¹ Among the many who have given in to the temptation are Rawls, John in A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), pp. 185–88Google Scholar; Firth, Roderick in “Ethical Absolutism and the Ideal Observer,” Philosophy and Phenomenological Research, 1952, pp. 336–41Google Scholar; Harrison, Jonathan in Hume's Moral Epistemology (Oxford: Oxford University Press, 1976), p. 114Google Scholar; and Glossop, Ronald in “The Nature of Hume's Ethics,” Philosophy and Phenomenological Research, 1967, pp. 527–36Google Scholar. When Ideal Observer theories are discussed, Hume is almost always cited as an early advocate of the view.

² Hume, , Enquiry concerning the Principles of Morals, p. 286Google Scholar. Likewise, when he holds that “everything, which gives uneasiness in human actions, upon the general survey, is call'd Vice, and whatever produces satisfaction, in the same manner, is denominated Virtue …” (Treatise, p. 499)Google Scholar, the survey that matters, one might think, is that taken by a suitably qualified judge. See also p. 591 of Hume, , A Treatise of Human Nature, ed. Selby-Bigge, L. A., 2d ed. with revisions and notes by P. H. Nidditch (Oxford: Oxford University Press, 1978)Google Scholar. A number of passages from the Enquiry also suggest this interpretation, usually in the context of emphasizing the proper role of reason in moral judgment. See, for example, pp. 173 and 290–91 of Hume, David, Enquiries concerning Human Understanding and concerning the Principles of Morals, ed. Selby-Bigge, L. A., 3d ed. with revisions and notes by P. H. Nidditch (Oxford: Oxford University Press, 1975)CrossRefGoogle Scholar. References in the body of the essay for the Treatise will appear parenthetically as (T.), while those for the Enquiry concerning the Principles of Morals will appear as (E.).

³ Hume, David, “Of the Standard of Taste,” in his Essays: Moral, Political and Literary, ed. Miller, Eugene F. (Indianapolis: LibertyQassics, 1985), p. 241Google Scholar. References in the body of the essay for “Of the Standard of Taste” will appear parenthetically as (“Taste”).

⁴ Hume, , Treatise, p. 619Google Scholar. Combining the explanatory and justificatory projects, of course, carries significant risks. An explanation of any fairly predominant moral view will likely be an explanation of a view many of us think is inadequate in some important way. If the theory offered is to be remotely plausible as a normative theory, the principles advanced must give us a purchase on actual practice that allows critical evaluation of what happens to be in place. Even so, there must be some explanation of why we hold the views we do see as justified, so the explanatory and justificatory projects cannot diverge completely (at least when the views being explained are our own). And the hope is that what explains our particular moral views and our practice of forming such views might simultaneously serve as a justification in our own eyes of both the views and the practice.

Christine Korsgaard does a nice job of articulating Hume's conception of normativity as reflective endorsement in the second lecture of her 1992 Tanner Lectures, “The Sources of Normativity” (manuscript).

⁵ I am thinking here, for instance, of his tendency to see human nature as extraordinarily and conveniently uniform.

⁶ His account of the artificial virtues, for instance, seems to tempt him in this direction (justice's silence concerning the weak, and modesty's especially strong claim on women, come to mind here).

⁷ Hume, , Enquiry, p. 279.Google Scholar

⁸ Thus, the idealizations involved are at least (i) the requirement of full knowledge, (ii) the complete impartiality of the responses, and (iii) the inclusion of the effects on everyone. Even more, or more specific, idealizations might be imposed. Roderick Firth, for instance, characterizes the Ideal Observer as not only omniscient with respect to nonethical facts, but as omnipercipient, disinterested, dispassionate, and consistent. Others, taking a lead from Richard Brandt, might suggest that an appropriate observer must have undergone cognitive psychotherapy. See Brandt, Richard, A Theory of the Good and the Right (Oxford: Oxford University Press, 1979)Google Scholar, where the condition is advanced as part of an account of what sets the standard for an individual's good. The contrast I will be pressing would only be heightened by adding some or all of these other idealizations. See also Railton, Peter, “Moral Realism,” Philosophical Review, vol. 95 (1986).CrossRefGoogle Scholar

