I. Conservatism and Traditionalism

First, the phrase “education for autonomy” generally refers to formal education and, more specifically, to the formal education that takes place in (American) public schools. That will be my focus here. ^{Footnote 8} But the claim that formal education should be education for autonomy reflects preferences for autonomy in the surrounding culture, and may reinforce informal cultural incentives and pressures to live autonomously. The operation of these cultural influences is thought to threaten traditional ways of life as well. By seeing what education for autonomy aims at, we may come better to understand what these informal pressures are pressures to do. That, in turn, will help make clear the threats that liberal culture does — and does not — pose to traditional ways of life, and may help identify legitimate ways in which those threats can be averted. Appraisal of arguments about formal education can therefore be expected to have an interest that extends beyond the purely academic or pedagogical.

Second, when education for autonomy is said to threaten traditional ways of life, what is often meant is that education for autonomy threatens to loosen or sever students’ ties to conservative religious communities of origin, and to communities of origin that conserve ethnic and cultural identities and practices. The charge is sometimes put forward as a general thesis, ^{Footnote 9} while at other times it is leveled by adherents of a particular way of life. The charge may be made most evocatively by members of one or another of the communities that are said to be threatened; for the most articulate of the accusers may provide thick descriptions of the modes of thought and practice that will be lost if their children are educated for autonomy. ^{Footnote 10} But despite the power of these particularistic descriptions to convey the sense of loss or threat that adherents feel, I believe that the features of one or another traditional life that seem to render it vulnerable are in fact shared by many such ways of life. Their commonality can be used to help us understand the more general charge that education for autonomy threatens traditional ways of life as such. Isolating those common features requires us to gain some distance from the fine textures and particularities out of which the thick descriptions are woven, and to abstract from the warp and woof of the ways of life in which complainants are enmeshed. Many of the rank-and-file adherents of traditions may not be inclined to such abstraction and may not recognize these abstract features as characteristics of their ways of life. The effort to identify them must therefore be vindicated, not by adherents’ explicit endorsement of them, but by the understanding their identification helps us to achieve.

Unfortunately, constraints of space — and lack of anthropological expertise — prevent me from moving from thick and particular descriptions of traditional ways of life to an abstract description that picks out shared features. Instead I will begin with what I hope are some plausible generalizations. I assume, first, that adherents of traditional ways of life — Catholic, orthodox Jewish, Old Order Amish, tradition-minded native peoples, revivalists of craft and agricultural techniques, and many others — prize patterns of behavior that have long histories of exemplification; and second, I assume that they prize them, in part, because they have long histories of exemplification. I also assume that adherents of these traditions take a long history of exemplification as a reliable indicator of some further good that is realized in the prized patterns of behavior ^{Footnote 11} .

We can identify further shared features of traditional ways of life by asking how these generalizations are to be explained. I suggest that we can explain them by supposing that adherents of the diverse traditions of which these generalizations are true accept — perhaps tacitly accept — a thesis about a source of value, a thesis which I shall call “traditionalism.”

Critiques of autonomy-based liberalism such as Michael Sandel’s may tempt us to identify the thesis of traditionalism negatively, as a denial that the moral significance of communal ties and the normative force of established patterns of behavior are derived entirely from individual choice. ^{Footnote 12} Traditional ways of life would then be identified as ways of life followed by those whose ties to their communities reflect their acceptance of the negative thesis. But the temptation should be resisted, for this characterization of traditionalism leaves us in the dark about what the traditionalist thinks the source of value is. Moreover, if autonomy essentially involves an act of will, then when we describe traditional ways of life as those which, in effect, downgrade the significance of choice, we make the charge that education for autonomy threatens traditional ways of life into a charge that is too easy to prove. I will therefore characterize traditionalism differently.

Traditionalism as I understand it takes the fact of long exemplification to be one source of the value attached to prized patterns of behavior. It also takes long exemplification to indicate that behavior that fits the pattern persisted because experience showed it to be a form of, or to contribute to, human flourishing. Conformity to traditionalist patterns requires acting for the right reasons, and among these reasons is the value that is taken to have been conferred on the pattern by its history. This motivational thesis explains why traditionalists assert, as I believe they typically do, that one good reason for someone to engage in the form of behavior in question — and one reason that engaging in it is an instance of flourishing — is that by engaging in it, the agent creates a link between present and past and thereby helps to keep the past alive. ^{Footnote 13} Some traditionalists, but not all, are authoritarians: among the patterns of behavior they value is that of deference to some authority. They differ from nontraditionalist authoritarians in thinking that traditionalist theses about value help to account for the value of deference. With these traditionalist theses in hand, we can describe traditional ways of life abstractly rather than thickly. A traditional way of life is a conception of the good which makes a prominent place for the values and motives that traditionalism of one kind or another identifies.

One way in which the foregoing analysis illuminates is by helping us distinguish two views that ordinary moral and political discourse often conflates: traditionalism and conservatism. To see the difference, note that some philosophers who have tried to distill the essence of conservatism have identified it with status quo bias. ^{Footnote 14} Others have said it favors preserving those things that exist and are of intrinsic value, even if their replacement would yield more value. ^{Footnote 15} Conservatism understood in these ways differs from traditionalism as I understand it in that it does not connect respect for its definitive value with human flourishing or claim that such value is importantly realized in patterns of behavior. Moreover, these analyses of conservatism preclude the possibility that a conservative, as such, could value and wish to revive patterns of behavior that have become moribund, since the instantiations of such patterns do not exist and hence are not elements of the status quo. By contrast, traditionalism as analyzed here leaves this possibility open, since it does not imply that the history of engaging in the valued behavior must extend to the present.

