Anarchy, State, and Utopia.
by Robert Nozick.
Basic Books. 367 pp. $12.95.

For better or worse, academic philosophers are intent on deepening our discussion of political and moral issues. After a generation in which questions of logic and scientific method have preoccupied the dominant, analytic school of American philosophy, tenured professors are now interested in applying their newly honed conceptual tools to the predicament of man in society, and there appears to be an audience for their writings. The publication of John Rawls's cumbersome A Theory of Justice surprised Harvard University Press four years ago by becoming its best seller. The success of the book in the marketplace, as well as the high esteem in which Rawls is held by his colleagues and students, has spawned a cottage industry of criticism and commentary on Rawls's “ideal contractualism,” which has been misread in some quarters as a doctrine for levelers.¹

Now one of Rawls's Harvard colleagues, singled out for mention three times in the final page of acknowledgments to A Theory of Justice, has entered the lists with a self-labeled libertarian work that joins issue with Rawls at several points but which, as befits an attempt at original theory, approaches the problems of justice and the state from a fresh perspective. Though Robert Nozick devotes some fifty pages to a critique of Rawls, his ambitious book, Anarchy, State, and Utopia, is intended to stand on its own. That it succeeds in doing so has been almost unanimously acknowledged. From the day of its publication the book has been celebrated. One of Nozick's colleagues on the editorial board of Philosophy and Public Affairs, the organ of the new wave of public philosophers, promptly acclaimed the author in Harper's for his unsurpassed dialectical skill. Another public philosopher, in a long essay in the New York Review of Books, applauded his “razor-sharp analysis.” Tardier reviewers in the general press have tended to confirm these judgments. Though some have disputed Nozick's libertarian ideology and regretted his ignorance of social theory, history, and political thought, not even the most hostile has contested his technical proficiency. The National Book Award of 1975 in Religion and Philosophy has completed the popular certification of Nozick's book as an enduring contribution to American political philosophy. Far more impressive in professional circles is the evidence in the acknowledgments and on the dust-jacket that W. V. Quine, the dean of American logicians, submitted helpful written comments on the entire manuscript and found the final product “brilliant and important.”

Hence it behooves us to consider Anarchy, State, and Utopia as a work of philosophy should be considered—not by whether we like its conclusions, nor by whether we approve of its author's bibliography, but by whether it forces its arguments upon us. In philosophy, it is not the destination that counts, but how one gets there. If Nozick in fact reaches a libertarian philosophy by a logically compelling route then we must, like it or not, entertain his conclusions until they are superseded. If he somehow fails, then the respectful reception accorded Anarchy, State, and Utopia by his professional colleagues must lead us to reconsider the terms on which academic philosophers have lately been addressing themselves to public affairs.

_____________

In its main outline, Nozick's case is not difficult to follow, though it is presented inelegantly, with long digressions (e.g., on vegetarianism), and an unflagging prolixity in refuting the kind of objection likely to be posed by the fatuous freshman or journalist. First, Nozick argues (contra the anarchist) that the modern state, with its claim to monopolize the legitimate use of force, is a moral necessity. If it did not exist, something like it would arise, as if by an invisible hand, from the moral requirement of protecting property and respecting individual rights. Property holders would hire a protective agency to guard their holdings. The agency would monopolize a territory, extending its services to those who do not pay for them to preempt the field from independent enforcers. It would then begin to feel toward its clients the same sense of moral obligation that they feel toward each other. In time, a minimal state—the “night-watchman state” of classical liberal theory—would emerge without any formal social agreement constituting its powers.

Second, Nozick contends that a more extensive state than this cannot be justified. The only redistributive taxation which is logically required by the nature of the state is for the protection of property. Any other redistributive activity is unjustifiably coercive. To base arguments for state intervention on some prior social compact is implausible, for an invisible-hand model of the state enables him to assert that the pattern of state power is no more the product of conscious, collective agreement than the pattern of traffic on a given Saturday night. He constructs a utopia, rather like the Marxian “withering away of the state,” to show that the minimal state is not at all disagreeable. In it people do their own thing, a thousand flowers bloom, and there are no laws or taxes beyond those required to support the police.

