The primary target of G.E. Moore’s criticism in Principia Ethica1 is John Stuart Mill’s defense of Utilitarianism. In this essay I will consider the possibility of defending Mill against Moore by focusing primarily—if not exclusively—on Mill’s account of psychological hedonism. In doing so, I will first offer brief summaries of Mill’s account of the principle of utility (PU) (§I) and Moore’s claim that in defending PU, Mill commits the “naturalistic fallacy” (§II). Then, I will discuss how Mill’s account of psychological hedonism may be used in an attempt to side-step Moore’s criticism (§III). Finally, I will show that because Mill’s account of psychological hedonism is inconsistent with remarks he makes elsewhere in Utilitarianism about the “ascetic on his pillar,” the method of defending Mill discussed in Section III fails (§IV).
At the outset of Utilitarianism Mill notes that “From the dawn of philosophy, the question concerning the summum bonum, or, what is the same thing, concerning the foundation of morality, has been accounted the main problem of speculative thought” (Mill, p.1). It is with this problem in mind that he attempts to establish PU—i.e. the theory that actions are “right in proportion as they tend to promote happiness; wrong as they tend to produce the reverse of happiness” (Mill, p.7)—as the first principle of morality. According to Mill, happiness—i.e. “an existence exempt as far as possible from pain, and as rich as possible in enjoyments, both in point of quantity and quality” (Mill, p.11)—is the only thing that humans do desire and, hence, the only thing that is desirable.2 Given this view of PU, Mill treats questions of right and wrong as purely instrumental questions about the consequences of our actions. In judging the moral praiseworthiness or blameworthiness of an action we need not focus on the character of the moral agent—contra Aristotle—nor on the subjective principle of volition whereby the action is performed—contra Kant—but rather, on the consequences of the action and whether they represent the greatest amount of happiness or pleasure for all those involved.
In analyzing Mill’s account of PU we must be careful to distinguish between the following logically distinct claims:
T1: Pleasure is something individuals desire.
T2: Pleasure is the only thing individuals desire—psychological hedonism.
T3: Pleasure is the only thing individuals ought to desire—ethical hedonism.
T4: The pleasure of the greatest number of people is the only thing individuals ought to desire—group hedonism.
Although Mill explicitly offers evidence in support of (T1), it is so trivially true as to require little discussion. And while it is clear that he argues for the truth of (T2), as well, it is entirely unclear whether he is also arguing for (T3) as is commonly assumed. In Section III, I am going to consider the possibility that Mill is solely concerned with establishing the truth of the descriptive claim (T2) that pleasure is—as a matter of psychological fact—the only thing we desire, and not with establishing the normative claim (T3) that it ought to be the only thing we desire. Nevertheless, determining the soundness of (T4)—where Mill moves from psychological hedonism (T2) to group hedonism (T4)—would take us beyond the scope of the present essay. But insofar as Mill’s ability to defend (T4) indirectly depends on his first demonstrating the truth of (T2), in focusing primarily on the merit of this claim, we will inadvertently be addressing the merit of (T4) as well.
In Principia Ethica, Moore attacks moral philosophers such as Mill who maintain that “ethics is an empirical or positive science: [that] its conclusions could all be established by means of empirical observation and induction” (Moore, p.25). Moore’s criticism—which at least in this one respect is simply an extension of Hume’s fact/value distinction—is that philosophers are not justified in deducing normative propositions about what we oughtto desire from descriptive propositions about what we do desire.3 It is not that Moore disagrees with the practical conclusions of Mill’s PU, he simply questions whether or not these conclusions follow from the premises offered in support of them.
