XI

Character and the Survival of Liberalism

"Liberalism--it is well to recall this today--is the supreme form of generosity; it is the right which the majority concedes to minorities and hence is the noblest cry that has ever resounded on this planet. It announces the determination to share existence with the enemy; more than that, with an enemy that is weak. It was incredible that the human species should have arrived at so noble an attitude, so paradoxical, so refined, so acrobatic, so anti-natural. Hence, it is not to be wondered at that this same humanity should soon appear anxious to get rid of it."

--José Ortega y Gasset, The Revolt of the Masses

1. Introductory

As a political reality, liberal democracy began in America and Europe with the revolutions of 1776 and 1789 and has since then settled securely in every inhabited continent. By both violent and peaceful means, it has dealt death-blows to its major ideological enemies, including slavery, authoritarian monarchy, fascism, and communism. Today, it looks as if the final and complete victory of liberal democracy is within sight at last.⁽¹⁾

Despite this fact--the fact, that is, that things look this way--I propose in this chapter to examine reasons for thinking that liberal democracy is in the long run not really a viable system. If a single political system is about to spread over the entire world, that system might not be liberal.

More precisely, I will entertain and take seriously the possibility that liberalism as a political system is inherently unstable, that it tends to undermine itself. In sections 3 through 6 I will argue that we have good reason to expect it to encourage the formation of a trait of character that drives individuals to do things that undermine liberal institutions and that prevents them from caring about certain values on which such institutions are based. It leads people to do illiberal things, moreover, precisely in the name of liberalism. However, I will also argue in sections 7 and 8 that there is at least one sort of liberalism that does not have this sort of instability. This is because it fosters traits of character that work in ways contrary to the vicious one encouraged by the other sort of political culture.

It will be evident that our own culture has, so to speak, not decided which sort of culture it will be. As Ortega y Gasset said in a related context:

"It is not that the present situation may appear to us good from one viewpoint, and evil from another, but that in itself it contains the twin potencies of triumph or of death."⁽²⁾

Before I try to show any of this, though, I must say something about what, as I take it, liberalism is.

2. Minimal Liberalism

Liberalism has taken many different forms, but it always rests on attachments to two goods that are taken to be of fundamental importance. These two basic values are liberty and equality. The relative strength of these attachments, and the political principles that articulate them, are among the things that distinguish one form of liberalism from another.

Both these basic values have been interpreted very differently by different liberal thinkers. The liberties affirmed may include extensive rights to do things that have profound effects on others, or they may be limited to the rights to regulate one's own purely "private" affairs and to express one's opinions. The equality involved might include rights of equal access to the education and medical care systems, or it might be far more restricted.

There is at least one limit, though, to the ways that the varieties of liberalism can differ one from another: though the extent to which each of these fundamental goods are affirmed may vary widely, in neither case may the affirmation be reduced to utter triviality. Some liberals seem inordinately attached to liberty, while others appear to be very partial to equality, but no ideology could be called liberal unless it affirms the value of both goods. At a minimum, liberalism must include something equivalent to two familiar principles: (1) people have a right to freely express their views, and (2) they have a right to the equal protection of the laws.

Further, liberalism does not treat the two fundamental goods in precisely the same way. In a certain sense, liberals, properly so called, value liberty (at least some liberties) above equality: whatever the liberties are that the liberal holds sacred, he or she will not sacrifice them for the sake of equality.⁽³⁾ At a minimum, this means that the liberal (in contrast to most nondemocratic socialists) will not try to bring about a more egalitarian society by suspending freedom of speech, not even as a passing stage on the way to a better world. These two principles, together with this constraint, comprise what might be called "minimal liberalism."⁽⁴⁾

The instability that tends to afflict certain sorts of liberalism has to do with the two principles of minimal liberalism. It arises from the fact that, in the sort of society liberalism brings with it, they can easily become incompatible, so that one of them operates to interfere with the secure operation of the other. One way to express the nature of this interference would be this: The first principle indicates that a liberal order rests on tolerance; if not on the virtue of tolerance, then at least on tolerant behavior in which one forbears to interfere with the expressive conduct of others.⁽⁵⁾ The second principle, however, can place such behavior in conflict with the natural and legitimate concerns of self-respect. When this happens, we can no longer expect people to adhere to the first principle.

I will only be able to make my thesis plausible if I first try to describe, briefly and impressionistically, what liberal society is like. The relevant features will perhaps stand out in clearest relief if I contrast it, also broadly and impressionistically, with the sort of institutional arrangements that liberal institutions have tended to displace wherever they have appeared.

3. The Preliberal World

For two hundred years liberalism has been making a revolution in the circumstances of everyday life so great that it is very difficult to understand it adequately. This is partly because the revolution is still well under way, and, in consequence, we are all participants and carry all the biases that participation brings. It is also due in part to the fact that the break with the past was deep and surprisingly complete. As Ortega says, "never in the course of history had man been placed in vital surroundings even remotely similar" to those that liberalism brought with it.⁽⁶⁾

Throughout history, most human beings have found themselves confronted by an array of practical requirements, of things that must be done. If they cared to consider how well their lives were going, they did so by considering how well these requirements were being met. Will I have enough food for the winter? Can I avoid angering the gods? Can I pay my taxes? These requirements they saw as placed on them by beings external and superior to themselves. Their superiority was established by, if nothing else, the fact that it was from them that these standards came.

Ortega describes the general character of the preliberal parts of human history like this:

"For the 'common man' of all periods, 'life' had principally meant limitation, obligation, dependence; in a word, pressure. Say oppression, if you like, provided it be understood not only in the juridical and social sense, but in the cosmic."⁽⁷⁾

What Ortega calls the "cosmic" sort of pressure is undoubtedly very important for understanding the difference between liberal and preliberal societies. Liberalism brought with it a striking degree of relative prosperity, which liberated huge masses of people from the need to devote most of their time to the simple end of producing what they needed to stay alive. However, the sort of pressure that I wish to single out and discuss is the other one mentioned by Ortega, the sort he calls juridical or social.

Most people in the societies that immediately preceded liberalism were confronted with a considerable and burdensome array of obligations that were placed on them by tradition. In one way or another, this is true of people who live in any sort of society, but certain of the traditional obligations in preliberal societies have two characteristics that, jointly, separate their world from our own. Together, they distinguish these social systems as instances of what, following a famous discussion by Sir Henry Maine, I will call "societies of status."⁽⁸⁾

First, in such societies, the traditional mores defined groups of people, and, in addition, they defined various obligations in such a way that, once an individual is a member of one of these groups, that individual cannot avoid a very significant burden of obligations of that sort. Second, membership in these groups was itself not a result of any voluntary action on the individual's part. Tradition placed obligations on individuals in a way that excluded individual choice. These obligations are the unchosen consequences of involuntary group membership. A few examples might help to reveal exactly what this means.

The example of a society of this sort that comes most readily to mind is a caste system. In a pure caste system, there are customs or laws that assign important obligations to people simply by virtue of their membership in their political or economic class. The obligations involved both require and enable these people to carry out the political or economic functions that, again simply by virtue of class membership, are properly theirs. Because they are attached to class functions in this way, these obligations are identical throughout a given caste and are not shared with the rest of the human race.⁽⁹⁾ To the extent that people inherit their caste membership from their parents, and cannot abandon it by individual choice (the peasants, for instance, are tied to the land), it is obvious that such systems fit the description I have just given of the assignment of obligations in societies of status.

Another example, somewhat less obvious, of a system that answers to this description is one in which the family structure is characterized by a certain sort of patriarchy. A rather extreme instance is to be found in the early Roman Republic as described by Maine. According to him, this was a system in which the law and its obligations only served to regulate the foreign relations between families. Within the family the only law, if we may call it that, was the will of the "despot enthroned by each hearthstone," the head of the family.⁽¹⁰⁾ In the earliest and purest form of this system, this meant that sons and daughters were subject for life to the arbitrary decrees of their father, which he could enforce with any sort of punishment, including death.⁽¹¹⁾ Thus, the individuals who occupied the status of son or daughter were vulnerable, just because they belonged to those categories, to a stream of unavoidable obligations. Of course, the status of son or daughter is, except in some cases of adoption, one that is occupied involuntarily.

As this example indicates, the sort of society I have described does not necessarily involve fixed duties that are unalterably built into the position occupied by the person to whom they apply. In principle, the obligations involved may be flexible, or even unpredictable. What makes a system a society of status in my sense is not the content of the relevant obligations but their source. The arrangements that Maine attributes to the early Roman Republic qualify by virtue of the fact that important obligations are based on group membership, and because membership in these groups is, at least typically, not the result of the voluntary conduct of the individual member. That is sufficient.⁽¹²⁾

4. The Liberal World

Liberal democracy has shown a powerful tendency to build a world that is very different from the ones I have just been describing. My brief discussion of liberalism in section 2 suggests one very likely reason for this: it is very difficult to reconcile these earlier arrangements with either of the two basic values of liberalism: liberty and equality.

Anyone who places a high value on liberty would, to the extent that they do so, be uncomfortable with the idea that there can be significant obligations unilaterally imposed on one, either by tradition or by a human overlord set up by tradition. Such authorities--for that is what they are--seem to function at the expense of individual freedom. People who are strongly attached to the value of liberty will sacrifice it only if it is necessary to do so. As liberal society has evolved, these particular sacrifices of liberty have lost their original appearance of necessity: other principles of social order seem to have served at least as well.

In much the same way, the old ideas have proved to be unattractive to people who value equality. When liberalism first became a formidable political fact in Europe, the serious competitors it encountered were examples of a particular sort of society of status: they were remnants of caste systems, in which many people were trapped in groups that played subservient roles in relation to other groups. This imputation of subordination seems to be an inevitable feature of such systems. The functions served by the various groups--such as ruling, fighting, tilling the land, and so forth--always seem to be such that people have more or less definite notions to the effect that some are superior to others. In practice, this has always given the impression that the members of at least one group are inferior to the individuals outside their group to whom they owe their obligations. This is true even when the social system is divided into groups each of which have duties to all the others, so that it realizes in some way the moral ideal of reciprocity.⁽¹³⁾ The subservience meted out by these systems was particularly severe when there was no way the individual could escape it.

Of course, the liberal conception of equality is not incompatible with hierarchy of every sort, but it does mean that, if we are to burden people differently, there must be some good reason for doing so. Over the years, people who live in the liberal nations have found it increasingly difficult to give good enough reasons for the traces of preliberal hierarchy that remained in their world.

