¹ For a standard introduction to the concept, see Weisstein, Eric, “Lexicographic Order,” in MathWorld (Wolfram Research, 1999, http://mathworld.wolfram.com/LexicographicOrder.html)Google Scholar.

² John Stuart Mill, Utilitarianism, chap. 2, paragraphs 5, 8, and 10 [X:211, 213f.]. In all citations of Mill's works, bracketed references indicate volume and page numbers in John Stuart Mill, Collected Works (Toronto/London: University of Toronto Press/Routledge and Kegan Paul, 1967–1989). Being a higher pleasure is, on Mill's definition, a relational property: a pleasure is higher with respect to a specific contrasting pleasure; that latter pleasure is lower with respect to the former. This means that, in principle, a pleasure could be higher with respect to another pleasure, while being lower with respect to a third. Mill, however, typically wrote as though pleasures fell into two classes, the higher and the lower (that is, he wrote as though “higher” and “lower” were monadic predicates). I will in due course consider both why he might have allowed himself this way of putting things, and what further issues might turn on it. Following up on my description of the standard version of the decided-preference criterion as “almost correct,” I will also explain why talk of amounts of a higher pleasure is importantly misleading.

³ Huxley, Aldous, Brave New World (New York: HarperPerennial, 1932/1998)Google Scholar; similar complaints can be found in Gunn, James, The Joy Makers (New York: Bantam, 1961)Google Scholar. For a representative acknowledgment of the problem by an academic, here writing for a popular audience, see Skorupski, John, Why Read Mill Today? (New York: Routledge, 2006), 61, 76Google Scholar.

⁴ Mill, Utilitarianism, chap. 2, paragraphs 23, 25 [X:223, 225].

⁵ Dostoyevsky, Fyodor, Crime and Punishment, trans. Garnett, Constance (New York: Bantam Books, 1962), 58fGoogle Scholar. During the twentieth century, the device most used to address this problem was rule-utilitarianism, the idea being that what got your moral theory into trouble was testing the particular action for its effects on utility, rather than testing the rule the action is subsumed by. (The rule “Kill miserly old women for their money” wouldn't look nearly as good, even to Dostoyevsky's student.) Act-utilitarians and rule-utilitarians argued for decades, among other things over which view to attribute to Mill. That debate was, in my view, a mistake: both interpretations are unhelpful anachronisms.

In Mill's scheme of things, whether a choice promotes utility depends on the preferences of experienced judges. Now, as a matter of psychological fact, the objects of preference are sometimes more and sometimes less particular. Around election time, for example, voters develop preferences over particular candidates; they also often have much more general preferences, exhibited in choices of rules about how to vote (e.g., straight ticket). Thus, the judgments derived from the preferences of the experienced will sometimes look more like act-utilitarian guidelines, and sometimes more like rule-utilitarian guidelines. (That is not to say that we will not sometimes find the contrast between “acting on general rules” and “measuring the consequences of each act”; these phrases are quoted from a letter Mill wrote to George Grote [Collected Works, XV:762].)

Here's why the debate was unhelpfully anachronistic. The twentieth-century debate came to a close with Lyons, David, The Forms and Limits of Utilitarianism (Oxford: Clarendon Press, 1965)CrossRefGoogle Scholar, which argued that, as rules get more contoured, rule-utilitarianism collapses back into act-utilitarianism. That argument works because there are no limits to the complexity of a rule, and that presupposition was allowable because twentieth-century ethics had taken the same antipsychologistic turn as twentieth-century logic. Because Mill's experienced judges will not form preferences over arbitrarily complex rules, the collapse of rule-utilitarianism into act-utilitarianism is preempted. That is, when you insist on framing your treatment of Mill in these anachronistic terms, you bypass the very material that allows Mill to do better than the parties to the twentieth-century debate did.

