Southern California Law Review

September, 1988

USC Symposium on Judicial Election, Selection, and Accountability

Comment

*1735 THE VIRTUES AND VICES OF A JUDGE: AN ARISTOTELIAN GUIDE TO JUDICIAL

SELECTION

Lawrence B. Solum [FNa]

INTRODUCTION

A core insight of the legal realists was that many disputes are indeterminate. [FN1] For example, in many appellate adjudications, respectable legal arguments can be made for both sides of the dispute. A contemporary reaction to the realist insight by critical legal scholars is expressed in the slogan 'Law is politics.' This critical slogan might be elaborated as follows: in openly political activities, such as the legislative process or partisan elections, debate centers on issues of value and social vision that are outside the scope of 'legal reasoning.' Judicial opinions merely dress up political decisions in the garb of legal reasoning.

The realist insight and critical reaction challenge conventional notions about the selection of appellate judges on the basis of merit--a combination of legal expertise and judicial temperament. [FN2] If appellate judges really render decisions on the basis of politics, then why should judges be selected (or elected) on the basis of merit? In his essay, Judging *1736 in a Corner of the Law, [FN3] Professor Schauer has gone so far as to suggest that appellate judges need not be lawyers and certainly need not be experienced or excellent lawyers. Moreover, Schauer maintains, the skills and knowledge desirable in appellate judges are not even taught in law schools. [FN4]

Why is Schauer's proposal so novel? Why do we find seemingly absurd the idea that lay people could be excellent judges in general and excellent appellate judges in particular? Why does one want to say that appellate judges, especially Justices of the United States Supreme Court, should not only be experienced in the law, but should be drawn from the ranks of our best and most experienced judges and lawyers? Why does Schauer himself feed constrained to make his suggestions only half seriously?

These questions reflect, I think, a strong intuitive discomfort with Schauer's suggestions. Indeed, we may be tempted to underestimate the importance of his suggestions or engage in facile refutation of his arguments. These responses would be grave errors.

First, we must resist translating our strong intuitive discomfort with Schauer's suggestions into a belief that they are not worth serious consideration. Key elements of Schauer's suggestion are widely reflected in the actual practice of judicial selection. The University of Southern California Law Center Symposium on Judicial Election, Selection, and Accountability was prompted, at least in part, by the injection of partisan politics into judicial elections. Judges in this country are selected on the basis of politics, [FN5] and they have often been politicians or the cronies of politicians. Vermont's superior courts include lay judges. [FN6] Other legal cultures have adopted some of Schauer's seemingly radical suggestions. For example, in Nicaragua, lay people constitute special courts which try overtly political crimes. [FN7] In England, lay magistrates have jurisdiction over some imprisonable offenses. [FN8] Moreover, the historical roots of the common law tradition lie in a system in which judges were openly the *1737 political agents of the sovereign. [FN9] Perhaps the most celebrated advocate of Schauer's suggestion was James I; the King's suggestion that his reason qualified him for the task of judging prompted Sir Edward Coke's rejoinder that only lawyers were skilled in the 'artificial reason and judgment of the law.' [FN10]

Second, we must resist overly idealistic assumptions about appellate judging. Judges, even Supreme Court Justices, are not Ronald Dworkin's Hercules. [FN11] Judges are real men and women doing a job with limited time and ability. If our intuitive discomfort with Schauer's suggestion is defensible, the argument must rest on an account of what judges actually do and are realistically capable of doing.

Schauer challenges us to explain our discomfort with the idea that nonlawyers could be good appellate judges. He tells us that familiar explanations will no longer do. We can no longer rely on the conventional notion that appellate judges must be selected on merit because only technical knowledge of the law is required to recognize (or discover) the legally correct solution to the questions judges usually face.

I take issue with Schauer's contention that a sophisticated understanding of the indeterminacy of appellate decisionmaking undermines old-fashioned notions that we should select appellate judges on the basis of experience and excellence in the practice of law. I will argue that judging in the special corner of the law reserved for appellate courts requires special intellectual and moral virtues that are best developed by immersion in the practice of law. I will discuss Schauer's suggestion that nonlawyers be selected for appellate judgeships as a focus for analysis, but my conclusions will be broader. The reasons for preferring lawyers to lay persons as appellate adjudicators will be the reasons for preferring experienced lawyers over novices and excellent lawyers over mediocre ones.

My argument will begin in Part I with a brief exploration of the idea that a good judge must possess judicial virtue. This account will relate my view of judging to the tradition of Aristotelian moral and political *1738 theory. Part II will sketch the distinctive intellectual and moral virtues required for excellence in judging. To be a good appellate judge, one must possess at least three virtues: a faculty for theoretical understanding of the law, a concern for the integrity of the law, and a practical ability to choose wisely in particular circumstances. In Part III, I will argue that the judicial virtues can, for the most part, only be acquired through participation in the practice of law. My conclusion directly follows: appellate judges should be selected on the basis of merit understood as the possession of these judicial virtues.

