Jan Narveson (essay date 1979)

Download PDF PDF Page Citation Cite Share Link Share

SOURCE: “Rights and Utilitarianism,” in New Essays on John Stuart Mill and Utilitarianism, edited by Wesley E. Cooper, Kai Nielson, and Steven C. Patten, Canadian Association for Publishing in Philosophy, 1979, pp. 137-60.

[In the following essay, Narveson explores the conflict between justice and utility in the thought of J. S. Mill.]


Few questions about utilitarianism have been more vexed than that of its relation to rights (and its associated notion, justice). It is commonplace to hold that there are nonutilitarian rights, rights not founded on considerations of utility. And it is even thought that the very notion of rights is inherently incapable of being significantly employed within the utilitarian framework. In the present paper, I wish to consider both of these matters. I propose to give reasons—mostly not really new—for rejecting the stronger, conceptual claim; and on the former, substantial question, I want to argue that the utilitarian at least has a respectable way of handling the vocabulary of rights, even a useful one. Further, I shall argue that what the utilitarian—in particular, J. S. Mill—has to say about rights is quite plausible, though I shall conclude with a question about how illuminating it is.

Ideally, we would begin with a detailed and thorough analysis of the notion, or perhaps we should say the notions, of rights. But this is scarcely possible within the limits envisaged here, and something more skeletal will have to do for present purposes. I will confine myself to three points here: correlativity, enforceability, and the distinction between positive and negative rights, this last being especially crucial in relation to the important subject of liberty. The three points are these.

(1) The normative content, as I shall call it, of rights claims is always to be found in statements about the obligations or duties of others: to say that A has some right or other is to say that some person or persons other than A have obligations or duties to do something or other in relation to A. (It might be in relation to some further person who is, e.g., a beneficiary, but if so, it is by virtue of A's relation to that person.) To attribute a right to A is not, as such, to say anything about what A ought to do, but only about what he may do; but the word ‘may’ here has its meaning in relation to the others in relation to whom A has the right: that A “may” means that these others must let him if he wants, or something further (cf. point (3) below).

(2) There is a connection between rights and enforcement or enforceability, which can as well be made stipulative for present purposes. Mill, as we know, claimed that morality in general was enforceable1, but if that is true, it will only be so by allowing rather attenuated methods of “enforcement” to count as enforcement.2 However, we can just say that what we are especially concerned with are rights which are genuinely enforceable, at least prima facie.

(3) We must be aware of the distinction between negative and positive rights, a distinction which is crucial to the whole subject, since ambiguities generated by talk of rights are largely due to failure to bring it out. The distinction is between (a) obligations not to interfere and (b) obligations to assist, help, or enable. This distinction is not the same as that between rights to liberty and rights to welfare or to property; rather, it cuts across that distinction. Whatever p may be, A's having the right that p

(This entire section contains 11526 words.)

See This Study Guide Now

Start your 48-hour free trial to unlock this study guide. You'll also get access to more than 30,000 additional guides and more than 350,000 Homework Help questions answered by our experts.

Get 48 Hours Free Access

p can consist either in others being obligated not to bring it about that not-p (negative right) or in their being obligated to help (to some degree) bring it about that p. My right to a house may be my right that you not damage, destroy, or use a house which I already have, or it can be a right that you help build one which I don’t have, or help buy it, etc. This significance of the distinction will emerge below, especially in sections 3 and 4 below.


With these preliminaries in mind, we now move to the question of how rights-talk fits into the utilitarian view. There are two general lines we can take on this, one a good deal more useful than the other. First, one can maintain that everybody (animals too, perhaps) has one basic right, namely the right to have one's utility counted equally with all others'. Second, and much more usefully, one can produce a schedule of more concrete rights which is to be grounded on the utility of having those rights recognized and adhered to in practice. The second is where “the action” is, but the first also deserves notice. I discuss it in this section.

In a lengthy footnote to the concluding chapter of Utilitarianism, Mill discusses an objection by Herbert Spencer to the effect that “the principle of utility presupposes the anterior principle, that everybody has an equal right to happiness”. He goes on to note that this idea “may be more correctly described as supposing that equal amounts of happiness are equally desirable, whether felt by the same or different persons”, and this, as he points out, is “not a pre-supposition … but the very principle of utility itself …”. He concludes that “If there is any anterior principle implied, it can be no other than this, that the truths of arithmetic are applicable to the valuation of happiness, as of all other measurable quantities.”3 (58) Now, Mill is surely correct on the two main points here. He is correct that everyone's having an equal right to happiness is not the most accurate way of expressing what the principle of utility either says or implies. Why? Because it seems to suggest that there is some equal amount of happiness to which everyone has a right; and this the principle neither says nor supports. Mill's correction of Spencer on this point may be taken to be definitive of utilitarianism as it is currently understood among professional philosophers: that equal amounts of happiness are equally desirable, no matter which sentient being experiences them. And he is correct in saying that this latter principle is not some further thing pre-supposed by utilitarianism, but rather “the very principle itself”. But it is less clear (a) that what is presupposed is only that arithmetical truths apply to quantities of utility, and (b) just where the notion of a right to utility, in some sense or other, should be brought in, if at all. A word about each of these is in order.

(a) Is anything “presupposed” by the principle of utility? If that principle is simply stated as Mill has it above, that equal amounts of utility are equally desirable, then it is arguable that a further assumption is needed, namely that desirability is all that counts, i.e., that right actions are actions which maximise what is, all things taken into account, desirable. This I take it, is what makes utilitarianism a “teleological” theory in the usual classifications of philosophers. A deontological theory, by contrast, is supposed to be one which says that something else besides desirability counts—that there are other sources of moral rightness than desirability (or, if you like, read ‘goodness’ for ‘desirability’). Now, if this is logically open to dispute, then on the face of it there are two claims made by utilitarianism, and not just one: (i) that desirability is all that counts, and (ii) that desirability is measured exclusively by utility. There is, of course, disagreement, or at least unclarity, about just what utility is, but those disagreements would issue in different versions of utilitarianism, rather than rejections of utilitarianism, whereas rejections of either (i) or (ii) would yield rejections of utilitarianism. And here again, those who argue against utilitarianism on the score of its handling of rights may take the line that rights are not essentially concerned with desirability at all; or they may (but less interestingly, in my view) argue that rights are resistant to utilitarianism on the second score, holding that they are responsive to values other than utility.

Why is the second score less interesting from the point of view of the status of rights in moral theory? Partly because when we consider putative candidates for values other than utility, it is too easy to reply that these, after all, are simply some sources of utility among others; but more importantly because any supposedly other values introduced will behave so much like utility. To take an example which is not entirely a caricature, suppose that one espouses, à la Nietszche, a concern for what we might call “supermanhood”. Now Nietszcheans, some of them anyway, seem to hold that supermen count for more than others, And they seem to think also that egoism has something to do with it; but unfortunately, what it seems to have to do with it is that supermen get to be egoistic, unlike sheeppeople, who do not. But what if one superman, S1, has it within his power to kill another superman, S2, or more generally to do something, x, which would cripple his superman potential? And suppose that this other budding superman would in fact become more of a superman than the first one if this crippling act x were not performed. Now, ought S1, as a good Nietszchean, to perform x or not? Well, if he thinks that supermanhood is an intrinsically good thing, whether or not he himself exemplifies it, then he may have to back off from x—to sacrifice a bit of his own superman-potential in the interests of greater supermanhood on the whole. Apart from the supposedly nonutilitarian value of supermanhood, its arithmetic may well be the same as the utilitarians'.

More likely, though, is that the theorist in question thinks he has an objection along the lines of (ii) when in fact he has one along the lines of (i). For he may introduce a supposedly evaluative notion, “human dignity” or “autonomy” or “personhood”, say, which in fact is not another evaluative notion at all, but rather, a claim that values aren’t all that count. What counts instead, he maintains, is whose values are relevant to a certain decision. People have, he will say, the right to make their own decisions, not because they will likely be better decisions if made by the agent rather than someone else, but just because they are the agents. There is, it will be maintained, a certain sphere belonging, so to speak, to each individual such that within that sphere he may do as he pleases without the intervention of anyone else, and this not because any values of any kind will be promoted by having it so, but simply because that sphere is that agent's. If it’s my arm, e.g., then I get to decide what is to be done with it, and this is so whether I use it stupidly or intelligently, whether I promote one value, another value, or none by my use of it. Clearly this is to introduce a purely deontic element into the picture: it is to say that whose values (if any) may properly be brought to bear on a given decision is to be determined by factors quite other than how much of any sort of value is thereby promoted. To this kind of view we shall return later in this paper.

(b) Meanwhile, what about a “right to have one's utility counted”? Is this something the utilitarian can countenance, or should? There are two aspects of this which we should distinguish. One is whether framing the principle of utility in such terms suggests considerations which go beyond the moral ambit of this theory. The other concerns what we might call the arithmetic of the theory. Both call for comment.

(1) Regarding the first, there is a continuance, in brackets, of the footnote by Mill which has occasioned this discussion, addressing itself to the question whether utilitarians, in addition to invoking “empirical generalisations from the observed rules of conduct”, may also avail themselves of “the laws of life and the conditions of existence”, thereby coming up with kinds of action which “necessarily tend to promote happiness”. And Mill agrees that they certainly may, taking exception, as a good empiricist, only to the term ‘necessarily’. Now, if there are fundamental “laws” of human nature relating certain kinds of conduct to happiness, those might be the foundation for specific concrete rights—a possibility we will advert to later in this essay. But what Mill does not do here is to address himself to the rather different question whether the principle of utility is basically empirical at all. The point is this. If we ask why an individual would or should be inclined to embrace considerations of pure utility, counting the utility of others equal in its impact on his own conduct to the utility of himself, then what answers might be produced? Mill's answer to this, insofar as he addresses himself to the matter at all, is that we have, or at least are capable of coming to have, a “feeling of unity with all the rest” (Ev. 30), stemming from a “deeply rooted conception which every individual even now has of himself as a social being” (31). But apart from the question whether we really do have this degree of fellow-feeling, and whether the kind of fellow-feeling we have really works in utilitarian fashion, there seems to be a further question: viz., why we should inculcate this feeling (which Mill does not argue is innate in us but only acquired, though “natural” (28-30)) rather than some other? The situation is especially acute in view of Mill's apparent embracing of what seems to amount to psychological egoism. In his notorious discussion of the proof of utility, as so many writers have pointed out, Mill's major premise is that “desiring a thing and finding it pleasant, aversion to it and thinking of it as painful, are phenomena entirely inseparable …” (36). But, although Mill does not come right out and acknowledge the matter, his terminology being apparently neutral on the score of whose pleasure can be in question in this premise, it would surely be extraordinarily implausible for Mill to be maintaining that for any given person, A, the thought that just anyother given person, B, is experiencing some pleasure is equivalent to A's desiring B to have that experience. Yet it is just that proposition which is required if we are to try to make the principle of utility fall out of the proof. Links seem to be required in order to connect the pleasure of an agent, which is what apparently motivates his actions on Mill's view (as it certainly was on Bentham’s), with the pleasures of all others. Well, what can provide such a link? One candidate would be an a priori idea that one ought to regard the pleasures of any person equally with those of any others. This in fact was the option embraced by Sidgwick, who arrives at “the self-evident principle that the good of any one individual is of no more importance, from the point of view of (if I may say so) the Universe, than the good of any other … And it is evident to me that as a rational being I am bound to aim at good generally … not merely at a particular part of it.”4 And one way of expressing this claim would be by saying that each person has a right that his utility be counted equally with that of all others.

What is of interest about this option is that it utilizes one of the ideas which has often been associated with the idea of rights, namely, that they are basic, a priori, and/or prior to considerations of the utility of others. The utilitarian cannot easily hold that the reason why we ought to account the utility of others as equal in value to that of our own is simply that the utility of others naturally and inherently must promote our own utility; both because that is implausible in itself and because it seems to make him a kind of egoist. It seems best for him to say that there is no further reason why the utility of others counts equally with our own—that it just does. But obviously this is more than the recognition that the “laws of arithmetic” also apply to quantities of utility. Some have, of course, denied that they do, in the sense that they deny the commensurability of interpersonal utility, the cardinalizability of measures of any utility, or both. But even if we grant both of those disputed points, there is still evidently a gap between my agreeing (a) that if I perform action x, the total utility for all persons concerned is n, whereas if I perform y, the total will be less than n, but the amount realized for me will be greater, and my agreeing (b) that I ought, then, to do x rather than y. If it makes any sense at all to “close” or “bridge” such a gap, the notion of ana priori right might be one way to bridge it. And it of course still leaves us with utilitarianism. So there is no conceptual impossibility about employing rights-talk as a utilitarian in this way. All that doing so will yield is room for a division between intuitionistic utilitarians and others (if there are any others); and that would be a dispute about meta-ethics, not about whether utilitarianism was correct as the substantial principle of ethics.

(2) A quite different issue, however, is also raised by the idea of expressing the principle of utility in terms of a right to have one's utility counted equally. For suppose we accept the idea that if A has the right to do x, then B must allow A to do x (at very least) and this entails enforceability; or, if this is different, that it entails the need for a substantial, rather than merely a very small, weight in whatever contrary consideration might override the right. This aspect of the idea of rights can, as will be discussed below, be effectively employed with regard to certain sources of utility rather than others. But can it be employed in a supposed right to have one's utility counted at all?

What we would need, to do so, is a distinction between cases where the agent goes wrong by virtue of having misestimated relevant utilities, or where although he estimates them properly he fails to act on the results, and cases where he simply doesn’t count some relevant people's utility at all. Now, perhaps there is such a distinction. But it is hard to see how it could be the basis for a major distinction in our response to such cases. After all, the difference in resulting utility might easily be just as great due to misestimation as to failures to estimate at all; and given the principle of utility, what must matter is just the size of the difference, rather than its source. What the source is will be relevant to the determination of just what kind of corrective action to take (if any), but it is not obvious that any major difference in the degree of severity (say) with which we should treat such agents could be correlated with that factor. Indeed, when we reflect on this, we can readily identify a disposition which would be worse, in utilitarian terms, than the disposition to fail to count some persons' utility: namely, a disposition to promote disutility rather than utility. What is wrong with the malevolent person is not that he fails to count some people's utility, but rather than he counts it in, so to say, the debit column rather than the credit column. Now, it’s not entirely clear what not “counting” someone's utility consists in. When people are close at hand, it must be a little difficult to ignore them entirely; and even if they aren’t, provided that one is aware that they are in the picture at all, it’s surely not too easy to pay no attention whatever to the question of how they might be affected by actions which do affect them. But one would still suppose that we could expect worse from someone who delights in the sufferings of others than from someone who doesn’t go out of his way to determine how others will be affected. And if we want to say that the “right to have one's utility counted” is really the right to have it positively counted, i.e. positively evaluated on an equal basis, then the difference between that and simply reasserting the principle of utility is evidently zero.

In this second aspect of rights-talk, then, it seems at least misleading, at best pointless, and at worst incoherent to talk of a right to have one's utility counted.


We turn, then, to the second and, I have said, “more useful” way of bringing rights into utilitarianism. This consists in identifying various types of actions, or dispositions, and signalling them out for special protection, or in identifying various sources of wellbeing which are thought to be more essential or more important than others, and granting positive rights to those who might readily fail to come in for enough of those goods otherwise. Both call for comment. (Indeed, both, but especially the former, could generate—have generated—books.); so too does the general idea. I shall discuss these in reverse order, first considering the general idea, then each category of rights within the utilitarian framework, following through with a case study of a putative nonutilitarian theorist who, I believe, turns out to be utilitarian despite himself.

(1) The general plan for utilitarian-based specific rights is implicit in Mill's treatment. To grant a right to someone is to grant protection to him in that respect: that is to say, to legitimize the use of some degree of coercion in preventing interferences with what he has the right to do or have. But these coercions have a cost: for those upon whom they fall are either injured or damaged in some degree, or at least deprived of some liberty, since they are either disenabled from doing certain things they want to do, or at least obstacles are put in their way. We are, in short, visiting some actual or potential disutility on violators or potential violators of the right. This is the cost of rights. In order for society to be justified in erecting them, therefore, the various actions which qualify for righthood must have at least enough utility (i.e., interference with them must have enough disutility to the subjects of them and to others) to offset the cost. It is for this reason that it would be a crude error to imagine that any action, however trivial the difference in net utility between doing it and not doing it, will automatically qualify for righthood on utilitarian theory, and partly for this reason that it is an equally crude error to imagine that actions which do qualify for rights on this theory may be interfered with, and hence the right in question infringed or violated, for any slight gain in net utility supposed possible by comparing simply that action, taken in the abstract, with some other and incompatible configuration of acts. For if the expected utility from a certain type of act as compared with some incompatible configuration of acts were so nearly the same, it could not have qualified for righthood in the first place.

There is another, more problematic reason for the latter. Once you institutionalize a particular right, you of course also alter people's expectations in regard to that kind of activity, and so you increase the loss in utility from that right's being violated. One can no longer simply turn around and say that the incompatible configuration of actions has more utility, viewed abstractly. For we can no longer view it so: now we have to take into account the disruptive effects of doing things differently from what has been expected. Even so, of course, the right can be outweighed sometimes. But that may be an advantage for the theory, not a disadvantage. Few theorists seriously maintain that no concrete right may ever be justifiably contravened. The question is only whether utilitarianism makes this too “easy”; which remains to be seen. It is obvious, though, that a right which can be routinely overridden is no right at all, practically speaking. So if utilitarianism were to lead to the implication that any right can be routinely overridden, then it would be correct to claim that utilitarianism should dispense with the notion altogether.

(2) Positive rights are conceptually more comfortable on the utilitarian theory than any other (plausible) one. To have a positive right is for it to be the case that others must, if needed, assist one in doing or having the thing in question. It is sometimes wondered just who are the subjects of the duties which positive rights impose. The answer is not far to seek, given utilitarianism: in general they fall on everybody, more or less in proportion to ability to help out and thus equally on those equally able. But of course it will be the case that division of labor is called for: some will help out directly, others by paying salaries (for instance) to those who do so, and by being ready to contact the right people when needed, and so on. In addition there is a specific answer, for those who are in the best position to help will, of course, have the duty to do so; though this could be onerous for certain persons, and for those persons, other people will have the duty to organize a system for relieving them or compensating them. The puzzles which some have alleged to find probably stem from failure to distinguish three relevant kinds of actions: (a) organizational and stimulative actions, that is, actions designed to get others into action as opposed to actions directly responsive to the needs in question; (b) those direct actions themselves, such as the application of artificial respiration or the supplying of food or shelter; and (c) supporting actions, such as the paying of money to professionals in the (a) and (b) departments. The duty of each and every individual will be to address himself to these rights and do what he can best and most easily do.

But when will there be positive rights? Again in form, the answer is simple enough. A will have a positive right against B if the amount which A gains by being enabled to do the action or be in the condition in question, less the amount which B loses by not only doing the enabling deed but also by being required to do it, gives more overall utility than would stem from not requiring B to do it. Well and good, but when is that? It must not be assumed that it is quite so easy for all those conditions to be realized as is often supposed by critics of utilitarianism. For to begin with, consider the prospect of spending one's entire life tending the sick, feeding the hungry, and so on, rather than the other things you could be doing with your time. The potential cost in quality of life may surely be enormous here. Of course, it may not. But suppose on the contrary that some people are attracted to these activities, would enjoy them and find their lives highly rewarding if devoted to them; and suppose that many others are so sympathetic with these activities that they will voluntarily give money and other kinds of help to make them possible. In that case, it may well be that the needs in question are pretty well catered to without imposing a duty on everyone to help out. Possibly it will be sufficient simply to inculcate a general disposition to approve and encourage those who do get more involved. It is never, of course, self-evident that any particular concretely specified action is right on utilitarian principles, and here is a case in point.

(3) It is fundamental in these matters that we realize that positive rights are incompatible with absolute negative rights. If you must do certain things, then you do not have the right not to do them if you don’t want to. This will therefore militate against positive rights to the extent that there is a case for negative rights. Our utilitarian case for positive rights, in particular, is constrained by the utilitarian case for negative ones. The more utility we suppose there is to be got from granting negative rights, the less there can be from granting positive ones.

It is true, of course, that any granting of rights at all deprives people of some liberty, namely the liberty to interfere with what the right is a right to. Still, if it is possible to make a clear formulation of there being rights to liberty, so that the liberty we are deprived of by granting these rights is only the liberty to interfere with others' liberty, then it can be coherently argued that pure negative rights maximize liberty insofar as it is possible to do this for more than one person.

But is it possible to do this coherently? J. S. Mill, as we know, attempted to formulate a principle of liberty which looks at first sight to be purely Libertarian, in the contemporary sense of that term: i.e., it seems to grant only negative rights. “That principle is, that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection. That the only purpose for which power can be rightly exercised over any member of a civilized community, against his will, is to prevent harm to others …” (Ev. 73). Though not devoid of unclarities, the force of this principle looks to be Libertarian, for it not only declares that the good of an individual, A, is not sufficient reason to prevent A from doing as A wishes, but it also implies that the good of any other individual or set of them is likewise not such reason; for it allows only the prospect of harm to others, rather than the prospect of less promotion of their good than otherwise, as reason to interfere. As D. G. Brown points out in his superb analysis of Mill's view on liberty and utility,5 Mill can only salvage certain things which amount to positive rights on his principle by fudging the distinction between harming and not helping. If we unfudge that, the Principle of Liberty appears to yield no duty to help others, even others in dire need and at trifling cost to oneself.6

There is another aspect of the Libertarian idea in Mill, however, which urgently requires attention if we are to get as far as the conclusion just noted. This has to do, in effect, with private property. That this is relevant to ideas of liberty may seem at first sight astonishing, but it surely is so, as can be seen by reflecting briefly on, for instance, the principle of free speech and discussion. For it will surely be agreed to by anyone, even the most determined socialist, that whether a given person, A, may hold forth with a given speech, S, depends on whether, for instance, he is proposing to do it in your bedroom at 2:00 A.M. A theorist may want to hold that it is the duty of society to provide places in which A may deliver his speech, either when he pleases or at least at some specified times; another theorist may even want to hold that it is the duty of society to provide an audience to hear him out, however puerile his tirade. But those theorists are not, any longer, defending free speech as a negative right. They are instead holding it to be, to some degree at least, a positive right. In its pure negative form, which I believe is equivalent to asserting the right of free speech strictly as a liberty and not anything else, we must surely hold that it consists in being allowed to say what one pleases to whomever is willing to have you do so at the places and times in question. And that depends on whether those so willing are the ones whose will properly determines the disposition of those places and times. Is it my living room, and am I the audience? If so, then if I have permitted you to produce speech S there at time t, with me and others I am willing to have there who are willing to hear you utter S, then perhaps, your liberty to utter S is assured and absolute given the Libertarian principle. But in order for it to be so, it must be the case that the living room in question is genuinely mine. If it is not, then others have, in principle, some voice in the matter of whether you are to be allowed to give your speech there; and to that extent, you will not be entirely free to do so, even in the circumstances just specified.

