|公 法 评 论||惟愿公平如大水滚滚，使公义如江河滔滔
et revelabitur quasi aqua iudicium et iustitia quasi torrens fortis
RONALD DWORKIN'S RIGHT ANSWERS THESIS THROUGH THE LENS OF WITTGENSTEIN
Louis E. Wolcher [n.*]
I. Introduction: the Bedeutungsk?rper Picture of Language and Philosophical Confusion
This is a case study of philosophical confusion. The case studied is Ronald Dworkin's 1991 essay Pragmatism, Right Answers, and True Banality, [n.1] and in particular the parts in which Dworkin once again describes and defends his thesis that hard cases have right answers. [n.2] I claim that Dworkin's confusion stems from an inability or refusal to resist the influence that is exerted on his philosophizing by a certain method of depicting language--one which insists always on portraying words like "right answers in hard cases" (hereafter "RA") as standing side by side with something else: namely, a thing called the words" "meaning." Dworkin is by no means alone in letting himself be transfixed in this way, and therein lies the general interest of the confusion he displays. For the picture of the way language works in which linguistic signs like "meaning" and "RA" are projected as having a Bedeutungsk?rper (meaning-body) [n.3] corresponding to them is a familiar and remarkably enduring one in philosophy. In The Blue Book, Wittgenstein identified our inclination to look for some thing corresponding to the noun in sentences like "What is length?," "What is meaning?," and "What is the number one?" (to which list I will add the question "What are right answers in hard cases?") as "one of the great sources of philosophical bewilderment." [n.4] Later, in the first paragraph of Philosophical Investigations, Wittgenstein elaborated on this picture of language by quoting from Augustine's description of how he learned to speak as a young child. [n.5] In The Confessions, Augustine recounts that his elders used words to name objects, and that "by hearing words arranged in various phrases and constantly repeated, I gradually pieced together what they stood for." [n.6] For Wittgenstein, Augustine's description offers "a particular picture of the essence of human language," which he characterized as follows: "the individual words in language name objects--sentences are combinations of such names.--In this picture of language we find roots of the following idea: Every word has a meaning. This meaning is correlated with the word. It is the object for which the word stands." [n.7] It is important to see that this passage locates in Augustine's primitive idea that a word's meaning is its bearer (the physical entity it names) the roots of a somewhat more sophisticated idea: namely, that the object for which a word stands is not (or at least not necessarily) its bearer, but rather another entity called its "meaning." In other words, Wittgenstein is telling us that the more sophisticated idea of the way language works is not different in kind from Augustine's, but is at bottom merely a variation of the primitive picture that Augustine draws. If we wanted to compare the production of sentences to the production of paintings, we could say, for example, that the sentence "The word "Rover" stands for my dog," and the sentence "The word "RA" stands for its meaning," are drawn in the same style.
The picture of language that Wittgenstein describes is a contingent one: we do not have to employ it. But if we do use it, it tends to take hold of our intelligence and produce a curious effect on our philosophizing. That is, if we insist on thinking of linguistic signs as things that do or do not have meaning in the same way that a sweater does or does not have the color blue, or a box of marbles has or does not have marbles in it, then certain puzzles can (and do) arise while we are philosophizing, and we can (and do) run around in circles trying to solve them. Thus we ask questions like "What is the meaning of "RA%?," and think that the answer must be a proposition asserting the existence of a relation between the sign "RA" and something else for which "RA" stands. The question presupposes, in other words, something like the following picture of the kind of answer that will satisfy it: "Answer: the meaning of "RA" is x." Now it turns out that many of us are content to live our lives happily accepting a particular value of x in what for us is the definitive answer to the question that our own picture of language leads us to ask. But for other, less contented, souls it can come to pass, sooner or later, that each in the series of answers we try out begins to feel more or less dissatisfying. Indeed, to some of us, the values of x that we have seen written down in various answers all begin to seem wrong, incomplete, or worse. Not only do we remain puzzled, but we begin to wonder if there is any answer at all to our question. But instead of changing our picture of the way language works, we decide to cling to it all the harder. We feel inclined to say things like "The word "RA" has does not have a meaning, only different meanings, depending on who is giving the answer." But even as we think or say this, we continue to imagine linguistic signs under the aspect of the Bedeutungsk?rper picture. It's just that now we imagine them as disembodied things next to which no meaning (or no one meaning) stands, in much the same way that no coat of wool stands next to the skin of a lamb that has just had its springtime shearing. But of course the naked lamb is naked only because it once had a coat of wool--if this were not the case for lambs in general, probably no one would ever have thought to call spring lambs "naked."
