Philosophical anarchism and the paradox of politics

Philosophical anarchism and the paradox of founding

Philosophical anarchism—as opposed to the “political” anarchism we find in writes such as Bakunin, Kropotkin, Goldman, and others—is the position that maintains that either there are, a priori, no possible, or there are or have never been, a posteriori, any legitimate states. Robert Paul Wolff’s work is an example of the “a priori” argument for philosophical anarchism. Wolff argues that the obligation to be an autonomous individual is incompatible with political obligations to a legitimate state because to be autonomous is to act only on those reasons that one can affirm as good reasons. However, to obey an authority is to “surrender” one’s judgment and obey another simply because of their authoritative position. Thus, one cannot autonomously obey authority. From the moment the argument emerged in book form it was quickly challenged and for many, including some anarchists, the a priori approach is a dead end (Wolff, 1998). ⁴

On the other hand, “a posteriori” philosophical anarchism has a number of important defenders in part because it accepts, in principle, a legitimate state, but then shows that there are no existing legitimate states (and perhaps never have been and never will be). Joseph Raz argues that “regarding those who consent to authority or respect the law, a reasonably just government has authority, the right to command. Not everyone consents or respects the law. Those who do not consent have no obligation to obey, but they owe the government the lesser obligation to support just institutions and, of course, they should comply with the law whenever there are good independent reasons for doing so (Raz, 1981, 2009). Others have argued in various ways to Raz’s conclusion, but it is the conclusion that there is no obligation to obey the law and/or the conclusion that there are (or cannot ever be) any legitimate states that ties philosophical anarchists together. To be sure, the debate between anarchists and those who argue that states can be legitimate continues. ⁵ My interest is not in that debate itself, but in how anarchists deal with the consequences that follow from their apparently radical conclusions. The philosophical anarchist I will be discussing in this essay, A. John Simmons, makes, to my mind, the most persuasive case that vis-à-vis most (but not necessarily all) residents within modern states, the state is illegitimate.

Simmons’ basic argument is straightforward enough: after surveying various grounds for the legitimacy of the state other than consent—including the duty of fair play; tacit consent; gratitude; a “natural duty of justice”; and more recently “hypothetical consent” positions of the Rawlsian and Habermasian variety—Simmons (1979, 2001) argues that only express consent given under specified conditions can generate political obligations. Alternative arguments for the existence of political obligations and/or state legitimacy are shown to: (1) collapse into an express consent position; (2) and/or are shown to fail in explaining why we must obey our own state rather than all just states; (3) and/or fail to be sufficiently general that they generate political obligations for the majority of citizens.

Take the example of “fair play”, argued for by Hart (1955), Rawls (1999) (for a time anyway), and more recently George Klosko (1987). The fair play view—if we neglect for the moment differences amongst its expositors—holds that insofar as members of a community understood as a cooperative venture receive the benefits of that cooperation, they are obligated to do their part, to obey. A legitimate state, then, would be one in which individual members of that state accept the benefits of life together in that state and are obligated thereby to obey the state, to fairly do their part to maintain and advance the cooperative scheme. Insofar as in modern states the benefits are distributed widely (if not always evenly and justly), then most citizens do indeed have political obligations.

The problem, Simmons argues, is that this requires both that individuals in a state view political community as a “cooperative scheme” and that they truly accept the benefits of that scheme rather than merely receiving them. Simmons (1979: 139) claims—to my mind rightly—that it is highly unlikely that many citizens both accept the benefits in the right way and think of the state as a cooperative scheme, and thus have obligations to that state. ⁶ Related justificatory failures can be found in the other proposed grounds for state legitimacy and political obligation. Insofar as very few people within any state have ever accepted benefits in the right way, or given their express consent, or have an obligation to their own state because of a natural duty to obey just institutions—and granting as Simmons (1979: 192) does that political obligations correlate with state legitimacy—then all existing states, vis-à-vis the vast majority of citizens, are illegitimate.

