Note: the following is a review of Hillel Steiner’s Essay on Rights that was originally published in the Canadian Journal of Philosophy, vol. 26, no. 2 (June, 1996), pp. 283-302. The criticisms it presents appear to apply to Steiner’s current views, including those presented in his essay on this blog. I therefore post it here in its original, unedited form. Sections VI through IX are most directly relevant to the debate over libertarianism and land.

I  Introduction

What is justice? And what is it for? Hillel Steiner says that his Essay on Rights will answer  the first question, in the process showing  that the second has no answer  (1). The book ends  by saying, ‘I’ve  offered no reasons as to why we should be just. Nor do I think that any can be found’ (282). Is this brutally self-effacing conclusion warranted?  Are the pros­pects for motivating what Steiner calls justice as bleak as Steiner himself seems to think?

Yes and  no. The conception  of justice set out in the opening  pages seems eminently well-motivated. ‘We unavoidably restrict one another’s freedom. And  justice is about  how those restrictions  ought  to be ar­ranged.’  Steiner immediately  adds,  ‘What it’s not about  are the ends which might be achieved by that arrangement’ (1). Yet what Steiner calls justice is very much about the end of mediating conflicts that arise as we pursue our personal projects in a social world. More perspicuously, he says ‘Questions of justice arise precisely where the moral permissibility of one person’s  restricting another’s freedom is not determined by the comparative  merits of the ends to which they are respectively commit­ted’ (2).

If the book is about justice, it is also about rights insofar as, ‘the elementary particles of justice are rights. Rights are the items which arecreated and parcelled out by justice principles’ (2). Accordingly, rights, as the elementary  particles of justice, have to be capable of resolving disputes.  In effect, rights supply  a verdict about who has the right of way. Rights (literally and metaphorically) allow us to navigate through traffic with minimal risk and inconvenience, conceding the right of way to other motorists when required by the rules of the road, without in any way implying that other motorists’ destinations are more important than our own.

Why not settle cases on the merits of the ends at stake? One problem is that comparative  merits of different ends are sometimes  hotly con­tested. A second consideration, not mentioned  by Steiner, is that even when no one would contest the comparative merits of the ends at stake, we often don’t want to settle things that way. Instead, we want to settle things in advance so that questions of comparative merit never arise. We want to get home at a reasonable hour. Stopping at every intersection to put our ends up for public inspection would slow us down, and might be rather undignified  to boot. We are better off with pre-established rights of way.

If that is what rights are for, then what happens when rights come into conflict? The crux of Steiner’s essay is that they cannot (2). Genuine rights are compossible, which is to say they form a set, the members of which cannot yield conflicting verdicts about who has the right of way. Readers who are familiar with Steiner’s work will know that compossibility has far-reaching implications regarding the structure, content, and authority of rights. The book’s main purpose is to fully explore those implications, showing how much can be gotten from such a theoretically spare begin­ning.

II  Freedoms and Unfreedoms

The first substantial chapter is a lengthy analysis of freedom. (Freedoms, Steiner argues later, are what  rights distribute.) Steiner begins with a methodological stipulation. ‘It’s assuredly not the job of philosophers to legislate on which linguistic (much less, moral) intuitions we may hold nor, therefore, on what conception  of liberty  we may employ. They utterly lack the authority to do so. Rather, their brief is the more modest one of indicating which set of intuitions can be held consistently’ (7). Well, then, let us ask whether Steiner’s methodological  intuition can be held consistently. Evidently not, for in the very act of proclaiming that phi­losophers can do no more than indicate which set of intuitions can be held consistently, Steiner himself goes well beyond  merely indicating which intuitions can be held consistently. And not for the last time.

Steiner says ‘a person is unfree to do an act if, and only if, his doing that action is rendered  impossible by the act of another person’ (8). He offers the definition again a bit later (32), claiming at that point to have defended it. The ‘if’ part seems obvious; the ‘only if’ part does not, and the ‘if’ part is what Steiner defends. Regarding the ‘only if’ part, it would seem that if I am locked in a prison cell, I am unfree to go to the theater regardless of whether  my going has been ‘rendered  impossible by the action of another person.’ If I accidentally locked myself in, I am as unfree to go to the theater as I would have been had someone else locked me in. Of course, it is Steiner’s prerogative  to settle the issue by fiat, and I see no harm in stipulating that the term ‘unfree’ applies only to inabilities caused by other people’s interference. I just find it odd that considerable effort goes into defending  one side of the biconditional when the other side is simply stipulated.

Steiner insists that no freedom (to perform a particular fully specified act-token) can be possessed by more than one person at a given time.

