II. The Formal Rationality of Judging

Equating rationality with the logical reasoning that produces correct answers has a long, august intellectual history, and legal realism, legal positivism, and postmodernism cannot be uncoupled from that tradition. Implicit in and common to the weaknesses of those theories of law is a familiar conception of rationality, the account of rationality that has formed Western analytic philosophy and Western thought for centuries. The bedrock assumption of analytic philosophy is that rationality is formal reason, the kind of inferential reasoning used in the paradigmatically rational domains of logic and mathematics. This assumption percolates through common understandings of rationality as well, and it is exemplified by the familiar use of the expression “That’s not logical” to mean “That’s not rational.” Moreover, analytic philosophy’s defining project of theory construction and demarcation, which in philosophy of law assimilates law to a theory of law, proceeds from this conception of rationality. Principles and rules are, along with logical relations, the elements of reasoning. Logical relations within and among those propositions plus inferential principles are needed to construct theories, classically by axiomatization, and to justify and apply those theories.

Yet when rationality is equated with the rule-constituted and rule-governed reasoning of logic, the limitations of logic become the limitations of rationality. Rationality runs out as quickly as logic runs out. Law, science, and ethics, as well as many other domains, would not be rational. Judicial decision-making would quickly go beyond formal reason. When the law consists of rules, “applying the law” becomes the logical maneuver of subsuming the facts of cases under rules to deduce decisions. As long as judges do just that, they are not pretending to apply rules. In logic, for example, constructing a truth table to determine whether an argument is valid requires only following the rules, not any judgment. In logic and mathematics, deduction produces correct answers; those conclusions that, given true premises, are necessarily, and therefore universally, true. Were that same exercise possible in law, it would produce correct judicial decisions. But applying rules to facts is not straightforward. Because real legal problems are messy, amorphous, complex, and dynamic, much work needs to be done before facts can be subsumed under a rule. The rule-application of formal reason assumes that the facts of legal problems can be adequately characterized and well formed; that relevant rules exist and can be readily identified; that the applicable rules are sufficiently robust and determinate to be helpful; and that the inevitable conflicts between rules can be resolved by applying higher rules. But the law and the legal theories proposed to represent the law lack the power and clarity needed to specify the facts of problems properly and productively and to ascertain the relevance of substantive rules to those problems; the ancillary bridging needed to engage the general concepts of its substantive rules with the particular facts of situations; and the higher-order rules needed to settle conflicts, as Karl Llewellyn’s famous long pairs of thrust and counter-thrust of canons of statutory construction vividly demonstrate.Footnote ⁴ The resources of both law and law-as-a-theory-of-law are missing when and where they are needed most.

Bereft of the panoply of rules needed to justify decisions in the logical terms prescribed by formal reason, judges have to pretend to be applying law or to confess that they are making law. So much for the rule of law. That regrettable dichotomy—judges either objectively apply the law or subjectively make the law—is imposed by the narrowness of formal reason. Logic runs out, and, for formal reason, when logic runs out, rationality perforce runs out.

What ensues is a jurisprudential salvage operation to delimit judicial indeterminacy. Leiter attributes that undertaking to Hart when he argues that the “central strategic move” in the chapter on formalism and rule skepticism in Hart’s classic text, The Concept of Law, is “to concede to the skeptic” but “to argue that this indeterminacy is a marginal phenomenon,” so that the skeptic can be “portrayed … as having unrealistically high expectations for the determinacy of rules, as being ‘a disappointed absolutist’ ….”Footnote ⁵ Legal realists largely adopt the different strategy of relegating indeterminacy to appellate courts.Footnote ⁶ They emphasize the problems of specifying the general terms in rules and resolving conflicts between rules and pay less attention to the manifold difficulties judges in lower courts encounter in framing problems and assessing the relevance of rules and facts.

Preserving the legitimacy of judging and the integrity of judges requires more than the stopgap containment of indeterminacy. It requires a more expansive conception of rationality that proceeds from two recognitions: that indeterminacy results from the limitations of logical reasoning, and that judgment is inevitable when judges use logical reasoning. Before turning to formal reason and non-formal reason, however, an important preliminary issue needs to be addressed. The charge of attacking a straw man, in the guise of what Leiter calls “Vulgar Formalism”, has to be repudiated.Footnote ⁷ Vulgar Formalism is “the idea that judicial decision-making involves nothing more than mechanical deduction on the model of the syllogism,” which, Leiter rightly emphasizes, “is not a view to which anyone today cares to subscribe.”Footnote ⁸ Sophisticated Formalists appreciate that “legal reasoning is not mechanical, that it demands the identification of valid sources of law, the interpretation of those sources, the distinguishing of sources that are relevant and irrelevant, and so on, and they offer a theoretical account of how these various bits of reasoning are done ‘rightly’.”Footnote ⁹ One tenet of formalism common to both Vulgar and Sophisticated Formalism, however, is that “the law is ‘rationally’ determinate,” which means that “the class of legitimate legal reasons … justifies one and only one outcome ….”Footnote ¹⁰ Leiter’s obvious suspicion of rationality is understandable because it is hard to see how doing “bits of reasoning … ‘rightly’” will produce “one and only one outcome.” Nevertheless, his Sophisticated Formalists need an account of rationality to support their own formalist doctrine that the law is “rationally determinate.” And where is the rationality of all those “bits of reasoning” that are done “rightly”? How do the Sophisticated Formalists’ conceive rationality?

