I do apologize for not being able to address everyone, and every theme, and I renew my thanks to all the commenters for their careful attention to the ideas and arguments I presented; even where I dispute the analysis, I have learned from the criticisms.

Reply to Kail: As befits a distinguished Hume scholar with (appropriate!) Nietzschean sympathies, Kail gives, in his first paragraph, an excellent summary of the position I was trying to defend, and states the central Nietzschean thesis well:  “The activity of the moral philosopher…is not a disinterested attempt to arrive at moral truth but a post-hoc rationalization of a general evaluative stance” arrived at because it is responsive to various psychological needs.  Kail then raises three issues.

First, he fairly asks what sense of objectivity is at stake in the argument for moral skepticism.  It is certainly not Mackie’s, which I agree is idiosyncratic.  The skeptic (or non-factualist, to borrow Loeb’s terminology) at issue here claims that there are no objective moral facts in the precise sense that there is no fact of the matter about what is morally right and wrong independent of the cognizing states of persons, even under idealized conditions of cognition.   Kantian constructivists and sensibility theorists will have quarrels with this, but let me assert, dogmatically and without argument (at least in this forum), that this skepticism about objectivity, if successful, reaches both the constructivists (who need a metaphysical view, their protestations to the contrary notwithstanding) and the sensibility theorists (whose view, even if correct, is tantamount to conceding that morality is not objective).  (I make the latter case in “Objectivity, Morality, and Adjudication,” reprinted in my Naturalizing Jurisprudence [OUP, 2007].)

Second, Kail articulates a concern raised by others (including Cholbi, Gowder, Zangwill and Donnelly), namely that “philosophical disagreement [might be laid] down to a distortion of moral phenomena that owes itself to the bias of philosophers.”  Nothing in the argument I have given rules that out, but this line of argument would require, first, abandoning the Parfitian optimism so characteristic of modern moral philosophy, and, second, some account of why it is philosophical methods of rational discursiveness are ill-suited to discerning the facts about morality.

Third, Kail correctly notes that the answer to the question what is the best explanation of intractable disagreement among moral philosophers is not one to be settled a priori, and will require more empirical, psychological details than I have offered so far.  We have a lot of empirical evidence about the sources of moral judgment—from Freud to Haidt—and it overwhelmingly fits the Nietzschean story.  But that story needs to be written in detail to be persuasive, as Kail rightly demands.

Reply to Clark & Dudrick: Clark & Dudrick deny that Nietzsche really argues from the fact of moral disagreement among moral philosophers to moral skepticism.  To be sure, they agree that Nietzsche rejects the arguments of moral philosophers, but this is “because their arguments fail.”  Why do they fail?  Clark & Dudrick, oddly, never say. To be clear, I think Nietzsche has other arguments for moral skepticism, besides the argument from moral disagreement:  for example, as I argued in Nietzsche on Morality [2002], Nietzsche holds the now familiar view that objective moral facts would not figure in the best explanation of our moral judgments.   But I do not see that Nietzsche actually directly engages the arguments of the major moral philosophers, and shows them to be dialectically unsound, and Clark & Dudrick adduce no contrary evidence.  But if Nietzsche has explanatory arguments for moral skepticism—including the argument from disagreement among moral philosophers—then we can see why he wouldn’t need to engage the details of, e.g., Kant’s argument for the categorical imperative.

Against the textual evidence I cite in support of my interpretation of Nietzsche’s position, Clark & Dudrick offer what we might call the ‘Argument from Semi-Colons’:  the fact that Nietzsche starts by appealing to the Sophistic “insight” that every morality can be “dialectically justified” and concludes that “it is a swindle to talk of ‘truth’ in this field”  does not support my reading because the various propositions are “separated by semi-colons”, which “opens the possibility that the claims of these clauses are simply a list of the things Nietzsche takes the Sophists to have claimed.  We need not think that Nietzsche any one of these claims as evidence for any other.”  Depending on one’s view of semi-colons, I suppose that is a possibility.  But to take it seriously, it would help to identify the actual argument for the skeptical conclusion, which Clark & Dudrick do not do.

Reply to Zangwill: Zangwill asserts that “Nietzsche consistently believed in objective orders of rank among human beings,” though he cites no texts in support.   The only supporting texts I can think of come from unpublished notes, and involve arguments so bad (as I discussed in “Nietzsche’s Metaethics:  Against the Privilege Readings,”European Journal of Philosophy [2000]), that one may hope Nietzsche recognized the position was not sound, or even consistent with his views about the objectivity of value in the books he published, and so left these thought experiments in his notebooks.   It is clear that Nietzsche has a higher opinion of Goethe, Napoleon, and Beethoven, than of the rabble and the herd, but much more is needed to show he thinks that judgments about value has any objective standing.  (I discuss these cases in some detail in Nietzsche on Morality [2002], esp. at 149-155.)

More interestingly, to my mind, Zangwill presses the important question (noted, in various ways, by several others) about “[w]hy…the academic philosopher’s position [is] somehow epistemologically privileged?”  As he notes, surely it “is not implausible that what is known as the ‘the ivory tower’ is a bad place for thinking about moral questions.”  Loeb’s comments bring out nicely why disagreement among moral philosophers is a particular embarrassment for moral realists, and I will not repeat what he says here.  I will only add that outside the ‘ivory tower,’ moral disagreement multiplies, and in some ways becomes more extreme, than it does within.  If the argumentative strategy of appeal to disagreement is sound, then going outside the disagreement among moral philosophers does not help the case of the moral realists.

Reply to Carson: Carson raises a number of pertinent issues, that I have dealt with in reply to others, so I will not rehash those matters here.  But he articulates an especially perspicuous version of another way in which my argument might be self-defeating (Donnelly also picks up on this). As Carson puts it:  “the existence of extensive disagreement about metaethical questions, including the question of whether there are objective moral truths, creates very serious problems for Leiter.  The logic of his view seems to commit him to saying that there are no objective truths about controversial metaethical questions and that there are no objective (metaethical) facts about whether or not there are objective moral facts.”   Here I will invoke something Loeb notes:  “We have to look at the total explanatory picture.”  The question is always what is the best explanation for intractable disagreement.  In the case of intractable metaethical disagreement, at least part of the explanation may be that those who are especially sanctimonious feel that their moral attitudes must have objective standing, and so they will not give up their metaethical realism no matter what.   In this respect, I suspect moral realists are a bit like smart philosophical deists, who can deploy a vast dialectical armory in support of their preferred metaphysics.  But this is speculative, and I concede I have not made the case, and Carson’s kind of self-defeatingness looms over the argument.

Reply to O’Brien: O’Brien rightly calls attention to the fact that philosophers like Anscombe and MacIntyre have endorsed, or noticed, similar facts about the intractability of disagreement among modern moral philosophers.  I am skeptical the story would become more hopeful if we included the pre-modern moral philosophers in the mix—the disagreement, both in breadth and depth, would only grow worse.  And I suspect the skeptical case would be even easier to make given that the pre-modern moral philosophers, and their nostalgic contemporary defenders like Anscombe and MacIntyre, rely on theistic and metaphysical suppositions that are discredited on independent grounds.

Reply to Donnelly: I have dealt with part of Donnelly’s objection in reply to Carson.  Donnelly correctly sees that I am committed only to the thesis of “moral anti-realism:  it is not the case that there are any moral facts.”  He wonders whether this is compatible with the view, defended by Pigden, among others, that Nietzsche is an error theorist.  I do not think Nietzsche is an error theorist, or a fictionalist, or a non-cognitivist, because I do not think he has any considered view about the semantics of moral expressions—indeed, it would be astounding if he did.  My critical discussion of Hussain’s fictionalist reading (which also requires that Nietzsche be a cognitivist about moral judgments) gives some indication of why I am an agnostic about Nietzsche’s moral semantics.

Reply to Loeb: I have no reply to Loeb, since everything he says strikes me as right and instructive.  I would encourage the critics of the argument at issue here to read Loeb’s comments with care.

Reply to Wong: Wong’s interesting response draws on his own development of a sophisticated form of moral relativism, which I can not hope to address adequately here.  I concur with Wong that differences in “cognitive styles…might partially account for differences between moral philosophers,” but I think, to the extent it plays an explanatory role, it will license the moral skepticism I take Nietzsche to be defending.

Reply to Langsam: Langsam, like Clark & Dudrick, suggests that Nietzsche does not need to appeal to the fact of intractable disagreement among moral philosophers “to discover the inadequacies in the arguments of moral philosophers.”  But his textual evidence, which overlaps with mine, only states Nietzsche’s conclusions about the inadequacies not the argument for why they are inadequate.

