Legal Realism

Brian Leiter

“Legal realism” refers to an intellectual movement in the United States that coalesced around a group of law professors and lawyers in the 1920s and 1930s, including Karl Llewellyn, jerome Frank, Felix Cohen, Herman Oliphant, Waiter Wheeler Cook, Underhill Moore, Hessel Yntema, and Max Radin. These writers thought of themselves as taking a realistic look at how judges decide cases, at 1. what the courts … do in fact,” as Oliver Wendell Holmes, jr (a major intellectual forebear) put it (Holmes, 1897, p. 461). What judges really do, according to the realists, Is decide cases according to how the facts of the cases strike them, and not because legal rules require particular results; judges are largely “fact-responsive” rather than ‘rule-responsive” in reaching decisions.

How a judge responds to the facts of a particular case is determined by various psychological and sociological factors, both conscious and unconscious. The final decision, then, is the product not so much of “law” (which generally permits more than one outcome to be justified) but of these various psycho-social factors, ranging from the political ideology to the institutional role to the personality of the judge. Thus, the legacy of realism in both the practice and teaching of law consists of phenomena like these: lawyers now recognize that judges are influenced by more than legal rules; judges and lawyers openly consider the policy or political implications of legal rules and decisions: law texts now routinely consider the economic, political, and historical context of judicial decisions. In this sense, It Is often said that “we are all realists now.”

The realists are by now the subject of a substantial historical literature (see the bibliography to Fisher, 1993, pp. 325-6). This article will concentrate, by contrast, on the largely neglected, but substantial, contributions of realism to a philosophical theory of law and adjudication, one at odds with the mainstream of the jurisprudential tradition. The realists, unfortunately, often expressed hostility to systematic theorizing and even denied the existence of a “realistic” school of thought; their own theoretical efforts were, at the same time, hindered by a lack of philosophical sophistication and control. These features of their work have led to a highly critical treatment of the realists In the work of later, more philosophically acute jurisprudents (cf. Hart, 1961). Almost despite themselves, however, the realists succeeded in developing a powerful and coherent theoretical view of law and adjudication.

261

Jurisprudential Methodology

Realists have a fundamentally different conception of methodology in jurisprudence, and it is this that puts them at odds with the mainstream of the tradition. Modern legal philosophy has, like most of twentieth-century Anglo-American philosophy, employed the method of conceptual analysis: hence the tide of the seminal work of this genre (Hart, 1961). In its simplest form, the method of conceptual analysis calls for the explication of the meaning of concepts (“morality,’ “knowledge,” “law”) that figure in various human practices; it Is an essentially armchair Inquiry. Such an approach does, however, aim to !Ruminate real social institutions, for ‘the suggestion that inquiries into the meanings of words merely throw fight on words is false” (Hart, 1961, p. v); rather one seeks “a sharpened awareness of words to sharpen our perception of phenomena’ (Hart, 1961, p. 14, quoting 1. L. Austin). In analysing the concepts, then, we Illuminate the social phenomena they describe – for example, our moral, epistemic or legal practices.

Among the features of the concept of law thought to require philosophical explication, two are generally taken to be central: (1) of all the various norms in a society, only some subsets are norms of ‘law” (“criteria of legality’); and (2) that a particular norm Is a “legal’ norm provides agents with special reasons for acting (“normativity of law”). An account of the criteria of legality demarcates the boundary (If any) between norms of law and all other norms in the society (especially norms of morality), and at the same time defines the scope of judicial obligation: judges must abide by and enforce the norms of law.An account of the normativity of law, by contrast, explains how or why law changes our reasons for action. It thus helps demarcate the boundary between group behavior that Is merely habitual and that which is genuinely rule governed; in the latter case, but not the former, the norm describing the behavior provides a standard of conduct to which people can legitimately appeal in justifying conformity with or criticizing deviation from the norm. Only when we understand norms from this ‘Internal” point of view – that Is, as providing agents with these special reasons for action – can we begin to understand the norms that comprise “law” (Hart, 1961, pp. 54-5). (See Article 15, LEGAL POSITIVISM.)

Realism has often been construed by its critics as a conceptual theory – what might be called “the predictive theory,’ often attributed to Holmes (1897, pp. 458, 461; see also, Llewellyn, 1930, esp. at pp. 3-4; Frank, 1930, p. 46; Cohen, 1935, pp. 828-9, 839). According to the predictive theory., a norm is a norm of law just In case It constitutes an accurate prediction of what a court will do; the claim that it is the law that a particular exchange of promises constitutes a contract is, on the predictive theory, equivalent to a prediction that a court will enforce these promises when called upon to do so by one of the parties. If a court declines to find an enforceable contract in the case at hand, then. on the predictive theory, there Is, as a matter of law, no contract. Thus. the final criterion of legality, for the predictive theory, Is what courts do In the particular case, and an accurate statement of law is equivalent to an accurate prediction of what the court will do. Because the

262

predictive theory understands by the concept “law” nothing more than a prediction of what courts will do, the only reason for action provided by the theory comes from the prudent concern of those subject to the “law” to avoid or engage the power of the courts. The statement, “As a matter of law, Mr Jones, you are bound to do X by this contract,” provides, on the predictive theory, only one reason for action for Mr jones: namely, his prudent desire to avoid sanction by the courts for failure to perform. The only normativity of law, then, is the type of normativity present for the person Holmes calls “the bad man” ‘who cares only for the material consequences which … knowledge enables him to predict’ (Holmes, 1897, p. 459).

The predictive theory, so construed, was famously attacked by H. L. A. Hart (1961, pp. 101-2, 132-44). According to Hart, to conceive of the normativity of law as the Holmesian ‘bad man” does is to adopt an “external point of view” on the law, “recording and predicting the decisions of courts or the probable incidence of sanctions”; but this Is to miss precisely the internal aspect of rules which Is distinctive of a legal system. namely that people “continuously express in normative terms their shared acceptance of the law as a guide to conduct” (Hart, 1961, p. 134). This particular criticism, however, seems to miss the mark: for why not see Holmes as also characterizing law from an “internal point of view” and simply contesting how the rules of law look from that point of view? On this Holmesian account of the “Internal point of view,” norms of law are seen by agents as providing – contra Hart – only prudential reasons for action (cf. Perry, 1995).

