A. Kantian Objections to Slavery Contracts

Kant's legal philosophy is grounded in the universal principle of right, which states that an “action is right if it can coexist with everyone's freedom in accordance with a universal law, or if on its maxim the freedom of choice of each can coexist with everyone's freedom in accordance with a universal law.”Footnote ⁹ Kantian freedom, according to Ripstein, is relational: each person is free to the extent that nobody else is his master.

Of course, no one can be protected from all the effects of other people's choices, as this would entail a unilateral right to constrain the freedom of others.

Each person is her own master, because no other person may choose for her how her powers are used. This innate right to one's own person is a basic constraint on the conduct of other persons, rather than a protection of some “nonrelational” aspect of oneself.Footnote ¹¹

On this view, no issues of justice could arise for a person completely alone, incapable of affecting, or being affected by, the actions of others. However, the innate right to freedom would still govern our actions if we were incapable of using external objects or engaging in cooperative activity. Each individual would still be obliged to respect every other person's right to her own bodily powers. Because we are, in fact, capable of using external objects, and engaging in cooperative activity, the principle of universal right must be extended to cover acquired rights.

In one sense, acquired rights are an extension of the universal principle of right, an extension required to deal with our ability to set and pursue purposes through external means. Contract must be possible because we are capable of cooperation with other persons. For Kant, it would be an “arbitrary restriction” on freedom if we could not pursue our purposes through contractual relations with other persons.Footnote ¹²

B. Rightful Honor

However, on Ripstein's Kantian view, the innate right to be one's own master within a system of equal freedom also generates an “‘internal duty’ of rightful honor,” which requires one to maintain her innate sovereignty—rightful honor precludes her from making herself subject to the choices of another person.Footnote ¹³ According to Ripstein, arrangements inconsistent with rightful honor are not a kind of “juridical possibility” that virtuous individuals should aim to avoid.Footnote ¹⁴ Rather, “your duty of rightful honor says that no such arrangement can be binding” and no person can be entitled to enforce a claim to such an arrangement against you.Footnote ¹⁵

Where rightful honor is maintained, persons are free to do as they please. Substantive evaluation of the merits of a person's purposes or choices has no role to play under the principle of universal right. Individuals choose their own purposes, and they are free to pursue those purposes in collaboration with other persons.

According to Ripstein,

On this view, consent is transactional, requiring an offer of consent and an acceptance by the party who may now do what would otherwise be forbidden. When I offer consent to an action, X, to be performed by A, and A accepts that offer, the resulting performance of X by A is an expression of my own freedom or purposiveness. Our wills are united, but our separate purposiveness is maintained.Footnote ¹⁷

For Ripstein, “consensual interactions are not wrongs at all, because they are consistent exercises of purposiveness.”Footnote ¹⁸ I might agree to give you all of my property, if it suits my particular purposes to do so. In the same way, when I agree with you that I will do something with my body, mow your lawn for instance, I make some of your purposes my own. Particular actions, like property, have a “mine and yours” structure. Just as my property might have been (or might still be) acquired by another, my particular actions might be acquired by another. But my general purposiveness is not the kind of thing that may belong to another, so an agreement to abandon it cannot be binding upon me.Footnote ¹⁹

Notions of purposiveness and separateness (and their loss) seem clear enough if we focus on the standard idea of a master/slave relationship. If I am your slave, then I am obliged to obey any and all commands that you make regarding my body. You might let me choose to do just as I please, and you might even choose to supply me with resources to support my choices and pursuits. But I have lost my independent right to set my own purposes, because our contract gives you the right to determine the use and disposition of my body. I have agreed to act as an extension of your will. Call this obligation slavery.

Such a contract renders one completely subject to the unilateral choice of another. Even if the contract provides the means to finish some important project, its completion seems to terminate the slave's right to set his own purposes. Moreover, as Ripstein points out, models of slavery that afford slaves some residual legal rights are still problematic, because the master's day-to-day control over the slave precludes effective exercise of these rights.Footnote ²⁰

A legal moralist such as Kristol might concede the basic point concerning slavery contracts and simply contest the extension of the antislavery argument to gladiatorial contests. But he might go further, insisting that the resistance to slavery contracts is itself an ad hoc restriction on freedom of contract. “Why should free persons be required to preserve their rightful honor?” Here again Kristol might insist that we cannot explain the wrongfulness of slavery agreements unless we appeal to the authoritative status of the majority's moral sentiments. So we must assess the antislavery argument itself, not just its extension to gladiatorial contests.

C. The Cogency of Slavery Contracts

I quote Ripstein's account of what I call the cogency argument at length.

According to Ripstein, a person can vary his rights and obligations to other persons only if he is a being entitled to set his own purposes. Yet “a slavery contract both presupposes and rejects that entitlement.”Footnote ²² Put another way, the contract purports to reduce the slave to a mere thing or chattel, but a mere thing can have no obligations of any kind, so the contract cannot be binding.

The notions of person, thing, and ownership require careful attention here. The Kantian starting point holds that a person begins with the rights and obligations of universal right, the right to independence from others and the duty to respect the independence of others. The question is whether we can make sense of a contract that would, in effect, give future control of my body and actions to another person. Of course, such control cannot be exactly like the ownership of mere things, since mere things have no obligations to respect one's ownership of them. Let us set this concern aside for the moment.

Ownership concerns one's relationship to other persons generally (not just the person or thing owned); it is a claim that other persons respect one's exclusive title to what is owned. If you own a chattel, A (a mere thing), then other persons must respect your control of A and compensate you for any loss of value should they damage or destroy A without your permission. Moreover, you can, through contract, transfer these rights regarding A to me or any other willing person. So if slave ownership is chattel ownership, then the slave is not the only one who must respect my ownership of him. I have a right against other persons that they recognize the slave as my chattel.

The cogency argument may establish the incoherence of slavery contracts that purport to generate both obedience rights against the slave and ownership rights to the slave against other persons. However, an obligation slavery contract, one that focuses on the master's rights against the slave (and the slave's corresponding obligations to the master) is perfectly coherent.

Suppose that I agree to become your slave on the condition that you first transfer a large sum of money to my children. It is a clear condition of the offer that your ownership of me does not engage until receipt of the funds has been confirmed. However, my untimely demise, maiming, or disappearance during the confirmation period vitiates the agreement and allows you to recover the funds that you have transferred. You have no right to control me during the confirmation period, but I am strictly liable for delivering myself in the pre-agreed condition after confirmation of transfer is completed. Presumably, this handles any concerns about the slave's ability to ascertain whether his own claims under the slavery contract have been honored.

Of course, I do purport to transfer to you the right to determine all of my future actions, including the right to dispose of my life at your discretion. I can no longer choose to enter any new contractual obligations with anyone else, or you for that matter. You determine all my future obligations for me. Again, this is obligation slavery.

