Dworkin's 'one-right-answer' thesis is crucial for understanding Dworkin's entire political and legal theory because of that theory's reliance on the status of value, particular moral value. Dworkin thinks that law is a branch of political morality, which is part of general morality (and which is, in turn, an extension of personal ethics). For him, legal argument is a special kind of moral argument -that concerned with morally justifying the use of institutionalised communal or state power against its citizens. I will mostly talk about moral value although I think that many of our aesthetic judgements are also moral (my judgement of the greatest music has a large moral component which I'm basically at a loss to describe- like many people) and our moral and legal judgements (on abortion and the aesthetic status of the foetus, on euthanasia and the aesthetic status of the 'living corpse'). I believe that although there are differences -for example, that it is more common to find evaluative judgements resulting in incommensurable ('no-right-answers') propositions in aesthetic judgements (eg. 'Beethoven was greater than da Vinci'), or that there are two or more ways equally unsurpassable ways of interpreting, say, the Bach Partitas- this is more the result of hunch.¹

The argument that law is a subspecies of morality is not, as Waldron surprisingly claimed recently, a significant extension of Dworkin's supposedly earlier position that moral propositions are part of legal claims.² The earlier and the later positions are identical, just that on the later position Dworkin is more explicit. If moral propositions are constitutive of part of law, then the universal nature of moral propositions and their independent authority must ground the whole of law because any conflict may only be resolved in favour of the moral position. Dworkin's argument against 'soft conventionalism' (now known as 'soft positivism') in Law's Empire surely hit that nail on the head.³

As moral value is the foundation of his theory (and for this read 'the most abstract' account or justification) of law, in a sense all the rest follows for Dworkin: the attack on descriptive, non-value, legal positivism and therefore his view that evil legal systems are relatively insignificant, followed by his particular take on what constitutes moral justification in law, namely, the centrality of moral democracy, the moral role of judges, the emphasis on non-utilitarian rights, the advocacy of judicial integrity and its relation to justice.

If there can't be right answers in matters of value judgement, merely because they are matters of value, such judgements must be seen as purely subjective matters of taste, or attitude, or emotion, or subjective feelings, or 'mere' belief or even whim. But to suppose this, as Dworkin says, is 'bizarre' for 'how can they be values if we can just make them up'.⁴ Morality cannot be just a matter of taste, mattering only to the person whose taste it is. A short way with this point is that value judgements are fallible and subject to reason.

If we think, instead, that there are right answers to questions concerning value, this must mean there is a possibility of error. It is not surprising therefore that people have thought Dworkin odd, even arrogant, in supposing that his elaborately structured account of law was more than just 'his view' for which there is no right or wrong, and for which no reasoned arguments could attract the appropriate objectivity. The charge of 'subjectivity' is false for, I think, obvious reasons. (First, though, I have to make couple of remarks on the sociology of thinking about value. i. The criticism often comes from people who adopt what they assume is a theoretical perspective and which flatly contradicts how they speak and act; this group includes both judges⁵ and philosophers (eg. Blackburn and Gray). ii. It also comes from an untheoretically inclined public imbued with the contemporary disposition of subjectivity ('that is just your opinion' or 'that is true for you') who in direct contradiction proclaims an objective view that people are 'entitled to their opinions'. Williams notably has pointed out the simple-mindedness of the latter view).⁶

Language, the way we think, our actions and projects -our 'forms of life'- and our concepts and ideas, all show how embedded, how necessary, value judgements are. The most common objection to the possibility of their truth, or objectivity, is the absence of an 'external' point of view from which their truth may be tested. The objection borrows from the science's luxurious model of empirical confirmation. But to deny the truth of moral propositions⁷ is to make a value judgement. If you deny that it is true that it is morally wrong ever to abort, you endorse the moral judgement that it is permissible. This argument shows that the logic of moral value requires a position or stance to be taken. In the real world -the real, real world of morality- you can't sit on a fence or throw up your hands. The same argument applies to aesthetic value. If it can't be true that St. Paul's in London is a great piece of architecture, then it follows that it is not a great piece of architecture. In other words, scepticism about the possibility of value judgement doesn't make sense.