One of Moore’s biggest criticisms of Mill concerns the following analogy from Utilitarianism:
The only proof capable of being given that an object is visible is that people actually see it. The only proof that a sound is audible is that people hear it; and so of the other sources of our experience. In like manner, I apprehend, the sole evidence that it is possible to produce that anything is desirable is that people do actually desire it. If the end which the utilitarian doctrine proposes to itself were not, in theory and in practice, acknowledged to be an end, nothing could ever convince any person that it was so. (Mill, p.34)
According to Mill’s analogy, to claim that an object is visible is to claim that the object is capable of being seen. And the only way to show that an object is capable of being seen is to show that people can and do see the object. That something is seen is proof that it is visible and likewise with things that are audible. In both cases, the appeal to empirical evidence proves that the respective objects are seen and heard—and ipso facto—capable of being seen and heard.
However, according to Moore, because Mill uses “desirable” normatively and not merely descriptively, his use of “desirable” is not analogous to his purely descriptive use of both “visible” and “audible.” As Moore says, “The fact is that ‘desirable’ does not mean ‘able to be desired’ as ‘visible’ means ‘able to be seen’. The desirable means simply what ought to be desired or deserves to be desired” (Moore, p.67). So, by Moore’s lights, Mill is not simply making the psychological claim that we do desire pleasure (T1) or that pleasure is the only things that we desire (T2). Rather, he is making the stronger normative claim that pleasure is what we ought to desire (T3). Given Moore’s reading of Mill’s analogy, even if we grant that pleasure is the only thing desired (T2), it would not logically follow from this fact alone that pleasure ought to be the only thing desired (T3). If Moore is correct in assuming that Mill is making normative claims about the desirability of pleasure based solely upon empirical claims about the desiredness of pleasure, then his Humean objection to Mill’s analogy is indeed a strong one. Of course, as I suggested earlier, this assumption may very well be incorrect.
Mill readily admits at the outset of Utilitarianism that insofar as PU is the first principle of morality, it is not susceptible to proof in the “ordinary and popular meaning of the term” (Mill, p.4). Normally, when we say something has been proven, we mean that it logically follows from other claims or principles. But, first principles, insofar as they are first principles, cannot possibly be susceptible to this type of logical proof. There can—by definition—be nothing prior to a first principle. And insofar as Mill believes PU is a first principle, he naturally rejects the idea that it could be proven in the strict sense of the word. Nevertheless, Mill believes that there is a “larger meaning of the word ‘proof’ (Mill, p.4),” and he claims that “considerations may be presented capable of determining the intellect either to give or withhold its assent to the doctrine; and this is equivalent to proof” (Mill, p.5). So, let us consider what I am going to call Mill’s indirect proof that “happiness is desirable, and the only thing desirable as an end” (Mill, p.34).4
Mill’s indirect proof of PU consists of two distinct steps: First, he uses the aforementioned analogy between “visible/audible” and “desirable” to establish the banal claim that we do desire pleasure (T1). Second, having established (T1), Mill attempts to show that pleasure is the only thing we desire (T2). This second and more controversial step of the indirect proof is fueled by the claim that people do—as a matter of psychological fact—desire nothing except to the extent to which it is conducive to pleasure (Mill, p.35). This is where we find Mill’s account of psychological hedonism—the truth of which he feels so strongly about he notes “that to desire anything except in proportion as the idea of it as pleasant is a physical and metaphysical impossibility” (Mill, p.38). By Mill’s lights, pleasure is not one of several ends or goals of human conduct, but rather, it is the only end. As he says, “Human nature is so constituted as to desire nothing which is not either a part of happiness or as a means to happiness, and we can have no other proof, and require no other, than that these are the only things desirable” (Mill, p.35). And because Mill believes that he has successfully shown that people desire happiness (T1) and that they never desire anything else (T2), he believes he has offered the only proof of PU that is possible.
