Liberalism is the dominant contemporary political philosophy. As such, it naturally attracts criticism. One of the more interesting critical debates in the last decade has been that between liberalism and communitarianism. Communitarian writers have argued that liberalism cannot support-or, worse, that it actively weakens-the communal bonds upon which so much of our society relies. According to some of these writers, liberalism depends on a spurious model of the individual as an "unencumbered self" which exists prior to its personal or social commitments. Liberalism's neutrality between different conceptions of the good life allegedly deprives community institutions of the support they may need from the state, and devalues the importance of political life for society as a whole. Social policies based on these principles weaken or destroy the ties that unite people into communities. Ironically, the argument goes, such policies may even be self-defeating; for communities are instrumental to the liberal goal of producing conditions under which individuals can live together peacefully and cooperatively. Communities help people develop the habits and relationships which allow them to function in liberal society; in so doing, they provide the necessary environment for the development of true individuality. By weakening community, liberalism thus undermines the conditions which make its own success possible. 
Liberals have responded, sensibly enough, that liberalism not only depends on community but regards it as one of the most common and important features of civil society: so important, in fact, that the foremost liberal thinkers have generally assumed that people can be counted on to form communities spontaneously, without any need for state involvement. Also, communitarians ignore the historical circumstances that led to the development of liberal government. Early modern European history demonstrates clearly that the public enforcement of a conception of the good life, far from being unifying, can in fact be deeply divisive, leading to unrest and even civil war. State sponsorship of some communities typically means restricting or disadvantaging other communities, which is virtually guaranteed to be a source of strife. Neutrality with regard to the good life, claim liberals, is the best response to the unavoidable fact of social plurality. Finally, say liberals, communitarians often ignore or gloss over the deep difficulties of putting into practice their positive suggestions, particularly the suggestions that would favor group obligations or group rights at the expense of individual rights. After all, limiting or abolishing individual rights is not a painless or victimless course of action: some people will lose out in the process. And even aside from these objections from the point of view of the individual, government intervention in communities can be damaging to the communities themselves. It can lead to undesirable rigidity and resistance to change, and it can raise barriers to the creation of new (and possibly superior) kinds of social groupings.
These responses suggest that liberalism is not as vulnerable to communitarian critique as may have appeared at first. Still, the communitarian arguments, even if they fail to demonstrate the ultimate inadequacy of liberalism, may be useful as a guide to revising or supplementing liberal theory. Allen Buchanan and Will Kymlicka both propose that each side in this debate may have ignored some salient points from the other side. 
Despite centuries of liberal insistence on the importance of the distinction between society and the state, communitarians still seem to assume that what is properly social must become the province of the political. ... Despite centuries of communitarian insistence on the historically fragile and contingent nature of our culture ... liberals still tend to take the existence of a tolerant and diverse culture for granted, as something which naturally arises and sustains itself, the ongoing existence of which is therefore simply assumed in a theory of justice. Communitarians are perhaps a bit hasty to conclude that liberal neutrality is a problem and state support of communities is the solution. But they may be correct in thinking that political theory needs to take more seriously the role of community in public life, and that theorists should examine that role in more detail.
I believe that this attitude is correct. In this dissertation, I argue that a currently existing political theory can provide us with the conceptual resources necessary to incorporate the key insights of both liberalism and communitarianism.  Although the formulations of that political theory that have been made up to this point need some revision, the fundamental insights of that theory remain steady. The political philosophy I have in mind is anarchism.
Anarchism has never been widely understood. When people think of it, they are likely to think of terrorism, revolution, and chaos. But that is only if they do think of it, which is unusual. As a political philosophy it is relatively unexplored, and not regarded as a serious rival to liberalism, conservatism, or socialism. Admittedly, the view is not as obscure as it was in the years following the Second World War, when it was thought of even by its most prominent scholars as an interesting but obsolete historical relic. The New Left of the 1960s and 70s inspired a new interest in both the philosophy and the practice of anarchism. Academically, its influence continues in areas such as feminist philosophy.
But a modest upsurge in the 1960s and some scattered influence among political thinkers do not, by themselves, mean that anarchism is a distinctive political philosophy, or that it is worth taking seriously. A viable anarchist political philosophy should demonstrably have at least the three following features:
Some recent studies of anarchism have made similar arguments. But much contemporary work has taken a piecemeal approach to the subject. There have been several historical studies of the classical anarchist thinkers; at least one notable (if brief) theoretical defense of anarchism, by Robert Paul Wolff; decision-theoretic arguments for the possibility of a peaceful society not governed by a state; and a few practical studies of actual anarchically run societies. These approaches are not complete in themselves, and each of them raises questions and difficulties for the proponents of other approaches. Some of the principles defended by the classical anarchists have uncertain applicability in practice; Wolff's justification of anarchism is brief and flawed; decision-theoretic explanations may have little relation to political or social practice; and anthropological studies of extant anarchic societies may need considerable interpretation in order to provide a model for modern industrial society. A complete theory will need to bring together the disparate strands of anarchist scholarship, demonstrating a broad compatibility between these strands and plausibility for the theory as a whole.
This dissertation is an attempt to begin this unifying task. By incorporating features of these approaches, I demonstrate that anarchism is a coherent and distinctive political philosophy which can and should be put into practice. I provide a distinctive and plausible view of human life and agency, drawn largely from the work of the eighteenth-century anarchist William Godwin. Unlike most previous anarchist writers, however, I do not regard the abolition of the state as the defining feature of anarchism, nor even among its most important tenets. This conflicts with most existing understandings of what constitutes anarchism. To avoid confusion, I call my theory "quasi-anarchist," as it calls for a radical decentralization of power while retaining a role for state government.
I should remark that I do not intend to cast aspersions on the quality of recent writing on anarchism. On the contrary, the important work by Mark Philp, Michael Taylor, and others is indispensable to this project. These fine writers, whose work converged on the same subject, seemed to have little mutual communication; this is an attempt to bring together in mutual conversation some of the insights that have been developed recently.
The first step in setting out on this task will be to provide some historical background and preliminary definitions.
Anarchism is often briefly glossed as the doctrine that opposes state power and authority. But this brief definition only hints at the rich content of the theory. The classical anarchists shared common ground with a surprisingly wide variety of political views. Like classical liberals, they viewed the state as frequently burdensome and potentially oppressive of individual rights. Like socialists, including Marx, they opposed the exploitative capitalist economic system.  And like some conservatives, they considered small local communities the ideal locus of social and political life. These apparently eclectic views have led some commentators to regard anarchism as an ad hoc mingling of fashionable positions; or, worse, an inherently confused jumble because of the internal tensions between the incompatible goals of liberalism, socialism, and conservatism. 
