SOURCE: "Cardozo and the Upper-Court Myth," in Law and Contemporary Problems, Vol. 13, No. 2, Spring, 1948, pp. 369-90.

[Frank was an American jurist who served on the U.S. Court of Appeals for the Second Circuit from 1941 until 1957. In the following excerpt from a review of Selected Writings of Benjamin Nathan Cardozo, Frank faults Cardozo's description of the judicial process because it ignores the operations of trial courts.]

The practical is disagreeable, a mean and stony soil, but from that all valuable theory comes.

[Oliver Wendell Holmes, Holmes: His Book Notices and Uncollected Letters and Papers]

[The] first step toward improvement is to look the facts in the face.

[Holmes, Rational Basis of Legal Institutions]

I

There has recently been published a volume, Selected Writings of Benjamin N. Cardozo, which every thoughtful lawyer and judge will want ready at hand. It will repay constant re-reading. It includes nearly all Cardozo's extra-judicial writings, notably The Nature of the Judicial Process, first published in 1921, and The Growth of the Law, first published in 1924. In these two books, one of our most eminent appellate judges set forth his legal philosophy. More important, he showed how this philosophy aided him in his judicial work, and, in that connection, disclosed some of the intimate details of upper-court techniques. I say "more important" because, before Cardozo, no judge, with the exception of Holmes, had been similarly candid. Cardozo's frankness emboldened others, lawyers and judges, to be less diffident in thinking about and commenting on courthouse ways.

He did not confine himself to a description of what appellate judges do. He told, also, what they ought and ought not do. While wisely indicating the proper limits of judicial legislation, he exhorted these judges to put moral ideals into practice. His descriptions and exhortations he both illustrated refreshingly by his own judicial opinions, and fortified by apposite reflections on a variety of legal and non-legal philosophies.

This sort of critical self-revelation by a judge was novel and exciting. Somewhat less novelty and inventiveness appear in the contents of his own philosophy—far less than in the off-the-bench writings of his great master, Holmes, all of which contain highly original insights. Cardozo, more productive than Holmes, was also more eclectic. But, as in the case of Cicero, brilliant eclecticism became itself creative.

It was tonic that a revered judge, by no means "radical," should boldly declare, "I take judge-made law as one of the realities of life" [The Nature of the Judicial Process]. To be sure, this had often been said previously, once forth-rightly (and with appropriate qualifications) by Holmes in a dissenting opinion to which Cardozo acknowledged his indebtedness. But many lawyers then regarded Holmes as queer and flighty. Holmes, too, had frequently noted that sometimes subconscious factors affected judges, that they sometimes decided cases intuitionally, that policy attitudes, too often entertained unconsciously, influenced decisions. "The very considerations which the courts most rarely mention, and always with an apology," Holmes wrote as early as 1879, "are the secret root from which the law draws all the juices of life. We mean, of course, considerations of what is expedient for the community concerned. Every important principle which is developed by litigation is in fact and at bottom the result of more or less definitely understood views of public policy; most generally, to be sure, under our practice and traditions, the unconscious result of instinctive preferences and inarticulate convictions, but none the less traceable to public policy in the last analysis" ["Common Carriers and the Common Law," American Law Review (1879)]. Again, however, it was important that such a judge as Cardozo should enunciate that thesis; especially valuable was it to have Cardozo say: "Deep below consciousness are other forces, the likes and the dislikes, the predilections and the prejudices, the complex of instincts and emotions and habits and convictions, which make the man, whether he be litigant or judge.… There has been a certain lack of candor in much of the discussion of the theme, or rather in the refusal to discuss it, as if judges must lose respect and confidence by the reminder that they are subject to human limitations.… The great tides and currents which engulf the rest of men, do not turn aside in their course, and pass judges by" [The Nature of the Judicial Process]. Following Holmes, he sagely added: "The training of the judge, if coupled with what is styled the judicial temperament, will help in some degree to emancipate him from the suggestive power of individual dislikes and prepossessions."

Interested in matters of policy, Cardozo was fascinated by cases presenting situations where gaps in the precedents existed, or by those which invited deviation from the precedents because the old doctrines had produced unduly harsh results. He delighted in portraying the "creative function" of the judges in such instances, a function Cardozo brilliantly performed in some of his landmark opinions. Wisely, however, he warned repeatedly that those were "exceptional cases."

Perhaps Cardozo's most original suggestion was that a judge, in determining what legal rules or principles to apply in deciding a case, can and should employ four methods. Cardozo called them the methods (1) of philosophy (or analogy or logic), (2) of evolution (historical development), (3) of tradition (customs), and (4) of sociology (justice, morals, and social welfare). This suggestion stimulates the advocate who aims to persuade judges, and supplies judges with some very useful tools in decision-making. These tools, [Selected Writings of Benjamin N. Cardozo], to be sure, have some rough edges. As Professor Patterson says in his Preface to the four methods "were not exactly phrased or clearly delimited" by Cardozo. In particular, Cardozo's use of the word "logic" sometimes confuses, as when he speaks of a choice "between one logic and another," although what he really means, as the context shows, is this: The choice relates to the principle which becomes the major premise of the judge's logic—whether that premise be found in historical development or in tradition or in social welfare, or in a compromise between two or more of those three.

The following comments, in 1918, by Hoernle (in a review of a book of legal essays by divers writers, a book from which Cardozo often quotes) point up this common fallacy in which Cardozo indulges: "What logician ever demanded of a lawyer a procedure yielding artificial, unreasonable, unjust results? It is the lawyer, not the logician, who has insisted … on torturing the ever-novel, ever-changing forms of social and industrial life into a strait-jacket of concept and rules, which are… treated as rigid and fixed … There is no logic known to me which forbids the recognition that 'legal systems do and must grow, that legal principles are not absolute, but relative to time and place'; or which demands the ignoring of the connection between law and economics.… But what makes the difference … between the kind of thinking that is rightly denounced as formalism, logic-chopping, hair-splitting,… and the kind of thinking which our authors seek to secure … by the introduction of historical, sociological, psychological considerations? The difference is material, not formal. Formally, thinking is good or bad, according as the conclusion does, or does not, follow from the premises. Materially, thinking is good or bad, according as the premises from which conclusions are drawn are more or less complete as measured by the whole range of the problem, or more or less relevant to the task and the purpose in hand. What our authors criticize as bad logic is really bad premises, not a faulty technique in deducing conclusions, but a faulty subject matter.… What our authors call for, and seek to secure, is better premises, a fuller and completer range of material considerations out of which to elicit 'substantial justice'.… The solution depends on what view we take of the … factors which ought to be considered, if concrete justice is to be secured in each particular case.… No logician can legislate for the lawyer on such material points" [review of The Science of Legal Method, in Harvard Law Review (1918)].

