|Top of Form
by Anthony D’Amato*, 85 Nw. U. L. Rev. 148-189 (Fall, 1990)
Abstract: If, as a result of taking Indeterminacy seriously, we revolutionize the way we teach law and the way we select judges, then we will also revolutionize the way cases are litigated (because the new judges will expect to hear a different kind of argumentation) and the way people order their lives in anticipation of the way their disputes will be decided by these new judges.
Tags: Indeterminacy, Pragmatic indeterminacy, Formalism, Judicial decisions, Benson, Hegland, Kress
[pg148]** The Indeterminacy debate—to which Professors Robert Benson, Kenney Hegland, and Ken Kress have made important contributions as well as being so kind as to respond to four of my “Aspects of Deconstruction” essays—is, according to Professor Steven Winter, the key issue in legal scholarship today. [FN1] The debate has intensified in the United States and will spread to other countries in the near future.[FN2] Professor Winter observes that “the legal academy is experiencing a state of epistemological crisis.”[FN3] We are indeed in the midst of an immense paradigm shift in the way we think about law, and no one knows where it may end. Many scholars are confused, disheartened, and anxious to retreat to a bunker mentality of vague and benign Formalism.[FN4] Yet if scholars are going to catch up to what experienced practitioners know—that you can win cases even though the law is clearly against you, and lose cases even though the law is overwhelmingly on your side [FN5]—we should boldly confront [pg149] the possibility of Indeterminacy and its implications for the way we teach and write about law.
In this essay I defend my version of Pragmatic Indeterminacy in response to the main points of my critics. I am indebted to Professors Benson, Hegland, and Kress for participating in the present dialogue and giving me the opportunity to continue it.
I. Rational Argument Is Alive and Well
There comes a point in every law school course when a student asks, “Can you tell us what you are looking for on the exam?” I realize the empowerment of that moment: whatever I say, I will get. If I say, “Good penmanship!” I will get flawlessly scripted exam papers. What I actually say is: “Persuasion—the best exam answer is one that best persuades me. Take whatever position you like, the important thing is to write the most convincing argument you can for the position you choose.”
I try to practice what I teach. I believe in the power of rational argument to persuade decisionmakers. I have never suggested that there is no forcefulness in legal rhetoric. Yet Professor Hegland says that “[t]o deny that there are ‘easy cases' is to deny the possibility of rational argument. . . . I have no real problem with this analysis except that it suggests that we are insane.”[FN6] Professor Benson contends that all deconstructionist argument, including my own, “is best explained not as an attempt to change the world, but as an act of intellectual self-gratification by the participants.”[FN7] And Professor Kress, charging me with taking the position that language lacks persuasive value, suggests that I write articles “in order to get a good raise next year.”[FN8] No one can accuse my critics of pulling their punches.
Their basic position assumes the following argument: (1) legal scholarship consists of rational discourse; (2) rational discourse assumes that one argument can objectively be better than another; (3) when one argument is better than another, it should win; (4) rational discourse is a better argument than Indeterminacy-deconstructionism; (5) rational discourse should thus win over Indeterminacy; (6) therefore, legal scholarship [pg150] should henceforth consist solely of rationalist discourse. Of course, this line of reasoning ends where it began, completing a harmless circle.
Yet even if logically tautological, my critics' positions may be psychologically necessary. It is hard for some people to accept the possibility that the rational discourse they have come to know and love—and that has garnered them academic and remunerative honors because of their superior verbal ability to engage in it—may have little or nothing to do with judges' decisions in particular cases. If in addition they are convinced, as Professor Kress appears to be, [FN9] that the legitimacy of the legal enterprise is jeopardized when the question of indeterminacy is raised, they are apt to circle their wagons.
The problem in today's legal discourse is impatience. Like judges, many scholars think that if someone has anything at all useful to say, it can be said in twenty minutes. And so Indeterminists resort to provocative statements, outrageous hypotheticals, weird counterfactuals—anything that within twenty minutes might shake loose the listener's moorings of objective meaning. We say, paraphrasing Lewis Carroll, that it is not so much that our words have meanings as that our words have us. But these gibes hardly ever win the day; rather, they encourage some “serious” mainstream scholars either to sneer and walk away, or to accuse us of being nihilists and irrationalists.
Some Formalist scholars, like Professors Hegland and Kress, have a gibe of their own—actually, an adaptation from the oldest “refutation” of skepticism in philosophy. The classic philosopher says to the Sophist: “If you are truly skeptical as you say you are, why are you not skeptical of your own claim of skepticism?” My critics in effect ask me: “If you don't believe that rational argument constrains judges, why are you resorting to rational argument to convince us?”[FN10] My answer is: “What would you have me resort to instead—irrational argument? [FN11] The only [pg151] way I have any chance of convincing you is to argue in a way that you think is rational, even if secretly I believe that my ‘rational’ may differ from your ‘rational.” [FN12]
My general reply to this riposte to skepticism is that a skeptic does not have to disbelieve everything including skepticism itself. It suffices simply to be unsure about what to be sure about, including skepticism itself.[FN13] If, as I shall argue in a moment, a skeptic does not know where, if anywhere, a line can be drawn between “easy” cases and “hard” ones, then every case is potentially indeterminate. That conclusion is enough to constitute skepticism, even though it falls short of claiming absolutely that every case is “hard.”
The proof for a legal Formalist who sincerely wants to understand the possibility that words do not have determinably precise meanings [FN14] is to read the exciting philosophical and artificial-intelligence literature that has enabled twentieth-century philosophical thought to overcome the conceit of the Platonic view that words have determinate meanings.[FN15] The philosophical studies that have helped me the most are indicated in [pg152] an extended footnote. [FN16] The real utility of reading this literature is that it [pg153] can open our own minds to the possibility that we are prisoners of the words we use. To escape from that prison we have to look at words and concepts in a new way. We must somehow get “outside” the words we utter in order to analyze them. Getting outside our words requires us to overcome our own minds' psychological temptation to smuggle in the very notions of meaning that we are trying to discover. What Wittgenstein said of philosophy is equally true of jurisprudence: “[p]hilosophy is a battle against the bewitchment of our intelligence by means of language.”[FN17]
My “long” answer to Professor Hegland is that, before pronouncing my views “insane,” he might do well to spend a few years perusing the development of twentieth-century linguistic philosophy. He will not necessarily find a cure for insanity. Rather, the process of working through [pg154] the literature is as much psychological as logical, showing us how to get outside our minds with the assistance of philosophers who have tried the same thing in their attempt to come to grips with words. As for Professor Benson's view that this was all said previously by the American Legal Realists, and that we are now only going through a rerun, I would concede that the Legal Realists gave us a taste of something profound. [FN18] However, they wrote with the disadvantage of predating the critical supporting contributions of Wittgenstein, Goodman, Quine, Kripke, and the artificial-intelligence researchers. This time around, I believe, Indeterminacy is here to stay.[FN19]
To respond directly to Professor Hegland's argument that the best “rational” argument must win, let me engage in the following thought experiment. Suppose we have a lawsuit with the ground rule that the decision will be awarded to the side which presents the most rational argument. Surely Professor Hegland cannot object to this starting point, because it is his own. And surely he cannot object if we further assume [pg155] that he is the designated attorney for one of the parties, and that he will devise a brilliant, thorough, scholarly, exhaustive, scintillating legal argument that all observers agree constitutes the epitome of rationality. I still maintain that a judge can decide the case against him. How? The judge says, “Professor Hegland, your argument was undoubtedly one of the most supremely rational arguments I ever heard in my life. But there was one problem: it was formally rational, whereas your opponent gave a more substantively rational argument. To be sure, your opponent stumbled and faltered, miscited some cases, and played loose with some statutes, and if I were grading her performance I would give her a C and you an A. But her position was more substantively rational than yours, and since I am interested in substantive justice and not in grading lawyers' rhetoric, I must award the decision to her.”