Our obligation slavery contract may even say something such as “M, your new master, has the right to determine all your future actions, including all your future contractual obligations.” However, such language is confused. We need not understand the slavery contract as transferring my moral capacity to enter agreements. Instead, we can treat the rights conferred on you as resulting from one rather comprehensive exercise of that capacity. I have entered a contract that leaves me no room for future contracts. The basic capacity for entering agreements remains, as it must, with me. My continued possession of this capacity explains why I must abide my end of the bargain once you have fulfilled yours. Since I have agreed that you are entitled to determine all of my future actions, I am precluded from entering into any future contractual relations with other persons. But my inability to enter contractual relations with others is a consequence of my agreement with you, not a transfer of my capacity to enter agreements. My slavery contract with you precludes other possible contracts for two reasons. First, it precedes them in time, so it has precedence over them in cases of conflict (I could not enter a complete slavery contract with you until such a time as my other contractual obligations were extinguished). Second, its content is such that no future contract can avoid conflicting with it.Footnote ²³

Suppose that I enter into a complex web of contractual agreements that leaves no room for any further contracts—virtually every moment of my day is determined by my contractual obligations. It is both physically and psychologically possible for me to satisfy the obligations under this web of contracts, but there is no time or discretion left that remains my own. I do retain enough time to verify that my various employers have satisfied their obligations under the contracts. Again, let us say that these obligations involve transfers of funds to my children. I enter into the various contracts one at a time, but with the last contract in the sequence I simply run out of room for any further contractual obligations. Of course, no one is my master in the sense that they can determine all of my future actions, and none of my employers can have the right to end my life as this would frustrate the other contracts that I have entered. Yet my capacity for entering contractual obligations is, in one sense, exhausted. There is no room for further exercise of that capacity. Even if it is morally impermissible for a person to complete this web of contracts, the idea of such a web is cogent. The completion of the web does not entail that I have lost or transferred my basic capacity for making agreements, and there is nothing incoherent in the claim that I am obliged to abide all the requirements of my web. I have neither lost nor transferred my capacity for making agreements. I have used it in a way that precludes further usage.

Of course, none of my employers enjoy anything like your master-type right to choose for me. The web of obligation, even if it binds me completely, can be seen as my own construction, evolved through my particular choices, in a way that your master's right to guide all my future choices is not. Likewise, as my master, you enjoy a kind of final disposition over my life that none of the web's contractual partners can hold. These considerations may make the web permissible or more appealing than the slavery contract, but they do not make it more cogent. The web is one way of rendering myself unable to make further contracts; the slavery contract is another. In both cases, I still have the moral capacity to make and abide contracts, but the obligations I have already undertaken prevent me from entering new agreements. In our slavery agreement, the fact that I am no longer in a position to make new agreements results from the scope of our agreement, not a transfer of my moral capacities.Footnote ²⁴

While it may appear that you now hold my capacity for entering agreements, any subsequent contract that you reach “on my behalf” is not a new contract for me. Any obligations to third parties I take on are themselves rooted in our original agreement. I am a subcontractor, constantly at your disposal. That is our agreement. I still am not a mere thing or chattel without obligations. Of course, in the most radical version of an obligation slavery contract, you may also dispose of my life anytime that you choose. This would destroy my capacity for abiding agreements, but you would not be seeking to rely on that capacity after my demise.

Now suppose that a third person, Bob, entices me to disobey you, damages me, or even kills me. What rights do you, as my master, hold against this person? You may think that Bob has taken or damaged your property. Here, the Kantian will insist that only things can be owned, and things, unlike persons, cannot have contractual obligations. Thus, a slavery contract, a contract to become an owned thing, is logically incoherent. But what if we say that I am not, in the strictest sense, a chattel that has been transferred into your ownership? Rather, I am a person operating under a contractual obligation (a very comprehensive one) to you. So perhaps your complaints against Bob must be made in terms of tortious interference with contract or loss of employee services, and not in terms of interference with property. That would be an unhappy result for you, as your claims would be hampered by the various limitations and theoretical quandaries of these torts. Moreover, since I am not a mere chattel, there may be no simple sense in which you may transfer your “ownership” of me to another person (could there be a no-trade clause?).Footnote ²⁵

You would, no doubt, prefer to advance your claim against Bob in terms of your property rights. That would make more sense if our original contract purported to transfer my very person into your rightful possession—the contract turned me into your chattel. Here, the cogency argument is in a stronger position.

If my personhood, or capacity to enter agreements, has been transferred to you, then I cannot be bound by the contract—so the contract destroys itself. We might bite the bullet here; we could accept that because I have transferred my personhood to you, I can no longer be expected to honor any agreements. I am now a thing or chattel, and I cannot wrong you by my actions. You would have no rightful claims against me for my disobedience, but you would be free to deal with me by whatever means were within your power. On the other hand, you would hold rightful possession of me, a right (which I once held myself) held against all other persons—you would hold property rights to me against third parties. Here, Ripstein is likely to respond that the purported transfer of my personhood undermines the contract itself, not just my obligation to abide it. Contracts are agreements between two persons with the capacity to enter binding agreements; once we speak of the transfer of personhood itself, we undermine a crucial precondition of the possibility of contracts. Hence, “a contract to transfer everything can transfer nothing.”Footnote ²⁶

This argument is sound, but only as applied to contracts that purport to grant you full property rights, against other persons, to my person. If we restrict ourselves to an obligation slavery contract, the alleged incoherence does not emerge. An obligation slavery contract can be understood as a comprehensive exercise of my capacity for contractual relations; we need not construe it as a transfer of my very capacity for contractual relations. Thus, an obligation slavery contract evades the cogency argument, but it still renders me subject to your choices. Even if you do not hold property rights to me against others, I have no rights against your use of my body, and I am still obligated to do as you please.Footnote ²⁷

Put another way, if I enter an obligation slavery contract, then I give a particular person the right to treat me as they may treat their chattels. The master has a right to treat me like a thing. Moreover, I also take on an obligation to respect my new master's treatment of me, and I agree to obey my new master's wishes. This may run afoul of rightful honor. However, agreeing to permit a particular person to treat me like a thing is different from agreeing to become a thing. The former may be impermissible, but only the latter is outright incoherent. As long as we accept that an obligation slavery agreement cannot provide the master with full ownership rights against third parties, the slave does not purport to become a thing. Thus, incoherence is avoided.Footnote ²⁸

D. A More Modest Account

So what grounds the duty to maintain rightful honor, a duty that precludes the juridical possibility of a voluntary obligation slavery contract? The answer must be more modest than the ambitious cogency argument, but Kantian liberals can still provide a reply that handles the legal moralist's complaint about ad hoc restrictions on consent.

The answer, I think, is simple: “That is not what contracts are for.” On the Kantian view, we begin with the idea that no one is entitled to choose for another; each person is entitled to set his own purposes. Contracts are admitted to the moral picture to prevent an “arbitrary” limitation on personal freedom—the introduction of contracts opens up projects that sovereign persons could not pursue without cooperation and agreement. But universal right remains the overriding Kantian commitment—contracts are introduced to expand the range of projects that sovereign individuals can pursue, not to threaten personal sovereignty itself. Call this the rightful honor argument.

