An Approach to the Problem of Authority

This paper examines the central problem of authority as illustrated through the lenses of philosophical anarchism. The justification of authority constitutes the paramount problem of political philosophy. The philosophical debate on political obligation, initiated by defenders of the state and anarchists, concerns the justification of a special bond between citizens and political institutions and lies at the centre of the discussion on authority. Philosophical anarchism’s criticism of political obligation creates great difficulties to state justification. Yet, this anarchist strand has been dismissed as a merely negative critique of political authority, with no alternative positive proposal of its own. In disagreement with this view, this paper aims at improving our idea of authority by revealing the value of the anarchist approach. For this I focus on the position of critical philosophical anarchism and on three theorists each of whom relates in a significant way to this position. I discuss Rousseau as a traditional theorist whose view is a basic inspiration for the anarchist approach to political institutions. Joseph Raz’s theory is analyzed as a view largely compatible with critical philosophical anarchism’s perspective on state authority. Finally, I discuss Simmons as a representative critical philosophical anarchist, from whose approach, however, I depart in my defense of critical philosophical anarchism. My conclusion is that critical philosophical anarchism sheds a new light on the important insights on authority offered by the above theorists and helps us adopt a new perspective on political institutions. In contrast with political anarchism, it does not propose to overthrow the state. Yet it justifies the former by establishing an everlasting criticism of authority. The absence of political obligation reaffirms the responsibility of citizenship. This entails a responsibility on our part, whether anarchists or not, to participate in evaluating continually authority’s decisions and practices, toward a better political world. Journal of Business & Financial Affairs J o u r n a l o f B usi ness & Finncial Af f a i r s


Introduction
In this paper I set out the central problem of authority as illustrated through the lenses of philosophical anarchism. For this I focus on the position of critical philosophical anarchism and on three theorists each of whom relates in a significant way to this position (a position which will be illustrated throughout my argument and will be defined clearly in the final section). I discuss Rousseau as a traditional theorist whose view is a basic inspiration for the anarchist approach to political institutions. Joseph Raz's theory is analyzed as a view largely compatible with critical philosophical anarchism. I use it to illustrate how accounts of state authority motivated by the anarchist perspective can be understood and improved. Finally, I discuss Simmons as a representative critical philosophical anarchist, from whose approach, however, I depart, criticizing it on central points in my defense of critical philosophical anarchism.

The paradox of authority
We live in a world dominated by political institutions. We find our lives ruled and controlled by them. We mostly take this situation for granted. How did we arrive at such a state of affairs? And is this how things should be? In many other areas of our lives we feel that things should be under our own control. We think that it is important to be able to decide and make choices for ourselves. We consider it important that we be free to act within a background of various options and free to pursue the best options for ourselves in life. We do not want other people to tell us what to do and to take control over what concerns us. So why in the case of the state do we take rule for granted? Even within the state, the desire for self-government survives in the form of dissatisfaction, when there comes a point that political interference feels unbearable. Why, then, do we so readily accept political power? Should we do so?
The appearance of this paradox is the starting point in this paper for examining the anarchist position and its approach to the problem of authority. The paradox reflects the idea that the best way to justify rule to individuals who can be, and have the right to be, self-ruled, and who thus find constraints undesirable, is to show that this rule is their own decision, that government is the result of self-government in the sense that we put constraints on ourselves. A decision to be constrained seems to be the most promising and comprehensive account of political constraints. What is required is to demonstrate how this can happen. The preceding explanation of the origins of the state suggests an answer, but we need to determine whether this is what actually happens. Indeed, as we shall see, the paradox in fact is not a paradox.
Nevertheless, its seemingly paradoxical character helps reveal an error in the defenses of political institutions. The problem lies in that showing the state to be good for us does not amount to showing that we accept it voluntarily. The state's being good can be a reason for our voluntary acceptance, but the two are not identical. Unless it is our own actual appreciation of the beneficial character of the state that leads us to accept the state, the state cannot be seen to be the result of self-government, namely the result of the participation, decision, and control of those who want to survive and live together freely. One could say that "being in our benefit" makes it a reason for us to want the state, and that a reason for us is our own reason and thus our own choice. But we still need to prove that the state is beneficial on the basis of reasons that actually are for us and thus that it can be, or is, our choice.
The error of the state defenders lies in the fact that, in demonstrating the merits of the state, they also thought that they had demonstrated its legitimate authority. But something that is potentially a benefit for a number of individuals is not so, unless it is based on their choice. In addition, something that was a benefit for some in the past is not necessarily a benefit for those concerned at present. For the latter to be the case, it needs to be actually a benefit at present and to be seen as such by those whom it concerns.
The defenders of the state-When I refer to "defenders of the state," I mean all those theorists who defended accounts of political obligation that reflect the approach I criticize in this paper from the perspective of anarchism. Such theorists range from traditional political philosophers, such as Hobbes, to contemporary theorists, such as George Klosko. But their approach might be adopted by any theorist or individual;committed themselves to a correct starting point when they attempted to defend the state on the basis of self-government. But this involves a continual assertion of choice. Instead, they provided reasons for seeing institutions as desirable and thus deserving of acceptance, reasons that could motivate choice but that are not themselves choice. They also thought that such reasons can be given once and for all, which ignores the continual need for the expression of choice. An argument for the benefits of political institutions could defend their existence, but not their authority. Nor could it defend their existence once and for all. Nevertheless, the defenders of the state thought that their argument did exactly this. The result is that, in addition to showing that the state is necessary for providing order and safety, they started out facing the state as a good in itself and as an entity over and above individuals, with independent existence. This might be described as a romanticized view of the state. This view may be what lies at the bottom of our unreflective acceptance of the state as an inevitable reality and what makes us forget its defects and our initial dislike of constraints. It has not yet been proven that constraints are beneficial for us as constraints we choose ourselves to impose on ourselves and on one another. Showing constraints to be self-imposed is the only way to demonstrate that the state exists for the sake of those it constrains and not at their expense.
The philosophical anarchist presses this point within the debate on political authority. In order to explain this position better, I want to illustrate the above argument about the seeming paradox. An effective argument for the state is based on the point that autonomy is extremely important and that, for the state to be justified, it needs to be shown to protect it. But, the argument goes, for this the state need not be autonomously chosen. This argument constitutes Raz's theory of the state, which is explored in the works just cited and is analyzed below [2][3][4]. People make mistakes in their choices, and the right kind of choice cannot be derived from the universal consent of actual individuals. Rather there are concerns that are primary for us, good reasons that apply to us, whether or not we are able to see them. When the agents of the state are sensible enough to find and serve these reasons, then we are better off if we let them do so. We should make sure that a state serves good reasons that apply to us, whether or not we can actually choose them. In this way we have our capacity to be autonomous protected and enhanced, through a good government built on such reasons. So a good government is justified because it protects autonomy, but it is not the case that we choose it autonomously. This is a good starting point for approaching our relationship to the state. It asserts the value of autonomy without facing the difficulties arising when we try to defend it through individual consent. It also clarifies why the protection of important benefits is essential for our self-government, even if we do not actually choose them. The anarchist will eventually concede that this argument can be a basis for defending the state. In order for the defender to make proper use of this perspective, however, he needs first to see the point of the insistence on choice that the critical philosophical anarchist maintains. The anarchist claims that for the state to be legitimate, it needs to be shown to have been chosen. He finds a point in a perspective that argues that a good government is good because we choose it; it is our choice that makes it so that it protects us. This idea implies that we are self-governed through the state: rather than claiming that the state is justified because it allows us to be self-governed and enhances this capacity, the argument holds that the state can be justified only if it is a way through which we govern ourselves. The difference between the two approaches is that, while the defender sees autonomy to be respected because it is protected and served through a good government, whether or not we choose that government, the anarchist claims that autonomy is respected only if we choose the state such that it becomes positively a way through which we govern ourselves: the state is good because we choose it and because it becomes a way through which we choose.
This difference is important, because each view has a different criterion of justification. For the one view, the respect for autonomy that government is expected to show consists in finding and serving what is right; freedom is realized through the realization of good reasons. For the other view, autonomy is shown to be respected by government only if government becomes itself a way through which we exercise autonomy, and maybe the best way. This distinction is crucial also because it points out a confusion that needs to be avoided in evaluations of the state. The romanticized view of the state exhibits this confusion: it emerges from an illegitimate move from unjustified identification and vacillates between one criterion and the other. It sees the service of good reasons on the part of the state as a way in which the state makes us autonomous, instead of just allowing us to be autonomous at best. It identifies rightness, or merit, with individual authorization. It is this assumption that leads defenders to see the state as an independent good in itself, as inherently connected with and as expressing the value of persons and of the interaction among them.
