A Theory of Ideology
Yale University Press
New Haven & London
Copyright © 1998 by Yale University. All rights reserved.
With the permission of Yale University Press, the author has made an online version of this book available under a Creative Commons Noncommercial Sharealike (by nc sa) license, accessible at the author’s website at JackBalkin.Yale.edu
The theory of ideology that I have been developing in this book is based on the principle of ambivalence. Our tools of understanding are partially adequate and partially inadequate to understanding the world and what is just and unjust within it. Yet this idea presupposes that there are degrees of greater and lesser adequacy. It assumes that our thought can be “good enough” under some conditions, even if, in other settings, it seriously misleads us. Similarly, I have defined ideological effects as those that tend to promote or sustain injustices. This definition presupposes that things can be more or less just.
Nevertheless, our judgments of what is just and unjust themselves depend on our cultural software, which is a result of memetic evolution. As I argued in Chapter 1, human values are articulated and refined through culture. Perhaps the concept of justice, like that of truth, is merely the product of a particular development of cultural software. If so, the theory of cultural software faces three serious problems.
First, the process of critical self-reflection would be not only endless but pointless. For our critical judgments would be the arbitrary product of accidental evolutionary developments. Self-reflection would simply be another version of the continuing struggle of different memes to gain ascendancy in our thought processes.
Second, each culture has its own peculiar memetic development. If the idea of justice is merely a product of memetic evolution, perhaps each culture has its own conception of justice or has no conception of justice at all. Ideological analysis requires that we try to see what is just and unjust in the thought of the analysand. But if analysands do not have the same conception of justice as we do, we may not be able to understand their actions properly, for their concept of justice may be entirely different from ours. At best we will simply impose our own standards of justice on others who do not share them. And it will come as no surprise that, from our perspective, the views of others are found wanting.
Third, just as there may be no common idea of justice between cultures, there may be no common idea of justice between persons within the same culture. No two people share the same cultural software. If “justice” is merely an evolutionary product–a concatenation of particular memes that we have assimilated in our heads–perhaps we are simply imposing our personal conception of justice on others when we criticize their thoughts and actions. Perhaps there is no idea of justice that applies to all human beings-just individual ”justice programs” in conflict with all of the others, trying to spread and take over as many different minds as possible.
In short, if we take the memetic development of culture seriously, perhaps justice is an arbitrary mutation, peculiar to each culture’s or even to each individual’s memetic evolutionary history. Asserting that our judgments of justice apply to other cultures or other persons merely reflects the power of our own memes over our own imaginations. Of course, we may insist that others should look at justice and injustice the same way we do. But that is only because our memes are attempting to dominate and replace the memes in other people’s minds. This domination can occur in many ways–by persuasion, by indoctrination, by physical force or economic conquest–but it is at its basis a struggle of memes for superiority and dominance in the minds of human beings. Eventually, perhaps, all human beings may share a common sense of justice, but it will be only as the result of an effective conquest by certain particularly aggressive and effective memes.
Questions like these are serious problems not only for any theory of ideology but for any conception of human morality and politics. Any theory of ideology and any theory of moral discourse must confront them. I have phrased them in the way they arise for the theory of cultural software, but it is clear that much more is at stake in answering them than the fate of this particular theory.
Although it is possible to imagine that justice is peculiar to each culture’s or each person’s memetic development, it is impossible to be morally engaged with others given this assumption. I shall argue that ideological analysis, and indeed all moral discourse, must presuppose a transcendent value of justice. Tools of understanding produced by cultures to pursue justice are articulations of this value. Because the conception of what is just is necessarily related to wha tis true–for example, with what has happened and what is happening in society–moral discourse also presupposes a transcendent value of truth.
Defining Transcendent Values
The word transcendent has many meanings. For some people it recalls Plato’s theory of Forms in the Republic: a determinate and universal norm of Justice by which human institutions can be judged and found wanting. Something is just to the extent that it follows the formula or resembles the Form of Justice. But I reject this view, for I do not think that our values of truth and justice are determinate.
By a transcendent value, I mean:
1. A value that can never be perfectly realized and against which all concrete articulations and exemplifications remain imperfect or incomplete. A transcendent value is also a transcendent ideal.
2. A value that appears to us as a demand or longing. A transcendent value seems to call out to us to enact it in our culture and institutions. Our sense of justice seems to demand that we correct injustices when we recognize them; our value of truth seems to demand that we correct falsehood.
3. A value that is inchoate and indeterminate, which human beings must articulate through culture but which is never fulfilled. Precisely because the demand of a transcendent value is inchoate and indeterminate, it can never be completely satisfied. We attempt to realize and understand a transcendent value through its articulations in culture: these include the positive norms of our culture, our technology, and our institutions. But these articulations are always incomplete and imperfect. Our institutions and theories of justice always fall short of what justice demands. Hence there is an ongoing dialectic between transcendent values and their cultural articulations.
4. A value whose existence is presupposed by some essential aspect of human life or some essential human activity. Thus the argument for the existence of a transcendent value is transcendental; the existence of the value must be presupposed given the nature of the activity. Hence we can also speak of transcendent values as “transcendental” values.
Not all human values and ideals are transcendent. Machismo and meekness, for example, do not fall into this category. Many human values and ideals satisfy some of the four conditions listed above but not others. It is by no means clear how many transcendent values there are. But I believe that moral and political discourse requires at least two: truth and justice.
For some the very notion of truth as a “value” will seem odd. A sentence in a natural language, they will say, can have the logical value of being true or false, but truth itself is not a value in the same way justice is. Many philosophers hold that a sentence in a natural language is true when what it says bears a certain relation to the physical world or to other beliefs we currently hold. These correspondence and coherence accounts miss the phenomenological dimension of truth. Truth appears to us not only as a property of sentences but as a demand for understanding and recognition. Thus when I say that truth is a value, I am not attempting to offer an analytic definition. I mean that human beings have an inexhaustible drive to understand what is the case and what is not in the world around them. It is this value that we experience as a demand.
Transcendental Arguments for Transcendent Values
Transcendent values are similar to what Kant meant by regulative ideals: these values are a necessary precondition to certain forms of thought and certain types of activity. Hence the argument for transcendent values is transcendental. A transcendental argument is a “can’t help it” argument; it claims that we cannot avoid presupposing something when we engage in a certain kind of thought or activity that we cannot help thinking or doing.
Transcendent values of truth and justice are necessary preconditions to ideological analysis, but one does not have to engage in ideological analysis. However, ideological analysis is really a special case of the more general activities of moral and political understanding and moral and political discourse. This is hardly surprising; the skeptical argument about justice that I offered above does not merely undermine the project of ideological analysis; it also undermines the possibility of moral and political judgment about other persons and other cultures.
To be sure, moral and political judgment and moral and political discourse are not logically necessary–people can live like hermits and have no contact with each other. But they are practically necessary. As soon as human beings come in contact with each other, live with each other, or affect each other’s lives, questions of justice between them necessarily arise.
People often like to say that certain beliefs are “true for them” or “right for them” but not necessarily for others. This is a simple way to avoid controversy and appeal to a sense of fair play and tolerance. Particularly if what one believes is likely to be thought unusual or odd, it is easy enough to deflect anticipated criticism by asserting that what one believes is “true for me” or ”right for me” but that one wouldn’t dream of insisting that the belief has to be true or right for others. This is especially so in a pluralist society like our own, in which respect for differences of opinion (or at least the appearance thereof) is thought to be a virtue.
But the practical difficulty of “true for me” or “right for me” arises precisely when our actions affect other people and come into conflict with other people’s values and goals. Then we have to defend what we are doing, either to those we affect or to someone else. At that point we can no longer treat truth and justice like a pie, from which everyone gets to take away his or her own personal and private share. We must regard truth and justice as something that has claims on others besides ourselves. We must abandon the convenient dodge that what we believe is true and right is true only for us and right only for us and for no one else. Of course, we can continue to insist that all individuals have their own truth and their own justice, and that all we are really asking for is tolerance. But then we must claim that our view of tolerance is one that isn’t just “true for us” but should be respected and accepted by others as well.
Transcendent ideals of truth and justice are presupposed in our understanding of encounters between people as encounters between subjects of justicethat is, as the sort of entities that can be treated justly or unjustly. Questions about what is true and what is just necessarily arise whenever people affect each other’s lives. They arise when people meet together by design or are thrown together by chance. They arise when people live in a single community or when they encounter each other through travel, conquest, or colonization. They arise when people meet face to face in open dialogue or when they affect each other’s lives without meeting, as when a bomber places an explosive on an airplane or a factory owner pollutes a river upstream from people he has never met. As soon as we encounter an Other, justice presses its demand on us, whether we respond to that demand or not.
