Perfectionism and Politics*
(published in Ethics, 111, No. 1 (October, 2000)
Richard J. Arneson
Philosophers perennially debate the nature of the good for humans. Is it subjective or objective? That is to say, do the things that are intrinsically good for an agent, good for their own sakes and apart from further consequences, acquire this status only in virtue of how she happens to regard them? Or are there things that are good in themselves for an individual independently of her desires and attitudes toward them? The issue sounds recondite, but has been thought to be pregnant with implications for politics as it ought to be. Plato vigorously insists that knowledge of the good is precious and the person who has it is uniquely fit to rule:
"In the world of knowledge, the last thing to be perceived and only with great difficulty is the essential Form of Goodness. Once it is perceived, the conclusion must follow that, for all things, this is the cause of whatever is right and good; in the visible world it gives birth to the light and to the lord of light, while it is itself sovereign in the intelligible world and the parent of wisdom and truth. Without having had a vision of this Form no one can act with wisdom, either in his own life or in matters of state."
More recently, John Rawls considers briefly a specific version of the doctrine that an objective and comprehensive account of human good should inform the theory of justice. He calls the view the "principle of perfection," and characterizes its strongest form as a "teleological theory directing society to arrange institutions and to define the duties and obligations of individuals so as to maximize the achievement of human excellence in art, science, and culture." If the standard of excellence is set high, we get a doctrine that says social life should be ordered so as to promote the highest achievements of a few geniuses. Rawls sees perfectionism as the enemy of the liberty and autonomy that are the birth right of all individuals in a just and liberal society.
The term "perfectionism" tends to convey the idea that the best life for humans is the most perfect they can live, the kind of life that is the best expression of their nature. I shall stipulate that for the purposes of this essay a perfectionist doctrine of human good holds that what is good for its own sake for a person is fixed independently of her attitudes and opinions toward it, that it constitutes an ideal way to live that an individual might attain to a greater or lesser extent, and that in principle cardinal interpersonal comparison of the amount of good that different individuals achieve for themselves over the course of their lives is possible. A version of perfectionism, like the one on which Rawls’s discussion focusses, that identifies the good with a high standard of excellent achievement that few can attain and sharply discounts the value of other putatively good experiences and attainments, I shall label "narrow perfectionism."
Plato embraces and Rawls rejects the project of basing the account of social justice on knowledge of the good, but both philosophers associate this project with distinctive political views, which many have found repulsive. Plato holds that perfectionism implies autocracy. Those individuals who can attain knowledge of the good are few in the best of circumstances, and those few are uniquely fit to rule. The knowledgeable few should be political rulers, not the ignorant many, as in a democracy, and moreover the knowledgeable few, not each individual herself, should be entrusted sovereignty over the administration of each person’s life. Rawls associates at least the narrow doctrine of perfectionism with elitism, the latter being the idea that the proper function of political society is to serve the interests of a minority of its members. Elitism follows from perfectionism if one supposes that society should be arranged so as to promote the human good, that human good is a rare and fine achievement, and that only a few people are capable of leading lives that achieve any human good. A society dedicated to promoting human good would then devote itself to promoting the goodness of the lives of a select few of its members. Even a broader perfectionism invites the same objection insofar as the official embrace by a society of any principles of human good that not all its members can achieve puts the state in the business of promoting the interests of some but not all members.
This essay examines the implications of perfectionism for politics with a focus on the issues of illiberal interference with individuals’ pursuit of their own chosen ends in private life and elitist ordering of society for the good of a few of its members. How might an account of what is objectively good for humans reasonably shape our views of just social and political arrangements? I pursue this question with a special concern for the degree to which there is conceptual room and also, as it were, some plausible perches, within the space of perfectionist views, for political principles that are nonelitist, recognizably liberal, and egalitarian. This concern leads to an assessment of the significant attempts by Amartya Sen and Martha Nussbaum to develop a liberal perfectionist egalitarian approach to politics that has been dubbed "Aristotelian Social Democracy."
The short answer to the question, what perfectionism implies for politics, is "not much." Or more circumspectly: It all depends. What kind of perfectionism is held to be true, what views of political morality this perfectionism is yoked to, and what empirical conditions hold in the setting to which perfectionism is to be applied all shape the doctrine’s political implications.
PERFECTIONISM AND ELITISM
Consider the effect of asserting one or another variety of perfectionism about human good against the background assumption that the correct master principle of political morality is Lockean libertarianism. According to the Lockean libertarian, each adult person has strong negative rights, which neither any other person nor any social agency may violate under any conditions. Each person has the right to do whatever she chooses, in concert with other persons on any mutually agreeable terms, just so long as the chosen action does not wrongfully injure nonconsenting persons in ways that would violate their Lockean rights. Each person has the right not to be wrongfully harmed by others in ways that would violate her Lockean rights, these rights being understood to include rights against coercion, physical assault, theft of one’s property, fraud, the suffering of physical damage in consequence of the actions of others without one’s consent, and the threat of any of the above. One’s right to act as one chooses in concert with others so long as one does not wrongfully harm others who do not consent to bear these costs includes the right to suffer any purported injury to which one voluntarily consents.
To a first approximation an objective account of human good would be an idle wheel if it was added to the Lockean account of rights. Assume a Nietzschean account of the good of the kind sketched by Rawls, which would hold that only the highest achievements of the highest specimens of humanity have any value. This assumption does not alter Lockean rights to any extent. A poor specimen of humanity who chooses cheap thrills on the low road in life has the same Lockean rights as any other person, even the highest specimen, no more and no less. Assume that some few special people have been blessed with a perfect vision of the Platonic Form of the Good, and are thus uniquely filled with wisdom and uniquely suited to rule the state. This assumption alters not a bit the distribution of Lockean rights. Those who have experienced the Form of the Good and those who missed the sight have exactly the same Lockean rights.
This is only so to a first approximation. For one might protest that what counts as wrongfully harming someone so as to violate her rights depends on what counts as harming someone, and what counts as harm depends on what counts as good. This is so, but does not really affect the Lockean theory, which already is committed by its own lights to an objective understanding of rights violations.
For the Lockean, vast stretches of human affairs proceed by way of mutual consent, and the moral permissibility of these transactions and their immunity from legitimate interference depend only on the quality of the voluntary consent each party gives to the transaction, not at all the character of objective goods and harms that may accrue to the parties. Also, my Lockean rights include no rights to be given positive assistance, aid or nurturance by others, so whatever an objective account of human good might correctly say about the myriad ways in which you could enhance my life and improve its quality by interacting with me in any of an indefinite number of ways has no bearing at all on what you owe me by way of conduct.