Of course, once I make this kind of argument, a critic has two surefire comebacks. The critic either says, “I had already thought of the formalist-substantive distinction and by not mentioning it I meant to exclude it by implication,” or “All right, let's change the ground rules—now the best substantive rational argument will win the lawsuit, and we assume that I make the best substantive rational argument.” This is like my example of the person who was referring to her finger and not pointing at a pelican;[FN20] a critic can always say that he meant to exclude all outrageous cases, or he can amend his proposition so that we exclude references to fingers. All I can do is to ask that if we make whatever amendments in the rules the critic wants, will the critic now be satisfied? If the critic says yes, then I move to a subsidiary thought experiment:
We now assume that Professor Hegland as attorney has made the best substantive rational argument. Our judge then replies, “Although you have made the best substantive rational argument from the point of view of goal-rationality, your opponent has made the best substantive rational argument from the point of view of means-rationality. Since law must concern itself with means, and cannot pick and choose among the different values that people assign to goals, I must award the decision to her.”
Let us again defer to Professor Hegland's amendatory powers, and assume that he replaces his previous argument with the best substantively means-rational argument. Does he stop here? Does he now claim that a judge would have to be wrong to deny him the victory? If so, our judge says: “Professor Hegland, although you have now made the best substantively means-rational argument, your argument is only best within the framework of Kaldor-Hicks optimal rationality. Your opponent's substantively rational argument exhibits a Pareto-optimal means-rationality. As between two substantive means-rational arguments, the [pg156] Pareto approach overcomes an otherwise better argument based upon the approach of Kaldor and Hicks. Sorry, you lose.”
One last time? Professor Hegland now presumably presents the best substantively means-rational Pareto-optimal argument. Again he could lose. The judge could rule that although his argument was more means-rational and Pareto-optimal than that of his opponent, his opponent's argument was closer to a Peircean instrumental rationality than his own argument, which was closer to a Deweyan pragmatic rationality. As between Peirce and Dewey, when Pareto only slightly eclipses Kaldor-Hicks, Peirce wins over Dewey.
The point, of course, is that this sort of thing can go on forever. No initial assumption of rationality can possibly constrain the outcomes of particular cases. The deconstructionist's best rejoinder to critics such as Professors Hegland and Kress is simply to accept all their starting points (such as “rationality,” “the power of persuasion,” and “reasonableness”), and argue within these categories for the opposite result. This is the approach I took in my sci-fi example of the under-aged President that bothered Professor Hegland.[FN21] I accepted his premises and then simply changed the context. Then, assuming he would object to the changed context, I shifted back to the present and gave a presently acceptable constitutional argument in support of the Presidency of eighteen-year-olds.[FN22] I tried to do the same with any posited theory in a previous article, contending that even a court's prior commitment to a particular theory leaves the court free to decide the case for either side.[FN23]
Bottom line, do I believe in the power of rational argument? Of course I do, for a very good reason: it often seems to work. Not that it's meaningful; not that it should work; only that pragmatically it seems to get results. More precisely, my position is like that of the Georgia deacon who was asked if he believed in baptism by total immersion. He replied, “Believe in it? Hell, I've seen it done!”
II. Why Dissent? Why Not?
Several scholars have counted the frequency of dissenting opinions in appellate court cases and found that the rate is rather low—in one [pg157] study, only four percent of the reported cases had dissenting opinions.[FN24] Some scholars have seized upon this statistic and said that it proves that only a minority of cases are indeterminate. I disagree. As Professor Kress fairly summarizes my position, a low appellate dissent rate is more likely to be the consequence of indeterminacy than determinacy.[FN25]
But then, because Professor Kress regards Indeterminacy as just another “theory,” he misstates my position. He argues that if I am correct that deconstructionism causes a decline in the dissent rate, then deconstructionist theory will itself have persuaded judges to alter their behavior, and hence at least one theory (i.e., the deconstructionist theory) can make a difference. Ingenious though this argument is on its own terms, it changes what I actually said. In the first place, I did not claim that deconstructionism is a “theory” in the sense that Professor Kress uses, which is that a theory is an explanation for events. In my view, deconstructionism is, among other things, a way of looking at the world that challenges the psychological need to explain events by theories. Secondly, I said that the spread of the deconstructionist outlook may cause some judges to become defensive about the way they present themselves to the outside world. Judges may decide that there is no point in arming critics of their decisions with well-reasoned dissenting opinions. If critics are going to attack the legitimacy of judicial decisionmaking by claiming that the case could have gone either way under existing law (judges may say among themselves), then why provide those critics with the ammunition of persuasive dissenting opinions? Hence institutional pressures for dispensing with dissenting opinions may increase in an atmosphere where deconstructionists question the competence of rules of law to decide specific cases. None of this is to say that deconstructionism is a theory that causes judges to dissent less frequently. Judges need not have a conviction that deconstruction is correct if they want to reduce their rate of dissent, but rather they need only to desire to accommodate a readership that (rightly or wrongly) has become deconstructionist-minded about the rationalizations they read in judicial opinions.
Professor Kress further misinterprets me when he says that I attribute no persuasive value to dissenting opinions in terms of their impact upon future judicial decisionmaking. I did not say they lack persuasive value, any more than I claimed that rational argument lacks persuasive value. I concede that I overstated my case when I said that there was “no point in dissenting” when I should have put it less absolutely by saying that no one can be sure that there is any point in dissenting, for it might just as well work against as for the dissenter.[FN26]
[pg158] Consider a hypothetical Supreme Court case in which the Court states, “We are now persuaded that Justice Holmes, dissenting, was right and the majority was wrong. Therefore we will adopt the reasoning of Holmes's famous dissent.” Does this prove that Holmes was vindicated, that his dissent persuaded a future Court? I maintain that there are millions of things that can have persuasive value to judges, and one of them might or might not be Holmes's dissent. The present Court just as easily could have said, “Although Holmes wrote a persuasive dissent, we are bound by stare decisis to follow the majority.” This shows that the “persuasive” value of Holmes's dissent turns on whether the present Court prefers to say it has “persuasive” value rather than saying that the theory of stare decisis has the greater persuasive value. Practicing lawyers are loathe to cite a dissenting opinion for fear of giving the other side a cheap rebuttal: “Our opponents have resorted to citing dissents, whereas we rely on holdings!” In short, the fact that favorable reasoning can be found in a dissenting opinion does not mean that it has “persuasive” value (unless we look for persuasiveness in the cogency of the reasoning itself—a point I shall return to at the end of this section).