The strength of Simmons’ position rests upon the plausibility of the assumptions he sees as undergirding consent theory, the intuitive power of the idea that consent (alongside promising) is the firmest foundation for the creation of obligations and authority, and the patently obvious fact that most of us have never expressly consented to any state. The assumptions are: (1) human beings are naturally free (within the bounds of natural law); (2) that only a voluntary act can generate political obligations, which are always artificial; (3) consent “protects” us from the State by ensuring that we “assume” the duties of citizenship knowingly and willingly; and (4) this helps to ensure that the State works for the interest of its citizens (Simmons, 1979: 61–69). Simmons (1979: 77, emphasis added) defines consent as “the according to another by the consentor of a special right to act within areas within which only the consentor is normally free to act; this is accomplished through a suitable expression of the consentor’s intention to enter such a transaction and involves the assuming of a special obligation not to interfere with the exercise of the right accorded”. While Simmons avoids a general discussion of the necessary conditions for felicitous consenting (which he likens to those of promising, as articulated by Searle in Speech Acts), he does make special mention of the intentional (and/or knowing) and voluntary nature of acts of consent. Consent cannot be given in “excitable” situations, i.e. under threat or duress, and tacit consent may only be construed as a “suitable expression” of consent if it meets certain strict requirements (Simmons, 1979: 81). Specifically, tacit consent cannot be understood as the performance of acts that “imply consent”; rather, it must be a non-verbal expression of consent in order to be consent. Thus, for consent to generate political obligations one must expressly consent even if that consent is expressed by silence. However, an extremely small number of individuals have ever expressly consented to their state, whether in words or silently and the consequence of this is that most states are illegitimate and most of us have no political obligations to obey the state. As we will see below, this doesn’t mean we have no other reasons to obey the State and follow its laws (as well as the right to punish those who break the law).

Simmons’ own commitment to a version of natural rights and natural duties/law is explicit, and his account of consent relies crucially on concepts central to speech–act theory: intention, context, express utterance, and so on. This has two consequences for his argument. First, a significant amount of weight is placed on the moral philosophical underpinnings of natural rights and natural duties (in Simmons’ case, as we will see, a version of moral realism). Second, the theorization of consent as a performative speech–act is susceptible to criticisms of the Austinian and Searlean account of performatives. If consent as a performative speech act can never, in principle, be “accomplished”, then even express consent given within a proper context is incapable of sufficiently generating political obligations (Derrida, 1988). I will avoid discussion of this second consequence, but suffice it to say that for Simmons, there is nothing, in principle, philosophically or practically difficult about consenting: it just rarely happens.

The in-principle possibility of fully legitimizing a state is contested by Connolly, Derrida, and Honig. In different but related ways they all attempt to show the existence of a “paradox of founding” or a “paradox of politics” (Derrida, 2002; Honig, 2007, 2009). ⁷ Given the existence of this paradox, all three argue that any attempt to found, once and for all, the legitimacy of the State or of law, is doomed to failure. I focus here on Connolly, however, for two reasons. First, Connolly and Honig share a conception of the paradox of politics, but Honig’s (2007) use of the paradox—unlike Connolly’s—is largely directed towards a criticism of Habermasian conceptions of the legitimacy of law within modern social-democratic states. Honig’s argument, in other words, concerns less a paradox of politics “as such” and more a paradox of liberal- and or social-democratic politics. Secondly, Derrida’s “Force of Law”—the essay in which he argues for the inescapable a-legitimacy of law on the basis of an ungroundable performative force of language—is connected to his earlier appropriation of Austin and the debate that ensued between Derrida and Searle, and that raises an entirely different—if not unconnected—set of concerns.

Connolly—drawing on Rousseau’s chapter “On the Legislator” in The Social Contract as well as a short essay by Paul Ricoeur, “The Political Paradox”—actually argues for two (in my view distinct) paradoxes that trouble State legitimacy. First, Rousseau (1978) argues in “On the Legislator” that good laws create good citizens. However, the adoption of good laws requires citizens to already be good, i.e. to appreciate in the proper way what is good law and to have the motivations and desires to only vote for good laws. Any newly formed polity therefore faces a real problem precisely when it comes to adopting its first laws: the General Will and the citizens who ought to opt for the General Will over against their own “private” will are often unsure of the best course of action (because the General Will is blind). Presumably, good citizens are trained through good laws to recognize good laws when they see them. However, even if citizens know which laws to select, they are often pulled by their own private wills towards adopting bad laws and convincing other citizens to do the same. Again, it is good law that generates the habits of mind and body necessary to choose good laws. Whether it is a matter of recognizing good laws or having the proper motivations to opt for good laws over against one’s own private will, good laws provide the necessary education. But adopting good laws depends on the knowledge and good habits produced by the law: hence, the paradox (Connolly, 1995). If the legitimacy of law rests, at least in part, on the normative legitimacy of individual laws—i.e. a post-positivist concept of law in which a law must be normatively justifiable if it is to be a legitimate law—then this paradox throws a wrench in the workings of liberal-democratic states. Rousseau’s own answer—that we require a Legislator to “persuade without convincing” by ventriloquizing the divine—appears to undermine democratic sovereignty, and thus we are left with an irresolvable paradox.