The 'external' argument is common. Perhaps Bentham began to downplay the role of value by appearing to psychologise it in the form of pleasure and pain. I doubt it, however. He constantly proclaimed that the first principle of his philosophy was the 'Principle of Utility' and he organised his political and moral concepts about it -explicitly in his Fragment of Government. That people should be happy came first; what made them happy, the empirical question, came second. Or perhaps the materialising of value came from lesser economists, and later than Bentham. But the recent arguments between 'realists' and 'anti-realists', or the 'mind' dependent and independent moral philosophers, I think only make sense on the supposition that supposes 'realism' to be the superior -the foundational- ground for knowledge that is governed solely by the empirical. What has more reality, one might ask: the wrongness of child torture, or the shape of the planet?

The word 'fact', too, is in general use skewed in favour of empiricism. Legal argument uses a distinction between questions of evidence -'fact'- and questions of law which concerns those value- driven arguments concerning statutes or cases (the domain of the advocate's 'opinion' and the judge's 'judgement', note). However, contemporary philosophical analyses of fact and truth claim coherence in a relationship between truth and fact that has nothing to do with the significance of empirical truth: eg. Strawson: 'facts are what propositions state when true', and Tarski: 'p is true if and only if p'. Truth, in other words, is merely a relation between propositions and what makes them true, and not an account of what grounds truth in any domain. I believe that some moral judgements are more striking and important than a large proportion of empirical facts, and so I have little difficulty with claiming there are moral facts. I believe, for example, that it is a true proposition that torturing children for pleasure is morally wrong; I believe it is a fact. In any case, if value propositions can be true, the assertion that 'only facts can be true' can't distinguish fact from value.

Nevertheless, the Humean principle which demarcates empirical propositions from value propositions is difficult to deny. The famous debates in the sixties (see Searle and others⁸) tested, and favoured that principle. 'Promising' was initially thought to be definable in the form of descriptive premises ('I promise' as a speech act in appropriate circumstances) from which a value proposition expressing an obligation could be defined. But through argument it became clear that additional moral premises were required thought indefinable purely as conventional, as it required additional moral premises to define the 'appropriateness' of the speech act. The principle is also in widespread use, embedded in our language, practices and concepts. Its precise origins in Hume do not bear on the general question, however, for it is widely believed that he was sceptical of moral truth, and Dworkin is not of course a moral sceptic.

Dworkin endorses the Humean principle. That principle declares that empirical facts can't be sufficient ground of the truth of value propositions. But there is a relationship between empirical facts and value in that it is almost impossibly abstract to show what value is without referring in the end to the empirical facts of what actually happened, eg. a killing, or the existence of sculptures, mountain ranges and paintings, or communal/state practices of coercion. I think Cohen is right in his paper 'Facts and Principles' to say that no fact figures in the ultimate expression of any value proposition although in showing what is morally or aesthetically valuable it is necessary to refer to empirical facts.⁹ That, say, 'we should respect people' is too abstract a moral principle in most cases to be of practical use, but at a lower level of abstraction reference to judgements about particular acts as unjust, or about what is beautiful, will be more natural. However, no empirical fact even at the lowest level is sufficient to establish the moral principle. Of course, the interplay of empirical proposition and principle/hypothesis between the two domains is subtle. For value in general, the truth of empirical propositions, or empirically determinable practices, is part of the process of making value judgements. We can't make value judgements about the worth of a painting unless we see the painting, and for that the painting must be there, before us. But the value of the painting is not the same as the physical painting.

The above discussion of value objectivity is intended to explain both how people could be conceptually justified in accepting a category of value that is distinct from empirical facts and that the concepts of 'the external' and 'the truth' are not barriers to understanding value as objective. (Note that the arguments are conceptual because they attempt to provide a coherent account of the way we actually think and act -when we are thinking clearly- and which are expressible in the language we use).