Now a Kristol-style critic is likely to cry foul here, claiming that the rightful honor argument against slavery contracts is, like resistance to gladiatorial contests, an ad hoc constraint on the general role of consent within liberalism. However, the rightful honor argument is not an ad hoc constraint on consent. Rather, one starts with the view that independence from the choices of others is sacrosanct—something that must be maintained. Contracts are allowed into the picture to the extent, but only to the extent, that they enable persons to pursue goals that could not be pursued without contractual relations. Contracts that cannot be reconciled with a person's continuing status as a being with the right to set his own purposes are invalid from this perspective. Thus, the prohibition against obligation slavery contracts is not ad hoc, because it squares with Ripstein's Kantian account of why we should recognize contracts in the first place. Less impressive than the cogency argument (or what that argument claims), but sufficient for a reply to Kristol. Thus, if gladiatorial agreements are either obligation slavery contracts or a comparable violation of rightful honor, Kantian liberals may reject gladiatorial contracts without opening the door to legal moralism.

Determining when a demanding contract crosses the line into obligation slavery may require further judgment. For rightful honor to be maintained, both parties must retain their entitlement to pursue their own purposes. I cannot abandon my right to choose future projects just for the sake of some current benefit (I become your slave if you pay my wife's current medical bills). Presumably, it is also unacceptable for me to lock myself into a comprehensive project that cannot be abandoned without your permission (where I spend all my remaining waking hours counting the blades of grass on your vast estate). Here, you are not entitled to make me do whatever you please, but my status as a sovereign person is compromised by the right of choice that our contract affords you. I hold no similar rights (the right to determine whether you continue with current projects) regarding your choices, so the powerful normative asymmetry between us makes you my master. I count such contracts as forms of obligation slavery.

But are gladiatorial agreements obligation slavery agreements? Ripstein considers consensual killing a kind of slavery agreement and gladiatorial contracts as a kind of consensual killing. I accept that some consensual killing agreements are also obligation slavery agreements, but gladiatorial agreements are rather special cases of consensual killing. Before proceeding to the discussion of consensual killing, we should consider some alternative forms of “slavery” agreements.

E. Choosing “for” Someone, Purposiveness, and Separateness

In its strongest form, obligation slavery places me under an obligation to accept all directives you, my master, issue regarding the use of my body. (A weaker but still objectionable version would give you rights of choice that compromise my right to set my own purposes.) Moreover, this obligation entails that I waive any rights that I would have against otherwise wrongful assaults. Note that the entailment does not run in the other direction—I might waive my rights against your attacks without accepting an obligation to obey you.

Rightful honor requires that I preserve my ability to choose for myself. I am not permitted to bind myself to a situation where others choose for me. However, the notion of having the right to choose for someone is somewhat complex. I have a right to choose for myself, as, absent particular agreements, I have no obligation to use my body in a way that satisfies other people's preferences. Moreover, I am wronged if anyone attempts to secure their preferences for the use of my body through force. I am entitled to enforce my right to choose for myself, and, where I am unable to do so, violation of that right grounds a claim to compensation. In the slavery contract discussed above, I purport to take on a totalizing obligation to obey you, so it is no longer possible for you to wrong me in any way.

In the alternative version of a slavery contract, the slave does not take on any obligations to his master. Instead, I as the slave simply waive my right against the use of force by you, my new master, against my person. I do not agree to obey your commands regarding the use of my body. Instead, I simply forgo any right against your use of force. There are two subversions of this contract. On the first subversion, you retain your rights against my use of force, so any physical resistance to your assaults is wrongful. Even here, I do you no wrong if I flee without damaging your person or property. I have no obligation to obey you—all decisions to obey are based on a kind of cost-benefit/risk analysis, though physical resistance that could harm you is not a legitimate option. Call this single waiver slavery. On the second subversion, we both enter the same normative situation, each abandoning any right of complaint about the other's use of physical force. The distinction between slave and master would not be normative—it would be a product of relative power and threat advantage. Call this mutual waiver slavery.

Suppose that I agree to enter a single waiver slavery contract not just for some material benefit, but because I want to subject myself to a challenge—the challenge of defying and frustrating a would-be master. Again, I do you no wrong in defying your commands (as long as I stop short of direct violence). Of course, the challenge of defying you is likely to dominate the rest of my life (unless your resources and talents present very little threat), and I have abandoned my entitlement to terminate this project. I have no enforceable right that you respect my decision to discontinue our game. I have lost the entitlement to choose for myself, and you enjoy the ability to control me through the sword of threat and the normative shield of the basic rights that you retain against me. Perhaps the normative asymmetry generated by your retention of your basic rights makes you my master in a way that rightful honor cannot permit.

But suppose that we enter a mutual waiver slavery contract. I want to take on the comprehensive challenge, but I do not care for the restraint imposed by your person (or property) remaining sacrosanct. Willing to accept a similar challenge, you agree to enter a bilateral state of nature with me. Neither party enjoys a unilateral right to call off this state of nature. In normative terms, mutual waiver slavery contracts do not generate masters, because there is no normative asymmetry between the parties.Footnote ²⁹ There is what we might call a loss of normative separateness, because a return to full personal sovereignty requires the agreement of both parties. If there is a master in mutual waiver slavery, that master is identified by empirical considerations—either the parties begin with a radical difference in power and threat advantage or such a difference emerges later.

Should Kantian liberals treat this loss of separateness as a violation of rightful honor? Even if the initial loss of separateness does not generate a master, both parties in a mutual waiver slavery agreement risk being enslaved (should their counterpart emerge as master). This is, I think, sufficient reason for treating each party's consent as defective—mutual waiver slavery is distinct from both obligation slavery and single waiver slavery, but all three are comparable violations of rightful honor.

Where does this leave us? Death within a voluntary gladiatorial agreement seems to be a killing sanctioned by a kind of mutual waiver, perhaps mutual waiver slavery. However, we still need to ask whether consensual killings must constitute slavery agreements. Likewise, we need to determine whether the particular form of mutual waiver found in gladiatorial contests is mutual waiver slavery. We also need to ask whether gladiatorial contests involve something other than waiver—perhaps a kind of mutual obligation (perhaps each gladiator is required, not just permitted, to attack her opponent).Footnote ³⁰

A. The Extension to the Slavery Argument

According to Ripstein, murder constitutes “biological slavery” because “the murderer decides whether the victim will continue to exist.”Footnote ³¹ Consent, he argues, cannot correct this problem:

For the moment, accept that the victim takes on an obligation to permit the killing. As I argue in Section I.C, such an obligation is not “normatively impossible” in the way that Ripstein claims. Again, the cogency argument conflates an agreement to permit a particular person to treat me like a thing with my agreeing to become a thing. Our obligation slavery agreement remains cogent as long as it does not purport to transform me into a chattel, something you can fully own (through rights held against other persons). An agreement to permit you, but only you, to kill me is not an agreement that purports to generate such ownership rights. I could still resist, by rightful force, anyone else's attempt to kill me. Likewise, our agreement may permit you to kill me, but only in a particular way at a particular time.

