So what happens when we try to take our reasons all the way down? When we continue asking why we should do anything? We begin to get to a complex meta-ethical question: what constitutes a reason for action? What is it to have a reason to do something? (Warning: this will be an abstract and theoretical post, but it is important to fundamental questions like why we should do anything at all.)

There are at least three things that this last question could mean, three things we could be saying when we speak of having reasons. I like to distinguish the different kinds of reasons in terms of grammar: it’s so far the most precise way I’ve found of spelling them out. There is on one hand a difference of case, between ablative and dative reasons; and on the other a difference of person, between third-person and first-person reasons. English has the second of these distinctions but not the first.

I know the distinction between ablative and dative from my study of Sanskrit and Pali (and to a lesser extent Latin) grammar; the same distinction, I believe, is there in Greek. (It’s not there in German, which has only a dative and no ablative.) In Sanskrit, ablative and dative case endings can both be used to express what we would normally call reasons; but they are very different kinds of reasons. The ablative case describes a cause; it describes the reason why we did something (or why we’re doing it or will do it). The dative case describes a purpose; it describes our reason to do something. The ablative in this sense is translated with “because”; the dative, with “in order to.”

So when we speak of reasons, it can be helpful to specify whether we’re speaking of reasons in the sense expressed by the dative, or only by the ablative. Ablative reasons are the reasons that natural scientists are best at expressing; they’re the only kinds of reasons discovered by chemistry or physics. Everything in the universe acts according to ablative reasons: the rock fell because it had been dropped (and because of gravity). Essentially, they are causes; the “why” in an ablative reason can be replaced with a “how.” In Aristotle’s scheme of four explanations, they are efficient explanations.

Dative reasons, by contrast, are final explanations; they have to do with purpose, aims, teleology. On Aristotle’s understanding, everything had a dative reason; for a modern scientific understanding, this is not the case. There is no purpose to rocks falling or the sun shining. There is, however, some sort of purpose in the biological action of lifeforms, even on a purely scientific explanation. We cannot explain the movements of, say, blood clotting in a wound entirely on the basis of chemical and physical movement; we explain the blood clotting much more effectively if we can talk about what it’s for, namely to protect the wound and stop bleeding. Purpose is such a central part of biological explanation that, until Darwin, it was the most obvious and preferred proof for the existence of God. Everything biological, from the cell to the ecosystem, acts with some purpose to the preservation and reproduction of life; how could this have happened without the action of a God? Nobody had a good answer to that question until Darwin; ever since then, evolution replaced God in explanations, and people have made attempts to base ethics on evolution (usually failing miserably).

It’s dative and not ablative reasons that are of interest to me here. Ablative reasons help us explain the action of the universe; but they tell us less of interest for ethics. To ask “why should we do something?” we need to ask about purpose.

Some – especially Kant – would step in and require a further distinction among dative reasons. The best way I’ve found of putting this distinction is also grammatical: third-person versus first-person reasons. (My grad-school colleague Drew Schroeder used this distinction to help explain Kant to me, though I don’t think Kant himself puts it in those terms.) When a biologist explains blood clotting in terms of its purpose, Kant would say, that explanation too has nothing to do with what actions we should actually take. The purpose of our action has to come from within us.

Sociologists and psychologists can easily explain actions in dative terms. This is clearest in the case of functionalists like Talcott Parsons, for whom basically every social phenomenon can be explained in terms of its purpose for society at large, but pretty much every social scientist will explain actions in terms of some sort of purpose, including individual self-interest or evolutionary fitness. But they’re still explaining action causally, looking at the social or biological variables that cause one course of action to be taken rather than not taken. In the end these third-person dative reasons still turn out to be efficient explanations: we ask what something is for only in order to explain what caused it. First-person reasons are different: they’re the reasons that we use for choice and deliberation in an action.

On the internalist view, I think, these distinctions look a bit less important. If our reasons for action come down to our existing desires or other motivations, then it may well be sufficient to say that we want X because it gives us pleasure, and it gives us pleasure because our upbringing predisposes us that way. But I think it’s that very way of phrasing the question that looks suspicious to the externalist. Should we really take a view that’s that conservative – that just leaves the preferences formed by genes and upbringing as they are? Don’t we want to have better reasons than just being slaves of our pasts? It’s the sorts of judgements implied in those questions – the idea that it is better to make a free and rational choice – that Kant appeals to, and I don’t think it’s too hard to see the appeal in his view.

Several comments from Jesse took this approach. “Morality,” he claims, “has existed in some form or other since the first self-replicating proteins formed in the primordial ocean.” Citing game theory, he notes that organisms which helped each other out would have been far more likely to survive and thrive. Ethan Mills, while somewhat skeptical of the game-theoretic explanation, still cites James Rachels for another kind of evolutionary explanation: at the social rather than individual level, societies wouldn’t have lasted long without morality.

Now I am not and was not speaking only of “morality” in the sense of aiding (or refusing to harm) others. (There was a reason the word “morality” didn’t appear in that post.) As I noted in my comment, I was also speaking of other kinds of value – including virtues like self-discipline and patient endurance that would be valuable whether or not anyone else is around, and for that matter of aesthetic value, the value in good art or the beauty of nature.

But that’s not the big issue here, for it’s not so hard to come up with evolutionary explanations for these other kinds of value either. Self-disciplined creatures would very likely have adapted better to their environments. There are plenty of people, perhaps most notably Denis Dutton, who have even tried to find evolutionary explanations for aesthetics.

I am not going to pass judgement here on whether evolution is a correct or adequate causal explanation for the origins of human value judgements. For the sake of argument, in this post, I am going to assume that such accounts get the causal origin of value judgements basically correct. Because far more important is a deeper criticism: they miss the point.

The error being made here is parallel to the one that tries to prove God’s existence merely as a First Cause of the universe, not as a First Explanation. In this bastardized version of the cosmological argument, the causal processes of the universe must have a starting point, identified with God – a rather useless God, one that doesn’t mean anything more than the Big Bang. But the intellectually respectable form of the cosmological argument isn’t just about causes, but about other kinds of explanation: not just where the world comes from, but what its essence is and what it’s for.

Return more directly to the present topic: to explain the causes of value judgements, to identify where they have their origin, is not actually to explain value. What this kind of explanation explains is the bare fact that people happen to make judgements of value. What it doesn’t and can’t explain is the truth or falsity of those judgements. To have an adequate account of ethics and values, we need to know not merely why people happen to think some things good and some things bad (or why they act accordingly). We need to know why things actually are good and bad. (Our mode of explanation needs to be ethics, not ethics studies.)

