If we have any rights, we surely have the right to self-defense. And yet self-defense has proven very puzzling to Rights theorists. To see why, take a simple case:
There are two agents, AGGRESSOR and VICTIM: AGGRESSOR resents VICTIM for his sauve good looks and skill on the dance floor and has made it clear that he intends to kill him. One day AGGRESSOR shows up at the dance hall, gun in hand. He takes a shot at VICTIM but misses narrowly. He prepares to fire again, taking more careful aim, but VICTIM too has a gun and his only hope of surviving is to return fire and kill or disable AGGRESSOR.
Is it morally permissible for Victim to shoot aggressor? Of course! But now here's the puzzle. Rights Theorists have wanted to say:
- It is permissible for VICTIM to shoot AGGRESSOR because VICTIM has the right to self-defense.
- VICTIM has the right to self-defense because he, like everyone, has the right not to be killed or harmed. (That is why, if you are attempting to kill or harm him, you are doing something wrong.)
- In defending himself, VICTIM will kill or harm the AGGRESSOR.
- If the AGGRESSOR, had a right not to be killed or harmed, then it would be impermissible to kill or harm him.
The puzzle is: what happened to AGGRESSOR’s right not to be harmed?
It seems we must say that, by launching his attack, the attacker somehow loses his right not to be harmed. But where does it go?
Note that wherever it goes it doesn't go very far: Suppose that AGGRESSOR takes his second shot and his defective gun explodes rendering it useless and leaving him wounded and helpless on the ground. AGGRESSOR is no longer a threat and now most people (well, anyway, most philosophers) think it would be wrong for VICTIM to draw his gun and execute the now helpless AGGRESSOR. Apparently, AGGRESSORs right against being harmed returns as soon as he is no longer a threat to VICTIM’s life.
And yet there must be more to AGGRESSOR’s losing his right against harm than his just being a threat. After all as soon as any victim decides to defend himself against any aggressor he becomes a threat to that aggressor but we don't want to say that deprives a defender of any rights.
So maybe we should say that the AGGRESSOR's right not to be harmed doesn't go away at all? Couldn't we say that while AGGRESSOR retains his right not to be killed, VICTIM’s right to self-defense is more stringent or weighty than a simple right not to be killed and so it overrides or trumps the AGGRESSOR's right?
But that doesn't seem right either. Even acting in self-defense you have to respect some people’s right not to be harmed. Thus it would not be permissible for VICTIM to defend himself by grabbing some innocent bystander and use him as a shield against AGGRESSOR's bullet. The bystander’s right not to be harmed is not trumped or overridden by VICTIM’s right to defend himself.
So it must be something specific to the AGGRESSOR that removes, diminishes or discounts his right against being harmed; something that distinguishes AGGRESSOR from the bystander. But what can that be?
Isn't the answer obvious? AGGRESSOR, unlike the bystander, is a bad guy. He is trying to do something morally impermissible and for that reason he forfeits his claims against others harming him.
Alas, this collides with the problem of "innocent threats". Two examples from the literature:
Out for a walk one day, this man was swept up by a tornado. Now he is plummeting towards earth. When he strikes he will fall directly on VICTIM. With VICTIM's body cushioning his fall the FALLING MAN may survive but VICTIM will certainly be killed. VICTIM cannot get out of the way, but is carrying a ray gun which he can use to vaporize the falling man and save himself. *
An enemy has sent an automated tank on its way to kill VICTIM. VICTIM has an anti-tank gun and can destroy the tank. To discourage this, the aggressor has strapped an innocent passerby to the front of the tank to act as a human shield. Victim can save himself only by killing this person. *
In both these cases most people would say that it is morally permissible for VICTIM to defend himself even though, regrettably, this would result in the death of an innocent person. And by "most people", I do not mean just "most philosophers" or "most undergraduates when given questionnaires". I mean that it is the nearly unanimous opinion of all mankind, in virtually every human society of which we are aware, in every legal system ever devised, that this is so. No court-- not now, not ever-- would hold VICTIM guilty of wrongful killing if he defended himself in these situations. No work of literature or drama has or would treat a victim’s refusal to defend himself in such circumstances as anything less than heroic; that is, as a sacrifice beyond the call of moral duty.