But if, in company with many other legal philosophers, Cardozo sometimes went astray in his conception of logic, that was a minor fault. Despite it, these books constitute invaluable treatises on upper-court judging, treatises which almost certainly will never be equalled. I can testify that they have proved an indispensable guide to the perplexed for at least one upper-court judge, and, for him, an inexhaustible source of moral inspiration. To avoid any possible misunderstanding of the balance of this article, let me say, then, that I join whole-heartedly with the many who hail Cardozo as one of our greatest appellate judges and who appraise his books as priceless contributions to a comprehension of what appellate courts do.

By the way of preface to what follows, I quote the following: "That he [Cardozo] was a great judge, that he advanced the progress of keen thinking about the purposes and workings of the courts, is beyond question. That he was a great person, too, is undeniable.… Surely here was a wise and good man, entitled to veneration. But he was neither an immortal nor a mortal god. Being human, he escaped perfection … No man is great in all his aspects. Diderot observed that 'everything even among the greatest of all the sons of men is incomplete, mixed, relative; everything is possible in the way of contradictions and limits; every virtue neighbors elements of uncongenial alloy; all heroism may hide points of littleness; all genius has its days of shortened vision.' Unmitigated or monolithic praise of the great departed often encourages imitation of their errors and weaknesses.… If it should be said that it is presumptuous for so unaccomplished a person to criticize … one so great as Cardozo, my answer will be this: It is a prized democratic maxim that even an alley cat may look at a king; it was an untutored boy who saw the true nature of the emperor's clothes" [Anon Y. Mous, "The Speech of Judges: A Dissenting Opinion," Virginia Law Review (1943)].

II

Since Cardozo, when at the bar, was principally an upper-court lawyer, and, in his long tenure on the bench, an upper-court judge (except for a few months), it would have been understandable if he had avowedly limited himself to writing of "The Nature of the Appellate Phases of the Judicial Process." In that event, these books would have deserved the all but uniform praise they have received. But I think they merit a marked criticism almost never voiced: unfortunately, Cardozo purported, without qualification, to describe the entire judicial process. Because of his reputation, the very excellence of his teachings in the narrow appellate field, to which they legitimately pertain, has tended to dampen inquiry in a far larger field where inquiry is far more necessary.

For Cardozo completely by-passed the operations of the trial courts, as if to say either that they had little significance or that their unique decisional activities and distinctive functions had no place in that process. Although, before he used the phrase, the "judicial process" had been defined to include "all the steps and proceedings in a cause from its commencement to its conclusion" [see United States V. Murphy, 82 F. CD. Del. (1897); State V Guilbert, 56 Ohio St., 47 N. E. (1897); Blair V Maxbass Security Bank 44 N. D., 176 N. W. (1919)]. Cardozo excluded, as if non-existent, the events occurring in the trial stage of thousands of cases, events which occur in trial courts but never in upper courts: the witnesses testifying, the lawyers examining and cross-examining the witnesses, the jurors listening to the witnesses and to the arguments of the lawyers, the trial judge (when sitting without a jury) passing on the credibility of the witnesses' oral testimony. The omission of all these phenomena—familiar to every trial judge, trial lawyer, and newspaper reporters who "cover the courts"—renders Cardozo's exposition, as a description of how courts work, seriously misleading. Eminently satisfactory as an account of appellate-court ways, it is bizarre as an account of trial-court ways—as bizarre as would be an account of manners at Buckingham Palace if taken as also applicable to rush-hour behavior in the New York subways. Nor was Cardozo's misdescription inadvertent. On the contrary, as we shall see, he grew irritated when his grave omission of trial-court happenings was called to his attention, and did his best to discourage efforts to correct his description.

The omission was grave for this reason: because of it, Cardozo, with seeming justification, could give credence to a gross over-estimation of the reliability and excellence of our courthouse products. "Nine-tenths, perhaps more, of the cases that come before a court," he wrote, "are predetermined—predetermined in the sense that they are predestined—their fate established by inevitable laws that follow them from birth to death" [The Growth of the Law]. Substitute "upper court" for "court" in that sentence, and it is perhaps not too wide of the mark. It cannot possibly stand up, however, as a characterization of our entire court system. In most instances, when a case has already been decided by a trial court, a capable lawyer can accurately predict what, if an appeal is taken, will be the decision of the upper court. But (for reasons I shall presently canvass) no such easy prediction can be made of most trial-court decisions. Cardozo, by restricting the judicial process to appellate courts, presented a picture of the workings of our court system as, in the main, just, reliable, and steady. That highly inaccurate picture afforded smug satisfaction to much of the legal profession. For, if Cardozo was correct, the judicial process needed comparatively little improvement.

Now the truth is that, for most persons who become involved in litigation, what trial courts do has far more significance than has the performance of upper courts. For not only is the overwhelming majority of decisions not appealed, but in most of the relatively few that are appealed—probably not more than 6 per cent annually—the appellate courts accept as final the trial-court findings of fact. This they do because of a circumstance which accounts for and derives from a unique characteristic of our trial courts: a jury, or a trial judge in a non-jury case, can observe the demeanor of the orally testifying witnesses. The appellate judges cannot. Observation of witnesses' deportment is by no means an infallible method of determining the accuracy of their testimony; but, no better method having been devised, such observation of witnesses, whenever it is possible, is generally deemed essential in our legal system. Judge Learned Hand summarized views often previously expressed by our courts when he said that "that complex of sight and sound, from which we make our conclusions in a courtroom, is in large part eviscerated when reduced to the printed word" [Petterson Lighterage & T. Corp. v. New York Central R. R., 126 F.2d (C. C. A.2d 1942)]. As Wigmore put it, "The witness' demeanor … is always assumed to be in evidence" [Evidence]. A "stenographic transcript," wrote Judge Ulman, "… fails to reproduce tones of voice and hesitations of speech that often make a sentence mean the reverse of what the mere words signify. The best and most nearly accurate record is like a dehydrated peach; it has neither the substance nor the flavor of the fruit before it was dried." [The Judge Takes the Stand].