As a new alternative, the critical philosophical anarchist accepts the approach that focuses on good reasons with regard to some justifications of the state, and yet insists on the value of choice for the purpose of political relations. Anarchism returns to choice in a way different from that of the second approach discussed above and with a criterion that becomes less problematic: it reminds us that it is important that through the state we remain free, rather than become free. What is the importance of such an insistence on choice?
At the beginning of this paper it was pointed out that in many areas of our lives we care most about being self-governed and about preserving this capacity and right for ourselves, because it is the only secure way we can survive and flourish. We do not want our lives to depend on others. The defenders of the state follow this natural way of thinking when they begin to show that the state is at our service. The idea then is generated that the state can be justified because it is created for the sake of, and on the basis of, our very capacity to be self-ruled. Yet now a paradox also seems to be generated: we create a condition in which we are ruled by others in order to remain selfruled. That is, we do exactly what we do not want to do in order to secure what we want the most. As indicated above, the paradox is not really a paradox, but its seemingly paradoxical nature helps reveal the confusion in the defenders' argument. It is not a paradox because, in order to be self-ruled in some ways, we need both ourselves and others to be constrained in some other ways. Most importantly, however, it is not a paradox because it is a consistent and reasonable idea that, in order to be self-ruled, we need to be self-constrained: it is possible and sensible that self-imposed constraints constitute appropriate conditions for individuals to enhance their capacity for choice. When, for example, one decides to quit smoking because it is better for one's health, one puts a constraint on oneself not to smoke again. This constraint enhances one's freedom by helping one to apply the decision with which one chooses to rule one's life in this respect. If the state is a way for us to be self-constrained, then it can serve as a way for us to be self-ruled. The impression of a paradox arises from the fact that it is in the nature of the state that some rule others, and so the state is offered as a way of our becoming self-ruled that involves being ruled by others.

Dissolving the Paradox. Rousseau as a paradigm of state justification
There is a way through which the seeming paradox is dissolved, and it lies in Rousseau's intuition that the state is justified only as a way of our being governed by ourselves. As Rousseau states in his The Social Contract: "The problem [of political justification] is to find a form of association which will defend and protect with the whole common force the person and goods of each associate, and in which each, while uniting himself with all, may still obey himself alone, and remain "as free as before" [5]. To create such a civil society, individuals must unite under an agreement the conditions of which are unanimously accepted and with the intention to hold each other to those conditions. As Arthur Ripstein puts it, "Rousseau's claim is that a community consists in a group of people in agreement both about the conditions of their interaction and their intention to hold and be held by each other to that agreement [6]. By becoming a member of a community created by such an agreement, each individual identifies with the general will, which is the united will of all self-legislated citizens expressing the choice applied within political society. Thus in order to be free, the citizen always needs to hold to the general will, "to will impartially with all the others [6]. So if he disobeys this will, others have the right, based on their mutual agreement, to coerce him to be free: whoever refuses to obey the general will shall be compelled to do so by the whole body. This means nothing less than that he will be forced to be free [5]. "Coercion is legitimate on grounds of freedom because the agent has chosen to be coerced" [6]. The idea defended here is that having others decide for us and serve us with regard to some matters is not a denial of freedom or choice, if the nature of their decisions and the content of their choices represent our own choice: we are the legislators, thus the creators of the ways we are to be constrained, thus self-constrained, and thus free. In order for the defenders of political institutions to preserve this argument, however, they need to attend to it properly as the only way of defending the state in terms of self-government. They need to attend to the proof and preservation of this kind of choice.
As established the previous section, there is a difference between seeing something as justified because it is rightly discovered to causally promote our autonomy and seeing it as justified because it constitutes a consequence of our autonomy. The anarchist insists on the need for the latter, which captures Rousseau's idea of self-government, to make the argument for the state a defense based on the idea of freedom through self-constraint. But the defenders of the state do not stay consistent in their use of this argument. By focusing on benefits of the state that make it good for us, they gradually change the initial argument from choice into an argument for the protection of choice through benefits. They are right to see the fact that the state provides safety as a strong reason to support it, but in the end they also believe that this makes the state a good in itself. Thus their defenses end up saying that we are ruled by others in order to be self-ruled. That their view allows the defense of the state to be rendered paradoxical is an indication of the mistake the defenders make in the process. The problem is reflected in the failure of accounts of political obligation that the anarchist stresses. And it is telling of something the anarchist observes and indicates with worry: that, by nature, the state involves subjugation in the disagreeable sense of some ruling others with the intention to subject them, which creates inappropriate relations among individuals. See discussion of Raz in the following section.
There is, however, a way of correcting the state defenders' error. By staying clear and insisting on the form that Rousseau gave to statedefense, we might come to realize that at one level it is unachievable for the state. Then we may legitimately move to the other argument -still in light of the importance of choice indicated by the argument that we cannot achieve in a direct form -which gives an alternative view of legitimacy in terms of autonomy, although clearly not choice-based. We can then pay attention to its specific implications for the state.
If, with Raz, we try to determine what we can tell about how the state works and focus on its protecting our capacity for autonomy through its service of good reasons, we need to see what direction such a defense takes. It cannot prove political obligation, nor can it justify the state once and for all. Rather it can concede the continual instability of the state and concentrate, as most important, on the attempt to justify the reasons why they are subjected to individuals themselves, not through actual choice, but through reasonable testing and evaluation in the light of the lack of such a choice. (This is facilitated, for example, by the enforcement of law through transparent and accountable mechanisms, which, at some level, is an expression of continuing choice). This is a demanding approach, yet one that is a consistent recognition of the value of choice, not departing from the initial argument about choice in a destructive way. The two arguments come together through a fair compromise: we assert the value of choice, while realizing that we cannot base the state directly on choice; in the light of this impossibility, we insist on the importance of finding acceptable ways of justifying reasons to the individuals subjected, making the defense of the state in terms of goodness a recognition of self-government. The idea of basing government on ethical concerns that we all share arises as the prominent task of justification for the defenders to pursue.

Raz's Theory as an Illustration
Raz's position helps illustrate the preceding discussion. I will use it as a representative example to show how we can better understand and improve the views on the state offered by those motivated by the anarchist position.
Starting from the central anarchist intuition that there is some puzzlement with the idea that "one person has the right to rule another," i.e., that the right to rule is deeply disturbing as a notion [4], Joseph Raz sees this puzzlement as rooted in the fact that authority involves a "dimension of subjugation" that is distinctive of it, namely it involves duties that are "deliberately imposed by one human being on another with the aim of subjecting that other to a duty [4]. Hence, he sees the anarchist complaint to be ultimately about "the problem of subjugation," that is, of the subjection of one person to another [4], where unequal dependence is the main aim, and this is facilitated by giving dependence a specific form. In essence this "is a problem of the relations between one person and another," which, as I will illustrate throughout this paper, concerns the anarchist most and remains vivid with regard to political authority [4]. Importantly so, it signifies the central role of domination within political authority, of "the capacity of one party to exercise control over another party [7]. The anarchist worries further about the fact that domination and hierarchy are definitive aspects of authority at the expense of freedom and equality, encouraging other harmful social phenomena such as exploitation and coercion. For a useful account of domination and exploitation in relation to authority and the prior role of domination in this regard [7]. Green's and McLaughlin's view that authority is problematic and potent [7,8]. Thus, Raz's account helps us illustrate how the concern with the creation of problematic relations between people, as it appears in subjugation, or domination, lies at the heart of the anarchist scepticism on authority. This paper also aims to demonstrate that it takes on special significance in the criticism of political obligation.