To be sure, people often try to avoid the mutual recognition of others as people who can be treated justly or unjustly. They may refuse to see the people they affect as subjects of justice. So conquering armies and slaveholders have often believed that their victims were less than human; they pretend that the people they subjugate are like inanimate objects to whom no justice is due. Yet even though conquerors and slaveowners refuse to recognize others as subjects of justice, we cannot understand the meaning–and the inhumanity–of their actions until we recognize these actions as an encounter between such subjects. It is ironic but true that we cannot understand the depth of injustice without an idea of justice.
Nevertheless, because the idea of justice is indefinite and indeterminate, the boundaries that demarcate subjects of justice are always contested and unclear. Today, after hard-fought battles over human equality, most people draw that line to encompass all human beings but no other entities. Perhaps in time we will draw it differently, and the nature of the “we” who draws it will change accordingly. But this potential for change simply reflects the fact that our notions of justice are always imperfect and incomplete. The indefiniteness of the boundaries of the subjects of justice is simply another way of expressing the fact that justice is a transcendent ideal.
We should note, moreover, that animal rights advocates might think it possible to act unjustly toward lions and bears without believing that these animals can act unjustly. They might compare the situation of animals to those of newborn infants and certain mental incompetents, who can be treated unjustly but cannot act unjustly toward others. Put another way, they can claim that an entity can be a subject of justice without being an agent of justice. The boundaries of justice must include both the question of who is a subject and who is an agent of justice. The interesting question of whether subject and agent can be separated in the way the animal rights activist describes is beyond the scope of this book. My point, rather, is that the indeterminacy of the boundaries of justice–the indeterminacy of who is a subject or agent of justice–is part of its transcendent character.
Transcendent ideals of truth and justice are transcendental because they frame the structure of our understanding of human action. We need them to understand the meaning of human action in encounters with others, whether this action is directed at us or at third parties, and whether the encounter is friendly or violent, fair or oppressive. Understanding others in dialogic encounters is a special case of understanding human action generally, and ideological analysis is a special case of understanding a dialogic encounter.
To take an extreme example, suppose that a conquering army finds a defenseless group of women and children huddled and starving in the cold. The army then proceeds to execute them and seize their possessions. In this encounter, there is no dialogue, no reasoned analysis, no self-critical doubt and reconsideration. There is only the brute act of power. How is a transcendent ideal of justice presupposed in this vicious act, which allowed no time for dialogue and involved no attempt at mutual understanding? It is implicated in our subsequent understanding of what has happened.
We cannot understand the meaning of this massacre as a human action except by reference to an ideal of justice that applies to both the victors and the vanquished. Even if none of the victims is alive to tell their story, we cannotunderstand what their murderers did–as the brutal actions of responsible individuals rather than as the random or determined actions of objects–without reference to a common and transcendent ideal of justice. We do not accuse stones and rocks of injustice when they fall in an avalanche and kill innocent people. We do not accuse lions and bears of injustice when they attack people. They cannot act unjustly because we do not regard their action as being of the same order as human action. What distinguishes the latter kind of action is precisely the fact that it can be just and unjust, and furthermore, that its meaning cannot be adequately understood except against this fact. Because human action is this kind of action, we must presume an idea of justice as part of our framework for understanding it. Finally, we cannot understand the meaning of the massacre unless we recognize that it happened to subjects of justice–to the sort of entities to whom it is possible to act unjustly. The idea of justice frames our understanding of the meaning of this encounter in terms of both the nature of the action and the nature of its results. That is why justice is a transcendental as well as a transcendent ideal.
Yet, one might object, why do we need to presuppose a transcendent ideal of justice to understand the meaning of what happened at the massacre? Why can’t we simply apply the standards of justice of our own culture? Often we do simply point to the positive norms of our culture to judge others. Yet these norms of justice themselves presuppose a transcendent ideal. And when our views are challenged by those who do not share our culture’s norms, we will inevitably be led to reassert this ideal.
Suppose, then, after we have condemned the massacre, that the conquerors could speak to us. “What right have you,” they might say, “to apply your standards of what is just and unjust to us? By our own culture’s standards, what we did was regrettable but necessary. We had the right to do what we did and so we did it. Your culture’s standards, parochial as they are, apply only to you and not to us. They can have no claim on us. You think you have understood what happened. But we think you have completely misunderstood and mischaracterized what we did.”
To respond to this argument, we must ascend from the positive norms of our culture to a transcendent norm. We must insist that what the conquerors did was unjust not only by our own standards but by a standard that they, too, should agree to; their failure to agree to it shows that they are mistaken, or wicked, or both.
At the moment we make this claim, we must acknowledge that our own views, and the views of our culture, might actually be limited or parochial in some respects. For we appeal to a transcendent standard that might judge andfind both cultures’ norms wanting. Nevertheless, our ascent to the transcendent norm allows us to turn the conquerors’ argument back on them. For we can say to them: “If standards of justice and truth are internal to each culture, you can have no objection to our characterization of you as war criminals. For just as our standards can have no application to you, your standards can have no application to us. We are as correct in proclaiming your evil in our culture as you are correct in proclaiming your uprightness in yours. But your very assertion that we have misunderstood you undermines this claim. It presupposes common values of truth and justice that we are somehow obligated to recognize. And on that ground we are prepared to argue for your wickedness.”
I have used an imaginary dialogue to show how our understanding of human action presupposes transcendent values. This use of the dialogic form was no accident. The rhetorical structure of dialogic encounters reveals the regulative nature of transcendent ideals in a particularly striking way. Suppose that we find ourselves in a debate with someone about a question of public policy. Her views are very different from ours. We attempt to persuade her; failing this, we try to persuade a third party that our views are more reasonable than those of our opponent. Our very attempts to convince the audience and justify our own position require that we appeal to common ideals of justice and truth that are binding on both ourselves and the audience. Moreover, we appeal to these common ideals even if we disagree among ourselves about what those ideals require.
We saw previously how our understanding of injustice presupposes an ideal of justice. A similar phenomenon is at work in dialogic encounters. Even when we accuse our interlocutors of great evils, we make reference to a common value of justice that we claim they have failed to live up to. And their defense, even if unconvincing to us, will appeal to reasons that they insist should persuade us and exculpate them. When we criticize our opponent to a third party, we invoke an ideal of justice that applies not only to ourselves and the audience, but to the person we criticize.
In short, transcendent ideals are presupposed by the rhetorical situation of having to persuade an audience. They seem to spring forth magically from the rhetorical encounter. Like a beautiful mosaic whose pattern emerges from the juxtaposition of diverse stones, the framework of transcendent ideals that undergirds the rhetorical situation emerges through the confrontation between different and conflicting perspectives.
Moreover, these ideals undergird the rhetorical situation regardless of our private intention to tell the truth or to act justly. People often use arguments to deceive each other and convince each other of things that are unjust. They bully and coerce each other with their words. They take advantage of their audience’s lack of information, or its emotional, political, or economic weaknesses. Nevertheless, even when we are being deceitful and trying to persuade the audience to believe what we know is not true, we phrase our appeal in terms of values of truth and justice that we claim are binding on both us and the audience. The ideal of truth frames even our act of lying, for the ideal is presupposed by our decision to lie. Without a notion of truth, the practice of deceit becomes incoherent, just as, without a notion of justice, the practice of injustice makes no sense.
The analysis of ideology is a special case of the dialogic encounter. In ideological analysis, we interact with a person, a text written by that person, or a culture. We try to understand their ways of thinking; through this processwe learn something about ourselves and our own judgments. This process creates a virtual dialogue with others, even if they are not physically present. Our ambivalent attitude toward cultural software means that although we criticize the others, we must also allow them to criticize us. Thus ideological analysis is a special case of the more general situation in which we are confronted by people who disagree with us about what is just, and we must deal with their objections through argument and persuasion. Because ideological analysis is a kind of dialogic encounter, it presupposes the same transcendent ideals.
As we saw in Chapter 6, any ideological analysis we apply to the thought of another could, in theory, be applied to our own thought. Because the position of analyst and analysand is symmetrical, we must assume that neither we nor the analysand has a completely accurate or just view of the situation, and that the cultural software of each is partially adequate and partially inadequate to understanding what is just. We must take an ambivalent attitude not only about the other party’s cultural software but also about our own.