As if this were not enough to render it undesirable as a political philosophy, anarchism also seemed fixated on an impractical and dangerous goal, the abolition of the state. Indeed, the attack upon state authority has generally been understood to be the single defining characteristic of anarchism, and the one from which most of its other positions could be derived. This critique of the state itself relied upon implausible assumptions about the malleability of human nature and people's ability to interact socially and politically with only reason to guide them.
If this popularly accepted view is correct, anarchism should be doomed by its impracticality and internal inconsistencies to, at best, an existence on the margins of social and political life. It would not be worth serious consideration alongside the mainstream political philosophies. But this view is inaccurate. As I will demonstrate in the following chapters, anarchism is grounded in a distinctive and original view of the values of society and individuality, and of the good life for humanity.  Its social and political recommendations have an internally compelling rationale, and are not merely ideas appropriated from different political theories. Furthermore, it is not inherently utopian; and its most important goals may be achievable within the confines of the state.
Modern political theories
Here are some basic definitions of the major modern political theories that
will occupy our attention: liberalism, libertarianism, and anarchism.
Of these theories, liberalism is probably the most resistant to being reduced to a simple definition. This is, in part, because of its long history and philosophical influence. The number of varieties and gradations of liberalism is enormous. For our purposes, then, it will probably be best to enumerate primarily the features that will receive the most discussion, particularly the ones that most strikingly distinguish liberalism from anarchism.
The essential characteristic of any form of liberalism is its dedication to liberty. This definition, of course, calls for explanation of just what is meant by 'liberty'. The most widespread interpretation focuses on individual liberty.  The state often has an equivocal status in modern liberalism. On the one hand, liberalism focuses on protecting individual liberty from encroachment by the state, by instituting legal protection for civil liberties such as freedom of speech, assembly, religion, and so on; and protection from the abuse of police power by requirements of habeas corpus, judicial review of legislative and executive decisions, etc. On the other hand, liberal theories often look favorably on the use of state power in the economy, to ensure that people will be able to take advantage of a range of significant options. (This is sometimes described as a principle of positive freedom. Negative freedom, by contrast, is the absence of restraints to one's behavior, without any guarantee that a significant range of options will exist or that one will be able to take advantage of any of those that do exist.)  Although liberalism traditionally supports the existence of a market-based economy, it recognizes the utility of various state actions that regulate or constrain the market: for example, redistributive methods of taxation to provide benefit to the poor.
Protecting individual liberties from state action is generally held to take strong precedence over state intervention. In the most important modern liberal theory, John Rawls's A Theory of Justice (1971), his principle of equal individual liberty is lexically prior to a second principle, which provides for regulation of social and economic inequalities. "This ordering means that a departure from the institutions of liberty required by the first principle cannot be justified by, or compensated for, by greater social and economic advantages." (Rawls 1971, 61)
Implicit in the discussion so far is the assumption that liberal theory applies to society in the context of a state; and that the state should conduct itself according to the rule of law, rather than (for example) by whim of the sovereign. Also usually implicit in modern versions of liberalism is support for democracy. Although a monarchy liberal in many of its features is not a theoretical impossibility, it is generally believed that one of the most important rights one can have is the right to a voice in selecting one's own government.
Another important element in this capsule summary of modern liberalism is the doctrine of state neutrality with regard to the good. This idea is not found in all versions of liberalism; chapter 5 discusses a perfectionist liberal theory proposed by Joseph Raz. But all versions of liberalism incorporate at least the notion that one of the primary reasons for the preference given to individual rights over state economic regulation is because the resulting individual freedom affords each person the opportunity to conduct "experiments of living": to seek, explore, and put into practice their own conception of the good.  This basic principle is sometimes interpreted as meaning that the state must remain neutral with regard to views of the good. Whether this neutralist interpretation is necessary for a state to be liberal will occupy some of our attention in chapter 5.
Finally, we may ask whether there is a distinctively liberal social theory. One of the most notable avenues of attack for rival theories, as we have seen earlier in this chapter, has been the claim that liberalism either lacks a coherent social theory; or that its social theory is de facto one of loosening or destroying traditional social structures by its preference for individual rights (rather than, for example, family or other group rights). Although this summary is not the right place to evaluate this debate, we can safely say that liberal social theory is largely laisser-faire. Under liberalism, people are free to enter into any social groupings they desire; and these groups are free to conduct their activities as they see fit, provided that they not infringe upon the basic rights of individuals, which are protected by the state. Liberal theory claims that civil society is the appropriate forum for debate over the good, for all the reasons mentioned above (pp. 1-3).
To summarize: liberalism maintains that an ideally just modern society is one within a state which operates by the rule of law. This state will be legally prohibited from interfering with basic individual liberties. Such a society will have a market economy subject to restraints that enhance the overall level of liberty in society; these restraints must not infringe upon basic individual liberties. The state's protection of liberties safeguards the opportunity for debate about the good to be carried out in civil society, a less coercive and divisive forum than the government. Some theories require that debate about the good must take place only in civil society, requiring state neutrality with regard to conceptions of the good.
The next theory of interest is minimal-state libertarianism. This theory is characterized by the principle of minimal state interference in the voluntary interactions of individuals. While modern liberalism permits economic restraints that enhance opportunities to take advantage of freedom, libertarianism requires a completely laisser-faire state. No state action is acceptable except for the minimum necessary to defend the nation from outside aggression, enforce freely adopted contracts, and punish people who commit infringements against anyone else's person or property. In capsule form: the state should provide military, police, and judiciary; these shall be used only to ensure that all and only voluntary transactions take place.
While some libertarian writers argue (or simply assume) that a condition of minimal state interference in choice will also be a condition of maximal opportunity for choice, this correlation is dubious. The aim of minimal interference in voluntary choice, if achieved, could produce a condition in which the actual opportunities for voluntary choice are less than those available in other political conditions. This is because some kinds of interference in negative liberty may be necessary to provide increased opportunities for choice in other contexts. The classic example of such an interference is state prohibition of slavery, even in the hypothetical case where slavery is voluntarily entered into. This state action, although it clearly prevents one kind of voluntary choice, insures that people will not find themselves in the position of having contracted away en masse their future opportunities to make voluntary choices. 
