MIDWEST STUDIES IN PHILOSOPHY, X m (1988)

Virtues and Their Vicissitudes AMELIE 0. RORTY

en does a set of habits, skills, or dispositions qualify as a virtue? What assures W - o r can assure-that such dispositions are well and appropriately exercised? The somber answer to the first question is that character traits are classified as

virtues whenever they are admired or thought beneficial, even though they some- times conflict with one another and often fail to secure individual thriving. There is considerable social pressure to acquire and exercise such traits-sets of voluntary and discriminating dispositions, habits, capacities, and skills-even though they are not always directly rewarding or rewarded. Typically such traits are admired when they are the expression of a cultural ideal that is thought to be relatively difficult to realize, an ideal that usually involves modulating some natural tendencies such as self- protection or the desire for whatever conduces to one’s own happiness. They are re- garded as beneficial when they are thought to serve social welfare, -especially when doing so appears to involve some cost to oneself A culture can of course be mistaken about the traits that serve its thriving, failing to identify characteristics that are cen- tral to social welfare and admiring those that damage it. Entrepreneurial traits might, for example, be valued in the mistaken belief that they tend to improve the standard of living, which itself might erroneously be believed essential to a culture’s (concep- tion of its) thriving.

But at least sometimes’the dark answer has a bright side: there is some relation between socially prized traits and flourishing, if only because social esteem is one of the goods of life. With moral and political luck, the connection is stronger: traits that are socially prized are often-though certainly not necessarily-connected to at least some aspects of social and individual flourishing. That connection is sufficiently strong so that both moral philosophers and those who on the whole want to live well can at least initially be guided by social conceptions of virtues. But where there is good luck, there is also bad luck: the connection between socially recognized virtue and flourishing is also generally su5ciently weak to allow the indignant and the vi- sionary grounds for reformist proposals. Even though individual conceptions of thriving are largely socially formed and individual thriving is socially controlled, in- dividual and social thriving can conflict.’

136

The somber answer to the second question is that nothing can assure that the traits constituting the virtues are always well and appropriately used, except by the strategic, unilluminating maneuver of not counting typically virtuous traits as virtues in a person unless they are well and appropriately exercised by that person.2 But the merit of such a strategy is largely limited to the benefits of elegant theory construc- tion; it leaves us in the dark about what assures or even conduces to the appropriate use of such traits.

However skeptical we are about the pretentions of purely philosophical ethical theory to serve as a guide for action, we nevertheless do want a philosophic ethics that does more than propose a general theory about right, good, duty, virtue. There are im- plicit practical constraints and directives on normative ethical theories: they should be capable of being psychologically and educationally action-guiding in structuring and restructuring our practices toward living well, and-since failure is nearly as common as success in this area-they should also explain why it is sometimes so diffi- cult for even the best of us to succeed.

I. CHARACTER AND THE CONTEXT OF VIRTUE The virtues hunt in pack.

Before moving to the large and difficult problems we have set for ourselves, we first need to describe how those traits which we call virtues function in forming actions. When a set of dispositions constitute a virtue, the thoughts and categorial preoccupa- tions that are central to that virtue form interpretations of situations; they focus the person’s attention and define what is salient, placing other concerns in the back- ground. To act well, and to do so reliably, a person must perceive and interpret situ- ations appropriately, and do so reliably. Without appropriate cognitive structures- thresholds of attentiveness that are sensitive without being hypersensitive, habits of salient focusing that are corrigible without being distractible, imaginative habits of association that elicit relevant material without being volatile-good will remains empty.

The cognitive dispositions that partly constitute the virtues are tropic or mag- netizing: perceptions and interpretations of situations elicit the responses that are im- mediately appropriate to them. Significantly, they also generate scenarios- sequences of events-that require the further exercise of the virtues. Because they generate situations that manifest the continued need for the exercise of strongly en- trenched virtues, such cognitive dispositions are self-reinforcing. Dispositions of inter- pretation structure patterns of salience and importance: they organize the dominant proper descriptions of situations. Without waiting to be called upon, a generous per- son is perceptively and interpretively sensitive to needs, even when such needs are unacknowledged by others. Where some might perceptively focus on rqlations of power, or on aesthetic compositions, a generous person notices how she can correct what is wanting. Tropic dispositions lead a person to gravitate to the sorts of situations that predictably elicit prized character traits. Often avoiding situations where she might herself require aid, a strongly generous person tends to move toward situations in which her contributions might be useful, even when she has no desire to find herself