Setting aside Vulgar Formalism, there are three normative accounts of judging. First, judging is not rational. That is the skepticism of legal realism, Critical Legal Studies, and versions of postmodernism. Second, a normative theory renders judging rational. That is Sophisticated Formalism, which adds theory to the rules of the law. Either rules are systematized into a theory, or an external theory is imported into the law. Leiter cites Ronald Dworkin as the “leading theoretical spokesman” of the former,Footnote ¹¹ and economic analysis of the law is the current paradigm of the latter. But such valiant attempts to achieve non-deductive rationality by expanding the array of substantive considerations and figuring out how to reason “rightly,” e. g., by moving from rules to principles and then to theories and by switching from foundationalism to coherentism, are futile because they simply shift the loci of indeterminacy. That is abundantly clear with respect to producing the “fit” required by coherentist theoretical approaches. The most popular of these methods, reflective equilibrium, has been criticized by D. W. Haslett for “the lack of any clear guidelines for making the many adjustment decisions that would have to be made in achieving a ‘fit’ among the various elements that go to make up a person’s equilibrium.”Footnote ¹² And that failure precludes rationality:

In the absence of “a satisfactory answer to this question by proponents of reflective equilibrium methodology, and so far we have not been given a satisfactory answer,” Haslett concludes that we can decide “only arbitrarily, or else irrationally by appeal to prejudices, intuitions, or revelations.”Footnote ¹⁴ Nor are attempts to incorporate external theories from the social sciences—nowadays economics—into the law successful because insofar as they are prescriptive, their prescriptivity pertains only to means, not to ends. Rationality is restricted to instrumental rationality, which leaves the determinations of ends subjective and arbitrary.

Consequently, theoretical brands of Sophisticated Formalism inevitably require judgment, and the need for judgment again vitiates formal reason’s claim to rationality.Footnote ¹⁵ Sophisticated Formalists have no answers to crucial questions about how matters of theory construction, theory justification, theory application, and theory assessment can be decided rationally, let alone be rationally determinate. So Leiter’s hesitation about the rationality of Sophisticated Formalism is warranted. That leaves the third position: there is a normative account of judging that renders it rational, albeit not in formal reason’s sense of being rationally determinate. Only this position has an answer to the question of how judges can decide rationally when there is no single, definitive, justified outcome. That answer is the more expansive rationality of non-formal reason.

III. Rational Deliberation

The hallmark of the Western philosophical tradition, which Socrates started, is a commitment to rationality. Rationality initiated the form of what Stephen Toulmin calls, “the Gospel According to Socrates” which emphasized “… the importance of ‘keeping the talk going’; and keeping it going in terms that did not discriminate, in principle, between the parties to any issue.”Footnote ¹⁶ Over the ensuing centuries, however, the theories propounded by philosophers crystallized that rationality into what Toulmin calls “logicality”:

Philosophers agreed that historically and culturally invariant principles had to ground a rational system because mere empirical universality could not provide the kind of authority needed to avoid relativism. They disagreed not about the nature of rationality or the nature of philosophy but about the a priori correctness of their competing principles.

The problem with this philosophical legacy is not logicality, the formal reason method it prizes and exalts, but the overweening narrowness of rationality that results when rationality is confined to this method.Footnote ¹⁸ Deducing theorems from axioms in logic or geometry is, of course, a paradigmatically rational activity, and the parallel exercise of deducing decisions from principles or rules in law, morality, and everywhere else is itself rational. But it is a mistake—in logic, law, morality, and everywhere else—to reduce rationality to this one kind of rationality. Even within formalist domains such as logic and mathematics, formal reason runs out because, e. g., it ignores the skill required to identify important theorems and construct proofs for them, neither of which can be automated (the tasks are largely formally undecidable), let alone the capacity to create relevant formal systems in the first place (cf. the many differing systems of logic now available).