Thus our ideological analysis assumes that neither we nor the analysand has a monopoly on what is true or just and that neither of our views offers a perfect, complete account. This already presupposes ideals of truth and justice that apply equally both to ourselves and to the analysand and that are not identical with either of our own views. The ambivalent conception of ideology presupposes common ideals against which both of our views might be found partially inadequate.
We must postulate transcendent norms whenever there is a clash or encounter between the positive norms of different cultures, different groups, or different persons. This encounter can be the virtual dialogue of ideologicalanalysis, the actual dialogue of debate and argument, or the physical encounters of politics, warfare, and economics. It can be a genteel discussion or a violent confrontation. In each case, our encounter with an Other causes the transcendent norm magically to spring to life.
Cultural Relativism and Imperialist Universalism
The idea of a transcendent standard of justice might seem to resemble another, importantly different position: I shall call this position imperialist universalism. This is the view that there are universal concrete standards of justice and human rights that apply to every society, whether pre- or postindustrial, whether secular or religious, and that it is the duty of right-minded people to change the positive norms and institutions of all societies so that they conform with these universal norms of justice and universal human rights. This position is worrisome to many people because they see it as a cover for the imposition of a particular set of standards of justice and a particular set of institutions on all of the peoples of the world, whether or not these standards and institutions are appropriate to all, and even if their imposition will result in considerable misery and human suffering. In other words, people are worried by claims of universal standards of justice because they are worried that some form of cultural or political imperialism, particularly from the industrialized West, will be furthered under the name of universal justice and universal human rights. Similar concerns apply within a single multicultural society like the United States. Ethnic minorities, religious minorities, and women may well be concerned that the attitudes, perspectives, and values of majority groups and of men will be imposed on them in the name of universal standards of justice.
All of these concerns are valid ones, but none of them is an argument against the existence of transcendent values. A transcendent value cannot be identical to the values of any particular group because it is an indeterminate standard against which the positive norms of all groups must be seen as incomplete and imperfect. To identify the positive norms and values of one’s own group with transcendent standards of truth and justice is already to misunderstand the nature of such a standard.
In fact, one needs to presuppose transcendent standards to make the case against imperialist universalism. The banner of universal standards of human liberty has often been waved in front of conquering armies, whether military, cultural, or economic. People have often invoked universal standards of reason and justice to promote unjust or inappropriate measures that are unfair to foreigners, women, and minority groups. But one can criticize these usurpations only if one presupposes a transcendent standard of justice. For what can the argument against such actions be other than that they are unjust according to a sense of justice that applies both to oppressor and oppressed? Our very notions of tolerance and respect must be based on values that apply both toourselves and to other peoples and lands.
To criticize imperialism we must argue that the concrete norms and institutions of the West are not universal standards but only imperfect articulations of justice. Applied thoughtlessly to other cultures, they will produce grave injustices. The argument cannot be simply that the imposition of “universal norms” is unjust because these norms conflict with the norms and institutions of the other culture; it must be because such an imposition offends a sense of justice that transcends both the positive norms of the West and those of the other culture, and against which each might be found wanting. The argument within a multicultural society is similar: We cannot simply argue that it is unjust to impose the values and perspectives of whites and males on women and persons of color because each group has its own values and perspectives. For if the perspectives and values of majorities can have no purchase on minorities, how can it be just to impose the perspectives of minorities on majorities? If the former are improper to apply to the latter, why are the latter proper to apply to the former? Rather, one must acknowledge that the perspectives of each group are partial and incomplete, and that each has the obligation to understand the possible claims of truth and justice in the other’s perspective. Like any dialogic encounter, this demand requires an ascent to transcendent values.
We might try to avoid this conclusion in two ways: First, we might argue that it is unjust from the perspective of the majority’s own values to apply its values to minorities. But then we have no response if the majority disagrees and insists that from its perspective it is being entirely fair. The majority is surely as good a judge as we are about what its values require of it. Second, we might argue that the majority should try to understand the perspectives and values of minorities because the majority is oppressive and the minority is oppressed. But this argument already appeals to a common idea of justice that applies to both groups. Moreover, we are surely not arguing that in all things the perspectives of minorities are right and the majority’s perspective must yield to it. Minorities can also have biased and parochial views of a situation. Moreover, there may be multiple minority perspectives, some of which are in conflict with one another. The argument must be rather that various minority perspectives have important elements of truth and justice in them that are likely to be overlooked by majority perspectives because of the majority’s position, interests, and cognitive framework. In other words, the multicultural situation we have been considering is the familiar one of ideological analysis, and the same arguments apply to it.
If a belief in transcendent values does not require a fixed and determinate standard of justice that applies to all cultures, why does it not collapse into cultural relativism? In fact, there are two forms of cultural relativism, one which I accept and another which I reject. The first claims that different cultures have different norms because they have different histories and have faced different problems, and that if we understood the history and problems of different peoples, many of their norms would not seem so strange to us and would even seem justified. This claim is not inconsistent with a belief in a transcendent value of justice; indeed, it proceeds from such a belief. The transcendent value is indeterminate; it has simply been articulated in different ways in different cultures.
This version of cultural relativism treats the other culture with hermeneutic charity-like a person or a text that has something to teach us. In this sense it takes an attitude not fundamentally different from textual interpretation orideological analysis. Hermeneutic charity toward the norms of another culture is a necessary admonition against rushing to judgment and may even enlighten our own views. It is not identical, however, to the fantastic claim that if we take enough time to understand another culture, we will agree that whatever that culture considers just will turn out in fact to be just. This conclusion is equally inconsistent with an ambivalent attitude toward cultural software. After all, our own culture’s norms are partially inadequate and unjust, and it is likely that the same is true of other cultures. It is entirely possible that when we learn more about another culture, we may conclude that some of its norms and customs are not justified by its peculiar circumstances and history, just as our own culture’s history does not excuse all of its present practices.
Nevertheless, even if we find that a culture’s norms are partially unjust, there is still the further question of whether it is right for others (including ourselves) to take active steps to abolish that culture’s practices. There may be good reasons to avoid doing so. First, intervention might greatly disrupt the society and cause even greater suffering and misery than is created by the current injustices. Second, our notions of tolerance and respect for political self-determination may counsel against intervening in the affairs of another country or another culture. Once again, none of these claims is inconsistent with the postulation of a transcendent value of justice. Indeed, all of them implicitly make reference to such a value.
Moreover, this version of cultural relativism is compatible with the recognition that some different ways of living may be incommensurable with each other, as long as not all are incommensurable. We may not always be able to say whether one way of life is more just than another. But this does not mean that no ways of living are more just than any other, or that no changes in a culture’s norms and mores would make it more just. Even if it is not clear how to compare a constitutional democracy in an industrial age with a close-knit agrarian farming community, it might still be possible to argue that the practice of appropriating lands from the weak and the defenseless in the latter community is unjust and should be changed.
A second version of cultural relativism, which I reject, holds that questions of what is just and unjust exist wholly within a given culture. One can judge a culture’s justice only by its own norms, because “justice” is by definition something relative to a system of positive cultural norms. Because of our own cultural upbringing we may not be able to help judging other cultures. But applying our ideas of justice to theirs is a category mistake. We are quite literally talking nonsense when we do this.
This conception of cultural relativism is inconsistent with the notion of transcendent values. For that reason it is also incoherent. If notions of justice are wholly internal to each culture, then no culture can meaningfully object when another takes it over, seizes its lands, and massacres its inhabitants. It can insist that under its own standards what the other culture is doing is wrong. But this should hardly bother the invaders, since these standards cannot by definition apply to them. If applying the aggressor’s standard of justice to the victims is a category mistake, then applying the victim’s norms to the aggressors is equally so.
The reason why this position makes no sense is that one can always object, and this objection is meaningful. But the very fact of raising a meaningful objection already places the victim in a dialogic encounter with the aggressor.And as we have seen, the rhetorical structure of this encounter presupposes transcendent ideals.
People may be drawn to this second form of cultural relativism because they feel it is important to respect the values of other cultures and as a perpetual reminder that their own culture does not have all the answers. But ironically,these laudable reasons require the presumption of transcendent ideals. We need a transcendent value of justice to respect another culture’s norms as well as to criticize them. To understand why the actions of other cultures make sense, we must already believe in a common and transcendent value of justice that their norms partially and incompletely articulate, as do ours.
