II. Two Kinds of Aggregative Views of the Social Good

There are at least two fundamentally different kinds of views that might be called “aggregative conceptions of the social good.” Good-based views purport to be grounded upon a fundamental insight into the human good. Right-based views, by contrast, begin with intuitions about justice and work up a conception of individual advantage that is suited to shed light on questions concerning the just distribution of the benefits and burdens of social cooperation. The main criticisms of this essay are directed only at good-based views.

To approach the distinction between good-based and right-based views, it will help to ask why aggregative views of the social good are attractive to begin with. I believe that their attraction lies partly in their ability to accommodate two deep-seated intuitions. The first intuition is that one can often speak sensibly of the good or the interests of a society, and even of the inclusive “society” of humankind. The second intuition is that, strictly speaking, a society is not the sort of thing that has a good of its own or interests of its own, distinct from the good or the interests of the individual human beings who are its past, present, and future members or citizens. When the good of the society or the nation is invoked in support of political agendas that seem not to serve or further the interests of individual citizens, many of us are inclined to regard this as a prelude to political oppression.

These intuitions, however, are not sufficiently precise to support a plausible aggregative conception of the social good. In the first instance, there are many nonaggregative ways of reducing the good of a society to the good of its individual members. The good of a society might, for instance, be equated with the good of the individual with the largest shoe size. Second, there are many implausible formulas for tying the social good to an aggregate of the goods of the individuals composing the society. It will not do, for instance, to equate the social good with the aggregate of the good of each individual weighted in order to reflect variations in shoe size. We can reveal something important about the conceptual foundations of aggregative views of the social good if we ask what exactly is wrong with these two patently inegalitarian conceptions of the social good.

Here we can distinguish two importantly different kinds of answers. The first and perhaps most natural kind of answer is that such views are wrong in the moral sense. If we use an inegalitarian account of the social good to assess and choose among proposed laws, policies, and institutions, this would be unjust to those whose good is given less than equal weight. This sort of answer, however, is open only to those who think that we can give content to the idea of justice without relying upon a determinate idea of the social good. If we accept the familiar consequentialist view that what makes an action or institution just is its contribution to the aggregate good, then clearly we cannot choose among competing accounts of the relevant aggregative procedure by reference to their justice or injustice. That would be viciously circular.

Orthodox consequentialists have a second way of explaining the problem with inegalitarian aggregative views. They hold that such accounts are wrong in the sense of being false to an independent order of facts about the human good. The truth is that whatever sorts of things are good in one human life are equally good when realized in the life of another person, no matter his shoe size. If we make use of an inegalitarian aggregative function, our decisions will be informed by false beliefs about which laws and institutions will bring about the most good. This might make these decisions unjust, but the injustice is explained by the mistaken view of the good on which they turn, not the other way around.

In attempting to make sense of the egalitarianism found in all plausible aggregative theories of the social good, we have tripped across a familiar distinction between two kinds of conceptual foundations that such theories can have. Some aggregative theories are “right-based” in the sense that they are shaped by prior convictions about the right, and are sculpted in order to play a circumscribed role in a broader theory of distributive justice. Other aggregative theories are “good-based” in the sense that they are sculpted so as to track the truth about what really is good. Such theories do not rest on prior convictions about the right; most such theories purport instead to provide us with independent insight into the right.

We catch sight of this same distinction if we ask why the best-known and most influential aggregative views are grounded in subjectivist conceptions of individual well-being, such as conceptions based on pleasure or desire-satisfaction or preference-satisfaction. A “right-based” theory might hold that justice demands that we take equal account of the preferences (or, less plausibly, the desires) of each individual when formulating a conception of the social good that is suited for the public assessment of alternative laws and institutions. Our reason for doing this is not that each individual is sure to be right about what really does conduce to his or her own good. Human beings are fully capable of making a mess of their own lives, even when left free to do what they prefer under conditions of full information. Instead, we take this approach because it would be unjustly paternalistic to overrule an individual's preferences (or, some would say, informed preferences) when assessing the impact of proposed laws and policies on the quality of that very individual's life. To substitute an expert opinion for the individual's own preferences would be tantamount to refusing to recognize that individual's status as an autonomous agent with equal standing in the social order.

There is another, somewhat more complicated justice-based reason for using a subjective conception of individual advantage in working up an aggregative measure of the social good. The subjectivism of the metric can be seen as deepening the egalitarianism of the resulting conception of the social good, and, by extension, of the social and political institutions that are sculpted so as to maximize the social good so conceived. To see the point, consider a conception of the social good that aggregates over the creation of great works of art. Such a metric might be deemed egalitarian in the limited sense that it makes no difference, from the point of view of this metric of social advantage, which individual creates the work of art. Resources go to the people most likely to create great art, whoever they might be. This metric of social advantage is antiegalitarian, however, because what it counts as good falls outside the interests and the capacities of many individuals. On this conception, the social good might well be undiminished if such people are unfed or unhealthy or unsheltered. By contrast, there are few human beings who will not be displeased and dissatisfied if they are ill, hungry, or homeless. By keying the social good to these simple and ubiquitous experiences or attitudes, subjectivist accounts yield antielitist assessments of alternative policies, laws, and institutions. Nor is it necessary to recognize the truth or superiority of a particular philosophical or religious conception of the good life in order to embrace such a metric of social decision or to accept its dictates as legitimate.

Once again, these right-based arguments are not available to those classical consequentialists who propose to ground claims about justice in conceptually prior claims about the good. Such theorists almost invariably favor one or another subjectivist conception of the individual good, but their defense of this stance must be grounded directly in claims about the nature of the good rather than in claims about justice. The general strategy is to argue that no achievement, however objectively valuable it might seem, can really add to a person's well-being unless that individual has some sort of favorable subjective attitude toward it (e.g., she prefers or desires it, or she likes or is pleased by it). An admirably clear version of this view is found in the following passage from Wayne Sumner's influential book on welfare:

There is an uncontroversial truth in the vicinity of Sumner's claim here. Even the best human activities are less good for us if we regard them as worthless than if we consummate them with a running appreciation of their intrinsic value. However, as I will argue below, this point can convincingly be accommodated by a thoroughgoing objectivism about value, and it does not distinguish the prudential value of activities from other, more impersonal sorts of value that they can have. At any rate, Sumner is not making this uncontroversial claim. He is making the stronger claim that we must have some positive attitude toward our activities and achievements if they are to have the least bit of intrinsic goodness for us. Indeed, by the end of his book, he stakes himself to the even stronger claim that a certain kind of subjective pro-attitude—a kind he calls authentic endorsement or affirmation—is not only necessary but also sufficient for conferring prudential value upon the conditions of one's life.

