 I'm going to be speaking this afternoon on a very famous book, Robert Nozick's Anarchy State in Utopia, which came out in 1974. I remember very well when the book came out in 1974. Of course, I was only in my early 50s at that time. To prepare for this talk, I re-read the book, and it brought back many memories to me of going over various points with Nozick. I would sometimes think, oh, I remember discussing this note with him and had a certain objection here. It was in part reliving the past when I re-read the book. The book many people find Anarchy State in Utopia a difficult book because Nozick moves extremely fast over a vast number of arguments. In that way, he was very much like Murray Rothbard. They were both extremely fast and covered great many arguments. As you may know, they didn't get along personally very well, but they were in that way very much alike. As I'll try to be showing, Nozick was very much influenced in the way he argued by his study of Rothbard. We can see a great deal of the book as a struggle, in my view, unsuccessful of Nozick to differentiate his own position from that of Rothbard. Some people will say about Anarchy State in Utopia. They don't like it because they say, oh, it's a very unsystematic book. It isn't well organized. Nozick just keeps coming up with one thing after another. You never know what's coming next. In fact, the book is quite well organized and systematic. In some ways, the strand of the book I'll be discussing today is a very intricate argument. It's rather hard to follow, but it's a lot of fun trying to figure it out. If you don't see it, it is a very systematic argument. The fundamental principle of the book is that individuals have rights. There are things that you can't do to individuals without violating those rights. So individuals have rights. Some people, like Thomas Nagel in a famous review of the book that came out in 1975, said, well, Nozick just says people have rights, but he doesn't give arguments that they have rights. In fact, he does have arguments that people have rights. One of the ways he tries to defend the view that people have rights is the criticism of utilitarianism or consequentialism. On this view, we can make consequentialism, suppose we're trying to determine what should we do, what is morally required in a situation. The only natural answer is that we ought to do what's best. We ought to take the best alternative. That would be what is morally required. It seems really common sense. Well, if we're trying to do what's good, we should do what's best. Nozick thought the plausibility of that view stems from an application of individual rationality, application of decision theory, which he was one of the pioneers. His 1963 doctoral dissertation was on decision theory. Supposing you're a particular person, you're trying to do what's thinking, what's best for you. You would think about the different options available, and you'd think, well, what would be the utility of each one, and then you would multiply that by the probability of that happening. The one with the highest utility multiplied by the probability was happening would give you your preferred option. This is basic decision theory. He thinks, Nozick thinks, well, the plausibility of the utilitarian view, which is you would pick the moral option, the one that has the best consequences overall, is that you would apply this principle of individual choice to the society. He'd be saying, what is the morally best thing to do is what would maximize the best consequences for the society. And Nozick said, well, what's wrong with that? He says, here he's following Rawls in The Theory of Justice 1971. He says, this doesn't take seriously that we're talking about separate individuals. There isn't a collective entity that we can talk about maximizing the welfare of that entity. Let's say, for example, supposing you went to the doctor, and the doctor said, unless I amputate your arm, you're going to die. So you would think, well, it's worth it to give up my arm to preserve my life. So Nozick said, but this isn't, when we're talking about society, we can't make trade-offs like that in the same way. We can't say sacrificing some people will make society as a whole better off that each person has an independent existence that's entitled to moral consideration. Now, there are some utilitarians who, like RM Hare, have tried to reply to this, so utilitarianism does take the separate existence of people seriously because it considers each person is counting for one when you're calculating the utility, so your utility count, amount of utility counts the same as mine does, but this is really Nozick's point that there isn't an entity whose, if we apply, we're talking about society, there isn't an entity that we can talk about whose utility is being maximized. So what we have is individuals, separate individuals, and Nozick thinks it's plausible to think that individuals have, if we consider the separateness of individuals, that individuals have rights, that limit our pursuit of consequences because individuals can't be sacrificed for the sake of some total amount of consequences for society. We have only separate individuals. So these rights are what he calls side constraints rather than maximizing principles. He took this term from work on computers. A good way to understand this for those of you like me who aren't very knowledgeable about computers is to think of these constraints as rules, somewhat like rules in a game. For example, in the recent soccer World Cup, players aren't allowed to grab the ball in their hands and throw it into the net. They can't say, well, if I did that, we'd get so many penalty points, but this is to be weighed against whatever the advantages would be in scoring. If I did that, it's just that's not allowed, that limits the action. So rights are, he knows if you like that, they're constraints on what we can do. They limit the choice of options. So if we take rights as side constraints, then we're not trying to minimize the number of rights violations by people. For example, here, this is a principle some people find hard to understand. They say, well, if rights violations are bad, doesn't this mean we should try to