138 AMELIEO. RORTY

in such situations and would prefer to be somewhere else, doing something else. Tropic dispositions can often work independently of, and sometimes in conflict with, a person’s strong desires. Self-activating dispositions promote or create the occasions that require their exercise. A heroic figure, for example, can sometimes structure his foreign policy and his personal relationships in such a way that others predictably come to depend upon and even to require his bold, imaginative leadership.

When the virtues are esteemed and rewarded, they rarmfy to develop and exer- cise associated supportive traits, while inhibiting other, often highly beneficial traits. In situations of conflict, publicly recognized virtues tend to determine priorities, sometimes at the cost of highly functional, but unacknowledged traits. The magnetiz- ing dispositions of centrally organizing virtues are focal and expansive in a person’s character: focusing on what is salient blurs what is in the background.3 (The persistent and tenacious are not, for example, normally sensitive to the ironies of the contingen- cies of practical life; rarely delighting in ludicrous turns of chance, they are often so in- tent on their purposes that they are not responsive to the unexpected.

Traits do not, of course, form actions in isolation. Individual virtues underdetermine appropriate actions. They only function within a supportive, direct- ing, and sometimes oppositional network. The attribution of any trait is made against the background of ceteris paribus assumptions about an interrelated network of stan- dard operating functions. Some of the interrelations among the virtues arise primarily from their cognitive components, others from the consequences of the actions they standardly generate. Art$cially, solely for the purpose of exposition, and not because they are psychologically separable, we can distinguish the cognitive combinatonal properties of the virtues from those that are normally formed by the dynamics of their habitual exercise in action.

The cognitive components of the virtues carry the whole range of combinato- rial logical and psychological properties: logically, their cognitive contents presup- pose and entail one another, complex virtues contain simpler virtues as ingredients; they can be contraries, and even contradictories. Psychologically, their associations can be law-like; the development of the cognitive components of one virtue can pre- suppose, enhance, or block another, they can reinforce or inhibit each other, they can combine in new virtues and form cyclical patterns of vacillation and ambivalence.

The virtues are also strongly, dynamically interCOMected in socially structured narrative^.^ There are culturally fixed and socially controlled expectations that pro- mote and then replace the sequence of virtues, as appropriate to age and to role. The virtues of youth are sometimes seen as dangers in the middle-aged, those of age as in- appropriate to youth. Sometimes the successes of one virtue generate situations which require not only the replacement, but also the checking of the original virtue. For instance, the bold, inventive entrepreneurial traits of an early mercantile society can so change material and social conditions that the original virtues are reclassified as vices unless they are strongly checked by prudential and sometimes even cautious calculation. Obviously, these relations and changes do not necessarily occur smoothly or automatically. Early virtues can continue to operate even when they are judged no longer appropriate; counterpoised virtues do not always achieve an appropriate bal- ance.

139

No virtues without virtue; no virtue without virtues.

With this characterization of the structures and relations among the virtues in mind, we are better equipped to return to our original question: What assures the virtue of the virtues?

(1) The most radical solution is that of contextualizing attributions of Virtue: since the same set of dispositions can sometimes be appropriately and sometimes in- appropriately exercised, even by the Same person, traits qualify as virtues only when they have been appropriately exercised. While this solution is elegant, it appears to make the attribution of virtue redundant, reiterating-in the form of a pseudo- explanation-the judgment that the action is meritorious. In any case, this is a solu- tion provided for and by a problem in theory construction: it sets the conditions for attribution virtue without providing the analysis of its conditions.

(2) The strictest solution is the introduction of a condition for self-modulation: no set of dispositions qualifies as a Virtue unless it includes its own appropriateness- assuring conditions. To qualify as a virtue, a set of traits must be discriminatingly, in- ternally self-regulating to determine the appropriate occasion, extent, and manner for its exercise (true courage does not lapse into bravado, generosity does not lapse into wastefulness). Significantly, it must also be externally self-regulating to determine the appropriate balance with other, sometimes competing sets of traits.