There are serious objections to this stance. We can begin to see them by noting that we choose certain activities, and indeed sometimes make them our life projects, because we think that there is something particularly valuable about them that merits this commitment. It would be hard to sustain such projects and commitments if we thought that they differed from such worthless activities as counting blades of grass and collecting saucers of mud only because we happen to affirm or endorse our engagement in them.Footnote ⁴ Indeed, it is hard to see what sort of affirmation or endorsement we could continue to extend to them in the face of this acknowledgment. The activities we choose as life projects seem from the participant perspective to meet our affirmation halfway. They seem to call for, or merit, our endorsement. To accept that this appearance is a mere illusion, and that one's life goes no better if one throws oneself into loving friendships or philosophy than if one throws oneself with equal enthusiasm into collecting saucers of mud, is tantamount to accepting that there is no particular point in doing anything. This is nihilism masquerading as a positive account of the sort of value humans might hope to find in their lives.

There is considerable pressure on the aggregativist to affirm this conclusion, despite its nihilistic undertones. As noted above, we cannot sum up the value of the motley array of beings and doings that enhance the lives of different human beings unless we locate this goodness in some common coinage. The attempt to make use of a literal coinage (i.e., willingness to pay) leads to an aggregation that gives far too great a weight to the interests of those who can afford to pay the most. This makes this approach implausible as an element of a broader theory of justice. Perhaps it goes without saying that it is also implausible as a grounding insight into the good life for human beings, since we apportion our willingness to pay to the value of that for which we are paying, not the other way around. A more plausible aggregative approach is to follow the classical utilitarians in identifying the goodness of various beings and doings in the fact that they correspond to subjective pro-attitudes or produce subjective pleasures. The latter, hedonistic view is counterintuitive enough that most contemporary aggregativists favor one version or another of the desire-satisfaction or preference-satisfaction account of welfare. This deprives them of the resources for distinguishing the intrinsic prudential value of poetry and pushpin, or friendship and collecting saucers of mud. As noted above, these views seem from the participant perspective to be close cousins of nihilism rather than fit guides to deliberation about what will conduce to one's own flourishing.

Aggregative views might perhaps be plausible if put forward as accounts of the end that society must promote if it is to treat each of its members as justice demands. Such right-based theories do not imply that when we deliberate about what sort of life would be best for us, our sole concern ought to be to get more rather than less of whatever it is that the aggregative theory aggregates. If an aggregative theory purports to be grounded in faithful insight into what makes lives better for those who lead them—that is, if it is good-based—then it does carry this implication, and it ought to be useful both for political deliberation and for individual prudential reasoning. I do not think that any aggregative view can meet this dual demand. If we find something that is a plausible candidate for aggregating, we will have found a concomitant of the many things that people rightly find valuable and not the thing that gives value to their lives.

III. The Person-Relativity of “Goodness For”

Those who regard welfare (or well-being or prudential value) as an independent subject of philosophical inquiry tend to think that the distinguishing mark of this sort of value is its irreducible person-relativity. What exactly is it supposed to mean to call this sort of value irreducibly person-relative? This is unclear, and I believe that certain theorists of welfare make use of this notion in a way that is essentially unclarifiable because it is incoherent.

The main contours of the obstacle to clarity can already be seen when we note that such theorists both do and do not seem to think of prudential goodness as a species of a larger genus called “goodness.” On the one hand, they do not regard it as a mere accident that the word ‘goodness’ appears in the phrase ‘prudential goodness’. Quite the opposite: they rely upon the existence of some sort of association between the good simpliciter and the prudentially good to explain how the prudentially good ought to figure in deliberation. They think that what is prudentially good for me ought to figure in my deliberation in something like the way that goodness simpliciter ought to figure in everyone's deliberation: it marks off something that is to be promoted. On the other hand, they do not think of prudential goodness as a species of a more inclusive genus called “goodness.” That is, they regard it as a controversial matter whether all things that are prudentially good, or good for a person, are good simpliciter.

Sumner gives voice to this paradoxical position when he describes welfare as a “peculiarly perspectival kind of value.”Footnote ⁵ What does it mean to say that prudential value is perspectival? In Sumner's view (which I have chosen as my stalking-horse because it gives clear expression to a thought structure that operates in the background of a great deal of contemporary philosophical work), when an occupant of a particular perspective makes claims about what is good for him, he is not making claims from a particular standpoint about what is good; his claims are irreducibly claims about what is good-for-him. The coherence of the subject matter, then, does not depend upon the innocuous observation that particular things can appear to be good from different perspectives, or even that which courses of action are correctly seen as good will depend in part upon one's particular circumstances. It depends, instead, upon the deeply paradoxical notion that things can appear to a person to be good-for-him without appearing to him to be good.

It might be thought that even the most objective and impersonal kinds of goodness exhibit a similar person-relativity. For instance, in certain circumstances it might be morally good for a parent to discipline a child but morally bad for anyone else to do so. On further reflection, however, this sort of person-relativity has only a superficial grammatical likeness to the sort that surfaces in our talk of prudential goodness. When we say that it is good (or morally good) for a parent to discipline a child but that it would not be good for others to do likewise, the words ‘good’ and ‘for’ are joined together only coincidentally. The ‘for’ is the beginning of a nominal phrase. We could rephrase our claim, without loss of meaning, as follows: For the parent to discipline the child is good (or morally good) simpliciter. By contrast, when we say that all those acts that further the interests of those who perform them are good for their agents, we cannot prise apart the ‘good’ and the ‘for’ without loss of meaning. To see this, consider that some think it possible to advance one's interests by theft or fraud, and this leads them to conclude that theft and fraud can be good for those who get away with them. Those who accept this claim are not thereby accepting the claim that for people to steal or commit fraud can be good simpliciter. When we use the phrase ‘is good for person p’ to refer to the prudentially good, the words form a two-place predicate, not a one-place predicate together with an incomplete subject.