So far we have only begun the Indeterminist account of dissenting opinions. A second point is that although the Court says that the Holmes dissent had persuasive value, the Justices may have made up their minds independently of Holmes and then cited the Holmes dissent as acceptable window dressing. Again, we cannot know what the Justices thought; the truth value of their claim that they were persuaded by Holmes must be indeterminate.[FN27]
Third, since all cases are inevitably distinguishable on their facts from any and all precedents, neither the majority opinion nor the dissent can be “persuasive” in the sense of intellectually constraining. To be sure, a great deal of legal scholarship has been expended on arguing whether case A is really distinguishable from case B, and qualifying terms will often be used to the effect that the distinction (if any) must be “non-trivial,” “relevant,” “significant,” “germane,” “pertinent to the theory of decision,” and so forth. But the real world of facts does not present itself to us pre-labelled by such terms. Rather, those terms are [pg159] the conclusions we reach—the rationalizations we verbalize—when we have decided that a difference in fact between two cases is sufficient for us to “distinguish” the cases on the basis of that difference. If making that distinction seems to entail changing or modifying the decisional rule,[FN28] then our ultimate refuge is to simply say that such are the workings of the common law.
Finally, my charge that dissenting opinions are futile is leveled against dissents that claim the majority misinterpreted the law. Since Pragmatic Indeterminacy contends that law cannot constrain individual case results, it follows that a Pragmatic Indeterminist would see no point in critizing a majority's opinion on the ground that it is contrary to law.[FN29]
But occasional dissenting opinions point out facts in the case that the majority either overlooked, misinterpreted, or suppressed—facts that, when fully acknowledged, indicate that the decision was unjust. Such dissents are indeed valuable as fulfilling a “whistleblower” function that can help keep courts honest.
I would not conclude, however, that dissenting opinions have no persuasive value whatsoever for subsequent courts. Rather, my claim is that millions of things, including a dissenting opinion here and there, have persuasive value. Many things have far more persuasive value. If, during childhood, a Supreme Court Justice witnessed his father beating his mother, and was deeply aggrieved by what he saw, this experience is likely to have a major impact upon the Justice's “reasoning processes” in a case where a battered wife is claiming a deprivation of her constitutional rights. Or if a Justice as a teenager was unfairly rounded up in a police arrest, and beaten in the police station, such an experience will undoubtedly be more “persuasive” in the Justice's attitude toward citizens claiming police brutality than any number of dissenting opinions pointing out that the police do not always tell the truth.
For an advocate to be persuasive, it is very important to know the attitudes and views of the judges. But the difficulty of discovering them is commensurate with their importance. Judges normally do not tell us their personal feelings and attitudes, although some trial judges—in the relatively less constraining oral proceedings of a trial—are often quite candid in referring to their own experiences as support for the numerous [pg160] evidentiary decisions they make. Appellate judges in their written opinions tend to write impersonally and cite “the law” rather than their personal convictions for their decisions. There are at least two reasons for the impersonality of written opinions. First, the judge's readership expects an impersonal, legalistic explanation of why the case had to be decided a certain way. Candid references to the judge's own idiosyncratic experiences could anger a losing litigant who demands impartiality and not whim.[FN30] Second, it is intellectually difficult for anyone, including judges, to fine-tune introspectivity to the point of recognizing personal motivation.[FN31]
When I begin an oral argument to a three-judge panel, I see before me three robed citizens who appear similar to each other in many respects, especially sartorially. But inside they have wildly divergent life experiences, and react differently to each word that I utter. Although my words may produce a surface conformity of judicial behavior, the meanings they evoke in each judges can be different. To take what should be the least controversial example, I start by stating my name. This will produce a certain degree of overt conformity throughout the trial; for instance, if the judges refer to me by name when they ask questions, they will produce sounds like “Mr. D'Amato” or “Professor D'Amato.” This surface conformity, however, masks potential whirlwinds of prejudicial thoughts. For all I know, my name conjures up in the first judge's mind reminiscences of the blue skies of Rome, the operas of Verdi and Puccini, and delicious Italian food. For the second judge, my name might trigger images of Chicago gangsters, Mafioso massacres on St. Valentine's day, and Marlon Brando in the role of Godfather. The third judge may silently be condemning me for the affectation of spelling my name with an apostrophe when I should have had it Americanized.
As I proceed with my argument, the judges' images may colorize the next few sentences I utter, until a word or phrase—or an inflection or bit of body language—in one of those sentences sends a judge down an unwanted byroad (perhaps the scene for the first judge is still Italy, but she now envisions Mussolini during the Second World War haranguing a [pg161] crowd with demagogic oratory).[FN32] Indeed, whatever words I utter—nouns, adjectives, legal phrases, names of other judges, case citations—each utterance will evoke for each judge the different contexts in which that judge has learned the word over time. [FN33] I might cite a case and what runs through one judge's mind is that he had to state that case when he was a student in law school and he was embarrassed and ever afterwards hated that case. And so I will have engendered a most unfortunate emotional response simply by referring to a case by its name. As Quine so well put it:
Different persons growing up in the same language are like different bushes trimmed and trained to take the shape of identical elephants. The anatomical details of twigs and branches will fulfill the elephantine form differently from bush to bush, but the overall outward results are alike.[FN34]
Thus I suggest that when academicians such as Professor Kress confidently assert that a dissenting opinion can persuade a future court, we should take it with a grain of salt. Of course a dissenting opinion, like anything else in the universe, might or might not persuade a future court. But that does not invest the dissenting opinion with any non-random claim to rhetorical power.
Yet the academy may issue a final rejoinder—one that has the particular echo of the classroom: “Not every dissenting opinion will persuade future courts; only dissenting opinions that are in themselves truly well reasoned, logical, beautiful, argumentatively tight—in short, only persuasive dissenting opinions—will have persuasive value for future courts.” If that is the academy's fail safe position, I cheerfully concede it on the ground that if you've seen one tautology, you've seen them all.
III. There Are No Easy Cases
Professor Kress asserts that he understands Indeterminacy because, after all, he is one of the few scholars who has taken the trouble to define it.[FN35] I find this claim unintentionally humorous, for if Professor Kress [pg162] actually thinks that Indeterminacy can be defined, then he cannot understand it! [FN36] Professor Kress calls his own position “moderate indeterminacy,” which he locates somewhere between determinacy and indeterminacy.[FN37] Although he insists that most cases are clearly determinate,[FN38 ]he does not give us an example of a clearly determinate case. If there are so many determinate cases to choose from, why does he not identify at least one? His failure to cite a single example may amount to a claim of inability to determine whether any specific, given case falls within the determinate category. But if that is his position, then he must concede that every case is indeterminate, because not knowing whether or not a given case is indeterminate is another way of saying that all cases are indeterminate.