Dworkin calls evaluative reasoning 'interpretive'. It is conceptual because it attempts to establish a coherent account of our concept use. But because the concepts concerned are evaluative, their use is unique. In some ways Dworkin might have been wiser had he preserved the term 'evaluative' since it draws attention to the significance of value in making interpretive judgements. But his reason is clear. Interpretation requires evaluation of common facts, those facts that constitute the practices of employing and engaging with the meaning of particular evaluative concepts. 'Justice' is an evaluative concept but it hardly helps that we know that we should be just if we don't know what the particular instances of justice are. Such instances are those that are recognisable within the practice - the facts- of justice. This tying to practice makes the evaluation an interpretation 'of something'. For Dworkin, all our ethical and moral concepts are like that. There is an attractive reality about his view, because it shows how difficult -requiring considerable creative effort- it is to extend our moral concepts. In one way it is conservative but, being tied to facts, it prevents, too, speculation about justice in strange utopias (eg. the world of angels). That feet on the ground approach is consistent throughout Dworkin's work. He often expressed general political and legal theory as that approaching philosophical accounts 'from the inside out', meaning that making judgements about existing practices, the more particular the better, is the place to start.

The entirety of legal reasoning is value-driven. Our sense of ourselves provides us with our first information about what is valuable in our lives. Or, our primary judgements are about ourselves (just an example: we can't understand what another's pain is unless we understand our own). Our recognition of others as equals to ourselves requires us to understand the value of the lives of others; thus Dworkin derives morality (comprising our duties to others) from personal ethics (and, I think, equality and freedom). Part of our duties to others comprises our communal/state duties so, in turn, that part of the morality of the state concerning the moral justification of the application of coercion is law. So morality stemming from personal ethics extends not just to the whole content of law, but to its structure as well.

When we make value judgements about what constitutes particular laws, we find moral coherence only by seeing the division of laws into legal systems and, within those systems, seeing differing kinds of laws such as tort and criminal law. Again, many people criticise Dworkin here for being 'parochial', by 'privileging' the Anglo- American systems, or the US system in particular. But this criticism is also unfounded because moral judgement is universal and can't be tied to particular systems. (This criticism is just another version of the argument against Dworkin from 'the actual existence' of evil legal systems).

How do we reason entirely within value, that is, without ultimate reference to empirical fact? That is the question Dworkin directly introduces us to (and is the 'research project' that Raz commends us to engage with¹⁰). He dignifies the presence of the evaluative concept in our discourse but, better, his argument asks for evaluative arguments that morally justify ordinary discourse. These, I believe, are what his theory of truth in value and his unity of value thesis are primarily about.

His theory of truth is that propositions are true if they best further the point of any practice. What is the point of law? Given the derivation of law ultimately from personal ethics, it is to justify morally a political structure of relationships between people. A quick way to this point is to ask what the opposite position would be. The point of law is to distribute injustice wherever possible? It answers itself. Law's moral point is morally to justify that general set of practices loosely connected by family resemblances -that practice 'we call' law and engage with in our use of 'the concept' of law- that invoke state coercion to redistribute wealth, to punish, to direct people for their own good, and so on. At a practical level, you might ask, when faced with only two possible ways to interpret a provision in a statute: What is the right interpretation? Is it the one that is more unjust than the other, or is it the one that is less unjust? As I said, it answers itself, and any practising lawyer would agree.¹¹ What is the point of morality? It is to establish and maintain the morally right relationships between people.

At the risk of being repetitive, the above is difficult to understand when you are convinced that the base for understanding truth and 'fact' is only empirical or analytic. It is a useful exercise to turn the debate in the right direction by trying to view truth in science from the standpoint of value.¹² Propositions of science are true only if they further the point of science? Is the point of science prediction? Leiter who claims a 'realist' position in denying objectivity to value judgements, justifies it by reference to what he obviously thinks is the point of science, which is 'to deliver the goods' such, eg. science 'sends planes in the sky' and 'has eradicated certain cancerous growths'.¹³ This is embarrassing. In disparaging truth in value, Leiter employs an evaluative judgement which he clearly thinks is true.

If truth in the domain of value can only come from that domain, what is the best account of truth for moral value? Consider first whether morality would be best served by all moral propositions being neither true nor false, without objectivity. That would reduce all moral propositions to matters of taste, or feeling. We could make them up! Does it make good moral sense to think of morality that way? It would mean we couldn't criticise our judgements about others, nor our judgements about ourselves, and we'd lose that check to our judgements that comes from constant re-assessment of our reasons. It would lessen the value of moral education. Our judgements about how to live our lives and how to treat others would rank the same as choosing ice-cream flavours.