Although it’s true that both Becker and McEnroe were free to compete in the 1990 Wimbledon Men’s Singles Tennis Championship, it cannot be true that they were both free to win it. Of course, it’s often difficult to predict which one of many possible worlds will become actual. But there’s no possible world in which two (or more) such attempters can both be unprevented.(41)

Is this true? Perhaps, if we lean hard enough on the word ‘attempters’, but that would  take us away from the real issue. A slightly different question  illustrates  the problem: are Becker and McEnroe both free to concede the match? The answer is that Becker and McEnroe are each free to concede the match even when it isn’t possible for both of them to concede. A world in which neither of them intends to concede (i.e., the real world)  is a world  in which each is free to concede. On Steiner’s definition of ‘unfree,’  Becker is unfree to concede only if his conceding is rendered  impossible by the act of another person. Now, Becker’s conceding was, as a matter of historical fact, not rendered impossible by the act of another person, and neither was McEnroe’s. Therefore, contra Steiner, there is a possible world (namely, the actual world) in which two people can each have the same fully specified freedom.

What, then, is the real issue? What motivates Steiner here, I think, is the idea that any fully specified freedom possessed by more than one person would  not be compossible. As I see it, there is a perfectly natural sense in which I can be free to do X (for example, take a certain seat on a certain bus at a certain time) not because no one else could possibly do X but simply because no one else will actually do X. But that sense of being free to do  X is not Steiner’s.  As Steiner conceives of freedom, compossibility is built in.

Steiner goes on to ask, ‘What counts as an increase in a person’s liberty?’ (42) One common-sense requirement  of any answer, Steiner believes, is that enlarging our repertoire of available actions must not count as an unequivocal increase in liberty, for such a view would fail to integrate unfreedoms into the rubric. In Steiner’s words,

 Residents of classical Athens could not perform the action (act-type) of going to the refrigerator to fetch a can of beer. Are we more free than they were? Possibly, but not for that reason. For as we’ve already seen, whether I’m free to do this act-type entirely depends upon whether I’m free to do at least one of its tokens. Suppose I am. Does this imply that I’m more free than the classical Athenian? No, because there are many of these tokens which I’m unfree to do. My cranky neighbor, for instance, would no more allow me to fetch a can of beer from his refrigerator than he would give me the time of day. Our  technologically enhanced repertoire  of act-types also increases our inventories of act-tokens which we’re  prevented by others from doing.(43-4)

At the same time, he says, ‘Another commonsensical requirement of any freedom-metric is that it implies that my total liberty increases whenever, ceteris paribus, I become free to do an act-token which I was previously  unfree  to do’ (44).[1. The ceteris paribus clause is meant to set aside cases in which I become free to fetch a can of beer but at the same time become unfree to fetch a can of my neighbor’s beer. We might add that there are cases where, in becoming free to do an act, I also become obligated to do it. When I suddenly acquire the option of helping to push my neighbor’s car out of a snow bank, I am less free in virtue of no longer having the option of permissibly spending my time in more pleasant ways. (Or  does my neighbor have to make me physically unable to spend my time in other ways in order to make me unfree?)] Steiner  then considers  how we might assign numbers to particular freedoms. The point is to arrive at a number that would  reflect the extent of our total freedom, once freedoms and unfreedoms are each factored in. However,

 Observe how this leads to very peculiar results. Suppose that Blue is free to do acts A and B, the significance of each of which is respectively valued as +10 and +8, and that she is unfree to do C which is valued as -9. On whatever formula these figures are combined to yield the extent of her total freedom with respect to these three actions, the implication of removing the restraint against her doing C – and thereby making her free to do it – is a reduction in that total. Blue’s acquisition of the added freedom  to do C would entail a decrease in her overall freedom! To avoid such contradictions, while still integrating valuational magnitudes into computations of personal liberty, it would  therefore be necessary to exclude the use of negative numbers from valuational assessments of actions. (47)

The contradiction is peculiar, for sure,but we hardly need to resort to the even stranger proposal, ‘simply and boldly, to delete all negatively valued actions from  the set of actions which  people  are describable as free or unfree  to do’ (47). Let C stand for the option of jumping  from Blue’s third-floor office window. Notice that the value of actually jumping is not the same as the value of having the option of jumping. The value of actually  jumping  is -9, but  the value of having  the option of jumping presumably is zero?[2. That is to say, the option  of jumping from Blue’s third-floor office window presum­ably is worthless. I am not saying this value necessarily is zero. (It could even have positive value in a case  where  threatening to  jump  would put  Blue  in a better bargaining position.) The point is only that it need not be -9.] (It should be obvious that  the two  values  are not likely to be identical. Actually  jumping  is painful; having the option of jumping  is not.) Since the calculation is about acquiring the option of jumping, and not about actually jumping,we add zero to the total rather than -9, which means that those who like to attach numbers to freedoms (or values thereof) can avoid Steiner’s peculiar result without having to deny what is  plainly true,  namely that jumping from Blue’s office window can have negative value.  Nor do we need to give up Steiner’s intuition that  my  total  liberty  increases, other  things  equal,  when  I acquire option C. The  additional liberty  may  be worthless, but  it is, nonetheless, an additional liberty.