The question at the heart of this essay might be put as follows: Does this dyadic predicate refer to an essentially person-relative kind of goodness or to the relation that people have to what is good simpliciter when it is their life that manifests or possesses this goodness? If we settle upon the latter answer, then we ought to disown the view that welfare forms a self-standing domain of normative theoretical inquiry. An account of welfare will be a one-sentence footnote to an account of the human good simpliciter. Hence, those who think that welfare is such a self-standing domain are rather deeply committed on the answer to this question: they must view welfare as an essentially person-relative kind of goodness. Sumner expresses this grounding stance as follows:

On this view, welfare is relativized to persons not because it depends upon how value (simpliciter) appears to them, nor in the sense that it always conditions the life of some particular person or another, but in some other more fundamental sense. For this reason, talk of welfare is taken to refer to a distinctive kind of goodness, one that might be theorized in isolation from a broader inquiry into the human good. I cannot say with clarity what this more fundamental sort of relativization might be, because I myself cannot manage to bring it into view. I will argue below that there is no clarifiable thought behind the welfare theorists' verbal gesture toward this fundamental sort of person-relativity. Yet if this is a confusion, it is a deeply interesting one. I think Sumner is right that one is driven to try to articulate a position of this sort if one attempts to make sense of ordinary thought about welfare. When philosophers fall into incoherence because of their faithful effort to make sense of the broader culture's prevailing views about how human beings ought to live, these confusions are pregnant starting-points for critical reflection on the broader culture.

IV. The Modern “Discovery” of Personal Goodness

The idea that we sometimes speak of goodness in a special, person-relative sense is hardly exclusive to those contemporary philosophers who offer specialized theories of welfare. Henry Sidgwick regarded this dualistic conception of the good and its relation to practical rationality as a crucial discovery of modern moral philosophy—one that freed philosophical ethics from a deep-seated error that had marked “the earlier age of ethical thought which Greek philosophy represents.”Footnote ⁷ According to Sidgwick, Plato and Aristotle had affirmed the mistaken view that ethically virtuous actions are “good for the agent” (Sidgwick's emphasis) even when their consequences “would be on the whole painful to him—as (e.g.) a heroic exchange of a life full of happiness for a painful death at the call of duty.”Footnote ⁸ They were able to make this mistaken idea look appealing by “playing on a confusion of thought latent in the common notion of good.”Footnote ⁹ In particular, their views betrayed “a confusion of thought between what it is reasonable for an individual to desire, when he considers his own existence alone, and what he must recognize as reasonably to be desired, when he takes the point of view of a larger whole.”Footnote ¹⁰

Sidgwick is quite right to suggest that the rise of the dualistic conception of the good marks one of the most crucial differences between ancient and modern ethical theory. It helps to explain why we often have the sense, when reading the ethical works of Plato and Aristotle, that their inquiries have a fundamentally different subject matter from the ethical theories of modern philosophers. Modern ethical (or moral) philosophers generally regard themselves as inquiring into a set of limits on each agent's pursuit of his or her own good. These limits are sculpted so as to leave others free to pursue their own, similarly self-interested ends, and perhaps also to help them to succeed in these self-interested pursuits. Those who conscientiously adhere to these limits are good in the moral sense, though their lives might well be less good in the prudential sense precisely because of their conscientiousness. It is only if we presume this fundamental schism between impersonal moral goodness and prudential goodness that the morally good and the prudentially good appear to protrude as distinct subjects, each of which might be theorized in isolation from the other.

It takes only a passing familiarity with the works of Plato and Aristotle to see that they have a wholly different idea of the subject matter of their ethical inquiries. In Aristotle's ethics, for instance, virtues that we would regard as straightforwardly ethical, such as justice, kindness, and truthfulness, are vindicated on the ground that they are essential to the enjoyment of a lifelong good that can be translated without obvious infelicity by the English-language term ‘happiness’ (the standard translation of ‘eudaimonia’). Aristotle does not seem to regard these virtues as different in kind from traits such as wit, mildness, and friendliness—traits that seem to modern readers to have little if any moral value, but that make an obvious contribution to the happiness of their possessors and a small circle of intimates. Aristotle regards the former, seemingly moral virtues and the latter, seemingly nonmoral virtues as harmonious elements of a single comprehensive sort of goodness to which humans ought to aspire. He does not recognize the existence of a realm of person-relative or prudential goods whose pursuit must be limited if one is to exemplify seemingly moral virtues such as justice, kindness, and truthfulness. For this reason, he seems both to be doing and not to be doing what we moderns think of as moral philosophy. As Elizabeth Anscombe once observed, “If someone professes to be expounding Aristotle and talks in a modern fashion about ‘moral’ such-and-such he must be very imperceptive if he does not constantly feel like someone whose jaws have somehow got out of alignment: the teeth don't come together in a proper bite.”Footnote ¹¹

If we turn our attention from Plato and Aristotle to the works of Immanuel Kant and the classical utilitarians, the dualistic conception of the good protrudes with special clarity as a distinctive marker of modern ethical thought. We can see it in Kant's idea of the human will as torn between two possible highest-level principles: the moral law and the principle of self-love, adherence to which makes for the efficient pursuit of one's own happiness. When the authority of the moral law is recognized, the pursuit of one's own interests is not abandoned but rather limited in such a way as to accommodate and even to enhance the similar pursuits of others, making one a fit member of a harmonious if merely notional “kingdom of ends.” We can also see the telltale marks of modern dualism in the works of J. S. Mill, Sidgwick, and other classical utilitarians. Here it surfaces in the form of a conflict between the impersonal demands of aggregate utility maximization and the personal demand to enhance one's own utility by furthering one's own projects. To reject the first sort of demand would be to reject morality entirely. Yet to insist that personal projects must be subordinated entirely to the first sort of demand would be to adopt a recipe for universal insipidity, since universal adherence to this stricture would leave nothing truly interesting or inspiring for human beings to help each other to do or to be. This is the impasse that prompts Sidgwick to conclude that practical reason is divided against itself, and can neither answer to all of the goods to which it ought to answer, nor formulate a sensible practical principle for settling these conflicts.

It seems clear, then, that Sidgwick has located a vitally important element of the sea change in Western thought that separates modern moral philosophy from the ethical inquiries of Plato and Aristotle. Yet despite the historical influence and contemporary prestige of dualist conceptions of the good, their rise seems to me to be a story of encroaching confusion and not (pace Sidgwick) of progress or discovery. Plato and Aristotle appear to be free of this confusion. Indeed, we can find in their works an insightful diagnosis of our propensity to fall into precisely this sort of confusion, but only if we cease to interpret these texts through the lens of the confusion itself.