For these reasons the assumption that we have available an account of what is objectively good for people for its own sake only very slightly affects what people may and may not do to one another with moral legitimacy according to Lockean libertarianism.
In some cases, as with libertarianism, the theory of right fences in the theory of good, so that its specific features hardly matter. With many other conceptions of right, this is not so, particularly for views that identify justice with promoting best outcomes rather than with respecting given moral constraints. Within this class of consequentialist views, the requirements and recommendations of the theory of right vary depending on the substance of the theory of the good that is conjoined to it, provided the good is allowed to influence what qualify as morally preferable outcomes.
For example, one response to Rawls’s quick dismissal of perfectionism as an option for the theory of justice is that he too quickly assumes that a perfectionist theory must assume a Nietzschean form, and hold that society should be arranged and individual duties set so that the level of the very highest perfectionist achievements of those whose achievements are highest is maximized. Perfectionism is taken to be tied to a maximax view of justice, which requires us to maximize the maximal achievements. But it is arbitrary to presume without argument that a perfectionist theory must take a teleological form that builds in a kind of elitism from the outset. On a maximax view, the lives and perfectionist achievements of all but a very few persons have no moral value. Those who cannot reach the highest achievements should be induced to lead their lives in ways that contribute to the highest achievements for the most perfect specimens and may legitimately be used as instruments to this end. Perfectionism, one might suppose, would not lead to these elitist recommendations if its political morality is egalitarian.
But this supposition is not correct for all significant cases. We might embrace a maximin theory of right. In the context of perfectionism, this yields the position that society should be arranged so that the perfectionist achievements of the person whose perfectionist achievements are least should be made as great as possible. Consider the maximin view in the leximin variant, which gives ancillary instructions as to what should be done if the perfection of the least perfect is maximized and there are resources to do more. Leximin perfectionism would hold that society should be arranged so that, as a first priority, the perfection of the person whose perfection is least is maximized, as a second priority, that the perfection of the second-least perfect person is maximized, and so on up to the person whose perfection is highest. Maximin (leximin) policy tilts to a maximal degree in favor of improving the human perfection of those who are worst off in this respect. But to this political morality one might add the theory of the good that Rawls mentions in connection with his discussion of perfectionism. This theory assigns no value to achievements and attainments except those at an extremely exalted level. On this view, for all practical purposes leximin and maximax could yield the same recommendations. If virtually all persons are incapable of attaining any perfections that have any positive value, then nothing can be done, from the standpoint of perfection, for these persons. If the level of significant and worthwhile achievement is set sufficiently high, then only the achievements of the very best register as having potentially any moral value, so urging us to give priority to boosting the perfections of the least perfect and urging us to give priority to boosting the perfections of the most perfect comes to the same.
This result does not mean that what we might call democratic or egalitarian versions of perfectionism must be indistinguishable in practice from versions that build in a highly inegalitarian goal directly into the theory of right. My point is simply that the adoption of an egalitarian principle of right does not suffice to determine that a perfectionist political morality on the whole will not be starkly elitist in its recommendations and assessments.
If perfectionism is to avoid recommending elitist policies, it must assert an egalitarian theory of right and a theory of human good that assigns significant positive moral value to the perfectionist achievements that the ordinary mass of human beings can feasibly attain. If a nonelitist and noninvasive perfectionism is to be a viable option in normative political theory, these components of the view must be defensible.
ILLIBERAL RESTRICTION OF FREEDOM
A political theory might be nonelitist in its view of the nature of human good and the proper principle for regulating its distribution, but illiberal in its conception of how politics should be arranged so that human good is achieved. A perfectionist doctrine could follow Plato but not Nietzsche. But it is immediately obvious that perfectionism can be yoked to many different moral, metaphysical, and empirical claims, only some combinations of which support illiberal governance. John Stuart Mill along with Thomas Hill Green and other nineteenth-century British writers sees a liberal political order as the best vehicle for delivering perfectionist values.
What type of politics makes most sense depends not only on the claim that human good is objective but on its content. One might hold with Mill that what is good for a particular individual depends on her nature, and individual natures differ. Even if there were one good for all persons, individual humans are so various that very different plans of life will suit different individuals aiming to secure as much of this common human good as possible. Mill adds that our individual natures are not transparent to ourselves, so we have to go through a process of discovery involving experiments in living, self-observation and self-culture, and general knowledge acquisition to learn about our nature and our good. Hence we need wide individual freedom.
Of course these assertions taken together do not suffice to rule out all paternalistic rule. From the fact that what is good for a person is relative to that person it does not follow that the individual herself is the best judge of her good. A wise statesman or for that matter an observant back-fence neighbor might discern what is good for Smith better than Smith. But the person other than Smith who knows best what is good for Smith might not be reliably motivated to seek Smith’s good, whereas, human nature being what it is, an individual can ordinarily be trusted to have a strong incentive to seek her own good.
The perfectionist cannot too easily rely on this motivational asymmetry, however, because the promptings of ordinary self-interest as ordinarily conceived and felt might not have much of a tendency to prompt me to seek my true perfectionist good. My self-interested desires might prompt me have prudent regard for the size of my bank account and the sheen of my fancy car, but not to pursue my true good, which might be stern medicine. But it is a difficulty for one who would design a blueprint for an illiberal perfectionist utopia that those who are cognitively wise might not be motivated to pursue wisdom for humanity at large according to the perfectionist vision.
Some liberals seek to buttress these doubts against the likely efficacy of coercion or manipulation directed against an individual in order to improve the quality of that person’s life, as assessed by perfectionist standards, over the course of her life. Their claim is that such paternalistic coercion and manipulation could not possibly succeed, because the perfectionist quality of an individual’s actions depends on the quality of her intentions, and threats and other manipulations (other than rational persuasion) cannot improve the quality of an individual’s intentions. Achieving the good is a perfectionist achievement only when one intends the good and intends it for the right reasons, not for ulterior motives such as threats and manipulation induce when they are effective. The claim is often stated by appeal to autonomy. An intervention that interferes with another agent’s choice of conduct cannot increase the perfectionist value of that conduct if the interference lessens the degree to which the conduct is autonomously chosen. An agent chooses autonomously to the degree that she chooses her action from a wide array of alternatives that has not been deliberately reduced by the actions of others, she chooses on the basis of her own perception of the strength of the reasons favoring the different courses of action available, and her perceptions of reasons are formed by background values that have emerged from the agent’s independent reflective scrutiny.