⁶ Here is an indication of just how difficult the puzzle has seemed. As thoughtful a reader as Himmelfarb, Gertrude (On Liberty and Liberalism: The Case of John Stuart Mill [San Francisco: Institute for Contemporary Studies Press, 1990])Google Scholar was driven to the view that On Liberty and Utilitarianism are irreconcilable, and that their incompatibility is to be explained by attributing the two books to different authors: the team consisting of Mill and Harriet Taylor in the one case, and Mill on his own in the other. Since the two volumes were in the works at about the same time in Mill's life, since On Liberty explicitly acknowledges “utility as the ultimate appeal on all ethical questions” (chap. 1, par. 11 [XVIII:224]), and since the last chapter of Utilitarianism takes the first steps toward protecting individuals from the kind of oppressive measures that might seem to follow from the Principle of Utility, and thus is naturally regarded as a segue to On Liberty, this is an exegetical last resort.

⁷ And that is the way reconstructions of Mill's arguments have, by and large, treated such claims. For the past few decades, the gaps have been filled in by appeal to what Rawls, John, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), 426fGoogle Scholar., called the Aristotelian Principle: the idea that, “other things equal, human beings enjoy the exercise of their realized capacities (their innate or trained abilities), and this enjoyment increases the more the capacity is realized, or the greater its complexity.” Rawls claims that “Mill comes very close to stating [the Aristotelian Principle] in Utilitarianism, ch. II, pars. 4–8,” and Rawls further describes it as “a principle of motivation … [which] expresses a psychological law governing changes in the pattern of our desires.” But the claim that humans prefer more complex activities is supported neither by argument for the Aristotelian Principle of the sort that would make it compelling to a present-day audience, nor by reconstructed Millian argumentation that would justify attributing it to Mill. Rather, it is introduced by Rawls as a platitude.

In a very similar vein, Gray, John, “Mill's Conception of Happiness and the Theory of Individuality,” in Gray, John and Smith, G. W., eds., J. S. Mill's On Liberty in Focus (London: Routledge, 1991), 200, 209Google Scholar, treats as obvious the claim that the “actualisation of [his unique range of] potentialities is indispensible for any man's greatest well-being,” allows that a similar claim regarding autonomy might, for all we know, be found false in the future, and states that autonomy and authenticity are required for a happy life—but without providing anything like a tight Millian argument for these claims (compare also Gray, John, Mill on Liberty: A Defense, 2d ed. [New York: Routledge, 1996], chap. 4Google Scholar).

Representative recent discussion of the distinction between higher and lower pleasures includes: Riley, Jonathan, “Is Qualitative Hedonism Incoherent?” Utilitas 11, no. 3 (1999)CrossRefGoogle Scholar; Riley, , “Interpreting Mill's Qualitative Hedonism,” Philosophical Quarterly 53, no. 212 (2003)CrossRefGoogle Scholar; Riley, , “On Quantities and Qualities of Pleasure,” Utilitas 5, no. 2 (1993)CrossRefGoogle Scholar; Arrhenius, Gustaf and Rabinowicz, Wlodek, “Millian Superiorities,” Utilitas 17, no. 2 (2005)CrossRefGoogle Scholar; Schmidt-Petri, Christoph, “Mill on Quality and Quantity,” Philosophical Quarterly 53, no. 210 (2003)CrossRefGoogle Scholar; Skorupski, John, “Quality of Well-Being: Quality of Being,” in Crisp, Roger and Hooker, Brad, eds., Well-Being and Morality (Oxford: Oxford University Press, 2000), 244fGoogle Scholar. I am unhappy with these treatments (to the extent that they frame themselves as reconstructions or explications of Mill—which, to be sure, not all do). Their focus tends to be on what the right mathematical model for Millian higher pleasures is: for example, whether higher pleasures are infinitely more valuable than lower ones. But, first, Mill did not himself think about these problems by trying to find mathematical models for them; rather, he explored psychological implementation issues. The complaint is not just that much of the current discussion is anachronistic, but that some of the disagreement in it is merely apparent. Since Mill did not think in these terms himself, any model that reproduces the right outputs, in this case lexical preferences, is as good as any other. Second, although Mill does occasionally appeal to infinities (especially in his treatment of natural kinds in A System of Logic, book I, chap. vii, sec. 4, in Mill, Collected Works, VII:122–26), our comfort level with the concept of infinity, and our willingness to treat it as a reliable mathematical tool, is a side effect of that set theory class we all took in college. Before Cantor, infinity was felt to be a philosophical and mathematical problem, not a resource. So appealing to its infinitely greater value to explain why an object of preference is lexically higher-ranked than another is perhaps useful shorthand for us, but unlikely to be following Mill's own train of thought. However, I will have occasion to discuss one of the themes of this literature below, namely, whether Mill's distinction is, as the parties to the dispute tend to put it, qualitative or merely quantitative.