We can see a similar difficulty in Jean-Francois Lyotard’s vision of justice as constituted by a multiplicity of different and incommensurable language games. According to Lyotard, discussions of justice are like moves in a language game. Each language game has its own standards of justice, and each grows and extends itself by the development of new moves and new rules created through playing the game. The problem, as Lyotard sees it, is the danger of imperialism-one prescriptive system may attempt to dominate and extinguish the others, or fail to recognize their singularity and their claims to exist and to develop on their own through their own internal logics. Thus, in addition to the multiplicity of justices, Lyotard concludes, we must also have a “justice of multiplicity.” We must have a justice, Lyotard argues, that “prescribes the singular justice of each game as it has just been situated.” The justice of multiplicity allows innovation within each game but forbids “terror”-which toLyotard means the attempt by one game to take over and dominate most of the others.
It is a tolerant vision, but as Lyotard’s interlocutor, Jean-Loup Thebaud, reminds him, it also undermines the notion of mutually incommensurable language games. For now Lyotard has himself become “the great prescriber,” ordering the various games of justice to respect each other’s boundaries. If justice is purely internal to each particular game, how can any game have the authority to prescribe the boundaries of the others? Such a game must be engaged in its own form of imperialism, like a superpower preventing wars of conquest between smaller countries by the deployment of its own troops.
To avoid this conclusion, we must acknowledge that some forms of oversight are just and others unjust, and this judgment must be able to be acknowledged within each distinct language game. Yet this means that the games of justice cannot be completely walled off from each other but must be interpenetrating. Discussions about justice between games must be recognizable and coherent moves within each game. So all the games must have something in common with each other; they must all be able to communicate with each other, at least on the subject of justice between them. Lyotard and Thebaud end their discussion in laughter, recognizing the incongruity of their situation. But this incongruity also shows us the inescapability of transcendent ideals.
Transcendent Values and Positive Norms
Not all arguments about what is just explicitly refer to a transcendent standard of justice. People often argue about justice by referring to the positive norms of their culture and its institutions. A culture’s positive norms of justice, even where they are quite unjust, constitute its attempt to articulate the transcendent value of justice. These positive norms and institutions always presuppose a transcendent value of justice, because it always remains possible for people to criticize their culture’s norms and institutions on the grounds that they are not sufficiently just.
In the Jim Crow South, for example, people might have argued that a black man should not sit next to a white woman in a public bus because it is disrespectful or because it is against the law. These arguments are appeals to the positive norms of that culture and to its laws. Yet it was also possible for southerners to understand that these norms and institutions were unfair and to criticize them, even while living in the same culture.
This ability to recognize and critique the injustices of our own culture is another reason why it makes sense to speak of transcendent values. It is hardly surprising that people are often critical of other cultures and their traditions. What is surprising is that people are able to apply their critical focus to their own culture’s norms and institutions. To be sure, it is always possible that these criticisms are misguided in any particular case. But the mere ability to articulate them already presupposes a sense of justice against which the norms and institutions of one’s society might be understood as imperfect and inadequate.
One might object that when we criticize Jim Crow laws, we are only offering criticisms based on some other aspect of our culture’s positive norms. Enforced segregation on public buses is unjust because it offends our culture’s commitment to equality, as evidenced in other customs and other laws. So perhaps criticism of our own culture can proceed without the need to postulate any transcendent value of justice. People simply engage in internal or immanent critique of their own culture and institutions by transferring judgments and norms from one aspect of that culture to another.
But this poses an interesting problem: If our judgments of justice come solely from the positive norms of our culture, how do we know that there is a conflict between segregation on public buses and our commitment to equality? Why do we not see the status quo as a perfect accommodation between competing interests, whose differences are resolved differently in different areas of social life? Thus blacks are permitted social equality with whites in some situations but not in others because the balance of competing considerations is different. The mere recognition of a problem as a problem already takes us beyond a simple application or transfer of cultural norms. Our very sense that there is a conflict between a culture’s norms already presupposes a transcendentstandard of justice.
Even if we concede that there is a conflict between different positive norms of our culture, both the customs and laws of Jim Crow and our egalitarian commitments are equally part of our culture and institutions. If the culture’s norms and institutions are the only standard of justice, what makes one norm or one institution more just than the others? By themselves the culture’s norms cannot tell us how to reconcile a conflict between them, for both antagonistic elements are present in the culture. To take another example, suppose current law protects blacks but not homosexuals from job discrimination. If we believe that this treatment is inconsistent, why does it follow that the just solution is to give homosexuals protection from discrimination? Why not remove legalprotection for blacks?
One might respond to this difficulty by arguing that our commitment to egalitarianism is a more central feature of our cultural norms than our commitment to racial separation. However, this claim can be understood in two different ways. If something is central because it is more prevalent, we refer only to a positive norm. (For purposes of argument I assume that there are noncontroversial ways of judging prevalence. If there are not, this simply reinforces my point.) Nevertheless, mere prevalence does not guarantee the worth of a cultural practice, unless our only goal is to reinforce the most prevalent positive norms simply because they are more prevalent. Racial inequality may be a central and pervasive feature of a regime of Jim Crow or apartheid, for example, but this does not make it a worthy basis for critique of the few egalitarian norms that might remain elsewhere in the culture. By reinforcing the most prevalent practices of a culture we may reinforce its most deeply unfair elements. On the other hand, by “central” we could mean “more valuable” or ”more just.” In that case our judgment must refer to a transcendent conception of value or justice that informs our notion of centrality.
We can resolve these puzzles only if we assume that there is a value of justice that our culture imperfectly responds to. Then we can sensibly say that our culture is more just in some regards than in others, and that all conflictsmust be resolved in favor of the more just features. Hence even when we engage in a so-called internal or immanent critique of our culture’s norms and institutions, we must presuppose a transcendent value of justice.
Transcendence of Value Versus Transcendence of Position
The notion of “transcendence” might suggest escaping or getting outside of culture in order to judge it. But this way of talking confuses two very different ideas: the transcendence of position and the transcendence of value. A transcendent position is a perspective that is not limited in any way; it is a sort of ”God’s-eye view.” A transcendent value is a value that cannot fully be realized; it serves as a regulative ideal to our understanding. The notion of transcendence is quite different in the two cases; the first refers to a perfect perspective for understanding, while the second concerns the necessary framework for normative understanding in all perspectives.
Because the word transcendent applies to both, it is tempting to try to connect them. From a transcendent position, perhaps we could fully understand a transcendent value. We could know what was really and completely just if we had perfect information and no impediments to our understanding. But in fact, the idea of a transcendent position is inconsistent with the idea of a transcendent value. A transcendent position makes justice determinate, and a transcendent value cannot be made determinate.
Moreover, the notion of a transcendent position is incoherent. To have a perspective is already to be limited in some way. To have a perspective is to have some apparatus for understanding the world. And to have an apparatus means already to see things in some ways rather than others. To avoid the limitations of each possible apparatus of understanding, one would have to have no apparatus at all. And then one would understand nothing, not everything.
Nevertheless, the idea of a transcendent position has been an attractive one in the history of human thought. The reason is not difficult to understand. Intelligent people understand that disagreements arise from differences in perspective, from lack of information, and from the cognitive limitations of position and interest. They recognize, moreover, that the thought of all human beings is limited in one way or another. Thus it seems natural to think that limitation of perspective is the cause of disagreement and uncertainty about justice. If so, then perhaps by removing the causes we could rid ourselves of the effects. We could put our knowledge about what is just on a sure footing by reference to a perfect perspective or a perfect system for judgment.
These considerations explain the appeal of two very common approaches to justice: ideal observer theories and ideal process theories. Ideal observer theories claim that justice is what an observer under ideal conditions would find to be just. Ideal process theories argue that correct judgments about justice are the product of what emerges from some ideal process of decisionmaking. Thus justice is the product of ideal decisionmaking conditions or a consensus reached under ideal conditions of dialogue.
Ideal observer theories attempt to avoid speaking in terms of a transcendent ideal of justice by speaking instead in terms of transcendence of position. Ideal process theories try to avoid reference to a transcendent ideal by manufacturing justice out of an ideal procedure. The attempt to avoid the transcendent nature of justice creates problems for each theory. Conversely, to the extent that they are successful articulations of justice, they presuppose a transcendent ideal.
Ideal observer theories face two problems: First, they postulate an observer with ideal characteristics working under ideal conditions. Yet our notion of what makes these characteristics and conditions ideal already presupposes transcendent values of truth and justice. Things are just, not because they are so judged by a person with ideal characteristics; rather, these characteristics are ideal because they help a person understand what is just.
