THE PRAGMATIC READING

At the core of Rawls' (1996, 1999b) political liberalism is his notion of public reason. Specifically, public reason is defined as citizens' concern to give reasons to one another that respect the substantive values manifested in each other's status as free and equal. Citizens who reason this way recognize a principle of reciprocity, giving the same types of reasons to others that they can expect in return, in virtue of their status as citizens. Public reason instantiates these ideals by requiring citizens to abstract from certain aspects of their moral identities, including what Rawls (1996, 175) labels their “comprehensive conceptions,” or notions of the good. When representatives deliberate using public reason, they take the perspective of all members of society but no member in particular. When citizens give reasons that recognize each other's status as autonomous and equal, they engage in reason-giving that respects the principle of reciprocity. I call this type of reason-giving “mutual justification.”

At points in Political Liberalism, Rawls (1996) seems to maintain that when citizens reason, they do so in a manner that is abstracted from their moral identity. Rawls distinguishes between the characteristics possessed by parties in the original position and the characteristics of actual people. To those who attempt to read a theory of the person into the original position, Rawls replies, “When in this way, we simulate being in the original position, our reasoning no more commits us to a particular metaphysical doctrine about the nature of the self than our acting a part in a play, say of Hamlet or Lady Macbeth, commits us to thinking that we are really a king or a queen engaged in a desperate struggle for political power” (1996, 27). Furthermore, Rawls is concerned to distinguish a purely “political” conception of the person or of the citizen from what he calls “noninstitutional or moral” identity (1996, 30). He is clear that individuals could undergo significant religious or moral transformations of belief and yet have their political identity as citizens unaffected. Rawls writes, “On the road to Damascus Saul of Tarsus becomes Paul the Apostle. Yet such a conversion implies no change in our public or institutional identity” (1996, 31).

These remarks, among others, can plausibly be interpreted to support the pragmatic reading offered by some feminist critics of Rawls. Specifically, because Rawls begins with the fact of pluralism among comprehensive conceptions, it seems that he posits a nonpolitical private sphere untouched by normative inquiry as a starting point in theorizing. On this account the public sphere of citizens serves to mediate between prior moral identities but does not define these identities. Citizenship here seems as though it is subservient to comprehensive conceptions and as a result is powerless to reform sexist power imbalances that are the result of some of these same identities.

In an article that exemplifies the pragmatic reading, John Exdell (1994) argues that Rawls' theory is powerless to effect egalitarian change within the private sphere as it is commonly understood in current society because of a more general move toward relativism at the heart of political liberalism. As he reads it, Rawls' Political Liberalism is characterized by an attempt to find uncontroversial premises as the basis for public reason defined in terms of current public opinion. As Exdell sees it, Rawls therefore must rule out disputed religious views as well as feminist accounts of politics (Exdell 1994, 495). Because public reason is informed by status quo views of what is reasonable, and this excludes controversial views, Exdell claims that Rawls is unable to criticize accepted inequalities within family life based on a feminist analysis. For Rawls, public reasons apply primarily to the “basic structure” of society, defined as its main “political, social, and economic” institutions (1996, 11). Although Rawls claims to include the family in the basic structure, on the pragmatic reading Rawls ultimately fails to follow through on this claim. Rather, because his theory seeks consensus in a conventional understanding of the political he is led to ignore the importance of political dynamics in what are commonly believed to be private spaces.

Given the pragmatic reading, Susan Okin (1994) seems to stand on solid ground when she accuses Rawls of falling victim to the “double life” of liberalism famously attacked by Marx in “On the Jewish Question.” Although persons qua citizens who engage in public reason need not take on the characteristics of those behind the veil of ignorance, for Rawls persons' public identities are distinct from the characteristics of actual people. Arguably, it follows that abstractly free and equal citizens might have opposite characteristics when it comes to the actual people they represent. In other words, Rawls posits a dichotomy between persons as private individuals and citizens qua political persons. In politics, citizens must interact as if all were equal, although the same need not be true in family life, for instance. As a result, according to Okin, Rawlsian politics leaves private life unaffected. It does women little good to have equal representation as citizens if in their actual life they are the victims of domestic inequality, which cannot but affect their political equality or inequality. I interpret Okin to mean not only that women subject to domestic abuse probably will not participate actively as public citizens, but also that more generally that their domestic inequality is a subset of their general political inequality. Formal political and legal equality does not mean substantive equality for women in actual life.

In an article on political liberalism, Okin argues that Rawls not only allows for a split between the public and the private realms, but that his concept of liberalism leaves the latter unaffected even in cases of extreme injustice. Okin writes: “Rawls also strongly implies that in the nonpolitical aspects of [citizens'] lives—personal morality or religion for example—they may hold views such as that there is a fixed natural order or a ‘hierarchy justified by religious or aristocratic values’” (Okin 1994, 28–29). She also quotes an earlier passage in which Rawls discusses the fact that his “political conception of justice” is freestanding from comprehensive conceptions of the good and says that as a result, he “does not deny there being other values that apply, say to the personal, the familial, and the associational” (Rawls quoted in Okin 1994, 29).