Few are seriously prepared to jettison this distinction, between what actually is good or bad and what people merely believe to be so. We want to say that Pol Pot was wrong when he thought it was a good thing to commit genocide on his own people. To consistently say such a thing requires that we believe value judgements can be correct or incorrect; they need to have a referent, to refer to the action having a real goodness or badness independent of whether the agent takes the action or believes in its goodness. (Some do try and advocate a thoroughgoing value relativism, of course; I have responded to some such arguments here, here and here.)

I think the early analytic philosophers, more than those who find evolutionary explanations sufficient, at least grappled with this problem – they just failed to solve it. They asked: what do we mean when we call something good or bad? Those who try to reduce judgements of good or bad to a simple descriptive property – good is what fosters the species, produces pleasure, etc. – run into trouble pretty quickly, for it’s pretty clear that a great many usages of “good” do not simply mean any of these things. One could try and argue that those who use “good” to mean anything other than species preservation or the production of pleasure are mistaken, but they’ll have a pretty hard time making the case. Neil Sinhababu made a valiant effort, but I have argued that he failed, with some additional thoughts.

I have many problems with G.E. Moore’s concept of the naturalistic fallacy, and especially with the inadequate alternative he provided (as I discussed last week) – but I suspect that this important point is where he was coming from when he came up with it. Moore took the idea of the naturalistic fallacy much too far; I think one can legitimately make inferences from “is” to “ought” statements, but one should still be careful about doing so, especially when one puts a particular kind of descriptive claim at the heart of one’s ethics. The problem is nicely illustrated by Ayn Rand in this deeply problematic passage:

the fact that living entities exist and function necessitates the existence of values and of an ultimate value which for any given living entity is its own life. Thus the validation of value judgments is to be achieved by reference to the facts of reality. The fact that a living entity is, determines what it ought to do. So much for the issue of the relation between “is” and “ought.” (The Virtue of Selfishness, p.17)

What Rand doesn’t seem to have thought of is that such a view can have absolutely nothing to say to the person who chooses to kill herself – in suicide, in war, in civil disobedience. If your system of values comes out of the desire to live, it is irrelevant to anyone who takes that desire as unimportant. I think there’s a similar problem with the point Ethan makes in his comment that Buddhism is ultimately based on our desire to end suffering; not everybody treats that desire as decisively important, and I think Buddhists have a problem addressing those who don’t. (I intend to take up this point more next week.)

How to get around this problem? I note that Ethan lists Aristotle as having a “naturalist” theory of ethics comparable to Rand’s or Sinhababu’s. But Aristotle’s theory is a bit different from theirs, in that he sees value as an inextricable part of the natural world. The idea of God as First Explanation ultimately derives from his thought, because for him explanation needs to be teleological as well as causal: you need to explain things in terms of their purposes, what they’re for, as well as their (efficient) causes, what put them there. Aristotle’s views of nature are of course tied up with beliefs about causes that we cannot share in a scientific age. But it seems to me that we may still need something like them in order to make sense of value.

Americans have typically greeted bin Laden’s death with jubilation and celebration, often waving American flags and chanting “U.S.A.” But some minority voices, such as Linton Weeks at NPR radio and Pamela Gerloff of the Huffington Post, have raised questions about this celebration. Is it really a good idea to celebrate a human death, even the death of one’s enemy?

This all makes a good occasion to revisit an earlier short post of mine, one of my favourites. The thing that affected me most at my one Goenka meditation retreat was not the meditation practice in general, but the closing practice of karmic redirection, because it specifically involved wishing George W. Bush well – and, more generally, wishing one’s enemies well. What applies to Bush here applies to bin Laden – the two men are of course enemies of each other, but I also consider them both enemies of mine.

A couple months ago, Thill questioned the value of Goenka’s practice – not over its efficacy, but over the values that underlie it. Thill asks: “Is wishing the enemy well actually a case of masochism since the enemy is a person who wants to harm us?… What if the enemy is a sadist whose happiness consists in seeing you suffer? Then, wishing this enemy happiness is tantamount to wishing one’s own suffering!”

As Jim Wilton rightly noted in his replies, wishing enemies well does not entail wishing them success in their aims, or wishing that their desires be fulfilled. This is as true of one’s friends as of one’s enemies. If my friend is addicted to crack cocaine, wishing him well does not mean that I wish he find more crack to smoke. Indeed I wish him the exact opposite. What he needs most is a change in the structure of his desires; he will probably be better off with the desires unfulfilled, as that would bring about the relevant change. And the same applies to people with evil or hateful aims: wishing them a good and happy life carries with it the wish that they improve and become better people. Thill’s comments here have assumed a simplistic understanding of happiness that equates it with the satisfaction of desire, when often what is needed for a long-term and stable happiness is the exact opposite.

In reply to Jim, Thill makes an important point: “note the element of self-interest in all this. In wishing all that for your enemy, you are also wishing a change in your enemy’s attitude towards you. It is all tantamount to wishing that he or she is in a condition in which he or she ceases to be your enemy!” That’s true. But even if one characterizes it as self-interested, one should notice what such wishing for one’s enemy’s virtue doesn’t imply: namely revenge. One wishes that, in spite of the bad things the enemy has done, he might still become better and happier, in the process ceasing to be an enemy. One does not take the enemy’s violent and painful death as an occasion for celebration.

Now let me clarify: this is not a call for pacifism. Shortly after the September 11 attacks, I sat in on a class at Harvard where the professor’s response to the attacks was “I think we should set up an exchange program, so that people in our countries can better understand each other.” (Students applauded.) I was stunned at the naïveté expressed there. We are not talking about people who express frustrating differences at the ballot box (like, say, Québec separatists – most of the time). We are talking about people who want to kill you, and have just killed several of your fellow countrymen simply because they were your fellow countrymen; they would do it to you if given the chance – like on an exchange program.

Gandhi, to whom Thill refers in this context, was considerably more sophisticated than said professor. Gandhi understood that his pacifism would cause great suffering, even many deaths, to his own side; but that it was worth it to achieve his goals in a morally upstanding way. It’s worth celebrating the success of Gandhi’s nonviolent methods against colonialism – and those of Martin Luther King, who derived many of his methods from Gandhi. But Gandhi and King were facing enemies who believed in justice over power, in the rule of law, in the value of human life. The goals of the British Empire and of the American South were to preserve an unjust and discriminatory social order which they believed to be benign. The goals of the Nazis, by contrast, were extermination. If an Indian stood fearlessly in front of a British soldier’s gun, the soldier would rightly fear the public repercussions of shooting. If a Jew stood fearlessly in front of a Nazi gun, she would merely save the Nazi the work of rounding her up. Bin Laden, in this respect, was far more akin to the Nazis – his attacks weren’t even to make demands, the destruction itself was the goal. (It is worth noting that Bush, however, would have been significantly more akin to the British Empire.) I agree with Thill on this much: one often must fight against one’s enemies, and sometimes this does require violence.

