Corps de l’article

Liberal culturalism is the now-familiar view that some minority cultural rights are requirements of liberal justice (Patten, 2014, p. 3, citing Kymlicka, 2001, chap. 2). The standard arguments for this position are subject to some powerful objections, such as that they assume an essentialist view of culture, and that they only justify a right to some culture not necessarily one’s own. Alan Patten’s Equal Recognition restates the case for liberal culturalism so as to avoid these and other objections, solidifying “the moral foundations of minority rights.” The objection that will be the focus on my comments concerns the idea of neutrality.

The liberal commitment to state neutrality originally seemed to undermine both majority nationalism and minority cultural rights. Citizens are free to pursue their cultural objectives within the framework of rules and institutions that establish fair terms of cooperation, but not to use the power of the state to advance religious or cultural goals shared by some but not all. One response was to reject neutrality, insisting that a state could be respectably liberal so long as its pursuit of less-than-universally shared goals took place within the limits set by basic individual liberties (Taylor, 1993, pp. 175-177). Will Kymlicka’s innovation was to argue that minority cultural rights follow from liberal principles themselves. If liberty means autonomy rather than non-interference, it requires an adequate array of choices. Culture is essential to this range of choice, and members of minority cultures can be at an unchosen disadvantage with respect to the reproduction of their culture. The existence of brute-luck threats to the conditions of autonomy does not justify a right to one’s own culture, however, because the solution could be subsidy for transition to the majority culture (Patten, 2014, p. 6).

This flaw in the argument from autonomy puts additional weight on a second liberal argument for minority cultural rights, which is that the state cannot avoid certain kinds of local non-neutrality. The state cannot be neutral with respect to language and culture in the same way it can with respect to religion, because it must conduct its business in some relatively small number of official languages, recognizing only some holidays, within certain geographical boundaries. Some minority cultural rights may therefore be called for as a matter of fairness, realizing cultural neutrality overall, though not with respect to each particular policy. This second liberal argument for minority rights (from unavoidable local non-neutrality) is open to the objection that it assumes neutrality of effects, while liberalism is committed only to neutrality of justification; policies that have differential impact may be legitimate if they are justifiable on neutral grounds (Patten, 2014, p. 8). In chapter 4, therefore, Patten reformulates the idea of liberal neutrality so as to salvage the argument from local non-neutrality. He argues that liberalism is not based on neutrality of justification or neutrality of effect but neutrality of treatment. Neutrality of treatment is a “downstream” conception of neutrality, in the sense that it is not a foundational principle, but a consequence of a more fundamental non-neutral commitment, which Patten calls “fair opportunity for self-determination” (Patten, 2014, pp. 108-109). The purpose of my comment is to get a better grasp on what neutrality of treatment involves, and what its relationship is with neutrality of justification and neutrality of effect. My main question is whether Patten is committed to a more foundational upstream conception of neutrality than he admits.

Neutrality of justification counts as neutral the establishment of a religion on grounds of maintaining social peace, because avoiding bloodshed is a neutral concern; neutrality of effect counts as non-neutral the legal protection of basic liberties, because such protection makes it harder for boring ways of life to flourish (Patten, 2014, pp. 112-114). The idea of neutrality of treatment is meant to avoid these counterintuitive implications. Even if it is publicly justifiable, religious establishment involves treating members of different religions differently. Conversely, the state treats everyone the same in protecting their basic liberties, even if the use people make of these liberties gives rise to unequal outcomes. To illustrate the distinction between neutrality of effect and neutrality of treatment, Patten gives the example of a philanthropist faced with a choice between donating to two worthy projects. I could give each project that amount of money it needs to succeed, or I could simply give each project the same amount of money. The same example can be used to illustrate the contrast between neutrality of treatment and neutrality of justification. Instead of giving each charity the same amount of money, I could decide how much to give based on reasons accepted by both charities.