Either way,--whether we think that words have meaning or we think that they don-t--we just can-t bring ourselves to let go of the picture! Rather than opening our eyes and looking at the world to see what is the case, we let our method of representation take control of our intelligence and dictate in advance what must be the case. It is as if someone saw M.C. Escher's lithograph Waterfall[graphic unavailable in this format], and then, instead of asking questions like "What technique did Escher use to achieve such a convincing optical illusion?," were to say: "Where on earth is this odd place that Escher found--a place where water runs downhill and uphill in a continuous stream?--I must set out immediately on an expedition to find it." And the skeptical turn in philosophy in general ("Words have no meaning") [n.8] and legal philosophy in particular ("There are no "right" answers to hard legal questions, only "different" answers") [n.9] can be compared profitably to what would happen if this person, after he looked all over the earth for a place corresponding to Escher's picture, but without success, were to throw up his hands in exasperation and exclaim: "Escher's picture is a false statement about the place that it depicts--there is no such place!" When we are philosophizing about law, we fail to see that the word "meaning" no more has to apply to "RA," the way lawyers use it, than Escher's picture has to portray a location in the physical world. We fail to see that no question of "RA-s" meaning arises unless we ourselves ask it, and that the sense of any question we do ask is intimately bound up with the method we have for answering it. We let the shadowy object that our prejudice tells us the word "meaning" must stand for specify our act of describing meaning, instead of investigating how we describe meaning by unpacking the descriptive techniques that we have laid down for ourselves well in advance of any particular act of description. [n.10]
Dworkin adheres to the Bedeutungsk?rper picture of meaning in Pragmatism, Right Answers, and True Banality. And this leads him into confusion even as--indeed especially as--he insists that the right way to ascertain the meaning of "RA" is not to look for objects existing "out there" in some transcendent plane, [n.11] but to look at the ordinary uses that lawyers make of that symbol. For Dworkin prescribes the investigation of ordinary use as a means of ascertaining the meaning of signs like "RA." That is, he makes the investigation of use into a technique for bringing words into contact with their meaning: a meaning that is not the same as use, but rather something else. Now Dworkin gives evidence in his essay and elsewhere of having read Wittgenstein with at least some amount of understanding, if not approval. [n.12] And although in such remarks as "the key to meaning is use" one clearly detects the influence of Wittgenstein's ideas on Dworkin, [n.13] one also sees the fulfillment of Wittgenstein's sad prophecy that "[t]he seed I'm most likely to sow is a certain jargon." [n.14] Dworkin turns the investigation of use into a handservant of his theory of meaning, instead of investigating the use of language for the purpose for which Wittgenstein prescribed it: to cure philosophy, once and for all, of the inclination to ask (and waste time answering) questions like "What does "RA" mean?" [n.15]
The assumption on which Dworkin's right answers thesis rests is this: if all or most lawyers and judges use the symbol "RA" the same way in their practice, then it follows that "RA" has something called a "meaning" that is given to it by that sameness of use. In what follows, I will try to demonstrate how this assumption, instead of being true or false, is without sense--that it is, in fact, an excellent example of the kind of nonsense that philosophers laboring under the influence of the Bedeutungsk?rper picture of language are prone to say. I will also try to show that if Dworkin were to engage his skeptical critics on a field where the determination of the sense of their no-right-answers thesis preceded any determination of its falsity, then his affirmative philosophical project--to build a program for judges and lawyers on the foundation of "RA" having a meaning--would be exposed as being no more than an idiosyncratic rule of grammar that he has laid down to govern the way he talks about law while he is philosophizing. Finally, I will argue that Dworkin's plea for a "more constructive jurisprudence," [n.16] unless we are inclined to dismiss it as just another bit of philosophical nonsense spoken by someone who is confused about how language works, is best understood as a call for philosophers of law to abandon philosophy and become lawyers.
II. The Right Answers Thesis and the Meaning of "Right Answers"
In his 1991 essay, Dworkin defends the thesis that there are right answers in hard cases by revealing that his motive for asserting it is to save legal practice from the falsehoods of skepticism. He writes:
The only reason anyone has to say that there can be right answers in hard cases is that this has been denied, sometimes in an interesting, internally legal way, by many legal philosophers, and it is of great legal importance that their skeptical claim be understood to be false because practice would have radically to change if it were true. I said that there are right answers, that is, only because and after others said that there are not. [n.17]
Dworkin characterizes his claim that there are right answers in hard cases as "a claim made within legal practice rather than at some supposedly removed, external, philosophical level," and he equates the meaning of the phrase "right answers in hard cases" with the "ordinary sense in which lawyers might say this." He then explains that the ordinary sense in which lawyers say that there are right answers in hard cases is captured in lawyers" opinions, held or expressed in the context of particular hard cases, that there is one legal argument that is "on balance the soundest." [n.18] That lawyers do not always (or even hardly ever) all agree on what the right answer in a hard case is in no way subverts, for Dworkin, the sense of their shared opinion and belief that there is indeed a right answer. It's like this, for Dworkin: lawyers can and do differ in saying who ought to win a particular case, but at some level the words "the right answer" plays a role in their practice--in their use of it--that is the same for all of them.
In A Matter of Principle, Dworkin describes this sameness of use in a bit more detail by saying that the way most lawyers talk and argue shows they assume that the bivalence thesis is true. [n.19] The principle of bivalence holds "[t]hat every proposition is true or false, none being neither and none both." [n.20] If a proposition that someone has a particular legal right is represented as p, and its negation as ~p, then even if lawyers have genuine disagreements about which answer--p or ~p--is the right one, they behave in their practice as though they accepted the truth of the following statement: " Either p or ~p (but not both of them) is true ." For example, if a system's legal grammar stipulates that the dispositive concepts are "legal right" and "no legal right," we would not expect to see lawyers constructing their arguments in order to appeal to a third conceptual space lying between or outside p and ~p, namely "neither p nor ~p." And indeed, in the American legal system we hear lawyers arguing in court about whether Mary's promise to John is or is not legally enforceable, but we hardly ever hear them arguing (at least in court) that Mary's promise is either: (a) legally enforceable; or (b) not legally enforceable; or (c); or (d) yet another movement in the ongoing dialectic of gender and class oppression. The formal expression of the bivalence thesis -- "The proposition " p or ~p, and not (p and ~p) " is true"--is essentially the same as Dworkin's thesis that there is a right answer to at least some hard cases, [n.21] and is taken by Dworkin to mean something that all lawyers agree about: their actions, he maintains, show that they agree that the bivalence thesis is true. [n.22] Indeed, his attack on the implicit empiricist metaphysics behind the demonstrability thesis [n.23] is premised explicitly on the view that statements are "given" whatever sense they have in an enterprise by their use in that enterprise, and not by virtue of the existence of some "hard fact" in the world that makes them true. [n.24] In other words, sameness of use implies sameness of meaning. Dworkin wants us to understand first and foremost that an investigation of the way lawyers use words shows that there is a meaning that "RA" stands for--something about which some legal philosophers assert, with sense, that it does not exist, and about which he asserts, with sense, that it does.
It is important to understand that Dworkin's claim about the meaning of "RA" hangs on a consensus of use, and not on a consensus of opinion or interpretation. A consensus of use is shown by what speakers do with a symbol; a consensus of opinion or interpretation, on the other hand, is shown by what speakers say they agree about. If Dworkin's criterion for the meaning of "RA" were of the latter type, one would expect to find him arguing that the meaning of "RA" is what all lawyers would say (in the form of their opinion or interpretation) when asked the question "What does "RA" mean?" But Dworkin does not argue this, and for good reason, for all that this criterion of meaning would get him is a regress of symbols, with nothing left over to establish the meaning of "RA" that he both wants and needs. For opinions and interpretations, if they are to do us any good as criteria of meaning, must first be represented. And if, in general, the meaning of a symbol is taken to be what a given group of people say it is, then even if they all say the same words, their common expression of meaning will take the form of yet another symbol: the symbol "b," for example, in the answer "Symbol "a" means b." But here it can be seen that a criterion of meaning that equates the meaning of a symbol with the expression of a consensus of opinion or interpretation ("b") cannot establish that those whose expressions of meaning are the same also all mean the same by "b" unless it asks the further question, "And what, then, do you mean by "b%?" And of course the answer to this question will take the form of still another symbol. As Wittgenstein says, we can extend the regress of symbols by this method just as far as we want, but at some point we will (and do) stop at a symbol (the last one) of which it must be said: "there will be no such thing as an interpretation of that." [n.25] In other words, there arises no question of whether there is (or is not) a consensus of opinion or interpretation on the meaning of the last in the series of symbols unless we ourselves ask it. But we do not ask it (we stop the regress), and so the meaning of the last symbol--the one that is supposed to deliver us meaning--is something about which our consensus-based theory has absolutely nothing to say.