The analysis draws some plausibility from the fact that it also helps us to understand the tension some writers have seen between education in a tradition and education for autonomy. Barry Chazan, for example, favors education in what he calls “the Jewish Tradition” and says the Tradition “implies a body of values, laws, norms, behaviors, and practices which are regarded as rooted in Jewish history and which are considered to be relevant and good for the lives of Jews today.” It is natural, I think, to read Chazan as saying that young people should be educated into patterns of behavior that are valuable because they have a long history of observance. It is also natural to read him as saying that they are “relevant and good for the lives of Jews today” in part because their history itself confers value and in part because collective engagement in them makes other goods available to the community, such as connection with its past. If this reading is right, then we can take Chazan to be endorsing an instantiation of the general two-part thesis I have called “traditionalism.” ^{Footnote 16} Furthermore, I believe it is natural to read him as endorsing some version of the motivational thesis I associated with traditionalism, a version according to which young people should be taught to participate in long-standing patterns of behavior for reasons associated with the history of those patterns. ^{Footnote 17} Since it is precisely an instance of the motivational thesis that raises the tension Chazan sees between education in the Jewish Tradition and education for autonomy, ^{Footnote 18} the rather abstract characterization of traditionalism given above helps us understand his argument.

Finally, the analysis of traditionalism suggests two ways of connecting education for autonomy with the threat to traditional ways of life without reducing the connection to triviality — ways that, if elaborated, yield just the two arguments that I promised to examine. Traditionalism identifies certain reasons as good ones for replicating patterns of behavior. Among those reasons, I said, are the venerability of the pattern and, sometimes, the venerability of an authority to which the traditionalist claims deference is owed. To adhere to the relevant tradition is, in part, to treat those reasons as good ones. Education for autonomy is, as we shall see, thought to require education for “critical reflection.” The crucial premise in the first argument can then be taken to assert that traditionalist reasons will not or cannot survive critical reflection in a pluralistic setting. A result of education for autonomy is that the force of traditionalist reasons gradually ceases to be felt and people cease to live out traditional ways of life. To see the second way of connecting such education with the threat, suppose the traditionalist conjoins claims about what reasons are good ones with the further claim that the reasons she identifies are to be responded to without further reflection by the agent on the fact that, by responding to them, he is sustaining or reviving a traditional way of life. According to this claim, traditional ways of life are, by their very nature, sustained unreflectively. Such ways of life are therefore destroyed by critical reflection on one’s own reasons for action. They are threatened by the encouragement of critical reflection that education for autonomy entails.

To show where these arguments go wrong, I shall now try to fill in the details of the outlines suggested by my brief analysis of traditionalism.

II. The Central Argument

The arguments that interest me are those for the conclusion below:

• C:Education for autonomy threatens traditional ways of life.

The arguments begin by supposing what I said at the outset many liberals accept, namely:

1. (1) One of the goals of education is to prepare students to live autonomously.

What is autonomy? What is the goal at which (1) says education should aim?

Autonomy figures prominently in contemporary moral philosophy and the vast literature on it has greatly deepened our understanding of this important value. It would be natural to think that the kind of autonomy that is said by (1) to be an educational ideal is borrowed or derived from that literature. But if this is so, the debt is not readily apparent since discussion of autonomy in the philosophy of education proceeds largely in isolation from current work in ethics. Even if it is assumed, as I do assume, that the interpretations of autonomy that are central to that work are generally too controversial to serve as the kind of publicly defensible ideal of state-funded education to which (1) refers, one way to define and sharpen the ideal that is deemed defensible would be to contrast it with these stronger interpretations. But because I am concerned with just one condition of autonomy rather than with identifying a set of defining conditions, I will abjure such comparisons here. As we will see, philosophers of education generally take the condition with which I will be concerned to be a necessary and it is the condition of autonomy that is widely thought to threaten traditional ways of life, and thus to lead from (1) to C. To see what that condition is, we need to see what philosophers of education have said about autonomy.

Many advocates of education for autonomy equate autonomy with self-government. Thus, Constance Kamii, for example, says that “Autonomy is being governed by oneself.” ^{Footnote 19} Josh Corngold says that “Personal autonomy generally refers to one’s capacity to be self-governing.” ^{Footnote 20} The word “self” presumably occurs in the phrase “self-government” to indicate that the self is the agent who does the governing. But to take it to indicate only that leaves open the question of who or what is governed. An alternative way of interpreting the phrase is as indicating that the self is both the agent who does the governing and is the subject who is governed. The second interpretation answers the question left open by the first. It also helps to explain the equation of autonomy with self-government. For while the word “governor” and its cognates have suggestive connotations drawn from the terms’ use in mechanics — in particular, they quite helpfully suggest that to govern oneself is to maintain oneself in a kind of equilibrium ^{Footnote 21} by both the conscious and the virtual monitoring of one’s own behavior ^{Footnote 22} — the most illuminating connotations are conveyed by the uses of the terms in politics and administration. Political and administrative governance, unlike mechanical governance, is typically exercised by legislation. To claim that “self-government” entails that the self both governs and is governed, and to adopt a legislative conception of government, together suggest that self-governance is a matter of giving oneself a law. Since that is just what the term “autonomy” denotes, the equation of self-government with autonomy can be explained in a natural way.

But while taking the “self” in “self-governance” to have a double-referent — to refer to both the ruler and the subject — makes the equation of autonomy and self-governance plausible, the plausibility comes at a price. For it is not clear what it is to govern a self. And so while I accept the nominal equation of self-governance with giving oneself the law, I shall assume that the underlying idea can be more felicitously expressed by taking the subject of self-legislation to be one’s own life rather than one’s self and by restricting the referent of “self” to the agent doing the governing or legislating. The equation of autonomy and self-government, clarified by these assumptions, says:

1. (2) To live autonomously is to live a self-governed life.

What is it to for someone to live a self-governed life?