⁹ “Morals,” he observes, “excite passions, and produce or prevent actions. Reason of itself is utterly impotent in this particular. The rules of morality, therefore, are not conclusions of our reason” (Treatise, p. 457)Google Scholar. This argument is just the first in a salvo Hume fires at attempts to found morality on reason. I will not here either rehearse or endorse the whole collection.

¹⁰ Hume's argument is set out nicely on p. 272 of the Enquiry.

¹¹ Hume, , Enquiry, p. 174Google Scholar; see also pp. 173 and 312. In the Treatise, Hume takes on the same project, asking what “distinguishes moral good and evil. From what principles is it derived, and whence does it arise in the human mind?” (p. 473). Interestingly, Hume is concerned primarily to establish the principles that govern our judgments of virtue, not to explain the origin of our idea of virtue. This contrasts intriguingly both with Hume's own discussion of causation and with Francis Hutcheson, who seems especially concerned to show that our idea of virtue arises from a moral sense. See Hutcheson, Francis, Inquiry into the Original of Our Ideas of Beauty and Virtue, selections of which can be found in Moral Philosophy from Montaigne to Kant, vol. 2, ed. Schneewind, J. B. (Cambridge, UK: Cambridge University Press, 1990).Google Scholar

¹² Hume, , Treatise, pp. 574–75Google Scholar; the same view is advanced on pp. 296, 471, and 499, as well as on p. 261 of the Enquiry.

¹³ Combining the first claim with the second, Hume writes:

Every quality of the mind is denominated virtuous, which gives pleasure by the mere survey; as every quality, which produces pain, is call'd vicious. This pleasure and this pain may arise from four different sources. For we reap a pleasure from the view of a character, which is naturally fitted to be useful to others, or to the person himself, or which is agreeable to others, or to the person himself. (Treatise, p. 591)Google Scholar

¹⁴ In a footnote in the Enquiry, Hume distinguishes two kinds of benevolence, general and particular. He does this in a way that, on the one hand, distinguishes both from the universal benevolence he ridicules (as nonexistent) in the Treatise (p. 481)Google Scholar and, on the other hand, treats as pretty much equivalent general benevolence, humanity, and sympathy (Enquiry, p. 298n.)Google Scholar. In this essay I will treat sympathy and humanity as interchangeable. There are, I think, some subtle and important differences, but not differences that matter to the issues I am exploring here.

¹⁵ See also, for instance, Hume, , Enquiry, p. 267.Google Scholar

¹⁶ See Hume, , Treatise, p. 582Google Scholar. As Hume emphasizes, all sentiments “whence-ever they are deriv'd, must vary according to the distance or contiguity of the objects …,” so the first is a problem not just for the appeal to sympathy: “if the variation of the sentiment, without a variation of the esteem, be an objection, it must have equal force against every other system, as against that of sympathy”—assuming, as Hume does, that “[t]he approbation of moral qualities most certainly is not deriv'd from reason, or any comparison of ideas” (ibid., p. 581).

¹⁷ This is because sympathy interests us in the welfare of others and “[t]he goodness of an end can bestow a merit on such means alone as are compleat, and actually produce the end.” This means that “when the cause is compleat, and a good disposition is attended with good fortune, which renders it really beneficial to society, it gives a stronger pleasure to the spectator, and is attended with a more lively sympathy” (Hume, , Treatise, pp. 584–85Google Scholar; see also the footnote in Hume, , Enquiry, p. 228).Google Scholar

¹⁸ The same point is made in Hume, , Enquiry, p. 228:Google Scholar

[T]he tendencies of actions and characters, not their real accidental consequences, are alone regarded in our moral determinations or general judgements; though in our real feeling or sentiment, we cannot help paying greater regard to one whose station, joined to virtue, renders him really useful to society, than to one, who exerts the social virtues only in good intentions and benevolent affections.