This other aspect of the libertarian idea emerges in Mill when he discusses the bearing of social interests on various of our actions, allowing that when our actions essentially affect these interests, we may then be held responsible for our inactions as well as our actions. But, says Mill in a notorious passage, “there is a sphere of action in which society, as distinguished from the individual, has, if any, only an indirect interest; comprehending all that portion of a person's life and conduct which affects only himself, or as it also affects others, only with their free, voluntary, and undeceived consent and participation … This, then, is the appropriate region of human liberty”. (75) Various critics have pointed out that it is not plausible to claim that there is any “region of action” which is inherently incapable of affecting others, and if allowing genuine liberty is contingent on there being any such region, then libertarianism is surely an ill-fated idea. But we should be clear that what Mill is in effect doing is arguing that the “inward domain of consciousness” is the private property of the person whose consciousness it is: that person, and no one else, may properly determine what goes on in that “sphere”. The region of “tastes and pursuits”, another on Mill's list, obviously can get us out into the public arena where these pursuits can bump up against the pursuits of others, and Mill's argument can only be construed to defend ‘absolute’ liberty in those actions up to the point where they do so; appropriately, he deduces a right of association next, for if I do x in your presence and you do y in mine, then this will be consistent with liberty only if we both agree to this—and only if our “presences” are ours. But what if the ground on which we stand belongs to someone else? To say that it does is to say that that someone, call him A, may declare us out of bounds and require us to move on, and no principle of freedom of tastes and pursuits can be invoked against this unless the property in question is not entirely that person’s.

Now, any action requires the use of something or other by the actor: perhaps only a mind, as in the case of thinking, perhaps also an arm and a head, as in scratching one's head, and so on. In every case, to assert that one may do as one pleases, i.e. that one's own decision is a sufficient condition for the action decided upon to be legitimate, is ipso facto to assert that the things involved in the action, at least during the time when the liberty in question is a right for the person in question, do belong to that person. And if for any reason one thinks they do not, then to that extent one cannot coherently hold that the person in question is fully at liberty, has an unqualified right, to do the thing in question.

If the doctrine that literally nothing but harms to others licence interference with liberty, i.e. Libertarianism, is to get off the ground, then what seems to be required is that everything which can figure in any human action be divided up and assigned to distinct individuals. In form, the doctrine will get off the ground if this is done in an arbitrary manner, but certainly the doctrine will lack credibility unless it is done in a non-arbitrary manner. Mill does not go into this, nor does Brown in his analysis. In part, of course, Mill doesn’t because he isn’t really a Libertarian, being instead a utilitarian bent on championing some liberty as being in the interest of mankind. The interesting question is whether he could still have been arguing plausibly if he had gone farther, sticking to the evident meaning of the principle of liberty as he states it and swallowing the implication that we do not have the sort of duties which he supposed it obvious that we had, such as “to give evidence in a court of justice; to bear his fair share in the common defence, or in any other joint work necessary to the interest of the society of which he enjoys the protection; and to perform certain acts of individual beneficence, such as saving a fellow-creature's life …” (74). Or at any rate, the implication is that these are not rock-bottom duties, but instead, people would have them only insofar as they have voluntarily taken them, for instance as members of some organization which assumes these responsibilities. To do all that, I think, would require completing the doctrine about private property. And to argue that Libertarianism is actually supported by utilitarianism, one must in effect argue that it is in the public interest to have everything private: that the sum of human happiness (or whatever one takes utility to be) will be maximized if with respect to every particular thing there is some particular person or voluntarily acting group of persons such that what is done with that thing, within the limits of the rights of others, is determined entirely by the will of that person or group.

I certainly have no intention here of arguing for that conclusion, but I do want to point out that it is not logically impossible, at any rate, to proceed in that fashion. It is not, for example, incoherent to argue from public morality to private property just because the former is public and the latter is private. All morality, I take it, is public, as a matter of form: for something to be morally wrong is for it to be a sort of thing which there is a publicly-considered reason for condemning. But this does not logically imply anything about how much, if any, of the domain of humanly usable objects should be reckoned as in the private sphere. Its being in the private sphere means that we all ought to respect the will of the person who is the thing's owner, and the reason, if there is one, why we all ought to do so will itself be a publicly-considered one. Further, that publicly-considered reason could be that it is in the public interest to run things that way rather than some other. This, indeed, was the sort of argument given by Adam Smith on behalf of noninterventionism in the economic realm; and whether the argument was correct or not is open to question, but if it is wrong, it is not so on purely logical grounds.

But how might such an argument go? There are two general lines available. First, one can argue on extrinsic grounds, for example by pointing out that when usable entities are “public”, then the determination of what is to happen to them is evidently going to have to be political, that politics is invariably messy, inefficient, even downright incompetent, that it involves enormous demands on the time and energy of the body politic which could well be better spent in other ways, and that it is fraught with dangers, including especially the dangers of harassment of innocent people and misuse of the overwhelming powers of government which is virtually necessitated by their being powers of that kind. One can argue that when activities are run as businesses, there are people who have a genuine and personal interest in the flourishing of those businesses, an interest, moreover, which compels them to cater to the people they serve rather than high-handedly telling those people what they ought to want and to remove their caps and stand humbly while waiting in line to get it. But even those arguments are not purely “extrinsic”, looked at at all closely. They edge up toward the second sort of arguments, those on intrinsic grounds. These are arguments to the effect that to be at liberty, to have a sphere in which others cannot tell you what to do (but can only suggest), a sphere in which you do what you please, in short, is a good thing; as did Mill. But there are two versions of this argument the difference between which we need to take careful note of. One way is also exemplified by Mill, perhaps, for Mill, having stated that he has no brief with “abstract right as a thing independent of utility” and regards utility as “the ultimate appeal on all ethical questions”, goes on to say “but it must be utility in the largest sense, grounded on the permanent interests of man as a progressive being” (74). But what about the interests of unprogressive beings, beings who would rather sit and drink beer and watch television than participate in the higher activities of progressive mankind? Mill's official doctrine, actually, quite plainly defends the right of the unprogressive to be so as well; but then we may have our doubts about the amount of weight we should want to put on considerations of man as a “progressive being” in the ultimate doctrine by which we want to defend that right. It would be better, I think, to leave ‘progressiveness’ out of the picture at that level, and instead merely talk of the actual preferences and satisfactions of people—their interests as seen by them, rather than their “true interests” as seen by some high-brow social theorist.

And if we take this second option—the “lower” as opposed to the “higher” way, as we might call it—then it seems to me that the defense of liberty on intrinsic grounds is likely to be easier rather than harder. People want to do what they want to do, and not (necessarily) what anybody else wants them to do; and prima facie, then, they shall have more of what they want if their wanting to do it is taken to be a sufficient reason for others not to interfere. Obviously, there will be cases in which a person's long-run wants conflict with his short-run wants, and cases where his ignorance of relevant information creates a problem, since it may be that what he would want given that information and what he now claims to want also conflict. Here, notoriously, we get into the manifold problems of aggregating utility even at the individual level. Nevertheless, we do not clearly need to defend liberty “as an abstract right independent of utility”. That there is a straightforward if prima facie argument at the ordinary level of what I have called “low-brow utility” is surely not to be ignored.

But it is, certainly, a good way from there to the defense on utilitarian grounds of liberty as a right. Granted that each person, A, seems prima facie to be in for more utility if A does what he wants rather than what others do, it does not follow that A wouldn’t be in for still more utility if other people also did what A wants rather than what they want to do; nor is it self-evident that what each person gains from accepting the liberty principle generally will out-weigh what he loses from doing so.

Not self-evident, agreed. But a couple of reminders will push things along in that direction some considerable distance. For there is, remember, freedom of association, which follows straight off from the general idea of liberty. Now if a whole lot of people should think that they would be better off subscribing to certain positive rights and duties, then why wouldn’t those people join together in a community, conditions of entrance into which involved likewise subscribing to those principles? And if those people did do that, while others who disagreed with the wisdom of those principles remained nonsubscribers in their Lockean state of nature, doesn’t that look like going in the direction of each sort of party being happier? Doesn’t that seem more plausible than cramming the duty down the throats of all?


Within the confines of this essay, it is, of course, not possible to flesh out arguments of the kind I have just been briefly sketching. Instead, I hope to advance our appreciation of the situation regarding rights and utility by examining with some care a recent contribution on the subject—in fact, a contribution to a recent volume in this series: “Rights, Utilities, and Contracts”, by Alan Goldman.7 This essay is intended to show “how rights are related to utilities by showing how they arise from interests”, the thesis being that the “first purpose of recognizing rights is to protect certain interests of individuals against utility maximizing or additive calculations regarding (other) interests of others”. (123)8 “… we consider certain interests qualitatively more important than others, e.g. our interest in being able to say what we please more valuable in its satisfaction than our interest in not being offended. This ranking holds even though utility may lie occasionally on the side of silencing someone when many others take offense at what he intends to say … Part of the rationale.. is to protect these important interests against being overridden by additions of those less important.” (123)

So it seems that what determines whether a given interest is also a right is how important it is. If so, of course, the next question would be what measures “importance”. Prima facie, it would mean something like “more valuable”—which in turn usually means that it has more of what utilitarians call utility. Which in turn, of course, would usually mean that the point of recognizing rights is that there is more utility in doing so than in not doing so. But then, how is utility going to “occasionally lie on the side of silencing” the rightholder, for instance?

Or is this the view? Well, “It seems then that we recognize a qualitative two-fold distinction between types of interests, some generating rights and others remaining mere interests or utilities (when satisfied)”. (123, my underlining) Perhaps what we have here is an espousal of turn-of-the-century G. E. Moore-type intuitionism, with some interests being not only more substantial but uniquely (and inexplicably) more “valuable” than others? Thus we have mentions of satisfactions of some interests being “intrinsically more valuable than the satisfaction of others”, and of “intrinsic differences among interests” (127). But I don’t believe Goldman really means this, and if he does, then he has some explaining to do to the agents who participate in the “contractarian framework” which, at the close of the article, he invokes to shore up these needed distinctions; for these agents, as usual, are first, a la Rawls, put into a condition of equality and then told to negotiate schedules of rights in the interests of what would be “prudent”, (133), i.e., rational, i.e., most in the interests of the agent. Which suggests that the appeal to special “intrinsic values” is really, as one would expect, a fancy way of saying that they are very sizable interests. But wait a minute. Mightn’t there be “trivial rights” which it would be silly to claim ought never to be overridden by mere utility? Take, for instance, my right to water my roses. Mightn’t that be overridden by, say, the greater need of others for the water? Trouble is, if it’s a really sizable need—e.g., if others would die of thirst if I exercised this right—then it will turn out that my right is overridden by a greater right, the right to life, say: find sizable enough utilities, and of course Goldman will discern a right hiding under them. However, he has two other gambits. One is to distinguish between one's having a right and its being right to exercise it on a particular occasion. But the example he chooses to illustrate this is rather infelicitous: Goldman's “right to eat dinner when I like” which, if exercised at four o’clock, say, might “greatly disrupt the activities of the rest of my family. … Must my right be honored at their expense?” No, because “I can have a right to eat dinner when I like and yet be wrong in choosing to exercise it at certain times when that would inconvenience others.” (128) But what kind of right is this? Is it a positive right against the other members? If so, we might have our suspicions about how Goldman runs his family life, since its being a positive right against them would mean that if he wants to eat at 4:00, then they must scurry off, at his bidding, to round up the dinner—greatly to their inconvenience, indeed! Presumably, however, he doesn’t intend that. He probably means only to assert a negative right. These, we will recall, entail only a duty on others not to prevent him from having dinner at 4:00; and this, it seems reasonable enough, they do have. But it also means that if he wants his dinner at 4:00, he’s going to have to cook it himself, and maybe clean up afterward too! Even that might inconvenience others somewhat, of course, and so we might still agree that Goldman probably ought not to exercise this right. But this could reasonably be phrased in terms of a right on their part not to be inconvenienced, something which Goldman says would make “talk of rights lose all meaning here”. But why would it? His right to have dinner when he pleases is surely exactly on all fours with their right not to be inconvenienced, if we are going to talk about rights in this family context at all. The one is as trivial as the other, and after all, Goldman is supposed to be producing an example of trivial rights at work. I conclude that Goldman still doesn’t have an example of a nonutilitarian right somehow grounded on interests. What he has, if a right at all, is two utilitarian rights, both grounded in the general interest.

Goldman's other argument concerning “trivial” rights also has its problems. The argument is that “although these rights may not seem important in themselves, as parts of the general right to do whatever we please that is not morally forbidden, or that does not interfere with the rights of others, they represent an extremely important interest. This is our interest in being free from tyrannical interference with the everyday course of our lives, one which is more important than the satisfaction of our occasional desires to interfere in the course of others' lives.” (128) This, unfortunately, is either to compound muddle or to play into Mill's hands. To say that a “trivial” right of his own may not be overridden by a (merely) greater utility of someone else's when he has a “trivial” right to do something “which is not morally forbidden, or that does not interfere with the rights of others”, is question-begging, since what we want to know is whether others do have the right to override a trivial liberty of his on account of some urgent need on their part. And on the other hand, to say that we have a nontrivial interest in “being free from tyrannical interference with the everyday course of our lives” is simply to assert that what looks trivial is not so, on closer examination. For it is a case of something nontrivial (according to Goldman), namely our liberty to do as we please. If that is nontrivial, though, it means that it will take a substantial utility of someone else's to justify his interference: but if it’s a substantial utility, it will then be important and he will have a right to do it, and mine will thus be overridden, and so we are back in the utilitarian ballpark.

What has happened here is, I fear, that Goldman has fallen into the error discussed in section III above, viz., of supposing that on the utilitarian theory, the question whether something or other is a right, is identical with the question whether it, taken by itself, has more utility than something else which might interfere with it. But this, as we saw, is not the question. Rather, we are asking whether the utility we would get from forcing the person to do something else instead would exceed the utility of that other thing. That is why rights look unassimilable to utilitarian analysis, because it looks as though to recognize a right is to recognize that somebody may do something even if something else would have more utility. That is why we might be tempted to say that the very purpose of rights talk is to disallow “mere” utilitarian calculations. But it is a temptation we needn’t succumb to, for we would simply have omitted a vital element in the calculation. Your need to do Y, which would require my assistance and thus collides with my doing X, which is what I want to do and can do without your assistance, does not override my right to do x just because it is more of a need; for what it must outweigh is not simply the utility of my doing X—which it might well—but the utility of someone’s, or your, or everyone’s, forcing me to help, this having a substantial amount of disutility, which must be subtracted from the utility you get from doing your act Y. And it will take a sizable difference to make up for that, especially when the forcing is done by means of the elaborate and expensive machinery of the law.

This may also be the place to make a point about contractarianism, and about lexically ordered interests and summed lesser interests of others. Goldman, like many other critics, is worried about the possibility, which seems to loom large on a utilitarian format, that the lesser utilities of many people will add up to something which overwhelms the individually greater utility of one or a few. And he evidently thinks, though it is not entirely clear from the text, that rights talk is designed to prevent this, in particular by looking at the interests involved only on a pairwise comparative basis and then granting a right to the higher rank-ordered interest, regardless of the relative incidence of the two. (…“we cannot allow [rights] to be overridden by interests in the lower category” (126), as well as the frequently reiterated claim that sums of smaller interests can’t outweigh more important ones.) Further, a contractarian base is invoked to help this out, as it is in Rawls, who hopes to achieve the same sort of thing (e.g., “no one has a reason to acquiesce in an enduring loss for himself in order to bring about a greater net balance of satisfaction” (TJ, 14)). But it won’t work. For what a rational individual will do, according to the standard view (accepted by Rawls and, one assumes, Goldman), is calculate the probability of his being the gainer in an arrangement whereby the larger individual utilities of one or a few can be overridden by the summed lesser utilities of many, times the amount of benefit in question; then subtract the amount one would stand to lose by being the one overridden times the probability of being that one; and then compare the resulting figure with the utilities of the alternatives. And of course it will sometimes be rational to accept membership in such an arrangement. It all depends on what the probabilities are, given your position—and of course, if you are behind something like a Veil of Ignorance, thus being required to choose a position where you don’t yet know what your place in the system will be, then you’ll end up always opting for general utility, with its implication that the summed utilities of many can outweigh the individually greater utilities of few. Naturally, one must include in one's calculation the fear of being the one overridden, for security may be (normally is, with most people and with respect to some things) an extra consideration, over and above the disutility incurred on a given occasion of being the one outweighed. Even so, nothing like a lexical priority rule will emerge. Nor is this unintuitive, carefully applied. The enormous convenience of being able to use automobiles is certainly fairly small by comparison with the interest in not being violently killed; but we grant the right to drive to virtually all and sundry though we know that a few thousand people will be violently killed annually. (It helps that we can, with care, reduce the probability of being killed quite a lot: but not to zero, or even a vanishingly small figure.) Invoking veil-of-ignorance contractarianism will therefore do no good: the utilitarian camel's nose crops up behind the veil, and soon it is apparent that the whole camel is grinning away at us!

The general conclusion is: once admit impartially considered interests as the basis of rights, and from there on in it’s a losing battle against the pressure of utilitarianism. Whether we should admit that starting point at all is, of course, another matter: but Goldman has made his bed, and I suggest that a clearheaded look at the matter may yield the conclusion that lying in it would actually be rather nice.


This survey has been both rather long and yet, as I hope will be agreed, quite sketchy and nondefinitive. I wish to conclude it with a reflection which may go some way to explain the power of the utilitarian idea—and, perhaps, its futility.

The power of it is not far to seek. Consider some theorist who claims that justice ought to be done though the heavens fall, and that there are “abstract rights distinct from utility”; and let us ask this theorist which scenario for the future of mankind he would prefer as between S1, in which all of his favorite rights were carefully respected but somehow nearly everyone was really quite miserable, and S2, in which they were pretty generally violated but somehow nearly everyone was very happy indeed—a pair of scenarios which ought in principle to be possible (shouldn’t they?) if indeed justice and utility are distinct or even, as such theorists tend to think, opposed. I think it unlikely that any serious philosopher, however enthusiastic for rights, will actually pick S1. Why is this? I suspect that it is because it really would be absurd to say that not only are justice and utility opposed, but so too are justice and goodness. A theorist urging us to embrace strong schemes of rights could hardly be understood if he did not also imply that a world in which those schemes were respected would be a better world. But if he does say that, then he will be hard put to resist the plausibility of utility as the measure of betterness. Is it really possible that a human world should be better than another, but more miserable?

The catch is that the rights theorist may hold that the reason why people would be happier if their rights were respected is precisely that rights are so important to them. This, indeed, seems to be what nearly all theorists in the past few years do say. When they are busy “refuting” utilitarianism, their favorite gambit is to envisage some serious violation of what we generally allow to be rights because of the “general good”. Yet if rights are so important as they suppose, how can serious violations of them be consistent with the general good? Anybody can see that if citizens are routinely able to be hauled off to concentration camps in the middle of the night on account of some supposed failure to regard the “public interest” which amounts to a lot less than a similar violation of the rights of still others by the victim himself, then their rights are not respected. And any reasonable person would be shocked and frightened by this. How, then, is a “general good” to be got off the launching pad which is great enough to offset this manifest disutility—as it quite obviously is? No theorist seems, after all, to have tried out this argument: “See here, utilitarianism implies that people would be allowed to cut right across your front yard, right after you’ve freshly replanted it too, and this merely in order to keep the other half of town from being flattened by a gas-main explosion!” Nor, curiously enough, have they tried out this one: Suppose there’s a move to have a big parade, which lots of people and their children will enjoy; and the only feasible route is such that they’ll have to block Jones' access to his own driveway for a couple of hours. Jones is against it—he doesn’t like parades, and he finds the driveway-blockage a nuisance—but although he minds, he doesn’t mind very much. There will be a violation of his rights, but it’ll be a violation which doesn’t bother him seriously—just a little. Will the theorist now say that it would genuinely be wrong to have the parade anyway, despite this unfortunate but unavoidable (without cancelling the parade) violation? (Or if they would, would they also regard its being wrong as any real reason not to have the parade in this case?) If so, we should like to know about it! But in fact, the sort of violations of rights “owing to the calculus of advantages” which these theorists seem to have in mind aren’t quite like that. They are, instead, serious ones, where a great deal is at stake. But of course, the more is at stake, the more difficult it is to work up a plausible amount of “public utility” to outweigh it. And so, the stronger this argument sounds, the less plausible must it actually be as an argument against utilitarianism.

Nevertheless, it is possible that utilitarians have the cart before the horse. Perhaps the reason why people would be so unhappy if their rights were violated is that they really do have those rights, “naturally” and antecedently to any considerations of utility from any other course; but it’s just that they are highly concerned about those rights, and so rights are the source of the utility, rather than the other way around as the utilitarian supposes. Well, suppose this is true. What follows? Why, it will then follow that the utilitarian will be enthusiastic about respecting those rights! His reason for being enthusiastic, of course, will be that those rights are the source of so much utility; but so long as they are, he can hardly ignore them. The effect, in short, is that the utilitarian can readily agree with the most telling things his opponent thought he was marshalling against him, leaving the opponent with arguments of the “let justice be done, though the heavens fall” form—arguments to which the deontologist is quite welcome, thank you!