Second, the notion of an ideal observer of justice begins to unravel as soon as we inquire into the characteristics of the observer. All observers have a perspective, but ironically, the ideal observer cannot. To have a perspective isalready to be a finite being, with a particular history and a particular set of needs, concerns, and desires. Moreover, having one perspective to some degree precludes having others, because some perspectives are mutually incompatible, or are produced by living different kinds of lives, all of which no single person could lead. Is the ideal observer white or black, pregnant or nonpregnant, untouched by violence or the victim of child abuse? It is clear that she or he can be none of these things, for to be any one of them would already shape and limit her perspective. And this is precisely the problem: to have a perspective is to exist as a finite human being with a particular set of commitments and a particular life history. The ideal observer, on the other hand, can have no gender, no history, and no group identification. The observer can have experienced no defining moments in life, can belong to no political party, can adhere to no ideology or worldview. The observer cannot, in other words, exist as a human being.
The great irony here is that our ability to understand justice stems from our situatedness and our finitude. Our life experiences are the raw materials from which we make sense of the normative demands of life. Without them we cannot understand anything at all. What we most have in common with other human beings is what separates us: our finitude, our inadequacies, and our limitations of perspective. We are able to make normative sense of the world because we exist as individuals with a history, who have experienced things and been changed by them, who have perspectives and cultural software that simultaneously limit and empower our understanding. This finitude, this historicity, this limitation, is what makes the transcendent appear to us as transcendent–beyond our grasp and full comprehension. Ideal observer approaches fail because they rid observation of its humanity, which is inextricably linked to conditions of human imperfection.
Ideal process theories describe justice as the outcome of an ideally fair process or decision procedure. Like utilitarian and deontological theories of justice, these theories are useful heuristics for articulating our sense of justice. But they never completely capture our sense of justice. Ideal process theories presuppose transcendent ideals, they do not produce them. Whatever procedures we employ cannot justify themselves; they must appeal to ideals of truth and justice in order to convince us that they are fair and likely to produce correct conclusions about what is true and just. This problem is similar to that faced by ideal observer theories: Things are just, not because they are the result of an ideal procedure; rather, the procedure is ideal to the extent that its conditions are fair and it leads to just results.
Procedures cannot be determinative of justice because we can always criticize them in terms of the results they produce: the best criticism of the fairness of a procedure is usually the injustice of the results it produces. So a procedure must at best be considered a way of approximating what is just; it is an articulation of justice, and like all articulations, it will necessarily be imperfect, producing results that are always subject to further criticism.
Ideal dialogic theories are interesting and important versions of ideal process theories. They hold that truth or justice is what people would agree to after a dialogue under ideal conditions. Earlier I noted that transcendent ideals of truth and justice seem to emerge naturally out of the structure of dialogic encounters. Hence it is natural to attempt to identify truth and justice with what results from ideal dialogue. One might thus identify truth or justice with the actual consensus of the community in the long run. Or, recognizing that many encounters are unfair and coercive, one might identify truth or justice with the consensus emerging from a dialogic encounter under ideal conditions.
Ideal dialogic theories are distinctive because they emphasize that truth and justice are linked to ongoing processes that involve both discovery and creation. The substance of an agreement about truth and justice is true or just not because it conforms to some preexisting test or criterion but because it is the result of a contingent process that results in agreement. The fact of agreement makes true or just what would not be true or just absent the agreement.
Yet even ideal dialogic theories presuppose a transcendent ideal of truth and justice. Such theories well understand that the brute fact of agreement does not make the substance of the agreement true or just. They must distinguish, in Habermas’s terms, between a rationally grounded consensus and a false consensus. The task of the ideal dialogic procedure is to make an agreement between finite human beings of limited perspective impervious to criticism on the grounds that it might be unjust or partial. We can try to solve this difficulty by postulating that the agreement takes place under ideal conditions. Yet as before, transcendent ideals of truth and justice are presupposed in articulating the ideal conditions. What makes these conditions ideal is that they lead the parties to an understanding of what is true or just. Thus the fact that the parties lack important information or suffer from unconscious needs to reduce cognitive dissonance tends to make the results of their deliberation suspect. But the reason why they are suspect is not simply that they deviate from the criteria of ideal dialogue. It is because they are likely to produce a consensus that is not true or just.
Ultimately, however, the problem is not simply that we need better procedures; the problem is that we need better people. The participants in any community are finite beings of finite intelligence whose understandings are shaped and circumscribed by their history. Their perspectives are necessarily limited by the partial inadequacy of their tools of understanding and by their inability completely to take into account situations and consequences beyond their apprehension as well as those which may arise in the future. No dialogue between finite human beings, whose understanding is constituted by the historical development of cultural software, can be an ideal dialogue under ideal conditions. For their perspectives are always limited by the fact that each has a perspective. The history of their discussions is always limited by the fact that each has a history. The only truly ideal dialogue would be one between gods. They would already understand everything, and therefore there would be nothing left to say.
I believe that there is a deep connection between being the kinds of finite beings we are–who have absorbed tools of understanding produced through evolutionary bricolage–and our experience of moral and factual truth as transcendent ideals. Although people may have used the fiction of a transcendent position to understand transcendent values, the two notions are actually opposed to one another. It is precisely because transcendence of position is impossible that we experience justice as a transcendent ideal. To have a perspective about human action is already to be imperfect, fashioned from what Kant called “the crooked timber of humanity,” imbued with tools of understanding that are the product of history and the object of ambivalence. Yet to have a perspective about human action is already to presuppose the transcendent. The transcendent exists because we are imperfect, because we have a perspective. The transcendent is the frame through which we understand the normative meaning of human action. The transcendent is the limit that shapes our horizon of moral experience.
The Muse of Justice
As limited and imperfect human beings, we cannot stand outside our cultural practices and our cultural software. And our values must be immanent, in the sense that we can express them only through the tools bequeathed to us by culture. How then can we make sense of the transcendence of human values while recognizing that we always make judgments within culture? How can human values be both transcendent and immanent at one and the same time?
In Chapter 2 I argued that we should think of value as a verb, not a noun, as something we do or feel, not something we have. Human values are inchoate and indeterminate urges or demands that are articulated and refined through culture. A transcendent value is a special kind of human value, a value that can never be fully fulfilled. A transcendent value is an inexhaustible demand.
This way of speaking conflicts with the standard metaphor that we employ to describe evaluation, the metaphor of measurement. We evaluate things by measuring them against our sense of justice, just as we measure the length of a table by laying a ruler against it. Hence we have the familiar metaphors of number (things are more or less just), weight (justice comes from balancing competing considerations), size (the lesser of two evils), and distance (coming closer to or diverging from justice). In the standard conception, values work like scales or rulers, and evaluation is a kind of measurement.
This conception has important metaphorical entailments. The first is a separation between value and the thing valued. Because a value is a standard of measure, it must exist apart from the thing that it measures. One cannot use a ruler to measure itself anymore than one can use a balance to weigh itself.
As a result, the metaphor of measurement also seems to suggest that we must somehow stand outside culture in order to evaluate it. We must use a determinate conception of justice existing outside our existing culture to evaluate it, in the same way that rulers must exist independently of the objects they measure. Thus the twin notions of determinacy of value and separation from the object of evaluation are yoked together under the metaphor of measurement.
The idea of justice as an indeterminate or inchoate urge does not fit this familiar metaphor. Instead of a standard of determinate measurement, the transcendent value of justice is an insatiable urge. Thus we have two contrasting metaphors of the value of justice: justice is like a ruler of determinate length that we use to measure the world, and justice is like an indeterminate demand that can never be fulfilled despite our best efforts. Each of these is a metaphorical account of human value. Each is helpful in its own way, but neither can be usefully employed in all contexts and circumstances. To understand the phenomenon of transcendence we must recognize the metaphor of measurement as a metaphor, and exchange it for a different figure.
The contrasting metaphors of determinate measure and indeterminate demand produce different accounts of why our cultural institutions are imperfect, why there is no example of justice in the world that is perfectly just. There are two ways of expressing this inadequacy. One makes use of the notion of a determinate measure, and the second makes use of the notion of an unfulfilled but indeterminate demand.
Under the first metaphor, a determinate conception of justice exists apart from individual examples of justice and is used to measure them. So we explain the fact that no example of justice is perfectly just by saying that each example is an imperfect representation of a determinate conception of justice. The justice of a law or an institution is a question of the quality of the correspondence between the determinate idea of justice and the concrete example. Thus virtue is a process of good copying, and the virtuous person is a good copyist. One makes an institution just by copying the determinate idea of justice as accurately as possible in all of its details. But because no copy can be perfect, there is no perfect example of justice in the world.
