When Mill published Utilitarianism as a book in 1863, hoping to further the acceptance of utilitarian morality, he was well aware of the doubts about whether people would care enough about morality under the utilitarian conception. He writes:

Mill identifies a concern here not with the moral norms espoused by utilitarianism (the importance of increasing pleasure, minimizing pain, and so on) but rather with the utilitarian view of where morality comes from, i.e. human sentiment. What effect would, and should, widespread cultural acceptance of this view of morality's origins have on moral commitment? If morality is not (in some sense) objective, does it follow that morality is therefore less important than it is typically and traditionally taken to be? Mill wants to reply, as the allusion to Kant in the opening quotation suggests, to those who worry about deriving morality's authority from the contingencies of human psychology. His aim is to repudiate, in particular, the suggestion that morality is in any sense less important if it is understood as rooted in human emotion.

Mill's discussion of these matters, as Roger Crisp among others has noted, has suffered neglect compared to the other chapters of Utilitarianism, notwithstanding the fairly prominent role Mill's discussion of these issues plays in Christine Korsgaard's The Sources of Normativity.Footnote ² However, philosophical discussions of the genealogy of morality, especially naturalistic accounts of the origins of morality, are now generating the kind of philosophical attention and excitement that they have not experienced for a long time. Furthermore, Mill is addressing a concern that continues to draw widespread support in the world today, namely that a secular and naturalistic conception of morality will somehow diminish morality's authority and that people will behave in worse ways as a result. It is illuminating and instructive to read chapter 3 in light of these contemporary interests and concerns.

I. SENTIMENTALISM AND MORAL AUTHORITY

Mill's aim in chapter 3 of Utilitarianism is to combat the idea that if morality is ‘merely’ a matter of feeling it follows that morality will lose its ‘binding force’.Footnote ³ Mill does not neatly distinguish whether the concern is that utilitarianism will undermine the motivational force or the normative force of moral claims: he tends, here at least, to associate the two senses of force. The close association, while perhaps not helpful for certain purposes,Footnote ⁴ makes considerable sense in this context as Mill's concern is morality's authority. If moral claims have authority then this certainly involves more than the idea that they are reliably followed: the fact that people are moved by a claim does not ipso facto legitimate it. But, on the other hand, it is tempting to think that motivating power is a necessary if not sufficient condition of morality having authority. To adapt some words of John Rawls, a conception of morality ‘is seriously defective if the principles of moral psychology are such that it fails to engender in human beings the requisite desire to act upon it’.Footnote ⁵ We can think of this as the demand that an adequate conception of moral norms accord with what has been dubbed the ‘practicality’ of moral judgment.Footnote ⁶

Mill's subject is the psychological or motivational implications of an adequate conception of moral authority. His concern, after all, is with the idea that if people come to believe that morality is rooted in human sentiment then they will feel less bound by moral obligation on those occasions when morality dictates something other than what they want. To adapt one of Mill's pointed questions: ‘If my own happiness lies in something else, why may I not give that the preference?’Footnote ⁷ And this has been the nagging worry about sentimentalist conceptions of moral authority: that they cannot adequately capture the sense in which one is beholden to do the morally right thing even when one is inclined to do otherwise. This is why there is a longstanding association of moral sentimentalism with moral skepticism.Footnote ⁸

Mill wants to show how a sentimentalist conception of morality does not undermine the binding power of moral duty. This does not mean that he has to provide an account of moral authority that gives moral demands such strong motivational power that they can never be resisted: the ‘magnetism’ of morality doesn't, as we know, attract everyone at all times. But he must show that taking on a sentimentalist account of the sources of normativity (to adopt Christine Korsgaard's nice phrase) would not have the effect of undermining people's commitment to morality. The moral skeptic, as Korsgaard puts it, thinks that once we see ‘what is really behind morality, we won't care about it any more’Footnote ⁹ and what this suggests is that a non-skeptical conception of morality will not leave us cold to its claims. That is to say, satisfactory conceptions of morality will not undermine the sense of morality's importance: indeed, they must presumably preserve something at least close to the general, pre-reflective, sense of the importance of morality. The question is: can Mill's account do this?