⁸ Mill, A System of Logic, book III, chap. xvii, sec. 1 [VII:525].

⁹ Mill, A System of Logic, book I, chap. v, secs. 4–7 [VII:97–108]. The replacement was not thorough, and I will return to the distinction between impressions and ideas in Section VII below.

¹⁰ Paul Thagard's quasi-connectionist networks, in which each node represents a proposition or goal, are very close in spirit to Mill's associationist models of cognition. (In full-fledged connectionist networks, contents are not located at single nodes.) See Thagard, Paul, “Explanatory Coherence,” Behavioral and Brain Sciences 12 (1989): 435–67CrossRefGoogle Scholar; and Thagard, , “How to Make Decisions,” in Varieties of Practical Reasoning, ed. Millgram, Elijah (Cambridge, MA: MIT Press, 2001)Google Scholar.

¹¹ More carefully, a desire consists in the idea of pleasure associated with the idea of the object of the desire. Mill took this definition over from his father, and noticed that it wouldn't do as it was. (James Mill, Analysis of the Phenomena of the Human Mind [London: Longmans, 1869], vol. 2, 191f.; I will refer to this work below as the Analysis. See p. 258 for a similar definition of “motive.”) Even if you specify that the object of the desire lies in the future, what you get looks like hope or wish or even just expectation, rather than desire. However, Mill never adjusted the account of desire he had inherited to handle the objection.

¹² Carlisle, Janice, John Stuart Mill and the Writing of Character (Athens: University of Georgia Press, 1991), 18fGoogle Scholar.; Mill, Autobiography, chap. 3, par. 5, in Mill, Collected Works, I:71, 73. A reading group in which Mill participated subsequently worked its way through the Analysis (Mill, Autobiography, chap. 4, par. 20 [I:127]). In a eulogizing passage in the Autobiography, Mill praises his father for “what he achieved in … analytic psychology…. [H]e will be known to posterity as one of the greatest names in that most important branch of speculation” (ibid., chap. 6, par. 11 [I:213]).

¹³ Mill, Analysis, vol. 2, 215, which also adduces power and dignity as examples; John Stuart Mill endorses them as “almost perfect” (233n.). See also ibid., vol. 2, 188 (on “money … hugged as a good in itself”), and 233n. on how “persons, things, and positions become in themselves pleasant to us by association; and, through the multitude and variety of the pleasurable ideas associated with them, become pleasures of greater constancy and even intensity, and altogether more valuable to us, than any of the primitive pleasures of our constitution … as the love of wealth….” (The point of the extra documentation here and below is to demonstrate that the account of miserliness is not just a throwaway; it turns up again and again, and it was evidently important in the younger Mill's thinking.)

¹⁴ Ibid., vol. 2, 266. Mill notes the qualitative difference in the resulting feeling at 321. (Because John Stuart Mill's notes are quite long—sometimes as long as twenty pages—in cases like these I will give the page rather than the number of the note.) Here is his father's version of the explanation: “Money, for example, instrumental in procuring the causes of almost all our pleasures, and removing the causes of a large proportion of our pains, is associated with the ideas of most of the pleasurable states of our nature. The idea of an object associated with a hundred times as many pleasures as another, is of course a hundred times more interesting” (ibid., vol. 2, 206f., endorsed yet again by the son at 236n.).

Discussion of the psychological phenomenon in question is not confined to the Analysis. In his Principles of Political Economy, John Stuart Mill gives a related explanation for the once-popular economic doctrine of mercantilism: “As it is always by means of money that people provide for their different necessities, there grows up in their minds a powerful association leading them to regard money as wealth in a more peculiar sense than any other article; and even those who pass their lives in the production of the most useful objects, acquire the habit of regarding those objects as chiefly important by their capacity of being exchanged for money. A person who parts with money to obtain commodities, unless he intends to sell them, appears to the imagination to be making a worse bargain than a person who parts with commodities to get money; the one seems to be spending his means, the other adding to them. Illusions which, though now in some measure dispelled, were long powerful enough to overmaster the mind of every politician, both speculative and practical, in Europe” (Mill, Collected Works, III:505f.).

