“Conceptions of justice must be justified by the conditions of our life as we know it or not at all.” Footnote 1
— John Rawls
I. Introduction
In A Theory of Justice and later works, John Rawls presents his vision of the ideal society — a society that is characterized by favorable circumstances and strict compliance with the principles of justice. Footnote 2 The members of this society have a sense of justice and can rely on each other to act justly. Footnote 3 The circumstances are favorable in that people have the education, economic means, and skills needed to establish a constitutional democracy. Footnote 4 The principles of justice in this “realistic utopia,” as Rawls calls it in The Law of Peoples, have been chosen by rational, self-interested, and mutually disinterested representatives behind a veil of ignorance in light of a realistic view of human psychology and sociology. Footnote 5 So although the assumption of strict compliance is unrealistic even given this background, the principles themselves are realistic because they take into account “[t]he general facts of human psychology.” Footnote 6 If people who have a sense of justice nevertheless lack the capacity for acting on a given conception of justice, the fault lies with the conception. According to Rawls, a political conception of justice (but not necessarily a moral conception) must be practicable, it must “fall under the art of the possible.” Footnote 7 Thus, even principles that are not strictly impossible for human beings to comply with, but that only saints or heroes can be expected to comply with, must be ruled out as principles of political justice. Principles that fail to engender widespread compliance even in favorable circumstances cannot be the right principles of justice.
Much of the criticism of Rawls’s theory of justice is directed at the ideal part. Footnote 8 G. A. Cohen, however, criticizes the theory’s realistic element: the view that the right principles of justice must take the general deep facts of human psychology and sociology into account. According to Cohen, justice is not constrained by the limits of our nature or the nature of human societies, even at their historical best. Justice is what it is, even if it will never be realized, fully or at all, thanks to our inherent imperfections. Footnote 9 David Estlund takes the same position, arguing that since our natural motivations can be justice-tainting, they cannot be a constraint on the right conception of justice. Footnote 10 Further, Cohen and Estlund both agree, the right conception of justice is not constrained by the likelihood that implementing it will have negative results, regardless of the social circumstances. If the attempt to implement a certain conception of justice is likely to result in widespread harm or in other injustices, then it should not be implemented. It does not follow, however, that the conception itself is false. Indeed, a conception of justice can be true even if no one can abide by it.
Estlund’s and Cohen’s claim that justice is what it is even if no one can satisfy its demands is reminiscent of Kant’s view that morality requires motivation by Pure Practical Reason, even if no human being ever has or ever will be motivated thus. Footnote 11 In this essay, I address the view that our natural capacities and the likely negative effects of implementing a conception of justice, are irrelevant to its truth or justification. If Estlund’s and Cohen’s arguments are successful, they will have upended not only a crucial part of Rawls’s justification for his conception of justice, but that of most philosophers, from Aristotle to the present. Fortunately, I think their claim is no more plausible than Kant’s. In the next section, I will discuss Cohen’s arguments against basing the fundamental principles of justice on human capacities, including motivational capacities, and in Section III, I will address Estlund’s arguments. Although there is an overlap between their arguments, they are different enough to merit separate treatment. My main argument against them is that (i) there is no way to judge if a principle of justice is sound independently of our motivational capacities or incapacities and the effects that the principle is likely to have if implemented; (ii) a principle of justice that, if implemented, makes it hard or impossible for individuals committed to justice to lead happy and worthwhile lives, even if the circumstances are favorable to living justly, cannot be sound; (iii) without the constraints noted in (i) and (ii), there can be no reason to reject racist, sexist, or other wrongheaded principles of justice that have been advanced as sound over the years, principles that even Cohen and Estlund would reject. In short, justice is justice only if kept within the limits of human nature.
II. The Human Nature Constraint: Cohen’s Critique
Cohen gives two related arguments against the claim that we must choose fundamental or ultimate principles of justice in light of (among other conditions) the basic facts of human nature and the likely effects of implementing them: a metaethical argument and a normative argument.
A. Metaethical argument
Cohen argues that the fundamental principles of justice — indeed, all fundamental moral principles — are fact-insensitive, and, hence, cannot be responses to or reflections of any (descriptive) facts of human nature. As he puts it, “all principles that reflect facts reflect facts only because they reflect principles that don’t reflect facts, and the latter principles form the ultimate foundation of all principles, fact-reflecting principles included.” Footnote 12 Rawls’s principles of justice are, therefore, not fundamental principles. At most, they are either nonfundamental principles derived from fundamental principles in light of relevant facts, or simply rules of regulation.