Yet rationality and objectivity are features of correct answers. Logic and mathematics are exemplars of both because they produce right answers for everybody. What is the source of this correlative universality? As Harold Brown points out in his analysis of formal reason, which he calls, “a classical model of rationality,”Footnote ¹⁹ universality derives from necessity, and necessity derives from the application of rules. He explains why “rules are at the heart of our classical model of rationality:” “if we have universally applicable rules, then all who begin from the same information must indeed arrive at the same conclusion, and it is these rules that provide the necessary connection between our starting point and our conclusion.”Footnote ²⁰ The theories that philosophers construct are systems of these putatively “appropriate” and “universally applicable” rules.

Deriving conclusions from rules is simple in theory but not in practice. In addition to the problems of the indeterminacy of general terms in norms and the conflicts between norms, there is the crucial and formidable challenge of ascertaining the pertinence of norms. Is a rule relevant to a particular issue, and if so, why is it morally relevant to this particular situation? Formal reason derives resolutions from rules to establish relevance, remove indeterminacy, and settle conflicts. This approach invites a vicious regress, though, because the rules are derived from higher rules. The essential work occurs before and beyond logic, thereby beyond the rationality imparted by formal reason. For formal reason everything that precedes and structures the subsumption of facts under norms—everything that goes into establishing the givens of the subsumption—must be non-rational because it is not rule-governed. The triumphal application of norms to facts is the last step in a complex deliberative process of analysis, critical assessment, and judgment that is not rule governed. Formal reason cannot incorporate the rationality of a complex, necessary, and manifold process.

A superficial depiction of judging and the commonplace assimilation of the rule of law to the “law of rules”Footnote ²¹ render law indeterminate because they manifest the constricted rule-governed rationality of formal reason. Because law is comprised of rules, judging is the application of rules. Frederick Schauer’s analysis of argument by analogy reveals just how deeply and irrevocably rules are entrenched in the law.Footnote ²² An argument by analogy identifies a number of properties that two (or more) entities possess, and based upon these similarities, infers that the second entity is likely to have an additional property that the first entity has. Unlike the certainty conferred on the conclusions of valid deductive arguments, the conclusions of arguments by analogy possess only probability. How strong an argument by analogy is—how probable the conclusion is—depends upon an array of judgments about the relevance and importance of the similarities, as well as the relevance and importance of any posited dissimilarities. To avoid that rampant irrationality, Schauer strives to efface the need for judgment by invoking a rule that determines the relevance of the properties in the premises of an argument by analogy. Even argument by analogy, an exemplar of informal reasoning, must be converted into some instantiation of rule-governed reasoning. The hegemony of formal reason renders that temptation irresistible, if not mandatory in the law, as Hart confirms: “… what conventional legal thought in all countries conceives as the standard judicial function … [is] the impartial application of determinate existing rules of law in the settlement of disputes.”Footnote ²³

Assimilating judging to the impartial logical application of rules conveys the appearance of correct outcomes. Yet lawyers and academics concede that there are no demonstrable right decisions for at least some cases. Schauer reports, “what most legal insiders and commentators now take to be commonplace,” namely, that “[n]owadays it would be hard to find very many dissenters from the view that when judges change the law, they base their decisions on a mix of policy and principle that can hardly be thought of as a deductive or logical exercise.”Footnote ²⁴ That recondite “mix of policy and principle,” however, promotes the belief that Hart calls the “Nightmare view of adjudication”:

Legal insiders, legal commentators, and legal philosophers have indeed moved beyond deduction and logic, but they have not moved beyond the constrained rationality of deduction and logic. For that reason, they remain ensnared in the trap of indeterminacy.

Both Llewellyn and Hart spy the trap. Llewellyn is frustrated that so many of his students fall for it: “You cannot watch generations of law students assume, two thirds of them, as of course and despite all your effort, that if the outcome of an appeal is not foredoomed in logic it therefore is the product of uncontrolled will which is as good as wayward ….”Footnote ²⁶ For Hart, the trap springs when a decision falls within the penumbra of a general term or a legal rule, and he recognizes the consequence: “… if legal arguments and legal decisions of penumbral questions are to be rational, their rationality must lie in something other than a logical relation to premises.”Footnote ²⁷ But what is that elusive “something other”?