This violence is, however, regrettable. In war, killing another human being can be – and often is – the best course of action. But it is a tragic right action, and one should be aware of this fact. Thill claimed in another context: “Even if you want to kill a dog or a horse in order to put it out of misery and you do it skillfully, it would still be a gross distortion to describe this act as one which gives pleasure to the agent.” That is, one feels compassion, a painful emotion occasioned by another’s suffering. I discussed compassion myself in response to Thill’s post, noting that because we are not perfect or ideal people, we need remind ourselves that others’ pain is a bad thing (even if a hypothetical perfect person might need feel no regrets). The killing of an enemy, it seems to me, fits under exactly this class of action: necessary but regrettable, a proper occasion for compassion. Finding and punishing bin Laden was an important goal, and it is good that the US government under Obama succeeded in accomplishing this goal. And yet even so, it is not an occasion for celebration, but for sadness that it had to come to this.

Suppose a trolley is hurtling down a track, on which are placed five innocent people with no chance to escape in time. You are standing beside a switch that will redirect the trolley onto a track where stands one innocent person, who also has no chance to escape. Should you flip the switch, and thereby kill one to save five?

Now suppose there is no track onto which the trolley can be redirected; the five innocents will be in its path no matter what happens. Instead of being beside a switch, you are standing on a bridge over the tracks, beside a very fat man looking down over the action. You can push the man over the bridge, knowing his enormous girth will stop the trolley’s movement before it hits the innocents. Should you push the man, and thereby kill one to save five?

Michael Sandel begins his famous course on Justice with this action scene, and it’s a great way to start such a course. This trolley problem, ingeniously introduced by Judith Jarvis Thomson and the late Philippa Foot, is a wonderful way to shock beginning students out of their ethical complacency. For nearly all people faced with this problem agree they would kill one to save five in the first situation but not the second. After hearing one case they think there’s an easy principle by which to decide the right action; after hearing the second, they are forced to admit that there isn’t.

Whether they are meek Stonehill students who feel uncomfortable disagreeing and asking hard questions, or cocky Harvard students who think they already know everything, the trolley problem forces undergraduates to think hard about ethics. It makes us realize that ethics cannot be limited to “common sense”: it makes us see that our untrained “intuitions” are not enough on their own, that there is something to be learned from studying ethics and having special training in the subject.

Still, the trolley problem can be overdone – and typically is. Many analytical ethics courses, including the one I took as an undergraduate at McGill, are effectively about nothing but the trolley problem. If one reads the great thinkers in ethics at all, one reads them merely to provide a theoretical justification for each main side of the problem: Kant for why we shouldn’t push the fat man, Mill for why we should flip the switch. There’s no place for Aristotle or Nietzsche in such a course, let alone Mencius or Śāntideva. The goal is only to hammer out some sort of principle that could allow one to be consistent in both cases. (The most common candidate for such a principle is Thomas Aquinas’s doctrine of double effect, although instructors in such courses usually skirt around giving credit for this principle to someone whose motive in coming up with it so obviously “religious.”)

Beyond introductory courses, many analytical ethicists spend their careers coming up with fanciful hypothetical cases of the trolley sort – “thought experiments,” they are typically called, ways of isolating principles behind our existing “intuitions.” Often more and more conditions are placed on the hypothetical situation in order to elicit the response one hopes for. For example, since one might avoid pushing the fat man for fear of legal consequences, one might instead speak of Thomson’s alternate “transplant” case:

Suppose you are a solitary doctor in a remote wilderness area, perhaps the Canadian Arctic. You have five patients in front of you who each need a different organ transplant: one needs a new heart, one needs a lung and so on. You have the knowledge and equipment to make the transplants, but you don’t have the donated organs, and there is no way they can be shipped to you in time. But a lone explorer walks into your clinic for comprehensive health tests that require he be sedated. The tests turn up very well: all of his organs are in perfect shape. And a thought occurs to you: you could cut him up and take out all of those organs to give to the other five who need them. Nobody would ever find out. Should you cut him up, and thereby kill one to save five?

But one begins to wonder: what sort of ethical exercise is this, anyway? What sort of situations – what sort of life – is one preparing to deal with? True, in most human lives one will face decisions about sacrificing some people’s interests for the sake of others. If one is a medical doctor or in the military, lives may very well be at stake. The trouble is, in these actual situations, the additional factors in one’s decisions – like the possibility of being caught – will be more rather than less. Hypothetical examples like the trolley problem are designed to bring in the economist’s method of ceteris paribus, “with other things being equal.” But in philosophy as in economics, the ceteris are never actually paribus. If one is to make the right decision in a real case, one can’t merely leave out the “extraneous” factors; everything must be part of the decision. Moreover, in many such cases – the trolley cases certainly suggest this – the decision must be made in a split second. If one is to be prepared to do the right thing when a real hard case comes up, shouldn’t one be thinking through similar real hard cases, rather than fanciful science-fiction scenarios? (The reliance on hypothetical cases may be one more reason why surveys find ethicists aren’t actually more ethical than anyone else.) I’ve suggested before that this sort of point is what underlies the contemporary resurgence of “virtue ethics”: a shifting of our philosophical concern away from rare or hypothetical cases to the difficult task of acting better in everyday life.

Sandel, I think, got this right. Begin ethical reflection with the trolley problem as a wonderful pedagogical device to shock students out of their complacency and get them actually thinking. But after that first introductory moment, get them thinking about real cases in their complexity, and the deep thinkers who are justly revered for their sustained reflection on that complexity.

In the prologue to The Sources of Normativity, probably her most important and influential work, Korsgaard provides what she calls a “very concise history” (her emphasis) of the connections between metaphysics and ethics in Western philosophy. I noted recently that the concept of obligation is central to Korsgaard’s philosophy, as it is to Lévinas’s; this prologue provides us with historical reasons why an obligation-centred philosophy might be a worthwhile project.