But this same argument will also explain, perhaps, the futility of utilitarianism in these matters. In order to see this, we must note a further point about the notion of rights, not brought out in the introductory Section I of this paper. This point is that there is a tendency to claim that talk of rights is not simply synonymous with talk of the obligations of others toward the right-holder. Some suppose that when Jones says, “Hey, you can’t take that—it’s mine!”, the ‘it’s mine’ part of his statement does more than just repeat the injunction not to take it. It is even thought that it does more than make some kind of reference to such things as how it came to be Jones' rather than someone else’s, or that it is Jones' rather than someone else's rights which are in question. Some think that the claim that something belongs to Jones, that Jones has a right to it, actually explains why others have the obligation not to use it without his permission. Others think that that is a mystification and an indulgence in obscurantist metaphysics. Whether or not it is that, though, it does bring up the point that when someone has a right, there has to be something about him or her in virtue of which that someone has the right in question. And in the case of utilitarianism in particular, we should be clear that this something cannot just be the fact that more utility on the whole would be expected by having people generally let Jones use or decide who uses the thing in question. For obviously there has to be some reason for supposing that there will be such greater utility, and this reason will have to be something about Jones. Now these facts about Jones, or whoever's rights are in question, these properties giving rise to rights within utilitarianism, cannot be “founded on” utility. On the contrary, utility must be “founded” on them. It is surely clear that we can have no utilitarian considerations unless there is a background of human psychological facts which are their basis: that X gives pleasure to A must, if true at all, be simply true, or due to some basic fact about our natures, and we can get results from a principle of utility only if we have intervening premises of that kind. Is it possible that premises about rights are also in that category? Not, if the above argument is correct, premises about rights which can genuinely override considerations of utility. But yes, if we mean premises about rights which identify, in effect, important sources of utility. If so, it is unclear whether the victory of utilitarians is hollow. Perhaps, instead, it would show merely that much of the argumentation about the alleged conflict of justice and utility is misplaced. And in any case, what Mill said would still be quite true—and worth quoting as the conclusion of this essay:

“It appears from what has been said, that justice is a name for certain moral requirements, which, regarded collectively, stand higher in the scale of social utility, and are therefore of more paramount obligation, than any others … Justice remains the appropriate name for certain social utilities which are vastly more important, and therefore more absolute and imperative, than any others are as a class … and which, therefore, ought to be, as well as naturally are, guided by a sentiment not only differing in degree, but in kind; distinguished from the milder feeling which attaches to the mere idea of promoting human pleasure or convenience, at once by the more definite nature of its commands, and by the sterner character of its sanctions.” (60)9


  1. Cf. Mill, Utilitarianism, Ch. 5: “We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; …” (Everyman edition of Utilitarianism, Liberty and Representative Government, p. 45.)

  2. Thus, continuation of the same sentence quoted in note 1: “… if not by law, by the opinion of his fellow-creatures; if not by opinion, by the reproaches of his own conscience.”

  3. Bracketed numbers attached to quotations from Mill all refer to pages in the Everyman edition cited above.

  4. Henry Sidgwick, The Methods of Ethics, 7th edn., (Macmillan, London, 1907), p. 382.

  5. D. G. Brown, “Mill on Liberty and Mortality”, Philosophical Review, 81 (1972), 133-58.

  6. Op. cit., pp. 142-6.

  7. In New Essays on Contract Theory, ed. Kai Nielsen and Roger A. Shiner, Canadian Journal of Philosophy, Suppl. Vol. III (1977), 121-35.

  8. Bracketed numbers in this Section refer to pages in the volume cited in the text.

  9. The ideas in this paper may be compared with some developed in my book, Morality and Utility (Johns Hopkins Press, Baltimore, 1967), especially Chs. VI-VIII, and in “Utilitarianism, Group Action, and Coordination”, Nous, 10 (1976), 173-94.

D. A. Lloyd Thomas (essay date 1983)

Download PDF PDF Page Citation Cite Share Link Share

SOURCE: “Rights, Consequences, and Mill on Liberty,” in Of Liberty, edited by A. Phillips Griffiths, Cambridge University Press, 1983, pp. 167-80.

[In the following essay, Thomas offers philosophical and moral justifications for Mill's liberty principle as contained in his essay, On Liberty.]

Mill says that the object of his essay On Liberty is to defend a certain principle, which I will call the ‘liberty principle’, and will take to say the following1: ‘It is permissible, in principle, for the state (through law) or society (through social pressure) to control the actions of individuals “only in respect to those actions of each, which concern the interest of other people” ’.2 The liberty principle is a prescription of intermediate generality. Mill intends it to support more specific political prescriptions, such as liberty of conscience, of expressing and publishing opinions, of framing a plan of life to suit our own character, and of combination for any purpose not involving harm to others (p. 75). The liberty principle is more general than these prescriptions but less general than its possible moral foundations, such as utilitarianism. My concern will be with attempts to defend the liberty principle by showing it to be supported by an acceptable moral position.

It is not my intention to offer a defence of a reading of the liberty principle in terms of ‘interests’, although I believe this is the most plausible construal of the text taking into account all of the formulations Mill offers. However, such a reading does call for an account of the term ‘interests’. A distinction will be made between ‘desire-based’ interests and ‘right-based’ interests. It will be argued that Mill must have intended ‘interests’ to be taken as ‘rights-based interests’.

To explain the notion of desire-based interests, let us suppose that each person has some set of non-instrumental desires: things desired not, or not only, as a means to something else. Let us further suppose that each can say which of his own set seem to him to be of greater significance. The most significant sets of non-instrumental desires may be expected to cluster around a person's more important activities, such as work (if not regarded merely as a means to other things) and various kinds of association with others. Such clusters of desires may be called a person's non-instrumental interests. Though based on desires, this account of interests allows for a distinction between a person's interests and his ‘mere’ likings, such as the desire to eat chocolate mousse, which may not be part of a cluster of non-instrumental desires a person regards as significant. And some of the things one is interested in (the career of Kiri te Kanawa) may not be part of one's interests, or something one has an interest (i.e. a stake) in. An account of desire-based interests must also include the notion of ‘instrumental’ interests: those things which, though they may not be desired as such, or even at all, are needed to secure a person's non-instrumental interests. Perhaps some instrumental interests are common to all people, such as health, while others vary from person to person according to what happen to be a person's non-instrumental interests. It is doubtful whether a person can be mistaken about his non-instrumental interests, but he can be mistaken about his instrumental interests: about what is in his interests.

Desire-based interests do not serve well for a reading of Mill's liberty principle. On this account actions concerning the interests of others would be those which impinge on another, either by way of affecting the pursuit of his non-instrumental goals, or by way of affecting his instrumental interests. Now whereas it is perhaps an exaggeration to say that any action would concern the interests of (some) others, some liberties Mill insists upon certainly would not be justified if we accepted this reading of ‘interests’. Many exercises of the liberty to express and publish opinions (e.g. those by Ralph Nader) concern the desire-based interests of others. As this reading of ‘interests’ leaves little of significance that may not be controlled by the state or society it is to be rejected, for it is clear that Mill intends his liberty principle to leave much of significance.

An alternative conception of interests is based, not on desires, but on rights. We think of persons as having rights to control themselves and (if we believe in private property) segments of the material world external to their bodies. A person has the right to decide who has access to his body, and no one else has this right, except with his consent. On the rights-based conception of interests, a person's interests are all those things over which he has rights of control. On this view every person has a sphere protected by rights in which choices about how things are to be in that sphere are reserved for that person.

The question ‘Does this action concern your interests?’ may receive different answers depending upon which conception of interests is being used. Whether your rich aunt will strike you out of her will is a matter that concerns your desire-based interests (unless you are already wealthy enough). But her decision does not concern your rights-based interests, for it is your aunt, not you, who has rights of control over her assets. In striking you out of her will she does nothing that concerns your rights-based interests, for by so doing she does not invade your protected sphere. If you steal a trifle from your aunt you do something that concerns her rights-based interests, but stealing so little may have no effect on her desire-based interests.

Reading ‘interests’ as ‘rights-based interests’, i.e. as those things over which one has rights of control, makes better sense of the liberty principle, because then it will rule out, in principle, external control of many significant actions. It is also a reading that fits several passages in the essay.

there is no parity between the feeling of a person for his own opinion, and the feeling of another who is offended at his holding it; no more than between the desire of a thief to take a purse, and the desire of the right owner to keep it (p. 140).

Holding an opinion is like owning something: you have rights of control over your opinions. That their existence is an annoyance (or worse) to me, and hence, possibly, in conflict with my desire-based interests, is of no more significance, so far as the rights of the situation go, than the fact that your owning a Gaugin I want has deshed one of my ambitions as a collector. One of a number of other passages which fit this construal of ‘interests’ is the following.

The only part of the conduct of anyone, for which he is amenable to society, is that which concerns others. In the part which merely concerns himself, his independence is, of right, absolute. Over himself, over his own body and mind, the individual is sovereign.3

If the term ‘interests’ in the liberty principle is to be read as ‘rights-based interests’ the next question is ‘What is the source of this account of rights?’ In at least one passage Mill appears to draw his account of rights from what are recognized as such in a particular society:

This conduct consists … in not injuring the interests of one another; or rather certain interests, which, either by express legal provision or by tacit understanding, ought to be considered as rights (p. 132).

But it is not clear that this passage must be read as having relativistic implications. For there are two ways in which it may be taken:

(i) Express legal provision or tacit understanding is what makes certain considerations into rights-based interests for the purposes of the liberty principle.

(ii) Certain considerations are interests, and their being so is not dependent on express legal provision or tacit understanding. These interests ought to be recognized for what they are—people's rights—either through legal recognition or tacit understanding.

The second way of taking the passage does not have relativistic implications.4

In any case, it is difficult to believe that Mill's considered position could be a relativistic one. For if it were it would follow that the significance of the liberty principle, in its applications, would vary from one jurisdiction (and set of social customs) to another. But Mill's stated objective is precisely to find a principle avoiding such variations.

… the practical question, where to place the limit—how to make the fitting adjustment between individual independence and social control—is a subject on which nearly everything remains to be done … Some rules of conduct … must be imposed … What these rules should be is the principal question in human affairs; but if we except a few of the most obvious cases, it is one of those which least progress has been made in resolving. No two ages, and scarcely any two countries, have decided it alike; and the decision of one age or country is a wonder to another (pp. 68-69).

There follows an attack on deciding the matter by reference to the influence of custom. Further, those rights-based interests recognized in some jurisdictions (even in England) would not allow Mill to draw from the liberty principle the more specific liberties he wants, for example, the liberty to utter words offensive to Christianity (p. 90). But clearly Mill thinks that his liberty principle requires liberty of expressing and publishing opinions under any jurisdiction, anyway under any in a society not in a ‘backward state’ (p. 73).

Rights drawn from law or convention traditionally have been contrasted with natural rights. It would appear, however, that natural rights cannot be Mill's basis for an account of rights-based interests, for he says:

I forgo any advantage which could be derived to my argument from the idea of abstract right, as a thing independent of utility (p. 74).

This passage is, I believe, incompatible with one account of natural rights. That account is to be found in places in Locke (though it is not his only account of natural rights), and also, more recently, in Nozick. On one Lockean view there is some characteristic(s) of each person such that possession of that characteristic implies each person's having a certain natural right. For example, that someone exists implies that he has a property in his own person. For this view of natural rights the label ‘perceived’ may be used, as it is thought that any rational person will be able to ‘see’ the connection between possession of the characteristic(s) and possession of the right.

This view of natural rights cannot be reconciled with the passage just quoted from Mill, nor with the more general consideration that Mill's ethical theory is consequentialist. A ‘perceived’ natural right requires no appeal to the consequences of its recognition as part of its justification, and therefore must be ‘a thing independent of utility’. But there is another conception of rights, with a good claim to the title ‘natural’ (in that it too may be contrasted with rights conceived as deriving from positive law or convention) which is compatible with Mill's position. Natural rights may be seen as related to justifiable practices. The requirements of a practice supply the connection between the characteristic(s) possessed by a person, and the claim that he has a certain right. The justification of the practice requires a comparison of the consequences there would be if that practice existed with the consequences if alternative practices existed, the comparison to be made in terms of some end(s), such as aggregate utility. (This way of justifying rights will be referred to as ‘practice consequentialism’.) On this view the connection between characteristics possessed by persons and rights must be made through a practice consequentially justified, whereas the justification of perceived natural rights does not require any reference to practices and their consequences.

We can use the term ‘assigned’ for natural rights justified in the practice consequentialist way. Persons are thought of as in an initial condition of ‘rightslessness’, and are then assigned rights on the basis of justified practices. A reluctance to allow that assigned natural rights really are natural rights is due perhaps to the belief that natural rights must be ‘natural’ in both of two ways. They should be ‘natural’ rather than ‘artificial’ where ‘artificial’ indicates a construction of positive law or social convention. They also must be ‘natural’ rather than ‘artificial’ where ‘artificial’ indicates that no ‘contrived’ processes of reasoning are necessary for their attribution: they are simply properties people have. Assigned natural rights are ‘natural’ in the first way, but not in the second.

As Mill does not forgo making use of the notion of abstract right, but only of abstract right ‘as a thing independent of utility’, he could make use of a conception of natural rights assigned on the basis of the consequences of practices assessed in terms of utility. The rights-based interests made use of in the liberty principle therefore may be those things over which a person ought to have rights of control by reference to the best assignment of rights from the point of view of the principle of utility. We may call such interests ‘utilitarian rights-based interests’. It should be noticed that there need be no connection between prescriptions justified by reference to utilitarian rights-based interests and prescriptions justified by the direct application of the principle of utility to desire-based interests. If the rule ‘Whenever one person's desire-based interests are in conflict with another’s, decide what is due to each by reference to what would maximize utility’, then no one would have any rights and, ergo, no utilitarian rights-based interests. Direct applications of the principle of utility to desire-based interests would leave few, if any, classes of action not in principle controllable by the state or society.

Should the term ‘interests’ in Mill's liberty principle be understood as ‘utilitarian rights-based interests’? Not if by ‘utilitarian’ is meant ‘desire-satisfaction utilitarian’, i.e. the view that a certain psychological state, satisfied desire, is the only intrinsically good thing, and that the goodness of anything other than satisfied desire is wholly dependent upon its relationship to satisfied desire, either by way of ‘producing’ it, being a precondition for it, etc.5

The reason why a conception of rights-based interests for use in the liberty principle cannot be grounded in desire-satisfaction utilitarianism is apparent if we consider, as an example, Mill's claim that we ought to have the liberty of ‘framing the plan of our life to suit our own character’ (p. 75). It follows from the formulation of desire-satisfaction utilitarianism that the consequential justification of a practice securing this liberty must be based solely on how granting or withholding it would affect the satisfaction of the desires of all concerned. The position over-all of granting or withholding that liberty may be regarded as the sum of how granting or withholding it would impinge upon each of the individuals concerned. So let us consider the case of one person, P, in isolation.

P's desires may be divided into two categories. Firstly, there is P's desire (or lack of it) to enjoy this liberty for its own sake, apart from the effect its possession may have on the satisfaction of P's other desires. Secondly, there is the incidence this liberty will have on the rest of P's desires (i.e. those other than the desire to enjoy this liberty as such). If it is likely that P's choice of a plan of life will be better (from the point of view of maximizing the satisfaction of P's desires) than someone else's choice on P's behalf, then there is a prima facie utilitarian case for P having this liberty. Thus a utilitarian defence of P's having this liberty will depend upon (a) how much P desires to frame her own plan of life, and (b) how well P can be expected to frame her own plan of life from the point of view of satisfying the rest of her desires.

Now clearly people differ in how much they desire this liberty and, if they have it, in how likely they are to frame a plan of life that will maximize the satisfaction of their desires. For some it is plausible that possession of this liberty will be for their utilitarian good, for they have a significant desire to enjoy it and can be expected to plan their lives better (from the point of view of satisfying desires) than anybody else could do it for them. Others, however, appear to have no strong desire to enjoy this liberty, nor much capacity to frame a plan suitable for maximizing desire-satisfaction. (They may prefer life in some clear institutional structure.) From a utilitarian point of view it would seem reasonable, therefore, to say that this liberty should be enjoyed by some but withheld from, or granted in only a qualified form to, others. But Mill believes that the liberty principle requires that this particular liberty should be enjoyed by all, at least in societies no longer in a backward state.

It may be objected to this argument that we think not only in terms of matching people's liberties to the desires they already have, but also of instituting a liberty in order to encourage a desire for it which does not as yet exist. If everyone is assigned the liberty to frame her own plan of life, then even if at present such an assignment is best only for some (perhaps a minority) it may encourage a desire for the liberty in many who do not as yet have it.

This objection fails, however, because the fact that a desire for such a liberty may be encouraged in this way can be of no significance for a desire-satisfaction utilitarian. To suppose that there is reason to encourage a desire that does not already exist is to assume that some objective value is placed on having that desire: a value not itself wholly desire-dependent.

To this it may be replied that there are cases where a utilitarian could advocate the encouragement of a desire in those who do not already have it, without assuming that some objective value is to be placed on having that desire. One example is the encouragement of a desire of a kind reasonably supposed to hold out particularly good possibilities of intense desire-satisfaction. This consideration does not, however, affect the present argument. While those deprived of liberty can desire it intensely, in general it would be implausible to encourage the desire for liberty as such on the ground that it held out particularly good possibilities of intense desire-satisfaction.

There is another case where a desire-satisfaction utilitarian can justify a policy not in keeping with the existing distribution of desires, but do this without assuming that an objective value is to be placed on the satisfaction of some desires. Suppose most people desire to be dependent upon an addictive and debilitating drug. In the short run a policy of discouraging dependence is not optimal in utilitarian terms, as it does not provide people with what they want. However, a utilitarian may argue that if the dependence persists future aggregates of satisfied desire will diminish, and hence that there is a utilitarian case for the policy.

This objection also fails. The example of drug dependence differs from encouraging a desire to frame one's own plan of life in that the former calls for no discrimination between the objective value of satisfying various desires: if most are drug dependent the satisfaction of desires of whatever character can be expected to reduce to zero, including desires related to drug dependence. No similar consideration applies in the case of fostering a desire for liberty to frame one's own plan of life: the reason for encouraging that desire is a non-desire-dependent preference for it over more passive inclinations.

Mill recognizes that the value of framing one's own plan of life is not wholly desire dependent:

He who lets the world … choose his plan of life for him, has no need of any other faculty than the ape-like one of imitation. He who chooses his plan for himself, employs all his faculties … And these qualities he requires and exercises exactly in proportion as the part of his conduct which he determines according to his own judgment and feelings is a large one. It is possible that he might be guided in some good path, and kept out of harm's way, without any of these things. But what will be his comparative worth as a human being? (p. 117).

Mill's utilitarianism is not, or not only, desire-satisfaction utilitarianism, and later it will be argued that the most interesting support Mill lends to his liberty principle does not derive from that form of utilitarianism. But it must be allowed that Mill sometimes places more emphasis on his Benthamite origins, as in this passage:

Such are the differences among human beings in their sources of pleasure, their susceptibilities of pain, and the operation on them of different physical and moral agencies, that unless there is a corresponding diversity in their modes of life, they neither obtain their fair share of happiness, nor grow up to the mental, moral, and aesthetic stature of which their nature is capable (p. 125).

Such a justification of the liberty to frame one's own plan of life is weak. At best it is only a prima facie case, and over-all the case against allowing persons the liberty to lead new or unusual ways of life may be stronger. The gain for a few in being able to better match their particular desires and dispositions to an appropriate way of life easily could be outweighed by the distaste felt by many at people living lives like that. Such gains also might be outweighed by the anxiety produced in many about the worth of their own more conventional ways of life.

An account of rights-based interests suitable for Mill's liberty principle is not to be found in desire-satisfaction utilitarianism. But it may be suggested that a more satisfactory account is to be found in a more complicated form of practice consequentialism. The consequences of the practices on which certain rights are based could be assessed by reference to certain supposed objective and intrinsic goods, such as individual autonomy, as well as by their tendency to maximize desire satisfaction. (To regard individual autonomy, say, as an objective good is to take account of it when estimating consequences in a way not wholly determined by its relationship to satisfying desire.) This amendment may repair the faults revealed in a justification based on desire satisfaction alone. For example, if individual autonomy is an objective and intrinsic good, it may be reasonable to assign a right to everyone to frame her own plan of life, not just those for whom it would be best in desire satisfaction terms.

But if this approach is taken some reason needs to be given why individual autonomy (say) should be regarded as an objective good—a reason not requiring Mill to seek support outside a consequentialist ethic. It will be argued that a suitable form of consequentialism (‘experimental consequentialism’) is to be found in Mill alongside the more familiar elements of desire-satisfaction utilitarianism. Experimental consequentialism does not give us precisely the conclusion that individual autonomy is an objective good, but it does provide reason for taking account of autonomy in a way not wholly dependent upon its effect on satisfying desires.

Experimental consequentialism may be introduced by way of noting an objection to desire-satisfaction utilitarianism. The claim that things other than desire satisfaction are of value only in so far as they are related to satisfied desire is plausible in some cases. The nicotine addict may allow that there is nothing good about his smoking other than that it satisfies one of his desires. If the desire to smoke were to cease he may then see nothing good about smoking. This is not, however, the way we view all objects of desire. With many we would not suppose that they ceased to be of value if our desires for them happened to cease. We separate the desirability of an object from its capacity to satisfy our desires. In some cases the occurrence of desire may be taken as a discovery of the value of the object of desire.

Experimental consequentialism is based on regarding the significance of some desires in this light. There well may be things other than the experience of satisfied desire which are intrinsically good. We come to believe that some things are intrinsically good by way of their capacity to arouse and to satisfy desire. The capacity to do this for some particular person is not, however, conclusive evidence that the object of desire is intrinsically good. Sometimes we revise our beliefs about the value of the object of desire, at first thinking that the object itself is of value, but later finding that its value lies only in its capacity to satisfy our desires. The extent to which people who have experienced the object of desire regard it as intrinsically good is an indication, though no conclusive indication, that it is so.

Experimental consequentialism differs from desire-satisfaction utilitarianism in its attitude towards certainty and evaluative knowledge. The latter assumes that we know what is intrinsically good. Such room as there is for new evaluative knowledge will be about the kind of thing which can give rise to satisfied desire. We may come to desire new kinds of thing, or come to recognize new sources of satisfaction of the desires we already have, but in all this we make no discovery about that which is intrinsically good. By contrast, experimental consequentialism, though an objectivist position, makes no claims to certain evaluative knowledge: the attitude towards whether this or that has intrinsic value is provisional.

Experimental consequentialism contrasts with desire-satisfaction utilitarianism in one further way. If desiring something is a possible indication of the intrinsic goodness of that which is desired, then although it is through your desires that you come to realize this, its goodness is not dependent on its happening to be your desires, in particular, that are aroused or satisfied. There are objects of possible desire desirable as such, though we may remain uncertain about what they are. For a desire-satisfaction utilitarian, however, all statements about the desirability of things (other than desire-satisfaction itself) are implicitly person-relative.

Mill appears to adopt experimental consequentialism in some places in his ethical writings. J. L. Mackie bases his account of the ‘proof’ of the principle of utility on a similar interpretation of Mill.6 There is further evidence that Mill took this view in the first two chapters of Utilitarianism and, especially, in the last chapter of A System of Logic. It is inappropriate to attempt to argue the case for attribution here. It is not claimed that this is Mill's only ethical position: frequently he appears to be defending desire-satisfaction utilitarianism.

