It is a scientific fact that a sudden magnetic field is demonstrably shown to coincide with a change in everyone's moral beliefs from time to time. People who have a strong conviction that abortion is morally prohibited before the magnetic field occurs have an equally strong conviction that abortion is morally permissible immediately after.

There is no such scientific fact, of course. But if it did exist, says Ronald Dworkin in his latest work, Justice for Hedgehogs, it would make no difference at all to the way we justify our change in moral belief. We would not say: “I think now, because of the magnetic field, that abortion is permissible.” Rather, we would give the moral reasons that we now think support our changed convictions, such as a perceived superior right of the mother or the lack of sentience of the fetus. Says Dworkin: “The only reason you could have to think that truth has caused your moral opinion is an independent belief that your conviction is true” (p. 78).

Dworkin also asks us to imagine the following: A gifted neuroscientist is able to make such accurate predictions of what an artist will paint that he can map exactly the sequence of brushstrokes, the precise use of color, and so on. Indeed, with the help of a powerful computer and robot, he produces a painting that is identical to the masterpiece that the oblivious brilliant artist shortly afterward finally finishes. The question is whether the value of that work of art diminishes in any way when we learn beforehand exactly what the artist was going to do. The artist acts against the background of his own conscious understanding of what he is doing; and that conscious understanding is sufficient in itself to resist any account that his work was of no value because his work was not “really” creative. Although we would applaud the scientific feat, Dworkin says it would have no relevance to the question of the value of the artist's work. The artistic endeavor is measured against what is important: that the artist strove to produce his painting in a remarkable combination of talent and conviction.

In this rich, complex, and closely argued book, Dworkin boldly affirms the independence of arguments of value, arguments that remain securely within their own domain. Mostly, but not at all exclusively, he is concerned with moral value. He claims it is wrong to assume that external forces could force conflict between moral values on us, as Isaiah Berlin and others have urged. Rather, he says we should be more confident in justifying our judgments of value by reference to the more abstract values we hold; we also have a personal responsibility for making our judgments coherent. “Value judgments are true, when they are true . . . in view of the substantive case that can be made for them” (p. 11).

These ideas together form what he calls the “unity of value” thesis, the “big thing” that the hedgehog knows in Archilochus's comparison of the fox and the hedgehog, famously used by Isaiah Berlin. Take the well-known supposed conflict between freedom and equality in the distribution of resources. If we think that people are of equal value as human beings but also that people should be free to keep what they have worked for, we must try to see in what ways equality and freedom need to be qualified to respect both values. We cannot simply discard either value if it is inconvenient, say, by denying people equality of bargaining power but affirming maximum freedom for the market: “You cannot determine what liberty requires without also deciding what distribution of property and opportunity shows equal concern for all” (p. 4).

Justice for Hedgehogs is a masterwork that offers many clear and confident solutions to major problems of morality. The first part of the book is a ringing endorsement of Hume's distinction between fact and value, and it claims for value both independence and objectivity. The second half takes up the challenge of justifying our moral judgments and is broadly an endorsement of Kant on morality, with a plausible addition (or extrapolation): that “living well” might responsibly include an integrated balance—even a trade-off—between personal ethics and morality. As usual, Dworkin engages with the best contemporary thinkers who address these sorts of questions and, hence, it is a book that tells us, too, what they have to say. And it is monumentally conceived, for Dworkin asks us to incorporate other large books of his within it, notably on legal philosophy (Law's Empire and Justice in Robes) and political philosophy (particularly Sovereign Virtue).

Because the range and ambition of this book are enormous, I can only concentrate on two of the more important themes to give prominence to the detail and force of Dworkin's arguments. They are, first, his defense of the unity of value thesis, from which all else follows; and, second, his striking argument that morality derives from personal ethics. I conclude, further, with an observation on legal argument.

THE INDEPENDENCE AND OBJECTIVITY OF VALUE

Dworkin concedes that if, as the principle of independence requires, value can only derive from value, the account is circular. But he says it is a nonproblematic circularity consisting only of a restatement that value exists in an independent domain from science. For it is also true that science derives from science, and, indeed, this is thought to be a virtue of science. The real problem is that our ways of arguing matters of value are much less developed; in argument about matters of value we therefore do not yet possess the confidence we have when we contest matters of science. No wonder, then, that science, with its astounding success in acquiring knowledge, has dominated thinking about truth for so long. Moral philosophers who suppose truth is science-based, says Dworkin, have all too often wrongly urged “a colonial philosophy setting up embassies and garrisons of science within value discourse to govern it properly” (p. 9).