These attempts to buttress the perfectionist case against paternalism fail, and fail in ways that suggest the issue is wide open. First, even if one accepts the claim that nothing can be valuable unless autonomously chosen, this does not rule out the possibility that perfectionism might justify paternalism. Restriction of choice now even to the extent of forcing a single choice upon the individual can pave the way to autonomous choice of that good or others later. This long-run fostering of autonomous choice of the good is easier to bring about if the coercion does not force upon the agent a single option thought to be good but instead prohibits some tempting bad options while leaving many other options open. Second, it is scarcely plausible that no perfectionist value can be attained except via autonomous choice. To undertake action involves exercise of agency, but agency can be exercised, and in ideally admirable ways, under conditions of coercion and duress and when the agent is moved to act by motives other than the perfectionist good that is being achieved. An actor might perform splendidly when motivated solely by the pay for the work done. The mercenary motive may subtract from the perfectionist value of the achievement but does not reduce it to nothing. A slave might be forced to do theoretical physics, and do it surpassingly well; again the conditions under which the achievement is made do not extinguish the value of the achievement.
One might try to avoid this result by insisting that even if the achievement is fine, it does not make the life of the agent better than it would have been absent coercion. But why not? One can achieve goods that are valuable and whose value enhances one’s life even if one fails to appreciate their value or recoils from one’s own achievement in response to its coerced or manipulated character. No doubt achievement does more to enhance an agent’s life, other things being equal, when the agent wholeheartedly endorses the doing and properly rates its value. Autonomy is a value, and it may well be a value that affects the value of any chosen activity, but it is still one value among others, and can be outweighed. The arguments that try to establish on conceptual grounds that coercion and manipulation of an agent cannot improve the perfectionist quality of her life do not succeed and to my mind do not suggest considerations that decisively strengthen the empirical presumption against perfectionist-inspired paternalism.
Whether a perfectionist theory of value should lead one to advocacy of judicious coercive or manipulative paternalism depends also on the theory of right one should embrace. Consider the example of Mill’s argument for an absolute prohibition on (certain kinds of) paternalism in On Liberty. Mill argues for the speculative conjecture that in the long run, the aggregate sum of human utility understood along perfectionist lines will be greater if policy makers and individuals adopt a strict and exceptionless rule against paternalism than if they adopt any other policy in this regard. Mill’s conjecture might be correct or incorrect. Just suppose for the sake of the argument that it is correct. This would not settle the issue concerning the justifiability of paternalism. For our political morality even if consequentialist should not affirm that one should always do what maximizes total human good, for the reason that we should care about the fair distribution of human good across persons as well as its aggregate sum.
Mill’s conjecture that strict no-paternalism does best to maximize the sum of human good, if correct at all, might well be correct because this policy, though it predictably tends to diminish the welfare of some agents who are poor choosers, achieves more than compensating gains for individuals who are good choosers. Individuals vary in their talents at value-forming, choice making, and choice executing. Call those who have more talent of this sort "good choosers." Some good choosers will lead bad lives overall either because their circumstances are bad or they suffer bad luck even though they always choose well (in this context we can ignore coming to a bad end because one chooses well for the sake of others). But on the average, good choosers will tend to do better in life and attain better lives than poor choosers. This means that even if Mill’s conjecture proved to be correct, this might well occur through boosting of the lifetime welfare of those who have more welfare to an extent that exceeds the effect of diminishing the lifetime welfare of those who have less. Any egalitarian principle of social justice that tells us to prefer gains in welfare for the worse off to gains in welfare for the better off will give extra weight to the welfare losses that a strict no-paternalism regime might inflict on the worse off sector of the populace and will therefore favor paternalistic policies as morally right in circumstances where Mill’s straight aggregation political morality would recommend strict no paternalism.
To summarize: if society is organized to maximize some function of the perfection its members achieve, in principle coercive state interference with individuals’ personal lives for their own good might be warranted, if it boosts the coerced individual’s achievement of perfection and is part of an overall best strategy for maximizing the appropriately weighted sum of perfection. But one may well have familiar liberal reasons for doubting that such paternalistic interference will very often do more good than harm, measured in perfectionist terms. The liberal perfectionist will be suspicious of invasive state interference into private life as a matter of empirical lore, not rigid constraining principle. Attempts by philosophers from diverse standpoints such as John Stuart Mill and Ronald Dworkin to defend a strict principle against paternalism to boost perfection are not plausible.
LIBERAL AND EGALITARIAN PERFECTIONISM
To this point in this essay the strategy has been to explore the possibilities for a nonelitist and liberal perfectionist politics by way of general and abstract considerations. However, some points representing possibilities in the space of perfectionist democratic theories have actually been developed in some detail. If one of these actual existing theories proves viable, then further worries about the possibility of liberal egalitarian perfectionism would be otiose. The next task then is to examine actually existing perfectionisms. For a start, this essay turns to assess the "Aristotelian Social Democracy" account as worked out in a series of writings by Amartya Sen and Martha Nussbaum.
The starting point is Sen’s analysis of capabilities and functionings conceived as the basis of interpersonal comparison for a theory of social justice. The need for some measure of interpersonal comparison arises for theories of justice that posit obligations that apply to individuals and to society (individuals regarded as a collective) to provide significant assistance to improve the life prospects of those who are unlucky and unfortunate and face grim life prospects absent such assistance. Any theory of this type that supposes the lucky and fortunate ought to aid the unlucky and fortunate needs some measure of people’s condition so that one can identify who is fortunate, who unfortunate, and thus who ought to be aiding whom.
Sen’s writings develop a sustained criticism of the two types of proposal for the role of interpersonal comparison measure that have figured prominently in recent theories of justice: resources and welfare. The details of these criticisms do not matter for our purposes. As we shall see, Sen’s view, though not explicitly perfectionist, points toward the perfectionist claim that the fair treatment of persons that social justice requires cannot be known except by appealing to a sound account of human good that enables us to determine what treatment enables the person to flourish, to lead a truly choiceworthy, genuinely good mode of life.