Given this problem, one should at least concede the idea that "no unlimited authority can be legitimate" and thus that "we need a doctrine of limited government, i.e., of the principled limitations on the possible scope of governmental authority" [4]. As we will see later in this paper, this requirement is not a demand simply and primarily to limit the scope of authority as much as possible. Rather, it is a demand to find reasons and principles that determine the nature and functions of authority in a manner that makes it justifiable to all reasonable, adult individuals subject to it. The suggestion I shall develop, to apply an ideal of legitimacy to further justifications of constraints in view of the results of the debate on political obligation, sets such a background as a primary condition. Although the concern with limited authority is a liberal demand and anarchism is defined by a prior concern with whether authority can ever be morally legitimate, once this concern is properly recognized, further evaluations of constraints can function within this background. Moreover, the defenses provided by various contemporary political theorists, such as Rawls and Dworkin, may be seen to work within these boundaries, in fact to work adequately only within them. Within this framework, Raz's reply to the anarchist challenge is that "the basis of legitimacy is relative success in getting people to conform to right reason" [4].
To explain: Raz's defense of "practical authority," that is, "authority with power to require action" [3], involves three main normative theses [3]. These work within the background of his general approach to the analysis of authority, which includes the idea that authority necessarily entails obligations to obey. Because the justified use of force would not be authority unless it included an appeal to compliance, which is meaningful only if there are things to comply to. In short, legitimate authority is usually exercised by giving directives, issuing instructions. It is much more than use of coercive threats, it imposes duties and confers rights and thus involves an obligation to obey [3,7]. It also involves the claim that the indispensable feature of authority is "surrender of judgment," which is not taken to mean that obedience erases personal deliberation, but rather that it is not conditional on personal examination of the thing prescribed. When individuals follow authoritative prescriptions, they do so on the understanding that, whatever their opinion is about the thing prescribed, after they have acknowledged the authority they are expected to follow its directions as given [3]. For this idea, the first distinctive feature of authority is "surrender of private judgment," which means that with regard both to conduct and to belief an individual who accepts authority recognizes that someone else's prescription is to be followed simply because it comes from one "acknowledged by him as entitled to rule," that that person is entitled to decide for him or her on certain specified areas without any need for persuasion through further argument whenever a relevant issue arises [9]. The second distinctive feature of authority, which is relevant to the present discussion, is the "mark" of authority, namely the need to provide the sign or credential of authority through "some public way of identifying the person whose utterances are to be taken as authoritative," which regards "the recognition and acceptance of certain criteria for designating who is to possess this kind of influence" [9]. It is also important that we concentrate on the source of authority, the source of the special sort of reason for action that authority is meant to denote, rather than its content, and that we find this on the person and their status [9]. For similar points, especially his reference to "peremptory reasons," [10]. Finally it involves an account of the nature of authoritative instructions as "dependent" and "preemptive" reasons for action, namely reasons meant to reflect the balance of reasons on which they depend, instead of being added to them, and reasons meant to replace the original reasons on which they depend, while not having absolute supremacy [3].
Accordingly, Raz's first thesis is the dependence thesis, concerning "the general character of the considerations which should guide the actions of authorities" [3], which is the position that authorities should act on dependent reasons in order to achieve an ideal exercise of authority [3]. That is, "all authoritative requirements should be based, in the main, on reasons which already independently apply to the subjects of the directives and are relevant to their action in the circumstances covered by the directive" [3]. Yet, while they reflect reasons that apply to the subjects of authority, authoritative reasons still "make a difference to what [the] subjects ought to do". The second thesis, the normal justification thesis, which "concerns the type of argument required to justify a claim that a certain authority is legitimate," the position that for authority to be justified it should be shown to be the best way for individuals to conform to reasons that apply to them, reasons to which they themselves are committed [3]. This paper maintains that the normal and primary way to establish that a person should be acknowledged to have authority over another person involves showing that the alleged subject is likely better to comply with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly [3]. This is the crux of Raz's defense of authority. Together, the dependence and the normal justification theses "articulate the service conception of the function of authorities, that is, the view that their role and primary normal function is to serve the governed" [3]. This leads to the third normative thesis that Raz adopts, the pre-emption thesis, which "concerns the way the existence of a binding authoritative directive affects the reasoning of the subjects of the authority" [3]. On this position, such requirements pre-empt the reasons they are intended to serve: when they guide action, they replace their underlying justifying reasons [3]. The thesis claims that "the fact that an authority requires performance of an action is a reason for its performance which is not to be added to all other relevant reasons when assessing what to do, but should exclude and take the place of some of them" [3]. The resulting idea is that, because authoritative reasons should, in order for authorities to be justified, be dependent on already existing underlying reasons, authorities "should have the right to replace people's judgments on the merits of the case" [3].
Let me note here that the contractualist approach to authority, to which Rousseau's theory belongs, adopts the service conception of Raz and the ideas reflected in the two theses that compose it. According to this approach, authority can be justified on the basis of reasons that represent, apply to, and are the best for the individuals subjected. Yet this approach departs from Raz's third thesis. In contractualism the central idea is that, in the case of practical justifications such as those concerning political authority, the relevant reasons have to be explicitly justifiable to those they concern. Thus in the social world it is important that individuals actually see the reasons that apply to them. Practical reason differs from theoretical reason, and in the case of the former individual judgment can never really be replaced. This is because in theoretical reason the point is to find the truth as it applies irrespectively of individual opinion (as is the case with axioms in mathematics), whereas in practical reason the truth is determined on the basis of individual evaluation and interaction itself. This position leads the contractualist to a distinctive method of justification that gives a special role to choice and is of great importance for the debate on political authority.
In contrast, within the context of his own theory as described above, Raz sees "theoretical" authorities, referring to "authority for believing in certain propositions," to be more likely to have the same structure as practical authorities and to be supported in the same way by his dependence thesis [3]. Yet there is crucial difference between theoretical and practical authority. To start with, the former is nonexecutive, while the latter is exactly "the right of A to issue practical directives and the correlative duty of B to follow them or to obey them" [7]. It is crucial to note that Raz relates his account strongly with the idea of theoretical authority when he focuses on authority in terms both of "its expertise (or that of policy-making advisers)" and, more importantly, of its "ability to secure social coordination" [4]. This assimilation seems to sit well with his normal justification thesis and the service conception as well as with further general ideas that motivate his view, ideas that I will discuss in my criticism below. It is also consistent with the anarchist attitude towards authority, a scepticism that does not reject forms of authority such as parental authority or the authority of the teacher and that demonstrates a complex attitude towards theoretical authority. For relevant and very clarifying accounts of theoretical authority and the anarchist position on it [7,11]. But it also brings to the fore the concept of "an authority" with its special characteristics. For the following analysis of this idea [9]. More precisely, this notion of authority maintains that deference to someone as authority is based on his or her special knowledge and the presupposition that others are debarred from such knowledge. The idea of authority here puts "the person prior to the system," the latter concerning established procedures for creating authority, which are prior according to the idea of "in authority" [9].
The concept of "an authority" also focuses on special capacities and the quality of one's decisions as what sets one apart from others, irrespectively of whether they accept that person or not, and as what makes one person an intermediary between the world and the rest of us, who adds something to it for us to take -it centers on differential access [9]. This notion of authority also presupposes a kind of inequality, namely personal differences (unequal capacities) and a hierarchy prior to the authority relationship on which that relationship is based [9]. A second condition, which however does not concern us primarily here, is that the knowledge available to the person of authority should be "in principle available -at least to some humans," i.e., that there exists an "'epistemological' framework," a "class of things capable to be known," this involving the second-order "belief that the mind of man can have contact with the reality on which [the relevant] authority speaks" [9]. Given the involvement of these features in the idea of theoretical authority, its use within Raz's account must be approached carefully. I will explain how in the following paragraphs.
Finally, Raz sees his account as making the surrender of judgment, characteristic of authority relations, compatible with keeping one's  [3]. He situates his position in relation to the liberal theorizing on authority, explaining that the liberals focus on the Rawlsian "duty to support and uphold just institutions" as a proper way of justifying limited government, while his own account is offered as an attempt to answer the prior question of an "understanding of which institutions are just" -or, to be "setting the question in a certain way, one has a duty to uphold and support authorities if they meet the conditions of the service conception" [3]. Raz's account also becomes a basis for the attitude of "respect for law" (it is actually what grounds its application, meaning, and validity), which he sees as an acceptable expression of the morally desirable sense of identification with our societies and thus one that binds to the authorities of their societies those who adopt it [2,12].