That question is important because the trap is not unique to the law. It plagues morality as well as other domains such as science, and it pervades our lives. Charles Larmore defends morality but has no explanation of how judgment can be rational: “… if moral judgment is not thoroughly rule-governed, it is not arbitrary either. Judgment certainly involves risk. Yet it does not resemble the flipping of a coin or a decisionistic leap of faith.”Footnote ²⁸ Logic is bounded, and judgment is inescapable; yet judgment is not arbitrary. How can that be? Jerome Frank identifies, in his exceptional style, the obstacle that needs to be overcome: “There can be no greater hindrance to the growth of rationality than the illusion that one is rational when one is the dupe of illusions. Man can invent no better way to balk any of his ideals than the delusion that they have already been achieved.”Footnote ²⁹ Rationality is one of our ideals, but we are duped by the illusion that rationality has to have the necessity, certainty, and universality of the formal reason of logic. So, in fidelity to the Socratic Gospel, we have to scrutinize our ideal of rationality.Footnote ³⁰

Writing about “the irrationalist fallacy” in judging, Richard Wasserstrom recognizes the mistake of confining rationality to logic and proposes a way of developing a more capacious account of rationality:

Judges make all sorts of decisions about all sorts of cases, and those decisions are judgments. A judgment is the outcome of a deliberative process or a collection of deliberative processes that, as Wasserstrom puts it, “permits of some kind of reasoned justification.” The notion of a “reasoned justification” is unfortunate, however, because it summons proofs in logic and derivations from principles in moral philosophy, which are reasoned and regarded as justifications. To avoid confusion with the orthodoxy of philosophy, when a judgment emanates from a rational process of deliberation, it should simply be described as “rational”.

IV. Non-formal Reason in Bioethics

A judgment is not an instinct, intuition, or opinion. A judgment is a decision or conclusion that is formed, developed, assessed, and adopted in a process of deliberation.Footnote ³² Judgment often is invisible because it is confused with intuition. Judgment also can be attributed to experience. Experience does not supplant deliberation, however, because the judgment that accompanies experience is formed by the rationality of the deliberative processes used in that experience. Doctors who have been practising for many years, e. g., can look at a patient and make a diagnosis that is likely to be accurate. Such a diagnosis is not an intuition. The clinical judgments of doctors have the three features of the rationality that Hart identifies, and those features also belong to non-formal reason. The diagnosis is fallible; it might be wrong. The diagnosis is normative; what the doctor ought to do for the patient depends on the diagnosis. The diagnosis does not derive from deduction; the doctor’s judgment results from a process of rational deliberation.

The goal of non-formal reason is the improvement of deliberation by making it more rational, thereby making more rational judgments and making judgments more rational. A judgment is rational when it emanates from a process of deliberation that is rational. So what makes a process of deliberation rational? A process of deliberation is rational when it incorporates and utilizes the resources of non-formal reason.Footnote ³³ The four fallible, improvable, learnable resources of non-formal reason are observation, creative construction, formal and informal reasoning methods, and systematic critical assessment. Process-based accounts of rationality have been developed by Harold Brown and Cliff Hooker to vindicate the rationality of science,Footnote ³⁴ and Frank invokes science as a model for judging throughout Law and the Modern Mind.Footnote ³⁵ The contents of the four resources that follow are examples from science. Observation obtains reliable information about empirical features of the world, thereby extending the range of relevant evidence that is considered. Creative construction develops new hypotheses and theories, new instruments and experimental procedures, new theoretical methods, and new institutional roles, e. g., new peer review processes. Formal and informal reasoning methods include various logics and decision theories, along with argument by analogy and other informal techniques. Systematic critical appraisal includes carrying out tests, proposing alternative explanations, checking performance and assumptions in wider ranges of conditions, improving the power of instruments or test methods, e. g., by adopting double-blind controlled experimental designs wherever appropriate, and by participating in peer review.

Non-formal reason is ubiquitous. The best way to understand and appreciate non-formal reason comes from practical and, whenever possible, real examples. Two curt illustrations of the four resources from the social sciences complete this section. Then, to return to the law, the next section shows how the deliberation of hypothetical judging utilizes the resources of non-formal reason.

Abby Lippman and F. Clarke Fraser asked women who are at risk of having a child with a genetic condition about how they decided whether to try to get pregnant.Footnote ³⁶ For academics the prescribed method for these women to make a rational decision is to apply the principle of maximized subjective expected utility, i. e., multiply the probability of an outcome with the utility of that outcome, and the right choice is the result that would produce the greatest subjective expected utility. What Lippman and Fraser discovered, however, is that the women dismissed the probabilities of having a child with a genetic condition. Instead, they framed their decision in binary terms. They knew that either they would have a child with a genetic condition or they would not. The women imagined “scenarios” of what it might be like to live with a child with a genetic condition. They wanted to see whether they could cope with the worst scenarios they could imagine and whether they could accept the responsibility for bringing an affected child into the world. For that, they used the resources of non-formal reason to construct and assess their scenarios.