One scholar who has tried to find a determinate case is Frederick Schauer.[FN39] Professor Schauer, who views an “easy case” as determinate and a “hard” case as indeterminate, has tried to find a line separating the two types of cases. If a line between the two can be drawn, then he will have vindicated Professor Kress's “moderate indeterminacy” position. The importance of finding such a line for both Professor Kress and me is critical. Professor Kress bases his entire refutation of Indeterminacy on the “pervasiveness” of easy cases as distinguished from the allegedly few controversial ones,[FN40] while I would concede defeat for Pragmatic Indeterminacy if such a line could be drawn. Much is at stake as we follow Professor Schauer in his valiant attempt to draw a line between easy and hard cases.
Professor Schauer starts his quest by attempting to draw a line at the top rung of the legal ladder—the Supreme Court. If he can find an easy case there, then many more easy cases abound below. But he soon admits defeat, explaining that the nature of the Court renders it unlikely that an easy case will ever get there. He concludes flatly that “there are no easy cases in the Supreme Court.”[FN41]
The second downward rung on Schauer's Ladder consists of appellate court cases, both state and federal. But Professor Schauer soon finds [pg163] himself unable to cite a single example of an easy case at this level. He concludes, “I will concede that there are few, if any, easy cases in any appellate court.”[FN42]
Professor Schauer's concession of all appellate cases to the Indeterminacy camp is sound. In any appellate case the parties have submitted briefs containing plausible legal arguments for their own positions. [FN43] Perhaps there exists an appellate court case in which one of the parties submitted a wholly frivolous brief, [FN44] but even then, a moderate amount of ingenuity could supply a plausible legal argument for that party. For example, one could contend that the precedents cited by the opposite side are distinguishable, that an applicable statute is unconstitutional, or that equity and fairness demand that the spirit and not the letter of the law be followed (or conversely that the letter of the law be followed even though the other side's position is morally attractive).[FN45]
Thus Professor Schauer must step down to the third rung: the trial level. But no easy case is anywhere to be seen in trial court. He writes:
In terms of cases that reach trial and decision, there are probably very few [easy cases]. Indeed, easy cases are most likely less prevalent in trial courts than in appellate courts. The appellate process narrows the issues, but, since trials take place prior to this narrowing, they raise a substantially larger number of factual and legal issues. And as the number of issues increases, the potential justifications for making a decision one way or another [pg164] also increase, thus making it more difficult to designate as “easy” any final trial court decision. . . . To Jerome Frank, for example, the injection of contested factual issues at any point in the trial process was sufficient to make uncertain the results in almost every case that in some way wound up in court.”[FN46]
Since the top three rungs of Schauer's Ladder comprise all contested cases, and since law students are fed an almost exclusive diet of reports and summaries of reports of such contested cases, [FN47] the conclusion of a well-known Formalist such as Professor Schauer is notable:
I will concede that there are few if any easy cases anywhere in the litigation process, and that any case filed in a court is capable of being decided one way or another relatively unconstrained by precedent or written law.[FN48]
In short, a Formalist has conceded that every case on the books—every single case that law students study for three years as well as every single case that they will ever litigate in practice—is not an easy case.
Many American law professors today would routinely agree with Professor Schauer's reasoning and conclusion, yet go about their business as if the point had never been made. The explanation may lie in the possibility that legal scholarship today is characterized by massive cognitive dissonance (which would not be surprising given Professor Winter's description of the present paradigm crisis).[FN49] Each law professor is quite aware of the plasticity of law, and yet writes about law as if it is determinate.[pg165] By talking as if law is determinate, a professor adds an aura of authority to her teaching and writing. She may not necessarily be dissembling when she writes articles criticizing cases on the grounds of correctness or validity; rather, as an author, she may have simply suspended her disbelief in Determinism. In the classroom she can switch back to Indeterminacy with ease; she can argue not only that the result in any assigned case could have come out the other way, but also demonstrate to any student who offers a hypothetical variation that his example contains implicit assumptions each of which can be deconstructed, rearranged, and made to produce the opposite result. Perhaps she rationalizes her formalist authorial behavior on the ground that publication in law journals is a highly constricted form of professional activity that will be judged by certain prevailing standards, which at the present time include mainstream Formalism. What she believes may not necessarily be the same as what she feels obliged to write. An example of this rationalization is the very article by Professor Schauer that I have been recapitulating in this section. Although Professor Schauer ends his article with the concession that there are no litigated or reported cases that are “easy,” earlier in the same article he characterizes the legal realist view of decisional indeterminacy as follows: “To take this view as an accurate generalization of all or even most judicial decisions, however, seems at least erroneous and at times preposterous.”[FN50] In brief, Professor Schauer early in his article labels preposterous his own findings later in the same article—a textbook case of cognitive dissonance.
Psychological denial of the truth of Indeterminacy clearly has functional utility in professional life. Determinacy is self-reinforcing, whereas Indeterminacy is dirty linen that is dangerous to expose in public. Even when a scholar such as Professor Kress is writing about Indeterminacy, his fears of illegitimacy appear to compel him to insist on the pervasiveness of easy cases although he does not produce an example of a single one.[FN51] At the other extreme from psychological denial of the truth of Indeterminacy is what might be called psychological acceptance. This acceptance is a convenient way of saying, “count me in” while going about one's business in the pre-Indeterminacy fashion. Thus Judge Richard Posner, in his most recent book, concludes “ today we are all skeptics.”[FN52] Interestingly, he does not aim this newfound skepticism at its most appropriate target—his own previous, Formalistic, economic-analysis-of-law determinism.[FN53] The Coasean cat may have departed, but its grin remains.
[pg166] We have not completed our descent to the bottom of Schauer's Ladder, for there is at least one more rung below the trial level where an easy case might be found: the unreported, unlitigated case. Here, at last, Professor Schauer purports to find a line between hard and easy cases. Easy cases, he says with an implied sigh of relief, are the cases that are never litigated. Since an unlitigated case is one that a potential plaintiff decides would be a clear loser in court, Schauer argues, it must be “easy,” and that is why the party never takes it to court. And, since unlitigated cases are more numerous than litigated ones, Professor Schauer and his followers are able to assert that much of the law—the law represented by cases that never get litigated—is made up of easy, determinate cases.
