I. Political Theory versus the Theory of Ideals

John Rawls describes his theory of a just society as a “realistic utopia,” a phrase that captures the dual nature of normative political theory. On the one hand, theory seeks a standpoint from which it can evaluate existing political arrangements and advance normatively attractive alternatives to them. On the other hand, that standpoint and the conclusions flowing from it must pass a reality test. If critical distance is to be maintained, the status quo cannot be our benchmark. Something can be realistic without being actual. But it must at least be possible, and the understanding of possibility that guides political theory must be appropriate to the political domain.

Implicit in Rawls’s conception of realistic utopia is an assumption I share: political theory is meant to be action-guiding. Recall the famous opening passage of A Theory of Justice:

The truth about justice, if we can arrive at it, is more than contemplative. It has imperative force, whence Rawls’s “must be.”

This conception of political theory as more than contemplative has a distinguished provenance. In the Nicomachean Ethics, Aristotle says that “our present study, unlike the other branches of philosophy, has a practical aim; we are not investigating the nature of virtue for the sake of knowing what it is, but in order that we may become good, without which result our inquiry would be of no use.” ^{Footnote 2} And political theory as Aristotle understood it is likewise a species of practical philosophy.

In recent years an alternative conception of political theory as purely truth-seeking and contemplative, sometimes called the “theory of ideals,” has emerged. According to Alan Hamlin and Zofia Stemplowska, the theory of ideals has as its distinct purpose to “identify, elucidate, and clarify the nature of an ideal or ideals” and to explore issues, such as commensurability, priorities, and trade-offs among the multiplicity of ideals and principles. Unlike theories designed to shape social arrangements, theories of ideals are not subject to considerations of feasibility. ^{Footnote 3}

For reasons related to Aristotle’s critique of Plato’s theory of Forms, I am not persuaded that the theory of ideals as a distinct enterprise makes sense. ^{Footnote 4} For example, it is doubtful that a free-standing inquiry into justice as such can produce determinate results. Consider three familiar concepts: the just act, the just person, and the just society. The adjective “just” does not denote the same idea in all three cases, reflecting differences among the nouns of which the adjective is predicated. We cannot investigate the nature of the just society without some understanding of what a society is, and that understanding will include some account of the limits of what a society can be. In short, we cannot develop a conception of a just society without at the same time raising issues of feasibility. That is why Rawls is right to insist that the utopia he seeks must be “realistic.”

The argument of this essay does not rest on the premise that the theory of ideals is an unworkable enterprise, although I suspect that it is. My focus, rather, is on political theory understood as practical philosophy, as potentially action-guiding. In this domain, even the staunchest defender of the theory of ideals concedes that questions of practical possibility are unavoidable.

II. Varieties of Possibility and Their Role in Political Theory

Some dimensions of possibility are unproblematic. Everyone would agree, I assume, that X is not politically possible if it is not logically possible. This principle, surprisingly, does some real work in political theory. For example, suppose a conception of utopia includes, among its other features, two principles: first, whenever a proposal can make some people better-off without making anyone worse-off, society should accept and implement it; and second, at least some individual choices should not be subject to collective determination. In a famous article, Amartya Sen demonstrates that given certain background assumptions, these two features contradict each other: society can embody either the Pareto principle or the liberal principle, but not both. ^{Footnote 5} Other well-established findings of social choice theory impose similar logical constraints on possible utopias.

Another uncontroversial dimension of possibility is nomological. Everyone grants, I presume, that compatibility with basic laws of nature marks the dividing line between utopia and science fiction. (Imagine a proposed utopia whose economic arrangements depend upon the existence of perpetual motion machines.) Granted, science progresses, and new discoveries often supersede long-held verities. Still, some laws are so fundamental to our understanding of how the world works that we cannot disregard them as long as there is no evidence that warrants doubt or disbelief.

A third uncontroversial dimension is biological. We are organic beings with distinctive physiological structures, and any theory of human society must take into account these basic features of our species. It would be interesting to speculate on how our social arrangements would differ if we nourished ourselves through photosynthesis rather than ingestion. The agricultural sector as we know it would not exist; nor would the culinary arts (or, I fear, dinner-table conversations). Such contrary-to-fact inquiries would be of theoretical utility, illuminating why basic features of our social life are what they are. But they would be of no practical or normative use, because there is no possibility that these features could be other than what they are.

