Toward the end of his talented, diverse, and very long book, Robert Nozick embraces the idea of philosophy as an art form, and of the philosopher as a literary creator who works with ideas. This reinforces an idea that may have already occurred to the reader; if this book is in some way like a literary work, it is clear what kind of literary work it is like.
Nozick, when young, wrote several articles of startling brilliance, originality, and, in some cases, formidable technical resource, in such fields as the formal discipline of decision theory. He then produced the notorious Anarchy, State, and Utopia, a lengthy book which advocated individual rights, libertarianism, and a minimal state; attacked public welfare; discouraged redistributive social justice; and represented taxation as forced labor. It was very clever and not very pious, and gave a distinct impression of hard-talking heartlessness (though the genuinely heartless publicists of the right who welcomed it as a philosophical liberation failed to notice that Nozick was committed, most probably, to believing that most of America belongs to the Indians).
After the hard, scandalous success we should not be too surprised to find now a book that has deeper and more spiritual ambitions, which covers many large traditional subjects, and which devotes the same outstanding talents not just to solving puzzles or showing off, and not at all to slamming the pious, but to reaching toward more speculative and awesome reflections on the meaning of life. This is Philosophical Explanations. It is an attempt at the Great American Novel of philosophy.
Like most other such attempts, it fails. It is an extraordinary mixture. In part, it is as brilliant and exciting as anything in contemporary philosophy. Quite often it is suggestive and interesting. Sometimes it is very bad, and at moments it is so deeply awful that it is only by considering the Great American Novel syndrome that one can see how it came about. A feature of that syndrome is the disposition to take the size of the attempt for success itself. That can affect readers as well as the writer. Many large things have already been said about this book by commentators. One is that it introduces new philosophical techniques, and is likely to influence the way in which philosophy is done.
Nozick does offer some claims to a distinctive method, but he sensibly makes much less of his originality in this respect than the commentators do. The idea, borne by his title, is that philosophy should try to explain things, rather than offer proofs. It is not altogether clear what he means by this distinction. He wants to avoid “coercing” people with attempted proofs to inescapable conclusions; he does not want to proceed by rigorous deduction from self-evident premises. But it is not obvious how that aim is related to offering explanations: some explanations (some mathematical ones, for instance) themselves proceed in that way. He recommends something else again when he tells us to proceed in a tentative rather than a dogmatic spirit—there can be tentative suggestions of proofs, and there can be dogmatic explanations. Nor is any of this the same as the aspiration, which he also expresses, to make a philosophical work that is more like the many-columned Parthenon than like a tall, thin tower, so that when bits fall down, as they will, something of beauty may be left. That analogy poses the question, not whether the work consists of proofs, but whether it all consists of one proof.
One kind of thing Nozick’s various distinctions do exclude: a philosophical work that tries to deduce all its conclusions from a few axioms. That is excluded, but then virtually no philosophy has been like that. Spinoza claims to have done that in his Ethics, but no one believes him. At the very beginning of Western philosophy, Parmenides’ poem (or half of it) may have tried to do that, but it is hard to tell from its ruins—except that they seem more like the ruins of a temple than of a tower.
In fact, almost all past philosophy has consisted of explanations, and has been known by its authors to do so. Plato’s did, and the word aitia, the Greek for “explanation,” is closely associated with its most notorious speculative construct, the Theory of Forms. Aristotle was obsessionally explicit in offering explanations. Even the rationalist Descartes can be seen as, much of the time, explaining—and he was aware, in the case of physics (at least), of something to which Nozick draws attention, the power of explanations that one knows to be untrue. Kant was centrally engaged in explanation: explaining, as Nozick reminds us, how something can be so (for instance, that every event must have a cause), but also, and very illuminatingly, how we can seem to be forced to believe things that are absurd or impossible. Kant and other philosophers, such as Wittgenstein, have believed something that Nozick mentions but does not in fact attend to very much in this book: that a good philosophical explanation, insofar as it corrects everyday belief, carries with it an explanation also of error, and of why incorrect everyday belief should seem true.