Langsam, echoing some earlier commenters, suggests that it is an “undisputed fact…that there is a consensus in certain societies with respect to certain nonfoundational moral issues,” and he adds that this “common sense morality” itself requires explanation.  He admits that “there is much moral disagreement in our societies, but there is much agreement, also.”  Given that caveat—that there is both disagreement and agreement—then I quite agree that there is an explanatory burden with respect to the latter as well as the former.  Of course, Nietzche’s explanation for the agreement is also not one that vindicates its factual status.  If the appeal of ascetic moralities is explicable in terms of their ability to seduce back to life those who suffer, that hardly shows moral judgments to be reliable trackers of moral reality.

Langsam also observes that Nietzsche’s “kind of psychological explanation of a philosopher’s view does not discredit it.”  Everything turns on what is meant by “discredit.”  Certainly Nietzsche thinks, as Langsam notes, that genuine philosophers are affectively engaged partisans, a role that philosophers usually disown.  So insofar as Nietzsche correctly diagnoses their real motives, he discredits their views by the very standards philosophers accept.  More importantly, the best explanation of the views of the philosophers discredits their objective standing.  Nietzsche, of course, thinks philosophical views are none the worse for that, but that is hardly a view the targets of his criticsm subscribe to.

I can not resist concluding by noting that no one gives a more delightful and evocative reading of Nietzsche than Harold Langsam, so if you have the opportunity, ask for a recital!

Reply to Pigden: I hesitate about using the label “error theory” for Nietzsche’s view, for reasons already noted in response to Donnelly.   But with that caveat, I am otherwise comfortable with Pigden’s restatement of the core argument.  Unfortunately, it seems to me that in his other examples, Pigden does not take seriously enough the requirement of his premise (2), namely, that the “best explanation” of the intractable disagreement includes the absence of any fact of the matter about which the parties are disagreeing.  Pigden deems it “crazy” and “patently absurd” to think there is no fact of the matter about free will, but we first need to settle what the best explanation for the persistence of debates about free will really is; Pigden does not even consider the question.  I have no quarrel with Pigden’s alternate argument for moral skepticism, though appeal to moral “perceptions” does open the door to the standard defusing explanations of disagreement in a way that appeal to disagreement among moral philosophers does not (again, videLoeb on this point).

Brian Leiter
University of Chicago

I am grateful for the overwhelming response to my essay, and to the multitude of interesting, illuminating and often important objections and observations offered by the nearly thirty readers who took the time to comment, often at length. I want to especially thank “On the Human” for making a forum like this possible—I am sure I never would have gotten so much valuable feedback from so many philosophers and scholars had it not been for this opportunity.

Since my essay of three pages generated over thirty pages of comments and discussion, I hope I will be forgiven for not responding to every issue raised. Don Loeb, in his comments, make many of the points I think are relevant, so I commend those to readers. Herewith my first set of replies to an excellent set of critics; the second set of replies will follow next week.

Reply to Wedgwood: Wedgwood grants that “moral philosophers disagree about foundational moral questions,” but denies that they disagree about “more concrete moral issues,” such as: “we are normally obliged to keep our promises, to refrain from telling lies and killing or torturing people, and to help those who are in need.” I, of course, conceded that moral philosophers often do agree about concrete cases, but what kind of agreement is this really? Importantly, Wedgwood has to preface each of these concrete moral propositions by saying this is what we “are normally obliged to” do, meaning, of course, that he recognizes that each moral precept has exceptions—and as soon as we press on the exceptions, we start to find the real disagreements about the foundational moral questions. I suppose no one would think that Mussolini and Roosevelt really converge on the same moral truths just because they both agree about the concrete question that normally the trains should run on time. Moral philosophers—at least the conventional kind who subscribe to the propositions in question—are surely less far apart than Mussolini and Roosevelt, but that doesn’t alter the fact that their apparent agreement on suitably general and hedged “concrete” moral propositions belies real disagreements, which come out as soon as we press on the concrete cases.

But Wedgwood also wonders whether the foundational moral disagreements are real. He suggests that most moral philosophers may not be “really convinced of the precise moral theories that they defend.” Perhaps this is right, but there is no evidence of this in the writings of most moral philosophers (with the exception, ironically, of Nietzsche).Wedgwood’s speculative hypothesis seems more wishful, than factual, but maybe he has evidence in mind of which I am unaware.

Wedgwood also raises doubts about whether intractable disagreement supports an inference of anti-realism about the subject of the disagreement. I would plead guilty to the charge of “Whiggishness” if it were my thesis that “sufficiently intelligent and conscientious inquirers” (to quote Wedgwood) will always discover the facts. But that is not my thesis. My thesis is that failure of convergence demands explanation, and sometimes that explanation will invoke the non-factual character of the disagreement—but other times it will be the absence of pertinent evidence, or the irremediable cognitive deficit of one of the parties to the disagreement, and so on. In the case of the debates among historians that Wedgwood fairly notes, surely an important part of the explanation for their persistent disagreements have to do with evidential problems, both what would count as decisive evidence for one proposition or another, and also our access to pertinent evidence. In the case of debates about the truth of string theory, the lack of pertinent evidence is the central problem, I thought, due to lack of equipment to test the predictions of the theory. Perhaps I have been misled by Steven Weinberg’s work!

Wedgwood concludes by suggesting the self-defeating character of my argument is much worse than I concede. Wedgwood thinks that careful philosophical attention to the metaphysical import of intractable disagreement has a longer history than I do. Here I take some comfort from having Paul Bloomfield, another moral realist, on my side—he notes: “I don’t recall reading any discussions about moral disagreement so precise and technical from the distant past.” Wedgwood is right to identify a debate on this issue in antiquity, and perhaps the spirited exchange here will propel the question to philosophical center stage, such that in ten or twenty or thirty years we will have a clearer handle on whether the disagreement is intractable and, if so, why. But I still think it is too soon to draw any inferences from the state of philosophical play at present.

Reply to Katsafanas: Katsafanas’s reconstruction of the argument usefully illuminates a point on which my original presentation no doubt created some confusion. As he puts it, the datum for which I purportedly demand an explanation is that: “ (1) there exist incompatible X theories providing dialectical justifications for incompatible X propositions.” Katsafanas agrees that it could be true that one can construct “ (2) apparent dialectical justifications for incompatible X theories” in terms of “(3) the agent’s psychological needs” to accept “a particular X theory,” but he denies this shows that there would be “no objective facts about X propositions.” He gives some nice examples—involving economic disputes—to illustrate the points, and concludes that “it is hard to see why (2) and (3) have any bearing on whether there are objectives facts about the topic in question.”

I agree that moral skepticism would not follow from (2) and (3), but that was not my argument. Michael Ridge (who has his own objections) gives a cleaner statement as follows: “(1) There has been no convergence on any substantive, foundational proposition about morality under what appear (and purported) to be epistemically ideal conditions of sustained philosophical inquiry and reflective contemplation across millennia; (2) The best explanation for such a failure in convergence…is that there are no objective moral truths; (3) Therefore, there are no objective moral truths.” I shall return, below, to Ridge’s different objections to this argument, but I believe he correctly capture the dialectical structure of the inference to the best explanation argument at issue (as does Loeb). Katsafanas, I think, misconstrues the explanandum: it is not that there exist incompatible X theories, but that there is no convergence in opinion among epistemically well-situated disputants about which X theory is true. It is that fact which demands explanation, and the proposal on offer is that there is no moral truth for an X theory to discover, which then facilitates the construction of “apparent dialectical justifications” for different theories.

The case of theological rationalizations for the existence of God is not, contrary to Katsfanas, an analogous case. It is not true that the non-existence of God “remains undetected by a large number of people who are otherwise rational and epistemically informed.” In fact, the overwhelming tendency of post-Enlightenment intellectuals for the last two hundred years is towards atheism, and the odd exceptions (e.g., certain Protestant American philosophers) admit of obvious, often embarrassingly obvious, explanations that are compatible with the non-existence of a deity.

Reply to Young: Young invokes what he takes to be obviously true concrete moral judgments to resist the anti-realist conclusion. His examples are like Wedgwood’s, and also Bloomfield’s (discussed below), and I will not repeat here what I say in reply to those philosophers. Young also suggests that to infer moral skepticism from intractable disagreement among moral philosophers “would be as fallacious as arguing from the fact that sometimes we can’t decide whether a man is bald or hairy to the conclusion that there is never a ‘fact’ of the matter as to whether someone is bald or hairy.” But Young’s analogy is confused. The actually analogous argument would not be about whether there are any facts about when any individual is bald or hairy, but about whether there is a fact about the baldness of some particular person, the one whose “baldness” or “hairiness” is in dispute. As I am sure Young knows, metaphysical skepticism about whether there is a fact of the matter about the application of a vague predicate in some instance is (in fact!) one of the positions widely defended about cases like this.

Reply to Knobe: Knobe does not dispute the argument on offer, but offers the possibility that “whatever methods moral philosophers have been using, these methods do not in any way help; people to converge on” the moral facts. I fully agree that this conclusion is compatible with my argument, and that one upshot of the argument (if one accepts it) is that the dialectical methods of moral philosophers are not the way to secure convergence in judgment.