Whatever the self is that does the governing, I take it that the relevant element of the self is that agent’s reason. Someone lives a self-governing life when she lives a life that is appropriately regulated or governed by her own reason. Living a life is an ongoing activity. Following Rawls, I take that activity to be the carrying out of a plan ^{Footnote 23} and I take it that an important truth about plans can be expressed by the metaphor of a center: there are principles and values which are central to the plan and their centrality consists in the guiding role they play in the agent’s deliberation about what to include in the plan and how to execute it. Some of these principles and values are associated with the right. Others are the defining or distinguishing elements of the agent’s conception of the good. There are typically a large number and wide variety of such elements. They include especially strong desires that may have any of a number of objects but help to define the person’s character, such as the desire to excel in an especially consuming pursuit, and commitments to political, ethical, religious, and aesthetic values. Especially important are higher-order desires to be a particular kind of person: to play certain social roles well, to live up to certain ideals of conduct and to have a character which is stable or which returns to equilibrium through the vicissitudes of life. ^{Footnote 24}

To lead a fully self-governed life would be to lead a life in which all of one’s values and principles, and all of one’s choices, were self-governed or appropriately regulated by one’s reason. For reasons that will be clear later, I shall concentrate on just one condition of a self-governed or autonomous life, a condition that applies to the values and commitments that are central to a conception of the good.

A condition that required the adoption of these values and commitments to be regulated by one’s reason would be too strong, at least if it meant that their adoption should be a process in which the reasons for and against the values were consciously tallied. Most of us acquire at least some of the commitments that are most important to us before we have reached an age at which we are capable of deliberating about them. But if it is too much to require that our adoption of a conception of the good should be regulated by our reason, it may seem much more plausible to require — as a condition of self-governance — that our rejection, modification or continued endorsement of their central elements should be vindicated by our reason once we reach adulthood. ^{Footnote 25} We can express this conclusion as:

1. (3) An agent lives a self-governed life only if the central elements of her conception of the good are vindicated by her own reason.

Claim (2) — that to live autonomously is to live a self-governed life — and (3) imply that:

1. (4) Someone lives autonomously only if the central elements of her conception of the good are vindicated by her own reason.

It may be possible to live a life that satisfies the condition laid down in (3) and (4) while being unaware that the condition is satisfied or without entertaining any beliefs about whether it is satisfied. Perhaps the condition can be satisfied by someone who accidentally vindicates his central aims, without knowing or believing that he has done so. But a person who satisfies (4) entirely in these ways does not seem to be autonomous. As I hinted earlier, I believe that self-government is consistent with a certain amount of virtual monitoring of one’s own behavior. Our intuitions about autonomy also seem to demand, however, that the ongoing process of self-government be at least episodically conscious and reflective. Someone who satisfies (4) without being aware that she does, or without believing that she does, does not seem to exercise enough conscious control over her life to satisfy our intuitive judgments about what counts as living a self-governing or autonomous life.

This line of thought does not itself constitute an objection to (4), which expresses a necessary and not a sufficient condition of living autonomously. It does, however, suggest a plausible way of strengthening (4). According to the strengthened version:

• (4')Someone lives autonomously only if the central elements of her conception of the good are, and are believed by her to be, vindicated by her own reason.

When I speak of the central elements of someone’s conception of the good being “vindicated by her own reason” I mean, as a first approximation, that they survive a process in which she tries to identify her reasons for having the conception of the good that she does, and in which she critically assesses the quality of those reasons and the degree of support that they provide. Since this process of reflection, identification, and critical assessment is often referred to as “critical reflection,” (4′) is equivalent to:

1. (5) Someone lives autonomously only if the central elements of her conception of the good survive, and are believed by her to survive, critical reflection.

Teaching students to engage in critical reflection obviously requires imparting the skills exercised in that process. I will say more below about what these skills are. They include the ability to apply argumentative standards, to assess evidence, and to see what the balance of reasons dictates. They also include the ability to see whether preferences and hopes are reasonable, to identify adaptive preferences and to distinguish adaptive preferences from those that result from the reasonable tailoring of ambition to circumstance. But recall that we want vindication by reason and self-government to be a reflective process, of which people are aware and that they voluntarily undertake. And so teaching students to engage in critical reflection also requires imparting the desire to exercise these skills and to guide oneself by the results. The desire to live such a life is a highest-order desire, a desire that governs whether and how to pursue the other central elements of one’s conceptions of the right and the good. And so these considerations, together with (5), imply that:

1. (6) Someone can live autonomously only if she has the highest-order desires (a) to subject the central elements of her conception of the good to critical reflection and (b) to follow that conception only if its central elements survive, and are believed by her to survive, critical reflection.

From (1) — the claim that one of the goals of education is to prepare students to live autonomously — and (6), we can infer:

1. (7) One of the goals of education is to inculcate in students the highest-order desires (a) to subject the central elements of their conceptions of the good to critical reflection and (b) to follow those conceptions only if those elements survive, and are believed by them to survive, such reflection.

We saw in the move from (4′) to (5) that critical reflection is a condition of autonomy. I said then that it contributes to autonomy because the person who tests her conception of the good by critical reflection and acts accordingly exercises some rational control over her life. Why does the exercise of rational control contribute to autonomy?

Autonomy is a form of freedom. It may seem intuitively plausible that the exercise of such control is freeing. But if it is true that traditional ways of life are threatened by an education for critical reflection — if it is possible to sustain the move from (4′) and (5) via (7) to the conclusion C about the threat that education for autonomy poses to traditional ways of life — then we seem to have a second and complementary reason for thinking that critical reflection contributes to freedom. For it is natural to suppose that the freedom in question is negative or has a negative component, and that this component can be identified by identifying what education for critical reflection frees the reflecting agent from. It may also seem natural to contend that such an education consists in part in learning to question received ways of doing things, so that students can free themselves from the dictates of custom, authority, and tradition.

I do not believe that the assumption of rational control over one’s life, and critical reflection on one’s conception of the good, must have the effect of freeing students in this way because I do not think that the moves that will take us from (7) to C are sound. But I think it is quite likely that education for autonomy and critical reflection will indeed have this effect. The reasons it is likely to have it are to be found in the ways an education recommended by (7) is likely to proceed.