But it doesn't work. For one thing, there are many reasons people choose not to litigate their case other than a calculation that they would lose on the merits. They may not have the money to pay lawyers or investigators.[FN54] If they have the money, their expected recovery may not be worth the expenditure. [FN55] If their cases are too “new,” they may be discouraged from litigating by unimaginative attorneys. Additionally, some of them may simply be wrong in their assessment that they would lose if they took their cases to court. [FN56] Finally, if Professor Schauer were correct that unlitigated cases are easy cases, all we would have to do to transform any unlitigated case he names into a hard case is to offer to one of the parties to litigate it for free. But surely the Formalists cannot [pg167] plausibly assert that a decision to litigate any unlitigated easy case they might choose would move it over the line into the domain of hard cases.[F57]
If any easy case exists at all, it can only be found within the category of unlitigated cases—at yet a lower rung on the ladder. Professor Schauer has not descended that far, but to conclude the present analysis we might well ask whether he could. Is there a line between easy unlitigated cases and hard unlitigated cases? The only possible distinction would have to be based on the type or content of the cases. But what type? What content? How could they be characterized?
Some of my colleagues are put off by such questions. Of course, they say, there are millions of easy unlitigated cases. Look in your wallet and take out a dollar bill. There is a statement on each bill saying, “This note is legal tender for all debts, public and private.” Cash is used as legal tender in millions of transactions each hour of each day. Surely these are all “easy cases” of law-application.
My reply: Have you ever walked into a Federal Express office and attempted to send a package by paying cash? Your cash will be refused; you must either present a valid credit card or already have an account with Federal Express. What happened to “legal tender”? Is the Federal Express case an “easy” case (easy because you have no right to force anyone to accept cash) or “hard” (hard because you might very well have a right to rely on the wording on the currency and insist that any publicly offered service be payable in cash)? How do we know? Must we wait until such a case is litigated? The problem is that such a case might never be litigated. Anyone in a hurry to send an overnight Federal Express package would not have the time to get an injunction against Federal Express ordering it to accept cash. So we have a potentially never-to-be-litigated case involving U.S. currency that is neither clearly easy nor clearly hard.
One reply to my Federal Express example is that it is unusual, weird, exceptional. Professor Hegland might rejoin: “Cash is always good as cash except for Federal Express and other exceptional cases.” But if we exclude exceptional cases, we are left with a tautology: cash is good as cash except for exceptions. Such a rule tells us nothing about law or the real world; it is purely an abstract logical construct. Recall the replies my critics made to my examples of the Underaged President[FN58] or the Pelican.[FN59] Once I've identified an exception, they say, “Well, it's [pg168] an exception, so we'll exclude it.” The more exceptions I identify, the more happy they will be to exclude them all. The problem in real life is that no one can specify the exceptions in advance[FN60]—and that, in a nutshell, is a major reason why law is indeterminate.
But is the Federal Express case really unusual? The real world has an uncanny ability to complexify and degrade legal rules.[FN61] We are beginning to see many diverse examples of cash not being accepted as legal tender. Major automobile rental companies are increasingly rejecting cash and insisting on credit cards; major hotel chains are starting a similar policy. The policy is not irrational: the customer may damage or destroy the automobile or hotel room and then run away. A credit card on file is some protection against these depradations, whereas an advance cash deposit by the customer may be insufficient to cover the damage (or if it is sufficient, there is nevertheless a temptation for an employee of the rental company or hotel to abscond with the cash). Other examples include convenience grocery stores with signs saying that they will not accept large-denomination bills, taxicabs saying that drivers cannot give change in excess of five or ten dollars, buses that refuse to give change, the New York subway system that at current writing is refusing to accept pennies, and even underworld refusals to take suitcases of hundred dollar bills in payment because spending them would attract attention.
Even in ordinary situations where cash is routinely accepted, I would assert to any colleague that he cannot tell me in advance for certain that cash will be accepted the next time it is tendered. For example, I ask a colleague to specify for me an easy case of cash being accepted as legal tender. She says, “This afternoon I will go shopping for groceries. I will pay cash, and it will be cheerfully and legally accepted.” But what is there to preclude the possibility that this very afternoon she will walk into the accustomed grocery store and see a sign saying, “Only checks and credit cards accepted”? She protests to her grocer; he says, “Sorry, there have been so many robberies lately that we have decided not to take in any cash.” Since my colleague is carrying only cash, she may well have to walk home without groceries. She may be outraged. She may threaten to sue. She may have a lawsuit alleging implied promissory estoppel, but it would not be an “easy case.”
Suppose instead she attacks my question as unfairly asking for a prediction instead of a retrodiction. Suppose she says, “I went shopping for groceries yesterday, and paid for them in cash, and so that was an easy case.” But my reply is that in fact it was not a “case” at all.[FN62] No controversy arose between her and her grocer. The idea of “case” or [pg169] “controversy” would be vacuous if we called every event and every transaction that happens in the world a “case.” Rather, a “case” is something where people go to enough trouble to make opposing claims against each other. When “law” is invoked to “resolve” such a claim-conflict, then we can properly call it a “case.” When no dispute arises, we can never know whether, if a dispute had arisen, what kind of case it would have been. (If one objects that the case was “easy” precisely because no dispute arose, the objection is tautological: we mean by a “case” a situation where a dispute arises.)
If Professor Hegland was shopping at my hypothetical grocery store, he might be tempted to amend his proposition: “All unlitigated cases involving the use of cash as legal tender are easy cases, except for those cases in which someone for a good reason refuses to accept cash as legal tender.” If Formalists want such a rule, I bequeath it to them. It either shifts attention to the phrase “for a good reason,” which is indeterminate, [FN63] or it is a tautology (similar to “except for exceptional cases”).
I put it to my critics to identify their own content-distinctions between easy and hard cases. I am confident that however they define it, a case can be invented that clearly fits their specified content for an easy case and yet is not, in fact, easy.
There is the exception of tautologies. A Formalist might specify, for example, “a case in which the plaintiff has no cause of action.” [FN64] Such a case, they may claim, is “easy.” I would reply with two arguments. First, if you start out by equating “no cause of action” with “zero possibility of winning a lawsuit,” then of course it is an “easy case” by definition. But it is only an easy case in a tautological world and not in the real world. It is only an easy case by the operation of language: if we assume X, and say that “X = plaintiff loses,” where X can take on any value including “no cause of action,” then it follows as a matter of language that the plaintiff must lose. Secondly, suppose we equate “no cause of action” with the empirical observation that, under existing law, there is no precedent to suggest that this plaintiff has a cause of action. I assert that it does not follow that the plaintiff must lose, for the obvious reason that new causes of action from time to time win their day in court. “Today's frivolity,” says Professor Risinger, “may be tomorrow's law. . . .” [FN65] If we start with the limited number of actions recognized in [pg170] the eleventh century, we will find many subsequent cases that created new causes of action—enough easily to invalidate the proposition that a plaintiff with no “cause of action” must lose.
An example of another kind of tautology is: “Suppose I simply refuse to pay the grocer for the groceries; doesn't he have an easy case against me for the money?” If the speaker is saying that she recognizes a legal obligation to pay but is refusing to pay just to prove there is such a thing as an “easy case,” then it is an academic or logical exercise and not a real-world case. She does not really have a dispute with her grocer; instead she is having a dispute with me. On the other hand, if she refuses to pay the grocer because she does not recognize a legal obligation to pay, we would then ask her to state her reasons. She might say a number of things: (a) I have a credit balance at this store; (b) I have a claim against this grocer and I am resorting to self-help; (c) my children are starving and I have no money; (d) I already paid for these groceries at another store; (e) I have been declared legally insane. None of these are “easy” disputes, for we can readily imagine judges who might rule either way on these claims and furnish plausible reasons for their decisions (a job which need be no harder than adopting the winning side's brief).