¹⁹ Hume, , “Of the Standard of Taste,” p. 229Google Scholar. The great uniformity in the general sentiments of mankind regarding the value of virtues might seem to reduce the importance of being able to distinguish a right taste from a wrong one. However, as Hume rightly points out, once we turn to particulars, dramatic differences emerge and loom large. The almost universal consensus concerning the value of the virtues masks deep differences concerning which particular character traits constitute the specific virtues. What one person counts as bravery, another sees as foolhardiness.

²⁰ Hume, , “Of the Standard of Taste,” p. 230Google Scholar. This goes a bit far, I think, since in the absence of any standard there seems no particular reason to refrain from regulating the sentiments of others to the extent one can.

²¹ Hume's appeals to “points of view,” not just in ethics but elsewhere, suggest that he sees a point of view primarily as a way of seeing or thinking of something, and not as the occupying of a particular position in the viewing of something. See, for instance, his appeal to the idea in Hume, , Treatise, pp. 169, 220, 356, 389, and 440Google Scholar. See also Hume, , “Of the Standard of Taste,” p. 239.Google Scholar

²² Hume's particular account of (uncorrected) moral sentiments, as having a distinctive qualitative feel and being the product only of sympathy is, I believe, quite implausible. But it plays an important role in his sentimentalist theory of morality, since any such theory must have an account of which sentiments come into play and which do not. In “A Humean Theory of Moral Judgment” (manuscript), I argue that a Humean, even if not Hume, can provide such an account. Relying on Hume's account of the indirect passions, the suggestion there is that a Humean can make sense of the relevant sentiments as responses-for-reasons (e.g., being angry-because-he-hurt-you) where the cognitive features of the reactions are not at the start moralized, but become so as those responses are themselves approved of by a judge properly situated.

²³ We confine our view in this way, according to Hume, in response to, and in recognition of, the limits nature has placed on the scope of human affections (Hume, , Treatise, p. 602Google Scholar, and Enquiry, p. 225n.)Google Scholar. The role played here by Hume's appeal to the “narrow circle” is in other places played by an appeal to those “who have a connexion” with the person judged (Hume, , Treatise, pp. 591 and 602)Google Scholar. Which group comes within our view depends on the context and character of the person judged.

²⁴ Hume, I should add, is extraordinarily (and uncharacteristically) careful about respecting the difference between what we feel and what we say. (See Hume, , Treatise, pp. 582 and 603.)Google Scholar His point is not that people are sometimes hypocrites, but that our moral judgments, though grounded in sentiment, are not a mere reflection of how we happen to feel. His account thus allows some slip between occurrent sentiments and occurrent judgments. This distinction raises some interesting complications when it comes to interpreting Hume's arguments against the rationalists. Elizabeth Raddiffe presses these difficulties in “Hume on Motivating Sentiments and Moral Reflection” (manuscript). I try to address them in “Practical Morality and Inert Reason” (manuscript).

²⁵ See Hume, , Treatise, p. 603:Google Scholar

The case is here the same as in our judgments concerning external bodies. All objects seem to diminish by their distance: But tho the appearance of objects to our senses be the original standard, by which we judge of them, yet we do not say, that they actually diminish by the distance; but correcting the appearance by reflexion, arrive at a more constant and established judgment concerning them.

Hume makes the same point, with the same example, in the Enquiry, pp. 227–28.Google Scholar

²⁶ It is worth emphasizing that the analogy is not just with “secondary properties,” but with all properties “discovered” by sense. See Simon Blackburn's discussion of the dangers of stressing an analogy between secondary properties and moral properties in interpreting Hume in “Hume on the Mezzanine Level,” Hume Studies, vol. 14 (11 1993).Google Scholar

²⁷ That we have settled, in making ordinary color judgments, on normally sighted human observers in daylight conditions, is no accident, though presumably it could have been otherwise. Were we to evolve so as to be visually sensitive to ultraviolet light, or were we to establish prevailing lighting conditions that allowed (with relative ease) a more articulate range of discriminations, we might well shift the standard we use in regulating our color judgments.

