight to autonomy and the basic interests of the group's members, in particular in ethical agency (the capacity to live one's own life as one sees fit) and in well-being (Song, 2006, 2), while being sensitive to the injustice and the history of oppression experienced by many indigenous minority cultures (Song, 2006, 17–23). In order to avoid the main difficulty faced by most autonomy-based approaches, an autonomy-based defence of exit rights should condemn any form of state intervention as long as there is no humanitarian urgency. Since non-liberal minority cultures form, not only contexts of choice, but also contexts of identity cherished by most of their members, and since the present disadvantage they experience in terms of individual liberties and opportunities usually stems from the long history of conquest and domination they have suffered, governments should minimally respect historical agreements, allow as far as possible compensation for past oppression and, for the rest, not interfere in their internal affairs. The only remedial action that would be recommended is to provide dissident members (mostly the women) with appropriate exit rights¹¹ to leave their culture and enter the mainstream liberal culture, which generally affords them more opportunities to lead a flourishing and self-directed existence. In other words, governments should be morally required, not only to make sure that dissident members have real exit rights (otherwise state action could be legitimate), but also to make the cultural transferral the least costly possible, and in the most gentle, gradual, and supportive fashion, for instance with long-term, cross-generation, and facilitating programmes. I think this kind of exit rights protection taken on by the state is likely to meet most worries about the usual defence of exit rights.

One must keep in mind, however, that the above defence of exit rights is an autonomy-based, not a tolerance-based or pluralism-based defence. Cultures are prominently valued, not as contexts of identity, but as contexts of choice, that is, for their ability to provide equal opportunities and to protect the basic interests of the group's members in ethical agency and in well-being. Exit rights are envisioned here only as a means of providing dissident members with a viable alternative designed to compensate for the limitations of their culture as a context of choice, not as a means of tolerating non-liberal cultures. In other words, once governments have done their duty to respect historical agreements and allow compensation for past oppression, they should no longer be morally required to assent to the demands of those minority cultures whose paternalistic and authoritarian traditions offer too much resistance to change, no matter whether they are threatened or not. On the other hand, liberal states should have a moral duty, within their limited resources, to be receptive to the demands, not of purely liberal minorities (no culture is purely liberal), but of minorities with traditions and ways of life whose central parts meet people's basic interests in ethical agency and in well-being.

The policy towards non-liberal minority cultures supported by Kymlicka (1995, 163–170) also seeks to avoid the pitfalls of many autonomy-based approaches, naturally not by tolerating illiberal practices prevailing in indigenous or national minorities, but rather by inducing internal reforms, without yet endangering the societal autonomy of the communities within which those reforms ought to be carried out. In other words, according to such a policy, government's action should avoid both mere passivity and forcible intervention. Rather than mere inaction, or the recourse to force and compulsion, the policy suggested by Kymlicka is one of diplomacy, dialogue, persuasion and education.¹²

I see one main difficulty with such a solution. In exchange for helping non-liberal minority cultures preserve their societal autonomy through diverse funding programmes and collective rights, it allows governments to be key agents, along with dissident minorities, for cultural change within those cultures by initiating their conversion to liberal principles. In my view, such a solution risks falling into the paternalistic trap most common to current autonomy-based approaches. In effect, most members of such cultures will very likely view any incentive for liberalization of their way of life, whether financial or legal, as undue pressure exerted on their culture, as nothing but the expression of the majority culture's hidden cultural biases and ethnocentrism. Given the historical oppression experienced by many minority cultures, such incentives end up very often, contrary to their purpose, in the radicalization by the elite of the least liberal trends in their traditions (Weinstock, 2005). The autonomy-based defence of exit rights avoids this important shortcoming. Letting dissident members of non-liberal minority cultures decide by themselves whether to leave their original culture or not, and helping them, if they so decide, to transfer to the larger culture through exit rights guaranteed by the state, surely constitutes a solution that does not risk being too paternalistic or disrespectful towards minority cultures' traditions and way of life. It combines the awareness of the past oppression of non-liberal minority cultures with a central concern for the basic interests of the group's members in ethical agency and in well-being.