My general position is that we can never know whether any actual case is easy. [FN66] Nothing in law[FN67] is determinate in the sense that the law constrains a judge to come out a particular way with respect to a particular fact situation where there are opposing parties each claiming a right to the judge's decision.[FN68] Yet it would be unfair to go to the other extreme and accuse Indeterminists of saying that law should therefore be ignored. For the Indeterminist, “law” assists a judge in locating similar cases and thereby helps the judge dispense societal justice. Thus the words of the law serve a retrieval function. They help call up situations (judicial cases as well as hypothetical events addressed by statutes) that may be similar to the present case. Once those precedential situations are called up, the judge examines the decisions reached in those cases and [pg171] decides whether justice compels a similar result in the present case. Although positivism errs in saying that law-words constrain judicial decisions, law-words do play an important heuristic role in initially identifying previous decisions so that they may be compared with present situations.[FN69]
For the Pragmatic Indeterminist, judges do not behave randomly or unpredictably over the long run of cases; far from it. As I will contend later in this essay, lawyers can predict how judges will probably decide a given case or line of cases. If lawyers did not have such an ability, we would have chaos, not law. This general predictability, I shall argue, is expectable in any system that is fairly characterized as legal. Thus, when a critic charges Indeterminists with saying that law or language does not matter, the critic indicates she is acquainted only with superficial criticisms of the Indeterminacy thesis, and not with Indeterminacy itself. Legal writing of all kinds—theorizing, analyzing, articles in law reviews including this one—may have a long-term impact on the direction law takes and on the attitudes that judges and legislators have about the law. Pragmatic Indeterminacy simply doubts that anything that is written can point to the correct answer in the next case. This includes statutes,[FN70] precedents, rules, regulations, and law-words generally.
IV. Cracks in the Formalist Edifice
A Formalist regards at least some part, and perhaps all, of law as a logically precise system. Statutes, rules, and precedents are premises, and a legal conclusion is a logical deduction from the applicable premises. The law-words that make up this system form the decisional basis for judges, and so long as the system itself is coherent, these law words constrain the judge.[FN71] But one thing has to be assumed in order to make Formalism possible—a fundamental postulate concerning the nature of language. The founders of legal positivism failed to discover it in the nineteenth century, [FN72] but in 1958 H.L.A. Hart found it: every word has a [pg172] determinate core meaning that is not reasonably disputable.[FN73] This is the Fundamental Formalist Postulate. Under FFP, if the sovereign issues a command, and your conduct falls within the core meaning of that command, you are constrained to act accordingly. [FN74] Anyone who says otherwise is, according to this view, simply wrong.[FN75]
I have already argued that the absence of “easy cases” deprives Formalism of any useful content.[FN76] But even on the philosophical level, substantial cracks within the Formalist edifice have been found in recent years that render Formalism shaky on its own FFP assumption that words have core meanings:
1. Contemporaneous with Hart's announcement of the “core” postulate, Lon L. Fuller invented a Godelian application of Hart's exemplary statute, one that was undecidable even though by hypothesis it fell within the core.[FN77] Fuller thus showed that determinate core meanings do not necessarily constrain decisions.[FN78] I have discussed this development at length elsewhere.[FN79] Suffice it to suggest here that the number (or more appropriately the density) of Godelian undecidable propositions in law [pg173] may vastly exceed the number in mathematics.[FN80]
2. Even though the FFP is designed for deductive application, the actual content of the “core meaning” of any word requires an inductive inquiry. We must look at all the previous instances of the speaker's use of the word in question—including, by implied reference, all previous uses of that word by any speaker throughout history. We then add up all these instances, get the average meaning, and “apply” it to the current case. For example, we find what the term “due process” meant throughout history by all speakers and in every single case in which the term was invoked, and then “apply” its “core meaning” to the present case. But this raises (apart from the unmanageability of the scholarly effort that would be required) the problem of the validity of induction itself. As Nelson Goodman demonstrated, no amount of historical evidence (i.e., not even all the evidence from the beginning of the universe to the present moment) compels the result in the next case. In fact, all the previous evidence can be reformulated so that it is compatible with any result in the next case![FN81] Wittgenstein arrived at the same conclusion in his study of ordinary language.[FN82] Saul Kripke has explicated and generalized Wittgenstein's result.[FN83] Kripke has shown that even if you have a rule as determinate as the mathematical law of addition, a person who adds two numbers she has never added before (68 + 57, to use Kripke's example) can give any answer (say, 2,381) and say that she has used “quaddition” which gives the same results as “addition” except when 68 and 57 are [pg174] being added.[FN84] “Law” is an fortiori case from these mathematical examples, not only because legal rules appear on their face to be less rigorous than mathematical rules of addition or multiplication, but also because time is a relevant differentiating factor in law. If someone added 57 + 68 twenty years ago and got 125, it would be hard to make a plausible claim that the same person adding the same numbers today can justify a total of 2,381.[FN85]
But in law, every new case is different from all previous cases at least in so far as time (and probably the identity of the parties) is concerned.[FN86] Thus, Wittgenstein, Goodman, and Kripke have shown that there can be no (legal) rule that compels a given result (no rule that constrains a judge), because any result can be incorporated in a reformulation of the rule.[FN87] Another way of stating this position is to say that no amount of [pg175] legislative history can ever constrain a present case result.[FN88] A decision either for the plaintiff or for the defendant can be justified, using the Wittgenstein-Goodman-Kripke analysis, on the basis of exactly the same legislative history.[FN89]
3. The law of a given legal system may be internally consistent, [FN90] but all the real-world facts and events to which the law refers can be consistently exchanged for an entirely different set of facts and events. This is the result reached by Lowenheim-Skolem, which, simply stated, is that ontology is indifferent to any formal system.[FN91]If we combine the previous Wittgenstein-Goodman-Kripke result with Lowenheim-Skolem, we realize that even a highly formalized set of rules, such as the Restatements of Contracts and Torts, can consistently be said to “apply” to mutually[pg176] inconsistent descriptions of fact situations.[FN92]
4. An important contribution to the breakdown of Formalism was [pg177] made by Professor Kress himself in 1984.[FN93] He showed that the time between when the facts of a case arise and when the case is adjudicated in a court will necessarily affect the state of the law applied to the case. For during the gap between facts and adjudication, intervening decisions in other cases will be handed down, and these will have some impact on the content of the law that will be applied to the instant case.[FN94] It follows from Professor Kress's significant thought experiment that the law can never be determinate at the moment that people act; or, in other words, that no one can know precisely what the law is when one needs to know it in order to act legally.[FN95]
5. A possible additional infirmity in the Formalist facade is the formal proof I worked out in 1983 that in all contemplated or actual cases, the party who challenges the applicability of a given rule (call him Jack) has an inherent advantage.[FN96] The central proposition in this argument is that Jack has an economic advantage over the person who is trying to utilize the state's judicial machinery to enforce the rule (call her Irma):
First, Jack has a net incentive to discourage Irma from bringing the legal action at all. Second, Jack can more efficiently widen the gap between his own conduct and the law on the books than Irma can restore it. Third, Jack's contemplated transaction costs will be less than Irma's, and hence he can spend the difference in the purchase of legal creativity to unravel the rules of law. All of these asymmetries lead to a systemic bias in favor of persons disadvantaged by legal rules to make the law less certain.[FN97]
This previously unnoticed asymmetry, which I argued is built into the structure of our legal system, leads to the result, entirely within FFP, that rules of law will unravel over time. The result is analogous to the Second Law of Thermodynamics: rules of law will inevitably lose their structural sharpness and tend toward randomness and unpredictability. Accretions made to the rules (such as legislative histories or dense codifications) do not reduce entropy—they actually hasten the race to randomness.[FN98] The conclusion I reached is that whatever our degree of certainty [pg178] might be about the meaning of any given legal rule today, the meaning of that rule will inevitably be less certain tomorrow.