Reply to Gowder: Like Cholbi, Gowder is sympathetic to a different explanation for intractable disagreement among moral philosophers: namely (to put this in my terms) that it reflects a kind of cognitive or epistemic defect—in Gowder’s terms, “the corruption of our capacity to make moral judgments.” What are the corrupting influences? Gowder suggests corruption of moral judgment by “self-interest, ideology (in the Marxist sense), by indoctrination and socialization, and most generally, by motivated cognition appearing in a multitude of forms.” I am sympathetic to the thought that moral philosophers are, in fact, subject to these influences, but it is not obvious that they will suffice to explain why bourgeois utilitarian moral philosophers like Parfit can’t agree with bourgeois Kantian moral philosophers like Korsgaard. We would need a more fine-grained explanatory story showing how these ‘corrupting’ influences affect the actual disagreements we find. And we would also need some reason for thinking that minus these corrupting influences, philosophers would converge on the moral facts.

Reply to Harman: Harman fairly asks for clarification about what is meant by “foundational propositions” about morality. I took foundational moral propositions to be those advanced by, e.g., deontological or utilitarian theories which specify the criteria in virtue of which “concrete” moral judgments are warranted: so, e.g., “it is wrong to break this promise” is a concrete moral judgment, while “the wrong-making feature of an action is its effect on utility” is a foundational proposition. In the sciences, “foundational” propositions would be, e.g., laws of nature in virtue of which particular true predictions are made. “Instrumentalist” versus “realist” interpretations of physics (per Harman’s example) would not be an example of disagreement about foundational propositions, whereas disagreement about whether the laws of Newtonian mechanics correctly describe the motion of mid-size physical objects would be (but there is no disagreement about the latter).

I claimed that “the disagreements of moral philosophers are amazingly intractable. Nowhere do we find lifelong Kantians suddenly…converting to Benthamite utilitarianism, or vice versa.” I do not see Harman’s purported counter-examples as probative. Utilitarianism and libertarianism (the Narveson example) are not opposite moral doctrines, but perhaps there are details of Narveon’s moral theory of libertarianism I am missing. Nor are Benthamite and Millian utilitarianism; if that is the most “dramatic” conversion story on offer, it seems to me to confirm the original hypothesis. Moral vegetarianism seems to me less clearly a foundational moral judgment, as opposed to a concrete or applied one. There is some evidence that philosophers, especially some utilitarians, have become vegetarians because of the force of utilitarian arguments, but it was not any part of my claim here to deny that foundational moral principles might influence concrete judgments. (As a side note, it is interesting that even Peter Singer has conceded that the real impact of Animal Liberation came from the description of factory farming, not the arguments he (and others) have given for vegetarianism. That the foundational utilitarian arguments have had little or no impact—as distinct from the ability of vivid descriptions of factor farming practices to arouse emotional responses that change behavior—should be clear from the fact that the same foundational premises lead to Singer’s notorious views about the permissibility of killing the severely disabled, views which have found almost no acceptance, even among the very same people who consider themselves ‘moral’ vegetarians.)

Reply to Berry: Berry suggests that it is more fruitful to think of Nietzsche as a “skeptic” in the Pyrrhonian sense, as someone who suspends judgment, rather than as someone affirmating an anti-realist metaphysical thesis. Her argument here is developed at greater length in her forthcoming book on Nietzsche and Ancient Skepticism (Oxford University Press, 2010), and so a full adjudication of its merits will have to await that book’s appearance. I will note that the textual burden for Berry’s view is heavy, considering Nietzsche’s penchant for strong statements (e.g., “there are altogether no moral facts” or “Christianity has not a single point of contact with reality”), which do not look like suspensions of judgment at all. She makes an intriguing case for the Pyrrhonian reading of one of the Nachlass passages on which I rely, though I am not convinced by her claim that the Sophists did not have philosophical views about, e.g., the objectivity of ethics. Plato certainly thought they did, and argued against them. The Sophistic way of opposing Platonism involves precisely a dispute about the objectivity of value, in a way the Pyrrhonian response does not.

Reply to Ridge: I have already availed myself (in the reply to Katsafanas) of Ridge’s crisp statement of the core skeptical argument. Ridge seizes upon my formulation of the position in the original posting in terms of there being “no objective moral truths,” and correctly notes that quasi-realists, given that their view that “truth is a mere grammatical device,” can both explain moral disagreements and retain the idea that there are moral truths. Truth is obviously a semantic notion, and my unqualified use of it in the original posting invites Ridge’s objection. But I did use “truth” interchangeably with “facts” for a reason (one that is clearer in the longer unpublished essay on which this is based): namely, that the skeptical thesis I am abscribing to Nietzsche is a metaphysical one (about what exists) not a semantic one. As a general matter, I do worry that the semantic tail too often wags the metaphysical dog in metaethical disputes, but whether or not that is right, it is clearly anachronistic to think Nietzsche has a view about the subtle semantic issues at stake between cognitivists and non-cognitivists, between emotivisits and quasi-realists, and so on. What matters for purposes of my argument is that the quasi-realist stands with Nietzsche (and Mackie and Gibbard and Stevenson) on the metaphysical question, i.e., that there are no moral facts.

Dancy’s particuralism raises a different kind of challenge, as Ridge interestingly notes. For the particuralist has no difficulty with the phenomenon of intractable disagreement among moral philosophers about foundational moral principles, since he is, like Nietzsche, a skeptic about the existence of the latter! Moral reality, however, is supposedly captured in the particularist and context-sensitive judgments, and so is untouched by the Nietzschean argument. That seems right, as far as it goes. But appeal to particular and context-sensitive moral judgments runs afoul of the more familiar anti-realist arguments that appeal to “folk” disagreement and disagreements between cultures. It may be, as Ridge says, that particularism “cannot reasonably be dismissed out of hand”—which positions can be, I wonder?—but absent a particularist explanation for the diversity of particular moral judgments, it is hard to see it being much help in resisting an argument from disagreement.

Ridge, playing devil’s advocate (since this is not really his view), borrows a page from Parfittian optimism, claiming that “until the twentieth century, avowing a purely secular moral theory was hazardous to your health.” The suggestion is that, therefore, we need to wait longer for the promised progress in moral theory. This is a common, though, curious refrain, and for three reasons. First, most fields with factual subject matters have usually managed to make progress, as measured by convergence among researchers, over the course of a century—and especially the last century, with the rise of research universities. Moral theory is, again, the odd man out, when compared to physics, chemistry, biology, or mathematics. Even psychology, the most epistemically robust of the ‘human’ sciences, managed to make progress: e.g., the repudiation of behaviorism, and the cognitive turn in psychology in just the last fifty years. Second, Spinzoa and Hume and Mill and Sedgwick may not have advertised their secularism, but the idea that their moral theories are for that reason discontinuous with the work of the past hundred years does obvious intellectual violence to the chains of influence of ideas and arguments. Third, and relatedly, so-called “secular” moral theory regularly conceives itself in relation to a history that stretches back in time (some times back to the Greeks), so that it becomes unclear why the bogeyman of the deity was supposed to have constituted the insuperable obstacle weighing down intellectual progress. Most contemporary deontologists may be atheists, but it isn’t obvious that their atheism enabled them to make stunning intellectual progress beyond Kant. (I also do not doubt that Parfit thinks all moral philosophers are converging on his view, as Ridge notes, but to call his claim “controversial” is, I would think, a significant understatement. )

Ridge is skeptical about whether all metaphysical and epistemological views can be traced to moral commitments of their proponents. In fairness to Nietzsche, I should note that his thesis applied to “great philosophies”—like Kant and Spinoza—and not to those “philosophical laborers” and “scholars” who possess “some small, independent clockwork that, once well wound, works on vigorously without any essential participation from all the other drives of the scholar” (Beyond Good and Evil, 6). Many professional philosophers may, indeed, be laboring away at problems in a “disinterested” way. Still, as the recent Chalmers survey of philosophers brought out, there are striking, and surely not accidental, correlations between philosophical views across different areas: e.g., theism and moral realism and libertarianism about free will. Even the “philosophical laborers” are not wholly disinterested inquirers!

Finally, Ridge makes the fair point that anti-realism about morality “need not imply the abandonment of serious first-order inquiry.” Ridge no doubt correctly describes the view of many moral anti-realists, but it is an important, and separate, question, whether they are right to be so confident. Ayer and Stevenson both thought, of course, that there was lots of “serious first-order inquiry” to be done about non-moral facts and logic, inquiry which might influence our moral attitudes. But have we really established that Ayer was wrong “that argument is possible on moral questions only if some system of values is presupposed”? As Ayer observes: “It is because [rational] argument fails us when we come to deal with pure questions of value, a distinct from questions of fact, that we finally resort to mere abuse.” That certainly seems to me an apt description of the state of foundational disagreements between Kantians and utilitarians.