The tendency of liberalism has been to develop moral traditions that are altogether inimical to traditions in which individuals are assigned to groups and obligations are assigned to them as group members. Naturally, there must be obligations that are not the result of voluntary actions of the individual who has them, but with the destruction of the old way the only such obligations that can remain are ones that apply to everybody.

Clearly, however, no moral tradition could consist solely of such universal obligations. A society thus limited would be one in which what each individual owes to all the others is absolutely identical; it would in this sense be devoid of what one philosopher has called "moral particularity."⁽¹⁴⁾For many reasons, including a sufficient number of obvious ones, the moral entanglements of unique individuals must be to some extent unique. Liberal societies have avoided unarticulated sameness by relying on certain moral practices that, while some of them can probably be found other sorts of social systems, have nonetheless assumed unprecedented importance in the liberal democracies. This transformation has resulted in the formation of systems that, again following suggestions from Maine, I will call "societies of contract."⁽¹⁵⁾

To the extent that individuals live in such a society, they see their differential obligations, the obligations they do not share with all of humanity, as foreseeable consequences of their own voluntary conduct.⁽¹⁶⁾ Sometimes, the conduct involved literally is the signing of a contract of some sort or other, but perhaps more often it is one of the many other sorts of intentional actions by which we become entangled in the lives of others, including joining a club, enrolling in a school, entering a profession, or taking part in the give and take of an intimate relationship over a long stretch of time.⁽¹⁷⁾

Such actions, unlike the preliberal sources of differential obligation, are not violations of our freedom: they are actually expressions of it. Further, the obligations involved arise from conduct that, in principle, any adult can do, and not from membership in a group that sets one above or below other people. Thus, moral particularity, on the face of it at least, is created in ways that are compatible with the liberal ideals of freedom and equality.⁽¹⁸⁾

5. Finding Out Who One Is

Insofar as earlier societies were based on status rather than contract, they presented people with a prefabricated array of moral obligations. They did so, moreover, in a particular way: by presenting them, in a certain sense, with a prefabricated identity. If I were born into the hereditary caste of priests--or musicians, or tillers of the soil--this classification would indicate to me a significant part of what I would be doing with my life, and of what would be important to me as I lived it. This classification would tell me something important about my obligations and--at least as important--these obligations would indicate how important this classification was. They would indicate to me that my caste membership is not just another fact about me, like my being tall, deep voiced, or blonde: it indicates to me what I am.

Admittedly, there is a sense in which, even after status has been replaced entirely by contract, one could say that society still presents the individual with a prefabricated identity. Convention gives a certain meaning to words and phrases such as "intelligent," "slovenly," "of Italian ancestry" and it may be that, independent of my own voluntary actions, these conventions indicate that I am all of these things. Someone who is willing to make a certain philosophical assumption might conclude that it is in that case true by convention that these things are true of me and that, consequently, it is society--the social group that is governed by these conventions--that makes it so.

Even if this is so, however, there is still a large difference between this social group and a status system. If these words and phrases apply to me by linguistic convention, they do so merely as an unassorted collection of epithets. These conventions do not indicate to me which of them are important and which are trivial. This is of course what people are wondering about when they seek to find out who or what they are, when they seek to discover their own individual identity: they wish to know, not merely which attributes do they have, but which are the important ones. This latter question is what, in large part at least, a status system settles for us. In such a system, there may be little need to seek to find one's own identity; it is, to some extent, settled.

What Maine called "the movement from status to contract"⁽¹⁹⁾ makes it possible for a social system to contain differential obligations without assigning individual identities. To the extent that this transformation has occurred, it is possible to sever being from doing, to separate the question of who or what I am from that of what I morally ought to do. This would mean that I am able to find out what special obligations I have toward others without knowing who I am--without considering my race, class, or gender, for instance--but simply by considering what I have done.

To the extent that a social system develops under the influence of liberal ideas, it seems inevitable that this separation should occur. As I have already suggested, it seems unlikely that any society can function unless its method of assigning moral obligations allows different people to have different obligations. We cannot bear our obligations simply as members of an undifferentiated mass. Yet liberal societies are constrained as to how they bring about the needed differentiation. They must avoid presenting their inhabitants with prefabricated identities. Dictating to people who they are to be represents a highly undesirable sacrifice of liberty. Moreover, since one's identity always seems to be in some way good or bad, it constitutes a very serious sacrifice of equality as well. Of course, liberal institutions must be based on assumptions about what people are like. It would be desirable, however, to make no assumptions about any one person, independent of their overt behavior, other than basic principles that apply to everyone: for instance, that people are able to think, to make choices, to act on their choices, and to accept responsibility for foreseeable consequences.⁽²⁰⁾ But a contractual system can be based on precisely such universal principles as these. Thus, it seems inevitable that, unless there is some powerful counteracting force at work, liberal societies will, so far as it is feasible to do so, complete the transition from status to contract, and thus free themselves from the necessity of socially assigning identities.

This suggests rather strongly that in such a society the solution to the issue of individual identity is entirely up to the individual. Finding the answer to it is a solitary enterprise, something the individual does alone. This suggestion, if anyone should take it and believe it, would be an error. The reason it would be an error is, simply, that it is doubtful that there is such an entirely solitary enterprise as (successfully) finding out who one is. At least, it is not an enterprise on which, in any society, the general run of humanity might be embarked.

This is a far from trivial fact, and it helps to explain deeply important features of liberal regimes. I will discuss these implications in the next section but, before I do that, I should briefly recount some of the reasons (in case they are not sufficiently obvious) why the search for identity is something individuals cannot not be expected to do entirely on their own.

First, there are probably many individuals who, to some substantial extent, inevitably absorb their ideas about themselves from other people. The opinions of others exercise an influence over them that they cannot come close to entirely resisting. They accept some of the opinions of others, not as conclusions that they are able to confirm somehow, but simply because they are opinions and are held by someone. Discovering their own identity, insofar as such individuals are able to accomplish it, is to a significant extent not their own doing.

Of course, it is likely that not everyone is like this. There probably are people who avoid, with at most trivial exceptions, this sort of uncritical absorption of the ideas that others have about them. Rather than substituting the judgment of others for their own, they think for themselves. Yet to the extent that it produces any real understanding, this thinking for oneself is seldom, if ever, something that is done entirely by oneself, at least when the thinking that is done is about oneself.

One reason why this is so can be found in the fact that the self, as a subject matter of inquiry, presents the inquirer with some serious difficulties. My own identity is something about which I have very strong motives for deceiving myself. As I have already suggested, it always seems to be either good or bad. The traits I am most likely to find important--traits of character, skills or incapacities, and emotional dispositions are obvious instances--are seldom if ever entirely neutral. They are usually something to be either proud or ashamed of. If I am left entirely to my own devices, the promptings not only of vanity but of the perfectly normal need for self-esteem will certainly lead me to suppress, augment, or slant the evidence.

Further, whether we distort it or not, this evidence is peculiarly difficult to collect. Simple introspection, direct inspection of the contents of one's own mind, is of far less value in these matters than one tends to think it is. The most important statements about ourselves do not take the form of descriptions of simple mental events.⁽²¹⁾ Consequently, to know whether some such statement is true requires that I do something more than to turn my mind's eye inward and see whether some event has occurred within me. To some extent they seem to function more like explanations of events rather than descriptions of them, and usually serve to explain an array, often a large one, of occurrences both mental and physical.

Knowing whether one of them is true has all the prerequisites that the art of explanation has. Among other things, I must be aware of other plausible explanations for the same facts, and I must have good enough reason to believe that the one I am giving is the best of them. Does my inability to accomplish much lately indicate that I have become self-indulgent, or does it mean that I have become understandably depressed because life has disappointed me so many times? Perhaps it is merely some undiagnosed physical illness that has lowered my spirits. The picture presented by the available evidence is often radically incomplete. It is understandable if my desire to think well of myself steps in and paints over the white spots on the canvas.⁽²²⁾

Clearly, substantial difficulties stand in our way as we try to acquire solidly based conceptions of who we are. The ones I have just described, however, are not hopelessly insurmountable. There are routes by which they can sometimes be circumvented, and the most readily available ones by far involve relying on the participation of other people.

This is probably most obvious with regard to the problem of resisting one's urge toward self-deception. With few exceptions, I can rely on others to be far less biased in my favor than I am. It is not, of course, that I am biased and they are not: their biases are rather in other directions than mine. We can rely on each other, as individuals with clashing and thus complementary prejudices, to test and check our ideas of ourselves.

Similar things can be said of our efforts to overcome the problems that are inherent in the explanatory nature of many of the relevant ideas. Finding all of the most plausible explanations for something and finding the best reasons for preferring one explanation to another are undertakings that rely on one's powers of imagination. In difficult cases, even when the thing to be explained is my own behavior, my best chance of success requires me to avail myself of imaginations that are not my own. Possibilities that are invisible to me might be visible, even obvious, to other people.⁽²³⁾

We learn from points of view that differ from our own, and this especially includes ones that are adverse to ours. This adversity is an indispensable source of insight. A look of surprise on someone's face as I say something that seems obvious to me can, and often should, lead me to reconsider my ideas and either abandon them or be sure I have good reasons for keeping them. It is at least arguable that no one ever arrives at a substantial sort of self-understanding without at some point gaining insight from collision with the viewpoints of others. What is more than arguable is the fact that, for most people, finding out who they are is not something they do on their own.

6. Identity as a Datum and as a Task

To the extent that it withdraws from people their socially prefabricated identities, liberal society has not compelled them to find out who they are on their own. The search for one's identity cannot be a solitary enterprise, not at any rate if it has a reasonable chance of being successful. Yet it seems it must be carried on somehow or other. People seem to have an ineradicable need to have some definite and well-grounded notion of who or what they are.⁽²⁴⁾ How then can the search be carried on in a liberal society? An answer to this question is already implied by what I have said in the preceding section of this chapter. I will now make it explicit.

The attempt to understand who we are, like any investigation, must be carried on by means of symbolic behavior, behavior (such as using words, making pictures, or putting together mathematical formulae) that communicates messages by representing things. This means that one of the ways in which investigations can be classified is in terms of this symbolic behavior: who is doing the representing, and to whom are these things being represented? Based on how these two questions are answered, we can classify these processes as "monological" and "dialogical."⁽²⁵⁾

To say that an investigation is dialogical is to say that at least some of the symbolic behavior by which it is carried out consists of an exchange between two different representers, in which each one serves both as author of a message (or messages) and as audience for the message(s) of the other. At least some of the messages sent by each side are received by the other and influence in some significant way subsequent messages sent by the other. (For these purposes, messages can be questions.) Examples of such symbolic behavior include courtroom cross-examinations, Socratic dialogues, most ordinary conversations, and many wordless encounters between individuals that transpire by means of looks and gestures only.