minimize the number of rights violations? But no, he says, no, that's not right. An example might be this, which I remember discussing my friend, Father James Sadowski, in the Catholic Church priests are forbidden under any circumstances to violate the secrecy of the confessional. If someone confesses a sin to a priest, the priest can never disclose what the person has told him. So I raised the issue with Father Sadowski, which has come up by a lot of people raising it. Supposing a priest confessed to another priest that he regularly violated the secrecy of the confessional and intended to do so in future, then if the priest that he confessed to reported him to the authorities, they'd stop him from saying confession. So wouldn't that minimize the number of rights violations? Because if he keeps it secret, then this priest is probably going to go around violating the secrecy of the confessional an indefinite number of times and you're not supposed to do that. But this is not the way the principle would be applied in the church. The priest is forbidden under any circumstances to violate secrecy of the confessional. So he couldn't report this priest from doing that. So that would be an example of the side constraint. It's a limit on what you can do. You're not trying to minimize number of rights violation. Then we could come up with an even more unusual example. Supposing if I violated rights in some way, I'd be preventing future rights violations by myself. I couldn't do that either. It's just that each time if the rights are side constraints, I can't violate the right. I'm forbidden to do anything against rights that they just a rule stopping me from doing things. So what happens if we have a really catastrophic situation supposing say like in the movie, Doctor Strange Love, unless somebody breaks into a Coke machine, then the whole world would be blown up. Then Nozick said, well, then maybe we could that rights would lapse, but he doesn't commit himself to that. Some people misinterpreted the book. They say, well, he says that if it's a catastrophic situation, then rights don't apply anymore. But he doesn't really say that he doesn't reject it either. He just says that's one thing that has to be, might be if it were really catastrophic, perhaps the rights do lapse. Now, we know if suppose we accept that individuals have such rights of this very strong form. What are the rights that people have? And what Nozick comes up with is a very Rothbardian view with some differences. He says, well, individuals do have rights over their own bodies. He uses the phrase self ownership just once in the book where he refers to the classical tradition of self ownership. But he doesn't generally use that phrase, but it's clear he accepts that principle. Then he also argues that property rights are individual. The world starts out unowned and people have to do things to the world in order to acquire that. He doesn't specify what the principle of initial acquisition is, what you have to do to acquire property. But it's clear he accepts some such principle. There are some differences from Rothbard in that I'm not going to go into in that he accepts something called the proviso or the Lockean proviso, and Rothbard doesn't, that Nozick thinks that before people started acquiring property, everyone was free to use certain objects where they couldn't stop other people from using them also. They couldn't exclude other people. So if you acquire property, you're making people worse off in the sense they can't use the objects anymore. But it's very easy in his view to meet that proviso. There's a difference. In Rothbard's view, using something just is acquiring it. The first user is the one who acquires it. So in practice, the views wouldn't differ that much. They both accept this principle of you acquire property by doing something to it on an individual basis. There's a famous case that Nozick discusses where he's talking about Locke's principle of labor mixture, and he gives a famous example. He says, well, supposing you throw a glass of tomato juice in the ocean, does this mean you acquire the ocean? Obviously not. A lot of people, unfortunately, have taken him to mean that he's rejecting some sort of principle of initial acquisition by doing something to the property. But he's just, in giving that, he's just saying we need to specify what the principle is, and he doesn't think people have really been able to do it yet. He's not rejecting such a principle of initial acquisition. On the contrary, he accepts it, and I can say I know that because he told me that himself. So people who argue otherwise are not correct. Now, Nozick's project is a very interesting one. He's starting off, he says, well, individuals have rights over their own bodies, plus strong property rights. So this is really identical, except for these minor differences in how we acquire property with Rothbard's anarchism. In Rothbard's view, we start off with individuals acquiring property in a very similar way to Nozick's view. What Nozick is trying to do is against Rothbard, Rothbard says, well, if we have these individuals with these strong property rights, then we can't have a state because a state is violating people's rights. Say if you have a protection agency, it's forbidding competition with that, with other agencies or independents with itself, plus it's taking resources from people by taxation. They have to pay for these protective services if they don't want them. So Rothbard says, well, you start off with individual rights and including strong property rights, then you can't legitimately get to a state. So what Nozick's project is, is to say, well, we can start off with Rothbard's anarchism. Then he can show, he tries to show by steps that would be to each person's advantage that are morally acceptable, that we would arrive at a state or at least what he calls a state-like entity from that. So you say, well, we can show there's a series of steps if we start off with what Rothbard postulates, we can get to a state in morally legitimate ways that are to everybody's advantage. Now, how does he think this process goes? Well, supposing people are trying to defend their rights, and they join protective organizations