Here is further discussion in An Examination of Sir William Hamilton's Philosophy: “[A]ssociation can generate new mental affections. Let us take, as one of the obvious examples, the love of money. Does any one think that money has intrinsically, by its own nature, any more value to us than the first shining pebbles we pick up, except for the things it will purchase? Yet its association with these things not only makes it desired for itself, but creates in many minds a passionate love of it, far surpassing the desire they feel for any of the uses to which it can be put” (Mill, Collected Works, IX:284n.).

Finally, in a letter to Theodor Gomperz, Mill recommends “arguing questions [in economics] at first on the supposition of barter, in order to adjourn the difficulties which arise from the wrong and confused associations which cling to the idea of money” (Mill, Collected Works, XV:859).

¹⁵ Is this enough to account for a miser's generally preferring money to what money can buy? After all, no one has had the opportunity to build up associative links of the sort we have just described around the ideas of each sum of money. I expect that we are to think of the miser either as having performed something like an induction (each amount of money that I have considered is more valuable than what it will buy, so all amounts are), or as entertaining a more indefinite idea of money (rather than one or another sum in particular). By Mill's lights, each of these options involves the miser in a further cognitive error—I mean, over and above the one I am about to describe.

Mill may be overlooking the supplemental reinforcement we are given for riches, power and reputation. Being wealthy has many social rewards, over and above the actual purchases it enables one to make, and these additional rewards ought, on the Mills' shared psychology, to have a further conditioning effect.

¹⁶ Mill, Analysis, vol. 2, 215; Mill, Collected Works, III:810.

¹⁷ If Mill were right about the psychological machinery, the mistake would be hard to avoid: the miser is falling afoul of the awkwardness with which connectionist networks generally handle trade-offs of this kind.

¹⁸ Albeit not quite the misunderstanding: the miser doesn't actually believe that he can buy more than his money is worth.

ATMs disburse one's cash, again and again and again, but we don't ever, as far as I know, find ATM-misers, people who would give up the money the ATM provides for ATM access privileges. On Mill's account, there should be at least some such people. The moral: Don't forget that this is the history of psychology, not necessarily a plausible psychology for us.

¹⁹ John Stuart Mill, Autobiography, chap. 5, par. 4, in Mill, Collected Works, I:143.

²⁰ How can someone who wants to use preferences as the bottom line, and who gives this sort of account of how they are shaped, be in a position to regard some of them as mistaken? Mill is in an awkward position, one which may or may not be sustainable. Since our concern here is whether the awkwardness is a reason to think we are misreading him, notice that it is a variant of a position standardly adopted by informed-desire theorists. Such informed-desire or informed-preference theorists take desires or preferences to be the bottom line: something is good for you because you desire it, and not, e.g., because it is objectively valuable. Typically, such theorists insist that practical reasoning consists exclusively in means-end reasoning, but they are unwilling to insist that your desires and preferences never need correcting. So they take the benchmark to be the desires you would have if, say, you knew more. Mill is executing the same maneuver, using the preferences of other, more experienced people as his way of allowing for a preference to turn out to be mistaken. Mill's reasons for doing it his way, rather than taking the currently more popular counterfactual-based approach, are reconstructed in Millgram, Elijah, Ethics Done Right: Practical Reasoning as a Foundation for Moral Theory (Cambridge: Cambridge University Press, 2005), 82f., at n. 27CrossRefGoogle Scholar.

²¹ Mill, Autobiography, chap. 1, par. 7 [I:13–15].

²² The younger Mill quotes his father on the role of praise and blame in parental pedagogy in the Analysis, vol. 2, 314, and remarks, in a tone that conveys the deep impression it made on him, on “the desire [James Mill] made [the minds he came in contact with] feel for his approbation, the shame at his disapproval” (Mill, Autobiography, chap. 4, par. 7 [I:105]). Compare also ibid., chap. 5, par. 4 [I:141], on his teachers' overreliance on “the old familiar instruments, praise and blame, reward and punishment.”