To be sure, some features of human life long taken to be fundamental have turned out to be mutable. Since the dawn of history it was assumed that fertilization — and hence human reproduction — could occur in only one way and that divine intervention was the only remedy for infertility. (Numerous Biblical stories turn on this assumption, as did some key moments in the history of European monarchies. ^{Footnote 6} ) The invention of in vitro fertilization expanded the scope of the possible, as have other reproductive technologies, and these changes have had important social consequences. In a different vein, gender differences in social roles, long thought to be rooted in human biology, have turned out to be more socially constructed than gender essentialists were willing to admit.

So we must be cautious about assuming that specific features of human life are unalterable. But caution should turn into doubt only when we have good reason to suspect that scientific advances or social changes could affect some aspect of human biology that we previously took to be fixed. A shift from ingestion to photosynthesis is not one of them; so as far as we know, agriculture is a permanent part of human life.

If the role of logical, nomological, and biological possibility in political theory is not essentially contested, when we argue about the role of feasibility conditions in ideal theory, what are we arguing about?

III. Conceptions of Possibility

One dimension of contestation is the conception of “possibility” appropriate to the political domain.

“The idea that something that has hitherto been unsuccessful will therefore never be successful does not justify anyone in abandoning even a pragmatic or technical aim,” Kant wrote in “Theory and Practice.” And he continued: “This applies even more to moral aims, which, so long as it is not demonstrably impossible to fulfil them, amount to duties.” ^{Footnote 7} As Juhu Raikka has observed, this conception of possibility has deeply influenced contemporary political theorists, especially those inspired by Kant directly or through Rawls. ^{Footnote 8}

While not being demonstrably impossible may be an appropriate conception of possibility for some domains, politics is not one of them. Demonstrating that X is impossible is rarely possible. Depending on how strictly we construe the criterion of demonstration, it could rule out nothing more than logical impossibility; scientific inquiry can and often does revise even well-established laws of nature. And even if we broaden the idea of demonstration to include premises based on scientific laws that we have no reason to doubt, it remains a criterion that excludes little of relevance to politics. Furthermore, it is bound to be underinclusive, unless one believes that the domain of the impossible includes only those matters that can be demonstrated to be so. We know that is not the case even for axiomatized mathematical systems, so we have no reason to believe that it is true for politics.

Demonstration is a standard that does not admit of degrees; either you have demonstrated a proposition or you haven’t. It excludes all consideration of probability. As David Estlund rightly says, “It is not the case that ought implies reasonably likely.” ^{Footnote 9} It does not follow that we are obligated to pursue aims that are highly unlikely. At some point, the degree of improbability of success becomes relevant to our moral obligation, especially when the pursuit of a moral aim entails moral costs.

This is true even when the improbability rests on features of human motivation. Suppose, arguendo, that Lenin and Stalin could have created a pure communist society if the system had succeeded in creating a population of New Soviet Men whose only desire was to maximize aggregate well-being as defined by the party leaders. The potential gains in equality, social harmony, and civic spirit could have been substantial. But the improbability of creating such a population was more than enough to vitiate whatever obligation there may have been to adopt such policies. (And the costs of trying, I would argue, were so high as to erect a prohibition against making the attempt.)

The consequences of improbability extend well beyond human motivation. For example, some things that science shows to be possible are so improbable that we would be crazy to bring them into our deliberations. Consider a box filled with two gases, A and B, separated by an impermeable partition. If we remove the partition and wait for a suitable length of time, a random distribution of A and B throughout the box is a near-certainty. The probability that all the A-molecules and B-molecules will remain separated is extremely low. But it is not zero. ^{Footnote 10}

Now suppose that B is deadly, even at the concentration that a random distribution of A and B would yield. It would be crazy to smell the side of the box in which A was initially located on the off-chance that random motion had left all of B on the other side. Even if an evil demon assures you that if you sniff just once and survive, Kantian perpetual peace (or if you prefer, Rawls’s law of peoples) would become the permanent condition of the human race and that under no other circumstances could that happen, you would have no obligation to do so. When achieving a moral aim becomes sufficiently improbable, duty ceases.

I agree with Estlund that we should reject what he calls “complacent realism.” ^{Footnote 11} Surely theory is useless if it merely describes and ratifies the status quo. He wants to defend the kind of theorizing that “holds the real world to higher standards than it actually meets.” ^{Footnote 12} I suspect we all do. The crucial question is why the real world fails to meet those standards.