Another account of truth we could consider is that based on the value of proof. Such demonstrable -or 'criterial'- type arguments seem inappropriate for value while common and appropriate for science and analytic truth. A paradigm view about scientific hypotheses is that they may be demonstrated to be false by the existence of contradictory empirical data. (The characteristic scientific hypothesis is 'if there are swans, they are white', which is 'falsified' by the existence of some black swans). But no empirical fact can show value judgements to be false because value propositions don't ultimately entail empirical propositions. They are of the form 'If foetuses are not sentient, abortion is permissible'. A judgement that it is morally permissible for a person to have an abortion according to such a principle is false if the foetus is sentient, but that doesn't render the moral principle false: only the empirical proposition used in applying the principle was false. In short, black swans make the scientific hypothesis false; the existence of sentience in any particular foetus does not make the moral principle false.

Would science's demonstrability, or provability, test for truth be valuable for the domain of value? There are two ways we can see how that proposal could work. i. It could have the consequence of reducing all moral judgements to subjective taste, in other words, the consequence that no value judgements can be made because none is demonstrably true. We shouldn't accept that. Thinking of morality as objective, that is allowing that our judgements can be wrong, and for reasons, is undeniably valuable. ii. It could also have the consequence of something worse, by creating a positivistic legislation of moral value, identifying what is morally right and permissible through demonstrable convention. We would retreat several centuries. It would mean the end of personal conscience and moral independence. I agree with Dworkin that it is of fundamental value to our human existence that our convictions are our own, not those implanted by someone else. So I would reject the implantation of science's conception of truth into the value domain for value reasons, namely those that the Humean principle informs us should come from that domain. In short, our personal moral life and our moral life within the community would become ossified if we used a concept of truth in value according to which moral truth required external demonstration. What would be real gains in relative certainty about what was required of us, and what we could reasonably expect from others, would be more than outweighed by a life of rigid conventionalism and conservatism. To sum up: no demonstrability morality would mean we could 'make values up'; demonstrability morality would mean living in the 12^th century.

Dworkin wants to raise the status of value arguments. He thinks that the development of methods of value argument is not as highly developed, or appreciated, as are scientific methods. He encourages us to argue matters of value, and to realise we are not doing something akin to science. This is the most abstract thrust of his argument against legal positivism. Doing law is not like 'painting with numbers' or applying algorithms or merely 'applying rules' with a rigid, pre-Wittgensteinian concept of 'rule' in mind. It is why he writes of law (particularly in Law's Empire) as defined, not as 'a set of rules' but as an 'argumentative attitude'. He bemoans our 'leaden age' in which methods of science have great status in popular imagination while the methods of value argument are unknown or disbelieved.

But people do engage with value. Art critics do with art. Practising lawyers, at least, do with law, and Dworkin's contribution in particular to case-by-case argument is striking amongst legal philosophers. (In which of the last ten monographs published by the major university presses of the UK and North American was more than a single case cited? The great jurists -Bentham, Austin, Kelsen, Hart and Dworkin- were all real lawyers). Dworkin began his academic life through direct engagement with particular arguments for particular cases in law, and then he began outwards for more abstract justifications, towards philosophy. It is a continuum of abstraction. Real cases pose philosophical questions and philosophy must engage directly with real cases. There is no 'special field' from which law, and morality, may be viewed. Why should there be if the tests for the best arguments for law are present at some level of abstraction in the cases. And so, in the end, he argues that there is no special 'meta-field' from which independent assessment of particular legal -or value- arguments may be made. (He once contemplated writing a book entitled 'Philosophy's Tutor' in which Law would be the wise tutor to Philosophy). He is critical of 'low level' justifications of the Sunstein and Sen sort because they both require a level of agreement that keeps the level low, meaning the justification, while pleasingly short, rests on consensus.¹⁴ Consensus is fact, and morality is not, and although a consensus may create a stable solution it is essentially a compromise that doesn't rest on a moral foundation. His original criticism of Rawls's idea of consensus as the basis of a 'political' concept of justice is basically the same.¹⁵