Given the preceding analysis, we can now examine how central aspects of Raz's view bear on our argument. Raz's normal justification thesis respects the anarchist idea that authorities, if they are to exist, can exist justifiably only if they are shown to be for the people and not vice versa. The service conception already claimed this. The position reflected in the combination of all his theses, that it is only on the basis of the independent reasons determining the moral responsibility of people that relations of authority can be accepted, corresponds to the anarchist perspective on justification. In the debate on political obligation, anarchism establishes a demand for justification that has been neglected, a demand activating an ideal of legitimacy as a constant test for any account of constraints. This throws new light on Raz's account. He offers his account as explaining which institutions are just. Although anarchism begins with the prior question of legitimacy rather than justice, as far as institutional evaluation is concerned we can see Raz's approach as functioning within the background of justification established by the anarchist: his theses set the terms for the moral acceptability of constraints. We can also read his claim that the dependence thesis articulates a condition for the legitimate exercise of authority in the language of the anarchist ideal of legitimacy: only institutions that serve generally acceptable moral values are justifiable. This correlation will be understood when we will see how the anarchist ideal of legitimacy arises.
It is also important that Raz recognizes that "a complete justification of authority has to do more than provide valid reasons for its acceptance," as is the case with the normal justification thesis [3]. Such an account "also has to establish that there are no reasons against its acceptance which defeat the reasons for the authority," namely that the justifiable ways in which it functions are not accompanied by aspects that defeat their acceptability. For example, an important reason against its acceptance would concern "the intrinsic desirability of people conducting their lives by their own lights" [3]. This recognition fits the fundamental anarchist idea that our approach to political institutions needs to start from a prior consideration of the undesirability of constraints, of the need to consider both the defects of political institutions and their merits, rather than focusing only on the latter, which, as I will argue, helps reformulate the debate on political authority. The focus on people's initiative, a strong motivating reason for such an outlook, expresses the very anarchist concern with freedom and the importance of creating a background of appropriate relations among them that enables people to control their lives meaningfully. In the light of these considerations, Raz's justification thesis qualifies as one that functions properly within the background of justifications of limited authority as determined by the anarchist ideal. The whole account of Raz, then, is thoroughly characterized by the anarchist perspective and the way that perspective determines our approach to political institutions. Each of the specific elements of this account acquires a clear meaning and position within the framework established by the anarchist as one settling the tasks of political theory and action.
In order for this account to work properly, however, there are certain aspects that should be approached with caution. The considerations shown in the previous paragraphs to be recognized by Raz can apply to his view, provided that it satisfies certain conditions. Although the idea that the justification of political authority lies in its being an efficient vehicle of the reasons of individuals as moral agents is intuitive, its real force within political reality lies in making sure that authority is actually and continually proven to be such a vehicle. We need methods for applying the ideals of legitimacy that ensure that people understand their reasons, that they see how those reasons apply to them, understand authorities to serve them properly, and also feel capable of overthrowing them the moment they cease to do so. The view that authority may do well in helping us satisfy reasons that apply to us is valuable when used in the critical spirit that the anarchist encourages with regard to political obligation, and which extends to every evaluation of constraints, rather than in a spirit of confidence that leaves authority unquestioned in these respects. In the end, the relation of the anarchist position to accounts of limited authority is a natural consequence of the central anarchist tenet that it is extremely difficult to legitimize political institutions.
Against this background, the idea of "an authority" is not helpful. The critical outlook that we need to adopt and apply with regard to political institutions contrasts with this conception. The concern with proper relations between individuals and with a sincere exercise of moral responsibility and control over their affairs, which makes active participation indispensable, opposes a view of expertise based on exclusion and differentiation. Rejecting the idea cultivated throughout history that there are areas of exclusion and groups of agents who can entirely and constantly decide for the rest, with its resulting differentiation among people and the subordination of some to others, is what motivates the anarchist challenge in the first place. The notion of "an authority" encourages that idea and raises expertise to the level of an exceptional capacity and a higher goal. Its connotations with differential access obscure the crucial understanding of social life as a practice that concerns us all equally and for which we are all qualified. When we accept that the anarchist challenge expresses a legitimate and valuable concern, we cannot attempt to meet it by applying such elements that it itself falsifies.
Raz's view constitutes a representative way of how to account for claims of authority within the framework of justification defended by the anarchist, but his understanding of authority must not contain the connotations indicated above. To avoid this, it is important not to identify practical authority with theoretical authority. In practical reason it is not enough to find and follow the right answer. Rather it is crucial that we participate in formulating and applying such answers to ourselves. Individuals must be able to see as justifiable to them the reasons that authorities served. This is the impulse involved in the contractualist rejection of Raz's pre-emption thesis and in the anarchist insistence on the role of choice. Ultimately, Raz's theory facilitates and should be interpreted to account for the idea that political institutions can be an authority because they are based on an authority that we all constitute. The anarchist forces on us the task of ensuring in every case and at every level that authorities qualify as ways of discharging our social responsibilities and goals. This makes justification a hard process. We should remember that in fact the anarchist critical impulse rejects the surrender of private judgment and endeavours to confine it to the largest possible extent when it is inevitable that we live with authorities, of which this aspect is a central characteristic. Making this feature as compatible as possible with the basic inalienable capacity of freedom and with the kind of social relations that it requires characterizes the demand for justification established by the anarchist. This makes it a central concern of this demand that the surrender of judgment, whenever it occurs, genuinely reflects an aspect of our authority. Thus the idea of expertise, if used at all, must be seen as a difficult exercise rather than a prior ideal -something for political institutions to be actually and continually proven to qualify as and to become, by approximating to the ideals of legitimacy in the way the task of justification demands. It itself becomes an issue of constant evaluation that concerns all of us.
In the end, the above remarks redirect us to the real issue of political legitimacy, which is the moral justification of practical authority, of the authority to issue practical directives and have them obeyed. The form of authority that anarchists find morally suspect and is discussed here is a dominative social power that is binding and content-independent, necessarily involving recognition and submission by its subjects, which uses coercion even if it is not defined by it. For a comprehensive account of all these features of authority [7].

The Argument for Critical Philosophical Anarchism
In the rest of this paper, I outline the main argument in defense of critical philosophical anarchism. This argument takes two directions. On the one hand, it examines how critical philosophical anarchism helps improve our perspective on the state by presenting an alternative to the dominant positions concerning it. On the other, it focuses on a way of understanding critical philosophical anarchism that departs from, and improves on, the perspective of its theoretical defenders.

An Alternative to Prominent Positions on the State
The aim of this paper is to demonstrate the positive character and contribution of critical philosophical anarchism, an anarchist view that argues that there is no adequate defense of political obligation and concludes that in this sense the state is illegitimate. Political obligation is understood as a special relationship between individuals and political institutions which is the correlate of political authority and expresses a central sense of political legitimacy. For an account of the "correlativity thesis" and of the whole debate on political obligation and authority [13]. The position of critical philosophical anarchism helps us understand that the reasons on the basis of which we set out initially to justify the state are correct. Indeed, they are exactly those motivating the anarchist in the first place. But the method the defenders of political institutions seek for satisfying these reasons is wrong. The anarchist reminds us of and focuses on what it leaves behind. In fact, it involves a perspective that we can all share. Anarchism is the only perspective that holds consistently the view that we want to be self-governed and that the only way to be so within the state is to prove it to be a self-imposed constraint. An analysis of the dialogue between the philosophical anarchist and the defender within the debate on political obligation is critical for this purpose. Its result will be that we have no political obligations. The anarchist can treat this as a demonstration that we do not have a comprehensive argument for such a special relationship to the state, and as a good reason explaining why we cannot have such an argument. This, in turn, will open the way toward seeing a higher challenge to the enterprise of justification as a whole: maybe we chose the wrong argument altogether. Does this lead to the conclusion that there is no way to ground the legitimacy of the state? In terms of political obligation, there is probably no way. The state offers no additional ethical concern over and above the ones that we can share with one another. Self-government and equality cannot be expressed by consent within the state.