II. AUTHORITY: PSYCHOLOGICAL NOT ONTOLOGICAL

Mill argues that for any individual, while there are a variety of possible sources of moral motivation external to the mind, what ultimately moves someone to act morally is a matter of feeling. He allows that moral motivation can derive from what he calls external sanctions but, as Roger Crisp rightly stresses, external sanctions ultimately consist in internal feelings: ‘the hope of favour from others, fear of their anger, and sympathy for them’.Footnote ¹⁰ Mill calls unmediated inner concern the ‘internal sanction’ of morality and identifies it with conscience, which he in turn identifies with ‘a pain, more or less intense, attendant on violation of duty’.Footnote ¹¹ The important point to stress here is that, for Mill, the distinction between internal and external sanctions is a distinction that takes its place within the larger view that all moral motivation is rooted in sentiment.

This last point needs emphasis because the claim that the main motive to morality comes from feelings and not from beliefs is at the heart of Mill's answer to worries about a sentimentalist conception of moral authority. He writes: ‘whatever a person's opinion may be on this point of Ontology [regarding the metaphysical nature of moral facts], the force he is really urged by is his own subjective feeling, and is exactly measured by its strength’.Footnote ¹² This is why, Mill claims, there is no reason to fear that a change in belief in the source of moral authority (such as from transcendental facts to human feelings) will have a drastic impact on moral behavior. His answer to the worry is very Humean: metaphysical beliefs are too motivationally inert to be the crucial element in explaining and understanding our motivation to be moral. It is the nature of our feelings that is of central importance.

The problem with Mill's answer is that it begs the question. The question at issue is to what extent the inner feeling of being bound by morality depends for its existence on the belief that certain external facts obtain (that there are moral norms, for example, that are not the artifact of human sentiments). To assert, in this context, that the moral feelings would be there anyway (however things turn out in the external world) is to assume the claim that needed to be defended.Footnote ¹³ Mill, to his credit, realizes this and acknowledges as much:

Mill expresses the worry well and, indeed, one would think that he would be well-placed to see it since his own idea of external sanctions depends on the idea that inner feelings can depend on matters external to the mind. The particular concern here is whether the sentimentalist can deliver an adequate conception of moral constraint. More specifically, as Mill observes, the concern is that if people come to a conception of the binding power of morality as self-imposed, doesn't this invite the thought that one can, as it were, unbind oneself the moment the constraints prove to be a nuisance?Footnote ¹⁵

Mill understands the problem but his response is somewhat disappointing. He simply points out that people can and do ask ‘Need I obey my conscience?’ regardless of their view about the origins of moral authority. Mill is right, of course, to stress how anyone can come to question whether he or she should act morally on one occasion or another: furthermore, we know that people act against their conscience regardless of their view (or absence of a view) regarding the metaphysical origins of morality. But, even if this is granted, one can still worry that sentimentalist theories leave moral authority in particularly bad shape,Footnote ¹⁶ and that if moral constraints are merely a matter of feeling then they are going to be weak in their binding power and especially subject to arbitrary or whimsical change. Hence the worry that, on the sentimentalist picture, if morality becomes inconvenient, it is likely to be dropped.

The underlying challenge for Mill is to explain in more detail how acceptance of the sentimentalist account of moral authority would not (and should not) have a debunking effect on the authority of moral demands. Perhaps many who espouse sentimentalist theories continue to live in a way that shows a vivid respect for the force of moral concerns (as Mill did himself): the question is whether the moral experience and action of such people simply outstrips the explanatory resources of their officially espoused view. Thus the fact that sentimentalists appear to be capable of living their sentimentalism is, an anti-sentimentalist might claim, a misperception. This also suggests a corollary thesis: that if moral sentimentalists really were able to integrate their espoused moral theory into their lived experience they would not be able to experience the authority of moral conscience as they currently do. Mill recognizes this concern about sentimentalism but, at least in setting out his first reply, says very little to counter the allegation that sentimentalism would subvert moral authority by undermining the experience of binding moral constraint.Footnote ¹⁷

III. EXTERNALITY AND AUTHORITY

It is tempting, as Mill does, and as I have done in elaborating Mill, to identify the problem of moral authority for the sentimentalist as fundamentally about the lack of external constraint on the person tempted to do something wrong. This is the language that J. L. Mackie uses when discussing concerns about the potentially skeptical effect of sentimentalist conceptions of ethics:

And Crispin Wright, likewise, identifies worries about the psychological upshot of non-realist conceptions of ethics with the loss of an external sanction for morality:

However, two points are in order here. First, we should note that for many people the focus on the importance of an external source for morality comes from the view that moral norms are not binding (or, at any rate, are greatly diminished in their authority) if they lack an external source of enforcement. Mill agrees with this, and grants the need for society to establish a system of rewards and punishments in order to establish incentives to moral behavior for those whose conscience is lacking. These are the external sanctions, and Mill argues, plausibly enough, that there is no reason why a sentimentalist account of morality cannot make use of them (although there is a significant question regarding religious sanctions, as we shall see below). However, an adequate conception of moral authority, and Mill clearly understands this, needs to make sense of the internal commitment to moral concerns that goes beyond compliance for the sake of carrots and sticks. This is the challenge he is taking on, and it takes us to my second point.

One danger when discussing moral norms, as Kantians have stressed, is that of conflating externality and objectivity. One can think that there are fixed and objective moral constraints – constraints that cannot just be wished away – but that these moral constraints are internal or found inside the self.Footnote ²⁰ We should therefore be wary of identifying the notion of binding norms with externality and lack of binding norms with internality. Mill and Kant are actually in agreement here: what needs to be offered is an account of the importance of morality that makes sense of why a person should feel bound by moral norms not just from without but from within.

IV. METAPHYSICS AND THE IMPORTANCE OF MORALITY

Mill's description of the sentimentalist weaseling his or her way out of the constraints of moral norms (‘This which is restraining me, and which is called my conscience, is only a feeling in my own mind’) of course brings to mind Hamlet's famous line that ‘there is nothing either good or bad, but thinking makes it so’.Footnote ²¹ The underlying insinuation of that line seems to be that if value originates in the human mind then there cannot really be anything that is truly good or bad. But the concern of Mill's with which we opened, remember, was not that the sentimentalist cannot provide an account of moral obligation, but rather that the conception of moral obligation which the sentimentalist offers somehow diminishes its importance. This is a not a worry about the reality of morality but about the preservation of its deliberative priority and importance.Footnote ²²

Why worry that a sentimentalist account of morality will undermine the kind of importance and priority traditionally accorded to moral concerns? After all, if a metaphysically deflationary account is given of moral values, then presumably the same kind of deflationary account will be given for other forms of value too (aesthetic, prudential, and so on) – and if all values are, as it were, deflated together, then why aren't their rankings in terms of importance vis-à-vis one another the same as they were prior to the deflation? If this is so, then although there might be an understandable sense of loss for the person who has come to a ‘disenchanted’ conception of the world, why should sentimentalism affect morality's importance in respect to other concerns?

Part of the answer, as Mill's quotation indicates, lies in the way the priority of moral concerns vis-à-vis other concerns has, at least traditionally, been precisely connected with the idea that moral values have a greater metaphysical depth than other values (most obviously in the form of religious cosmologies). Thus if the deeper metaphysical status of morality adds to morality's authority over more metaphysically ‘shallow’ concerns (such as selfish pleasure or personal advancement), then metaphysical deflation threatens to affect moral value more than other values. Think about the case of motivation for reasons of self-interest. It is hard, one might think, to see how people's tendency to be moved by their self-interest would be changed by a metaphysically deflationary account of self-interest: this is something that will be pursued anyway.Footnote ²³ In the case of morality, however, the concern is precisely that morality's ‘metaphysical gravitas’ is doing normative and motivational work.Footnote ²⁴ This explains the concern that the metaphysical deflation of value does not affect all values equally – the worry is that since moral value was more metaphysically inflated than other values to begin with it stands to suffer more from deflation.Footnote ²⁵

This does not yet provide an explanation of what it means for moral values to have greater metaphysical depth, and how exactly (if at all) metaphysical depth adds to the motivational or normative power of moral reasons. Is it supposed be just a brute fact that, say, a realist account of the ontology of moral facts has normative significance? How does the metaphysical constitution of moral reasons (say, natural versus non-natural) by itself do anything?