Moral judgement requires reference to our convictions as well as rationality. But of course rationality and convictions are not exclusive, one reason being our strong conviction that we should be rational. (Ask yourself whether you ought or must conclude, knowing that all men are mortal, and that Socrates was a man, that Socrates was mortal). You must face your convictions and your convictions in the end define your moral being (or 'dignity'). Further, it is against the background of moral value, not science, that we should develop methods of moral argument. Since there is no value reality of the 'external' scientific sort, our 'ultimate convictions' are not like external empirical propositions. Rather, our convictions are a constitutive part of our moral experience. But even to say that they are 'part of us' has the aroma of the subjective to it. It is wrong to go anywhere near thinking that value judgements based on our convictions are 'subjective' or 'perspectival'¹⁶ or 'mind-dependent' or 'non-realist' for that also sees value argument from the point of view of 'object' or 'angle' or 'the outside looking in' or the 'unreal'. (Dworkin's idea of working from the inside out does not employ the external metaphor but rather that he works from individual cases to more general propositions that unify and unite different individual cases).

Dworkin's method of value argument is at its most abstract a more worked out version of Rawlsian 'reflective equilibrium'. This method of value argument requires interplay between one's convictions and rationality and contains the seeds of the unity of value thesis. It is not merely a rationalisation of our own convictions because of our likely confrontation with inconsistency, with reliance on false or irrelevant historical facts including one's personal history, or otherwise irrational ways such as emotional reaction. (Dworkin's 'authenticity') This confrontation with one's own ideas may lead us to modify or abandon our convictions, or hold them less strongly than others. It doesn't follow, of course, that there is some criterial method implied here that ensures the right answer (Raz's suggestion).

Dworkin discussed 'reflective equilibrium' years ago in a commentary on Rawls's moral methods, in the context of how to 'construct' a moral justification.¹⁷ The idea that one might change one's convictions is important because it marks the difference between scientific reasoning and value reasoning (in this case, moral reasoning). His point is that our intuitions should not be thought of as analogous to external empirical data. A 'natural' model of reflective equilibrium would incline us to assume that our moral faculty might 'outstrip' our explanatory powers, in the same way as a scientist would suppose that some empirical data -an incomplete set of bones, eg- could not be reconciled with the existence of a particular genus until new data was discovered and so was incomplete. Dworkin's preferred 'constructive' model means that we must make the most coherent sense of the totality of the intuitive 'data' -our convictions- that we have. Imagine applying that way of reasoning to the scientist. He would have to 'construct' a skeleton only out of the bones he possesses, thus creating something at odds with possible evidence that science has not yet discovered.

Why does Dworkin propose the constructive model? Again, the answer lies in moral justification. The constructive model allows a better account of moral responsibility than that implied by the natural model. The natural model permits a reneging on a particular sort of responsibility. The scientist once he has done his duty of completing and understanding his experimental work can throw up his hands and say 'insufficient evidence'. The moralist cannot do this, as all the arguments are available to him. This distinction is particularly apparent in courts. A judge can say there is insufficient evidence, and acquit on that basis, but can't say there is insufficient legal argument, or that he can't make sense of the arguments put to him. He can't throw up his hands and decide nothing.¹⁸

Raz suggests one interpretation of Dworkin's view (the 'objectivity' interpretation) is that if we act responsibly, we will find the right answer to the moral question.¹⁹ I disagree. Pretty clearly someone can be as diligent as they can using the constructive method and come to the wrong conclusion. That is, they can get it objectively wrong but as they've tried their best they can't be said to have acted irresponsibly. Is this more of the scientific bias on Raz's side? ²⁰ He seems to think there is 'something there' that makes morality true independently of any moral judgement. His alternative account of Dworkin's view is what he calls the 'constructive' or 'innovative' approach to moral reasoning. This is clearly the right account, but again, Raz betrays his own belief in the quasi-scientific approach for he says that Dworkin must be rejecting the constructivist view when he says that 'Morally responsible people may not achieve truth, but they seek it'.²¹ I can't see the incompatibility between the 'objectivity' approach and the 'constructive/innovative' approach.