Under the contrasting metaphor, justice is an inchoate yearning that we attempt to articulate through our cultural constructions. To be just we must construct examples of justice using the indeterminate urge for justice as our goad rather than as our guide. This means that the virtuous person is not a good copyist but a good architect. She attempts to satisfy her sense of justice by constructing just institutions. There are many different ways of constructing a just institution, depending upon the situation in which she finds herself and the resources she has available to her. Nevertheless, she responds to an indefinite and indeterminate value that can never be fulfilled. Her constructions cannot exhaust justice’s demand. Thus human cultural creations will always fail to be perfectly just, but not because they are defective copies of a determinate standard. Their imperfection arises from the necessary inadequation that must exist between an indeterminate and inexhaustible urge and any concrete and determinate articulation of it. This relationship of inadequacy between culture and value is what we mean by transcendence.
Note that unlike the metaphor of measurement, the metaphor of the indeterminate demand does not suggest that we must go outside our culture to evaluate our institutions. Rather, we feel the demand of justice as we construct and reconstruct our institutions using the cultural tools bequeathed to us. To feel the demand of justice we do not have to travel to a place beyond culture; the demand presents itself as a sense of the inadequacy of our tools that we experience as we work with them.
We might understand this idea better through an analogy to the myth of the Muse. In Greek mythology, the artist created works of beauty because of a Muse, who not only inspired but also demanded the creation of the work. But there are two different conceptions of the artist’s relationship to the Muse. In the first, the artist is no more than an amanuensis who copies out what the Muse dictates to her. Artists often talk this way out of a sense of modesty: they tell us that they merely wrote down what a higher intelligence created. Thisversion of the myth, however, disguises important features of human artistic creation. Most artists are not obedient copyists; they create only as a result of practice and hard work, and they suffer greatly for their art.
In the other conception, the Muse is a harsh taskmistress who relentlessly drives the artist to create the beautiful, often to the point of madness. The Muse demands enormous sacrifices of the artist but is never satisfied with the results, and so artists live their lives in a sort of perpetual bondage to their Muse. The copyist has the advantage of knowing what the finished product will look like; the servant of the Muse does not have this luxury. The servant must turn her inchoate sense and drive for beauty into a work of art, always with the risk that it will not please the Muse who goads her. In this story, the Muse is the mythological externalization of the human drive to value and create works of value. Thus we might say metaphorically that transcendent human values are like Muses; and that there is a Muse of justice as well as one of truth.
Do All Cultures Have a Concept of Justice?
So far I have assumed that when we discuss questions of justice with another person, both of us are speaking about the same concept. But what happens if we relax this assumption? Does this undermine the idea of a transcendent ideal of justice?
Suppose that we come across a culture that lacks a word for justice. Instead, they have a concept they calldharmatzedek, a term that I borrow from the Sanskrit word for duty (dharma) and the Hebrew word for righteousness (tzedakah). (I combine the two words because I do not want the reader to assume that I refer to the concepts of justice in either Hinduism or Judaism.) According to the views of this hypothetical culture, dharmatzedek is a cosmic order of the universe. Social order is a special case of the cosmic order. Things adhere to dharmatzedek when they reflect the proper order of nature, an order that includes not only human beings but animals, gods, and even inanimate objects.
Can we still say that a transcendent ideal of justice is presupposed in our conversation with members of this culture? This is really the question of whether it is possible for us to have a conversation with them about what is just and not just. The fact that they have no such word does not mean that such a conversation is impossible. We modify our existing cultural software all the time in order to understand what others are saying. For example, I have just introduced the concept of dharmatzedek into the present discussion.
Let us suppose that our communications with this culture lead us to believe that by dharmatzedek they mean the concept of natural order that I have described above. Then we will be able to have a conversation with them, for our notion of justice is a notion of achieving appropriate social order and rectifying inappropriate social order. It overlaps with their concept of dharmatzedek even if it is not identical to it. Their sense of social order will be very different from ours, and they may use very different ways of expressing it and making claims about it. But once we create a theory of what their concept means to them, we can begin the process of understanding how the world makes sense to them. Conversely, from their point of view, we will have a very strange concept called ”justice,” which concerns the order of society and the entitlements of individuals but does not concern their relationship to other things in the universe. They will see our concept of justice as a truncated and mangled conception of dharmatzedek, with an exaggerated focus on the concerns of individuals. But as soon as they formulate a notion in their own language that makes intelligible what we are talking about, they can begin to see that what we are saying makes sense from our perspective, even if they do not entirely agree with it.
Now suppose further that we discover that this culture is engaging in very inefficient forms of agriculture. Not only that: they refuse to engage in practices that would increase productivity. They believe that these practices would show improper respect for the land and disturb the moral order of the universe. They hold this view even though their forbearance means that many of their people will starve. And when we suggest new methods, they refuse to accept them because our methods are against dharmatzedek.
Note that my very description of their objection shows that we can understand why it makes sense for them to object to agricultural innovation. It is intelligible even if we think it mistaken. Conversely, their conversations withus enable them to recognize that our notion of “justice” is an impoverished version of dharmatzedek this will allow them to understand why we think that one should adopt the new methods, even though they are convinced that we are quite wrong. Once again, the very fact that we can describe our differences from them means that some sort of mutual understanding is possible, even if it is not a perfect understanding.
Nevertheless, we should not assume from our ability to communicate that the other culture really has the same concept of justice as we do and that their concept of dharmatzedek is parasitic on it. We should not assume that our ”justice” corresponds to some sort of “natural kind” and their concept of dharmatzedek does not. But if that is so, doesn’t this undermine the notion of a transcendent conception of justice? Does it mean that we must acknowledge a separate, transcendent ideal of dharmatzedek, and so on, for each different conception in each different culture?
Before answering this question, we might raise the stakes even higher. Most people think that Plato offered a theory of justice in his Republic. But the concept of dikaiosune (justice) in fourth-century B.C. Greece is hardly the same notion as the concept of justice we have today. Today in Western democracies we tend to think of justice in terms of getting what we are entitled to, fulfilling our duties to others, and avoiding injury to them. Our contemporary notion of justice is largely organized around the language of individual rights and focuses on interpersonal injury and benefit. Plato does not have this conception. His major concern in the Republic is how the individual fits into the social order. Whereas many (though by no means all) political theorists in the contemporaryWest tend to view the state as an instrument for fulfilling individual needs and protecting individual rights, Plato’s conception regards social order as primary and the well-being of individuals as secondary. For Plato, dikaiosune is satisfied when the individual exists in the right relation to his or her society. That is one reason why, in the language of contemporary conceptions of justice, Plato’s scheme for an ideal city seems so authoritarian to us.
Nevertheless, Plato is one of the founding voices in the Western discussion about the nature of justice. And he is also identified with the notion of a universal transhistorical conception of justice. Yet if Plato’s dikaiosune is notthe same as our contemporary conception of justice, perhaps we face the same difficulties in understanding his conception in the Republic as we face in understanding the meaning of dharmatzedek. How then can we claim that there is a single transcendent conception of justice when each culture seems to have a different conception, including those cultures that form the wellsprings of our own?
We can solve this problem by recognizing that our contemporary rightsbased notion of justice is not a transcendent conception. It is itself an articulation of a transcendent ideal. It has come into being at a certain point in history and will no doubt be replaced by some other normative conception in the future. Plato’s society-based notion of dikaiosune is also not a transcendent conception but an articulation that reflects the cultural software of his time. So, too, the natural order-based conception of dharmatzedek represents that culture’s articulation of a transcendent value. Each culture (and each person) shares this transcendent value, but each articulates it in a different way.
Throughout this chapter I have called this transcendent value “justice” because that is the word that comes most easily to me, given my cultural situation, my cultural software. And it is also the easiest way to explain the idea of transcendence to an audience that shares most of my assumptions. But if justice, dikaiosune, and dharmatzedek are all articulations of some higher transcendent ideal, what is the nature of that ideal? What is the common ground of all three notions? What is the cognitive framework that each of the three fleshes out partially and imperfectly, and that makes it possible for each of us to understand the other?