Cohen explains his view with the help of the principle of promise keeping. Footnote 13 What justifies this principle? Let’s say that it is the fact that promisees cannot pursue their projects if promisers do not keep their promises. But this fact gains whatever normative force it has only thanks to (i) the fact that the ability to pursue their projects is essential to people’s happiness, in conjunction with (ii) the principle that we ought to care about people’s happiness because they merit such concern or because it shows respect for them and they have respect-meriting properties. Cohen proposes that underlying this is the fact-free principle, P4: “one ought to respect beings, human or otherwise, who have the relevant characteristics.” Footnote 14 This principle is true even if no one has such characteristics, although, of course, it applies to human beings or other beings only if they have these characteristics. The truth of a basic principle is not the same thing as its applicability. Basic normative principles are fact-free or a priori, like the truths of arithmetic. Footnote 15 The most general such principle is “One ought to do A if it is possible to do A,” where A is something worth doing. Footnote 16 This is true even if we do not have the capacity to do A, whether this incapacity is organizational or technical, or something inherent in our nature. Footnote 17 In the latter case, our “very makeup” is flawed, although we cannot be blamed for it. Considerations of feasibility cannot constrain the content of basic normative principles.
Everyone would agree that organizational or technical incapacity to act on a moral principle does not show that the principle is false. What is controversial is Cohen’s view that even an innate incapacity does not show that the principle is false. I discuss this below in connection with Cohen’s second argument.
Someone might object that the very terms involved in P4 show an understanding of the relevant facts of human nature, so Cohen has not shown that fundamental principles are fact-insensitive. But this, according to Cohen, is to misunderstand his thesis, which is not a temporal, causal, epistemic, or psychological thesis, but a logical one holding that “affirmations of fact-insensitive principles are logically prior to affirmations of [fact-sensitive] principles.” Footnote 18 It is a thesis about “the structure of belief in principles.” Footnote 19 We may become aware of these principles only as a result of our encounters with facts, but “commitment to fact-sensitive principles carries with it commitment to fact-insensitive principles.” Footnote 20 This is how a clearheaded person must think of her “own normative beliefs.” Footnote 21 The inability to obey a fundamental principle is not a problem because “subscribing to a principle is not an action but the having of a belief or an attitude.” Footnote 22 The important question for political philosophy is “not what we should do but what we should think, even when what we should think makes no practical difference.” Footnote 23 Fundamental principles of justice are not necessarily for guiding our actions, but for “spelling out what justice is,” and justifying the derivative principles and rules of regulation that we need for action-guidance. Footnote 24 Fundamental principles “formulate our fundamental convictions.” Footnote 25
But why should we accept that the fundamental principle that justifies promise keeping is P4 or anything with the logical structure of P4? A more plausible candidate is the partly fact-based P4*: “We ought to respect human beings (only) because they have respect-meriting properties.” This principle is true only if human beings do have such properties. This seems to be the way most clearheaded human beings actually think of their fundamental convictions. And it is enough both for understanding and for justifying promise keeping. We do not achieve greater understanding or justification of why we should keep our promises to other human beings by generalizing to all beings with the relevant properties. But if we did recognize the broader implications of our principle of respect for humans, there is no reason why we couldn’t or shouldn’t commit ourselves to a generalization that is analogous to our generalization for humans: “We ought to respect both human beings and other beings if and only if they have respect-meriting properties.” And this is not a fact-insensitive principle.
My P4* is akin to the fundamental principle implicit in Rawls’s conception of human beings as individuals who are, and who regard themselves as being, free and equal in their moral powers: the principle that human beings deserve to be treated with equal justice (only) because they are free and equal persons. Footnote 26 Since this conception of persons, and thus the principle, underlies Rawls’s choice of the original position and his conception of the right principles of justice, it follows that his principles of justice also presuppose a fundamental principle that is both normative and factual: if persons were not free and equal, this principle would not be true. Footnote 27
B. Normative argument: The irrelevance of human capacities
Cohen’s second argument for his thesis that our fundamental convictions transcend human capacities is that when a human incapacity is said to rule out a moral principle, we need to ask what we would say about the principle if it were within our ability to obey it. Footnote 28 It is only when we “clear the decks of facts about capacity” and answer that question that “we reach the normative ultimate.” Thus, if we exclude “one ought to do A” only because no one can do A, we are bound to agree that “One ought to do A if it is possible to do A.”