But while this solution is philosophically ingenious in defining conditions for identifying the virtues, it is not psychologically illuminating, or, for that matter, psy- chologically convincing. In practice, while the individual Virtues, charactenied as a set of intellectual and practical dispositions to typical actions, are internally self- regulating, they are rarely externally self-regulating. When the action claims of var- ious virtues compete, individual virtues do not themselves determine the appropriate balance. While kindness might be self-regulating in situations in which no other vir- tues compete with its action-claims, what is it about kindness that checks its exercise when it competes with truthllness? The claim that the requirements of kindness and those of truthfulness must in the end always coincide has all the air of denying the phenomena to save the theory.

(3) The most familiar and the most thoroughly explored solution is the master virtue solution. Since definitions of the various individual virtues do not by them- selves give rules for determining their appropriate relative priority, an independent external condition, a regulative master virtue such as phronesis or caritas or Kantian good will-virtues that unite the practical and intellectual traits required to deter- mine the priority and balance among the several virtues-seems also to $ necessary. Since the master virtue solution has been developed in some detail, we should look at some of its most refined presentations.

It was because he thought that wisdom provides the necessary and sufficient condition for the development and the proper exercise of the various virtues that Soc- rates argued for the unity of the virtues in knowledge. No virtues without virtue, and

140 AMELIE0.RORTY

no virtue without knowledge. To defend this intellectualist account of virtue, Socrates had to introduce non-intellectualist conditions on wisdom: the wise must not only ap- ply what they know in argument and discussion, but also in the minutiae of action and practi~e.~ Notoriously, the Socratic condition is circular: the various traits that assure wisdom are both intellectual and practical. The wise are both good and clever: besides giving each dialectical inquirer the logos appropriate to his understanding, they also teach by good example. None but the wise are virtuous; yet none but the vir- tuous qualify as wise. A good formula, perhaps also true, but not really helpful for those of us who need to know how to become wise and good.

It was because he thought that the proper exercise of presumptively virtuous traits requires both cognitive and character dispositions-well-formed, discrimina- ting habits directed to good ends, appropriately understood-that Aristotle located the master virtue in phionesis. The internal self-regulation assured by locating each of the virtues in a mean defined by logos does not automatically assure the appropriate exercise of that virtue, in relation to all others, all things considered. But phionesis is an umbrella term for a wide range of independent traits that enable a person to see, and to actualize the goods that can best be realized in extremely varied, particular contexts. While Aristotle avoids some of the problems of the Socratic solution, he only postpones others. Since phionesis combines a range of independent intellectual and character traits-ingenuity, insight, perceptual sensitivity, acuity in inference, a sound sense of relevance, an active understanding of the relative importance of het- erogeneous and sometimes incommensurable ends, allocating different priorities to the various components of phronesis could sometimes lead to different action out- comes. A phionimos whose ingenuity was more acute than his sense of relevance, might form different actions from those performed by a phionimos with a somewhat different balance of traits. How does phionesis assure that the individual virtues- including those that compose and constitute it-are appropriately exercised, in the right way, at the right time?

It was because he thought that only a good will could assure that the various vir- tues would be rightly exercised, that Kant held that the only thing good in itself, with- out qualification, is a good will, that is, an autonomous, rational will. Kant hoped to avoid Aristotle’s difficulties in determining the priority among the various traits that compose phionesis by showing the unity of the rational good will. While in principle, the conditions that assure the will’s freedom can be distinguished from those that as- sure its goodness, Kant attempts to secure the autonomy of the will by identifying its conditions with those for practical rationality. Because the wi l l is reason in its prac- tical employment, it is self-legislative, free of external determination when it con- forms to the requirements of rationality. And since rationality requires self-legislated impartiality, practical reason assures both the freedom and the goodness of the will. Because Kant proposed to give an analysis of the conditions that make morality pos- sible, he was concerned to locate the unconditioned origin of action, the absolute lo- cus of responsibility. While his preoccupations were quite different from those of Socrates, his solution to the problem of assuring the virtues is nevertheless a variant of the Socratic solution: Kant’s claim that the commands of practical rationality are the claims of morality is parallel to Socrates’ identification of knowledge with virtue.