In a more abstract sense, we are talking about the transcendent ideal of a normative order. In each culture the members have a conception of a normative order that includes a notion of the subjects and agents to whom this normative order is relevant. Our notion of justice, the concept of dharmatzedek, and Plato’s dikaiosune are all ways of expressing the normative order that exists between the relevant subjects and agents. The subjects and agents of each normative order are “individuals,” and the normative order concerns their proper relation to each other. By “individuals,” however, I mean only the sort of sentient beings who can be subjects or agents, not the full-bodied conception of rights-bearing individuals that we associate with contemporary liberalism. In some cultures, animals and gods can be subjects or agents of the normative order. Moreover, in some cultures, what we call inanimate objects can also participate in the normative order because they are not, strictly speaking, inanimate-they are embodiments of or connected to gods and spirits.
Thus each culture recognizes an idea of a normative order, but each articulates it in a different way. There are as many ways to articulate the idea of a normative order as there are possible ways of articulating the relationships between individuals, society, and the universe. In Plato’s dikaiosune, for example, but not in a modern rights-based conception of justice, the normative order refers to the individual’s right relation to the state. Notions of individual entitlement are mediated through the language of this relationship.
It is equally important to recognize that some cultural articulations recognize only some of the possible elements of the normative order as salient. The concept of dharmatzedek, for example, includes our normative relationship to the universe. But the modern conception of justice makes a tripartite distinction between human beings, other living things, and inanimate matter. Justice is a relation between human beings (and possibly some animals). It does not extend to inanimate objects. The contemporary West deemphasizes the possibility of a normative order between ourselves and what our worldview sees as inanimate matter, even though there is currently much interest in protecting the environment. We in the West are more likely to argue for environmental protection because of the ways it will affect future generations of human beings or other living creatures than to claim that we have ethical obligations to rocks, stones, and bodies of water. Even if we recognized a normative relation between ourselves and inanimate matter, I suspect that we would not call it justice. We would give it some other name, like the “sanctity of nature.”
Comparing our contemporary notions of justice with other possible concepts like dharmatzedek or dikaiosune suggests the many different ways that cultures can articulate the transcendent ideal of a normative order between individuals that we understand as justice. Even though justice is an abstract and indeterminate concept, it already articulates and restricts that ideal, reflecting the concerns and attitudes of our cultural moment. Through culture we divide and distribute our sense of the normative order into multiple values and virtues, of which justice is only one among many. Not all cultures will do this in the same way and so we should not expect that their moral language will share the same distinctions as ours. But because all of our moral discourse presupposes the idea of subjects and agents in a normative order, we can be intelligible to each other even if we do not always agree. Indeed, if we could not understand the speech and actions of others as presupposing a normative order with subjects and agents of some kind, it is likely that we would not even understand them as being rational agents.
Pragmatism and Historicism
My argument that different cultures articulate transcendent values in different ways brings me at last to a final objection-one that should be particularly important in a book whose vision of culture is largely historicist and whose view of knowledge is essentially pragmatist. The objection is that one cannot meaningfully speak of ideals or values that transcend cultures because the idea of transcendence is itself wholly peculiar to a particular cultural tradition of
discourse–the discourse of Western philosophical thought. The concept of transcendent ideals has a specific history and genealogy that stretches backward from the present day to Kant and to Plato. The idea of transcendence is itself merely one set of tools of understanding that have been developed at a certain point in history to solve particular kinds of problems. Thus, the argument goes, the view of human values as a series of articulations of a transcendent framework neglects the possibility that the very idea of a transcendent conception isitself one of the contingent artifacts of a particular culture. And once we recognize the contingency of the idea of “transcendence”-as a historically produced tool of understanding–we can no longer take seriously the notion that justice and truth are values that transcend all cultural traditions.
The response to this pragmatic objection is entirely pragmatic in spirit. Surely the idea of a transcendent value is a product of a particular cultural history. We can trace its development from Ancient Greece to the contemporary West, and so the particular shape it has taken is contingent in the sense of having been the product of memetic evolution. But it does not follow that the features of the human predicament expressed through this theoretical concept are themselves wholly contingent. Rather, I argue, the concept of “transcendent value” is the best way, given who we are and where we are now, to make sense of these features of human existence–our experience of justice as an inexhaustible demand, and our sense of the inadequacy of all attempts at capturing this value and making it determinate. The concept of transcendence is the most adequate way of describing this inadequacy.
As tools of understanding, all of our ideas are imperfect, and this holds true even for our ideas of perfection. The idea of transcendent values is itself merely an articulation of that which it purports to describe. As an articulation, it is surely subject to revision. And perhaps someday we will exchange the notion of transcendence for another that will be more successful. But this does not mean that the features of human life that our ideas attempt to express are themselves wholly contingent and wholly internal to our discourse. Our conceptions are revisable only because there is something against which we revise them.
Moreover, there is a curious sense in which even the pragmatist objection to transcendence must make use of transcendent ideals. The pragmatist objection is that there cannot really be transcendent ideals because of the historical emergence of the concept of transcendence. Yet such a claim seems to hold itself apart from its own pragmatist scruples. For the objection must surely apply to itself; it is made wholly from within the discourse of a particular culture-and therefore can hardly serve as a judgment about the thought of other cultures. Moreover, the pragmatist objection seems to present itself as an assertion about the way things “really are” that applies with equal force to claims about truth and justice made in other discourses from other cultures with other histories. It offers an impossibility theorem applying to all cultures from within a particular culture. It makes a transcendent claim about the impossibility of transcendence.
The pragmatist thus ends up in a curious reversal. What I have dubbed the pragmatist argument turns out not to be so pragmatic at all, for this argument wants to see behind the illusion of adequacy of a particular conception. It believes in the reality of this illusion and thus in the reality of the state of affairs that the illusion conceals. The pragmatist argument wants to insist that, despite the comfort that the notion of transcendent ideals might give us, they are products of a cultural moment. Hence they cannot describe what is really the case; they cannot truly apply to any other culture than our own.
Conversely, as I have suggested, the argument for transcendent ideals is more truly pragmatic in temperament. Given who we are and where we have come from, the language of transcendence is the best way to explain our ability to discuss questions of truth and justice with other cultures and other persons. It is the best way to understand the phenomenological demands of truth and justice. It is the best way to describe the relation between human values and the felt imperfections of this world. Moreover, transcendent concepts are implicated by many other beliefs about ourselves and our world that we would find it hard to jettison. In other words, the pragmatist argument for transcendent values is that one should accept these concepts and this way of talking because they work.
For my part, this response to the pragmatist objection is as conclusive as I need it to be. It is, as I like to say, good enough for the purpose at hand. I am happy to acknowledge that talk of transcendent ideals of truth and justice is a part of our cultural software that arose at a certain point in history to understand the nature of human action, ideological analysis, and moral discourse. Like other cultural software, it may be revised, sharpened, and even discarded in time. But as of now, I argue that this way of talking is the most adequate way of describing the human predicament. More than that a pragmatist surely cannot demand.
As for the historicist, I would go even further: A historicist conception of human culture and human values not only is consistent with the notion of transcendent ideals, it requires them. By “historicism” I do not mean a theory which holds that the content of substantive values is successively revealed to us through the progress of history. Rather, I refer to the view that people’s values are shaped by the historical moment in which they find themselves. Hence as the problems people are faced with change, so, too, do their responses. Historicism in this sense is the temporal counterpart of cultural relativism. The historicist wants us to understand how people in different times and places could have held such radically different views of the world and of human values. She wants us to grasp how it made sense for people to believe in things and hold values that we today find curious or even reprehensible.
Implicit in this project are two assumptions. First, the historicist may wish to present the past as strange and even alien to us, but she cannot present it as utterly unintelligible. Rather, to learn the lessons of historicism, we mustseek to understand the past in all of its strangeness and alterity. By definition, an unintelligible past can make no sense to us, and therefore we can learn nothing from it. Indeed, the discovery of an unintelligible past simply leads us to the entirely sensible conclusion that we have not done the work necessary to understand it. For otherwise we cannot know whether the unintelligibility lies in the past or is due to the clumsiness of our efforts to comprehend it. The irony of historicism is that it presupposes the basic intelligibility of the past in order that we may experience its strangeness and difference. Moreover, the strangeness and difference that it presents is only one step in a larger dialectical maneuver. For historicism also wishes to show us how what we find strange and alien made sense to the people who lived through these times.