Plato and Aristotle, Korsgaard notes, had a philosophy focused on excellence (aretē, often translated “virtue”) rather than obligation, as do most of those who today reject Kantian and utilitarian ethics and are therefore usually lumped into the catch-all category of “virtue ethics.” Their ethics had much more to do more with what is good, what we should care about, than with what others oblige us to do. But, Korsgaard adds, in Plato and Aristotle this account depends on metaphysics, on a view of the way things really are. For them, a thing’s highest perfection and potential – its form – was in some sense more real than the existing particular thing as it actually is.

Korsgaard correctly notes that Christianity changed Western philosophy’s emphasis, away from excellence and toward obligation and law, with God as the lawgiver. But what if we no longer assume that God is the source of ethics? What we cannot do, she says, is go back to Plato and Aristotle’s world of excellence. “Because for us, the world is no longer first and foremost form. It is matter.” (4) By identifying ultimate reality with matter, we have separated the real from the good; we no longer look at actual things as reflecting a higher and better potential. And this means that a Platonic or Aristotelian ethics of excellence is no longer available to us.

What Korsgaard does not say, however, is that this new, hard, scientific world is entirely bereft of value. Indeed, she sees that it cannot be. (Although she does not put it in these terms, science’s claims to truth are themselves grounded in value.) She says:

If the real and the good are no longer one, value must find its way into the world somehow. Form must be imposed on the world of matter. This is the work of art, the work of obligation, and it brings us back to Kant. And this is what we should expect. For it was Kant who completed the revolution, when he said that reason — which is form — isn’t in the world, but is something that we impose upon it. The ethics of autonomy is the only one consistent with the metaphysics of the modern world, and the ethics of autonomy is an ethics of obligation. (5)

Now I think there is something very wrong with this paragraph. Korsgaard has accepted that value has a real place in the world, even the world of a modern scientific metaphysics; and she then claims that value’s place in the world is one of obligation (as opposed, by implication, to excellence). The next parts of the book flesh out her account of the ethics of obligation, but let us leave that aside for the moment. Let us assume for now that Korsgaard, in the rest of the book, succeeds in founding ethics on obligation. Isn’t there still something missing?

Korsgaard’s account of value, as provided here, derives that value only from obligation. If her account in the rest of the book were correct, it might be the case that all moral value comes from obligation. But is that the only kind of value in the world? Korsgaard never tries to argue that, and it’s hard to see how she could. She opens the prologue by saying: “It is the most striking fact about human life that we have values. We think of ways that things could be better, more perfect, and so of course different, than they are; and of ways that we ourselves could be better, more perfect, and so of course different, than we are.” (1) But things are not obliged to do or be anything, certainly not on any Kantian account of morality. Indeed if one were to imagine obligation being applied to things, it would likely have to be on something like the Greek teleological metaphysics that Korsgaard explicitly rejects: it is the purpose of a knife to cut well, therefore it is that knife’s duty to cut well.

There is, then, a yawning gap in Korsgaard’s historical account of value, even if we take her account of morality and obligation to be true. At a minimum, this ethics must be accompanied by an aesthetics. Some accounts of ethics – including those I’m most sympathetic with – do not restrict their concern to morality in the strict sense, and might therefore include aesthetics, but this appears not to be the case with Korsgaard’s. And while Korsgaard’s quote above tantalizingly lists “the work of art” along with “the work of obligation” above, suggesting the importance of aesthetics, it seems on a fuller reading that this is only apparent: when she uses the word “art” elsewhere in this passage, she contrasts it with what is natural, and so appears to mean only “artifice,” the Xunzian point that we are not naturally good but need to work on it.

And so it seems that aesthetics, at least, is missing from Korsgaard’s account. Just as we need an account of how people’s actions can be right and wrong, so we need an account of how things can be beautiful and ugly. Kant did not have this problem since he had a highly developed aesthetics, but it is not clear whether Korsgaard buys it. But it would seem, on Korsgaard’s account, that one must either adopt something very much like Kant’s aesthetics (as she does with his ethics) or return in some respect to a semi-premodern metaphysical account that sees value in the world while still taking science into consideration – as Hegel tried to do, for example. If one takes this latter route with aesthetics, however, it would seem that one is compelled to do so with ethics too.

I recently noted the strong similarities between Korsgaard’s philosophy of obligation and that of Emmanuel Lévinas. Lévinas, in one of his better-known essays, tells us that “ethics is first philosophy” – and by “ethics” he means obligation. But, I’m told, Speculative Realist Graham Harman retorts that “aesthetics is first philosophy.” I’m wondering if issues like this are what Harman has in mind: we don’t just need an account of moral value, we need an account of value as such.

In his excellent post which quotes Harman to this effect, Skholiast adds a quote from Wittgenstein that “Ethics and aesthetics are one.” I’m not sure I would go that far; but it seems to me that there must be some sort of connection between the two, a connection that Korsgaard implies only to ignore. We could, I suppose, say that axiology is first philosophy – “axiology” meaning the study of value – though that phrase doesn’t sound nearly as cutting as either Lévinas’s or Harman’s.

Ethical egoism of some description – say, as advocated by Epicurus – is a perfectly respectable philosophical position. One can say that one’s reasons to benefit others are all ultimately based on benefit to oneself, if one’s own self-interest is rightly understood. Neither Paul has a great deal of sympathy for this position, as far as I can tell, but it is not what they take as a target for their attack.

Rather, they reserve their greatest ire for a position that derives other-orientation from ātmanism – or at least from nondualism. Though Śāntideva is the last to believe in an ātman, he, like Vivekānanda, nevertheless gets to altruism by deconstructing the self, saying the differences we perceive between selves are not ultimately real. Śaṅkara and Buddhaghosa would likewise have taken the first step and deconstructed the self, saying the different human selves we perceive are; but what they would not have done would have been to take this as a justification for altruism. As with Epicurus, our primary goal needs to be our own liberation from suffering. This conclusion, the Pauls take as logically acceptable, though they disagree with it.

But the next step that Śāntideva and Vivekānanda take and Śaṅkara and Buddhaghosa do not – to say that Epicurean egoism is not acceptable because the individual self it defends is unreal – is a step too far, in the Paul’s eyes. For by deconstructing egoism, they reason, Śāntideva and Vivekānanda also effectively deconstruct altruism. (Williams’s chapter is entitled “How Śāntideva destroyed the bodhisattva path”!) If there is no self, there can be no other about which to be concerned; nor can there even be suffering to be prevented.