All of which seems to leave the right theorists with a choice of bullets to bite.
Judith Jarvis Thomson and Frances Kamm conclude that, given that it is permissible to kill innocent threats, we just have to say that that FALLING MAN and HUMAN SHIELD, because they are threatening VICTIM’s life, have forfeited their rights not to be harmed no less that the villainous AGGRESSOR, albeit through no fault of their own.
This seems odd and, again, opposed to common sense. No jury-- not now, not ever-- would convict these innocent threats of any crime were the Victim to be killed. The Shield and the Falling Man do not choose to cause death Yes, VICTIM’s death would be a bad thing, but it wouldn't be their fault any more than it would be a bystander’s fault. How can their actions, if they can even be called that, represent a forfeiture of their rights?
But then what is the alternative?
Michael Otsuka and David Rodin argue that if we follow Thomson and Kamm then we will be left with no arguments against killing innocent bystanders and will lose our grip on the commonsense distinction between innocence and guilt. They conclude that, contrary to general belief, it is impermissible after all to defend yourself against an innocent threat.
As Whitley Kaufman observes, this seems a position so wildly at odds with settled law and common sense morality that it could only seem plausible to a philosopher in the grips of a theory. But then Kauffman's own solution to what he calls "The Paradox of Self-defense" is to invoke the hoary doctrine of Double Effect. On this account, aggressors, innocent or otherwise, do not forfeit their rights not to be harmed. Nevertheless it is still okay for the victim to blow them to smithereens because, in so doing, the victim doesn't really intend to harm them, just to protect himself.
This, we may observe, is something only a philosopher could say with a straight face.
All of this seems an embarrassment to the philosophy of rights. Judith Jarvis Thomson acknowledges as much:
"Many people who do moral philosophy these days appeal to rights to explain why this or that piece of behavior is or is not permissible. For example, it is common to say that the reason why you cannot maximize utility in such and such a case is that fact that the utility -maximizing course of action would involve infringing a right --- indeed, violating a right, since the right in question is a stringent one, and the utility to be got not sufficiently great to override the right. But when we say that, in that case , the utility-maximizing course of action would involve violating a right, are we saying anything more than that, in that case, it is not permissible to take the utility-maximizing course of action? If not, then we can hardly take ourselves to have explained why it is not permissible, in that case, to take the utility-maximizing course of action. It is arguable that if there is any point at all in appealing to rights in such a discussion , there had better be something independent of permissibility's and impermissibility's which fixes their existence and degree of stringency. It is not obvious that this is true. It might be that to attribute a right is only to talk about permissibilities and impermissibilities, but in a way that groups or collects them and brings whole clusters of cases to bear on each other. I do not for a moment think it a novel idea that we stand in need of an account of just how an appeal to a right may be thought to function in ethical discussion. What strikes me as of interest, however, is that the need for such an account shows itself even in a case which might have been thought to be transparent"
I think Thomson is right here. The “Paradox of Self-defense” is really a reductio of a certain conception of rights. It is a conception which treats rights as it they were noumenal entities which hover about persons, somehow endowing some actions but not others with moral impermissibility. It is a view which looks to rights to explain the impermissibility of actions. The problem is, as Thomson says, rights cannot play that role unless there is some way to discover and weigh rights independently of our antecedent judgments about what is or is not permissible, and there is not. We need another account of the role of rights in ethics.
The Retributive Theory of rights offers such an account.
Retributivism begins with the idea that there is really only one kind of action that is morally impermissible: other things being equal, it is morally impermissible to harm others or coerce them with threats of harm. But other things are not always equal: there are situations in which it is morally permissible to harm other people and those are the occasions when rights are relevant. A retributive right is a moral permission (in Hohfeld’s terms, a privilege) to do harm to another. Retributive rights differ from one another in respect of the circumstances in make such violence permissible. Retributive Ethics is the view that retributive rights are the only rights there are.