On the other hand, an investigation is monological to the extent that the relevant symbolic behavior is not dialogical: if either (a) the representer and the audience are identical or (b) one functions only as representer and the other only as audience. An example of this sort of symbolic behavior can be found in Descartes's Meditations, in the thoughts the author attributes to himas he sits quietly by his stove, finding out what he knows and what he does not know.

Among the processes in which people putatively find out who they are, an obvious example of a monological one would be the sort that is carried out (assuming this ever happens) when individuals figure out who they are entirely on their own. This of course is the asocial process that I have said is not on its own a feasible way to find out who one is. Another instance of a monological process, however, is social in nature and has proven over the centuries to be entirely feasible. This is the one that is distinctive of preliberal societies.

In these cases, the party that sends the messages, and in so doing attributes an identity to someone, is not an individual human being but some superindividual entity, such as law, custom, or tradition. The audience is the individual to whom the identity is attributed. If the relevant entity informs me that, in one important respect, what I really am is a farmer or the son of a living father, there is no possibility that I will be able to send messages--such as questions, objections, or denials--back to it.

Things are very different when the process by which identity is found is, as I am putting it, dialogical. In that case, no one involved in the procedure is elevated to the position of an unanswerable speaker and no one is trapped in the role of a passive audience. On the contrary, any message can be answered, and the answer can influence subsequent messages from the other side. The outcome of the process--the message that the participants come away from the process believing or acting on--is determined by the give and take among all who have participated in it. It is a cooperative enterprise in which no one participant has the power to determine the final outcome. Indeed, there is no outcome that is final, in the sense of being unrevisable; there is only the outcome of the moment. Further, that outcome, for each participant, consists only in what he or she is actually convinced is so. Thus, the result of any one process is not only revisable but also (at least potentially) plural.⁽²⁶⁾

It seems obvious that liberalism requires the dialogical approach. As I have said, identity must be sought either monologically or dialogically. For all the reasons I gave in section 5, the sort of monological process in which the definer and the defined are the same person is not feasible, at least most of the time and for most people.

The other sort of monological procedure cannot coexist with liberalism for very long. This is the one in which the definer and the defined are different persons, the latter occupying the position of mere audience in relation to the former.

One reason for this incompatibility has to do with the first of the two principles that I attributed to "minimal liberalism" in section 2, the one that maintains that people have a right to freely express their views. This right opens all important subjects to dialogue with other individuals, including the subject matter of the monological process. An American teenager's immigrant parents tell him that ancient tradition dictates that his place in society requires that he obey his mother's older brother or prepare for the priesthood, he may find this fact impressive. In a liberal society, however, he will hear a clamor of many voices speaking on this issue, and the other voices will, unlike the voice of ancient tradition, answer questions and give reasons if challenged. As many immigrants to liberal countries have testified, the eventual effect of these voices is to take away the tendency to believe what the monological messages say and, with it, the will and faith needed to help transmit them eventually goes as well. The liberal dialogue washes away the preliberal monologue.

This means that, in a liberal society, people must for the most part seek their identity dialogically.⁽²⁷⁾ This in turn indicates that life in a liberal society will not be experienced by all as an absolutely unmixed blessing. Previously, one's own identity was to some extent a given; now it is simply a task, and one that to a significant extent is carried out in concert with others. Why this should diminish the blessings of liberalism can be seen by analogy with another important feature of liberal society as we have known it: the creation of differential obligations by means of contract.

The last time I lived in an apartment, the rent I owed for my home was not set by obligations I had to a lord to whom I was bound by the unavoidable accident of birth, nor was it imposed by tradition. Nonetheless, my obligation to pay the rent was obviously not entirely my own doing. The matter of whether I owe rent, and how much, and to whom, was settled by me and by my landlord. I could not have had any particular obligation in the matter unless we both agreed that I had it.

This means that we both acted under constraints that were sometimes strongly felt. The rent always seemed high to me and went painfully higher every year. I regretted that I could never find a landlord who would rent a similar apartment to me for less money. My landlord, on the other hand, often complained that the rent was so low that he could barely make money on our building after his expenses were paid. He probably was sorry that he apparently could not find an equally reliable tenant who would pay more than I would. My obligation to pay the rent, and my landlord's right to that rent, was the outcome of a process that we both found painfully frustrating at times. The reason was that we both participated in the process fully: unlike the lords of old, neither one of us could have things his own way.

What I have just said about the task of determining one's obligations by means of contract can also be said, at least if we speak on a sufficiently high level of abstraction, of the task of discovering one's identity dialogically. That is, in each case the task is carried out by means of an activity in which we cooperate with others and, in roughly similar ways, the necessity of the other's participation can be painfully constraining.

I want to have a low price for my home and for everything else I must pay for, and in a contractual system I am free to influence my trading partners in that direction, but they will not prove to be completely malleable. I make certain conjectures about who I am, mainly very flattering to me, and I try to convince my fellow participants in the dialogue to confirm these conjectures. They, however, have perspectives of their own, and are not biased in my favor in just the way I am. They are not likely to confirm my preferred self-conception in every detail. Their behavior may even attribute to me a failed or deficient identity. They might, for instance, attribute vices to me or imply that I lack admirable skills. I may not be able to think that I am who I would like to be. Just as the contractual system does not allow me to pay just what I want, so the liberal dialogue may make it difficult to think just what I want about myself. I may even find it very difficult to achieve self-esteem.

In a liberal society this a constant danger. This fact poses a problem for the stability of liberal societies. If what I have been saying here is substantially true, liberal society thrusts us into a situation in which the words of other individuals, and all their conduct capable of communicating thoughts about us, have great power both to help and to harm us. How will the inhabitants of such a world be apt to use this power?

In such a society most people, or most influential ones, accept the first principle of "minimal liberalism," the one that grants them freedom of expression. To the extent that they do accept this principle, they will apparently feel free to use their considerable power as they see fit. This would mean that they will feel free to withhold from us the support without which self-esteem may be difficult or impossible to achieve.

However, the constant threat of damage to which the first principle thus exposes us runs counter to the values represented by the second. The second principle, which promises the equal protection of the laws, identifies as a basic function of the liberal system the provision of security from harm by others.

The problem for the stability of liberal society lies in the fact that, to the extent that people are influenced by this idea, they will tend to react in a particular way to the insecurity created by the first principle. Looking at the world from the vantage of the second principle, we tend to view acts that harm others with suspicion, and we tend to view acts that do serious harm with strong suspicion. What we would suspect is that these acts are injustices, for what the principle promises is a certain security from harm as a matter of right. Of course, against harms to oneself one will be ready to respond with particular intensity.

The effect is reinforced by the liberal value which the second principle serves to formulate: the value of equality. The liberal dialogue exposes the individual to the possibility of a daily bruising. The acts by which the bruises are inflicted are challenges to one's self-respect: they are degrading or demeaning. On the other hand, those who do such acts often appear high-handed, as if they are somehow lording it over us. Thus the insults, derogatory comments, and the mere cold refusal to support one's preferred self-conception are typically felt by those who suffer them to establish a certain relation between higher and lower, in which the sufferer is subordinated to the one who is inflicting the suffering. Often, the bruised individual feels an emotion that is the inner correlative of a shouted accusation: "Who do you think you are!?"⁽²⁸⁾

7. A Vice Without a Name

Part of the problem is this. These suspicions, the suspicions that other people are committing acts of injustice simply by expressing their views, together with the fact that the apparent injustices are degrading to oneself, place terrific pressure on the principle of forbearance that underlies the virtue of tolerance. Together, they can make it seem necessary to react to such acts in ways that include censuring them and, often, go a good deal further than that.

What this means is that a liberal society will tend to produce a trait of character that acts in ways that are directly contrary to tolerance. This trait will be a particular sort of resentment. It will differ from other sorts of resentment in that it would be directed against the communicative behavior of others, and in particular against failures to support one's preferred conception of oneself and against expressions of opinions that conflict with this conception. It will differ from the emotion of resentment in that it includes notions that appear to ensure the legitimacy of the anger that accompanies it. The liberal notions of equality and immunity from harm are notions about what is just and what is unjust; when we are angry against those who violate them we are angry on principle and not merely out of personal vanity.

Though this trait seems clearly to be one that some people do possess, it does not seem to have a name in our culture. For lack of a more appropriate term, I will call it "hypersensitivity." We should not be surprised if people who are hypersensitive in this sense sometimes fail to support liberal institutions that guarantee freedom of expression. Such institutions serve, as part of their normal functioning, to permit others to cause them pain, to protect the acts that do so as a right. Their fondness for these institutions is apt at least to be dampened by the realization that this is so. The problem, however, is deeper than that: hypersensitive people are not merely hurt, they are offended. Their resentment is based on certain notions about justice. These notions include, in some form or other, the principle that symbolic behavior that damages the self-esteem of others is wrong. This would mean that, to the extent that institutions protect people who do such painful things, they serve merely to protect wrongdoing.

People who hold this view might well decide that these institutions are simply unjust, reasoning that institutions that protect injustice are themselves unjust. Those who arrive at this conclusion can be expected to denounce such institutions and actively undermine them when they have a chance to do so.

The object of such active attempts at undermining would be to control the dialogue through which, in liberal society, people discover who they are, to determine what participants are to say or refrain from saying. To the extent that they succeed, the search for identity will take place monologically and not dialogically. The monologue will be delivered by whoever does the controlling. What hypersensitive activists seek, in part, is a self-monologue conducted through the controlled behavior of others.

Exercising this sort of control is just what people who are afflicted with hypersensitivity will be inclined to do, unless some other factor intervenes and prevents them. Those who are in the grips of this trait see what some people say as an injustice, as violating the liberal proscription against harm and the ideal of equality. Anyone who is faced with such injustice would obviously be justified in using more or less coercive means to put an end to it.⁽²⁹⁾

Of course, this would not necessarily mean making threats of physical sanctions such as beatings, incarceration, or the expropriation of the offender's wealth. At a minimum, however, it would mean the use of psychological manipulation or threats (perhaps veiled) of informal sanctions such as shunning and other non-violent expressions of hostility and moral blame. Such methods can be very effective. If they work their effects often enough, dialogue on relevant subjects will come to an end, replaced by monologues or by silence. Perhaps a more likely result, if there are many hypersensitive people trying to control each other, is a social world that simmers with snarled threats and hints of reprisals, a world in which dialogue is intermittent and dangerous.