²⁸ “We know, that an alteration of fortune may render the benevolent disposition entirely impotent,” but this reduces not at all our regard for the benevolent person, since “we separate, as much as possible, the fortune from the disposition” in reflecting on the value of the benevolent character (Hume, , Treatise, p. 585).Google Scholar

²⁹ Juster, Norton, The Phantom Tollbooth (New York: Random House, 1961), pp. 104–5.Google Scholar

³⁰ Hume, , Treatise, p. 603Google Scholar. Virtually the same sentence occurs in Hume, , Enquiry, p. 228.Google Scholar

³¹ Even to communicate concerning how things seem, we need a shared standard for distinguishing how things seem from how they are. For instance, for me to understand your report that a box looks red, I need to know what it is for something to look red. And to do this, we need to have picked out a class of things we together denominate “red” and to have privileged certain conditions as setting the standard. We can then (but only then) understand what it is to look red as looking the way these things do to an appropriate observer under those standard conditions. The same shared practice allows us to communicate successfully under nonstandard conditions and without knowing the particular situation we each face as long as we both can figure out how the things we do see would look under standard conditions. We introduce a standard for distinguishing how things are from how they seem, and then “correct” how they seem in making judgments about how they are, in order to communicate effectively. And it is against that background that we can say that a box that is red will look brown under certain light, or that a coin that is circular will look elliptical when viewed from a certain angle.

³² Hume, , Treatise, pp. 581–82Google Scholar. Hume talks of general points of view here, rather than of the general point of view, because the problem he is pointing to is not unique to morality but arises “with regard both to persons and things.” In different areas, different points of view will be relied upon to resolve the “contradictions” that inevitably emerge. In any case, “[w]hen we form our judgments of persons, merely from the tendency of their characters to our own benefit, or to that of our friends, we find so many contradictions to our sentiments in society and conversation, and such an uncertainty from the incessant changes of our situation, that we seek some other standard of merit and demerit, which may not admit of so great variation” (Hume, , Treatise, p. 583Google Scholar; see also pp. 228 and 272 of the Enquiry).

³³ Hume, , Treatise, p. 602Google Scholar. Hume observes that “every particular person's pleasure and interest being different, ‘tis impossible men cou'd ever agree in their sentiments and judgments, unless they chose some common point of view, from which they might survey their object, and which might cause it to appear the same to all of them” (ibid., p. 591). Hume makes the same point on p. 272 of the Enquiry.

³⁴ At least not at this point. Propositional contradictions will arise from each speaking from her own point of view once a regimented moral language is introduced—for then when you say someone is virtuous and I say she is not, we will be contradicting each other. But the regimentation that makes sense of the prepositional contradictions comes only with the introduction of the general point of view, so the threat of such contradictions cannot be the grounds we have for introducing the regimentation.

³⁵ See Hobbes, Thomas, Lewathan (Harmondsworth: Penguin, 1968), esp. ch. 13.Google Scholar

³⁶ My sympathetic engagement with my own child's welfare, for instance, may well lead me to object to your treatment of him, even as that treatment by you is prompted by your own sympathetic engagement with your child's welfare. The conflict that emerges, though it reflects our own peculiar positions, in no way depends upon either of us being moved in this instance by self-interest.