The right to self-defense is a paradigm retributive right: it is a moral permission for a victim to do as much harm to an attacker as may be required to defend himself. Note, the retributivist does not say that there is a thing, called “The Right to Self-defense”, which makes defending oneself permissible. On the Retributivist view the right is the permission. The right no more explains the permissibility of the action than saying that an animal is mammal explains why it gives birth to its young alive; though it remains perfectly apposite to say that a certain action is permissible because it exercises the right of self-defense in the same way it can be pointful to say that a particular animal gives birth as it does because it is a mammal. Rights talk plays the same role of organizing and connecting moral permissions in the “realm of rights” as taxonomy does in the animal kingdom.
Retributivist rights, then, are ontologically unproblematic. The most convinced skeptic about rights talk cannot deny that there are Retributive Rights so long as he or she acknowledges that it is sometimes morally permissible for people to hurt or threaten to hurt other people. Those permissions are the retributive rights acknowledged by his or her moral theory.
In other posts I’ve tried to show how Retributivism can shed light on a variety of philosophical questions—about the moral force of promising, and the foundations of property rights. But our current topic is self-defense and you might wonder how Retributivism can be useful here: If the retributive right to self-defense cannot explain why it is permissible to defend yourself (because the right is the permission) what use is it? The answer is that Retributivism is useful here not because it explains where the right to self-defense comes from but where it doesn’t.
Recall that our puzzle about self-defense began with the assumption that VICTIM gets his right to self-defense from a prior right: a right not to be harmed or killed. VICTIM’s right to self-defense was supposed to arise from the Aggressor’s infringement of this right. “Paradox” threatened to set in when we wondered what happened to the AGGRESSOR’s right not to be harmed.
Is there a right not to be harmed? Note that, if there is, it is not a Retributive right. Having it doesn’t give you permission to harm someone else. Indeed it does not seem to give its bearer permission to do anything. Even when victim is attacked, this story goes, it is not the right-not-to-be-harmed that makes it permissible for victim to act defensively, but another right-- the right to act in self-defense-- to which the right-not-to-be-harmed’s violation somehow gives rise
Even if there is a right not to be harmed, it does not seem necessary for the right to self-defense. Consider the boxer, the cage fighter or the antique duelist. Each gives his opponent-- for the duration of the contest and with the limits prescribed by the rules-- permission to do him harm. And it is a moral permission: a boxer is not doing something morally wrong when he bloodies his opponent’s nose. If it’s the right not to be harmed that makes it impermissible for others to harm you then it must be that these combatants have given up or limited that right, at least against their opponents. But none of these – not the boxer, or the cage fighter or the dualist -- have given up the right to defend themselves. If the characters in the “Villainous Aggressor” story had been called “Hamilton” and “Burr” instead of Aggressor and Victim you might think the Aggressor less villainous. Burr had given up his right not to be killed by Hamilton when he challenged him to the duel. Even so, Hamilton having fired first and there being every reason to think he would fire again, Burr could still plead that his replying fatal shot was self-defense simpliciter. *
In these cases the defender has voluntarily given up his right not to be harmed. As we have seen, some rights theorists hold that one can lose the right involuntarily. Thomson thinks it permissible for VICTIM to kill the FALLING MAN and the HUMAN SHIELD because they have lost their rights not to be killed. If that were so, and it was really the right not to be harmed that generated a right to self-defense then we would predict that both these innocent threats would have lost the right to defend themselves. But this doesn’t seem to be so. Imagine yourself in the situation of FALLING MAN or HUMAN SHIELD. You see VICTIM raising his weapon to blow you away and you realize what he intends. Suppose that you happened to have a weapon of your own which you could use to preempt VICTIMs act of self-defense? It does not seem to me that this would be impermissible. True, VICTIM is not a villain, but neither are you.
The connection, then, between the right not to be harmed and the right to self-defense seems increasingly unclear. Why should we believe in this right in the first place? The answer is that we will be driven to it if we accept a conventional assumption about the way in which rights connect to the moral impermissibility of actions. That assumption-- one shared by most of the rights theorists we have been considering-- is that:
(M) A has a right that B do/not do φ iff It is impermissible for B to not do/ do φ
So, in particular:
(M’) A has a right that B not harm A iff It is impermissible for B to harm A.
If (M’) were correct then we could not deny the existence of VICTIM’s right not to be harmed without also saying that it is permissible for him to be harmed which would be absurd.