Step (7), just above, refers to highest-order desires to subject the central elements of [one’s] conceptions of the good to critical reflection and to follow that conception only if its elements survive such reflection. These are desires to live or to conduct oneself in a certain way. They are desires to be a certain kind of person — the kind of person who scrutinizes his conception of the good and who pursues it only if its central elements stand up to his own critical scrutiny. Thus, they are desires to be a person who decides for himself how to live on the basis of his own reason. American adolescents, having reached an age at which they yearn for independence, are strongly disposed to some form of the desire to be self-determining persons in any event. Their desire will be heightened and given focus by an education that aims at making them the kind of person (7) implies they should be. One way that aim might be pursued is by exposing students to literature that encourages identification with and emulation of characters who rebel against convention. I do not mean to criticize this way of educating students, but merely to point out that it is one way in which education for autonomy might threaten traditional ways of life.

To see another, consider that whatever else critical reflection involves, it is supposed to involve the agent’s reflecting on her conception of the good and asking herself whether she has reasons that justify her acceptance of it. Such critical reflection need not take place in a social setting. But education for critical reflection takes place in classrooms. Students learn to identify the reasons that support their views, to recognize that views are unsupported or inadequately supported by reasons, and to justify the views they hold, all by exchanging reasons with other students under the guidance of a teacher.

The fact that education for critical reflection takes place in a social setting, in which students learn to assess the justifiability of their views by seeing how well they can justify them to other students, encourages students to equate the justifiability of their views with the justifiability of their views to others. It may encourage them to equate justifiability with justifiability to the reasonable members of their peer-group. Or, recognizing the shortcomings of classroom deliberations and the vulnerability of its outcomes to contingency, they may equate justifiability with justifiability to some idealized group. Either way, it encourages students to accept an interpersonal conception of justification. To someone who accepts this conception:

1. (8) A conception of the good will be believed to survive critical reflection only if it is believed to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

An education that successfully pursues the goal expressed in (7) will inculcate a desire to pursue conceptions of the good that satisfy the condition expressed in that step and in (6) — steps that expressed conditions on living autonomously. And if (8) is right, then:

1. (9) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

To see whether, as C says, such an education threatens traditional ways of life — to see whether there are plausible steps from (9) to C — we have to know something about the results of critical reflection in a group setting. It is virtually axiomatic in liberal theory that critical reflection about the good among reasonable people leads to disagreement. Charles Larmore explains the axiom especially clearly and persuasively. He says:

This tendency to disagreement is responsible for the ethical pluralism of free societies. Such pluralism will be reflected in the composition of public school classrooms and in the results of the collective critical scrutiny that takes place in those classrooms. So the interlocutors referred to in (8) and (9) are bound to disagree, whether they are reasonable classmates or are notional interlocutors whose reasonability and “general capacities of reason” have been idealized. Everyday experience of life and exposure to entertainment and social media will have taught students the inevitability of disagreement and difference about how to live. And so it may seem that:

1. (10) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection, but who disagree among themselves “about the fulfilled life, the human good [and] the nature of self-realization.”

Recall that step (7) says:

1. (7) One of the goals of education is to inculcate in students the highest-order desires (a) to subject the central elements of their conceptions of the good to critical reflection and (b) to follow those conceptions only if those elements survive, and are believed by them to survive, such reflection.

As with step (7) so with step (10), an education that inculcates the highest-order desire to which it refers is an education that encourages students to think of themselves as, and to aspire to live as, a certain kind of person. It encourages them to aspire to a certain kind of freedom — to autonomy, understood as requiring rational governorship of their own lives. Successfully encouraging convergence on these ideals and aspirations is itself a challenge in the face of pluralism. If such an education succeeds, it still encourages students to think of themselves as members of society in which disagreement about at least some other elements of the good is inevitable — for that, as Larmore insists, is “an essential ingredient in our modern self-understanding.” ^{Footnote 27} And it encourages each to accord the others the status of a moral agent equal to himself, since each refrains from endorsing a conception of the good that cannot be justified to the rest despite their deep and enduring lower-order disagreements.

Encouraging students to think of themselves and to conduct themselves as free and equal persons so understood may seem to be a laudable educational goal. Indeed, it may seem to be what education for autonomy is really supposed to accomplish. The laudability of inculcating the desire referred to by (10), and its aptness as an aim of education for autonomy, may seem to be confirmed by the similarity between this desire and the desire with which T. M. Scanlon famously identified moral motivation: that is, the desire to be able to justify one’s conduct to others who also desire that their conduct be interpersonally justifiable. ^{Footnote 28}

The self-conception that would be encouraged by such an education is very powerful, in part because it draws on very attractive and laudable elements in liberal political culture. But in practice, if students do not adopt it with care and discrimination, it can undermine or corrode other self-understandings and identities, prominently including those that are part and parcel of traditional ways of life. And I think reflection on how an education of the sort described in (10) would play out confirms this. It suggests steps leading from (10) to C and shows how such an education threatens traditional ways of life. To see what those steps are, consider how students with the desire referred to in (10) might react to classroom disagreement, perhaps vigorous classroom disagreement, about what is of value and about the ends and purposes of human life.

One response that I believe is likely, and that I believe we teachers often see, is the assumption that the only justification worthy of the name is that which meets a very strong standard of interpersonal justification. Some students think that conceptions of the good are not justified unless their defining claims — their claims about what is valuable or about the ends and purposes of human life — can be backed up by arguments that are recognized on the whole as defeating all actual, reasonable interlocutors. Other students adopt an even stronger position, according to which a conception is justified only if such claims are backed by arguments that defeat all reasonable interlocutors, whether actual or imagined. If some of these students think that moral skepticism is a reasonable position, then they will think that they ought not endorse a conception of the good unless it can be backed by an argument that vanquishes the skeptic. ^{Footnote 29}

Another possibility, and one that I believe we also often see among students, is adoption of a somewhat weaker standard of justification. For someone who adopts the standard I now have in mind, the desire referred to in (10) is a desire to endorse a conception of human fulfillment for oneself only if one thinks that others with whom one is engaged in critical reflection could also find fulfillment in it. It is a desire to endorse a conception of the good only if one thinks those others could also endorse it or accept it as theirs. What this comes to, in practice, is the requirement that a conception of the good be endorsed only if it is not too opaque or incomprehensible to one’s interlocutors, or is one that they indicate they find attractive or could imagine themselves living out, even if they in fact do not.