Reply to Suikkanen: I am grateful to Suikkanen for posing the question of how the metaethical view defended here squares with my expression of moral and political opinions on my blog. Again, following Ayer and Stevenson, it is important to distinguish between genuine moral disagreements, and disagreements about non-moral facts. Especially in the reactionary and cognitive impaired political culture of the United States, many political disagreements really trade on ignorance of, or indifference to, non-moral facts. So, to mention Suikkanen’s example, most disagreement about the teaching of intelligent design trades on non-moral ignorance about the evidence for natural selection and its ability to explain complex organic phenomena. If someone agreed with me that “Intelligent Design” were simply creationism for those who had consulted a lawyer and a public relations consultant, but still thought it should be taught in the public schools, then our disagreement would be a “purely” moral or evaluative one. I do not know what rational considerations could adjudicate that dispute. But my “pure” evaluative commitments (to borrow Ayer’s phrase) are none the worse for being mine—indeed, I’m rather attached to them!

Reply to Bloomfield: Like some other commenters, Boomfield would locate the evidence for moral realism in “our engaged moral experiences.” That, of course, invites the skeptical rejoinder that people’s moral experiences differ quite a bit (at least prima facie), and that the experiences in question have dubious epistemic credentials (think of “engaged religious experiences,” which are no doubt also vivid to those who have them). Bloomfield, however, borrowing from Sharon Street, adduces examples of moral facts on which there is purportedly convergence—such as “The fact that something would promote the interests of a family member is a reason to do it” and “The fact that someone is altruistic is a reason to admire, praise, and reward him or her”—and suggests that the fact that “moral philosophers cannot agree on a justification for how they base their judgments and actions on these facts is an embarrassment to moral philosophy, but not to morality itself.” As I noted above, in reply to Wedgwood, this apparent convergence is only apparent, since the moral ‘truths’ in question are either reasonably disputed (La Rochefoucauld and Nietzsche reject the second one!), or need to be severely qualified (it is isn’t a reason to hide an escaped prisoner just because he is your brother, is it?) in ways that will implicate broader principles, precisely the terrain on which moral philosophy operates. Again, apparent agreement on concrete moral judgments can easily belie disagreement on foundational moral principles.

Like Harman, Bloomfield invokes disagreements about the correct metaphysical interpretation of science, though rightly concedes that “even if the philosophy of science is problematic this doesn’t imply that science itself has any worries.” (I note in passing that the right analogy to this point would not be, as Bloomfield suggests, how moral philosophy stands to ordinary morality, but how meta-ethics stands to ordinary morality.) He goes on to claim that “there is little agreement on the fundamental issues of science,” but the examples he gives, in fact, seem to involve metaphysical disputes about science, not debates internal to scientific practice. Perhaps the point trades on the ambiguity about “foundational” to which Harman called attention, and on which I remarked in reply to him.

Is Nietzsche a “moral philosopher” like those at whom he “hurls…insults” as Bloomfield amusingly observes at the end of his comments? I think the answer is actually ‘no,’ for two reasons: (1) he does not present his basic ethical views as having any rational warrant—he is self-conscious about their lack of objective standing; and (2) he does not suppose that rational arguments could change anyone’s ethical views. In the spirit of Ayer, however, he is rather good at “abuse” and polemics, precisely because those rhetorical devices are far more likely to have an impact on the emotive and non-rational foundations of our moral convictions.

The tricky part is the analogy, which is obscured a bit by Gray’s statement, “He does think there is a fact concerning morality — There is no right or wrong at all.”

What is a “fact concerning morality”?

The phrase could refer to a moral fact—like that killing people for fun is wrong or that utilitarianism is (in)correct.

Or it could refer to a metaethical fact—like the fact that moral realism is wrong (or right) or that Brian’s argument against moral realism succeeds (or does not succeed).

The distinction is crucial. Leiter argues from there being persistent disagreement about how to answer a question to there being no answer to the question. Specifically, he argues from persistent disagreement among philosophers over substantive moral questions to the metaethical claim that there are no moral facts–moral anti-realism. Call this, “Leiter’s inference”.

Analogously, one might try to reason from persistent disagreement among philosophers over whether or not there is free will to there being no fact about whether or not there is free will. Likewise, one might reason from persistent disagreement among philosophers over the whether or not moral realism is true to there being no fact about whether or not moral realism is true.

But the conclusions of these analogous arguments are implausible, say the critics, and so we should resist Leiter’s inference (from disagreement about what the moral facts are to the claim that there are no moral facts) as well.

Indeed, they claim, Leiter’s argument is self-defeating, since we could as well argue from disagreement among philosophers about the correctness of Leiter’s inference to there being no fact about whether Leiter’s inference is correct.

Leiter responds that we need to look in each case for the best explanation for persistent disagreement. The best explanation for persistent disagreement (among philosophers) about substantive moral questions, he thinks, is that these questions do not have correct answers.

In the case of disagreement about the correctness of Leiter’s inference, however, Leiter thinks that the disagreement is too new to count as persistent. Thus, he suggests, the best explanation for that disagreement does not suggest that there is no fact about whether Leiter’s inference is correct.

In other cases, where disagreement about certain philosophical questions does appear to be persistent and irresolvable, the skeptical conclusion that these philosophical questions have no answers might well be correct, he says.

Here are the relevant passages by Leiter:

That still leaves a slightly different version of the worry that the argument “proves too much.” For surely most philosophers will not conclude from the fact of disagreement among moral philosophers about the fundamental criteria of moral rightness and goodness that there is no fact of the matter about these questions, as I claim Nietzsche does.  But why not think that this meta-disagreement itself warrants a skeptical inference, i.e., there is no fact about whether we should infer moral skepticism from the fact of disagreement about fundamental principles among moral philosophers, since philosophers have intractable disagreements about what inferences the fact of disagreement supports?

Even if, after extended critical discussion, the meta-disagreement continues to persist, that still would not support the meta-skeptical conclusion that there is no fact of the matter about whether or not disagreement in foundational moral theory supports skepticism about moral facts.

Leiter makes it clear that he does think there is a true position concerning meta-ethics. Anti-realism is true.

Harold Langsam confirms this reading of Leiter:

He denies that we should infer from this meta-disagreement that there is no fact of the matter about what we should infer from the original disagreement. On the contrary, he suggests that the moral skeptic can legitimately argue that he is correct and the moral realist is incorrect and there is a good explanation of why the moral realist persists in his incorrect views.

Here are the passages by Charles Pigden and Michael Ridge that I find confusing or mistaken:

Charles Pigden:

The conclusion I suggest is crazy. How could there be no fact of the matter about whether determinism of one of its rivals is true? I suppose a radical Dummettian or an extreme postmodernist might flirt with such a radically anti-realist line of thought, but that way madness lies. On this issue there just has to be a way the world is. How about this argument (call it L/N**)?

This argument does not seem analogous. If anything, an analogous argument should be — Philosophers have always disagreed about foundational questions concerning free will, so we probably don’t have free will. An anti-realist position concerning free will should be supported by the argument.

Michael Ridge:

Even in those cases in which Leiter’s hypothesis has some plausibility, it seems a bit much to jump from that hypothesis – that the debate is driven by the moral agendas of the partisans – to the conclusion that there is no fact of the matter as to who is right in the debate in question. Suppose that the debate over mind/body dualism is driven by the moral agendas of the partisans. Should we infer from this that there is no fact of the matter as to whether mind/body dualism is true? Again, this seems highly implausible.

Yes, there is a fact of the matter about whether mind/body dualism is true. Leiter would probably want to side with anti-realism about mind-body dualism and say it’s probably false.

I agree with Brian Leiter about three things:

A) Nietzsche was a radical moral skeptic or, as we would nowadays say, an error theorist about ethics (a view he may have derived from Max Stirner).
B) The error theory is correct. There are no moral facts, or more precisely, no non-negative atomic moral facts, facts constituted by a specific item or person possessing a specific moral property (such as goodness) or standing in a specific moral relation to something else (such as being obliged to bring it about). And the reason for this is that the are no such properties or relations.
C) Something can be made of the ‘Argument from Relativity’ – that is, the prevalence of moral disagreement affords some support for the error theory.

However I don’t think that Leiter’s Nietzsche-derived version of the argument will do the trick. Here is Leiter’s argument in a nutshell (let’s call it L/N):

(1) As of now, there are extremely persistent disagreements between rival moral philosophers giving radically different and incompatible accounts of the moral facts.
(2) The best explanation of (1) is not an epistemic deficiency on the part of at least some of the rival philosophers – a deficiency which deprives all or most of them of access to the facts – but the thesis that there are no moral facts to disagree about.
So [probably]
(3) There are no moral facts to disagree about.