Perhaps the idea is that certain kinds of metaphysical picture are more inspiring, as some have found claims to the effect that we are created to be moral. Another alternative is that certain metaphysical pictures provide greater self-interested reason to be moral, as in claims to the effect that bliss unfailingly awaits the righteous. This last kind of claim is obviously an aspect of some religious conceptions of moral authority: God is always present to enforce the moral law, and to provide powerful motivation to humans to choose the morally right path. It is significant that both kinds of picture (obviously, the two are often combined) provide reasons to be moral that are free of contingency in the sense that (at least within the terms of the conception) one always has overriding reason to think of morality as in one's self-interest and/or of central importance to one's life.Footnote ²⁶

However, notice that both pictures provide sources of moral motivation that are very indirect and are closer to what Mill had in mind by external sanctions. Mill, regrettably, does not directly engage with the worry that morality's authority will be diminished if it loses the metaphysical resources offered by the external sanctions of religion. Indeed, as mentioned above, Mill argues that utilitarianism can help itself to all the external sanctions for morality, including religious sanctions.Footnote ²⁷ Mill suggests that there is no reason why ‘hope of favour and the fear of displeasure from our fellow creatures or from the Ruler of the Universe . . . should not attach themselves to the utilitarian morality . . . as powerfully as to any other’.Footnote ²⁸

Mill's phrasing here is rather slippery (indeed, the whole paragraph from which this sentence is drawn needs to be read closely), and it is worth noticing that Mill is not exactly saying that his moral theory is compatible with religious belief. After all, he is a moral naturalist, is he not? What he says is that utilitarianism can still draw on the fear of (or love of) God. But this takes us back to a now familiar dialectic: would and should the inner emotional state (fear or love) exist without the belief regarding the outer world (i.e. in God's existence, and so on)?

Mill does not say, and his discussion of the external sanctions of religion thus comes off as evasive, or disingenuous, or potentially elitist and cynical.Footnote ²⁹ He is clearly advocating a secular and naturalistic moral theory and, indeed, much of the interest in what he says in this chapter comes from the fact that he is (implicitly) defending the self-sufficiency of such a view. Hence a fuller and more satisfying answer to the skeptic about moral naturalism (and moral sentimentalism in particular) would, as Mill does not, explicitly respond to the objection that ethical thought and practice cannot get by without drawing on belief in the supernatural resources of religion.

V. SENTIMENTAL EDUCATION

Mill has a second form of reply to those who worry about whether utilitarianism will undermine the authority of morality. He observes that, whenever a conception of morality is revisionary, it is likely to appear as having less authority than the ‘customary morality’.Footnote ³⁰ He argues that this is a fact that one can expect to change over time as the new morality itself becomes part of common sense.

Mill, at least when foregrounding this reply, tends to emphasize the malleability of human nature. He suggests that humans are sufficiently flexible that there ought to be no concerns about whether they can be raised to care about morality under a new conception. Never underestimate, in other words, the power of upbringing. As Mill says: ‘there is hardly anything so absurd or so mischievous that it may not, by means of these influences [those of upbringing], be made to act on the human mind with all the authority of conscience’.Footnote ³¹ This suggests that, for Mill, the question of sustaining commitment to morality is in large part a practical, social question of inculcating and sustaining sociably minded feelings. There should be no problem in accommodating people to the idea that the authority of moral norms derives from the power of human sentiment and not from a human-transcendent source.

However, one can also see here some internal tension in Mill's view, one which derives from Mill's conflict over how much to attribute moral feeling to nature or culture. His initial remarks regarding upbringing suggest that we can inculcate desires, through use of early example and sanctions, to do more or less anything. But Mill pulls back from this radical conclusion, and suggests that, nonetheless, if the motivation to utilitarianism is not rooted in a deep aspect of human nature, then there should be serious concerns about whether it can sustain reflective commitment. He writes:

Mill needs moral sentiments to be malleable but not too malleable. He needs malleability if he is to defend his moral revisionism and, in particular, his claim that people can be educated to regard a utilitarian conception of moral norms as having all the binding power of traditional morality. But too much malleability threatens to aggravate rather than ease the worry about whether sentimentalism can provide an adequate conception of moral constraint: how can moral sentiments anchor moral norms if sentiments themselves are so easily recast?Footnote ³³ Note that this is a particularly tricky problem for Mill the liberal, for in his moral and political philosophy he wishes to celebrate the flexibility of human nature and to champion the importance of ‘experiments in living’.Footnote ³⁴ Mill qua liberal individualist and moral revisionist wants human malleability; Mill qua moral sentimentalist needs fixity, if moral norms are not going to dissolve upon reflection, and take on the authority-undermining appearance of arbitrariness.