The two great antimoralists in Plato's dialogues, Callicles and Thrasymachus, both claim that the conventional virtues serve the interests of the many at the expense of the strong or well-constituted.Footnote ¹² The Sidgwickian dualist would accept the charge that the moral virtues serve the aggregated interests of the many and sometimes undermine the interests of their possessors, yet would deny that this undermines the authority of morality, since we have reason to further the interests of the many even at considerable cost to ourselves. This clearly is not how Socrates responds to Thrasymachus or Callicles. How, then, should we understand Socrates' response? Sidgwick suggests an answer when he claims that, in Greek ethics, “it was assumed on all sides that a rational individual would make the pursuit of his own good his supreme aim.”Footnote ¹³ On the suggested understanding, Socrates and his interlocutors agree that genuine virtues must serve the interests of their possessors, and disagree only about whether the conventional virtues pass this test. Socrates argues that they do, while Callicles and Thrasymachus maintain that they do not.

Sidgwick is not alone in reading the Greeks in this way. The prominent contemporary scholar Terence Irwin regards this interpretation both as the traditional reading of Plato and Aristotle and as the correct one.Footnote ¹⁴ Yet this common interpretative approach does not fit the confrontation between Socrates and his fiercest interlocutors in the Gorgias and the Republic. Both Callicles and Thrasymachus argue as if it can be taken for granted that a self might be wholly virtuous or wholly vicious. That is, they take for granted that we can speak intelligibly of a self and of the interests of the self without knowing whether the self in question is virtuous or vicious, and hence that we can intelligibly ask whether the virtues or the vices would better serve the interests of the self. Socrates argues, by contrast, that in the utter absence of the virtues, the self dissolves into a multiplicity to which one cannot coherently assign interests. On Socrates' view, as on the view of the mature Plato and of Aristotle, the virtues are good in the sense that they actualize the proper potentiality of their possessors, pulling them together as full-fledged instances of the human kind. This sort of goodness does not correspond to either the personal or the impersonal pole of Sidgwickian dualism. It represents a monistic alternative to this sort of dualism, but not the sort of alternative at which one might arrive simply by discarding one or another of the dualistic poles.

Plato's diagnosis of vice as self-dissolution is given particularly memorable expression in Books VIII and IX of the Republic, where Socrates portrays a series of increasingly vicious kinds of psyches, each marked by increasing levels of intrapsychic conflict and disorder. The most vicious soul is pictured as a fusion of three warring parts: a multiheaded beast (representing an undisciplined array of conflicting desires), a large and well-fed lion (representing an unreasoning thumos—that is, an anger-tinged thirst for conquest and esteem), and a starved and feeble human being (representing an atrophied capacity for reasoned thought and action). The capacity for reasoned self-direction is dragged about and bitten ceaselessly by its bestial “soul-mates,” and this keeps it in a wholly passive and undeveloped form, unable even to mount active resistance to the psychological forces that master it.Footnote ¹⁵ We have the illusion that vicious people have coherent and unified selves only because the warring parts of their psyches are enclosed within the outward form of a human being. But the interests that it is most tempting to assign to them—the interests in whose name they seem to act—are, in fact, the interests of no single agent or self. And while there is a sense in which the true interest of the vicious person lies in the possession of the virtues, even this claim must be accepted only with qualifications. The problem is that in the utter absence of the virtues, we cannot speak straightforwardly of a self whose interest lies in being virtuous.

It is tempting to think of the remnant of humanity in the Platonic picture of the vicious psyche as itself a fully constituted human self to whom interests can unproblematically be assigned. This would be a mistake. Plato's point in crafting this image is to help us see the radically dissolute constitution of the tyrant, not to suggest that the tyrant's true self is an inner homunculus that might itself be perfectly well-constituted but that has the misfortune of being fused to a bevy of ravenous beasts. The tyrant's predicament is not that he has been abducted by wholly alien forces that are subjecting him to a kind of covert torture—a fate consistent with a greatly admirable character. The problem is that he has fractured into a nexus of warring drives—or, better, that where his self ought to be, there is nothing but a nexus of warring drives that keep the potentiality for selfhood unactualized.

We see now why it is both correct and also potentially misleading to assign to Plato the view that the virtues are good for their possessors. For Plato, the virtues are states of the psyche without which one cannot realize the human telos. The value of the virtues is lit up by the Form of the Good itself, in whose light we are able both to discern the form of the human and to distinguish between good and bad instances of the human kind. The value of the virtues, then, will be discernible by anyone who has a proper understanding of the human kind. To say that one's own virtues are especially good for oneself is not to say that one properly regards them as having a special, person-relative kind of goodness that the virtues of others do not have, but rather that one has an especially intimate relation to their goodness. One's virtues are essential to one's own constitution as a human being capable of self-initiated activity, and by extension capable of leading a coherent life. One lacks this relation to the virtues of others, but those virtues have the same kind of goodness: they are good because they are essential to the constitution of human beings whose activities are self-initiated and—in the ideal case—harmonious and unreserved.

If Plato and Aristotle recognized only a prudential or self-interested conception of the good, as Sidgwick and Irwin suppose, then presumably they would have portrayed the ideally virtuous human being as someone who always reasons in a recognizably self-interested or prudential light about what to do. It is clear, however, that they do not. Consider, for instance, Socrates' portrayal of the true philosopher's reasoning about political power. Socrates (here, I think, the mouthpiece for Plato) argues that, in the ideally just city, the philosopher would owe his wisdom to the education bestowed upon him by the city, and for that reason cannot justly refuse a demand to put his wisdom to use in ruling the city.Footnote ¹⁶ When he accepts the demand, he does not do this in the name of his own good. He does it for the good of the city.Footnote ¹⁷ He is fit for rule only if it is true of him that he would prefer to have been left alone to spend his time in pure contemplation.Footnote ¹⁸ His grasp of the Form of the Good is put to use in grasping a fact about justice in the city—namely, that his role as a trained philosopher is to rule the city if required to do so, and that no city can count as just unless each citizen plays his or her properly prescribed civic role. Refusing to rule after having seen this would be tantamount to refusing the verdict of the reasoning part of his soul, and this would constitute the (partial) disordering of his psyche, since it would amount to the temporary unseating of the reasoning part of his soul from its proper place of rule within his own psyche. This means that refusing to rule would be bad for him. It is crucial to notice, however, that this act would count as the disordering of his psyche, and hence as bad for him, only because he judges truly that it would be wrong on other grounds. The ground-level practical judgment here does not directly concern the proper ordering of his soul but rather the proper ordering of the city. The proper ordering of the soul conduces to this judgment and is perpetuated by making and acting on this judgment, but it is not the subject matter of this judgment. It could not have been Plato's view, then, that the ideal deliberator would choose only in the name of his own good.