Dworkin seems to understand that a theory of meaning predicated on consensus of opinion or interpretation will not accomplish his purpose, for he turns a variation of the argument just made against his pragmatist critics, and especially Richard Rorty, who he accuses of improperly distinguishing between two levels of discourse. According to Dworkin, Rorty-s distinction between the "internal level at which some practical enterprise like law . . . is carried on" and the "external level at which philosophers and other theorists talk about these enterprises rather than participate in them" presupposes the intelligibility of the language that is used to express the distinction. [n.26] But in expressing the distinction, Rorty and his pragmatist allies "would have to paraphrase the philosophical statements in some way to bring out their supposedly special meaning, and in doing that they would have to fall back on other words and ideas that also have a perfectly ordinary and clear use, and they would then have to tell us how those words mean something different from what they do in that ordinary use." [n.27] Dworkin's attack on the "false distinction between ordinary and philosophical discourse" leads him to say, of Rorty's brand of pragmatism, "there is nothing in it to accept." [n.28] Although Dworkin calls Rorty's pragmatism "banal," [n.29] it is unclear whether he is asserting that Rorty's expressions mean something banal that is not worth accepting, or that they are banal because they mean nothing at all. In either case, the most important point Dworkin is making is that one does not (by his lights) satisfactorily answer questions like "What is the meaning of the symbol "RA%?" by equating it with another symbol. Rather, one needs to be attentive to how symbols are used--only after the use is understood will the meaning be understood. It all hinges, for Dworkin, on the idea that the use of a symbol endows it with meaning.
Dworkin's claim about what "RA" in particular means coheres completely with his general theory of meaning:
Language can only take its sense from the social events, expectations, and forms in which it figures, a fact summarized in the rough but familiar slogan that the key to meaning is use. That is true not only of the ordinary, working part of our language, but of all of it, the philosophical as well as the mundane. Of course, we can use part of our language to discuss the rest. We can say, for example, what I just said: that meaning is connected to use . . . . But we cannot escape from the whole enterprise of speech to a different and transcendent plane where words can have meanings wholly independent from the meaning any practice, ordinary or technical, has given them. [n.30]
At the root of this passage we find the idea that words have meanings that owe their existence to what we do with them. This human activity of doing-things-with-words (using them) is offered as the "key to meaning:" that is, the uses we make of what are otherwise lifeless signs endow those signs with a property (meaning) that they then can and do "have," and that gives them what they need to come to life as language that is meaningful, in the sense of being full-of-meaning. Although Dworkin intimates that the correct way to understand a word's meaning is to look at its use in context, he does not prescribe the investigation of use as a cure for what he later calls "philosophical bewitchment." [n.31] Rather, he offers the investigation of use as the means to a different end: the end is to know the meaning of words in this or that context, and knowing the meaning is something towards which the investigation of use is oriented as a means. The Bedeutungsk?rper picture of language continues to grip his imagination even as he thought he had shaken it off by turning his eyes from the Platonic realm of transcendent meanings to the real world of actual use, for even now it prompts him to say that a symbol's meaning must be what his investigation of actual use shows it to be.
The philosophical confusion lying at the foundation of Dworkin's claim--that because lawyers use "RA" in the same way, therefore, that symbol must have something called a "meaning" with which their use endows it--can be summarized as follows: he thinks that a description of use (his only criterion of meaning) can deliver him something more than just what it describes. His confusion is like that of the man who thought he was on to something important about the essential nature of human goodness by announcing a program in which he would describe whatever human beings happened to do, and then call it all "good."
The source of Dworkin's confusion is best exposed by inverting the premise of his argument. Instead of "lawyers use "RA" in the same way," let the premise be "lawyers use "RA" in different ways." Suppose, for example, that lawyer A uses "RA" to mean the legally correct answer in a case, and lawyer B uses "RA" to mean the politically convenient answer. Then, following Dworkin, we would be inclined to say that "RA" must mean something different to the two lawyers. But it would be illicit to defend this inclination simply by saying that the phrase "legally correct answer" means something different than the phrase "politically convenient answer," for this would leave open a question that leads to the unhelpful regress described earlier: What is it, then, that these two phrases mean that is different? So let us ask a different question instead: taking the root term "RA," used by A and B alike, to be merely a naked symbol, what does it (including any symbol that is substituted for it) mean that is different in the two cases? To ask this question at all is to presuppose a method or methods for answering it. Just as there are various methods of answering the question "How tall is the Empire State Building?," so too there are various ways of getting an answer to the question "What does "RA" mean to lawyer A and lawyer B?" For present purposes, we need only consider three of the many possible ways for settling this question: (1) meaning is given solely by a representation of the picture that the symbol suggests to the imagination of those who use it; or (2) there are two criteria for meaning: the picture that a symbol suggests to the imagination (as in (1)) and how the symbol is used by the person to whom the picture is suggested; or (3) meaning is given only by how the symbol is used (Dworkin's view).