Nevertheless, this permission gives you the right (against me) to decide whether I continue to exist. I obligate myself not to interfere with your choice to kill me, as long as that effort respects the permission's requirements. In fact, I may desire to die in the prescribed fashion, just as I may continue to desire counting every blade of grass on your vast estate after I give you the right to determine whether the grass counting project ever ends. But in each case, the direction (or existence) of my purposiveness is determined, as a matter of right, by your choice. In each case, I take on an obligation that permits you to choose for me. Thus, consensual killing agreements appear to be forms of obligation slavery, violations of rightful honor that Kantian liberals can reject through the rightful honor argument. Again, contracts enter the Kantian moral picture to expand the range of purposes that persons can pursue, not to compromise individuals’ basic right to choose for themselves.

Similar considerations apply to consensual maiming in Wright's Case.Footnote ³³ Here, a beggar asked the accused to maim him, hoping that disability would increase his intake of charitable donations. Here, the victim gives another person permission to destroy an aspect of his purposiveness. Likewise, the victim accepts an obligation not to interfere with action on this permission. Thus, the contract in Wright's Case made the accused the victim's master—it gave the accused the right to determine whether the victim would retain (or lose) the relevant capacities.

Ripstein insists that his objection to consensual murder does not depend on Kant's ethical objections to suicide. Suicide may be unethical, but it does not empower another person to choose for me. Likewise, rightful honor does not preclude an assisted suicide “in which one person provides another with a mechanism that enables the other to take his own life.”Footnote ³⁴ (Whether Kantian liberalism has the resources to distinguish between permissible and impermissible assisted suicides is another important question.)Footnote ³⁵

Ripstein does not attempt a detailed Kantian analysis of assisted death practices, but he does give brief consideration to the distinction between assisted suicide (AS) and voluntary euthanasia (VE) in a footnote. In his view,

In all other cases, one person can act on another's behalf only if she acts so as to preserve the other's purposiveness or capacities. The Kantian view demands preservation of personal sovereignty here—a person capable of ending her life must make that choice for herself. Of course, this neglects an important case, a person who is able to articulate clear preferences but unable to use suicide machines. Persons suffering late-stage Lou Gehrig's disease (ALS), for example, may be able to communicate (though slowly) but unable to effectively activate a machine.Footnote ³⁷

B. Voluntary Euthanasia as Slavery

For the sake of argument, assume that persons seeking death can be assisted in a way that keeps the final act in their own hands. Many authors maintain that there is still no principled distinction between VE and AS. According to Dan Brock, we may tell ourselves that a doctor who provides a patient with a self-administered drug does not kill. However, as Brock points out, if A provides B with a lethal drug, knowing that B will inject it into an unwilling C, B is not the sole killer. Rather, A and B are joint killers of C. Call this joint murder. Nothing changes, Brock insists, if we make B the person to be killed. Here, A and B are joint killers of B, even if B performs the final act in his own AS. Doctors are killers in both AS and VE, but both are acceptable because of the patient's informed consent.Footnote ³⁸

Ripstein, however, would respond that A is culpable in joint murder, because A knows that his actions will facilitate B’s violation of C’s personal sovereignty. In other words, B purports to take the authority to decide whether C dies, and A knows this. However, B’s self-injection does not compromise personal sovereignty. B is informed about the nature of the drug or assisting technology, and he may also consent to a physical connection to that technology, but he does not authorize another person to initiate a lethal dosage. B reserves for himself the final decision about his continued existence, so A does no wrong in giving B the drug for self-injection.

However, on the Kantian view, just as B cannot take final authority over C’s life, C cannot give that final authority over himself to B. The consent in VE purports to transfer such authority to another person, and this transfer makes a slave.

I may have good reason to think that you will kill me only if you believe that this is what I truly want (in a sense, I am asking for your validation of this desire).Footnote ³⁹ But a master can choose to act on benevolent motivations, and a would-be slave can choose a benevolent master. The right to choose for someone is the mark of a master.

The parity of an obligation slavery agreement with the performance of VE for an ailing loved one may still seem dubious. Consider a more specific contrast. A pauper gives a rich man permission to kill him on a specific date if the rich man first pays the medical bills of the pauper's children. Such an agreement constitutes obligation slavery, as the pauper gives the rich man the right to decide his fate. Clearly, in asking you, a close family member, to end my life, I need not intend to offer a similar right over my person. Our agreement involves permission, because I appreciate that you may, when the time comes, find the task too difficult. I may be unwilling to burden you with a definite obligation, and you may want to back out at the last moment. Likewise, the permission is revocable. I am asking you to kill me because I want to die, not because I want to give a family member the right to decide the matter (or so it seems). Physical resistance to your attempt would be a withdrawal of my permission, not a failure of obligation (quick aggressive resistance might constitute a wrongful assault on your person, but, in most VE cases, the permission giver would be unable to offer such resistance).

The rich man does not afford the pauper an opt-out clause; he pays the medical bills in exchange for nonrevocable power of life and death. As I understand him, Ripstein holds that the nonrevocable character of the pauper's permission only worsens the core problem, which is still present in VE. Suppose that I do not withdraw my VE permission, and you kill me. By maintaining my permission, I validate your decision to act (or not) on the permission. My position in VE is comparable to my offering to be your slave in a more comprehensive obligation slavery contract. As long as I maintain that offer, I give you the opportunity to take permanent control of my person through acceptance. In VE, as long as the permission is maintained, you hold the opportunity to make a permanent decision about the disposition of my person. If you kill me while the permission is maintained, I have validated both your right to kill me and your particular decision under that right—just as the maintenance of my offer validates your subsequent right to command me, if you accept my comprehensive obligation slavery proposal.

I may want you to kill me. But we should not conflate desiring (wanting or wishing) with willing (or decision making). I might have a desperate desire to die, but still refuse to take my own life because I still have reservations about the moral validity of terminating my own life. By giving you permission to kill me, I empower you to determine the validity of my desire. Your will determines whether my desire is a legitimate reason for terminating my life; you may refuse to kill me, kill me for reasons entirely your own, or kill me because of my desire to die. Even in this last case, however, you determine whether my desire for death should be satisfied. You choose for me.

Ripstein has a plausible case against VE. It remains to be seen, however, whether gladiatorial contests involve comparable consensual killings. Ripstein insists that the problem with gladiatorial contests is only exacerbated if gladiators sign forward contracts that preclude withdrawal of their consent.Footnote ⁴⁰ In other words, even if gladiators have a right to surrender they enslave themselves by maintaining a permission that would legitimate death at the hands of their opponent. In Section III, I argue that gladiatorial permissions are not analogous to VE permissions. However, to conclude this section, I consider one last problem with Ripstein's discussion of consensual killing.

C. An Obligation to Kill?

Obviously, consensual killing involves permission. Minimally, this permission must include the victim's waiving his right to complain about the relevant physical intrusion. It may also involve obligations on the part of the victim—obligations to refrain from physical interference with the relevant contact or intrusion. As noted above, however, physical resistance probably functions as a withdrawal of consent in VE cases, and, obviously, gladiators would not take on such obligations with their permissions. Still, permission, as discussed so far, legitimates whatever decision you, the would-be killer, make—you are within your rights to kill me or at least attempt my killing. With the permission in hand, you get to decide.