There are perhaps two distinctive things that Nozick particularly has in mind in emphasizing explanation as against “coercive” proof. He is telling us, for one thing, that philosophers should entertain ideas in a patient and imaginative manner, and not treat them, or those who offer them, as the local heavies in a Western movie treat the stranger in town. It is not the most creative approach in philosophy to shoot an idea out of someone’s hand as soon as he picks it up. Here Nozick’s conclusion is fine and good. Not everyone needs to be told it.
The second thing that Nozick encourages goes further and is less good. This—which emerges much more in the later part of the book, about value—is the idea that a philosophical conception recommends itself if, as well as fitting a number of our spontaneous practices of thought and feeling, it would be nice if it were true. At times, Nozick starts at quite a distance from the facts and constructs a model of how things might be, the main point of which seems to be that the world would be a better and less squalid place if that were how things turned out to be. Thus he offers a picture in which we have a dignity that lies in possessing genuine free will; in which there is an objective world of values that are not created, though they are in some way brought to life, by our thought; in which these cosmic values are what give meaning to life.
These are not contemptible ideas, and Nozick does not merely assert them. He puts them to some explanatory work—though that work is rather narrowly conceived—and he argues about them in some detail. But what gives the ideas their initial impetus does not lie in any explanatory power—at least any beyond their effortless ability to match the conviction we may have, some of us more than others, that we have the dignity of undetermined free will and that there are objective values. What launches them is the feeling that everything would be higher and more inspiring if things conformed to Nozick’s model, and it is that feeling, together with a certain tone (of which more later), that grounds the fear that this book is trying to do a dreadful thing: to lead philosophy back to an aspiration from which the work of this century has done so much to release it, the aspiration to be edifying.
There are some important kinds of philosophical explanation that do not figure enough in these pages. One is that Kantian type of explanation I have already mentioned, the explanation of why what seems to us to be so should seem to be so if in fact it is false. There is a good example of this lack in Nozick’s first section, on the identity of the self. Philosophers have much discussed issues of what differentiates one person from another and what it is for a particular person to continue to exist: is it continuity of body, of character, of memories, or what? An important slant to these questions is given when the perspective is the person’s own. Suppose I am invited to reflect what person in the future would, under various imaginable transformations—e.g., if my memories or character or body should be radically altered—be me. Some philosophers (including myself) have claimed that the following principle applies to such cases: that the question whether a certain person in the future, let us call him Y, is the same as a person present now, X, cannot be a question of whether Y is merely the best candidate available at the time for being X.
Imagine two different states of affairs five years from now. In one of them, a certain living body, Y (not your present one), has been programmed to have your present memories and character (more or less), and your own present body has been destroyed. In the other state of affairs, Y exists, just the same, but your present body, X, also continues, with the same memories and character. Clearly in the second case, X is you, and Y is not. Y is just a copy. Does that mean that Y cannot be you in the first case either? I and others have argued that is does mean that. Nozick denies this, and claims that who one will be in the future depends on the available candidates. In the first case, you will be Y, since there is no better candidate; in the second case, you will not be Y, since there is a better candidate. He calls this the “closest continuer” theory of (personal) identity.
Nozick actually claims to show that the principle that he rejects is false, by producing a counterexample. This strays a bit, early in the book, from the protocol of eschewing proof, but in the event he does not stray far, since the counter-example is feeble and proves nothing. (It concerns a group of émigré philosophers who will be the Vienna Circle if there is no other and larger group of émigré philosophers elsewhere, but will not be if there is. But a group of persons is not the same sort of thing as a person. “Are you the Vienna Circle?” the welcoming committee asks. “At least some of it,” the émigrés might reasonably reply.)
More interestingly, Nozick proceeds to find other kinds of philosophical issues which, he claims, have a similar structure, for example, the analysis of justice, or the meaning of sentences. This is worthwhile, and sheds light on a number of questions. But even if this is so, it will not answer the question that many will feel most needs answering: how can this be the right answer for personal identity? How can it be that the question whether I should fear for what happens to some future person (supposing that I care about myself) should depend on whether someone else has been killed off who would have had a better claim to being me if he had survived? Perhaps the question can be answered, but it is not answered here, and it certainly needs to be. If what Nozick claims is true, then some things that we very deeply believe about personal identity are false. We believe, for example, that our claims to personal identity do not depend on who else is around. And what needs philosophical explanation is why we believe them. That explanation would also help us to think straight about our future without. believing those things.