Aristotle follows Plato in the main particulars just emphasized. He too sees the virtues as constituents of a well-ordered psyche, without which the human self dissipates into a disordered multiplicity, and with which human beings are capable of full and unimpeded engagement in the intrinsically valuable, self-directed activities that compose a good human life.Footnote ¹⁹ There is a recognizable strain of Platonism, too, in Aristotle's insistence that genuine self-love does not consist in a determination to get as large as possible a share of what one desires or finds pleasurable, but rather in an abiding concern to act in accordance with the dictates of right reason.Footnote ²⁰ This counts as self-love because the reasoning part of the psyche is the part that is most truly and properly one's own, and because this part must be ascendant if one is to pull oneself together into a unified and coherent self. Hence, the true self-lover is the person who loves himself because and to the extent that he thinks and acts as he should, and not merely out of a kind of brute self-preference. This sort of self-love leads one to do nothing but what is objectively fine and hence properly praised by all. That is, it wholly defangs the potentiality for conflict between the demands of self-love and the practical exigencies associated with the objectively good.Footnote ²¹

Plato and Aristotle cannot be accused of having simply overlooked or ignored the view that there is an essentially person-relative species of goodness. We can find in their texts the makings of an interesting account of the persistent temptation to misconceive of the good as relative to subjective desires or appetitive pleasures. If we accept Plato's conception of the soul and its fate in nonideal political circumstances (which is to say, in this world as we know it and as it is likely to remain), we would expect most actual human beings to view the good through the distorting lens of overindulged appetites and/or an overweening thumos. Nor would Plato expect those who call themselves philosophers to see things more clearly. It is a pervasive theme of Plato's dialogues that the same deformations of the psyche that lead to errors about the good also place limits on the therapeutic power of philosophy to correct these errors. For instance, those with a thirst for power or public approval will have too calculating a relation to their own words and thoughts, and too indignant a reaction to illuminating objections, to learn readily from philosophical dialogue.Footnote ²²

Aristotle, for his part, offers an explicit account of the role of desires and pleasures in structuring the appearances (phantasmata) of goodness and badness that provide initial guidance to practical thinking.Footnote ²³ If one's characteristic desires and pleasures have been misshapen by a bad upbringing, one will lack a reasoned appreciation of virtuous action, and philosophical inquiry will be powerless to set things aright. The therapeutic powers of philosophy extend only to the stabilization and deepening of evaluative outlooks that are already tolerably virtuous.Footnote ²⁴ Those who have been badly brought up will see goodness in what they desire or what pleases them, but there will be no deeper commonality in these appearances of goodness that might be uncovered by sustained philosophical reflection. If such people attempt to put their conceptions of the good into words, their efforts might be expected to come to a rapid end with the general claim that the only thing shared in common by all good things is that they are desired by oneself or pleasing to oneself, and that this commonality must explain their goodness. Since it would be baldly implausible to suppose that only one's own desires or pleasures have the capacity to confer goodness, this thought leads naturally to an essentially person-relative conception of goodness: what is pleasing to or desired by a person is good for that person. Thus might one diagnose the rise of the subjective pole of modern dualism.

What might explain the rise of the objective pole? If one were tempted to affirm a brutely person-relative conception of goodness, one might well encounter phenomena that resisted explanation within its parameters. For instance, we often have arguments about the good, and we generally think that something is at stake in such arguments. If we interpreted each interlocutor as making speaker-relative claims about what is good for himself or herself, we would have to disown the stubborn intuition that such arguments involve genuine disagreements. Likewise, in the course of deliberation, we often wonder what should please us or what we should desire, and the answer seems to require a conception of the good that is not anchored directly in actual pleasures or desires. These recalcitrant phenomena might induce us to reject subjectivism itself. I think they should. But they might also prompt us to supplement a subjectivist conception of the good with another, more objective idea of goodness, perhaps by aggregating across subjective goods (à la utilitarianism) or by elaborating transcendental conditions for the possibility of coherently valuing subjective ends (à la Kantianism). We would then have arrived at modern dualism.

V. Against Brutely Person-Relative Notions of “Goodness For”

There is nothing distinctively modern about the mere thought that there is some important distinction between what is good for oneself and what is good simpliciter. As noted above, Plato and Aristotle have the conceptual resources to make this distinction. They can explain it as nothing less than the difference between those things that make my life good simpliciter and those that make anything else good simpliciter. This general approach has few friends in modern philosophy, but it was affirmed in the following passage from G. E. Moore:

If we were to accept Moore's analysis of talk of “goodness for,” there would be no point in the extended theoretical analysis of this sort of goodness. A footnote would do. Not surprisingly, friends of this theoretical enterprise have maintained, as against Moore, that what contributes to the objective goodness of a person's life need not be good for that person. As Sumner puts it:

Sumner is undoubtedly right that it is possible for a person to be unable to appreciate certain respects in which his own life has been objectively good, and to wish that his life had been different even in ways that would have made it objectively worse. But if we turn to this passage after a fresh reading of the works of Plato and Aristotle, we will find it natural to ask why this detracts only from the value of the person's life for him and not also from its goodness in what Sumner calls the “perfectionist” sense. An inability to appreciate the defining activities of one's life must imply either that those activities are unworthy of appreciation or that one is unable to appreciate that which is worthy of appreciation. In other words, it implies either a deficiency in one's activities or in one's running apprehension of the value of those activities. Either way, it implies that one's life has been marked by an objective imperfection. Hence, those who really have measured up fully to the standards of perfection proper to human lives, including those standards that apply to evaluative thought itself, cannot at the same time regret this accomplishment.