²³ Mill, Analysis, vol. 2, 298f.

²⁴ Mill, Autobiography, chap. 1, par. 22 [I:35].

²⁵ The quotes in this stretch of argument are from Mill, Utilitarianism, chap. 5, par. 24 [X:250f]. As the gesture at informed-desire theory in note 20 may remind us, the appeal to what, as a matter of psychological fact, people prefer is meant to sidestep question-begging appeals to what they should prefer. Nonetheless, some such circularity may remain in these arguments. Presumably, there are conceivable environments in which justice and liberty would not turn out to be higher pleasures: perhaps those in which justice and liberty are consistently associated with electric shocks. (Think of the dogs in learned helplessness experiments; see Peterson, Christopher, Maier, Steven, and Seligman, Martin, Learned Helplessness [Oxford: Oxford University Press, 1993]Google Scholar.) Why do we disregard those environments when we are considering which pleasures count as higher, especially since those classifications are being deployed in an argument about what social environments are to be brought about?

²⁶ Mill, An Examination of Sir William Hamilton's Philosophy [IX:183, 197]. The antecedents of these counterfactuals—in this example, going back into the room—are naturally to be understood as bearing phenomenalist analyses also. See Lewis, Clarence Irving, Mind and the World Order (New York: Dover, 1956), 135fGoogle Scholar., for a brief statement of that more recent reformulation of Mill's view.

²⁷ Mill, An Examination of Sir William Hamilton's Philosophy [IX:179f.].

²⁸ They do not always matter more: we often trade in a material object for the sensations, as when one consumes a piece of cake. So material objects and sensations do not quite or always stand in the higher pleasure/lower pleasure relationship. But here there is a straightforward explanation: once the piece of cake has delivered the sensations and pleasures of taste, it is gone.

²⁹ In Proust's more exotic variation on this theme, even something on the order of the memory of a baked good can, if sufficiently connected to other hedonically charged ideas, be made into the focus of a personality and a higher pleasure. More familiarly, gourmets enjoy their food as they do because of the discriminations they are able to make; they read Bon Appetit the way other people go window-shopping; it is common enough in some circles to find oneself at dinner with people whose culinary enjoyment is inseparable from the conversation they are having about the food they are eating, the food they have eaten on other occasions, and the food they mean to eat in the future. For the importance of the ability to construct new associative links to an idea, see Millgram, Elijah, “On Being Bored Out of Your Mind,” Proceedings of the Aristotelian Society 104, no. 2 (2004): 163–84, secs. 2–3CrossRefGoogle Scholar.

³⁰ Mill, Utilitarianism, chap. 2, par. 5 [X:211].

³¹ This is perhaps the reason that Mill is willing to treat the higher pleasures as a class: since they do not come in amounts, Mill may be concluding that they do not trade off against each other. If this is what Mill was thinking, however, I am not happy with it: different higher pleasures do trade off against one another, and instances of the very same higher pleasure can trade off against each other (as when you have the option of sacrificing some people's liberty to allow liberty to many others). Many people surely have preferences over such trade-offs, and Mill discusses one such trade-off himself, in On Liberty, chap. 3, par. 9 [XVIII:266].

I hope to discuss this subject further on another occasion. For now, notice that one sort of standard objection is evidently miscast: the objection that no one is always going to give up eating for Mozart. Mozart isn't normally a generic means to eating, and so Mozart and eating don't stand to each other in the relation we have been examining.

³² Mill, On Liberty, chap. 1, par. 11, in Mill, Collected Works, XVIII:224.