On its face, this argument is seductive. For surely anyone with a moral sense would agree that, for example, if we could wave a wand to banish all undeserved suffering from the world, we ought to do so. Or that if we had the powers of Superwoman or Superman, we ought to use them to combat evil everywhere. So far we are in agreement with Cohen. But we also think that since, being mere mortals, we cannot banish suffering or combat evil everywhere, we are not unjust not to do so. To put it differently, in the face of our incapacity, we believe that there is no “ought” remaining — or, at least, none that applies to us. It applies only to fictional heroes with wands or the powers of Superpeople. For the rest of us, what might remain is only an occasional, rueful wish. So if we want a principle that imposes an ought on us, the correct fundamental principle is “One ought to do A if and only if it is possible to do A,” where A is something worth doing. There is no warrant for Cohen’s view that “claims of the form: if it is possible to do A, then you ought to do A” embody “fundamental justice.” Footnote 29
Perhaps, however, when Cohen says that the correct fundamental principles of justice can exceed our human capacities he has in mind only our psychological capacities — that is, human motivations — not our physical capacities. For example, if our examination of our beliefs about justice and our judgments of particular situations reveals that the socialist principle of equality is a sound fundamental principle of justice, then it’s a sound principle, even though we do not know how to make it work, or even if we are too selfish or greedy to live by it. Footnote 30 Since this kind of examination of our beliefs and judgments is the only way to reveal fundamental principles, the fact that socialist societies tend to be tyrannical and poverty stricken does not show that the principle of socialist equality is wrong. It merely shows that we ought not to implement it, at least until such time as we have improved our nature and solved the calculation problem — that is, learned how to run an economy without the price system. Footnote 31 We “do not decide what to believe, whether about fact or about value and principle, in the light of what we expect the effect of believing it to be.” Footnote 32 We decide what to believe by discovering what is true or justified.
Some effects are, indeed, irrelevant to truth or justification. For example, creationism cannot be true just because it makes large numbers of people feel more secure in their religious beliefs, nor evolution false because believing it makes them feel insecure. But a medical principle that, when acted upon, regularly leads to the death of patients is just plain false. This effect is central to its truth and justification. No one would say that the fault is in the patients rather than in the medical principle. Footnote 33 And since principles of justice have practical significance, they are more like medical principles than like principles of arithmetic. Hence, a moral principle that, if implemented, would regularly lead to mutual distrust and unhappiness in even the most favorable circumstances, is also, it seems reasonable to say, false.
Moreover, contra Cohen, it seems unlikely that when we examine our particular judgments of situations and relationships to discover the right principles, Footnote 34 our judgment of what is right can be independent of all considerations of what humans are like, and what the effects would be of instituting certain principles. This point can be illustrated by a discussion of Cohen’s argument for socialist justice — specifically, for the luck egalitarian principle of justice — in Why Not Socialism? Before I start, however, it is important to note two things: (i) Although Cohen believes that socialist justice would be true justice even if we were not good enough for it, or even if we could not figure out how to harness our innate generosity to the task of production in a complex economy, he also states in Why Not Socialism? that basic human nature — that is, human nature minus capitalist corruption — is good enough for socialist justice; Footnote 35 and (ii) Cohen is a pluralist, holding that there are many fundamental moral principles, including many different fundamental principles of justice, with luck egalitarianism being the correct principle of distributive justice. Footnote 36
Cohen’s task in Why Not Socialism? is to get us non-socialists to agree that socialist principles and behavior on such principles are more attractive than capitalist principles, all without appealing to human capacities. Using the analogy of a camping trip, Cohen seeks to convince the reader that just as a camping trip in which everything is freely shared and people relate as friends or family is better than one that uses market principles, so a society in which people have equal material prospects and a “reciprocating attitude” toward each other is better than one in which they are unequal, and in which market participants are motivated by greed and fear and typically treat each other merely as a source of profit. Footnote 37,Footnote 38 A socialist camping trip is better than a capitalist one both “on grounds of efficiency” and “on grounds of fellowship,” because people try to make sure that everyone has a more-or-less equal chance to flourish, so long as everyone contributes to the flourishing of others. Footnote 39
If the way people typically behave on camping trips exemplifies socialist justice, as Cohen thinks, then socialist justice is both undemanding and attractive — at least on camping trips. Cohen is right that socialist camping trips as described so far beat capitalist ones hands down. Socialist distributive justice, according to Cohen, requires a rough equality of material prospects — rough, because inequalities that stem from people’s preferences across income and leisure, rather than from bad luck in social circumstance or inborn nature, are acceptable. Footnote 40 In other words, the correct fundamental principle of socialist distributive justice is luck egalitarianism. Footnote 41
However, the impression that camping-trip socialist justice is undemanding and attractive turns out to be an illusion when Cohen gives examples of the inequalities that have to be made up, for these go far beyond material inequalities. He argues that if some people on a camping trip just naturally enjoy nature more than others, they must equalize the situation by doing more than their share of unpleasant tasks. Socialist distributive justice dictates a radical equality — an equality that involves redistribution not only of material goods but even psychological ones. Footnote 42
Now Cohen’s appeal to the way people typically behave on camping trips no longer serves his purpose, for people decidedly do not typically behave the way he says they do. On actual camping trips, people who enjoy nature more than others do not think they must “make up” for this “unfairness” by, for example, doing more than their equal share of cleaning up. Footnote 43 Instead, they might think that those who do not enjoy nature very much should cultivate their capacity to enjoy it more, or cultivate an enjoyment of other things, such as cleaning up, instead of dragging other people down to their own level of enjoyment. This dragging down is what most people would regard as unjust, not only those who are dragged down but even, I believe, most of the people for whose sake they are supposed to be dragged down. For most people do take responsibility for their own well-being and are too proud to impose a burden on those whose only fault is the ability to enjoy themselves more. Footnote 44 At the same time, those who do not mind imposing such a burden on others — those who buy into this kind of egalitarianism — are hard to see as just or benevolent. “It’s only fair that you should have to do the dirty work since you enjoyed the trip more than I did” seems like a pretty ugly attitude to have. A socialist camping trip that requires a redistribution of enjoyment — and requires it as a matter of justice — is neither actual nor, in most people’s eyes, ideal.