The second paradox identified by Connolly is different. Along with other philosophers who ground state legitimacy in an act of consent, Connolly (1995: 139) claims that for Ricoeur all political acts must “reflect” a previous act of consent or agreement. The problem Ricoeur raises, however, is that the nature of the social contract and of the consent that institutes it is necessarily virtual, ideal, an expression of the Truth of polity (and all of this in an unstated but broadly Husserlian framework of the relation between Ideality and Truth). But all polity includes politics, and politics is the realm of contingent historical decisions, concrete instantiations of power, the use of violence under the cover of legitimacy but in violation of the ends of political community, and so on (Ricoeur, 1992). These political acts rely for their legitimacy on the legitimating force of a virtual, ideal act of consent, but precisely because they are acts of real, contingent politics, there is a necessary disconnect that undermines the legitimacy of the state. Thus, if the act is political then it necessarily fails to be fully legitimate. For Connolly, translating this idea somewhat, a political act, if it is to be political and not merely administrative or executive, cannot simply follow from a previous act of consent, implying that a political act is always (if not in precisely Schmittian terms), a “decision”, an exceptional act that can be guided but not controlled by rules, standards, and judgments. Moreover, Connolly continues, because of Ricoeur’s phenomenological reading of consent, the act of consent is itself accessible only in the form of memory and reflection of a “virtual” act, i.e. an act that has never taken place. Actual consent, “pure” consent, never happens. Thus, in addition to the Rousseauian paradox we have a Ricoeurian paradox, the former showing the disjunct between normative and democratic legitimacy, and the latter describing a temporal and, as it were ontological, paradox in which political acts never acquire their full legitimacy because their justification always requires a reference to a prior legitimizing act that has never taken place and necessarily exists virtually and as Ideal.

The plausibility of Connolly’s argument rests—just as in Simmons—on a set of assumptions and/or defensible positions. First, all consent thinkers—as far back as Hobbes and Locke—are aware of the criticism that Simmons grants as obvious: consent rarely, if ever, happens (especially at the founding of political communities). ⁸ Like Raz, Ricoeur, and Connolly account for the idea of consent and its importance (as well as its limitations) in terms of, to use Raz’s (1981: 128) terminology, its “expressiveness”. The social contract and consent express a vision of what political community ought to be, what would make a political community just and legitimate, and sometimes consent can express one’s identification with their society. Unlike Raz, however, Connolly and Ricoeur are not interested in how an act of individual consent can express a sense of belonging to a political community, but rather in how a politics of consent employs a never pure or real consensual moment to legitimize political decisions that often exclude, discriminate, employ unjustified violence, and so on. While many would no doubt agree that claims to legitimacy can mask illegitimate actions, Connolly goes further and argues that no political act can ever be fully legitimate because a legitimating moment is necessarily virtual, posited retrospectively. This claim is surely contestable, but no more contestable—which is to say, no more implausible—than the hypothetical (read: virtual) consent positions Rawls and sometimes Habermas endorse.

Similarly, the Rousseauian paradox rests upon the assumption—common to natural law theorists as well as Habermas and other Kantians—that the legitimacy of a law rests at least in part upon its normative acceptability. It also shares with Habermas the idea that laws are educative and help to create a set of cultural and political practices that enable the adoption of further good laws. ⁹ But this creates a tension between democratic sovereignty (whatever the people choose is legitimate) and a just legal order (only those laws which are normatively justifiable are legitimate). Connolly differs from Habermas and others precisely because he denies that any final reconciliation between the two poles of modern liberal-democracy is possible. Moreover, Connolly, Honig, and Derrida are thankful for the paradox of politics/founding because it leaves open spaces of contestation that would otherwise be closed by a perfect fit between normative and democratic legitimacy.

Although I have left out a great deal of argumentation, the difference in approach and philosophical justification for philosophical anarchism (as found in Simmons) and the paradox of founding (as found in Connolly) should be clear. Simmons’ argument rests on the assumption of an unproblematic account of consent as a speech-act and a version of natural law and natural rights, and it ends with the conclusion that state legitimacy is finally a matter of an individual’s “transactional” relation to the State: if I have consented, then I am obligated and the state is legitimate vis-à-vis me. If I have not consented, then I have no political obligations and the state is illegitimate (at least vis-à-vis me). Connolly locates the reasons for the ultimate impossibility of fully grounding state legitimacy in a tension internal to democracies between the sovereignty of the people and the normative justification of the laws as well as in a temporal and ontological paradox wherein any particular act of the State must rely on a prior, but absent, founding act for its justification. What both philosophical anarchists and “paradox of founding” theorists share, however, is the argument that existing States are very likely a- or illegitimate, and that those who defend the legitimacy of the State need to respond to these provocative, plausible, and to my mind persuasive arguments. ¹⁰