It follows that the decisions of trial courts—in which courts alone can the witnesses be seen and heard—determine the fate of, say, 98 per cent of all litigated cases. That 98 per cent Cardozo usually disregarded. For him, a 2 per cent tail wagged a 98 per cent dog. It was as if a meteorologist had founded his studies of weather solely on weather conditions at the equator, or as if a physician had insisted that human health must be studied exclusively in terms of the health of twenty-year-old males. Cardozo, generalizing from wholly insufficient material, was guilty of an unwarranted extrapolation. He basically erred in considering the "facts" of cases as "data," as "given" to the courts. That is true, on the whole, in the appellate phase of that process. It is emphatically not true in its trial phase. The "facts" of cases (as I shall try to show) are not the facts as they actually occurred, not things which exist outside of court. They are usually processed by the trial courts, are their peculiar products. As Judge Olson recently said, "trials are commonly called law suits, but it often seems they might better be called fact suits" ["Observations from a Trial Bench," Washington Law Review (1947)].

Upper courts concern themselves chiefly with the legal rules and principles. So, too, did Cardozo in his books. As a consequence, for him the judicial process signified little more than the application to facts, already "found," of (I) established rules and principles, or (in occasional "exceptional cases") of (2) new rules and principles brought into being through the "creative" activities of the courts. The same was true of Cardozo's conception of "jurisprudence." His definitions of "law"—composed in part of ingredients taken from Holmes and from Pollock and Maitland—included non-legal materials, but did so only in so far as they contributed to such legal concepts or generalizations: "We shall unite in viewing as law," he wrote, "that body of principle and dogma which with a reasonable measure of probability may be predicted as the basis for judgment in pending or future controversies" [The Growth of the Law]. I underscore "the," as it high-lights Cardozo's perspective: The "body of principle and dogma," he is saying in effect, should alone be regarded as "the basis of judgment." Nothing is said to indicate that the ascertainment of the facts of a lawsuit also enters into the making of a judgment, so that to predict future judgments one must be able to predict what facts will be ascertained in future cases.

What I am driving at grows clearer in his other definitions: "A principle or rule of conduct so established with reasonable certainty that it will be enforced by the courts if its authority is challenged, is …a principle or rule of law" [The Growth of the Law, (emphasis added by Frank)]. Subsequently he expanded this statement into a general definition of "law." After noting that "that word stands for a good many notions," he added: "I find lying around loose, and ready to be embodied into a judgment according to some process of selection to be practiced by a judge, a vast conglomeration of principles and rules and usages and moralities. If these are so established as to justify a prediction with reasonable certainty that they will have the backing of the courts in the event that their authority is challenged, I say that they are law.…" ["Jurisprudence," emphasis added by Frank]. It would waste time to argue whether these definitions of the hopelessly ambiguous word "law" are preferable to one of a dozen or more others. But it is distinctly worth while to point out that his definitions contain phrases which Cardozo never bothered to explain although they shriek for elucidation. I refer to the phrases, "will be enforced by the courts," and "will have the backing of the courts." If you peer behind those words, you will be gazing at an immense legal jungle (partly explored by Wigmore, in his Principles of Judicial Proof, and by others in books written for practicing trial lawyers), a jungle Cardozo disdained to enter—the jungle of trials and trial-court fact-finding.

Without trial-court fact-finding, judicial "enforcement" of the legal rules seldom gets under way. For a court "enforces" or "backs up" a rule only if the court, in some specific lawsuit, holds that the facts which invoke that rule are the facts of that specific case. If, and only if, the court so holds, does it apply that rule to those facts. It follows that a rule is not "enforced" unless a trial court has "found" or purports to have "found" the pertinent facts, i.e., those facts to which the rule applies. It also follows that, in almost any particular case—and therefore in almost all cases—trial-court fact-finding is fully as vital as any legal rule. To the human beings whose specific lawsuits the courts decide, the determinations of the facts have the utmost significance. If the facts are found against a party, he loses. The facts of a case, as found, furnish the ticket to the decision. "No tickee, no washee." For a legal rule is only a conditional statement. It says, "If the facts are thus and so, then these legal consequences ensue."

One would suppose, then, that a study purporting to cover the judicial process, a treatise on "jurisprudence," would include an extensive discussion of the methods of trial-court fact-finding and of the influences that affect it; would explain the multitude of factors involved in it; would emphasize its ineradicable chanciness and uncertainty in most cases, due to the unavoidable fallibility of witnesses, jurors, and trial judges.

Cardozo, however, ignored those topics, and without apologies. This is the more remarkable since, in the single brief passage in which he mentions the function of facts in litigation, he acknowledges their undeniable significance: "In what I have said," he wrote in 1924 [The Nature of the Judicial Process], "I have thrown, perhaps too much into the background and the shadow the cases where the controversy turns not upon the rule of law, but upon its application to the facts. Those cases, after all, make up the bulk of business of the courts. They are important for the litigants concerned in them. They call for intelligence and patience and reasonable discernment on the part of judges who must decide them. But they leave jurisprudence where it stood before." A few lines later, he says that "jurisprudence remains untouched, regardless of the outcome" of such cases.

Truly, an astonishing attitude. "Jurisprudence," as Cardozo envisions it, stands aloof from those decisions, "important to litigants," which, according to Cardozo himself, "make up the bulk of the business of the courts." Why this snobbish aloofness on the part of "jurisprudence" when so much is at stake for our citizens? Because, in Cardozo's view, "jurisprudence" is indifferent to anything other than the legal generalizations, the legal rules and principles. Since the "judicial process," for Cardozo, is substantially co-extensive with "jurisprudence," his exposition of the nature of that process likewise cold-shoulders, as unworthy of consideration, the "bulk of the business of the courts."