Hypersensitivity need not, however, lead one to become an open foe of the legal and political institutions of liberalism. Some people who feel its promptings can also be strongly attached to the first of the two principles of minimal liberalism. They can, in other words, both believe in free expression and in their right to curb expression. Such a position is not a comfortable one, since it seems clearly contradictory, and people who find themselves in it will (assuming they are averse to self-contradiction) believe or hope that there is some way to interpret the two principles of liberalism that will strike some reasonable compromise, a balanced view of their relative importance.

Despite its appearance of moderation, this position is in its own way just as opposed to liberalism as the openly hostile one. As I have already said, an essential element of liberalism is the idea that the freedom of expression protected by the first principle is not to be sacrificed for the sake of the values protected by the second. Such a sacrifice is, of course, what the balanced view sought by those who hold this position amounts to.⁽³⁰⁾

If a society constituted by liberal ideas tends to produce people who suffer from the principled resentment I have been describing here, it plainly is planting the seeds of its own destruction, whether the stance that this resentment takes in relation to liberal institutions is straightforwardly hostile or whether it is ambivalent and reformist. The system would, in either case, be an unstable one.⁽³¹⁾

8. A Viable Liberalism

We have reason to think that the sort of resentment that I described in the preceding section is a good deal more than a theoretical possibility. There are a number of recent developments in current liberal culture that seem to indicate a heightened sensitivity to degrading expressive conduct by others and a spreading belief that forcibly silencing the expressions is a just cause.

Perhaps the most important single development of this sort is one of which many Americans seem to be unaware. Many liberal democracies currently have laws on the books that prohibit the expression of views or feelings that are offensive to members of various races or ethnicities, or to people with certain sexual preferences. In Austria one can be sent to prison for denying or "minimizing" the Nazi genocide. In Denmark one runs a serious risk of criminal prosecution by saying in strong language that homosexuality is morally wrong. Laws punishing racist speech are very widespread. The United States seems to be the only industrialized nation in which it is still entirely legal to express racist opinions.⁽³²⁾

The United States does not yet have such laws, largely because it has a written Bill of Rights guaranteeing freedom of speech and of the press, and because of the way its legal system presently interprets these guarantees. However, it has been the setting for a number of pervasive cultural developments that seem to arise from relevantly similar principles and motives. Probably the most obvious and widely discussed instance are the campus speech codes in force at many colleges and universities. These codes often give administrators wide discretion to use sketchily defined powers in disciplining students and professors for saying things about various protected groups (including women, racial minorities, and homosexuals) that are deemed to be derogatory. That is, they punish expressive conduct, and do so on account of the views that the conduct expresses.⁽³³⁾

Other recent developments that are fundamentally of the same sort can be found in recent changes in rules governing the work place, in both the public and private sectors. The rules regarding sexual harassment, which in a great many places of employment have been changed in recent years, are very much to the point. Traditionally, harassment, as a prohibited act, did one of two things: either it involved offers of work-related benefits in exchange for sexual favors, or it involved unwanted bodily contact. The former sort of act obviously violated the contractual rights of the employer and the latter in many cases just as obviously constituted a hurtful incursion against another's body. Both are the sort of act that may be prevented by more or less coercive means within a liberal order.

Recently, however, such means have been used to contend with "harassment" that consists simply in expressive conduct that creates a "hostile" work environment for some individual. In a great many cases this conduct need not express any hostility on the part of the individual that is doing it: it is enough if they express attitudes or feelings that are demeaning to certain people. In such cases the rules will be seen from a liberal point of view as doing precisely what a campus speech code does, as abridging the freedom of people to express their views.⁽³⁴⁾

A distinct but related development in the American workplace is the widespread introduction of compulsory "diversity training." This involves requiring workers to attend sessions that are explicitly meant to look into and improve their attitudes and feelings about their minority and female co-workers. In the course of doing so, the people who run these sessions have been reported as using methods that are psychologically intrusive and painful to the workers who are subjected to their training.⁽³⁵⁾

My point here is not that all or any of these policies are bad.⁽³⁶⁾ It is that each of them represents in some way an abridgment of the freedom of expression in order to protect the values embodied in the second principle of minimal liberalism, especially immunity from harm. Moreover, it is just the sort of harm that prompts hypersensitive people to attempt to silence the expressions of others. That is, each of these policies is a step away from liberalism, and it is just the sort of step that a hypersensitive person would take. I cannot argue here to what extent they actually are due to hypersensitivity--it would take me too far into the fields of journalism and sociology--but one thing that can be learned from what I have said already is that liberal institutions definitely are vulnerable to the sort of damage that hypersensitivity will cause if it is not somehow prevented from doing it.

This raises the question of how a liberal culture can protect itself against such damage. As long as it consists simply of what I have called minimal liberalism, there is no such protection at all. The question, then, is: What can be added to minimal liberalism to provide such protection?

One plausible solution to this problem would be to rely--perhaps more strongly that we have--on a certain institutional arrangement that is already part of our culture, and is probably part of all cultures. The most important part of the dialogue in which we find our identity is the stream of day-to-day, face-to-face encounters we have with other individuals. In it, we are able continuously to develop and learn about our individual natures. One thing that helps to prevent hypersensitivity from putting a halt to this process is the fact that it takes place within an ancient institutional framework that serves to protect individual identities from the sort of bruising to which hypersensitivity is a destructive reaction: namely, what is sometimes called etiquette.

This framework includes the ceremonial rules I discussed in an earlier chapter (VII.6) together with many other rules that seem to proscribe all the ways in which we might give offense to others, including being too blunt, taking more than one's share of something, getting out of turn, and so forth. These rules enable us to show (or give the impression) that we value the people we encounter, and care enough about their dignity to take their feelings into consideration. Emily Post said:

"Rule of etiquette the first--which hundreds of others merely paraphrase or explain or elaborate--is: Never do anything that is unpleasant to others."⁽³⁷⁾

Regarded as a solution to the problem of hypersensitivity, what etiquette does is to protect the dialogue by preventing us from provoking the hypersensitive into damaging it. As such, however, it has at least one serious drawback, which is suggested by Post's broad statement of the function of etiquette. As it stands, this statement is an obvious exaggeration, since it is clearly unrealistic to enjoin people to avoid ever displeasing anyone. But it does come close to stating what codes of etiquette are for, and this very fact indicates a hazard that they all bring with them. The fact that some action that causes pain to someone--especially if the pain consists in some sort of trauma to their self-esteem--very often makes it possible for the pained individual to object to the action as rude or, in more recent parlance, "insensitive." The danger is that, if such objections were raised frequently enough, our day to day encounters with others would be seriously damaged.

In particular, they would be damaged as means to revealing the truth. The truth is often unpleasant in one way or another, and the truth about who or what we are is especially likely to hurt. If every plausible charge of rudeness (or insensitivity) were raised and pressed vigorously, we would have very strong reasons to express ourselves--at least when our message has any bearing whatever on the conception our auditors have of themselves--with politely tortuous indirectness or not at all. In such an environment, the truth is not likely to flourish.⁽³⁸⁾

Though etiquette is no doubt part of the solution to the problem of hypersensitivity, it can also become part of the problem itself. Undeniably, this is something that does sometimes happen. Most of us have had the experience of knowing someone who is so quick to object to slights that talking to them feels like a hazardous undertaking. Usually, we simply avoid such people or restrict our talk with them to impersonal or superficial matters. Someone might wish to remind us at this point that such problems are the exception rather than the rule, that the dialogue, at least so far, is for the most part not sabotaged by accusations of rudeness and insensitivity. This is true, and it gives us some ground for hope. But then the question is: given the pains that can be caused by self-discovery through dialogue, and given that hypersensitivity seems to be on the rise, what reason can we have to expect that this will continue to be true?⁽³⁹⁾

There is a brief answer to this question that is, it seems to me, somewhat too brief. It is that, in a certain way, etiquette is a self-correcting system. Accusing people of breaches of etiquette is unpleasant to others and, for that reason, the rules prohibit doing so too readily and too directly. To do so is itself a breach of etiquette. The system prevents those who adhere to it from using it as a club with which they can beat dialogue to death.

The reason this answer is too brief is that it assumes that people who act according to the strictures of etiquette will do so consistently. If they do, this itself requires an explanation. How might we explain the fact that people generally adhere, as consistently as they do, to a system that assures them that various acts that cause them pain are wrong and also heavily constrains their efforts to defend themselves against such acts?⁽⁴⁰⁾

Good manners as we know them do not seem to be self-supporting. They seem to rest on something outside themselves. One support that seems very important is a certain cultural value, one that is not itself a matter of etiquette. What I have in mind is the value that is embodied in a saying that parents sometimes tell children to say to themselves when they feel an urge to hit someone for something they have said: "Sticks and stones may break my bones but names will never hurt me."

This idea has been ridiculed by advocates of restrictions on free speech.⁽⁴¹⁾ It is not difficult to see why: it seems on its face to be false. If, as I have argued here, liberal society consigns us to developing and discovering our identity in dialogue with others, then clearly we will be profoundly affected by the contributions others make to the dialogue, and this will certainly include the names they apply to us. I have already given reasons for thinking that these effects can be both good and bad. To deny this seems to be foolish, and worse: it seems to discourage people from protecting themselves against the bad effects.

But there are other aspects of the sticks and stones proverb that are not so easy to dismiss. First, it clearly implies that the effects of being hit with stones is radically different from the effects of being pelted with words. Second, because it is something that people are supposed to remember and repeat to themselves, it suggests that the things the would-be victim thinks are relevant to the effects of the latter sort of pelting. Words do not have effects by themselves, but depend on a contribution from their audience.

This suggests an alternative way of reading this quaint aphorism, not as an incredibly naive and inaccurate statement of fact, but as an attempt to change the facts, a thought that makes itself true. As such, though it is clearly optimistic, it can be regarded as resting on a principle that is no more obviously false than its most popular alternative. Even its optimism may be methodologically sound.