Mill hopes to steer between this Scylla and Charybdis by emphasizing that the core sentiments of morality rest on the ‘firm foundation’ of our ‘social feelings’ towards others.Footnote ³⁵ This means that the malleability of our moral sentiments, for Mill, must always been seen against the backdrop of what is not malleable (or, at least, is limited in malleability): our deep-seated concern for others, especially when it comes to their experience of pain or pleasure. This concern with the pain or pleasure of others, in Mill's view, is always present in any conventional human morality: it is the seed from which they have all grown.Footnote ³⁶ This allows Mill to conceive of moral progress as an immanent process and he is thus able to soften the contrast between moral tradition and moral change: the envisaged progressive changes in morality emerge out of the utilitarian principle which has been there all along. Moreover, even if this concern for others is an ultimately contingent fact about human nature, it is not something which we are likely to experience as arbitrary.

VI. MILLIAN OPTIMISM

Mill's first reply to the skeptic about utilitarianism's ability to preserve the kind of importance typically and traditionally associated with morality in many ways circumvents the question by simply asserting that we just do care about it (which is to say, for Mill, that we care about the well-being of others). But, if his second reply is right, and sentiments are radically malleable, then it cannot simply be taken as a given that humans give priority to moral concerns. If human sentiments are as malleable as the second reply implies, then whether (and how much) we care about morality is up for grabs. The concern to establish morality on the firm foundation of human nature thus leads him to downplay his initial emphasis on the malleability of sentiment.

Chapter 3 does not provide a non-question-begging response to the worry that human care and concern for others will diminish without an attendant conception of metaphysical moral order to structure and reinforce these other-minded sentiments. This explains, in part, why chapter 3 is followed by a longer defense of the Humean position that subjective concern explains the existence of moral norms rather than vice versa. If such a defense is convincing, then concerns about abandoning a metaphysical conception of moral objectivity will be significantly allayed. If moral commitment does not depend on metaphysical beliefs, then removing those beliefs will leave moral motivation intact.

The word ‘intact’, however, is worth stressing, for the skeptic about Mill's sentimentalist grounding for morality need not hold (to reiterate an important point) that all care and concern for others would go away without a metaphysical grounding for moral objectivity. The skeptic need only hold that metaphysical beliefs play some significant role in structuring and reinforcing moral motivation. Bernard Williams was clearly sympathetic to this worry about the potentially deflating effects of naturalistic explanation, and this explains the skeptical turn that moral naturalism takes in his hands. Williams writes:

Part of the issue here is that Williams does not share Mill's optimistic and ‘progressive’ conception of human nature, and much of Williams' work is devoted to exposing the problems that open up for moral naturalism without such a conception.Footnote ³⁸ The general thrust of Mill's response to such worries about moral naturalism is in keeping with a recent claim of Scanlon to the effect that the ‘kind of gravitas [moral judgments] require is . . . not metaphysical but normative’.Footnote ³⁹ We could put the Millian response in rhetorical form: How could (and should) metaphysical models of moral objectivity have more normative and motivational weight than the substantial and earthly matter of people's happiness? No metaphysical authority for being concerned with others is, or should be, needed. Furthermore, no change in metaphysical beliefs is likely to stop our care for the pain or happiness of others. Utilitarianism's moral authority (or the authority of any other position for that matter) consists in the authority of the substantive answers that it gives to questions about how to live.

The substance of Utilitarianism's guidance falls under the broad heading of the value of happiness yet Mill also suggests a more specific value that underlies the value of being moral: the value of being in unity and harmony with other people. This value has its roots in human nature, and when properly cultivated by education and celebrated by society can inspire people to promote the general happiness with (as he says of Comte's view) all the ‘psychical power and social efficacy of a religion’.Footnote ⁴⁰