Suppose that lawyer A would associate "RA" with a picture of a criminal defendant being released from jail, whereas lawyer B would associate "RA" with a picture of the same defendant being hanged for his crime in front of a cheering and approving crowd. Let A and B draw or point to their respective pictures. In that case, if the meaning of a symbol is taken to be the picture that it suggests to the imagination--possibility (1)--then it would make sense to say that "RA" means something different in the two cases, for A and B do not point to the same picture. But now suppose that there are two criteria for the meaning of the symbol "RA"--the use of the symbol and the representation (by words or otherwise) of the picture that is suggested to the imagination by that symbol (possibility (2)). In that case it is clear that we would need to know how these two criteria work together. We would have to have a way of deciding, for example, what the meaning of "RA" is when we observe two lawyers using it the same way in practice but also giving different representations of the image that is suggested to them when they are asked the question, "What is the RA?" But of course Dworkin takes pains to deny that any part of his criterion for the meaning of "RA" is that lawyers agree in their representations of which answer to a hard case is right, and so his theory allows him to avoid having to resolve problems like the one just mentioned. [n.32] He makes use alone his criterion of the meaning of the symbol (possibility (3)). But like consensus of opinion, consensus of use must be represented (or at least be representable) to be of any value as a criterion of meaning. If Dworkin were challenged to support a particular claim of meaning, for example, his commitment to a use theory of meaning would require him to describe use in order to make out his case. This means that the role that a representation of sameness of use plays in his right answers thesis is to establish what the meaning of "RA" is.
We are now in a position to see plainly what the source of Dworkin's confusion is: if the meaning of a symbol is represented by the use of that symbol, and only that, then it is otiose to say, "The use of "RA" is the same (or different), therefore the meaning is the same (or different)." Consider a parallel case: If our only criterion for the meaning of the sign "one meter long" is that a body comes up exactly to the height of our measuring rod, then we gain no philosophical ground by saying "All of the pieces of wood in this room come up exactly to the height of the measuring rod, therefore they are all one meter long." [n.33] This sentence is essentially a rule for the substitution of signs: all that it does is allow us to employ the words "one meter long" as a label in expressing the circumstance that all the pieces of wood in the room come up exactly to the height of the measuring rod. Assuming that Dworkin is right in reporting that the use of the symbol "RA" is the same for all (or most) lawyers, [n.34] therefore, we could read his theory as a rule for the substitution of signs. We could imagine a role for the description of sameness of use in his theory as follows: Dworkin might be calling his description of use "the meaning of "RA%"--he might be read to have laid down a rule in the syntax of the language that expresses his theory that authorizes him to substitute the linguistic sign "the meaning of "RA%" for the linguistic signs that make up his description of the way lawyers use "RA." Roughly speaking, this would be like the case of the man who told his neighbors that henceforth he was going to start calling everything they did "friendly deeds." Dworkin undoubtedly would reject this characterization of his theory--would deny that he uses "RA" as a mere label for whatever lawyers happen to do with that symbol. But this way of putting it does have the advantage of showing that if we lay it down as a rule that a description of use is what gives us meaning in our theory, then the question "What is the meaning of "RA%?" asks exactly the same thing as the question "Can you tell me, without paraphrasing, what the description of the use of "RA" says?" And of course the correct answer to questions of the latter sort is always this: "I will read the description out loud to you, and then you will be able to hear for yourself what it says."
It is as if Dworkin were saying that the meaning of "RA" is the hollow of a mold into which its use (a solid) fits exactly. [n.35] But we really do not see any hollow of meaning before the use fits into it--we do not see two things, the hollow and the solid, before they are fitted together (as in i), such that we could compare the shape of the one with the shape of the other. Rather,
i)[graphic not available in this format] ii)[graphic not available in this format]
all we see in Dworkin's theory is the description of use, and it already lies in, and completely fills, the hollow of meaning (as in ii). So here there is no question of comparing the one with the other. If Dworkin held that meaning is use (which he does not) then we would be entitled to rewrite his phrase "the key to meaning is use" [n.36] as "the key to use is use." Since we cannot rightly call his expression a tautology, and since reading his theory as a mere rule for the substitution of signs makes it trivial, we are left wondering what other role the word "meaning" could play in this and kindred phrases that he writes. Well, here is another possibility: to say that a description of use shows what "RA" means could be read as a kind of emphasis. Dworkin might be using the word "meaning" to emphasize or accentuate whatever descriptions of use he does give. On this reading, we could give sense to the various expressions of Dworkin's theory of meaning by hearing him say something that is the equivalent of "Look at the way lawyers use "RA%!" (spoken in a loud tone of voice). But while this interpretation does describe a possible use for Dworkin's words--the phrase "Hard cases have right answers" would be like an exclamation point, on this reading--it does not deliver him a Bedeutungsk?rper to go along with the symbol "the meaning of "RA%." The little word "meaning" (in my dictionary it comes between "meander" and "meanwhile") is just too slim a reed on which to float a philosophical project as ambitious as Dworkin's. [n.37]
III. Bipolarity: Sense and Nonsense Before Truth and Falsity
In the Wiener Ausgabe, Wittgenstein sums up his philosophical method as a transition from the question of truth to the question of sense. [n.38] But in Pragmatism, Right Answers, and True Banality, Dworkin once again passes up the opportunity to engage skepticism about right answers on a field where the determination of sense precedes the determination of truth or falsity. On such a field, sentences that are ruled nonsense never earn the right to be called "false," because they are not understood to represent any state of affairs of which it could be said, "This is what it would be like if the sentence were true; but that's not what it is like, and therefore the sentence is false." One way to deal with a philosopher who asserts a proposition of the form "x does not exist" is to ask him to give a representation of what it is that does not exist. In the case of someone who asserts "Unicorns do not exist," for example, the speaker could be asked to give a description or draw a picture of the thing that he would call a "unicorn." His representation (UNICORN [graphic unavailable in this format]) could then in principle be compared with reality (according to some method of comparison) as a means of deciding the question "Do unicorns exist?," and hence of determining whether the answer "Unicorns do not exist" is true or false. But the skeptical philosopher of whom Dworkin writes--the sort who asserts "Right answers in hard cases do not exist"--does not want to say that there is something describable or depictable which corresponds to the symbol "RA" in her proposition, and of which she is saying that although this thing-s existence is imaginable, it just so happens, empirically, that it does not exist (although some day it might). This kind of skeptic wants to say, rather, that any answer that anyone could give to the question "What is the right answer?" would automatically amount to a different (including historically and politically contingent) answer. She would reject a priori any answer, in other words, as not being what she means to exclude when she says "Right answers do not exist." [n.39] Here the skeptic reveals that, although she stipulates in advance that neither she nor anyone else could represent what "RA" means, she would nonetheless like to say that what "RA" stands for (its Bedeutungsk?rper) does not and cannot exist. But if that is so, how could anyone even try to give a right answer? What is it that the skeptic rules out as absolutely impossible here? I know what it would look like for me to sprout wings and fly--it is this imaginable and representable state of affairs that I mean to exclude when I say "You know, it really is impossible for me to sprout wings and fly." But in the case of the skeptic who says that there are and can be no right answers, only different ones, what is it that is being excluded? What is it that I know there cannot be, if I know there cannot be right answers? If the skeptic lays it down in advance that no one can represent what she means when she excludes the possibility of "RAs," then she can be asked to explain how the question about right answers--to which the skeptical claim is supposed to be the answer--can be asked at all. She can be asked, for example, to respond to Wittgenstein's critique of skepticism in the Tractatus: "Scepticism is not irrefutable, but obviously nonsensical, when it tries to raise doubts where no questions can be asked. For doubt can exist only where a question exists, a question only where an answer exists, and an answer only where something can be said." [n.40]
To deal with the skeptic in the manner just described would amount to an investigation into what the young Wittgenstein called the "bipolarity" of propositions like "Right answers in hard cases do not exist," [n.41] and would serve the purpose of exposing the nonsensicality, not the falsity, of what the skeptic asserts. "There are no right answers in hard cases" would be shown as standing on the same level as "There are no grynxs" (a silly word that I just made up). But this method of argument is dangerous from the perspective of those who want their own reports of the world not just to make sense (to have a use), but to be true, for the method, once unleashed, very quickly turns back on its user. If the skeptical claim is not seen to be false, but rather just nonsensical, then the negation of the skeptic's claim ("It is not the case that right answers in hard cases do not exist") is not a report about the world at all. This negation is exposed as a rule of grammar according to which meaning is established--it is not itself an expression that is answerable to any meaning that it could contradict. [n.42] It is a rule of grammar in this sense: just as we may have no use (as things stand now) for syntactically correct strings of words like "The number five has green complexion," and thus may tell a child who utters these words, "This sentence is nonsense--you must not say it anymore," so too an investigation of the non-bipolarity of the legal skeptic's claim might reveal that "Right answers in hard cases do not exist" is a syntactically correct sentence that is not false, but rather a string of signs for which we have no use (as things stand now). Thus we might lay down a stipulation like this excluding it: "The sentence " There are right answers in hard cases " is a rule in the syntax of our legal language which does not permit anyone to say "There are no right answers in hard cases%." Seen from this perspective, Dworkin's long-standing argument "against the positivist's claim that there cannot be "right" answers to controversial legal questions, but only "different" answers" [n.43] is just another form of his argument for there being a Bedeutungsk?rper corresponding to a description of the way lawyers use the symbol "RA." It is not enough for Dworkin's purposes that legal skeptics are uttering the philosophical equivalent of gibberish--he must read them to sensibly deny the existence of something that "RA" means. Enemies need something to fight about. And an enemy who is merely confused about grammar would not give Dworkin anything to fight about--only someone to pity.
IV. Using "Right Answers" the Way Lawyers Do: Philosophers as Lawyers
Dworkin issues "a plea for a more constructive jurisprudence" at the beginning of Pragmatism, Right Answers, and True Banality. [n.44] What he would like to be able to tell his critics is that they should give up the errors of skepticism and join with him in the project of figuring out what the symbol "RA" means in particular hard cases. Thus, at the end of the essay he writes:
We should now set aside, as a waste of important energy and resource, grand debates about whether law is all power or illusion or constraint, or whether texts interpret only other texts, or whether there are right or best or true or soundest answers or only useful or powerful or popular ones. We could then take up instead how the decisions that in any case will be made should be made, and which of the answers that will in any case be thought right or best or true or soundest really are. [n.45]
But what Dworkin would like to say to his critics is just what he cannot say--or rather, he can and does surely utter these words, but what he cannot say, with sense, is that "RA" means something that skepticism could be wrong about. For "RA" is a symbol that lawyers use this way and that way, and a complete description of the facts of their use of it would deliver to Dworkin and his critics nothing more than what the description says. Such a description would be a book in which not a single word is written about what "RA" means. A special chapter describing the uses of "RA" in the context of asking and answering questions like "What is the meaning of the symbol "RA%?" would exist only if the book's author found any lawyers and judges who actually talked that way. But of course even then the chapter would describe the use of these words, and not their meaning.
In Wittgenstein's later philosophy, the word "meaning" hangs closely together with the concept "form of life." And Dworkin clearly shows signs of having understood and embraced the latter concept, at least at some level. For participants in a linguistic practice to understand one another, Dworkin writes in Law's Empire, "means not just using the same dictionary, but sharing what Wittgenstein called a form of life." [n.46] Accordingly, if we wanted to be charitable to Dworkin, we could give the invitation at the end of Pragmatism, Right Answers, and True Banality sense by reading it under the aspect of Wittgenstein's concept of a "form of life." We could read it, in short, as a call for philosophers of law to begin participating with lawyers in the same form of life. But such a call would not come down to asking philosophers of law just to talk to lawyers, or to talk like lawyers talk. It would come down to asking them to be lawyers. For lawyers are the only kind of people who could use "RA" in the same way that lawyers do.