I want to begin by raising some questions about the dilemma neutrality of treatment is meant to solve. Consider the first horn, about neutrality of justification and religious establishment justified by social peace. The problem is that neutrality of justification is overinclusive, which suggests that neutrality of justification is not enough, not that it is unnecessary or pernicious. However, since neutrality of treatment is a downstream approach, it seems that it must eschew neutrality of justification. Neutrality of treatment is justified by a “particular, justifiable, liberal value”—namely, fair opportunity for self-determination; Patten says that his account is “not embarrassed” that this value is “quite particular and nonneutral” (Patten, 2014, p. 109). If Patten is not committed to neutrality of justification, then in developing the alternative conception of neutral treatment, he is not articulating an additional requirement for policies to count as neutral, but an alternative to the requirement of neutrality in justification. The rejection of neutrality of justification might open the door to perfectionism, or a comprehensive/non-political approach to justice (meaning one that claims correctness rather than multiperspectival acceptability), an issue I’ll return to below. If Patten were committed to neutrality of justification, in the form of a Rawlsian principle of public reason, the fact that neutrally justified policies can seem non-neutral would present no puzzle. We simply need to distinguish the principle of public reason from specific public reasons, such as equal treatment of religions, and recognize that the balance of public reasons may favour a policy that has unequal impact.

Turn now to the second horn of the dilemma, concerning neutrality of effect’s failing to count legal protection of basic liberties as neutral. The objection works if neutrality of effect is interpreted as involving a guarantee of equal outcomes regardless of the choices people make, but no one has ever believed in that. Moreover, the idea of the effect of an action or policy is normally understood in comparative or counterfactual terms, as the difference that the policy makes relative to some alternate course of action. If there is more than one alternate policy, there is more than one effect, and if the effects of a policy on different ways of life are equal compared to one alternate policy, they won’t in general be equal compared to others. To make the basic-liberties criticism stick, Patten has to define the inequality as unequal impact relative to a particular alternative policy: locking people into their existing conceptions of the good (Patten, 2014, p. 114). Yet every policy has unequal effect relative to some alternate policy. As Robert Nozick argued, it is irrelevant that the law against rape has a differential impact on men and women as compared to a state of nature, because the law in question is independently justified.[1] Similarly, laws against assault benefit the weak more than they do the strong, but that does not make them non-neutral. The fact that legal protection of basic liberties has unequal impact relative to some alternate policy is irrelevant unless that alternate policy is the appropriate baseline. The conclusion one might draw is not that liberals are uninterested in neutrality with respect to effects, but that they are interested in neutrality relative to the appropriate baseline, a property that is in any case shared by neutrality of treatment. Although the example of the philanthropist makes it seem as if the assessment of neutrality is non-comparative, Patten says that the state violates neutrality of treatment “when, relative to an appropriate baseline, its policies are more accommodating of some conceptions of the good than they are of others” (Patten, 2014, p. 115). Neutrality of treatment is not meant to be an effects-based conception of neutrality, so what role is the idea of a baseline playing here?

Patten identifies two further problems with neutrality of effect, apart from the alleged non-neutrality of basic liberal rights, but I don’t think they bring us beyond neutrality of effect so much as they specify which effects are of concern. The first is that the effects in question might be understood as total effects, over the long term, including effects that arise by way of the different responses people freely make to the policies in question. The second is that even if we limit our attention to direct effects, the size of these effects might depend on background factors that are not a matter of public responsibility—e.g., the case of opening up a field for soccer and cricket, where field availability is not a binding constraint for the cricket players because there are so few of them that they can’t play anyway (Patten, 2014, pp. 115-117). These points specify which effects are of interest—roughly speaking, direct effects controlling for the right background variables, relative to the right baseline package of policies.

So what is the appropriate baseline? Patten denies that it should be “no policy”—i.e., state inaction, a “do-nothing point” (Patten, 2014, p. 118). Nozick’s rape example explains why; differential impact relative to a baseline in which people are free to violate each other’s rights is not a bad thing. Instead, Patten favours “fair opportunity for self-determination” (Patten, 2014, p. 118). Yet fair opportunity for self-determination is also the value that grounds neutrality (Patten, 2014, p. 109). Thus it seems that fair opportunity justifies equal treatment relative to the baseline policies required by fair opportunity for self-determination. Does this mean that we could dispense with the idea of neutrality, employing only the idea of equal treatment relative to the baseline policies required by fair opportunity for self-determination?