It follows from Nozick's argument that the whole panoply of state activities as we know them—from conscription to child-labor laws to welfare benefits, public services, economic management, and rockets to the moon—are illegitimate. Such activities, Nozick argues, can be supported by voluntary subscription of free property-owners but not by taxation, which infringes on the individual's property rights, and hence his liberty.

It speaks well of the reception of this radically non-interventionist book that it has been civil and intellectual, a marked change from the often vituperative treatment that a few years ago greeted Edward C. Banfield's considerably more moderate The Unheavenly City. But it speaks less well for the professors who have given it lengthy reviews in popular journals that in the effort to be gentlemanly they have for the most part neglected the threshold question of whether Nozick's book deserves to be taken seriously as an important theoretical contribution.

Every writer in philosophy must subscribe to a few ground rules, and Nozick accepts the most basic of these: that one must articulate an understanding of what one is doing and then do it in a coherent way. He devotes the better part of his first two chapters to telling us what he thinks an adequate theory must accomplish in any realm of human conduct. Before dealing with his explicitly political arguments, therefore, it is essential to give his view of the proper method the prominence he gives it. Then we shall see not only what he attempts but how utterly he fails.

_____________

According to Nozick, we cannot hope to understand any realm of human conduct by dealing with it on its own terms. We must explain it in terms of “fundamental” concepts that make “no use of any of the notions of the realm” itself. “Only via such explanations can we explain and hence understand everything about a realm; the less our explanations use notions constituting what is to be explained, the more (ceteris paribus) we understand.” Especially illuminating, Nozick believes, are “invisible-hand” models which explain outcomes as generated in a patterned way by the mechanics of unintended consequences. Even if one knows an invisible-hand model to be based on error (“fact-defective”), on false generalization (“law-defective”), or on a fictitious linkage of events (“process-defective”), it is better to posit a hypothetical mechanism than to try to explain any realm of human conduct in terms of “desires, wants, beliefs, and so on.” “Even wildly false initial conditions will illuminate, sometimes very greatly,” for they at least give us a hypothetical pattern against which to view the otherwise seamless web of human intentions. Only such an approach, Nozick assures us, promises “a full understanding of the whole political realm.”

This view gains authority for Nozick by its fruitfulness in the exact sciences where it has the virtue of avoiding circular reasoning. If we reduce politics to something other than the political, we are, if nothing else, sure to avoid tautology. Nozick asserts the scope and power of reductionist methods sweepingly—the only gesture at judiciousness is a footnote excluding the invisible hands of witches, ghosts, and goblins—yet, as it turns out, he follows his strictures for less than half the book. He uses his reductionist techniques only to revise John Locke's derivation of the minimal or “night-watchman” state from the acts of isolated individuals in a state of nature. The methodological weakness of Locke, by Nozick's standards, is that he posits a conscious political act—the social contract—as a prelude to legitimate political community. In so doing he errs doubly, by explaining a political institution in terms of something political and by using a conscious social agreement to explain a phenomenon which, like all phenomena, can only be satisfactorily understood in terms of an invisible-hand mechanism.

_____________

As we have noted, Nozick corrects these Lockean lapses by showing that a “legitimate” minimal state might arise from the minute acts of individuals who pay a dominant protective association to safeguard their property. The six chapters devoted to this demonstration and the method behind it are a tour de force of microeconomic construction with only one serious flaw: the assumption that a Wackenhut enterprise in a state of nature will feel the same sense of obligation toward the consumers of its services that they as moral individuals feel toward one another. It is far more likely, as the example of the Sicilian Mafia suggests, that it would treat only its paying customers “morally,” and those in accordance with their most recent cash payments. But since the morality of the protective agency is clearly labeled as an assumption, its use affects only the plausibility, not the internal consistency, of Nozick's model, and does nothing to undercut his claim: he has provided a fact-defective, law-defective, process-defective, invisible-hand model of the state as emerging without either a social contract or an intelligent design.