I. THE ROLE OF THE VIRTUES IN JUDGING

My mention of virtue may sound almost quaint to some sophisticated readers. The brands of modern moral philosophy that dominate contemporary jurisprudence are more likely to emphasize moral rules, rights, or consequences than virtues. I do not want to deny that consequentialist or deontological theories of morality can contribute importantly to our understanding of morality, [FN12] but I do want to emphasize the value of a virtue-centered account of morality in enabling us to understand judicial selection. My claim is that it is precisely by examining judicial virtue that we can understand our intuitions about what makes for a good judge. [FN13] Before I redeem this claim by offering an account of the specific virtues that make for good judging in Part II of this Essay, I will briefly explicate Aristotle's theory of the virtues and relate that theory to the practice of judging.

In recent years, there has been a revival of interest in Aristotelian moral theory, and especially in Aristotle's theory of the virtues. [FN14] For Aristotle, the virtues are acquired dispositional qualities [FN15]--they are *1739 potentialities or powers which are states of character or of mind. [FN16] Aristotle characterizes the virtues as intellectual or moral, [FN17] and his views can be sketched by examining these two categories.

The moral virtues are states of character concerned with choice; examples include courage, temperance and justice. [FN18] Aristotle thought that each of the moral virtues could be seen as the mean between two opposing vices: thus, courage is a mean between the vices of timidity and recklessness. [FN19] Moral virtues, says Aristotle, are acquired as a result of habit; one must act courageously in order to become courageous. [FN20]

The intellectual virtues are practical and theoretical wisdom. Practical wisdom or phronesis is excellence in deliberation: the person of practical reason is able to choose good ends and the means to achieve those ends. [FN21] Practical wisdom operates in the realm of praxis: action in particular situations. Theoretical wisdom or sohpia, on the other hand, operates in the realm of theoria; abstract thinking, science and theory. [FN22] A person begins to acquire the intellectual virtues through education; these virtues mature through experience. [FN23]

The virtues, then, are those characteristics of mind and will that are conducive to a good life--for Aristotle a life of happiness or faring well and doing well. For example, the person who possesses the virtues of temperance, courage, and wisdom will likely flourish, as will a society composed of such persons. But the person who possesses the corresponding vices of intemperance, cowardice, and stupidity will likely not be happy and will not contribute to the happiness of others.

Finally, a word about the distinctive nature of the virtue-centered account of morality and its relationship with a theory of judicial selection. Aristotle's theory focuses on states of character and mind. In contrast, a consequentialist theory like utilitarianism and a deontological theory like Kant's categorical imperative each focus on rules for decision. *1740 For example, the utilitarian rule is maximize utility, and Kant's rule is act so that the maxim of your action could be willed as a universal law of nature. Likewise, a theory of appellate judging could provide a decision procedure for good judging: decide appeals in accord with the original intentions of the framers or decide appeals so as to achieve efficient outcomes. Judicial selection might be a matter of selecting judges who adhere to the correct decision procedure. My suggestion is that a virtue-centered theory will provide more fruitful insights into our intuitions about what makes for a good judge. In the next Part, I consider the specific virtues that are conductive to good judging.

II. A SKETCH OF THE JUDICIAL VIRTUES

It this section, I discuss three aspects of judicial virtue. [FN24] The first of these virtues is judicial intelligence--excellence in understanding and theorizing about the law. [FN25] The second virtue is judicial integrity: the good judge must have a special concern for fidelity to law and for the coherence of law. The third virtue is judicial wisdom: the good judge must possess practical wisdom in the choosing of legal ends and means. [FN26] The first and third are what Aristotle called intellectual virtues; the second is what Aristotle might have called a moral virtue.

There is a fourth aspect of judicial virtue that I will not discuss in detail: the virtue of justice. Clearly, a good judge must be just. Sometimes justice consists in applying the legal rules: my discussion of judicial intelligence and integrity will elaborate on this aspect of justice. Sometimes justice consists in departure from the legal rules: my discussion of judicial wisdom will touch on this topic. Schauer's essay, however, poses the question whether the virtue of justice is a virtue that must or should be developed through the practice of law. I believe that an understanding of that question is best approached by first considering the role of judicial intelligence, integrity, and wisdom.

*1741 A. JUDICIAL INTELLIGENCE

The realists observed that in many legal disputes, especially disputes that are adjudicated or appealed, the results are not determined by the law. Frequently, very attractive legal arguments exist on both sides of a dispute. In such cases, the realists urged judges to decide on the basis of policy considerations. Schauer builds on this realist observation. Citing the law and economics literature on the selection of cases by litigants for formal adjudication and appeal, Schauer argues that almost all the cases presented to appellate judges will be indeterminate. As a result, the cases will be decided on the basis of nonlegal factors, such as morals, politics, efficiency, and so forth. Appellate judges, therefore, need not be experts in the law. Perhaps they should be politically accountable; perhaps they should be trained in public policy science, economics or moral philosophy, but they need not be trained in law. Schauer, like James I, believes that judging does not require 'the artificial reason of the law.'

In a sense, this is Schauer's strongest argument. If appellate cases do not turn on what the law is, then it follows that appellate judges do not need to possess special intellectual virtue in the law. I will challenge Schauer's argument at several levels. The first challenge isto the contention that appellate courts face exclusively, or even primarily, 'hard' cases. The second challenge is more serious: even in indeterminate cases, a correct decision may require substantial legal expertise. The third challenge reveals a crucial problem with the argument: some cases are perceived as indeterminate not because the law is indeterminate, but because the case itself is complex. The upshot of these three challenges is that good appellate judges must possess legal intellectual virtues; they must have the ability 'to think like a lawyer' which is acquired through legal education and participation in the practice of law.