I think the answer must be “no,” because questions of neutrality can be raised with respect to the baseline itself. Consider the legal protection of basic liberties, one of which is the right to vote and run for office. The choice of official languages will advantage some people and disadvantage others, with respect to their exercise of political liberty, and their pursuit of their way of life more generally. The problem here is that many of the policies and institutions that constitute the baseline (the package of policies that is the benchmark for assessing differential impact of all other policies) must have a particular cultural “format” (Patten, 2014, pp. 169-170). Some elements of the baseline institutions required by justice will therefore be unequally accommodating of cultures. As a result, some minority cultural rights are justified as a way of attaining neutrality overall.

How are we to determine whether a particular component of the baseline package of policies is unequally accommodating of different conceptions of the good? What is the metric of accommodation? One possibility is to think of the various possible packages of policies as arrayed in a space, and to assume that each culture or conception of the good prefers some ideal point in this space. Neutrality would then consist in maintaining equidistance between ideal points, or, where there are more than two conceptions and hence more than two ideal points, minimizing the total distance between the baseline package of policies and the various ideal points. Neutrality in this sense (splitting the difference between policy preferences) cannot be the main criterion for selecting the baseline package of policies, however, for the reasons discussed above in relation to Nozick. Some conceptions of the good are inhospitable to freedom of religion. Others are inhospitable to gender equality. Policies protecting people’s basic rights and liberties will not be equidistant from all conceptions of the good, nor should they be.

The alternative is to think of equal accommodation as a matter of substantively equal treatment. In the case of the philanthropist, it initially seems obvious that giving each project the same amount should count as neutral treatment, because each party is getting the same amount; neutral treatment means treating people the same. However, there are lots of familiar cases in which treating groups the same does not count as equal treatment, meaning treatment that shows equal concern and respect: people with and without disabilities; children vs. adults; rich vs. poor. Treating these different groups the same might constitute unequal treatment, if not a violation of their rights. Suppose that I, the philanthropist, have to choose between donating to two elementary school associations to cover costs of extracurricular activities. One school is located in a wealthy neighbourhood and will have no trouble raising a lot of money. The other is located in a poor neighbourhood and will struggle to raise any. Treating these two groups the same might not count as substantively equal treatment. We presumably need a theory of justice that tells us which equalities are required by justice, and which are not.

In sum, although policies required by justice should not be equally accommodating of conceptions of the good that conflict with justice, equal accommodation of conceptions of the good that are compatible with justice is pro tanto a good thing. Justice does not in general require that a country have an official language; it simply requires political liberties, which must be exercised in writing and speech. Due to current constraints on human cognition, technology, and resources, this communication must take place in some small number of languages, within courts and parliaments. The choice of an official language is therefore non-neutral where justice does not require non-neutrality; that is to say, it is non-neutral between conceptions of the good that are fully justice-compatible. Other things equal, that’s a bad thing, and since it can’t be avoided in this dimension of policy, it may merit some form of compensation or accommodation for disadvantaged cultures in other dimensions of policy.

A further question that arises at this point is whether the conception of justice that sets the baseline for determining what counts as equal treatment must be political in Rawls’s sense of being acceptable to all reasonable comprehensive doctrines, or if it can simply be true, despite being reasonably contestable. I see at least two ways of interpreting Patten’s rejection of upstream neutrality (Patten, 2014, p. 109). The first involves rejecting justificatory neutrality across all views, while holding on to justificatory neutrality between an appropriate subset. Fair opportunity for self-determination is not and need not be neutral among all moral views; it is simply true. However, fair opportunity for self-determination might require neutrality of justification among all views accepting this value, and the resulting requirement of neutrality in justification. The model here might be David Estlund’s defense of a qualified acceptability requirement, which he claims must be defended as true (against those who have different conceptions of the right standard of qualification); all other reasons invoked in political justification need only be acceptable to qualified points of view.[2] However, it is possible that Patten means to reject any requirement of neutrality in justification. The model here would be Ronald Dworkin’s use of the “endorsement constraint” to yield a kind of practical neutrality out of a form of argument that involves no attempt to avoid disagreement or bracket controversial claims (Dworkin, 1983, pp. 25-30).[3] The endorsement constraint is a generalization of the view that, because belief is not subject to the will, forced worship is pointless. People’s lives don’t go better when engaged in valuable activities unless the people in question recognize this value. Thus, even though everyone’s fundamental interest is in leading a truly good life, and politics should promote the leading of better lives, the state should be neutral between conceptions of the good, at least with respect to its coercive polices.