Conclusion

Culture has become a focal point of debate among political philosophers over the past 15 years, given the number and importance of claims to cultural recognition made by various minority groups in our pluralistic societies. The core issue remains, however, as to how an appropriate account of the importance and role culture has for individuals can be provided from a liberal point of view. In this paper, I have tried to offer what seems to me the most viable defence of the worth of cultures from a liberal point of view, which I have called the liberal defence of the intrinsic worth of cultures, while distinguishing it from the most currently discussed liberal defence, the instrumental one advanced by Will Kymlicka. While each of them recognizes individual autonomy as an intrinsically valuable good, both differ not only as to the role (causal or logical) culture has to play in a self-directed existence, but also as to why cultures deserve to be protected. For the intrinsic view, the first and primary reason for protecting a culture lies in its intrinsic value defined in terms of its ability to foster individual well-being and autonomy. Since, for the instrumental view, membership in a culture is only a means among others in order to make valuable choices, the reasons provided for protecting a culture are that it would be too costly for people to leave their culture and enter another, and that such a culture forms a context of identity that is worthwhile for them. While naturally taking these reasons into consideration, the intrinsic view considers them as only additional reasons for protecting a culture, no longer as core reasons intended to compensate for the shortfalls in making a convincing instrumental case for the protection of cultures. A final difference between the two liberal defences of the worth of cultures concerns their policies towards non-liberal, indigenous or national minority cultures. For the liberal defence provided by Kymlicka, liberal societies should shoulder the burden both of securing the societal structure of non-liberal minority cultures and initiating their conversion to liberal principles via internal reforms. The liberal defence provided in this paper leads rather to an autonomy-based defence of exit rights that recommends, in order to avoid the underlying paternalism of Kymlicka's approach, that liberal states focus primarily on demands for protection made by minorities already open to liberalization, and leave dissident members of non-liberal minority cultures free to decide by themselves whether to leave their original culture or not, while helping them, if they so decide, to transfer to the larger culture with a long-term, gradual, supportive and respectful process of integration.

Notes

¹ Réaume (2000, 246–247). I examine her criticism more thoroughly in part III of my paper.

² The position upheld by Habermas in his last essay on cultural claims (2005) is, at least at first glance, uneasy to grasp correctly. On the one hand, and unlike many liberals, he contends that culture has an intrinsic, non-instrumental value for its members (Habermas, 2005, 17). On the other hand, however, cultural rights should, according to him, not be restricted to access cultures of origin but rather give access to a variety of cultural goods and traditions (Habermas, 2005, 18), otherwise it would be tantamount to committing the essentialist fallacy and treating particular cultures as 'endangered species'. On this point, he explicitly endorses in a footnote Jeremy Waldron's view of culture. This is why I think the most coherent reading of Habermas's position is the one outlined in the present article, which I call the cosmopolitan reading.

³ I say 'probably' because the choice/circumstance distinction on which Kymlicka's theory of minority rights relies, mainly inspired by the luck egalitarian trend in liberal egalitarianism, could also be called into question. For such a criticism, see Quong (2006).

⁴ Kymlicka (1989, 176) makes no distinction between 'enforced' and 'facilitating' assimilation, which he combines. But in light of Réaume's convincing analysis, it is far from sure that such a distinction is really superfluous.

⁵ Another, and more straightforward way to respond to the cost-of-transfer argument would be to say that, since liberals typically base their defence of cultural rights upon respect for individual autonomy, if individuals choose to perpetuate their culture, these choices deserve respect, whatever the value, intrinsic or instrumental, one ascribes to culture. But this response only begs the question. The question is: On what grounds can people legitimately demand protection of their culture? It is not enough that people want to preserve their cultural heritage for it to be protected. The necessity of resources allocation to a minority culture must also be publicly demonstrated by governments. What if the protection of an endangered linguistic or cultural heritage turns out to be, in a context of scarce resources, too costly for the whole society? Should not the transfer of its adherents to the dominant language or culture be an alternative to contemplate? What if the prospect of leading a decent life for children is far greater in the mainstream society than in the threatened culture? In such situations, the will of people is naturally to be considered but only as one element among others, given that the choices and preferences people have in a society cannot all be satisfied and taken at face value. And among the issues to be pondered, the question as to whether language and culture have an intrinsic or merely an instrumental value is surely likely to be central.