What might the relevant difference between a word and a stick be? A word, as applied to me, can only hurt me (in the relevant sense) insofar as it purports to represent what I am, and only insofar as its being applied to me leads to my accepting this representation as just. Nothing like this is true of the way that sticks hurt us.

One respect in which words seem to be different is that between the hurling of the word and the harm that it does me are a number of things that I can be said to do: I listen to it, understand it, see myself in light of it, and so forth. Of course, we can also say that being hit, being cut, and bleeding are also things that we do. They are also alike in that both sort of events are normally more or less involuntary. But one important difference, and one reason why the former sort of events are more action-like, is that they and only they can become voluntary. One cannot bring oneself to bleed by an act of will, but one can bring oneself to understand or to fail to understand.

If that is true, then the sticks and stones principle, as I have read it, is a plausible one. It rests on the idea that we are able to withhold some step that is essential to our being seriously harmed by what others say about us, and that such harm only comes when we believe what is said simply because it is said. What the principle says, in effect, is that we ought exercise this sort of control, that we should not absorb the thoughts that others have about us but rather should form and act on our own individual judgment about such things. What the naive-sounding proverb serves to do is to inculcate a certain trait of character, in which giving in to the inaccurate and adverse opinions that others have of oneself is regarded as wrong. Obviously, this trait is closely related to self-respect, and is probably a more important basis for it than conduct in which we stand up to tyrants and quarrel with our enemies. I will call it "independence."⁽⁴²⁾

Independence, in this sense, is one thing that can help etiquette to continue to exist. It can do so by, in a way, actually reducing the importance of etiquette. Those who possess this trait are to that extent less likely than others to see the value of and--this of course is the point--insist on the protections for delicate feelings that etiquette affords. Thus they will not mind as much as they otherwise would constraints on their power to seek that protection.

In a way, however, independence merely pushes our problem back another step. The problem is, how can a liberal culture avoid spawning a hypersensitivity that undermines its own institutions and end the dialogue that they normally foster? If independence is part of the answer, then it, too, requires some explanation. Its exercise is difficult and painful, and silencing others is a genuine alternative to it, one that offers obvious advantages. What can be added to our picture of liberal culture that might help independence to grow and do its work?

A partial answer to this question is already before us. Independence is based on the sticks and stones principle, which is itself a part of our moral and political culture. It is an idea that people accept, in part, because it is part of this culture and continues to be passed on from parent to child. If it exists together with the liberal notions of liberty, equality, and protection from harm, the result is a particular form of liberalism. In it, independence serves to protect liberty from being engulfed by equality.

I will call this sort of liberalism, including any culture that shares the same principles, "individualist liberalism." It is distinguished from the other sorts of liberalism by the fact that it possesses a certain source of immunity to the sort of instability I have been describing in this chapter. It is possible to select from among the other varieties of liberalism one which probably cannot have any such immunity at all. This is the variety that results if one takes the distinguishing ideas of individualist liberalism and replaces them with ones that are directly opposed to them.

The distinguishing ideas of individualist liberalism are the sticks and stones principle and the assumption on which it rests, that people can distance themselves in some way from what others say about them. Suppose that these ideas are replaced by a way of thinking that begins with the ideas that the identities of individuals are a product of the groups to which they belong, that this in turn is decisively influenced by the way other people classify them, and that the influence of the classification depends on the way those others characterize the groups involved. For want of a less awkward term, we can call this way of thinking "identity collectivism."

People who have absorbed a liberal political culture that includes this notion would face an excruciating dilemma to the extent that they find, as we all do sooner or later, that others characterize them in some unflattering ways. They would view the liberty of others as a source of serious damage to themselves, and respect for that liberty as acceptance of the inevitability of such damage. No one should be surprised if such people try to seize some sort of coercive control over what other people say and, perhaps, think. The society that such a culture would produce would of course not be a liberal one.

9. Prospects

It is not obvious whether our own political culture more closely resembles the individualist sort of liberalism or the non-individualist sort I have just described. It is certain that it has contained strongly individualist elements, at least until recently, but it is also well known that identity collectivism has been gaining influence, especially among intellectuals.⁽⁴³⁾

The coexistence, within the borders of the same national culture, of the individualist variety of liberalism together with a variety that I have argued is unstable raises a certain problem about the viability of the individualist variety itself. The advantage, as to stability, that individualist liberalism enjoys rests on independence and the sticks-and-stones principle. The problem is that, in the event that a culture with individualist elements also has aspects that are hostile to this principle, the principle and the trait of character that is based on it might not survive. If they die, the system seems doomed to develop into some newer, less stable sort of system.

So far, the institutional support for the survival of liberalism consists in the fact that this principle has the status of a social convention. Such conventions, as I have pointed out (VII.3), do seem to provide some degree of support for the widespread acceptance of principles of action, but it is doubtful that the sort of support they can provide is sufficient to solve the problem we are now confronting. They support principles by presenting us with a certain reason for accepting and following them, the reason being that others accept and follow them. The problem at present is that here we are dealing with principles which some people have keenly felt reasons not to accept or follow. Such reasons might well be stronger than the one the convention presents on the other side. Is there any way a culture could provide some additional reason for accepting the principle, in such a way that the convention itself has some further institutional support?

Actually, what the sticks-and-stones principle counsels us to do is something we already have additional reason to do. It tells us to protect ourselves from a certain sort of harm that others can cause us. We already have reason to avoid harm. But it also counsels us to use certain methods for avoiding this harm. It advises us to avoid the harm the opinions of others might do by refusing to accept those opinions. Someone has no reason at all to accept this advice if they think that this method cannot actually be used. If they think that they cannot avoid absorbing the adverse opinions others hold about them, then the sticks-and-stones tradition and the rugged independence that is based on it will make little sense to them. There seem to be a significant number of people who do think this way.⁽⁴⁴⁾ To the extent that this way of thinking spreads, the sticks-and-stones tradition will necessarily lose ground.

People whose minds function in this way believe they lack a certain skill: namely, the ability to understand the opinions of others without at once succumbing to them. The problem we are now considering would be alleviated if there were institutional arrangements that somehow teach people this skill or encourage them to acquire and use it. To the extent that people do learn this lesson, an otherwise insuperable obstacle to accepting the sticks-and-stones principle has been removed.

One institutional arrangement that does perform this task can be found among the systems that I have called societies of contract. These are societies in which the differential obligations of individuals are foreseeable consequences of their own voluntary acts. The institutional arrangement I have in mind is a certain variety of society of contract, one that has historically been closely related to liberal democracy: namely the competitive market. This is a contractual system in which people who might become obligated to us have alternatives to doing so. Those who are indebted to me were not stuck with me, but could have dealt with someone else instead.

In a way, it might seem paradoxical to claim that such systems can help people to gain independence from the opinions of others. In any sort of contractual system, the opinions of others, at least some others, are very important. To the extent that we live in such a system, we can only deliberately acquire claims against other people by offering them something they want. In order to do this, we must be able to understand--and be interested in understanding--the wants of others, and to do this we must to some extent be able to understand and be interested in their thoughts as well.

This is true especially of the competitive market. Under competition, the other person, as I say, has alternatives to dealing with us. Such a system penalizes us, not merely for failing to divine the wants and thoughts of others, but for failing to do it fast enough, for delaying until someone else discovers the unsatisfied desire and satisfies it. Such a system exerts a continual pressure on us to attend to the possible desires and thoughts of potential trading partners. For that reason it might discourage the emergence of character types that essentially involve contemplating one's own thoughts and completely withdrawing from the points of view taken by other human beings: including proud Byronic heroes, Walden Pond hermits, and vision-questing shamans. It is more likely to produce test-marketing consultants, pollsters, and organizational yes-men.⁽⁴⁵⁾

This is the truth--or one of them--behind the doux-commerce thesis I discussed earlier (VI.2). Important as this truth undoubtedly is, however, it should not blind us to another one, one which is even more important in the present context. This is the fact that, while competitive markets penalize us for being oblivious of the facts about other people's thoughts and wants, they also penalize us, and with the same sort of rigor, for taking them to be more than facts. An entrepreneur who always accepts, as if they were the word of God, the beliefs and desires of the consumers regarding his or her product would be at a severe disadvantage in facing a more independent competitor. For instance, if the product I offer is a very popular one and I simply adopt the consumers' attitude toward it--which in this case is strong approval--as my own, I would be unable to detect the shortcomings in that product and would be at the mercy of a competitor who could detect them and offer a product the consumer would prefer. Investors during wartime who are swept up by a general mood of optimism into thinking that their country will win are apt to lose money if this thought proves to be a self-gratifying illusion, while those who are immune to such collective fervor stand to profit.

These observations can easily be generalized. Competitive markets, and contractual systems in general, reward us for satisfying the desires of others, but only for satisfying desires they in fact do have at the time we attempt to satisfy them, and only for doing things that in fact do satisfy those desires. Any susceptibility to factors that cloud one's awareness of these facts is a liability, even if these factors are the thoughts and wants of the people we seek to please. If a nation's wartime optimism proves to be tragically misguided, and its rulers must find ways to pay reparations extorted by a ruthless victor, they will be eager to borrow money from whoever can help them pay. People who foresaw this need are, other things being equal, more likely to be in this position than those who succumbed to the popular belief that this day would never come.

Things look quite different if we imagine moving from such a society to one that lacks the one feature that is essential to contractual systems. Such systems are constituted by the fact that, to some extent or other, obligations in such a system are voluntary. If I have an obligation to you which, in contrast to this, is completely independent of my ability to choose, then you are not and never were under any pressure to entice me into accepting or keeping this obligation. This removes from your shoulders the peculiarly strong reason you would ordinarily have, in a contractual relationship, to take my desires and beliefs seriously.

I, on the other hand, may have such a reason, a stronger one than exists in a contractual relationship, and fundamentally different in kind. Suppose that the obligation is to perform some set task in a manner to be specified by you. Then I have a good reason to be alert to information about how you want this job done. Depending on the exact nature of our relationship, I can probably avoid problems if I can somehow collect information about your preferences. But I so far have no reason at all to discover a method of performance that, in your terms, is better than the one you want, one that would make you even happier. That is, I have a reason to take your wants and beliefs seriously without the reason that is present in contractual relationships, and heightened in ones that are disciplined by competition, to approach those wants and beliefs at some critical distance from them.