Hart is more successful in his attack on the predictive theory as an account of the criteria of legality. We may summarize the two central difficulties as follows. First, the predictive theory has no satisfactory account of judicial mistake, by equating the law with what the court does (or will do), it makes it impossible – indeed unintelligible – to complain that a court is mistaken about the law. Second, the predictive theory simply fails a conceptual analysis “.” Take, for example, a judge trying to decide what the “law” Is on some point; according to this theory, what she Is really trying to do is decide what it is she will do, since the law on this point Is equivalent to a prediction of what she will do

The manifest absurdity of the realists’ purported conceptual theory might have suggested that Hart had misinterpreted the realists; yet, on the whole, Hart’s criticisms have been widely embraced. In fact, however, there Is a better explanation for the absurdity: as a methodological matter, the realists were not engaged In conceptual analysis. Indeed. Holmes makes clear on the very first page of “The path of the law” that he is talking about the meaning of law to lawyers, who will .appear before judges, or … advise people in such a way as to keep them out of court’ (1897, p. 457), and not aiming for a generally applicable analysis of the concept of law. So, too, Frank cautions that he “is primarily concerned with ‘law’ as it affects the work of the practicing lawyer and the needs of the clients who retain him” (Frank, 1930, p. 47 n.). (Hart’s criticisms may be more apt with respect to Cohen, 1 9 3 5.) In fact, it will become clear as we consider the realist arguments for legal indeterminacy (below) that the realists are, Inconceptual matters, tacit legal positivists with respect to the criteria of legality.

263

If the realists are not engaged In conceptual analysis, what methodology are they employing? Interestingly, conceptual analysis has fallen out of favor In philosophy since the late 1960s, except in jurisprudence. This general development marks what might be called “the naturalistic turn” in philosophy, and it is here that we will find the key to realist methodology: for the realists are not bad legal philosophers, as Hart’s analysis might suggest, but prescient ones, philosophical naturalists before their time.

Naturalists In philosophy all share the following methodological view: philosophical theorizing ought to be continuous with and dependent upon empirical inquiry in the natural and social sciences. It will not do to seek an account of phenomena through an armchair analysis of concepts; we must begin, instead, with the relevant empirical data about these phenomena provided by the various sciences, and construct our philosophical theory to accommodate them.

W.V. Quine (“Epistemology naturalized” in Kornblith, 1994) provides one important contemporary paradigm of philosophical naturalism, what we may call “Replacement Naturalism.” According to Quine, epistemology studies the relationship between evidence (in the form of sensory input) and our various theories about the world (the cognitive “output” as it were). Traditional (non-naturalized) epistemology wants to find a normative, foundational relationship between evidence and theory: It aims to show which of our theories are really justified on the basis of Indubitable evidence. Quine argues that the foundationalist program is impossible, in part because evidence always underdetermines the choice among theories, and thus does not Justify only one of them. From the failure of the normative, foundational project, Quine draws the conclusion that the only fruitful study of the relation between evidence (sensory input) and theory (cognitive output) is a descriptive account of what Input causes what output, of the sort provided by psychology. Thus, says Quine, “Epistemology … simply falls Into place as a chapter of psychology” (Komblith, 1994, p. 25). The science of human cognition replaces armchair epistemology: we naturalize epistemology by turning over its central question – the relation between theory and evidence – to the relevant empirical science.

The dominant strand of naturalism in realism is a type of replacement naturalism. Indeed, Quine’s famous slogan – “Epistemology … simply falls Into place as a chapter of psychology” – echoes Under! Moore’s own jurisprudential credo 2 5 years earlier; his work he says “lies within the province of jurisprudence. It also lies within the field of behavioristic psychology. It places the province within the field” (Moore and Callahan, 1943, p. 1). jurisprudence – or more precisely, the theory of adjudication – falls into place, for the realist, as a chapter of psychology (or social science generally); we abandon the normative ambition of telling judges how they ought to decide cases in order to undertake the descriptive study of the causal relations between Input (facts and rules of law) and outputs (judicial decisions). This yields a fully naturalized descriptive theory of adjudication, rather than a conceptual theory of the criteria of legality or a conceptual theory of adjudication, as a by-product of the former (cL Hart, 1961).

Yet Moore, it may seem, does not speak for all realists, some of whom appear to

264

retain the ambition of formulating a normative theory (as do many naturalists epistemologists, cf. the essays by Alvin Goldman in Kornblith, 1994). All the realists endorse the following descriptive claim about adjudication: in deciding cases judges respond primarily to the stimulus of the facts. The question, to which w return below, Is whether this descriptive claim can be parlayed into a normative theory, the traditional ambition of jurisprudence (to tell judges how they ought t decide).

Legal Indeterminacy

Why abandon the normative ambitions of traditional jurisprudence in favor o replacement naturalism? Here we need to understand the influential realist arguments about the Indeterminacy of law.

The law on some point is rationally indeterminate when the “class of legal reasons” (hereafter “the class”) Is insufficient to justify a unique outcome on that point. The class encompasses those reasons that are proper justificatory grounds of judicial decision: for example, that prior, analogous cases have held a similar way or that a relevant statute requires the outcome are legitimate reasons for a judicial decision. The law is locally indeterminate when it is indeterminate only in some select range of cases (for example, those cases that reach the stage of appellate review). The law is globally indeterminate when it is indeterminate in all cases. (Strictly speaking, we are concerned only with the underdeterminacy of law, not its indeterminacy: we are concerned that the class justifies more than one, but perhaps not simply any, outcome.)

The law Is causally indeterminate if the class is insufficient to cause the judge to reach only one outcome in that case. More precisely, suppose that relevant “background conditions” obtain: judges are rational, honest and competent, and they do not make mistakes. The law is causally indeterminate just in case the class together with relevant background conditions is still insufficient to cause the judge to reach only one outcome in that case. One reason this might be true is, for example, because the law is rationally indeterminate on that point: if the class justifies more than one outcome, then even rational, honest and competent judges will not be caused by applicable legal reasons to reach the decision they reach; we must look elsewhere to find out what caused them to do what they did. (On this way of conceiving the varieties of indeterminacy, see Leiter, 1995.)

All realists defend the following two theses about indeterminacy: (1) the law is rationally indeterminate locally not globally; and (2) the law is causally indeterminate in the cases where it is rationally indeterminate. Some realists defend this additional thesis: (3) the law is causally indeterminate even where It is rationally determinate and the background conditions obtain.

