We are all enormously grateful to Christine Korsgaard for her characteristically lucid and insightful lecture on the basis of animal entitlements. Her paper will be appearing in the Handbook on Ethics and Animals, ed. T. Beauchamp, Oxford University Press, alongside a paper of mine in which I sympathize with many of her major conclusions and with her critique of Utilitarianism, but then reflect on subtle differences between her Kantian position and my own neo-Aristotelian position. I post here three sections of that forthcoming paper (whose title is "Animal Thinking and Animal Rights"). I hope this exchange will give readers a richer sense of the theoretical options in this area and also of the important distinction (important to both Korsgaard and me) between an ethical view and a political view:
III. Utilitarianism: Strengths and Problems
The philosophical school that has, until now, made the largest contribution to thinking about the ethical treatment of animals is classical Utilitarianism. Both Jeremy Bentham and John Stuart Mill were passionately interested in the lives of animals and both thought that human treatment of animals was ethically unacceptable. Bentham -- noting that Hinduism and Islam are ahead of Christianity in their recognition of ethical duties to animals -- famously predicted that a day would come when species difference would seem to all as ethically irrelevant, in the context of bad treatment, as race was by then beginning to be agreed to be:
The day may come when the rest of the animal creation may acquire those rights which never could have been withholden from them but by the hand of tyranny. The French have already discovered that the blackness of skin is no reason why a human being should be abandoned without redress to the caprice of a tormentor. It may come one day to be recognized that the number of the legs, the villosity of the skin, or the termination of the os sacrum, are reasons equally insufficient for abandoning a sensitive being to the same fate. What else is it that should trace the insuperable line?...The Question is not, Can they reason? Nor, Can they talk? But, Can they suffer?
Mill, noting that this passage, written in 1780, anticipates many valuable legal developments that make at least a beginning of protecting animals from cruelty, responds in similar terms to Whewell's dismissive statements concerning duties to animals. Whewell argues that it is "not a tolerable doctrine" that we would sacrifice human pleasure to produce pleasure for "cats, dogs, and hogs." Mill responds: "It is 'to most persons' in the Slave States of America not a tolerable doctrine that we may sacrifice any portion of the happiness of white men for the sake of a greater amount of happiness to black men." He adds a comparison to feudalism. At his death, Mill left a considerable portion of his estate to the Society for the Prevention of Cruelty to Animals.
Both Bentham and Mill felt not only that large conclusions for our treatment of animals followed from their Utilitarian principles, but also that the ability of those principles to generate acceptable conclusions in this area was a point in favor of those principles – by contrast, for example, with the principles of vulgar Christianity (represented in Whewell's hostile reaction to Bentham), which made species difference all-important. For both Bentham and Mill, Utilitarianism, with its commitment to treat all sufferings and pleasures of all sentient beings on a par, had made decisive progress beyond popular ethics in just the way that abolitionist views were then making progress beyond popular racist views. Seeing how the view enabled one to cut through unargued prejudice, and to treat subordinated beings with due concern, one saw a strong reason,they thought, in the view's favor.
There is no doubt that Utilitarian thought has made valuable and courageous contributions in this area, and that it still does so today, in the work of preference-Utilitarian Peter Singer, one of the leading voices against cruelty to animals. I now want to argue, however, that Utilitarianism cannot meet the challenge of animal complexity, as we currently understand it. (It has related problems with human complexity).
Utilitarianism can be usefully analyzed, as Bernard Williams and Amartya Sen have analyzed it, as having three parts. The first is consequentialism: the best choice is defined as the one that promoted the best overall consequences. The second is “sum-ranking,” a principle of aggregation: we get the account of consequences by adding up all the utilities of all the creatures involved. Third, the theory invokes some specific theory of the good: pleasure in the case of Bentham and Mill, the satisfaction of preferences in the case of Peter Singer. Looking at animals, Utilitarians begin from the understanding that they, like human beings, feel pleasure and pain, and they argue that the calculus of overall pleasure cannot consistently exclude them. The right choices will be those that produce the largest aggregate balance of pleasure over pain – or, in Singer’s case, the largest net balance of satisfaction over dissatisfaction.