The last sentence of the previous paragraph is a tautology. It is meant to draw attention to the fact that the concept "use of language" stands in an internal relation [n.47] to the concept "form of life." The difference between internal and external relations derives from the difference between rules that make up a given grammatical system and statements that are made within that system. For example, the proposition that the color of my shirt is lighter than the color of my car asserts the existence of an external relation between the colors of two particular surfaces: these surfaces could be colored differently than they are, and it is this contingency (which is none other than the possibility that the proposition is false) that gives sense to the idea we have that someone could inspect my shirt and my car and determine whether the proposition is true. But the assertion "The color white is lighter than the color blue" is not about any colored surface; indeed, Wittgenstein would say that it is not about anything at all. Rather, it is a rule of grammar that establishes an internal relation between the color-words "white" and "blue." On the one hand, it is a rule that establishes a use for propositions like "This bit of cloth, which is white, is lighter-colored than that metallic surface, which is blue"; and on the other hand, it rules out (as nonsensical rather than false) statements like "White is darker than blue." [n.48] In other words, the proposition that something is white in color, and something else blue, entails (already includes) the grammatical relations of the former being "lighter," and "not darker," than the latter. [n.49]
It is likewise the case that, in the grammar of Wittgenstein's philosophy, a description of how a given group of people use a bit of language is also a description of an aspect of their form of life. Since "use" and "form of life" stand in an internal relation, it makes no sense to say, in the grammar of his system, "I am using language the same way these people do, but at the same time I am not participating in their form of life." Dworkin's use-driven theory of language neglects the most important sense of Wittgenstein's concept "form of life," which comes through in the following passage from Philosophical Investigations: "%So you are saying that human agreement decides what is true and what is false?%--It is what human beings say that is true and false; and they agree in the language they use. That is not agreement in opinions but in form of life." [n.50] As this passage suggests, to use a symbol the same way that a given group of people do (by Wittgenstein's lights) has nothing to do with plucking something called the "meaning" of that symbol out of one's observation and description of the people's use of it, and then talking about it (the meaning). To use it the same way they do is to use it as they do, all the way down; it is to agree in one's actions with their form of life, and not just to agree, in one's words, with something called "RA-s" "meaning." Dworkin's effort to turn lawyers" uses of "RA" into a way of use, and the way of use into something called "RA-s" meaning that philosophy could then talk about, is reminiscent of the hungry man who tried to turn his loaf of plain bread into something more by calling it a banquet.
This last point can be illuminated by considering the mistake that Dworkin makes in his discussion of lawyers" use of the grammatical distinction between following and ignoring precedent. After reporting that lawyers do in fact draw distinctions in these terms, he asks, rhetorically, "If the ordinary distinction can-t be used in [a] descriptive and critical way, then how can it be used?" [n.51] Dworkin says that there is a way in which lawyers use the distinction between following and ignoring precedent in talking about judges, and that the fact that lawyers can be seen as making interpretive claims in using this distinction does not lessen the force or cogency of their claims. [n.52] But what misleads him here is the word "way," for his description of how lawyers use the distinction does not allow for the possibility of a different way. "Maradona kicks a football, but not in the same way as Ronald Dworkin does:" here I can sensibly use the words "in the same way" because different ways of kicking have been or can be described independently of the concept "kicking a football." But if we all agreed in advance that there is only one series of motions that we would ever call "kicking a football," then my sentence would come down to saying that Maradona kicks a football and Dworkin does not: the idea of a "way" of kicking would drop out as a useless appendage. Likewise, where the description of the use of words is also supposed to give us our only access to the way they are used (as in Dworkin's theory of meaning), it makes no sense to talk about same or different ways, because the description of use leads to only one end. Wittgenstein revealed the source of this kind of confusion in one of his lectures:
The word "way" corresponds to the word "analogous," which means "something else," not the same thing over again. We cannot wish to explain "a way" independently if the way is included in what is being described. . . . What seems to be assumed is that a way has been described and that one could not get to the end of it. In fact a way has not been described. One cannot reasonably object to not reaching the end of the way if in giving the way one gives also the end of the way. There is no sense in talking of a way if there is only one end and a different end is precluded. [n.53]
Dworkin's error about the distinction between following and ignoring precedent is just like the error he makes in thinking that the meaning of "RA" is given by a description of its use: he imagines that he can describe how lawyers use a given set of symbols, and that something extra will then pop out of his description--namely, the way they use the symbols. But the concept of a "way of use" is a loophole that he himself provides when he says he is only using words the way lawyers use them, and that he himself closes when he says that the distinction between following and ignoring precedent is drawn by practice alone (by use), and not at some "different, strange, and disengaged level of discourse." [n.54] What is left of the loophole after it has been both made and then closed by Dworkin himself is exactly nothing, and certainly not anything that would give "way of use" the Bedeutungsk?rper that he needs for his theories. What is left, in short, is just the description of use.
A complete description of lawyers" uses of the symbol "RA" would be a description of their words and deeds woven together in a million different strands to make up that complex and multi-hued tapestry we call legal practice. But the tapestry's description would describe the tapestry--it would not describe the tapestry's meaning. There would certainly be things to admire and things to despise in what the tapestry's description describes. But a dogmatic precommitment to the view that, for all of its subtle and complex variations, the tapestry as a whole must mean just one thing would greatly hinder our ability to see it clearly in the first place.
If Dworkin and his critics followed this advice--%use "RA" as lawyers use it"--they would turn to writing legal briefs instead of scholarly books and articles, and they would start practicing law instead of describing how law is or should be practiced. This is not to say that the use that philosophers make of the symbol "RA" could not develop in this direction, or that lawyers might not welcome the competition if it did. But if the use of "RA" by philosophers were to come down to this--if we read Dworkin's article is a prediction of things to come--then the closing line of his essay, "Legal theory already has enough ways of wasting time," [n.55] although he meant it as a reproach to his critics, had better be rewritten to read: "Legal theory is a waste of time." This is what Stanley Fish has been saying for years. [n.56] I doubt very much that Dworkin wants to agree with him, but then again it often happens that what we want and what we say are at odds with one another.
* Professor of Law, University of Washington School of Law. An earlier version of this paper was presented at the 1997 annual meeting of the Critical Legal Conference held at University College, Dublin.
1 . Ronald Dworkin, Pragmatism, Right Answers, and True Banality, in Pragmatism in Law and Society 359-88 (Michael Brint & William Weaver eds., 1991).
2 . Earlier versions and defenses of the thesis can be found in Ronald Dworkin, Law's Empire 225-75 (1986), Ronald Dworkin, A Matter of Principle 119-45 (1985), and Ronald Dworkin, Taking Rights Seriously 279-90 (1977). In Pragmatism, Right Answers, and True Banality, Dworkin says that he has not changed his mind "about the character and importance of the one-right-answer claim." Dworkin, supra note 1, at 382 n.1. This still seems to be the case, judging from Ronald Dworkin, Freedom's Law: The Moral Reading of the American Constitution (1996). There Dworkin extends the claim to cover the case of legal and academic debate about constitutional values, which he recasts as "a debate not about how far democracy should yield to other values, but about what democracy, accurately understood, really is." Id. at 15.