Both these features intensify as the relationship becomes less and less voluntary. At the limit, in a completely involuntary relationship--slavery --there is probably little incentive to operate as an independent mind at all. If you are literally my master, your whims are the most important single factor in my life (or nearly so). Perhaps the most obvious reason for this is the fact that I can avoid savage punishments by divining your whims in advance. But I gain little or no benefit from thinking about whether they are right or whether your beliefs are true.⁽⁴⁶⁾

An uncritical sensitivity to the inner experiences of others, at least of one other, is a valuable trait for a slave to have. This is not true of the inhabitants of a contractual system. It encourages sensitivity of a sort, but it rewards those who combine it with a certain critical distance. I should probably emphasize, since there is sure to be misunderstanding otherwise, that I am not saying that this feature of contractual systems is uniformly beneficial as far as its effects on human character are concerned. An organizational yes-man who is taken by his boss's delusions is at a disadvantage when competing with one who is merely (but convincingly) pretending to be, but he may be a less despicable person. As far as anything I have said in this section is concerned, competitive markets might have bad effects on character, and might even, on balance, do more harm than good. The trait I am talking about here is a neither a virtue nor a vice but a skill, and can be put to either virtuous or vicious uses.

In the present context, what matters is that this skill removes an otherwise insuperable obstruction in the way of acquiring the cognitive independence that I have argued is crucial to the survival of liberalism. To someone who can understand the opinions of others without succumbing to them by mere contagion, the sticks-and-stones principle will seem as sensible as it did to many of us when we were children, but to someone without it, it will seem as cruelly nonsensical as its critics are now saying it is. To the extent that someone does follow its advice, however, they will have no reason to defend their self-esteem by silencing others.

Postscript: Toward a Society of Status?

I would like to close by saying a few more words about the non-individualist liberalism I described in section 8, in which the sticks-and-stones principle is replaced by the collectivism of identity. I will call the resulting sort of culture "collectivist liberalism." It deserves our attention if only because it may represent an important part of the future of our own political culture. If what I have said in this chapter is close to the truth, we have reason to be very worried about this.

As I have described it, it represents a return to the society of status. It seems to represent status renascent and reformed along egalitarian lines. Traditional societies of status were hierarchies, in which a small and fixed collection of groups were clearly ranked, with some groups higher and others lower. Because of liberalism's commitment to equality, collectivist liberalism is permanently wedded to the idea that all groups are on the same level. In addition, in contemporary liberal society there are an indefinite number of relevant groups. Where the traditional systems were hierarchical and closed, the newer one is egalitarian and open-ended. Yet it shares with them one of their essential characteristics: in each case, the individuals who inhabit them acquire their identity monologically from some human agency or other.

Collectivist liberalism, as we see it in our culture, also possesses the other essential feature of societies of status: membership in the relevant sort of group brings unchosen obligations. Being liberals of a sort, adherents to this way of thinking are apt to think of the regime in favor as granting rights rather than imposing obligations. Various groups are to have an enforceable right to be spoken of in ways that are not demeaning to them. But of course in each case this means that the members of another group--that is, all non-members of the group that possesses the right--have the obligation to speak as the right prescribes.

This combination of egalitarianism, open-endedness, and status, could represent a particularly explosive combination. It is very doubtful that it could possibly achieve the static sort of order that pre-liberal societies sometimes came close to attaining. Unless it abandons its commitment to equality, it cannot allow one group to impose its view of the social world on all the others, as societies of status have done in the past. Given that, there seems to be no way that it could arrive at some official truth about who everyone is and what is their worth.

The result would seem to be an analogue, within the realm of symbolic behavior, of the politico-economic system I described at the end of the last chapter (X.5): an endless squabble between increasingly numerous groups, each of whom presses for their right to speak while seeking to control what others are saying about them. Advocates of collectivist liberalism need to explain what would prevent such a regime from disintegrating into what one critic has called "a simmering sort of mutual dislike on the level of everyday discourse,"⁽⁴⁷⁾ and perhaps into something much worse on the level of political activity.

When people think about these issues they often arbitrarily limit the groups who will press claims to some finite list. They often ignore, for instance, creationists who try to compel schools to eliminate textbooks that describe evolution as scientific fact, on the grounds that such ways of speaking demean them by branding their views as unscientific, or Moslem fundamentalists who seek to suppress books that wound them by satirizing the Prophet Mohammed.⁽⁴⁸⁾

Ignoring them, and the indefinite number of others who have equally sincere and well-founded complaints, amounts to assuming that there could be some equivalent of the constitutional provision I discussed earlier (X.5), in which the groups that are going to count is strictly limited. But this seems clearly inconsistent with identity collectivism itself, which indicates that we are all radically vulnerable to the symbolic behavior of others and have a profound and ineradicable interest in controlling that behavior.

1. In case this claim seems too apocalyptic to be very plausible, the following more reserved one will do as well, for the present purposes: "Now, outside the Islamic world, there appears to be a general consensus that accepts liberal democracy's claims to be the most rational form of government." Francis Fukuyama, The End of History and the Last Man (New York: Avon, 1992), p. 211. Fukuyama's book, however, is in part an argument that the more apocalyptic claim is actually true.

2. José Ortega y Gasset, The Revolt of the Masses (London: George Allen and Unwin, 1961), p. 59. The quotation at the head of this chapter is from p. 58 of the same edition. For anybody interested in the subject I am treating here, most of Ortega's book--chapters 1 through 13--will repay a close and thoughtful reading.

3. John Rawls can be interpreted as adhering to this requirement when he gives his first principle of justice (which specifies the nature of his commitment to liberty) "lexical" priority over the second one (which gives the content of his egalitarianism). See his A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), p. 61.

4. I emphasize that the characterization of liberalism that I have given covers very widely disparate views, probably including both the extremely egalitarian liberalism defended by Ronald Dworkin in his "Liberalism," in Public and Private Morality, ed. Stuart Hampshire (Cambridge, England: Cambridge University Press, 1978), and the laissez faire, market- oriented liberalism defended by Ludwig von Mises in his Liberalism (San Francisco and Irvington, N.Y.: Foundation for Economic Education and Cobden Press, 1985). Some people (though not the same ones) would argue that one or the other of these authors is not a genuine liberal, but both are probably covered by the description I have given. Equality is a much less important idea for von Mises than it is for most liberals, but he does insist on a right to equal protection and consequently does not fall outside the boundaries of liberalism as I have described it. See Liberalism, pp. 27-30. Dworkin is more problematic, since he explicitly argues that equality is a more fundamental value than liberty, and that we actually have no right to liberty at all. See, with regard to this last assertion, "What Rights Do We Have?" in Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1977), pp. 266-78. However, Dworkin does think we have important rights to various liberties (in the plural, not the singular: a distinction that he makes much of), which include extensive rights to free expression, and he apparently would not sacrifice such expression rights for the sake of equality. Thus, he seems to be, at any rate, what I would call a minimal liberal.

5. For an interesting discussion of related issues, see Stephen Macedo, Liberal Virtues: Citizenship, Virtue, and Community in Liberal Constitutionalism (New York: Oxford University Press, 1990).

6. Ortega, The Revolt of the Masses, p. 42.

7. Ibid., p. 43.

8. Henry Sumner Maine, Ancient Law (New York: Henry Holt and Company, 1888), p. 165.

9. Here I am using caste in the sense employed by Ludwig von Mises. See his "The Clash of Group Interests," in Money, Method, and the Market Process: Essays by Ludwig von Mises, ed. Richard M. Ebeling (Norwell, Mass.: Kluwer Academic Publishers, 1990), p. 204.

10. Maine, Ancient Law, p. 162.

11. "The parent . . . has over his children the jus vitae necisque, the power of life and death . . .; he can modify their personal condition at pleasure; he can give a wife to his son; he can give his daughter in marriage; he can divorce his children of either sex; he can transfer them to another family by adoption; he can sell them." Ancient Law, p. 133.

12. It is also possible, though I have no reason to stress this point, to interpret the gift-exchange systems I discussed in chapter VIII as societies of status. As I pointed out there (VIII.4), such systems inevitably generate significant obligations (obligations to make return gifts) by a process that is substantially outside the individual's control. As anthropologists have noted, in such systems gifts are reserved for exchanges between members of one's own group. To the extent that people are perceived as falling outside one's group, commercial exchange is used, and, in some cases, people who are completely alien are treated without moral constraints at all. See Marshall Sahlins, Stone Age Economics, (New York: Aldine de Gruyter, 1972), p. 162. To be a member of the group is, in part, to participate in a tradition-defined process that, by its nature, brings a flood of obligations with it. Since membership in such groups is typically by birth or adoption--the former always and the latter usually an involuntary one on the recipient's part--such societies are clearly ones in which an important body of obligations are the unchosen consequences of involuntary group membership. Gift-exchange systems differ from many societies of status in one important respect: in many societies of status, the obligations are owed to someone who is either above or below oneself in some hierarchical order. As I use the term, however, societies of status does not refer to systems that are necessarily hierarchical. The same is true of Maine's usage.

13. A certain hierarchical attitude is often openly displayed in the writings of the classic apologists for such systems. One clear example is Plato's Republic. Another is John of Salisbury's defense of the medieval caste system in the Policraticus. The latter author compares the commonwealth to a human individual, with some parts performing the functions of the feet and others doing the work of the soul. The commonwealth is healthy when "the higher members shield the lower, and the lower respond faithfully and fully in like measure to the just demands of their superiors, so that each and all are as it were members one of another by a sort of reciprocity. . ." The Portable Medieval Reader, ed. J. B. Ross and M. M. McLaughlin (New York: Viking Press, 1949), pp. 47-48. Such reciprocity is, as John says, between superiors and inferiors, not between equals.

14. Alasdair MacIntyre, After Virtue: A Study in Moral Theory (Notre Dame: University of Notre Dame Press, 1984), p. 220.

15. See footnote 8, above. Friedrich Hayek gently criticizes Maine's formula as follows: "The true contrast to a reign of status is the reign of general and equal laws, of the rules which are the same for all, or, we might say, of the rule of leges in the original meaning of the Latin word for laws--leges that is, as opposed to the privileges." The Constitution of Liberty (Chicago: The University of Chicago Press, 1960), p. 154. While Hayek's comment is quite true, I think it misses the point of Maine's distinction, which is to explain how people in a given society come to have differentialobligations. General rules equally applied (as long as they are described no further) cannot explain that.

16. I should emphasize that a given society is a society of status or of contract to the extent that it answers to one definition or the other. It is no part of my thesis that all systems answer to one or the other, or that a pure instance of either sort of society has ever existed.