One of Nozick’s objectives is to respect and make sense of deep metaphysical and, in some cases, religious motivations that express themselves not only in some unschooled reflection but also in a variety of religious and philosophical traditions (he refers quite often to Indian thought). His way of confronting these traditional questionings, however, is often somewhat oblique. He takes up the question “Why is there something rather than nothing at all?” This has certainly in many traditions been felt to be a real and puzzling question: to such an extent that well-known attempts in philosophy to show that the question is misconceived tend to seem shallow and unconvincing. Nozick turns the issue into a subtle discussion of the theory of explanation. If it is claimed that a principle is explanatory of everything, can it also explain itself?
This is, once more, an interesting discussion, but the reader may feel that he or she has been left alone to work out the relevance of Nozick’s immense ingenuity and logical inventiveness to the archetypal metaphysical concern which started the discussion off. (The reader, incidentally, will need not only close attention, but also some technical knowledge, to follow Nozick down some of his paths in this book—what a Gödel numbering is, for example.)
By far the most successful section of the book is that on knowledge. This is very vigorous and resourceful, wonderfully clever, and, at a rather technical level, highly instructive. It is also academically the most conventional part of the book and very directly rooted in the recent literature. The argument is at the same time very original in its application. The basic idea is that someone knows something if he “tracks” the truth, where this means, very roughly, (1) that he believes something true, (2) that he wouldn’t believe it if it were not true, and (3) that if it were true, he would believe it—where the last clause adds the idea that not only is it true and he believes it, but his belief is “sensitive” to the truth. By “sensitive,” Nozick refers not to a special capacity of the believer but to the existence of a factual link between a belief and that which the belief is about. This is all worked out in considerable detail and with a subtlety that does not, as often happens in discussions of this subject, lose touch with the point of our having such a concept of knowledge.
The most spectacular result is achieved when this definition is applied to dealings with the skeptic. Nozick writes, “We do not seek to convince the skeptic but rather to formulate hypotheses about knowledge and our connection to facts that show how knowledge can exist even given the skeptic’s possibilities.” The skeptic says, for instance, that we do not know that we’re not in a tank having our brains stimulated by a mad scientist in such a way that it seems to us just as though we were in a room, seeing a table, and so on. So we don’t know that we are in fact seeing a table, even though it seems to be there, in perfect view, and everything seems as it usually does. Nozick says, and derives it from his definition, that we don’t know that we’re not in the tank; we do know that if we’re in the tank, then we are not seeing the table; yet we do know that we’re seeing the table. (Knowledge, in the technical phrase, is “not transitive under known logical implication.”)
This is the one argument that practically no one has used against the skeptic, and it testifies to Nozick’s remarkable discussion that this conclusion, which might seem a desperate device, begins to look natural and plausible. If one wants to know how to deal with skeptics, or indeed what knowledge is, this discussion must be read.
I am less certain that Nozick’s dealings with the skeptic get their strength, as he suggests, from his general objective of substituting explanation for proof. His point, as I’ve said, is that he is not trying to refute the skeptic, that notoriously discouraging task. But even if he is not trying to show the skeptic that the skeptic should not believe (or rather disbelieve) everything that he does, he is trying to show us that we need not accept what the skeptic says. If the skeptic is right, however, we should accept what he says; so the skeptic can’t be right. It is not obvious that those who have attempted to refute the skeptic (notably Descartes) have wanted more than that. But this is not a simple question, and this section is not only one in which the argument comes vividly to life, but also one in which the distinction between explanation and proof may for once be doing some helpful work in its own right.
When this high point of the book ends, we are less than halfway through. All the rest belongs to a division called “Value,” in which such things are discussed as free will, punishment, moral motivation, self-improvement, “the value of valuers,” and the meaning of life.