3 . It appears that Wittgenstein first used the term Bedeutungsk?rper in a manuscript he wrote during 1932-34, and that was published posthumously as Ludwig Wittgenstein, Philosophische Grammatik 54 (1969). The term is translated as "meaning-bodies" in Ludwig Wittgenstein, Philosophical Grammar 54 (Rush Rhees ed. & Anthony Kenny trans., 1974).
4 . Ludwig Wittgenstein, The Blue and Brown Books 1 (2d ed. 1960).
5 . Ludwig Wittgenstein, Philosophical Investigations ?ì 1, at 2e (G.E.M. Anscombe trans., 3d ed. 1968).
6 . Saint Augustine, Confessions 29 (R.S. Pine-Coffin trans., 1961).
7 . Wittgenstein, supra note 5, ?ì 1, at 2e.
8 . See, e.g., Kripke's succint summary of the "sceptical argument:" "There can be no such thing as meaning anything by any word." Saul A. Kripke, Wittgenstein on Rules and Private Language: An Elementary Exposition 55 (1982).
9 . Cf. Dworkin, Law's Empire, supra note 2, at viii-ix.
10 . Cf. Wittgenstein's Lectures on Philosophical Psychology 1946-47 48 (P.T. Geach ed., 1988) ("But I say: "Don-t try to specify the act of description by means of the object that is to be described; but by the technique of description.%")
11 . See Dworkin, supra note 1, at 364 (asserting that advocacy of this kind of metaphysics is a charge that some of his critics have unfairly leveled against him).
12 . For example, he relies on the authority of Wittgenstein's name twice in the essay, id. at 360, 380, three times in Law's Empire, supra note 2, at 63, 69, 315, and twice in Freedom's Law: The Moral Reading of the American Constitution, supra note 2, at 177, 304.
13 . Dworkin, supra note 1, at 362. In Philosophical Investigations, for example, Wittgenstein writes: "For a large class of cases--though not for all--in which we employ the word "meaning" it can be defined thus: the meaning of a word is its use in the language." Wittgenstein, supra note 5, ?ì 43, at 20e-21e.
14 . Wittgenstein's Lectures on the Foundations of Mathematics: Cambridge, 1939 293 (Cora Diamond ed., 1976).
15 . See Wittgenstein, supra note 5, ?ì 133, at 51e, where he writes:
It is not our aim to refine or complete the system of rules for the use of our words in unheard-of ways.
For the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems completely disappear.
The real discovery is the one that makes me capable of stopping doing philosophy when I want to.--The one that gives philosophy peace, so that it is no longer tormented by questions which bring itself in question.
See also id. ?ì 309, at 103e ("What is your aim in philosophy?--to shew the fly the way out of the fly-bottle.").
16 . Dworkin, supra note 1, at 359.
17 . Id. at 367-68.
18 . Id. at 365.
19 . Dworkin, A Matter of Principle, supra note 2, at 120-21.
20 . A.R. Lacey, A Dictionary of Philosophy 24 (2d ed. 1986).
21 . In Pragmatism, Right Answers, and True Banality, Dworkin maintains that his right answers thesis is "characteristically or generally sound in hard cases," but he limits his defense of the thesis to "some hard cases" in order to achieve clarity, as he says, about "the kind of claim I am making." Dworkin, supra note 1, at 365.
22 . Compare Dworkin's view that once a participant in a practice begins speaking at all, "the law of the excluded middle is a necessary truth" that would be inconsistent for him to deny. Dworkin, A Matter of Principle, supra note 2, at 134. He takes the expression of the law of the excluded middle -- "p or ~p" -- to mean something that could be contradicted by what another expression ("~(p or ~p)") means. But here it can be asked, what would it look like for the negation of the law of the excluded middle to be true? For example, what would it look like for Ronald Dworkin to be neither alive nor not alive? What state of affairs is the proposition "Ronald Dworkin is neither alive nor not alive" even trying to describe for us? If the law of the excluded middle is a necessary truth, then it is not the kind of truth that has some imaginable and representable antithesis. Instead of thinking of the laws of logic under the aspect of "truth and falsity," they can be compared with rules of grammar that we use to construct propositions. It is the propositions we construct with the help of the laws that then have the possibility of being true or false. In the light of this comparison we no longer feel the overwhelming urge to call the laws themselves either true or false, because they themselves are not propositions that are trying to say something about the world. It appears that Dworkin fails to see the possibility of saying that the law of the excluded middle and the law of contradiction do not mean something that could be true or false, but are grammatical rules that "determine a meaning and are not answerable to any meaning that they could contradict." Wittgenstein, Philosophical Grammar, supra note 3, at 29. See Wittgenstein's Lectures: Cambridge, 1932-1935 71 (Alice Ambrose ed., 1979) ("The laws of logic, e.g., excluded middle and contradiction, are arbitrary. This statement is a bit repulsive but nevertheless true.").
23 . This is the view that his right answers thesis must be wrong because he cannot produce the kind of argument that would settle which answer is right in a particular hard case in such a manner that "no one who is rational can or will resist it." Dworkin, supra note 1, at 385 n.20; see also Dworkin, A Matter of Principle, supra note 2, at 137.
24 . Dworkin, A Matter of Principle, supra note 2, at 138, 141.
25 . Wittgenstein, supra note 4, at 34.
26 . Dworkin, supra note 1, at 361-62.
27 . Id. at 363.
28 . Id. at 369.
29 . Id. For his part, Rorty agrees that pragmatism is banal in its application to law, but he also says that "Dworkin's criticism of pragmatism amounts to little more than "gerrymandering the word "pragmatism" to mean crass instrumentalism%." Richard Rorty, The Banality of Pragmatism and the Poetry of Justice, in Pragmatism in Law and Society 89, 89 (Michael Brint & William Weaver eds., 1991).
30 . Dworkin, supra note 1, at 362.
31 . Id. at 380 (citing Wittgenstein).
32 . Id. at 383 n.5 ("Notice that I do not claim that lawyers all agree about which side is favored by the best arguments. (I could hardly claim that because a hard case is one in which lawyers do disagree.)").
33 . Or consider the case of double negation, a variation of which Wittgenstein discusses in Lectures on the Foundations of Mathematics, supra note 14, at 80-82. Suppose that A uses the symbol "not" in the sentence "The soup is not not delicious" to indicate to his host that it is mildly delicious (~~p = p), while B uses it in uttering the same sentence as an intensifier to indicate that the soup is very definitely not delicious (~~p = ~p). If our theory of meaning takes meaning to be represented by the use of the symbol (and only that), then after we give our description of the different uses of the symbol "~", Wittgenstein tells us, "it would be very queer to say that negation was a different thing in the two cases." Id. at 82.