Having thus artificially circumscribed the judicial process (by admittedly excluding from it the bulk of judicial business), Cardozo reaches a conclusion as to its workings which necessarily is artificial but which affords him much comfort: since (1) but a small percentage of the legal rules lack certainty, and (2) certainty in the judicial process means certainty in rules, it is demonstrable (according to Cardozo's reasoning) that (3) the extent of uncertainty in the judicial process is small. This conclusion (which is correct in respect of appellate courts but otherwise false) has a logical corollary; i.e., experienced, able lawyers, Cardozo implies, can usually predict court decisions with accuracy: he declared that most rules, even those which allow courts some discretion, "have such an element of certainty that, in a vast majority of instances, prediction ceases to be hazardous for the trained and expert judgment" ["Jurisprudence"]. In other words (so Cardozo apparently maintains) competent lawyers can predict the outcome of most lawsuits before trial.

Not only that. "Law," as Cardozo defines it, shows up as largely stable, since "law" consists of the rules. Here, apparently, is cheer for the layman, since, says Cardozo, the "law" will, on the whole, conform to the layman's "reasonable expectations" ["Jurisprudence"].

The fatal vulnerability of that thesis,—i.e., that legal certainty and the predictability of decisions are measurable by, and correspond to, the certainty of the rules—can be made clear by exposing the error in one of Cardozo's remarks about facts. After noting that the issues in most lawsuits relate "not to the law, but to the facts," he adds: "In countless litigations, the law is so clear that the judges have no discretion." But trial courts—juries or trial judges trying cases without juries—have an amazing discretion in finding the facts. When, as happens in most trials, the testimony is oral and the several witnesses disagree concerning the facts, the trial court's discretion in the determination of the facts—based on a selection of some of the witnesses as credible and others as not—usually is utterly uncontrollable. Indeed, the jury system has often been praised just because juries, through general-verdict fact-finding, have a virtually unregulated power to nullify the legal rules; and much the same power is possessed by a trial judge who sits without a jury.

The exercise of this discretion will often paralyze prediction. In the first place, witnesses, being human, are humanly fallible. No one has discovered or invented any instrument or objective method by which a jury or trial judge can pick out those of the witnesses, if any, who accurately observed the facts in dispute, accurately remembered what they observed, and accurately (without bias or prejudice or perjury) report in court their memories of what they observed. Conventional jurisprudence, in turning its back on that difficulty, relies on an implicit postulate or axiom—the Truth-Will-Out Axiom, i.e., that bias, mistakes and perjury are infrequent, abnormal, and that, when they occur in litigation, they are usually uncovered, so that they have slight effects on the outcome of lawsuits. That axiom, completely out of line with observable trial courtroom realities, should be repudiated. Once it is repudiated, it becomes obvious that, in the selection of portions of the testimony on which to rely, the jury or trial judge must make a guess. In this guessy choice, on which fact-finding is constructed, there inheres much inscrutable, un-get-at-able subjectivity: Not only are the reactions of the witnesses, to the past events about which they testify, shot through with subjectivity. So, also, are the reactions of juries or trial judges to the witnesses. For the juries and trial judges are themselves but fallible witnesses of the witnesses. Thus fact-finding encounters multiple subjectivity.

The truth, neglected by Cardozo & Co., is that the facts in dispute in a lawsuit are past events which do not walk into the courtroom; that those actual past "objective" events can be ascertained, at best, only through subjective reactions to the testifying witnesses' subjective reactions; and that, therefore, to speak of "finding" the "facts" is misdescriptive. Therefore, my description of the nature of a legal rule needs revision: a legal rule means, "If the jury or trial judge (expressly or impliedly) says that it believes the facts are thus and so, and if there is some substantial evidence to justify that statement of belief, then these legal consequences ensue."

No third person can tell whether such statements correctly report the beliefs of the juries or trial judges (i.e., whether the statements match their private beliefs); for it is not permitted to examine or cross-examine juries and judges. But even if any such statement did correctly report such belief, no appellate judges could probe the belief to determine whether it matched the actual past facts, the "facts in themselves." For no one can formulate, in the form of rules, the bases of such a belief. The belief is "unruly," one might say. Long ago, Sir Henry Maine, in a passage which has largely escaped attention, pointed to the delicate task of a trial judge "in drawing inferences from the assertion of a witness to the existence of the facts asserted by him. It is in the passage from the statements of the witness to the inference that those statements are true, that judicial inquiries break down." It "is the rarest and highest personal accomplishment of a judge to make allowance for the ignorance and timidity of witnesses, and to see through the confident and plausible liar. Nor can any general rules be laid down for the acquisition of this power, which has methods peculiar to itself, and almost undefinable" ["The Theory of Evidence," Village Communities (1872)]. Wherefore, as there is no yardstick for measuring the accuracy of a trial court's finding of facts in a case where the testimony is oral and credibility is in issue, often the discretion used in reaching that finding cannot be controlled.

You will search in vain in Cardozo's discourses on the judicial process for any mention of the huge measure of discretion vested in juries and trial judges with respect to the facts. (Indeed, nowhere did he refer to juries.) He wrote as if discretion in the judicial process consisted solely of discretion inhering in the rules or in the selection of rules. That perhaps explains why it never occurred to him to note that it is peculiarly true of jurymen and trial judges—when reacting at trials to witnesses who testify orally—that they are affected by "forces" which lie "deep below consciousness," the "likes and dislikes, the predilections and the prejudices, the complex of emotions and habits and convictions." In the upper courts, as Cardozo noted, those "forces" sometimes somewhat influence the choice of rules. In the trial courts, those "forces" influence, often immeasurably, the choice of facts which is within the wide discretion of jurors and trial judges.

Since that choice is a guess, frequently induced by inscrutable factors, the lawyer, trying to predict a decision, engages in a baffling undertaking: he is trying to predict what facts will be "found," and therefore guessing the future guess of a jury or trial judge. Moreover, before a case is tried, the lawyer's guess often must be about the future guess of a jury, or trial judge, as yet unknown to the lawyer. Necessarily, such a guess is wobbly. It would be wobbly even if all the legal rules and principles were as precise as a table of logarithms, as fixed as the North Star.