Minimal-state libertarianism is based on the principle of voluntary interaction. But even a minimal state necessarily involves some involuntary interactions; if one takes the principle of voluntarism seriously, it may seem desirable to seek a way to do without even those few involuntary relations. This brings us to the first of the two broad categories of modern anarchism. I will refer to this category as 'right-wing' or 'capitalist' anarchism. The best known writer in this field is Murray Rothbard.  Capitalist anarchism, like statist libertarianism, is based upon an ideal of negative liberty.
Right-wing anarchists believe that the functions currently in the hands of the state could be adequately (or better) provided by private businesses. This would require the parceling-out of some of the authority currently in the hands of the state to the businesses which would provide public services. Some, but not all: although in the right-wing anarchist picture the state would be entirely dismantled, not all of its authority would be placed into the hands of private organizations. Since right-wing anarchists are dedicated to the libertarian ideal of a society based entirely on voluntary relations, some of the state's powers, such as eminent domain or military conscription, would be done away with entirely. This view is discussed in more detail in chapter 4.
The second category of anarchism, 'left-wing' or 'communal' anarchism, presents a striking contrast to the right-wing view. The connections between this category and the preceding one are minimal, based almost entirely on a similar antagonism to state rule. Aside from this shared enemy, the two views have little in common, as they are based on profoundly different views of the good life. As we shall see in chapters 2 and 3, communal anarchism is grounded in an ideal of positive liberty, not negative liberty; and its goal is the maximization of opportunities for autonomous choice, not the minimization of state interference in choice.
Communal anarchism traditionally opposes not only state authority, but all forms of oppressive authority, including the authority held by the owners and managers of businesses.  During the late nineteenth century, its time of greatest influence, leftist anarchism was known as much for its opposition to capitalism as for its opposition to statism. The right-wing proposal of parceling-out of state power to corporations is thus unacceptable to the communal anarchist, for it does nothing to solve the problem of unequal power relations between employer and worker. Rather, communal anarchists have generally adopted an ideal of a society in which ownership and social power are decentralized.
This remark may call for some comment before proceeding. Libertarians and right-wing anarchists sometimes assert that a libertarian society will tend to be a decentralized society characterized by relations of equality rather than dominance. Liberals and left anarchists wonder whether this amounts to lip service to the goal of egalitarianism. Libertarian arguments in favor of voluntary slavery (see chapter 5) clearly seem to favor pure negative liberty over equality. Many libertarian proposals that do not amount to outright slavery also would erode the cause of equality. To take one example, libertarians regard discrimination in housing as within the protected sphere of individual liberty. While discrimination may be objectionable, they assert, the best protection for individual rights is the free market.
A landlord might "discriminate," for example, by insisting, as George Pullman did in his "company town" in Illinois in the late nineteenth century, that all his tenants appear at all times dressed in jacket and tie; he might do so, but it is doubtful that many tenants would elect to move into or remain in such a building or development and the landlord would suffer severe losses. (Rothbard 1978, 207)Even if we suppose that Rothbard's economic prediction is correct (it may well not be), he fails to consider the possibility that a very rich landlord might be willing to suffer economic losses, regarding them as part of the normal expenses incurred in the quest to control other people's lives.
Libertarianism, both statist and anarchist, has considerable difficulty dealing with the likelihood that the rules of the libertarian society would allow some people to exert much more power over others than they can under a liberal government. There are three likely responses to this worry: first, that a libertarian society, with its respect for individualism, would not be the sort of environment in which controlling urges would be prone to develop or flourish. Second, that the free market would reliably provide alternatives to indenture and to restrictive labor and housing contracts. Or, third, that these results are actually desirable, given the basic presuppositions of libertarianism. The first two claims seem dubious. The third essentially concedes my basic claim: that the libertarian value system places insufficient stress on the need for equal social relations and decentralization of power.
To return to our outline of communal anarchism: communal anarchists do not seek a society of individuals isolated in their autonomy. They have had much to say about the best ways for people to work cooperatively in the absence of relations of authority and obedience. They place great emphasis on the need for collective decision-making, both because this is practically necessary and because human beings can only flourish in the context of society. They claim that the best way to produce efficient and fulfilling cooperative effort is to insure that everyone involved in that effort has a hand in the major decisions regarding it, in a context of equality and free expression. This approach thus would seek to limit or negate the power of rulers and owners and place power instead in the hands of relatively small, egalitarian communities.
This dissertation will develop a left, communal anarchist position. Furthermore, it will do so while rejecting the apparently central criterion of anarchism: the abolition of the state. My theory will argue for some role for state authority. This last fact will raise the question whether the view deserves the name 'anarchism' at all, even if it is prefaced by 'quasi-'. That name may mislead; but I wish to retain it for two reasons. One is simple convenience: 'quasi-anarchism' is considerably shorter and less pompous than something like 'individualist decentralized communitarianism'.
Convenience, of course, is probably not a sufficient reason to keep the label. My second reason is that, as I will show, the theory is firmly grounded in the anarchist tradition. Although few anarchists have gone so far as to defend the state, most have resigned themselves to its short-term continued existence, and many have concentrated on the best form of community life even within the state. This is particularly true of the work of William Godwin, my main source and inspiration. My theory follows in this tradition. I hope, then, that keeping the name of anarchism, suitably qualified as 'quasi-anarchism', will inform more than it misleads.
The classical heritage: William Godwin
Godwin developed the first distinctively anarchist political theory, one
which contains virtually all the elements found in the succeeding works of
Proudhon, Bakunin, and Kropotkin. But Godwin's theory was not only the first;
it was also the most philosophically profound and adept.  He presented a detailed and sophisticated view of moral
ideals and the good life which provides a stable grounding for the details of
the theory. And although his reputation as a utopian is well-deserved, the
most notoriously unrealistic features of his work are rarely central to the
I have chosen to adapt Godwin's theory in this dissertation precisely because it provides a detailed theory of value and the good life. Current anarchist philosophy typically lacks this feature. This weakens the force with which it can reasonably respond to the other major political philosophies, for leaving one's underlying value theory unarticulated will tend to limit the scope of one's arguments to matters of strategy or policy. And although Godwin's theory has been extant for just over two hundred years, most political theorists, assuming they have heard of Godwin at all, do not have a clear understanding of his view or of its consequences.
Before I proceed to the exposition of Godwin's views, I will examine two recent approaches to anarchism. That discussion will illustrate how a neo-Godwinian approach can address the most crucial missing elements in these recent anarchist doctrines.