Rational Indeterminacy

The class includes legitimate sources of law (such as statutes, prior court decisions) and legitimate ways of interpreting and reasoning from those sources (for example Interpreting statutes by the ‘plain meaning”, reasoning by analogy). Someone might think the law is Indeterminate because there are too few sources of law (so that there are no legal reasons for decision on some points) or too many conflicting sources (so that the conflicting sources provide reasons fconflicting decisions). The realists, however, argue that rational indeterminacy results from there being too many conflicting but equally legitimate ways of interpreting and reasoning from the sources, thus yielding conflicting legal rules. Thus, for example, Llewellyn argues that it is equally legitimate for a court to treat precedent ‘strictly” or “loosely.’ On the strict view, a later court can reexamine the [earlier] case and can invoke the canon that no judge has power to decide what Is not before him, can, through examination of the facts or of the procedural issue, narrow the picture of what was actually before the court and can hold that the ruling made requires to be understood as thus restricted. In the extreme form this results in what is known as expressly ‘confining the case to its particular facts’. (1930, p. 72)

The strict view of precedent, says Llewellyn, ‘Is applied to unwelcome precedents” (1930, p. 73), as a way of distinguishing them from the case at hand. But there Is another approach to precedent, the ‘loose view,” which “is like the other, recognized, legitimate, honorable” (1930, p. 74). On this view the earlier court ‘has decided, and decided authoritatively, any points or all points on which It chose to rest a case” so that In ‘Its extreme form this results in thinking and arguing exclusively from language that is found In past opinions, and In citing and working with that language wholly without reference to the facts of the case which called the language forth” (1930, p. 74). But if ‘each precedent has not one value [that Is, stands for not just one rule], but two, and … the two are wide apart, and … whichever value a later court assigns to it, such assignment will be respectable, traditionally sound, dogmatically correct’ (1930, p. 76), then precedent, as a source of law, cannot provide reasons for a unique outcome, because precedent can be Interpreted to stand for more than one rule, and so justify more than one outcome.

As with precedent, Llewellyn argues that with respect to the Interpretation of statues, “there are ‘correct,’ unchallengeable rules of ‘how to read’ which lead In happily variant directions’ (1950, p. 399). By mining the cases, Llewellyn shows that courts have endorsed contradictory ‘canons of construction’ like ‘A statute cannot go beyond Its text” but also “To effect its purpose a statute must be Implemented beyond its text” (1950, p. 401; Llewellyn adduces 28 contradictory canons at401-6., cf. Llewellyn, 1930, p. 90). But If a statute can properly be construed In contradictory ways to stand for different rules, then reasoning from the statute will not justify a unique outcome in the case at hand (cf. Radin, 1930).

Indeterminacy enters not just In the Interpretation of statutes and precedents, but also In the wide latitude judges have in how to characterize the facts of a case. After all, rules – what we get by Interpreting precedents and statutes – must be applied to facts; but the facts of a case do not come with their own descriptions, and must be characterized In terms of their legal Import. Many realists argued here, as well, that judges could legitimately characterize the same facts in differing ways. and thus even with a definite rule, the judge could still be justified in reaching more than one decision depending on how he characterized the facts. (See Frank, 1930, pp. 108-10; Frank, 1931, p. 28; Llewellyn. 1930, p. 80.)

Local Indeterminacy

The Realists, unlike the later writers of critical legal studies, defended only the view that the law was locally Indeterminate, that is, that the class only failed to provide a justification for a unique outcome in some circumscribed class of cases. Most, but not all, realists were concerned with appellate litigation, and with the opinions of appellate courts; all confined themselves to cases that were actually litigated before courts at some level. Thus, Llewellyn explicitly qualified his defense of the Indeterminacy of law by saying that “in any case doubtful enough to make litigation respectable the available authoritative premises … are at least two, and … the two are mutually contradictory as applied to the case at hand’ (1931, p. 1,239, emphasis added).

Now the evidential base of cases actually litigated clearly could not support the inference that the law is globally indeterminate, for it would omit all those “easy” cases in which a clear-cut legal rule dictates a result, and which, consequently, no one (typically, at least) bothers to litigate. In any event, it is far less controversial, and certainly familiar to all practicing lawyers, that the law is locally indeterminate, even if this fact conflicts both with the popular perception In the realists’ own day as well as our own. Moreover, if the law is locally Indeterminate in some or most of the cases actually litigated that still raises the troubling specter of judges deciding cases unconstrained by law. The realists, unlike many contemporary political and legal philosophers, were not concerned with this Issue, and in some respects even endorsed the practice unguardedly (see Cohen, 1935).

Causal Indeterminacy

All the realists make the point that the law (as the putative cause of decision) is causally indeterminate where It is rationally indeterminate. Indeed, this follows immediately given two assumptions: first, that law exercises Its causal influence through reasons; and second, assuming the background conditions obtain, that reasons cannot be the sole cause of a decision if they do not uniquely justify that decision. But if the law is rationally Indeterminate on some point, then legal reasons justify more than one decision on that point: thus we must look to additional factors to find out why the judge decided as he did. As Radin remarked, “somewhere, somehow, a judge Is Impelled to make his selection” of an outcome (1930, p. 881). And as Holmes observed more than 30 years before: “You can give any conclusion a logical form. You always can imply a condition In a contract. But why do you Imply It?’ Holmes Is quick with an answer: the basis for the decision is to be found in ‘a concealed, half-conscious battle on the [background] question of legislative policy’ (1897, pp. 465-6, 467; cf. Llewellyn, 193 1, p. 1,252). Other

267

realists, as we shall see shortly, looked to an array of psychological and sociological factors as the real causal determinants of decision.

Some realists made the point that the law was causally indeterminate with respect to how a court would rule on a particular dispute precisely because the background conditions often do not obtain (for example, Frank, 1931, p. 240). This raises very different sorts of questions about the legitimacy of the adjudicators process, and to the extent it is true (and it often Is), It Is of considerable importance to lawyers and litigants.

But some realists held a further, more startling thesis: that the law is causally indeterminate even where it is rationally determinate and the background conditions obtain – precisely because reasons per se are causally inefficacious This view Is clearest in Moore, who took most seriously the naturalistic imperative to make jurisprudential theorizing continuous with empirical inquiry in the social sciences (see his 1923, 1929, 1943). For Moore, the relevant science was (usually) psychology, In particular Watsonian behaviorism, which viewed human beings, like rats and dogs, as complex stimulus-response machines.