For most of us, then, an examination of our judgments about an ideal camping trip leads to the conclusion that if, contrary to fact, people demanded or expected equal psychological goods, camping trips would be terrible. Hence, luck egalitarian socialism would be a failure even on camping trips. This evaluation appeals to what we think people are like and how they would feel about Cohen’s conception of socialist justice. Insofar as Cohen himself finds his imagined camping trip attractive, he must assume that benefactors on camping trips do not resent being burdened with unequal obligations toward others on account of the latter’s deficiencies, and beneficiaries do not become grasping and demanding. Thus, Cohen’s own examination of his judgments implicitly appeals to what he thinks people are like and how they (would) feel about socialist justice on camping trips. How could it be otherwise? It is not possible to bracket all of our beliefs about human psychology when we examine our intuitions about a supposedly ideal relationship in order to arrive at the fundamental principles governing it.
If I am right that Cohen’s conception of luck egalitarian socialist justice would strike most people as absurd on a camping trip, then it will surely strike them as even more absurd in a large political-economic system. For this conception implies that it is unjust if it turns out that people’s preferences do not lead to “a comparable aggregate enjoyment of life” where the preferences themselves are the result of genetic or social luck. Footnote 45 If they don’t, are other, luckier people then to be prevented from acting on their preferences in order to equalize the resulting enjoyment? Cohen does not say. The implications are equally absurd if we apply luck egalitarianism to other sorts of relationships, for example, that of two students. If one student just naturally enjoys learning more than the other student, should she have to do more of something that she does not enjoy in order to equalize the distribution of enjoyment? In schools and universities, should good grades and good publications be distributed equally, insofar as they are the result of innate intelligence or fortunate circumstances rather than effort or choice? And what about victories and honors in sports? Unless Cohen can show that such redistribution would be incompatible with other fundamental requirements of justice or of some other part of morality (as a “radical pluralist,” Cohen believes that moral principles can conflict), Footnote 46 the logic of his position commits him to a redistribution of all these goods in all fields — enjoyment, grades, publications, victories, and honors. In any case, it is easy to see that most people would find the very idea that this is justice absurd and the system it proposes unjust. For instead of treating everyone equally, it yokes those who are naturally happier or more talented to the less happy or less talented simply in order to bring the former down to the level of the latter and create an equal outcome. Only those who deeply resent the happier and more talented would welcome this view of justice. Luck egalitarianism would not fare much better if it trimmed its demands and required only equality of material prospects.These judgments, again, rely on intuitions about what people are like and what the effect would be of implementing Cohen’s conception of justice.
Of course, Cohen would argue that, regardless of what we would say or do as presently constituted, it would be better if we were the kind of creatures who were willing to keep producing unstintingly even if most of our earnings were taken away and redistributed and if we were happy to equalize unhappiness. A world in which human beings had such abundant generosity would be the ideal world. But there are other imaginary ideal worlds that, in my view and many other people’s views, are as attractive to us as Cohen’s is to him. In my ideal world, it would be better if we were the kind of creatures who happily produced as much as we wanted, did not resent anyone for having more or being capable of producing more, did not feel contempt for anyone who was not as smart or productive as we were, respected each other’s rights, took responsibility for our lives, helped each other when necessary, settled disputes amicably, and did all this without an authority figure with the power to tell us what to do. Footnote 47 This world seems to me to be far superior to Cohen’s socialism or Joseph Carens’s market socialism (of which more below). And no doubt there are other people whose conception of ideal human beings and systems is something else again.