According to Okin, this passage is “reminiscent of those aspects of liberal theory that Marx criticized for splitting into ‘abstract citizens’ and ‘human beings’” (Okin 1994, 29). In her view, the passage from Rawls is one of many demonstrations of the fact that his claim that the political conception is “freestanding” means that it neither affects the private realm nor is affected by it. Despite the fact that Rawls relies on a family he assumes to be just in his account of moral development in Part Three of A Theory of Justice, ultimately the family is neglected by Rawls. On the pragmatic reading, because Rawls' conception of citizenship is solely public, it does not penetrate the moral identity of private persons or deal with fundamental injustices in the family. Okin quotes Rawls to illustrate this point: “The political is distinct… from the personal and the familial, which are affectional… in ways the political' is not” (Rawls quoted in Okin 1994, 26).

The division between actual person and citizen is a problem for Okin because many real inequalities exist in the private realm. For example, Okin asks how real people can view themselves as equals if they come from families that have rigidly sexist hierarchies. Despite any formal equality in the public realm, a real individual will not see herself as an equal if she is raised in a family where she is not treated as one. In Okin's view, by leaving public reason out of family relationships, Rawls renders his own theory powerless to create real equality. Rawlsian politics, therefore, remains an abstraction that does not affect the real characteristics of real people (Okin 1989, 1994).¹

If Rawls were merely responding to the problem of toleration among bearers of different comprehensive conceptions, as the pragmatic reading suggests, then he would have little to say to monolithic societies whose views allow for the domination of women. If no person in a society objected to women's subjugation, then there would be no reason to seek a change in that society. Furthermore, if the notion of the reasonable does not affect actual identity or comprehensive doctrines except to the extent that they challenge specifically public conceptions of justice, then Rawls might allow inequalities within distinctly nonpolitical realms. The family is the paradigmatic example of such a nonpolitical realm.

THE STRONG PUBLIC RECONSTRUCTION

I take Okin's critique to be among the foremost challenges for political liberalism. Answering it requires re-reading Rawls and indeed abandoning the pragmatic reading, which his text often seems to support. When we read them closely, the passages that Okin cites can be reinterpreted to support the strong public reconstruction. Reading Rawls this way would clarify how Okin's objections could be incorporated into political liberalism and Rawls' theory could ultimately be saved from the attack on abstract citizenship. On the strong public reconstruction, free and equal citizenship and public reason are not devices for avoiding public condemnations of private inequality. Rather, free and equal citizenship and public reason are the ideal standards, which should determine the boundaries between the public and the private spheres in the first place. When privacy is understood with reference to these ideals, it does not protect domestic inequality.

Consider the full version of the first passage Okin quotes about private hierarchy existing side by side with public equality:

Rawls' distinction between these two types of reasoning does not mean he need endorse a situation in which illiberal private doctrines merely coexist with liberal public ones. In fact, Rawls claims that within the realm of political thought and discussion, when public reasons and comprehensive doctrines conflict, the latter must give way to the former; in other words, because citizens engaged in political discussion must be able to envision alternatives to the social status quo, they cannot be paralyzed by a view of the current social order as “fixed” or as a “hierarchy justified by [their] religious or aristocratic values.” One implication of this way of resolving such conflicts is that in liberal societies comprehensive doctrines themselves should become more liberal when they encounter the notion of public reason. Specifically, the concept of public reason necessitates that part of liberal citizens' moral identity must be as public reason givers, even if this conflicts with what might seem subjectively more important convictions. I contend that this tension should invoke a transformation in moral identity for the person with an illiberal disposition and an illiberal comprehensive doctrine.

In his chapter “The Overlapping Consensus,” Rawls (1996) presents some evidence for the strong public reconstruction. Going a step farther than he does in the passage above, Rawls suggests there that public reason must always trump private ways of reasoning when they conflict. He describes the process by which a community moves from what he calls a modus vivendi to an overlapping consensus, which is necessary for a society to be just. The modus vivendi in his conception is a form of truce, in which for pragmatic reasons political actors put their genuine differences aside and agree to live in peace together. Here the bonds of the community are entirely based on self-interest, and if all parties are no longer benefited, the community is likely to come apart. In an overlapping consensus, by contrast, “a balance of reasons as seen within each citizen's comprehensive doctrine and not a compromise compelled by circumstances, is the basis of citizens' respect for the limits of public reason” (1996, 169).

The fact that these reasons come from within citizens' comprehensive doctrines might make it seem as if Rawls would leave these doctrines in place. What happens, however, if an aspect of one's comprehensive doctrine is incompatible with public reason? Rawls answers that to the extent that it is incompatible, it is unreasonable. At this point, Rawls' answer might be interpreted as merely definitional: a doctrine is unreasonable to the extent that it does not exhibit public reason. On this account, the reasonable would be merely a system of classification with no normative implications. As a result, it might be argued that Rawls is merely distinguishing between reasonable and non-reasonable doctrines, and thus it would be correct to say that Rawls does not privilege public reason over comprehensive doctrines.