For the behaviorist, the content of the mind is a black box, not to be Invoked in explaining behavior. On this view, reasons are causally relevant only as certain types of (aural, visual) stimuli, but are not causally relevant In virtue of their rational content or meaning (“[A] proposition of law,” says Moore “is nothing more than a sensible object which may arouse a drive and cue a response” (1943. p. 3).) So, for the behaviorist, the fact that reasons justify a decision Is not a causally significant fact, because justification involves a relation between the rational content of different propositions, and such content is off-limits for the be. The law Is causaIndeterminate on this picture even when it is rationally determinate, because rational determinacy (that is, justification via rational contents) is causally irrelevant for the behaviorist. Thus, Moore can say that the “logical processes of the institution of law. . . throw[ 1 no light on any pertinent question as to what the institution has been, Is or will be” (1923, p. 611) ~ that is, such rational processes make no causal difference. Similarly, Moore complains that to move ” [from necessary logical deduction to necessary behavior” is “an easy step” but a misstep, presumably because rational determinacy does not entail causal determinacy for the behaviorist (1929, p. 704). And even Llewellyn observed that for Moore’s account of adjudication, “all reference to the actor’s own ideas Is deprecated or excluded” (193 1, p. 1,245). This final thesis about Indeterminacy is not only the most unfamiliar but also the least Important for purposes here.

Notice now that the realist argument for Indeterminacy turns on a conception of what constitute legitimate members of the class, that Is, what count as legitimate legal reasons. This, of course, Is just to presuppose some view of the criteria of legality, and it is a deficiency of realist jurisprudence that It has no explicit theory on this score. Yet when Holmes chalks up judicial decision not to law but to a half conscious judgment of policy, he is plainly presuming that such considerations of policy are not legitimate sources of law. And in demonstrating the indeterminacy of law by concentrating on indeterminacy in the Interpretation of statutes and precedents the rest seem to be supposing that these exhaust the authoritative sources

268

of law, a thesis easiest to justify On Positivist grounds. Indeed, Llewellyn even says at one point that judges take rules “in the main from authoritative sources (which In the case of the law are largely statutes and the decisions of the courts)” (1930, p. 13). The realists did not develop a conceptual analysis of “law,” but it appears they may actually need the positivist analysis!

Finally, we are now in a position to see the motivation for replacement naturalism In jurisprudence. Recall that Quine had argued for replacement naturalism as follows: the central concern of epistemology is the relationship between evidence (input) and theory (output), if a normative, foundational account of this relationship is unrealizable (because, for example, evidence underdetermines theory), then there is only one fruitful account of this relationship to be given: namely, the purely descriptive. causal account given by the science of human cognition.

But now we can see that the realists have made the very same argument about the theory of adjudication. Its central concern is the relation between facts, rules of law, and legal reasoning (that Is, the class) – the “Input” – and judicial decision – the ,,output.” If the law is always rationally indeterminate, then no normative, foundational account of this relationship is possible; there is no normative, foundational relationship between the class and a particular decision, because the class Is able to justify more than one decision (that is, it underdetermines any particular decision). Given the failure of the traditional normative project, the realists propose seeking a fruitful descriptive account of what input causes what decisions, by subsuming the theory of adjudication within a scientific account of judicial behavior.

Descriptive theory of adjudication

The realists, as philosophical naturalists, sought to make their theorizing about adjudication continuous with scientific inquiry. While all the realists were scientific in attitude and method, only Moore pursued the link with social-scientific Inquiry systematically. Indeed, Llewellyn aptly described Moore’s position as “semibehaviorist, via cultural anthropology” (1931, p. 1,243 n. 50). Note, however, that the model of psychology, anthropology and social science at work in Moore and in realism more generally is positivistic (in the scientific, not legal, sense), not hermeneutic: we seek to study human (or judicial) behavior as we study the rest of the natural world, relying on detached observation in order to formulate causal laws. The hallmark of the naturalistic impulse in realism is this attempt to “formulat[el laws of judicial behavior” (Moore and Hope, 1929, p. 704) based on actual observation of what it Is courts do in particular cases. As Cook put it: legal scholars must eschew a prior! methods and “observe concrete phenomena first and … form generalizations afterwards” (1924, p. 460).

The central proposition that issued from this inquiry is what we may call the “core claim’ of realism: in deciding cases, judges respond primarily to the stimulus of the facts of the case. Observation of court decisions, in other words, shows that judges are deciding based on their response to the facts of the case – what they think would be “right” or “fair” on these facts – rather than because of legal rules

269

and reasons. (Recall that because of the rational indeterminacy of law, judges can justify post hoc the decision that strikes them as “fair” on the facts.) The challenge for at least some realists was to correlate facts (“input”) with decisions (‘output”), in order to “observ[el and stat[el the causal relation between past and future decisions” (Moore and Sussman, 1931, p. 560).

The core claim is stated neatly by Oliphant: judges “respond to the stimulus of the facts in the concrete case before them rather than to the stimulus of over-general and outward abstractions In opinions and treatises” (1928, p. 75). Similarly, Llewellyn cautions that, in looking at the pronouncements of appellate courts, one must understand “how far the proposition which seems so abstract has roots in what seems to be the due thing on the facts before the court’ (1930, p. 33). Later Llewellyn would speak of ‘the fact-pressures of the case” (193 1, p. 1,243; cf. 1960, p. 122) and “the sense of the situation as seen by the court’ as determining the outcome (1960, p. 397). Max Radin suggested that the decision of a judge was determined by ‘a type situation that has somehow been early called up In his mind” (1925, p. 362), where “” situations” were simply “the standard transactions with their regulatory incidents [which] are familiar ones to [the judge] because of his experience as a citizen and a lawyer” (1925, p. 358) (for example: “the situation of a person bargaining for actual wares, agreeing to pay a certain amount for them and carrying them off on a promise to pay at a future time, Is a common situation” (1925, p. 357)).