VIRTUES AND THEIR VICISSITUDJ3 141

The analogue of the Socratic solution inherits an analogue to the Somtic problem: as Kant himself was the first to acknowledge, the condition that assures morality is judi- cial rather than generative. Acting fiom a good will requires testing the rationality, the universalizability, of maxims of action; but practical rationality neither develops the motives nor discovers the appropriate empirical maxims for any given action. While a good will assures the possibility of acting purely fiom a conception of the moral law, it does not by itself determine actions or, for that matter, empirical motives.

Both Socrates and Kant extend their views by including the individual virtues within virtue. As the Socratically wise are also temperate, courageous, and just, so too Kant's person of good will actively commands himself to possess the individual vir- tues. In doing his duty for its own sake, the moral agent does what each duty ration- ally commands. Consistent rationality recognizes that in order for a moral intention to issue in well-formed actions, the will must necessarily be supported by the virtues. Since Kantian morality commands the acquisition of the several virtues, Aristotle's problems reappear within the Kantian frame; something like ghronesis is required to assure the appropriate connection between the purity of the moral intention and the appropriateness of a particular action.

It is because she hopes to combine the advantages of Aristotle's complex psy- chology with Kant's strict account of responsibility that Philippa Foot distinguishes the virtues from other practical dispositions and skills: they are, she says, controlled by the will, directed toward what is good.6

But the virtuous cannot assure the appropriate use of their traits by willing good ends. A villain and a good citizen can, and often do, will the same ends. To begin With, they have the same species-defined ends; they are constituted so as to have at least some of the same central needs and desires whose satisfaction constitutes thriving. Besides the necessities and comforts of life, they want esteem and friendship, and want their friends and families to thrive. The virtuous and the vicious can be ambi- tious for fame, respect, and fortune; they can even sometimes have virtually identical intentional descriptions of their general aims; and both can be prudently courageous in pursuing their ambitions. Nor is the difference that the good citizen is reliable, while the cad is not; they can both be counted on to behave in character. Sometimes it is just their petites habitudes-the configuration of their ends and h t s , the way they are courageous, prudent, tactful in pursuit of their common ends-that differen- tiate the cad and the good citizen.

Although each of the familiar versions of the master virtue solution encounters somewhat different difficulties, we can generalize that there are roughly three major problems with that solution. There is, first, the problem of selecting among the sen- ous candidates for the master virtue, particularly since each can, in principle, com- mand different patterns of dominance and recessiveness among the various virtues. When there is uncertainty and conflict in action, phronesis might well farm different priorities from those proposed by caritas, envisaged by magnanimity, commanded by the good will, or assured by justice. Is there a master master virtue?

Second, there is the problem of regression: If the master virtue is to be action guiding in particular, variable, contingency-ridden contexts, it does not act from, nor is it characterizable by, a set of rules. But then just how does the master virtue regulate

142 AMELIEO. RORTY

or check competing first-level virtues? How does it hse intellectual and practical traits to guide its determinations? Just identitjing the will as reason in its practical ap- plication does not help us understand how that identity works, particularly if the will could in principle fail to affirm what rationality commands. Even if there are some very general principles that define right or good, what determines the appropriate ap- plication of these principles through the contingencies and vagaries of particular sit- uations?

Third, there is the problem of determining the appropriate balance among the component dispositions that themselves constitute the master virtue. Phronesis-or good sense, magnanimity, caritas, or justice-are each composed of a complex set of skills, capacities, and traits. The primary candidates for the master virtue involve an appropriate level of acuteness and precision in perception, the ability to focus-and stay focused-on what is important, despite irrelevant attractions. They also involve impartiality, a capacity to form well-structured, valid inferences, ingenuity and tact, and the open-minded traits required for corrigibility combined with firmness of pur- pose. Because the master virtue itself encompasses a wide variety of potentially coun- terpoised intellectual and character traits required to coordinate first-order virtues, all the problems about the appropriate coordination of virtue-assuring dispositions reap- pear. The regulative intellectual and character virtues are also action-forming; and first-order virtues can sometimes also function as regulative master virtues. The dis- tinction between first- and second-order virtues appears to be a contextdependent difference in functional role rather than a distinction between types or traits.