Two recent approaches
Two important recent books, Robert Paul Wolff's In Defense of Anarchism
(1976) and Michael Taylor's Community, Anarchy, and Liberty (1982) exhibit
roughly complementary approaches. Wolff's argument tries to show, on the
basis of a few simple and widely accepted premises, that anarchism is
desirable-indeed, that it is the only acceptable political system. Wolff
makes some salient points, but in the end his argument is moral rather than
political; it fails to justify the full extent of anarchist social and
Michael Taylor's work, on the other hand, starts with the assumption that a stateless society is desirable, and argues at length that a just, peaceful stateless society is both possible and consonant with the pre-eminent value of liberty. His argument discusses the nature of actual stateless societies, providing decision-theoretical analysis and considerable empirical information to convince the reader of the feasibility of an anarchic society.
Taylor's important contribution is lacking in one crucial respect. It explicitly does not seek to justify the basic assumption that statelessness is desirable. The focus of his work, and of Community, Anarchy and Liberty in particular, is on the necessary conditions for successful anarchy, without providing a deep justification of that goal. This is not a crippling flaw; after all, one has to start somewhere, and earlier anarchist writers provide much argumentation that Taylor can legitimately take as given. Still, that tightly defined aim threatens to limit the utility of Taylor's conclusions if we find reason to doubt the ultimate worth of the goal of complete statelessness.
The remainder of this chapter examines both of these theories in more detail.
Robert Paul Wolff
Wolff's In Defense of Anarchism argues for what might be called 'moral
anarchism'. Wolff's claim is that the nature of moral autonomy forbids
accepting any authority as binding. His argument is almost this brief: to
retain one's moral autonomy, one must always reserve the right to decide for
oneself on any moral issue. But one cannot both be autonomous and accept
someone else's authority; for to accept another's authority over one's
actions is heteronomy. He therefore concludes that only anarchism is fully
compatible with individual moral autonomy.
If all men have a continuing obligation to achieve the highest degree of autonomy possible, then there would appear to be no state whose subjects have a moral obligation to obey its commands. Hence, the concept of a de jure legitimate state would appear to be vacuous, and philosophical anarchism would seem to be the only reasonable political belief for an enlightened man. (Wolff 1976, 19)Wolff's argument is concerned with the relation between autonomy and authority. To be autonomous is to be "self-regulating," reserving to oneself the power of making important decisions about belief and conduct. Ever since Kant placed it at the center of his moral theory, autonomy has had an important role in most philosophical conceptions of moral agency: it is a necessary feature of human maturity and dignity. To lack autonomy is to fall short of being a fully realized person. Crucially, Wolff stresses the role of autonomy as obligation. To maintain my status as a moral agent, I have the duty to reason autonomously and, as much as possible, act on the basis of that autonomous reasoning. 
Although authority is a familiar idea, it has been less well examined than autonomy. Finding an acceptable conception of authority will occupy a great deal of chapter 5. For the present, though, let us work with Wolff's definition:
Authority is the right to command, and correlatively, the right to be obeyed. It must be distinguished from power, which is the ability to compel compliance, either through the use or the threat of force. (1976, 4)Wolff's thesis is that autonomy and authority are incompatible. If I am truly autonomous, I cannot accept authority; that would be tantamount to abandoning my paramount duty of self-regulation.
This point may be clearer if we contrast it to the case of power, which is not similarly incompatible with the duty of autonomy (as Wolff understands it). Admittedly, autonomy and power may come into practical conflict. If Don overpowers me and forcibly seizes my lunch money, my ability to control my own conduct is impaired. Nonetheless, Don has not deprived me of my moral autonomy, as Wolff conceives it.  My will has not been controlled; he has interfered only with my ability to act according to the dictates of my will. Even while I am being overpowered, I can still carry out my duty to reflect upon the best use of the money. When not under direct coercion I can plan how to avoid future coercion, and so on. Indeed, boldly retaining one's independent will in the face of overwhelming force is a Romantic cliche: "My head is bloody, but unbowed." 
So although the powerful may seek to crush autonomous thought, the existence of power poses no logical threat to the carrying out of the duty of autonomous reasoning. But Wolff claims that authority does pose a logical threat to autonomy. For in Wolff's definition, authority is more than power: it is the right to be obeyed.  On its face, there does seem to be a conflict between the individual's duty to follow the dictates of her own reason and the authority's right to determine what that individual shall do. While Wolff's analysis of the problem does not search deeply enough to find a sensible way of reconciling the difficulty, he does seem to be correct that there is a genuine difficulty confronting us.
Furthermore, it is not a trivial difficulty that can be disposed of quickly and tidily. Wolff himself deals with several unsuccessful objections. For example, he points out that he is not making the implausible claim that one ought never to do what authority demands, that one's reactions to authority should always be contrary and rebellious. Quite the opposite: authorities may demand actions that are also called for by our own rational reflection. This does not count against Wolff's main point at all. For if I follow someone's directives because I am independently persuaded of the sensibility of that line of action, I can hardly be said to be obeying authority.
If that is the case, then I am not, strictly speaking, obeying a command, but rather acknowledging the force of an argument or the rightness of a prescription. ... The person himself has no authority-or, to be more precise, my complying with his command does not constitute an acknowledgment on my part of any such authority. Thus authority resides in persons; they possess it-if indeed they do at all-by virtue of who they are and not by virtue of what they command. My duty to obey is a duty owed to them, not to the moral law or to the beneficiaries of the actions I may be commanded to perform. (6)Political authority must amount to more than just moral exhortation; for exhortation is always possible, even when there is no political structure and hence no political authority. Wolff's definition requires that political authority imply the right to command, which is something more than the opportunity to admonish. Can an autonomous person recognize such a right?
For the autonomous man, there is no such thing, strictly speaking, as a command. If someone in my environment is issuing what are intended as commands, and if he or others expect those commands to be obeyed, that fact will be taken account of in my deliberations. I may decide that I ought to do what that person is commanding me to do, and it may even be that his issuing the command is the factor in the situation which makes it desirable for me to do so. ...But insofar as I make such a decision, I am not obeying his command; that is, I am not acknowledging him as having authority over me. (15-16)Wolff concludes that autonomy means taking commands from only one authoritative source, oneself, and rejecting all attempts to override one's autonomy. 
Since rejecting all authority means rejecting state authority, apparently being an autonomous agent implies anarchism: or, as Wolff puts it in the passage quoted above (p. 18), "philosophical anarchism would seem to be the only reasonable political belief for an enlightened man." (Wolff 1976, 19)
Wolff's brief book provoked a remarkably large number of responses, virtually a cottage industry for a few years, including a book of equal length (Jeffrey Reiman's In Defense of Political Philosophy, 1972).  Most of these responses assert, correctly, that Wolff does not prove his point. The overridingness of moral autonomy does not automatically establish the need for political anarchism. But Wolff's error may put us on the right track. A successful argument for political anarchism will depend on appropriately valuing individual autonomy, though not in quite the way that Wolff thinks.