On my reading of Rawls, however, both the overlapping consensus and public reason are ideals necessary for the just society. He argues that reasonableness is not only a desirable characteristic of a society but also essential to justice. Unreasonable comprehensive doctrines, therefore, must change to include actual public reasons. Rawls is clear that although these reasons are often linked to comprehensive doctrines, they also must exhibit the independent characteristic of being addressed solely to free and equal citizens. For instance, it is not enough for an individual to turn to the Bible to defend civil rights laws. Rather, one must derive one's arguments for civil rights from public reason. But the claim that public reason is the reason of citizens does not entail, as the pragmatic reading advanced by Exdell suggests, solely looking to commonly accepted principles found in the culture. Rather, Rawls argues that it necessitates looking to what citizens, understood as “part of a public ideal [my emphasis] of democratic citizenship,” would accept.²

The pragmatic reading defended by Exdell and Okin is flawed because it fails to recognize that Rawls subjects society to ideal moral principles. Several of Rawls' terms might, however, lead readers to believe that he prioritizes the empirical facts of a given society over moral principles. Rawls claims, for instance, that open societies are “marked” by the “fact of reasonable pluralism” (Cohen 1993; Rawls 1996, 63–64). At first glance this might seem to rest his theory in contingent empirical facts about a given culture. But Rawls is clear that there is a distinction between a nonprincipled “fact of pluralism” and a principled “fact of reasonable [my emphasis] pluralism” that is relevant to political liberalism (1996, 63).

In addition, Rawls' claim that he “hopes to articulate a public basis of justification … working from the fundamental intuitive ideas implicit in the public political culture” might seem to rest his project on notions present in existing empirical societies (1996, 192). But for Rawls, beliefs in the public political culture are a mere starting point in thinking about public justification. Later on, in that same paragraph, he is clear that intuitions need to be tested by “abstracting” from one's particular comprehensive view and cultural standpoint in order to ask whether these intuitions conflict with public reason. Proponents of the pragmatic reading might draw from those passages where Rawls indicates that his theory only requires such beliefs to be translated into the form of public reason to emphasize that public reason is not transformative; they could argue that because translation is a minimal requirement, it would not necessarily change the content of the reasons (Rawls 1999a, 616). On the strong public reconstruction, however, the test for the reasonableness of comprehensive doctrines is substantive and not merely formal. Plenty of conceptions that violate the ideal of public reason can be expressed within the form of public reason. One might, for instance, try to justify abolition of basic liberties or the institution of theocracy, which clearly violate Rawlsian ideals, while relying on the rhetoric of freedom and equality. However, nonjustifiable principles expressed in the language of public reason are still nonjustifiable.

One possible source of ambiguity in the conflict between the pragmatic and strong public understandings of the political conception of justice lies in Rawls' idea of the political conception as “freestanding.” Freestandingness is characterized by its not entailing any “specific metaphysical or epistemological doctrine beyond what is implied by the political conception itself” (Rawls 1996, 10). However, Rawls is also clear that a political conception can still be compatible with some comprehensive doctrines, which can serve to support it. About this there is no dispute between the pragmatic reading and the strong public reconstruction.

The conflict between the pragmatic reading and strong public reconstruction becomes evident on examination of whether comprehensive doctrines or the freestanding political conception have priority. The freestanding political conception can be understood as a normative standard for evaluating comprehensive doctrines. On the strong public reconstruction, doctrines conflict with the freestanding ideal when they thwart the realization of the political conception of the person. Thus, to evaluate comprehensive doctrines, we should work backwards from the political conception. When comprehensive conceptions conflict with a political conception, citizens should aim to transform them.

The freestandingness of the political conception of justice and public reason requires that citizens have the capacity to abstract from aspects of their comprehensive doctrine to inquire into whether that doctrine is mutually justifiable. Here, public reason, which is itself freestanding because it is a component of the broader political conception of justice, is central. Regardless of the way citizens talk about politics in their daily lives, ideally they would be able to use public reason as an independent test for their belief's compatibility with the freestanding political conception of justice. Such ability requires that citizens have some critical distance from aspects of the particular culture that they find themselves in and a capacity for reasonable judgment that can serve as the basis for a critique of citizens' own culture. Certainly, nothing about this need for abstraction requires abandoning comprehensive doctrines altogether. For example, Martin Luther King's “Letter from a Birmingham Jail” interweaves religious and nonreligious arguments, both of which reinforce the ideals of free and equal citizenship. Although decisions about state policy should not depend on exclusively religious arguments, these arguments can play a powerful role in supporting public reason. In cases unlike King's, however, in which the freestanding political conception of justice is not built into a comprehensive doctrine, the requirement that citizens abstract from aspects of their moral identity through the use of public reason will result in a dialectic of transformation. Here the political conception of justice requires that citizens endorse the overlapping consensus; as a result, their comprehensive doctrines must evolve to fit the demands of public reason.

Before I go on to explore how the strong public reconstruction addresses some of Okin's concerns, I want to point to more textual evidence in support of the strong public reconstruction. In his reply to Habermas, Rawls clarifies that the overlapping consensus is an ideal versus an “overlap that is already present or latent” (Rawls 1996, 389). The idea here is that “the political conception of justice is worked out first as a freestanding view that can be justified pro tanto without looking to, or trying to fit, or even knowing what are, the existing comprehensive doctrines” (1996, 389). Rawls then suggests that “When the political conception meets these conditions and is also complete, we hope the reasonable comprehensive doctrines affirmed by reasonable citizens in society can support it, and that in fact it will have the capacity to shape those doctrines toward itself.” It is clearly the normative attractiveness of the political construction which leads Rawls to “hope” that the position in actual societies conforms to the normative ideal. This suggests that in endorsing the overlapping consensus, citizens have to work out how to incorporate the requirements of the freestanding political conception as an ideal.