But if judgments of value could not be true or false, there could be no unity of value. Moral beliefs would be merely matters of taste, not judgment. Dworkin points out, however, that we talk and act as though there were moral truth. Ordinarily, we regard torturing babies to be wrong independently of the views of others (we would not think twice about forcing our judgment on the torturer), and that its wrongness can be supported by reasons independent of our personal feelings. On the other hand, there appears to be no evidence for the truth of such moral judgments. Historical practice, feelings, and what other people believe are often irrelevant and uncertain guides. It is not surprising, then, that there are skeptics about the existence of moral truth.

Dworkin nevertheless claims that there is no threatening skeptical position against the objectivity of value. Either skepticism is just another way of making a straightforward moral judgment or it is contradictory. There is, however, a genuinely skeptical position that is not threatening because it derives from value; for that reason Dworkin calls it “internal” skepticism. An example is the view that morality is not universally applicable because of cultural difference. That states a moral view—one he thinks is wrong, however—because it claims that a statement of morality is false if it claims universal applicability. He emphasizes that he has “no quarrel” with any form of internal moral skepticism; it expresses a genuine hypothetical moral view.

Dworkin, however, concentrates his energy on the “external” skeptics. They believe either that there is no moral objectivity because it is an error to suppose that there are “things in the world” (Dworkin calls such things “morons”) according to which moral judgments could be objective, or because ordinary moral judgments can be recast in a way that shows their true status to be incapable of objectivity. Both the “error” and “status” skeptics are external skeptics because they each believe themselves to be looking at morality “from the outside” in a detached and disengaged manner. Each attempts to disparage morality without relying on it in the same way as internal skepticism does. Here matters become more complicated.

The first position—error skepticism—is, according to Dworkin, self-defeating because it relies on the claims that (1) moral judgments can be true only if they are proved by empirically determinable facts, and (2) there are no such facts. If so, the position is merely an internally skeptical one and is not skeptical in the disengaged way it was intended to be. Take the issue of abortion. The external error skeptic says that abortion is neither morally right nor morally wrong. True, there is nothing in the world (no morons) that, scientifically, could make it so. But, says Dworkin, if there is nothing to make abortion right or wrong, no other conclusion is possible but that it is morally permissible; so the external error skeptic contradicts his own claim that first-order moral judgments are not objective.

Status skepticism, however, is the dominant form of skepticism among philosophers. It relies on a distinction between first-order and second-order claims of morality. First-order claims are ordinary moral judgments; second-order claims are claims about such judgments (it is just the same as the difference between doing arithmetic, as in “2 + 2 = 4,” and talking about arithmetic, as in “arithmetic is taught in schools”). According to the status form of external skepticism, first-order claims of morality cannot be true or false because—and this is the second-order claim—they are not descriptive of anything but are only “expressions” (of feelings or emotions, say).

Thus, the difference between error and status skepticism is that the error version says ordinary moral claims are “misconceived” (there are no morons out there) and the status version says such claims are “misunderstood” (they are more accurately expressed in other ways). Dworkin argues that status skepticism is the most popular form because it is thought to allow us—unlike error skepticism—to keep our own convictions, such as that abortion is wrong.

Although popular among philosophers, status skepticism is also self-defeating, he says, because nothing by way of meaning is added by saying “abortion is objectively wrong” or “that abortion is wrong is true” or “it is an external fact that abortion is wrong,” and so on. These expressions, rather, are about emphasis. Thus, since the status skeptic denies them, just like the error skeptic, these further claims are first-order moral claims. This argument, like that against the error skeptics, is breathtakingly simple. Dworkin employs no special logic or sleight of hand to establish it. There has to be, in other words, an extra argument showing why moral claims are different from philosophical claims. If the status skeptic claims that these moral expressions are part of our semantics, our semantic practices disprove it because people who say that torturing babies is wrong mean it as something different from a mere expression of their attitudes. They certainly do not mean it is true for them but not true for others.