There are three hopeful explanations. One is the lack of understanding or imagination: if people could be convinced that certain changes they hadn’t previously considered were possible, they would go along with them. Another is that people harbor prejudices about other human beings or unfamiliar social arrangements that time and persuasion can change. The American people changed their minds about gay rights and same-sex marriage remarkably quickly, largely because they became more familiar with their gay fellow citizens and became decreasingly comfortable with the arguments against according them equal rights. A third possibility is that unequal power relations allow some to oppress others or to advance their interests at others’ expense. Here again, savvy political strategies can sometimes bring about fundamental change. Once the Voting Rights Act empowered African Americans, white southern politicians had no choice but to take them into account, and changed practices led to shifts in attitudes.

In cases like these, the reasons why the world fails to meet higher standards turn out to be malleable. What is infeasible today may be feasible tomorrow or the day after. These are the circumstances in which undemanding norms are the enemy of achievable moral progress.

This sense of moral possibility animated the man who became America’s greatest president. During his debate with Stephen Douglas, Abraham Lincoln offered a memorable interpretation of the Declaration of Independence. When its authors asserted that all men are created equal, Lincoln said, they knew that not all were enjoying that equality or would do so anytime soon:

Lincoln assumed, or hoped, that the principle of moral equality would effect a change in the beliefs and sentiments of peoples around the world and that progress toward this principle would set in motion a virtuous cycle leading to further progress. Experience has vindicated his belief in the malleability of entrenched prejudices. (It has also shown that some individuals will not change their behavior without being compelled to do so.)

But are all the morally relevant features of the human species as malleable as are their beliefs and sentiments? It is tempting to reason from the true proposition that some individuals are capable of meeting demanding standards to the conclusion that all are and to shape a conception of realistic utopia on that basis. The unexamined background assumption is that human beings are endowed with substantially equal moral capacities. But why should we accept that assumption when there is so much evidence against it?

Consider the case of a grenade thrown into a foxhole. We know that it is possible for individuals to fall on it, saving their comrades as they sacrifice themselves. We also know that it is highly desirable for individuals to do so. But we neither expect all to act in this manner or consider the failure to do so as a sign of moral defect. Instead, we give individuals who perform such acts special (often posthumous) recognition.

Similarly, some citizens will regularly act on their understanding of what’s best for the country, regardless of its effect on their own well-being. All honor to them. But in what sense is it reasonable to expect that everyone should do so? Yes, one can imagine a world in which they do, and that world looks much more attractive than does the one in which we live. But that world is the equivalent of a self-enclosed mathematical system in which the conclusions follow from the axioms but have no bearing on anything outside that system.

If you are looking for one climber to scale a rugged mountain, the fact that what’s feasible for him would be much too difficult for others does not matter. But if you’re leading a hike for a diverse group, you cannot take your bearings from the person who is in the best condition. When it may not literally be impossible that an out-of-shape retiree could keep up with a lean and fit college student, it is improbable enough to be excluded as the basis of a course of action for the entire group. Possible for some does not mean possible for all.

For this reason, societies draw rough and ready distinctions between actions thought to be within the capacity of all or nearly all and those that are not. Moral requirements are limited to the former, while the latter shapes societal understandings of human excellence and supererogatory conduct. It is not unrealistically utopian for a society to hope for, or actively create the conditions for, the development of unusual capacities; it is utopian to expect that excellence, moral or otherwise, will pervade the entire populace.

The proposition that X is both desirable and possible (for some human beings) does not warrant the conclusion that X ought to be an aspirational standard for individuals or institutions generally. The exemplary should not set the standards for the ordinary. Political communities comprise individuals whose moral capacities differ significantly. That is why the category of the supererogatory does real work in political as well as moral theory.

We need not take people just as we find them, of course. Political communities confirm what many political theorists assert: morally grounded civic education can make a difference. But its efficacy has limits. Even if I had devoted my entire life to physical training, I never could have run a four-minute mile. Only a small fraction of the population begins with the capacity for excellence in running that expert training can develop. Similarly, some individuals have an unusual capacity for altruism or for devotion to the common good. But there is no reason to believe that even the best form of civic education would make such behavior pervasive throughout the population.