The Utilitarian approach has the merit of focusing attention on something of great ethical importance: suffering. Humans cause animals tremendous suffering, and much of it is not necessary for any urgent human purpose. Animals would suffer a great deal without human intervention, but there is no doubt that much of animal suffering in today’s world is caused, directly or indirectly, by human activity. So the focus on animal suffering is valuable, and these philosophers deserve respect for the courage with which they put this issue on the agenda of their nations.
Five problems, however, can be seen, if we hold this theory up against the complex cognitive and social lives of animals. The first point is that pleasure and pain, the touchstones of Utilitarianism, are actually disputed concepts. Bentham simply assumes that pleasure is a single homogeneous type of sensation, varying only in intensity and duration. But is he correct? Is the pleasure of drinking orange juice, for example, the same sort of sensation as the pleasure of listening to a Mahler symphony? Philosophers working on this question, from Greek antiquity to the present day, have, on the whole, denied this, insisting that pleasures vary in quality, not just quantity. Moreover, Mill himself insists on this point in Utilitarianism. A second point on which Mill insists – along with many other philosophers, past and present -- is that pleasure is a type of awareness very closely linked to activity, so that it may be impossible to separate it conceptually from the activities that are involved in it. We do not need to resolve all these issues in order to realize that they arise in animal lives as well as human lives, once these lives are seen with sufficient complexity. Happy's pleasure seeing herself in the mirror seems unlikely to be the very same sensation as her pleasure when she eats some nice bananas, or hugs her small baby elephant with her trunk.
Pleasures, second, are actually not the only things relevant to animal lives. These lives consist of complex forms of activity, and many of the valuable things in those lives are not forms of pleasure. Happy’s self-recognition in the mirror, the mourning of elephants for their dead, are not pleasures; the latter may even be deeply painful. Nonetheless, such meaningful elements in animal lives should, we intuitively feel, be fostered and not eclipsed – for example, eclipsed by raising animals in isolation so that they don’t have contact with fellow group members and so are unable to mourn. Animals want much more than pleasure and the absence of pain: free movement, social interactions of many types, the ability to grieve or love. By leaving out all this, Utilitarianism gives us a weak, dangerously incomplete way of assessing our ethical choices.
Third, animals, like human beings, can adjust to what they know: they can exhibit what economists call “adaptive preferences”. Women who are brought up to think that a good woman does not get very much education may not feel deprived if they don’t get an education, so Utilitarian theory would conclude that education is not valuable for them. This means that the theory is often the ally of an unjust set of background conditions. Much the same sort of thing can be said about animal preferences. If animals are given a very confined life, without any access to social networks characteristic of their species, they may not actually feel pain at the absence of that which they haven’t experienced, but this does not mean that there is not an absence or that it should not be taken seriously. By refusing to recognize value where there is not pleasure or pain, Utilitarianism has a hard time criticizing bad ways of treating animals that have so skewed their possibilities that they don’t even hope for the alternative.
Fourth, a familiar point in criticism of Utilitarian theories of human life, Utilitarianism’s way of aggregating consequences doesn’t treat each individual life as an end; it allows some lives to be used as mere means for the ends of others. If it should turn out that the pleasures of humans who exploit animals for their use are great and numerous, this might possibly justify giving at least some animals very miserable lives.
Finally, all Utilitarian views are highly vulnerable in respect of the numbers. If the goal is to produce the largest total pleasure or satisfaction, then it will be justified, in the terms of the theory, to bring into existence large numbers of animals whose lives are extremely miserable, and way below what would be a rich life for an animal of that sort, just so long as the life is barely above the level of being not worth living at all.