Now of course, this is not supposed to be a deductive argument. Premises (1) and (2) could be true – at least (1) could be true (2) could be true in the sense that the non-existence of moral facts could apparently be the best explanation of persistent disagreement – and (3) could still be false. (If an explanation is false there is a sense in which it is not really the best explanation though it may be the best as judged by conventional epistemic criteria.) Neither Nietzsche nor Leiter supposes that the persistence of moral disagreement among philosophers and the apparent excellence of the error-theoretic explanation necessitates the non-existence of moral facts. The facts could be lurking about inaccessibly (so to speak) in which case the interminable disputes would be due to the epistemic fact that the moral facts are difficult to access. But the error-theoretic explanation is a better explanation than the rival explanations of the moral realists because none of the rival explanations of the moral realists can convincingly explain why, if the basic moral facts are – and can be known to be – of kind K, they can have been so widely believed to be of kinds X, Y and Z.

Is L/N a good argument? I think not. For there are parallel arguments with equally plausible premises (or premises which would have been equally plausible when they might have been put forward) leading to patently absurd conclusions. Let’s start with L/N*:

(1*) As of now, there are extremely persistent disagreements between rival philosophers giving radically different and incompatible accounts of the facts about free will (that is whether determinism or some version of the liberty of indifference is true).
(2*) The best explanation of (1*) is not an epistemic deficiency on the part of some of the rival philosophers – a deficiency which deprives all or most of them of access to the facts concerning free will – but the thesis that there are no facts about free will to disagree about.
So [probably]
(3*) There are no facts of the matter with respect to free will. There is simply no answer to the question of whether determinism or one of its rivals is correct.

The conclusion I suggest is crazy. How could there be no fact of the matter about whether determinism of one of its rivals is true? I suppose a radical Dummettian or an extreme postmodernist might flirt with such a radically anti-realist line of thought, but that way madness lies. On this issue there just has to be a way the world is. How about this argument (call it L/N**)?

(1**) As of now [1600] there are extremely persistent disagreements between rival philosophers giving radically different and incompatible accounts of the large scale structure of the universe.
(2**) The best explanation of (1**) is not an epistemic deficiency on the part of at least some of the rival philosophers – a deficiency which deprives all or most of them of access to the facts – but the thesis that there are no facts about the large scale structure of the universe to disagree about.
So [probably]
(3**) There no facts about the large scale structure of the universe to disagree about.

Here again the conclusion is crazy. And it would have been equally crazy in 1600. How could there be no facts of the matter about the large scale structure of the universe? There has to be a way the world, at large, really is, even if it is forever inaccessible to human enquiry.

I could multiply examples but you get the point. If parallel arguments from plausible premises lead to crazy conclusions, there has to be something wrong with Leiter’s argument.

But perhaps I have been cheating. With L/N* and L/N** it is scarcely conceivable that there could be no facts of the right kind. But it is perfectly possible that there might be no moral facts so long as we understand such facts in something like my sense as facts constituted by a specific item or person possessing a specific moral property or standing in a specific moral relation to something else. Thus a better parallel to Leiter’s argumet would be the following (L/Nt):

(1t) As of now, there are extremely persistent disagreements between rival theologians giving radically different and incompatible accounts of the divine facts (where a divine fact is a fact constituted by an item’s possessing a godhead-conferring property).
(2t) The best explanation of (1t) is not an epistemic deficiency on the part of at least some of the rival theologians – a deficiency which deprives all or most of them of access to the facts – but the thesis that there are no divine facts to disagree about.
So [probably]
(3t) There are no divine facts to disagree about.

The argument may be dodgy and the conclusion debatable but at least it isn’t crazy or inconceivable. We can deny divine facts without absurdity. This suggests a reply on behalf of Leiter. L/N style arguments are only cogent if it is conceivable that there are no facts of the kind that are denied in the conclusion. But it is quite conceivable that there are no moral facts (construed as non-negative atomic moral facts). Hence my ‘parallel‘ arguments L/N* and L/N** are not really parallel to Leiter’s L/N and thus do not discredit its cogency.

However, I still think it is a bad argument. Suppose that human beings never make it out of this solar system (a plausible scenario given that our civilization is likely to collapse owing to global warming). And suppose we never discover a method which we know to be reliable for detecting earth-type planets orbiting other stars. Then the following would be an argument with true premises and an obviously silly conclusion (L/Ne):

(1e) As of now, there are extremely persistent disagreements between rival astronomers giving radically different and incompatible accounts of the other-earth facts (where an other-earth-fact is a fact about the number > 1 of earth-type planets not identical with the earth in the Milky Way).
(2e) The best explanation of (1e) is not an epistemic deficiency on the part of at least some of the rival astronomers – a deficiency which deprives all or most of them of access to the facts – but the thesis that there are no other-earth facts to disagree about.
So [probably]
(3e) There are no other-earth facts to disagree about.

Here the conclusion is equivalent to the quite conceivable claim that there are no earth-type planets in the Milky Way besides the earth. Thus the argument L/Ne is a genuine parallel to L/N. But though quite conceivable the conclusion of L/Ne is wildly unlikely. Hence L/Ne is not a good argument. And since L/Ne is parallel to L/N, L/N is not a good argument either.

RELATIVITY REFURBISHED
Here’s a characteristic pattern in human enquiry: broad agreement with respect to the evident and the observable, disagreement with respect to the theoretical and the occult. Everybody agrees that the sun seems to rise in the East, to proceed across the sky and to set in the West, but there has been an enormous amount of disagreement about the explanation of these evident facts. When this happens we may have reason to be suspicious of any given explanatory theory but we have no reason to deny the existence of any explanatory facts. For the best explanation of the persistent disagreement may simply be that the explanatory facts are difficult to discern. Explaining the evident and the observable is often an extremely challenging task. If the world were transparent to us there would be no need for science. A moral realist might maintain that the situation in ethics is similar to that which existed before science really got going in the seventeenth century. There is a lot of agreement with respect to the ‘appearances’ but a lot of disagreement with respect to the explanation of those appearances. In the course of enquiry the correct explanation (whatever that turns out to be) will eventually win out. Our problem is that the moral sciences have not really gotten into their stride.

But is this really the way it is in ethics? It seems to me that we have disagreement both with respect to the ‘evident and the observable’ and with respect to the theoretical and the occult. I need not enlarge on the disagreement with respect to the theoretical and the occult since that is the theme of Leiter’s paper. But there is also disagreement with respect to the ‘evident and the observable’. People disagree about the rights and wrongs of slavery and of war, about marital obligations and women’s rights (if any). Lots of people don’t hold it to be self-evident that all men (and women) are created equal and are endowed by their creator with certain inalienable rights We should not forget that genocide and state terror (not to mention the terroristic actions of non-state agents) are typically perpetrated by conscientious agents, people who really believe that they are doing the right thing. Slavery without guilt was once the order of the day and it is only relatively recently that we have developed a conscience about such things. Ethnic cleansing is now widely – though not universally – regarded as a Bad Thing though in New World societies most of us are the beneficiaries of acts of ethnic cleansing. Thus there is an enormous amount of disagreement about the observable ‘facts’ to be explained (the appearances to be ‘saved’) as well as about the theoretical explanations. Of course, some of this moral disagreement can be put down to differences as to the non-moral facts on which morality partly supervenes. If you believe in God, then what he wants has implications for ethics. And there is a lot of disagreement about which God exists and what it is that he wants. But there is also a large residue of moral disagreement that cannot be put down to differences as to the non-moral facts.

Let’s consider a relatively trivial and hence ungruesome case of moral disagreement. G.E. Moore thought it self-evident that ‘self-abuse’ is intrinsically bad. Many of his fellow Apostles thought otherwise. (‘The Apostles’ was the nickname of a secret Cambridge debating society of which Moore, Russell, Lytton Strachey and Keynes were members.) On both sides this was regarded as a matter of immediate perception not the outcome of a theoretical argument. Thus each thought the other had misperceived the moral facts. How was this possible? Well, Moore presumably put down his opponents’ misperceptions to a taste for self-abuse and his opponents presumably put down Moore’s misperceptions to the remnants of an excessively puritanical upbringing. Thus each explained the other’s perceptions without recourse to the moral facts. Moore could explain his opponents’ reactions without appealing to the goodness of self-abuse and his opponents could explain Moore’s reactions without appealing to its badness. And the two sets of explanations seem to be on a par. Moore’s explanations of his opponents ‘perceptions’ seem just as good as his opponents’ explanations of Moore’s.

This suggests an argument for the error theory which combines moral ‘relativity’ with ‘explanatory impotence’.

1) There is a great deal of moral disagreement at the level of alleged ‘perceptions’.
2) Hence many moral ‘perceptions’ can be explained without recourse to corresponding moral facts (since everyone agrees with respect to some moral ‘perceptions’ that there are no corresponding moral facts).
3) But ‘true‘ moral ‘perceptions’ can be explained without recourse to the moral facts in much the same way as ‘false‘ ones.
4) Thus moral ‘perceptions’ generally can be explained without recourse to corresponding moral facts.
5) The only good reason for positing moral facts would be to explain our moral ‘perceptions’.
6) Thus there is no good reason to posit moral facts.
7) If there is no good reason to posit facts of a certain kind, then probably there are no such facts.
8 ) So probably there are no moral facts.