³³ To be fair, however, “here” is the System of Logic (book VI, chap. X, secs. 2–3 [VIII:912ff.]). Mill does go on to affirm his “belief … that the general tendency is, and will continue to be, … one of improvement…. This, however,” he insists, “is not a question of the method of the social science, but a theorem of the science itself.” The concept is introduced in the chapter of the System of Logic titled “Of Progressive Effects; and of the Continued Action of Causes” (book III, chap. xv [VII:509–15]). Compare also ibid., book V, chap. v, sec. 4 [VIII:790f.], where we are told that history shows “Man and Society” to be “actually undergoing a progressive change”; Mill uses the point as a premise in an argument against overreliance on empirical laws. In On Liberty, chap. 3, par. 17 [XVIII:272f.], Mill tells us that “the progressive principle … in either shape, whether as the love of liberty or of improvement, is antagonistic to … Custom.” Although he uses “progress” and “progressive” in their ordinary senses later in the paragraph (“we flatter ourselves that we are the most progressive people who ever lived”), the point is that “the despotism of Custom,” by imposing stasis on a society, may make it no longer subject to forms of development characteristically produced by progressive causes. Compare the related use of the term in Utilitarianism, chap. 2, par. 24 [X:224], where the fact that the human mind is “in a progressive state” explains improvement in our practical arts; it is not identical to the improvement.

³⁴ Mill, System of Logic, book III, chap. xxiv, sec. 9 [VII:620]; compare ibid., book III, chap. xxiii, sec. 3 [VII:593f.]: “To enable us to affirm any thing universally concerning the actions of classes of human being, the classification must be grounded on the circumstances of their mental culture and habits, which in an individual case are seldom exactly known… .”

³⁵ The implementation account blocks the obvious objection to the conclusion: that if pleasures are unpredictable, then, for all we know, it is constraint rather than liberty that maximizes pleasure. As before, we are bound to have our own doubts about whether a variety of options makes it more likely that people get what they want; it has become an iconic complaint that there are five hundred channels, but nothing to watch. But, as before, we should not forget that associationism is a defunct psychological theory. How realistic should we expect the psychological process we have just sketched to be?

³⁶ This is not, by any means, a fully satisfactory account even of the partial solution: it is difficult to cash out in terms of Mill's psychological machinery, because we do not have good associationist models for pleasures being linked to indefinitely many other unspecified pleasures. We do not have a model for a higher pleasure persisting because one expects older and fading lower pleasures to be constantly replaced by equally ephemeral pleasures that are still vivid. And we do not have a good explanation for why, once all the lower pleasures have been introspectively sand-blasted away, the generic means to them should persist as a pleasure. That is, we do not have the right psychological gloss on the “permanent” in Mill's famous phrase, “the permanent interests of man as a progressive being” (On Liberty, chap. 1, par. 11 [XVIII:224]). For perhaps the best discussion of Mill's attempts on the philosophical puzzles he took the Crisis to have raised, see Vogler, Candace, John Stuart Mill's Deliberative Landscape (New York: Garland, 2001)Google Scholar; for another interesting reading of it, see Paul, Laurie, “The Worm at the Root of the Passions: Poetry and Sympathy in Mill's Utilitarianism,” Utilitas 10, no. 1 (1998)CrossRefGoogle Scholar.

If only the higher and hardwired pleasures persist, why isn't the effect of analytic introspection to produce a sort of scaffolding of a personality, while stripping away the personality itself—that is, to leave in place a concern for large abstractions like liberty, justice, and so on, while removing any concern for the personal and concrete goods (other than eating and the like) one would use one's liberty and security to obtain? That prospect gives Mill a stake in understanding great literature, poetry, art, and so on to be higher pleasures. Doing so gives him an initial answer to the charge that this solution to the problem of analytic introspection leaves only hollow lives. And it gives him an answer to the worry that sticking with the higher pleasures, which are shared by everyone, amounts to an endorsement of conformity: even if everyone loves liberty, and even if everyone loves great literature, they will have different favorite works of great literature.

For the meantime, and in the service of sharpening the remaining difficulties, Mill announces, in the course of introducing the decided preference criterion, that “[i]t is better to be a human being dissatisfied than a pig satisfied; better to be Socrates dissatisfied than a fool satisfied” (Utilitarianism, chap. 2, par. 6 [X:212]). The human and Socratic faculties are presumably valued in virtue of the processes we are discussing: because they function as generic means to further pleasurable activities and experiences. But the faculties are presumed to be valued even when the pleasurable activities and experiences do not (or no longer) ensue. And that is presumed to be the case even though, as we saw, most people do not become misers, because they come to see that they are not actually going to get most of the goods of which the money makes them think.