If the desirability of anarchism can be established at all, it will be based at least partly upon the overriding importance of individual autonomy. But Wolff's attempt to ground anarchism on each person's moral imperative to be autonomous is insufficient to the task. Since Wolff's definition of autonomy is expressed in terms of each individual's moral duty, it fails to provide a clear prescription for society as a whole. More specifically, it fails to establish anarchism in the following three ways:
Claim i. Many commentators, particularly Reiman, have noted that one can carry out one's moral duty of autonomy under any political circumstances, including the most authoritarian tyranny. The requirement of autonomy as Wolff construes it does not require more than that I deliberate fully on my duties, and carry out my autonomous decisions to the best of my ability. Although my capacity to carry out some of the decisions I make may be infringed upon by a restrictive government, I can nonetheless still give myself the laws which I would live by, if I were free to do so. The existence of a hostile government does not relieve me of this responsibility. Indeed, I may find that one of the duties prescribed by my autonomous will is to oppose the government which is tyrannically interfering with my deliberation and the carrying out of my other duties.  As Reiman says,
Moral autonomy entails making the final decisions about what one should do. Political autonomy entails having the liberty to act upon the decision one has made. A defense of political autonomy can be an argument against coercion, against the authority of the state, but a defense of moral autonomy cannot-because one is not more morally autonomous when one is less coerced. I am not less able to make the final decision about what I should do when I can do nothing. (Reiman 1972, xxiv)The duty of individual autonomy seems to be constant across one's political surroundings, and does not necessarily imply the goodness of any political system. It would take additional principles to determine which system is the best.
Perhaps Wolff could respond as follows. Admittedly, the duty of autonomy binds each person under any political condition. But this ignores the question of what political condition is most compatible with the duty to be autonomous. If we take the "view of the universe," the fact that each person has an overriding duty to think and act autonomously might lead us to think that the proper political system is the one which interferes least with the ability of each citizen to carry out that duty. For it seems odd for an autonomous individual to support a government which will inevitably tend to interfere with that same autonomy.
Claim ii. The argument from the "view of the universe" still does not establish anarchism as the preferred political system. For Wolff's theory of autonomy treats it only as a duty which applies to each individual; and the results of an individual's autonomous reflection may well conflict with the neutral and noncommittal perspective of "the universe". For example, my duty to think and (if possible) act autonomously, as Wolff states it, may not automatically result in the conclusion that I should respect anyone else's duty of autonomy. My conscientious reasoning might convince me that I should override others' autonomous wills, perhaps even everyone else's autonomous wills. If I decide upon careful reflection that my moral duty is to enforce my will upon others, then apparently I have fulfilled my duty to be autonomous and need not worry about fulfilling it further. Wolff's minimally described duty of autonomy thus seems, on its face, to be potentially compatible with a considerable degree of personal or political tyranny; this is hardly an anarchist stance. 
This is not to say that an autonomous deliberator is inherently likely, simply in virtue of her autonomy, to view her duty as requiring her to become a tyrant or to disregard other people's duties to be autonomous. Someone who regards personal autonomy as a crucial personal goal will probably have reason to regard others' autonomy as a worthy end, and an important factor in determining her own duty. As likely as that is, though, it is not guaranteed by the Wolffian requirement to be autonomous. Furthermore, even if she recognizes that others have a duty of autonomy, that fact may be just one factor among many in her own autonomous deliberation. We have not established that she has any reason, in that deliberation, to treat it as overriding any other factors. If the duty of autonomy is to be used to demonstrate the desirability of anarchism, we must show that one ought to put a substantially high weight on everyone's, not just one's own, autonomy.
Finding a way to do this is not a trivial matter. Here is an example of a way that does not work. Suppose we try to extend to each person the principle of absolute non-interference in autonomy. This would produce something like the following principle: act upon your autonomous deliberations in all circumstances except those in which you would be infringing in any way upon any other person's autonomous deliberations. While this bears superficial resemblance to the thesis I will ultimately be defending, its stress on the overridingness of autonomy renders it a blunt instrument when applied as a social principle. At the very least, we can see that it would involve serious problems of practical coordination. Absent a highly articulated procedure for deciding which situations fall within my proper domain of action, I would face serious difficulty when deciding upon any action that might infringe, however slightly, upon anyone else's autonomy. Furthermore, we can see that this principle has great potential for abuse: people who obey it will be preyed upon by those who do not. 
This extreme principle is not consistent with any existing anarchist theory or historical anarchist practice.  Every society, whether under state government or not, has recognized the occasional need to interfere with some people's behavior-including, at times, people who sincerely believe that they are acting for the best. Wolff, of course, did not propose the principle and would almost certainly reject it. I have mentioned it, at the risk of attacking a straw man, in order to show the limitations of arguing for anarchism on the basis of autonomy conceived solely as a duty.
Since Wolff's argument stands or falls on the basis of his conception of autonomy as a duty, it thus seems that his account ultimately fails. Nonetheless, as I have remarked, it seems correct that an argument for anarchism can proceed most fruitfully by appeal to some notion of autonomy. An appropriate definition of autonomy will need to avoid the extremes we have mentioned. On the one hand, we will need to find a principle that places considerable weight on every mature person's ability to deliberate autonomously and to act (within reason) upon the results of that deliberation. For as we saw, my imperative to protect my own autonomy does not necessarily imply an imperative to protect everyone else's autonomy. On the other hand, the principle should not lead us to think of other people's autonomy as an absolute boundary which allows no infringement. If someone else's autonomy is a goal worth pursuing, it may be that the best way to maximize that person's overall autonomy is to infringe on it on some particular occasion. Restricting ourselves from ever interfering with another's will would sometimes thwart our intention to respect and promote that person's autonomy throughout his or her whole life.
Chapters 2, 3, and 5 carry on the search for a suitable principle of autonomy. Also, as I mentioned earlier, in chapter 5 I will argue for claim iii, which asserts that obedience to authority may sometimes be the best way to fulfill one's duties, including the duty of autonomy. Before we consider these matters, however, let us examine another of the significant recent discussions of anarchism: an examination of the ways of life of actual anarchic societies, accompanied by a decision-theoretic model explaining how this is possible. This is the work of Michael Taylor, in Anarchy and Cooperation (1976), The Possibility of Cooperation (1987), and Community, Anarchy, and Liberty (1982).