(M) would have to be true if the impermissibility of actions were explained by rights. But, as Thompson observed, no such explanation is on offer. Until it is, (M)’s bi-conditional threatens to collapse to a mere definition: so that “A has a right that B not…” becomes just an uninteresting paraphrase of “It is impermissible for B to.”
Retributive theory rejects (M) and offers instead:
(R) A has a right that B do/not do φ = it is permissible for A to force B to do/not do φ
Always with the stipulation that it is only permissible for A to use the minimum force required to force or prevent φ. In the case where B threatens to harm A we get.
(R’) A has a right that B not harm A = it is permissible for A to force B not to harm A
Thus on the Retributive account, the right not to be harmed just is the right to self-defense.
The retributive right (R’) does not entrain the non-retributive (M’). As we saw in the case of our duelists: that it is morally permissible for A to forcibly defend himself against B, does not mean that it is impermissible for B to harm A. Having the retributive right against being harmed does not make it impermissible for others to harm you.
You might ask, then, what right, according to the Retributivist, do people have that makes it impermissible to harm them? The answer is “No right does or could do this”. Rights, in the manner of (R), reduce to permissions, they do not create them. It is impermissible to harm other people, other things being equal. This root moral fact is not explained by invoking rights and to suppose that it is leads a fortiori to confusion and paradox.
The “Paradox” of self-defense arose from a bad theory about what makes self-defense permissible. But, when we remove ourselves from this muddle, a more interesting question remains:
If we assert a general Retributive right against harm, we would be saying that it always permissible to forcefully prevent another from harming you. But any potential victim who acts in self-defense aims to harm the person who is threatening him. Does the threatening agent then have the right to self-defense?
It is tempting to think not when one thinks of cases like this:
Baker is angry at Able. So angry he starts throwing punches. Able hits back. Baker blocks the punch hard, breaking Able’s arm. At that point others intervene to stop the fight.
Suppose that the broken arm was the only injury in the fight. If Able demanded compensation from Baker for the injury could Baker argue that the injury was inflicted by a permissible act of self-defense? Surely not.
So should we say it is always permissible to defend yourself against another provided they are not defending themselves against you? No, not if we want to say that it would be permissible for FALLING MAN and HUMAN SHIELD to defend themselves against VICTIM’s self-defense.
The problem here is a general one and not just for Retributivism but for any rights theory which hopes to explain the moral permissibility of enforcing any of our rights. Suppose I am standing on your property-- or on your foot-- without your permission. You ask me nicely to step away. I refuse. In that event it may be morally permissible for you to push me off. But if it is morally permissible for me to push back, given my general right not to be pushed around, the morally permissible mayhem is likely to get out of hand.
Confronting this problem it may be tempting to reach into the rights theorist’s standard bag of tricks. To invoke “degrees” or “degrees of stringency” of rights, to distinguish between “infringing rights” and “violating them”; perhaps we’ll need the “doctrine of double effect” or even -- if all else fails-- to introduce considerations of general utility. But I think we can do better than that with the materials already at hand.
Let us do some thought experiments.
At the saloon Baker hears an ugly (and, in fact, false) rumor about Able and Baker’s wife. A notorious hot-head, Baker straps on his pistol and tells everyone he intends to kill Able. He marches through town to Able’s house and bangs on the door. When Able answers, Baker draws his gun and announces that he is there to kill Able like the dog he is. Fortunately for Able, he is armed and a quick draw. Able shoots. Baker falls dead.
Able’s action certainly seems permissible; a paradigm of self-defense. But let me give you more background to the story.
Baker heard the ugly rumor at the saloon early in the day and spent all afternoon working himself up into a liquor fueled frenzy. The Bartender (as it happens an employee of Able’s) called Able asking if he wanted Baker cut off. Able told him “No”, in fact, he said to let Baker run a tab.
Baker left the saloon waiving his weapon around, telling anyone who would listen that he was on his way to gun Able down. Overhearing this, several friends of Able phoned to warn him. They offered to call the police who would certainly have thrown Baker in jail to sleep it off. But Able tells his friends not to call the cops. He tells them he doesn’t want Baker to get into trouble with the law and he is able to defend himself.