Nor are our beliefs merely desires. Some philosophers argue that since beliefs do not motivate and only desires do (my knowledge that aspirin will cure my headache does not motivate me to take it, only my desire does this), a first-order moral judgment is only about a person's desire, and so cannot be true or false. In response to this, Dworkin simply denies that our moral beliefs are “really” desires. They do not always motivate, he says, pointing to Richard III's desire to do what he knew to be wrong: “I am determined to prove a villain.” In any case, Dworkin says we quite naturally discover a person's beliefs from their desires.

Yet another form of status skepticism asserts that second-order statements are not rewrites of first-order statements but exist in a completely different philosophical discourse. Dworkin takes Richard Rorty as an exemplar of this sort of argument. It works quite well for explaining fiction, too. For example, Lady Macbeth makes true statements from within the play; but in another “discourse,” because she is fictional, her statements cannot be objectively true. Rorty applies this argument more generally. He says, for example, that we can play a mountains language game, which is independent of whether mountains actually exist, and so we can do the same with a morality language game. The philosophical language game is independent of whether morality is objective.

Another such line is taken by the so-called “projectivists,” who appear to say that we can be committed to morality and act according to its dictates, but that these are only commitments that are “projected” onto the world, and thus without attendant objectivity. In these cases, Dworkin says the language-game strategy fails because it does not supply an argument showing why the alternative language is not a restatement of the original first-order claim. Unlike in fictional literature, we cannot show that there are different language games whose difference prevents us from asserting that moral judgment is true beyond its own respective game. By contrast, we know very well why Macbeth is fiction.

It is useful to be reminded that Dworkin's life work has been to convince us of the importance (Taking Rights Seriously and A Matter of Principle) and the pervasiveness and coherence (Law's Empire and Sovereign Virtue) of value. To him, it is not just that we cannot rewrite our moral convictions into expressions, desires, or appearance, but that it is morally ridiculous to think we could: it is not something that we should want to do. The skeptical attempts fail to secure the significance and seriousness of moral conviction; instead, they belittle it. As he says, if there are no objective values, there would be nothing bizarre about my believing abortion wrong but encouraging a friend to have one.

Dworkin convincingly undermines the threatening forms of skepticism, and he is right to characterize them as assuming that value objectivity, to be successful, should have science at its root. This is not to say that scientific-causal type arguments have nothing to do with value judgments. If a person has the frontal lobe of his brain missing we know that he is likely to be incapacitated. But we should not say a person is excused from murder because his frontal lobe was missing when he killed; that account would not fully state his excuse. Rather, it is that he is excused if the moral requirement that he have a particular capacity is not there (say, that he lacked the capacity to know what he was doing or if, knowing that, he could not judge whether it was wrong).

There is a subtle distinction Dworkin draws here; it is between a moral principle of responsibility that refers all questions of responsibility to whether a person could “causally control” his action, and one in which the determining factor is that he had the “moral capacity” to control his action. To suppose that physical causes explain our responsibility does not integrate with the rest of our beliefs about responsibility; instead, it appears to rest a lack of responsibility on “a haphazard piece of quantum whimsy.”

In many cases, we are subject to moral censure even though facts determine the outcome of what we do. An obvious example is an unsuccessful attempt to do something that was in fact impossible to achieve (there was no money in the pocket we attempted to pick, say). We have no problem with ascribing blame in such circumstances; but the causal control principle would suggest we would be blameless. Moreover, we do not have control over our beliefs because they are controlled by how the world in fact is (and it would be impossibly odd to suppose that we are free because we can make up our beliefs about the world). And some of us cannot act otherwise than the psychological facts of our personalities allow (for example, Mother Teresa could not help being selfless, and Hitler had a disposition to evil), but such facts do not withhold us from judging them. Nor is it the case that we assume lack of responsibility when it comes to people who are mentally incapacitated or are simply immature (as with young children). The mentally incapacitated and children are often just as capable as others of initiating action, and the problem is instead that of identifying the extent to which incapacity has distorted their judgment.

The capacity principle allows us to make these fine-tuning adjustments to responsibility in contrast to blanket responses imposed by the causal principle. Since the matter is an interpretive one—it is a question of value—in choosing between the two accounts of responsibility, we should choose the one that makes the most sense. In short, the causal control principle makes little sense of how we actually think.