Taylor's early work in Anarchy and Cooperation (which was later heavily
revised and re-issued as The Possibility of Cooperation) addresses the
well-known problem of public goods. A good is said to be public if its
enjoyment is non-excludable: that is, if the good, once produced, is
necessarily available to all
or most of the public. Classic examples of public goods are national defense
and clean air. One cannot produce these goods and reserve their enjoyment to
oneself.  The problem of public goods, sometimes also
called the free rider problem, arises because individuals have an economic
incentive to avoid paying the cost of producing the good, since the enjoyment
of those goods cannot be restricted. If other people are already producing
the good, then each individual has an incentive to "free-ride" on their
efforts. If no one else is producing the good, then there is no incentive, or
indeed a negative incentive, to offer up one's own share of effort. Solitary
effort cannot be expected to benefit the individual: indeed, there is little
likelihood of breaking even on that transaction, since everyone else has an
incentive to free-ride on the amount of the good one produces, without
The standard solution to this problem is the use of coercion. Most modern commentators assume that this coercion must be applied by the state. But Taylor denies that state action is necessary for the production of at least one public good: namely, public peace and order. Anarchy and Cooperation and The Possibility of Cooperation use decision theory to demonstrate, first, that a high level of concern for others can be sufficient incentive to provide one's share of the good; and, second, that when concern for others is not present, social structures other than the state can provide the coercive means to insure compliance in production of the good.
The notion of an alternative means of preventing free riders, which is explored primarily in Community, Anarchy and Liberty, will occupy most of my attention. The main conclusions of this book are, in my paraphrase:
There is considerable evidence that social order is possible without the state. Taylor points out that recent economic and anthropological studies suggest that people in pre-agricultural societies (including some societies that existed well into the twentieth century) were typically healthy, had relatively long life spans compared to the average throughout most of recorded history,  and worked less than the average modern worker.  Deaths from injury (including violent injury) were, as far as we can tell, relatively uncommon.
Taylor argues that these studies provide evidence of the importance of community, which is a prerequisite for social order in the absence of government. The three essential characteristics of community, in his definition, are
The tools allow community members to present each other with what Taylor calls a "throffer": a condition that combines the motivational effects of threat and offer.  Favored (socially responsible) behavior produces results that are better than one could expect if one were left entirely to one's own devices. Socially destructive behavior, on the other hand, results in payback, leaving one worse off.
The offer inherent in much of community life is the prospect of generalized reciprocity. Community members have considerable motivation to share food and labor with their neighbors, in the reasonable hope that eventually, though usually not immediately, some return kindness will be forthcoming. The threat consists in a variety of social responses that one's neighbors can undertake in order to make one's life worse. Some examples of these responses are gossip and ridicule; rituals of public denunciation and shaming; and, ultimately, threat of ostracism and exile.
Since most anarchic societies up to the present have been pre-agricultural, individuals were highly dependent upon neighbors for their subsistence. So while sharing food with neighbors was more of a sacrifice than it would be for the modern middle class family, the prospect of receiving food in return was a powerful motivator. Correspondingly, attacks upon one's public reputation and the threat of ostracism involved more than emotional deprivation. Staying in public favor was often a matter of survival. Under these circumstances, it is not surprising that community-based societies were able to keep the peace without military or police forces.
Taylor also argues that the existence of community, with its opportunities for cooperative effort, enhances the individual's ability to do important tasks without placing serious limitations on liberty. This remains true even if one's fellow community members place intrusive and restrictive conditions upon their cooperation. This argument relies on the claim that one is still free to do the things one would have done in the absence of community, provided one is willing to accept the mode of living one would have had in the absence of community (which may include starving). In other words, such communities may not place physical barriers in the way of community members who wish to pursue unpopular goals, but they can make such pursuit practically infeasible by means of economic or emotional blackmail.
Taylor's analysis is very important for the future of anarchist philosophy. It demonstrates how some societies have functioned in the absence of the state, and it suggests mechanisms that may be useful in envisioning ways to do without the state in the future. But it is by his own admission (and intention) incomplete; it assumes in advance the desirability of doing without the state. It does not explore the underlying reasons why such a condition should be thought worth having.
The small, stable community which is at the center of my argument is nowhere in this book assumed to be good or right in any sense; it is nowhere defended as an ideal, or viewed as a happy and continuously harmonious place, free from conflict or from constraint and coercion. The contention is simply that community is necessary-if people are to live without the state. (Taylor 1982, 3)This and other features of Taylor's approach leave unanswered some issues of primary importance. First, Taylor glosses over the most considerable difficulty his position faces: namely, the problem of intercommunity relations. Second, he fails to discuss the profound barriers to using the pre-agricultural model of intracommunity social control in a market-oriented, technologically advanced society. Third, Taylor's treatment provides us with little basis for preferring one form of community organization over another. Finally, his argument assumes the desirability of doing without the state without explaining why that goal is fundamentally worthwhile. While this is not necessarily a fatal flaw, it means that Taylor fails to consider the possibility that the ideals which provide the best underpinning for an anarchist position may also be compatible with a limited role for the state in promoting social order and justice. 
Here, in a bit more detail, is my analysis of these difficulties. The first and second of these problems are addressed in detail in chapter 4, so my remarks about them here will be brief.
1. When members of one community are in conflict with members of a different community, the disputants have little (if any) recourse to the non-violent means of maintaining order which are available within community walls. Non-violent means depend on connectedness of some sort: kinship, frequent social interaction, or economic interdependence. If the feuding individuals are not connected by such ties, then the inhibitions against violence are greatly reduced. Fear of retaliation in kind remains a powerful disincentive, but this varies with the likelihood of retaliation.
Taylor realizes that this is a problem:
We have no grounds for believing that growing up and living in a community necessarily engenders a tolerant, pacific and cooperative disposition towards outsiders. It is true that many primitive anarchic communities lived at peace with their neighborsx but many did not, and the world is a great deal more crowded now. (Taylor 1982, 167)His response, however, is brief and unsatisfying.
I have not seen a plausible solution to this problem of inter-communal relations in the literature of anarchism or anywhere else, and I do not have a solution which I find persuasive myself. But it should be remembered that the problem is altogether less severe than that of the relations between nations, for there is the fundamental difference that the communities we are talking about would be stateless as well as small, so that the potential for inter-communal aggression and exploitation would be limited. In view of this, the goal of a radically decentralized world of small communities is not rendered less attractive than the current state of affairs solely in virtue of the lack of a solution to the problem of inter-communal relations. (Op. cit., 168)One might (equally briefly) respond that a social situation marginally better than the present may still not be the best workable system we can imagine; comparisons to the current state of affairs may not quite meet our expectations for a political theory. What is more, the communities we discuss might be small and stateless, but there is no guarantee that they would be unarmed or peaceable, or that they would stay small or stateless. Gang rule is not unknown in history, and it is particularly undesirable.