Aristotle observes in the Politics that “What belongs in common to the most people is accorded the least care: they take thought for their own things above all, and less about things common, or only so much as falls to each individually. For apart from other things, they slight them on the grounds that someone else is taking thought for them . . . .” ^{Footnote 14} Although some individuals will exhibit the same concern for public spaces as for their own homes, there is no reason to expect that most people most of the time will be indifferent to the difference between what is theirs and what isn’t, whether we are considering physical property or children, and a realistic utopia will take this into account.

The same holds for democratic citizenship. “There is no reason to believe that democratic procedures would lead to anything but disaster if voters looked out exclusively for their own interests or those of people they are especially close to,” Estlund says. ^{Footnote 15} I agree, but the operative word is “exclusively.” A requirement that voters give no more weight to their own interests or to those of their near and dear than to the interests of other citizens would be utopian in the wrong way. Most people most of the time will act on self-preference to some extent, and within limits they are not wrong to do so. The challenge is to define a demanding but reasonable balance between self-preference and civic-mindedness.

Two considerations shape this balance. First, a degree of preference for one’s own well-being and that of one’s family is morally defensible. As citizens, we are obligated to give great weight to the interests of all children in our community, but we are not obligated to give other people’s children just the same weight as our own. And second, whatever we may wish and whatever moral theory may say, a degree of self-preference is baked into our nature. We can restrict it through reflection and choice, but we cannot hope to expunge it. So the question voters should ask is not the unmodified “What ought we do as a community?” but rather, “What ought we do consistent with my own defensible interests?”

It is in this light, I suggest, that we can understand and apply Onora O’Neill’s important distinction between abstraction and idealization. As she formulates the difference,

The problem for both empirical and normative theories is that contrary-to-fact premises often lead to unsound conclusions. For example, economists often assume that agents are rational, self-interested, and in possession of all relevant information. Elegant theories can be constructed on this basis, but the predictions they generate are likely to be wide of the mark. By incorporating more realistic assumptions, such as cognitive distortions and motivation dispositions that violate rationality, behavioral economics can give a more accurate account of the decisions we actually make.

It is true, as Andrew Mason suggests, that we can remove the taint of falsehood from idealizations by representing them as counterfactuals: If we were rational in the economists’ sense (even though we aren’t), what predictions would follow? ^{Footnote 17} The answers this question generates may be of theoretical interest. But if we have good reason to believe that the counterfactual cannot become actual, then these answers are neither empirically nor normatively useful. “What would justice be if we were all disposed to act justly and needed only to agree on principles of justice?” is an interesting theoretical question. But because the assumption is deeply and permanently counterfactual, it cannot yield sound, let alone workable, conclusions for politics as we know it, dominated as it is by beings whose motivational imperfections are irremediable.

It is also true, as Zofia Stemplowska suggests, that by assuming the “fanciful” in our normative theories, we can learn “just how crucial certain constraints are to shaping what we consider desirable or just.” For example, “assuming that human nature was more malleable than we think it is would allow us to see how our conception of a relatively rigid human nature shapes what we consider to be just.” ^{Footnote 18} This is of theoretical interest, but again, if we have good reason to believe that the assumption of greater malleability is deeply counterfactual, we have no reason to accept as normatively binding or as action-guiding the conclusions that may flow from it.

Political theories that are unrealistically utopian often elide the distinction between what is possible for some and what is possible for all. In defending his conception of “hopeless realism,” Estlund says that even when there is good reason to believe that a vision of a good society will never be realized in practice, it can still qualify as a normative standard:

The operative assumption in this account, it appears, is that all citizens can be publicly and privately virtuous in the required sense, not just in moments of civic danger, but steadily and enduringly. This I deny. Can I demonstrate the truth of my denial? Of course not; I fail Kant’s test. But so what? That is not the right test. The burden of proof lies on those who assert that observations confirmed by the preponderance of human experience since the beginning of recorded time are irrelevant to establishing reasonable limits of human possibility. ^{Footnote 20}

It is dangerous to assume that the failure of most people to live up to the standards that exemplary individuals can meet is a “defect of people.” That formulation opens the door to political schemes designed to make people better than they can be — and to increasingly harsh measures when they fail (as they inevitably will) to meet that test. ^{Footnote 21}

To recognize the inequality of moral capacities as a permanent feature of the human condition is not to discard, or even jeopardize, the principle of human equality. Despite differences of moral capacity, we can continue to see human beings as equal in moral worth or, if you prefer, as equally endowed with fundamental rights. ^{Footnote 22} The respects in which we are unequal shape what we can reasonably expect from political life and, to an extent, how we should shape our political institutions. These inequalities do not allow some to regard others as being of lesser worth or to place them in a position of structural subordination. As Thomas Jefferson understood, democracy and inequality of capacities can coexist, in theory and in practice.