REAL FREEDOM, CAPABILITY, AND FUNCTIONING
Sen proposes that the theory of social justice should assess people’s condition in terms of the real freedom they have to live their lives in ways they have good reason to value. We can list the doings and beings of different kinds that people achieve. Call these doings and beings "functionings." A person’s achievement is a vector of functionings. A person’s capability to function is the set of vectors of functionings that the individual could choose. Each set is a combination of the different functionings that an individual could choose that are co-possible. Sen plausibly identifies the person's functioning capabilities with her real freedom to choose one or another mode of life. We may be especially interested in the person’s capabilities over the course of her life, the various life options among which she can pick one. We should also distinguish the person’s broad capability to do or be whatever she has reason to value and the narrower capability to do or be in ways that constitute her well-being.
The set of functionings that an individual manifests at a moment is huge, wildly heterogeneous, and mainly consists of trivialities such as the slight twitch of my nose to the right. For almost any such trivial functioning there will be an infinite number of alternative trivially varying functionings (nose-twitchings) that might have been reached, that constitute a small aspect of the person’s overall capability. Some regimentation is needed to separate the significant from the trivial and highlight the former. Sen tends to be non-commital about which of the various ways this regimentation might be made is to be preferred. Nussbaum explicitly advocates taking the relevant capabilities for social justice purposes to be those picked out as having value by an objective theory of human good.
The principle of justice that Nussbaum espouses asserts that the first priority of justice is to bring it about that every person gets a decent level of capability for each of the functionings that are needed for a genuinely good quality of life. For each person, what justice requires is not maximizing any aspect of her condition, but satisficing: making sure that she has a sufficient level of capability.
The Nussbaum-Sen use of perfectionism in a theory of social justice thus directly addresses the worry about elitism associated with Nietzsche and the worry about elitism associated with Plato. The satisficing egalitarianism aspect assuages the former worry and the focus on capabilities or real freedom rather than achieved functionings assuages the latter worry.
There are two distinctive features of Nussbaum’s understanding of the capability approach to justice that merit discussion. On her view, we identify the list of important functionings that are essential to a life that would qualify as flourishing . The capability to achieve each and every one of these functionings at an adequate level must then be secured for every individual. The second noteworthy feature of the account is that the functionings on the list are specified at a high level of generality. She describes the Aristotelian account as the "thick vague theory of the good."
The list of functioning capabilities includes items such as these: living out a normal lifespan, having good health and its requisites, freedom of movement, freedom from assault, freedom of choice in sexual matters, the capacity and opportunity to engage in imagining, thinking, and reasoning.
I have no quarrel with any of the items on Nussbaum’s list. Each seems to be an important human functioning, a plausible component of flourishing. But a structural worry obtrudes. I doubt that a list as expansive as hers is really a list of functionings all of which any person must achieve at some threshold level if her life is to count as attaining a decent or adequate level of well-being. This issue seems undecidable until one has a grip on how to determine where the level of a "good enough" life is to be set, but even if one sets that issue aside, a problem looms, because one can imagine lives that are high in well-being despite failing to attain any positive amount of some items on Nussbaum’s list. Subpar scores on some dimensions can be offset by very high scores on other dimensions of functioning. Someone might lead a monkish life with no sexual expression, but the life might still rank high in other values, and be an unqualifiedly good life overall. This example generalizes.
One might respond by paring down the list, but the pared down list might be inadequate to guarantee that gaining just barely adequate amounts of the goods on this list qualifies one’s life as adequate. Since the goods of human life conflict, and more of this usually means less of that, we are used to the idea of sacrificing some elements of the good to gain others so as to fashion as good a life for oneself overall as one can. This suggests that the idea of a life that is adequate or good enough in well being is the idea of a life whose combined scores on many dimensions of human flourishing is at an adequate level on the whole. The idea of a precise numerical aggregate well-being score may presuppose more commensurability than the subject allows, but even if commensurability is only partial, our vague assessments of a person’s lifetime well-being are all things considered aggregate judgments.
This is not to deny that there might be some crucial functionings all of which a satisfactory life must include at some satisfactory level. A life of unrelieved severe chronic pain, I am inclined to think, is a bad life for the person living it no matter what else is true of her doings and beings. The same might be claimed of a life utterly devoid of the exercise of significant cognitive functioning. It might also be that some minimal level of desire satisfaction is a requisite for an overall satisfactory life. Other goods in high degree can partially offset even deficiencies in these core functionings, but there may be deficiencies too great to be offset no matter what. Or perhaps for some core functionings, as functioning drops toward zero, further decreases in that functioning have increasing disvalue for the person’s overall well-being.
The second noteworthy feature of Nussbaum’s account is that the items on the list of objective goods that she deems essential to flourishing are specified at a high level of abstraction. Agreement that the items on the list are objectively good leaves it entirely open how good they are or how weights should be set fixing the value of the different items or various specifications of any given item. There is probably general agreement among people that various abstractly described items are good but immense disagreement as to how particular kinds and amounts of these goods should be valued. But a perfectionist doctrine cannot go too far in the direction of tolerating conflicting valuations of goods and lives and denying that there are correct and incorrect valuations.
Nussbaum suggests that the pluralism about value and limited commensurability in her account of the good support liberal tolerance. This would be nice, but is not so. If there are many goods in life whose relative value is impossible to rate, and many types of life that are incommensurable, then many different life choices by an individual will not be subject to the criticism that a better choice is identifiable, but by the same token, different types of regime, illiberal and liberal, providing different combinations of kinds of lives distributed differently across persons will be invulnerable to criticism to the same extent. From lack of commensurability what follows is not liberal tolerance but lack of a basis for critique of regime choices as well as individual choices.
THE SHADOW OF ELITISM
The thick vague theory of the good that Nussbaum asserts is thick in propounding the ends of a genuinely good life and vague in its specification of those ends. The difficulty with this conception as so far elaborated is that the list is slanted entirely toward commonplace and ordinary good that just about anyone could achieve in feasible circumstances. This leaves entirely undiscussed and unanswered the worry that perfectionism leads unavoidably to elitism. Some warrant is needed for the assurance that if the list were extended by addition of higher achievements, the theory would not alter its character radically. The threatening possibility is that it might turn out that the higher perfectionist goods are only available to an elite few persons under the best of circumstances but that securing that these goods are attained by somebody should be the top-ranked priority.
This essay has argued that this problem cannot be addressed just within the theory of right, here by asserting a satisficing conception of right. Espouse satisficing if you like. But if what qualifies as a morally valuable perfection is set at a sufficiently high level, satisficing with respect to capability for the good will imply Nietzschean politics—in practice, only the attainments of an elite few will matter at all.