6. To the extent that the Critical Legal Studies movement is linked with the Indeterminacy thesis, the linkage stems from Duncan Kennedy's “fundamental contradiction” which purportedly demonstrated that there was in every case an unresolvable tension between the forces of altruism and selfishness.[FN99] This tension, in Professor Kennedy's view, rendered every right or principle indeterminate and useless as a constraint upon judges in any given case. But now, over a decade since Professor Kennedy announced his theory, it has become evident that no theory, rule, or principle can constrain particular judicial decisions,[FN100] and hence Professor Kennedy's “fundamental contradiction” is not needed. Moreover, his “fundamental contradiction” is itself just a pair of theories—one a theory of “altruism” and the other of “selfishness.”[FN101] Thus, instead of telling us that all theories are indeterminate, he replaced them with a reductionist account of two overarching indeterminate theories. Although Professor Kennedy's “fundamental contradiction” was of historical importance in the development of Indeterminacy, with the passage of time it appears somewhat less significant as a logical crack in the Formalist edifice.[FN102]
Although the preceding demonstrations may not prove conclusively that Legal Formalism is moribund,[FN103] they should be discouraging [pg179] enough to a Formalist to suggest the desirability of seriously investigating the alternative of Pragmatic Indeterminacy.
V. Signs along the Pragmatic Route
Indeterminacy rejects FFP.[FN104] A Pragmatic Indeterminist would show that FFP is not the way the mind works, that it is not the way that we learn to speak and to use language. An asserted “core meaning” of a word is a matter of degree and not of kind, and in any event may vary from person to person. Whether real-world conduct “falls within” a “core meaning” of some word or phrase is a matter of human judgment.[FN105]Pragmatic Indeterminacy not only rejects Formalism, but is a substitute for Formalism.
For lack of space if not of ability, I cannot give an extended account here of Pragmatic Indeterminacy. No account of the subject can be complete, because Indeterminacy cannot be bounded or circumscribed—that word (or concept) like all other words (or concepts) does not have a determinate core meaning. However, let me try to sketch what at the present time I think are its main procedures and goals: [FN106]
[pg180] 1. The starting point can be labeled the Non-Anarchic Postulate. Law is a predictable phenomenon. Experienced lawyers can make educated guesses about how judges will probably decide many cases and controversies. A client walks in the door with a legal problem, and within a few minutes, an attorney can start giving probabilistic legal advice that the client can use to good advantage in planning further behavior. Of course, because each fact situation is unique, we can never know as a matter of “objective probability” that the lawyer's advice is accurate. But “subjective probability” works quite well in these cases, as I have tried to demonstrate in an earlier writing.[FN107]
If the results of courts were not fairly predictable, then lawyers could not give probabilistic legal advice and pretty soon no one would pay lawyers for their advice. Lawyers would be out of business, and society would become chaotic.[FN108] For Pragmatic Indeterminacy, some nontrivial degree of judicial regularity and predictability is required. So long as we have a non-chaotic society, we can be fairly confident that “law” works in that society.
2. What is our evidence that law works? The evidence, I submit, cannot be found in legal materials. All the reported cases in the world cannot give us any assurance about what courts will actually decide. This apparently startling proposition can be readily tested by a thought experiment. Suppose you come across fifty volumes of the reported cases of a particular jurisdiction. You read all the cases and note an amazing consistency: there are no overrulings, each case cites precedents, and when you look up the precedents, you find that they are “on all fours” even by your own standards, no matter how rigorous your standards are.[FN109] None of this means that real people have in fact been treated predictably or consistently. For all you know, the statements of facts in the reported cases are actually misstatements, invent the facts, misinterpret[pg181] the facts, or are so highly selective of the facts as to constitute uncandid accounts of what really took place.[FN110] In particular, all the “law-facts” (such as “the plaintiff was also negligent”) can be wildly at variance with the real-world facts. You might even find, on further investigation, that all the reported cases in this jurisdiction were plagiarized from case reports of another jurisdiction. The judges in “our” jurisdiction simply copied the opinions of other judges and tacked them on to their own cases in an attempt to “look legal.” I am not saying that this strange state of affairs is likely, only that it is conceivable.[FN111] The thought experiment simply shows that we cannot use court reports as evidence of what is happening in the real world.[FN112]
Our evidence that law works must instead come from lawyers and the public. If most people, including lawyers, believe that litigated cases are not decided randomly but can be predicted at a confidence level ranging, say, from fifty-five to ninety percent depending on the case, then they are probably right because they are talking from experience with decisions the courts have made in their own jurisdictions. Societies that view their court systems as generally issuing unsurprising rulings are probably accurate in viewing their courts as behaving within predictable limits. Our evidence for popular belief in judicial predictability is the care people take to structure their transactions by taking into account the risk of having them upset by challenges in court. If the risk were wholly unspecifiable, people would not make plans in light of what courts might decide.