The problem of state violence

For Simmons, as for theorists of state legitimacy generally, one central reason why we need the state to be legitimate is to distinguish the force it uses in executing laws and punishing offenders from the violence used by criminals and others. The stability of the legitimate force/illegitimate violence distinction is one of the central reasons why we desire (a theory of) state legitimacy. For those who think the state can be legitimate, on whatever grounds, the problem of the morality of state violence is, in principle, easily solved. The legitimacy of state violence, once secured, does not mean that there are no moral problems left for particular practices of state violence. The moral legitimacy of the death penalty, to take just one example, does not follow from the state’s legitimate use of force. Moreover, that the state has a general (and usually exclusive) right to use force doesn’t mean that we cannot criticize abuses of state violence; on the contrary, we can find in theories of state legitimacy moral grounds for precisely such criticism.

For Simmons and other philosophical anarchists, however, the problem of state violence is not so easily solved. For, having denied the legitimacy of the state, Simmons appears to leave the State’s right to use force ungrounded, morally illegitimate. Thomas Senor, in a reply to Simmons, argues that “if in order to do anything that can correctly be called ‘punishing,’ the punisher must be in authority over the one being punished, then governments do not punish at all, on Simmons's account, and so their fining and jailing law breakers cannot be just. If, on the other hand, one can punish even if one is not in a position of authority, then the governmental punisher must be behaving unjustly, since justice requires that the punisher be in a position to punish if the punishing is to be legitimate” (Senor, 1987). ¹¹ Simmons is in the unfortunate position of having to concede that when the state punishes, it is only “punishing”, insofar as it is part of the meaning of punishment that the one who punishes is authorized to do so. The state, lacking such authority, is engaging in a paradigm case of injustice.

Simmons’ (1987, 1992 especially the third chapter) response to these worries is to turn to Locke’s (1988) “strange doctrine”, i.e. the argument in the Second Treatise that there is a “natural right to punish”. The complete alienation of this right is part of the transfer of rights to the political society and then to the government that takes place when individuals consent to form or join a political society. The state’s right to punish, then, like all of its political rights, comes from the act of consent. However, even if the state lacks consensual authority, the state is not necessarily without the right to use force (Simmons, 1987: 276). Insofar as the natural right to punish is natural, hence equally held by all individuals and retained unless one expressly transfers it, any individual can rightly punish another person who has broken the law of nature, including members of governments.

The natural right to punish is strange, but Simmons attempts to make its existence plausible. He does so by arguing that we can reconstruct Locke’s position to be that acts that violate the law of nature are punishable by others because the criminal forfeits some of their rights against attacks on their person for a period of time. The forfeiture of rights is the result of something like a “fair play” view of life under the law of nature, in which “one’s status under the rules depends on respecting them. Protection under the rules is contingent on our obeying them; any rights the rules may define are guaranteed only to those who refrain from violating them” (Simmons, 1992: 153).

If this argument is persuasive not only as a reconstruction of Locke’s position but also as an account of the existence of a natural right to punish, then we have, first, a moral obligation to obey a number of civil laws, no matter how illegitimate, because many civil laws are also laws of nature and, further, we all have pre-political duties to, for example, support just institutions (Simmons, 2001: 112–118). Second, with the natural right to punish, any individual (and individuals as members of existing governments) can enforce these laws and obligations even if the state is not legitimate, so long as they punish justly and for the right reasons. The consequences of philosophical anarchism, then, for our actual political and moral lives are, while real, not that radical (Simmons, 2001: 112–118). Presumptions to obey certain laws, or pay certain taxes, may very well no longer exist; but moral obligations suffice, and the natural right to punish exists, to (at least potentially) ensure civil order.

What happens to the moral problem of state violence in Simmons’ account? Simmons acknowledges that punishment by illegitimate states is not legitimate; but he quickly—too quickly—forecloses further questions about the morality of state violence by referring the problem back to a largely undefended moral realism that serves as the ground of natural right to punish. For no matter what one makes of the reconstruction of a natural right to punish in Locke’s text, the crux of Simmons’ argument for the existence of such a right is the moral realism that underlies it. In the next section, I will look more closely at Simmons’ minimal defense of moral realism.