A moderate version of Nietzscheanism is taken seriously and ultimately embraced by Hastings Rashdall, a perfectionist theorist who wrote in the early twentieth century. Rashdall suggests that there may well turn out to be "a final irreconcilability between the higher Well-being of the few and the lower Well-being of the many." The problem Rashdall sees is straightforward. If most people under the best feasible circumstances are only capable of lower goods such as playing checkers and watching simple comedies on the television and working at socially productive but rote and unskilled jobs, and a very few people are capable of exquisite goods such as proving interesting mathematical theorems and understanding and appreciating the world’s best literature and composing beautiful art and contributing significantly to basic science, then it may turn out that the resources that an Aristotelian Social Democracy could devote to providing fair opportunities for the many to achieve lower goods would be better spent, according to a perfectionist accounting, by providing expensive opportunities for the few to generate more of the very highest accomplishments. Having in mind the experience of late nineteenth century England, Rashdall supposes the distributional problem to be compounded by the fact that to sustain a class of people capable of producing and consuming higher goods one must sustain them as a leisure class at great social expense. Rashdall notes that progressive-minded people who value high culture often reassure themselves that this difficulty is not severe in practice because over time the cultural goods produced by the few are diffused among the general populace. But he suspects this reassurance might be glib and that social progress measured by egalitarian standards comes at a cost in perfectionist values that may sometimes be excessive. The "price of a real advance towards social equality" might be "vulgarizing rust," and though one may hope this rust will only eat away "at the mere polished surface of our intellectual life," one should acknowledge the real possibility that social progress would ultimately destroy its inner core. Rashdall ultimately resolves the conflict in his own mind by affirming perfectionism over egalitarianism, so that Bentham’s "Every body to count for one, nobody for more than one" must be amended to "every one’s good to count as much as the like good of any one else."
For a literary depiction of what is essentially the same conflict between human perfection and egalitarian versions of social justice, one might consult the novel The Princess Casamassima by Henry James. (Nussbaum has written an extended interpretation of this novel, but she oddly recruits James as her comrade in the project of Aristotelian Social Democracy, whereas my sense is that James tends to favor the viewpoint of the hero of the novel, who espouses something very close to Rashdall’s creed.)
A perfectionist theory that disavows the elitist and illiberal views that are associated with this tradition of thought needs to explain the disavowals. What is wrong with the position that on a scale of human perfection, the achievements of a Michaelangelo or an Einstein or a Michael Jordan so dwarf the possible attainments of almost everyone else that the potential moral value of the lives of the latter should count almost as nothing by comparison? Commenting on the idea that any political morality worth taking seriously must be egalitarian at a fundamental level, Sen writes, "it is difficult to see how an ethical theory can have general social plausibility without extending equal consideration to all at some level."
However, part of the challenge of the perfectionist tradition is that it suggests a break with this consensus. Rashdall suggests a puzzle by contrasting our reaction to conflicts of interest between animals and humans with our quite different reaction to conflicts of interest among human persons. The non-human animals known to us, even the higher mammals that are our close relatives, count for something morally, but their interests (in gaining pleasure and avoiding pain, say) do not count the same in moral calculation as the comparable interests of humans, because we have intellectual capacities they lack. But the intellectual capacities of humans differ as well. If my interests should count for more from the moral standpoint and deserve greater consideration than the comparable interests of a chimpanzee or bonobo, because I have greater intellectual capacities, why do not the interests of Michaelangelo count for more than my comparable interests, and deserve greater consideration than mine should get, because he has greater intellectual capacities than I have? (We would then revise Rashdall’s doctrine to "everyone’s good to count vas much as the like good of anyone else who possesses the same level of rational capacity.") I do not raise this question as though it is unanswerable, but some answer is needed. The perfectionist propounds her doctrine in the shadow cast by Plato and Nietzsche.
So far as I can see, Nussbaum and Sen do not take up this challenge which is internal to the perfectionist tradition of thought but just brush it aside. But their view contains resources for a response.
Nussbaum makes a suggestion as why one should not think a perfectionist politics would in practice recommend redistribution toward the better off that is interesting but unsuccessful. She observes that resources beyond a certain level can make one’s life go worse rather than better. This is surely true. With too much wealth I might be tempted to pursuits that would be my ruin. But that is so because I am not an especially astute and careful chooser. If those most capable of achieving the greatest goods tend to be among the best choosers, they would not be tempted into bad choices by largesse of possessions, but would use them wisely. Moreover, there are many worthwhile pursuits that are expensive. Cutting-edge physics requires particle accelerators and other expensive gadgets. The fact that resources can be distracting for the foolish does not show they could not be put to excellent perfectionist use in the hands of the wise (though perhaps paternalistic monitoring of some errant geniuses would be necessary).
A PARTIAL RESPONSE
A better though partial response to the elitism worry is available to the egalitarian. It begins by distinguishing a life that is good in the sense of fine or admirable or noble and a life that is good for the person who is living it. A life can be filled with admirable deeds that are noble but self-sacrificing, and involve the knowing renunciation of the agent’s self-interest for the good of others or for some good cause. Such a life might be good in the sense that it is filled with good deeds but not good for the agent. A life might be good abstractly considered in virtue of its excellent accomplishments (whether or not they did anybody any good) without thereby being good for the person whose life it is. The theory of objective human good should deliver an account of what constitutes a life that is good for the person who lives it.
The theory of human good that I have called "narrow perfectionism" identifies the human good with the perfection of capacities in human nature that are deemed estimable. But this narrow perfectionism from the start is an implausible account of what makes a life go intrinsically better for the one living it. One could develop and exercise the significant capacities to a high degree but also suffer from chronic severe pain, the frustration of all of one’s strongest aspirations and aims, loneliness and utter lack of significant love or friendship or other human ties, the instability of distressing mental illness, no enjoyments or pleasures of even the most simple kind, and so on. This would not be a life one would see as high in well-being or welfare, these being just other names for what one attains when one is prudent and lucky or just very lucky and gets a life that is good for oneself.
Once one distinguishes a good life and a life that is good for the one living it, it becomes immediately plausible to maintain that even if great accomplishment is one dimension of human good or well-being, there are other important dimensions that are objectively important constituents of it. These other components include having relations of love and friendship, having experiences that are interesting and pleasant, fulfilling one’s important reasonable life aims or at least a subset of them, having a rudimentary understanding of the world one inhabits including its people, having ordinary bodily vigor and good health, and the sustaining of all of the above through a life whose span contains more rather than fewer years. The items just mentioned are important components of a good life and they are all quotidian, even humdrum, well within the range of most humans given reasonably favorable life circumstances. To these items one should add significant accomplishment relative to one’s native capacities. This is a separate and distinct good from great accomplishment. I might be a bad accountant, football player, dancer, scientist, poet, linguist, creative scientist, and so on, yet still take good measure of what I am relatively best at and where my comparative advantage lies and manage over the course of my life to achieve at a commendable level given my capacities. This can be an important component of well-being for me even though genuine significant accomplishment is beyond my reach.