Yet there is a way of looking at the state that may be helpful. We can bracket the question of political obligation -take into account its results, but leave aside any effort to change them -and concentrate on something else: ethical reasons that we share are expressive of autonomous and active participation, and we should try to prove the state to be based on these. It can be legitimate in terms of not violating such ethical concerns and of being compatible with them. We can see the state as an association that we create for goods that, we find, ordinary associations do not provide. We are social beings. Yet our societies are the result of our collection and interaction. The state is not something mysterious, an entity over and above us that raises special demands. Once we say, as the traditional defenders have, that the state frees us, we adopt a romanticizing view. Instead, we need to remember that the state is not a good in itself, that it does not have an independent existence and value of its own. We need to see that the dominant defenses of the state have gone too far in attempting to identify beneficial order as a source of choice. Constraints are always constraints, and their being good for us does not change that. The real challenge is to see whether the political world in which we find ourselves living and which we might not want to abandon is one that deserves to continue to exist, on the basis of reasons that apply to us. This sense of legitimacy is our aim.
Political anarchism says the state is an evil. Naturally, on this basis, it rejects any effort to justify authoritative and coercive institutions, their existence and legitimacy, and our obligation to them. Critical philosophical anarchism does not reject the state in such an absolute way. (Note that strong philosophical anarchism such as the view of Robert Paul Wolff is more directly connected to the strong demands of political anarchists, but this is not the view I examine). What critical philosophical anarchism does is reject the legitimacy of existing political institutions, by proving that there is no adequate moral defense of political obligation, given its idea of what a legitimate state would be like. It also stresses the distinction between different kinds of evaluations of political institutions. In the face of these characteristics, critical philosophical anarchism has been criticized as a purely negative view, one that is skeptical of any positive effort of justification without providing its own alternative solutions to social problems. I disagree with this criticism. Critical philosophical anarchism shows dissatisfaction with the defenders of the state, but it does so by addressing positively the fundamental question about the very existence of political institutions, which these defenders neglect.
There are two prominent, opposing positions on the state that are defended, and they seem to be the main options available. On the one hand, the defenders of the state focus on the fact that it can provide necessary peace and order. From this they move further to declare the state an end in itself, having an independent ethical status of its own and, for some, being over and above individuals. We are urged to believe that we should have the state as a matter of moral necessity. On the other hand, the second position starts from the idea that we are rational adults who do not need the state to tell us what to do. Its advocates are motivated by an opposition to extended state interference, even for reasons of welfare. They even aim at rolling the state back, creating as minimal a state as possible, one that employs just the army and the police to protect individual property. The first, which was mentioned earlier in this paper, is the romanticizing approach to the state, a kind of idealism. This approach obscures the fact that the state is created by individuals and is a collection of people who come together in a particular way for their own good. The second is the (Right) libertarian position, which focuses on undermining the state and limiting its role to practices that can elicit individual consent. It nonetheless sees autonomy only as the independent interest of self-contained individuals and understands human relations only in terms of negative rights and duties, neglecting thus the sociability and interdependence of human beings and tending to endorse great inequalities. But are these the only defensible options we have? Should we either consider the state as an imperative moral ideal or try to reduce it as much as possible?
My argument is that there is a third option that has been neglected by these two perspectives: the position offered by the critical philosophical anarchist. This anarchism agrees with the libertarian that we are selfruled and we do not need the state to decide for us. From this, it points out the mistake of romanticizing the state. The anarchist criticism of political obligation reinforces the claim that the state should be seen as a means for our ends as morally developed (adult) agents, that it exists to serve us and has no ethical value over and above us. But seeing this, the solution is not to create a minimal state or eliminate the state altogether. Philosophical anarchism offers a perspective from which we can be entirely distrustful of the state while at the same time accepting a full welfare state. This means that the existence of the state does not replace our critical assessment of it, but, at the same time, when it exists, it is better that it covers more areas of social need rather than fewer, otherwise its functions are characterized by self-generated and unjustified domination, exploitation, and coercion. In this light, the anarchist ideal of legitimacy stresses that our focus should be on quality, rather than quantity: on the legitimacy of the character that state authority takes when exercised. This is the most reasonable position for us to adopt while living in the state. As a result of this position, the anarchist stands with the friend of the state in that we should help out and offer our positive participation while living in it.

Improving the way critical philosophical anarchists see their position: Simmons's theory as an illustration
According to the preceding argument, prominent positions on the state are unsatisfactory. But the view that philosophical anarchists themselves have about their position is also unsatisfactory. This part of my argument is necessary for a demonstration of any positive contribution on the part of critical philosophical anarchism. I will examine Simmons' position as representative of the literature on philosophical anarchism that I find unsatisfactory. Simmons misses nothing of the main characteristics of critical philosophical anarchism. Indeed, these are the features on which I base my account. My aim is to detect and defend something about them that has been neglected.

Simmons' theory:
Critical philosophical anarchism is involved in a "subversive campaign" [14]. For representative bibliography, applying to the rest of this paragraph, see [8,[15][16][17][18][19]. That campaign aims to demonstrate, through criticism of the state, that the non-existence of a general political obligation affects broadly our thinking and acting in the public domain. By leading us to the conclusion that there is no general political obligation, the anarchist criticism "removes any presumption in favour of obedience," and with this it "forces us to view the position of man in political society in a different way" [17]. It forces us to cease to think and act on the assumption that there is a morally binding special political relationship and to approach cases of obedience and disobedience on the basis of independent moral grounds. It invites us to distinguish moral reasons for action from grounds related to a distinctively political status and their political formalization, calling on us to accommodate ourselves to a careful weighing of the various moral considerations that are at issue in social situations. We need then to cultivate within the political structure the way we think about political institutions. We should begin to consider instances of illegality under a different light. In everyday life, we ought to appeal to particular laws as long as they overlap with morality and to recognize that many of them are arbitrary (for example, policies interfering with private conduct that is harmless to others, like the criminalization of drugs). In this way, we may gradually get used to depending less on authority in most of the practices and interactions of our social life.
The critical outlook that the anarchist project forces upon us involves a more difficult way of discharging our social responsibilities. It suggests "that we are more thoughtful about and more sensitive to the particular moral issues in our lives" [19]. We are reminded that we owe it both to ourselves and to others qua persons to take a responsibility for our social lives that goes beyond blind obedience, that the latter would be unacceptable even if we had political obligation, and that "citizenship does not free a man from the burdens of moral reasoning" [17]. Because in practice we are used to acting as if the directives of political authorities are beyond question, despite the fact that we now recognize that all relationships of authority need justification, "it is this widespread habit of compliance that the philosophical anarchist is trying to subvert" [14].
Correspondingly, the anarchist perspective and its accompanying ideals of legitimacy insist on the evaluative role of a set of generally acceptable moral standards, thus functioning as a principled reflection on the seriousness and the implications of an attempt to justify political authority. In this, the anarchist position becomes a suitable standpoint for the political theorist and for the wider population to adopt in order to conceive and apply adequately political ideals in an attempt to construct proper forms of social organization. It thus constitutes a strong basis for a deeper understanding and improved conception of our social relationships and lives. On the whole, as philosophical anarchists themselves claim, the anarchist scepticism challenges social order "from within," it forces us to reassess the moral status and significance of social order, and it "makes a difference to the nature of social relations" [8].
Nevertheless, critical philosophical anarchists also claim that the widespread adoption of such an attitude does not challenge the existence of political institutions or lead necessarily to a significant change in our existing social reality at the practical level. These positions relate to a list of arguments that these anarchists employ against the accusation that their position leads to an extreme and invariable radicalism and entails disorder. For such accusations [20,21]. As Simmons claims, the view that there is no political obligation does "not entail that disorder or revolution is justified" [17]. He advances his claim mainly on the basis of three lines of argument: first, that to be a correct position, critical philosophical anarchism should also be "weak" and adopt a "balance-of-reasons" approach concerning the weight of "judgments of state illegitimacy" [22]. Second, that there are various classes of moral reasons that individuals may have for complying with the law, even in the absence of political obligations [17][18][19]22]. Third, that political obligation is only one aspect of the justification of political institutions, and the kind of evaluation properly involved in it does not replace the kinds of evaluation involved in other justifications of them [17,22,23]. These arguments interact with each other.
According to the first argument, philosophical anarchism should be seen as giving to its conclusion against state legitimacy a "minimum content," namely that "the subjects of illegitimate states have no political obligations" [22]. This conclusion does not extend to a claim that either these subjects or people generally have an obligation to oppose and eliminate the state, although defenders of weak anarchism may hold such a view on independent grounds. Thus the anarchist conclusion of the debate on political obligation "does not translate into any immediate requirement of opposition to illegitimate states" [22]. This feature is also usually seen to be what differentiates philosophical anarchism from political anarchism. In addition, the judgment that there is no political obligation is not held to provide final reasons for action, which outweigh all other moral considerations, but rather to present the relevant rights and obligations within a "balance-ofreasons" approach, to be examined in view of other good reasons [22]. On balance there may be good moral reasons for not opposing the state, even if it is illegitimate.