This leads to the second assumption implicit in historicism. Morally speaking, the historicist does not want to let us off the hook. She wants to upset our smug assurance that the real reason why the views and values of the past seem ugly and ignorant is that they really are ugly and ignorant. Behind this project is usually a further, deeper agenda: the hope that we can take some critical distance from ourselves, that we can understand that people in future generations will find certain aspects of our practices as strange and abhorrent as we find those of the past. And this agenda in turn harbors two equal if opposite hopes: The first is that the present will come to see that it does not hold all of the answers to questions of value. The second is that if we can learn to be charitable to the strangeness of the past, we may merit an equal charity from the future.
The assumptions of historicism open a virtual dialogue between ourselves and the past, a dialogue that has much in common with the critical approach to ideology that I have offered in this book. To make the past intelligible to us, we must understand why the actions of previous generations made sense to them. We must attempt to see the truth and the justice in what they thought and what they said. And this project brings us inevitably back to the postulation of transcendent values of truth and justice.
Our recognition of historical changes in values requires ideals against which to understand this change. We can describe the history of people grappling with successive tools of understanding that reflect the periods of their emergence because we have a backdrop against which we can describe the limitations of this grappling. We are able to observe the parade of human conceptions passing through history, mutating and reversing themselves, because we have a language for describing their relative adequacy and inadequacy. In this way the concepts of historicism and transcendence are interdependent and intertwined. Each supports the other as its necessary adjunct and companion. We understand the transcendent as transcendent because we can see its articulations vary in history. The variance of history is coherent because we understand it against the background of the transcendent. The experience of historicism makes the concept of transcendence emergent; the concept of transcendence makes the language of historicism coherent.
 Here again we should note the potential distinction between subjects of justice, who can be treated unjustly, and agents of justice, who can act unjustly.
 Here again we should note the potential distinction between subjects of justice, who can be treated unjustly, and agents of justice, who can act unjustly.
 This is one reason, I think, why theorists like Bruce Ackerman and Jnrgen Habermas have turned to idealized forms of dialogue as means of explicating concepts of justice and truth. See Bruce Ackerman, Social Justice in the Liberal State (New Haven: Yale University Press, 1980); Bruce Ackerman, “Why Dialogue?” Journal of Philosophy86 (1989): 5-22; Jnrgen Habermas, Knowledge and Human Interests, Jeremy J. Shapiro, trans. (Boston: Beacon, 1971); Jnrgen Habermas, Legitimation Crisis, Thomas McCarthy, trans. (Boston: Beacon, 1975); Jnrgen Habermas,The Theory of Communicative Action, Thomas A. McCarthy, trans. (Boston: Beacon, vol. 1, 1984; vol. 2, 1987). Like other philosophical theories, dialogic theories of justice and truth are articulations of our transcendent ideals. As articulations, they presuppose the existence of transcendent ideals rather than produce them.
 Habermas has argued that certain ideal criteria are presupposed in communicative encounters; he has tried to capture them in his notion of an “ideal speech situation.” Jnrgen Habermas, “Discourse Ethics: Notes on Philosophical Justification,” in The Communicative Ethics Controversy, Seyla Benhabib and Fred Dallmayr, eds. (Cambridge: MIT Press, 1990), 60-110, at 85; Jnrgen Habermas, “Warheitstheorien,” in H. Fahrenbach, ed.,Festschrift fur W. Schultz (Pfullingen: Neske, 1973), 211-65; Habermas, Legitimation Crisis, 110. Thus Habermas argues that “participants in communication cannot avoid the presupposition that the structure of their communication … rules out all external or internal coercion other than the force of the better argument, and thereby also neutralizes all motives other than that of the cooperative search for truth.” Habermas, “Discourse Ethics,” 86.
My argument differs from Habermas’s in two important respects. First, Habermas relies on procedural and substantive criteria of an ideal speech situation instead of transcendent ideals of truth and justice. Indeed, he tries to derive ideals of factual and moral truth from the results of an ideal rational consensus. Later in this chapter I shall argue that a theory of ideal consensus presupposes these transcendent ideals and that an ideal speech situation is at best a heuristic for articulating them.
Second, I do not believe that when people engage in discourse they must presume that their discourse either does or can approximate the criteria of an ideal speech situation. I seriously doubt whether the notion of an ideal speech situation involving finite human beings with limited perspectives and historically generated cultural software is acoherent one. If the idea is incoherent, there is no reason to think that it is presumed in people’s speech acts.
 Sometimes we and the analysand will agree totally about what is right in a particular situation. Even so, we must still acknowledge that our views of what is just are revisable, incomplete, and imperfect. From a larger perspective what we think to be unambiguously just may be much more complicated and problematic. To acknowledge this we must still postulate a regulative ideal of justice against which our current judgments might be found wanting.
 Many philosophers have advanced various versions of moral relativism. See Gilbert Harman and Judith Jarvis Thompson, Moral Relativism and Moral Objectivity (Oxford: Blackwell, 1996); David B. Wong, Moral Relativity (Berkeley: University of California Press, 1984). But these theories often make some accommodation for principles of tolerance and for possibilities of moral dialogue between peoples of different cultures. For example, Wong argues that certain moral principles of tolerance apply to all agents even if they are not “universally justifiable to all agents” (189). So his theory is not strongly relativist in the sense that I discuss in the text.
Harman comes closer to that position. He argues that people can evaluate the actions of a person either relative to their own values or relative to the values of the other person. Similarly, we can either offer reasons for action that make sense from our perspective or offer reasons that would carry weight with the other person. But there is notranscultural notion of morality. It is true that many people believe in tolerating the views of others, but if a principle of tolerance is widespread, it is because from different perspectives many people have good reasons to abide by it; it is not because a principle of tolerance applies to all people generally.
Harman recognizes only one way of criticizing the views of others if what they did was right from their perspective: Harman argues that although we cannot say that it was wrong of a person to do an act that is consistent with that person’s values but not our own, we can properly say that it was wrong that the person did the act. We can say that what a person did was wrong in the same way that we can say that it was bad that a tiger mauled children, or that it was bad that an enemy took steps that worked against our interests. Thus, although it makes no sense to say that it was wrong of Hitler to exterminate Jews (assuming that Hilter had good reasons from within his own value system), we can say that it was wrong that Hitler exterminated Jews (49, 59-61). It is by no means clear how much work Harman thinks this distinction can do in dealing with problems of justice between cultures.
 Jean-Francois Lyotard and Jean-Loup Thebaud, Just Gaming (Minneapolis: University of Minnesota Press, 1985), 100.
 This argument is taken from J. M. Balkin, “Transcendental Deconstruction, Transcendent Justice,”Michigan Law Review 92 (1994): 1131-86, at 1175.
 The most famous ideal process theory is John Rawls’s theory of the original position. John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971).
 The best examples are Jdrgen Habermas’s and Bruce Ackerman’s work.
 Charles Sanders Peirce’s view of truth as the eventual consensus of a community of investigators can also be understood as an ideal process theory because the consensus is never the actual consensus of any given time but is always deferred. See Charles Sanders Peirce, “How to Make Our Ideas Clear,” in Philosophical Writings of Peirce,Justus Buchler, ed. (New York: Dover, 1955), 23-41, at 38.
 John Rawls’s theory of the original position is probably not an ideal dialogic theory in the sense I have described. Rawls imagines his participants in mutual discussion, but the principles of justice that emerge are not contingent upon the actual results of any sustained dialogue between people in the original position. Rawls claims that he can already show us the results that the participants will necessarily arrive at. His ideal procedure grounds justice not on dialogue but on rational decision.
Moreover, Rawls’s veil of ignorance produces agreement by stripping away so much information from the participants that they are for all practical purposes identical. They agree on the principle of maximin-maximizing benefits to the least advantaged–because they have insufficient information about themselves to gain a strategic advantage by refusing to agree. Thus only one rational person is really necessary in the original position, because all rational agents under the veil of ignorance will decide to do the same thing. As a result, not only can we not call the decision a result of dialogue, we cannot even call it an agreement. It is indistinguishable from a single individual’s decision of instrumental rationality. This is the ultimate consequence of Rawls’s attempt to convert questions of justice into questions of rational decisionmaking. See T. K. Seung, Intuition and Construction: The Foundation of Normative Theory (New Haven: Yale University Press, 1993), 17.
 Thus there is an analogy to the earlier criticism of Rawls. A truly ideal consensus under ideal conditions would require only one ideal participant, because each person in an ideal consensus would know everything (including the perspectives of all of the other parties) and would presumably have the same moral reactions to this knowledge. If the participants did not have the same moral reaction to the same information, it is not clear why they would agree.
 This argument is taken from Balkin, “Transcendental Deconstruction, Transcendent Justice,” 1139-40.