Observation: Women who have the genetic condition or have family members with the genetic condition have accumulated extensive, meaningful observations of the nature of the condition. They know what it is like to live with the condition and the impacts they have on their own family and other families. Women who do not have personal experience visit homes, clinics, schools, and support groups to see what impacts the condition has on people and families. These women also observe themselves—what they are feeling and how they are reacting emotionally.

Creative construction: Women use their imaginations to construct, revise, and vary scenarios of what it might be like to live with and raise a child with a genetic condition and to anticipate how they and others could respond to these scenarios. Most important were the scenarios of how their family life would alter, from installing aids and wheelchair ramps in their houses, to budgeting more strictly, revising vacation plans, and adopting new and demanding care roles.

Formal and informal reasoning methods: Formal reasoning and all sorts of informal reasoning pervade the women’s deliberations. The major premises of deductive arguments raise important issues, e. g., “If the costs of care are substantial, we will not have enough money to put our other children through college.” Inductive arguments abound, e. g., drawing on how well they have dealt with past challenges or crises. Argument by analogy compares the strengths and weaknesses of their family to families with an affected child that they have visited. Metaphors are invoked to try to understand what is required of a “good mother” or a “good parent.”

Systematic critical appraisal: In envisaging themselves in their scenarios, the women are identifying and scrutinizing the values in their lives, assessing the genuineness and importance of their commitments, and gauging their capacities to cope. Systematic critical appraisal has a vital social dimension in that the women can obtain an understanding of the related experiences of others, from which they can refine and expand their scenarios, and from which they can assess their own capacities, discern the social boundaries within which decisions may be made, and appraise the collective sense of what it is reasonable to do.

There is no certainty, no necessity, and no universality for these women. Their use of the four resources renders their deliberation rational, assuages their guilt, and improves their judgment. That is the best they can do and the best they can be expected to do.

Myra Bluebond-Langner researched how children found out that they had leukemia and were dying despite a concerted conspiracy of silence by their parents and their health care professionals.Footnote ³⁷ These children were eminently rational in discovering what they wanted to know, and their rationality resided in their use of the four resources of non-formal reason.

Observation: Much of what the children learned came from careful, intense observation. On their initial entrance to the oncology clinic to which they had been referred, they became engrossed in watching the children at play. In the hospital the children detected a hierarchy of authority among the physicians and were able to construct that hierarchy. Residents were differentiated because they were younger than the attending hematologists, who were in charge of the leukemic children, but older than the interns, and because, like the hematologists but unlike the interns, they generally wore lab coats over their street clothes.

Creative Construction: The creative construction of the children often took typically childish forms. They would congregate in places where they could talk privately, and they would devise strategies for eavesdropping on conversations. Because the children knew that information exchanged between adults was more valuable and more reliable than information adults conveyed to them, they surreptitiously sought that precious information. They would try to delay separations from their parents in the clinic’s waiting room so that they had more opportunities to overhear adult conversations; they would hide behind office doors; they would pretend to be asleep; they would turn down the volume on the TV; and they would ask people to be quiet when doctors assembled by the door of their room to talk to their parents. Sometimes they were more daring: one child planted a tape recorder in the doctors’ office, which served as a conference room.

Formal and Informal Reasoning Methods: The children used deduction to learn what they wanted to know the most—what was going to happen to them:

The children employed familiar reasoning methods to understand their surroundings and what was happening and was likely to happen to them. To begin to comprehend who they were, they quickly differentiated themselves from children with other diseases:

They used induction and analogies: a boy’s loss of hair indicated that he was on the drug vincristine because other boys had lost their hair after being put on vincristine. Their reasoning often relied on assessments of other people’s behavior: “Alan … knew that things were not going well when, in spite of how good he felt, his mother would cry. ‘In my house, crying and bad things go together.’”Footnote ³⁹

Systematic Critical Assessment: The children critically assessed what was most important to them—information. They astutely realized that information is reliable only if it comes from trustworthy sources. So they tested people who served as their sources, including Bluebond-Langner:

On another occasion Bluebond-Langner passed a test by watching TV with a patient for an hour in silence.

Bluebond-Langner discovered that all of the children she studied learned that they were dying. They wielded the resources of non-formal reason deftly, and they found out what they wanted to know.

If children use non-formal reason, so do judges. Hart repudiates the false and pernicious dichotomy that formal reason propagates in judging: “Neither in interpreting statutes nor precedents are judges confined to the alternatives of blind, arbitrary choice, or ‘mechanical’ deduction from rules with predetermined meaning.”Footnote ⁴¹ Hart understands that a judicial decision can be rational even if it cannot be proved to be correct. He recognizes the judicial virtues that constitute the role of a judge: “impartiality and neutrality in surveying the alternatives; consideration for the interest of all who will be affected; and a concern to deploy some acceptable general principle as a reasoned basis for decision.”Footnote ⁴² These judicial virtues belong to the deliberative process of judging and differentiate that process from making law capriciously. A rational deliberative process is the source of Hart’s “informed impartial choice” that yields a “reasoned product.”