One point at which narrowly perfectionist theories clearly go awry if intended as theories of what is prudentially good or good for a person is their treatment of pain, pleasure, and desire satisfaction. If one identifies the human good narrowly as excellence of attainment, then clearly whether an individual suffers pain or gets pleasure in her life is of no account except insofar as the experience of pleasure or pain might impede or facilitate the stretching of one’s talents to their greatest significant attainments. In his discussion of this issue the narrow perfectionist Thomas Hurka takes this line. The relevance of desire satisfaction is dismissed on the same basis. But this seems perverse, and lacking plausibility once one considers what sort of theory of human good one is seeking. A rational person with full information and after careful deliberation might well prefer a life with less accomplishment and less pain to a life of more accomplishment and more pain. This choice would not be a misguided choice to toss away one’s good.
Arguing along the same line as Hurka and Sher take, Richard Kraut asserts that only if what is desired has genuinely desirable features is the satisfaction of desire a contribution to one’s good, and on this basis he holds that pain is not intrinsically bad, because it is just disliked sensation, something we in fact want to avoid, and it lacks features that make it worthy of avoidance. But contrary to Kraut I submit that there is a feature of pain that is bad: the way it feels. If an individual has malfunctioning internal wiring so that he desires pain, but what he is experiencing is the same as what I feel when my hand touches a hot stove, then his desires are misleading him, and he would be better off forming a desire against it and acting on it—not in virtue of further consequences such as tissue damage, but just considering the experience on its merits. It should be a test case of a candidate objective theory of human good that it can deliver the result that pain is intrinsically bad and a life with more pain rather than less is, other things being equal, a worse life for the one who is living it.
The split between objective and subjective theories of human good is that according to the former, what is good for me intrinsically is so independently of my desires and attitudes and opinions. But care is needed here. An objective theory can say that pain is intrinsically bad, and for that matter could also say that desire satisfaction is intrinsically good. On an objective theory, to say that pain is bad is to say it is bad independently of my attitudes, desires, and opinions concerning the badness of pain. These could all be misguided or incorrect. To say that desire satisfaction is intrinsically good is to say that the satisfaction of my desires is good independently of my desires, attitudes, or opinions concerning desire satisfaction.
The "wider" a perfectionist doctrine, the more it values goods that virtually all humans can reach, and the smaller is the gap between the value assigned to these lower goods compared to the value assigned to the higher goods that few can attain. Wide perfectionism need not be elitist as narrow perfection is.
EGALITARIAN SATISFICING VERSUS PRIORITY
Sen proposes equality of capability as a social justice goal but allows that equality must be balanced and traded off against other worthy social aims. Nussbaum more firmly champions an egalitarian ethic as a first principle of justice that must be fulfilled. The principle of equality she affirms is that everyone must be brought to a threshold level of capability to achieve each and every one of the important human functionings that she supposes to be needed for human flourishing. The distinctive character of her doctrine is that the principle of equality is affirmed in this satisficing version. Each and every person must be provided basic capabilities, the ones needed for a satisfactory level of functioning for all. The satisficing view if acceptable would obviously help answer the elitism worry, and would address the worry about illiberal restriction of freedom by leaving a wide space of above-threshold matters wherein individual freedom is not constrained by social justice.
In Nussbaum’s view, then, the egalitarian demands of justice are strict but also limited. Once everyone is brought to the threshold of basic capability, inequalities among persons above this level are a "don’t care" from the standpoint of justice.
Taken dead literally, Nussbaum’s view would seem vulnerable to a familiar objection. Some individuals are extremely unable, but can benefit even if only slightly from great infusions of assistance. Some of these severely disabled but assistable individuals cannot be brought to the threshold of basic capability (unless the threshold is set at an absurdly low level, which would create obvious problems of another sort) no matter what amount of assistance at whatever cost is applied to them. Some of the severely disabled can be brought to the threshold of basic capability but only if given enormous levels of assistance that if supplied would press the society that supplies the aid to a level of wealth such that no above threshold levels of capability could be sustained for anybody. The severely disabled then become basins of attraction taking in enormous levels of social resources even though their capability levels improve only slowly as a result.
This objection would probably be an unfair one to lodge against Nussbaum. At least in her most recent writings she seems to acknowledge the problem and suggests that her insistence on basic capability for all is not unqualified. But this acknowledgement leaves a gap to be filled: What are the specific principles of justice we should affirm and what exactly is the role of satisficing egalitarianism within them? I shall suppose that Nussbaum continues to affirm satisficing egalitarianism hedged by the condition that its requirements are to be relaxed if aggregate human flourishing would otherwise be greatly diminished. The discussion below considers objections that would apply to this qualified view.
Satisficing egalitarianism of basic capability of the sort Nussbaum embraces (with some qualification) seems to me to be open to objections that sharply limit its appeal.
One difficulty is how one nonarbitrarily sets the threshold level. Why here and not higher or lower? What we have is a smooth continuum of possible levels of overall capability for flourishing. Higher capability is always better than lower capability. But I do not see how any unique level (not even a broad thick line) can be picked out such that if a person has that level, she has "enough."
Further objections follow from this one. Let us consider the version of the satisficing view that says that the first priority of justice is to move as many individuals as possible just past the threshold that marks sufficient basic capability. Now consider situations in which we must choose between helping individuals who are far below the threshold level and individuals who are very close to it. Call the former group the "unavoidably below threshold group," because with available resources we can bring none of them to the threshold, and the latter the "avoidably below threshold group," because all of them can be boosted past the threshold. We have sufficient resources to give either the members of the "unavoidably below" group a significant gain in capability for flourishing—a one unit gain for each--or the members of the avoidably below group an identical one unit gain for each, which would push all of them past the threshold. The Nussbaum satisficing ethic unequivocally must hold that we ought in this case to help the better off not the worse off. But why? In the example, the badly off can be given just as much of a capability boost as the better off group. Why not help the worse off in these circumstances?