How true that is becomes obvious when one considers that in the great majority of suits (e.g., negligence actions and the like) both sides agree as to the applicable rules, the disputes relating to the facts alone. In the light of such cases, it passes understanding that many distinguished legal thinkers, who are rule-obsessed (i.e., who want to believe that decisions are readily foreseeable whenever the pertinent rules are precise), absurdly prattle that clear and definite rules prevent much litigation because (so say these thinkers) most men will not be so foolish as to begin suits in which the relevant rules have such definiteness. Nor can it be ignored that no legal rule, no matter how exact, precludes the injection, by one of the parties to a suit, of an issue of fact which throws open the doors to the reception of oral testimony, and thus to a choice of the "facts" by a trial judge or jury.

These choices of the "facts," resulting from the exercise of the trial courts' guessy discretion, may "leave jurisprudence untouched," but, if so, they leave it looking pretty lifeless, indeed inhuman. The legal rules and principles, the sole foundation of that artificial, ghostly, Cardozian jurisprudence, are stable for the most part, as Cardozo correctly maintains; and, often, predictions of what rules and principles the courts will employ can be fairly exact. Trained lawyers, for example, know the "jurisprudence" relevant to murder trials or automobile accident trials, and can prophesy, with a high degree of reliability, what rules will be applied in such litigation. But what of it? The layman wants to know whether those rules will be applied to the actual facts. If, in a murder case, the correct rule is applied to the wrong facts, an innocent man will be killed by the state. If a trial court, by believing a perjured witness, decides that a deed is a mortgage, the court's legal rule may be impeccable, but the wrong litigant will win.

If, as Cardozo suggests, it be the function of the courts to ensure that, in general, decisions conform to the layman's "reasonable expectations," will a layman consider that a court has discharged its obligations if the legal rules thus conform but, because of an error in fact-finding, the decision in his case does not? As I have said elsewhere: "When the actual facts of a case are not ascertained by a court, its decision may be completely erroneous while yet appearing to be correct. It matters little to a citizen, when he wrongfully loses a lawsuit, whether the decision is the product of the application of the 'wrong' legal rule to the 'right' facts, or the product of the application of the 'right' legal rule to the 'wrong' facts. It is a basic tenet of our theory of justice that cases which are substantially alike should, usually, be decided the same way—according to the same legal rules—and that cases which are substantially different should be decided differently—according to different legal rules. Defects in the ascertainment of the actual facts may result, therefore, in a denial of justice. Because of such defects, cases which, in truth, are very different may seem to the courts to be virtually identical and are decided identically. To the extent, then, that removable defects in fact-finding are not eliminated, proper and practicable individualization of cases does not occur, and avoidable injustice is done" [If Men Were Angels].

When one discusses legal certainty and predictability, is it not misleading to discuss merely the certainty of the rules and of predictions as to what rules courts will employ, while refusing to talk of decisions? If a man, defeated in a suit because of a mistake of the trial court about the facts, goes to jail or the electric chair, or loses his business, will it solace him to learn that there was no possible doubt concerning the applicable legal principle?

Usually a client wants his lawyer to prophesy a specific decision relating to a matter in which that particular client has a specific interest. The likelihood that the lawyer will successfully predict such a decision will vary with the stage at which he is asked for his opinion:

1. When the client, having just signed a contract, asks what are his rights thereunder, at that time neither the client nor the other party to the contract has as yet taken any steps under the contract. The lawyer's prediction at this stage must include a hazardous guess as to what each of the parties will do or not do in the future. The prediction must frequently be so full of if's as to be of little practical value.
2. After events have occurred which give rise to threatened litigation, the client may inquire concerning the outcome of the suit, if one should be brought.
   1. Before the lawyer has interviewed prospective witnesses, his guess is on a shaky foundation.
   2. After interviewing them, his guess is less shaky. But, unless the facts are certain to be agreed upon, the guess is still dubious, if the lawyer does not know what judge will try the case; it is more so, if there may be a jury trial, since the lawyer cannot know what persons will compose the jury.
3. After the trial, but before decision, the lawyer's prophecy may be better. For he is now estimating the reaction to the testimony of a known trial judge or a known jury. Yet, if the testimony was oral, that guessing is frequently not too easy.
4. After trial and judgment by the trial court, the guess relates to the outcome of an appeal, should one be taken. It therefore usually relates solely to the rules the upper court will apply to the facts already "found" by the trial judge or jury. At this stage, a competent, trained lawyer can often predict with accuracy. Only this last prediction situation, no others, does Cardozo discuss.

Surely his discussion is altogether too restricted. And surely those interested in the judicial process ought not to disregard such factors, producing uncertainties in decisions, as, inter alia, the following: perjured witnesses, coached witnesses, biased witnesses, witnesses mistaken in their observation of the facts as to which they testify or in their memory of their observations, missing or dead witnesses, missing or destroyed documents, crooked lawyers, stupid lawyers, stupid jurors, prejudiced jurors, inattentive jurors, trial judges who are stupid or bigoted or biased or "fixed" or inattentive to the testimony. Nor ought humane men ignore the fact that a party may lose a suit he should win because, in preparation for trial, he cannot afford to hire a detective, an engineer, an accountant, or a handwriting expert. Are we, judges and lawyers, to give no heed to such matters and, because of such heedlessness, to do nothing to improve, so far as practicable, the methods of fact-finding in trial courts?

III

None of these matters did Cardozo deign to consider in these books. In banishing such matters from the province of "jurisprudence," and, correlatively, in excluding them from the judicial process, he did a marked social disservice. For if they are thus ostracized, if eminent judges—setting an example to the bar—will not soil their hands with them, but regard them as legal bastards beyond the pale of proper professional notice, who will attend to them?