A consequence of Steiner’s view is what he calls the Law of Conserva­tion of Liberty: ‘there can be no such thing as an absolute loss of (or gain in) individual liberty’ (52). The idea (I think) is that, if we all suddenly acquire the freedom to brew beer in our basements, bottle it, and  put it in our  refrigerators, then as fast  as I can  put  beer in my refrigerator, thereby  acquiring a freedom to fetch it from my refrigerator, my neigh­bor also acquires unfreedoms to fetch that same beer. The total gain in freedom thus  comes out to zero. Would  Steiner  say that  there can be a net change in the total number of options? Or would he say that, as people in general come to have more options than  they  once had, there is a corresponding increase in ‘unoptions’ such that the net change is always zero?[3. Steiner would say the latter (personal communication, cited by permission).] I find  the zero-sum conception of liberty uncompelling, but it is interesting insofar  as  it  directs all  of our  philosophical attention to questions about how liberty is distributed. If sum-total liberty  is con­stant, then Steiner  does not need to defend  his rights theory  against the consequentialist argument that  invading some  people’s liberty  can be justified  in  virtue  of increasing the  total  sum.  The  total  sum  cannot increase, on Steiner’s view.

III  Compossibility

The heart of the next chapter is a comparison of two accounts of the nature of rights. Benefit Theory maintains that to hold a right is to be the person who stands to benefit from its enforcement. Steiner opts for Choice Theory,which maintains that the true rights-holder is the person who has the power to waive the claim. Blue may appear to have an unwaivable right (i.e., a right that Blue is not at liberty to waive) not to be enslaved. Steiner finds it more accurate, though, to look at Blue as a third-party beneficiary (62). Blue cannot waive her right not to be en­slaved because Blue has no such right in the first place. By analogy, if your florist promised you to deliver flowers to me, then I am a third party beneficiary of a right held by you. I may appear to have an inalienable right to the flowers insofar as I could not release the florist from his duty to deliver flowers to me even if I wanted to. But that appearance is created by the fact that the right, and consequently the power of waiver, is held by someone else. Blue’s apparent right not to be enslaved, for example, is actually a right ultimately held by a state official!

As evidence that some state official has such a right, Steiner notes that it is within the power of certain state officials to pardon offenders, to accept plea-bargains, to decide not to prosecute, and so on (69). One might object that surely it is ordinary citizens who possess rights like the right not to be enslaved. Steiner’s reply is that this ‘relies on the unstated assumption that a necessary condition of being a right-holder is being the beneficiary of a duty’  (66). It doesn’t really, though. Steiner’s own unstated assumption is that Benefit Theory is the only alternative  to Choice Theory. Steiner argues (successfully) that being a beneficiary is not a necessary condition of being a right-holder, but he makes an equally vulnerable assumption himself, namely that a necessary condi­tion of being a right-holder is having (and being able to exercise) the power to waive the right. Neither condition seems necessary. What we call rights do not comprise so neat and tidy a category. Most but not all of them are waivable. For that matter, in most cases holding a right goes hand in hand with being its ostensible beneficiary.

In one of my favorite parts of the book,Steiner offers an illuminating distinction between vested and naked liberties (75). A vested liberty is surrounded by a protective perimeter formed by the duties of others, effectively prohibiting interference by others. A naked liberty, such as the liberty to use a public telephone, lacks such protection, though it’s rarely without any protection. ‘This liberty is surrounded by a rather  more penetrable perimeter which both allows me to make the calls and allows others to prevent me from doing so by using that phone themselves, though not by assaulting me, taking my coins and so forth’ (75).

In that  case, your florist has taken on duties that turn out to be mutually exclusive. He has liberties that are both committed and naked. Your right that the flowers be delivered and the van owner’s right to the immediate return of the van are incompossible. Steiner says, ‘the salience of nakedness in the creation of incompossibilities is clear enough. Any duty which depends for its fulfillment on the exercise of a naked liberty stands  in danger of being  non-fulfillable’ (87). Moreover, committed naked liberties can come into conflict through entirely permissible or even obligatory actions. I may have a naked liberty to catch the only cab in town and may be committed to doing so in virtue of needing to make an appointment (88), but that is incompossible with your similarly naked but committed  liberty to take the cab for an appointment somewhere else.

The solution, in theory, is to not have liberties that are both committed and naked. Intensionally described liberties, though, inevitably are na­ked in some respects. Goal-based rights lack full extensional specifica­tion, thus leaving open possibilities for conflict. Their perimeters are not impenetrable (92). The trick is to contrive a way of describing domains of rights, i.e., sets of vested liberties (90), that avoid the potential for conflict inherent in intensionally described liberties. Accordingly, intensional descriptions of duties must be viewed as, at best, surrogates. They must be translatable  into extensional specifications. The duty to make one’s patient comfortable is not a duty to do whatever it happens to take. It is a surrogate  for an extensionally described duty to (upon request) help one’s patient sit up in bed, give him a pillow, and the like (100).

For any set of rights, ‘Extensional differentiability is the necessary and sufficient condition of their being compossible’ (99). The solution, then, lies in title-based domains, the elements of which are property rights­ time-indexed rights to physical objects -rather than rights to achieve goals (91). Because no two persons can simultaneously have rights to the same  physical  thing,  title-based  domains  effectively partition  action space  so that  no one’s  domain  overlaps  anyone  else’s  (93). In other words, ‘In a compossible set of rights, all rights are funded. The set of resources  respectively  required  for  the performance of each of their respective duties are specifiably distinct from one another’  (101). Any conflicts that do arise among time-indexed  rights to physical things can be adjudicated  by inspecting  their extensional specifications. Such in­spection will reveal who in fact has the right of way in any given case.