who join a defense agency to help enforce your rights. So it might turn, there are several possibilities, but it might turn out that one agency is much better than other agencies in protecting people's rights. So if that happens, there'll be a cascading effect where people will tend to join that agency and this agency is better than the other, so people will try, will converge to that. Nozick compares this to the process that Mises writes about in theory of money and credit and human action and how if you start off with commodities and have barters, some media of exchange will be more better than others, people will converge on them and then will eventually have money. So Nozick suggests that a protection agency might turn out to be something like that and one is better, people will converge on it, then we'll join that, so we'll just end up with one protection agency. Now, one thing we have to be careful here, supposing Nozick's right and everybody joined one protection agency, then we wouldn't have a state, we just have everybody a member of the same protection agency. The problem with Nozick's argument is that he needs to be the case that there are some people who don't join the protection agency because as we'll see later, there are things the dominant agency can do to those people that result in a rather state-like entity, but if everybody joins the agency then the whole process wouldn't result in a state. So it's rather a tension, not a contradiction. Nozick said, well, here's a motive for people that would apply for people joining this single agency, but at some point it's not, this motive won't work fully, so some people won't join it. It's a bit of a tension there. I suppose that doesn't happen though supposing there isn't a single agency that is dominant in that sense. Then Nozick suggests, well, it's likely either that there'll be some agencies that are dominant in one area and others will be dominant in a different area and there'll be borderline areas where there will be struggles or differences in how dominant each one is. So then yet another possibility is that we could have a series of agencies that they each or they have a procedure. What happens if they come into conflict? They'd have a procedure on what to do in case of conflict. Say they'd agree on an appeals procedure and then Nozick thinks then in that case they'd be forming a single agency just because they'd established a process for what to do in case of appeals. It could be questionable why does this make them one agency? Maybe because that's very similar to what Rothbard says. There'd be an appeals procedure if different protection agencies came into conflict, but why should we call that one agency? But I don't think much turns on that. That's just a semantic point. So Nozick really has rather, if you take this last option, Nozick has really defined the dominant agency into existence and says that he's calling an appeals arrangement a single dominant agency. I don't find that plausible because I say it doesn't affect the substance of the argument very much. So it says I've already covered that. He thinks the reason for the, if you have this balance of agencies where one isn't strong with the other, they'll find it in their interest to come up with an appeals procedure rather than fighting it out among themselves. That'll be very costly so they'll come up with an appeals procedure. Now, now we come to the fun part, which is, okay, we've got this dominant agency. Now he embarks on an extremely complicated argument to show that the dominant agency will first transform itself into what he calls the ultra minimal state. Once it's done that, or at the same time, it's obligated morally to go from ultra minimal state to the minimal state. And if it does that, then you'll have the state like entity. So now we start the complicated argument. What happens if somebody violates your right supposing say somebody just comes up to you and somebody comes up and steals your car while you're at this lecture, you're parked outside and you go out and you've discovered someone has driven away with your car? What happens then? Well, you're clearly entitled to compensation. You can get your car back plus, or if your car isn't available anymore, somebody's already dismantled it and sold all the parts. You're entitled to the value of the car plus various other items such as whatever, the cost of searching for the person who's done this plus maybe psychological costs to you. So whatever you want to compute these, but you're entitled to compensation. And the question comes up, is this enough? Is it enough that you get compensation for a rights violation? Now here we arrive at a central misunderstanding of the argument, anarchy, state, and utopia. Some people say, look, Nozick really gives up on rights because he started off. He has an extremely strong view rights or side constraints. Once you have rights, that's it. You can't violate them except maybe in a catastrophe, but then he abandons that. Then he says, well, it's all right. You can violate rights as long as you compensate people as long as you restore them to their previous utility or welfare level. So he's really abandoning his position on rights because he's saying, as long as you're free to violate rights as long as you compensate the person for the rights violation. And ironically, some of the people who raise this objection to Nozick say, oh, Nozick is terrible because he's abandoning rights. Some of those people favor a pure restitution theory of punishment. So they're the ones who are in fact guilty of what they're accusing Nozick of. But that isn't Nozick's view. Rather, what he's not saying, you're free to violate rights as long as you compensate people. He's asking the question, what happens if you do violate rights? And he says compensation usually isn't enough. Why not? Supposing somebody stole your car. Why wouldn't it be enough if the thief had to pay you back for your car plus everything you've lost by it? Well, first there's a problem we could call gains from trade. Supposing, say, if the thief hadn't stolen your car, say, he just wanted the car and he approached you and said, I like to buy this car. You might not be willing to trade