Seemingly aware of the initial implausibility of (T2), Mill anticipates the likely objection that humans do desire other things—e.g. virtue, money, power, and fame—besides pleasure as ends in themselves. First, he considers the example of virtue. According to Mill we did not originally desire virtue disinterestedly as an end in itself, but rather, we desired it solely as a means to the attainment of happiness. Through time, virtue has “come to be itself a principle ingredient of the individual’s conception of happiness,” (Mill, p.36) and is now desired not merely as a means to happiness, but is desired “for its own sake” (Mill, p.36). But doesn’t this undermine the truth of psychological hedonism? Mill tries to resolve this tension with his account of associationism, whereby we come to desire virtue as a “good in itself” through the direct association of virtue with pleasure.
When Mill concedes that virtue is desired “for its own sake,” he is simply acknowledging that virtue has gradually been adopted as a “part of happiness” (Mill, p.36). He relies on this associationism to explain the other possible counter-examples as well. Money, power, and fame have all become “part of happiness” via their direct association with pleasure. Mill concludes, “It results from the preceding considerations that there is in reality nothing desired except happiness. Whatever is desired otherwise than as a means to some end beyond itself, and ultimately to happiness, is desired as itself part of happiness” (Mill, p.37). We can assume that by “the preceding considerations” Mill means (T1) and (T2).
On this reading, the indirect proof is not an attempt to demonstrate that we ought to desire pleasure (T3) at all, but merely an attempt to show empirically that humans do desire pleasure alone as an end-in-itself (T2). As Mill says:
I believe that these sources of evidence…will declare that desiring a thing and finding it pleasant…are phenomenon entirely inseparable, or rather, two parts of the same phenomenon—in strictness of language, two different ways of naming the same psychological facts. (Mill, p.38)
Further comments such as “there is in reality nothing desired except happiness (Mill, p.37)” and “people never desire anything else [except happiness] (Mill, p.35)” lend additional support to this interpretation of the indirect proof. Given the truth of (T2), pleasure is the only thing that we can and do desire, and hence, it is the only thing that is desirable (i.e. capable of being desired for its own sake). If this is correct, Mill is not using “desirable” normatively—as Moore assumes—but merely descriptively. And if his indirect proof is an attempt to demonstrate the truth of (T2) and not the truth of (T3), then Moore’s criticism loses much of its force. After all, (T2) doesn’t say anything about what we ought to desire. It is solely concerned with what humans seemingly must desire.
Perhaps a useful analogy could be drawn between Mill’s indirect proof and Aristotle’s “demonstration by refutation” of the principle of non-contradiction (PNC).5 With his demonstration of PNC, Aristotle is not trying to convince someone that they oughtto believe in the truth of PNC, but rather, he is trying to show them that they already do and must so believe. Similarly, Mill’s indirect proof could be viewed as an attempt to demonstrate that everyone necessarily desires pleasure alone as an end in itself; after all, according to Mill, it is a “physical and metaphysical impossibility” for anyone to desire anything else. If this is the correct reading of Mill’s indirect proof, then the proof is not and could not be aimed at someone who genuinely doesn’t believe in or act according to PU, precisely because it is impossible for anyone to actually be in such a position. So, Mill is not trying to persuade us that we ought to act in accordance with PU, but rather, he is simply reminding us that we already do act this way.
And if (T2) is true, what use could Mill—or anyone else for that matter—have for ethical hedonism (T3)? If we must necessarily desire pleasure alone as an end in itself, as Mill explicitly suggests, what sense does it make to say that what we ought to desire pleasure? After all, “ought” and “ought not” only make sense in the absence of necessity or compulsion. So, if Mill’s indirect proof is simply aimed at demonstrating the truth of psychological hedonism (T2), then he is not subject to Moore’s charge that he fallaciously derives an “ought” from an “is.” After all, on this interpretation, Mill’ claim that it is both physically and metaphysically impossible to desire anything except pleasure seemingly renders any subsequent concerns about what we ought to desire superfluous. Given the truth of (T2), Mill has no reason to argue for the truth of (T3). Nevertheless, if we are going to side-step the “naturalistic fallacy” in this fashion, we should make sure that Mill’s account of psychological hedonism is both coherent and consistent.