³⁷ Hume, , Treatise, p. 591Google Scholar. Which group is relevant will of course depend both on which type of character is being evaluated and on what information is mutually accessible. Hume notes, for instance, that the character traits of statesmen, given their role, must often be evaluated with an eye to the welfare of whole countries (though only their own), even as he emphasizes that the relevant group will usually be those within a “narrow circle” (Hume, , Enquiry, p. 225n)Google Scholar. Annette Baier suggests that the process of correction is one of selectively sympathizing either with those who are close (when we are distant) or with those who are distant (if we are close). I think, in contrast, that the object of our sympathy is supposed to be always the same group of people—“the person himself, whose character is examin'd” and those “who have a connexion with him” (Hume, , Treatise, p. 591)Google Scholar. See Baier, Annette, A Progress of Sentiments (Cambridge, MA: Harvard University Press, 1991), pp. 181–82.Google Scholar

³⁸ In “Hume and the Bauhaus Theory of Ethics” (manuscript), I argue that Hume's functionalist account of morality in general, and the virtues in particular, accommodates the importance of utility while setting him apart from utilitarians in a distinctive and fundamental way.

³⁹ See Hume, , Treatise, pp. 477–513Google Scholar, and Enquiry, pp. 183–204, 303–11Google Scholar. J. L. Mackie argues that a proper appreciation of the parallels undermines Hume's distinction between the artificial and natural virtues. But I think this is a mistake. Within this account of morality, it is still possible and important to mark the difference between artificial and natural virtues in just the way Hume does—that is, by noticing the difference between those character traits that one can specify and approve of only within the context of a set of conventions (e.g., justice) and those that one can specify and approve of absent an appeal to convention (e.g., benevolence). The natural virtues naturally engage approbation in a way that artificial virtues do not, although the natural virtues no less than the artificial ones count, in the end, as virtues only because they are properly approved of within a conventional system of approbation. See Mackie, J. L., Hume's Moral Theory (London: Routledge & Kegan Paul, 1980), p. 123CrossRefGoogle Scholar. For an account that is more sensitive than Mackie's to the sort of artifice involved in moral judgment, see Baier, , A Progress of Sentiments.Google Scholar

⁴⁰ Whether embracing the general point of view, and not the Ideal Observer's, as the standard really does offer the best available solution is, certainly, open to debate. The Ideal Observer's inaccessibility does not automatically settle the issue, since it is possible, at least, that each of us doing the best we can to estimate the Ideal Observer's response might in itself be enough to resolve our conflicts. Still, Hume is on reasonably solid ground, I think, since our different perceptions of the actual effects of someone's character on everyone are virtually certain to find an echo in our estimates of what the ideal Observer would perceive.

⁴¹ For someone already convinced of the value, say, of pleasure no matter whose, any standard for moral judgment that stops short of taking into account the pleasures and pains of all sentient beings will no doubt seem a standard we have reason to move beyond. The burden, though, falls on that person to explain the value of pleasure (or whatever). And Hume will press the issue. We are, as a matter of fact, concerned with the welfare of those with whom we can sympathize, and we have reason to regulate our judgments by how we would feel from the general point of view, but what reason is there for us to care, in the absence of sympathy or some other sentiment, about those who fall beyond the pale?

⁴² Or they might differ because they take as relevant different “narrow circles,” though presumably this difference is fairly and easily handled within the theory as a case of the two not both succeeding in taking the general point of view.

⁴³ Simon Blackburn has pressed this suggestion in conversation. This would mark a significant divergence from standard Ideal Observer theories, since on those the proper standard is set by an observer responding to all the actual consequences of someone's character, rather than to that character type's consequences. Suppose, though, that the change is made in the name of avoiding unnecessary inaccessibility.

⁴⁴ “One man's ambition is not another's ambition: But the humanity of one man,” Hume insists, “is the humanity of every one; and the same object touches this passion in all human creatures” (Hume, , Enquiry, p. 273).Google Scholar

⁴⁵ Hume, , Enquiry, p. 226Google Scholar. However, he thinks that “[a]bsolute, unprovoked, disinterested malice has never perhaps place in any human breast” (ibid., p. 227). And he maintains that “none are so entirely indifferent to the interest of their fellow-creatures, as to perceive no distinctions of moral good and evil …” (ibid., p. 225).