This is not to say that those who have done ideally well at their life activities—including that central life activity by which we apprehend and appreciate the goodness manifest in our own lives and in the world around us—must think of their lives as having been ideally good. The monist can side with Aristotle and with common sense in recognizing that the lives of such people might be marred by bad fortune, and that the same events that mar their lives might benefit others. Furthermore, the monist can accept the truism that there is an important difference, for me, between my bad fortune and that of a stranger, or between my life's going well and a stranger's life's going equally well. Only one of these lives is mine, and I must be actualized in this life if I am to be actualized at all. The difference here lies in how we are related to various goods and not in what sorts of goods they are. It is, at heart, nothing less than the difference between persons. This difference can shrink almost to an irrelevance in the context of intimate love or family bonds. Yet in many cases it appears to loom large, and we can accept this appearance at face value without committing ourselves to the existence of an irreducibly person-relative kind of goodness.

It might seem that there must be a difference between “perfectionist value” and welfare because a person can have any array of excellences of character without having a life that is ideally good for her. Even if the Platonic thought that the good person cannot be harmed is not absurd on its face, still it is a controversial claim about the coincidence of two conceptually distinct kinds of value, not a straightforward conceptual truth. But while this point seems extremely plausible, it implies only that one's welfare is conceptually distinct from the impersonal goodness of one's character, not from the impersonal goodness of one's life. In my view, we should follow Aristotle in holding that a good character is necessary, but not remotely sufficient, for an ideally good life. Such a life requires many other goods, including health, fulfilling loves and friendships, engagement with the arts and sciences, and the good fortune not to be laid low by oppression, abuse, hunger, or physical or psychological illness, not to mention earthquakes, tidal waves, land mines, military invasions, or oversized asteroids. Without such good fortune, a good character cannot express itself in a full life of valuable activities. Hence, we can accept the common-sense thought that being an ideally good person is no guarantee of living an ideally good life, without insisting upon the existence of a special, essentially perspectival kind of goodness that a life can have for the person leading it.

On the Moorean approach, we cannot say which goods are good for X unless we can say which goods X possesses. Sumner suggests that there is no plausible way for the Moorean to spell out what it means for someone to possess a good, such that each person possesses all and only those goods that conduce to that person's well-being. The most plausible approach would be to hold that intrinsically valuable experiences or activities are good for those who experience or engage in them. Sumner thinks that this sort of approach reverses the order of explanation between intrinsic value and well-being. “It is believable,” Sumner writes, “that intimate connections with others make the world a better place by virtue of enriching our lives, but it is not believable that they enrich our lives by making the world a better place.”Footnote ²⁷ This objection misses the mark. What the Moorean must accept is that our lives are made richer by experiences or activities that are intrinsically good. We can affirm this plausible-sounding claim without settling the question whether a clear sense can be attached to talk of the value of the world.Footnote ²⁸

Moore's deepest reason for analyzing talk of “goodness for” in terms of goodness simpliciter is that otherwise the concept becomes unintelligibly speaker-relative. When I speak of that which is good for me, Moore writes, “The good of it can in no possible sense be ‘private’ or belong to me; any more than a thing can exist privately or for one person only….”Footnote ²⁹ I believe that Moore is right about this. His argument might be reconstructed as follows. If something is good for me, this implies that I would be right to regard it as good. But the proposition that the thing is good cannot very well be true when I entertain it and false when other people entertain it. The proposition is either true or it isn't. Hence, anything that really is good for me must be objectively good. This is not to say that it must be good for everyone. It could, after all, be good for me in the sense favored by Moore—i.e., an objectively good occurrence arising in my life rather than in the lives of others. It is only to disown the thought that its goodness is mine in the sense of being something that I can affirm and others can deny without someone's being mistaken.

We might try to preserve a distinctive realm of person-relative goodness by denying that when something is good for me, I would be correct in regarding it as good simpliciter. But how, then, am I supposed to see those things that are good for me, if not as good? Presumably, I see them aright when I see them simply and irreducibly as good-for-me. If this is an intelligible way of seeing them, then we sidestep the threat of person-relative truth. Others can agree that things that are good-for-me are indeed good-for-Tal-Brewer—provided that they have some way of understanding what they are agreeing to. Yet if goodness for is not a kind of goodness, they will be unable to bring their understanding of goodness to bear on the task of understanding what this phrase means. But how, then, are they to cotton on to its meaning?

No concept can be brutely person-indexed. If I attempt to introduce a new person-indexed concept with the phrase ‘nose-for-X’, and tell you that what is a nose-for-me need not be a nose-for-you, you might suppose that this new phrase names the particular nose among all noses that is found on the face of the speaker. But suppose I go on to say that a nose-for-X need not be a nose at all. I will have to tell you what something must be and how it must be related to a particular person if it is to be a nose-for-that-person. If I cannot do that, I will have succeeded in introducing only a meaningless string of vocables.

The question, then, is what nonindexical anchoring relation is supposed to give sense to ‘goodness for X’ if this phrase does not pick out a kind of goodness with a certain relation to X. Sumner holds that it must be some subjective pro-attitude, and this does seem like the most appealing candidate. (It is also just the sort of view that we are led to expect by the Platonic and/or Aristotelian diagnosis of the appeal of subjectivist understandings of goodness.) However, Sumner doubts that traditional candidates such as preferences or desires can do the trick. His own favored candidate is a pro-attitude that he calls personal life-satisfaction, where this involves “an endorsement or affirmation of the conditions of one's life.”Footnote ³⁰ If it is supposed to be a substantive claim that there is a congruence between those things that I affirm (or desire or prefer) and those things that are good for me, then we have simply postponed the puzzle. We still have not seen what it could mean for something to have the irreducibly indexical property of goodness-for-me. The “subjective pro-attitude” approach will help us to get a grip on this strange property only if what it means for something to be good-for-me is for it to be authentically affirmed by me (or desired by me, or preferred by me). Yet those who deny that it is invariably good for me to perform those actions that I authentically affirm (or desire, or prefer) do not seem simply to be ignorant of the meaning of the phrase ‘good for X’. If there is an objection to what they are saying, the objection is not that it is self-contradictory but that it is false.