Upon close examination, we find that Mill’s defense of (T2) contradicts claims he makes elsewhere in Utilitarianism about the practices of the ascetic. Ascetics, by Mill’s lights, are individuals who believe that self-mortification and self-torture should be practiced for their own sake, and not for the sake of any useful end:
Unquestionably it is possible to do without happiness…it has often been done voluntarily by the hero or the martyr, for the sake of something which he prizes more than his individual happiness. But this something, what is it, unless the happiness of others or some of the requisites of happiness? ...but, after all, this self-sacrifice must be for some end…All honor to those who abnegate for themselves the personal enjoyment of life when such a renunciation they contribute worthily to increases the amount of happiness in the world; but he who does it or professes to do it for any other purpose is no more deserving of admiration than the ascetic on his pillar. He may be an inspiring proof of what men can do, but assuredly not an example of what they should. (Mill, pp.15-16)
Here, Mill explicitly distinguishes between two different types of renunciation: the first—and morally praiseworthy—type involves sacrificing one’s own happiness for the greater happiness of others. The other type of renunciation involves sacrificing one’s own happiness without increasing the “amount of happiness in the world” (Mill, p.15). It is this latter type of renunciation that is practiced by the ascetic and morally condemned by Mill. But, if, as was suggested earlier, the truth of psychological hedonism renders talk of how we ought to behave vacuous, on what grounds can Mill claim that the “ascetic on his pillar” ought not to desire self-torture for its own sake? If Mill is conceding that the ascetic is capable of renunciating pleasure—and it seems that he is—this admission undermines not only his own account of psychological hedonism, but the indirect proof of PU that depends on it as well.
The most likely response to this objection is going to involve an appeal to Mill’s aforementioned associationism. However, the example of the “ascetic on his pillar” is not analogous to the earlier ones. The virtuous person, for instance, could only come to desire virtue “for its own sake” by first desiring it as a means to pleasure. The same can be said about the lovers of money, power, and fame, as well. Associationism only makes sense when the objects in question are themselves conducive to pleasure. But it is seemingly unreasonable to suggest that the ascetic originally desired self-mortification and self-torture as a means to pleasure. Indeed, torture and mortification are paradigmatic examples of things we would normally say are contrary to pleasure and happiness. According to the common use of “pleasure,” the ascetic receives no pleasure from his unusual practices. In fact, the very goal of these practices is to cause pain and suffering. But, if the admitted existence of people who willingly torture themselves does not undermine psychological hedonism, what would? Just as Mill is seemingly unable to directly prove psychological hedonism, we are seemingly unable to falsify it. And if nothing is going to count against the truth of psychological hedonism, one must begin to wonder whether the theory can serve as the foundation of an empirical moral science as Mill hopes.
So, where has this discussion left us? On the one hand, Mill explicitly states that it is a physical and metaphysical impossibility for someone to desire anything other than pleasure as an ultimate end. On the other hand, he explicitly allows that the “ascetic on his pillar” does just that. If I am right, then anyone who intends to defend Mill will have to do one of the following: a) rescue Mill’s psychological hedonism by explaining the paradox of the ascetic who can seemingly do the impossible, or b) undermine the force of the “naturalistic fallacy” without appealing to what would otherwise be an internally inconsistent account of psychological hedonism.
1 G.E. Moore, Principia Ethica, Prometheus Books (New Jersey, 1988). Principia Ethica was originally published in 1902. All references to Mill will be to this work.
2 Throughout this paper I will be following Mill in using “pleasure” and “happiness” interchangeably.
3 Ultimately, Moore’s objection hinges on the indefinability of “good,” but for our purposes we need only be concerned with his objection to Mill’s attempt to derive an “ought” from an “is.”
4 Although I am calling it an indirect proof, there is really nothing indirect about it. The name is simply intended to reflect the fact that Mill explicitly denies that PU is subject to any type of direct proof.