1. Judicial Intelligence is Required in Easy Cases

Is it really true that only cases in which the law is indeterminate will be adjudicated? It is certainly true that appellate cases are to some extent selected for indeterminacy. [FN27] This point is, however, easily exaggerated. Even appellate courts decide easy cases, and easy cases do require knowledge and reasoning skills that are developed through the practice of law.a. *1742 Appellate courts frequently decide easy cases:This point can be demonstrated both empirically and through the application of rational choice theory to litigation behavior. A brief survey of the arguments follows.

i. Empirical evidence: Many litigated cases, even appellate cases, are relatively determinate. There is considerable evidence that appellate courts, even the United States Supreme Court, decide a large number of easy cases. For example, a very large percentage of cases decided by the United States Court of Appeals is disposed by unanimous unpublished memoranda. One of the criteria for memorandum disposition is that the decision not resolve a previously 'indeterminate' legal question. [FN28]

Even more striking is the caseload of the United States Supreme Court, which is usually assumed to consist entirely of hard cases. The Supreme Court, however, decides an enormous number of easy cases. Consider the actual disposition of cases by the Court. [FN29] During last term, the Court had 4,339 cases on its docket. Only 175 of these cases resulted in written opinions, and 102 of these were decided by per curiam or memorandum decisions. The thesis that the Court is presented only with hard cases finds its strongest support in the large number of 5-4 decisions by the Court; there were 45 last term. Surely these are hard cases. More troubling for this thesis is the only marginally smaller number of unanimous decisions. Last term 18.4% of the cases with full opinions were unanimous decisions, and 92.2% of the Court's memorandum orders were made without dissent or concurrence. While, these unanimous opinions raise a question whether all the Supreme Court's cases are hard, it may well be that these unanimous decisions reflect political or moral consensus on the Court about cases in which the law is indeterminate. So far, the Supreme Court/hard case thesis seems to be in good shape.

Similarly, the large number of cases in which certiorari was denied does not pose an insurmountable problem for the thesis. A denial of certiorari does not constitute a decision on the merits and hence may simply indicate that the Supreme Court was not interested in what was *1743 actually a hard case. It is at least possible that this was the case in all but a handful of the many hundreds of cases in which certiorari was denied. Dismissal of a case brought under the appellate jurisdiction for want of a substantial federal question is, however, a decision on the merits and is in fact a decision that the case was 'determined' by existing law. [FN30] The hundreds of cases which are disposed of in this way pose a very thorny problem for the thesis that the Supreme Court does not face easy cases. There is one line of defense: it is often charged that the Court dismisses for want of a federal question in a case where a thorny question of federal law simply does not interest the Court. [FN31] It is possible that the Court deliberately thwarts the law in almost all of the hundreds of cases in which it dismisses an appeal, but at this point the theory that the Court faces only hard cases seems to be a degenerating research program, as more and more ad hoc explanations are added to refute the seemingly contrary data.

My conclusion is that the empirical evidence provides strong support for the conclusion that appellate courts confront a significant number of cases in which the results are relatively determined by existing law.

ii. Rational choice theory: Schauer, however, did not rely on empirical evidence. The primary support for his contention was found in literature which suggests that litigants settle cases in which law is determinate, and only indeterminate cases proceed to the appellate level. Most prominent in this literature is the work of George Priest and Benjamin Klein, who argue that the process of selecting cases for litigation will result in a plaintiff success rate that approaches fifty percent, irrespective of the favorability of the substantive law to plaintiffs or defendants. [FN32] Schauer elaborates on the Priest and Klein model of the process and concludes that almost all the appellate cases will be legally indeterminate. [FN33]

*1744 As a matter of rational choice theory, this conclusion is questionable. It is based on a model of the appellate process which includes a number of simplifying, but, highly questionable, assumptions. Let me mention in passing a tension between Schauer's rational choice argument and his conclusions: if incentives to settle ensure that only legally indeterminate cases are appealed when the decisionmakers use legal criteria for decision, it would follow that only morally (or politically or economically) indeterminate cases will be appealed when lay judges use moral (or political or economic) criteria for decisionmaking. [FN34] I will leave some of the more detailed criticisms to a footnote [FN35] and briefly survey some of the more important difficulties with Schauer's argument.

Initially, litigation decisionmakers may have preferences which cannot be accommodated in settlement before appeal. The Priest and Klein model, which is limited to personal injury disputes, explicitly assumes that the determinants of settlement are solely economic. [FN36] This assumption is dubious outside the personal injury context. For example, some *1745 'public interest' litigation may have ideological motivations. Those who make the decision to appeal may have a strong preference to make their arguments, even if they do not believe the arguments will succeed. Likewise, attorneys for government agencies which are not directly subject to market pressures may wish to litigate only when they have an extremely high probability of victory; they may believe that promotions or other rewards will be tied to their victory rates. On the other hand, the political system may reward the decision to advocate a politically popular position which has a low probability of legal success. Individuals also have ideological motivations; an example is the tax protest movement.