Yet what if our consensual killing agreement involves not only waiver permission and noninterference duties for me, but also an obligation for you? What if our agreement obligates you to kill me? Consensual killing violates rightful honor because it gives the would-be killer the right to determine whether the victim lives or dies. If our contract demands that you kill me, I must waive any complaint about your performance. However, if you are also contractually obliged to kill me, it is hard to see how our agreement makes you my master. Of course, you can, in one sense, decide whether to kill me. Any person with the physical ability to end my life holds that power. Likewise, people can decide whether to abide their bargains. I suggest, however, that if our contract obliges you to kill me, it does not provide you with the right to determine whether I live or die. By killing me, you do not choose for me in this sense. In normative terms, you have no choice. You must fulfill your contractual obligation.

The idea of an obligation to kill presents a serious problem for Kantian liberalism's account of consensual murder. Ripstein, I think, needs to show that an obligation to kill is incoherent. Otherwise, consensual murder agreements can be formulated in a way that avoids violations of rightful honor.Footnote ⁴¹

Presumably, an argument against the cogency of an obligation to kill would take the following form:

This argument is similar to the cogency argument against obligation slavery agreements, and it suffers from similar problems. In Section I.C, I argue that obligation slavery agreements do not transfer or destroy the slave's capacity for making or abiding bargains. A master would destroy that capacity if he killed his slave, but, to be crude, that capacity is no longer relevant after the slave is dead. Likewise, while A’s capacity to demand (or refuse) performance is destroyed by B’s act of killing, that capacity is no longer needed once the agreement is complete. The obligation slavery agreement ends with the slave's death, and the consensual killing agreement ends when B kills A.

An obligation to kill seems to pose similar problems for Kantian liberalism's treatment of gladiatorial contests. A gladiator cannot test or prove his deadly skills against an opponent who refuses to provide a deadly challenge. So a gladiator might make his offer of waiver permission for attack contingent on the acceptance of an obligation to act on that same permission. However, for reasons related to each gladiator's right of surrender, it is not clear that gladiatorial agreements can involve genuine obligations to attack.

A. Consent to Death and Consent to the Risk of Death

According to Ripstein, there is a crucial difference between consenting to the infliction of death or serious injury and consenting to the risk of death or serious injury.

By contrast, Ripstein tells us, the gladiatorial contest cannot be understood simply as a risky consensual endeavor. Since each gladiator gives the other the power of life and death over himself, the contest reduces each participant to the status of a thing, a thing without a right of self-defense.

On the Kantian view, gladiators may engage in the same physical actions as a person exercising a right to self-defense. However, gladiators cannot exercise this right, because they have abandoned their rights against being killed. According to Ripstein, the gladiator, being reduced to a mere thing, cannot use rightful force to defend his person. Rather, a gladiatorial agreement

The gladiator sanctions his own death, though he has also committed himself to physical resistance. By contrast, a person who is incapacitated by an assailant is unable to defend himself, but he has the right to do so, because he retains his personhood.

The boxer's agreement to fight is consistent with his right to self-defense, but he does not exercise this right during a rule-governed boxing contest. He has sanctioned the blows directed at him, even if he offers physical resistance to them. His right to self-defense engages precisely where he is unlikely to be able to exercise it: when he is down. Here, he retains his right against further assaults intended to kill or further injure him; even if he is unable to exercise that right, others might exercise it on his behalf. The gladiator's contract, by contrast, does not preserve this right to self-defense.Footnote ⁴⁴

Ripstein does not intend to provide a decisive defense of professional boxing, which, he notes, may be more brutal than his discussion suggests.Footnote ⁴⁵ His account “is abstract, and speaks only to the factors relevant to classify particulars….”Footnote ⁴⁶ The crucial question is whether an individual consents to death/maiming rather than risk of death/maiming.Footnote ⁴⁷ A ban on boxing would be legitimate if boxers consented to maiming, like the victim in Wright's Case Footnote ⁴⁸ (as opposed to a serious risk of maiming). Gladiatorial contestants, Ripstein insists, consent to more than the risk of death/maiming; they validate their opponents’ right to inflict death/ maiming.

Ripstein maintains that the core problem is only worsened if the gladiatorial agreement involves a form of forward consent that cannot be withdrawn. On my understanding of his view, there is a close parallel between a consensual murder case where a victim maintains (does not withdraw) the permission to kill and a gladiator who chooses to fight on (refuses to preserve his life through surrender). A fighter that leaves this permission in place legitimates his opponent's decision to kill him, making him a slave to the eventual victor. By providing and maintaining the opponent's permission to kill, a gladiator violates his own rightful honor, even if the contract allows him to withdraw that permission.

B. Voluntary Gladiators as Mere Things?

According to Ripstein, the fact that both fighters consent to a gladiatorial contest does not correct the crucial defect in each participant's consent. Instead, he maintains that both fighters reduce themselves to mere things, like animals that fight to the death in an arena for the amusement of others.

In Section I.C, I argue that we should distinguish between agreements that give you the right to treat me like a thing and agreements that purport to transform me into a thing that you own. Obligation slavery and single waiver slavery agreements are instances of the former, but even these contracts do not transform me into a mere chattel (over which you enjoy full rights of ownership against other persons). We might think of mutual waiver agreements in a similar fashion. Mutual waiver agreements allow each party to do (or attempt to do) things to the other that, under normal conditions, the parties could do only to mere things or chattels. But part of you treating me like a thing involves acting as if you are my owner, my master. Both obligation slavery and single waiver slavery generate normative asymmetry—they afford you a superior normative position that makes you my master. Under obligation slavery, I must obey your commands and suffer your uses of my body without resistance. Under single waiver slavery, I am not bound to obey your commands. However, I have no right to complain about your predations, and my right to physical resistance is compromised by the personal rights that you retain. I may refuse to obey you, but I have no right to fight back.

By contrast, a mutual waiver agreement does not make you my master. Nor does it make me your master. Each of us may attempt to dominate the other, but neither enjoys any normative advantage or right of control. Ripstein says that each gladiator gives his opponent the power of life and death, but this is misleading. During the contest, both of us may attempt to kill, but neither of us has a right to the desired result. Likewise, neither of us enjoys a secure position of choice, because our deliberations are harried by the threats that we pose to each other. You might enjoy something like the power of life and death if you knocked me unconscious during our contest. However, in the version under discussion, the referee would declare an end to the contest under these circumstances. (Once more, this still leaves plenty of room for death and maiming.)