Federal District Court judge, Joseph Hutcheson, affirmed that “the vital, motivating impulse for the decision Is an Intuitive sense of what Is right or wrong for that cause” (1929, p. 285). Frank cited ‘a great American judge, Chancellor Kent” who confessed that, “He first made himself ‘master of the facts.’ Then (he wrote) ‘I saw where justice lay, and the moral sense decided the court half the time; I then sat down to search the authorities … but I almost always found principles suited to my view of the case”‘ (1930, p. 104 n.). The same view of judging Is presupposed In Llewellyn’s advice to lawyers that, while they must provide the court ‘a technical ladder’ justifying the result, what the lawyer must really do is ‘on the facts … persuade the court your case Is sound’ (Llewellyn, 1930. p. 76). As Frank pointed out, the very same advice had been offered by a former president of the American Bar Association (Frank, 1930, pp. 102-3 n.). It Is no small virtue of the realists’ core claim that it constitutes what every practicing lawyer knows.

Notice that the core claim forms the crux of the realists’ notorious ‘rule skepticism” (Hart, 1961, pp. 132-44). The realists were skeptical not about the existence or conceptual coherence of rules, but about whether rules make any significant causal difference In judicial decision making. It Is what judges think would be “right” or “fair” on the facts of the case – and not legal rules – that generally determines thecourse of decision according to the realist. (It Is striking. too, that on this central issue Hart can do no better than assert what the realists deny: “it is surely evident that for the most part decisions … are reached either by genuine effort to conform to rules consciously taken as guiding standards of decision or, If intuitively reached, are justifed by rules which the judge was antecedently disposed to observe and whose relevance to the case in hand would be generally

270

acknowledged” (1961, p. 137, emphasis added). The real dispute between Hart and realism, then, Is not conceptual, but empirical: it concerns how often rules do or do not matter (causally) in adjudication.)

While all realists accepted the core claim, they divided sharply, however, over the Issue of what determines how judges respond to the facts. One wing of realism, represented by Frank and Hutcheson, held that what determines the judge’s response to the facts of a particular case are idiosyncratic facts about the psychology or personality of the individual judge (the “idiosyncrasy wlng’). Another wing of realism, represented especially by Llewedyn and Moore, held that judicial response to the facts was “socially” determined such that these responses fall into very particular patterns, making generalization and prediction possible (the “sociological wing”). Over time, there has been a gradual “Frankification’ of realism, with the views of the idiosyncrasy wing coming to stand for realism itself, but, in fact. prominent realists had a very different view about the determinants of judicial decision.

The Idiosyncrasy Wing

Judge Hutcheson’s confession that he reached his decisions by getting a “hunch” about what would be right or fair on the facts of a given case (1929, p. 278) provided the foundation for the idiosyncrasy wing of realism. Frank specifically endorsed Hutcheson’s view and declared: “the way in which the judge gets his hunches Is the key to the judicial process” (1930, p. 104). While conceding that “rules and principles are one class of… stimuli’ producing decisions, Frank claimed that it Is ‘the judge’s Innumerable unique traits, dispositions and habits” which are decisive, which ‘shape his decisions not only In his determination of what he thinks fair or just with reference to a given set of facts, but in the very process by which he becomes convinced what those facts are” (1930, pp. 1 10~1 1). Political and economic biases, often thought to be important in adjudication, in fact only .express themselves in connection with, and as modified by. these Idiosyncratic biases’ (1930, p. 106). Thus, concludes Frank, “the personality of the judge is the pivotal factor In law administration” (1930, p. 1 1 1; cf. 193 1, p. 242). (Note, however, that no one. In the idiosyncrasy wing adhered to the view, often wrongly attributed to realism, that “what the judge ate for breakfast” determines the decision.)

But if ‘the ultimately Important influences in the decisions of any judge are the most obscure, and are the least easily discoverable” (1930, p. 114) precisely because they are these idiosyncratic facts about the psychology of the individual judge, then how will it be possible to formulate laws of judicial behavior? In fact, Frank repudiates this ambition of the behaviorists when he declares that, ‘The truth is that prediction of most specific decisions … is, today at any rate, impossible’ (1931, p. 246; cf. 1930, passim). The qualification (‘today at any rate’) is important: for the problem Is not that there are no determinants of decision (Frank does accept a type of Freudian psychic determinism) but rather that such determinants are epistemologically opaque: we have no reliable way of knowing what they are. But even Frank allowed that If judges became suitably self-aware – undergoing, say, psychoanalysis – then they could provide us with the Information about their personalities that would make prediction possible (1930, p. 163).

Even if later images of realism seem to conform to the Frankian model, Frank himself was aware that his was a minority view, both among lawyers and among realists. Fond of armchair Freudian speculations, Frank charged that the continued demand for certainty (via predictability) in law was the product of an infantile longing for the protection of a father-figure (1930, p. 34). More interestingly, he suggested that the reason he and Hutcheson had “far less belief in the possibility of diminishing the personal element In the judge than Oliphant or Liewellyn” was due to his and Hutcheson’s greater experience with trial courts, where, Frank suggested, the personal element was oninipresent (1931, p. 30, n. 31).

The Sociological Wing

Writers in the sociological wing of realism did not deny the relevance of the fact that judges are human beings with individual personalities (see Llewellyn, 1930, pp. 80-1, 1931, pp. 1,242-3); rather they insisted that the relevant facts about judges qua human beings were not primarily Idiosyncratic ones. As Cohen aptly put it (in answer to Hutcheson and Frank): “Judges are human, but they are a peculiar breed of human, selected to a type and held to service under a potent system of government controls … A truly realistic theory of judicial decision must conceive every decision as something more than an expression of individual personality, as … even more importantly … a product of social determinants” (1 9 3 5, p. 843). If writers like Frank emphasized the psychological profile of the individual judge, writers like Lleweflyn, Cohen and Moore emphasized the “sociological” profile of the judge, one he had in common with many others. judicial decision is still primarily explicable in terms of psycho-social facts about judges (that determine how they respond to the facts of particular cases), it is just that these psycho-social facts are held to be general and common, rather than idiosyncratic.

Unfortunately, the realists did not have a rich sociological theory of judicial personality. Their strongest argument was an inference to the best explanation: given that judicial decisions can be correlated with the underlying facts of the cases decided, it must be the case that there are “social” determinants of decision that force decisions of individual judges into these predictable patterns.

The realists tended to draw their best examples of this point from the commercial realm (rather, say, than constitutional law). Here they commonly advanced two sorts of claims: with respect to the underlying facts of the case (whether stated in the opinion or not), what judges do is either (1) enforce the norms of the prevailing commercial culture; or (2) do what is best socio-economically under the circumstances.