In the example, one can either provide a certain gain in capability for individuals unavoidably below the threshold or an identical gain for the same number of individuals who are already better off and can be pushed past the sufficiency level by the provision. I have criticized Nussbaum’s view for failing to favor the worse off on these assumptions, but a stronger claim can be made. Suppose one must choose between supplying a one unit gain in capability for an individual whose prior capability level is very low or instead a greater gain for an individual whose prior capability level is already much higher (even though below the sufficiency level). Although the claim is controversial, I would hold that justice demands giving priority to achieving gains for those who are badly off. In the context of the capability approach, the prioritarian would hold that the moral value of securing a one unit gain in capability for a person is greater, the lower the person’s capability level prior to receipt of this benefit (and similarly for the moral value of avoiding losses). The priority view of justice asserts that the calculation of what should be done should tilt in favor of those who are worse off, so that in choosing policies and actions we should select the one that maximizes morally weighted value not aggregate value. The priority approach and the satisficing approach squarely disagree.
One could avoid the result criticized above by a reformulation of satisficing egalitarianism. Say that social justice demands as a first priority (unless the cost in terms of the aggregate capability enjoyed by all is "too great") that all people be provided a sufficient level of capability, the level of basic capability. This is to be interpreted as follows: Give priority to gaining greater capability for those below the threshold, giving greater priority to helping an individual, the further she now is from the threshold level. This interpretation of the satisficing view avoids the oddity of giving priority to helping the better off just in order to move more individuals past a specified threshold. (Priority here does not mean strict lexical priority but some weaker relation that gives extra weight to gains depending on how badly off the proposed beneficiary of the gain would be prior to its receipt.)
So construed, satisficing egalitarianism differs from prioritarianism in two cases. In both cases, I would contend that the priority view yields more plausible judgments. But here I am content to point to the different implications the two conceptions would reach. Consider cases in which we have to set policy or choose action when the interests only of those people already past the threshold of sufficiency can be affected. For the sufficiency doctrine, all such decisions are a "don’t care" from the standpoint of justice. At least, the egalitarian component of justice is fully exhausted by the sufficiency idea, which holds that inequalities above the threshold are not morally undesirable provided the resources that sustain them could not be redeployed to boost the long-run numbers of people who enjoy capability past the threshold level. But the priority view pays no attention to the threshold. If we have a choice between helping those just past the threshold (but guaranteed to stay above it, whatever we do) or helping the same number of people to the same extent who are far above the threshold, priority says unequivocally that we ought to favor the worse off.
The degree of tilting towards the worse off that priority imposes depends on the weighting that is imposed. At one extreme, prioritarianism gives absolute priority to the worse off, so that securing any gain of any size however tiny for a worse off person is morally to be preferred to securing a gain however huge for any number of persons who are better off. Prioritarianism here becomes leximin. At the other extreme, prioritarianism gives no extra weight at all to securing benefits for a worse off person rather than to securing benefits for a better off person. Here prioritarianism becomes utilitarianism, if utilitarianism is interpreted as holding that we should maximize the aggregate sum of benefits (different versions giving different construals of what qualifies as a benefit or loss for a person). So prioritarianism is the name of a family of positions, not a single view. By prioritarianism I intend to refer to a position roughly midway between leximin and straight aggregation, but I have no precise view about how the weighting should be set.
Given that the weighting does not render prioritarianism equivalent to leximin, then there will be cases in which prioritarianism favors helping the better off at the expense of the worse off, because the partially discounted sum of benefits that can be achieved for the better off exceeds the weighted sum of benefits that can be achieved for the worse off. Hence there is another generic case in which satisficing and prioritarianism will disagree. Suppose we can do something that would achieve a benefit or avoid a loss for people below the threshold of sufficiency, and we have a single alternative option, to use these same resources to achieve a benefit or avoid a loss for some people who are well above the threshold. If the benefit we can bring about for the better off is great enough, and the gain we can bring to the worse off is small enough, then prioritarianism will favor helping the better off, even if they are past the level of sufficiency. This result will of course continue to be possible no mater how badly off the worse off are and no matter how well off the better off are, so long as the difference in the gain we could create is sufficiently large. I have to say this result seems to me to strike another blow in favor of prioritarianism. If I am worse off than Smith, but the flower seeds in my possession will do me barely any good at all but would provide Smith a wonderful and profound aesthetic experience, then Smith, not I, should have these flower seeds, and if our only options are (take seeds from me, give to Smith ) and (leave seeds in my possession) then I should favor the former. The same point would hold if the calculation favors the better off because the numbers of the better off who can be aided greatly exceeds the numbers of the worse off who could benefit from a choice to assist them. It is reasonable to tilt in favor of the worse off but unreasonable to tilt in favor of the worse off to the extent of strict lexical priority.
One might reject satisficing egalitarianism on the ground that one rejects any egalitarian theory of justice, that is to say, any theory that holds that sometimes those who are fortunate through no merit of their own should sacrifice their interests to some extent to improve the life prospects of those who are objectively unfortunate through no choice or fault of their own. But within the class of broadly egalitarian theories of justice, the satisficing view embraced by Nussbaum seems to me to be inferior to its prioritarian rival.
CAPABILITY VERSUS FUNCTIONING
Should a moral principle that asks us to promote human perfection take as its task the provision of opportunities for perfection or the achievement of the thing itself? Nussbaum consistently advocates the former. Sen also leans toward the capability-oriented view, but less definitely than Nussbaum. According to the capability approach, social justice requires that priority be given to bringing about a condition in which all persons have the capability or real freedom to achieve all of the valued functionings deemed central to human life.
Nussbaum provides examples that are intended to suggest the appropriateness of the focus on capabilities. Having adequate opportunity for sexual expression, a person may for various reasons choose a celibate life. Having adequate opportunity for play and recreation, a person may singlemindedly devote herself to work and take no advantage of her play and recreation opportunities. Having food available, a person may prefer to fast, whether for religious reasons or as part of a political protest or for some other reason. In all of these cases, the idea is, social justice would be violated if people lacked the capability to be adequately nourished, enjoy normal amounts of play and recreation, and have a satisfying sex life. But it is no social injustice in these cases that the functionings of nourishment, play, and sexual expression are not achieved, and indeed it would be unjust to force or pressure individuals to turn their capabilities into functionings.
The insistence that justice is done when individuals have capabilities whether or not they are exercised might be thought to find support in a norm of responsibility and further support in an ideal of freedom. The latter ideal could invoke the value of freedom or a right of personal sovereignty.