34 . Let it be understood that I have been assuming this all along, and that I feel no overwhelming need to say that what Dworkin means by his report is true or false.
35 . This exploits a metaphor and images that Wittgenstein uses in The Brown Book. Wittgenstein, supra note 4, at 170.
36 . Dworkin, supra note 1, at 362.
37 . The foundation of Dworkin's concept of "law as integrity," for example, is the existence of rights and responsibilities which flow from past decisions (right answers), and his advocacy of an "interpretive, self-reflective attitude" in legal practitioners (including judges) presupposes that it is their job to seek right answers. See, e.g., Dworkin, Law's Empire, supra note 2, at 95-96, 225-75, 411-12. But here it can be seen that Dworkin requires the symbol "RA" to mean something--to have a Bedeutungsk?rper--that corresponds to the concept "rights and responsibilities [which] flow from past decisions," id. at 96, and that corresponds to what legal practitioners are supposed to be looking for when they look for "the best route to a better future, keeping the right faith with the past." Id. at 413.
38 . Ludwig Wittgenstein, Wiener Ausgabe, Band 1, at 177 (1994).
39 . The best example in American legal theory of this kind of person is Stanley Fish, who is one of Dworkin's most ardent critics. See, e.g., Stanley Fish, Almost Pragmatism: The Jurisprudence of Richard Posner, Richard Rorty, and Ronald Dworkin, in Pragmatism in Law and Society 47, 47-81 (Michael Brint & William Weaver eds., 1991). When Fish asserts that "[f]ormalist or literalist or "four corners" interpretation is not inadvisable . . . it is impossible," for example, he is not saying that there is a kind of four-corners interpretation that is imaginable and representable, but that is empirically impossible. Id. at 56. His sentence does not play the same kind of role in his article as the following sentence would, for example: "It is impossible for Ronald Dworkin to lift a 500-pound boulder over his head." We can imagine, and represent in words or pictures, what it would look like for Ronald Dworkin to lift a 500-pound boulder over his head, and it is this state of affairs that such a sentence would try to exclude as impossible. But anyone who is reasonably familiar with Fish's work knows that he is not saying that there is some describable practice that the words "four-corners interpretation" stand for, and that engaging in this practice is so very, very difficult for human beings to do that it comes down to being impossible in fact.
40 . Ludwig Wittgenstein, Tractatus Logico-Philosophicus ?ì 6.51, at 73, (D.F. Pears & B.F. McGuinness trans., 1961). Although nothing of consequence turns on it here, I agree with S.G. Shanker that Wittgenstein never wavered from the position he first announced in the Tractatus that philosophical scepticism is not false, but nonsensical, and that many of his critics and admirers alike have mistakenly read his life-long strategy of asking what seem to be maddeningly skeptical questions as covertly advancing a skeptical theory. S.G. Shanker, Wittgenstein and the Turning-Point in the Philosophy of Mathematics 2 (1987).
41 . See, e.g., Ludwig Wittgenstein: Cambridge Letters 47 (Brian McGuinness & G.H. von Wright eds., 1995), where Wittgenstein writes to Bertrand Russell, in 1913: "What I mean to say [by the word "bipolarity%] is that we only then understand a prop[osition] if we know both what would be the case if it was false and what if it was true."
42 . Grammatical rules in this sense are like stipulations concerning units of length: we can use a stipulated length (a bar we call "one meter," for example) to measure with, but it is hard to see what the point would be of saying that the bar we have chosen as our standard of measurement is itself one standard unit long. Consider Wittgenstein's remark about what was then the standard meter-bar in Paris: "There is one thing of which one can say neither that it is one metre long, nor that it is not one metre long, and that is the standard metre in Paris.--But this is, of course, not to ascribe any extraordinary property to it, but only to mark its peculiar role in the language-game of measuring with a metre-rule." Wittgenstein, supra note 5, ?ì 50, at 25e.
43 . Dworkin, Law's Empire, supra note 2, at viii.
44 . Dworkin, supra note 1, at 359.
45 . Id. at 360.
46 . Dworkin, Law's Empire, supra note 1, at 63.
47 . Glock defines Wittgenstein's concept of "internal relations" as "relations which could not fail to obtain, since they are given with or (partly) constitutive of the terms (objects or relata), such as white's being lighter than black." Hans-Johann Glock, A Wittgenstein Dictionary 189 (1996). He continues:
[T]here is no such thing as justifying or doubting an internal relation. Since the relation is (partly) constitutive of the relata, one cannot coherently deny that it obtains without ceasing to talk about those relata. Consequently, a skeptic cannot meaningfully deny that the relation obtains. At most, he could reject the practice which treats the two as internally related.
Id. at 191.
48 . See Ludwig Wittgenstein, Remarks on Colour 17e (G.E.M. Anscombe ed. and Linda McAlister & Margarete Sch?ttle trans., 1978).
49 . See Wittgenstein, supra note 40, ?ì 4.123, at 27 ("A property is internal if it is unthinkable that its object should not possess it. (This shade of blue and that one stand, eo ipso, in the internal relation of lighter to darker. It is unthinkable that these two objects should not stand in this relation.)").
50 . Wittgenstein, supra note 5, ?ì 241, at 88e.
51 . Dworkin, supra note 1, at 380.
52 . Id.
53 . Wittgenstein's Lectures: Cambridge, 1932-1935, supra note 22, at 73. The particular case under discussion is the use of the expression "I like it and don-t like it" to indicate ambivalence. Wittgenstein tells his students that it would be a mistake to say that this expression is being used in a "different way" than the law of contradiction is used in logic, because one cannot hope to describe the law of contradiction independently of its application. Once the law's application has been described, whatever meaning there is in the law is already included in the description, and does not come out as something extra that takes the form of the "way" the law of contradiction is used.
54 . Dworkin, supra note 1, at 381.
55 . Id. at 382.
56 . See, e.g., Fish, supra note 39, at 80; Stanley Fish, Doing What Comes Naturally: Change, Rhetoric, and the Practice of Theory in Literary and Legal Studies 372-98 (1989).