In addition, it may not be possible to structure a settlement that compensates the appellant for the expected untility of an appeal that has a positive but small probability of success. For example, in criminal cases there may be incentives for losing defendants to appeal even when the chances of success are low. The value of freedom from incarceration may be high enough to justify an appeal even when discounted by a small chance of success on appeal. The prosecution is not likely to be able or willing to bargain with the defendant for a reduction in sentence in exchange for the plaintiff's abandonment of an appeal that is unlikely to succeed. Jailhouse lawyers may actually prefer work on successive appeals and habeas petitions to the other pursuits available in prison.

Finally, there may be an asymmetry of stakes in appellate litigation which makes it impossible to reach an efficient settlement. The Priest and Klein model assumed that there was symmetry of stakes, i.e. that the plaintiff had as much, and only as much, to gain as the defendant had to lose. [FN37] On appeal, however, this symmetry may not exist. For example, one side may have an interest in setting a precedent that will yield benefits in future disputes but the other side may have a stake only in the dispute on appeal. In such circumstances, Priest and Klein predicted that the victory rate for plaintiffs on appeal would not approach 50 percent. [FN38]

*1746 In sum, it is very difficult to argue that rational choice theory predicts that appellate decisions will be legally indeterminate when realistic assumptions about motivations and options is incorporated into a model of the choice situation.

b. Decision of easy cases requires judicial intelligence: All of these theoretical considerations coincide with the empirical evidence. There are many appeals in which the law is relatively determinate. This is the first reason for my contention that appellate judges must possess knowledge of the law and skill in legal reasoning. At all levels of the appellate process, judges confront cases in which decision requires an intellectual virtue that is distinctly legal. Appellate judges must read and understand statutes, regulations, and cases. They are required to identify the legally significant aspects of a factual situation and apply the law to those facts.

Of course, none of this comes as a surprise. Anyone who has participated in the appellate process knows that judges do these things. Although an advocate of lay judging may disagree with the number of cases in which these skills are required, the occasional easy case must be acknowledged and some means to deal with it must be provided. Schauer, for example, argues that the easy cases can be separated from the normal appellate process and handled by legal technicians who would screen the cases and decide them by virtue of a summary procedure. The Ninth Circuit has adopted some aspects of Schauer's proposal. Staff Attorneys screen out easy cases and write memoranda and proposed opinions; judges then decide these cases without oral argument and frequently the court will adopt the recommendations of the staff attorneys without modification. A similar premise may offer support to arguments in favor of the creation of a National Court of Appeals or an Intercircuit Court to handle the more routine appeals directed at the Supreme Court. The new court could be staffed by legal technicians, while politicians or philosophers appointed to a Supreme Court would be freed from the mundane chores of legal reasoning.

Even if there are more easy cases than ad advocate of lay appellate judging admits, these sorting mechanisms may suffice to preserve a role for lay judges in the decision of the hard cases. The next section of this essay argues against that conclusion. I will accept the premise that appellate courts (as presently constituted or as they would be constituted after a screening mechanism had been instituted) will decide only hard cases. This premise does not lead to the conclusion that lay persons should be selected as appellate judges.

*1747 2. The Resolution of Hard Cases Requires Judicial Intelligence

There is an obvious reason why a large but limited category of hard cases requires specifically legal intellectual skills. I will discuss this category briefly and then consider who very general reasons behind the difficulty that nonlawyers would experience in resolving appeals.

A. Some hard cases involve issues that are conceptually legal: Many appeals which are legally indeterminate are nevertheless resolvable only by specifically legal considerations because the issue being appealed cannot even be conceptualized without a detailed knowledge of the legal process. Procedural issues are obvious examples. Consider appeals that pose questions about pleading under the simplified procedures in the Federal Rules. For example, how would a lay person decide the issue of what constitutes sufficient particularity in the pleading of fraud under Rule 9(b)? Think about a lay person deciding a hard case concerning the applicability of statutory interpleader to an insurance dispute. Imagine a lay judge attempting to resolve an appeal where there were conflicting precedents about the impact of the Erie doctrine on Rule 19 in a diversity case that has been removed from a state court system permitting Doe pleading. Even cases that pose important social policy questions may be so intertwined with legal technicalities that lay judging would be difficult. The eleventh amendment, for example, has important implications for civil rights actions, but the impact of changes in doctrine in this area of the law is difficult even for lawyers to understand. Standing issues may pose similar problems for lay judges.

Even if these 'legal' hard cases are indeterminate, the considerations of policy or value or politics which must inform their resolution do require judicial intelligence. This may not pose a fatal problem for the advocate of lay judging. Again, mechanisms could be adopted for screening out the hard cases involving legal technicalities. A special court of legal technicians could hear the 'legal' hard cases. This solution is, however, problematic. A large number of the social policy cases may also present 'legal' issues on appeal. The court of legal technicians might actually adopt a deliberate strategy of creating legal technicalities in order to retain jurisdiction, and hence power, over as many of these disputes as possible. The advocate of lay judging owes a great deal of explanation about how such a proposal would work.

b. Outcomes are constrained by law in all hard cases: Assuming such an explanation is forthcoming, there is a related and more pervasive reason why the adjudication of hard cases requires the virtue of judicial *1748 intelligence. Schauer contends that in appellate cases in which there is an outcome-determinative issue which is not itself determined by the existing positive law, the decision will require a choice that can be made on moral, political, or economic grounds. This contention assumes a simplified model of the output of an appellate court as a pure decision: to affirm or reverse the lower court. When the actual output is viewed it becomes apparent that legal factors do constrain appellate decisionmaking, even in hard cases.