Again, this gladiatorial contest is a form of mutual waiver, albeit a constrained one. I give you a right to try to kill me subject to the contest rules, and vice versa. If we agree to a death match using only trident and net, you would break the rules by making use of an additional net cutting blade. Moreover, in the right-of-surrender contest, we may continue our attempts to strike killing blows only until one of us invokes surrender (or a referee intervenes to preserve an incapacitated fighter). Our match might also include a pre-agreed time limit in addition to a right of surrender. Now one might say that such a contest still violates rightful honor because, even if neither participant has a master, each participant still surrenders too much control over their person. In other words, both participants have agreed to be treated too much like a thing. I will deal with this issue in the following sections. First, I want to dispose of the idea that gladiatorial contracts transform persons into things.

Arguably, gladiatorial contracts of the late Roman Republic should be understood as reducing participating gladiators to mere things. In these agreements, a free person or citizen would, in effect, sell himself to a particular lanista (owner of a gladiatorial school). Such a gladiator would then be forced to accept the training regimen and fight selections arranged for him by his lanista. The lanista would in turn be forced to accept the producer's rules, including the producer's right to order the execution of a gladiator for poor performance. In this tradition, gladiator chattels could surrender, but survival would be contingent on the producer's assessment of their performance. Likewise, the referee would halt the contest once a contestant suffered a grievous wound. The wounded gladiator would be declared the loser, and, once again, his survival would be contingent on the producer's assessment of his performance (though the producer usually deferred to the audience's judgment). Fighting chattels had no complaint of their own about rule violations; they had no right against attacks that exceeded contest rules. Such rights were held by their owners (and the producer) against other owners. Victorious gladiators would execute underperforming opponents, but not as a matter of rightful choice. Rather, as slaves, they would be bound to follow the commands of both the producer and the referee.Footnote ⁴⁹

Our gladiatorial agreement need not emulate this model (though our actual contest, except for the possible postfight execution scene, might look quite similar). It can be an agreement between the two of us, an agreement for just one deadly contest. Even if our agreement contains no unilateral right to surrender, the claim that we are forced to fight to the death is misleading. No one else has a right to demand that we fight on; we have no owners. If, at some point, we both agree that honor is served, we can agree to terminate our agreement. As a matter of practicality, this would require some break in the conflict—a moment when we have broken apart or made distance between us. Likewise, it would be best to have some kind of prearranged word or phrase that one fighter could offer and the other repeat. “Honor served,” perhaps.

Suppose that we agree to a trident-and-net contest, which precludes any other weaponry and requires us both to wear the same minimal armor. You, however, weave a few small but very sharp bards into your net. These additional weapons are likely to catch and tear any exposed flesh trapped within your net. This breaks the contest rules, and any wounds inflicted by these barbs constitute an illegitimate assault on my person—your covert weaponry violates my rights, not the rights of some promoter or slave owner. A third party who detects your malfeasance could intervene against you, exercising my right to self-defense on my behalf.

Now if our contest is going to protect incapacitated opponents, secure respect for contest rules, and provide an individual right to surrender, third parties must be involved. An unscrupulous gladiator might kill an opponent he knows to be incapacitated, but even thoroughly honorable contestants cannot be expected to show too much concern for their opponents. To protect themselves, contestants would have to empower a referee to oversee the contest.

To make much money as gladiators, we will require promoters, and interested promoters will likely demand that we fight for them. That is how the world of fighting (boxing and MMA) works. However, we need not delve into the details of fighting promotions and pay-per-views here. Imagine a well-heeled promoter with a great love of the fighting arts. He provides fighters with a large venue for fighting; he collects some funds through live audience, web streaming, and downloads. He employs excellent referees and high quality medical personnel. He takes his chances on the quality and duration of fights. Fighters agree to fight other fighters; the promoter takes the bouts that come to him, and there are not enough bouts or contestants for him to worry about arranging the “best” matches. Even if the promoter arranges the bouts himself, his agreements with fighters maintain protection for incapacitated fighters, individual rights of surrender, and strict enforcement of contest rules (though if you surrender too quickly too often, he may not allow you to fight in his venue again, and a quick surrender may cost you your purse). The emergence of such a promoter is not likely, but it would handle concerns about fighters being reduced to chattels, forced to fight for the entertainment of others. Kristol's objections to gladiatorial contests are not contingent on the absence of such promoters, and Ripstein, as I understand him, would still insist that such contests violate rightful honor.

Gladiators in the contests I have just described do not become things. Rather, their fighting agreements involve a mutual waiver, an exercise of their capacity to make agreements, not a transfer of that capacity to other persons. Of course, Ripstein might respond that such agreements still allow persons to be treated too much like things.

C. Risk and a Gladiator's Goals

According to Ripstein, a gladiator abandons his right not be maimed or killed, thus legitimating whatever wounds his opponent inflicts on him. A boxer, on the other hand, consents to the risk of serious injury, but he does not give his opponent the right to maim him.

Ripstein indicates that this would change if the boxing match were sufficiently brutal. His thinking seems to work along the following lines:

Three points should be conceded here: (1) gladiators who fight with weapons similar to ones found in Roman gladiatorial contests intend to kill or maim their opponents (their broader motives for attempting to kill or maim their opponents may vary, but they are attempting to maim or kill); (2) the physical resistance permitted by the contest rules is not an exercise of the right to self-defense; (3) there is a sense in which each gladiator's consent legitimates the death or wounds that they suffer in the contest (as long as death or wounds are inflicted within the contest rules).

However, these elements do not establish that gladiatorial consent compromises rightful honor. My consent to our contest legitimates my death (should it occur) in the following sense. If you kill me by acts within the contest rules, I suffer no wrong. I have no complaint about your intent/attempt to kill, and I have no complaint about the outcome. But that is not the same thing as empowering you to choose for me.

In Section II.B, I argue that Ripstein has a plausible case against consensual murder and VE, even if the permission to kill is revocable. By maintaining the permission to kill, I validate whatever choice you make about the disposition of my person. Here, I am empowering you to choose for me. Gladiatorial consent is different. I accept the risk that your attempt may succeed in order to test my own skills as a gladiator, and you accept the same risk in relation to my efforts.

My resistance to your attempt is not an exercise of my right to self-defense, but it is part of a project that I have chosen for myself. I could not pursue that project without your deadly intent, so I have contracted with you to permit your attack. I consent to the risk posed by your intent, not your right to give it effect. However, because both parties agree to the same serious risks, neither starts out as subject to the other's choices in any straightforward sense.

As we have already established, mutual waiver slavery does not generate the kind of normative asymmetry found in obligation slavery and single waiver slavery. The problem in mutual waiver slavery is the risk of empirical circumstances that turn one party into the master of the other. Is our mutual waiver also mutual waiver slavery? At the outset, our gladiatorial agreement creates no master, so it creates no slaves. And, in the course of our deadly contest, neither of us holds a secure position of choice from which she may decide whether to end another's life. Failing to press a fatal advantage may impose a fatal cost.

Matters are different in the case of an incapacitated opponent. Imagine your trident propping up my chin as I kneel, weaponless and clutching at a grievous wound. Ripstein has something like this in mind when he contrasts boxing with gladiatorial matches; he describes the latter as giving each contestant the right to strike at a downed opponent. Here, my consent to your death strike cannot be understood as a risk undertaken for the sake of testing my own skills. If I am completely unable to react, there is no longer any skill for me to test. And you now occupy a secure position from which to choose my fate. If fighters consent to this kind of death, then they consent to a form of mutual waiver slavery—the parties begin as equals, but risk being reduced to slaves at some point in the contest. However, as we have already discussed, gladiatorial contests can be quite deadly, even if they do not allow contestants to kill utterly prone opponents.

