We discussed Nozick’s claim that rights are what he calls “side-constraints,” meaning that they constrain actions: no one can do anything that would violate another person’s rights.
I characterized utilitarianism as having two parts:
A hedonistic account of good and bad, according to which things are good (or bad) only because they produce pleasure (or pain).
A consequentialist or maximizing account of right and wrong, according to which the right action is the one that produces more good than the available alternatives (after their bad effects are subtracted, of course).
In the chapter we read today, Nozick takes on both parts.
His example of the experience machine is meant to challenge the hedonistic account of good and bad. Roughly, if you think it matters whether you are really doing things in the world, then you think there is something good (or bad) other than pleasure (or pain): really doing things or interacting with others (rather than merely thinking you are doing so).
But most of his fire is aimed at the consequentialist account of right and wrong. Nozick thinks it cannot accommodate individual rights.
The problem with consequentialist theories of right and wrong like utilitarianism is that they treat right and wrong as goals: an action is right if it fits the goal of promoting utility and wrong if it does not. The problem, as Nozick sees it, is that we do not treat rights as goals. He illustrates this with one of the stylized examples that are used to raise objections to utilitarianism: suppose the town sheriff can execute one innocent person to save ten innocents from dying in a riot that will happen if the innocent person is not executed. Obviously, if we are working only within a system of goals, and one of those goals is to save innocent life, then we should kill the one to save the ten. But Nozick says that this ignores the rights of the one not to be killed. The sheriff should respect rights first and doing so rules out killing the one innocent person.
How would you respond if you were a utilitarian? The first thing you should do is deny that violating individual rights really would maximize utility. The example as given ignores all the bad consequences of executing an innocent person and it makes a highly artificial assumption that killing the one innocent person is the only thing that can be done to save the ten. In a more realistic example, utilitarianism would not recommend violating individual rights, a utilitarian will say.
But suppose we insist on the facts as stipulated in the example: it really is true that there is nothing else to do to save the ten innocent people. The ten orphans are innocent too and, if utilitarians are really genuinely honestly cornered into making a choice between one innocent life and ten innocent lives, they are comfortable with using an old standby: arithmetic. The choice is terrible, but if it’s forced on you, it’s better to lose one life than it is to lose ten.
This is a powerful one-two punch for the utilitarians. Still, Nozick points out that this is not the way we think about rights. We do not think that whether someone’s rights are to be respected or not depends on calculating the effects of doing so. And we are not willing to engage in the kinds of trade-offs that the utilitarians claim are sensible.
Nozick thinks this shows that we do not treat rights as goals, where we would try to maximize achievement of the goal. Instead, we treat rights as what he called side-constraints. In this case, the innocent person’s rights act as a constraint on what the sheriff may do: the sheriff may not do anything that would violate the innocent person’s rights. If that is incompatible with reaching the goal of saving as many innocent lives as possible, that’s the way it goes. Individual rights take priority over the achievement of even worthy social goals.
Why doesn’t our poor sheriff face a dilemma? The side constraint “never kill” tells him not to kill the innocent guy and the side constraint “protect the innocent from being killed” tells him that he must kill the innocent guy (in these admittedly weird circumstances).
The answer is that Nozick only believes there is one side constraint at play in this case. We are forbidden from killing, but we are not required to prevent others from killing. Thus there is no dilemma here. The sheriff may not kill the innocent person. If he can save the innocents without violating a side constraint he may do so. But he is not required to do that and so there is no conflict of requirements here.
This would be irrational if rights were based on goals, such as the goal of minimizing the number of innocent deaths. But, again, Nozick does not think rights are based on goals. His case against utilitarianism is that the utilitarians improperly treat rights as if they were derived from goals.
Suppose Nozick is correct and that rights are not derived from goals. What are they based on?
Nozick’s attempts to explain why we have rights that are side constraints take him through several long and interesting digressions. He takes it for granted that our rights must be based on some natural features that we all share and he went on a search for those features.
He concluded that the natural features in question are ones that enable us to guide our lives by plans. That, in turn, is said to be important because it is how we give meaning to our lives (Nozick 1974, 48–50).
While suggestive, Nozick’s arguments are thin. As James noted, he sometimes tries to get by with a quick reference to a phrase from Kant about treating people as ends and not means. We will have to pay attention to this later. Nozick’s version of libertarianism rests heavily on his conception of rights. We will need to see whether it is enough for him to show that rights take the form of non-utilitarian side constraints or if he needs a deeper explanation of what rights we really have.
Nozick and Mill endorse quite similar principles. Mill’s harm principle holds that individuals should be left free to do anything that does not harm others: “the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection” (Mill  2000, ch. 1, par. 9).
Similarly, Nozick holds that “there is no justified sacrifice of some of us for others” and that there is “a libertarian side constraint that prohibits aggression against another” (Nozick 1974, 33).
However the philosophical bases of their respective versions of libertarianism are quite different. Nozick thinks the state is limited to protecting rights. Mill has to think it is committed to maximizing overall utility, even if this comes at the cost of rights. Furthermore, Nozick maintains that individual rights are derived from features of the individuals themselves and that one person’s rights cannot be balanced against another’s. Mill, on the other hand, derives individual rights from overall utility. If protecting individual liberty were not the best way of promoting the social good, Mill would have to reject it.
Furthermore, they have different opinions about what kinds of liberty are most important. Nozick emphasizes economic liberty. He thinks that “the state may not use its coercive apparatus for the purpose of getting some citizens to aid others.” Its sole function is to protect people “against force, theft, fraud,” and violations of contractual agreements. Anything more is “redistributive,” limiting one person’s liberty for another person’s sake (Nozick 1974, ix). Mill believes that there is a utilitarian case for economic liberty. But he also believes that it is separate from the one he made in On Liberty (Mill  2000, ch. 5, par. 4). On Liberty defends individual liberty of thought, expression, association, and determining how to live. Mill regards commercial activity as something different.
Mill, John Stuart. (1859) 2000. On Liberty. Edited by Mark C. Rooks. British Philosophy: 1600-1900. Charlottesville, VA: InteLex Corporation.
Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books.