³⁷ This sense of “scalar” is being borrowed from Williams, Bernard, “Persons, Character, and Morality,” in Williams, , Moral Luck (Cambridge: Cambridge University Press, 1981), 6–8CrossRefGoogle Scholar.

³⁸ Mill claims that it is “practically important to consider whether the feeling itself, of justice and injustice, is sui generis like our sensations of colour and taste, or a derivative feeling, formed by a combination of others” (Utilitarianism, chap. 5, par. 2 [X:240f.]). Mill's attempt to trace the special qualities of the intense feeling to, especially, a qualitatively distinctive thirst for revenge (ibid., chap. 5, paragraphs 18–22 [X:248–50]; Mill, Analysis, vol. 2, 325f.) looks like a throwback to Humean analyses of complex ideas, which claimed to reveal the impressions of reflection embedded in them (see Millgram, Ethics Done Right, 222–24, for examples from Hume). For other points at which Mill toyed with the thought that the higher pleasures (i.e., the mental states involved in the enjoyment of higher pleasures) might be qualitatively distinctive, see also note 14 above.

³⁹ Vogler, Mill's Deliberative Landscape, 74–77, 80–82. For a quick overview of Vogler's agenda, see my review of her book in Ethics 112, no. 4 (2002): 880–83.

⁴⁰ Mill, Analysis, vol. 1, 90–91 (James Mill's discussion), and 108–9n. (John Stuart Mill, quoting his own treatment in the Examination of the Philosophy of Sir William Hamilton).

⁴¹ Although he still seems to have thought of ideas as fainter or weaker than sensations, Mill thought it necessary to amend Hume's view that impressions and ideas are distinguished only by their degree of vivacity; Mill's official view is that the difference between them (as between beliefs, memories, and other sensations or ideas) has to be treated as a primitive.

⁴² In the Analysis, vol. 2, 252n., 254n., Mill emphatically recommends volume 2 of Ruskin, John's Modern Painters (1843–1860; London: George Allen, 1906)Google Scholar. That is, Mill was apparently also exploring art criticism for guidance on the higher pleasures, but must not have found the solutions he was seeking. If he had, we would see more of Ruskin's quite startling account worked into Mill's later theorizing.

⁴³ Mill, Autobiography, chap. 7, par. 20 [I:259].

⁴⁴ Mill, On Liberty, chap. 3, par. 2 [XVIII:261].

⁴⁵ Ibid., par. 10f. [XVIII:267].

⁴⁶ Ibid., par. 6 [XVIII:264f.]; compare John Stuart Mill, The Subjection of Women, chap. 3, par. 14, in Mill, Collected Works, XXI:312f.

⁴⁷ Elements of the argument are scattered through On Liberty, chap. 3, paragraphs 6, 14–16 [XVIII:264f., 269–72]. Perhaps this is one reason why Mill fretted over the workings of attention. Vogler, Mill's Deliberative Landscape, 94ff., usefully discusses Mill's worries about that subject.

As the previous section suggests, there may be other Millian explanations as well. When you find something pleasurable—not independently, as it were, but merely because you had desired it—the psychological machinery presumably operates as follows. To desire a slice of mince pie, for instance, is to associate an idea of the slice with the idea of pleasure. Now, the idea of pleasure is not the sensation of pleasure that you would have, if you liked the pie on its own. When you taste the pie, some of the vividness or energy of the sensation (of the pie, not of pleasure) is transmitted to the idea of pleasure, and is likely to make it somewhat more strongly felt; but even a more vivid idea is much weaker than a sensation. To desire something because it is what other people want is to have, when such a desire is satisfied, a pleasure of this secondhand sort. So preferences or desires that arise merely from conformist tendencies will eventuate in weak satisfactions.

A caveat: Recall that Mill was aware that his father's analysis of desire was unsatisfactory; it is thus hard to know how much weight can be rested on the workings of the admittedly flawed psychological machinery.

⁴⁸ Mill, On Liberty, chap. 3, par. 4 [XVIII:263].

⁴⁹ We should not forget, as commentators all too often do, that the argument of Mill's On Liberty is therefore not an allowable part of the liberal arsenal until the archaic machinery has been satisfactorily replaced.