The connection between experimental consequentialism and Mill's liberalism may be traced through these steps:

1. Possible objects of human desire may be divided into those whose existence is not dependent on human activity, and those created, partly or wholly, by human activity. Obviously most objects of human desire fall within the latter group, including what Mill calls ‘a pattern of human existence’ or a ‘mode of life’ (p. 125).

2. From experimental consequentialism we have the claim that the only way of recognizing that which is objectively good is through its becoming for us an object of desire. Though preferences may be expressed between ‘modes of life’ known to us only through abstract characterizations or hearsay, such preferences are of little significance for a determination of their desirability.

3. The best indication, though still a fallible one, that a ‘mode of existence’ is intrinsically good (or contains elements that are so) is that many who are well acquainted with it come to desire it.

4. The foregoing considerations indicate two political desiderata:

(a) People should be free to participate in the creation of as great a variety of ‘modes of existence’ as possible.

There is no reason that all human existence should be constructed on some one or some small number of patterns (p. 125).

(b) People should be free to experience as many ‘modes of existence’ as possible.

As it is useful that while mankind are imperfect there should be different opinions, so it is that there should be different experiments of living; that free scope should be given to varieties of character, short of injury to others, and that the worth of different modes of life should be proved practically, when anyone thinks fit to try them (p. 115).

5. Thus, if a state or a society takes a constricted view of desirable ‘modes of existence’, making it difficult for anyone to live in other than one of a few acceptable ways, members of that society will be prevented from experiencing new ‘modes of existence’; ones which may have contained something of worth. This could be a reason why Mill is, if anything, more concerned about freedom from social pressure than from illegality: a ‘mode of existence can survive illegality in a tolerant social environment, but it is difficult for it to last in a hostile one, even if it is not threatened by illegality.

Two further assumptions must be added to arrive at Mill's political position: his commitment to specifically individual liberty, and his belief in progress.

The reason for Mill's preoccupation with the liberty of individuals becomes clearer in the light of the ethical theory attributed to him. In the chapter ‘Of Individuality’, Mill several times affirms that initiatives for experiments in new ‘modes of existence’ come from individuals. For example:

… there are but few persons, in comparison with the whole of mankind, whose experiments, if adopted by others, would be likely to be any improvement on established practice. But these few are the salt of the earth; without them, human life would become a stagnant pool (p. 122; see also pp. 124 and 128).

Belief in progress is the remaining component in the relation between Mill's experimental consequentialism and his liberalism. For a desire-satisfaction utilitarian progress would be a process of reforming social and political institutions so as to make possible ever larger aggregates of desire satisfaction. His belief in progress would be a matter of supposing that scope for such improvement exists and is likely to be realized. But his commitment to maximizing desire satisfaction remains viable even with pessimism: his principle provides a basis for social choice even if it is believed that the best option available now is worse than the best option available in the past, and that future prospects will become worse still. By contrast, belief in progress is essential to Mill's position. He supposes that a process of individuals creating and choosing between new modes of existence under circumstances of freedom will have a direction. It will tend to throw up a wider variety of more significant modes of existence, and they will become available to a greater proportion of the population.

Belief that such progress will occur under circumstances of freedom is indispensable to Mill's position. If it were thought that choice under such circumstances would lead to the adoption of ever more insignificant modes of existence, this case for individual liberty would collapse: it then would be more reasonable to preserve such significant modes of life as already existed. A belief in progress given circumstances of freedom is essential to Mill's position, but a belief that there will be progress towards those circumstances is not, as is evidenced by his remarks on Chinese civilization (p. 129).

This defence of liberalism meets two conditions indicated earlier. It is consequentialist: persons ought to be free from control by the state and society because this is necessary for progress towards more significant modes of existence. Mill is not represented as regarding individual freedom as an end in itself. But although the defence is consequentialist, the importance of individual autonomy is not dependent on the consequences of protecting it in terms of desire satisfaction. Indeed the argument given for individual autonomy does not depend upon adherence to any particular conception of what is good, but on a view of the method by which we discover what is good.

There is a problem, however, about using this approach to defend Mill's liberty principle. What has been justified is a presumption favourable to experimentation with new modes of existence. The liberty principle is both more general and more specific than that. It is more general in that it allows liberty to persons and groups even if it is unlikely that they will participate in any experimentation. The liberty principle is also more specific. It says that people should be free from constraint in any activity not concerning the rights-based interests of others: it proposes specific freedoms and absences of freedom, thus indicating what experiments in modes of existence are permissible. Any particular system of rights-based interests defines and protects certain modes of existence and the values exemplified in them. The cost of this protection is that people may not pursue modes of existence incompatible with that system. For example, the conception of rights-based interests currently accepted in the West is incompatible with the pursuit of most private ways of life involving violence and its associated values. Mill recognizes that any given mode of existence requires restraint as well as freedom.

All that makes existence valuable to any one, depends on the enforcement of restraints upon the actions of other people. Some rules of conduct, therefore, must be imposed, by law in the first place, and by opinion on many things which are not fit subjects for the operation of law (p. 69).

The assignment of any specific set of rights-based interests presupposes a definite view of what modes of existence are more, and what are less, worthwhile. It is implausible that all of those modes of existence made difficult or impossible to pursue because of our current conception of rights-based interests are ones in which there is simply nothing of value to be discerned. The attempt to formulate a liberty principle, however, supposes that there is a fixed framework of rights-based interests and that people should be free to choose their own plans of life provided they respect that framework. This supposition is consistent with a Lockean natural rights view, and it perhaps also could be justified by a form of consequentialism which assumed that we know that certain ends, other than desire satisfaction, have value. But if our consequentialism takes an experimental attitude towards what is intrinsically good, the supposed fixed framework of rights-based interests dissolves. There is no definitive knowledge of values by reference to which a fixed conception of rights-based interests can be assigned.

If experimental consequentialism is incompatible with a determinate conception of rights-based interests, and hence with even the project of formulating a liberty principle of the kind Mill sought to defend, what can liberalism amount to? One option would be to take as a starting point the set of rights-based interests implicit in the law and morality of a particular society. To be a liberal would be to take a certain attitude to that received conception: an awareness not only of the modes of existence protected by the existing assignment of rights, but also of the cost of that protection: the difficulties the present assignment of rights puts in the way of experiments in some new modes of existence. This way of formulating liberalism gives it one feature in common with conservatism: both are attitudes towards an existing social and legal structure rather than a definitive body of principles.

This characterization of liberalism cannot be formulated in ‘one very simple principle’ allowing us to read off stances on practical issues. Any modification to the existing conception of rights-based interests probably will involve costs for some modes of existence as well as possible gains for others. Nevertheless, one concession to the wish for principles can be made, though it is principle quite distinct from, and less ambitious than, Mill's liberty principle. Sometimes the existing conception of rights-based interests is unnecessarily restrictive, in that a more permissive one would protect present modes of existence just as well, while allowing people to follow new ones if they wished. Perhaps a conception of rights-based interests permitting only heterosexual forms of sexual activity would be an instance. If it is possible to change current conceptions of rights-based interests so that the position for the present modes of existence is not significantly worsened, and experiments in new ones are made possible, such a change is warranted. That these changes may result in new modes of existence offensive to others (though they do not threaten their modes of existence) would be no reason against the change. But admittedly this is a limited outcome, and much more has to be done to explain the permanent significance of the traditional liberal freedoms from the approach of an experimental ethic.


  1. This is a revised version of a lecture given in the Royal Institute of Philosophy series, February 1981, and at the Political Thought Conference, New College, Oxford, January 1981, under the title ‘Mill's Kantian Liberalism’. My thanks are due to those present on both occasions, and to Jerry Cohen, John Gray, John Halliday, John Kelly and Anne Lloyd Thomas for their helpful and encouraging comments.

  2. J. S. Mill, On Liberty (Everyman Edition), Utilitarianism, Liberty and Representative Government, 74. All page references in the text are to this edition of On Liberty. For a similar reading of Mill's principle, see J. Rees, ‘A Re-reading of Mill on Liberty’, in Limits of Liberty, Peter Radcliff (ed.) (California: Wadsworth, 1966), 100.

  3. p. 73, emphasis added. Other passages favouring an interpretation in terms of rights-based interests are to be found on pp. 120, 121, 132, 136 and 137.

  4. On the other hand Mill says at p. 132 that the conduct each is bound to observe includes ‘bearing his share (to be fixed on some equitable principle) of the labours and sacrifices incurred for defending the society or its members from injury and molestation’. This suggests that there is a choice of acceptable ‘equitable principles’. So perhaps Mill's view is that some of the interests he refers to in his formulations of the liberty principle are constant for any of its applications, while others are partly a matter of which of a permissible set of options a particular society chooses to take up.

  5. The argument to follow would not be affected if we adopted the more plausible position that ‘satisfied desire’ covers a set of distinguishable psychological states. ‘Desire-satisfaction utilitarianism’ also could be adopted as a label for the different view that what is intrinsically good is the world coming to conform to how we desire it to be, whether or not this gives us ‘satisfaction’.

  6. J. L. Mackie, Ethics—Inventing Right and Wrong (Harmondsworth: Penguin Books, 1977), 142.

Mark Strasser (essay date 1991)

Download PDF PDF Page Citation Cite Share Link Share

SOURCE: “Mill's Moral View,” in The Moral Philosophy of John Stuart Mill: Toward Modifications of Contemporary Utilitarianism, Longwood Academic, 1991, pp. 23-53.

[In the following essay, Strasser evaluates Mill's moral stance and characterizes Mill primarily as an “act-utilitarian.”]

Both the quantity and the quality of pleasures must be considered in utility calculations. However, Mill's theory needs further explication, since we must discuss whose happiness should be promoted. For example, Mill might claim that an individual acts rightly if her action promotes her own happiness. Or, he might claim that an individual acts rightly if her action promotes the happiness of the society in which she lives. Mill chooses neither of these, instead suggesting that the promotion of humankind's utility is paramount.

To make matters more complicated, Mill writes ambiguously when he says that actions are morally right insofar as they promote happiness. He might mean that an agent acts rightly if she performs an action which will promote utility. Or, he might mean that an agent acts rightly if she performs an action which is in accord with a rule, the general acceptance and following of which will promote utility. Most of the time, this distinction is unimportant. An action which will promote utility will also be an action which is in accord with a rule, the general following of which will promote utility. Sometimes, however, the action which will promote utility is not in accord with a rule, the general following of which will promote utility. At these times, a choice must be made.


Whether one is an act-utilitarian or a rule-utilitarian, one believes that the principle of utility has a special status. It cannot be viewed as on a par with other moral rules because it is the principle upon which the whole moral system is based. The act-utilitarian believes that the principle of utility has a special status in that it is the principle by which all actions should be judged—if there is ever a conflict between the dictates of the principle of utility and any other principle or rule, the dictates of the principle of utility must be followed.

The rule-utilitarian believes that the principle of utility has a special status in that it should never be directly applied to actions. Instead, it should only be used to determine which rules people should accept and follow, i.e., those rules the general acceptance and following of which would promote the most good.1 Once these rules are ascertained, people should always act in accord with them.

Some writers insist that Mill is an act-utilitarian (someone who believes that the rightness or wrongness of an action depends solely upon the consequences of performing that action). Act-utilitarians claim, for example, that the “rightness or wrongness of keeping a promise on a particular occasion depends only on the consequences of keeping or of breaking the promise on that particular occasion.”2 If keeping the promise would maximize utility, then the promise should be kept. If breaking the promise would maximize utility, then the promise should be broken.

Others argue that Mill is a rule-utilitarian (someone who believes that a particular action is right if it is in accord with a moral rule, wrong if it is against a moral rule). A moral rule is a rule which, if generally followed, maximizes utility.

There seems to be evidence to support both sides. Yet, most of the statements which allegedly support Mill's rule-utilitarian position either give outright support to or are consistent with an act-utilitarian position. Ultimately, this debate will be seen to have been given undue importance in the recent literature. Mill would not have ascribed to either act-utilitarianism or rule-utilitarianism as they are currently understood. Act-utilitarians are too willing to break moral rules in their quest to maximize utility.3 Rule-utilitarians are too rigid either in not building in enough exceptions into their rules or in insisting on following those rules when doing so would clearly be disutility-producing. Mill does not want secondary rules broken readily and hence does not want to characterize them as merely ‘guides’ or ‘rules of thumb’. He also does not want secondary rules followed if doing so would clearly be disutility-producing, which is why he is more appropriately labeled an act-utilitarian than a rule-utilitarian.

There is an even more important reason why Mill is not accurately labeled as a standard act-utilitarian, much less a standard rule-utilitarian. While he believes that the act which maximizes utility is right, he does not believe that an action must be utility-maximizing in order to be right. That point will be explained … [elsewhere]. In this [essay], there will be an examination of Mill's view of the relations among acts, rules, and the principle of utility.


In A System of Logic, Mill denies that we can construct a perfect moral system. Basically, he offers two points:

1. Moral rules are extremely difficult to formulate because we cannot anticipate all of the circumstances which might obtain in future situations. If we cannot thus anticipate, then we will be unable to devise rules to cover all cases and our rules will be imperfect.

2. Even if we could formulate rules to include all of the appropriate exceptions, these rules would be so complicated that most people would not be able to learn and remember them.4

For example, generally, one should not steal. However, stealing is sometimes permissible; indeed, “to save a life, it may not only be allowable, but a duty, to steal, or take by force, the necessary food or medicine.”5 Doubtless, there are other extenuating circumstances in which stealing would also be permissible. Thus, because of the difficulties involved (a) in formulating perfect rules, and (b) in learning and remembering them, we must view moral rules as very good guides, as very useful rules of thumb. Moral rules are not principles which must always be obeyed, they are rules which should be obeyed most of the time. However, it is the act-utilitarian rather than the rule-utilitarian who uses moral rules as (even very useful) rules of thumb.6

One must “know what are the practical contingencies which require a modification of the rule, or which are altogether exceptions to it [my italics].” Mill argues that “rules of conduct point out the manner in which it will be least perilous to act, where time or means do not exist for analyzing the actual circumstances of the case, or where we cannot trust our judgment in estimating them.”7

When we are considering whether to follow a rule, we must consider that:

1. Rules are imperfect. There are times at which we should disobey rules in favor of following courses of action which will produce much better results.

2. Our present rules have stood the test of time—they have been proven to be effective through the ages. We must think twice (and, perhaps, twice again) before disregarding a rule.

Mill argues that rules should occasionally be disobeyed and thus he cannot be accurately labeled a rule-utilitarian. However, he gives some very helpful advice: rules can be wrong but usually are not.

When one, as an unbiased, knowledgeable, careful deliberator, is certain that following the appropriate rule would produce bad results, one should definitely disobey the rule. Mill criticizes those who would insist upon following a rule despite their knowing that a different action would yield much better results. The person “who goes by rules rather than by their reasons, like the old-fashioned German tacticians who were vanquished by Napoleon, or the physician who preferred that his patients should die by rule rather than recover contrary to it, is rightly judged to be a mere pedant, and the slave of his formulas.”8

Mill is not merely arguing that improper rules should not be used; he is suggesting that the world is such a complicated place that we will be unable to devise rules which will cover all of the situations which a doctor might face. So, too, we will be unable to devise rules to cover all of the situations which a moral agent will have to face.9 Our complicated world sometimes warrants ignoring the rules and judging the particular case according to the principle of utility. Rules of thumb are good for most cases most of the time. In those cases in which one knows that the application of the ‘appropriate’ rule would not promote utility, one should ignore the rule and promote utility instead.

Mill clearly favors the act-utilitarian view. In “Bentham,” he writes, “Insofar as Bentham's adoption of the principle of utility induced him to fix his attention upon the consequences of actions as the consideration determining their morality, he was indisputably in the right path.”10 In a review of Taylor's Statesman, Mill comments, “To admit the balance of consequences as a test of right and wrong, necessarily implies the possibility of exceptions to any derivative rule of morality which may be adduced from the test.”11

These passages explicitly show Mill's commitment to a type of act-utilitarianism. Further, when we consider that Mill is accepting the “creed which accepts as the foundation of morals, Utility, or the Greatest Happiness Principle [my italics],”12 we have reason to believe that Mill's principle involves the maximization rather than merely the promotion of utility, i.e., Mill believes that action morally best which maximizes rather than merely promotes utility. He argues that the utilitarian standard involves the “greatest [my italics] amount of happiness altogether,”13 and that “[a]ccording to the Greatest Happiness Principle, … the ultimate end … is an existence exempt as far as possible from pain and as rich as possible in enjoyments [my italics].”14 Here, too, Mill is offering a maximizing condition. However, Mill does not believe that only those actions which maximize utility are morally right.15


Some critics interpret Mill to be claiming that an act is right if it is in accord with a rule the general acceptance and following of which will maximize utility, i.e., they claim that Mill is a rule-utilitarian.16 They support their interpretation by citing Mill's claim that one can accept the principle of utility as the ultimate principle and yet still accept secondary rules as the rules by which one should live.17

Mill offers an often-cited analogy. “To inform a traveler respecting the place of his ultimate destination is not to forbid the use of landmarks and direction-posts on the way. The proposition that happiness is the end and aim of morality does not mean that no road ought to be laid down to that goal, or that persons going thither should not be advised to take one direction than another.”18 Just as a traveler can use landmarks to reach his final destination point, one can use secondary rules to help one maximize happiness.

Yet, Mill's example is somewhat misleading, since one's always using landmarks or secondary rules might be counter-productive.19 From a special vantage point, one might see a much better route than is indicated by the direction-post. So, too, for example, although lying is usually immoral and we are ‘directed’ to speak truthfully, we might realize that, because of certain extenuating circumstances, lying would be morally permissible. Our special vantage point, i.e., our knowledge of the extenuating circumstances, would allow us to ignore the rule and instead to maximize utility.

We should not ‘forbid the use of landmarks and direction-posts’. However, act- and rule-utilitarians agree that such aids should not be forbidden; they disagree about whether their use should be required. Quite simply, if the traveler knows of a much better route, the act-utilitarian will suggest that he take that route, while the rule-utilitarian will suggest that the traveler follow the direction-posts.

Individuals should not “test each individual action directly by the first principle [i.e., the principle of utility],” both because such a process would be too time-consuming and because such a process would sometimes result in the individuals' acting wrongly, e.g., because they miscalculated the utilities involved. Generally, secondary rules are the ones by which individuals should live.20 However, if there is a conflict between a secondary rule and the principle of utility, one must follow the dictates of the principle of utility.

In a different passage, Mill seems to explicitly support a rule-utilitarian position. “In the case of abstinences indeed—of things which people forbear to do from moral considerations, though the consequences in the particular case might be beneficial—it would be unworthy of an intelligent agent not to be consciously aware that the action is of a class which, if practiced generally, would be generally injurious, and that this is the ground of the obligation to abstain from it.”21

Mill seems to be saying that an action is wrong if the class of actions to which the action belongs should not be performed. A class of actions should not be performed if, were the actions generally performed, bad consequences would result. Mill's looking at whether an action is “of a class which, if practiced generally, would be generally injurious” as “the ground of the obligation to abstain from it,” is similar to one's looking at an action to see whether it is in accord with some rule and then looking at the rule to see whether its generally being followed would be harmful.22

This reasoning seems rather confused. He acknowledges that the ultimate goal is to promote utility and that the action in question would be beneficial. Then, he objects that the action is of a class which, if practiced generally, would be harmful. However, the agent is not wondering whether she, generally, should perform this action; she is wondering whether she should perform it this particular time. Thus, as soon as Mill raises his objection, the agent will rightly respond, “But I don’t want to do it all of the time. I want to do it just once.”

Mill might then explain that the agent's performing this action would increase the probability that other people would also perform it. Others might see her performing the action or they might see the results of the action. In either case, they, too, would want to perform the action in question. Mill would then point out that many people's performing the action would be harmful. Thus, for fear of promoting the performance of the action at non-beneficial times, the agent should refrain from performing the action this time. In order for anyone to be justified in performing the action this time, the benefits must outweigh the harms produced by the agent's promoting others to perform the action at non-beneficial times.

The above analysis renders Mill's reasoning more understandable. Mill is not saying that the action is forbidden merely because it belongs to a class of actions which, if generally practiced, would be harmful. Such a claim is not very convincing. If an action were acceptable only if its general performance would produce good consequences, then many noble and praiseworthy activities, e.g., dying for one's family, country, religion, etc., would not be acceptable. Indeed, the uniqueness of an action can be part of its utility. For example, in military maneuvers or card play, a ploy may be successful in part because it is so unexpected.

Mill is pointing out that actions have various subtle effects which individuals often overlook when trying to ascertain the effects of a particular action. For example, someone might not consider that:

1. Other people will see that he has broken the rule. Their respect for the rule will be lessened and they will become more likely to break the rule at non-optimal times.

2. Seeing that the rule has been broken, other people will feel less secure that the rule will be followed when it should be. Even if these people realize that the rule was broken to promote greater utility and thus they will not be tempted to break the rule when such an infraction would cause disutility, they might not be so confident in their neighbors' perspicacity. Even though they understand why this particular rule-breaking was proper and thus would not be more likely to break the rules improperly, they could not be sure that others would so understand. Insecurity would result—an obvious disutility.23

3. Seeing that this rule has been broken, people will be more likely to break other rules at inappropriate times.

Mill is not as clear about the act/rule distinction as one might like. He argues that there is a “necessity that some rule, of a nature simple enough to be easily understood and remembered, should not only be laid down for guidance, but necessarily observed, in order that the various persons concerned may know what they have to expect.” Mill believes that “the inconvenience of uncertainty [which results when rules are viewed as merely providing guidance] … [is] a greater evil than that which may possibly arise, in a minority of cases, from the imperfect adaptation of the rule to those cases.”24

Mill seems to be offering a rule-utilitarian account in that the rules should be “necessarily observed.” However, the reason that rules should necessarily be observed is that their non-observance will cause insecurity. If Mill is afraid that a particular act of rule-breaking will cause net disutility, then he should not claim that rules must be observed, but that rules should be broken only when the various disutilities (including the increased insecurity) would be outweighed by the utility of the rule-breaking. Indeed, in the next paragraph he amends the claim that rules should necessarily be observed by arguing that “the license of deviating from [the rules], if such be ever permitted, should be confined to definite classes of cases, and of a very peculiar and extreme nature.”25

In a different work, Mill argues that the “admission of exceptions to rules is a necessity equally felt in all ethical systems of morality. To take an obvious instance, the rule against homicide [and] the rule against deceiving … are suspended against enemies in the field, and partially against malefactors in private life.” The particular circumstances may require one to do something which is normally forbidden. “That the moralities arising from the special circumstances of the action may be so important as to overrule those arising from the class of acts to which it belongs, perhaps to take it out of the category of virtues into that of crimes, or vice versa, is a liability common to all ethical systems.”26 Here, Mill seems quite clearly to be advocating an act-utilitarian line. However, on the next page, he claims, “The essential is, that the exception should be itself a general rule; so that, being of definite extent, and not leaving the expediencies to the partial judgment of the agent in the individual case, it may not shake the stability of the wider rule in the cases to which the reason of the exception does not extend.”27

Again, Mill is afraid that the stability of the rule will be undermined and that people will disobey the rule at inappropriate times. “No one who does not thoroughly know the modes of action which common experience has sanctioned is capable of judging of the circumstances which require a departure from those ordinary modes of action.”28 Agents must not haphazardly ignore rules, especially considering that their judgment might be biassed.29 Were secondary moral rules simply rules of thumb, people adopting this theory would break rules too often.