V. Non-formal Reason in the Law

Judgment is unavoidable in the law for the same reason it is unavoidable outside of the law—rules run out quickly and rules need to be selected and interpreted. Jerome Frank recognizes the similarity between the judge and the ordinary person: “Judging begins … with a conclusion more or less vaguely formed; a man ordinarily starts with such a conclusion and afterwards tries to find premises which will substantiate it. If he cannot … he will, unless he is arbitrary or mad, reject the conclusion and seek another.”Footnote ⁴³ Real problems, in the law as in life, are complex, messy, and malleable. A deliberative process has to begin somewhere, and that starting-point can be an initial impression, inkling, hunch, or preference. The search for mutually supportive principles and facts involves analysis, interpretation, and creativity, and the outcome then has to be critically assessed, all in a rational process that is dynamic and fallible. The process is reflexive because how it is being conducted, how productive it seems, and how it might be improved are themselves subjects of on-going scrutiny. Choices about what materials to examine, which methods to use, and how to assess outcomes, as well as determinations of whether an approach is working and when to stop, are all matters of judgment. Because judgments abound, judgment is the gist of judging.

In “Toward a Critical Phenomenology of Judging” Duncan Kennedy imagines how, if he were a judge, he would work through a case,Footnote ⁴⁴ and his hypothetical deliberation manifests the sort of judging that Frank prescribes. Supposing that he is a U.S. District Court judge, Kennedy has to decide whether to issue an injunction that would stop striking bus drivers from lying down in the street in front of a bus station to prevent replacement drivers from taking the buses on their routes.Footnote ⁴⁵ His initial impression of the case is that “the law” does not allow the workers to prevent the use of the buses during their strike, so the employer will get the injunction. Kennedy dislikes that outcome because his strong moral and political convictions make him sympathetic to the striking workers. In addition, he would like to change “the law” generally to make it favor the employees. So, Kennedy perceives a conflict between what “the law” seemingly requires and what he would like to achieve, which he dubs his “How-I-want-to-come-out” (HIWTCO).

In his book, How Judges Think, Richard Posner dismisses Kennedy’s phenomenological foray into the mind of a judge with the “damaging criticism” that “judges no more identify with Kennedy’s imagined judge than they do with Dworkin’s fictional ‘Judge Hercules.’”Footnote ⁴⁶ Were that comparison apt, Posner’s criticism would be damning. But unlike Dworkin, Kennedy does not make the judicial task superhuman. Kennedy’s depiction of judicial deliberation is thoroughly practical, the antithesis of Dworkin’s abstractions. More potentially damning is Posner’s portrayal of Kennedy as a judge “torn between” the law and his personal sense of “social justice,” a conflict that “impels him to search for ways in which he can deny the injunction,” to which Posner objects that “[j]udges do not think the way he imagines them to ….”Footnote ⁴⁷ Posner is right that Kennedy’s moral and political commitments provide the motivation for his judicial investigation, but what is wrong with that? Kennedy’s impression is that the case for issuing the injunction is compelling, probably decisive. He could search for weaknesses and flaws in that argument, but there is a stronger critical strategy—the one Kennedy adopts. He constructs a case for not issuing the injunction. Then the cases for issuing and not issuing the injunction would have to be compared, and a judgment about which case is stronger would have to be made. Kennedy’s HIWTCO might not prevail, but, regardless, developing and assessing it instills rationality into the deliberation, thereby establishing the legitimacy of whatever judgment results.

Kennedy goes to work. He has to see whether the positions that are likely to prevail—that the law gives the employer the right to use the means of production (m.o.p.) during a strike; that the employees interfered with that right; that an injunction is an appropriate remedy for that interference; and that there is jurisdiction to issue an injunction—can really withstand critical scrutiny. His two goals are to see how these specific workers could be allowed to obstruct the buses and how the law could be changed to give workers in general as much control as possible over the m.o.p. during a strike. Kennedy’s initial sense of the law is not optimistic, but the best way to critically assess that impression is to construct the strongest possible position for the union. Although doing that conforms with his personal moral and political views, he is not violating “the rules of the game of legality.”Footnote ⁴⁸ As Kennedy explains: “What would betray legality would be to adopt the wrong attitude at the end of the reasoning process, when I’ve reached a conclusion about ‘what the law requires’ and find it still in conflict with how-I-want-to-come-out.”Footnote ⁴⁹ For the purposes of legitimacy and rationality, it does not matter whether he is called an “activist judge.” He is doing what judges and lawyers alike are supposed to do. Assessing the strength of a position ultimately requires a judgment about how it fares when the arguments for it are compared to the best arguments for an opposed position. That critical assessment does not apply to just the law. In praising the “Socratic dialectics,” John Stuart Mill emphasizes the importance of what he calls “negative logic”: “ … no one’s opinions deserve the name of knowledge, except so far as he has either had forced upon him by others or gone through of himself the same mental process which would have been required of him in carrying on an active controversy with opponents.”Footnote ⁵⁰