Some anarchists have suggested ways to deal with such a situation without invoking the state. In chapter 4, I argue that these solutions are generally unacceptable and that the solution that makes sense is simply to allow for the existence of a state. Luckily, Taylor's observations will still have considerable application in such a theory.
2. In addition to the problems of intercommunity relations, there is a corresponding problem of intracommunity relations. If the role of the community is to provide order in the absence of the state, then we must be sure that this instrument can carry out its primary function. I argue in chapter 4 that modern circumstances render unavailable the most powerful persuasive tools of the pre-agricultural communities. Ostracism and expulsion no longer mean deprivation and possible death for the exile. Even if they did, there would be no guarantee that these extreme measures would be used in a just manner. Judicial restraints such as due process are important precisely because courts can mete out severe punishments: restraints help insure that they do not assign these punishments arbitrarily or vindictively. We would have good reason to fear communities which were able to decide on questions of life and death in the absence of the procedural guarantees we find in good judicial systems.
Thus communities, if they are to function properly, must rely on other, less deadly, persuasive tools. Chapter 4 argues that these persuasive tools are not sufficient by themselves to prevent all free riding; so the state remains necessary. But chapter 6 argues that these persuasive tools are still powerful enough to take a good deal of enforcement out of the hands of the state.
3. Although Taylor's purpose is primarily descriptive, one might hope for more attention to recommendations for the best way to organize and run communities. But Taylor gives little attention to this matter, treating community as a necessary instrument whose drawbacks must be excused because of its unavoidable role in providing social order.
But there are many different ways to organize and run communities. Giving such little attention to the question of which ways are preferable is analogous to demonstrating that the state is a necessary instrument of social order and then treating this result as indifferent between liberal government and authoritarian government. Taylor's presentation seems to allow for complacency or quietism regarding these issues. But I will argue, on the contrary, that they should be one of the main concerns of anarchist philosophers: for if the community is to be our main source of social order and the primary locus of much lived experience, surely it will make a very big difference just what sorts of communities carry out these functions. It seems reasonable to ask for a more detailed examination of the preferred way of organizing such societies. It is sensible to want to know what ideals they should seek to fulfill and how they should operate. We would expect no less of a statist theory of social order and justice.
Taylor could respond that this extensive discussion of the actual working mechanisms of communities is an enormous task, far outside the scope of his slender book. This is true; Taylor himself was under no obligation to undertake this task. Still, someone must eventually try to produce a description of the ideal community structure. While I do not intend to produce an exhaustive analysis, I can at least start on the task that Taylor has left us. I carry this out in chapter 6.
4. Taylor provides useful information about how social order can be created without state intervention. But he does not tell us why that ultimate goal is worth having. He explicitly disavows any intention to provide a philosophical argument for the thesis that anarchism is a highly desirable form of social life.
I believe, however, that there is a genuine need for an explicit defense of a set of principles that explain the desirability of anarchism. For it is far from clear that statist society is necessarily the worst that we can do. The state does serve, under many circumstances, to insure justice, order and public safety. This is a truism, but it is no less true for being a truism. Why, then, should we seek to abandon or limit state authority?
This question, of course, cannot be answered simply by looking at what various anarchist writers have said. One's picture of the ideal post-state society-and one's notion of which features of present-day society are most objectionable-depends on which set of values one holds. As I indicated in my brief discussion of traditional anarchist categories above, the capitalist anarchists place ultimate value on pure negative liberty, and hence their theory has no intrinsic objection to oppressive power relations that are the result of voluntary choices. Communal anarchists, on the other hand, have strong objections to oppressive relations regardless of whether they result from state or voluntary interactions. These differences in values lead to deeply divergent pictures of the ultimate goal of post-state society; and there are other differences, only slightly less striking, between different varieties of communal anarchism. If we are to have any idea about what anarchist society should look like, we need to know more about why authority is problematic in the first place.
Taylor could argue that agnosticism about this matter is preferable to basing his argument on ideals which may not be shared by everyone. And a condition of statelessness would, of course, provide opportunities for development of many different kinds of anarchic societies. It is reasonable to avoid second-guessing the means by which individual communities would encourage their members to be socially responsible.
Still, keeping the theory reasonably flexible does not require us to be completely undiscriminating. Presumably, a community should not have any of the fatal features that make states so unappealing. But in order to say just what those features are, we will need to know just what it is that we are objecting to. And it is easy to come up with examples of horrendous injustice in small, locally-based communities, even when they avoid using direct physical coercion. Given these concerns, it is sensible to try to set forth a broadly acceptable set of values that will help us have a better idea of what our communities ought to look like at their best. I will attempt this task throughout the dissertation.
The structure of this dissertation
This dissertation has the following basic structure. In chapters 2 and 3, I discuss William Godwin's political theory. I believe that this theory can provide us with a model for an acceptable modern anarchist view. These chapters present a summary of Godwin's most important claims, and discuss some central issues of Godwin interpretation.
Chapter 4 addresses the role of the state. As I have mentioned, my view diverges from those of most of the nineteenth- and twentieth-century proponents of anarchism. I claim that there is no feasible way to do away with the state and still maintain an acceptable level of justice.  Chapter 4 investigates two different proposals for providing justice in the absence of the state. I conclude that although these proposals are the most plausible yet constructed, neither gives us any realistic hope for a modern society that is both stateless and just.
This conclusion rules out anarchism as it is usually understood. But in chapter 5, I argue that on the contrary, the theory proposed by Godwin can support the existence of the state and, under certain circumstances, an obligation to obey the law. My theory assumes the need for some state authority, abandoning the traditional emphasis on complete statelessness. Rather, it focuses on the need, within a state-governed society, to redistribute much social authority into the hands of communities. Even under state authority, small-group units (families, schools, workplace offices, and so on) have a major role to play in the establishment and maintenance of social order; and I will argue that placing more emphasis on these small units will enhance their ability both to provide social order and to allow each individual the opportunity for complete personal and individual development.
Finally, chapter 6 takes up the question of the means of maintaining order in communities. Placing more power in the hands of communities to promote peace and order does not by itself guarantee that these groups will be fair, just, or productive of individual flourishing. I examine some of the many ways in which small communities fall short of the ideal. I then suggest some ideals that these societies should use to guide themselves, and investigate some of the ways in which those ideals can be put into effect.