IV. Two Strong Theoretical Alternatives
Seeing these problems helps us think: it informs us, I believe, that we need a theoretical approach in ethics that can do two things. First, the approach must have what I would call a Kantian element: that is, it must have as a fundamental ethical starting point a view that we must respect each individual sentient being as an end in itself, not a mere means to the ends of others. (I’m simply helping myself to an extension of Kant’s approach to human beings at this point, not offering any story about how one might use Kant’s own actual views to generate obligations to animals.) Second, the approach must have what I would call a neo-Aristotelian element, the ability to recognize and accommodate a wide range of different forms of life with their complicated activities and strivings after flourishing. I’ve suggested in writing about this that for this part of the view we can turn to a version of the Aristotelian idea that each creature has a characteristic set of capabilities, or capacities for functioning, distinctive of that species, and that those more rudimentary capacities need support from the material and social environment if the animal is to flourish in its characteristic way. But of course that observation only goes somewhere in ethics if we combine it with the Kantian part, the idea that we owe respect to each sentient creature considered as an end. Putting these two parts together, we should find a way to argue that what we owe to each animal, what treating an animal as an end would require, is, first, not to obstruct the creature's attempt to flourish by violence or cruelty, and, second, to support animal efforts to flourish in positive ways (an analogue of Kantian duties of beneficence.)
In the case of humans, as Kant and Aristotle would agree, our beneficence is rightly constrained by concerns about autonomy and paternalism: rather than pushing people into what we take to be a flourishing life, we ought to support, instead, ample space for choice and self-determination. In the case of animals, by contrast, although we should always be sensitive to considerations of choice, to the extent that we believe an animal capable of choice among alternatives, for the most part we must and should exercise informed paternalistic judgments concerning the good of the creature, and our duties of beneficence will be correspondingly more comprehensive, as they are in the case of human children.
There two recent ethical approaches that both contain these two elements. One is Christine Korsgaard's Kantian view, developed in her recent Tanner Lectures, "Fellow Creatures: Kantian Ethics and Our Duties to Animals." Another is the extended version of the neo-Aristotelian capabilities approach that I have described in Frontiers of Justice.
Kant's own views on animals are not very promising. He holds that only humans are capable of moral rationality and autonomous choice, and that only beings who are capable of autonomy can be ends in themselves. Animals, then, are available to be used as means to human ends. Kant thinks that we do have some duties with regard to animals, but these, on closer inspection, turn out to be indirect duties to human beings. In particular, Kant holds that treating animals cruelly forms habits of cruelty that humans will then very likely exercise toward other human beings. This, rather than any reason having to do with respect for animals themselves, is his reason for imposing a range of restrictions on the human use of animals.
Korsgaard's view is subtle and difficult to summarize, but let me try to state its essential insight. For Korsgaard as for Kant, we humans are the only creatures who can be obligated and have duties, on account of our possession of the capacity for ethical reflection and choice. Korsgaard, however, sees that this fact does not imply that we are the only creatures who can be the objects of duties, creatures to whom duties are owed. She also puts this point another way. There are, she argues, two different senses in which a being can be an "end in itself": (a) by being a source of legitimate normative claims, or (b) by being a creature who can give the force of law to its claims. Kant assumes that these two ways in which something can be an end in itself pick out the same class of beings, namely all and only human beings. Korsgaard points out that a being may be an end in itself in the first sense while lacking the capacity for ethical legislation crucial for the second sense.
Korsgaard's conception of animal nature is Aristotelian: she sees animals, including the animal nature of human beings, as self-maintaining systems who pursue a good and who matter to themselves. She gives a fine account of the way in which we may see animals as in that sense intelligent – as having a sense of self and a picture of their own good, and thus as having interests whose fulfillment matters to them. We human beings are like that too, she argues, and if we are honest we will see that our lives are in that sense not different from other animal lives.