The nightmare, as Steiner calls it, is that ‘if a set of rights is incompos­sible, then Adjudicator will have in effect to supply such specifications ex post’ (101). Is that so bad? Must a moral system be fully compossible (i.e., guaranteed never to yield conflicting  judgments) right from the start, or is it okay to start with principles that may or may not be compossible, refining them so as to render them compossible in particu­lar respects as the need arises?[5. As Steiner uses the term, if rights are compossible, it isn’t just that people can avoid having their rights get in each other’s way. Rather, they could not put their rights in each other’s way even if they tried (89).]

Consider  that, as children, we embrace simple and straightforward moral rules. As we grow up,we modify them when circumstances force us to modify them. For example, the commandment ‘Do not lie’ that we embraced as children eventually is replaced by a more nuanced commit­ ment to honesty, one that gives way in cases that invoke certain compet­ing commitments, such as to avoid indiscretions that might gratuitously hurt innocent people. Our principles are as complicated as they need to be in order for us to deal with our actual circumstances – no more, no less.

Incompossibility, as a theoretical potential for conflict, is unimportant. Actual conflict is what forces us to modify our principles, not theoretical potential for conflict. If our principles are compossible from the start, then by hypothesis conflict will never actually occur. Nevertheless, the process of internalizing principles is not guided by the issue of theoreti­cal compossibility. Nor should it be. It is guided,as it should be, by actual experience. Our principles of justice take shape as means of resolving actual disputes, not theoretically possible disputes. Thus we embrace, and have excellent reasons to embrace, principles that are not quite compossible.

Steiner wants an escape from retrospective legislating, but such an escape is precisely what  an evolving precedent-based  legal system is supposed to provide, albeit gradually and in piecemeal fashion. When conflicts occur, judges need to refine the contours of rights so as to create compossibility  in that particular  region of the rights landscape. Moral agents sometimes need to refine the contours of their internalized moral commitments in a similar way. That’s  how it works in the real world. That’s how we find out where the incompossibilities are, and that’s how we know which ones are worth fixing.

IV Moral Dilemmas

The next topic is moral dilemma, as exemplified by Sartre’s story of the student  in Nazi-occupied  France who  must  decide  between  staying home with his ailing mother or leaving home to join the French resis­tance. How to make a consistent judgment about what to do? How can a moral code be structured so that an agent will be able to comply with it? As Steiner notes,

 J.O. Urmson has usefully distinguished three broad types of theory concerning the structure of moral  judging. Mononomic theories (such as utilitarianism) hold  that there’s no plurality of first principles or primary rules, that  there’s only a single primary rule, and that all moral judgments are inferable from it. Hierarchical theories allow that there can be a plurality of primary rules, and assert that there are ordered by priority rules (which may include ones demanding decision  procedures). And intuitionist theories, following Rawls’s  description, also embrace the plurality of primary rules, but deny that these are ordered. (113)

 Intuitionism, Steiner claims, cannot be a correct account of coherent moral judgment (114). Why not? Steiner’s conjecture is that ‘what intui­tionists refer to as “weighing up the pros and cons” or “judging the case on its merits” can amount to nothing other than the identification of a priority rule which orders the conflicting primary ones’ (115). So sup­pose that Blue, having read Sartre, eventually  makes up her mind and decides that the student should leave his mother and travel to England where he can join the French resistance. Steiner claims that ‘Blue’s moral judgment in favor of Sartre’s student going to England trivially implies that it’s not the case that he ought to have remained in France’ (116).

To those who believe in the reality of moral dilemmas,  though, the whole point is that this is not trivial. The student’s deciding to join the resistance will of course be extensionally equivalent to his deciding that joining the resistance is his only obligation. That much is indeed trivial. His felt experience, though, will be that of having  two compelling obligations and not being able to meet both.

Steiner argues that all dilemmas can be precluded from occurring. How? The answer is simple. ‘A necessary condition  of a situation’s confronting someone with a dilemma is that the primary rules conflict­ing with it are ones which he already affirms’ (119). To avoid dilemmas, then, one need only be sure not to affirm primary  rules that can conflict with each other.

The surest way to do avoid conflicts among one’s primary moral rules, though, is to avoid affirming any moral rules at all. Just say no. No moral commitments; no moral dilemmas. Steiner’s stated reason for requiring sets of rights to be compossible is that incompossible sets are self-con­tradictory. Now,if that were the only reason, then rejecting moral com­mitments altogether would be the way to go. It is by far the surest and probably the only genuine guarantee against incompossibility.[6. After mentioning an analogous proposal (in the context of interpersonal conflict) to do away with rights altogether, Steiner says, ‘Since the reasons for rejecting such a reform can only be moral rather than philosophical, I shall leave aside any further consideration of it here’ (82).]