Nonetheless, Mill ultimately holds that utility must be promoted. He believes that “where time or means do … exist for analyzing the actual circumstances of the cases, [and] where we [can] trust our judgment in estimating them,”30 we must ignore the rule and promote utility. We should never perform an action which we (reliably) know to be disutility-producing. A man who does so “cannot discharge himself from moral responsibility by pleading that he had the general rule in his favor.”31

Mill explains his position in a letter. “I agree with you that the right way of testing actions by their consequences, is to test them by the natural consequences of the particular action, and not by those which would follow if everyone did the same. But, for the most part, the consideration of what would happen if every one did the same, is the only means we have of discovering the tendency of the act in the particular case.”32 In order to look at the effects of a particular action, “[w]e must look at [the action] multiplied, and in large masses.”33 Thus, the insecurity which Mill fears would be produced by rule-breakings must be attributed to particular acts of rule-breaking. The destabilizing of rules, i.e., the increasing of the likelihood that rules will be broken at improper times, is an effect of particular acts of rule-breaking. These effects must be considered whenever one proposes to break a secondary rule.

When Mill considers consequences, he includes the often under-appreciated considerations of character. He criticizes Bentham for not having considered whether “the act or habit in question, though not in itself necessarily pernicious, may not form part of a character essentially pernicious, or at least essentially deficient in some quality eminently conducive to the ‘greatest happiness’,”34 and Sedgwick for not having recognized that utilitarians also believe that “an essential part of the morality or immorality of an action or a rule of action consists in its influence upon the agent's own mind: upon his susceptibilities of pleasure or pain, upon the general direction of his thoughts, feelings, and imagination, or upon some particular association.”35

Secondary rules are important because their generally being accepted and followed will tend to promote good character. They will also promote security, which is quite important. A “change, which has always hitherto characterized, and will assuredly continue to characterize, the progress of civilized society, is a continual increase of the security of person and property.”36 Mill writes, “Security of person and property and equal justice between individuals are the first needs of society and the primary ends of government.”37

Another important feature of secondary rules is that they give specific dictates about how to act, which is necessary because “utility, or happiness, [is] much too complex and indefinite an end to be sought except through the medium of various secondary ends.”38 Individuals “who adopt utility as a standard seldom apply it truly except through the secondary principles.”39

Yet, one would be wrong to think that it is obvious what the secondary rules should be, even given a particular end (e.g., the promotion of general utility). “The grand consideration is, not what any person regards as the ultimate end of human conduct, but through what intermediate ends he holds his ultimate end is attainable, and should be pursued; and in these there is a nearer agreement between some who differ, than between some who agree in their conception of the ultimate end.”40

Once the secondary principles have been formulated, Mill claims that one can pretty readily decide which principle is appropriate in any particular case. “There is no case of moral obligation in which some secondary principle is not involved; and if only one, there can seldom be any real doubt which one it is, in the mind of any person by whom the principle itself is recognized.” Indeed, he somewhat surprisingly further argues that “only in these cases of conflict between secondary principles is it requisite that first principles should be appealed to.”41

Yet, one's having the ability to pick out the appropriate secondary rule does not imply that one has the ability to apply it correctly. The “common error of men [is] of sticking to their rules in a case whose specialties either take it out of the class to which the rules are applicable, or require a special adaptation of them.”42 Thus, it “is one thing to be master of general principles, and to be able to reason from them under assumed hypothetical circumstances: it is another thing to possess the talent of justly appreciating actual circumstances, so as to regulate the application of principles to any given case.”43

Mill is a little misleading when he claims that the principle of utility should be appealed to only in cases involving conflicts between secondary rules, since such conflicts are easy to generate. We consider the case of lying. One secondary rule is ‘Never lie’. Another secondary rule is ‘Never lie unless doing so would promote utility and telling the truth would promote disutility’. In a case in which lying would promote utility and telling the truth would promote disutility, we have a conflict between rules and must appeal to the principle of utility.

“If evil will arise in any specific case from our telling truth, we are forbidden by a law of morality from doing that evil: we are forbidden by another law of morality from telling falsehood. Here then are two laws of morality in conflict, and we cannot satisfy both of them.” When there is a conflict between two rules, we must “resort to the primary test of all right and wrong, and … make a specific calculation of the good or evil consequences, as fully and impartially as we can.”44

Basically, Mill is suggesting that whenever one would produce net disutility by following one secondary rule, one's act would not be in accord with a different secondary rule, viz., ‘Do not produce evil’ (‘Do not produce net disutility’). Whenever there are conflicts between secondary rules, one must appeal to the principle of utility. Thus, rule-utilitarians are not representing Mill's position when they argue that one may morally permissibly produce net disutility as long as one's act is in accord with a rule the general following of which would maximize utility. Mill would argue that in such a case one's act would not be in accord with the secondary rule ‘Do not produce net disutility’. He would then claim that one would have to appeal to the principle of utility to settle the conflict, and that the principle of utility would not allow the individual to perform a (net) disutility-producing act.

In general, a few principles emerge from Mill's writings with respect to the issue(s) at hand:

1. We should use secondary principles in our moral decision-making. Further, we should not be willing to break these secondary principles as readily as we would, were they merely guides to action.

2. These secondary principles should be simple enough so that they can be readily remembered and applied.

3. If, after including all of the relevant disutilities, we, as unbiased, knowledgeable, careful deliberators, are certain that following the rule would promote disutility and ignoring the rule would promote utility, then we must disobey the rule.45


Some critics believe that maximizing act-utilitarianism is untenable,46 since agents who seek to maximize utility may be unable to do so because they lack information about what other people will do. This can easily be illustrated.

Suppose that Whiff and Poof each have to decide whether to push the buttons in front of them. The utilities can be summed up in the following diagram.47

Push Not-push
Push 10 0
Not-push 0 6

The optimal situation would involve both Whiff's and Poof's pushing the buttons. However, if one of them is not going to push a button, then neither of them should.

Suppose that neither knows what the other will do. What should Whiff do? If he pushes and Poof not-pushes, then no utiles will be produced.

As long as Whiff knows that Poof knows the possible outcomes and Poof knows that Whiff knows the possible outcomes, they will both push, since they are both seeking to maximize utility and their both pushing would produce optimal results. However, Whiff and Poof would not be helped by this solution if their situation were a little different. Suppose that as long as they both pushed or not-pushed, they each would produce ten utiles; otherwise, they each would produce minus five utiles. If neither has any idea what the other will do, neither can say what he himself ought to do.

Obviously, Whiff and Poof should talk to each other, if possible. If they cannot, then they are in a difficult position and each will simply have to make a guess about what the other will do.48

In discussing these issues, we must be careful to distinguish among several issues:

1. Did the agent act rightly?

2. Should the agent be blamed?

3. How much blame should he receive?

In the problem posed above, the agent does not know what to do. Should Whiff push the button? Yes, if Poof is going to push, and no, otherwise. If Whiff does not know what Poof is going to do, Whiff will have to guess, taking into account all that he knows about Proof. If Whiff guesses correctly, he will have acted rightly; otherwise, he will have acted wrongly.49 Co-ordination problems do not prevent agents from acting rightly. They merely prevent agents from knowing what to do.

Should we blame Whiff for guessing wrongly? This is a separate issue. Whiff's lack of knowledge may affect whether or how much he should be blamed for his bad guess. It does not at all affect whether he in fact performed the right action.

Current theorists are correct that co-ordination problems may affect act-utilitarian analyses of the proper courses of action. Mill would welcome the suggestion that agents should try to be well-informed before making their decisions. However, the co-ordination utilitarians are in no better position than the act-utilitarians. Both should find out all they can about the situation at hand, including what other people are planning on doing, and then act accordingly.


When Mill talks about the effects of an action, he includes many of the consequences which one might not (at first) attribute to the performance of a particular action. For example, Mill believes that one should be truthful because “the cultivation in ourselves of a sensitive feeling on the subject of veracity is one of the most useful, and the enfeeblement of that feeling one of the most hurtful, things to which our conduct can be instrumental.” Thus, one's lying has a negative effect on one's character. Also, one's lying will help to “deprive mankind of the good, and inflict upon them the evil, involved in the greater or less reliance which they can place in each other's word.”50 People will feel less secure if I lie to them, both because they will trust me less and because they will tend to trust others less as well.

When Mill is trying to decide whether a particular lie would be justified, he does not merely look at the immediate effects of the lie, e.g., my being very pleased versus someone else's being slightly displeased. My lying will hurt my character, will make the other person less likely to believe people, and will make me less likely to believe people.

Mill argues that the consequences which rule-utilitarians claim should be traced back to rules can in fact be traced back to actions. By taking a rather broad view of which effects are correctly deemed the effects of particular actions, Mill can account for the considerations which a rule-utilitarian might cite, e.g., effects on character, security, etc., and Mill is able to secure an additional benefit. He can claim that lying is sometimes morally permissible,51 for example, at those times when calamities might be averted by a little ‘judicious’ lying.


As a type of act-utilitarian, Mill believes that an action which maximizes utility is morally right. However, his view is somewhat unusual with respect to whose utility should be promoted.

Mill is quite concerned about promoting the happiness of the individual. He believes that one of the reasons that individuals should pursue the higher pleasures is that, by so doing, they will be both better and happier people.52 Mill is also quite concerned about promoting the happiness of society, arguing that even if an individual, himself, is not happier as a result of his becoming nobler, other people will thereby be happier.53 Yet, neither the promotion of individual happiness nor the promotion of societal happiness is Mill's ultimate concern. Mill's ultimate concern is to promote the utility of humankind.

In On Liberty, Mill says, “I regard utility as the ultimate appeal on all ethical questions, but it must be utility in the largest sense, grounded on the permanent interests of man as a progressive being.”54 Mill's talking about “man as a progressive being” does not make each individual's interests of primary importance. Rather, the permanent interests of humankind are important. Each individual's utility is promoted as a fortunate by-product of the promotion of humankind's utility.

When talking about ‘man’, Mill might seem to be referring to the interests of society. Such an interpretation is incorrect.

The permanent interests of man are not simply the permanent interests of a particular society, e.g., Greek, Roman, or British. Rather, they are the interests of the Society of Man, i.e., Humankind. When Mill talks about Humankind, he is not merely talking about all of the people who happen to be living at a particular time. He is talking about “the Human Race, conceived as a continuous whole, including the past, the present and the future.”55

Mill agrees with Comte that “reflection, guided by history, has taught us the intimacy of the connection of every age of humanity with every other.”56 We are narrow-minded if we think that we should promote the good of a particular society of a particular time period at the expense of the good of humankind. It is “the good of the human race [which] is the ultimate standard of right and wrong.”57

Mill complains that Bentham's doctrine, which emphasizes individuals' acting out of self-interest, is very destructive of “all natural hope for good for the human species [my italics].” Bentham undermines “our hopes of happiness or moral perfection to the species [my italics].”58

In A System of Logic, Mill says that the ultimate standard is “conduciveness to the happiness of mankind.”59 In Utilitarianism, Mill also speaks about promoting the happiness of all humankind.60

Mill believes that we are able to predict what will promote the happiness of humankind, because “social phenomena conform to invariable laws.”61 We merely need to figure out what would be the ideal setting, e.g., which attitudes people should have and which actions people should perform, for happiness to be promoted. We can examine the history of humankind to see which rules and actions further that goal.62

When Mill talks about promoting the happiness of humankind, he does not specify whether he is talking about the average happiness or the (net) total happiness. The difference may be illustrated in the following way.

Suppose that there are four individuals in World W, each of whom is fairly happy. For the sake of illustration, we shall assign each of them a “happiness level” of 25. The question at hand is whether the world would be a better place were there an additional person, Jones, who over her lifetime would enjoy a happiness level of 10.

Were Jones to live in W, the (net) total level of happiness would be raised to 110 happiness units. However, the average happiness level would be lowered from 25 to 22. Thus, on the (net) total happiness account, W would be a better place were Jones also to live there. On the average happiness account, W would be a worse place were Jones to live there.

Family planning practices might differ, depending upon whether one adopted the former or latter account. Mill was clearly interested in promoting a decline in the population rate. “It appears to me impossible but that the increase of intelligence, of education, and of the love of independence among the working classes, must be attended with a corresponding growth of the good sense which manifests itself in provident habits of conduct, and that population, therefore, will bear a gradually diminishing ratio to capital and employment.”63 Here, he might be inferred to be promoting the average happiness. However, Mill would not be in favor of destroying all but two people (who happened to be very happy and to require no one else but themselves), even though the average happiness might thereby be increased.

Clearly, Mill wants future generations to be considered in calculations of utility. However, he does not specifically address some of the issues raised if one is indeed to consider future generations, e.g., whether one has a duty to procreate, whether future generations' interests should be discounted, if so, by how much, etc.64

An additional complicating factor which must be considered is that Mill does not believe that only rules and actions should be judged by the principle of utility. Attitudes should also promote utility. Mill criticizes Whewell for assuming that people's moral attitudes are above reproach. “The point in dispute is, what acts are the proper objects of those [moral] feelings; whether we ought to take the feelings as we find them, as accident or design has made them, or whether the tendency of actions to promote happiness affords a test to which the feelings of morality should conform.”65 Indeed, “intellectual progress [is] in no other way so beneficial as by creating a standard to guide the moral sentiments of mankind, and a mode of bringing those sentiments effectively to bear on conduct.”66 Thus, when Mill talks about promoting utility, he is talking about the utility of humankind. He is interested in changing both attitudes and actions in order to bring about his goal.

Merely because Mill is primarily interested in promoting the interests of humankind does not mean that Mill is uninterested in protecting the individual. The protection of individual liberties is extremely important, because protecting them is the best way to promote the interests of humankind. On Liberty is written to demonstrate on empirical grounds that the protection of individual liberties will ultimately be the best way of promoting the interests of humankind.67


It might seem unimportant to ascertain whether Mill was an act-utilitarian, since even if this controversy could be resolved the result would seem to have little bearing on any contemporary philosophical discussions. Such a view is mistaken. In understanding Mill's act-utilitarianism, we shall see how contemporary act-utilitarians can refute some of their severest critics.

Act-utilitarians are often charged with being unable to show that justice will be preserved or that promises will be kept. Mill deflects these charges by talking about the hidden disutilities caused by the performance of such actions. If justice is not preserved, insecurity will result.68 Insecurity will also result if people habitually lie. Mill is including these results as part of the total consequences of particular actions.

Mill is an act-utilitarian who looks at immediate and eventual consequences of actions in light of how these actions will affect the utility of humankind. He thus disagrees with act-utilitarians who only look at the immediate consequences of actions or who only look at consequences in light of how they will affect a particular society or even the world at a particular time.

When theorists talk about the remote effects of actions, e.g., “possible effects on the agent's character, and effects on the public at large,”69 they are talking about effects which, according to Mill, are effects of particular actions. While even Mill admits that it is difficult to determine exactly how much disutility is promoted by a particular lie, he still believes that the disutility is produced by the action. We may have to look at many actions of a similar type in order to ascertain all of the effects, but it is the effects of the action that concern Mill.

Mill would be confused by a critic who claimed that effects on character or on public feelings of security are reasons not to support act-utilitarianism. For Mill, these would be reasons to support his type of act-utilitarianism.

Since Mill's act-utilitarianism takes into account many of the factors that are taken into account by various forms of rule-utilitarianism, it might seem unimportant to make the distinction. Such a claim is false because of the way that rule-utilitarianism tends to be formulated.

Suppose that even after one has taken all of the hidden disutilities of a particular action into account, one still sees that one should perform what is normally called an unjust action. The act-utilitarian will claim that the unjust action should be performed. She will explain her position in one of two ways:

1. She will claim that the performance of justice is not always morally correct, opting for a rather paradoxical view of both justice and morality.

2. She will adopt Mill's view and claim that because the dictates of justice cannot be disutility-producing, what is normally just (i.e., what the secondary rules of justice dictate to be just) is not just in this particular case.

The rule-utilitarian would say that even if the dictates of justice do not promote utility in this particular case, the dictates generally promote utility and thus must always be followed. There is a major difference between the act- and rule-utilitarian. In a situation in which the dictates of justice do not promote utility, the act-utilitarian claims that the dictates must be ignored, while the rule-utilitarian claims that the dictates of justice must be followed.

Mill argues, “In such cases [i.e., where justice and utility conflict], as we do not call anything justice which is not a virtue [e.g., which is disutility-producing], we usually say, not that justice must give way to some other moral principle, but that what is just in ordinary cases is, by reason of that other principle, not just in this particular case.” By doing this, we will avoid conceptual difficulties. “By this useful accommodation of language, the character of indefeasibility attributed to justice is kept up, and we are saved from the necessity of maintaining that there can be laudable injustice.”70

Mill is not merely making a semantic point here. He is claiming that since justice is founded on utility, justice cannot dictate that we perform disutility-producing actions. In cases in which the apparent rule of justice dictates that we do something which is obviously disutility-producing, we must either amend the rule or ignore it altogether.

Mill has little patience for those who would claim that not following the normal dictates of justice to promote utility would involve doing an ‘evil’. He agrees with Whately's analysis of such a case. “[W]hen in a discussion, one party vindicates, on the ground of general expediency, a particular instance of resistance to government in a case of intolerable oppression, the opponent may gravely maintain that ‘we ought not to do evil that good may come;’ a proposition which of course has never been denied, the point in dispute being, ‘whether resistance in this particular case were doing evil or not’.”71

If truth-telling would be disutility-producing and lying would be utility-producing in a particular case, then lying would be morally required. Lying and acting ‘unjustly’, i.e., not following the normal dictates of justice, may not have (net) negative effects on one's character, on public security, or on others' lying or breaking other moral rules at inappropriate times. Lying has a bad effect on one's character only when the lie would promote one's acting in disutility-producing ways. By the same token, acting unjustly would have a bad effect on one's character only when the ‘injustice’ would promote one's acting in disutility-producing ways. If I lie or act unjustly because doing so would slightly benefit me and would greatly hurt others, then my so acting would be wrong—I have counted my own interests too heavily. My character will suffer in that I am promoting my own tendency to act selfishly and in that I am promoting my own and others' tendency to lie or act unjustly at inappropriate times. However, if I lie to promote utility, then my character may not suffer. I should lie when doing so would promote utility, and my character is helped rather than harmed insofar as my tendency to lie at utility-producing times is promoted. True. If I lie to promote utility, then my character might suffer in that the tendency to lie in general rather than at only utility-promoting times might be strengthened. Further, my character might suffer in that the disposition to break other moral rules might also be strengthened. Nonetheless, the disutility produced might not be sufficient to outweigh the utility of telling the lie.72

Suppose that Ann and I agree to meet somewhere. I go there but she fails to keep the appointment. She has broken her promise. Much disutility can result from promise-breaking, e.g., I may grow to distrust her and others.

Suppose, however, that she has very good reasons for her promise-breaking. Once I have been apprised of the facts, I might have two very different reactions:

1. I might grow to trust her as much as or more than I did before. I would realize that she has exercised good judgment. Her action would promote feelings of security, both in myself because I know that she has acted as she should have, and in others, especially those who directly benefited from her good use of her judgment.

2. I might still be angry and resentful, despite my knowing that her action promoted utility. Mill would condemn my resentment rather than Ann's action. He would say that “just persons [resent] a hurt to society, though not otherwise a hurt to themselves, and [do] not [resent] a hurt to themselves, however painful, unless it be of the kind which society has a common interest with them in the repression of.”73

Mill argues persuasively for the protection of liberty and the preservation of justice on utilitarian grounds. We learn from him, not (as some rule-utilitarians and deontologists seem to imply) that the preservation of justice is usually but not always utility-producing, but that the preservation of justice is necessarily utility-producing. When one, as an unbiased, knowledgeable, deliberate calculator, takes into consideration the effects of following the standard rules of justice and morality and one sees that following those rules will be disutility-producing, one will understand that those rules must be amended or ignored entirely.

It is important to understand Mill's act-utilitarianism because we can then see that some of the arguments (e.g., that insecurity and the formation of bad characters will result from performing certain actions) used by rule-utilitarians and deontologists to support their own theories, are used cogently and convincingly to support an act-utilitarian position. However, there is an important respect in which act-utilitarianism and rule-utilitarianism are not so far apart after all.

We should not conform our acts to socially useful rules when doing so would promote disutility. However, when making the appropriate calculations, we must include the remote effects of rule-breakings, e.g., effects on character, effects on general security, etc.

If we really know of a case in which following the rule would promote disutility and breaking the rule would promote utility (including all of the considerations above), then the rule-utilitarian will want to modify his rule to account for the exception, since the rule-utilitarian cannot justify a rule when the adoption of a different rule in its stead would produce more utility.74 If we know that following a rule would be disutility-producing, then an act-utilitarian will claim that we should not follow that rule, i.e., he will say that we should disobey it. By the same token, if we know that our following a rule would be disutility-producing, then the rule-utilitarian will claim that we should not have that rule—she will say that we should instead have a different rule to follow.

Usually, agents suspect rather than know that following a particular rule would be disutility-producing. In such situations, the rule-utilitarian and the act-utilitarian face similar problems. The act-utilitarian must decide whether he should disobey the rule of thumb which is usually effective to follow. The rule-utilitarian must decide whether this rule is in need of an (or another) exception.

The claim here is not that act-utilitarianism and rule-utilitarianism are co-extensive. A rule-utilitarian might argue that we have to keep moral rules fairly simple if people are going to be able to accept and follow them correctly. Mill and act-utilitarians in general whole-heartedly agree with such a claim. The rule-utilitarian might further argue that these rules must be obeyed, even if on a particular occasion the rule's being obeyed would produce net disutility. At this point, Mill and act-utilitarians in general would disagree. They would argue that one must promote utility, even if one must break a secondary rule to do so.