⁶ This is the core of the criticism made by Margalit and Halbertal (1994).

⁷ It is worth noting, however, that Réaume makes no distinction between Kymlicka's and Raz's respective account of culture, which she mistakenly confounds and interprets as one and the same intrinsic account.

⁸ In effect, intelligence, sensitivity, extroversion and imagination might form the psychological profile of a perfect serial killer. These traits are valuable only insofar as they contribute to, or make possible an altruistic behaviour.

⁹ Naturally, if it appears that people no longer have any attachment to their culture and let it decay, there is no longer reason to protect it since such a culture simply becomes unsuited as a context of choice. One cannot 'oblige' people to maintain their culture simply because it provides a valuable context of choice. Contrary to value-collectivism for which a culture is valuable independently of its contribution to the well-being of people, the value-individualism espoused in this paper contends that culture has intrinsic value only when and insofar as it contributes to people's well-being and autonomy. If an autonomous life has ultimate value, then people must be free, not only to make choices within the range of options provided by their culture, but also, at a second level, to revise this range of options, to transform their cultural heritage and even ultimately to abandon their culture if they find it no longer worthy of allegiance. For this reason, it would be self-defeating to paternalistically oblige people to maintain against their will their culture on the grounds that it is deemed intrinsically valuable. As Kymlicka and Ronald Dworkin have emphasized on many occasions, an autonomous life cannot be led from the outside, in accordance with values or principles people do not endorse.

¹⁰ It could be objected that the intrinsic value of a culture does not constitute a 'sufficient' reason to protect it, any more than the importance of culture as a context of identity or the costs of integrating into another culture. As the example of the decaying culture shows, other issues (availibility of resources, social costs, common interest, viability, etc.) should also be pondered. This is undeniable. I nevertheless consider the intrinsic value of a culture as a sufficient reason to protect it from a moral point of view: it suffices that a culture provides its members with the enabling conditions for their personal flourishing, well-being and autonomy for it to deserve protection. By contrast, it is not sufficient, morally speaking, that a culture forms a lively context of identity and shared values, or that the transferral to another culture be too costly for its members, for such a culture to deserve protection. In effect, a culture could be oppressive, and the transferral costs for dissident members could be considerably alleviated. It remains, however, that even if one grants that the intrinsic value of a culture is a sufficient 'moral' reason to protect it, moral reasons are only one kind of reasons among many others (economical, political, etc.) that legislators should balance against one another when making a decision.

¹¹ I mean by 'appropriate exit rights' exit rights that satisfy minimal standards (freedom from physical abuse, a minimal education, a mainstream liberal society, etc.), which ensure that people have real right to exit. On this point, see Spinner-Halev (2005, 160).

¹² For a more explicit and thoroughgoing explanation of this policy, see Tan (2000, 32–34, 59–64).