Connolly (1995: 41–74) evades the problem of the legitimacy of state violence even as he continuously and insightfully investigates the assemblages of affect, belief, institutionalization, and law that underpin a “desire to punish”. Drawing on Nietzsche, René Girard, and Foucault, Connolly argues that our practices of punishment in modern society might be inseparable from a desire for revenge linked to the existential conditions of human life, specifically its vulnerability to forces beyond our individual (and collective) control. The locus of the tensions within the politics and practices of punishment is the concept of “responsible agency”, despite the recognition by many philosophers—from Augustine to Kant and beyond—of a “divided will” and conflicts between rational freedom, animal desire, and a causally determinative world.

Like Bernard Williams (1981) and Thomas Nagel (1976)—who are keenly aware of how the existence of “moral luck” drives skeptical questions about the existence of a responsible agent—Connolly is attuned to the difficulties of sustaining theoretically a conception of the responsible actor that can withstand our knowledge that forces beyond human control shape in intimate and often unknowable ways what we do and why we do it. Connolly’s concern is not only philosophical but political, for he is more interested in how a questionable responsibility is repeatedly imposed on certain political subjects (often along racial, gendered, classed and sexual lines). For Connolly (1995: 46) “there are powerful cultural pressures, then, to obscure this tension between the social indispensability of responsibility and the problematical desert or merit of those to whom it is applied”. Punishment must be understood not only in consequentialist or deontological terms, but also as a complex political and ethical practice that manifests and reflexively intensifies affective as well as cognitive dispositions characterized by what Nietzsche called ressentiment (Connolly, 1995: 49–58).

However, Connolly’s analysis of the desire to punish leaves the question of the legitimacy of punishment, and of law and the state, unasked. This is important because Connolly acknowledges his recourse to a “deterrence” theory of punishment: “I continue to prize my security and mobility. But now a larger part of me thinks and (sometimes more precariously) feels that Carter [a young African-American in Baltimore accused and convicted of murdering a white man] should be imprisoned only to protect others from the fate suffered by his victims, and only as long as the probability is high that he will repeat such acts in the future” (Connolly, 1995: 72). For Connolly, a questionable punishment is a necessary, hence tragic, response to the actions of some individuals. The virtues of Connolly’s argument lie in its embrace of the messy world of desire, agency and revenge, a world that is resistant to the justifications of punishment we find in retributivism and consequentialism. But one can, and must, ask: is my “security and mobility” enough to make punishment morally legitimate, especially if the State is, at best, a-legitimate? Why rely on the state to punish? What authority do they have in these matters if, as the paradox of founding shows, such authority cannot be fully grounded? If, at least on occasion, we must put people in jail, do they not have the right to ask us what authority we have to do so, and do we not have a responsibility to provide such a justification?

Ten years after examining the desire to punish, Connolly argues—this time in conversation with Agamben’s conception of sovereignty—that although the paradox of founding one sees in Rousseau cannot be overcome, it can be better negotiated. In the course of this discussion he argues that when a “positional sovereign” such as a court decides, “We owe positional respect to the institution called upon to make authoritative decisions when the pressure of time compels them. That responsibility is met by a presumption of obedience to Court decisions and by public admission that democratic constitutionalism needs such a body in a non-parliamentary system. But to express partiality for democracy is to accept other responsibilities as well” (Connolly, 2005). These other responsibilities include active struggles to identify and publicize moments when sovereign authorities employ un- or anti-democratic ethoi such that sovereignty undermines the possibility of a pluralist democratic ethos and politics. Thus, “We meet our first responsibility to sovereignty by a presumption to obedience that might be overridden on some occasions; we meet the second by contesting publicly the partisanship against democracy exercised by the Gang of Five [i.e., the Supreme Court justices who decided Bush v. Gore]” (Connolly, 2005: 146).

What about cases in which judicial decisions employ state violence against convicted individuals? Do we have a presumption to obedience in such cases? Do criminals (Hobbes, for example, argues that we have a right to resist the sovereign when he tries to punish us)? The moral problem of state violence has not been adequately addressed in Connolly’s discussions of the paradox of founding, punishment, and presumptive obedience to courts. The question, given what we know about the desire to punish and the paradox of founding/sovereignty/politics is: why still support punishment? What makes punishment morally acceptable? What are the grounds of presumptive obedience as well as disobedience? Simmons turns to an account of natural law and rights that is problematic, but Connolly leaves the status of the legitimacy of law and its violence unclear.