But neither Paul says this because they wish to advocate an Epicurean egoism, to take us back to the egoistic nondualism of a Śaṅkara. They want us to be altruistic – but only on the right grounds, and these grounds are grounds of encounter. For there to be real altruism, there must be real others; and therefore altruism must come out of encounter and not out of ātmanism or nondualism.

And while up to now I’ve discussed this issue in the sectarian terms of Catholics attacking Buddhists, I think the distinction made here also shows up in contemporary analytical ethics. Derek Parfit has argued for altruism on grounds which even he identified as analogous Buddhist non-self – the self is not a real entity from moment to moment, and so we should not privilege it over others. Mark Siderits has recently taken up, at book length, the similarities between Parfit’s view and those of Buddhist thinkers like Śāntideva.

I used to think there were close similarities between Parfit’s (and Śāntideva’s) view and that of Christine Korsgaard, who – like them – argues that full-blown egoism is not rational. But the Catholic Pauls pushed me to see the differences between them. For Korsgaard criticizes egoism in a very different way, one that they could endorse.

Korsgaard, it turns out, does not deconstruct the ego itself – only egoism. The self, on her account, is quite real; but its reasons for action are not fundamentally egoistic. In everyday life, “We do not seem to need a reason to take the reasons of others into account. We seem to need a reason not to. Certainly we do things because others want us to, ask us to, tell us to, all the time…. We respond with the alacrity of obedient soldiers to telephones and doorbell and cries for help.” (The Sources of Normativity 140-1) Korsgaard tries to argue that reasons for action are public in their very nature; each individual’s reasons for acting are not separate from the reasons of other individuals. And one of the fundamental ways in which reasons apply to others is obligation, which comes out of respect for others’ humanity or personhood. If I am blithely torturing a stranger (Korsgaard’s example, derived from Thomas Nagel) and the stranger asks “How would you like it if I did that to you?” I can continue to torture the stranger, but not in the way I did before, for the stranger has now obligated me.

There are very strong echoes here – possibly uninentional – of Emmanuel Lévinas, the Jewish archetypical philosopher of obligation and encounter. Obligation is not a concept that shows up in Śāntideva – or, for that matter, in Aristotle. Korsgaard’s own introduction notes that it was the Christians – surely under the influence of Jewish law tradition – who began to move the mainstream of Western philosophy away from concepts of excellence (or virtue) and toward concepts of obligation. And this obligation always seems to be an obligation toward someone irreducibly different from oneself. The Advaitic ātman might have good reason to reduce its own ignorance, but it is not obligated to do so.

So, leaving aside egoistic philosophies for the moment, we can draw boundaries between two quite different justifications for altruism, two different ways in which egoism can be considered an error. In Korsgaard, Lévinas and I think the Catholic Pauls, we get an encounter variety of altruism, where each separate and individual self is in part constituted by binding obligations to others (whether other people or God). Whereas in Śāntideva, Parfit and Vivekānanda, we get a nondualist variety of altruism, one based on the idea that the selves themselves are not really real. The Catholic Pauls attack the second because they wish to move us toward the first.

So far, it’s pretty much the usual story of 19th-century reform. But Hacker takes his critique much further than the basic historical point, and this is where it gets interesting to me. Hacker’s special ire, beyond his general disdain for modern Hinduism, is reserved for the “tat tvam asi ethic”, the idea that because we are all ultimately one infinite spirit (“you are that,” as the Chāndogya Upaniṣad supposedly claims), we should help each other because we are really helping ourselves. For Hacker, it is not merely the case that classical Advaita Vedānta thinkers never adopted an altruistic or activistic ethics based on the tat tvam asi of the Chāndogya Upaniṣad, but that they could not have. For, Hacker claims, “From the philosophical point of view, to base the tat tvam asi ethic on the foundation of the Vedāntic monism of consciousness is a logical impossibility.” (“Schopenhauer and Hindu ethics,” p. 305) On the next page he goes on to describe the tat tvam asi ethic not merely as a “logical impossibility” but as a “logical monstrosity.” (p. 305, my emphasis) Hacker wants to show the tat tvam asi ethic is a modern invention because, in his mind, the great Vedāntic sages of old were way too wise to ever have fallen for such a load of garbage.

What is it about Vivekānanda’s tat tvam asi ethic, in Hacker’s mind, that makes it logically impossible and even monstrous? For Hacker, genuinely ethical behaviour – by which he means altruistic behaviour – depends on the existence of separate persons, whose differences are irreducible:

Ethical behavior presupposes an interpersonal relationship, which loses its metaphysical justification if individual personhood has no ultimate reality…. Neither the monism of will nor the monism of consciousness or spirit has a real place for the concept of person. But when this concept is not taken seriously, ethics remains on a naturalistic level; that is, there is no true ethics, good and evil have no truly metaphysical relevance, and ultimately there are only ways of realizing or veiling the impersonal universal One…. There is no sense in which an identification of a “that” with a “thou,” such as we have in tat tvam asi, can explain why good and bad behavior exist. Interpersonal relationship is not identity, and it is certainly not identity of a person with an impersonal being.

As philosophical argument I do not think this goes very far, not by itself anyway. Much of it depends on the semi-tautological identification of “ethics” with altruism. If one acknowledges that an ethics can be based on self-interest and that other-interest can be grounded in self-interest, then there seems little logical problem here: the tat tvam asi ethic might not really or ultimately be altruistic, but so what? Even in historical terms, Hacker seems to be on poor ground in believing that such a monistic ethic is purely modern. Hugh van Skyhawk, replying to Hacker in the 74th (1993) volume of the Annals of the Bhandarkar Oriental Research Institute, argued that a similar view was found in the sixteenth-century Marathi poet-saint Eknath (also spelled Ekanāth or Ekanātha). Eknath told his listeners (in Skyhawk’s translation) that the true yogī “immediately gives up his own interests and ventures into difficulties for the sake of others”; and argues for such altruism on strongly nondualist grounds:

He, for whom there is no more “I” and “mine” and “thee” and “thine” by virtue of the contact with the worship of the divine non-duality and the Self is called the highest bhakta. If he gives his fortune (nijavitta) to another, no misgivings arise in his citta. He does not even sense a trace of alienation. No feelings of doubt arise. The object in the right hand is given to the left hand. Who is the giver here? Who is the receiver?

Overall, then, Hacker’s arguments against monist ethics aren’t particularly persuasive. What excites me about Hacker’s arguments is his reasons for making them. Wilhelm Halbfass’s introduction to his collection of Hacker’s writings stresses the increasing importance in Hacker’s work of his conversion to Roman Catholicism. And Catholicism, it seems to me, stresses encounter over ātmanism: it is all about one’s relationship to a God with whom one is not identical.