And so he is, for Able is very wealthy. His mansion is a virtual fortress with high fences, guard dogs and security guards. Strangely though, while his friends continued to call and report Baker’s progress through the town, Able did not phone the police nor did he take any measures to protect himself. Quite the reverse: he threw open his gates, lowered the draw bridge, kenneled his dogs and dismissed his guards. Then he got his pistol and waited patiently for Baker to arrive.
Now, Able’s behavior looks less innocent. It is still true that at the moment that Baker stood before him gun drawn Able had no choice but to kill or be killed. But Able could easily have avoided this awful choice and had he acted differently, Baker might have suffered, at worst, only a night in jail.
When the whole story comes out the police will surely have some questions for Able. It is no longer clear that this was a simple case of self-defense. Why so? Because it is part of everyone’s conception of the right of self-defense that it only entitles you to use as much force as is required to defend yourself. If an attacker can be disarmed as easily as he can be killed; if setting the phaser to “stun” rather than “kill” will do the job; then it is impermissible to use the excess force. Of course, reasonable people might disagree about how much force is really required in any given case and when the aggressor is malign we prefer to err on the side of the innocent victim. But Able’s behavior was at least reckless, not just in putting his life at risk, but also, it turned out, Baker’s.
So we might judge Able’s behavior over all as impermissible. But what about its culminating action? Was it impermissible at that final moment, for Able to fire? Before you answer, let me give you one more piece of information.
Able never liked Baker. It was Able himself who started the rumor and arranged for Baker to hear it hoping for exactly the course of events that subsequently transpired.
Now Able’s fatal shot looks less like an act of self-defense and more like the last step in a premeditated murder. And we do not need to look for new reasons why this is so. Once again: the right of self-defense is only a permission to exercise the least amount of force you can reasonably believe is required to save you. In Able’s case, it now appears, that amount was zero. He could have prevented Baker’s assault by not starting the rumor in the first place. His shooting of Baker was morally impermissible. Were all the facts to come out Able would certainly be charged with a crime. Perhaps not first degree murder, because Baker’s bad behavior made him complicit in his own demise—more likely something like depraved indifference manslaughter. Whatever the legal charge, “self-defense” would be no defense.
FALLING MAN 2
The situation is much like FALLING MAN, but there is some history. The falling man is known to VICTIM. He is in fact a hated business rival. This falling man wasn’t taken up by a great wind, he has fallen off a tall building. In fact, he didn’t just fall, he was pushed. He was pushed by a confederate of VICTIM’s who was hired to do the killing. Knowing the time and place of the planned assassination, VICTIM has arrived on the scene to enjoy the sight of his rival’s brains splattered on the pavement. But VICTIM has miscalculated and now he sees his rival plummeting towards him. Luckily for him, VICTIM has brought a ray gun which…
I hope we agree that it is impermissible for Victim to kill this falling Man. If there is any doubt about this consider
FALLING MAN 3
Exactly like FALLING MAN 2, except that VICTIM has not miscalculated. He planned to be in precisely this situation so that he could have the pleasure of blasting his rival out of the sky and then pleading self-defense.
I think it clear that in both FALLING MAN 2 and 3 that it is impermissible for VICTIM to vaporize the falling man even though the physical circumstances at the fatal moment are identical to those in the first FALLING MAN story. And the reason is not that in the later stories that the VICTIM is a bad guy doing a bad thing, though he is surely that. What we must explain is why in the latter cases but not the former ones the retributive right (R’) does not make it permissible for VICTIM to save himself. The explanation, once again, is that VICTIM is morally permitted to use only as much force as is necessary to save himself and when we consider how much force that is we take account of VICTIMS options in the whole history of the fatal event, not just at the last second. A victim who uses more force than is reasonably necessary by this standard is culpable for the harm the excess inflicts. In cases 2 and 3 no force at all was necessary and VICTIM is culpable for any harm done the falling men.
And so it is in other cases of enforcing your rights. You were pushing me off your land or your foot. Given it was yours and you had no other remedy, it was morally permissible for you to do so. Likewise, it would have been permissible for me to push back, provided I had no less violent way of preventing your violence against me. But in our story I did have a less violent way: I could have refrained from stepping on your land-- or your foot-- in the first place.