As I see it the position is relatively straightforward if we ignore science, quasi or not. Dworkin is saying -for the moral reason behind his preferring the constructive method- that the constructive method imposes on us the responsibility to make the most coherent sense we can of our moral convictions and be prepared to abandon some or all of them, or modify appropriately, when rationality requires it. We might make a mistake but that makes no difference to our discharge of responsibility. The mistake doesn't arise from an external moral reality but instead from amenability to reasons and the possibility of error. Convictions are crucial to moral argument but they are not the only argument. There is no demonstrability. These complexities are embedded in the way reasonable people argue. Although I think this is a lot -a breakthrough- I don't think it is much more than this. Dworkin's focus and concentration is on the arguments that we already engage in, on personal ethics, on morality, in politics, and on law. For it is these arguments that tell us how to argue, not quasi- meta-theorising. It is difficult not to see the moral force of his injunction to us.

I also think that Raz makes more of the unity of law thesis than he need, perhaps again because he seems concerned to map and 'individuate' the world of values. The unity of value thesis finds its expressed origins in the 1973 paper. My sense of Dworkin is that his theory, not just on this point, is remarkably consistent over many decades. I often feel that he'd worked most of his general theory out - perhaps with the exception of equality of resources- in his thirties. What he means by unity is just an elaboration of our responsibility to be rational, that responsibility requires us to be consistent and coherent (for consistency -mere 'elegance'- as Dworkin says many times does not ensure coherence). Thus there are no genuine conflicts between values. If you take the view, as I suspect Raz does, that values exist in some external world where they can be 'individuated' and shown to have properties that in their nature conflict, then the idea will seem mysterious.²²

Dworkin in fact doesn't deny possible conflicts between values where they can be explained by some abstract account that reconciles them. He is explicit that we may rightly conclude that there are incommensurable values in some cases. But he doesn't want conflict or incommensurable values to be a default position when the argument gets difficult. We have a responsibility to push on. What he is against is the supposition that there are conflicts and incommensurables 'out there' that impinge on our ability -or outstrip our ability- to make rational moral judgements. We should not compromise our values, or trade them off, or 'balance' them. Again, it is his dislike of the 'throw up our hands' position, which as I have said he sees as a reneging of responsibility. It is this view that led him to criticise Berlin's idea that liberty and equality were in fact in conflict. Berlin's bleak view arose from his observation of empirical reality, 'the world that we encounter in ordinary experience'.²³ Did Berlin really think that the morality of freedom and equality could be founded on our empirical experience of 'ordinary experience', by the way? He wasn't doing sociology!

Two examples to finish with to show first, the power of the unifying type argument and second the importance for Dworkin of responsibility in constructing an argument:

i. In Fisher v Bell²⁴ the interpretation of a statute making a criminal offence to offer to sell flick-knives was held not to apply to a shopkeeper because he had flick-knives for sale in his shop-window. The argument was that under the law of contract, it was well established that having goods in one's shop window was not an offer to sell but an 'offer to consider an offer' made by a customer. And so it seemed that there was a possible 'conflict' existing in the English legal system between its criminal law and its contract law. This case was widely criticised for hampering police control of a flood of cheap flick- knives circulating in Britain in the late fifties. An explanation of the case is that the judges sought unity of value by deciding that 'offer for sale' meant the same in all compartments of law. But this would be to ignore another possible unifying interpretation I recall his saying that the judges had to 'look until they saw something', that different compartments of law serve different purposes, and what contract law said was irrelevant. But there is a third possibility which might have been the underlying one. This was to resolve the interpretive conflict by favouring the more abstract principle that ambiguity in statute should be resolved in favour of the defendant. It seems to me that this case shows the force of the unifying argument against the face of conflict.

ii. This is an example Dworkin used several times during seminars he gave with Gareth Evans in Oxford in 1973 on objectivity. Dworkin imagined a newspaper competition for which readers had to offer jingles on the quality of a particular brand of cornflakes. There were other tasks (eg matching parts of photos) but this was the tie- breaker. Dworkin thought that however bland, however boring, however superficial the jingles were, the judges nevertheless had the responsibility to make their decision on what they thought was the best one. He thought it would be reneging responsibility to announce that there was too little there to base a decision, that there was no 'objectivity' to such trivia. He also thought the judge must attempt to understand the reasons that competitors have to suppose that one entry might be better than another and, if necessary, go back and look at the entries with the mission of deciding which was the best. I recall his saying that the judges had to 'look until they saw something'.