One could object to an account of political liberalism that requires transformation when conflicts arise between comprehensive doctrines and freestanding political conceptions on the grounds that such a view would not ensure stability in the overlapping consensus. But as is the case with the overlapping consensus itself, Rawls helpfully distinguishes between empirical and normative understandings. He differentiates between stability, as it exists for instance in a modus vivendi, and “stability for the right reasons” (Rawls 1996, 392). Ideally regimes will be both stable in the sense that there are no empirical conflicts and also stable in their reliance on good public reasons. However, to achieve this ideal state of affairs, some empirical instability, as might be required by the transformation of comprehensive doctrines, is justifiable.

The process of transformation can happen in two distinct ways. First, citizens can begin by holding their comprehensive doctrine and their capacity for reasonable judgment as distinct conceptions. Indeed, individuals' private beliefs could remain incompatible with public reason as long as their public deliberations and politically relevant decisions were guided by the principle of public reason. This would require the cognitive dissonance of a scientist who publicly followed scientific method as a matter of profession but privately doubted the validity of the scientific method. Yet, to become publicly justifiable, private beliefs that clash with public reason must largely remain at the level of conscience; they should not influence publicly relevant action, including, as I will argue, certain actions within the family. Indeed, to be publicly justifiable, these beliefs should not be communicated to children, for instance, for fear of affecting them.

Although it may be a conceptual possibility for a person to live a publicly justifiable double life of private person and public citizen, some commentators have argued that this position would be quite difficult to maintain psychologically. This double life would deny the individual a coherent moral identity and would require a significant divide between the public beliefs that guide action and private beliefs inconsistent with those actions (Wenar 1995, 32).³

Both accounts of transformation resulting from conflicts between public reason and private comprehensive conceptions have important implications about the effect of public reason on moral identity. At least in the public realm, citizens must reason in a very specific way that is incompatible with many comprehensive doctrines. Fundamentalists, who believe that all political reasoning should be based solely on God's will, must be able to bracket their beliefs when they are discussing politics. In some realms they must embrace a way of thinking contrary to the notion that religion, exclusive of public reason, answers all questions. Therefore, even on the first conception of transformation, fundamentalist moral identities will be transformed. Persons with strongly faith-based or text-based religious beliefs who are unable to stand back from their conceptions of the good should learn to do so. It is important to note that although a political conception of justice requires personal moral transformation, it does not give the state license directly to coerce citizens into such a process. State action of this variety would violate the freedom of conscience and privacy rights also required by public reason. This ideal of transformation thus has important implications for the duties of citizens; it does not call for coercive law.

Does the notion of the reasonable thus risk transforming certain cultures? Does it interfere with religious beliefs and family structures often thought to be private? It is true that Rawls thinks public reason can be “freestanding,” and that one need not have a comprehensive conception that agrees with it. This is the case in the first process of transformation. However, Rawls by no means indicates that all identities must be unconnected to a comprehensive conception. It is more likely that comprehensive conceptions will follow the second process and be changed to accord with public reason. The fact that a conception has been transformed also does not indicate that one's identity has been destroyed. As H. L. A. Hart (1963) pointed out in his famous retort to Lord Devlin, transformation of a culture is not equivalent to destruction of a culture. Similarly, the transformation of individual identity is not equivalent to its destruction. Moreover, nothing in the suggestion that comprehensive conceptions need to be consistent with public reason—and may even need to be transformed to become consistent—suggests that individuals cannot have “private” reasons for their substantive political beliefs. Surely individuals can have a variety of ways of understanding their political views. The point is just that these private beliefs should not undermine public reason.

Even people who retained the belief that hierarchy was a natural factor in all areas of ethics would have to bracket that belief when discussing politics if they were to become reasonable. In this minimal sense, their moral identity would change such that they could justify bracketing their comprehensive conceptions. Okin could object here that this fact does not negate her main point about the Rawlsian double life. She could say that it is not enough for people to cease to have beliefs about hierarchy in public discussion, because the fact that they still hold these beliefs will influence their behavior in private life. In this case, the public persona does not succeed in altering an area of private life that is fundamentally political. For example, although these individuals could vote for equal sports funding for women in school because in public they operate as public reason givers, in private they might still raise their daughters to think that they are inferior to boys. In this case, as Okin sees it, private life has more of an effect on these women's political identity than does a public action such as voting. The daughters in this example might be eligible for equal sports funding, but if they were raised to think women should not compete in sports, they might not participate. Furthermore, their lack of participation might hinder their capacity to compete in general, and this in turn might affect their political equality. Although their fathers endorse political structures that support equality when they act as citizens, their actions as private individuals would in fact undermine these structures. Okin could argue that public reason does not affect private reason enough because it is limited to the political realm.

This example relies on the role sports play in socializing children to see themselves as equal competitors, which in turn can impact the public roles they play in the future. But it does not follow that all differences in how children are raised will have a similar effect. For instance, a time may come in which boys and girls are socialized to make differential use of sports but this process has no link to their political equality. In such an instance private beliefs about whether boys or girls should play sports will arguably not have a political meaning and thus will not require transformation. Private beliefs about different roles for different genders are not in all empirical circumstances incompatible with the type of equality required by public reason. In this sense, they need not always mirror public reason.