The point is highlighted in the much more powerful arguments of another Catholic Paul, Paul Williams. Williams, to my knowledge, says nothing about Hacker in his work; since Williams is a Buddhologist, he may well be entirely unaware of Hacker. And yet Williams’s criticism of Śāntideva (in the final chapters of his Altruism and Reality parallels Hacker’s criticism of Vivekānanda in remarkable ways. Among Śāntideva’s most famous passages (now even excerpted in an introductory ethics text) is his “equalization of self and other” in verses VIII.90-119 of the Bodhicaryāvatāra, in which he argues that, since the self is an illusion (a standard Buddhist view), egoistic action does not make logical sense and we should be altruistic (an innovation of his). Śāntideva is not a monist like Vivekānanda; he is strongly opposed to the Vedāntic idea of a universal cosmic self. Nevertheless, there is a close parallel in that both Śāntideva and Vivekānanda try to deconstruct our ideas of self in order to deconstruct ethical egoism and urge altruistic action. And so Williams’s criticisms of Śāntideva turn out on similar lines to Hacker’s criticisms of Vivekānanda.

Unlike Hacker, Williams makes no attempt at historical criticism; Williams has no doubt that Śāntideva actually believed all this. He simply thinks that Śāntideva is dead wrong. In thinking and arguing this, he has provoked a strong reaction among Buddhologists, no less than five of whom (Barbra Clayton, John Pettit, Jon Wetlesen, Mark Siderits and José Cabezón) have tried to refute him in print. I’m not going to examine today whether Williams is right or wrong (it is a complex question); but I want to explore important points in his arguments.

What Williams claims, against Śāntideva, is that there can be no compassion unless there are persons feeling the compassion for other persons. Compassion requires the existence of persons feeling suffering; without sufferers, there is no suffering and no compassion. (T.R. (Thill) Raghunath made a similar argument in a recent comment.) If the self is deconstructed, so too is suffering, and indeed perhaps all reasons for action.

Both Paul Hacker and Paul Williams, then, are trying to tell us: you cannot have it both ways. Either you can have a nondual view (monist or otherwise) that deconstructs our everyday selves, or you can have the commitment to altruistic alleviation of others’ suffering. The two don’t make sense together; and the first certainly isn’t an argument for the second.

Such a view seems to me to have profound roots in the Abrahamic monotheisms; while the Pauls in question are Catholic, one could surely also imagine it being made by a Jew. For indeed the criticism reminds me strongly of Emmanuel Lévinas and his insistence on the irreducible otherness of other people – with God as the ultimate other. (For breaking down the distinction between himself and God, al-Hallāj was tortured and killed.) The ethical deconstruction of self seems important to a nondual view of the world; but to refute such nonduality seems central to theism. (But not only Abrahamic theism: the nineteenth-century Bengali devotional poet Ramprasad Sen criticized nondualism by saying “I want to taste sugar, not to become sugar.”)

Ni’s talk focused on the Chinese concept of gongfu 功夫, dating from the early centuries CE and meaning any practical art – it could include calligraphy, sports, cooking, good judgement or statecraft. (Although the word gongfu has long ago passed into English with an alternate spelling, it is probably best to keep using the Pinyin spelling rather than confuse people with a term most associate with goofy movies about roundhouse kicks.)

Gongfu as Ni understands it then bears some resemblance to the Greek concept of technē, or Alasdair MacIntyre’s concept of practice, with one crucial difference. Aristotle’s technē involves a telos; it is embedded within a larger determinate framework of human flourishing. With gongfu, on the other hand, Ni agreed with my earlier characterization of the process as a technique. It is open to us to choose our aims; gongfu merely allows us to achieve those aims. There is a gongfu of killing as well as a gongfu of saving. (Ni effectively uses the concept to expand his previous characterization of Mencius into a constructive position.)

Ni urges us to a conception of practical philosophy in which gongfu, thus conceived, takes centre stage. Theoretical philosophy, especially metaphysics, then serves the function not of description but of recommendation. Philosophy is a way of achieving our chosen ends, a set of instructions rather than responsibilities. Philosophies, like other practices, can be evaluated as techniques – on their effectiveness at achieving their aims.

There’s a word for the kind of philosophy Ni is describing, and it’s relativism. Ni’s gongfu is not relativistic with respect to means; a philosophy can be discredited if it fails to achieve its goals. It is, however, entirely relativistic with respect to ends; ultimate ends are up to our decisions and choices, and there is no rational basis on which to criticize them. The value of each gongfu is relative to the incommensurable ends it aims to achieve.

As such, Ni’s approach seems vulnerable to the standard criticisms levelled at relativism. One asks: does this philosophy have any grounds on which to criticize evil actions – of which we might often take Adolf Hitler’s as the paradigm? Ni’s first answer was, to my mind, entirely unsatisfactory: that Hitler’s project failed on its own terms, that he committed suicide and ended his life in misery. This claim is of course true as far as it goes, but it doesn’t go far. It is not too difficult to imagine a Hitler who succeeded, perhaps by reining in his ambitions a little bit and maintaining the Nazi-Soviet pact. Such a Hitler, maintaining his reign of terror for decades or more, seems worse than the Hitler we know.

Ni then proceeded to offer a strong perspectival defence of sorts: criticism would be part of our own gongfu. We can criticize Hitler from our side, within our own ends; we can and should take this a step further and fight him. Action against Hitler is a part of achieving our aims; it’s just that there’s no objective ground from which to criticize him.

Against such a view, I developed some of the arguments I made in my critique of postmodernism. Relativism privileges the strong. It is no coincidence that Gandhi and Martin Luther King Jr. believed in universal, objective truths; for it was only on such a basis that they could nonviolently shame their oppressors into relenting. Imagine King standing up and proclaiming: “I have a dream that my children will one day live in freedom and justice and brotherhood. But I know that you have a dream of maintaining this world of segregation, and I know that objectively my dream is no better than yours. So I will fight for my dream, and you fight for yours.” If civil rights leaders had all talked that way, even thought that way, it’s easy to imagine the South remaining segregated for centuries.