Moreover, the notion of an incapacitated opponent requires clarification. We should not conflate being on the ground with being incapacitated. To avoid a thrust from my trident you step back with your lead foot, but just as your foot leaves the ground I crouch low and sweep out with my net. Your foot is caught and you are upended as I pull back. You land hard on one shoulder, and I spring forward again to attack. I have the advantage, and you trained to avoid being caught in this fashion. But you also trained for being caught like this, drilling the relevant defenses, ground attacks, and standing recoveries over and over again. My advantaged position does not make me your master—while the odds now tilt in my favor, there is still no simple sense in which I get to decide your fate. You are on your back, but I can die on my feet. As precarious as your situation is, consenting to be struck on the ground can still be understood as consent to a risk that tests your fighting skills. And consenting to be struck on the ground here is not a matter of giving me a right to determine life and death.

D. The Opponent's Choice

If the idea that a gladiator subjects himself to his opponent's choices in a way that compromises rightful honor remains attractive, then consider the following examples of risk taking:

Perhaps people should not be allowed to take the risks involved in climber vs. mountain and man vs. machine. And perhaps those who take these risks should suffer criminal penalties. However, Ripstein's Kantian liberalism cannot explain criminalization here, because the participants do not subject themselves to another person's choice. Justifying the criminalization of climber vs. mountain and man vs. machine would require substantive evaluation of the chosen activities—we would need to say that the chosen activities lack significant value or that their value is outweighed by their risks.

Ripstein insists that by consenting to your attack, I make myself thinglike, an object of your choice. And you make yourself thinglike in consenting to my attack. However, there is an important way in which this gets matters backwards. In consenting to your attack, I am treating you as thinglike, like a cliff face or a killing machine that I intend to overcome. You are, of course, a person; you are a self-directing agent in a way that cliff faces, killing machines, and even slaves are not. But does the kind of choice that distinguishes you from the slave make a significant difference between me and the master in master vs. slave?

Of course, master vs. slave is impermissible, but this is because the slave is forced to participate in the contest—slavery compromises his rightful honor even if his master never orders him to fight. However, the master's rightful honor is not compromised in master vs. slave. In climber vs. mountain and man vs. machine, rightful honor is not compromised because participants choose to face risks posed by things; things do not choose to kill, because things do not choose. In master vs. slave, the slave has either the status of a thing or, if his slavery results from an obligation slavery agreement, his master is entitled to treat him like a thing. Moreover, the slave is obliged to obey the master's orders. No doubt, the slave chooses whether to abide his obligations; likewise, to the extent that one makes deliberate decisions during a fast-paced deadly contest, the slave chooses his moves. But, even if he kills his master, the slave does not choose, as a matter of right, the master's fate. Rather, the master chooses to accept the risk posed by the slave's assault.

Master vs. slave lacks the mutuality of our contest. Neither of us is obliged to enter the fight, and we are both protected by a rule that safeguards utterly prone opponents. Likewise, while there has been little discussion of this element so far, we both enjoy a unilateral right to surrender. In contrast to master vs. slave, neither of us seems to have a strict obligation to attack, as our mutual waiver simply permits our deadly attacks.

E. The Right to Surrender

The role of surrender in protecting rightful honor requires further discussion. We have already established that a referee must be authorized to protect prone fighters. Moreover, the two fighters could agree, together, to end their contest. Does rightful honor require an additional individual right to surrender?

Presumably, gladiators should reserve the right to surrender in order to reinforce the protection already provided by the referee. Fighters may detect the infliction and/or severity of a wound before the referee; likewise, they may realize that they are trapped in a fatally compromised position before the referee becomes aware of their vulnerability. On this account, the right to surrender is just another device for protecting incapacitated fighters. Call this protective surrender. But we can make sense of this right being abused. Protecting vulnerable opponents might require that referees respect such declarations of surrender, but subsequent analysis (through replays or medical examination) might reveal that the surrendering gladiator feigned his incapacity.

Moreover, a gladiator can exercise a de facto right to surrender by making himself prone or defenseless. Suppose that, just as the contest begins, I cast aside my weapons and fall on my face with my arms outstretched, refusing to move. It is child's play for you to drop your heavy net upon me and jab away at my back with your trident.Footnote ⁵¹ Even if such a maneuver is a violation of our fighting agreement, you cannot be permitted to strike when I am vulnerable in this way. Allowing you to strike when I am in this position affords you a secure position of choice from which you may determine my fate.

Gladiators who abuse protective surrender or exercise de facto surrender cannot be killed for their refusal to fight (they cannot consent to this remedy), but perhaps they have still violated their agreement. What compensation might be appropriate for such a violation is a question that I cannot pursue here. The important question is whether rightful honor requires a stronger right to surrender—an unproblematic right to withdraw from the contest just because one decides that he does not wish to fight (or fight on).

Suppose that we break apart for a moment during our fight. You make distance because your attack has been stifled; you wish to refocus and begin again. “Honor served?” I ask. You wave off my offer by shaking your head and pointing your trident at my face. I want to leave the contest, but our agreement does not permit an individual right of surrender. I could feign injury or throw myself upon the ground, but I refuse to violate our agreement by such cowardly tactics. I agreed to a contest that does not permit a physically game gladiator in a feasible fighting position to leave the contest unless both fighters agree that honor has been served. Have I violated (or worsened an existing violation of) my own rightful honor by not reserving an additional (nonprotective) right of surrender for myself? Call this a proper right of surrender.

You want the contest to continue. I remain “locked in.” But you are keeping yourself “locked in” as well. A slave can be happy with the current state of his life, and he may have good empirical reasons for his prediction of continued happiness. He is no less a slave. You may be happy to fight on, and you may be correct in predicting that you will want to fight on no matter how the contest goes. Yet by waving off my “Honor served?” you preserve a situation that can only be exited with my assent. My offer may not come again, and I may decide to mirror your disdain if you offer the question later in our contest. You are not a slave, but you are not my master either.

Here we have no master/slave relationship, but there is a loss of separateness. Neither of us has a unilateral rightful choice to exit the contest, but no one else holds that choice for us. We are not things; no one else is entitled to make us fight to the bitter end, and no one else is entitled to attack us the way we are entitled to attack each other. As I see it, there are different possible conceptions of rightful honor. One holds that individuals must fully preserve their separateness; individuals cannot make agreements that give other agents partial rights over the disposition of their person. Another holds that such agreements are permissible as long as they do not generate master/slave relationships. Such relationships are avoided when the kind of rights that A gives to B, B also gives to A. (There is a strange echo of Kant's own views about marriage here.)Footnote ⁵²

I am not sure that there are rational grounds for preferring one model of rightful honor over the other. It does seem to me that there is a kind of mutual respect or recognition that can be shown between two gladiators only in contests where rightful exit requires the assent of both fighters. But I will not pursue that point here. For the sake of argument, I accept that we should follow the first model of rightful honor, which requires a proper right of surrender.