Oliphant gives this example: looking at a series of conflicting court decisions on the validity of contractual promises not to compete, Oliphant observed that, in fact, the decisions tracked the underlying facts of the cases:

All the cases holding the promises invalid are found to be cases of employees’ promises not to compete with their employers after a term of employment. Contemporary guild [that is, labor union] regulations not noticed in the opinions made their holding eminently sound. All the cases holding the promises valid were cases of promises by those selling a business and promising not to compete with the purchasers. Contemporary economic reality made these holdings eminently sound. (1 928, pp. 15 960)

Thus, In the former fact-scenarios, the courts enforced the prevailing norms (as expressed In guild regulations disfavoring such promises); in the latter cases, the courts came out differently because it was economically best under those factual circumstances to do so.

Llewellyn provides a similar illustration (1960, pp. 122-4). A series of New York cases applied the rule that a buyer who rejects the seller’s shipment by formally stating his objections thereby waives all other objections. Llewellyn notes that the rule seems to have been rather harshly applied in a series of cases where the buyers simply may not have known at the time of rejection of other do where the seller could not hcured anyway. A careful study of the facts of these cases revealed, however, that in each case where the rule seemed harshly applied what had really happened was that the market had gone sour, and the buyer was looking to escape the contract. The court in each case, being “sensitive to commerce or to decency” (1960, p. 124) applies the unrelated rule about rejection to frustrate the buyer’s attempt to escape the contract. Thus, the commercial norm – buyers ought to honor their commitments even under changed market conditions – is enforced by the courts through a seemingly harsh application of an unrelated rule concerning rejection. It Is these ‘background facts, those of mercantile practice, those of the situation-type” (1960, p. 126) which determine the course of decision.

Moore tried to systematize this approach as what he called “the institutional method’ (1929, 1931). Moore’s idea was this: identify the normal behavior for any “Institution” (for example, commercial banking); then identify and demarcate deviations from this norm quantitatively, and try to Identify the point at which deviation from the norm will cause a judicial decision that corrects the deviation from the norm (for example, how far must a bank depart from normal checkcashing practice before a court will decide against the bank in a suit brought by the customer?). The goal is a predictive formula: deviation of degree X from “Institutional behavior (that Is, behavior which frequently, repeatedly, usually occurs)’ (1929. p. 707) will cause courts to act. Thus, says, Moore: “the semblance of causal relation between future and past decisions Is the result of the relation of both to a third variable, the relevant Institutions in the locality of the court” (1931, p. 1,219). Put differently: what judges respond to Is the extent to which the facts show a deviation from the prevailing norm in the commercial culture.

The theory of the sociological wing of realism – that judges enforce the norms of commercial culture or try to do what is socio-economically best on the facts of the case – should not be confused with the idea that judges decide based. for example, on how they feel about the particular parties or the lawyers. These ‘fireside equities” (Llewellyn, 1960, p. 12 1) may sometimes influence judges; but what really determines the course of decision is the “situation-type”, that is, the general pattern of behavior exemplified by the particular facts of the disputed transaction, and what would constitute normal or socio-economically desirable behavior in the relevant commercial context. The point is decidedly not that judges usually decide because of idiosyncratic likes and dislikes with respect to the individuals before the court (cf. Radin, 1925, p. 357).

But why would judges, with some degree of predictable uniformity, enforce the norms of commercial culture as applied to the underlying facts of the case? Here the realists did little more than gesture at a suitable psycho-social explanation. “Professional judicial office,’ Llewellyn suggested, was “the most important among all the lines of factor which make for reckonability” of decision (1 960, p. 45); “theoffice waits and then moves with majestic power to shape the man” (1960, p. 46). Echoing, but modifying Frank, LleweUyn continued:

The place to begin Is with the fact that the men of our appellate bench are human beings … And one of the more obvious and obstinate facts about human beings Is that they operate in and respond to traditions … Tradition grips them, shapes them, limits them, guides them … To a man of sociology or psychology … this needs no argument. (1960. p. 53)

Radin suggested that “the standard transactions with their regulatory incidents are familiar ones to him [the judge] because of his experience as a citizen and a lawyer” (1925, p. 358). Cohen, by contrast, simply lamented that “at present no publication [exists] showing the political, economic, and professional background and activities of our judges’ (1935, p. 846), presumably because such a publication would identify the relevant ‘social” deterrminants of decision.

Of course, by the time of The Common Law Tradition, Llewellyn had actually repudiated many of the naturalistic ambitions of early realism, remarking, for example – and with Moore obviously In mind – that the judge is not a “Pavlov’s dog’ (1960, p. 204). The final collapse of naturalism in realist jurisprudence comes with Ilewellyn’s Introduction of ‘situation-sense,” a mysterious faculty that permits judges to detect the “natural law which is real, not imaginary … [that Is] indwelling in the very circumstances of life’ (1960, p. 122, quoting Levin Goldschmidt). Plainly, though, the ambition of making theories of adjudication continuous with social scientific inquiry has been abandoned in favor of rank mysticism when the explanation for the correlation of decisions with underlying facts is the operation of a non-naturalistic faculty, ‘situation-sense”I (But for a different understanding of later Llewellyn, see Kromnan, 1993, pp. 209-25.)

One final difficulty may seem to plague the realists’ descriptive theory. For surely It is obvious that some cases that come before courts are easy (the rules clearly dictate a certain outcome), that judges often appear to strive to conform to the demands of rules, and that judges often decide in ways that are consistent with rules being causes of decision. Call these phenomena “the rule truisms.” How can the core claim of realism be compatible with the rule truisms.

The key here Is to remember that the central naturalistic commitment of realism is to explain judicial decision in terms of the psycho-social facts about judges that account for how they make decisions. But judges, as the sociological wing emphasizes, are a special breed of human being, and this too counts as a relevant psycho-social fact. ‘Judges,” says Cohen, “are craftsmen, with aesthetic ideals” (1935, p. 845) ~ surely a relevant fact about the psychological profile of the judge. And Llewellyn concedes that In trying to get a court to decide in your favor, 11 rules’ loom Into Importance. Great importance. For judges think that they must follow rules, and people highly approve of that thinking” (1930, p. 4). judges, In short, are guided according to the realists by “a certain ideal of judicial craftmanship’ (Kronman, 1993, p. 214). Let us call this the idea of “the normative judiciary” – of how judges ought ideally to decide cases. To the extent that human beings qua judges have a conception of the normative judiciary, then to that extent the psychological fact about them explains why they are sometimes rule-responsive in the way the rule truisms suggest. Theoretical coherence, a virtue not much prized by the realists, is, nonetheless, preserved in the face of the rule truisms – by showing that even “rule-responsiveness,” however infrequent, is, nonetheless, explicable within a naturalistic account of judicial decision.