it for whatever the current blue book value of the car is. You might want much more for the car than that. Why should you just be put in the same utility level as before from an exchange? Why shouldn't you be able to say, look, I'm not going to give you the car unless you give me more than that. You could insist on more of the gains from trade than simply putting you back to your initial level of welfare or utility. So that would be one central reason against saying compensation is enough. Another problem is a certain kind of fear. Supposing you knew that someone in your neighbor was going around breaking people's arms. You just like going around snapping people's arms. But you also knew that if the person did that to you, you would get full compensation for the person's breaking your arm. Again, including all the psychological costs. You say you made you afraid because you were just thinking about being injured. You made you physically afraid, so you'd be compensated for that too. But so supposing though you know someone's in the neighbor is going to do that. But in fact, the person doesn't break your arm. You'd still feel the fear you'd be afraid he's going to break your arm, but he wouldn't have done anything so you wouldn't get any compensation for your fear. So this is a reason it's very important knows its argument. Why he thinks compensation in this kind of case isn't enough for rights violation because some people would be subjected by certain kinds of rights violation. Some people would be subjected to a particular kind of fear that it isn't morally legitimate to impose on them, but they wouldn't be compensated for that because their rights wouldn't in fact be violated. So given the problems posed by full compensation, then the question comes up. Why don't we just prohibit all rights violation? And here again, this is another very big misunderstanding of knows it. He isn't asking me, why can't we just prohibit all rights violations? He isn't saying, well, why can't we say violations of rights are always wrong? That isn't what he means by prohibition. Prohibition means the imposition of an additional penalty in addition to compensation. So the question he's asking is why can't we say that besides having to pay the full compensation for rights violation, why shouldn't we always say that the rights filer has to suffer some other penalty as well? Why don't you have to pay something extra? And he thinks that in many cases, this extra penalty's prohibition is justifiable. But to answer the question though, why don't we always prohibit rights violations? Again, in the sense of imposing an additional penalty for rights violation, his answer is that it isn't always possible to get prior consent of the person whose rights you might have reason to violate. Supposing, say, take a famous example, you're lost, you're climbing a mountain, you're lost in a snowstorm and you see someone's cabin, you need to break into the cabin in order to stay alive, but you can't locate the owner. Would we say you could be, if you do break in, you'd have to compensate the owner for any damage you do to the cabin? But would we want to say you can have an additional penalty imposed on you because you didn't get his permission, even though that you weren't able to get the permission. It doesn't, at least it knows if you always seem plausible to say we should have this additional penalty besides full compensation. So now we have, after going through this material about risk, now we can apply this to the question of the competing protection agencies. Now what happens if you're in a client of this dominant agency? We can imagine there's a single agency or group of agencies that is the strongest one in a particular area, and you get into conflict with someone who isn't a client of that agency. So, Nozick says, well, the people, the one who isn't in that agency might decide to impose what Nozick calls a risky decision procedure on you. That's to say, if an agency, you imagine someone's violated your rights and you want to take action against this person, you think, say, somebody's stolen your car, you go after the person who you think has stolen it, then how do you determine the guilt of that person before you start taking action against them? You might employ a procedure that is risky in the sense that it has what some people would consider too much of a chance to find guilty people who aren't guilty. There are various procedures people could use, each one will balance finding some people, finding people who are guilty, who aren't guilty with finding people innocent, who in fact are guilty. So there'll be various ways of balancing those factors. So it might be that the people will impose a procedure on a client of the dominant agency that is too risky from the point of view of people in the dominant agency. So what happens then? Now in Rothbard's view, just the procedures doesn't matter. The question just is, is the person guilty or not? If you apply whatever procedure you want, the person turns out to be not guilty, then you're liable for compensation or punishment or whatever for violating that person's rights. But the procedure doesn't matter, there aren't any procedural rights. But Nosey questions this, he thinks that if someone applies a risky decision procedure to you, you would have the right to resist it. And why is that? This is precisely because this problem of fear that I mentioned, you might think that if you know that an agency might apply risky decision procedures, you might feel fear because suppose they say, if we find you guilty of crime will execute you or say you're guilty of theft will impose these, will require you to pay compensation. If you can't, we can force you to labor till you can pay. So you might feel fear that they're going to apply the risky decision procedure to you and you will feel this fear even if they don't. So if say they apply the procedure to you and in fact you turned out to be not guilty, then you'd be compensated for it. But what if they didn't apply the procedure to you, then you'd still feel fearful and you wouldn't be compensated because they haven't done anything to you, just you'd feel fearful about it. So in Nozick's view, the dominant