One huge problem with rejecting moral commitments altogether, though, is that having the kind of commitments we have may serve an important purpose in our lives, despite the fact that those commitments might some day come into conflict and leave us facing a dilemma. If we should  find ourselves in a dilemma, one way to deal with it is to meet one commitment, fail to meet the other, and get on with our lives, simply accepting that we have failed to meet a commitment. Another  way to deal with the dilemma is to tell ourselves that the commitment we failed to meet was conditional. We can tell ourselves that it was understood from the start that we had other priorities as well, and that some of them were more important. But the latter story often would be false. Sartre’s student may have had reasons to be committed to his family. Such purposes may not have been well-served by a commitment like this: ‘Mom, I swear I’ll always be here for you, unless, of course, something  more important comes up.’ On the contrary, an unconditional commitment was (or could have been) exactly the kind of commitment he had reason to embrace.

Nor is anything  built into the fabric of human  life to preclude our having good reasons to have more than one unconditional commitment. Suppose I want to be loyal – really loyal – to my country and  to my family. It would be a mistake to say I shouldn’t want to be both on the grounds that we can conceive of situations where I would have to choose between them.[7. Steiner replies: Surely the assignment of categorical status to a commitment implies that the assignor has already anticipated the possibility of its conflicting with other commitments (personal communication, cited by permission). Perhaps Steiner is right. (If anything is clear, it is that reasonable people can disagree on the subject of moral dilemmas.) My view, though, is that a person could with utmost sincerity make two unconditional promises, not noticing the extremely remote possibility of their coming into conflict. As I see it, the unnoticed and extremely remote possibility of conflict does not change the fact that the two commitments are unconditional.]

If we anticipate conflicts between  our  prospective  commitments  to country and family, then we cannot embrace both in a genuinely uncon­ditional form. If we do not anticipate conflict, we are entirely capable of embracing both. If conflict unexpectedly materializes, we will have a dilemma on our hands. Be that as it may, the price of insisting that our moral code be guaranteed never to leave us in a moral dilemma is that we fail to embrace the kind of commitments that human beings have reasons to embrace and we fail to be the kind of moral agents that human beings have reason to be. Or so a pluralist might argue. If we are in a dilemma, such that we are forced to act in a way that is extensionally equivalent  to having no more than one unconditional commitment, we will feel a sense of loss, and  rightly so. If we insist on having  no more than one unconditional  commitment in the first place, then we pay the price up front, perhaps  needlessly.

V Gridlock

Justice is about who ought  to stand  down in cases of moral deadlock. Rights supply a reason for one of the adversaries to yield, without implying that the adversaries have reached agreement about the sub­ stantiveissue. The Adjudicator in Steiner’s dialogue informs the disputing parties that,

 We get the rights rule solution by asking “Who should have the freedom here?” and not by asking “Which one of your respective opposing actions is the morally better one?” We know that there’s no agreed answer to the latter question. That’s why, even if you both accept the rights rule solution, you’re not going to leave the shop in perfect rapport with one another. But each of you can agree to the same answer to the former question despite, or consistently with,your disagreement on the latter. (210)

But how to guarantee that my reason to yield the right of way will trump all my reasons not to? Steiner answers that, on pain of inconsis­tency, I can have no primary rule specifically enjoining me to violate rights per se. Accepting an injunction to violate rights contradicts accept­ing rights as rights in the first place (198). This seems correct, as far as it goes. The problem is that it does not guarantee the primacy of rights. For one thing, it does not rule out having primary rules enjoining me to do things that incidentally violate rights, as Steiner admits (198).

What if the only way to avoid  catastrophe  is to do something  that incidentally would violate someone’s rights? Steiner’s answer, in effect, is that we have to get our story straight (199). Look at it from the point of view of people whose rights we want  to violate so as to avoid catastrophe. ‘For them, whatever  would occur as a result of their rights not being violated simply does not amount  to a moral catastrophe. Because if it did, they would thereby  have sufficient reason to stand down and waive the correlative duties owed to them’ (200).

For several reasons, this will not do. To mention just one, consider that people whose rights stand in the way of averting a catastrophe may not know it, and there may not be time to explain. I hasten to admit that Steiner has a point of practical significance. As a matter of fact, people typically do waive their rights in emergency situations, and go to great trouble and sometimes great risk to be of help. Certainly they avoid getting in the way. Regarding the theoretical issue, though, the fact that looming catastrophes give people reason to waive their rights is beside the point. The theoretical issue is about what  happens when, for any number of reasons, people have not waived their rights. The original question has been left untouched: What if the only way to avoid catas­ trophe is to violate rights?[8. Another way to duck the question would be to claim that, if Blue forgives us for breaking her arm  in the course  of preventing a train  wreck, then it retroactively becomes  the case that we did  not violate her rights. (Really? Suppose we steal her silverware and  she forgives us for that too.) In any case, however many ways there may be of ducking the question, we still need an answer to it.]

For that matter, what if all of our alternatives involve violating rights? Steiner’s answer is that this is impossible. ‘The rights rule has to be such that, in any conceivable deadlock, only one of the parties is within his/her rights’ (201, emphasis added). The only way to guarantee this is to insist that rights must be compossible (202). We guarantee that there can never be conflicting rights by stipulating that genuine rights never conflict. The things we think are our rights may, for all we know, be prone to all sorts of conflict, but genuine rights cannot be.