Although the phrase “public reason” seems to suggest an inherent divide between public and private life, on the strong public reconstruction there is no such clear or prior divide. Rather, to the degree that privacy exists at all, its boundaries must be determined by and normatively argued for through public reason. This has a major implication for traditional understandings of privacy: domestic life is not immune from political examination. This means that even if an individual is able to maintain a double life, the “private self” for such a person cannot be defined by his or her role in the family. On the strong public reconstruction, the “public” commitment to the overlapping consensus and public reason must itself penetrate all aspects of a citizen's interactions with his or her family because other family members' self-conceptions are at stake. The double life might allow citizens to retain private beliefs at odds with public reason because of political liberalism's commitment to freedom of speech and conscience, however, public reason requires that citizens in the political liberal state be committed to transforming these beliefs to make them consistent with public reason's demand for equality. Although freedom of conscience is therefore essential on a political liberal account, it does not serve to justify cordoning off the family as a protected private space immune from the requirements of public reason.

Although this point is implied in the strong public reconstruction's requirement that comprehensive doctrines must change to accommodate the reasonable, Rawls makes it explicit in “The Idea of Public Reason Revisited,” written in part as a response to Okin's critique (Rawls 1999b, 157). Notable especially is his refusal to exclude the family from public scrutiny on the grounds that it is not part of the basic structure:

There will no doubt be substantial debate about how much private life affects our ability to function, and see others, as free and equal citizens. This is but one easy example of a blurry line. The challenge will come from those who want to defend private spaces: is there any limit on public reason's influence over domestic life?

The requirement of freedom of speech and conscience that is central to Rawls' account of political liberalism suggests that there are limitations on how the project of transformation should be achieved. Some means of transformation are clearly ruled out. For instance, it is a paradigmatic violation of free conscience to require that citizens be prevented from voicing views about the inequality of women. Although sexist doctrines are contrary to the ideal of free and equal citizenship and their content, they are political viewpoints and should be protected as such. Free and equal citizens should not have their political beliefs dictated to them. Similarly, any attempt to legally limit what can be said in the home, even to children, would impose an intolerable restriction because of the requirements of freedom of speech and freedom of conscience. Issues of surveillance aside, the protection of conscience and speech is not limited to public spaces; it extends to domestic life as well.

Given the requirement that public justification cannot be instilled in citizens through the blunt sanction of criminal law, how can the process of transformation be achieved? The answer lies between two approaches. On the one hand, political liberalism cannot embrace methods of transformation that would violate freedom of conscience. On the other hand, it should reject a concept of neutrality that equates freedom of speech and conscience with the idea that the state has no interest in transforming the beliefs of its citizens. The most obvious means by which the state could effect such a transformation is by ensuring conditions that empower those who might be regarded as unequal within the family. Okin's suggestion about structuring divorce laws to ensure an exit option for women is paradigmatic. Similarly, state actions to ensure women's financial independence would also help to support structural conditions of equality without violating freedom of conscience. In regard to female children, the state should also provide them with opportunities that counterbalance attacks on their equality from within the family. As suggested earlier, ensuring equal funding and education in extracurricular activities such as sports is important, not only because sports teach girls competitive drive, but also because participation teaches them that they are equals to men in all areas of life, including political life.

In addition to creating these conditions, the state and its citizens have an obligation to be explicit about why these conditions are established. Public leaders and citizens should link these opportunities to their commitment to ensure that girls are raised as equal citizens, regardless of the beliefs of their parents.

Although the ideal of transforming families to ensure girls' equality might seem an appealing aspect of the strong public reconstruction, a harder case concerns public reason's influence on issues of education. The idea of public reason is arguably inconsistent with a particular pedagogy for teaching public high school students that evolution is false. In teaching “creation science,” some educators promulgate the belief that fundamental truth is found exclusively in the literal meaning of the Bible, suggesting that nonbelievers cannot hold plausible views on issues of science or morality. This is a problem on the grounds of public reason because a teacher—who serves as a representative of the state—is allowing a comprehensive doctrine to drive the curriculum. Another teacher might believe—because of her own comprehensive doctrine—that creation science is true, but would still appeal to public reason when creating curriculum.

But is the implication here that public reason requires restrictions on teaching creation science, not only in public schools, but in private schools as well? Is this is not an intuitively problematic limitation on the private sphere?

Educators in both public and private schools face a challenge in combining a commitment to public reason with a respect for freedom of conscience. It is contrary to the spirit of political liberalism, not to mention bad pedagogy, to restrict a belief or to dogmatically declare its falsehood. Restriction of creation science in this sense would be a mistake. But teaching evolution in science classes and requiring it in both public and private schools is not a restriction on belief. Teaching evolution as good science might lead those who have adopted creationism to rethink their beliefs or to transform them so as to be consistent with evolution. Such a process is not a restriction on freedom of conscience but stems from the right of future citizens to have information about basic science. Although it is not a direct attack on comprehensive beliefs, the transformation of these beliefs might be one indirect effect of public reason (see also Macedo 2000; Tomasi 2002).