Someone may say in Cardozo's defense that the followed an established tradition, since few books or articles on jurisprudence had mentioned the vagaries of trials and the obstacles to prediction inhering in fact-finding. But, alas, Cardozo lent his imposing authority to the strengthening of that tradition. Worse, in 1932, in his last published address on the subject. ["Jurisprudence"], he severely criticized those writers who, about that time, were calling attention to the grave deficiencies and unfortunate consequences of that tradition. He ascribed to these writers "anarchical professions," expressions manifesting "a petulant contempt" of "order and certainty and rational coherence," an attempted "degradation of the principles, rules and concepts," "a tendency to exaggerate the indeterminacy … or chance element" in the decisions of cases, and an advocacy of the position that "conformity and order are to be spurned by the judge as no longer goods at all." He wholly disregarded the fact that several of those he thus criticized were trying to describe, not to praise, the current workings of the trial courts, that their purpose was to show that the description of the judicial process in conventional books on jurisprudence had grossly exaggerated the extent of legal certainty and had led to gross over-estimations of the capacity of lawyers to predict many decisions in particular lawsuits, because, no matter how certain most of the rules and principles might be, the fact-determinations in many lawsuits could not be fore-told. He erroneously asserted that these writers regarded as desirable the amount of existing legal imprecision they described. He took their descriptions of existing conditions as manifestations of their aims and desires, of their ideals, their program. When they said, "This is the way things are," in effect he read them as saying, "This is the way things ought to be, the way we want them to be."

I shall dwell on this startling misreading, because it is the key to an understanding of Cardozo's attitude. At first glance, his strictures may appear to have been provoked by his irritation that the descriptions these writers gave of courthouse ways were strikingly at odds with his own relatively tranquil picture of the judicial process. I think, however, that the explanation is more complicated and runs thus: In this 1932 paper, he criticized all those persons unfortunately called "legal realists" but who might better have been labeled adherents of "constructive legal skepticism." Although they had in common a skeptical attitude towards traditional jurisprudence, by no means did they constitute a homogeneous movement, since they disagreed sharply with one another. Nevertheless, they may be roughly divided into two groups:

1. The first and larger group (of whom Llewellyn is representative) may conveniently be labeled "rule-skeptics." They resembled Cardozo in that they had little or no interest in trial courts, but riveted their attention largely on appellate courts and on the nature and uses of the legal rules. Some (not all) of this group (Oliphant being the most conspicuous here) espoused the fatuous notions of "behavioristic psychology." Some (not all) of these "rule-skeptics" went somewhat further than Cardozo as to the extent of the existent and desirable power of judges to alter the legal rules.
2. The second and smaller group may conveniently be labeled the "fact-skeptics." They importantly diverged not only from conventional jurisprudence but also from the "rule-skeptics." So far as appellate courts and the legal rules are concerned, the views of the "fact-skeptics" as to existent and desirable legal certainty approximated the views of Cardozo, Pound, and many others not categorized as "realists." The "fact-skeptics'" divergence sprang from their prime interest in the trial courts. Tracing the major cause of legal uncertainty to trial uncertainties, and claiming that the resultant legal uncertainty was far more extensive than most legal scholars (including the "rule-skeptics") admitted, the "fact-skeptics" urged students of our legal system to abandon an obsessively exclusive concentration on the rules.

Cardozo, in his 1932 paper, did not differentiate between the "rule-skeptics" and the "fact-skeptics." Someone might conceivably think that he deemed the "fact-skeptics" not worthy of his attention. But several times he cited the writings of at least one of them, and singled out those writings for special censure. It occasions surprise, therefore, that he said nothing responsive to the features of those writings which distinguished the views of the "fact-skeptics" from his own—the stress on the facts of cases, on juries, on trial judges, on conflicts in testimony, on the difficulties of lawyers trying to prophesy decisions in lawsuits turning on disputed fact-issues, on the frequent inability of citizens to rely on decisions because of the unknowability of future fact-findings. In this paper, Cardozo uttered not one syllable about those subjects. Why, even when specifically criticising the "fact-skeptics," he maintained such silence, I can only explain as follows: so ingrained had become his habit of assuming that legal uncertainty stemmed exclusively from rule-uncertainty, that, when the "fact-skeptics" spoke of legal uncertainty, he jumped to the conclusion that they, too, must mean solely rule-uncertainty. Wherefore, he was deaf to their reiterated assertions that they had chiefly in mind the uncertainties resulting from the contingencies which affect findings of fact.

For instance, they had said that anyone who visited trial courts would see that the outcome of most trials was chancy because of the numerous chance circumstances influencing fact-determinations; that description Cardozo translated as an expression of a preference for "the random or chance element as a good in itself… exceeding in value the elements of certainty and order and rational coherence." Again, as a part of their description, they had made the statement that, as trial courts have immense discretion in finding the facts of cases, those courts can, and sometimes do, nullify rules by deliberately or inadvertently finding the wrong facts. Cardozo interpreted that descriptive statement as disclosing a desire that a judge should have the power to overturn or nullify any legal rule he disliked. The "fact skeptics" had maintained that stare decisis often yields a certainty that is only an illusion, not primarily because of the instability of the rules, but because—even if all the rules were indubitably clear and fixed, and even if always the courts slavishly adhered to the precedents—the facts found by trial courts would frequently be unforeseeable. This, Cardozo charged, was an expression of a disdain for precedents, of a program for wiping out stare decisis.

In short, he did his utmost to bring into disrepute these efforts, based on observations of trial courts, to revise the traditional, misleading description of the judicial process. He stood steadfastly by his implicit position that that process and jurisprudence begin and end in the upper courts.

You see the tragic consequence of such a position: if, at any time, the legal rules and principles of a legal system are in pretty good shape, then (according to Cardozo), so also is the judicial process of that system, regardless of whether the decisions of the courts are needlessly unfair or unjust. By shrinking the scope of the judicial process, by resisting those who would have it mean the administration of justice in all the courts, Cardozo blocked inquiry into the actual performances of courthouse government.