Across the divide

At this point, then, we have seen two positions prominent in contemporary political thought that both put into question the legitimacy of the state and, presumably by extension, the moral legitimacy of state violence. And yet both positions, I have argued, too quickly solve or evade the problem of state violence: Simmons by recourse to a moral realism that is largely undefended; Connolly by appearing to endorse a “deterrence” justification of punishment as well as claiming a presumptive obligation to obey legal authorities without offering reasons for such an obligation. I hope to show now that when we push both Simmons’ and Connolly’s arguments we find that we have to, as it were, cross (or construct) a bridge over the divide that separates the analytic and continental traditions they inherit. In other words, internal limitations and lacunae in each text force us to raise questions normally asked only in the “opposed” traditions, and thus we must find our way into the opposed tradition in such a way that we can also find our way back (and again forth).

For Simmons, the loss of political legitimacy does not mean that individuals or states cannot punish violations of (moral) laws. There are good reasons, Simmons argues, in support of the claim that a natural right to punish exists. The problem is that the best defense Simmons provides, not of his reconstruction of the Lockean natural right to punish but of the moral theory upon which it rests, is that moral realism is not obviously wrong, is intelligible and, moreover, is a defensible if controversial position with contemporary moral philosophy. In defending his position against “natural rights skeptics”, Simmons argues that “the ‘common core’ of natural law doctrine seems to amount to little more than this: that there are universally binding (‘objectively valid’) moral rules, knowable by use of our natural faculties, which may be appealed to to evaluate and criticize … Natural law theory, in this sense, implies some form of value objectivism—a position that is controversial enough, of course, but one that is still within the mainstream of active theoretical debate.” For this reason, “the defender of natural rights is not obviously committed to any more than the existence of objective (not essentially conventional) moral rules defining rights (although the content of these rights might be conventionally determined). The standard (but not only) epistemological and metaphysical positions accompanying these views are admittedly realist in character … But since moral realism and objectivism are clearly still live issues in moral philosophy, neither of these commitments for the natural rights theorist is obviously damning” (Simmons, 1992: 103–105). ¹² No other justification for Simmons’ moral realism is given. The difficulty with this defense is that in any use of (state) violence there is a real need for moral justification because violence is morally questionable, perhaps even prima facie wrong. If so, then, as Simmons (2001: 124) himself argues, there is a call for justification because justification “is in large measure a ‘defensive’ concept, in that we ask for justifications against a background presumption of possible objection”. Not only is violence morally questionable; the state’s acting violently against another person or people without right is a paradigm of immoral, unjust action.

I would suggest that whenever we consider normative justifications of violence we must imagine them as addressed to, at the very least, the potential victim of violence. ¹³ In other words, we have to imagine ourselves justifying our violent acts to the object of that violence, not in an ideal speech situation that assumes a non-violent sphere of communicative action, but precisely in, to borrow a phrase from Robert Cover (1986), a field of pain and death.

If we begin by imagining the situation in which we have to justify violence to the potential object of violence, Simmons’ defense of the use of state violence to someone who has violated the (moral) law appears, ultimately, to be: “moral realism is a defensible and defended position within moral philosophy.” I hope this is not unfair to Simmons, nor do I think this answer, in itself, is “obviously damning.” But compare this defense to a possible defense in a state legitimate on consensual grounds. Such a justification might go as follows: “We all have a natural right to punish in the state of nature, and we are all bound by a natural moral law. When you and I formed this community we agreed to transfer that right to the State, and we also agreed to be bound by the laws that we, all together, authorized the State to make. We had and have your word, your consent, to engage in specific acts of violence, including what we are about to do to you.” Between “moral realism is a defensible position within moral philosophy” and “you consented …” there is a vast difference in justificatory force. The latter justification speaks directly to the individual person and their obligations and commitments and refers as well to the speaker’s obligations and commitments, in a specific set of circumstances. The former justification answers to nobody in particular (for it is a claim addressed to anyone interested in philosophy) and makes no reference to any obligations or commitments (save perhaps to a moral theory). Even if a defense of moral realism were provided, natural rights defended, etc. isn’t a sufficient rejoinder on the part of the potential object of violence merely to cite any number of arguments against moral realism, against natural rights, and against the legitimacy of the state; or at the very least, to claim (with equal validity) that non-cognitivism, for example, is a defensible moral theory too?

The justificatory gap described above is opened by the difference between a justification that refers to claims made by some moral philosophers and a justification that refers to the voluntary, intentional act of consent by the employer and object of state violence. Of course, one could always object that here two issues are being confused: the justification of a moral theory and the justification of a specific moral judgment and act. However, if my argument seems to obliterate the distinction between philosophical debate and concrete justifications of particular practices, that is precisely the point. If violence always stands in need of justification; and if there are no legitimate states; and if the only defense of state violence relies quite directly on a philosophical defense of moral realism in general and a theory of natural rights and laws in particular; then the argument for moral realism and a theory of natural rights ought, at the very least, actually be defended; to the object of state violence; in a field of pain and death.