“Law” in this view means only a lawyer's prediction that on such-and-such facts a court will decide with a certain percentage probability for the client.[FN113] The “law” can be different depending on the facts, the client, and the lawyer. For example, a client tells two attorneys, A and B, the same story. Attorney A says the client's chance of prevailing in a lawsuit is fifty percent, attorney B says it is seventy five percent. Both may be precisely correct; the law can be different for the client depending upon choice of counsel. For lawyer A may be inexperienced in this area of the law, and may have a poor record of dealing with similar cases. Lawyer B may have more talent, experience, and success, fully justifying a seventy five percent confidence level of winning the case.[FN114]
[pg182] 3. What reason does an Indeterminist give for the fact that law works in non-chaotic societies? A clue can be found if we look closely at what we mean by predictability. If I can predict that a certain set of real-world events—assuming competent description of those events in court clothed in appropriate legalist rhetoric—will have an eighty percent chance of obtaining a favorable judicial decision for my client, then I must have some elementary notion of similarity in my ability to make this prediction. I must have mentally compared my client's set of real-world events to other similar sets, and made the judgment that the result in my client's case will track the results previously obtained by similarly event-situated persons.[FN115]
I did not necessarily need words to go through the mental process of similarity-comparison. A word is a conclusion I reach when I have already decided that two things are more similar to each other than they are to a third thing. For example, if the color I now see is similar enough to red colors I have seen in the past, I call it “red.”[FN116] If my client's contract is sufficiently similar (in my own view) to adhesion contracts I have seen or read about, I venture to call it an “adhesion contract” (of course, my label is a shorthand for a prediction of how a court might view the contract).[FN117] Our mental ability to sense similarities is prior to our ability to learn a language.[FN118] Indeed, our language ability itself [pg183] could not have arisen if we lacked a prior mental ability to determine similarities and differences. Rudolf Carnap noticed this fundamental point in his masterwork The Logical Structure of the World.[FN119] The basic unit of his system is the erleb, a momentary cross section of the total stream of experience. Erlebs are related to each other by “part identity” and “part similarity.”[FN120] The relationships of erlebs are preverbal.[FN121] A word is simply a label we learn to attach to erleb relationships that we notice.[FN122]
[pg184] The importance of similarities for law study has, of course, been noted long ago. Reasoning by analogy is said to be the basic way that we “think like a lawyer.”[FN123] Lawyers are able to make predictions for their clients because of their ability to notice similarities across real-world situations. This may seem quite obvious and elementary, but it is a long way from the Formalist notion that lawyers learn rules of law and then apply them to the facts of a case (and even farther away from Dworkin's view that judges find the best theory of law and apply that).
Decisionmaking by reference to similarities in prior decisions is not just an attribute of the legal system. It applies in nearly all social situations: in the family, in the office, in voluntary organizations. The Similarity Engine (real-world similarities fueled by our inductive expectation that similar situations should yield similar results) keeps society running with a minimum of unfair surprises. The legal system is simply a more formal example of the workings of this Similarity Engine. By becoming familiar with it, lawyers can make useful predictions to clients, and judges can maximize their own effectiveness and power by behaving in a manner that actualizes the utility of lawyers' predictions.[FN124]
4. Law, however, works only probabilistically. Similarities, basic though they are, can only be a matter of more-or-less. Compounding the difficulty is the possibility that differently situated observers may draw different distinctions between similar cases. [FN125] The result is that law-words, absolute though they may appear when printed on paper, can at best point to likely or probable results in particular cases (which is another way of saying that there are no easy cases). Pragmatic Indeterminacy not only regards each judicial decision as indeterminate, but also—if it were possible to exclude the normative dimension of justice or morality[pg185] from law and mean by “law” only the words of the law (statutes, precedents, etc.)[FN126]—considers it to be a category mistake to use notions of “legally right” or “legally correct” in criticizing a given decision.[FN127] Although we often use the words “correct” or “incorrect” in criticizing judicial decisions, the fact that any decision is a matter of probability combined with the fact that no case is “easy” means that either side can legitimately win any individual case.[FN128] I maintain not only that our real world is one of probabilities and not certainties, as quantum theory has [pg186] amply demonstrated, [FN129] but also that our legal world is made up only of probabilities and not certainties.
The realization that law only tells you what might happen to you and not what will necessarily happen [FN130] is perhaps a “soft” view of law. I think it is more humanistic than the dictatorial view of legal positivism which regards laws as determinate commands that bind judges as well as us.[FN131] The soft view of law considers human deciders (judges) as being part of, rather than apart from, the law.[FN132]
5. Yet my contention that the language of law at best serves a retrieval function[FN133] is admittedly a deflationary view of law.[FN134] Many people, accustomed to nobler sentiments where law is concerned, may regard the retrieval thesis as a disaster. I suggest that it is a gain. If we have been fooling ourselves that law words constrain judges in individual cases, then we can dispense with a considerable amount of time-consuming theorizing about law and about whether specific decisions were correctly[pg187] decided under the law. We will then have time to turn our attention to things that matter, such as justice.[FN135]
Justice matters not only in itself, but also because it may be able to point us to the right decision in any given case—something the words of the law cannot do. A judge ought to decide a case in the interests of justice. Where justice is concerned, there can be easy cases. Not all cases are easy; some are very hard in terms of sorting out the justice interests. But I think it can be done, although the task is formidable. In law school education we have barely scratched the surface.
6. How should law be taught? First, by giving a lot more attention to the facts of cases and to how those facts are ascertained and proven.[FN136] The tendency of recent casebooks to truncate the statement of facts of cases and instead pile on legal rhetoric is a move in precisely the wrong direction.[FN137] Second, attention should be paid to morality and justice. These are the engines that move impartial judges (if anything moves impartial judges). Law-words do not constrain judicial decisions, but sentiments of what is moral, fair, and just may have a powerful, normative impact upon decisionmakers. Law school curriculums need to focus on how normative elements are best identified and argued—instead of assuming that students pick it up through osmosis.[FN138]
7. Perhaps the most important nonacademic consequence of Pragmatic[pg188] Indeterminacy is the issue of how we select judges. Societal justice depends on having good and fair-minded judges making decisions. The more convinced we are about the inability of law-words to constrain particular judicial decisions, the less we should require technical legalist competence in candidates for judgeships. Although media commentators urge us to appoint judges who are learned in the law, who did well in law school, and who have the ability to craft sophisticated, persuasive opinions, the media is as usual a couple of decades behind. For the ability of a judge to state the law in a sophisticated way has, in my opinion, practically nothing to do with what we should really be concerned about—fairness and justice.[FN139] The more we require our judges to be verbally skilled practitioners of the legal art, the less we can expect them to have found room in their lives for actual empathic experiences, for the wisdom that comes from contemplating the human condition, and for the maturity of judgment that comes from reflecting upon what to do in thousands of daily interactions with other people in diverse contexts. What we need on the bench are qualities of compassion, fairness, mercy, good judgment, experience of many walks of life, sensitivity, humanism, and empathy.[FN140] Since judges cannot be constrained by law-words,[FN141] they should be the kind of people who feel constrained by justice.[FN142]
[pg189] VI. Conclusion: Consequences of Pragmatic Indeterminacy
When a sympathetic commentator such as Professor Benson suggests that the Indeterminacy debate is a tempest in an academic teapot, reflective of a mix-and-mash culture and useful only for the purpose of self-indulgence, I would like to underline for him the last two consequences of Pragmatic Indeterminacy that I have listed in the preceding section. If, as a result of taking Indeterminacy seriously, we revolutionize the way we teach law and the way we select judges, then we will also revolutionize the way cases are litigated (because the new judges will expect to hear a different kind of argumentation) and the way people order their lives in anticipation of the way their disputes will be decided by these new judges. I doubt that a debate about law can be dismissed as academic if it entails societal consequences such as these.