Given that Patten is appealing to a “distinctly liberal” value called “self-determination,” one might think that the underlying consideration has something to do with the conditions necessary for individuals to reflect about what is valuable, and to form, adjust, or revise their plans of life on this basis. One could be self-determining, in this sense, even if one did not fully realize one’s conception of the good (perhaps because it is a conception that is very hard to realize, such as making a great work of art). Alternately, one could fulfil one’s conception without engaging in reflection about what is valuable, as in the case of someone who unquestioningly leads a fulfilling family life in a society in which powerful social norms encourage conformity with traditional values. However, Patten says that people’s interest in self-determination is “their interest in being able to pursue and fulfill the conception of the good that they, in fact, happen to hold” (Patten, 2014, p. 125). In other words, “self-determination” means fulfilment of one’s conception of the good, whatever it is (though it can’t be worthless (Patten, 2014, p. 131), and it has to be consistent with everyone having a fair opportunity for self-determination (Patten, 2014, p. 109)). Thus, despite the connotations that self-determination has, Patten is not attributing any particular importance to the capacity for revision and reflection. This impression is supported by Patten’s sceptical comments about the intrinsic importance of autonomy (Patten, 2014, p. 132), and by his account of the connection between self-determination and well-being, which makes use of the endorsement constraint (Patten, 2014, p. 131). The endorsement constraint does not identify a value but a feasibility constraint, which is that without buy-in even objectively valuable activities don’t make a person better off. Patten’s conception of self-determination is thus quite minimal, in the sense of appealing to relatively uncontroversial normative premises.

However, if there really is no upstream constraint on the reasons we can appeal to, it is open to perfectionists to concede that self-determination in Patten’s sense is important, but to insist that it is not all that matters, and therefore to respecify the baseline of comparison for determining what counts as equal treatment. I might be committed to equal treatment of conceptions of the good relative to a baseline of just institutions where justice is defined according to a particular comprehensive doctrine. That is to say, I may accept the pro tanto value of equal treatment defined relative to a background of just institutions, but define justice according to (not so as to promote) a particular comprehensive doctrine—fair opportunity to flourish, according to what I take to be the correct conception of the good.

Patten argues that a perfectionist definition of the baseline would undermine self-determination (Patten, 2014, p. 147). The situation Patten has in mind is one in which the state adopts a set of policies intended to promote superior conceptions of the good, then measures inequality of direct impact relative to this baseline (e.g., it builds promotion of art and culture into the baseline, but is still able to condemn the privileging of hockey over basketball). I think Patten is attacking the weakest form of perfectionism here. The perfectionist is taken to be someone whose fundamental principle is “promote flourishing,” and who defines basic rights and liberties and settles other matters of justice based on this goal. The perfectionist subordinates justice to maximizing excellence, which seems obviously wrong. A more plausible version of perfectionism would recognize that respect for people’s rights is an independent moral value, distinct from the goal of promoting well-being, excellence, or achievement, but would maintain that the identification of the rights people have as a matter of justice depends on what the truth about human flourishing is (Wall, 1998, p. 12). Such a conception of justice need not be committed to equalizing well-being regardless of the choices individuals make. For example, assume that the view in question is that interesting, meaningful work is an important component of a good life, much more important than wealth. An opportunity-focused perfectionist would conclude that we need to assess the economic system partly in terms of the distribution of opportunities for meaningful work across different social positions. If institutions that are just according to this metric are in place, and I nonetheless choose to focus on attaining wealth rather than meaningful work, there is no cause for the state to try push me into a better way of life, nor to compensate me for my lack of true well-being.