Now when a human being legislates, she does so, according to Kant and Korsgaard, in virtue of a moral capacity that no other animal has. This does not mean, however, that all human legislation is for and about the autonomous will. Much of ethics has to do with the interests and pursuits characteristic of our animal nature. When we do make laws for ourselves with regard the (legitimate) fulfillment of our needs, desires, and other projects issuing from our animal nature, it is simply inconsistent, and bad faith, Korsgaard argues, to fail to include within the domain of these laws the other beings who are similar to us in these respects. Just as a maxim cannot pass Kant's test if it singles out a group of humans, or a single human, for special treatment and omits other humans similarly situated, so too it cannot truly pass Kant's test if it cuts the animal part of human life from the animal lives of our fellow creatures.
I have saved until last a part of Korsgaard's conception that lies at its very heart. We humans are the creators of value. Value does not exist in the world to be discovered or seen, it comes into being through the work of our autonomous wills. Our ends are not good in themselves; they are good only relatively to our own interests. We take our interest in something "to confer a kind of value upon it," making it worthy of choice. That, in turn, means that we are according a kind of value to ourselves, including not only our rational nature but also our animal nature. Animals matter because of their kinship to (the animal nature of) a creature who matters, and that creature matters because it has conferred value on itself.
Korsgaard's conception of duties to animals has what I demanded: a Kantian part and an Aristotelian part. It says that we should treat animals as ends in themselves, beings whose ends matter in themselves, not just as instruments of human ends. And it also conceives of animal lives as rich self-maintaining systems involving complex varieties of intelligence. So far so good. Now I shall describe the way in which my own conception articulates the relationship between the Kantian and the Aristotelian. Then, more tentatively, we can ask what reasons there may be in favor of choosing one rather than the other.
Because my view has been advanced as a political doctrine rather than a comprehensive ethical doctrine, I have not developed the view's metaphysical/epistemological side. In keeping with my espousal of a Rawlsian "political liberalism," I have expressed the relevant idea of intrinsic value in a nonmetaphysical and intuitive way. However, were I to flesh out the view as a comprehensive ethical view, I would insist that the lives of animals have intrinsic value. This value is independent of human choice and legislation, and it is there to be seen. If humans had never come into being, the lives of other animals would still be valuable. We humans are, fortunately, attuned to value, so we are capable of seeing what Aristotle's students saw, that there is something wonderful and awe-inspiring in the orderly systems characteristic of natural end-pursuing creatures. To this sense of awe, I suggested in Frontiers that we must add an ethical sense of attunement to dignity. What is wonderful about an animal life is its active pursuit of ends, so our wonder and awe before such a life is quite different from our response to the Grand Canyon or the Pacific Ocean: it is a response to the worth or dignity of an active being who is striving to attain its good. Wonder and awe before the dignity of such a life would be inappropriately aestheticizing, would fail to recognize what, precisely, is wonderful about the creature, if it simply said "Ooh!" "Aah!" and saw no implications for the ethics of animal treatment. If we have appropriate wonder before an animal life, wonder that homes in on what the creature actually is, a self-maintaining active being, pursuing a set of goals, then that appropriate wonder, I argue, entails an ethical concern that the functions of life not be impeded, that the life as a whole not be squashed and impoverished.
Let me put this point another way. When I have wonder at the Grand Canyon, it would seem that I have appropriate wonder, wonder that is appropriately trained on the relevant characteristics of the object, only if I form some concern for the maintenance of the beauty and majesty of that ecosystem; even here, then, wonder has practical consequences. If I say, "How wonderful the Grand Canyon is," and then throw around, I am involved in a contradiction: my actions show that at some level I really do not think that the Grand Canyon is very wonderful. With animals, all this is true, but also much more. Animals, because they are active sentient beings pursuing a system of goals, can be impeded in their pursuit by human interference. In Frontiers, I argued that it is this quality of active, striving agency that makes animals not only objects of wonder but also subjects of justice. The way we wonder at the complexity of animals, if it is appropriate, really trained on what they are, includes a recognition that they are active, striving beings, and thus subjects of justice. The right sort of wonder (not pretending that an animal is like a fine chair or carpet, but seeing it for what it is) leads in that sense directly to an ethically attuned awareness of its striving.