But he abandons this style of presentation abruptly when he comes to discuss justice within the state. Here he slips unawares into a rival doctrine of explanation. To understand justice, he now avers, we must know what people mean and have meant by it. We must see it in terms of past desires, wants, beliefs, and so on. A division of property is just if it has come about in a just manner: “whatever arises from a just situation by just steps is itself just.” Specifically, Nozick proposes a three-pronged “entitlement theory” of justice. An individual is entitled to holdings if they were, in the first place, acquired in accordance with principles that seem just. (Nozick declines to elaborate whether these principles are eternal or whether it is sufficient that they were thought just at the time the property was acquired.) Second, property is justly held if it was transferred to an individual by just means by someone who acquired it justly. (He sees force and fraud as unjust means of transfer.) Third, a holding is just if it is acquired by a rectification of past injustices in acquisition and transfer. (He is aware that the question of who is entitled to carry out this rectification is highly problematical.)

The entitlement theory of justice, sketched in three short pages, is Nozick's beacon for the entire second part of the book. He believes it enables him to prove that no state more extensive than the minimal state can be justified. It is based on principles of reasoning that he calls “historical,” by which he presumably means that they make use of fact and precedent after the manner of a lawyer, not that they enable us to interpret and understand events after the manner of a historian. He uses his “historical” principles to criticize “structural,” “current-time-slice,” and “patterned” approaches, of which John Rawls's theory is his prime example. Such theories err in proposing a model for distributive justice in terms of some fundamental principle (equality, merit, etc.) other than past judgments about justice. Justice, like love, Nozick believes, can only be understood in terms of “what has actually occurred.” He also finds of insufficient historical accuracy the Marxist theory which explains the existing distribution of property in terms of the mechanism of economic exploitation. The more a theory of distributive justice uses notions of procedural justice, he now seems to be saying, the more (ceteris paribus) it enables us to understand.

Though he prides himself on his candor in acknowledging small weaknesses in his arguments, he seems genuinely innocent of the theoretical thrust of his two main points: the fact-defective habits of model-building he has blithely pressed upon us in the first part of his book are at the root of the ahistorical mentality he attacks in the second part. And the kinds of explanation he brands as tautological in discussing the state are precisely those he introduces in discussing justice. If the history of the acquisition and transfer of property is essential to understanding legitimate entitlements, why isn't the history of the state essential to understanding legitimate taxation and intervention? And if, per contra, we must use a reductionist method to lay bare the fundamental structure of the state, why not use a similar method to view justice within the state? Since Nozick proposes no adequate answer to these questions, one cannot accept his work as a coherent, enduring, or even plausible contribution to political philosophy. To the extent that either his “fundamental” theory of the state or his “historical” theory of justice cut deeply, they cut against each other.

_____________

Would that they were as effective in advancing the cause of liberty as they are in devastating each other. For there would be nothing more salutary than to establish on philosophical grounds a presumption against the unexamined accretion of state power and state spending. But the most disappointing thing about Nozick's book for the reader inclined toward limited government is that it also falls short of providing an adequate exercise in libertarian argumentation. Here looser standards apply. Instead of ascertaining whether Nozick's arguments hang together, we can ask whether, taken individually, they enable him to advance the cause of libertarianism.

That some aspect of the cause is worth advancing few would deny, for libertarianism taps deep strains in American political culture—the suspicion of bureaucracy, the safeguarding of individual rights, the respect for privacy and private property, and the sanctity of free expression. Libertarians of the Left tend to emphasize civil liberties, those of the Right free-market economics, but all share the enduring American faith in inalienable rights and a horror of coercive government. Most libertarians are comfortable with Lockean individualism in political thought and respect a mathematically formulated reductionism in philosophical analysis. (Locke himself accepted mathematical statements as the most precise, simple, and universal.) As a gifted decision theorist accustomed to writing more mathematico, Nozick holds out the prospect of providing analytic underpinnings for libertarian thought. If Nozick's arguments are not telling, then there is some deep deficiency in the habits of rigorous logical analysis in dealing with issues of moral and political philosophy.