I don’t pretend this argument is decisive. But it seems to me a lot better than Leiter’s.

Let’s put Nietzsche aside for the moment and consider the merits of Leiter’s “moral disagreement” argument directly. Leiter’s argument is a specific instance of a more general line of reasoning: intractable disagreement about the nature of a certain kind of fact shows that there are no facts of the kind in question. He acknowledges that not all instances of this general argument schema are legitimate: “the question is always what is the best explanation for the disagreement in question, given its character and scope.” Thus Leiter assumes for the sake of argument an intractable meta-disagreement between the moral skeptic and the moral realist about what to infer from the fact of intractable disagreement among moral realists as to the content of foundational moral propositions. He denies that we should infer from this meta-disagreement that there is no fact of the matter about what we should infer from the original disagreement. On the contrary, he suggests that the moral skeptic can legitimately argue that he is correct and the moral realist is incorrect and there is a good explanation of why the moral realist persists in his incorrect views. I agree that the moral skeptic has a plausible response here. But then why not explore the possibility that participants in the original intractable disagreement about the content of foundational moral propositions have similar responses available to them? Why not explore the possibility that a consequentialist, say, can legitimately argue that he is correct and the Kantians and virtue theorists are incorrect, and here is why Kantians and virtue theorists intractably hold on to their misbegotten views. After all, this kind of argumentation is employed by philosophers all the time. A good philosopher doesn’t merely present arguments for her own views and respond to the arguments of her opponents; she will also attempt to show why the mistaken views of her opponents have appeal; she will attempt to explain why intelligent people might be attracted to such views, even though such views are mistaken. For example, skeptics about knowledge attempt to explain why nonskeptics think that we have knowledge, whereas nonskeptics attempt to explain the appeal of arguments for skepticism.

So what should we conclude from the intractable disagreements among consequentialists, Kantians, and virtue theorists about the content of foundational moral propositions? Let me suggest that we can gain some perspective on how to approach this question if we keep in mind that the moral theorists in question are not merely disagreeing about the content of foundational moral propositions; they also disagree about how best to explain a fact whose existence should not be in dispute. This undisputed fact is that there is a consensus in certain societies with respect to certain nonfoundational moral issues; this consensus is sometimes referred to as “common sense morality”. Of course there is much moral disagreement in our societies, but there is much agreement, also. Consequentialists, Kantians, and virtue theorists can all be seen as attempting to explain this consensus. Being moral realists, they agree that this consensus should be explained in terms of the idea that members of certain societies have discovered certain moral truths, but they disagree as to the nature of these truths. Given that the existence of the consensus is not in question, we cannot explain away the intractable disagreements among moral realists in terms of the idea that there is no fact here to be explained. But of course it does not follow that one of the moral realist positions is correct. For moral skeptics have their own explanations of this moral consensus. Nietzsche, presumably, would explain it as the triumph of slave morality; in the words of GM I 16, Judea has triumphed over Rome, presumably because Judea is “cleverer” than Rome (GM I 10). Other brands of moral skeptics, such as Freudians and Marxists, would have their own explanations of this consensus.

So now we have not merely a three-way intractable debate among consequentialists, Kantians, and virtue theorists, but at least a four-way intractable debate among the three kinds of moral realists and the Nietzschean as to how best to explain our moral consensus. How should we react to this intractable debate? We cannot just say that all four positions are mistaken, for we have a fact here that needs to be explained, and presumably there exists some true explanation of this fact. All we can do is allow each side to continue to develop their explanations (what Leiter refers to as the “‘progress’ within traditions”), with the hope that one explanation will emerge as the “best.” The question of what explains our moral consensus is a difficult one, so perhaps we should not be surprised that we have not yet reached agreement on the answer to this question.

Note that the Nietzschean cannot easily claim an advantage here by attempting to discredit his opponents’ explanations through showing how these explanations reflect the “affects” (BGE 187), or “drives” (BGE 6), or “desire[s] of the heart” (BGE 5) of their proponents. Put aside the issue of whether the moral realists can give similar discrediting explanations of Nietzsche’s views. What is noteworthy here is that for Nietzsche, this kind of psychological explanation of a philosopher’s view does not discredit it. (In BGE 5, Nietzsche does not criticize philosophers for having views that have an origin in the affects; he criticizes them for not “having the courage of the conscience that admits this.”) On the contrary, Nietzsche ranks the philosopher over the “scholar” precisely because in the philosopher, “there is nothing whatever that is impersonal” (BGE 6; see also the “We Scholars” chapter in BGE). Because the philosopher’s views “reflect who he is” (BGE 6), he can employ his views to “create values” (BGE 211), which for Nietzsche is the ultimate exercise of the will to power. Nietzsche’s criticisms of philosophy and philosophers are as extreme as anyone’s, but unlike some critics of philosophy (Rorty, perhaps?), his goal is not to end philosophy, but to make it better; he wants to encourage the emergence of more creative philosophers.

When there is some fact we want to explain, all we can do is try our best to explain it. If we don’t reach agreement on what the correct explanation is, we just have to try harder. Perhaps some new philosopher will emerge with some new, creative explanation that will secure agreement. But if no agreement is ever reached, so what? No one knew better than Nietzsche that there are things more important than reaching agreement. Let a thousand explanations bloom!

Any fair debate about what constitutes the best explanation would have to involve very long explanatory stories from the different contenders. I am in partial agreement with Leiter in thinking that the persistence of disagreement among smart and informed people over foundational normative questions—e.g., whether moral truth is best captured by Kantian values or utilitarian values or by a set of virtues or by some hybrid of these—is a significant reason for thinking that there are no facts that could settle the matter. It isn’t a sufficient reason by itself, however, because any plausible explanation of persistent and fundamental moral disagreement will rest on many different theoretical assumptions, including assumptions about what an adequate explanation will have to look like (e.g., whether it should be “naturalistic” and what that means). In this forum, let me just express why I think that someone who agrees with Leiter on (1) and (2), and who furthermore is also disposed to accept a naturalistic explanation of what a morality is, could reject (3) and assert that psychological needs, varying cognitive styles, social processes, and normative constraints would have to share more or less equal roles in the best explanation of disagreements among moral philosophers; moreover, these factors interact in ways that make them an ensemble cast. The individual and his psychological needs is just one player, not the star.

Much of the content of morality across different historical periods and cultures is common or at least overlapping in concerning duties such as ones that parents and children have to one another, ones that specify what sort of cooperation or help one owes to others, and ones that specify when and with respect to whom one should refrain from harming. I hold that this is no accident because I defend a constructivist view of the origin of morality, according to which the criteria for applying moral concepts are laid down not to identify independently existing moral properties but to provide a normative structure to their cooperative activities. Given widespread propensities of human beings (I avoid the term ‘human nature’ because of its essentialist connotations), cooperative structures effectively designed for them should possess certain broad features. For example, a norm requiring agents to reciprocate help from others plays a crucial role in reconciling the strong self-interest of human beings with the demands of cooperation; it is not that human beings lack other-regarding concerns, but reciprocation from people they help reduces conflict between these types of motivation and therefore the cost incurred by acting on the other-regarding concerns. Possession of a norm requiring reciprocation is one of the normative constraints on what could constitute an adequate morality for human beings.

If morality does have the function of providing a normative structure for social cooperation, there is pressure for convergence in judgments. Children are instructed in the criteria for the correct or truthful application of moral concepts. They are not taught that they may employ criteria of their own choosing, and tolerance of difference or individual creativity in how these criteria are interpreted or revised by individuals is limited. But if morality is a social construction, we can expect that such pressure for convergence results only in rough and approximate agreement. A morality could meet a function such as structuring and facilitating social cooperation and still allow for plenty of differences among its adherents. This is because the sort of constraint of which I have just given an example (possession of a reciprocity norm) is something like a mid-level generalization: it is consistent with a variety of different specifications; it can be given different rankings relative to other kinds of moral requirements (in comparison to most other cultures, for example, reciprocity in traditional Chinese culture is quite central and takes distinctive forms as its required forms are specified more concretely); and it can be derived from different sets of moral general values or moral norms. Under this sort of naturalistic picture, we can expect moralities to be dynamic, with some broad features remaining pretty constant but otherwise evolving in the way cultural norms tend to evolve: their interpretation and actual constitution can get revised through normative deliberation, but they are also shaped by nonnormative factors such as the influence of competing interests belonging to different individuals and groups.