Given these differences between different people’s convictions about distributive justice, no amount of examination of “our convictions” alone (as Cohen labels his own convictions) can settle the matter of what justice requires. We also need to expose our convictions and implicit views of human nature to the facts about our capacities, about the requirements of autonomy and self-respect, about the sources of happiness, and about the likely effects of implementing given principles of justice in a society in which people have a strong sense of justice. And for this we need both careful everyday observation, and solid social science. Perhaps if we examine our convictions in light of the facts of human nature, we will see that because people are so different, many of the ideals that appeal to different people are equally worthwhile, or at least permissible, so long as they do not use coercion or fraud to attract adherents or to keep them in. Then the right thing to do will be to leave people free to live as they like: as capitalists, or market socialists, or believers in some god, or even as full-blown socialists in their own communities. Footnote 48 We will then have utopia à la Robert Nozick’s in Anarchy, State, and Utopia. Footnote 49 Nozick envisages the ideal world as a framework for utopias, a framework that allows socialist, capitalist, and other communities to flourish within its boundaries, so long as no one violates anyone’s rights. (Of course, only a liberty-loving, free market capitalism allows such communities to flourish within its borders, so in a real sense capitalism conceived thus still wins on grounds of both desirability and feasibility.)
To summarize this section: Cohen fails to show that fundamental principles of justice must be fact-insensitive, or that our human capacities and the likely effects of implementing these principles are irrelevant to the truth of such principles, or that we can examine our judgments on particular issues in order to uncover our convictions about justice without implicitly consulting our view of human nature. I argue for the importance of the human nature constraint further in the course of discussing Estlund’s arguments.
III. The Human Nature Constraint: Estlund’s Critique
Estlund does not purport to know what justice requires, being concerned to show only that, whatever the truth be about justice, justice cannot be constrained by our natural motivational capacities or incapacities. Estlund is willing to concede “for the sake of argument” that if a society or the humans that make it up simply cannot meet the standards of a particular theory of justice, that theory is false. Footnote 50 So he would agree “for the sake of argument” that if we cannot banish suffering with a wand because we don’t have Superpeople powers, any theory that requires that we do is false. But widespread — or even universal — noncompliance does not show lack of ability to do as justice requires. For the noncompliance could be due to humans’ inability to “muster the will to do . . . what can soundly be prescribed or morally required,” and their “abilities and inabilities to muster their will are subject to moral evaluation in their own right.” Footnote 51 Estlund rejects the “human nature constraint” according to which “[a] normative political theory is defective and thus false if it imposes standards or requirements that ignore human nature — that is, requirements that will not, owing to human nature and the motivational incapacities it entails, ever be satisfied.” Footnote 52 For example, if “it is in the nature of humans to be more selfish and partial than socialist or egalitarian (or whatever other) theory would need them to be,” this does not refute such normative political theories. Again, if justice requires that people sacrifice their children to the state, then that is what justice requires, even if they cannot bring themselves to do it. Footnote 53 For its not as though they literally cannot do it. The right “oppressive threats and incentives” can bring about the right behavior. Footnote 54
Like Cohen, Estlund concedes that it would be useless to try to build institutions on the basis of a theory that imposes standards or requirements that ignore these facts of human nature, but the theory itself would not be false or unrealistic in any objectionable sense. If implementing a principle of justice would invite widespread noncompliance, then we ought not to implement it, but it can still be sound. Estlund explains: if people are not able to will to act justly according to the right principle of justice, then although it remains true that it ought to be implemented and obeyed, it is also true that if it will not be obeyed, it ought not to be implemented. “Ought” does not distribute over conjuncts. Footnote 55
Estlund’s first argument in support of his view that widespread motivational inability to comply with a theory or principle does not make it false starts with the example of Bill. Bill cannot bring himself to refrain from dumping his trash on the street instead of putting it in the trash can — he is too weak-willed or selfish. Footnote 56 Clearly this motivational inability does not block the requirement that he stop dumping trash on the street. His motivational inability makes it impossible for him to will to do the right thing, but it does not make doing the right thing impossible. This verdict does not change if no one can bring himself to will to refrain from dumping his trash on the street, even if this motivational inability is part of human nature, so long as he can, in fact, act otherwise. The same goes for a man who cannot stand being at a business meeting with a woman. Footnote 57 Even if all men had this motivational inability by nature, it would still be wrong to act on it, and they would still be required to act differently. Even if these traits and behaviors were excusable, it would not be simply because they were part of human nature. Footnote 58
Estlund’s examples of the weak-willed or selfish Bill and the male bigot are persuasive because we already know that these traits are bad and these behaviors wrong; that these traits are acquired traits, part of Bill’s and the bigot’s character, not innate; and that people have some responsibility for their character and behavior and some ability to change them or at least to behave differently. Footnote 59 We all have the capacity to become weak-willed or indifferent to others’ rights and needs or bigoted — this capacity is certainly part of our nature as human beings — but the capacity awaits development into a moral trait by our upbringing and our own choices. Bill’s inability to will himself to do the right thing does not block the requirement to do the right thing, or excuse his wrong action, because he is capable of doing the right thing and avoiding the wrong. Footnote 60 The same goes for the bigot.