Moral persuasion works by imagining ideals larger than one person’s given ends. Without it, there is only violent persuasion, persuasion by force – which, by definition, favours the strong. It is no accident that the most powerfully expressed relativist position in Plato’s Republic – the one which ends on a note of “you have your position, Socrates, and I have mine” – is expressed by Thrasymachus, who has argued that justice is merely the interest of the stronger. Without an ability to cross paradigms and argue about ends, the interest of the stronger is what prevails. When the weak prevailed and achieved a more just world, as they did in Gandhi’s and King’s cases, they could only do so because they had on their side a conception of the good beyond their own limited paradigms, one which had a binding authority on everyone.

Knowing this point, those aiming for change could certainly try to lie – to proclaim universal ideals they did not themselves believe in, as itself part of the technique, the gongfu, for achieving their individually derived goals. (I believe that Gayatri Spivak has argued for a “strategic essentialism” that bears a strong resemblance to this approach.) An outsider might refer to such a person as a liar and a hypocrite, but such outside criticisms do not of themselves need to bear any weight on the relativist individual who disregards outsiders’ ends. More important is that such an approach can itself be rather self-defeating – public figures aiming for social change have their words and actions relentlessly dissected and examined. If King or Gandhi had really believed that what they were doing was only best for them and not universal – but proclaimed the opposite – their lies would have stood a good chance of being exposed.

Or, pushing the point further, one might even try hard to believe in a universalist view in order to advance one’s own pragmatic goals. Ni’s interpretation of Mencius (about which I hope to say more) suggested such an approach: rather than deriving one’s ethical or political practice from a metaphysics of the world’s nature, one starts with the practice and employs the metaphysics as a part of it. So one might try to take on a universalist metaphysics in order to advance one’s pragmatic goals, even though one is convinced that there is no such universal metaphysics that transcends each individual’s given ends. I have somewhat more sympathy for this possibility, as I have explored a similar possibility with respect to hedonism. But I concluded there that such an attempt is self-defeating. More generally, from a commonsense point of view, it is bad to believe things one knows to be false; from a philosophical point of view, it is bad to avoid thinking too hard lest one think the wrong things. More specifically, contradictions get in the way of one’s own practice, whether personal or political: when one believes a contradiction, one cannot – pretty much by definition – believe either side of the contradiction wholeheartedly. It is much more difficult to fight for justice (or anything else) when one is already at war with oneself, for such a fight must be fought on two fronts.

Ni made one final reply before the lunch ended: he noted that I was myself arguing merely based on pragmatic effectiveness, not on the grounds of the larger metaphysical truth I hope to proclaim. He was absolutely right about this, I think, but in a way that does not undercut my position. I’ve said a lot here already; this point deserves enough attention that I will save it for another post.

The argument to this point is, I think, in perfect sympathy with Augustine. Human beings for him are invariably and inevitably flawed, in a way that makes them unworthy of our trust. Instead, Augustine wants to argue, we must place our trust in a truly perfect being, God. Augustine’s argument here underlies a great deal of conservative Christianity: even if church institutions and/or biblical scripture appear wrong to us, they are a better guide than our own weak and easily misled intellects.

For the moment, let us leave aside the question of how we know Church or Bible embody God, or even whether God exists. I think there is a far deeper question at issue here: even assuming he exists, how can we trust God?

Most of the answer to the question will hinge upon how we define God. But let us assume that God has one characteristic attributed to him by almost every believer, even by deists: that he is the creator of all that is, directly or indirectly responsible for everything that happens except (perhaps) those events caused by human free will, and perhaps the will of other free beings like angels.

If that is so, the verdict is severe: God’s track record is no better than ours. Too often we think of the “problem of evil” rather than, more correctly and appropriately, of the problem of suffering. And then we neglect to think the problem through, and blame it all on human free will. For when we live so close to the twentieth century, the readiest examples of grave horrors are human-caused; the mere mention of the names Hitler, Stalin, Pol Pot make it easiest to question God. But this version of the question is also the easiest to answer: the universe would not be as good if we were not free, and this freedom is worth the possibility of evil.

But how small this human-caused misery begins to look compared to the misery caused by God. In Hurricane Katrina, the Haiti earthquake, the Asian tsunami we have plenty of recent examples of suffering not caused by humans. Smallpox, malaria, tuberculosis, cancer have killed more than Hitler or Pol Pot ever did. The tortures of ALS make the gas chambers look humane. Crippling diseases, natural disasters, animal attacks: we didn’t do that. God did.

And that’s just a deist God, a God inferred from creation. The evidence against the God of scriptures is worse still. In the book of Exodus, God punishes every Egyptian family with ten “wonders” – diseases, crop failures and more – culminating in the deaths of all their firstborn children. They are punished not for their own actions, but for the actions of their Pharaoh – even though the text explicitly says that God “hardened Pharaoh’s heart,” God deliberately caused the Pharaoh to do the very thing that Pharaoh is punishing him for. Later God sends horrible afflictions, including the death of his children, on his most faithful servant Job, just in order to win a bet with the Accuser (“the Satan” in Hebrew). Worse even than all this is the idea of a literal hell, not necessarily attested in the scriptures but widely believed in the traditions, including by Augustine himself. Whatever Pol Pot did to his victims, it always ended with death. God keeps going, tormenting people for all eternity, with no deterrent purpose whatsoever, leaving sheer vengeful retribution as an end in itself.

It seems to me the evidence against God is, quite literally, damning. Augustine, it seems to me, is right that humanity is fallen and sinful, not worthy of trust. The problem is that God is worse. (And let me stress again that it is not God’s existence I’m addressing here. Like Ivan in Dostoevsky’s Brothers Karamazov, I would not trust a creator God even if he did exist. Maybe especially if he did.)

It is not only Western traditions that face this problem. These reflections came to me when I began reading Rāmānuja‘s commentary on the Bhagavad Gītā. Rāmānuja begins the text with a long homage to Vishnu as the creator of all things, who appears in the Gītā in the form of Lord Krishna. The purpose of life, according to Rāmānuja, is to reach knowledge of and devotion to this Lord. But Krishna always appears as a morally questionable sort of deity, from his childhood stealing butter, through his adulterous sexual affairs – to the advice he gives in the Mahābhārata itself. In the Gītā, Krishna tells Arjuna to kill his cousins and their armies because he should always do his duty (dharma) irrespective of the consequences. Even if one thinks this morally sound advice, the same Krishna later tells Arjuna to kill his rival Karna while Karna is fixing his chariot – an act that clearly violates all applicable rules of dharma – in order to achieve the consequence of winning the war. So too, it is Krishna who tells Yudhiṣṭhira to mislead Drona about Aśvatthāman the elephant, an act for which Yudhiṣṭhira later receives a karmic punishment – again, breaking the duty of truthfulness in order to bring about the best consequences. Krishna tells others to break the rules he himself sets out, and does so with impunity. Krishna’s bad deeds might not quite reach the scale of the Judeo-Christian God, but he is far from a moral paragon. He may be better than Pol Pot, but a human saint could surely outdo him.