The Attack on Formalism

Whatever their differences among themselves, the realists were united in their opposition to a very different descriptive theory of adjudication, often called “formalism” or “mechanical jurisprudence.” According to the formalist, “the judge begins with some rule or principle of law as his premise, applies this premise to the facts, and thus arrives at his decision” (Frank, 1930, p. 101). judges, of course, write their opinions in this “formalistic” mode, but the realists want to Insist precisely that, “the decision often may and often will prove to be inadequate if taken as a description of how the decision really came about and of what the vital factors were which caused It’ (Llewellyn, 1930, p. 3 7; cf. Cohen: “The traditional language of argument and opinion neither explains nor justifies court decisions” (1935, p. 812)).

“Formalism” as an epithet has actually been widely applied. In a very strict sense, the formalist holds that decision flow (or ought to flow) from certain axiomatic definitions. Thus, In the notorious (formalistic) opinion in United States v. E. C. Knight Co. (1 5 6 U.S. 1 (1 8 9 5)), the US Supreme Couheld that the regulation of a sugar manufacturer (responsible for 90 percent of sugar production in the United States) was not within the power of Congress to regulate “interstate commerce,” since, by definition, interstate commerce did not include manufacturing, which takes place only within a state.

In a looser sense, formalism names any view in which authoritative legal sources

275

together with the “methods” of legal reasoning are sufficient to provide a watertight justification for a unique outcome to any dispute. The deductive model of decision just described by Frank is the most familiar form.

The realists were intent to deny the descriptive adequacy of formalism on both counts. We shall return momentarily to the realist attitude toward formalism as a normative theory.

Formalism owed its intellectual underpinning to the work of Christopher Langdell, Dean of Harvard Law School in the late nineteenth century, who, along with certain followers (like joseph Beale), was a figure for whom the realists reserved a special antipathy. Langdell aimed to make law “scientific” in a different sense than the realists. As one commentator explains:

To understand a given branch of legal doctrine in a scientific fashion, one must begin … by first identifying the elementary principles on which that field of law is based (for example, in the case of contract law, the principles that the minds of the parties must meet for a contract to be formed and that each must give or promise to give something of value to the other in return). These elementary principles are to be discovered by surveying the case law in the area. Once they have been identified, it is then the task of scholars to work out, in an analytically rigorous manner. the subordinate principles entailed by them. When these subordinate principles have all been stated in prepositional form and the relations of entailment among them clarifled, they will … together constitute a well-ordered system of rules that offers the best possible description of that particular branch of law – the best answer to the question of what the law in that area is … [individual cases that cannot be fit within this system must be rejected as mistakes.(Kronmati, 199.3, p. 171; for an eloquent discussion of Langdell and realism, see generally pp. 1 70-99)

The realists, as we have seen, rejected this picture wholesale. In particular, they denied that the sort of categories adduced by the Langdellian scholar were really descriptive of the bases of decision. Notice, however, that the realists repudiate primarily the methodology, not the aspiration, of Langdell: they object to Langdell’s notion that decisions track abstract principles of law, rather than particular patterns of facts (cL Oliphant, 1928; Llewellyn, 1931, p. 1,240). The mistake of Langdell is in thinking we can learn the law, in the sense the lawyer needs to learn it (that is, in order to be able to predict what courts will do), by examining the opinions and the reasons given therein; rather, the realists qua naturalists discover that decisions track underlying fact patterns (“situation-types”), not published rationalizations (or the “axioms” to be adduced from them).

Normative Theory of Adjudication

All realists share a commitment to the core claim: in deciding cases, judges respond primarily to the stimulus of the facts. If this is a true descriptive thesis about adjudication, what room does it leave for anormative theory of adjudication? The realists are not always clear on this issue, but the majority of them endorse a type of view we may call “quietism.”

276

Quietists hold that since the core claim reports some irremediable fact about judging, it makes no sense to give normative advice – except perhaps the advice that judges “ought” to do what it is that they will do anyway. So if judges, as a matter of course, enforce the norms of commercial culture or try to do what is socio-economically best under the circumstances, then that is precisely what realists tell them they ought to do.

Thus, Holmes complains that “judges themselves have failed adequately to recognize their duty of [explicitly] weighing considerations of social advantage.” But having just noted that what is really going on in the opinion of judges anyway is “a concealed, half-conscious battle on the question of legislative policy,” it follows that this “duty” is in fact “Inevitable, and the result of the often proclaimed judicial aversion to deal with such considerations is simply to leave the very ground and foundation of judgments inarticulate, and often unconscious” (1 897, p. 467). Thus, what Holmes really calls for is for judges to do explicitly (and perhaps more carefully) what they do unconsciously anyway. (Contrast the non-quictistic Cohen (1935, p. 810), who recommends that judges address themselves to questions of socio-economic policy instead of the traditional doctrinal questions that they often address.)

In a similar vein, Radin suggested that the decisions judges make on the basis of the type-situations into which they put facts essentially track the sorts of decisions one would get by demanding explicitly that judges do the “economically or socially valuable thing” (1925, p. 360). Frank observed that with respect to what he dubbed “Cad! justice” – justice by personal predilection essentially – “[t]he true question. . .is not whether we should ‘revert’ to [it], but whether (a) we have ever abandoned it and (b) we can ever pass beyond it” (1 9 3 1, p. 2 7). Advocating a 111 reversion to Cadi justice’ . . . is as meaningless as [advocating] a ‘reversion to mortality’ or a ‘return to breathing”‘ (1 9 3 1, p. 3 1). This is because “the personal element is unavoidable in judicial decisions” (1 9 3 1, p. 2 5).