Apprehensive that a rule affirming the employer’s right to use the m.o.p. during a strike might be intractable, Kennedy pushes on. He considers several weak options, then, spurred by analogies, an idea emerges. What may these striking bus drivers do? Like other strikers, they may picket, and they may use various publicity measures to dissuade people from riding the buses, both of which interfere with the company’s use of the m.o.p. But picketing does not physically interfere with the company’s use of the m.o.p. and is justified by the First Amendment’s protection of free speech. Appreciating those differences, Kennedy strives to preserve the analogy. The workers laid down in the street; they did not use force; they submitted to arrest peacefully; and the press was everywhere.Footnote ⁵¹ Yet lying down in the street did delay the buses and disrupt the company’s service. That success derived from an important disanalogy with picketing.

Kennedy persists. Although the actions of the striking bus drivers went beyond conventional picketing, they nevertheless were a symbolic protest directed at the company, the public, and the replacement drivers. With respect to the company, it was an attempt to spur moral sympathy and companionship from the management. With respect to the public, it was an attempt to attract publicity and garner support. And with respect to the replacement drivers, who, after all, are the ones who actually stop the buses, it was an appeal, a declaration, and a warning. Kennedy portrays the conduct of the strikers as non-violent civil disobedience and an exercise of their First Amendment rights. The legal consequence of his interpretation is that “injunction of a non-violent civil disobedience protest would be an unconstitutional restriction of expression ….”Footnote ⁵²

What did Kennedy do? He transformed a labor law and tort law case that the strikers most likely would lose into a constitutional law case that the strikers might win. How did he do that? He expanded the scope of picketing in a way that encompassed the facts of this case:

Kennedy’s critical acumen and creative brainstorming reframed the case. He rejected prosaic options that were not plausible or would not produce the outcome he desired. He assumed that the employer’s freedom from interference was not unlimited, and he tested it by envisaging what the workers might be allowed to do. They could picket. What else could they do that would be relevantly similar to picketing? What about lying down in front of the garage? He marshalled the law that would protect such behavior, formulated the facts of the situation, and applied the new law to the new facts. He constructed a case for the pickets that is grounded in constitutional law and is at least a credible, if not formidable, competitor to the case for the employer that is grounded in labor and tort law and initially seemed invincible.

Kennedy’s deliberation is by no means arbitrary, capricious, or haphazard. It is eminently rational, but its rationality is not the rule-governed reasoning of formal reason. The rationality of Kennedy’s deliberation inheres in a process that was rationally developed and deployed and that used the four resources of non-formal reason. His brainstorming is a dedicated, persistent exercise in creative construction, motivated by the need to subject his impression that “the law” does not allow the workers to interfere with the owner’s use of the m.o.p. during a strike to systematic critical assessment. Systematic critical assessment examines the strengths of the case for the employer, the apparent weaknesses of the case for the workers, and various strategies for developing a stronger case for the employees. Creative construction shifts the case away from labor and tort law and subjects it to constitutional law, which renders a different array of compelling features relevant and generates a surprisingly strong case for the employees. Formal and informal reasoning methods pervade the entire process, but argument by analogy is most notable in identifying the similarities between picketing and lying down in front of the garage. Observation does not play as obvious a role in judging as it does in science, but it is there, e. g., in scrutinizing the demeanor of witnesses at a trial. Direct personal observation is not prominent in Kennedy’s deliberation, but the amassing of relevant, reliable empirical information ultimately gleaned from observation elsewhere is crucial. Vicarious visualization is evident in Kennedy’s restatement of the facts when he employs an image, not of a dirty, smoke-spewing city bus swathed in advertisements but a shiny Sceni-cruiser. Kennedy’s analogical thinking was, and probably needed to be, visual as well as verbal, picturing strikers marching with picket signs, blocking entrances to company parking lots, and pounding their fists on the vehicles of replacement workers. And had Kennedy been thinking about this case after Tiananmen Square, his route to the First Amendment undoubtedly would have been more direct.