On reflection, there are strong reasons for supposing that what they are saying is true. This is perhaps easiest to see in the case of Sumner's proposal. The fundamental problem with this proposal is that one begins to fall into circularity as soon as one tries to say what it is to affirm or endorse the conditions of one's life. Presumably, this is a matter of seeing something good or valuable about them. This raises the question what sort of value we see in our life conditions when we affirm or endorse them. If the answer is goodness simpliciter, then it would be surprising in the extreme to suppose that our lives go better for us whenever we affirm or endorse them. After all, we can presumably be mistaken about the goodness simpliciter of our lives. It is hard to believe that things go well for us when we perform bad activities under the mistaken idea that they are good, and that things begin to go worse for us as soon as we gain an inkling of their badness. We might well say of people who are blissfully ignorant of their own badness that for them—that is, according to them—their lives are going perfectly well. But it would be a mistake to regard this manner of speaking as an indication that there is some genuine yet essentially perspectival form of goodness that they currently enjoy but would lose if they were suddenly made aware of the badness of their lives.

Presumably, then, when we endorse our activities in Sumner's sense, we see them as good-for-us rather than as good simpliciter. But if goodness-for-us just is the property that our activities have when we affirm or endorse them, then to affirm or endorse an activity is to view it as something that one affirms or endorses, and to ponder whether to affirm or endorse an activity is to ponder whether it makes sense to see it as an activity one affirms or endorses. This is, of course, nonsense. But it is just the sort of nonsense we ought to expect when we try to ground the value of some array of our life activities in our subjective approbation of them rather than understanding our approbation as answerable to their value. The former alternative leaves nothing for approbation to be—or, at least, nothing that could sensibly be seen as an apt lodestone for key life decisions.

The same fundamental objection arises for other, more familiar subjectivist analyses of ‘goodness for’, including those that hold that what it means for something to be good-for-me is for it to be desired or preferred by me. In the course of deliberation, we ordinarily take great pains to formulate our preferences, even in cases where nothing but our own interests are at stake. This would make little sense if our preferences had the power to confer value upon their objects. We take pains to strike upon the right preferences because we seek to prefer that which is worth preferring. To prefer one thing to another is not merely to have a brute behavioral tendency to “go for” one thing rather than another. Preferences are not mere behavioral obsessions. To prefer one thing to another is to see it as better or more worthy than the other, and to stand ready to act in accordance with this assessment. Here again, we must ask: Better in what sense? And here again, the subjectivist analysis faces a dilemma: if the answer is “better simpliciter,” then the view is deeply implausible; if the answer is “better for us,” then we are no further along in our efforts to understand this puzzling phrase.

Unlike the term ‘preference’, the word ‘desire’ might perhaps be used without infelicity to refer to mere urges, or obsessions, or other dispositions to move one's body in certain goal-directed ways. However, if we count such things as desires, then it is wholly implausible to suppose that our lives go better in any sense whenever any desire is satisfied. If it makes our lives better to act on an obsessive urge that we correctly regard as entirely pointless—an urge, for instance, to wash our hands or collect saucers of mud—this is not because it satisfies a desire but only because and to the extent that it silences the urge and frees us to engage in something genuinely worthwhile. The equation of what is good for us and what we desire becomes remotely plausible only if we assume that having a desire to do something involves having at least an inchoate sense that there is some point or value in doing that thing. This is the position associated with Thomas Scanlon and Dennis Stampe, both of whom hold that desires—or, at least, a crucially important class of desires—are subjective seemings of goodness or of reasons for action.Footnote ³¹ They are, in other words, subjective outlooks on what is sometimes called “the space” of practical reasons or goods.

This conception of desires (for which I have argued elsewhere)Footnote ³² does guarantee an important connection between desire and value, but the connection is not one that will suit the purposes of those who equate what is good for a person with what satisfies that person's desires. Indeed, this equation will confront what will now begin to sound like a familiar dilemma. On the one hand, if desires involve the appearance of objective or impersonal values or reasons, then presumably it is possible for these appearances to be misleading, and it would be highly implausible to suppose that our doings can be made good-for-us (rather than merely good according to us) by the mistaken sense that they have another, more objective sort of value. On the other hand, if desires involve appearances of goodness-for-the-desirer, then it would be strange to suppose that this appearance is self-vindicating, and viciously circular to attempt to explain the very idea of person-relative value by holding that it is possessed by those things that we desire.Footnote ³³

It would be somewhat more plausible to hold that what is good for a particular person is what that person ought to affirm or prefer or desire, or has reason to affirm or prefer or desire,Footnote ³⁴ or would affirm or prefer or desire under epistemically ideal circumstances,Footnote ³⁵ rather than what the person actually does affirm or prefer or desire. Yet none of these positions can provide us with insight into the nature of a special, person-relative kind of goodness. After all, if there are objectively good states of affairs, then presumably we all have reason to affirm or prefer or desire them, and would do so under epistemically ideal circumstances. Likewise, if there are objectively valuable virtues or objectively valid moral principles, then presumably we all have reason to desire to exemplify those virtues or conform to those moral principles, to prefer to do so, and to affirm ourselves when we do.

It might be objected that I have sidestepped the most plausible subjectivist conception of prudential value—namely, the conception that locates prudential value in pleasure and prudential disvalue in pain. It is true that I have not discussed this sort of view, but this is because it strikes me as implausible in a number of ways. While there is considerable plausibility to the claim that extreme physical pain is always bad for those who experience it, it seems misguided to suppose that pleasure is the only thing that makes a life go well for the person leading it.Footnote ³⁶ The problem is not merely that our lives can be made better by the completion of arduous challenges whose painfulness is not compensated by some more durable or intense pleasure. The more fundamental problem is that the pleasure we take in our activities cannot always be conceived in isolation from those activities, and at any rate does not have the sort of “portable” value that would be retained in isolation from the activities that occasion it. There is no way, for instance, to isolate the pleasure we take in a conversation with an old friend from our wholehearted or unreserved engagement in the conversation itself. The pleasure is not the sort of thing that could occur in isolation from the activity, nor can we plausibly regard such a pleasure as valuable if we find no value whatsoever in the activity in which the pleasure is taken.Footnote ³⁷

What we have, then, is a powerful intuition that seems at first blush to support the existence of an essentially person-relative or perspectival goodness, but that cannot be clarified in any way that sustains this first impression. Theories of well-being or welfare that seek to make sense of this intuition at “face value” amount to attempts to shed light on a kind of thing that could not exist—or, at least, that could not exist while also having the normative status intuitively assigned to it. These theories tend to trace welfare to the satisfaction of desires or to the fulfillment of chosen projects. Because of this, they suffer from a single recurring defect: they are reflectively unstable from the point of view of the first-person deliberator. These theories relativize prudential value to subjective states that are wholly unsuited to settling deliberative questions concerning what it would be good for us to do, because they themselves are subjective outlooks on value and their dependability is itself fair game for deliberative review. These kinds of theories might perhaps provide us with a sensible way of thinking about individual advantage for purposes of helping those we care about or making public policy decisions. That is, they might provide a workable basis for a right-based aggregative theory of the social good, but they are wholly unsuited to providing the foundation for a good-based aggregative view.