Accordingly, as Simmons' second argument goes, there are three classes of such reasons. One such class includes moral duties to others as persons, such as duties not to harm others, which cover acts that are malum in se, and "wrongs of coordination" [18]. The latter are not wrong in themselves, but they become morally forbidden within contexts that make them harmful (e.g., when we endanger others by driving on the left in a society where the common practice is to drive on the right [18,19]). In cases where political institutions prohibit these kinds of wrongs, their laws overlap with morality and thus citizens are morally required to obey them. Also, if persons have a natural right to enforce these duties on others, then governments themselves (as "sets of persons") must have the same natural right, even if the absence of political obligation deprives them of a civil right. In this case coercion arises as part of a non-political context and is not seen to be an exclusive function of government [18]. On the whole, the arguments of this paragraph can be understood better within the context of Lockean political philosophy and its account of natural rights and duties, to which Simmons adheres [18,19]. The present argument in particular derives from Locke's doctrine of the natural right to punish [24]. Furthermore, we have a "natural duty of justice" to support institutions that exhibit certain qualities, such as benevolence or the promotion of happiness [18]. Such merits relate to "dimensions of 'justice'" that might counterbalance a government's "coercion without right," thus constituting grounds that provide government action with justification even where it functions without right [18]. Finally, there are weighty moral reasons for acting that do not have the status of duties but play a role in determining our judgments about action, for example when we would inconvenience others or disturb their plans when disobeying just because we are entitled to disobey [18]. Thus, Simmons concludes, the lack of political obligation does not entail that the state's right to command and be obeyed disappears in every particular case, nor that illegitimate states always act without justification in particular instances, nor further that we have a conclusive right to disrupt their functioning and oppose their laws [22]. All the reasons just discussed limit instances of disobedience and encourage support, even though the proof of no general political obligation shows that there are areas where the state is not entitled to require our support. Governmental action for preventing harmless private conduct, laws enforcing conduct that serves the protection of the state, and those imposing payments that finance government operations constitute examples of the latter [19]. These considerations bring us to Simmons' third argument. For the following presentation of this argument [23]. Simmons claims (a) that political obligation concerns only one area of justification of political institutions, namely their right to rule and its correlative obligations, and (b) that this needs to be assessed in terms of a particular relationship created on the basis of significant elements of specific interaction between governments and each of their citizens. I am in agreement with Simmons on the importance of such transactional evaluations for the problem of political obligation and the anarchist position. For one thing, they help avoid the derivation of political obligation from evaluations of political institutions that might cover for different claims about justification. For Simmons, the distinction between different areas of justification itself is of great importance. As he argues, the state may be defended for having a right to exist, which refers to a kind of justification other than state legitimacy (qua obligation) and which can be used to support the state independently of and despite the conclusions concerning its legitimacy. In this context, considerations about the general virtues of institutions play a primary role, generating the ones referred to as generic evaluations. Such evaluations can be seen to play a role within the problem of political obligation, but not to be primary and sufficient to generate this special political relationship. Given the implications of the problem of political obligation for further justifications of institutions that the anarchist criticism stresses, these evaluations can be reintroduced in the political debate and be useful.
At this point, considerations about the general qualities and accomplishments of institutions can be seen as the elements that justify the general moral duties to comply with political institutions discussed in the preceding paragraphs. In the hands of the anarchist theorist, they become reasons for showing that the anarchist perspective does not dictate widespread disobedience, retaining its critical value within a context that carefully separates various moral assessments and allows evaluation to take multiple directions and create various avenues of support. Significantly, critical philosophical anarchists can use these reasons to distinguish good from bad governments, contrary to the accusation that their view implies that all illegitimate states are morally equal and should be treated in the same way [17,19,22]. Another "dimension of evaluation" that is distinguished in this context is that states may sometimes "act with justification," i.e., have some of their particular actions or policies justified on moral grounds, even when they lack the justification to exist and the legitimacy to rule [23]. For example, a thoroughly non-egalitarian government is justified in prohibiting murder even though it neither is virtuous to merit support nor has a right to issue directions and use coercion to back them up. The separation of at least three dimensions of evaluation of political institutions corresponds to the different classes of moral reasons for complying with them, functioning according to the rationale of "weak" anarchism and of a "balance-of-reasons" approach. The combination of these aspects allows for great flexibility in the political debate. Thus certain critical philosophical anarchists think this combination characterizes the anarchist perspective, offering it in defense of their claim that their view does not have dramatically counterintuitive implications.
However, I think that we need to examine the resulting anarchist position more carefully. Simmons is right to claim that the anarchist conclusion about political obligation provides reasons for action that can be overridden in light of other serious moral considerations. I also concede that the lack of political obligation does not directly challenge the existence of institutions, given the independent function of generic evaluations of them and the indispensable moral force of such evaluations. Moreover, I find the idea that philosophical anarchism does not entail widespread disobedience and chaos a legitimate conclusion of these arguments, compatible with my support of the positive contribution of critical philosophical anarchism. The title of Simmons's book On the Edge of Anarchy has the positive meaning that this is exactly where we should be, that anarchism is not something we have to escape, that the edge of anarchy is not the verge of chaos -and it is on this idea that his work concludes. Nevertheless, I find this conclusion compatible with a radicalism that is central in the contribution of philosophical anarchism, a radicalism that Simmons's line of argument seems to neglect. The distinctive role of the anarchist is not to distinguish good from bad governments, nor to merely put limits on political institutions. Its edge lies somewhere else, and this is what I argue below.
To support my view, I begin with an estimation of specific claims made by Simmons in relation to the preceding account. Then I will follow this estimation with an account of my broader departure from his position.

Specific arguments against Simmons: Simmons closes his book
On the Edge of Anarchy with an account of the wrongs done to us by illegitimate yet benevolent states, given the non-existence of any special political obligation, deriving his Lockean anarchist conclusions about the position individuals should take with regard to this phenomenon [19]. At one point he confirms that "most of us in the 'free world' are in Lockean terms just persons in the state of nature (simpliciter), subjected by our governments to a variety of (usually) relatively minor, but frighteningly regular, wrongful acts and policies [19]. This explains why "good governments might merit our support, but they are not entitled to require it (without our free consent)" [19]. Later he suggests that certain "moral facts [that oblige us independently of any special legal requirement to obey, facts as those discussed above], plus considerations of simple prudence (i.e., our interest in avoiding legal punishment), seem to dictate that moderately good governments, which violate our rights only in ways such governments typically do, ought not to be resisted in ways that threaten to destroy them or to replace them with distinctly inferior alternatives" [19]. He then concludes that, "in the world of illegitimate states that will continue, moral persons must cast off their childhood lessons in good citizenship, and proceed by selectively supporting or opposing their governments' actions and policies solely according to the particular moral standing of each governmental move" [19]. Simmons offers these points as representative of the critical philosophical anarchist position, which comes out of the debate on political obligation with the message that we should adopt a more skeptical attitude towards existing governmentsan attitude, however, that does not involve rejecting them altogether, but focuses rather on the quality of their particular functions. Yet Simmons is missing something. First, his account of our moral obligations is determined by his advocacy of Locke's philosophy, which is not a view that someone, including an anarchist, has to be committed to. Even if his idea of moral responsibilities is an acceptable one, however, the problem with his approach still remains. With his first claim cited above, he dismisses as minor wrongs done to us by governments that are, nevertheless, frighteningly regular. He then supplements this claim with his further suggestion that good governments that violate our rights in ways that such governments typically do should not be dangerously resisted. But these statements constitute a very incomplete representation of the anarchist criticism that preceded them and its results. Indeed, this is obvious in the very form they take: how can a politically informed and active person, or just a reasonable one, let alone an anarchist, consider as minor violations acts and policies that are at the expense of the individuals concerned in a frighteningly regular manner? What is the habit of compliance if not such an untroubled acceptance of continual violations? How can a critical approach to political institutions accept their violations as those typically committed by such institutions? Is this not an unquestioning concession to the existing status quo rather than a challenge to it, as misguided as the romanticizing view of the state? Simmons's relevant appeal to prudence and to the dangers of possible destruction of good illegitimate governments and their replacement by inferior alternatives is a concession concealed by an air of political realism, rather than a reasonable pragmatic realization. Furthermore, these claims contradict the concluding demand for a critical and selective approach to governments, based on the quality of their particular policies. This attitude involves much more reflection and scrutiny, as well as abstraction from existing determinations, than a tendency from the very beginning to make concessions to them involves.