With that intuitive picture as my starting point, I then go on in Frontiers of Justice to argue that our ethically attuned awareness of the value of animal striving suggests that we ought to promote for all animals a life rich in opportunities for functioning and lacking many of the impediments that we humans typically put in the way of animals flourishing. Since my views on the content of our duties lie very close to Korsgaard's, I need not enumerate them here.
What might lead one to choose one of these views of our duties to animals rather than another? It is obvious that some people find realism about value implausible, and others find the idea that all value is a human creation implausible. The choice between the two views on this score must await the much fuller development of arguments for and against realism. Here Korsgaard has gone a lot further than I have, since I have deliberately avoided defending realism, given that I am trying to advance the capabilities view as a non-metaphysical political view. There would be a great deal of work for me to do were I to try to work out the view as an ethical doctrine comparable to Korsgaard's in its detail and completeness. The Aristotelian approach to value does involve a large measure of reliance on intuitions, as Korsgaard justly argues in The Sources of Normativity. It will not satisfy all people. By contrast, the Kantian account of normativity is intricate and philosophically rich; it does not seem to rest on such a fragile empirical foundation.
On the other hand, I think that the Aristotelian view has at least some advantages, albeit subtle and not decisive. Korsgaard does not exactly make the value of animals derivative from the value of human beings. Instead, her picture is that when we ascribe value to ourselves, we ascribe value to a species of a genus, and then it is bad faith, having once done that, to deny that the other species of that genus, insofar as they are similar, possess that same action-guiding value. However, there still seems to be a strange indirectness about the route to animal value. It is only because we have similar animal natures ourselves, and confer value on that nature, that we are also bound in consistency to confer value on animal lives. Had we had a very different nature, let's say that of an android, we would have no reason to value animal lives. And, so far as I can see, the rational beings recognized by Kant who are not animal (angels, God) have no reason to value the lives of animals. For me this is just too indirect: animals matter because of what they are, not because of kinship to ourselves. Even if there were no such kinship, they would still matter for what they are, and their striving would be worthy of support. For Korsgaard, it's in effect an accident that animals matter: we just happen to be pretty much like them. But I think that the value of animal lives ought to come from within those lives; even if one doesn't think of value as eternal and immutable, one still might grant that it comes in many varieties in the world, and each distinctive sort is valuable because of the sort it is, not because of its likeness to ourselves.
So, while I agree with Korsgaard that we are the only creatures who have duties, and while I think that she has argued in a way that puts the Kantian view in its best form, and, indeed, in a very attractive form, I still feel that there's something backhanded about the route to animal value, and that it would be good to acknowledge that this value is there whether or not these creatures resemble us. (Whether one could acknowledge that without relying on intuitions as the source of normativity is a further question that I shall not try to answer here; clearly I am less worried about reliance on intuition than Korsgaard is, or else I would not be willing to venture ahead at this point.)
There is another point of interesting difference, pertinent to our concern with animal thinking. Korsgaard, as I said, makes a very compelling case for recognizing in animals a range of types of awareness; even those who can't pass the mirror test are held to have a point of view on the world, and ends that matter to them. All this seems to me just right. So, while one might have expected that a Kantian view would draw a too-sharp line between the human and the animal, that seems not to be true of Korsgaard's view. In another way, though, I wonder whether there is not after all a bit too much line-drawing.
Korsgaard rightly says that we are the only truly moral animals, the only ones that have a full-fledged capacity to stand back from our ends, test them, and consider whether to adopt them. She does, however, say of children and people with mental disabilities that they too are rational beings in the ethical sense, it's just that they reason badly. If she once makes that move, I do not see how she can avoid extending at least a part of ethical rationality to animals. Animals, as we saw, are aware of their place in a social group. Many of them have the capacity for a type of reciprocity, and some, at least, seem to be capable of positional thinking, thus understanding the impact of their actions on others. At least some varieties of shame and even guilt figure in some of these animal lives, in ways related to their awareness of the rules that govern social interactions. So it seems that what animals most conspicuously lack is the capacity for universal ethical legislation, but that there's a part of ethical capacity that at least some of them have already. By splitting humans from all the others, as the only rational legislators, Korsgaard seems to have drawn a line that is not that sharp in reality.