In order to discuss this point, I need to develop some conceptual apparatus. Thus far, I have accepted the implicit assumption that indeterminacy and determinacy are exhaustive categories; in other words, the decision of an appeal is either determined by the law or it is indeterminate. In fact, this is not the case. A legal dispute may be constrained by the positive law, but not determined by it. Thus, I add to determinacy and indeterminacy a third category: underdeterminacy.

Roughly, an appeal is underdetermined by the law if the outcome including the formal mandate and the content of the opinion, can vary within limits that are defined by the legal materials. [FN39] The notion of a 'hard case' can now be explicated with reference to the category. An appeal is a hard case if (but not 'if and only if' as I will argue below [FN40]) the outcome is underdetermined in a way that an appellate judge must choose from among legally acceptable outcomes that constitute victory for different parties to the appeal. The point is that the outcome of an appeal need not be completely indeterminate in order for the case to be a hard case; an appeal in which the results are underdetermined by the law will be 'hard' if the legally acceptable variation creates the difference between loss or victory for the litigants. [FN41]

A moment's reflection on the nature of appellate cases will confirm that even the 'hardest' appellate case is likely to be underdetermined by *1749 the law and not completely indeterminate. Many appellate dispositions require more than a simple decision to affirm or reverse; for example the decision whether or not to remand may well be constrained by the law.

Perhaps even more pervasive are legal constraints on the grounds chosen for a decision. In our system, the rule of stare decisis makes the reasoning of an opinion an authoritative legal taxt that will have influence on future opinions. Although American judges typically do not attempt to formulate a single proposition as the authoritative holding, there can be no question that the reasoning of a case has a substantial effect on its future legal effect. Appellate judges must choose the grounds for their decision as well as a result in the narrow sense, and the choice of grounds will almost always be constrained by the positive law, even if the result itself is not. Is the contract invalid because the offer was not validly accepted, or because there was no consideration, or because it was contrary to public policy? The answer to such questions will invariably be constrained by the law. This notion of constraint that does not determine outcome suggests a third reason that decision of hard cases requires judicial intelligence.

Thus, the virtue of judicial intelligence will be required even in hard cases which do not involve technical legal issues because appellate judges are required to write opinions which explain the grounds for their decisions. Furthermore, in the writing of such opinions these judges must make choices that are constrained by the law. The reasons provided in the particular case being decided must be consistent with the law at large, not only with the statutes, constitutions, and regulations, but also with reasons and justifications provided in other cases. It would be difficult for lay appellate judges to write good opinions, even if they had an excellent grasp of the moral or political reasons that grounded their preferences.

c. The synthesis of legal and nonlegal grounds for decision requires judicial intelligence: There is an additional related reason that decisions in hard cases require judicial intelligence. Schauer argues that resolution of hard cases will require consideration of factors that are not legal in the narrow sense. Indeed, he suggests that a broad range of moral, political, economic, or cultural factors might be included. This suggestion poses the question of how these diverse factors are to be integrated with each other within the law. A partial answer lies in the particular skill of synthesizing the nonlegal with the legal that is part of judicial intelligence. In other words, lawyers are generalists. They are trained to apply considerations of policy, morality, economics, and culture to legal problems.

*1750 For all these reasons, a good appellate judge must possess the virtue of judicial intelligence in order to do a good job of resolving hard cases in which the law does not determine the result.

3. Hardness Sometimes Results from Complexity Rather Than Indeterminacy

There is another and perhaps especially important reason that appellate judges should possess the intellectual virtue that I have called judicial intelligence. Not every case of perceived indeterminacy results from actual indeterminacy of the legal materials. In at least some cases in which there is a 'legally correct outcome,' the reasons for this outcome will be sufficiently complex that some litigators simply will be unable to understand the law. I want to make a disclaimer about the frequency of such cases. Law professors may well overestimate the frequency of cases in which deep study and understanding of the law will enable a judge to recognize that a result is compelled by a complex but nevertheless determinate application of the existing law. Given this caveat, however, I believe that it is indisputable that such cases exist.

The nature of these cases might be clarified by an analogy with another practical discipline: engineering. A particularly difficult problem in engineering--the design of a bridge over a long and unusually windy gorge--may be 'hard' in the same two senses that an appellate case may be 'hard.' Some engineering problems allow several equally good solutions that are permitted by the various laws or rules of thumb that engineers derive from material science. In such cases, the engineer may choose from among the possible solutions on the basis of some nonengineering criterion such as aesthetic value. On the other hand, some engineering problems are hard because they are very complex. The best solution is not easy to derive. In such cases, there may be disagreement among practicing engineers about which solution is the best (or only) solution, but the mere fact of disagreement does not entail that no solution is in fact best. The more brilliant or more experienced engineer may be able to discover a solution which his or her more mundane colleagues cannot.