Our greatest judge, Learned Hand, after a long period of service on a trial bench, remarked in 1926: "I must say that as a litigant I should dread a lawsuit beyond almost anything else short of sickness and death" ["The Deficiencies of Trials to Reach the Heart of the Matter," in Lectures on Legal Topics]. Cardozo took no note of that remark. If he had, it might have given him pause; for he knew that Judge Hand, in so reporting conditions in our trial courts, was not maintaining the desirability of those conditions or expressing "a petulant contempt" of "order and certainty and rational coherence." Consider this statement by Cardozo: "I sometimes think that we worry overmuch about the enduring consequences of our errors. They may work a little confusion for a time. In the end, they will be modified or corrected or their teachings ignored. The future takes care of such things" "[The Nature of the Judicial Process]. Apply that statement to an error about the facts in a trial, as a result of which a man loses his life's savings or is hanged. Did Cardozo mean that "the future takes care of such things?" Of course not. He was not thinking of trials and trial-court decisions but of the legal rules. His was a Hamlet-less Hamlet.

The clue to Cardozo's impatience with those who sought to include descriptions of trial-court operations in the description of the judicial process appears in several passages in his books. "Judgments themselves," he wrote, "have importance for the student so far, and so far only, as they permit a reasonable prediction that like judgments will be rendered if like situations repeat themselves.… When the uniformities are sufficiently constant to be the subject of prediction with reasonable certainty, we say that law exists." "[The Growth of the Law]. Cardozo sought solely uniformities, knowing that, if un-uniformities also exist, the predictor is in trouble; but since in most trial-court fact-finding, much of un-uniformity is present, that trouble often arises. In another passage, he said: "One of the most fundamental social interests is that law shall be uniform and impartial. There must be nothing in its action that savors of prejudice or favor or even arbitrary whim or fitfulness" [The Nature of the Judicial Process]. Yet decisions based on jury verdicts notoriously lack uniformity and impartiality, are replete with prejudice and fitfulness. You see why Cardozo failed to discuss them. Again, he said, "The eccentricities of judges balance one another." [The Nature of the Judicial Process]. So they often do in appellate courts composed of several judges. But the same cannot be said of the divers trial judges and juries when, in separate courtrooms, they hear witnesses and find facts. As I have noted elsewhere, "There is no standardization of judges and jurors so that all of them will be sure to react in identical fashion to any given body of conflicting testimony. Judges and juries vary in their respective intelligence, perceptiveness, attentiveness and other mental and emotional characteristics which are operative while they are listening to, and observing, witness" [If Men were Angels]. From all such disturbing thoughts Cardozo screened himself, by rejecting consideration of nisi prius courts. One cannot believe that he would have written as he did if he had had several years of experience as a trial judge. He had a blind spot. He suffered from an occupational disease to which upper-court judges are susceptible—appellate-court-itis. (That is why, one surmises, his jurisprudence omits, among other things, what might be termed "juriesprudence.")

At one point, for a moment, he almost gives the show away. He quotes Haldane's comment that his own philosophic interest in principles had been valuable to him in his practice at the bar. "It did not help in the work of cross-examination," Haldane says in these quoted comments. "I was never good at that, nor in the conduct of nisi prius cases. But it was invaluable in the preparation for the presentation of great questions to the Supreme Tribunals, where the judges were keen about principles and were looking out for help from the advocate." ["Jurisprudence"]. That hint Cardozo did not follow up. It cannot be overlooked that never does he cast a glance at the many books on trial techniques, "hints to advocates," and such. He cites Wigmore's Evidence, but not Wigmore's Principles of Judicial Proof… which contains a wealth of detail about vicissitudes of trials. Perhaps the secret of Cardozo's avoidance of such books is that, as Morgan suggests, they should not be read "by those who want to believe, and want others to believe, that a lawsuit is a proceeding for the discovery of truth by rational processes" [in a review of Goldstein's Trial Technique, Harvard Law Review (1936)]. It is worthy of remark that Chief Justice Taft, doubtless more conservative in most respects than Cardozo, did interest himself in the deficiencies of trial courts, did try to better the litigating chances in those courts of the underprivileged. What explains this difference between those two men? Probably the fact that Taft, from personal experience, knew more about the lower courts.

In the kind of courtroom where Cardozo spent most of his professional life, the atmosphere is serene—stratospheric. There, lawyers alone address the court; and they must do so with decorum, in an orderly, dignified manner. Not so in the trial courtroom. Absent there the stratospheric hush. It is, as Wigmore notes, "a place of surging emotions, distracting episodes, and sensational surprises." The drama there, full of interruptions, is turbently enacted by flesh-and-blood witnesses and lawyers. In the upper court the clashes between witnesses and counsel appear only in reposeful printed pages. Little wonder if a judge, after many years in such a serene court, grows forgetful of the unserenity characteristic of trials, develops a myopia which limits his range of vision.

Relevant here is the attitude of persons accustomed to years of near-sightedness. "When," we are told, "finally they decide for themselves that they ought to wear glasses, they are often disappointed.… The following conversation is fairly common: Patient: 'Doctor, I think these glasses are wrong. I can't wear them.' Doctor: 'You seem to be able to see quite clearly with them. What is wrong?' Patient: 'I can't bear it. They make all my friends look so ugly. They have spots on their faces and their collars are dirty, and all the houses look so untidy and old. I don't like it. I don't want to see all that detail.' All one can say is that it is a pity such a patient was allowed to grow up so out of touch with reality, as this attitude of not wanting to see what is ugly may extend to one's attitude towards life and lead to a shelving of responsibility" [Mann and Pirie, The Science of Seeing]. This attitude of not wanting to see reality produces "shortsighted policies" [The Science of Seeing]. So, I think, with Cardozo. He exploited his prestige to block a movement aimed at preventing the legal profession from shelving its responsibility for conditions in the trial courts. That was indeed shortsighted policy. He had a "want to believe" and "a want to have others believe" that, on the whole, our courts employ thoroughly rational procedures. He treated his "want," his wish, as if it were a statement of existing courthouse realities. In other words, often when he said, "This is the way things are," he should have said, "This is the way things ought to be." Suppose, now, that he had so recognized, that, correcting his shortsightedness, he had perceived that his purported description of the judicial process was but a want, a wish, an ideal, an objective to be achieved by that process. What then? Then he would have converted his wishful thinking into "thinkful wishing." Then he would have felt eager to investigate our entire court system, including our trial courts, to ascertain how far it fell short of his ideal. Having ascertained, as he would have, its obvious shortcomings, he—a moralist and keen for justice—would have made detailed suggestions to bring that system closer into line with his aspirations.