When Steiner takes up the topic of duties to enforce rights, though, he seems to accept the possibility of unresolved  deadlock. Certain rights­ violations cannot be redressed. The victims can never be made whole. Suppose I can stop one such impending  rights violation or another, but not both. (Perhaps I am a police officer and  have a duty to stop such things.) Steiner concedes that, in that kind  of double-duty case, ‘it’s correct to describe my choice as defaulting on one of my two enforcement duties. Can any theory of justice do better?’ (206) Steiner may be right; perhaps no theory  of justice can do better.  But then, in light of this concession, it is no longer open to Steiner to insist on compossibility as the only way to guarantee that rights cannot conflict, for it turns out that compossibility is no guarantee.

VI  Equal Shares

Any workable dispute resolution mechanism has to begin with the idea that other people, right or wrong, sometimes have the right of way. Everyone has the right to be wrong (208). ‘Justice is a rule investing each person with a right to equal freedom…. A rule that distributes freedoms equally is obviously untainted by tendentious instrumental considera­tions of what that freedom is going to be used for…. Its freedom-alloca­tions can be used for actions which are good, bad, or morally indifferent. So it can underwrite a right to do wrong’ (216).

Abruptly, the conversation turns to apples. Steiner asks us to imagine that we are called to a meeting to decide on a principle of distribution. ‘Suppose what’s to be distributed  is a bundle of apples. Apples are a lot like freedom, as will soon become evident’ (217). The ensuing argument, though, is not about distributing freedom. It is about distributing natural resources. What would count as a neutral, impartial, and principled resolution, in keeping with our right to equal freedom?

Since each person  is invested with the same  power  as everyone else, each person’s antecedent title must  be the same as everyone else’s. Now, unless I’ve missed something, there are only two possible interpretations of what those titles are title to. One is that each person has a title to the whole bundle of apples. And the other is that each has a title to an equal share of apples. (219)

The first possibility is ruled out because the set of duties it entails are not compossible (219). Hence, by elimination, ‘each person has an ante­cedent right to an equal share of the apples’ (219). This is not to say there is an end-state  principle of distributive  justice mandating  that we have equal numbers of apples. As Steiner points out, such a principle would be violated every time I ate an apple (224). The right  to equal shares cannot be a right to an ongoing relation of equality. Instead, it is a right to be given, at the age of majority, an equal share ofthe world’s unowned resources.

Steiner later explains what equal shares means (268-72). If there are two people and twelve acres, an equal share is six acres. If we add a third person, then an equal share is four acres. And so on. Now suppose  the third person is a latecomer, arriving after the twelve acres have long since been divided and cultivated. In that case, the third person is entitled to the value of four (undeveloped) acres.

But wait. The introduction of latecomers changes everything. Go back to the original meeting, where people are discussing how to distribute a bundle of apples. Steiner said there were only two alternatives, and of the two, only equal shares was credible. Now we learn that his actual proposal is that shares be perpetually  held  in suspense, pending  the arrival of latecomers. Meanwhile, back at the meeting, the folks are saying, ‘If that’s what  he means by equal shares, then I can think of a third alternative: equal shares for us. To hell with latecomers.’ Let the record  show that I’ve repeated  this thought experiment many times under strictly controlled conditions, and the third alternative  wins by acclamation every time.[9. Incidentally, latecomers get most of the benefits under this regime, despite having officially been left to fend for themselves. Under the regime implied by the third alternative, the later they show up, the luckier they are. People who show up really early get to forage for nuts and berries, in the snow, with their bare hands; whereas we know from our own experience, people who show up late enough get refrigera­ tors, hot running  water, and pizza delivery. See David Schmidtz, ‘The Institution of Property,’ Social Philosophy and Policy 11 (1994) 42-62.]

VII Self-Ownership

Steiner takes himself to have established  that assignments  of freedom take the form of property  rights, and  that property  rights consist  of original equal-share rights and created derivative rights – an important distinction given that it is only original rights that justice requires to be equal (229). How then do we create derivative rights? Steiner begins by saying  that  slaves have  no rights  (231). To have any  rights  at all, a person’s self must not be part of someone else’s bundle. Steiner imme­diately concludes, ‘It follows fairly readily from this that our respective bundles of original property  rights must include at least ourselves. We must each be self-owners’ (231).[10. Steiner would  not prohibit our selling ourselves into slavery, though. Our right to ourselves is no exception to the rule that all rights are waivable.] The argument is painfully thin, but the conclusion is plausible enough. Suppose  we accept it for argument’s sake. The next step is to explain how people could derive rights to things other than their selves. Ownership is, in general, transitive. That is, if I make X from things I own, then as a general rule, I thereby own X. Thus, self-ownership grounds title to things produced entirely from self­-owned things (235).