It is important to understand Mill's act-utilitarianism because it is important to understand that he really is committed to the promotion of the utility of humankind above everything else. However, precisely because he is committed to the promotion of utility, he is justifiably worried that his calling secondary moral principles mere guides or mere rules of thumb would result in too many disutility-producing rule-breakings. As to whether Mill is better interpreted as claiming that the promotion of humankind's utility can best be achieved by ignoring disutility-producing rules or by simply amending them, this is a question which is, perhaps, neither answerable nor important to answer.


  1. Another form of indirect utilitarianism would involve motivations rather than rules. Brandt writes, “An optimific indirect theory is a normative theory which roughly holds that any other-person-involving act is morally permissible if it would be best for the moral motivations of (roughly) all agents to permit acts of that type in its circumstances, and that an other-person-involving act is an agent's moral duty if it would be best for the moral motivations of (roughly) all agents to require acts of that type in those circumstances.” He explains that he means “by a person's moral motivations a complex consisting of a desire/aversion for some kind of action for itself, in certain circumstances, a disposition to feel guilty if one does (does not) perform an act of this sort, and a disposition to disapprove or be indignant if someone else does (does not) perform an act of that type, in those circumstances.” R. B. Brandt, “Fairness to Indirect Optimific Theories in Ethics,” Ethics 98 (1988), pp. 342-343.

  2. J. J. C. Smart, “Extreme and restricted utilitarianism” in Mill: Utilitarianism with Critical Essays, ed. Samuel Gorovitz (Indianapolis: Bobbs-Merrill Co., Inc, 1971), p. 195.

  3. Alexander believes that act-utilitarianism's mandating that one break rules if doing so will be utility-maximizing may produce so many unwarranted rule-breakings (because agents would sometimes miscalculate and would thus break rules at inappropriate times) that one's adopting act-utilitarianism might produce less utility than would one's adopting rule-utilitarianism. See Larry Alexander, “Pursuing the Good—Indirectly,” Ethics 95 (1985), p. 324. For an illustration of how these miscalculations might occur, see F. H. Bradley, “Pleasure for Pleasure's Sake” in Ethical Studies (first published 1876), Introduction by Ralph Ross (Indianapolis: Bobbs-Merrill Educational Publishing, 1951), p. 49.

  4. John Stuart Mill, A System of Logic, Book 6, Ch. 12, Sec. 3, CW Vol. 8, p. 945.

  5. Utilitarianism, Ch. 5, Par. 37, p. 259.

  6. See Smart, “Extreme and restricted utilitarianism,” p. 199.

  7. A System of Logic, Book 6, Ch. 12, Sec. 3, pp. 945-946. For a similar point, see John Stuart Mill, The Subjection of Women, Ch. 3, Par. 10, CW 21, p. 307.

  8. A System of Logic, Book 6, Ch. 12, Sec. 2, p. 944.

  9. A number of theorists talk about the importance of rules' being fairly simple. See R. M. Hare, Moral Thinking (Oxford: University Press, 1981), p. 33; and John Rawls, “The Independence of Moral Theory,” American Philosophical Association Proceedings and Addresses 48 (1974-1975), p. 14.

  10. John Stuart Mill, “Bentham,” CW Vol. 10, p. 111.

  11. John Stuart Mill, “Taylor's Statesman,” CW Vol. 19, p. 638.

  12. Utilitarianism, Ch. 2, Par. 2, p. 210.

  13. Ibid., Ch. 2, Par. 9, p. 213.

  14. Ibid., Ch. 2, Par. 10, p. 214.

  15. Rem Edwards claims that Mill was “neither a maximizing act nor rule utilitarian.” See his “The Principle of Utility and Mill's Minimizing Utilitarianism,” Journal of Value Inquiry 20 (1986), p. 132. See also his “J. S. Mill and Robert Veatch's Critique of Utilitarianism,” Southern Journal of Philosophy 23 (1985), pp. 181-200. Here, it is argued that Mill agrees with standard act-utilitarians that those actions which maximize utility are right. However, he disagrees insofar as they claim that an action must be utility-maximizing in order to be right. For a related point, see Marcus Singer's “Actual Consequence Utilitarianism,” Mind 86 (1977), p. 69.

  16. See J. O. Urmson, “The Interpretation of the Moral Philosophy of J. S. Mill,” Mill: Utilitarianism with Critical Essays, p. 170; and H. J. McCloskey, John Stuart Mill: A Critical Study (London: Macmillan and Co., Ltd., 1971), p. 58.

  17. Utilitarianism, Ch. 2, Par. 24, p. 224.

  18. Ibid., pp. 224-225.

  19. This is Mabbott's example. See J. D. Mabbott, “Interpretations of Mill's Utilitarianism” in Mill: A Collection of Critical Essays, ed. J. B. Schneewind (Garden City: Anchor Books, 1968), p. 194.

  20. Utilitarianism, Ch. 2, Par. 24, p. 224.

  21. Ibid., Ch. 2, Par. 19, p. 220.

  22. A class of actions is merely a set of actions, all of which meet one or more criteria. Or, to use a different description, a class of actions is a set of actions, all of which correspond to some rule(s). To be concerned with the consequences of performing actions of a certain class is to be concerned with the consequences of performing actions of a particular set, all of which correspond to the same rule(s). Thus, Mill seems to be espousing a rule-utilitarian position—one should not perform an action unless the rule with which it is in accord would promote the general good. The consequences which result from not following the rule seem to be of concern, not the consequences of performing the action.

  23. For a similar argument, see Bernard Semmel, John Stuart Mill and the Pursuit of Virtue (New Haven: Yale University Press, 1984), p. 87. For related comments of Mill’s, see “Whewell on Moral Philosophy,” p. 182; and On Liberty, Ch. 4, Par. 8, CW Vol. 18, p. 280.

  24. A System of Logic, CW Vol. 8, Book 6, Ch. 11, Sec. 6, Par. 3, pp. 1154-55.

  25. Ibid., Par. 4, p. 1155. These two passages were deleted after the 1846 edition. Mill may have deleted these passages because he simply changed his mind and wanted to espouse a much more act-utilitarian line. However, passages in other works would not support this explanation. I suspect that he deleted these passages, not because his position had changed but rather because they sounded too rule-oriented and thus did not represent his position. One can easily understand Mill's difficulty. He does not want to advocate blind rule-following, but also does not want to advocate a position which would result in frequent and inappropriate violations of secondary rules. Of course, it may be that the passages were deleted because of printing costs. However, even were that true, other passages in Mill's writings support the position offered here.

  26. “Whewell on Moral Philosophy,” p. 182.

  27. Ibid., p. 183.

  28. Considerations on Representative Government, Ch. 5, Par. 8, p. 425.

  29. “Whewell on Moral Philosophy,” p. 183.

  30. A System of Logic, Book 6, Ch. 12, Sec. 3, p. 946.

  31. “Taylor's Statesman,” p. 640. In Chapters 4-7, there will be a discussion of how much net utility one has a duty to promote.

  32. CW Vol. 17, p. 1881. Brown points out this letter. See D. G. Brown, “Mill's Act-utilitarianism,” Philosophical Quarterly 27 (1974), p. 68.

  33. “Whewell on Moral Philosophy,” p. 181.

  34. “Remarks on Bentham's Philosophy,” CW Vol. 10, p. 8.

  35. “Sedgwick's Discourse,” p. 56.

  36. Principles of Political Economy, Book 4, Ch. 1, Sec. 2, Par. 2, Vol. 3, p. 706.

  37. Considerations on Representative Government, Ch. 15, Par. 11, p. 541. See also “De Toqueville” (1), CW Vol. 18, p. 80.

  38. “Bentham,” p. 110.

  39. Ibid., p. 111.

  40. “Blakey's History of Moral Science,” CW Vol. 10, p. 29.

  41. Utilitarianism, Ch. 2, Par. 25, p. 226.

  42. The Subjection of Women, Ch. 3, Par. 10, p. 307. Mill seems to believe that the moral judgment of individuals improves as they become more ‘civilized’. See “Civilization,” CW Vol. 18, p. 132.

  43. “Taylor's Stateman,” pp. 622-623.

  44. Ibid., pp. 638-639. See also Utilitarianism, Ch. 2, Par. 23, p. 223 in which Mill writes that even the rule of truth-telling, “sacred as it is, admits of possible exceptions.”

  45. The position described here is quite compatible with Berger's “strategy” conception of moral rules. See Berger, Happiness, Justice and Freedom, pp. 72 ff. The position here is also compatible with Henry West's. See his “Mill's Moral Conservatism,” Midwest Studies in Philosophy 1 (1976), pp. 71-80.

  46. See Lars Bergstrom, “Utilitarianism and Future Mistakes,” Theoria 43 (1977), pp. 84-102; Jordan Sobel, “Utilitarianism and Past and Future Mistakes,” Nous 10 (1976), pp. 195-219; Brian Ellis, “Retrospective and Prospective Utilitarianism,” Nous 15 (1981), pp. 325-339; Hector-Neri Castaneda, “On the Problem of Formulating a Coherent Act-Utilitarianism,” Analysis 32 (1972), pp. 118-124. For related comments, see Fred Feldman, Doing the Best We Can (Dordrecht: D. Reidel Publishing Co., 1986), Ch. 1.

    These theorists make interesting and important points which will not be addressed here both because these points do not directly address Mill's system and because insofar as they do indirectly address his system, they (analogously, if not directly) might be made against many, if not all, moral systems. For example, there are serious problems raised by the notion of group actions for moral theorists in general, at least in part, because such a notion may force theorists to reassess the basic unit of moral evaluation. There are also serious problems raised insofar as one should consider others' and one's own probable future immoral actions.

    Sympathizing with some of these criticisms directed at act-utilitarianism, Holbrook argues that “an action is right if and only if it is part of that series of actions that has the greatest overall pleasure as its result.” See Daniel Holbrook, Qualitative Utilitarianism (Lanham: Press of America, 1988), p. 42. He does not seem to appreciate that on this account an agent may be held morally accountable for others' moral misdeeds. While we should try to anticipate and consider our own and others' future actions when we deliberate about what to do currently, a theory like Holbrook's demands much more than that and seems to destroy the notions both of individual agency and of (individual) moral responsibility. In any case, the kind of analysis required to handle these problems would have to be too extensive and too foundational to be appropriately handled here.

  47. This is Regan's example. See Donald Regan, Utilitarianism and Co-operation (Oxford: Oxford University Press, 1980), p. 18

  48. Information constraints create difficulties for a variety of theories. Suppose that we adopt Regan's position and say, “What each agent ought to do is co-operate, with whoever else is co-operating in the production of the best consequences possible given the behaviors of the non-co-operators.” Regan, Utilitarianism and Co-operation, p. 124. Or, we can adopt the proposal suggested by Postow and say, “In any given situation, any group of one or more agents ought to follow a course of action by means of which the group would produce the most good that it can produce in that situation.” B. C. Postow, “Generalized Act Utilitarianism,” Analysis 37 (1977), p. 51. (For a discussion of how agents who lack perfect knowledge should act if they wish to be rational, see David Gauthier, Morals By Agreement (Oxford: Oxford University Press, 1986), especially Ch. 3.) If there is no available information about what others are doing, the agent may have to flip a coin. Whether one is an act-utilitarian or a co-operative utilitarian, one will simply have to hope for the best.

  49. In the next chapter, we shall see that Mill is sometimes thought to evaluate actions in terms of their foreseeable consequences. If Whiff has no reason to believe that Poof will act one way rather than the other, then neither of Whiff's alternatives is foreseeably better and he acts rightly no matter which of the two options he chooses, regardless of Poof's choice.

  50. Utilitarianism, Ch. 2, Par. 23, p. 223.

  51. Ibid.

  52. Ibid., Ch. 2, Par. 6, p. 212.

  53. Ibid., Ch. 2, Par. 9, p. 213.

  54. John Stuart Mill, On Liberty, Ch. 1, Par. 11, p. 224.

  55. “Auguste Comte and Positivism,” p. 333.

  56. Ibid., p. 334.

  57. Ibid., p. 335. Levi argues that ‘the permanent interests of man as a progressive being’ are related to the actualization of certain human potentials. See Albert William Levi, “The Value of Freedom: Mill's Liberty (1859-1959),” Ethics 70 (1959), p. 40. That interpretation is quite compatible with this one, as long as the actualization of those potentials would indeed promote the general happiness of humankind.

    By the same token, theorists are correct that Mill argues for the great importance of the promotion of autonomy and individuality, as long as they keep in mind that Mill so argues because he believes that the promotion of these qualities is necessary for the progress and well-being of humankind. See On Liberty, Ch. 3, Par. 1 ff, pp. 260 ff.

  58. John Stuart Mill, “Remarks on Bentham's Philosophy,” CW Vol. 10, p. 15. For related comments, see Richard Arneson, “Mill Versus Paternalism,” Ethics 90 (1980), pp. 470-489.

  59. A System of Logic, Book 6, Ch. 12, Sec. 7, p. 951.

  60. Utilitarianism, Ch. 2, Par. 10, p. 214.

    In the passages above, Mill is clearly interested in promoting humankind's utility. Sometimes, Mill writes as if he is interested in promoting the happiness of all sentient beings. It is not clear how the happiness of each sentient creature should be weighed, e.g., whether each counts for one and no one for more than one. See “Whewell on Moral Philosophy,” pp. 186 ff. Mill writes that the maximization of the happiness of all sentient creatures would be best, “as far as the nature of things admits.” Utilitarianism, Ch. 5, Par. 10, p. 214. It is not clear how this qualification should be treated.

    In this work, it will be assumed that Mill believes that the act which maximizes humankind's (foreseeable) utility is morally best. The account can be suitably modified, depending upon how animals' interests should be weighed.

  61. “Auguste Comte and Positivism,” p. 290.

  62. Utilitarianism, Ch. 2, Par. 24, p. 224.

  63. Principles of Political Economy, Book 4, Ch. 7, Sec. 3, Par. 1, CW Vol. 3, p. 765. See also “Chapters on Socialism,” CW 5, p. 729. For a brief, related (historical) discussion of Mill's views on the dissemination of birth control information, see Michael Packe, The Life of John Stuart Mill (London: Secker and Warburg, 1954), pp. 56 ff. See also Gertrude Himmelfarb, On Liberty and Liberalism (New York: Alfred A. Knopf, Inc., 1974), p. 121.

  64. A vast literature has been devoted to the topic of our obligations to future generations. See, e.g., R. I. Sikora and Brian Barry (eds.), Obligations to Future Generations (Philadelphia: Temple University Press, 1978).

  65. “Whewell on Moral Philosophy,” p. 172. Brandt seems to want to attribute his theory of indirect utility to Mill. See Brandt's “Fairness to Indirect Optimific Theories in Ethics,” p. 342. As will be explained more fully in later chapters, Mill evaluates an action's rightness according to the principle of utility and he evaluates a motivation's goodness according to the principle of utility, but not the moral goodness of an action in terms of the goodness of the motivation which, itself, is evaluated according to the principle of utility—Mill argues that the rightness of an action and the goodness of a motivation are each directly evaluated according to the principle of utility. The theory which Brandt describes is more appropriately attributed to Hutcheson. See my Francis Hutcheson's Moral Theory: Its Form and Utility.

  66. “Auguste Comte and Positivism,” pp. 322 ff.

  67. See my “Mill and the Utility of Liberty,” Philosophical Quarterly 34 (1984), pp. 63-68. See also Ted Honderich's “The Worth of J. S. Mill On Liberty,Political Studies 22 (1974), pp. 463-470 and his “‘On Liberty’ and Morality-Dependent Harms,” Political Studies 30 (1982), pp. 504-514.

  68. Utilitarianism, Ch. 5, Par. 25, pp. 250-251.

  69. See Williams's claim in J. J. C. Smart and Bernard Williams, Utilitarianism For and Against (Cambridge: Cambridge University Press, 1973), p. 100.

  70. Utilitarianism, Ch. 5, Par. 37, p. 259. Narveson criticizes rule-utilitarianism because it might morally require that one perform what one knows to be a non-utility-maximizing action. See his Morality and Utility (Baltimore: Johns Hopkins Press, 1967), p. 128.

  71. A System of Logic, Book 5, Ch. 7, Sec. 3, Par. 4, p. 829.

  72. For a related discussions, see Copp on “generic effects.” David Copp, “The Iterated-Utilitarianism of J. S. Mill,” New Essays on John Stuart Mill and Utilitarianism, p. 94.

  73. Utilitarianism, Ch. 5, Par. 21, p. 249.

  74. Goldman points out that there is insufficient appreciation of the difficulty of providing the relevant description of the appropriate moral rule within certain forms of rule-utilitarianism. See Holly Goldman, “David Lyons on Utilitarian Generalization,” Philosophical Studies 26 (1974), pp. 78 ff. Feldman claims that Lyons uses the wrong criterion for relevance. See Fred Feldman, “On the Extensional Equivalence of Simple and General Utilitarianism,” Nous 8 (1974), pp. 191 ff. See also Gertrude Ezorsky, “A Defense of Rule-Utilitarianism Against David Lyons Who Insists on Tieing it to Act-Utilitarianism, Plus a Brand New Way of Checking Out General Utilitarian Properties,” Journal of Philosophy 65 (1968). For an informative discussion on this and related matters, see Gerald Gaus, “Mill's Theory of Moral Rules,” Australasian Journal of Philosophy 58 (1980), pp. 265-279.

Roger Crisp (essay date 1997)

Download PDF PDF Page Citation Cite Share Link Share

SOURCE: “Utilitarianism and Equality: The Subjection of Women,” in Mill on Utilitarianism, Routledge, 1997, pp. 201-15.

[In the following essay, Crisp considers the implications of Mill's utilitarianism with regard to the equality of women.]


… [The] cornerstone of Mill's practical view is the principle of utility. According to this principle, the right act is that which maximizes overall welfare. Some of our acts involve our taking part in the practices of everyday, or ‘customary’, morality. Because my child is less likely to attack others if I encourage her to feel proud at her self-control and kindness, and shame and guilt at her cruelty, it makes utilitarian sense to bring her up to feel these emotions at the proper times, and thus to guide her conduct in a utilitarian direction.

… Mill accepts that some parts of customary morality may well be grounded on the promotion of human welfare (The Subjection of Women: SW 1.5; cf. SW 1.4). These include certain ‘secondary principles’, such as principles of justice, which have been initiated and continued reflectively, and tested against alternatives. Other parts of customary morality, however, such as what he saw as the common readiness to permit interference with individuals for their own good, he finds abhorrent, and wishes to see replaced. In other words, the mere fact that a certain moral principle is widely accepted does not justify it. Custom itself has no authority.

Another area in which Mill thinks the customary morality of his day is in need of radical reform concerns the relations between the sexes. Here Mill recognized that he had a great battle on his hands, and it was one he fought throughout his life, from the time he was arrested at the age of 17 for handing out leaflets on contraception until the final period of his life, in which he attempted in Parliament to extend suffrage to women. The battle, as he sees it, is against custom, and the intense and irrational feelings which protect it (SW 1.2). One of Mill's primary tasks in The Subjection of Women is to reveal the reality of oppression behind the genteel appearance of Victorian chivalry:

It was inevitable that this one case of a social relation grounded on force, would survive through generations of institutions grounded on equal justice, an almost solitary exception to the general character of their laws and customs; but which, so long as it does not proclaim its own origin, and as discussion has not brought out its true character, is not felt to jar with modern civilization, any more than domestic slavery among the Greeks jarred with their notion of themselves as a free people.

(SW 1.6)

What, then, was the origin of the relative positions of men and women? According to Mill, fully to understand this requires one to grasp the history of morality itself, so that one can see how those relations are still governed by a morality which has in other areas of life become obsolete.

Mill suggests that in the very earliest human societies each woman, because of her physical weakness and sexual attractiveness, would be enslaved to a man. Out of this primitive Hobbesian state of nature developed legal systems which merely legitimized these ownership relations:

Slavery, from being a mere affair of force between the master and the slave, became regularized and a matter of compact among the masters, who, binding themselves to one another for common protection, guaranteed by their collective strength the private possessions of each, including his slaves.

(SW 1.5)

Primitive slavery, then, is the source of the inequality of women. And in the legitimation of this primitive relation is to be found the origin of social or legal obligation: those who disobeyed the ‘law of force’ were thought guilty of the most serious crime, and duly punished with great cruelty (SW 1. 7). Here we must not forget Mill's brief but important account of morality's origin in the desire to punish in Utilitarianism 5.14.

Gradually, a certain sense of obligation developed in some superiors towards some inferiors (though not slaves), as the result of the development of consciences which motivated the keeping, from a sense of duty, of promises originally made for convenience. This evolution out of the pure law of force was the beginning of what we might recognize as a morality:

[T]he banishment of that primitive law even from so narrow a field, commenced the regeneration of human nature, by giving birth to sentiments of which experience soon demonstrated the immense value even for material interests, and which thenceforward only required to be enlarged, not created.

(SW 1.7)

Then the view arose among the Stoics, and was taken up in Christianity, that there are obligations to slaves (SW 1.7; cf. SW 2.12). Though not widely adopted, this was the clearest example of a new stage in morality—the morality of chivalry, in which the strong would be praised for refraining from oppressing the weak (SW 2.12).


Women, in fact, had from very early times played a rôle in creating the conditions for the morality of chivalry (SW 4.8). Because of women's vulnerability, they encouraged men not to be violent, and because of female weakness, to settle disputes by non-violent means. Women did not, however, wish their men to be cowards, because, after all, they themselves needed protection. So courage was always encouraged in sons by their mothers, and in young men by their female lovers. A possessor of military virtue would also be admired by other men, and such admiration would provide another opportunity of entry into the affections of women. The morality of chivalry, then, consists in an apparently odd mix of militarism and gentleness, especially towards women, who used their sexual power to their best advantage.

The morality of chivalry, however, is a thing of the past (SW 4.9). The individualism of the warrior has been replaced by cooperation in business and industry: ‘The main foundations of the moral life of modern times must be justice and prudence; the respect of each for the rights of every other, and the ability of each to take care of himself’ (SW 4.9).

Chivalry was not entirely successful. It depended on praise rather than blame and punishment, and for that reason left untouched the behaviour of the majority, unimpressed by the rewards of honour. With the morality of justice, society has the collective means to enforce decency on the part of the strong without relying on their higher feelings, and indeed it has done so—except as regards the relation of women and men. The behaviour of most men towards most women is governed not by the morality of justice, nor even the morality of chivalry, but the morality of submission—the law of force.