17. Admittedly, my (and Maine's) use of the word "contract" here is somewhat inexact. Yet, though many of the moral and legal practices toward which I am gesturing here are not contracts, still they resemble contracts in three important respects. Consider the following example. In our culture, the obligations one has toward one's children are not based on one's status as their biological ascendant. This is shown by the fact that we see exactly the same obligations as attaching us to our adopted children, while on the other hand we see none of these obligations as attaching us to children we might have given out for adoption. Yet such obligations could not be based on contract either. The only contract that could provide such a basis would have to involve the children as parties, and many of them are not competent to enter into contractual relations. Apparently, the obligations are based on an act on the part of parents or guardians (perhaps having the children in their homes and not taking steps to find other homes for them) that differs from entering a contract in that it can be unilateral. But it otherwise resembles a contract in that (1) it is voluntary, (2) it has an obligation as its foreseeable result, and (3) any adult in principle can do it.

18. I should point out that, though I use a phrase from MacIntyre here I am radically disagreeing with what he uses it to say. Consider this important passage: "We all approach our circumstances as bearers of a particular social identity. I am someone's son or daughter, someone else's cousin or uncle; I am a citizen of this or that city, a member of this or that guild or profession; I belong to this clan, that tribe, this nation. ... As such, I inherit from the past of my family, my city, my tribe, my nation, a variety of debts, inheritances, rightful expectations and obligations. These constitute the given of my life, my moral starting point. This is in part what gives my life its own moral particularity." After Virtue, p. 220. What he is expressing here is clearly a preliberal way of creating moral particularity. Both in this passage and throughout this enormously influential book, MacIntyre ignores the liberal way of achieving the same result. This omission is essential to the argument of After Virtue. If I understand it rightly, the relevant part of that argument goes as follows. MacIntyre eloquently presents, in broad outline, various features of preliberal moral systems which established differential obligations among the people who lived within them. He then notes that this description does not apply very well to liberal democracies. He concludes that the moralities on which such societies are based must be ones in which the only obligations that exist are those few that can be attributed to everyone. Within the wide limits established by these bloodlessly universal obligations, little of moral importance happens: mainly, life is a series of brutal conflicts between nakedly self-interested individuals or factions. See After Virtue, Ch. III and pp. 250-51. This argument, if it really is his argument, simply assumes, without any proof, that preliberal ways of achieving differential obligations are the only ones there are.

19. Maine, Ancient Law, p. 165.

20. See John Rawls, Political Liberalism (New York: Columbia University Press, 1993), pp. 29-35.

21. Gilbert Ryle, The Concept of Mind (New York: Barnes and Noble, 1949), p. 15 and passim.

22. Traits of character present an additional difficulty. When we want to know whether we have a given trait of character, we are mainly interested in whether the trait we have is an instance of virtue or of vice. This requires us to find out whether the principle it is based on is true or false. To have any well-grounded ideas about this at all, we must to some extent be able to solve problems about the nature and value of the proper ends of human action and about the legitimate constraints on the means we should avail ourselves of as we pursue our ends. I can only have good reason to deny that my actions are cowardly if I have good reason to deny that they reveal that I value safety too highly. Such reasons, in turn, can only be mine if I have some notion, more or less well founded, of what would constitute overvaluing safety. Such issues of course are complex and endlessly contestable.

23. Again, essentially the same things can be said of the problems arising from the notions about good and bad, right and wrong, that undergird ascriptions of character. The fact that, here too, our self-understanding depends on others is strongly suggested by the history of ethics. All these notions are the work of a long line of people working with and against each other. One reason why this is so is that it is only in the context of discussion and debate that human beings can gain an insight into the troubling question of how adequate are their notions about what is good or right. It is possible to have a rational basis for such ideas, but this is secured mainly by taking positions on relevant issues in the context of reactions of other people who have positions of their own. Under the pressure of comments both friendly and hostile (especially the latter), our ideas have been constantly checked, altered, and--we hope--improved.

24. No doubt there are many reasons for this, some of them profound and poorly understood, but one grossly obvious one is simple self-interest. Whatever is good for me is good for just the sort of person I am, and will often not be good for other sorts of people. This means that I am not likely to find out what it is that is good for me without finding out what sort of person it is that I am; and if I do not find out what is good for me, I am not likely to get it.

25. Here I am borrowing my terminology from Charles Taylor, who in turn borrows it from M. M. Bakhtin. See Taylor's "Multiculturalism and 'The Politics of Recognition,'" Multiculturalism and the Politics of Recognition: An Essay by Charles Taylor with Commentary by Amy Gutmann, Steven C. Rockefeller, Michael Walzer, and Susan Wolf, ed. Amy Gutmann (Princeton: Princeton University Press, 1991), pp. 25-73. Taylor uses "monological" and "dialogical" to say things that are very different from and to some extent opposed to the things I am using them to say, and it is quite possible that he does not mean the same things by these words as I do. For that reason, I hesitate to use the same terminology he uses, since I might thereby lead readers who are familiar with Taylor's essay to misunderstand what I am saying, or to think that I have misunderstood what he is saying. However, these words so perfectly express what I have in mind--better, indeed than they fit what he is saying--that I can see no better alternative. I will comment on the differences between Taylor's message and mine later on. That, I hope, will prevent some confusions.

26. One difference between Charles Taylor's use of "monological" and "dialogical" on the one hand and my own on the other (see footnote 25, above) lies in the fact that he restricts the use of monological to investigations of identity that are asocial, in which it is "something each person accomplishes on his or her own." ("Multiculturalism and 'The Politics of Recognition'," p. 32.) Other differences are to be found in his suggestions that throughout history the search for identity has always been almost completely dialogical (see p. 34) and his tendency to speak of the monological as an "ideal" which is inadequate because it "seriously underestimates" the inevitable "place of the dialogical in human life" (p. 33). All of this suggests that for him "monological" and "dialogical" mean simply "asocial" and "social," respectively. That, of course, is not what they mean when I use them. But these terminological differences, if that is what they are, seem to be connected to deeper ones. A central claim of his essay is that liberalism, when it arrived, stepped into a world that was thoroughly dialogical (in his sense) and then proceeded to interfere with the smooth working of this process (pp. 34-5). Another is that the only solution to the problems caused by this interference is to change liberal society so that it is more like preliberal society (as I have characterized it) (see pp. 51-61). I think that the first of these two claims has more than a little truth in it. However, given that the word "dialogue" and its cognates inevitably suggest a dynamic process involving an open exchange of conflicting views, it is also profoundly misleading to state this idea as he does, by describing preliberal society as dialogical. I shall presently argue that the second claim--that only the preliberal approach offers a hope of a solution--is not true.

27. I should emphasize something that I have already suggested but perhaps have not yet made explicit enough. The enterprise of seeking self-understanding through dialogue is not unique to liberal society. It must have existed throughout human history and at times was widespread even before the birth of liberalism. Henri Pirenne argued eloquently that, by the late twelfth century, relations among the merchants and artisans of the time took place in conditions that were remarkably free of the restrictions that characterize societies of status. This, however, was apparently only true of relations among these individuals. Relations with individuals outside the group were a different matter. Pirenne argued that these people regarded their freedom as a caste privilege and jealously guarded it as such against the rural peasants whom they held in subjection and regarded as their inferiors. See his Medieval Cities: Their Origins and the Revival of Trade (Princeton: Princeton University Press, 1952), Chs. VII and IIX. See also his Economic and Social History of Medieval Europe (New York: Harcourt Brace, 1937), Ch. II. The point, however, is that most people, including most of these lucky individuals, were given some portion of their identity by the moral and legal structure of their societies, and they have generally felt that this was an important portion of it. In addition, those who were not so fortunate had little opportunity or inclination to inquire what, in addition to this portion, constituted who they were. They knew who they were and that, at least, was the end of that. In liberal societies, this source of putative insight into one's own identity is for the most part not present at all, for anyone.

28. Later, I will briefly discuss some authors who claim that such acts really do in some way subordinate people. Here I am only saying that this is how such actions are often perceived. See footnote 43, below.

29. "'What is a sensitive person?' said the Cracker to the Roman Candle. 'A person who, because he has corns himself, always treads on other people's toes.'" Oscar Wilde, "The Remarkable Rocket," in The Fairy Stories of Oscar Wilde (New York: Peter Bedrick Books, 1986), p. 68.

30. Traditionally, countries that consider themselves liberal democracies have curbed speech of various different kinds, but their legal theories have taken sometimes heroic measures to make the limits imposed entirely compatible with freedom of discussion. Sometimes the speech that is curbed is interpreted as part of an action in which something illegal is not merely said but done, as in the cases of conspiring, inciting, or making fraudulent promises. Other communicative acts are made actionable because they have demonstrably bad effects on the victim's reputation or on some other thing that can be thought of as the victim's property, as in the law of libel and slander. Still others are interpreted as the appropriation of facts that do not rightfully belong to the speaker, invasions of privacy being one of several possible examples. Finally, as in the case of obscenity, they are sometimes prohibited on the basis of principles (involving, for instance, community standards or appeals to prurient interest) that are meant to single out for legal prohibition the communicative act itself or the way it is expressed, as contrasted with any thoughts or feelings that might have been expressed in them. Of course, it is debatable whether these theories are all fully plausible or genuinely consistent with liberalism. All I wish to insist on here is that they represent a great effort to avoid curbing communicative action simply because it expresses an opinion or feeling that is deemed to be bad. If we attempt to amend liberalism to find a more reasonable balance between freedom of expression on the one hand and equality and harm-avoidance on the other, we do precisely that. What we are striking at is not the communicative act itself, nor its physical consequences, nor the misappropriation of information. The enemy we are trying to neutralize, the source of the painful and degrading threats to one's identity, is the human mind that lies behind the utterance, with its uncontrollable power to form opinions, have emotions, and weigh the values of things. This is an undertaking that liberal societies have tried very hard the avoid.