That suggests a problem. Steiner notes that ‘each of us is the fruit of other  persons’ labor. How  can we each own what  we produce if we ourselves are others’  products?’ (237) If people own what they beget, then how can members of the next generation be self-owners? Steiner argues his way out of this quandary by invoking an assumption  tradi­tionally viewed as an objection to the Lockean story, namely the idea that labor-mixing does not suffice to establish unencumbered title to things produced from unowned inputs (247). If our parents do not acquire unencumbered  title to things produced  from unowned  inputs, then we have a way out if only we can argue that we are not entirely the products of our parents’ labor. Next, note that we are products  of our parents mixing their labor with their germ-line genetic information, and suppose that germ-line genetic information is a natural resource and that natural resources are unowned. Given all that, it follows that mixing labor with germ-line genetic information does not generate unencumbered title to one’s children (248).

What  form does  the encumbrance take? Following Steiner, let us simply assume it takes the form of parents owning their children only until the children reach the age of majority, after which children become self-owners. Children, Steiner says, have no rights (245). Minors are ‘at the disposal’ of their parents (248). When we cease being minors, though, we become self-owners, and acquire rights in the process.[12. Another way to look at it is to say children are self-owners from the start but  their self-ownership is temporarily encumbered by legitimate interests of parents. I thank Cecilia Nalagon for this suggestion. In any case, parents are not owners of their children. They are not at liberty to dispose of their children. Perhaps they are stewards. Their mandate is to do what they think is best for the child. Part of what is involved in carrying out that mandate is relinquishing control to the child as the child becomes capable of making his or her own decisions. And if parents do not carry out that mandate in good faith, they lose it; other adults (and the child too, I would  say) acquire the right to remove the child from  the parents’ custody.]

If Steiner’s argument that parental title is encumbered  by the contri­butions of unowned germ lines goes anywhere, it may not go in a direction he intends. If our parents’ germ lines are an unowned resource, and if mixing their labor with such a resource does not generate unen­cumbered  title to the products (i.e., to us), then we seem to inherit a bigger problem. Presumably,  we don’t own our germ lines either. In which case, how do we get unencumbered title to anything  at all? Our ownership of our labor is in doubt, since our labor power is the outcome of a process that uses unowned  germ lines as inputs. For that matter, our bodies themselves are phenotypic expressions of those germ lines. On Steiner’s assumptions, it is unclear how there can be any such thing as unencumbered ownership – even  of our own bodies. In any case, Steiner has given the old  Lockean argument  a new  twist. If the new argument  is no less vulnerable than the original, working through  it is at very least an interesting exercise.

I am left with the following thought. Do we really need a tortuously complicated story about self-ownership?  Why not just say that if you make something  with materials to which no one else has a claim, it’s yours? Likewise, if you make something  with materials freely given to you (or given up) by someone else, then you are making something with materials to which no one else has a claim. In that case, it’s yours. Further, if your person constitutes something made of materials freely given to you (or given up) by your parents or anyone else, then you constitute something to which no one has any claim. And if and when you make something out of yourself, then that too is yours.

VIII  But Who Pays For the Funeral?

Recall that, after proposing a questionable dichotomy between Choice Theory and Benefit Theory, and eliminating the latter, Steiner concluded that the power of waiver is a necessary condition of having a right. At the time, the argument seemed unmotivated, but it later turns out to be pivotal. From it, Steiner infers that the dead lack rights, since they, being dead, lack powers of waiver. ‘In short, there can be no moral counterpart to the legal power of bequest. So the justification of bequest, if there is one, cannot lie in the demands of justice. And the property of the dead thereby joins raw natural resources in the category of initially unowned things: things to an equal portion of which, as we’ve seen, each person has an original right’ (258).

Suppose we lose our rights when we die. What happens to the duties that were once correlative to those rights? Suppose I mail a parcel to you, but I die before the parcel reaches you. Would that give Steiner the right to intercept  the parcel and reroute it to a global fund?[12. Steiner says it would  (personal communication, cited by permission).] Why can’t I sell or give my house to you while I am still living? Does any moral or legal obstacle arise if I direct that the actual transfer will take effect at some future date? Does any moral or legal obstacle arise if I name as the future date the occasion of my death? Steiner says a moral obstacle does arise. There ‘can be no moral counterpart to the legal power of bequest’ because the legal power is merely a legal fiction (258). Really? Suppose I have the legal and moral right to give my stereo to my brother. Given that supposition, if I do in fact direct that my stereo is to belong to my brother, such transfer to go into effect at the time of my death and to be revocable before then, the fact of my so directing  is legal and  moral reality, not fiction. Or so it seems to me.

What could Steiner’s  proposal  mean in practical  terms? In the real world, people would not wait for the government to put their resources into a global fund. One way or another, people typically make sure their resources are either exhausted or else are in the hands of people they care about  by the time they die. Steiner gives no indication  of thinking it should it be illegal to sell property to one’s children, or to set up college trust funds for one’s grandchildren, or to give money to one’s favorite charity. Yet, unless  prohibitions  of such transfers  were ruthlessly  en­forced, the estates that are supposed  to bankroll the global fund would be nonexistent.