There is, we must report, a deep deficiency in the habits of rigorous logical analysis in dealing with moral and political philosophy. The mark of the careful logician is to define precisely his terms and to build up a system from discrete and mutually exclusive categories. Within such a system statements can be judged either true or false, with no shadings or anomalies between these two poles of truth value. Unfortunately, the concepts of political and moral philosophy have precarious boundaries. They are less like separate pieces on a checker board than overlapping circles, or, better still, interpenetrating spheres of uncertain dimension. If one defines concepts like legitimacy and coercion or liberty and equality as mutually exclusive, one is led into elementary errors. Moreover, in politics and morals we deal not merely with different preferences of individuals but with wildly different conceptual frameworks. A proposition that will seem false, meaningless, or trivial from one point of view will seem a fundamental truth from another (for example, Nozick's own dogmatic assertion that “all individuals have rights” would be counted unprovable from a strictly empirical point of view). Hence a two-valued logical system, which attempts to distinguish unequivocally true propositions from false and to put all basic concepts in watertight compartments, will lead one not merely into an inability to see the nuance and texture of the moral life but also into a trained incapacity to weigh the conflicting claims that are the stuff of politics.

_____________

Modern mathematics is not without tools with which to represent the complex reasoning that prevails in practical life. But W. V. Quine, who has helped develop some of these tools, has asserted in his own Philosophy of Logic that real set theory and multivalued logic must be excluded on formal grounds from orthodox texts. Thus, if one limits oneself to the approved textbook methods of a Quine, one is “logically” unfitted for dealing philosophically with questions of morals and politics.

It was an appreciation of such difficulties that led Ludwig Wittgenstein in his formalist phase to advise careful logicians to remain silent on issues for which they had no adequate notation. And it explains why any serious attempt to deal philosophically with politics must involve either a revolt against logical formalism or the acceptance of the more demanding task of broadening the canons of formal discourse Nozick has attempted neither of these projects. Instead, he has simply extrapolated into the political realm the habits that have served him so well as a decision theorist.

A thoroughgoing libertarianism has the virtue of enabling him to stay closer to his training than almost any other political position, save perhaps thoroughgoing skepticism. The libertarian's belief in the inviolability of moral personality translates nicely into a view of the individual as a “circumscribed area in moral space.” Relations between individuals then become “boundary crossings.” Any attempt to aggregate individuals and treat them as members of “society” can be rejected as a fallacy of composition. And any intervention by society in the rights of individuals can be seen in both a moral and logical sense as untenable.

Up to a point such simplifications are useful. They result in an increased sensitivity to that realm in which individuals may be viewed as self-contained, in which trust, community, and solidarity between them are either impossible or irrelevant, and in which all civil relations must be formally specified. Since politics must deal with this realm, a political philosopher can go far by combining textbook logic with a doctrine of atomistic individualism.

But not quite far enough. For just as philosophically there is a gray area in which concepts inter-penetrate, there is also sociologically a gray area in which individuals share common beliefs, interact, and shape each other's personalities and preferences. One might argue that this realm is not relevant to political philosophy, which must confine itself to formal legal relations, but one cannot plausibly make this argument without setting one's initial terms to take account of what one is excluding. To neglect such niceties is not to provide a purer but a more vulnerable statement of the libertarian position.

_____________

Were these conceptual faults unique to Nozick, his book would hardly be worth mentioning. But the reception of the book suggests that it is part of a clearing of the throat in which others will present their positions in like manner. Two errors which Nozick makes are so widely shared that they stamp not only the character of his book but of much American intellectual discourse about politics: he rigorously uses legitimacy as a concept which excludes coercion, and, more loosely, treats liberty as a concept which excludes equality. The flower-child in each of us will share with him the first error and the jejune logician will be attracted to the second.

On the nature of coercion, Nozick is an acknowledged authority. His widely admired essay on the subject, first published in 1969, shows how rigorous logical analysis may properly be used to sharpen the philosopher's use of individual concepts. The essay, which he views as preparation for the maturer formulations of his current book, distinguishes the conditions of coercion from those of inducement and deterrence. Coercion, Nozick says, always involves a threat, which is understood as such by the person coerced and is intended to alter his conduct. Where offers rather than threats are used to induce the alteration in conduct we have manipulation of incentives without coercion. Where a combination of threats and offers is used the situation can be called partially coercive. Where the threat of force is present but is part of an impersonal system not directed at altering the conduct of a given agent, we have deterrence but not coercion. Nozick is so meticulous in establishing these and related distinctions that one could hardly expect him to address explicitly in his essay the more complex conditions under which law is coercive.