The most important example of normative quietism in realism comes from Liewellyn’s work on Article 2 of the Uniform Commercial Code. For how can a realist, one might wonder, tackle the enterprise of designing rules for what ought to be done in commercial disputes? The answer should be obvious: tell judges that they ought to do what it is they will do anyway, that is, enforce the norms of commercial culture, of the prevailing mercantile practice. Thus, the Code imposes an obligation of “good faith” in all contractual dealings (Sec. 1-203) which means besides honesty. “the observance of reasonable commercial standards of fair dealing in the trade” (See. 2-103). But for a court, then, to enforce the rule requiring “good faith” is just for that course to enforce the relevant norms of commercial culturel The reliance of the Code throughout on norms of “good faith” and “reasonableness” is a constant invitation to the judge to do what he would, on the realist theory, do anyway: enforce the norms of the prevailing commercial practice.

A final worry might arise about this normative “program.” For the motivation for the quietism is the thought that since judges will decide in accordance with the core claim anyway, it would be futile or idle to tell them they “ought” to decide in some other way. But why think judges must be fact-responsive? One possibility is that the core claim of realism is supposed to report a brute psychological fact about human judgment: indeed, Frank often presents it that way (1930, p. 100). A perhaps more plausible hypothesis is that the core claim is inevitably true of common-law judges, who, by role and tradition, are invited to examine and rework the law In ways that are responsive to the changing circumstances In which legal problems arise. Thus, it is not because of a “deep” wired-in fact about the human psyche that the core claim is true, but rather a contingent, but still obstinate, fact about adjudication in the common-law system. Indeed, we may detect a further, non-quietistic, normative element in realism: namely, to the extent, however small, that judges are not fact-responsive and fairness-driven in their decisions (to the extent, for example, that they are sometimes formalistic or Langdeflian in their mode of decision), then to that extent they ought to decide as the core claim says mof them ordinarily do. This additi. more ambitious. normative demand does not, however, receive the sustained defense one would hope to find.

Other Themes from Realism

Writers often associated with realism have been the source of other intellectual themes that have recently overshadowed the distinctive realist contributions in philosophy of law. Primary among these is the purported argument against the public/private distinction generally attributed to the economist Robert Hale and the philosopher Morris Cohen (both contemporaries of the realists) (see the selections in Fisher et al., 1993; cf. Llewellyn, 1 9 3 0, p. 10: Cohen, 1935. p.816), and brought to prominence by the critical legal studies (CLS) movement in the 1970s and 1980s (see the introduction to chapter 4 in Fisher, 1993, a volume which generally views realism through a CLS lens). The argument runs as follows: since it is governmental decisions that create and structure the so-called “private” sphere (that is, by creating and enforcing a regime of property rights), there should be no presumption of “non-intervention” in this “private” realm (for example, the marketplace) because it is, in essence, a public creature. There is, in short, no natural baseline against which government cannot pass without becoming “Interventionist” and non-neutral, because the baseline itself is an artifact of government regulation.

This general argument has been widely influential (see, for example, Sunstein, 1987, pp. 917-19): unfortunately, it is based on a non sequitur. From the fact that a “private” realm is a creature of government regulation it does not follow that government action in that realm is normatively indistinguishable from government action in the “public” realm: for the key issue is the normative justification for drawing the baseline itself, not simply the fact that one has been drawn by an exercise of public power. If the underlying normative reasons for the baseline are sound (that is, for demarcating a realm of “private” transactions), then these reasons provide an argument against intervention. Hale and many of his contemporary followers are, of course, correct that many proponents of the baseline (wrongly)

278

regard its existence per se as a reason against government action (cf. Sunstein, 1987); but this is simply to repeat in reverse the non sequitur of those who think the regulability of the “private” sphere follows from recognizing that its very existence depends on public regulation.

References

Cohen, F. 1935: Transcendental nonsense and the functional approach. Columbia Law Review, 35, 809-49.

Cook, W. W. 1924: The logical and legal bases of the conflict of laws. Yale law journal, 3 3, 457-88.

Frank, 1. 1930: Law and the Modem Mind. New York: Brentano’s.

– 19 3 1: Are judges human? Parts 1 & 11. University of Pennsylvania Law Review, 80, 1 753, 233-67.

Holmes, jr, 0. W. 1897: The path of the law. Harvard Law Review, 10, 457-78.

Hutcheson, jr, J. 1929: The judgment intuitive: the function of the “hunch” in iudicial decision. Cornell Law Quarterly, 14, 274-88.

Llewellyn, K. 1930: The Bramble Bush. New York: Oceana.

– 19 3 1: Some realism about realism – responding to Dean Pound. Harvard Law Review, 44, 1,222-64.

– 1950: Remarks on the theory of appellate decision and the rules and canons about how statutes are to be construed. Vanderbilt Law Review, 3, 395-406.

– 1960: The Common Law Tradition: deciding appeals. Boston: Little, Brown & Co.

Moore, U. 1923: Rational basis of legal institutions. Columbia Law Review, 23, 609-17. Moore, U. and Hope, T. 1929: An institutional approach to the law of commercial banking.

Yale Law journal, 38, 703-19.

Moore, U. and Sussman, G. 1931: Legal and institutional methods applied to the debiting of direct discounts – I. Legal method: banker’s set-oft, 11. Institutional method: VI. The decisions, the institutions, and the degree of deviation. Yale Law journal, 40, 381-400, 555-75, 1.219-50.

Moore, U. and Callahan, C. 1943: Law and learning theory: a study in legal control. Yale Law Journal, 5 3, 1-1 3 6.

Oliphant. H. 1928: A return to stare decisis. American Bar Association journal, 14, 71-6,

107,159-62.

Radin, M. 192 5: The theory of judicial decision: or how judges think. American Bar Association journal, 1 1. 3 5 7- 6 2.

– 1930: Statutory interpretation. Hartard Law Review, 43, 863-85.

Bibliography

Fisher. W. W. et al. (eds) 19 9 3: American Legal Realism. New York: Oxford University Press.

Hart. H. L. A. 1961: The Concept of Law. Oxford: Clarendon Press.

Komblith, H. (ed.) 1994: Naturalizing Ppistemology. Cambridge, Mass.: MIT Press, 2nd edn.

Kronman, A. 199 3: The Lost Lawyer. Cambridge, Mass.: Harvard University Press.

Leiter, B. 1995: Legal indeterminacy. Legal Theory, 1.

Perry, S. 1995: Interpretation and methodology in legal theory. In A. Marinor (ed.), Law and Interpretation. Oxford: Oxford University Press.

Sunstein, C. 1987: Lochner’s Legacy. Columbia Law Review, 87, 873-919.