⁵⁰ Mill, The Subjection of Women, chap. 2, paragraphs 4, 12, and chap. 4, paragraphs 4f., 10, 13f. [XXI:288f., 293–95, 324–26, 329, 331–33].

⁵¹ John Stuart Mill, Considerations on Representative Government, chap. 10, paragraphs 1–3, in Mill, Collected Works, XIX:488; Collected Works, XV:558.

⁵² Ibid., chap. 2, par. 20 [XIX:390].

⁵³ See, e.g., Mill, On Liberty, chap. 1, par. 10 [XVIII:224]; compare Mill, Autobiography, chap. 5, paragraphs 15, 19 [I:169, 177].

⁵⁴ And political theorists do get called on such pronouncements. See, e.g., the complaint, in Berkowitz, Peter, Virtue and the Making of Modern Liberalism (Princeton, NJ: Princeton University Press, 1999), 30fGoogle Scholar., that Judith Sklar's claims about the “salutary effect [of life in a liberal regime] on the characters of citizens” are backed up “with scarcely a shred of empirical evidence.”

⁵⁵ Tinbergen, Nikolaas, The Study of Instinct (Oxford: Oxford University Press, 1976)Google Scholar.

⁵⁶ For the differing national characters of Frenchmen and Englishmen, see, e.g., Mill, Autobiography, chap. 2, par. 12 [I:59f.].

⁵⁷ Mill does say, in a letter to Alexander Bain, that “[e]thology [is] a subject I have long wished to take up, … but have never yet felt myself sufficiently prepared” (Collected Works, XV:645; the letter is dated 1859).

⁵⁸ Mill, On Liberty, chap. 3, paragraphs 1–2 [XVIII:260f.].

⁵⁹ There are traces in Mill of a further and spectator-driven train of thought: that original and uncustomary characters will be “noble and beautiful object[s] of contemplation,” and that “in proportion to the development of his personality, each person becomes … more valuable to others” (On Liberty, chap. 3, par. 9 [XVIII:266]). That is, Mill seems to be toying with an argument that experienced judges would find other uncustomary individuals to be a pleasure, and perhaps even a higher pleasure. Notice that the ethological sciences would not be in a position to supply the evidence for this argument, either. If you cannot say what original characters are like in general, you cannot demonstrate what the reaction to them will in general be.

There is a half-hearted attempt to show that originality is an all-purpose means, and thus, presumably, a higher pleasure: everything we value now was once an innovation, and so we should value innovators very highly indeed. As Mill noticed (On Liberty, chap. 3, par. 12 [XVIII:268]), and as Vogler, Mill's Deliberative Landscape, 108, has also pointed out on his behalf, a conformist will not find this a persuasive argument: future innovation is a means to what a conformist does not want. Think about hip-hop, from the perspective of a music lover of fifty years ago: an offensive vehicle for misogyny and braggadocio, confused about whether a record player is a sound-reproduction device or a musical instrument, it's not (the music lover would have insisted) even music. When novelty really is unpredictable, it does not, ahead of time, look like a benefit.

⁶⁰ Williams, Bernard, In the Beginning Was the Deed, ed. Hawthorn, Geoffrey (Princeton, NJ: Princeton University Press, 2005), 45fGoogle Scholar. Recent attempts to make the inculcation of virtue an object of public policy display what some of these obstacles look like in real-life politics. For some of the discussion by policy-oriented academics, see Galston, William, Liberal Purposes (Cambridge: Cambridge University Press, 1991)CrossRefGoogle Scholar; Berkowitz, Virtue and the Making of Modern Liberalism; and, for an overview, Kymlicka, Will, Contemporary Political Philosophy, 2d ed. (Oxford: Oxford University Press, 2002), chap. 7Google Scholar.

Notice that the civic virtues actually being promoted are not nearly as hard to describe concretely as those we have been considering. “Autonomy” and “toleration” are cookie-cutter objectives compared to fostering genuinely surprising personalities. Where creativity is the announced policy goal, we get phenomena like the National Endowment for the Arts; even the tiny sliver of the federal budget allocated to the NEA is controversial and consequently tenuous.