The pressure for convergence in judgments comes from the advantage that convergence provides to social cooperation, but social cooperation may only require a degree of convergence that falls far short of unanimity or even widespread consensus. Countervailing pressures work in favor of divergence. An important factor is the many-sidedness of human psyches, which tends to create some degree of internal tension and the possibility of oscillation between conflicting values. Another factor is that a society may need to draw from different psychological types because different types are suited for different roles (Wong, forthcoming), and these types will be attracted to different configurations of values. Differences can be functional for social cooperation but also lead to disagreement that has to be (and often is) managed for the sake of continuing cooperation between those who disagree.

Moral philosophers typically try to regiment and reconstruct the diverse, unruly and vague values of lay morality. Some attempt to show that one kind of value, say, from consequentialism or Kantian ethics or virtue ethics, underlies and makes the best sense of that morality (this is the kind of project I interpret to be the sort of “foundational” project that Leiter says we moral philosophers cannot agree upon). The pressure for convergence in lay morality and the presence of certain constraints (such as the need for a reciprocity norm) on what the content of an adequate morality could look like, I submit, accounts for a measure of broad agreement among moral philosophers. Very often, consequentialists, deontologists, and virtue theorists are not in radical disagreement on what duties people have at the mid-level of generalization. For example, advocates of strongly impartialist theories such as consequentialism or Kantian ethics often try to make room for the apparently disproportionate (from an impartialist viewpoint) importance that personal relationships and personal commitments have in lay morality. At the same time, there are philosophers who resist the reconciling move and declare lay morality wrong where it conflicts with their favored normative approach. In general, I find arguments for the superiority of one or another of these contrasting approaches quite unconvincing. The result is incomplete consensus even on this mid-level of generalization, and continued and deep dispute at the foundational level. The constructed nature of morality and the strong pressures for divergence doom attempts to provide some single, correct, deep theoretical basis that will regiment our messy lay moralities. Undoubtedly the psychological needs and temperaments of individual philosophers play a role in influencing which normative theories philosophers find deeply congenial, or on the meta-ethical level, their positions on the question of whether such regimenting foundational projects are a good idea or not, but deliberation on normative constraints, social processes, and the differing cognitive dispositions or styles of individuals also play roles.

Let me focus on the last factor. The psychologists Geoffrey Goodwin and John Darley have conducted studies in which they asked their participants whether they thought that there could be a correct answer as to whether a particular ethical belief was true (or alternatively, whether they thought that a particular belief that they held was true, as opposed to being an opinion or attitude); and whether they thought that a person who disagreed with them about the belief was mistaken, or whether instead neither party need be mistaken in the face of disagreement. These questions were directed at tapping into the participants’ perceptions of the objectivity of ethical beliefs. They found that most, but not all participants attached a strong degree of objectivity to ethical beliefs, but perhaps most interestingly, found that the degree of objectivity attributed varied with the type of belief, e.g., a belief about the morality of abortion was attributed a considerably lesser degree of objectivity than other beliefs such as those concerning the wrongness of inflicting harm (Goodwin and Darley, 2008). In another study, Goodwin and Darley (2010) found some evidence that those more likely to take a subjectivist stance showed a disposition to try to explain why there is disagreement over an ethical issue in terms of the parties holding different values. Those who tended to take the objectivist stance were less interested in explaining disagreement and tended either to disbelieve that someone else could disagree with them or put the disagreement down to some moral defect of the other. Goodwin and Darley suggest that the cognitive tendency (“disjunctive thinking”) to active unpack alternative possibilities while reasoning lies behind the tendency to explain disagreement.

Differences in cognitive styles might underlie differing reactions to an argument that I have made (2006), to the effect that when fundamental values conflict, some of us experience what I call “moral ambivalence:” in coming to understand the reasons why the other side in a disagreement holds the position it does, I claim, some of us might see a path we might have taken and our confidence in the exclusive correctness of our own position gets shaken. Moral ambivalence appears, I claim, when we look at severe conflicts between the claims of special relationships on our time, resources, and energies, and the claims that strangers can have on us in virtue of their humanity and need. Ambivalence appears, I claim, when we look at moralities that are most focused on the value of relationship and community and moralities that are most focused on individual autonomy. However, it is clear that this appeal to a certain kind of ambivalent experience does not resonate with some critics: they have not had it, and they don’t see why they should have it.

I am not proposing that Goodwin and Darley’s explanation of differing reactions among laypeople to moral disagreement can be transposed in any simple way to the differing theoretical reactions among philosophers. Presumably, we’re all pretty good at unpacking alternative possibilities while reasoning, at least at some levels of reasoning. Furthermore, a perennial complaint of universalists and moral realists is that relativists are not able to explain moral disagreement because relativists interpret parties to fundamental disagreements to be talking past one another. On the other hand, much philosophical energy seems directed towards showing that the arguments of proponents of rival normative theories are wrong. There is much ingenuity in anticipating and refuting possible argumentative countermoves by one’s opponents, but perhaps not as much persistent effort to attain some sympathetic insight into why the other side might have the normative commitments it has. The general idea that cognitive styles of some kind might partially account for differences between moral philosophers is not an implausible one and worth investigating. The idea that our profession rewards the products of the more pugilistic styles is also worth investigating.

References

Goodwin, Geoffrey P. and John M. Darley (2008). “The Psychology of Meta-Ethics: Exploring Objectivism.” Cognition 106: 1339-1366.

Goodwin, Geoffrey P. and John M. Darley (2009). “The Perceived Objectivity of Ethical Beliefs: Psychological Findings and Implications for Public Policy.” Review of Philosophy and Psychology. Published online October 11, 2009 @ http://springerlink.com/content/121596/?Content+Status=Accepted.

Wong, David B. (2006). Natural Moralities: A Defense of Pluralistic Relativism (New York: Oxford University Press).

Wong, David B. (forthcoming). “Individual versus Group Disagreement and Their Implications for Moral Relativism.” Blackwell Companion to Relativism. Ed. Steven Hales.

Regardless of this, it seems much harder to be a “compatibilist” about the debate in the philosophy of science, between instrumentalists and realists, than in the debate over free will; it is hard to make sense of the claim that the best theory about science does not depend upon the facts of its subject matter.

To begin, I am not too concerned about the dispute over whether the view Leiter (and I) wish to defend is called moral skepticism or moral irrealism (or anti-realism for that matter). To many of us, the word “skepticism” suggests an epistemological position, but there is another, easily recognized sense (whether etymologically fitting or not) in which the words means pretty much the same thing, and that is the sense Leiter is employing (as did Mackie).

Actually, I think that if the argument from (moral) disagreement should worry anyone, it should worry (in addition to moral realists) some who do not classify themselves as realists: constructivists, some expressivists (pace Mike Ridge), Kantians, and anyone else who claims that moral questions (often, or maybe typically) have right answers—people holding a collection of views much wider than the set containing only those who are thought of (or think of themselves) as moral realists. I think that the divide between these factualists (as I call them) and non-factualists (error theorists, classical subjectivists, simple non-cognitivists like Ayer, and incoherentists—ok, “ist”) is in many ways more interesting and useful than the distinction between realists and irrealists (if the latter distinction can be made clearly at all). For starters, acceptance or rejection of moral facts, truths, or correctness (I’ll just say “facts” henceforth) is itself a pretty significant fault line in metaethics. Intuitively, views that have a place for such things have much more in common with one another, on the whole, than they do with views that do not.

It would take reasoning —maybe fairly fine grained reasoning—to show that the argument from disagreement applies to each view in this category, but any such reasoning would have to proceed from an understanding of the core argument from disagreement itself, and getting that clearer will also help us to see the structure of Leiter’s argument. In particular, it is puzzling to many of the commentators why moral disagreement among anyone is supposed to count against moral factualism. An explanation is wanted.

There is more than one way to fill out the core argument. But one of the most promising approaches claims that moral disagreement is strong evidence against our having epistemic access—a way of knowing or forming justified beliefs about—moral facts, and from there argues that we have no good reason to believe in such facts. The argument might focus either on disagreement about the methodology of moral reasoning, on disagreement in moral belief itself, or more likely on both. Indeed, disagreement in the methodology of moral reasoning might help to explain disagreement in moral belief. But there are still many unanswered questions. For one thing, why does disagreement count against access? For another, why (if at all) does lack of access count against factualism?

A couple of analogies might help us better to understand the argument. Suppose I claim that most people have a faculty of extra-sensory perception that allows us to “see” the rank and suit of an overturned card. (I am not saying that moral knowledge is like this in more than the one salient respect.) To test the claim, I have each of twenty people write down what she believes to be the rank and suit of a particular card. Suppose that all twenty subjects give different answers. In such a case I have some reason to doubt that the faculty in question exists, at least within this population.

Note that my reason for doubting that the faculty exists seems to be largely independent of whether I myself know the card’s rank and suit. And even if one of the twenty subjects does give the right answer, the experiment still gives me warrant (albeit to a slightly lower degree) for thinking that this was merely a coincidence. It could be that only that subject (among those tested) actually has the faculty. Indeed, even if none of the subjects gives the correct answer, it could be that someone else has the faculty. Likewise, even if no one gives the correct answer, it’s still possible that several of these subjects do have the faculty but that for some reason it is not working well on this occasion.