The morality of justice is grounded on the equality of human beings. One's birth does not determine rigidly the course one's life must take; rather one is free to employ the talents one has in whatever career one wishes (SW 1.13; SW 4.5). And this freedom … protects autonomy, one of the most important sources of welfare. It is recognized both as an injustice to the individuals concerned, and as damaging to the interests of society as a whole, to place needless obstacles in the way of individual advancement: ‘Nobody thinks it necessary to make a law that only a strong-armed man shall be a blacksmith.’

There is clearly some hyperbole in Mill's claims about the morality of justice, and indeed he does become more sanguine when reflecting upon the constraints placed on its development in society by the present relations of the sexes. But it is no doubt true that Mill is writing at a time in which sexual inequality is not widely recognized as a serious injustice. What is Mill's explanation?

One main cause is the continuing physical weakness of women (SW 1.6). Men have nothing to fear from women, and so have seen no good reason to concede ground to them. Mill notes that less than forty years before the publication of The Subjection of Women, British citizens could own other human beings, and that absolute monarchy or military despotism as systems of government had only recently begun to decline in Europe. ‘Such is the power of an established system’ (SW 1.8). Since every man has the advantage of despotism over those closest to him, is it any wonder that he has not resigned his power?

More support for the customary morality of sexism, Mill believes, is provided by the sentiments and feelings (SW 1.2). Given that men have inculcated in women duties of submission and self-abnegation (SW 1.11), it is no surprise that many women do not desire any reform.

But Mill realizes that many women have not been taken in (SW 1.10). Many women complain of sexism in their writings, and are petitioning vociferously for the vote and for education. In addition, even more women, though they do not complain of men as a whole, resent the tyranny of their own husband. Yet more would complain if the hold of husbands over wives were not so tight.

Mill believes that wives are in a position of slavery (SW 2.1). His hope is that once this is recognized, radical changes will follow. What those changes will be, and how they are to come about, we must now begin to consider.


Mill is appalled by the position of women in his society. He is particularly concerned by the nature of marriage, which he sees women as being coerced into by the lack of any serious alternative (SW 1.25). The history of marriage as he describes it is part of the history of morality. It has its origin, as did ordinary slavery, in the law of force, and women are owned either by their husbands or by their fathers (SW 2.1). A woman has no legal rights against her husband, who has literal sovereignty over her; under ‘the old laws of England’, murder of a husband had been considered high treason, the penalty being death by burning. Even now, Mill notes, a woman's avowal of obedience at the altar is upheld by the law, her property passing to her husband (but not vice versa). A woman has no time of her own; she is constantly at her husband's beck and call. Nor does she have any rights over her body; there is no crime of rape within marriage in the law of England. The husband has legal rights over the children, and most women have no real opportunity of escaping the tyranny of their husbands.

Sexism, of course, is not confined to the private sphere. Here Mill finds it unnecessary to go into detail (SW 3.1), since it is an undeniable fact that women are prohibited from applying for places at university, and from the ‘greater number of lucrative occupations, and from almost all high social functions’. In particular, of course, women can neither vote nor stand for parliament (SW 3.2).

What is to be done? Mill says that the object of The Subjection of Women is to advocate ‘a principle of perfect equality, admitting no power or privilege on the one side, nor disability on the other’ (SW 1.1). Marriage should be on equal conditions (SW 1.25). As a voluntary association, it should be governed jointly, with a division of powers between wife and husband analogous to that between partners in a business (SW 2.7). Women should of course be given the vote, and be allowed admittance to the same positions in government, business and so on as men.

Mill's views, especially those on marriage, were considered dangerously radical even by most of those otherwise sympathetic to him. James Fitzjames Stephen, one of Mill's ablest critics, thought his position on female equality to be verging on the indecent (Stephen 1874: 134-5). It is no surprise, therefore, that Mill delayed the publication of the book, and strove to avoid particularly delicate issues such as divorce (SW 2.1).1

Nevertheless, a common criticism of Mill by modern writers is that he does not go far enough. Mill can be seen at best as a proponent of the rights of women, not their liberation (see e.g. Goldstein 1980). He advocates mere equality of opportunity for women, not seeing that the social conditioning of women, of which he was so aware, would prevent the majority's taking up any opportunities made available to them.

This criticism might be thought a little unfair to Mill. We know that he considers both law and customary morality to be important social tools for the maximizing of welfare, so it is to be expected that he will concentrate on reform within these institutions as a means to progress. He is quite aware of the indoctrination in women by society of exaggerated views about the moral and sexual importance of female meekness and submissiveness (SW 1.11). He believes, however, that equality in law and customary morality will result in important changes in the nature of the family and the relationships within it, and indeed in the character of society at large:

The family, justly constituted, would be the real school of the virtues of freedom … What is needed is, that it should be a school of sympathy in equality, of living together in love, without power on the side or obedience on the other. This it ought to be between the parents. It would then be an exercise of those virtues which each requires to fit them for all other association, and a model to the children of the feelings and conduct which their temporary training by means of obedience is designed to render habitual.

(SW 2.12)

Removing sexual discrimination would also allow women ‘a life of rational freedom’ (SW 4.19). Here Mill speaks of the ‘liberation’ of women, and he is clearly to be understood as speaking of women's being not only uncoerced by men, but positively free to make for themselves important decisions about the shape of their lives.

So Mill is concerned with the liberation of women, and with enabling them to exercise equal rights, not merely to possess them. And there is much in what he says. Were most in society, whether female or male, to begin to see women as the legal and moral equals of men in both private and public spheres, and not to view sex as relevant where it clearly is not, then the position of women would be greatly advanced. Nevertheless, as I shall show in the next section, the modern objection to Mill I have been discussing is not completely off target.


Mill recognizes the authority of custom. His error, perhaps, is not to see how entrenched are the views of women in modern society. In an early essay, Mill had claimed that there was no natural inequality between the sexes, except perhaps in physical strength (O 21.42). In The Subjection of Women, Mill is more circumspect. The debates about the nature of women raged during his day as they do in ours, and he was concerned to sidestep them. As a strict empiricist, ready only to accept the evidence of experience, Mill says that we can know nothing of the nature of women, other than, presumably because we are aware of the forces of distortion at work, that their present state is not the natural one (SW 1.18). History teaches us of the great openness of human beings to external influence (SW 1.19). Domination has always appeared natural, and indeed the unnatural is little more to most people than what is unusual (SW 1.9). Views of what is natural vary wildly (SW 1.9; SW 3.14). Women themselves have been permitted to say little of their own experience, and frankness is anyway hardly a common quality in unequal relationships (SW 1.21). But, Mill says, from the moral point of view this is all by the by: ‘For, according to all the principles involved in modern society, the question rests with women themselves—to be decided by their own experience, and by the use of their own faculties' (SW 1.23).

Unfortunately, however, in speaking of women as they are at present, Mill is less careful. While attempting to explain why women should be permitted to enter business and public life, Mill offers a set of generalizations about women's capacities which are unsupported in just the way he elsewhere criticizes (SW 3.8-13). Women are said to be more practical than men, and to possess ‘intuition’ and a sensitivity to particular facts instead of the male capacity to reason abstractly. They tend to be more nervous, their minds more mobile than men’s, but less able to concentrate.

Admittedly Mill takes pains to stress that he is not making claims about the universal nature of women, but only women as they are. Nevertheless, his ‘mere empirical generalizations, framed, without philosophy or analysis' (SW 3.14) not only make Mill the first to enter the quagmire of the sameness/difference debate which concerns many contemporary feminists, but suggest that it was perhaps fortunate that time did not permit him to carry out the programme of ‘ethology’ he describes at the end of the System of Logic. Whether Mill's claims be true or false, they are both irrelevant to the practicalities of his argument and based on no more evidence than claims about the nature of women as, say, instinctively maternal or sexually insatiable.

Even worse, Mill allows himself to express opinions about which choices women should make once liberated, and they are somewhat conservative:

In an otherwise just state of things, it is not … a desirable custom, that the wife should contribute by her labour to the income of the family … Like a man when he chooses a profession, so, when a woman marries, it may in general be understood that she makes choice of the management of a household, and the bringing up of a family, as the first call upon her exertions.

(SW 2.16)

One might think Mill says these things merely to ingratiate himself with his sexist audience, and so perhaps win them over to a more liberal position (compare his comments on Muslims and on common decency in On Liberty). But this is probably not the case. Mill believed that mothers were closer to children than their fathers (‘Letter to Hooker’ (1869) 17.1640), and does not raise the possibility that men could be involved in the raising of children (SW 3.1).

Mill's mistake is not just to generalize about women without good evidence when he has explicitly warned against so doing. He makes two further errors analogous to some I [have elsewhere] outlined in On Liberty. … [Concerning that] discussion of our attitude to self-regarding faults in On Liberty, Mill failed to see that social despotism might be based on the criticism of others which he permitted. Here, he fails to note that, if it is widely believed that women should raise children and not work, this will create an atmosphere in which it is difficult for women to make a free choice. Other options may just not occur to them, and there may be social pressure to conform. Further, in the passage of On Liberty on self-regarding faults, Mill seems temporarily unaware not only of the authority of custom, but also of the authority of his own text. The same is true in The Subjection of Women. By advocating a domestic rôle for women, Mill was playing into the hands of those sexists who wished to prevent women from competing on equal terms with men by any means available.

The tension in Mill's writing on women demonstrates the power of the very ideology he attempted to reveal. On the one hand, he sees very clearly the intricate methods of subjection employed by men against women over the ages. On the other, he fails to recognize that he himself has been taken in by the myth that the rôle of the vast majority of women is to raise children, along with other myths, such as that the age of a husband, or the fact that he earns money, entitles him to a greater say in marriage (SW 2.9).

These blemishes, however, should not blind us to the force and eloquence of Mill's case for the equality of women. He does anyway allow that ‘the utmost latitude ought to exist for the adaptation of general rules to individual suitabilities' (SW 2.16). And we find in his spelling out of the morality of justice a position which not only provides a resource for modifying his own position on child-rearing but is still important in debates about sexual equality:

[I]t is felt to be an overstepping of the proper bounds of authority to fix beforehand, on some general presumption, that certain persons are not fit to do certain things. It is now thoroughly known and admitted that if some such presumptions exist, no such presumption is infallible. Even if it be well grounded in a majority of cases, which it is very likely not to be, there will be a minority of exceptional cases in which it does not hold: and in those it is both an injustice to the individuals, and a detriment to society, to place barriers in the way of their using their faculties for their own benefit and for that of others. In the cases, on the other hand, in which the unfitness is real, the ordinary motives of human conduct will on the whole suffice to prevent the incompetent person from making, or from persisting in, the attempt.

(SW 1.13; cf. SW 3.1)

In 1869, Elizabeth Cady Stanton, a pioneer of women's liberation in America, wrote to Mill concerning his book:

I lay the book down with a peace and joy I never felt before, for it is the first response from any man to show he is capable of seeing and feeling all the nice shades and degrees of women's wrongs and the central point of her weakness and degradation.

(Lutz 1940: 171-2; cited in Rossi 1970: 62)

In the final section of this essay, let me show how Mill employed the theory he had developed in Utilitarianism and On Liberty to argue for women's equality.


Mill advocates that, rather than allow the question of women's position in society to be decided by prevailing custom, one should think of it

as a question of justice and expediency: the decision on this, as on any of the other social arrangements of mankind, depending on what an enlightened estimate of tendencies and consequences may show to be the most advantageous to humanity in general, without distinction of sex.

(SW 1.17)

The morality of justice, then, is the customary morality which Mill believes suits modernity. In the circumstances in which we find ourselves, equality is to govern our relations with one another, because in this way human welfare will be best promoted.

In his discussion of justice in Utilitarianism, Mill suggests that there is more disagreement about equality than about any other component of justice (5.10). It is clear, however, what he means by equality between the sexes: a substantive equality of rights, allied with an attitude of equality of respect by each citizen towards every other, male or female. Equality for Mill is closely tied to desert: ‘We should treat all equally well (when no higher duty forbids) who have deserved equally well of us' (5.36). And he is acutely aware, as we shall see below, of the distress caused to women who saw that the arrangements in their society prevented their exercising their talents as they wished.

What, then, would be the consequences of equality? The most important, of course, would be for women. First, they would be relieved from the terrible suffering inflicted by many husbands upon their wives (SW 2.1; SW 4.2). Mill, a husband himself, is of course aware that not all husbands are tyrants; but he is equally, and painfully, aware of the level of violence against women in the home and the lack of protection offered them by society and the law. Nor is it any surprise that women love their husbands whether they batter them or not: similar attachments were forged between slave and master in Greece and Rome.

Secondly, women might find a positive benefit in a marriage based on equality (SW 4.15-18). There might be far greater similarity in interests, tastes, wishes and inclinations, and a consequent decrease in painful disagreement. Mill compares the ideal marriage to friendship, in which association and sympathy enable each partner to enrich himself or herself through insight into the other's view of the world. There is little doubt that Mill is speaking here from the experience of his relationship with Harriet Taylor, and, given his view of sexual intercourse as an ‘animal function’ (SW 2.1), it is to be expected that he emphasize the non-sexual aspect of the marriage relationship.

Thirdly, and no less importantly, Mill's reforms would allow women a life of rational freedom in place of one of subjection to the will of men (SW 4.19-20). Here we see that Mill's case depends not only on Utilitarianism, and the conceptions of justice and human welfare there developed, but on the understanding of liberty and individuality worked out in On Liberty. Mill argues that, after food and clothing, freedom is the most important want of any human being. Mill asks his reader not to concentrate on how much freedom matters to others, but on how much it matters to the reader themselves—and then to base their conclusions about others on their own case:

Whatever has been said or written, from the time of Herodotus to the present, of the ennobling influence of free government—the nerve and spring which it gives to all the faculties, the larger and higher objects which it presents to the intellect and feelings, the more unselfish public spirit, and calmer and broader views of duty, that it engenders, and the generally loftier platform on which it elevates the individual as a moral, spiritual, and social being—is every particle as true of women as of men. Are these things no important part of individual happiness?

(SW 4.20)

The experience of freely running one's own life, then, is a central component of human welfare. Mill stresses two further elements which women will be able to incorporate in their lives through exercising this freedom: achievement, and the enjoyment of contemplating it (SW 4.21-2). As we have seen, Mill believes women should remain in the home, so his proposals as he conceives them concern mainly women who have brought up a family. He suggests that the only active outlet available to such women at present is charity, in which the effect is merely to remove the capacity for autonomy from those who are helped (SW 4.11). Not only could women achieve in the spheres in which men currently achieve, but they could enjoy their achievement, something again ‘vitally important to the happiness of human beings’, and be spared the undeserved disappointment and dissatisfaction of a wasted life.

Mill realizes that he cannot rest his case on the advantages of equality to women alone. Some benefits, he argues, will accrue to men. Like women, they will be able to enjoy a marriage of friendship with an equal. And there will be less obvious improvements to their position. For example, a man who wishes to live his life according to his own conscience and in accordance with the truth as he sees it will be able to do so, even if his opinions are at odds with those of the masses (SW 4.13-14; cf. S 2.5). At present, he would probably not attempt it, because of the social sacrifice he would be imposing on his family: ‘Whoever has a wife and children has given hostages to Mrs Grundy.’2 Mill was quite aware of the effects of social ostracism after the way he and Mrs Taylor were treated by society at large.

But the main advantages for men will come from those to society as a whole. Society will benefit from the constructive, as opposed to the mischievous, influence of women (SW 4.8-12, 20). At present, women denied liberty seek whatever power they can, without regard for its wider social implications. Sheer power depraves, whether it be held by a woman or a man. With equality, not only will women's interests move beyond the sphere of Mrs Grundy, but they will extend to combine the virtues of gentleness with a disinterested concern for society at large.

Discrimination is inefficient, resulting in the failure to realize or use the potential of those discriminated against (SW 1.13-14, 24; SW 3.1; SW 4.6-7). The equal education of women and their entry into professions previously prohibited to them would make available huge mental resources, as well as spur men to great competition.

Equal education would also affect the characters of both men and women: women would become more assertive, men more self-sacrificial (SW 2.10). And, as we saw above, there would be other effects on the morality of society, emerging from the family as a ‘school of sympathy in equality’ (SW 2.12). Present institutions result in men's becoming ‘depraved’ (SW 2.13; cf. SW 2.4). Mill is quite adamant about this: ‘All the selfish propensities, the self-worship, the unjust self-preference, which exist among mankind, have their source and root in, and derive their principal nourishment from, the present constitution of the relation between men and women’ (SW 4.4).

It might be thought that Mill is exaggerating. Certainly he is again sticking his neck out further than his empiricism might allow. But it is not implausible to suggest that were sexism to be rooted out, for the right reasons, from the family at all levels, many of those brought up within the institution would live up to the morality of justice elsewhere: ‘Though the truth may not be felt or generally acknowledged for generations to come, the only school of genuine moral sentiment is society between equals' (SW 2.12).

The vision of a society in which all are granted the rights and respect they deserve is at the heart of the liberalism and egalitarianism which Mill's act utilitarianism supported. That vision is as important now as it ever was, and modern utilitarians have stressed that the boundaries of such a society should extend to include not only women, but human beings in all countries, and indeed non-human animals. Because of the depressing fact that Mill's ideal remains in many respects as distant as it did in his time, his moral and political writings, all of which must be read in the light of Utilitarianism, will rightly remain at the centre of moral and political debate for many years to come.


  1. A tricky subject for him anyway, given his relationship with Mrs Taylor: see ch. 1.

  2. Mrs Grundy, in Thomas Morton's Speed the Plough (1798), represents a narrow-minded sense of propriety.

Henry R. West (essay date 1997)

Download PDF PDF Page Citation Cite Share Link Share

SOURCE: “Mill's ‘Proof’ of the Principle of Utility,” in Mill's Utilitarianism, edited by David Lyons, Rowman & Littlefield Publishers, Inc., 1997, pp. 85-98.

[In the following essay, West asserts the logical plausibility of Mill's proof of his utility principle.]

Utilitarianism, in every one of its forms or formulations, requires a theory for the evaluation of consequences. Whether the units of behavior being judged are acts, rules, practices, attitudes, or institutions, to judge them by their utility, that is, by their contribution to good or bad ends, requires a theory of what count as good or bad ends. In the philosophies of the classical utilitarians, Jeremy Bentham and John Stuart Mill, some variety of hedonism served this purpose. Mill calls this

the theory of life on which this theory of morality is grounded—namely that pleasure, and freedom from pain, are the only things desirable as ends; and that all desirable things (which are as numerous in the utilitarian as in any other scheme) are desirable either for the pleasure inherent in themselves, or as means to the promotion of pleasure and the prevention of pain.1

In Chapter IV of his essay entitled Utilitarianism, Mill addresses himself to the question of what sort of proof this principle is susceptible. The twelve paragraphs of the chapter present an argument which, if successful, is one of the most important arguments in all of moral philosophy, for, although Mill says questions of ultimate ends are not amenable to “direct proof,” he believes that considerations may be presented capable of determining the intellect to give its assent to the doctrine.2

Unfortunately, few commentators on Mill have had their intellects convinced, and perhaps even fewer have agreed on the correct interpretation of the argument. J. B. Schneewind says of it:

A greater mare's nest has seldom been constructed. It is now generally agreed that Mill is not, in this chapter, betraying his own belief that proof of a first moral principle is impossible, but there is not a general agreement as to what he is doing. In the last fifteen years there have been more essays dealing with the topic of “Mill's Proof” than with any other single topic in the history of ethical thought.3

Mill does say of the Utilitarian formula of ultimate ends that it is impossible to give a proof in the “ordinary or popular meaning of the term,” but he continues,

We are not, however, to infer that its acceptance or rejection must depend on blind impulse or arbitrary choice. … The subject is within the cognizance of the rational faculty, and neither does that faculty deal with it solely in the way of intuition. Considerations may be presented capable of determining the intellect either to give or withhold its assent to the doctrine, and this is the equivalent of proof.4

Furthermore, at the end of Chapter IV, he says that if the doctrine he has argued for is true, “the principle of utility is proved.5

Given the claim, that he has proved his principle or presented something equivalent to proof, I think it is worthwhile to lay out the structure of the argument in deductive form. In that way we can determine the nature of the premises he introduces, locate the gaps that prevent it from being a valid deduction, and see if plausible assumptions can be formulated or interpretations offered that will support the premises and bridge the gaps. I shall present what I believe to be a reasonable interpretation of what Mill had in mind, and I shall claim that the assumptions necessary to make the argument sound are—though controversial—at least plausible.

The conclusion he is seeking is stated in Paragraph 2: “The utilitarian doctrine is that happiness is desirable, and the only thing desirable, as an end, all other things being only desirable as a means to that end.” The connection between this idea and morality is mentioned at the end of Paragraph 9, where he says that the promotion of happiness is the test by which to judge all human conduct, “from whence it necessarily follows that it must be the criterion of morality, since a part is included in the whole.” Mill has a complex view of the way the ultimate standard of the promotion of happiness is to be applied in morality, and morality is only one of the “three departments” of what he calls the “Art of Life.” These general teleological views on testing conduct, and the place of morality within this teleological framework, are found elsewhere in the essay and in the Logic.6 They are not part of this “proof,” and I shall not be discussing them in this paper. I shall be examining only how he gets to the conclusion that happiness is desirable and the only thing desirable as an end.

The structure of the argument is very simple. In Paragraph 3 he argues that happiness is desirable. In the remainder of the chapter he argues that happiness is the only thing desirable. The outline of the argument can be given in Mill's own words:

(1) “The sole evidence it is possible to produce that anything is desirable is that people do actually desire it.” (Paragraph 3)

(2) “Each person, so far as he believes it to be attainable, desires his own happiness.” (Paragraph 3)


(3) “Happiness is a good.” (Paragraph 3)

He substitutes the expression “is a good” for the expression “is desirable,” but I presume that this is only for stylistic reasons. I think he would regard his use of these two as interchangeable in this context.7

The argument to show that happiness is the only thing desirable is likewise based on the evidence of actual desire:

(4) “Human nature is so constituted as to desire nothing which is not either a part of happiness or a means of happiness.” (Paragraph 9, but argued throughout Paragraphs 5-10)


(5) “Nothing is a good to human beings but insofar as it is either pleasurable or a means of attaining pleasure or averting pain.” (Paragraph 11)

Here his use of “pleasure” and “pleasurable” instead of “happiness” is merely stylistic. Throughout the essay he says that by happiness he means pleasure and freedom from pain.