Korsgaard will surely say at this point, as she does early in her lectures, that if the definition of rational being turns out to fit some nonhuman creatures, all very well and good, she is only focusing on our obligations to those whom it doesn't fit. But I'm not altogether happy with that reply. It seems that we need to understand our moral capacities as well as we can, making use of all the scientific information that is available. Information about animals is very helpful to us, in showing us how the capacities we have, which we might have thought transcendent and quasi-divine, are a further development of some natural capacities that we share with the animals. In understanding ourselves that way, we also attain a fuller understanding of how what Korsgaard calls our animal nature is related to our moral nature: our moral nature is actually one part of our animal nature, not something apart from it. Our moral nature is born, develops, ages, and so forth, just like the rest of our capacities, and, like them, it has an evolutionary origin in "lower" animal capacities. I feel that Korsgaard has pushed Kant to the limit in giving her extremely sensitive and appealing picture of how Kant and Aristotle may cooperate, but there is really no way, without departing from Kant rather radically, to acknowledge that our moral capacities are themselves animal capacities, part and parcel of an animal nature. I think that any conception that doesn't acknowledge this is in ethical peril, courting a danger of self-splitting and self-contempt (so often linked with contempt for women, for people with disabilities, for anything that reminds us too keenly of the animal side of ourselves). Although Korsgaard heads off this peril sagely wherever it manifests itself, she still doesn't altogether get rid of it; it is still lurking, in the very idea that we are somehow, in being moral, above the world of nature.
Those, then, are my reasons for tentatively preferring my own conception to Korsgaard's as a basis for the ethics of animal treatment. Nobody could doubt, however, that hers is considerably more finished than mine with respect to its metaphysical/epistemological side, which I've deliberately left uncultivated; nor should anyone doubt that her view provides a very good basis for thinking about our duties toward animals.
V. Political Principles: An Overlapping Consensus?
Now let us turn to political principles. Here, as I've said, agreeing with Rawls, we want to be abstemious, not making controversial metaphysical or epistemological claims. We are seeking an overlapping consensus among citizens who hold a wide range of reasonable comprehensive doctrines – including comprehensive Korsgaardianism and neo-Aristotelianism, but also including Buddhism, Hinduism, Christianity, and much else. So, we do not say that the human being was created on the seventh day of creation, or that humans will be reincarnated into animal bodies. By the same token, we do not say, with Kant, that human beings are the sole creators of value or, with Aristotle, that human beings discover a value that exists independently.
At this point, then, the major difference between Korsgaard's view and my imagined comprehensive neo-Aristotelianism has been bracketed. There are subtle differences that may remain, concerning the relationship between ethical rationality and other aspects of animals' good. It is difficult to say whether these differences really do remain: the idea that we are the creators of value goes very deep in Korsgaard's view, and colors every aspect of it, so it is very difficult to know exactly what her view would look like when recast in the form of a political doctrine appropriate to grounding a form of political liberalism in Rawls's sense. Certainly there would remain the idea that every sentient being has a good, consisting of a range of (non-commensurable) activities that are the activation of its major natural capacities, and that each animal is entitled to pursue that good. There would also remain (or so I believe) the sense that this good exacts something from human beings who are capable of choice: we have duties to protect and promote the good of animals. In these two respects, the imagined Korsgaardian view overlaps pretty completely with my neo-Aristotelian view, which borrowed the Kantian notion of dignity to ground ethical duties to the forms of life that wonder already singled out as salient. The emphasis on capacity and activity, the emphasis on a plurality of interrelated activities, the emphasis on ethical duty – all of this seems shared terrain between the two approaches.