On the whole, in dismissing the importance of certain facts about governments that the anarchist view brings to the fore, Simmons commits philosophical anarchism to a view much less radical than really entailed by the anarchist perspective. To be sure, to believe that for specific instances minor violations of rights are better than major ones, and also that it is not bad to accept minor violations of rights, is reasonable. But when these views are used as indications of an approach to the very problem with authority, they render an account that is incomplete and misguiding. The problem is that from such a perspective they seem to suggest that it is fine if our relations to the state are declining a bit and that the whole issue is to establish a minimal state. Simmons's statements lead him to this approach.
A more general departure from Simmons's approach: My following arguments for the significance of the question of obligation, for a departure of the anarchist view from supporting the minimal state and, and for the role of the anarchist ideal of legitimacy are meant to demonstrate where my approach differs fundamentally from Simmons's and where I find his approach to be more generally inadequate.
The Significance of the Question of Political Obligation. When the anarchist says that we cannot ground the state on voluntariness, justice, reciprocity and association, he reveals a gap in the justifications of the state. This is his message derived from the debate on political obligation: we do not participate in creating and managing the state and there are no ethical concerns that arise distinctively from it as such. This makes the state illegitimate in terms of political obligation, although it is not a positive proof of its illegitimacy in general: we have no special ethical relationship to the state, although for it to be illegitimate, or wholly unjustified, we need to show that it does something wrong. Thus, Simmons thinks that this conclusion is perfectly compatible with keeping the state. It encourages a critical outlook toward it and an independent approach to our obligations to others, perspectives that nevertheless continue to be expressed within a framework widely determined by the state: the state can have the right to impose specific duties, it should be supported when it serves ethical duties, and it can very well continue to exist in a justified manner (when it is a good rather than a bad government). This view stresses the importance of differences between various ways to evaluate political institutions: in terms of their existence, in terms of their legitimacy qua obligation, and in terms of specific obligations that these institutions might be justified to impose even when they are neither justified nor obligatory. It also asserts that political obligation is just one criterion of evaluation. It is the latter claim that I challenge in my approach to critical philosophical anarchism. This paper accepts that the difference between kinds of evaluation of the state matters, and so, in this light, the state can exist even when there is no political obligation. Yet although the separation between different kinds of evaluation is central in the anarchist perspective, its value does not lie in permitting different kinds of justification to proceed independently of one another. It rather lies in helping us see the force of the considerations provided for and against political institutions in each case by drawing attention to the elements that characterize primarily each of these considerations, according to the issue we want to examine. For example, it reminds us to look for morally significant features of specific interaction as the elements relevant for generating the particular relationship characteristic of the problem of political obligation, instead of trying to derive argumentative force from general qualities of institutions, which play a secondary role in creating such a relationship. This does not mean that the different avenues for the justification of political institutions do not affect each other considerably.
The result of the anarchist criticism is not that political obligation is just one evaluation among others with no distinctive effect on the justification of the state. I argue instead that, by showing the state to provide no distinctive ethical concerns, to be based on no special relationship of political obligation, the anarchist uncovers a very serious gap: lacking political obligation is a defect in the very nature of the state. How can political institutions relate to their subjects if they lack political obligation? And how can they function and be distinguished from their alternatives if they lack such a relationship? That is, how can they be (considered as) permanent and exclusive (characteristics that are intrinsic to their political nature) if they have no right to command and be obeyed? These questions lead to more general doubts about political institutions. They make us examine their very nature and discover defects that have been neglected. The coercion that makes political institutions efficient is also a problem, and this needs to be addressed. It is at this point that the romanticized view of state-order starts to look out of place. In this way, the challenge to political obligation is not just a restricted and harmless criticism of the state, but a viewpoint from which the state is seen as a defect (even if not as totally evil) and something difficult to justify.
Political obligation cannot be derived simply from arguments for the existence of political institutions, for the latter might survive the debate on political obligation. Nevertheless, the character of institutions is a relevant and important condition for finally validating the right to rule. It is not likely that morally unacceptable (e.g., extremely bad) governments would allow their right to rule to be valid even if the required specific transactions could apply to them. As this paper supports, the criterion of the moral quality of institutions is quite important in the debate on political obligation. While the considerations and results of the debate on political obligation crucially affect further defenses of political constraints. The particular and actual relationships required for political obligation to exist are not indispensable elements for deciding the moral value of institutions more generally, but the absence of such an obligation constitutes a serious gap in their status. The different view of our position in society that the anarchist subversive campaign creates with the removal of the presumption in favor of political obligation constitutes a serious change for social life in the presence of political institutions. But further, the absence of this central relationship that characterizes their status as political raises doubts about the very plausibility of their validity, function and viability. Even more importantly, the difficulty with proving political obligation that the anarchist criticism reveals, redirects the debate to the deeper concerns that need to determine our approach to political institutions. I argue that the anarchist position on the problem of political obligation brings back a more fundamental question that underlies every approach to political institutions. The perspective that every theorist needs to adopt is one characterized by the question whether political institutions should exist at all. The anarchist indication that political obligation cannot be established by appeal to the general merits of institutions, the basis required for it making it a matter of continual justification, leads to a wider consideration about whether appeal to general virtues is enough to motivate constraints and whether it motivates them once and for all. For such evaluations to be effective they need to apply within the background set by the fundamental question that the anarchist brings to the fore, and such a question makes the demand for justification constant. This clarifies the complexity of avenues towards a defense of political institutions, indicating what the proper way of using their merits to support them is and how difficult this may be. Subsequently, in the light of the debate on political obligation and its results as affected by the anarchist criticism, every attempt at justification is reformulated and the task of justification becomes harder.
This way, we can claim that indirectly the justification of political obligation has a significant effect on other dimensions of the justification of political institutions and thus on their overall justification. Ultimately, the outlook of every theorist and every person is determined by the demand that, rather than considering the merits of political institutions on the basis of an assumption that we need them and desire them, we start to ask whether we should need them at all, appealing to their merits in the light of this question. This is the anarchist perspective. Moreover, the way it is activated within the debate on political obligation makes that problem, if not decisive for the content of other kinds of defense, still totally decisive for their structure and force -decisive in the way it leads to a proper reformulation of the considerations applying to them and a correspondingly different estimation of their input. The skeptical approach and a careful weighing of the independent moral grounds that bear on action in different political circumstances that the philosophical anarchist recommends should be seen in this light.
On the whole, we can see a parallel between the position that the anarchist establishes within the debate on political obligation and the one advanced more generally with regard to political institutions. The removal of a presumption in favor of obedience involved in the subversive campaign of philosophical anarchism corresponds to a removal, through the fundamental question that the anarchist perspective raises, of a presumption in favor of the existence of political institutions. The fundamental concern becomes the very possibility of political obligation, or the very possibility of political institutions. The subversive campaign of philosophical anarchism within the debate on political obligation also necessitates a different outlook on our position in political society, which corresponds to a different outlook with regard to the status and stability of political institutions. Again, through the fundamental question that the anarchist criticism brings to the fore and the demand for justification that this raises, their existence is not taken for granted, their desirability becoming a matter of constant justification.