A better candidate for a trait that is both bad and part of human nature is cruelty. Estlund points out that a certain amount of cruelty is a widespread element of human nature, but that the conclusion that we should not require people to not be cruel does not follow. Footnote 61 An argument to the contrary is “absurd.” Some motivations are “justice-tainting rather than justice-shaping.” Footnote 62 Hence, a mere appeal to human motivations cannot constrain the truth about justice.
Estlund is right that a mere appeal to innate human motivations — which I will now call “propensities” in order to keep them distinct from acquired motivations, that is, motivations that are part of our acquired moral traits — cannot justify or refute theories of justice or, more generally, theories of ethics in general, because some of them are justice-tainting. To build on his point, consider the propensity to control others, or to scapegoat them for our mistakes or misfortunes, or to be pleased by their misfortunes because they make our own lives seem so much better by comparison (schadenfreude). One or more of these propensities seem to be present in nearly everyone, and entirely — or on balance — for the worse. Perhaps these propensities served our survival needs at one time, but they no longer do. Nor are they sources of enduring happiness or of peaceful and trusting relations with others. Clearly, then, they cannot be a constraint on the right conception of justice or morality just because they are innate.
Another reason that a mere appeal to our propensities is not enough to justify or refute a normative theory is that most of them are neither justice-tainting nor justice-shaping, neither inherently good nor inherently bad. Most of them are morally neutral. Even the in-group bias and the egoistic (partiality-to-self) bias, both of which are often decried because they are so often among the building blocks of bigotry or indifference to others’ claims, are at the same time necessary for survival as well as for relations of love and friendship. Many morally neutral propensities come in opposing pairs. If there is a propensity to cause others pain, there is also a propensity to help them and to want their happiness; if there is a propensity for stating facts, there is also a propensity for making up stories; if there is a propensity for acting fearlessly, there is also a propensity for acting fearfully; if there is a propensity to obey authority, there is also a propensity to defy authority, and so on. In themselves, these propensities are neither good nor bad.
It is because we have some bad propensities and some morally neutral ones that no one subscribes to the sweeping thesis that Estlund rejects, namely, that a mere appeal to human propensities can constrain the truth about justice. What philosophers who appeal to human nature hold is that a sound theory of justice or other moral principles cannot be constructed without taking fundamental human propensities into account, propensities that play a central role in most people’s well-being, such as love of our children, attachment to our property, and partiality to ourselves and those we love. Footnote 63 An appeal to human nature in a theory of justice or morality in general has to tell us which traits to cultivate out of our propensities in order to have good — that is, desirable and worthwhile — lives in a good human society, a society marked by mutual trust, cooperation, and widespread prosperity. Since the answer is an answer for everyone, what counts as a good life must be a life that each person with the capacity for it can, in principle, try to achieve compatibly with other persons’ attempts to achieve it. Hence a sound principle of justice cannot require people to act in ways that are contrary to living good lives or sustaining a good society, such as requiring us to give up our children or our property to the state, or being completely impartial between those we love and strangers. Footnote 64
If this is right, then Estlund’s claim that it could be just for Plato’s communism to require parents to give up their children to be brought up by the state, Footnote 65 or for Joseph Carens’s market utopia to require people to maximize production even though their earnings will be distributed equally to everyone, regardless of people’s propensities, and regardless of the materially counterproductive effects of such requirements, begs the question of what makes these requirements just. Footnote 66 If the well-being of individuals and society at large are irrelevant to the truth about justice, what is relevant? Calling a theory that thus goes against “the grain of human nature,” and that, if implemented, would lead to widespread unhappiness, poverty, and mutual distrust a theory of justice, sounds like an oxymoron.
These requirements are wrong for yet another reason: in assuming that some people have the right to force other people to live in certain ways and not others, they assume that these people are endowed with a special authority, an authority based in the wisdom to make and enforce just laws, to manage the economy well, and even to raise other people’s children well. This belief in the superior global wisdom of some might have been believable in Plato’s time, but cannot withstand scrutiny in ours. And the reason why it can’t has everything to do with our inborn cognitive and emotional nature. Footnote 67
The importance of taking human propensities into account in a theory of justice can be supported further by considering how our views of justice would change if our psychologies were radically different. Suppose, for example, that giving up our children to state bureaucrats were better for us and our children because trying to raise them ourselves predictably led to rotten children and miserable parents. Then most of us would be motivated to give them up and we would think that doing so was both just and rationally self-interested. Or suppose that living in a system in which the total product was redistributed equally turned out to be more personally fulfilling and socially beneficial than living in a system in which we could keep (the bulk of) what we produced. Then most of us would be motivated to do it and bring about either full-scale socialism or Joseph Carens’s market utopia.