The first standard of justification undermines the view that any conceptions of the good at all can be rationally justified. It therefore makes embracing a conception of the good a matter, not of endorsement on the basis of vindication by reason or rational scrutiny, but of caprice or emotivism. The second undercuts attempts to justify conceptions which are seen as demanding, outmoded, unfashionably conspicuous or moralistic — as many traditional religious and ethnic conceptions of the good are. Strictly applied, it would probably result in students being willing openly to avow only members of a small and anodyne set of fundamental values — a probability that the facts bear out. ^{Footnote 30} Adoption of either standard of interpersonal justification therefore threatens the justification of traditional ways of life. If either one is adopted by students who receive the kind of education referred to by (10), then:

• C:Education for autonomy threatens traditional ways of life.

This is what I referred to in my introductory remarks as the first argument for C, the argument that goes by way of students’ likely response to pluralism and disagreement.

III. One Argument from (7) to C

The argument for C can be divided into two parts. The first part, running from

1. (1) One of the goals of education is to prepare students to live autonomously.

through

1. (7) One of the goals of education is to inculcate in students the highest-order desires to subject the central elements of their conceptions of the good (a) to critical reflection and (b) to follow those conceptions only if those elements survive, and are believed by them to survive, such reflection.

is conceptual. The second part, beginning with the transition from claim (7) to

1. (8) A conception of the good will be believed to survive critical reflection only if it is believed to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

and running through C depends upon a series of claims about how students are likely to respond to education for critical thinking in a pluralistic classroom. The conceptual claims in the first part of the argument seem to me to be correct, and so I grant the truth of steps (1) through (7). Moreover, as I said in my opening remarks, the empirical claims between (7) and C seem to be supported by the facts and I will grant that they are more or less correct. The argument from (1) to C therefore seems to me to succeed.

But it need not. If the reasoning leading from

1. (10) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection, but who disagree among themselves “about the fulfilled life, the human good [and] the nature of self-realization.”

to C is sound, it is so because students make mistakes about the demands of interpersonal justification. And if the argument leading from (7) to

1. (9) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

and (10) is sound, it is so because students make mistakes about the place of such justification. C is sometimes pronounced with an air of resignation that I would like to dispel. The way to dispel it is to see how these mistakes can be avoided.

It seems clear enough that the standards of justification that are said to be applied in the move from (10) to C are too strong. Whatever justification of a conception of the good requires, it cannot require an argument that can defeat all comers, reasoned or imagined, let alone an argument that can defeat the nihilist or the skeptic. The other standard, while weaker, seems to conflate a conception’s ability to survive critical reflection with its ability to survive peer pressure. Indeed, the standards ascribed to students in the move from (10) to C may seem so obviously wrong that it would be doubtful that they should be ascribed to students at all, but for the fact that doing so explains some observed student behaviors so well. And so it seems clear enough that the way to block the argument for C is to teach students not to rely on these standards. To do that, it will help to say something about the standards they should adopt. It will also help to see why they are tempted to adopt the wrong ones. Let me begin with the latter question, deferring the former until the next section.

I said that (10) might be thought to derive some support from the fact that the desire it says an education for autonomy will inculcate — the desire to act only on an interpersonally justifiable conception of the good — is reminiscent of the desire with which Scanlon persuasively identifies moral motivation. The desire to justify oneself to other people — to have other people understand why we do and believe what we do and believe — are deep and common human desires. It would not be surprising if students developed the desires and, indeed, I believe they should be encouraged to develop them in some form. Nor would it be surprising if students who are aware of the moral and religious pluralism to which (10) refers should acquire that part of the “modern self-understanding” of which Larmore speaks — an understanding of themselves as members of societies in which disagreements about value are inevitable and intractable, and in which diverse interlocutors agree on very little. But which elements of conduct and belief should be justified, and how they are to be justified, must be specified with care. The similarity between the desire to act on a mutually justifiable conception of the good and the desire Scanlon identifies as basic to morality masks an important difference. The desire referred to by (10) should not be equated with moral motivation as Scanlon understands it.

Moral motivation as Scanlon understands it is the motivation to comply with principles “for the general regulation of behavior” that it would be wrong to violate. Conceptions of the good may include such rules, but they are hardly exhausted by them. For conceptions of the good may also include moral ideals, such as ideals of heroism, supererogation, and saintliness. They may include claims about what kinds of acts are virtuous or vicious even if not required or impermissible. They may include personal aims and plans, and hopes and dreams that are lofty and mundane. To endorse and follow a conception of the good is to endorse and to be moved by all of these ideals, claims, and values. Conceptions of the individual or communal good may entail elaborate ways of life for those who accept them. But large parts of them may have no implications at all for those who do not.

If that is right, then it is hard to see what argument there could be for requiring that conceptions of the good in toto be interpersonally justifiable across the divides to which (10) refers. This conclusion underlines the importance of distinguishing those pieces of an ethical view which have implications for the conduct of nonadherents, which impose burdens on them, or which will be coercively enforced, and those pieces whose application is more narrowly circumscribed. ^{Footnote 31} I grant for present purposes that the former must meet stringent conditions of interpersonal justifiability and that the desire to meet those conditions is a basic form of moral motivation. The latter do not need to meet those standards. If students do develop the desire referred to in (10), it is because the object of a desire that is commendable in its place, the place that Scanlon puts it, has been illicitly extended or blurred so that it encompasses far more than it should.

Is there some reason, aside from a failure to draw key distinctions, why students might suppose that their conceptions of the good must be interpersonally justifiable if they are to be justifiable at all?