IV. Human Nature

As John Rawls understood better than many of his followers, political theory did not begin with A Theory of Justice. The construction of ideal polities — utopias — goes back at least to Plato’s Republic. And for at least half a millennium, thinkers have objected to this practice. As Machiavelli famously said in The Prince, “It appears to me more appropriate to follow up the real truth of a matter than the imagination of it; for many have pictured republics and principalities which in fact have never been known or seen, because how one lives is so far distant from how one ought to live, that he who neglects what is done for what ought to be done, sooner effects his ruin than his preservation.” ^{Footnote 23} Machiavelli’s logic is straightforward: because there is far more evil than virtue in the political arena, and because virtue limits what we are allowed to do to further our interests, virtuous individuals are bound to lose out to those with fewer scruples. We cannot be required to conspire in our own downfall; what we ought to do is significantly shaped by our moral environment, in which immorality predominates.

It is interesting to speculate about how Machiavelli would have reacted to Rawls’s theory, especially the assumption that nearly all of us possess a sense of justice adequate to secure compliance with principles of justice that we all accept. I suspect that the Florentine would have asked why anyone should assume anything of the sort, and why those who make this assumption think that its implications should bind us in the real world.

Whatever the merits of this critique, Rawls invites it, or at least opens the door to it, when he describes his theory of justice as a “realistic utopia,” implying that there can be unrealistic utopias as well. And he offers a general account of the line separating the two. In A Theory of Justice, following Hume, he depicts the “circumstances of justice” as including the assumption of moderate scarcity: “Natural and other resources are not so abundant that schemes of cooperation become superfluous, nor are conditions so harsh that fruitful ventures must inevitably break down.” ^{Footnote 24} In short, just political communities require reasonably favorable material conditions, which obtain in some situations but not others. In The Law of Peoples, he follows Rousseau, whose Social Contract proceeded by “taking men as they are and laws as they might be.” The “men as they are” proviso, Rawls says, “refers to persons’ moral and psychological natures and how that nature works within the framework of political and social institutions.” ^{Footnote 25}

There are, then, three dimensions along which the distinction between realistic and unrealistic utopias may be determined. The first two are relatively unproblematic. Whatever may be the case in the least fortunate regions of sub-Saharan Africa, moderate scarcity seems a reasonable description of circumstances in Europe and North America. And while the nature and circumstances of law impose some limits on what laws can be, most of our debates about what the law should be take place within those limits. To be sure, the balance of political power in specific situations may make desirable legal reform impossible, but that hardly rules out the possibility that reform will become possible as the community’s political circumstances change. Three decades ago, the legalization of same-sex marriage seemed impossible; today it is inevitable. The plasticity of law reflects the malleability of public opinion and political arrangements.

It is the “men as they are” principle that does most of the work and occasions most of the controversy. As Rawls puts it, “Surely he [Rousseau] doesn’t mean people as he sees them now, with all the vices and habits of a corrupt civilization (as described in the Second Discourse). Rather, he means human beings as they are according to the basic principles and propensities of human nature.” But how can Rousseau (or we) determine those principles and propensities? Rawls continues: “These principles and propensities as those by reference to which we can account for the kinds of virtues and vices, aims and aspirations, final ends and desires — in short, the kind of character people have under different social conditions.” ^{Footnote 26}

The link between character and social conditions provides the clue to Rawls’s bottom line. In what is admittedly an optimistic reading of Rousseau, Rawls sees human beings as naturally good. “To say that human nature is good,” he says, is to say that “citizens who grow up under reasonable and just institutions . . . will affirm those institutions and act to make sure their social world endures.” Living in a just society, then, suffices to expunge antisocial sentiments from the panoply of human desires. That is part of what Rawls means when he insists that just social arrangements can be stable because they can generate their own support.