What Dworkin has achieved is of considerable significance. He has reinterpreted Hume not as moral skeptic but quite the opposite, as someone who supported the independence of value from science. The Humean principle, as Dworkin calls it, tells us that we can have confidence in looking to value to justify our moral judgments and therefore not be skeptical of morality at all. The independence of value is important for his book: it requires that values are “interconnected and mutually supporting.” And what Dworkin has also done is demolish the idea of second-order theorizing about morality. That domain is captured by morality and so Dworkin is free to develop a full account of morality based entirely on evaluative judgments. This cuts a lot of deadwood from the so-called metaphysics of morality or moralizing.

For too long, debate about the credibility of morality itself has existed under the shadow of “reality” determined by science, and this has been to the disadvantage of morality. The realists among moral philosophers try to show the interaction between morons and ourselves, and fail; while the anti-realists, who believe in mind-dependent morality, exist under the same shadow, for they have to say, bearing the realist camp in mind, that morality is somehow “made up,” coming entirely from within. This is, as Dworkin says, “an entirely bizarre assignment. How can they be values if we can just make them up?” (p. 9). Morality cannot be just a matter of taste, mattering only to the person whose taste it is.

THE DERIVATION OF MORALITY FROM PERSONAL ETHICS

Dworkin is strikingly original in his attitude to moral questions as he approaches these from the point of view of ethics. Ethics, he says, concerns those principles we should follow in order to make our lives “valuable.” Among these he unsurprisingly includes Aristotelian virtues. We must, he says, live lives of dignity, meaning not only that we must take our own life seriously (in other words, accept that it is objectively important how we live), but we must take responsibility for it as well. Living a life of dignity means that we can have self-respect. Moral questions, on the other hand, concern our duties to one another.

The originality of Dworkin's approach is that he takes moral questions to be extensions of his account of ethics. In recognizing the importance of self-respect in our own lives, we must as a matter of logical consistency recognize the importance of self-respect in the lives of others. If we have a duty to live valuable lives, then it is important to live well; thus, our personal lives need not be dictated solely by the requirements of morality. Although Dworkin's move from ethics to morality embodies something of Kant's requirement that we act in accordance with rules that would apply to others as well, the idea that our ethical lives are lives of dignity merely because we have lived in accordance with the requirements of the universal law is something with which Dworkin markedly disagrees. Here Dworkin supplies the warmth that is at least not clearly discernible in Kant: what is the life worth living? It cannot consist exclusively in acting according to moral law. That seems to be too austere (even boring) and at odds with what we ordinarily suppose is the “good life.” Dworkin finds it very implausible that our ethical life is fulfilled in this way. To make our lives valuable may require, if not selfishness, at least some attention to ourselves that can be traded against our moral duties.

All this looks at first sight to be an invitation to immorality, but it is not. Morality at root concerns dignity, and dignity is not defined in terms of morality but in terms of ethics. What is particularly wrong with utilitarianism is its inability to distinguish the self from others. There is no point in utilitarian calculation if it excludes, by definition, the extra importance that each life has to the particular person whose life it is; that is the gravamen of the charge that utilitarianism does not take the distinction between people seriously. In the move from personal dignity to morality, Dworkin considers the conflict we may have between our own and another person's interests. We can regard other people's lives as equally valuable to our own but not suppose that we have a duty to “make others equal,” whatever that would mean without our having any sort of “metric” (p. 275). For example, it does not appear to follow that if we win the lottery we thereby have a duty to give the money away to good causes. We need to draw lines, and Dworkin suggests that these will arise from an examination of the likelihood of the deprivation of “the ordinary opportunities people have to pursue whatever ambitions they choose,” rather than a subjective assessment of what people would like, the relative risk to oneself in averting harm to a stranger, and the fact of “confrontation” (we should pay a boatman to save the drowning man but not give the boatman money in advance in order to save men drowning on subsequent occasions; p. 278).

In other cases, the principle that our lives are of equal objective importance will determine the morality of helping others in need; if I can rescue only one of two people drowning, that principle will allow me to rescue my wife (of course), or choose the better musician, but not the white rather than the black person merely because he is white (for that would affront the equal worth of lives). On the other hand, where I can save either a large number or a small number even though I could give each an equal say in deciding, I could bypass that and just save the larger number: “The principle that it is better to save more rather than fewer human lives, without regard to whose lives they are, is a plausible even if not inevitable understanding of what the right respect for life's importance requires” (p. 283).