Even if Simmons’ defense of moral realism and natural rights is forthcoming, it is—as I suggested—susceptible to an equally plausible and defensible argument drawn from anti-realist, non-cognitivist, or rights-skeptical positions. If there is a stalemate here, with a body on the line, then this raises the question: why still punish, why risk such grave injustice? I suspect that Simmons would agree—if he agreed with the line of criticism, I have been following—that his reference to moral realism is a “spade-turning” moment in his arguments about and within anarchism, the moment where he must make recourse at least in part to “intuitions” rather than arguments. If so, then we have good reason to ask “where do such intuitions comes from?” Are the intuitions moral and political philosophers rely on discoveries of Reason? Or are they historical, cultural, temporal, contingent? Do our moral intuitions develop naturally, or are they the products of specific practices in particular cultures, practices that produce certain modes of thinking, habits of the body, predispositions to specific judgments? ¹⁴ Do we punish (only) in order to prevent further crime and ensure societal stability, or because the act was wrong; or is there a “desire to punish” that exceeds such justifications and is developed in response to existential conditions of human life that can and have led to the emergence of resentful affective and cognitive dispositions which shape and fuel practices of punishment?

These questions, of course, are precisely questions most often asked and answered on the “other path”, the continental path. This is not to deny that a non-cognitivist or a moral relativist might ask and answer yes to some of these more commonly “continental” questions. However, insofar as continental political theorists tend to evade meta-ethical debates common in analytic moral philosophy, the point of asking questions about the lived, embodied, and culturally shaped practices of ethical and moral life is not to even raise, much less answer, questions such as “can moral judgments be true or false?” Rather, one can broadly suggest that thinkers such as Foucault and Connolly are interested in tracing the historical emergence, development and justificatory role of our moral intuitions and practices in order to show both what is gained and lost in the development of late modern political orders. They draw on these genealogies in order to argue for alternative ethical and political practices that may modify, help to realize the ideals of, or escape, contemporary liberal-democratic, capitalist societies.

To ask whether a desire to punish can be found in Simmons’ argument is not to judge his argument (morally) wrong. It is to suggest that even a legitimate state, consented to by all, might still be left with the question: are our practices of punishment, our uses of state violence, fully justified, fully moral, if the intuitions on which we rely are shaped by and contribute to a desire to punish that precedes and exceeds philosophical justification? Further, what if liberal democracy itself is shaped by the desire to punish, resting as it is on a conception of the human subject as a free, responsible, consenting, intentional—and thus punishable—actor? If, as Simmons argues, the vast, vast majority of us do not live in legitimate states, then the moral question of state violence, of punishment, is even more pressing, for now we are left not with consensual relations amongst citizens of a particular polity and between them and a state, but only a de facto set of connections between some individuals and between them and some institutions. Given Simmons’ denial of other grounds of legitimacy, it is unsurprising that we have to turn to the debates of moral philosophers in those moments where we have to justify the violence of the state. However, whatever the value of moral philosophy, it is difficult to imagine how it could, directly or indirectly, justify the use of violence under conditions of non-consensual relations so satisfactorily that we have good reasons to overcome our intuition that violence is prima facie wrong.

If difficulties in Simmons’ arguments force us to raise questions asked more often by continental theorists, Connolly’s arguments force us towards problems still encountered frequently in analytic political theory, e.g. in Simmons. Connolly claims that we have a presumptive obligation to obey the decisions of courts, i.e. to respect the legitimacy of law. This obligation is only presumptive, however: if, as sometimes happens, the court makes a clearly illegitimate (or anti-democratic, anti-pluralist, resentful and vengeful) decision, then we are not bound to respect such a decision, and indeed Connolly argues that we have a responsibility to criticize and politically challenge such decisions. But the question emerges: what is the basis for our obligation to obey the decisions of courts and—if I am not generalizing too much from this single claim—a prima facie obligation to obey the law? That there are any number of good reasons, both instrumental and moral, to obey the law is no doubt true; but that doesn’t mean I have an obligation to obey the law. A similar argument can be made vis-à-vis Connolly’s tempered, tragic acceptance of the need to incarcerate at least some individuals for the sake of security and mobility. I’m not sure the resources for developing an answer to these questions can be found in Connolly’s work because Connolly distrusts the justificatory project of modern political philosophy. Answers to questions such as “what are the necessary and sufficient conditions for state legitimacy” are hard to find in the work of Connolly because the very project of asking these types of questions and providing such answers is under attack. ¹⁵