References

1. Binder, G. (1993) 'The case for self-determination', Stanford Journal of International Law 29: 223–270.
2. Buchanan, A. (1991) Secession: The Legitimacy of Political Divorce, Boulder: Westview Press.
3. Buchanan, A. (1998) 'What's So Special About Nations?', in J. Couture, K. Nielsen and M. Seymour (eds.) Rethinking Nationalism, Calgary: University of Calgary Press, pp. 283–309.
4. Forst, R. (1997) 'Foundations of a theory of multicultural justice', Constellations 4(1): 63–71. | Article |
5. Galston, W. (1995) 'Two concepts of liberalism', Ethics 105: 516–534. | Article |
6. Galston, W. (2002) Liberal Pluralism: The Implications of Value Pluralism for Political Theory and Practice, Cambridge: Cambridge University Press.
7. Habermas, J. (1996) 'Kampf um Anerkennung im demokratischen Rechsstaat', in J. Habermas (ed.) Die Einbeziehung des Anderen. Studien zur politischen Theorie, Frankfurt am Main: Suhrkamp, pp. 237–276.
8. Habermas, J. (2005) 'Equal treatment of cultures and the limits of postmodern liberalism', The Journal of Political Philosophy 13(1): 1–28. | Article |
9. Hartney, M. (1995) 'Some Confusions Concerning Collective Rights', in W. Kymlicka (ed.) The Rights of Minority Cultures, Oxford: Oxford University Press, pp. 202–227.
10. Kukathas, C. (1992) 'Are there any cultural rights?' Political Theory 20: 105–130. | Article |
11. Kymlicka, W. (1989) Liberalism, Community and Culture, Oxford: Clarendon Press.
12. Kymlicka, W. (1995) Multicultural Citizenship, Oxford: Oxford University Press.
13. Kymlicka, W. (2001) Politics in the Vernacular. Nationalism, Multiculturalism, and Citizenship, Oxford: Oxford University Press.
14. Margalit, A. and Halbertal, M. (1994) 'Liberalism and the right to culture', Social Research 61(3): 491–510.
15. Okin, S.M. (1999) 'Is Multiculturalism Bad for Women?', in J. Cohen, M. Howard and M.C. Nussbaum (eds.) Is Multiculturalism Bad for Women?, Princeton: Princeton University Press, pp. 9–24.
16. Patten, A. (2003) 'Can the Immigrant/National Minority Dichotomy be Defended?: Comment on Ruth Rubio-Marin', in S. Macedo and A. Buchanan (eds.) Secession and Self-Determination, New York: New York University Press, pp. 174–189.
17. Quong, J. (2006) 'Cultural exemptions, expensive tastes, and equal opportunities', Journal of Applied Philosophy 23(1): 53–71. | Article |
18. Raz, J. (1986) The Morality of Freedom, Oxford: Clarendon Press.
19. Raz, J. (1994) 'Multiculturalism: A Liberal Perspective', in J. Raz (ed.) Ethics in the Public Domain. Essays in the Morality of Law and Politics, Oxford: Clarendon Press, pp. 170–191.
20. Réaume, D. (2000) 'Official Language Rights: Intrinsic Value and the Protection of Difference', in W. Kymlicka and W. Norman (eds.) Citizenship in Diverse Societies, Oxford: Oxford University Press, pp. 245–272.
21. Reitman, O. (2005) 'On Exit', in A. Eisenberg and J. Spinner-Halev (eds.) Minorities within Minorities. Equality, Rights, and Diversity, Cambridge: Cambridge University Press, pp. 189–208.
22. Song, S. (2006) 'Justice as evenhandedness: an egalitarian argument for multiculturalism', Paper prepared for the Spring 2006 Political Philosophy Colloquium; University of Wisconsin-Madison (unpublished manuscript).
23. Spinner-Halev, J. (2005) 'Autonomy, Association, and Pluralism', in A. Eisenberg and J. Spinner-Halev (eds.) Minorities Within Minorities. Equality, Rights, and Diversity, Cambridge: Cambridge University Press, pp. 157–171.
24. Tan, K.-C. (2000) Toleration, Diversity, and Global Justice, Pennsylvania: The Pennsylvania State University Press.
25. Waldron, J. (1995) 'Minority Cultures and the Cosmopolitan Alternative', in W. Kymlicka (ed.) The Rights of Minority Cultures, Oxford: Oxford University Press, pp. 93–119.
26. Weinstock, D. (2005) 'Beyond Exit Rights: Reframing the Debate', in A. Eisenberg and J. Spinner-Halev (eds.) Minorities within Minorities. Equality, Rights, and Diversity, Cambridge: Cambridge University Press, pp. 227–246.