Rousseau distinguishes between amour de soi — concern for one’s own interests — and amour propre — concern for one’s status relative to others. Interpreters of Rousseau have often seen this latter concern as the source of the desire for superior rank and respect, inherently scarce positional goods that occasion endless social conflict. This is a misreading of Rousseau, Rawls insists. There are two forms of amour propre, the natural and the unnatural or “perverted.” The former is the “natural concern for a secure standing in relation to others and involves a need for equal acceptance with them.” The latter — unnatural — form of amour propre shows itself in “such vices as vanity and arrogance, in the desire to be superior to and to dominate others, and to be admired by them.” It’s unnatural or perverted object, Rawls concludes, is to be “superior to others and to have them in positions beneath us.” ^{Footnote 27}

Self-respect is perfectly compatible with recognizing others as equals, and thus with the reciprocity at the heart of just social arrangements. Only the desire for equality-seeking honor is innate, because self-respect is a fundamental human good. The desire to dominate others and to be recognized as superior is the consequence of flawed social institutions, and so is destructive competition. ^{Footnote 28} We need not be locked in an endless battle for material and social superiority; the desire for equality has its roots in human nature. ^{Footnote 29}

Rawls is aware that these propositions may sound unrealistically optimistic about human beings and social possibilities. Anticipating this critique, he quotes Rousseau, who asserts in the Social Contract that “The limits of the possible in moral matters are less narrow than we think. It is our weaknesses, our vices, our prejudices, that shrink them.” ^{Footnote 30} But Rawls acknowledges that he has adopted the only interpretation of Rousseau’s view of human nature that avoids dark pessimism. If we accept the view, favored by many Rousseau scholars, that unnatural amour propre is what human nature inevitably becomes in society, then “the kind of political society depicted in the Social Contract appears utterly [i.e., unrealistically] utopian.” If human nature is not naturally inclined toward equality, then an egalitarian society constituted by principles of justice such as justice as fairness becomes “unworkable.” ^{Footnote 31}

What are we to make of this?

On the one hand, Rawls’s optimism is clearly warranted — to an extent. Over and over again, reformers have shown that injustice can be overcome and that society can change for the better. In every case they have contended with skeptics who argued that these changes were not possible because they flew in the face of entrenched prejudice. But bad opinions can give way to better ones when circumstances change.

On the other hand, to allow the conditions for a realistic utopia to drive our understanding of human nature is to put the cart before the horse. If human nature is a meaningful concept and not simply a social construction, then its limits should shape our understanding of social possibility, not vice versa.

One way of putting the horse back in front of the cart is to wonder whether human motivations are as malleable as are their beliefs. A substantial tradition of thought, secular and religious, claims that they are not and that our innate motives are a mixed bag. Judaism distinguishes between good and bad urges (yetzer ha-tov and yetzer ha-ra), and a well-known Talmudic saying links government to the latter: “Rabbi Chanina, deputy to the kohanim [priests], would say: ‘Pray for the integrity of the government; for were it not for the fear of its authority, a man would swallow his neighbor alive.’” ^{Footnote 32} St. Augustine saw the libido dominandi — the lust for power, rule, and domination over others — as a fundamental human drive. James Madison argued that to ward off tyranny, republican institutions should force individuals and institutions to vie against one another: “Ambition must be made to counteract ambition.” Anticipating the charge that this represented an unduly negative view of our species, Madison asked, “What is government itself but the greatest of all reflections on human nature?” and continued: “If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary.” But neither antecedent is or will be the case, which is why it is dangerous to presume more virtue than there is in either rulers or citizens. “Man’s capacity for justice makes democracy possible,” said Reinhold Niebuhr, “but man’s capacity for injustice makes democracy necessary.” Neither is the product of social institutions; the capacity for injustice is just as deeply entrenched in our nature as is the capacity for justice. That is what led Niebuhr to declare, with the jaunty blend of pessimism and optimism that became his trademark, that “[o]riginal sin is the only empirically verifiable doctrine of the Christian faith.”

Neither secular nor religious thought is of one mind on this question. Within Christianity, the doctrine that became known as the Pelagian heresy denied original sin and affirmed that human beings could realize their innate capacity for goodness by their own efforts, without divine grace. American thinkers from Ralph Waldo Emerson to Abraham Maslow have denied the innateness of evil. “Our young people,” said Emerson, “are diseased with the theological problems of original sin, the origin of evil, predestination, and the like. . . . These are the soul’s mumps, and measles, and whooping cough. . . . A simple mind will not know these enemies.” ^{Footnote 33}

My point is not that Rawls is obviously wrong to follow Rousseau in endorsing the doctrine of man’s goodness. It is rather that by positing the capacity for a sense of justice as part of our moral nature while regarding injustice as the product of social institutions, Rawls is taking sides in one of the long-running debates in history. Those of us who believe, with Madison and others, that the desire for domination is as primordial as the capacity for cooperation are bound to regard Rawls’s premise, and the picture of society that flows from it, as utopian in the bad sense. And Rawls gives us no compelling reason to take his side in this debate.