What are the implications of the existence of 'hard cases' that are 'hard complex' rather than 'hard indeterminate' for the selection of appellate judges? I believe that one of our hopes is that appellate judges will be able to provide the right answers in very complex cases. Of course, this does not always happen. Appellate judges are not always selected for judicial intelligence, and we are all familiar with appellate *1751 decisions where the court got it wrong because the judges did not understand a complicated legal issue. My point is a modest one: other things being equal, we have a very good reason to select judges who possess the intellectual virtue of judicial intelligence--Sir Edward Coke's 'artificial reason of the law.'

So far, I have established the proposition that appellate judges ought to possess the virtue of judicial intelligence--a substantial obstacle in the path of an advocate of lay appellate judging. Still, it might be suggested that lay appellate judges could be given a 'crash course' that would enable them to gain an intellectual grasp of the law. More importantly, judicial intelligence can be used for good or ill. A very intelligent unprincipled judge could use legal reasoning skills to provide clever but sophistic demonstrations that a personal preference is required by the law; a nihilist judge could use judicial intelligence to create indeterminacy where stability could be found. In part because of these possibilities, the intellectual virtue of judicial intelligence is necessary but not sufficient for good appellate judging. I will now consider a second, specifically moral virtue: judicial integrity.

B. JUDICIAL INTEGRITY

A good judge should have a special fidelity to the law and its coherence. I call the judicial character trait that expresses this fidelity judicial integrity. Judicial integrity is what Aristotle would call a moral virtue: it is an aspect of character that creates a disposition to choose in a certain way.

Why should appellate judges possess judicial integrity? Why should appeals be decided by persons whose choosing is disposed towards fidelity to the law and its coherence? The basic answer is expressed in the ideal of the rule of law. The rule of law is a complex notion with a long history; [FN42] my exploration begins with the publicity requirement--the principle that the content of the law should be known or knowable by those who must live with its consequences.

One of the virtues of judging is suppressing one's political or moral preferences and deciding on the basis of the law. If all appeals were decided on the basis of politics or morality, the grounds for appellate decisions might well become more comprehensible, but the law, in the *1752 end, would become less certain. [FN43] The decisional law would truly shift with the political winds and as a result become less public. Less concern with fidelity to the law in appellate judging would threaten a second aspect of the rule of law: fairness demands that like cases be treated alike. Both aspects of the rule of law support the conclusion that appellate judges should have a special fidelity to the law.

There is a second and more controversial reason to value judicial integrity that derives from the principle that like cases should be treated alike. It is at least arguable that appellate decisions in what seem like distinct doctrinal areas may be strongly interconnected at the level of justification. This phenomenon is expressed in the maxim that the law is a seamless web. In a sense, however, these interconnections results from the choices of appellate adjudicators. Judges can try to shape the law so that it reflects, so far as possible, consistent conceptions of morality or politics, or they can treat each area of the law as an opportunity for expression of their own values, regardless of the consistency of such values with the values expressed in prior decisions. The hope that judges will strive for consistency is reflected in the postulate that the law works itself pure. If we believe that fairness requires consistency in the law as a whole, then we should value the concern for such coherence that is expressed in the virtue of judicial integrity.

C. JUDICIAL WISDOM

The final virtue I call judicial wisdom. I mean to refer to the particularly judicial variety of the virtue which Aristotle termed phronesis--the virtue of practical wisdom. [FN44] Practical wisdom is the virtue that enables one to make good choices in particular circumstances. The person of practical wisdom knows which particular ends are worth pursuing and which means best achieve those ends. Judicial wisdom is simply the virtue of practical wisdom applied to the choices which judges must make. *1753 The practically wise judge has developed excellence in knowing what goals to pursue in the particular case and excellence in choosing the means to accomplish those goals.

I will begin to make this abstract account of judicial wisdom more concrete by considering the contrast between practical wisdom and theoretical wisdom in the judicial context. The judge who possesses theoretical wisdom is the master of legal theory with the ability to engage in sophisticated legal reasoning and insight into subtle connections in legal doctrine. Such a judge, for example, would be able to penetrate the mysteries of the eleventh amendment or the Rule Against Perpetuities. I want to suggest that a judge who possesses judicial intelligence and judicial integrity is still not necessarily a good judge. Even a brilliant judge, dedicated to the law, could be a diaster in the decision of particular cases if lacking in the ability to appreciate what was truly important in the particular case and to devise the means to achieve those ends. Such a judge might fashion a rule that is theoretically sound but poses intractable problems when actually applied. For example, a theoretically sound burden of persuasion may be a practical disaster if it is too expensive for parties to develop admissible evidence which meets the burden.

The judge who possesses practical wisdom in addition to theoretical wisdom has a different set of abilities. We might say such a judge has common sense. Such a judge knows whether a particular doctrinal formulation will work in practice and can grasp which aspect of the law is most important in a particular situation. The wise judge not only understands the mysteries of the eleventh amendment, but can apply the eleventh amendment to particular case in a just fashion. The wise judge knows how real lawyers and parties will react to appellate decisions.