But these hypotheses seem desperate, and not simply because this sort of ESP already seems unlikely. When it comes to a (largely) empirical issue like this, we shouldn’t let the tail of theory wag the dog of evidence any more than we have to. The experiment itself provides evidence against the claimed faculty. Of course the evidence is only probabilistic, and should be seen as supporting only a somewhat tentative inference to the best explanation. Twenty cases and one card are nowhere near enough. Further testing could greatly strengthen or weaken the case. But if many people in lots of circumstances get the answer wrong, and get it right only in roughly the same proportions as would be produced by random chance, then we have powerful evidence that we lack epistemic access to facts about the rank and suit of overturned playing cards, and quite likely to other, similar matters as well.

Even so, we know that there are facts about the identities of playing cards, and our inability to discern these identities when the cards are face down does not count against there being such facts. We can modify the argument to avoid this worry. A better analogy, then, involves so-called auras. Suppose I claim that humans have a different sort of ESP—a faculty for discerning one another’s auras, distinguished by their colors. If, when I ask the same twenty people what color a particular person’s aura is, they all give different answers, I have the beginnings of a pretty good reason to doubt, not simply that we can discern what people’s auras are, but that people have auras at all. For the only marginally plausible reason we might have for believing in auras in the first place is that people believe that they are detecting them.

Although I think a central version of the argument form moral disagreement is best understood along lines somewhere in the neighborhood of these, there are still (or in some cases may be) big disanalogies between the ESP cases and that of moral disagreement. In the moral case, a central question will be whether we have independent grounds for believing that moral facts exist. Perhaps even more important is the fact that there so much more agreement (and so much less disagreement) over moral matters than there is in the two toy cases I have been discussing. This convergence that does exist on moral questions requires explanation in its own right. And a prominent candidate explanation involves moral facts together with some (imperfect) ability in us to discover what they are

On the other hand, holding the argument to a standard of total or near total moral disagreement seems unwarranted. What the anti-factualist needs is enough disagreement of the right sorts so as to suggest that alternative explanations of convergence in moral beliefs and attitudes, explanations that are incompatible with or obviate the (explanatory) need for moral facts, are more plausible than ones involving people’s coming to be aware of such facts. In this regard, I suspect that mere agreement in moral judgment should not be as much comfort to factualists as it is often thought to be, since disagreement about why an act is morally required [say] is moral disagreement too, and whether convergence among people with widely different theories is after all best explained by truth tracking or by other factors is as yet not established.)

Even in the card and aura cases, the evidence against the ability to discover the facts (and hence, in the latter case, against the facts themselves) would be undermined if there were highly plausible explanations for why the capacity should not have been expected to work in this particular population—if, for example, the failures typically involved intoxicated subjects, and success rates among the few sober people tested are significantly higher). But note that the mere possibility of such “defusing explanations” is not enough to undermine the evidence disagreement provides against the ability and the facts in question. And we can gather evidence about the defusing explanations themselves. There is, of course, nothing privileged about the no-access explanation. We have to look to the total explanatory picture. As I’ve argued, all too often, people on both sides of the debate have relied on anecdote or speculation, no substitute for good evidence (and probably of little probative value at all.

The point Leiter (and perhaps his German friend) make about disagreement among philosophers on fundamental moral questions (like utilitarianism vs. Kantianism vs. other approaches) is that the traditional defusing explanations seem obviously not to apply. It seems unlikely that one side to these disputes is typically more biased or prone to special pleading or less morally sophisticated or developed than the other, or that the two sides have important and relevantly different beliefs on non-moral matters, that they are applying the same moral principles to different circumstances (which call for very different sorts of actions), or that the disagreement concerns matters on which we should (arguably) expect disagreement such as cases involving incommensurable but real moral values, equally weighted moral considerations, or indeterminacy. And while, generally speaking, differences in background theory are well known to cause differences in “observation” or judgment, the moral differences among advocates of these conflicting approaches seem to reflect intuitive disagreements rather than to generate them. (At least, that’s the way it feels, for what it is worth).

One way to try to answer this type of argument is to echo Parfit’s famous claim that ethics is a very young discipline, as some here have. Shafer-Landau, for example, notes that many more people have been working on science than have been working on substantive ethics. Such claims, however, may unfairly downplay the continuity of substantive ethics as explored outside of more narrowly philosophical traditions (or prior to the very productive years that began around the time A Theory of Justice was published and Philosophy and Public Affairs founded) with the ethics about which disagreement among philosophers remains. Brushed aside, for example, is ethical thinking done within religious traditions, perhaps on the ground that the framework, methodology, or presuppositions employed there are so different from “ours” that we ought not to count the religious traditions as reasoning about ethics at all, but about the interpretation of religious texts and dogma, for example.

This reply seems to me to distort both the history and the issues, however. (Here I find myself echoing that Nietzsche guy again, who, as Leiter shows, scoffs at the more than two millennium long embarrassment of philosophical disagreement on morality.) How much of the moral reasoning done within a religious framework, for example, actually depends on non-moral factual beliefs about God or other religious matters? The answer is some, but certainly not all, not even most (pace Mike Ridge, again). That Pharaoh mistreated the Israelites by enslaving them, that Cain acted wrongly by killing his brother in anger, that Adam and Eve were wrong to attempt to avoid responsibility by blaming others (Eve in Adam’s case, and the serpent in Eve’s) do not depend on any hypotheses about our being God’s property or it being wrong to try to fool God with shoddy excuses, for example. Similarly, anyone who has ever taught Locke on property knows that, agree with him or not, little of interest there (the law of waste or the proviso, for example) depends on the theistic framework in which it is presented.

The argument also seems overly optimistic. That if we just work long enough on substantive ethics, we’ll resolve enough of our differences to put ethics on a par with science is a possibility, but an unrealized one on which factualists are not yet entitled to rely, especially when deep questions about the methodology of moral reasoning themselves look to be no more tractable than the moral questions themselves.

If factualists instead (or in addition) claim that sometimes apparent disagreement shows that we are talking about something different than and incommensurable with, say, what the Homeric Greeks were talking about (and so, not disagreeing at all), then so much the worse for moral factualism. That morality is something only recently (and locally) discovered ought to be viewed as anathema to any robust version of that approach.

I conclude that disagreement among philosophers is, as Leiter and Nietzsche argue, a potentially serious problem for moral factualism. The reason, as I said, is that traditional defusing explanations for moral disagreement do not seem to apply to this dispute. What is interesting, and perhaps a bit ironic, is that Leiter’s paper itself suggests the beginnings of a strategy for responding to this very argument—a family of defusing explanations that might apply especially to moral disagreement among philosophers. The hint is summed up nicely in Leiter’s bitingly Nietzschean final paragraph, though it is directed at the meta-disagreement over whether moral disagreement counts against moral factualism:

“[W]hat is the best explanation for the meta-disagreement . . . ? Surely one possibility—dare I say the most likely possibility?—is that those who are professionally invested in normative moral theory as a serious, cognitive discipline—rather than seeing it, as Marxists or Nietzscheans might, as a series of elaborate post-hoc rationalizations for the emotional attachments and psychological needs of certain types of people (bourgeois academics, ‘slavish’ types of psyches)—will resist, with any dialectical tricks at their disposal, the possibility that their entire livelihood is predicated on the existence of ethnographically bounded sociological and psychological artifacts.”

If deep and substantive moral disagreements among philosophers can also be explained as owing to “elaborate post-hoc rationalizations for the emotional attachments and psychological needs” of those philosophers, then perhaps the disagreements can be defused after all. We might add (or perhaps just make explicit) that in our profession we put a premium on ingenious defenses of contrasting positions, so it is no wonder that we continue to disagree. We get paid to.

My own view is that this cynical picture is not entirely fair, and that the disagreements among philosophers reflect the difficulty of coming up with answers that recommend themselves as definitively better. As my son, Isaac, pointed out to me, philosophers would not have been likely to dig in so hard on certain normative questions and not others if each side had not antecedently found the varying positions to be both plausible and reflective of deep seated normative commitments and beliefs. There are plenty of controversial positions to spend one’s time defending, without inventing ones that do not touch normative (or other) chords in us. You do not find philosophers disagreeing (much, anyway) about the law of the excluded middle, the wrongness of killing for hire (at least among those who think anything is ever right or wrong) or about whether animals even feel pain. In each case, part of the explanation is that the other side of the question simply isn’t very plausible to practically anyone. More often than not we pick philosophical issues to (get paid to) disagree about because there are genuine differences to start with, and that seems to be the case with respect to Kantian and utilitarian intuitions, and other deep-seated disagreements among moral philosophers to which Leiter and Nietzsche point.