Given our nature, however, all the evidence points in the opposite direction, from the small private plots in the Soviet Union that produced more food, proportionately, than the huge collective farms, Footnote 68 to the failure of all large socialist systems, and even most small ones. Footnote 69 The lesson of the seventeenth-century Plymouth Colony is especially instructive. The residents worked the land communally (according to ability), and shared the harvest equally (according to need). Soon, however, as Governor William Bradford recorded in his memoirs, people’s needs on the one hand, and their inabilities on the other, each multiplied exponentially, with the result that soon they had a famine on their hands. This, followed by a plague, led to the death of half of the members of the colony. Footnote 70 Equally important, everyone — the young, the old, wives, husbands — complained that the system was unjust to them. The experiment ended when the Governor gave every family private plots of land and the colony saw abundant times again. It is hard to see what can support the view that these examples only show that the problem is human nature and not the theory of justice in question. What justifies calling it justice?
The objections to these conceptions of justice do not apply to most commonsensical requirements to, for example, respect other people’s rights, or not be cruel. If cruelty is understood as causing someone pain for no good reason, or taking pleasure in his pain, then requiring people not to be cruel does not violate central human needs or lead to unhappiness, resentment, or poverty. Indeed, cruelty causes unhappiness to the victim without doing much for the perpetrator, and a society of cruel people would lack mutual trust and, therefore, the ability to cooperate to mutual gain. By contrast, most people do recognize, if only implicitly, that respect for each other’s rights and benevolence toward each other are crucial to a good society — that is, a society of mutual trust, cooperation, and widespread prosperity — and that a good society is likewise crucial to a desirable and worthwhile life.
Some of Estlund’s statements suggest that he would agree that such considerations are relevant to sound principles of justice. For example, he says that “it is never a feature’s status as characteristic of humans by nature that constrains the concept of justice, since a judgment is also required in every case about whether the feature’s moral value or significance suits it to have this kind of weight.” Footnote 71 He even allows that it might be permissible for parents to refuse to give up their children to the state if “a deep moral case” can be made for permitting it. This implies that he would regard the considerations I’ve adduced above about the connection of certain demands and certain traits to mutual distrust and unhappiness as relevant to sound principles of justice, because happiness and mutual trust are morally significant issues. But he never mentions happiness or unhappiness, or mutual trust or, for that matter, widespread prosperity, or love and friendship, as moral considerations. And many of his statements, and the general thrust of his argument, show that, by the logic of his position, he cannot regard these issues as morally significant. For he holds that the content of justice “might . . . precede any facts about what humans or any beings happen to be like, including whether limited altruism is part of their nature.” Footnote 72 Like Cohen, Estlund holds that principles of justice are (or “might” be) a priori. If the circumstances of justice, such as limited altruism, did not obtain, then “justice would fail to apply” to us, but the mere fact that we cannot live up to a certain conception of justice does not show that it doesn’t apply to us.
Unfortunately, as we’ve seen, on this view justice can require us to act in ways that are contrary to the conditions of a good life and good society. And this begs the question of what makes it justice. The truth about justice, after all, is quite different from the truth of, say, the origins of the universe, or some abstruse mathematical theory. Footnote 73 A theory of the origins of the universe can be true regardless of its effects on our lives, if any. Similarly, a mathematical theory can be true even if it fails to promote our good either directly or indirectly (say, through helping to solve a scientific problem that promotes our good). But the whole point of justice is to enable us to live good lives in a good society. For whether it’s the case that justice must, first and foremost, protect liberty, or secure equality, or promote happiness, it is surely the case that a just society must allow people to lead desirable and worthwhile lives through mutual trust, cooperation, and widespread prosperity. So those who think that a conception of justice can be true even if its implementation would cause widespread unhappiness, resentment, and mutual distrust, regardless of the circumstances, need to say what makes it true. And they cannot do this simply by consulting their own intuitions or convictions (Cohen). Nor can they simply ignore the question by speculating that the content of justice might even “precede any facts about what humans . . . happen to be like” (Estlund). Footnote 74
There is another implication of this rejection of the human nature constraint and endorsement (in principle) of theories that most of us would find repugnant that should trouble even Cohen and Estlund, namely, that their theories leave them unable to reject conceptions of justice that even they would find repugnant. Consider, for example, the view that justice requires naturally inferior people to be the slaves of naturally superior people. Or that it requires women to be the handmaidens of men because women lack the capacity for autonomy. Both of these views have been held by learned, thoughtful people, and the one about women is held even now by hundreds of millions of people around the world. To be sure, the proponents of these views claim to be basing their conceptions of justice on the different natures of these various groups: this suggests caution about appeals to human nature. But the problem is not their appeal to these groups’ natures, but their wrong conceptions of these natures. So it is not surprising that we can reject their claims as unjust only by pointing out (i) that their justification in terms of this group’s inferior nature and that group’s superior nature is based on false premises — premises that are themselves the product of ignorance, prejudice, or self-deception; (ii) that, in fact, all these groups share the same human nature; and (iii) that these conceptions lead to lives of severe deprivation and lack of dignity for the oppressed, and feed the self-delusion of the oppressors. Footnote 75 As far as I can see, there are no other reasons for rejecting such repugnant conceptions of justice — or for finding them repugnant. But if Estlund and Cohen are right to hold that a theory of justice can be true independently of our psychologies or of the effects of implementing it even in favorable circumstances, then they must admit that these theories also could be true.