Our understanding of the role of judicial wisdom in good judging can be sharpened by considering its connection with Aristotle's concept of equity. The law, according to Aristole, is defective because of its generality. No general rule can guarantee the best outcome in all particular situations. Equity allows the judge to dispense with the general rule in the particular case and reach a just result. [FN45] In modern legal parlance, we refer the same idea as informed judicial discretion or equitable discretion. In order to do equity and depart from the general rule when the circumstances demand, the judge must possess practical wisdom.

*1754 Consider in contrast the accounts of judicial discretion by Schauer and Dworkin. Schauer would have us believe that in the zone of discretion results are not determined by law and thus must be determined by something else entirely. For Schauer, law operates only in the core and it is politics, morality, or economics that rules the penumbra. Schauer believes that judicial intelligence does not determine the results in hard cases, and therefore legal expertise is simply not relevant to their resolution. For Dworkin, even the penumbra is part of law's empire. Dworkin believes that if the legal rules narrowly conceived do not determine the results in particular cases, then the judge must develop a theory that yields principles which will do the job of determination. [FN46] We might say that Dworkin believes that the law is entirely a matter of judicial intelligence.

In a sense, Dworkin and Schauer are making the same mistake because they both are assuming that legal knowledge is all a matter of theoretical wisdom. For Dworkin, appellate decisionmaking is a matter of theoretical wisdom about the principles that underlie the law. For Schauer, appellate decisionmaking is a matter of theoretical wisdom about politics, economics or culture. I want to suggest that, at least in some cases of discretion, Aristotle's conceptions of equity and practical wisdom provide a more fruitful explanation of how decisions are and should be made. An equitable decision is not determined by the rules or by any set of principles derived from a grand theory of the law, nor is it determined by some larger theory of morality or economics; rather, equity is a disciplined judgment about a particular case that is informed by an ability to select ends and means in particular cases--an ability that has been developed by experience in the practice or praxis of law.

In sum, a good judge must possess the judicial virtues: judicial intelligence, judicial integrity, and judicial wisdom. No formula that specifies the correct decision procedure for the resolution of appeals will guarantee good judging unless the judge possesses these virtues. Moreover, these virtues cannot be obtained through a sheer act of will; the best way for a judge to develop the judicial virtues is by practicing law. This claim is my final topic.

*1755 III. THE JUDICIAL VIRTUES ARE DEVELOPED THROUGH PARTICIPATION IN THE

PRACTICE OF LAW

Aristotle believed that the intellectual virtues were inculcated through training and experience and that the moral virtues were developed by habit. His view provides insight into the question of the judicial virtues as well. In order to acquire judicial intelligence, integrity, and wisdom, judges must be trained in the law, experience legal practice, and develop the habit of respect and concern for the law. For this reason, the best judges are likely to have experience as practicing lawyers, and the very best judges reach their full potential for judicial excellence only after they have spent some time on the bench. This conclusion is supported by consideration of each of the three judicial virtues.

Judicial intelligence is developed through legal training and experience. One acquires the ability to engage in legal reasoning by studying law; one develops excellence in legal reasoning through attendance at law school and by practicing law. In other legal systems, judges are civil servants who are recruited directly from law school; these judges may hone their legal reasoning skills on a lower court before promotion to more important judgeships. Lay people, however, are not likely to possess excellence in legal reasoning.

Similarly, judicial integrity is fostered by habit. Aristotle thought that a moral virtue such as courage was acquired by engaging in couragoeous actions; acting courageously enables one to become a courageous person. Likewise, the attitudes of respect and concern for the law are fostered by acting with respect and concern for the law. Such actions are required of legal practitioners, and it is through participation in the practice of lawor judging that one begins to develop judicial integrity. Lay people are unlikely to have developed the virtue of judicial integrity because they will not have had the experiences that enable one to respect the value of the law and care for its coherence.

Finally, it is quite clear that judicial wisdom can only be acquired through experience as a lawyer or judge. Legal practice allows the potential judge to develop a sense for what does and does not work as a practical matter. Legal practice exposes the potential judge to the people who are affected by the law and enables the potential judge to develop a sense of when the law is achieving something that is really worthwhile. Even if a nonlawyer could develop a sophisticated theoretical understanding of and real respect for the law, the lay judge will not possess the ability wisely to choose ends and means in particular situations.

*1756 If we return at last to our discomfort with Schauer's provocative suggestion that nonlawyers could be good appellate judges, we can see that there are many good reasons for our intuitions. Even if many appellate cases are hard cases, in which nonlegal factors such as morality, politics, or economics must play a role, it does not follow that nonlawyers could do a good job of making the decisions. Moreover, my analysis of judicial virtue, which revealed the difficulties with lay judging, has broader implications for judicial selection. The reasons for favoring the selection of lawyers over lay people as judges are also reasons for favoring the selection of excellent lawyers over mediocre ones, and for the selection of experienced lawyers over novices. In the end there are some very good reasons to favor merit selection of judges. That is not to say that we should not also consider whether the candidate believes in a particular procedure for appellate decisionmaking, nor is it a reason to reject consideration of a judge's political or moral views; many theories of judicial decisionmaking require judges who are adept at moral reasoning, economic theory, or political judgment. It is to say that even in the special corner of the law that is appellate adjudication we ought to select judges who are adept at legal reasoning, who have concern for the integrity of law, and who possess the experience to make sound practical judgments. First and foremost, judicial selection should be a matter of virtue and vice.