Consider responsibility first. The idea evidently is that the obligations of society to help an individual lead a good life are limited. Provided that an individual has adequate opportunity to lead a good life, whether he actually leads a good life or not is up to him—his responsibility and not that of society.
However, the thought that the obligations of society to assist individuals to lead a good life are not strict obligations to ensure some assured level of functioning come what may does not require opting for the capability approach. One might rather hold that the obligations of society toward an individual are limited in extent but are obligations to help the individual achieve functioning not just capability.
To see that the issue of capability versus functioning involves more than metaphysical hairsplitting, consider the case where one can at some cost enable an individual to achieve a capability but one knows for certain that the capability will not be exercised. In this circumstance the individual might value the unexercised capability, but this is not the case we are considering. The individual has no use for the capability and regards its provision as a matter of indifference. If one’s ultimate ethical concern is the quality of the lives that people lead, then capability provision that in no way enhances anyone’s life is pointless. But if the ultimate justice value is provision of a fair level of capability to each individual, then it should not matter that one knows the capabilities will not enhance anyone’s functionings. If on the other hand we are looking beyond the capabilities that people have to assess the functionings they get, then capabilities are being regarded either as means to some further goal or at least as valuable only on the condition that they enhance some functioning.
Another way to see that the idea that society has limited responsibility with respect to assuring individual well-being does not support the capability approach is to notice that given limited responsibility, a society may not be required to ensure capability at any given threshold level come what may. Consider my capability to move my body from place to place in roughly the ordinary way. I might have this capability, then lose it in an accident sustained in the course of some reckless activity, then regain it by social intervention, lose it again in some equally reckless fashion, then receive it again, and so on. Surely at some point in this cycle society’s obligation to me is fully discharged even though I now lack the capability to move about normally.
Responsibility aside, the issue whether justice requires society to secure individual functionings or individual capabilities turns in part on considerations concerning freedom. No doubt possession of freedom is an aspect of the good life. If my choice set contains just one option, which I must take, I am worse off, other things being equal, than I would be if I had additional options, even if the singleton option is still the best option in the expanded set. In this way freedom can be seen to be intrinsically, not just instrumentally valuable. Choosing freely can also increase the value of whatever of value is chosen. But none of these perfectly reasonable claims is of the right type to justify the position that the fundamental concern of justice is to provide freedom not achieved good. The claims insist that freedom is one important good among other goods. So to do the best we can to facilitate people’s achieving good lives we will inter alia have to provide them wide freedom. There is no consideration here that even suggests an exclusive focus on freedom as in the view that social justice entirely consists in fair provision of capabilities or real freedoms.
Perhaps the best interpretation of the norm that society owes its members capability for flourishing rather than flourishing itself is to hold that each person has a right of autonomy or personal sovereignty in self-regarding matters. One has the right to lead one’s life as one chooses within moral constraints even if it can be ascertained with certainty that one will lead one’s life in an inferior way and that the loss in well-being that one will suffer is not compensated by gains in well-being that will accrue to other persons (or animals).
There is no inconsistency in combining a perfectionist standard of human freedom (the extent of a person’s freedom is the degree to which she can attain the perfectionist goods) with a right of personal sovereignty that forbids society from seeking to advance an individual’s achieved perfection against her voluntary choice. But one may wonder whether the compromise is stable. Notice that an individual under a capability regime is entitled to pursue, say, a life of trivial pleasures rather than one of perfectionist achievement. But the individual, rather than be allotted circumstances that give her a specified level of capability, might well prefer that she be allotted circumstances that would enable her to gain a higher level of trivial pleasure over the course of her life. Justice according to the capability approach obligates society to "help" her in ways that do her no good by her own lights and do not advance her achieved perfectionist good or anyone else’s. This result looks to be socially wasteful by any reasonable standard.
Return to Nussaum’s examples where fixation on functionings rather than capabilities seems dubious. These examples are one and all cases that share the feature that it is very hard to see how a society, particularly by coarse-grained measures such as law and social policy, could do anything to promote functioning beyond providing capability. Coercing a person who wishes to remain celibate to have sex is very unlikely to enhance the coerced person’s well-being. Forcing a person singlemindedly dedicated to work to spend a certain proportion of her day in specified leisure and play activities is not a promising strategy for enhancing her achievement of valuable play but is very likely to interfere significantly with the fulfillment of the work-related goals that she is wholeheartedly committed to achieving. In order to clarify the choice between functioning and capability we should consider examples in which it is very clear that the two approaches do yield conflicting judgments about what should be done.
We should consider, then, cases in which it is clear that providing an individual capability for perfection (at some level our theory of justice deems fair) would result in her attaining overall a lower quality of life, a lower level of human flourishing, than would be attained if society were committed to seeking to promote people’s achieved perfection, not merely their opportunity for perfection.
Suppose that we could use funds from general taxation to subsidize opera (substitute some other good you deem genuinely valuable if you like) for adult citizens. We could do this in either of two ways. The first option devotes all available resources to enhancing people’s capabilities (e.g. providing opportunities to see performances) and achieves a higher level of capability for all. The second option devotes some resources to propaganda aimed to persuade individuals to avail themselves of the provided opera-going opportunities. The first option provides more capability; the second secures more achieved functioning.
Nussbaum explicitly states that the aim of Aristotelian Social Democracy is "enabling people to live well and do well." The point if to gain for everybody "a certain sort of capability—the capability to function well if one so chooses." She is then committed to favoring the first option in the example, which strikes me as an implausible commitment. I submit that when we imagine a clear case in which it is clear that by providing less capability for flourishing we could get more flourishing fairly distributed, one ought to opt for more flourishing.
So far as the arguments of this essay go, egalitarian liberal perfectionism remains a viable option within the theory of justice. The elitism charge can be met, at least in part, by defending an account of human good that accords significant value to ordinary achievements reachable by almost all persons. The objection that perfectionism must countenance illiberal restriction of individual freedom in the self-regarding sphere can be met by a two-part strategy: on the one hand, by a Millian perfectionist critique of paternalism, and on the other hand, by acknowledgement that when (if) paternalism succeeds in bringing about a greater human good that is fairly distributed, restriction of liberty is unobjectionable, call it illiberal if you like. The Nussbaum-Sen affirmations of egalitarian satisficing and of the limitation of the purview of social justice to concern for provision of capability not functioning cannot be sustained, so they are not available to be deployed against the objections that perfectionism must tend to be elitist and illiberal. But these Nussbaum-Sen claims criticized here are also unnecessary to the defense of liberal egalitarian perfectionism, so their inadequacies are not politically fateful.