He does however provide a reason that compels him, and it offers a clue about the basic motivation of his theory. To repeat: Rawls advances the conception of human nature that allows him to regard his account of a just society as feasible. The opposite conclusion — that justice cannot be realized on earth — would be a moral and human disaster, or so he believes. I reproduce his case at some length:

The stakes are very high, Rawls insists: if a reasonably just society is not possible, and if human beings are too self-centered and amoral to be capable of acting on a sense of justice, “one might ask, with Kant, whether it is worthwhile for human beings to live on the earth.” ^{Footnote 35}

This claim, I believe rests on two premises that we have no good reason to accept. In the first place, if human beings vary in their capacity for morality and justice, as I have argued they do, then a reasonably just society as Rawls understands it is not to be expected. Even so, a conception of justice can guide feasible reforms that improve the world without perfecting it. That is the point of Lincoln’s “standard maxim of free society” quoted earlier. I do not see why working to make our society and the world better would not suffice to provide meaning to our actions, even if we know that the gap between our ideals and reality is permanent.

Second, I do not see why the possibility of a just polity is needed to ward off resignation and cynicism. Surely there are other sources of meaning and purpose — art, philosophy, religion, or doing our best to be a good spouse and parent and to live a decent life. I would turn Rawls’s proposition on its head: it is when we invest too much hope in the political domain that we are most likely to fall prey to resignation and cynicism. Disillusion is the fruit of illusion, not of low expectations. A sober recognition of the permanent gulf between improvement and perfection offers the best chance of keeping our moral and emotional balance.

V. Conclusion: Balancing Political Risks

There is a political backdrop to this seemingly academic debate about ideal theory. Many realists — not least liberal realists — fear that unrealizable demands provide a predicate for government coercion when change inevitably falls short of high hopes. Many ambitious reformers — especially liberal idealists — fear that an experience-based concept of feasibility will preserve an unjust status quo that could be changed through determined action guided by more ambitious aspirations. Each group rests its stance on incontrovertible evidence. Those who fear ambition can point to any number of utopian revolutionary movements since the French Revolution; those who fear caution can cite successful social movements of the past two generations, which were undertaken in the face of claims that radical changes in race and gender relations or in the legal status of same-sex orientations were impossible.

Estlund worries that “We sometimes expect too little precisely because we have no normative standard that forces the question of whether more can realistically be expected.” ^{Footnote 36} But there’s a countervailing worry: we sometimes expect too much because we have no standard of possibility that forces the question of whether what we want is feasible.

In a similar vein, Zofia Stemplowska asserts that “[U]nless we know what is desirable when there is full compliance, we could adopt a direction of reform for nonideal circumstances that unnecessarily moves us away from the ultimate aim of full compliance.” ^{Footnote 37} Indeed we could. But if, as I have argued, full compliance is not a feasible standard, then a policy based on the assumption that it is feasible could prove unreasonably demanding, moving us further away from a less demanding but realistically achievable aim.

My point is simple: whichever way we turn, there are moral risks as well as potential gains. We can defeat ourselves by attempting too little but also by attempting too much. By being as precise as possible about the line that separates the feasible from the unattainable for the kinds of beings we are, living in the world as we know it, we can reduce these risks. But it is of the essence of the political domain that we can never be completely sure about the location of that line. Whatever our political ideals may be, the actions they guide can never be altogether free of moral risk.

The inevitability of risk does not prove that my account of realistic utopia is correct, of course. But it does show why the claim that my approach may expect too little and risks leaving remediable injustices intact is not compelling: the opposite risk is also possible and (at least) equally dangerous. Political practice that takes political theory seriously understands this. There is a zone within which the feasibility of our aims is a matter of judgment and conjecture. In the name of doing good, we may inflict harm. In the name of avoiding harm, we may neglect the good we could have done. Politics is not a one-way bet, and theory should not encourage politicians to believe that it is.