agency can prohibit other agencies or independence from applying these risky decision procedures to its clients. It can say to them, if you apply these principles to our clients, we're going to impose a penalty on you for doing that because of this problem of the risky decision procedures will impose risk. Fear on our clients. Now, an objection here that I'm sure will have occurred to nearly everybody is why does the dominant agency have more rights than other people? Why can't other agencies say, well, we're going to prohibit the dominant agency from imposing what we consider risky decision procedures on us? Why this asymmetry? But as usual with Nozick, he's thought of this objection. He says, no, this isn't right. He's not claiming that the dominant agency has more rights than other people. It isn't the dominant agency can impose prohibit risky decision procedures on its clients, but other agencies or independence can't. They can operate in terms of their own conception of risky decision procedures. But if there's a conflict, the dominant agency by hypothesis is going to win because it's the strongest one. So it will be able to impose its view of what the risky decision procedures are. So see, it isn't that it has more rights. It's just that its view of the matter will prevail. So if a dominant agency does this, it will be what Nozick calls an ultra minimal state because it will be deciding on what the acceptable decision procedure are for determining whether people are guilty of rights violations. So it can stop other agencies or independence from applying these procedures that it doesn't accept to its clients. But here we have a problem because the independence in applying such procedures aren't doing anything morally wrong. They're just following their own ideas of what proper decision procedures are. It's just they differ from those of dominant agencies. And Nozick suggests, well, since they're not doing anything wrong, they're owed some kind of compensation. The dominant agency has to compensate them in some way because it's prohibited them from applying what their conception of risky decision procedures to its client where they're doing so isn't doing something that's morally wrong. Now, as if this weren't complicated enough, we get into even more complications. As you can see, this is really tremendous amount of fun. There are few books that really is enjoyable as anarchy, state, and utopia. I must tell you, how do we determine what the compensation is? The dominant agency doesn't have to negotiate a market price in Nozick's view for prohibiting independence, meaning by independence people aren't clients of the agency from applying risky decision procedures to its clients. Why not? Here we get into another concept, what Nozick calls a productive exchange. Productive exchange is kind of a standard exchange where both people are better off, say, if I have oranges and you have apples, say, we exchange my oranges for your apples, we're doing this because we have reverse preferences. So we're better off as a result of the exchange. We both want to make this change. We're better off. So that's what Nozick calls a productive exchange. But there are also cases of what Nozick calls unproductive exchanges, which is you're making an exchange and you're better off making an exchange than not making an exchange, but you'd be still better off if the other person didn't exist at all. For example, this is one of Nozick's objections, a blackmail. He says, supposing he mentions Rothbard's argument, in blackmail the blackmailer is selling the service of withholding information for a price, so aren't the two people better off because of the exchange? Nozick says yes, but the blackmail victim would be still better off if that exchange weren't offered, the blackmailer didn't exist at all. So I don't think this is really a valid restriction on exchange, but that is the view Nozick takes. So he says, well, here the dominant agency would be better off if these other independents didn't exist with their own conceptions decision procedures, so they don't have to pay a full market compensation. What they can do instead is offer cut rate policies to these independents. They can say, we'll protect you from with our agency because you won't be able to protect yourself. You don't have to buy our policy, but if you do, you won't be very well off because then you won't be able to protect yourself. I'll just conclude, there are quite a few problems with Nozick's derivation of minimal state. Supposing he's right that people would be fearful if risky decision procedures were imposed on them, so they have this fear, this doesn't show what Nozick needs, that isn't sufficient to get to Nozick once for his conclusion, because he has to show not just that people would be fearful if risky decision procedures were imposed on them, but he'd have to show that the marginal reduction of fear from prohibition as against full compensation would make a significant difference. And he hasn't proved that would be the case. He hasn't shown that prohibiting these procedures would rather than having full compensation would make people substantially less fearful. Another problem would be the ultra minimal state might find if its clients wanted these procedures to be prohibited, it wouldn't pay them to do that. Worth more to the agency to keep allow risky decision procedures rather than provide services, low cost services to the people whose procedures are prohibited. The last problem is that supposing Nozick's argument is all right, he still wouldn't gotten to a state because there might be people who are not able to protect themselves to threaten other people at all, so they're too weak to credibly to threaten other people, so they don't pose any risk of imposing risky decision procedures on them. So in that case, the state would be under the ultra minimal state would be under no obligation to protect them at all, and we wouldn't have the universal coverage that Nozick requires for the existence of the state. So in conclusion, I don't think Nozick's argument for the minimal state works, but you can see I hope the great complexity and analytic interest of his argument in Anarchy State Utopia. Thank you.