In this case, then, we may not even need to talk, as Rawls did, of the overlapping consensus as consisting of a family of liberal political doctrines. We may be able to agree on a single political doctrine. If Korsgaard judges it important on balance to endorse a Kantian over a neo-Aristotelian political doctrine (Rawls's view, for example, over mine), the reasons for this difference would not come, I believe, from this particular area of the political doctrine, where Korsgaard has rightly seen the importance of invoking Aristotelian ideas.
The core idea of the political conception is the one I have already mentioned in talking about the ethical conception: that animals have characteristic forms of dignity that deserve respect and give rise to a variety of duties to preserve and protect animal opportunities for functioning. With this starting point, I then go on to envisage the general shape of a constitution for a minimally just multi-species world.
The political conception I have articulated seems like one that will be able to achieve an overlapping consensus among neo-Aristotelians and Korsgaardian Kantians. I conjecture that many other reasonable comprehensive doctrines will also support it: Buddhism, Hinduism, and, with time and persuasion, many varieties of Judaism, Christianity, and Islam.
What, however, of Utilitarianism? I have argued strenuously against accepting Utilitarianism as a comprehensive ethical doctrine concerning animal treatment, but I do not think that we should, without extremely strong reasons (such as have not yet been presented) conclude that Utilitarianism is not among the reasonable comprehensive doctrines that should be part of any political consensus. Since, however, the political principles I advocate are grounded in Kantian and Aristotelian ideas that are not as such part of Utilitarianism, it might seem that Utilitarians will have difficulty accepting that political conception. John Rawls argued that Utilitarians could form part of an overlapping consensus supporting his own political doctrine, and yet not all readers of that argument have been convinced by it. So we must ask ourselves: what reasons do we have to think that Utilitarians concerned with the ethical treatment of animals will accept the principles I have proposed?
The first point to be made here is that most of the points to which I've objected in Utilitarianism are already noted by John Stuart Mill, who proposed a variety of Utilitarianism in which qualitative distinctions among diverse life-activities plays a central role, and in which activity is understood to be valuable in its own right, not simply as a means to pleasant sensations. Mill's utilitarian view, notoriously, is rather Aristotelian; his arguments against simple Benthamism are so cogent that anyone who ponders them is likely to be strongly swayed in that direction. A Mill-style Utilitarianism can easily sign on to the overlapping consensus I have proposed.
Even were a Utilitarian to refuse to accept Mill's reformulations, another route of accommodation awaits us. Henry Sidgwick, while insisting that the correct ethical principle was the unmodified Utilitarian principle, also thought that this principle would not be a good one for most people to apply: better results, from the point of view of that principle itself, would be obtained by encouraging most people to follow a more conventional ethical code based on non-commensurable principles of virtue and vice. Now Sidgwick also thought that for this reason some top government officials should operate, meanwhile, with the correct principle, but his conception of government has been widely criticized for its undemocratic character and its insistence that we ought to conceal from most people the grounds of the political choices that govern their lives. If a modern Utilitarian believes, with Sidgwick, that most people should not try to use the Utilitarian principle, but also believes, unlike Sidgwick, that political principles should be based on ideas that can be publicly stated and that all citizens can understand and accept, then such a Utilitarian, while continuing to prefer the Utilitarian principle to others as the source of a comprehensive ethical view, will gladly accept my Aristotelian view for political purposes.
Much more would need to be done to show that, and how, each of the major reasonable comprehensive doctrines could support the political consensus proposed here. At this point, however, I believe we may conclude that there are no evidently overwhelming obstacles to that agreement. The transition from our current immoral situation to an ethical/political modus vivendi, and from that modus vivendi to a constitutional consensus, and, one may hope, from a constitutional consensus to a genuine overlapping consensus, is a development we may seek and foster without feeling that in so doing we are working in vain.