The first task, then, will be to investigate the classical theory that I think has the best prospects in the modern world: that found in William Godwin's Enquiry Concerning Political Justice.
 Some of the most prominent communitarians are Alasdair MacIntyre, Michael Sandel, and Charles Taylor. For some examples of communitarian criticisms, see MacIntyre 1981, Sandel 1982 and 1984, Sandel (ed.) 1984, and C. Taylor 1985a.
 For examples of responses to communitarians, see Gutmann 1985 and 1989, Holmes 1989, Kymlicka 1989 and 1990, Buchanan 1989, Friedman 1992, and Rawls 1993.
 See Buchanan 1989 and Kymlicka 1989.
 Kymlicka, op. cit., 990. Note that Kymlicka pointedly remarks on the next page, "While both sides have something to learn from the other, that is not to say that the truth is somewhere in between the two. ... state perfectionism would have undesirable consequences for our society."
 See chapters 5 and 6 for further development of this line of argument.
 Naturally, the view may not accord well with all of our deeply held political and social values. What is important is that the view be capable of a plausible defense that does not clash too sharply with the majority of our concerns. A good anarchist argument should meet the standards of reflective equilibrium proposed by John Rawls (1971): if the argument works well, we may be led to adjust some of our ground-level values, just as the values helped shape the theory at its origin.
 Godwin was not as much the economic revolutionary as later theorists, as we shall see.
 This view is not uncommon in the tradition of Marxist attacks on anarchism. See comments by Hobsbawm cited in Marshall 1993, 663. One can also find it, though, in the work of James Joll (a commentator sympathetic to anarchism): see Joll 1971. See also works cited by Ritter (1980), esp. in the introduction.
Still, the most common criticism of anarchism has been an implicit one: simply ignoring it.
 This claim is the central thesis of Alan Ritter's very useful Anarchism: A Theoretical Analysis (1980). Ritter surveys all four of the classical anarchists (Godwin, Proudhon, Bakunin, and Kropotkin), while I concentrate on Godwin.
 As distinct from, for instance, 'liberty' interpreted as collective self-determination. (Though collective self-determination does play an important role in many modern liberal theories.)
 For an introduction to the contrast between negative and positive freedom, see Berlin 1969 and C. Taylor 1985b.
 The phrase 'experiments of living' is John Stuart Mill's, from On Liberty. See Mill 1972, 124.
 For more information on this topic, see chapter 5. As I mention in chapter 5, Nozick (1974, at 331) and Philmore (1982) are, rather strikingly, open to the notion of enforceable contracts of indenture or voluntary slavery.
 David Friedman is another well-known capitalist anarchist. Some right-wing anarchists from earlier in the century were Benjamin Tucker and Albert Jay Nock.
 Saying that the communal anarchist opposes oppressive authority leaves open the very significant question of whether the anarchist can admit that any form of authority is not oppressive. This matter is addressed at some length in chapters 3 and 5.
 This is not to say that Godwin was the most politically astute of the anarchist philosophers. His theory was notoriously utopian in many respects. But this is not terribly surprising: as philosophers know well, philosophical sophistication may have little to do with practical utility. Part of my project will be to pare away some of the less realistic features of Godwin's system.
 This only hints at the large literature on or related to autonomy, about which much more will be said in chapter 3.
 Thomas E. Hill's interpretation of Kantian autonomy has similar implications for this sort of case. See Hill 1989, 93-4.
 This is a line from a well-known poem by Victorian poet W. E. Henley, sometimes titled "Invictus." See Henley 1908, v. 1, 125.
 Perhaps this is better regarded as a feature of legitimate authority. Although it may be prudent to obey an illegitimate de facto authority, we do not think of such an authority as having the right to command.
 Apparently this need not rule out taking advice from those more knowledgeable, and perhaps even treating that advice as authoritative: for example, by acting on the advice immediately, without seeking out other opinions or trying to confirm the adviser's information. But this still does not cede the right to make the ultimate decision, which remains in the hands of the autonomous agent.
 See also articles by Reiman (1978) and Pritchard (1979).
 Steve Darwall stressed this point to me in conversation.
 Nor, apparently, is it a Kantian stance. As Elizabeth Anderson has stressed to me, Kantian autonomy-rational autonomy-requires thought and action on the basis of the categorical imperative, which means acting only on maxims which can be willed as a universal law. So my autonomy is no more or less a duty than is anyone else's, and no more or less worthy of respect. (Kant's understanding of autonomy as rationally obeying the categorical imperative is not common to all views of autonomy.) For more on this subject, see chapter 3, Gerald Dworkin 1988 and 1989, and Feinberg 1989.
 Interestingly, this principle (or something like it) may underlie some of the arguments for libertarianism. If we were to limit the bounds of legitimate application of deliberation to one's own property, and add a principle of punishment for boundary-violators, we would have something very much like libertarianism. These additional principles, of course, are not derivable solely from the doctrine of the overriding importance of autonomy.
 D. O. Thomas considers and rejects a similar argument in his commentary on Richard Price, The Honest Mind (1977), 122.
 This explanation of the public goods problem will be brief; I assume that most readers will have some prior knowledge of it. For more information, see any of Taylor's three books, or virtually any intermediate micro-economics textbook.
 Taylor gives the term 'community' a special definition, which is discussed in more detail in chapter 6. An important feature of Taylor's notion of community is that it implies a high degree of personal acquaintance and interaction. So although we sometimes speak of a large city, or a scattered group of people linked by some cultural characteristic, as a community, these would not normally count as communities in Taylor's sense.
 Though not longer than average North American life spans today.
 On average, it only took four to five hours per day to produce necessary shelter, clothing and food for one's family. Most of this time was spent on food collecting and preparation. See Sahlins 1972, 56-57.
 Direct relations are those that are not mediated by other people or (more importantly) institutions. Many-sided relations means that people relate in more than one social circumstance of set of roles. For example, in pre-agricultural societies the links connecting community members are many, including close residency, kinship, cooperation in food preparation and other work, and so on.
 This term originated with Hillel Steiner. See M. Taylor 1982, 12; and Steiner 1975.
 Taylor does express doubts about the viability of pure anarchy throughout his work, starting most notably in Community, Anarchy and Liberty and increasingly in The Possibility of Cooperation. In the latter work, he is more concerned to promote a principle of federation, which is roughly compatible with my overall thesis.
 There may also be other public goods which cannot be provided by a stateless society, but I will focus my attention on justice, the public good which all theories of the state require states to provide at a minimum.
Proceed to Chapter 2
Return to the Table of Contents
Return to Chris Roberson's home page