But this issue haunts him in his new book, much of which is a truly eccentric attempt to cast law into a coercion-free mold. Using his old terminology, we can see that if law could have the character of an impersonal system of rules-of-the-game it would “deter” without coercing. If, on the other hand, it were a direct exercise of collective will designed to alter the conduct of a specifiable class of individuals it would partake of coercion. To be legitimate—and Nozick is consistent in using legitimacy to exclude any prior element of coercion—law must somehow be purged of the exercise of willed negative incentives, a problem different from eliminating the exercise of arbitrary will. Nozick seems aware that his old terms are not quite equal to this. Much of what he was once able to call deterrents he now recognizes as legal “prohibitions” which fall into a category that he continues to view as at least potentially coercive. The new problem for him is to turn negative incentives into positive ones and thereby transmute potential coercion into inducement or some other morally neutral process. This he does by introducing a “principle of compensation” which he puts at the foundation of the legal system in opposition to the “principle of fairness” and other juridical doctrines conventionally used in discussing legal obligations. He leaves his quasi-economic principle sufficiently ill-defined to be insusceptible of refutations, though not of implicit contradictions contained in his own later arguments for procedural justice.

We can now understand why Nozick has taken such elaborate pains to replace Locke's derivation of the minimal state with one that excludes the social compact. A social compact makes the law an act of popular will and hence an ongoing exercise in coercion. By obeying the law one is submitting to the will of the people: the many coerce the few, the majority the minority. By contrast, if the law were limited only to what would come into being in any case as the unintended consequence of non-coercive acts, one could obey it, much as one submits to the laws of physics, without feeling the visible hand of coercion. Nozick's animus against the social contract leads him to underline this point with a deeply felt satiric passage, the most powerful piece of writing in the book, suggesting that the ability to cast a vote on one's taxes is barely distinguishable from slavery. In both cases the product of one's labor is expropriated by a higher will over which one has no real influence.

Now, unlike Nozick's views on philosophical method, this one provides the real unifying thread of the book, which is less a celebration of liberty than a trembling at governmental coercion. Yet the more one sympathizes with the desire to limit state power the more one will see wisdom in John Locke's approach. For, even if maintaining order is the sole duty of the state, there will remain the questions of who is to define a threat to order, the means by which to deal with this threat, and—what is essential to law enforcement—the means by which to instill a general sense of obligation to assist the authorities. The difference between a night-watchman state and a police state lies in how these questions are resolved. Locke's answer is now an American commonplace: he saw the state as based on specific, limited, conscious, and (in extremis) revocable grants of political authority. These are symbolized in a social compact which provides not so much a license for ongoing coercion as a standard for ongoing popular participation in state power, and for constitutional limitations on its exercise. Locke saw that for law to appear as admonitory and non-coercive rather than arbitrary and threatening it must, in Jefferson's paraphrase, derive “its just powers from the consent of the governed.” Better a government of laws established consciously and authoritatively than an unconstituted invisible-hand regime depending on some professor's postulate about the felt moral scruples of a monopolistic policing-services corporation.

A reflective libertarian will see Nozick's book not as an improvement but as a travesty on Locke's restrained constitutionalism. To posit a night-watchman state without a conscious agreement on the sources and limits of its authority is a prescription for order without law. Better to admit that legitimate state action presupposes an element of coercion, which must be constituted and limited, than to pursue a chimera of non-coercive government which leaves open the door for the well-known fascist excesses in the name of protecting property. In the end, disjoining coercion from legitimacy results in an inability to deal with the phenomenon of law, an anomaly from which Nozick escapes in his last chapter by envisioning a utopia in which there is no need for “adjudication.”