[FNa] Professor of Law, Loyola Law School, Los Angeles, California. B.A. 1981, University of California at Los Angeles; J.D. 1984, Harvard Law School. I owe thanks to Ken Anderson, Bob Benson, Don Brosnan, and Nancy C. Brown for comments on earlier versions of this essay.

[FN1]. See J. FRANK, LAW AND THE MODERN MIND 27-31 (1930); Bingham, My Philosophy of Law in MY PHILOSOPHY OF LAW: CREDOS OF SIXTEEN AMERICAN LEGAL SCHOLARS 5 (1941); Llewellyn, A Realistic Jurisprudence--The Next Step, 30 COLUM. L. REV. 431 (1930).

[FN2]. There is a vast amount of literature devoted to the conventional notion of 'merit selection.' See, e.g., Gottschall, Carter's Judicial Appointments: The Influence of Affirmative Action and Merit Selection on Voting on the U.S. Court of Appeals, 66 JUDICATURE 165 (1983); Vandenberg, Voluntary Merit Selection: Its History and Current Status, 66 JUDICATURE 265 (1983); Winters, The Merit Plan for Judicial Selection and Tenure: Its Historical Development, 7 DUQ. L. REV. 63 (1968).

[FN3]. Schauer, Judging in a Corner of the Law, 61 S. CAL. L. REV. 1717 (1988).

[FN4]. Id. at 1732.

[FN5]. See Wall St. J., Feb. 1, 1988, at 1, col. 1.

[FN6]. See N. Y. Times, Oct. 27, 1983, § A, at 24 col. 2.

[FN7]. These people's courts were established by decree in 1983; the judges are Sandinista party activists. The activity of these tribunals has recently been suspended by President Ortega as a part of ongoing efforts to comply with the Arias Peace Plan. L.A. Times, Jan. 14, 1988, at I, p. 14, col. 1; see also Benson, Was Nicaraguan Tribunal a Kangaroo Court? Nonsense, L.A. Times, Nov. 18, 1986, at II, p. 5, col. 1.

[FN8]. England's lay magistrate system dates back to 1361. Today's English lay magistrates are required to take 40 hours of instruction. They have jurisdiction over criminal cases in which the potential penalties include a year's imprisonment or $2,200.00 in fines. Lay magistrates handle over two million cases a year. L.A. Daily Journal, Mar. 13, 1985 at 4, col. 4.

[FN9]. See J. BAKER, AN INTRODUCTION TO ENGLISH LEGAL HISTORY 15-16 (2d ed. 1979) (describing early royal courts as composed of advisers and courtiers to the king); T. PLUCKNETT, A CONCISE HISTORY OF THE COMMON LAW 104 (5th ed. 1956) (indicating that many early traveling royal judges were not professional lawyers).

[FN10]. Prohibitions Del Roy, 12 Coke Rep. 63, 65, 77 Eng. Rep. 1342, 1343 (1608).

[FN11]. R. DWORKIN, LAW'S EMPIRE 264 (1987); R. DWORKIN, TAKING RIGHTS SERIOUSLY 105 (1978).

[FN12]. Indeed, later in this essay I rely on such theories when I discuss the notion of the rule of law and its specific requirement that the law be public. See infra text accompanying notes 42-43.

[FN13]. This essay starts with the intuitions about judicial selection that are taken for granted in ordinary discourse about good and bad judging. This method falls within the broadly Aristotelian and Wittgensteinian approach to moral philosophy represented by such diverse contemporary figures as Elizabeth Anscombe, Phillippa Foot, and John Rawls.

[FN14]. See, e.g., P. FOOT, VIRTUES AND VICES (1978); P. GEACH, THE VIRTUES (1978); A. MACINTYRE, AFTER VIRTUE (2d ed. 1984). This development is in part traceable to Elizabeth Anscombe's essay, Modern Moral Philosophy. Anscombe, Modern Moral Philosophy in ETHICS (1968).

[FN15]. See W. HARDIE, ARISTOTLE'S ETHICAL THEORY 107-08 (2d ed. 1980).

[FN16]. Id. at 99. I am inclined towards the view that such states are at bottom not only states of the mind, but states of the brain. My approach to this issue would be within the camp in philosophy of mind camp that is self- identified as 'functionalist.' See generally P. SMITH & O. JONES, THE PHILOSOPHY OF MIND (1986); D. ARMSTRONG, A MATERIALIST PHILOSOPHY OF THE MIND (1968).

[FN17]. ARISTOTLE, Nichomachean Ethics in 2 THE COMPLETE WORKS OF ARISTOTLE 1751 (J. Barnes ed. 1984) [hereinafter EN].

[FN18]. See W. HARDIE, supra note 15, at 116.

[FN19]. EN supra note 17, at 1115a 6-7; W. HARDIE, supra note 15, at 118.

[FN20]. EN, supra note 17, at 1103a 14; see also W. HARDIE, supra note 15, at 99-100.

[FN21]. EN, supra note 17, at 1140a 25-28.

[FN22]. W. HARDIE, supra note 15, at 116.

[FN23]. EN, supra note 17, at 1103a 14; see also W. HARDIE, supra note 15, at 99-100.