VI. A Monistic Defense of Personal Projects and Commitments

It might seem that we must affirm a distinct person-relative kind of goodness if we are to have sufficient reason to have a special concern for our own future good, or to make efforts to advance our own plans and projects. After all, if similar future goods in the lives of others are appropriately regarded as good to the same degree and in the same sense as future goods in our own life, it might seem that one would lack any grounds for giving greater weight to the latter. Given our technologically enhanced capacity to do good for those in far-flung locations, those who are not in need can usually bring about more good in the lives of others than in their own lives. If all reasons for action were grounded in impersonal goods, the implication would seem to be that it would be unreasonable for those who are not in need to act in the name of their own idiosyncratic projects and commitments. This, however, would be tantamount to saying that they cannot reasonably have such projects or commitments, since there is no such thing as a project or commitment upon which one never acts. Only dualism, it might seem, can explain the rational priority of one's own projects and commitments while also acknowledging the rationality of moral concern for others.

This objection represents an attempt to generalize a potent criticism of utilitarianism first set out by Bernard Williams. Williams plausibly argues that strict universal adherence to the act-utilitarian goal of maximizing the aggregate good with each and every action would have the self-defeating effect of undermining the idiosyncratic projects, relationships, and commitments that give unity to our lives over time, make them worth living, and serve as invaluable sources of the sort of utility that the utilitarian seeks to promote.Footnote ³⁸ In other words, the universal adoption of selflessness would result in universalized banality, and this is a consequence that ought to be deplored even by those who are perfectly selfless.

It might look as if Williams's criticism applies to any theory that regards what is good for person X as a local instance, occurring in X's life, of a generic kind of goodness that can arise in an undifferentiated form in the lives of most other human beings. That is, it might be thought that if X must view what is good for him as having the same kind of goodness as what is good for others, then he has no reason to do or bring about what is good for him in preference to what is good for others. I believe, however, that this attempt to generalize Williams's argument rests upon a mistake. It rests, in particular, upon the mistaken thought that the good always figures in practical reason in a single, uniform way—namely, as something to be promoted or brought about. It is this conception of the good that makes it tempting to suppose that we should respond in the same way to all prospective goods, whether they would occur in our own life, in the lives of others, or in events or states of affairs that are not plausibly placed “in” any life at all. We should promote them.

This objection looks far less impressive once we arrive at a proper understanding of the place of apprehensions of intrinsic value in practical thought. When we are fully and properly engaged in unfolding activities that we find intrinsically valuable—an engrossing conversation, for instance, or a walk in the woods with a friend or loved one—we grasp the value of the activity not (merely) as something to be brought about but as something already present that is to be savored as it unfolds. A vivid sense of the value of our own unfolding activities does not leave these activities unchanged. We are not awake to our own activities, not engaged in them in such a way as to render them more complete and perfect of their kind, unless we have a vivid apprehension of their value. This does not mean that we must be entertaining the explicit thought that our activity is valuable—indeed, conscious rehearsal of such a thought might well interrupt the requisite appreciation. The point, instead, is that intrinsically valuable activities remain incomplete and imperfect unless we are drawn into them by an immediate appreciation of their value. This is what occurs, for instance, when we are drawn into a conversation by a rapt attention to, or effortless interest in, what is being said.

When we have this activity-completing sort of attentiveness to our unfolding doings, we see value not solely in some future state of affairs that is to be produced, but in an unfolding activity that is already under way. Such attention alters the activity itself by making our engagement in it stimulating and intensely active rather than lethargic, mechanical, or uninspired. That is, it makes the activity pleasurable in Aristotle's sense.Footnote ³⁹ Such attention also lights up certain ways of carrying forward with the activity as better than others, and it typically incites one to carry forward with the activity in just these ways. For instance, our appreciative attention to an evolving conversation brings with it an appreciation of the value of certain next steps that are made possible by the conversation, as when one feels an excited wish to try out a new idea on one's conversational partner.

This sort of practical thought has an intriguingly recursive structure. It involves a running appreciation of one's activity, but it also alters one's activity in such a way as to deepen or perfect the goodness that it brings into view. This recursive structure is, I think, what lies behind the experience of getting lost or wholly absorbed in one's own activities. Our attention enhances the value of the activity toward which it is directed, thus attracting a more unreserved appreciative attention that once again enhances the value of its object. This recursive amplification continues until one's appreciative attention is wholly claimed by what one is doing.

This sort of thought seems proper to all of the many and varied human activities that philosophers commonly count as intrinsically valuable, ranging from contemplative thought and the creation or appreciation of works of art, to sports, to friendship and sexual intimacy. If we accept the Aristotelian thesis that these activities are crucial elements of a good human life, then an answer to Williams's challenge is ready to hand. We have reason to direct our evaluative attention at what is good for us, rather than at what is good for others, and to be induced to act by that attention, because only the former sort of thought can consummate and complete the best and highest human activities. Furthermore, we will often have reason to devote ourselves in particular to those activities that we most desire to engage in, since these are the activities that are most likely to inspire the continuous and spontaneously appreciative attention that constitutes them as fully valuable instances of their kind.

It appears, then, that proper responsiveness to the impersonally good does not alienate us from our own distinctive projects and commitments, nor does it stand as a barrier to their reasonable development. On the contrary, it often serves to draw us into intrinsically good activities and renders them complete and unreserved, and hence fully enjoyable.Footnote ⁴⁰ It is true, of course, that we can also appreciate the intrinsically valuable activities of others, and we can and sometimes should seek to help others to engage in such activities. What we cannot do is to lend to the activities of others the sort of appreciative attention that constitutes them as inspired and fully valuable rather than lifeless and only marginally worthwhile. Thus, we can see, at least in outline, how a proper responsiveness to the good conceived along monistic lines can support a considerable measure of altruism, while also providing powerful reasons for each agent to throw herself into intrinsically valuable projects and pursuits of her own.