This is the simple outline of the argument. It is complicated by the fact that each individual's desire is for his own happiness, whereas the utilitarian doctrine that Mill is seeking to establish is that the general happiness is the foundation of morality.8 In Paragraph 3, this distinction is explicit. Having said that each person desires his own happiness, Mill says we have all the proof it is possible to require

that happiness is a good, that each person's happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons.

These propositions we may state as separate these:

(3A) “Each person's happiness is a good to that person.”


(3B) “The general happiness [is] a good to the aggregate of all persons.”

The distinction also can be introduced into the second part of the argument. Without doubt the psychological premise (4) means:

(4′) Each person desires nothing which is not either a part of his happiness or a means of his happiness.9

And parallel to (3A) and (3B) the distinction between each person and the aggregate of all persons can be introduced. This would give:

(5A) Nothing is a good to each person but insofar as it is either a part of his happiness or a means of his happiness.


(5B) Nothing is a good to the aggregate of all persons but insofar as it is either a part of the general happiness or a means of the general happiness.

From (3B) and (5B), we can then deduce an interpretation of the “utilitarian doctrine,” as follows:

(6) “[The general] happiness [or a part of the general happiness] is desirable, and the only thing desirable, as an end, all other things being only desirable as means to that end.” (Paragraph 2)

Examining the argument, it can be seen that (1) is a methodological premise; (2) and (4) are factual psychological premises; (3) or (3A) and (5) or (5A) are supported by (2) and (4), respectively. (3B) and (5B) putatively follow from (3A) and (5A), respectively. The premise (2) is probably non-controversial. The controversial premises are (1) and (4), and the controversial steps are from the fact of happiness being desired to its being normatively desirable, and from each person's happiness being desirable to that person to the conclusion that the general happiness is desirable to the aggregate of all people. There seem, then, to be three central issues: (A) Mill's methodology, which is to argue for what is desirable on the evidence of what is in fact desired; (B) his psychological hedonism, that each person desires his own happiness as an end and nothing as an end which is not a part of his happiness; and (C) the argument that if each person's happiness is a good and inclusive of the only good for him, as an end, the general happiness is a good, and encompasses the only good, as an end, to the aggregate of all persons. I shall take these issues up in turn.


I think it is hardly necessary to point out that Mill did not say that “desirable” or “the good” means “desired,” as Moore says he does.10 He is not committing a naturalistic or definist fallacy.11 Mill is quite explicit in demarcating factual and normative propositions.12

I also think it is not likely he was misled by the similarity of the verbal endings of “visible” and “audible” into thinking that “desirable” means “able to be desired.”13 The significance of the analogy with visible and audible is announced in the first paragraph of the chapter. The first premises of our knowledge do not admit of proof by reasoning, but are subject to a direct appeal to the senses; he suggests that the first premises of conduct are subject to a direct appeal to our desiring faculty. It does not follow that he regards what is desirable as a “permanent possibility of desire.” That would be to regard it as a matter of fact. The analogy is that as judgments of existence are based on the evidence of the senses and corrected by further evidence of the senses, so judgments of what is desirable are based on what is desired and corrected by further evidence of what is desired. The only evidence on which a recommendation of an end of conduct can be based is what is found appealing to the desiring faculty.

Mill also supports his appeal to desire by what is a pragmatic argument:

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.14

Nothing about the logic of recommending an end of conduct prevents any end whatsoever from being recommended, but only one based on actual desires will be convincing. This is one respect in which the “proof” is not a proof in the ordinary sense. There is no necessity about accepting desire as the sole evidence for desirability. It is logically possible that ends of conduct that are not in fact desired be recommended without contradiction. The force of the appeal to what is desired is only to convince, not logically to rule out all other possibilities.

The import of premise (1), however, is primarily negative. It denies the existence of an intellectual intuition of the normative ends of conduct. The faculty to which Mill appeals is one that takes “cognizance” of practical ends, but by way of feeling or “sensibility” rather than intellect or moral sense. He is denying that we intuit what is intrinsically a good in some directly cognitive way.

The only way to argue a negative claim such as this conclusively would be to take each putative intuition and examine it critically to try to show that it can be reduced to desire or else to absurdity. I can obviously not do this to defend Mill's proof. I can only state that I find unconvincing all claims to intuit values, independent of desires (or likes and dislikes); so I find his skeptical starting point a plausible one. However, the desires we do have provide practical ends that will be pursued unless frustrated by the pursuit of the ends of other desires. This provides an arena in which practical reason can seek to bring order out of disorder by analyzing desires to determine which, if any, are illusory; which, if any, are fundamental; what, if anything, is the common object of them all. It is this last question that Mill's psychological hedonism claims to answer.


Mill's argument for (4) has two parts. One is in Paragraph 11, which classifies as mere habit those ends of conduct that are sought neither as means to happiness nor as a part of happiness. He claims that they have become ends of conduct:

Will is the child of desire, and passes out of the domination of its parent only to come under that of habit. That which is the result of habit affords no presumption of being intrinsically good.15

That argument leaves him with everything that affords any presumption of being intrinsically good an object of conscious desire. I think it is unnecessary to accept Mill's Associationist accounts of all habitual conduct or his account of all nondeliberative nondesiring willed behavior as habit. It is enough if actions that are not the result of deliberation and conscious desire afford no presumption that their ends help to identify for us what is desirable. Evolutionary ethicists, natural law theorists, and many others would perhaps deny this, but I find it a plausible skepticism.

The other part of his argument is to claim that every object of conscious desire is associated with pleasure or the absence of pain, either as a means or an end. Many desires are acquired, such as the desire for virtue or for the possession of money, and have come to be desired through the mechanism of their association with pleasure or the absence of pain. Whether acquired or not, the ultimate ends of the desires can be regarded as experiences or states of affairs with a pleasure component: They are pleasures or “parts of happiness.” Although they may fall under various other descriptions, it is the fact that they are ingredients of happiness that provides a common denominator and a unified account of desire.

If Mill is correct that there is a pleasure component to the ultimate end of every desire, and if no other common denominator provides a unified account of desire, then Mill has a persuasive claim that it is the pleasure component (i.e., being part of happiness) that is the element in all objects of desire that makes them desirable as ends, which recommends them to the intellect as the basis for action and which can provide for critical assessment in case of conflict between desires.

It is tempting to read into Mill the claim that it is the agreeable quality of the state of consciousness desired that is the real object of desire. Just as the sense-data theorist claims that one sees only sense data, although it is palpable that he sees things that, in common languages, are decidedly distinguished from sense data (i.e., from whatever common language words are sense data words—“sights,” “sounds,” “appearances,” or whatever), so Mill might be thought to hold that one desires only the pleasure component of desired experiences, although it is palpable that people “do desire things, which, in common language, are decidedly distinguished from happiness.”16 I think this is impossible to reconcile with Mill's talk of the objects of desire as being “music,” “health,” “virtue,” “power,” “fame,” “possession of money,” not just the agreeable feeling attending these. Furthermore, he does not need to make such a strong claim. He only needs to claim that, as a psychological fact, music, health, etc., would not be desired if no pleasure or freedom from pain or past association with these were connected with music, health, etc. Desire is evidence of desirability, but it does not confer desirability. This is obvious in the case of things desired as means. On reflection, it is obvious in the case of things desired as ends. The possession of money is desired by the miser. This desire does not make the possession of money a normative object of action for a reasonable person. The evidence furnished by desire must be analyzed; it is only by analysis of the fact that the miser desires the possession of money as part of his happiness—that he would be made happy by its possession—that the evidence of desire fits into a comprehensive theory. It is this comprehensive theory that identifies the pleasure inherent in desirable things as what makes them desirable. The pleasure inherent in them does not itself have to be discriminated as the object of desire.

Some commentators also have thought that Mill reduces the relation between desire and pleasures to a trivial one in the passage where he says that:

desiring a thing and finding it pleasant, aversion to it and thinking of it as painful, are phenomena entirely inseparable, or rather two parts of the same phenomenon—in strictness of language, two different modes of naming the same psychological fact; that to think of an object as desirable (unless for the sake of its consequences), and to think of it as pleasant, are one and the same thing; and that to desire anything, except in proportion as the idea of it is pleasant, is a physical and metaphysical impossibility.17

This statement is certainly puzzling to the twentieth-century reader, but in context Mill is asking the reader to engage in “practiced self-consciousness and self-observation.” If the terms were reducible to one another independent of observation, it is hard to see why he would invite one to attempt what appears to be an empirical discovery. A clue to interpretation is that “metaphysical” means approximately “psychological” for him.18 In his notes on his father's Analysis of the Phenomena of the Human Mind, Mill takes issue with his father's statement, “The term ‘Idea of a pleasure’ expresses precisely the same thing as the term Desire. It does so by the very import of the words.”19 J. S. Mill says that desire:

is more than the idea of pleasure desired, being, in truth, the initiatory state of Will. In what we call Desire there is, I think, always included a positive stimulation to action.20

According to J. S. Mill, then, a distinction is to be made between desiring a thing and thinking of it as pleasant. Desire is psychologically more complex and conceivably could have an object not thought of as pleasurable. It is obvious that it may have a more inclusive object, as is the case in desiring the means to an end when the means are unpleasant. In any case the question is a psychological, not a linguistic one.

Mill's substantive claim is that desire and pleasure (or the avoidance of pain) are psychologically inseparable. If this is true, two things follow: First, (4) is established—each person desires nothing that is not either a part of his happiness or a means of his happiness; second, since attainment of pleasure and avoidance of pain are the common denominators of desire, the evidence of desire supports the theory that it is the pleasure and pain aspects of the objects of desire and aversion that make them desirable and undesirable and that should serve as the criteria of good and bad consequences in a normative theory of conduct.

An adequate defense of Mill's position would require a more thorough analysis of desire and of pleasure and pain, happiness and unhappiness. I believe, however, that the interpretation given above shows that the position does not completely lack plausibility and that it might be supported by such a refined analysis.


If the intellect is convinced that the person's happiness is a good and the sole good to that person, does it follow that it will be convinced that the general happiness is a good and the sole good to the aggregate of all persons? Mill presumably thinks this is obvious, simply asserting it without argument. He apparently thinks he has practically established (3B) when he has established (3A), and (5B) when he has established (5A). I think Mill has been misinterpreted in this argument because commentators have thought his conclusion to be a much stronger claim than it is. He is making a very weak claim, which is seen when we notice what he means by “the general happiness.”

According to Mill “the general happiness” is a mere sum of instances of individual happiness. Just as personal happiness is not a “collective something” but simply a sum of pleasures,21 so we can take Mill as holding that the general happiness is simply a sum of individual pleasures. There are still two ways of understanding the argument. One is that “to that person” represents the point of view of the agent when he is making prudential decisions; “to the aggregate” represents the point of view of the benevolent man when he is acting morally. Some things can be said in support of this interpretation, but I do not think that it is the correct one. I think rather that he believes that his analysis of desire shows that happiness is the kind of thing that constitutes intrinsic welfare, wherever it occurs. All instances of happiness will be parts of the personal welfare of someone, that is, “a good to someone,” but being instances of happiness, they have a common denominator that makes them the same kind of thing wherever they occur—whether in different experiences of a given individual or in the experiences of different individuals. Moreover, Mill assumes that the value of different instances of happiness can be thought of as summed up to generate a larger good. These assumptions are explicit in a letter Mill wrote regarding the move from (3A) to (3B):

As to the sentence you quote from my Utilitarianism, when I said the general happiness is a good to the aggregate of all persons I did not mean that every human being's happiness is a good to every other human being, though I think in a good state of society and education it would be so. I merely meant in this particular sentence to argue that since A's happiness is a good, B's a good, C's a good, etc., the sum of all these goods must be a good.22

His assumptions are even more explicit in a footnote to Chapter V of Utilitarianism. There, answering the objection that the principle of utility presupposes the anterior principle that everybody has an equal right to happiness, Mill says:

It may be more correctly described as supposing that equal amounts of happiness are equally desirable, whether felt by the same or by different persons. This, however, is not a presupposition, not a premise needful to support the principle of utility, but the very principle itself; … If there is an anterior principle implied, it can be no other than this—that the truths of arithmetic are applicable to the valuation of happiness, as of all other measurable quantities.23

It seems clear, then, that the “to each person” in (3A) and (5A) does not represent a “point of view,” but simply the location or embodiment of welfare that cannot exist without location or embodiment, and the “to the aggregate of all persons” in (3B) and (5B) refers to the location or embodiment of welfare in a group of individuals, not a point of view. A good to the aggregate of A, B, C, etc., is interpreted by Mill to be a sum of goods to A, plus goods to B, plus goods to C, etc. He assumes both that happiness is arithmetical, capable of being summed up to find a total, “general” happiness, and that goods to different people are arithmetical, capable of being summed up to find a total good “to the aggregate of all persons.”

With these assumptions, (3B) does follow from (3A), for to say that the general happiness is a good to the aggregate of all persons is merely to say that A's happiness, plus B's happiness, plus C's happiness, etc., constitutes a good to A, plus a good to B, plus a good to C, etc. And (5B) follows from (5A). If nothing is a good to each person but insofar as it is a part of his happiness (or a means to it), then nothing will be part of the sum of goods to A, plus goods to B, plus goods to C, etc., but insofar as it is a part of the happiness of A or part of the happiness of B, etc., or a means to these. This interpretation explains why Mill did not bother to state (5A) and (5B) explicitly and why he passed from (3A) to (3B) in one sentence. The evidence of desire shows that happiness is the kind of thing desirable as an end. It is not a different kind of thing when it is located in A's experience from what it is when it is located in B's experience. Thus, whether or not any single individual desires the general happiness, if each of its parts is shown to be desirable by the evidence of desire, because of the kind of thing each part is, then the sum of these parts will be desirable because it is simply a summation of instances of the same kind of thing.24 Given this interpretation the “utilitarian doctrine” represented by (6), is perhaps better stated by making clear that Mill believes happiness, wherever it occurs, is what is desirable as an end. This could be restated with the following reinterpretation:

(6) Happiness is [the kind of thing which is] desirable, and the only [kind of] thing desirable, as an end, all other things being desirable only as means to that end.

From this, the connection with morality is said to follow:

(7) The promotion of [happiness] is the test by which to judge of all conduct from whence it necessarily follows that it must be the criterion of morality, since a part is included in the whole.25

If my earlier elucidation of Mill's argument for happiness as the kind of thing that makes the objects of desires desirable was convincing, then the argument has some plausibility. Desire does not confer desirability; it is evidence for what kind of thing constitutes welfare. Thus, that one desires only his own happiness does not restrict the desirability of happiness to one's own happiness. If the desirability of happiness as such is identified (and not created) by one's own desire for it in one's own experience, its desirability—wherever it is located—can be admitted by the intellect.

That the value of different instances of happiness is arithmetical is certainly controversial, but not, I think, indefensible. Without an operational definition for measurement, it is difficult to know how happy two different individuals are, but it seems plausible that if two people do happen to be equally happy, then twice as much happiness exists. Mill recognizes the difficulty in determining how happy a person is. He thinks Bentham's measures of intensity and duration are inadequate to capture the complex hedonic dimensions of experience, asserting that the only test of the comparative pleasure of two experiences is the unbiased preference of those who have experienced both. This is not a direct measurement of felt experience, since the experiences are seldom if ever simultaneous. It is a judgment based on memory. In making interpersonal comparisons, it is even less reliable, since one can assume only a rough equality of sensibility between persons or make a rough estimate of difference in case evidence based on behavior or physiology shows a basis for difference. Thus summations of instances of happiness will be imprecise, but we do make judgments that one course of action will make oneself or another person more or less happy. These are not meaningless judgments; even if only rough estimates, they assume (and, I think, justifiably) that different instances of happiness are commensurable.

That the general happiness is simply the sum of the happiness of all individuals and that the good to the aggregate of all is simply the sum of the goods to each is, like Mill's methodological principle, primarily negative in its import. He is denying that there is any happiness or any value that cannot be analyzed without remainder as the happiness or the good of some individual or individuals. To prove this would require refutation of every claim to such an irreducible social good. So, again, I simply assert that I find his skepticism plausible.

If Mill's proof is plausible, as I have suggested, it does not follow that anyone will act on it. The intellect may be convinced that it is plausible or even that it is correct, without one thereby being moved to conduct his life in such a way as to maximize his own happiness or thereby being moved to identify the general happiness with his own and become a practicing utilitarian. That, according to Mill, requires a good state of society and education. But convincing the intellect may be a first step.


  1. John Stuart Mill, Utilitarianism, Chapter II, Paragraph 2. Utilitarianism, published in 1861, is reprinted in Collected Works of John Stuart Mill, Volume X: Essays on Ethics, Religion and Society, ed. J. W. Robson (Toronto: University of Toronto Press, 1969), pp. 203-259, and in various other editions. References will be to chapter and paragraph. Unless otherwise noted, references will be to paragraphs of Chapter IV.

  2. Chapter I, Paragraph 5.

  3. “Introduction” to Mill's Ethical Writings, edited with an introduction by J. B. Schneewind (London: Collier-Macmillan Ltd., New York: Collier Books, 1965), p. 31. In the years following 1965 when Schneewind wrote this, the frequency of essays on the topic even increased.

  4. Chapter I, Paragraph 5.

  5. Paragraph 11. (Emphasis added).

  6. Book VI, Chapter XII. For an analysis of these complex views, see D. P. Dryer, “Mill's Utilitarianism,” in Collected Works of John Stuart Mill, Volume X: Essays on Ethics, Religion and Society, pp. lxiii-cxiii, especially xcv-cxiii, and David Lyons, “Mill's Theory of Morality,” Nous 10 (1976): 101-120.

  7. In her essay, “Mill's Theory of Value,” Theoria 36 (1970): 100-115, Dorothy Mitchell makes a distinction between “desirable” and “good” based on an analysis of the use of “desirable” in contexts of ordinary language. I think she is correct that they are not synonyms in English, but I think, nevertheless, that Mill is using them as such in this essay.

  8. For a discussion of this point, see H. R. West, “Reconstructing Mill's ‘Proof’ of the Principle of Utility,” Mind 81 (1972): 256-257.

  9. This part of Mill's psychological doctrine is stated explicitly in his essay on “Whewell's Moral Philosophy.” He quotes Whewell as saying that “we cannot desire anything else unless by identifying it with our happiness.” To this Mill says he should have nothing to object, “if by identification was meant that what we desire unselfishly must first, by a mental process, become an actual part of what we seek as our own happiness; that the good of others becomes our pleasure because we have learnt to find pleasure in it; this is, we think, the true philosophical account of the matter.” (“Whewell's Moral Philosophy,” Collected Works, Volume X: Essays on Ethics, Religion and Society, p. 184 note; Schneewind, ed., Mill's Ethical Writings, p. 192 note.)

  10. George Edward Moore, Principia Ethics (Cambridge: Cambridge University Press, 1948 [first edition 1903]), p. 66.

  11. Moore's interpretation of Mill as committing the “naturalistic fallacy” is analyzed and refuted by E. W. Hall in “The ‘Proof’ of Utility in Bentham and Mill,” Ethics 61 (1950-51): 66-68. R. F. Atkinson in “J. S. Mill's ‘Proof’ of the Principle of Utility,” Philosophy 32 (1957): 158-167, calls attention to the continuing difficulty presented by a footnote in Chapter V where Mill says “… for what is the principle of utility, if it be not that ‘happiness’ and ‘desirable’ are synonymous terms.” This is a puzzling use of “synonymous” but I do not think that it has to be interpreted (absurdly) as claiming that “Happiness is desirable” is a tautology. He may simply mean that the two terms are applicable to the same phenomena, one descriptively, the other normatively.

  12. This is found in Book VI, Chapter XII, Section VI of the Logic where he says that a first principle of an Art (including the Art of Life, which embodies the first principles of all conduct) enunciates the object aimed at and affirms it to be a desirable object. It does not assert that anything is, but enjoins that something should be. “A proposition of which the predicate is expressed by the words ought or should be, is generically different from one which is expressed by is, or will be.

  13. Another of Moore's charges, in Principia, p. 67.

  14. Paragraph 3.

  15. Paragraph 11.

  16. Paragraph 4, J. S. Mill's father, James Mill, apparently did hold such a view: “… we have a desire for water to drink, for fire to warm us, and so on.” But, “… it is not the water we desire, but the pleasure of drinking; not the fire we desire, but the pleasure of warmth.” (James Mill, Analysis of the Phenomena of the Human Mind, 2nd edition, ed. John Stuart Mill, Chapter XIX.)

  17. Paragraph 10.

  18. For example, he says that “the peculiar character of what we term moral feelings is not a question of ethics but of metaphysics.” (“Whewell's Moral Philosophy,” p. 185.) This interpretation of the term “metaphysical” is argued forcefully in M. Mandelbaum, “On Interpreting Mill's Utilitarianism,” Journal of the History of Philosophy 6 (1968): 39.

  19. James Mill, Analysis of the Phenomena of the Human Mind, 2nd edition, Chapter XIX. The passage continues: “The idea of a pleasure, is the idea of something good to have. But what is desire, other than the idea of something good to have; good to have, being really nothing but desirable to have? The term, therefore, ‘idea of pleasure,’ and ‘desire,’ are but two names; the thing named, the state of consciousness, is one and the same.”

  20. James Mill, Analysis of the Phenomena of the Human Mind, 2nd edition, Volume II, Note 36.

  21. Paragraph 5 and Paragraph 6.

  22. The Letters of John Stuart Mill, 2 volumes, ed. H. S. R. Elliot (London: Longmans, Green, and Co., 1910), Vol. 2, p. 116; Collected Works of John Stuart Mill, Vol. XVI: The Later Letters of John Stuart Mill 1819-1873, edited by Francis E. Mineka and Dwight N. Lindley (Toronto: University of Toronto Press, 1972), p. 1414. Quoted in Schneewind (ed.), Mill's Ethical Writings, p. 339.

  23. Chapter V, the second from the last paragraph of the chapter.

  24. John Marshall, in “The Proof of Utility and Equity in Mill's Utilitarianism,” Canadian Journal of Philosophy 3 (1973-74): 13-26, especially p. 16, points out the ambiguity in the question, “What is desirable as an end?” It can be taken to ask “What kind of thing?” or “What specific thing?” He interprets Mill as thinking more in terms of the first question in arguing that each person's happiness is desirable. I am claiming that Mill's proof is concerned only with the first question and believe that Marshall is reading in too much in finding a proof of equity as well. For a further statement of Marshall's position, see “Egalitarianism and the General Happiness” in The Limits of Utilitarianism, ed. by Harlan B. Miller and William H. Williams (Minneapolis: University of Minnesota Press, 1982).

  25. Paragraph 9.


Representative Works


Criticism: Jeremy Bentham's Utilitarianism: The Science Of Happiness