Perhaps the thought is that conceptions of the good by their nature commit people to moral judgments about the value of the lives of those who reject them. Adverse moral judgments carry offense, and offense, like any other burden, must be justifiable to those who suffer it. ^{Footnote 32} Hence the conceptions of the good that entail them must also be justifiable. But conceptions of the good are not inherently judgmental; they need not entail judgments impugning the value of alternatives. Here — as elsewhere where ethical pluralism is at issue — the history of religious toleration is instructive. While one response to religious pluralism is to deny that “justifiable” and cognate notions apply to religious claims, another is the attempt to develop conceptions of such notions that do not entail religious exclusivism. ^{Footnote 33} The success of this project would show that it is possible justifiably to hold one subset of conceptions of the good — namely religious ones — without lapsing into the judgmentalism fear which I have conjectured leads students to apply interpersonal conceptions of justification to their conceptions of the good. ^{Footnote 34} Success would therefore undercut the motivation for thinking that conceptions of the good are justifiable only if they are interpersonally justifiable. Moreover, though I cannot argue the point here at any length, I do not believe that judgments about the value of someone’s life necessarily convey the kind of burden that requires justification. My own conception of the good entails the judgment that Rawls’s grass-counter, who opts to spend his days counting blades of grass, is wasting much of his life. ^{Footnote 35} While I think it would be impolite for me to tell him so, I do not believe that the entailment provides me with any reason at all to reject my conception of the good or to modify it so that it does not entail the judgment.

Another reason students might be drawn to a strong interpersonal justification condition is suggested by

1. (8) A conception of the good will be believed to survive critical reflection only if it is believed to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

This is the step at which the notion of interpersonal justification enters the argument. In introducing (8), I said that students come to think that conceptions of the good must be interpersonally justified because they are taught to engage in critical reflection in a classroom setting. That setting may be indispensable in acquiring the requisite skills and dispositions, but it can also be misleading. The point of teaching students to engage in critical reflection about the demands they should place on others, including the demands they should place on them in politics, is to ascertain the reasonability of those demands. In these cases, mutual acceptability may be constitutive of justification; indeed, I am prepared to grant that it is. But the point of teaching students to reflect critically on their conceptions of the good in toto is quite different. It is to teach them that conceptions of the good can be supported by rational argument, to think of themselves as the sort of persons who believe on the basis of reasons and to teach them to assess the reasons for their beliefs. In a classroom setting, this teaching takes place through activity that is collective. While this might suggest that being able to justify one’s conception of the good to others is necessary for it to be justified at all, that suggestion is mistaken. But the mistake may well be one that students make.

Thus, while students who apply the overly strong standards of justification referred to by (10) and C make an obvious mistake, they make a more subtle mistake if they suppose that conceptions of the good in toto require interpersonal justification at all, as (8) implies that they will. And they will go wrong if, in acquiring the desires referred to by

1. (7) One of the goals of education is to inculcate in students the highest-order desires (a) to subject the central elements of their conceptions of the good to critical reflection and (b) to follow those conceptions only if those elements survive, and are believed by them to survive, such reflection.

they acquire the desires referred to by

1. (9) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

and

1. (10) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection, but who disagree among themselves “about the fulfilled life, the human good [and] the nature of self-realization.”

It may be true, as (9) and (10) assert, that students who are educated for autonomy will develop those desires, but this is a problem in their moral education that needs to be, and can be, avoided.

IV. Justification and Conceptions of the Good

I have ventured a number of conjectures about how education for autonomy might endanger traditional ways of life, given certain mistakes that students are inclined to make. Aristotle famously said that human beings are naturally political animals. By that he meant that the human good is most fully realized in a polis. But another construction that Aristotle’s claim will bear is that human beings have a natural tendency to extend norms and ideals that properly apply to the political or public realm to other areas of life as well. ^{Footnote 36} If this tendency is indeed natural to us, then it is natural for students to make just the mistake I have claimed they do — namely, applying standards of public justification to their conceptions of the good. How can students be taught to avoid that mistake?

One of the most depressing findings reached by sociologists who study the religious lives of American teenagers is teenagers’ inability to articulate reasons for believing what they do ^{Footnote 37} — an inarticulacy the sociologists attribute to their never having been educated to do so. ^{Footnote 38} If this inarticulacy is to be overcome, students will have to learn to develop justifications for their conceptions of the good, as well as standards by which those justifications are to be assessed. The justifications students offer for their ways of life, including traditional ways of life, will include the goods those lives make available and reasons they are worth pursuing to the exclusion of others. While a good deal of the education in the relevant reasons and values should take place outside a common school setting, in venues like those in which religious education takes place, schools can provide valuable support. For I have granted steps

1. (1) One of the goals of education is to prepare students to live autonomously.

through (7). And so I have granted that education should be for autonomy, that it should encourage the desires to which (7) refers, and that students should be taught to examine the central elements of their conceptions of the good to see whether they are supported by what they take to be good reasons. These are goals at which public schools can aim.

Moreover, students can be taught to share their reasons for their ways of life in classroom discussion so that they develop some fluency in presenting them. Thus I assume that when students learn history, literature, social studies, and current affairs, they learn something about traditional ways of life. They should also learn about the contemporary adherents of those ways of life and, if possible, hear from some of them — including students and their parents who are attached to them. ^{Footnote 39} Managing such discussions will, of course, require that teachers exercise considerable care. Teachers should not be afraid to let students say why these ways of life might be considered valuable, to help them develop and articulate reasons and arguments for their ways of life as for any intellectual position, to help one another think their way into them and question their value, provided the questions are asked — or can, under the guidance of the teacher, be rephrased — respectfully.

A further difficulty facing teachers stems from the fact that no matter how much adherents of one or another way of life may have to say about the value of their life, that value may be fully appreciable only from the inside and so may be only partially communicable. The lapse of communication when reasons come to an end should not be mistaken for the inarticulacy I mentioned just above, and students may have to be taught how to draw the right conclusions when conversation comes to an end. Thus the young woman who finishes discussion of a ritual initiation into adulthood such as a bat mitzvah or confirmation by saying “But there is only so much of this the rest of you are going to get” may already have communicated quite a lot. Students should be encouraged to appreciate the value, and not just the limits, of communication. If at the end of a discussion they are inclined to say “To each her own,” they should also be taught that what makes a way of life one’s own is itself a matter about which a great deal can be and has been said.