The perspective which Cardozo's attitude obscured, I recently summarized in an opinion [In re Fried, 161 F.2d (C.C.A.2d 1947)]. as follows: "The 'substantive' legal rules, civil or criminal, embody social policies ('social value judgments'). To enforce, and thus give effect to, such policies is considered one of the principal duties of the courts. They discharge that duty, however, not at wholesale but at retail, by applying those rules in specific lawsuits to the particular facts of those respective suits as 'found' by the courts. As a 'substantive' rule merely declares that specified legal consequences will be attached to a specified state of facts, the rule should be operative only in particular instances where those facts actually occurred. Accordingly, the social policy embodied in any such rule is not actually enforced when, in deciding a case, a court, through misapprehension of what actually occurred, applies that rule to facts which in truth never existed. The whole job then miscarries: Mistakenly to apply a rule to non-existent facts—to facts mistakenly 'found'—is no less unjust, no less a defective operation of judicial administration, than to apply an erroneous 'substantive' legal rule to the actual facts. Either way, the policy expressed in the correct rule is frustrated. An error in 'finding' the facts thus yields what might be called 'injustice according to law.' The facts involved in any case are past facts. They do not walk into the courtroom. Judicial fact-finding, a human process by which a man or some men attempt to reconstruct a segment of an 'objective' past, is necessarily fallible. For it is a job of history-writing, and, like all history-writing, inescapably involves 'subjective' factors and encounters other obstacles sometimes insurmountable. But courts cannot shirk that job.… Unfortunately, the major efforts of those who have tried to improve our legal system have been devoted either to improvements in other phases of 'procedure' or in the 'substantive' legal rules. Those improvements will be needlessly nullified just to the extent that the fact-finding process remains insufficiently scrutinized and, consequently, needlessly defective. Fact-finding is today the soft spot in the administration of justice. In considerable measure that is true because the reformers have largely disregarded the actual fact-finding methods used by the trial courts which, as they are the chief fact-finders, and for other reasons, constitute the most important part of our judicial system; even the procedural reformers have restricted their attention chiefly to those phases of trial court 'procedure' which manifest themselves in upper-court, and occasional trial-court, opinions. It has been too little noticed that a 'substantive legal right'—an 'interest' said to be 'legally protected' by a 'substantive' legal rule—has no practical value when a court by mistakenly mis-finding the facts—because of missing witnesses or documents, or because it believes the testimony of witnesses who in truth are inaccurate, etc.—decides that the claimant has no such 'right' or 'interest.' Doubtless, for analytic purposes, there is often much utility in formally differentiating between 'substantive' and 'procedural' rights (or 'primary' and 'secondary,' or 'antecedent' and 'remedial,' or 'telic' and 'instrumental' rights). Once, however, it is stated, in terms of this formal analysis, that a judicial decision is the 'result of the application of the [substantive] rule of law to the facts procedurally established,' it becomes clear that a mistaken 'procedural establishment' of the facts destroys, for court-room purposes, the asserted 'substantive right,' from which it follows that, so far as courts are concerned, the effective assertion of any 'substantive right' depends entirely on the claimant's ability to maintain his so-called 'procedural right.' The Roman lawyers perhaps sensed this truth when they spoke of the 'procedural consumption' of a 'right of action' by which it was transformed into a 'right to judgment.' In other words, for practical court purposes, no 'substantive' right exists—whether it be a right asserted by a private person or by the government in its role of vindicator of a 'substantive' criminal rule—unless a court gives an enforceable judgment in favor of the alleged right-holder; and, ordinarily, a court will not give such a judgment, even when it uses a seemingly 'correct' rule, if it goes wrong on the facts. Of course, similarly a mistake in fact-finding may cause an erroneous judgment adverse to one who defends against an asserted claim. This, perhaps, appears more clearly if we crudely schematize the formal theory of the decisional process (i.e., the theory that a judicial decision or judgment is the product of a 'substantive' legal rule applied to the facts of the case) by saying: R X F = D—when R is the rule, F the facts, and D the decision or judgment. On that basis, an erroneous F will lead to an erroneous D. As the F consists of the trial court's belief as to what were the actual past facts, the F, and therefore the D, will be erroneous if the court reaches its F by reliance on inaccurate evidence.

"No matter, then, how excellent the 'substantive' legal rules (the R's) and the social policies they embody, specific decisions will go astray, absent competent fact-finding. (Holmes, J., once said that 'the only use of the forms is to present their contents, just as the only use of a pot is to present the beer …, and infinite meditation upon the pot will never give you the beer.'). All of which, I think, goes to show that our trial courts should assume a larger responsibility for the ascertainment, as near as may be, of the actual facts of litigated disputes."

IV

In an address in 1931, Cardozo spoke of "the myths that gather around institutions," saying that often such "myths are really the main thing," and "greater than the reality" ["Faith and a Doubting World"]. Maybe those observations were revelatory. For, in all his writings, Cardozo helped to perpetuate what I would call the Upper-Court Myth, the myth that upper-court opinions are "the main thing" in courthouse government. That myth I think deplorable. It be stows upon us appellate judges too much public kudos. It obscures the transcendent importance of trial courts, and the fact that trial judges encounter far greater difficulties than we do. In part, those difficulties inhere in the character of their job; in part they derive from our antiquated trial procedures. To improve the administration of justice we need, at a minimum, to overhaul our jury system; to revise our evidence rules; to give special training for the trial bench; to augment (without displacing the essential aspects of the adversary procedure) the responsibility of government for insuring that all important and practically available evidence is presented in trials. Those and other improvements will not be achieved as long as judges of great eminence, like Cardozo, continue to induce belief in the Upper-Court Myth.

It is high time that we apply to that myth all the skills of what Samuel Butler called the Art of Covery. "This," he said, "is as important… as Discovery. Surely the glory of finally getting rid of and burying a … troublesome matter should be as great as that of making an important discovery" [The Note-Books of Samuel Butler].