There are good reasons for the dismissal or “Foucauldian reversal” of some of the classical questions of political philosophy—specifically the question concerning the legitimacy of the state. However, do these good reasons for evading the question of legitimacy at a theoretical level prevent us from raising the question in specific circumstances, such as when the justification of state violence is at issue? If we take the standpoint of the object of state violence, then I see no good reasons for dismissing the legitimacy question because this question—“what right do you [the State] have to use violence against me/him/her/them/us in this case?”—is not only a reasonable question, but an urgent ethical and political matter. It may be that such a question, when asked in the course of philosophical investigations about the State as such, presumes too many answers to too many philosophical issues (e.g. the nature of reason, action, language, etc.) and even presumes too much about philosophical investigation itself (i.e. how to raise and answer philosophical questions). But in the concrete instance where a demand for justification is raised by the potential object of state violence, the legitimacy question doesn’t seem unintelligible, impertinent, or presumptuous. To the contrary, the question seems perfectly in order, necessary, and presuming no more than the intuition that violence is prima facie wrong. In such a situation, it would be unethical to dismiss the demand for justification, for that would be to deny both the moral pertinence of the demand (against a background where violence is always questionable) as well as to renounce any ethical relation to the individual making the demand, as it were disavowing our ethical and political relation to him/her.

If the legitimacy question is in order in a concrete situation, that doesn’t mean that it can be solved without recourse to more general philosophical considerations. For example, we may once again be led back to investigations into the authority and/or legitimacy of the State as such, investigations very much like Simmons’ in Moral Principles and Political Obligations. Or we might be led to consider the nature of a right, and what rights we may have simply in virtue of being human. Or, if we conclude or assume that there are no natural rights, then we may consider what justifications we can give for the all too human rights that are part of the constitutional structure of many contemporary political orders. In these, and in other instances we might imagine, justifying will play a central role; indeed, it is the purpose of the exercise, for the whole point of turning to the questions about state legitimacy still raised by analytic political philosophers is to respond to a concrete request for justification. Just as it will not suffice to explain to the victim of state violence that moral realism is a defensible position within moral philosophy, it will also not do to inform the victim that, for example, the dynamics and productivities of power have shifted historically, shaping and being reflexively shaped by punitive practices executed by the State. That the object of state violence is to be punished in order to reestablish in spectacular form the integrity of sovereignty; to extract economically productive effects from a disciplined body; or to manage and “make live” the life of the individual, is, obviously, no answer to the question of justification.

To justify practices of state violence is clearly something different from describing them, even if there is no clear line between facts and values, and thus descriptive and normative concerns. For this reason, then, Connolly’s arguments compel us to “leap” onto the path travelled by Simmons and others.

Conclusion

I have argued that when we examine philosophical anarchism and the paradox of politics together, difficulties in both positions push us towards questions asked and (however problematically) answered by the other position. Focused on the relationship between legitimacy and state violence, I have also tried to implicitly demonstrate when, how, and why a conversation between analytic and continental political thought should proceed. I have not argued for any resolution to the impasses I have tried to describe, primarily because I do not think there are any clear resolutions to be found. This is not, I would suggest, (only) because political philosophical problems are intractable; (some) political phenomena are intractable. The legitimate use of violence by the state is one such phenomenon. The use of violence by the state is neither necessary nor likely to disappear. It is not necessary because its use is neither a goal of politics as such (except under extremely perverse political conditions), nor is it the only means to achieve our political goals, (even those most closely connected to violence, e.g. internal security; it might, of course, be the most effective way). It seems as if we must have a normative justification for state violence—if violence is always morally questionable—but all normative justifications rely on prior philosophical commitments that are themselves constantly contested. While a contested philosophical position harms few, if anyone, a contestable normative justification of violence can become deeply implicated in harming many. We cannot stop trying to justify state violence within political and legal life, nor can we actually succeed at such justifications.

It is precisely because of the intractability of some political phenomena that we should construct bridges between traditions of political thought that ignore each other. This bridge—if we push the metaphor—does not get us to the phenomena themselves, which, as it were, are located in the gap the bridge crosses. But from these various perspectives we learn not only more about the philosophical problems, but we also learn more, if still from a height and distance, about the phenomena of politics. And if we do not learn how to solve either the problems or to satisfactorily account for the phenomena, we learn much more about why our best efforts are frustrated. And that might turn us towards resources we have yet to discover or create, resources that may be between traditions, or ultimately beyond them.