I have stressed that conceptions of the good are complex and that they encourage conceptions and ideals of the self. If these views of the self are central, as I believe they often are, then the central elements of a conception of the good will survive students’ critical reflection only if the ideals it encourages can somehow be squared with the conception of the self that is encouraged by public education for autonomy. It will therefore survive only if someone’s ideal of herself as the kind of person who engages in critical reflection about the good can be squared with the view of herself that her conception of the good actually encourages. Can identities associated with traditional conceptions of the good be squared with the identity suggested by (7)? Can the central elements of those conceptions survive critical reflection?

It might be thought that some cannot. For traditionalist conceptions identify reasons on which it is appropriate or desirable to act. Among those reasons are reasons stemming from the venerability of certain patterns of behavior. Traditionalist conceptions that are authoritarian take it to tell in favor of performing some action that it is enjoined by a venerable authority; indeed some may recommend the practice of giving oneself further reasons by taking oaths or promises of deference to such authorities. Communities living traditionalist ways of life may also instruct their members so that they comply with norms of traditional behavior out of habit and are uncomfortable acting against such norms. Acting for all these reasons might be thought incompatible with autonomy ^{Footnote 40} and, it might be argued, students should be educated to reject conceptions that identify them as good reasons for action.

There may be conceptions of autonomy that are incompatible with acting for such reasons. But all that I have granted, and all that is required by steps (3)

1. (3) An agent lives a self-governed life only if the central elements of her conception of the good are vindicated by her own reason.

through (7), is that students be encouraged to follow conceptions of the good that are such that their central elements, and not all the reasons for action they commend, survive critical reflection. It is far from obvious that a conception would fail the condition imposed by

1. (5) Someone lives autonomously only if the central elements of her conception of the good survive, and are believed by her to survive, critical reflection.

and that its adherent would fail to be autonomous, just because it commends acting for reasons that violate a standard of autonomy that applies to reasons rather than to conceptions. Indeed, it seems at least possible that someone could have good reason to follow a conception that commends subjection to religious or cultural authority, and hence the alienation of some kinds of autonomy. It also seems at least possible that following that conception could itself meet a defensible conception of freedom — albeit a conception which, I grant, would require exposure to at least some prominent alternative ways of living and knowledge of politically guaranteed exit conditions. ^{Footnote 41}

A standard of critical reflection strong enough to rule out those possibilities would have to be defended, not just for its philosophical cogency, but also for the legitimacy of aiming at its inculcation through public education. That, I believe, would be very hard to do. The pursuit of public educational aims must, since they are public, be politically legitimate. The political legitimacy of public educational aims requires that they be acceptable to all reasonable persons despite the pluralism noted between (9) and (10). I have supposed that adherents of traditional conceptions of the good are reasonable insofar as they accept the need for interpersonal justification of measures that impose burdens on others. The aim would have to be acceptable to them. Since (I assume) it would not be, that aim would be illegitimate and illegitimate because the standard of critical reflection now in view would be too strong.

I have acknowledged that teaching students not to apply public standards of justification to their conceptions of the good will be difficult. So, too, will teaching them to avoid what is virtually the opposite mistake, that of drawing on their conception of the good as the primary justification for political positions. I have discussed this challenge elsewhere. ^{Footnote 42} I shall not pursue it here, except to note that it requires a further act of reconciliation on the part of traditions that make room for political participation, for it requires students to reconcile the self-conception encouraged by traditionalist conceptions of the good with that of the conception of oneself as a citizen of a pluralistic democracy.

It follows from this requirement that students being educated for autonomy need the skills required to bring the demands of their various self-conceptions into harmony. This implication is just what we should expect. Because of great social differentiation of modern life, ^{Footnote 43} all of us occupy a variety of roles and have a number of identities. One way to impose some unity on our lives would be to reduce the number of identities that matter to us. This is something that education could be designed to encourage. But such reduction is at least sometimes procrustean and, hence, deforming, and often results from a false essentialism that can be politically toxic. ^{Footnote 44} Better to educate students to confront their plural identities, to determine which ones they have good reason to hold onto, and to learn the skills they need to negotiate and balance the demands of those identities as they wish. I recognize that this will not be an easy task and that it may be one to which public education can make at most ancillary contribution.

V. Another Argument from (7) to C

I have argued that students should be taught to engage in critical reflection on their identities and on the persons they aspire to be. But according to the second argument for C that I referred to at the outset, encouraging that kind of reflection is incompatible with traditional ways of life.

To see how this second argument goes, recall that in Section II I said that traditionalism is a thesis about value and the reasons for action to which it gives rise. Among the sources of value that traditionalism identifies is the venerability of a pattern of behavior including, in the case of some traditions, the pattern of deferring to a venerable authority. The value of venerability gives adherents of a traditional way of life reasons for action. But it might be thought that traditionalism includes a further thesis according to which the fact that one is moved by the force of those reasons, and so adheres to a traditional way of life, is not itself to be subjected to the kind of critical reflection that is undertaken to justify adherence. More crudely put, the thesis is that reflection-for-justification destroys adherence of the right kind. I believe that G. A. Cohen relies on this thesis when says:

What is the argument for the thesis? Cohen does not say, but perhaps it is the following. Someone who reflects on her adherence to her own religion or culture, and sees that she has reason to treat it as part of her good, thereby acquires a new, higher-order reason for adhering to it: namely, the fact that she sees that she has good lower-order, traditionalist reasons to treat her adherence as good for her. That higher-order reason is not just one more reason to adhere to her religion and culture, and to engage in the behavior it prescribes. Rather, it preempts or replaces some or all of the reasons that stem from traditionalist values. So a result of critical reflection and affirmation of a traditional way of life is, paradoxically, that someone who engages in it thereby ceases to adhere to a traditional way of life as such. ^{Footnote 46} It follows that to be a traditionalist Catholic or orthodox Jew or member of the Old Order Amish requires that one accept the guidance of one’s tradition unreflectively or unquestioningly.