LEGAL ARGUMENT

Justice for Hedgehogs shows why Dworkin was naturally drawn to law (in a published conversation in the 1970s with Brian Magee he said that he was primarily a philosopher before he discovered what a wonderful subject law was), and why his early work is almost entirely concerned with the detail of criticism of American constitutional cases, mostly published in the New York Review of Books, and the development of his anti-positivist legal theory. It explains exactly why Dworkin preferred the characterization of law as an argumentative “attitude” rather than a “model of rules,” for it is in legal argument that the most consistent, coherent, and advanced systematization of real moral argument takes place. Even now most moral philosophers know nothing about law. Yet almost any hypothetical example that a moral philosopher could think up will have occurred in some form in real time at a real place, and very careful (first-order) thought will have been put into resolving the problem. What is more, the form of moral knowledge the law embodies aims at overall integrity through the operation of statutes, and the use of law reports to enable judicial hearings to be informed not only of how previous careful decisions have been made but to ensure fair uniformity across the community. It can be difficult to convince a moral philosopher of this fact; he might need to be a lawyer or read Law's Empire, perhaps. But acquaintance with some reported cases would probably be sufficient. He would discover that for every clear law that is made, some human being, somewhere in the world, will throw up a situation that is baffling. Judges and lawyers will consider the moral arguments for an ultimate proposition either way. What the legislature has decided, what previous judges have decided, will supply only part of the reasons each way. Real decisions will also throw up questions of principle that need to be compared with hypothetical situations (the judge's common question, common also in lawyers' offices, “What if . . . ?”). Dworkin once said he had at times contemplated a book called Philosophy's Tutor, which would concern this same moral point and also consider problems of truth and language that legal argument routinely generates.

In my view, Dworkin's approach to law determined the shape of the later development of his thought. He wants moral argument on the ground to be more sophisticated and confident, and so it is natural for him to look to legal argument—that is, moral argument within the sphere of political morality. Justice for Hedgehogs therefore provides a more abstract justification for what he has been doing for years, beginning with his famous attack on the theory of legal positivism of H. L. A. Hart. At first his criticism was that the Hartian model did not allow moral argument to play any part in the determination of propositions of law. In Dworkin's view, Hart ignored the role of principles—that is, value judgments—in law because of the requirement that laws be recognized as valid according to an empirical/factual test of concurrence with official practice. Later Dworkin developed the idea of “interpretive” concepts, the essential point of which is that they concern value concepts. Using his well-known metaphor to describe interpretation, he says we should cast our legal and moral practices in their best moral light. If Dworkin is right it helps enormously in understanding the justification that Hart offered for his theory of legal positivism: that distinguishing law from morality, as Hart proposed, would have great moral benefits in enabling us to confront the official abuse of authority—we would not be fooled by natural law, say, into supposing that merely because the authority declared something, it was thereby morally required or permitted. Legal positivism would allow us a safe conceptual distance from authority. In Dworkin's terms, therefore, Hart provides a moral justification for seeing the interpretive concept of law in a particular way; the explanation for seeing the law separately from morality is itself a moral one.

This line of interpretation of Hart (which is at odds with Hart's own stated intention of his work as descriptive of legal practice) is wonderfully supported by chapter IX of The Concept of Law, where Hart specifically chooses his “conception” of law on the grounds that it will help subject official pronouncements to moral scrutiny (see p. 204). It also points to a more plausible connection between Jeremy Bentham and Hart, for Bentham always embedded his legal positivism within his general account of morality, the philosophy of utilitarianism. When Bentham, in his characteristically thorough fashion, was considering how he would proceed to shape his theory of law, he denied that the jurist should take the line of an historian describing current legal practices; instead, he said, the jurist should arrange the institutions and ideas about law according to the single principle of utility (see the introduction to his A Fragment of Government). In other words, he thought it was morality that should determine the proper understanding of law. In Dworkin's terms, Bentham unconsciously acknowledged that law was an interpretive concept, one that had to be understood evaluatively.