The section on free will is at several places uncharacteristically hesitant, not just in conclusion, but in direction. This is particularly so in the passage on retributive punishment, a subject on which, it seems to me, if Nozick had applied the same trenchancy as he has done to banging more fashionable views, he might well have given up. Here he attempts another application of the idea of explanation. The aim is not to justify retributive punishment (or to give a retributive justification of punishment), but to give an account of what its underlying ideas are. Nozick starts with some distinctions between revenge and retribution, and gives an elegant account of what an act of retribution is, revealing it as a form of communicative behavior which does not merely hurt but shows someone something. He then considers the simplest view of punishment as “showing” someone—a teleological version of it, in which the aim is to produce a change in the offender by which he comes to recognize his guilt.
There are well-known problems with theories of this kind, which tend to go around in a circle. The painful element, essential to punishment, in such an instructional act seems gratuitous. If the point of hurting a man is to tell him, why not just tell him? Or, if the aim is not just to tell him, but to convince him, the rationale may well become the same as that of reformative punishment. Anyway the rationale will not apply to the unreformable—apart from the point that it will always be an open question whether he might not be more effectively convinced by other means.
So Nozick moves to a wholly nonteleological interpretation, as he hopes, of retributive punishment. The aim of retribution is not merely to produce certain consequences but to do something “right or good in itself.” The aim, as he puts it, is to “connect the offender with correct values,” where this is not an aim that is added on to the act of punishment, but inherent in it. Nozick’s idea is to produce an effect on the offender, not necessarily in him: correct values are given “some significant effect in his life.” But, once again, why painful punishment rather than something else? This, replies Nozick, is necessary for it to be a significant connection with values. Why then should only those who have actually done wrong be so connected with values? Because, Nozick would answer, the wrong they have committed shows them to be not just unlinked, but “anti-linked,” to values. Is this, however, not also true of the dispositionally wicked? But, Nozick claims, if they have not committed acts they have not flouted values, and the aim of retributive punishment is to “replace flouting with linkage.”
The effect of Nozick’s criticism of the old arguments against retribution is simply to have reduced to the smallest possible circumference the circle in which the argument moves. He is not trying to produce a justification of retribution, but he does aim to produce its rationale, and the rationale provided moves round in so tight a circle that it is hard to see how the justification of retribution might eventually be applied. The “role of suffering in punishment,” Nozick says, is “to negate or lessen flouting by making it impossible to remain as pleased with one’s previous anti-linkage.” That version, in fact, is not yet entirely purged of the teleological—for suppose he remains pleased with his previous anti-linkage? But even with that impurity, the account is so utterly wrapped in itself that it is not going to give much insight to anyone who was previously puzzled by the idea of retributive punishment.
Another feature of the discussion of punishment is shared by all the material on value: it is vastly removed from any actual social institution. There are one or two desultory references to the law, but virtually nothing that focuses on the fact that punishment is inflicted by some actual authority in some actual social circumstances. This characteristic, spread over the whole large discussion of value, issues in a level of abstraction that is often bewildering, and which utterly discounts many obligations which one might have thought to be precisely obligations of philosophical explanation.
Value, we are told (I suppose one should rather say: “it is suggested”), consists basically in the degree of “organic unity.” We value, he implies, that which unifies diverse and apparently separate materials or experiences. Not much is done to give content to this old and uninformative proposal. For Nozick, as for earlier writers, it is based on the differing complexities of higher and lower organisms and on some selected considerations about aesthetic objects—despite well-known and indeed obvious difficulties about measuring “organic unity” anywhere, and, in the case of art, when we can supposedly detect it, doubts about necessarily preferring its increase. But such problems do not seem to concern Nozick very much. He is more interested in a set of logical devices dealing with the notion of value in the abstract. He discusses in detail such topics as what would be involved in “tracking bestness,” as he puts it, and “why is being a value-seeker and responder to value qua value itself valuable?”