To summarize the discussion so far: Estlund is right that not all innate propensities are a constraint on the right conception of justice. But this does not entitle him to conclude that no innate propensities are such a constraint. Some propensities, such as love of our children, attachment to what we have produced, and partiality to self and those we care about are central to desirable and worthwhile lives and to a good society, whereas some, such as the propensity to control others, or to scapegoat them, are conducive neither to a good society nor to worthwhile lives. Widespread noncompliance with a theory of justice by people who have a sense of justice and who live under circumstances favorable to justice shows a defect in the theory, not in the people, because only a theory that violates the conditions of a good life and good society can fail so drastically. Such a theory is unsound because it is unrealistic in its conception of justice and its expectations. The same is true of some principles of personal morality, such as Kant’s principle of honesty. This principle allows no exceptions to telling the truth, not even to prevent a murderer from murdering your mother. Footnote 76 We can be pretty sure that no one accepts or lives by this Kantian principle, because doing so would be destructive of their happiness and their deepest values. Equally important, it would leave those who are conscientious, as judged by this principle, vulnerable to the wicked. Living by this principle would thus promote the bad at the expense of the good.
But what should we say about theories or principles that do not have any such defect, and still do not command full compliance by most people, or perhaps anyone? Estlund claims that any principle or theory that “implies criticism of actual institutions or behavior is not as realistic as it could be,” Footnote 77 hence there is no good reason to reject a principle or theory that is unrealistic in demanding something that will almost certainly not be met. Footnote 78 For example, the citizens of a democracy may not satisfy the requirement that they be “publicly and privately highly virtuous (but not more than people are capable of),” and that institutions be designed such that “laws are just, rights are protected, the vulnerable are cared for, minorities are embraced and respected, and so on.” But that does not mean that the theory itself is unsound. Footnote 79
In describing a theory that implies criticism of actual institutions or behavior as not being as realistic as it could be, however, Estlund identifies a realistic theory with one that blesses the status quo. But a theory that blesses the status quo, no matter how morally lacking that status quo, is not much of a normative theory. Every normative theory, whether moral or political, must provide standards for a society or individuals to live up to. Hence it must imply criticism of some institutions and behaviors. What this shows is not that no such theory is as realistic as it could be, but that no such theory can be entirely descriptive. It is realistic and sound if it requires people to be and do what they are able to be and do compatibly with the requirements of a good life and good society.
It is true, as Estlund points out, that most (all?) people will violate even the correct principle of honesty or justice, whatever that might be, sometimes over a lifetime. Likewise, even on a sound conception of justice, most (all?) people will still sometimes fail to be just. This is true even if they accept what is in fact the correct principle. In these cases, it is people that are at fault, not the principles. For, by hypothesis, the principles do not violate the conditions of a good life or good society. What would be unrealistic is actually expecting any individual or society to be perfectly virtuous, building institutions on this assumption, and having no institutions or mechanisms for dealing with unjust laws or people. Footnote 80 The principles in question present us with ideals to aspire to, ideals that most people can and do live up to much of the time over their lifetimes, but not without fail. Why this is so is an important question for moral psychology, but not a question we need to answer here. Footnote 81 The important point for now is that our imperfections have important implications for how we should respond to occasional moral failures in ourselves and others, Footnote 82 and what kind of institutions we need to have to deal with failures of justice. No realistic moral or political principle or theory expects perfection in setting out an ideal to aspire to, whereas every unrealistic moral theory, personal or political, must expect that, in the absence of massive coercion and propaganda, everyone will fail to be even minimally decent by its standards. An unrealistic theory of justice must also expect that, if implemented, it will inevitably thwart our attempts to lead desirable and worthwhile lives in a good society. This is the real test of a sound theory of justice, and an unrealistic theory fails this test. Hence such a theory, to paraphrase David Schmidtz, is not even worth aspiring to. Footnote 83
IV. Conclusion
No sound theory of justice — more generally, no theory of personal or political morality — can dispense with the facts of human nature. If no one has the capacity to live by a principle of justice, or if no one can be motivated by it because it is contrary to the requirements of a good life and good society, the principle ceases to be normative for us. Implicit in the judgments we make about the nature of a good society are beliefs about human nature and the likely effects of certain requirements on individuals and society. Without the human nature constraint, there is no limit to the variety of repugnant conceptions of justice that can be dreamt up and justified.