This cannot be what Steiner intends, since he says his proposal ‘man­dates a far more extensive  redistribution, and one more globally di­rected, than anything contemplated by recent, much less earlier, classical liberals’ (282). How would it do that, though? Steiner says the theory doesn’t depend on premises about incentive structures (282). Perhaps the conclusion that his proposal mandates extensive redistribution does not depend on incentive structures. On the other hand, if we were to predict that repudiating rights of bequest would bring about extensive redistribution by freeing up large estates to feed the global fund, our prediction surely would very much depend on premises about incentive structures. Indeed, we would be counting on incentive structures being unlike anything ever seen in the real world.

Does Steiner think the world would be a better place if we gave up the right of bequest? If so, why not just say so, explain why, and give us a better alternative? Instead, Steiner argues that we have to adopt his equal shares proposal, and apply it equally to latecomers, not because it is a good proposal but rather because at the imagined town meeting there were only two alternatives and the other one, namely ‘everyone owns everything,’ was incompossible.

IX Redressing  Overappropriation

The epilogue defines two types of redress. Bilateral redress undoes unlawful transfers from specific owners. Multilateral redress takes from those who appropriate too much and gives to those who appropriate too little (268). ‘In a fully appropriated world, each person’s  original right to an equal portion of initially unowned  things amounts to a right to an equal share of their total value’ (271).[14. Suppose I own six acres of downtown Manhattan and the time comes for me to give up the value of two undeveloped acres. Do I have to give up the value that two acres would have now? Or do I give up the value that two undeveloped  acres had centuries ago before the area was settled? If the latter is what a pro-rata equal share of natural resources is supposed  to be about, then most of us get the equivalent of that by the time the doctor cuts our umbilical cords. If the former, then what the latecomer is really asking for is not a slice of natural resources, but rather a slice of an economy the overall value of which is more a function of other people’s labor inputs than of natural resources.] When he says each person, he means each person in the world. Money taken from overappropria­tors goes to a global fund for compensating  underappropriators (270).

Note that if, at the outset, there are two people and twelve acres, and we each take six acres, we are not overappropriators. Indeed, if we take five acres each, we are  underappropriators. Even so, when the third person shows up, we pay. Thus, Steiner is not proposing to take only from overappropriators. Underappropriators lose as well. Thus, it is misleading  to characterize  this redistribution as a form of redress, for resources  are  taken  from people whether or not they did anything wrong.

Would people be friendly to strangers in such a regime? Probably not. The mere existence of other people would be a constant  threat. Steiner might not feel that way. I might not either. But the average person would. To some extent, the average person  already does, even now. Imagine taking the amount of xenophobia  we already  have in the world and making it, say, a hundred times more intense. Imagine a world of one hundred people, each with a one hundred foot wide lot. Your lot shrinks by a foot or so whenever someone new shows up, and expands by a foot or so  whenever someone dies (or dies  unexpectedly, with  no time  to disperse their assets). It sounds like Hobbesian war to me. It is staggering to think  we got to this conclusion after beginning with a conception of justice as a mechanism for resolving disputes.

Some egalitarian theories are repugnant even in theory, attractive only to those who  resent other  people being free. Steiner’s book is a sympa­thetic reply to egalitarianism as much as anything else. Steiner is trying to define  an egalitarianism geared  to giving  people more opportunity rather than less. He envisions a society in which people get a leg up when they  reach  adulthood, and then are pretty much left to stand or fall according to their own  merit – until they die, when the fruits  of their labors become seed capital for the next generation.

Existing institutions of inheritance, though, are meant to do exactly that. And in terms of inducing people to set something aside for latecom­ers, they have been enormously successful, although they hardly result in a ‘level  playing field.’ I don’t blame egalitarians for considering them manifestly imperfect. On the other hand, if the proposed alternative is to pool our assets in a global fund, consider this: In the United States, existing  government schemes for pooling assets, whatever their intent may have been, have had the actual effect of bequeathing to future generations a debt that is now five trillion dollars, increasing at  the moment by roughly a billion dollars a day. This hardly counts as setting aside equal shares for latecomers, yet that is what government schemes for pooling assets tend to do. They go bankrupt, leaving  latecomers to pick up the pieces.

So I agree that a world in which everyone starts out at the age of twenty-one with a substantial equal bankroll (and eventually returns the favor by letting their estates  revert to the common fund) is an attractive world, perhaps even a just world. If I could push a button to realize that vision,  I would be seriously tempted to  do  it. However, repudiating inheritance and  establishing a global  fund  is not that  button. It would not begin to realize Steiner’s vision.

Nowhere in the book is there evidence of how Steiner thinks  his conception of rights could be put  into practice. There is no history, no empirical research, and  there are hardly any real examples. Instead, the argument is carried by imaginative but misleading thought experi­ments. On the back cover, Brian Barry is quoted as saying that ‘those who are  unpersuaded will learn a lot by being forced to figure out where the rabbit was introduced into  the hat.’ With this, I agree. It is not a persuasive book. That does not seem  to be its aim. It is, however, a creative and  provocative piece of philosophy. It is intricate and eccentric.  The arguments are unlike anything I’ve seen  before.  Realis­tically, no one can expect  to cover as much  ground as Steiner without false steps, but in this book even the false steps are  instructive.  For anyone seriously  interested  in moral and social philosophy,  An Essay on Rights is well worth reading.