Indeed, it seems as if the master virtue is not one trait, not even one way of a p propriately coordinating capacities, dispositions, traits: it is nothing in particular, over and above having a well-constructed character that tends to act well and appro- priately. If excellence of character is assured by. anything, it is assured by an appropri- ate configuration of traits, rather than by a single trait or by a conjunction of traits. But since each situation requires a slightly different configuration of traits to produce an appropriate action, and since different people might require different configura- tions in different situations, the appropriate configuration of virtue-assuring traits cannot be specified.

The difficulties with the master virtue solution suggest another, less familiar so- lution to the problem of assuring the virtues. (4) The next solution might be described as a checks-and-balances solution, for it locates the virtues in a system of supportive and tensed traits. A set of traits qualifies as a virtue only when it is supported and bal- anced against other traits in an appropriate pattern or configuration. The traits that constitute the virtues hunt in role-differentiated packs, not only requiring one another to determine particular actions, but also, darkly and significantly, to modify, to check and balance, one another's exercise. Long after moral philosophers abandoned So- cratic theories of the unity of the virtues, they still retained the Platonic assumption that the various virtues form a harmonious system directed to the same general ends. However differentiated in function, the virtues were assumed either to coincide in forming the same extensionally identified action or-more frequently-typically to coordinate and support one another's exercise and action outcomes.

VIRTUES AND THEIR VICISSITUDES 143

By contrast, the check-and-balances solution does not assume that the virtues form a mutually supportive, harmonious system directed toward realizing compati- ble ends. On the contrary, according to this solution, at least some of the virtues are assured because their typical cognitions and actions are dynamically opposed to one another.' Virtues can check one another in a number of ways. (1) When the cognitive component of the traits that constitute them are contraries or contradictories, they will constrain each other. (The cognitive set of a particular devotion or commitment can, for instance, sometimes be contrary to that of impartiality.) (2) Virtues can check each other when the traits that compose them are typically exercised in actions whose outcomes standardly undermine or flustrate one another's intention, direction, or satisfaction. (The habits of consensual sociability are, for instance, tensed against those exercised in strong independence.) (3) The development of one set of traits can typically inhibit the development of another. (The habits developed in the service of a rule-bound bureaucracy tend to block those required for improvisatory re- sourcellness.) (4) The exercise of one set of traits can produce conditions that require the exercise of an opposed set. m e combative soldierly Virtues that sometimes bring peace can endanger the fragile trust on which the preservation of that peace depends.)

It is of course not possible to give a general rule or principle to determine the proper mecLpure of any particular balance among tensed or opposed virtues. Some- times one virtue might entirely and appropriately block another, sometimes two virtues modulate or diminish one another's force; sometimes one virtue appears re- cessively within the action determined by another, modifying the way the dominant virtue is expressed, hesitantly or ambivalently. The whole range of combinatorial properties of the cognitive and habit-based traits are brought into play.

Unfortunately, despite the advantages of its dark and rare truthfulness, the checks-and-balances solution seems as regressive as the master virtue solution and nearly, but not quite, as uninformative as the contextualizing solution. Since there is no general rule or principle to determine it, what assures the appropriate balance in each situation? To make matters worse, there seem to be competing criteria for a well- ordered appropriate configuration, which varies with the weight assigned to each vir- tue. Do the difticulties of the system of checks and balances lead back to the master virtue solution in order to evaluate competing criteria for appropriate balance- which in turn requires a system of checks and balances among the various compo- nents of the master virtue?

While there are undeniable problems for the checks-and-balances solution, I believe that it nevertheless presents an advance over the other solutions. For one thing, it is psychologically truthful, for another, it presents some rudimentary heuris- tic guidance for maintaining the virtue of their virtues by locating them in a logical and psychologically dynamic field.

But here we are, once again, in the old familiar circle. Nothing in particular seems to assure virtue-nothing less than the whole of a virtuous character, well formed in a system of checks and balances, so as to assure its proper activity, acting appropriately as the situation requires, i.e., acting virtuously. Where can we go from nowhere?