There are parallel cases in other curricular areas, in which political liberalism would bring about transformation in a manner consistent with freedom of conscience. For example, as part of learning the history of this country, students should learn about struggles for equal citizenship. This should affect private religious beliefs about the inequality of women. Some skills, such as learning to interpret a text in a metaphorical rather than a literal manner, might risk that these children abandon certain forms of religious fundamentalism, which disdain anything but literal interpretations of texts. I think these “critical thinking” skills are also essential to liberal identity, even if they serve to displace certain religious views. Such requirements in education are not value neutral in that they promote the basic values of public reason. But they do so in a way that respects the freedom of conscience and the rights of parents. Parents do have an interest in participating in the education of their children. This interest, however, is not as fundamental as the interest future citizens have in developing the characteristics necessary to see themselves as free and equal. Although this is not an argument against private education per se, it is a position consistent with the idea of strong regulation of private schools (see Macedo 2000). Now that I have discussed how the transformation of the private sphere can be achieved while respecting freedom of conscience, I will return to developing the strong public reconstruction and its relationship to the public/private distinction.

I suggest that the result of the strong public reconstruction and the principle of publicly justifiable privacy is not only the blurring of private and public spaces but also, ultimately, a new understanding of how liberalism conceives of the interaction between the public and private spheres. What emerges is a conception of privacy reformulated from previous liberal conceptions. For Rawls, the private sphere is not a “space” into which politics cannot intrude. Rather, public reason and mutual justification demand that the moral identity of individuals as citizens is given priority over their private identities. This often means that our public commitments are more fundamental than private ones. Rawls writes, “The adult members of families and other associations are equal citizens first: that is their basic position” (Rawls 1999b, 160–61). When public reason demands it, individual moral identity must be completely reformulated. For Rawls, political liberalism

Some conceptions of liberalism conflict with Rawls's formulation. For some liberals, private space is not granted by politics. Rather, politics serves only as a pragmatic method by which to protect a prepolitical notion of family and individual privacy (Barber 1984). On my reading, however, Rawls (1999b) has not only blurred the boundary between public and private but also has eliminated the private realm as a place distinct from public scrutiny. Some traditional liberals will claim he has destroyed the very space liberals attempt to protect.

This claim is true in one respect. Rawls does rest the very existence of private spaces on public reason. However, this does not mean he calls for an end to the concept of privacy altogether. Public reason guarantees citizens' privacy rights to the degree they are justified. This is no small point, especially when we consider the type of language free and equal citizens should use when they deliberate about what types of institutions are in fact just. For Rawls, public justification proceeds largely in the language of individual rights. Families with a sexist comprehensive doctrine violate the very right of female children to learn to see themselves as free and equal citizens. On a Rawlsian account, this is a fundamental aspect of the right to education. Moreover, respect for this status is a fundamental basis for all rights on a Rawlsian account. Indeed it might be called the “right to have rights” (Trop v. Dulles, 356 U.S. 86 [1958]). The public culture should encourage and provide conditions that make it possible for these values to be reformed. Here, a new conception of privacy of individuals trumps the rights of families.⁴

Drawing from Rawls, I want to suggest that the boundaries of mutually justifiable privacy cannot be found by simply looking to the way family life is commonly conceived in a particular culture. Although a right of privacy must be protected, the definition of the private sphere must be found through mutual justification. This interpretation of privacy is liberal in the classical sense that it provides for the protection of individuals through a guaranteed right. However, it differs from traditional liberal theory in that, like republican theories, it suggests that these rights can be elaborated through a political process of public justification. Private life is not a priori distinct from or “protected” from the political. Rights and privacy are grounded rather in the politics of mutual justification among citizens.

CHALLENGES TO THE STRONG PUBLIC RECONSTRUCTION: THE POLITICAL/COMPREHENSIVE DISTINCTION AND THE ROLE OF THE BASIC STRUCTURE

So far, I have worked to reconstruct political liberalism to incorporate feminist concerns about injustices in spheres traditionally considered private. I have argued that although any intervention into family life must be justified by public reason, theorists of political liberalism should not regard the family as a private sphere immune from political justification. Before I conclude, I want to further develop my own reconstructed account of political liberalism by examining its responses to prominent critiques of political liberalism that have bearing on it.

One such critique challenges Rawls's contention that a political conception can be kept distinct from comprehensive conceptions, arguing that Rawls' own political conception unwittingly collapses into a type of comprehensive conception (Callan 1997, Ch. 2). Given the broad scope of my reconstructed political liberalism, one might attempt to apply the same critique to this view, suggesting that my argument for the application of political values to the family conflates political and comprehensive conceptions. Another way to put the challenge is to ask whether a political conception can serve as a freestanding standard for evaluating aspects of family life and still remain freestanding from comprehensive doctrines.

My suggestion that the ideals of free and equal citizenship extend to family life will certainly be controversial from the perspective of some comprehensive doctrines. The strong public reconstruction thus might invite comparison with the views of political liberals such as Steven Macedo, who argues in the provocatively titled “Liberal Civic Education and Religious Fundamentalism: The Case of God vs. John Rawls?,” that certain religious worldviews are in tension with some political liberal commitments and should be transformed in liberal societies. As Macedo (1995) acknowledges, this position also opens him to the charge that he is committed to a comprehensive doctrine.

In answering the question of whether my own reconstructed political liberalism is itself comprehensive, it is helpful to introduce a distinction between two types of comprehensive doctrines. Rawls distinguishes between “fully comprehensive” and “partially comprehensive” conceptions as follows:

Specifically, I contend that my reconstructed political liberalism is not fully comprehensive for two reasons. First, it is compatible with a range of comprehensive doctrines. As I have argued, the commitments entailed by my account of political liberalism might conflict with some comprehensive doctrines in regard to certain issues internal to the family. But even in those cases for which a commitment stemming from a political conception like mine should trump a commitment stemming from a comprehensive doctrine, that comprehensive doctrine need not be destroyed. On my view, many comprehensive doctrines can evolve in various ways to become compatible with a legitimate political conception.