So whether we are speaking of Vishnu or Jehovah, I do not think Augustine’s answer to human fallibility is acceptable. Perfect goodness is not to be found in men or in gods. But a chastened intellectualism without God seems to leave us with two unpalatable alternatives: a tyranny like Xunzi’s, or a life of miserable neurosis like Freud’s. I think this may be why Nietzsche and the existentialists view life without God as a terrifying (if perhaps ultimately fulfilling) “abyss”: if you don’t trust in God, you have to trust in man, and that’s not very comforting.

Or do you? I wasn’t thinking of it this way at the time, but I suppose all this might be part of the reason why, when I needed to pray, I turned to the bodhisattva Manjuśrī rather than to a God or Goddess as such. For Mañjuśrī, while perhaps omniscient, is not omnipotent. He lets much of the world suffer not because he chooses to – as God does – but because there’s too much he can’t prevent. A being who is omnibenevolent but not omnipotent – you can’t completely trust in such a being, because he might let you down; he can’t do everything. But if he exists – and maybe even if he doesn’t – he is at least more worthy of trust than either a human being or a creator God.

Sandel, however, claims that Kant’s position is not quite as counterintuitive as it seems. Sandel regularly makes this claim in his Justice course, which I taught for as a teaching fellow, and which Sandel has now made available to the public as a course as well as in a book. While Kant brooks no lies, Sandel says, he is quite happy with misleading truths. As evidence Sandel points to Kant’s own life:

Kant found himself in trouble with King Friedrich Wilhelm II. The king and his censors considered Kant’s writings on religion disparaging to Christianity, and demanded that he pledge to refrain from any further pronouncements on the topic. Kant responded with a carefully worded statement: ‘As your Majesty’s faithful subject, I shall in the future completely desist from all public lectures or papers concerning religion.’ Kant was aware, when he made his statement, that the king was not likely to live much longer. When the king died a few years later, Kant considered himself absolved of the promise, which bound him only ‘as your Majesty’s faithful subject.’ Kant later explained that he had chosen his words ‘most carefully, so that I should not be deprived of my freedom… forever, but only so long as His Majesty was alive.’ By this clever evasion, the paragon of Prussian probity succeeded in misleading the censors without lying to them. (Sandel, Justice, p. 134)

I was reminded of Sandel’s position recently while leafing through Śaṅkara‘s commentary on the Yoga Sūtras – specifically sūtra II.30, which speaks of the “restraints” (yama) that a yogin must practise, similar to the Buddhist Five Precepts and identical to the five Jain precepts. On both of these lists is satya, truthfulness. And in expounding the concept of truthfulness, Śaṅkara specifically warns us against the kind of misleading truths that Sandel describes.

Śaṅkara refers to a famous episode in the great Mahābhārata epic, an episode I greatly enjoy. The Pāṇḍavas – the Mahābhārata’s “good guys” – are at a stalemate against their Kaurava foes, because the Kauravas are led by the great general Drona. As long as Drona commands the Kaurava forces, the Pāṇḍavas cannot win. But the Pāṇḍavas have a plan, hatched by Krishna (Kṛṣṇa), that great trickster of a god. Drona cares more than anything for his son Aśvatthāman, and if he thinks Aśvatthāman is dead, he will lose the will to fight. But the Pāṇḍavas have no way to kill Aśvatthāman. What they can do is make Drona think Aśvatthāman is dead. Drona will believe Aśvatthāman is dead, Krishna notes, if he hears news to that effect from Yudhiṣṭhira, the leader of the Pāṇḍavas. For Yudhiṣṭhira, Kantian before the fact in his unswerving regard for the moral law, never tells a lie.

The problem, of course, is just that Yudhiṣṭhira never tells a lie! So he refuses to tell Drona that Aśvatthāman is dead – until Krishna comes up with a different plan. The Pāṇḍavas will kill an elephant named Aśvatthāman, and then it will no longer be a lie to say that Aśvatthāman is dead. It will merely be a misleading truth.

Yudhiṣṭhira agrees. The poor beast’s head is smashed in with a mace, and once Yudhiṣṭhira is within earshot of Drona the Pāṇḍavas call out “Aśvatthāman is dead!” Drona turns to Yudhiṣṭhira and asks, “Is this true?” And Yudhiṣṭhira calls out “Yes, it is true! Aśvatthāman _{the elephant} is dead” – adding the words “the elephant” in a voice too soft and hasty for Drona to have any chance of hearing it. Sure enough, Drona loses the will to fight, and the Pāṇḍavas begin to triumph.

Śaṅkara calls this story to mind in order to rule it out as an option for the yogin. The author of the Yoga Bhāṣya commentary says “The speech spoken to convey one’s own experience to others should not be deceitful”; to the end of this sentence Śaṅkara adds: “as when one states what one knows to be a fact, but this very truth is being spoken with the aim of tricking some other person. So Yudhiṣṭhira said ‘Aśvatthāman is killed – _{I mean the elephant.}‘” Even though it was for a good end, what Yudhiṣṭhira did was wrong; one should not do it.

Śaṅkara, then, explicitly rules out “misleading truths” as an option. Does Kant? I’m not sure. I didn’t buy the account when Sandel made it in class; it seemed to me that Kant was merely being a bad Kantian. But in his book Sandel explains in more detail. “A carefully crafted evasion pays homage to the duty of truth-telling in a way that an outright lie does not. Anyone who goes to the bother of concocting a misleading but technically true statement when a simple lie would do expresses, however obliquely, respect for the moral law.” More importantly, the misleading truth can be universalized. One would not be able to lie to the murderer if everyone told lies in such situations, for the claim could never be believed. But one could tell misleading truths if that’s what everyone did; it’s not that the claims would not be believed, but that people would have learned “to listen like lawyers and parse such statements with an eye to their literal meaning.” (Justice, p. 137) And because it can be universalized, one is treating one’s interlocutor as an end and not merely as a means; the misleading true statement “does not coerce or manipulate the listener in the same way as an outright lie. It’s always possible that a careful listener could figure it out.” (pp. 137-8) I’m still a little skeptical of the point, but Sandel might just be right. With these arguments, he’s managed to do what I thought was impossible: to defend Kant’s actions on Kantian grounds.