144 AMELIE0.RORTY

III. COMMUNITY AS THE CONTEXT OF CHARACTER No action without interaction; no interaction without politics.

But perhaps our somber reflections are premature. Perhaps we have been looking at the wrong place. Virtue is, after all, primarily and fundamentally an attribute of hu- mans rather than of actions or sets of traits. It is the man, Aristotle remarks, who is the source of action: the regression of traits that assures the appropriate balance of the virtues stops at the person, the total configuration of character, rather than at a master regulatory virtue or principle.8 But if the man is the source of action, we need to un- derstand what forms the man, what leads to his being the sort of person who is capa- ble of virtue. According to Aristotle, a certain sort of psycho-physical constitution is required, as well as appropriate economic and socio-political conditions and the good fortune of a person’s early models. The assurance of a person’s virtues cannot occur in a social or political vacuum.

Action takes place in a social world. It is, in the end, our social and political re- lation to others that keeps our virtues in whatever precariously appropriate balance they have. Minimally the cooperation and esteem we require from our fellows elicit the appropriate balance among supporting and opposing virtues. Significantly, our ac- tions have their sense, their meaning, and their direction in a public, interactive world. This is not just a consequence of the fact that political institutions shape our attitudes and social structures form our views, although this is of course true. Nor is it solely a consequence of the fact that our actions often so change our social and po- litical circumstances, that they in turn require us to change, although this is of course also true. In addition, we have been culturally influenced by the ideal of offering a so- cial justification of the principles that guide our actions.

More to the point of the assurance of the virtues, our actions are dramatically and substantively formed by our minute interactions with others. The details of most of our actions are determined through a subtle process of interaction. Think of the way in which a conversation, a real conversation, a common investigation rather than a comforting ritual or an exchange of monologues, takes place. In a real conversation, the participants do not know, ahead of time, what they will say, or even sometimes what they think. To be sure, there are constraints: interlocutors want (among many other things) to arrive at what is true, and they are guided by what they presently be- lieve is true. But at any moment in the conversation, there are an indefinite number of relevant, consecutive true things they could say and think. Closure is given by the minutiae of interactions: the look of puzzlement on an interlocutor’s face, the excite- ment of common pursuit, an ironic remark. The more subtly partners in a conversa- tion understand each other-the more they are familiar with one another’s gestures, facial expressions, and reactions-the more condensed and improvisatory their con- versation is likely to be. Like jazz musicians, they sometimes lapse into a familiar riff for a little rest, finding something in that riff that leads them in a new direction. Not only conversations and music-making, but many of our central actions-designing a playground or a curriculum, cooking a meal, selecting a Supreme Court Justice, hang- ing paintings for an exhibition-take this form. Even when we act in solitude or in

VIRTUES AND THEIR VICISSITUDES 145

character, from our deeply entrenched traits, the actions we perform emerge from an interactive process that sometimes takes place in for0 interno. Some of our interactive partners elicit our (very own) boldness, others elicit our (very own) caution. The con- figuration of a person's traits-the patterns of dominance and recessiveness-that emerges in any given situation is affected by her interactive company.9

We characteristically respond to a skeptical interlocutor in one way, to a con- firming interlocutor in another, to a co-explorer in yet another. The views we form as a result of our common investigations-views that each interlocutor genuinely holds as her own-are, in their fine grain details, co-produced, even when the interlocutors end with markedly different views.l0 As with views, so with the balanced pattern of our virtues. Character, in all its constitutional and socio-political configuration, reg- ulates the particular virtues; and community regulates character. It is these that, taken together, hold the virtues in check. But then it is these that, taken together, can also lead to vice.

Obviously, the interactive context of the formation of action cannot assure vir- tue, cannot determine the appropriate configuration of our typically virtuous traits. Even the shiningly virtuous are sometimes corrupted rather than supported by the company they keep. Nevertheless, once we know the patterns of our interactions, we can-bearing in mind our patterns of contrariness-strengthen an appropriate config- uration of traits by being carehl of the company we keep. But since we can also vol- untarily erode what we take to be a relatively virtuous configuration, all the problems of assuring the virtues are again postponed. What determines the company we choose to keep? Have we returned tophronesis?Or to a checked-and-balanced cycle of domi- nance and recessiveness in the configuration of traits, this time extended to the com- pany to which we gravitate, a characteristic cycle of high-minded and low-life company, or of solitude and random sociability? More soberly, does the interactive social determination of the checks and balances of the virtues throw us into the power politics of the control of the virtues? Aren't we playthings of the moral luck of our po- litical and social situation, the luck of the draw of our interactive community? After all, we are subject to the power of charisma and of interest groups all contending to define the dominance and recessiveness of their prime candidates for our virtues: there is no assurance that the outcome of such struggles issues in an appropriate bal- ance.