Second, although my reconstructed political liberalism incorporates many feminists' concerns about life in the family, there remain many significant life decisions about which political liberalism has no definitive requirements. Following Nancy Rosenblum, we might say that the values of political liberalism do not go “all the way down” into every single issue of personal choice (Rosenblum 1998). Rather, political liberalism creates a minimum standard of justice for evaluating decisions within the family without dictating answers to every issue within the family. Thus the reconstructed political liberalism I have developed can be distinguished from the more radical feminism of thinkers such as Carole Pateman (1988) and Catharine MacKinnon (1993). For these thinkers, many of whom suggest that the public/private distinction is untenable, feminism is a theory of complete human emancipation, a comprehensive and political conception. For instance, Pateman has famously argued against not just the legitimacy of some contracts, but more generally against the very idea of contract, labeling it a liberal device that invokes autonomy and “individuality” to undermine women's equality. But as Okin (1990) herself argued in her review of Pateman's (1988) The Sexual Contract, “There are other important distinctions to be drawn within the category of employment contracts. These may separate legitimate forms of work contract, possibly including some forms of prostitution, from illegitimate forms, including most forms of prostitution” (Okin 1990, 667). For instance, given a just distribution of wealth and reasonable limits on coercion within the family, we can imagine that a woman might make a reasonable decision to sell sex for money. In such circumstances there might be reason from the perspective of some comprehensive feminist views to oppose such a decision, but I believe political liberalism, even of the reconstructed kind, to be silent on the matter. A similar thing might be said about marriage more generally. Although as a political matter some marriage arrangements might be ruled out as undermining women's free and equal citizenship, the institution of marriage is not per se incompatible with a reconstructed political liberalism. On my view, the values of free and equal citizenship are relevant to many decisions once widely thought private, but they do not apply to all decisions regarding one's conception of the good life.

The distinction between comprehensive feminist theories and reconstructed political liberalism has potential implications for education. An education based on a comprehensive account of women's equality might weigh in on any and all matters of the good life. For instance, it might stress the importance of avoiding marriage on the grounds that it is inconsistent with women's full human emancipation. In contrast, my reconstructed political liberalism would encourage educators to stress how some institutional arrangements and some intrafamily decisions threaten women's equal citizenship, but it would avoid taking positions on issues of the good life that are not matters of justice. For instance, it would avoid dictating whether women should marry given conditions consistent with the values of free and equal citizenship.

The idea that my reconstructed political liberalism does not “go all the way down” and dictate answers to each question about the good life also makes it consistent with the view that other “affectional” values besides those associated with free and equal citizenship have a place in the family. The internal regulation of a family need not be governed solely by the principles of public reason; these principles are not sufficient to account for many of the subtleties present in family life. Concerns about equality, for instance, should not drive every interaction among family members; expressions of affection or humor, although integral to the family dynamic, are not reducible to questions of equal citizenship, nor are they necessarily inconsistent with this ideal. In short, a distinction can remain between the styles of governance in the family and the public realm, but this need not mean that any aspect of the family is immune from the scrutiny of public reason.

Because it is compatible with a range of comprehensive doctrines and does not dictate how persons should make each decision in their lives, my reconstructed political liberalism is distinct from fully comprehensive conceptions, despite its incursion into the realm of the family. A much harder question concerns whether my theory is partially comprehensive. Okin's critics, most prominently Martha Nussbaum, argued that her work could be characterized as comprehensive (see Okin et al. 1999, 108). In response, Okin herself suggested that her view was “in between” comprehensive and political liberalism (1999, 129). Although its meaning is not clear, theorists might interpret this comment as a concession that Okin's theory of liberal feminism was a partially comprehensive conception.

Regardless of Okin's own conception, however, one might plausibly argue that my reconstructed political liberalism is not fully but partially comprehensive. Again, Rawls describes a view as partially comprehensive “when it comprises a number of, but by no means all, nonpolitical values and virtues” (Rawls 1996, 13). These may include ideals of personal character, as well as ideals of friendship and of familial and associational relationship” (1996, 13). If we follow Rawls's definition as stated here, the strong public reconstruction would be partially comprehensive because it suggests a normative ideal that impacts familial relationships directly. Moreover, the strong public reconstruction challenges the idea that our family roles can be regarded as private aspects of moral identities, distinct from political commitments (also see Callan 1997, 30).

However, I believe that we should not accept without argument Rawls's implication in this definition (written before his “Public Reason Revisited”) that because partially comprehensive views address familial relationships, they are therefore “nonpolitical.” I have argued throughout this article that the very realms we often define as nonpolitical or private should be examined in light of conceptions of free and equal citizenship and are thus political in an important sense. By defining partially comprehensive views that examine the family as nonpolitical in Political Liberalism, Rawls (1996) risked trying to resolve the central issue of what is political in the first place with a pre-normative definition. This is precisely the move that Rawls himself warned against in “Public Reason Revisited.” As Larmore suggests, the domain of the political is not a “prepackaged sector of society, inherently distinct from the other areas of social life” (Larmore 1999, 607). I have suggested that the question of whether the domain of the political extends to family life is open to debate through public reason. As Okin argues, the coercion latent in the family lends credence to the idea that it is an institution central to politics and subject to the standard of free and equal citizenship. Because some issues surrounding family life are political, my strong public reconstruction of political liberalism is not rightly characterized by the Rawlsian term “partially comprehensive.” Here I invoke notions of the political and of free and equal citizenship that go beyond common definitions.