Very little of this offers a direction to a psychology of moral action. Nor does it stoop to recognize any of the wellknown problems raised by social explanations of actual systems of moral value, for example the conflicts between different systems of value in different societies and the difficulty of achieving objectivity in describing or explaining them. Indeed Nozick barely pays much attention to any distinctively human category. The picture we are left with is of the objective cosmic dimension of value as organic unity, with our having the opportunity to link ourselves with this, and by that be transformed and elevated. These salutations to independent values, alternating as they do with very detailed formalistic arguments (on the conflict of obligations, for instance), give the impression of some beautiful nebula, at one time seen through the optical telescope, at another with its light broken down through the spectroscope. There is a difference, however, since a nebula sends light to us, but values are inactive, unless we make them active in and through us. As Nozick puts it, in a very typical passage:
Value seekers and responders have a cosmic role: to aid in the realization of value, in the infusion of value into the material and human realm.
Can that possibly, one begins to wonder, be the way to put it—to put anything?
If we do take up our cosmic role, then we may, according to Nozick, be transformed. One of the most important features of his account, and one that he seems least prone to regard as peculiar, is his emphasis on the notion of spiritual superiority, and indeed on a kind of ressentiment, a destructive and belittling envy that can be aroused by the spiritual superiority of people who, as Nozick sees them, have excelled in this cosmic role. Many of his passages that are most genuinely passionate are about this subject. They have the same note, and the same animus, as was to be found in Anarchy, State, and Utopia on the subject of envy directed toward more mundane forms of superiority. The tone of the present account introduces more than a hint of a competition in spirituality.
There is something very wrong with the way in which these notions are deployed in Nozick’s writing: that is to say, there is something wrong with the writing, and also something wrong with them. The theory itself does not give much content to these ideas of spiritual self-improvement, and the level of generality prevents his giving us any side illumination from the merely historical or psychologically particular. He tries to give us the idea directly, so to speak, and the fact is that we cannot trust him as a messenger from any spiritual height, because by the time that he is fully launched on his cosmic role he not only cannot hear anything that is happening on Earth, but cannot properly listen to himself.
It is not merely a matter of the occasional sententious banality: “It is better and lovelier to be moral”; “imagination has always been a faculty prized by students of literature and art”; “in our own biographies, at least, each of us is the leading character.” Such phrases are alarming enough—as Anthony Burgess’s narrator says in Earthly Powers, “If I could write so blatant a tautology, I could write also of the goodness of evil or the badness of good, and probably, somewhere or other, did.” But it is more generally true that the nearer we get to Nozick’s image of what he calls “spiritually advanced persons”—the arousers of ressentiment in those unwilling to be, as he puts it, “helped along” in this direction—the more his capacity to listen to himself seems to lapse:
There are some individuals whose lives are infused by values, who pursue values with single-minded purity and intensity, who embody values to the greatest extent. These individuals glow with a special radiance. Epochal religious figures often have this quality. To be in their presence (or even to hear about them) is to be uplifted and drawn (at least temporarily) to pursue the best in oneself. There are less epochal figures as well, glowing with a special moral and value loveliness, whose presence uplifts us, whose example lures and inspires us.
There is a lot more of this. There is even more about people glowing. Such writing sounds like Close Encounters of Some Yet Higher Kind, or a commercial for breakfast food. Whether its defects derive from the state of the modern world or from the English language now or (as I believe) from confusion about values, it will not do. And Nozick, even if he has switched off the monitor to Earth, will know, since he is a very good philosopher, that the problem cannot just be a matter of finding some happier way of saying the same thing.
In fact I think that the part of this book that is about value is utterly misconceived, and that the whole enterprise of approaching the problems in this abstract way, virtually unrelated to human psychology or society, and assuming ill-defined and suspect notions of spiritual superiority, is a large error. I do not believe that there is any cosmic role, and I think that Nietzsche was right when he said that philosophy should stick close to the Earth. But even if I am wrong, and there is something to be recovered from such conceptions, it is a basic truth, which Nozick seems not to have encountered at all, that it could not be recovered in these terms, or by such ingenuous methods. It is not a new truth. The wise men to whom Nozick sometimes refers, rabbis, gurus, Zen masters, have known and shown that the cosmic role, if there is one, needs a high measure of irony, personal and indeed cosmic, in order to be presented or interpreted. A craftier route to the beyond will always be needed, if there is a beyond, and the mere combination of cleverness with earnestness is not going to find it.
February 18, 1982