It is indeed just for this reason that some moral philosophers attempted to re- place theories of virtue with theories of rules and principles. Complex, pluralistic, and dynamic societies with genuinely opposed needs, values, and interests are likely to have competing models of virtue. Acknowledging the motto "No virtues without vir- tue," they add: 'No virtue without rules and principles." Sometimes, when such soci- eties recognize that they require a variety of opposed models, they succeed in formulating general procedural rules and principles to adjudicate among-the claims of competing models by placing them in a system of checks and balances. But there is no guarantee that there will be agreement on such procedural principles: when there are competing models for the proper balance of the virtues, there is also usually dis- agreement about principles, including those procedural principles regulating the ad- judication of disagreement.11 Finding an overlapping consensus on rules for adjudi-

146 AMEllEO.RORTY

cating disagreement introduces justice as the master Virtue. A set of principles governing procedures for adjudicating disagreements appears to have the special ad- vantages of being self-referentially capable of arbitrating competing claims about its own formulation. But the self-referential closure of procedural rules and principles is broken when there are competing claims about their priority and formulation, par- ticularly when differences in formulation and priority among procedural principles would issue in distinctive policies and actions, or form a different community. Al- though it seems initially to be a liberating solution, the attempt to provide procedural rules and principles for resolving conflicts has all the advantages and disadvantages of the master virtue solution.

It seems that we have, despite everything, returned 111 circle to the somber an- swers to our original questions. Traits are called virtues when they are culturally re- garded as admirable or beneficial, when there is social pressure to develop and exercise them. When-as is normally the case in complex and dynamic societies- there are competing and sometimes conflicting models of benefit and admirability, we can attempt to find or form a consensus on procedural principles ofjustice, or de- velop a solidly operating system of checks and balances among the contending mod- els of virtue. Nothing can assure that competing models of virtue in a polity will be appropriately balanced, rather than merely determined by the accidental play of power politics. We can try to characterize the master virtue and the particular system of checks and balances among the virtues that, given existing conditions and domi- nant motivational structures, are most likely to assure the appropriate exercise of the various virtues. The proposal to make proceduraljustice the master virtue is just such an attempt.

But sometimes there just is no appropriate way to assure the balance among virtues. At worst, the emergent balance is determined by the power of charisma and the power of interest groups. With luck, the distribution of power is so structured as to produce a cycle rotating the benefits of each model of virtue.

These dark conclusions have led many moral theorists to elect to do program- matically normative moral theory, arguing that before we can direct operative psy- chological processes for assuring the appropriate balance of the virtues, we must determine the most general conditions of appropriateness. Unfortunately, the con- straints set by characteristic social and psychological process-the laws describing the conditions and effects of various power relations, for instance-directly enter into the determination of the criteria for appropriateness. The constraints of applicable prac- ticability, of psychological and politid realizability, appear within the determination of appropriateness. The difficulties of defining the appropriate balance among com- peting virtues reappear as difficulties in arbitrating among competing normative moral theories.

Reconstructivist moral philosophers who propose to assure Virtue by offering either rules for or an imitable model of acting well tend to bracket the contingencies that affect moral luck the luck of a person's constitutional and intellectual traits; the luck of having appropriate formative models to imitate; the luck of living in histori- cal, economic, and socio-political conditions that are consonant with one's own dis- positional directions; the luck of good company. But it is just the strength of practical,

VIRTUESANDTHEIRVICISSITUDES 147

descriptively oriented virtue theories that they acknowledge the pervasive presence of moral luck The virtues are, among other things, the range of skills that enable a per- son to cope with luck, to deal with the contingencies and vagaries of the particular sit- uations.