Kennedy’s process of investigation and deliberation manifests the sweep and the flexibility of non-formal reason. He analyzes the total situation, making judgments about all sorts of value-driven substantive and procedural matters that emerge, e. g., not just what considerations are relevant and how important they are, but what strategies and methods to adopt, how to improve them, how far to pursue them, and, eventually, when the inquiry should be terminated. It is not a constrained exercise in subsuming factual premises under legal premises but a search for propositions that are, in Dewey’s words, “worthy to serve as premises.” That is a dynamic, reciprocal inquiry, as factual premises and legal premises are tested against each other to see how mutually supportive they are. As an understanding of the facts develops, rules can be modified, or new rules can become relevant. And vice-versa, as Kennedy discovered. As his recognition of potentially relevant rules expanded, new facts that supported the new rules became relevant. Kennedy had to “choose intelligently” as he searched for “rules which will substantiate a certain way of looking at and interpreting the facts,” and in doing that, he discovered facts that could substantiate the rules his analogies produced, e. g., the strikers were not violent and submitted peacefully to arrest, and the strike was widely covered in the media.

But what exactly has Kennedy accomplished? Using non-formal reason, he has constructed a respectable legal argument for not issuing an injunction. With that accomplishment, he has transcended his bias in favor of the striking workers and advanced from a mere “How-I-want-to-come-out” to a provisional “How-I-might-be-able-to-come-out.” That is a start, but only a start, as Kennedy appreciates. He is acutely aware of the uncertainty and tentativeness of his possible defense of the strikers: “Will this fly? I have no idea …. It is part of the work of producing lots of alternative ways at the problem, hoping that one of them will break through. I am already wondering whether it’s even worth the time to pursue this one further.”Footnote ⁵⁴ Despite all the work he has done so far, much work remains. To decide the case, he has to make the even more difficult transition from “How-I-might-be-able-to-come-out” to “How-I-should-come-out.” To do that, each side of the case needs to be fully developed and critically assessed. So, Kennedy has to make a judgment about whether his First Amendment argument is sufficiently promising to devote more of his judicial resources to developing and assessing it. If he embarks on that task, he then will have to develop the strongest argument he can for the rule that workers may not interfere with the owner’s use of the m.o.p. during a strike. And, ultimately, he will have to make a judgment about which position is more rationally and legally compelling. The process leading to that culminating judgment would be a continuation and elaboration of the creative construction and critical assessment that has brought Kennedy to where he is now.

Kennedy does not know how he should decide this case because much depends on the state of “the law.” He is “free to work in the legal medium to justify the workers’ actions against the company,” and how his argument “will depend in a fundamental way on the legal materials—rules, cases, policies, social stereotypes, historical images—but this dependence is a far cry from the determination of the outcome by a process of legal argument that could only be done correctly in one way”.Footnote ⁵⁵ He repudiates the “fancy argument” of a professor of jurisprudence: “… I can’t show an ‘objective basis’ or a rationale or an explanation of rule application that will prove that any particular application was ‘correct.’ Indeed, the notion of correctness, at least as we usually use the word, say in math or science or logic, just isn’t applicable.”Footnote ⁵⁶ Kennedy is right. Kennedy’s deliberative process possesses the rationality of non-formal reason, and the realistic objectivity of that kind of rationality is the transcendence of bias.

VI. Conclusion

This analysis of judging is at odds with Kennedy’s own culminating assessment of his project because he, too, falls, with a tip of his hat to postmodernism, for the irrationalist fallacy.

His “humble artificer’s idea of reason” is formal reason, and although he is under its theoretical spell, his imaginative judicial deliberation ironically dispels that spell.

Formal reason casts indeterminacy as an inevitable and irremediable failure of judging, and the unpalatable result—making law—is characterized as a judicial misdeed. Non-formal reason removes the false dichotomy that creates this dismal portrayal of judging by providing a more expansive conception of rationality. Judgment is pervasive and ineliminable in judicial deliberation, as it is everywhere in life, but judgment need not be capricious, subjective, or arbitrary. With non-formal reason rational judgments can emanate from rational processes. That is why, when the rational processes of non-formal reason are operating in a well-designed legal system, confidence in the rule of law is warranted, and those who decide cases may rightfully and respectfully be called judges.

The goal of rationality is, as it is for all intelligent human activities, to improve our capacity to make judgments by overcoming the limitations and defects of our inherent finitude and fallibility. Reason is the capacity in virtue of which, within our finite resources, we transcend our imperfections through a process that improves our judgments. This transcendence includes, reflexively, the capacity of reason to improve itself.