_____________

The second conceptual difficulty is not central to Nozick's world view but it does lead him down a wrong path. He accepts uncritically the widespread use in libertarian circles of liberty and equality as if they were mutually exclusive and conflicting concepts. This is one standby of partisan politics that it should be the aim of a libertarian philosopher to dispel. In many Western countries there is a self-styled party of liberty (called a party of greed, reaction, or privilege by its opponents) which sees itself as opposed to one or another party espousing equality (also called envy, mob rule, or bouleversement) . But as normally used the two concepts interpenetrate. If we agree that Nozick is correct in opening his book with the reaffirmation that all individuals have certain inalienable rights, then we have already accepted the principle of equality (equal rights) into the definition of political liberty. As a matter of practical politics this relation helps to explain why policies which extend liberty can simultaneously advance equality.

To see the two ideals as polarities leads one into a misdirection of polemical energies of the kind that occupies Nozick in the second part of his book. Here he undertakes to explode egalitarian and other arguments for redistribution of wealth and income. Egalitarian arguments for distributive justice, he thinks, are “generally acknowledged to be the most weighty and influential” for a more than minimal state. In this he is mistaken: the case for distributive justice has been and remains a good deal less weighty and influential in justifying statism than arguments under many other headings: war, order, nation-building, economic development, productivity, job security, cultural unity, public education, and modern sewage treatment.

In the English-speaking world the ideal of liberty itself has provided some of the most popular justifications for the redistribution and expropriation of property. Some, adapting an insight from Aristotle, have seen a large class of middling property owners as the best guarantor of a regime of liberty against mob tyranny, on the one hand, and oligarchy, on the other. They have advocated a middleward distribution of wealth and the political power flowing from it, as a means of preserving liberty by preventing class warfare between the rich and poor. The Founders of the American Republic were all aware that the balance of property was closely related to the balance of power and the prospects of liberty in a republic, and this theme has never been absent from American political thought or legislation.

It has also been in the name of liberty that other writers have advocated a minimal level of education, services, and material security. Adam Smith and John Stuart Mill, neither of them “statists” or “egalitarians,” urged compulsory universal education on the grounds that rights are meaningless if one leads a brutish life without the capacity to enjoy them. Many who argued for child-labor laws used similar arguments, and many current arguments for a minimum-income floor and “equality of opportunity” proceed in the area where liberty and equality overlap. It is not, of course, essential that Nozick feel bound by the works within the tradition in which he is writing, but it is disturbing that he writes as if arguments which derive from intuitions much more penetrating than his own were, like Locke's views on the social compact, of small account. By disposing delightfully of the more fatuous arguments of the levelers he believes that he has disposed of all the weightier justifications for a more than minimal state.

In one sense he is correct. While scoring some very telling points against John Rawls, he argues, convincingly, that the attempt to produce any specified distribution of property or income requires continuous state interference in people's liberties. His point applies across the board. The attempt to pursue any objective through the apparatus of government—whether to redistribute property or to protect it, whether to impose a specified pattern on the future or to preserve a vague set of values—involves continuous interference in individual liberty, some of it by manipulation, some by deterrence, some by outright coercion. The minimal state, therefore, cannot be defined in terms of any purpose save a dedication to the procedures of its own self-limitation. Nozick puts us on the track of seeing this, even though he is prevented from formulating it by the very habits of thought that he views as his entitlement to speak authoritatively on politics.

There are, of course, many pilpulistic virtues to this book that will make extracts from it pedagogically useful (notably his oftcited pages demonstrating why Wilt Chamberlain is entitled to a higher salary than his teammates). But however devoted one is either to popularizing philosophy in America or to giving libertarianism a fair hearing, one should not mistake the essential thrust of this work. Nozick has given us a vision of a libertarianism without law, order without consent, and political philosophy without coherence. He thus fails by his own ideals, and there are none higher: truth and liberty.

_____________

¹ The one book-length critique of Rawls, however, Brian Barry's The Liberal Theory of Justice, points out accurately that despite the book's egalitarian rhetoric it explicitly ranks the principle of liberty ahead of all others, so that libertarian conditions must be fully satisfied before any others can come into play.

_____________