F. Does the Right of Surrender Preserve Rightful Honor?—Master in the Moment

Kantians might argue that a proper right of surrender is necessary, but not sufficient, to preserve the first sense of rightful honor. Even if a master is not established at the outset of a gladiatorial agreement, Ripstein might argue that one emerges once a fatal blow is struck. On this view, voluntary gladiatorial matches are impermissible, because they carry a risk of being reduced to slavery at some point in the contest. Thus, even if you kill me, your own consent remains defective, because we both took the same illegitimate risk in entering the contest.

Suppose that we fight for several minutes before your trident is caught up in my net. I smile for a brief moment, but then realize you have more slack than I thought. You twist your trident further into the net and stab toward my neck. I manage to say, “Yield!” just before an outer prong of your weapon pierces my throat. My right to surrender is not like my right to withdraw your permission to enter my house. Your acting on the latter permission does not, by itself, put my house at risk. By contrast, whenever you act on permission to attempt my death, you jeopardize my ability to withdraw that permission. Your choice to attack always threatens my ability to withdraw from the contest. Yet your attack cannot be faulted under our agreement.

A similar result might obtain absent a failed surrender attempt. Perhaps I never tried to surrender. When your trident pierces my throat I might be in the grip of overconfidence, acting on instinct, attempting a countermove, momentarily stunned, momentarily paralyzed with fear, or just feeling too stubborn to surrender. In some of these cases, I might attempt to surrender if I realize my peril. But a fight is not a game of chess. I cannot easily pause and reflect on the merits of surrender after your individual moves. Again, this reflects the instability of my right to surrender. In the other cases, the permission to strike me with deadly force (within the contest rules) remains in effect in a more straightforward sense.

Do you become my master in choosing your fatal strike? You are not a thing, and you are not my slave. The slave in master vs. slave does not have a right to choose the master's death—he strikes because he is the master's chattel or because he is obliged to do whatever his master commands. You, however, have a right to make deadly strikes against my person.

There are two lines of response here. First, a gladiatorial agreement may involve an element of mutual obligation, not just mutual waiver—you have an obligation to pursue your attack, just as I have an obligation to pursue mine. Second, even if you have no obligation to kill me, you still do not act from the secure or privileged position of a master.

G. An Obligation to Attack?

In Section II.C, I argue that Ripstein's account of consensual murder is threatened by the idea of an obligation to kill. If you accept not just a permission to kill me, but a definite obligation to do so, your actions are normatively constrained. In killing me, you simply fulfill your contractual obligation. Presumably, gladiators want something similar—I cannot show that I am a better trident-and-net player than you if you make yourself an easy target. I want you to fight.

So if I am proven wrong, because your trident ends up in my neck, your fatal attack does not make you my master. Rather, you, like the slave in master vs. slave, simply honored your obligation to provide me with the deadly attack that you owe me. I act under the same obligation, so our contract is a matter of mutual obligation, not mutual waiver. Even the master lacks a perfect right of surrender against the slave, though the slave is bound to obey his will. Suppose the master dies because he calls, “Yield!” just as the slave initiates a fatal strike. The slave is almost automatic in his response to the word, but his strike still pierces the master's heart. The master, like unfortunate climbers and machine fighters, dies because safety measures (or exit strategies) are never perfect, not because another chooses for him.

The slave is obliged to attack his master, but I am not sure that we can make sense of two voluntary gladiators holding a genuine obligation to strike each other. First, if we assume that there must be a proper right of surrender, the obligation to attack is conditional at best. I do not have to attack you where I have a realistic opportunity to surrender; I can choose to exit the contest instead. You are not really obliged to attack if you have another legitimate option.

Moreover, it is not clear that the notion of a mutual obligation to attack is coherent. If you have an obligation to clean my house, I can relieve you of that obligation. Your obligation is just the obverse of my right to your performance, and I may refuse that performance. In doing so, I rescind your obligation to perform, your permission to enter my home, and your permission to touch my property. In a voluntary gladiatorial contest, neither party has a right to the other's performance in this sense. If I can revoke your obligation to attack in this way, I can also deny you the permission that helps to protect you from me (if only momentarily before you rescind my obligation to attack). This is unproblematic in master vs. slave, because the slave is always in this position. The master always enjoys a right to strike the slave, and the slave never had a permission (a right to choose) to strike the master. However, the permission to strike me is something that you bargained for, and you cannot agree to allow this permission to be revoked where my permission to attack you remains intact.Footnote ⁵³

The idea of an obligation to attack is sensible within master vs. slave, but not within a fighting agreement that abides the first sense of rightful honor, which demands full separateness. However, an obligation to attack could, I think, arise within a fighting agreement that follows the second sense of rightful honor. The second sense of rightful honor does not require full separateness, so it permits fighting agreements that do not include a proper right of surrender. Even if such fighting agreements are matters of mutual waiver, they still produce an obligation to attack, though each fighter owes that obligation to himself. In such contests, a fighter who refuses to fight (and cannot rightfully exit the contest by his own choice) makes himself a mere victim. He must fight to preserve his own rightful honor.

H. Secure Choice

Even if we follow the first sense of rightful honor, gladiatorial contests that respect a proper right of surrender remain legitimate. You have a right to make deadly strikes against my person, but you have no right to my death as an outcome. Likewise, you do not occupy a secure position from which you may choose whether or not to press a deadly attack. Not pressing your advantage, or what you perceive as your advantage, might prove fatal if I am not inclined to show reciprocal mercy. Neither of us has a right to unharried deliberation on surrender, and neither of us has a right to unharried decisions about whether to press the attack. There is something to the idea that fights are deadly chess matches, but the similarities are easily overstated (you may have limited time to decide your chess moves, but I am not permitted to physically harass you while you deliberate).Footnote ⁵⁴ All your deliberate decisions are risk laden and made under duress. As are mine. Moreover, in the midst of such a contest, many if not most of our moves and countermoves are the product of instinct and conditioned training, not deliberate calculation (the same is true of many tests of skill, including mountain climbing, downhill skiing, and race car driving).

Unlike the slave, you are not obligated to attack. Your presence within the contest is a matter of choice—but you are bound to respect the right of surrender (along with the referee's authority) to the best of your ability. Like the slave, however, there are limits to your moment-to-moment ability to restrain your attack—in this respect you are thinglike, similar to the cliff face and the killing machine. You might intentionally carry through with your attack in spite of my clear surrender—but our agreement recognizes this as a wrongful assault (of course, it would sometimes be difficult to determine whether failure to comply with a surrender was intentional).

We are both thinglike here, but we are thinglike in this regard when we choose to face challenges posed by nonagents. The primary expression of our respective agencies is the deliberate choice to accept an agreement to fight. At this starting point, there is no master. The kind of choices that occur within the bounds of the contest do not establish a master, not even if death results.