Another issue that the family raises for theories of political liberalism concerns the role of the basic structure as the primary or “first” subject of political justice (Rawls 1996, 257). In one of her last articles, Okin acknowledges that Rawls has made inroads toward incorporating her critique of his account of political liberalism (Okin 2004). However, she suggests that Rawls's response to her remains ambiguous, particularly his position regarding the role of the family in the basic structure. Although Rawls says that the family is part of the basic structure of society and thus is subject to the principles of justice, he also claims that the principles do not “apply directly to the internal life of the many associations within [the basic structure], the family amongst them” (Rawls 1999b, 158). Here Rawls creates a distinction between the principles of justice, which constrain the family as part of the basic structure, and the internal principles of family governance. Specifically, he suggests that we

This suggestion, however, raises a complication regarding the meaning of the basic structure. As Gerry Cohen points out, “it is seriously unclear which institutions are supposed to qualify as part of the basic structure” (Cohen 2000, 136). On the one hand, Cohen points out that the basic structure is defined as including “major” institutions, including the family. On the other hand, the basic structure sometimes seems to refer only to coercive state institutions. The family rests in an uncertain place within this paradigm. The state influences the family through divorce and child custody laws, but many of the decisions within families are made without direct government control. One reason to narrow the meaning of the basic structure to comprise only institutions routinely linked to direct state coercion—and thus not to include all aspects of family decision making in it—is conceptual clarity. Although there are instances in which intrafamily dynamics are affected by the law and state coercion, the difference between state coercion and much family coercion is significant. For instance, fathers who prevent their children from leaving the house exert coercive control, but this is distinct from state coercion. If we define the basic structure in terms of coercive state institutions, does this mean political liberalism can say nothing about family life, given that the basic structure is the “first” subject of justice? Cohen himself thinks that Rawls cannot merely extend the notion of the basic structure to include the family and that therefore his account is not a complete account of justice. Cohen's critique of political conceptions that apply only to coercive institutions certainly resonates with my own emphasis on the relevance of the values of free and equal citizenship to intrafamily dynamics. Moreover, I think Cohen is right to identify some confusion in Rawls' definition of the basic structure. There is not room here to address all of Cohen's objections to the basic structure. However, before I conclude, I want to suggest why I can endorse applying the values of free and equal citizenship to aspects of family life not related to state coercion without committing to the idea that the basic structure must be entirely jettisoned.

If we accept the idea that the basic structure refers only to coercive state institutions, we can still view the values of free and equal citizenship as applying beyond the basic structure. Indeed, Rawls's claim that the basic structure is the “first” subject of justice implies that the basic structure need not be its exclusive subject (Rawls 1996, 257). My view is that although coercive aspects of the state are at times relevant to family injustice and rightly regarded as part of the basic structure, some decisions within the family are distinct from the basic structure. For instance, the ideal of free and equal citizenship suggests that parents should teach their children, both male and female, that gender is not an indication of fundamental superiority or inferiority. Although this requirement is based in political morality and has implications for the duties of citizens, it should not be directly mandated by law and is not rightly regarded as an aspect of the basic structure. I therefore reject Okin's suggestion that we should think of the family in its entirety as part of the basic structure and directly subject to the two principles of justice. Such an application of the two principles would raise a slew of complications. For instance, the first principle's commitment to freedom of speech would certainly complicate child rearing. The value of free speech might trump the value of mutual respect among adults debating politics, but in family settings where children are still learning the basics of social interaction the inverse is arguably true. It is also hard to understand conceptually how the difference principle—Rawls' view that primary goods such as income should be distributed to the advantage of the least well-off in society—might be applied within the family, for instance, among children. Moreover, the second principle's requirement regarding “offices open to all” seems to have at best an awkward fit with familial roles. The two principles are crafted to govern state institutions; thus, applying them to the family, a nonstate institution, would appear misguided.

However, even if we reject Okin's idea that the family in its entirety should be viewed as part of the basic structure, the values of free and equal citizenship can still be used to evaluate specific intrafamily dynamics that are not implicated by the basic structure. As I have argued throughout this article, the values of free and equal citizenship, including the requirement of reasonable treatment, should serve as a way of judging justice within the family. For instance, I have argued that the principles of political liberalism are relevant to personal decisions within the family in matters such as whether to allow female children to play sports. Decisions like these are not part of the basic structure, but they are, I have suggested, within the realm of normative political argument. Even in instances in which the state should not directly intervene to ensure that such decisions are consistent with the values of free and equal citizenship and public reason, these values should inform decision making within the family. A reconstructed political liberalism can thus define the basic structure narrowly in terms of state coercion while still insisting on the political relevance of injustices within the family. This approach avoids the ambiguity about the basic structure raised by Cohen but does not necessarily require that we abandon the idea of the basic structure entirely. We thus can rely on the basic structure as a way of evaluating family injustices related to state coercion, but this need not be the exclusive political means for evaluating all injustices within the family.