So although the anarchist position does not privilege one specific avenue of justification to the detriment of all other avenues, it nevertheless reintroduces an approach that constitutes a unified challenge to every avenue at the deepest level. Even if not obvious to its advocates, the perspective of critical philosophical anarchism carries with it and remains faithful to the classical position of anarchism that there is nothing lovable about external constraints and that the state remains a problem. It is in this respect that philosophical anarchism remains thoroughly radical. It challenges social order from within [8], and helps us press for the respect of self-government and equality within social life. In the criticism of Raz above I stressed the importance of voluntary choice in securing freedom and equality within political society. In the hands of anarchism, the force of consent is expressed in a negative form: we could not possibly agree on having a state; if we were given the possibility to agree on it, we would not. But I agree with Raz that it is wrong to believe that political societies can become voluntary associations. Simmons is wrong to believe this [19]. It is because they cannot become such associations that the anarchist opposes them and sees them as destructive to proper relations. In light of this problem, we need to find another way of asserting self-government, equality, and legitimacy within political societies. The anarchist perspective and ideal of legitimacy work in this direction -should be seen in this new light. Correspondingly, by making the demand for justification harder, the anarchist position does not render the various expressions of the critical approach it recommends vehicles of an immediate and radical change. For example, "Even if we find that we can seldom justify or forbear the consequences of disobedience or substantial opposition, we can at least lobby for the elimination of those laws that interfere with harmless choices, impose needless regimentation of behavior and lifestyle, limit personal liberty without securing important social benefits [We can ask] questions about the moral merits or defects of the individual laws, actions, or policies of our governments" [19]. It surely renders them, though, more obviously representative of a serious challenge to actual societies and more effective as indications of active citizenship. More precisely, the demand to show in every instance that the existing political constraints respect the values they are held to help secure creates a central role in the political debate for the task assigned to individuals: to think carefully about the relevant moral reasons for support or rejection. In this sense, critical philosophical anarchism has a much stronger link with the political anarchist criticism of the state as an evil than first thought. Horton is right to claim that the challenge to political obligation can change much in our political lives, since a central part of our view of the political world is shown to be a myth [25]. In the end we have moral reasons to be more independent in our reasoning about social behavior and also to develop non-political forms of solutions to social problems, or, at least, to understand their intuitiveness. Yet this view does not adopt the political anarchist demand for the removal of the state, not as the initial and primary anarchist goal anyway. But then, what does it help change?
The anarchist does not provide a broad criticism concerning different variables, which would have no strong overall impact. Rather, the anarchist criticism reflects a unified demand for justification that affects our perspective on political institutions. And it is not the case, as some philosophical anarchists claim, that the anarchist position within the debate on political obligation and the attitude it recommends will not change much in practical terms. On the contrary, this position and attitude reflect a significant implication of the anarchist outlook with regard to real institutions and our lives in relation to them: a gradual but stable effort to make substantial, actual changes. By rendering a principled support of the nature of state authority problematic, the anarchist position invigorates the debate on political obligation; it presses for the exercise of our critical powers; it allows the construction of improved accounts of the authority of the state and of more satisfactory political arrangements, imposing more demanding criteria on wider evaluations of constraints; it also paints a picture of the political that stimulates an innovative conception of our political relationships and that inspires more informed views about the role of public institutions. These aspects suggest substantial changes to our conception of political societies and to our political lives, which might neither be based on an instant reconstruction through revolution nor be a desperate expression of our dissatisfaction with authority and the present character of public life.
Anarchism becomes a reminder that we should now recognize and apply what we are more qualified and seem in fact to recognize, to wit, that all relations of authority are in need of justification and that the proper approach to political authority is one that regards its scope as limited on the basis of considerations of quality. This is the way the anarchist critique helps remove the habit of compliance. More importantly, the anarchist position remains consistent. Its fundamental concern is still not to establish limits on political domination. It rather remains the claim that domination is always problematic. All this involves a proposal that is as radical as it is valuable: it testifies that anarchism suggests and remains the continual source of a move towards a fundamental reconstruction of our social relationships and lives. The possibility of such a reconstruction is still to be proven and its realization is necessarily gradual. Yet it remains a desirable end and an alternative worthy of taking its place in our moral and political history.
Departing from the Minimal State. But, how do the anarchist perspective and its accompanying ideal of legitimacy differ from already existing views for and against political institutions? The anarchist message is that we do not have a comprehensive argument for political obligation and that there are good reasons why we cannot have such an argument. This is a way of understanding the gap in justification instead of trying to fill it: it prepares a perspective for, rather than a ground for, obligation. Beginning from the gap created from the absence of political obligation, critical philosophical anarchism suggests that we leave aside the attempt to answer the question of obligation, that political obligation concerns an everlasting effort of justification that is not susceptible to a final resolution. Indeed, the anarchist skepticism is not about providing such a justification for the state. Rather it is about bracketing the question of obligation and concentrates on something else, which is nevertheless motivated by the difficulty of this question. Instead of attacking the existence of the state and trying to roll it back, the critical philosophical anarchist claims that, whether we love it or not, when we have it, the state is not a matter of magnitude, or quantity, but rather of quality. To be critical toward the state in the way that the philosophical anarchist suggests means to see whether it matches ethical concerns that we need, whether, that is, as it exists, it is acceptable in view of justifiable claims we have toward one another. I call this "the quality thesis" on the state.
Such an approach departs both from romanticizing accounts of the state and from those against the welfare state. With regard to the latter, it is important to see how the anarchist approach differs from defenses of the minimal state. In contrast with what Simmons's account suggests, for the critical philosophical anarchist there is no way of filling the gap that the lack of voluntary participation creates by diminishing the tasks of the state to the minimal. A state that does not support education and healthcare and does not provide distribution and general protection, but merely polices property, is more unjustifiable than a full state. Such a state is even further from a condition of liberty without inequality that all forms of anarchism desire: instead of being an establishment that protects individuals without undermining their equal right to self-government and participation, it gives liberty to the few, whose interests it protects and perpetuates, at the expense of the many, who remain unsatisfied and unequal. It cultivates division and conflict by supporting a society where competition and social discrimination thrive. Thus, in fact it is very far from what is supposed to be the primary function of the state, for which it is claimed to deserve justification in the first place, namely to serve its citizens.
The Anarchist Ideal of Legitimacy. This guides the perspective on the state that the anarchist adopts. In its different forms, this ideal represents what a society characterized by appropriate relations between persons would be like. And because the state is not such a society, i.e., it is not the ideal, as the anarchist criticism of political obligation shows, what it can do instead is approximate it -to prove in every instance that its functions are compatible with the moral criteria of the ideal. This does not mean that it proves to have political obligation, but rather that, in the absence of it, this is the only way of ensuring active participation, not in the making of the state, but in the process of evaluating its functions and in being able to restrict it to what it can justifiably demand from us. This is what results from the evaluation in terms of political obligation and what makes that evaluation more important than Simmons thinks, not merely one among other evaluations that is moderate and limited, but rather the basis for a substantive transformation of our view of the state. It shows that the state has gone too far and is taking too much from us, that it offends self-government, equality, and proper relations in the name of a good that it is not. But instead of either overthrowing it or trying to fill the (unbridgeable) gap of justification that its defenders have attempted, we can become those who determine in a justifiable way where and how it should stop: when its functions are of a character that is justifiable to us and not at our expense, when by protecting through coercion it does not overdo the latter at the expense of the former. The perspective that the philosophical anarchist offers shows us a way of being entirely distrustful of the state while at the same time accepting the welfare state. This view does not depart from my claim that critical philosophical anarchism is still linked with political anarchism: at every instance there is the possibility of becoming dissatisfied with the state in terms of the ideal of legitimacy, and this endangers its existence. In light of the results of the debate on political obligation that the anarchist brings to the fore, the undesirability of illegitimate constraints becomes categorical. And the ideal of legitimacy becomes a constant guardian against abuses of the state, not by providing a form of consent, but by testing continually the quality of state functions. The anarchist thus reminds us that the dissatisfaction and lack of patience that we feel toward the state in times of crisis should be the characteristic attitude and the starting point for us to view our social responsibilities. This is the way for the state to be an instrument at our service, compatible with individual self-government and equal positive participation.

Conclusion
We therefore have to be committed neither to an inevitable acceptance of the state nor to a complete rejection of it. The anarchist approach offers an option that has been neglected and that is the most reasonable one: we can stay within the state and participate in advancing the social aims it is meant to serve, and yet always keep an eye on the way its dominative tendencies and its coercion might overstep its initial task.
But having adopted this view, the philosophical anarchist has further to answer the question of how we fulfil our obligations to others without the state within a background where the state exists. How do we do without the state apparatus in a situation where we do not have perfect abundance? That is, how do we manage independently of the state to coordinate and cooperate towards an effective and fair satisfaction of our needs in a world where goods are not abundantly available? This is a legitimate question to ask the anarchist who does not insist on removing the state and yet finds its defects in terms of illegitimacy to be a good reason for independence from it. He has to answer this question in order to prove that he remains an anarchist and can convince others of the merits of his position.