If this line of thought is right, then educating students for critical reflection — and encouraging them to think of themselves both as critical inquirers and as adherents of traditional ways of life — will indeed threaten those ways of life. And indeed the alleged tension between education for autonomy and adherence to traditional religion is sometimes presented in almost exactly these terms. ^{Footnote 47} Thus the traditionalist thesis that Cohen suggests, if it is right, would make it possible to bypass the empirical claims that supported steps

1. (8) A conception of the good will be believed to survive critical reflection only if it is believed to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

2. (9) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection.

and

1. (10) Education for autonomy will inculcate in students the highest-order desire to endorse a conception of the good only if they believe it to be justifiable to reasonable interlocutors who are engaged with them in critical reflection, but who disagree among themselves “about the fulfilled life, the human good [and] the nature of self-realization.”

and to move to

• C:Education for autonomy threatens traditional ways of life.

from

1. (7) One of the goals of education is to inculcate in students the highest-order desires (a) to subject the central elements of their conceptions of the good to critical reflection and (b) to follow those conceptions only if those elements survive, and are believed by them to survive, such reflection.

via a further conceptual claim about adherence. Is the thesis correct?

The argument for the thesis depends upon the assumption that reflection on the merits of a tradition, and the reasons for the patterns of behavior it identifies as valuable, cannot itself be a venerable pattern of behavior that is partially constitutive of that tradition. There are many reasons to doubt that assumption, some provided by conceptions of the good that serve as counterexamples. Reflection on the goods of ritual observance has long been a part of Christianity. Reflection on how the goods of the Amish community can be accommodated and preserved in the face of modernity seems now to be a settled feature of Amish life. ^{Footnote 48} And Alasdair MacIntyre has, of course, argued at length that traditions, including the Thomist-Aristotelianism he favors, can only be vindicated by extended reflection on and argument about their own intellectual foundations and about the goods of the social practices the traditions structure. ^{Footnote 49} In fact, I think that traditional ways of life need to encourage and to draw on the results of critical reflection. I want briefly to explore one reason why they should do so.

I have said that traditional ways of life foster higher-order desires to be a certain kind of person: a person who performs some actions and not others, who acts for certain reasons and not others, who strives to conform her life to models identified by the tradition as especially worthy of emulation. These desires and their objects come into play in the practical reasoning of a tradition’s adherents. In deciding how to act, they may ask themselves what action is right or wrong, permissible or impermissible. But they may also ask themselves what action would instantiate virtues singled out by their tradition as important; what an exemplary adherent of their tradition would do; what action best comports with the essentials of their way of life; what behavior would best instantiate it and carry it on or, on the contrary, what action would constitute a betrayal. In asking these questions they are asking, in effect if not explicitly, how to maintain and act on their identity as adherents of their way of life. Reflection on this question is something bearers of any tradition can normally be expected to engage in.

Sometimes such reflection yields an answer that is unwelcome or demanding. Adherence to a traditional way of life may require action that is not easily undertaken, and the temptation to defect may be strong. Faced with such temptation, someone may ask why she wants to adhere to the tradition. The results of critical reflection on the value of her way of life and her reasons for adhering to it help to answer that question. Cohen may well be right to imply that “the more you have to [rely on this answer], the more you have to reaffirm the tradition, the more likely it is that the grip of tradition [on you] is relaxing.” But the relaxation of that grip might not be a permanent or irreversible loosening. It might instead be the temporary groping that is part and parcel of coping with temptation. A picture of traditional ways of life according to which they are incompatible with explicit critical reflection on reasons for maintaining or adhering to them is a picture of traditions that deny their adherents the ability to respond to the temptations that are an ordinary part of human life. It is an unrealistic picture to which I suspect that few ways of life — and no enduring ones — have ever conformed.

The need for such reflection is also apparent when we shift attention from first-personal to interpersonal deliberation. Many ways of life depend for their integrity on adherents’ holding one another accountable for their behavior. This is true of occupational ways of life such as the professions; it is true for many religious ways of life as well. The act of holding accountable normally includes appealing to the identity that adherents of the way of life share and to the ideals to which they are assumed to be committed. Thus someone may say to the person she is counseling or calling to account that the conduct he is considering would be a disgrace (or a credit) to their profession, or that a faithful and observant member of their community would never engage in it (or would be proud to do so). Faced with the question of why that should matter, of why he should care about being a faithful adherent of their shared way of life, it will be necessary for the person holding him accountable to draw on the results of critical reflection about why that way of life is valuable and what reasons there are for him to carry it on.

In these cases of interpersonal deliberation, unlike the cases of first-personal deliberation, the identity at stake need not be thought to have loosed its grip even temporarily on the person drawing on the results of critical reflection. Cohen may be correct to say in these cases that “the more you have to [rely on the reasons turned up by critical reflection], the more you have to reaffirm the tradition, the more likely it is that the grip of tradition [on someone] is relaxing.” But that is not to say “the more likely it is that the grip of the tradition [on you] is relaxing” since it is relaxing, not on the person who is adducing the reasons, but on the person to whom the reasons are addressed. The person bringing forward the higher-order reasons to adhere to the tradition must be aware of those reasons and aware of the fact that she can deploy them with reason-giving force. But her own reasons for adhering to her way of life may be those that stem from traditionalist values.

If I am right about first-personal and interpersonal deliberation, then reflection-for-justification is not just compatible with the kind of adherence that traditional ways of life presuppose. It is necessary to sustain such adherence. The additional traditionalist thesis that promised to underwrite the move directly from (7) to C is false, and C is unsupported by the second argument. The first argument does support C, but it does so because it imputes to students certain interesting and understandable mistakes about the kind of justification autonomy demands. I believe that once those mistakes are recognized as such it should be clear that education for autonomy, properly understood, need not threaten traditional ways of life even if simultaneously educating students to appreciate both tradition and autonomy is a task of extraordinary complexity.