It seems we end where we began, recognizing that although there is much to be said about the detaiIs of the appropriate balance of checked and opposed traits, there is nothing in particular that assures virtue in general. Another equally sober reflection emerges from our investigation. Even action-guiding practical theories cannot-and indeed should not-provide moral solutions where virtuous moral agents have moral problems. The more practical an ethical theory, the more it reflects the sorts of diffi- culties that virtuous agents have, and the more clearly it locates and explains our fail- ures. Ethical theories designed to be practical and action-guiding cannot reasonably be expected to provide salvation where none is to be had.I2

Notes

1. Cf Michael Walzer, Interpretation andSociaf Criticism (Cambridge, Mass., 1987); JOM- than Lear, "Moral Objectivity," Objectivity and Cultural Divergence, edited by S . C. Brown (Cambridge, 1984).

2. Philippa Foot, "Virtues and Vices," Virtues and Vices (Berkeley, 1978), 7 K Following this strategy, Foot says, "The villain's courage is not a virtue in him."

3. Cf. "The Two Faces of Courage," Philosophy (1986): 5fF. 4. Cf Michael Slote, G o d and Virtues (Ithaca, N Y , 1983), chap. 2; Alasdair MacIntyre,

5. "The Limits of Socrates' Intellectualism: Did Socrates Teach Virtue?" Proceedings of the

6. Foot, "Virtues and Vices." 7. There is a tradition according to which virtue is assured by the balance among opposed

virtues. That tradition is represented by Plato, in The Statesman 31 1B-C: "Those who are care- N, fair and conservative-those of a moderate temperament-are not keen; they lack a certain sort of quick, active boldness. The courageous on the other hand are fkr less just and cautious, but they are excellent at getting things done. A community can never function well. . . unless both of these are present and active . . . woven together by the ruler." Cf. also Hume, Treatise of Human Nature 11.2. 1-12 and Spinoza, Ethics I11 Scholia and Defmitions, for detailed accounts of the ways in which the passions function within a dynamic system of support and opposition. The opposition of virtues is explored in Book I11 of The Fairy Queen where Spenser describes a duel between the Knight of Temperance against the Knight of Chastity.

"How Virtues Become Vices."

Boston Area Colloquium in Ancient Philosophy, vol. 3, edited by John Cleary (1986).

8. I am gratehl to Alasdair MacIntyre for stressing this point. 9. Stephen White pointed out to me that it might seem that interactive action formation

could be analyzed as a sequence of microactions. Msays x in mannerf; which serves as a stim- ulus for N to say y in manner g, which serves as a stimulus for M to say z in manner h. While a conversation or an improvisation could be analyzed in that way, such an analysis would miss the formation of the action as a whole. Of course any action can be broken down into a series of mi- croactions: not only such complex actions as emigrating, undertaking to follow a course of study, but also swimming across or walking around a pond can be decomposed into micromovements. But while such an analysis might explain the details of each micromovement, it would not ex- plain the form of the sequence, taken as a whole, in relation to other action sequences, taken as a whole. To understand the structured sequence of microactions as forming a complex whole, we refer to a shared general intention that integrally encompasses the interactive process: partners making music together or having a conversation.

10. David Wong pointed out to me that it might seem as if the greater the role assigned to the interactive structuring of actions, the less of a role does the configuration of character play

148 AMELIEO.RORlY

in acting well. It is true that we often not only have Merent types of conversations with Merent interlocutors, but even &rent types of conversations on the same subject. The emergent de- tails of a person's views on a topic are strongly influenced by the views and the characters of her interlocutors. But both are required: a particular interactive response is drawn from a person's repertoire of character traits and it is, as we say, characteristic. Cf Jonathan Adler, "Moral De- velopment and the Personal Point of View," Women and Moral Theory.

1 1. Cf. John Fbwls, T h e Idea of an Overlapping Consensus" and 'The Priority of Right and Ideas of the Good in Justice as Fairness." See also Jiirgen Habermas, The Theory of Communi- cative Action (Boston, 1987).

12. This paper grew out of many conversations, held over a long period of time: Rudiger Bittner, Larry Blum, Sissela Bok, Owen Flanagan, Genevieve Lloyd, Alasdair MacIntyre, Georges Rey, William Ruddick, Michael Slote, Stephen White, and David Wong have helped shape it, as did the participants in a colloquium at the University of Maryland.