31. It would also seem to be flawed in another way, since hypersensitivity appears to be a vice. It is related to but distinct from what I called "quarrelsomeness" and "petty vanity" in II.7. Those traits involved taking the concerns of self-respect too far. Hypersensitivity need not involve an overvaluation of such things. It is defined rather in terms of what the individual is willing to do to other people about such things. While the other traits are faults in the agent's view of the good, hypersensitivity raises grave doubts about the agent's justice. Providing an argument that it does involve such a breach would require a lengthy argument and one that I am not prepared at this time to give. There is another reason for thinking this trait is a vice, one that is probably more obvious and can no doubt be stated more briefly: it is the same reason that, as I have claimed earlier, envy is a vice (X. 2 and 3). The hypersensitive person does not merely believe that it is permissible to prevent, by means more or less coercive, the damaging communicative act. Such a person positively wants to do so. Where does this desire to indulge in coercion come from? It springs from a wish to maintain one's preferred conception of one's own identity. The point of the coercion would be to protect one's self-esteem from erosion caused by the expressive acts of others. But why would the expressive acts of others constitute a threat to one's self-esteem? The most obvious reason is that the opinions that others have about us as are in such cases experienced as evidence about who we are. But the coercive acts involved here only aim at stopping people from communicating their opinions, and obviously do not destroy the opinions themselves. What could explain why people would want to protect their self-esteem by such means? It is only because of the opinions, which are untouched by the act of suppressing their expression, that these expressions are at all relevant to one's own conception of oneself. Since the goal that is sought in it is the preservation of one's own conception of one's identity, the point of the coercion would seem to be to hide the evidence from oneself.

32. For more examples of national speech codes, see Jonathan Rauch, Kindly Inquisitors: The New Attacks on Free Thought (Chicago: University of Chicago Press, 1993), Ch. I.

33. For examples of such rules, see Dinesh D'Souza, Illiberal Education: The Politics of Race and Sex on Campus (New York: The Free Press, 1991). A more recent treatment of the subject, with much material on other, closely related cultural developments is Richard Bernstein, Dictatorship of Virtue: Multiculturalism and the Battle for America's Future (New York: Alfred A. Knopf, 1994). Both authors are critical of these phenomena, but there seem to be no book-length journalistic descriptions of the subject of which this is not true.

34. The speech code now in effect at my own university is a harassment rule. It defines harassment in such a way as to include the making of comments which are demeaning to members of one of a list of protected groups.

35. See Heather MacDonald, "The Diversity Industry," The New Republic, vol. 209 no. 1, July 5 1993, pp. 22-25. According to MacDonald, p. 23, one of the goals of the "diversity consultants," as they are called, is to detect and remove undesirable ethical beliefs, such as the idea that "fairness equals treating people the same."

36. I do think that the argument of this chapter could be used as part of a larger argument against all policies of this sort, but that would require lengthy discussions that would take us far away from the subject of this book.

37. Emily Post, Etiquette (New York: Funk and Wagnalls, 1937), p. 43.

38. Partly on the basis of his own observations, Jonathan Rauch argues at some length that, at least as regards its achievements in science and scholarship, Japan is a case in point here. In Japanese culture, criticizing other people's ideas is considered rude. As a result, ideas are seldom debated openly. Instead, they are traded on "a kind of gray market" in which "people criticize privately." The result, Rauch says, is a culture that is very good at following the lead of others but surprisingly poor (considering its large, well-educated, and extremely hard-working population) at producing new ideas. Kindly Inquisitors, pp. 126-7. See also his The Outnation: A Search for the Soul of Japan (Cambridge: Harvard Business School Press, 1992).

39. A closely related problem would be to explain why a polite aversion to offending others does not, by itself and aside from charges hurled by putative victims, stifle vigorous debate. Whatever it is that prevents this from happening, Jonathan Rauch's view seems to be that it is absent from Japanese culture: the Japanese are too polite to accuse one another of being offensive, but their horror of giving offense often has the same suffocating effect on discussion. See footnote 38, above.

40. The thought of people rebelling against such constraints is far more than a theoretical possibility. This, in fact, is one of the differences between traditional etiquette and the new standards of "political correctness," or what might be called the new ethic of sensitivity. Racist, sexist, or homophobic insults are obvious breaches of etiquette, and people who advocate new standards of linguistic sensitivity to combat such insults are sometimes called the "new Victorians," suggesting that what they are trying to do is simply to bring back etiquette as we once knew it. Yet there is at least one very large difference between etiquette on the one hand and these new standards on the other. Intemperate accusations of rudeness are always violations of etiquette, but accusations of racism, sexism, or homophobia, no matter how reckless they might be, are never "politically incorrect." None of the recent campus speech codes contained any explicit instructions to the effect that such epithets should not be flung about carelessly. What we are witnessing is the formation of codes of conduct that resemble etiquette except that their enforcement is unconstrained in precisely the way I am considering here. For a discussion of other ways in which these new codes differ from traditional etiquette and morality, see Gertrude Himmelfarb, The De-Moralization of Society: From Victorian Virtues to Modern Values (New York: Knopf, 1995), "Postscript: The 'New Victorians' and the Old."

41. "A child may chant 'sticks and stones may break my bones, but names will never hurt.' Names do hurt, though. That is why she chants. ... Words can break bones. 'Shoot her' might break a few...." Rae Langdon, "Speech Acts and Unspeakable Acts," Philosophy and Public Affairs, vol. 22, no. 4, Fall 1993, pp. 302. Interpreted in this way, the idea becomes a mere piece of wishful thinking, so deeply confused that it is contradicted by the very reason for wishing for it: it is precisely because words do hurt that the child tells herself that they can't. For similarly dismissive comments, see Stanley Fish, There's No Such Thing as Free Speech and It's a Good Thing Too (New York: Oxford University Press, 1994), p. 110.

42. After I had chosen this word for this trait I noticed that Ayn Rand used the same word for what is almost the same trait, the only difference being that the trait she is naming seems to govern one's thinking about everything, and not just oneself. "The virtue of Independence," she says, is "one's acceptance of the responsibility of forming one's own judgments and of living by the work of one's own mind." The Virtue of Selfishness: A New Concept of Egoism (New York: New American Library, 1964), p. 26.

43. Charles Taylor seems to adopt some form of this idea in his "Multiculturalism and 'The Politics of Recognition'," cited in footnote 25, above. It is also, in one way or another, fundamental to much of the recent literature that calls for new restrictions on "hate speech." Much of it insists that, in the words of Stanley Fish, epithets directed at a group to which one belongs cause "lacerating harms" that are "grievous and deeply wounding," and are not a matter of hurt feelings. This idea, as used by Fish and other authors, seems to involve some very direct connection between speech and the harms it does, as if people are affected simply by being expressly classified in unfavorably characterized groups. Stanley Fish, There's No Such Thing as Free Speech and It's a Good Thing Too, p. 109. See also the first three contributions to Words that Wound, ed. by Matsuda, Lawrence, Delgado, and Crenshaw (Boulder, Colorado: Westview Press, 1993). In recent years some philosophers have advanced a claim that can be interpreted as an interesting variation on this idea. The claim is that, in characterizing a group adversely, one is not merely saying something about them, one is performing an act, what J. L. Austin called an "illocutionary act," and that there is reason for legally forbidding such acts. See Andrew Altman, "Liberalism and Campus Hate Speech: A Philosophical Examination," Ethics, 103, January 1993, pp. 302-17, and Rae Langdon, "Speech Acts and Unspeakable Acts," cited in footnote 41. One of the illocutionary acts one is performing, according to these authors, is the act of "subordinating" someone. The argument for forbidding them seems to involve the notion that this act is done to someone--the group's members--and is thus not a victimless offense. Since it would make little sense to legally forbid an act on such grounds if it has no effect on the person to whom it is done, this notion seems to involve the additional idea that the act itself must somehow alter the people toward whom it is done: if I subordinate you then you have (in some non-trivial sense) been subordinated. The idea that one can be changed by such purely symbolic means is what identity collectivism amounts to. I personally doubt that this version of the idea of group responsibility is philosophically defensible, but that is beside my point here. The point is that the idea itself seems to be gaining adherents, and that we should expect it to have certain institutional and characterological effects unless other factors intervene.

44. A number of authors have suggested that--contrary to what I am assuming here--that this way of thinking, or something relevantly like it, is actually true, and that it is true of everyone. Mari Matsuda offers the following anecdotal argument regarding racist opinions: "At some level, no matter how much both victims and well-meaning dominant group members resist it, racial inferiority is planted in our minds as an idea that may hold some truth. ... In conducting research for this chapter, I read an unhealthy number of racist statements. A few weeks after reading about a 'dot busters' campaign against immigrants from India, I passed by an Indian woman on my campus. Instead of thinking, `What a beautiful sari,' the first thought that came into my mind was 'dot busters.' Only after setting aside the hate message could I move on to my own thoughts. The propaganda I read had taken me one step back from casually treating a fellow brown-skinned human being as that, rather than as someone distanced from myself." Words that Wound, p. 26. Matsuda's anecdote suggests an interesting idea, one that may well be true, but it does not support the conclusion she draws from it. The possible truth is the idea that, in understanding the statements of others, we often see the world momentarily as they do--this in fact may be an essential part of understanding what another person is saying. It may also be the case that--as this story suggests--such statements sometimes irresistibly compel one to see things from the other's point of view. But this is an as-if sort of seeing, similar to the "interpretations" that I argued in Chapter V are an essential part of the emotions, and not the sort of vision of the world that is involved when one actually believes something. It may be that, when I try to understand a racist idea, I experience (in an as-if sort of way) a psychological distance from members of the race that the idea is about, something that mimics the distance experienced by the racist. But this is not the vicious sort of alienation that I would presumably undergo if I really thought that these people are, or really might be, my inferiors. That is, what might be true of everyone is not the idea that the statements of others have an irresistible control over our beliefs, but something quite different.

45. As Georg Simmel put it, competition tends to bring people to an awareness of "the innermost wishes of the other, even before he himself becomes aware of them." This fact is somewhat obscured, he points out, by the tendency to think of competition as a sort of fight. If it is a fight, it is a "simultaneous fight against a fellowman for a third one." Conflict: The Web of Group Affiliation, trans. and ed. by Kurt Wolff and Reinhard Bendix (Glencoe: The Free Press, 1955), p. 61-4. Competitors do tend to trample one another, but the ultimate reason for this is that they are all rushing to offer their customers whatever they want.

46. In such a situation, there would probably be considerable advantage to me in absorbing your beliefs and desires as my own as much as I can. I might be able to anticipate the behavior of an agent who has the same beliefs and desires that I have myself than that of an agent with a mind really different from my own. I would have no need to imaginatively put myself in your shoes, I would already be wearing your shoes.

47. Richard Bernstein, Dictatorship of Virtue, p. 9.

48. See Jonathan Rauch, Kindly Inquisitors, Ch. I.