Utilitarianism and Slavery
By Alec Rawls © 1994/1998. To be published in a book of perspectives on slavery, I think.Warning: this piece is longer than my journalistic articles (20,000 words).

I Introduction

One of the most damning criticisms of utilitarianism has been its seeming incompatibility with any kind of indefeasible rights. Because utilitarianism fixes on the sum of utilities, it calls for freely trading off one person's happiness for another's. Any imposition of misery on a few must be acceptable under utilitarianism so long as it leads to an increase in the sum of happiness. Accordingly, it seems that utilitarianism must not be able to rule out any practice as in itself unacceptable. Utilitarianism, for instance, might call for a universally disliked person to be put to death, just for being disliked. Any argument against such an outcome would be contingent on the ways of people's pleasure and hence, it seems, not indefeasible. If this criticism were sound, it would render utilitarianism unacceptable, since there are things that in themselves surely must be ruled out -- things like slavery and murder and knowing punishment of the innocent. Further, the whole idea of basing a moral theory on mere likes and dislikes seems to be misguided. There are reasons why certain things are wrong. There are principles for valuing things that wrong actions violate. A moral view must be based upon those reasons and articulate those principles. It is not enough to tally up feelings and argue against barbaric practices on the grounds that they don't produce enough good feelings to offset the bad. If utilitarianism really does try to substitute an inarticulate tally of feelings for the variety of reasons that we are able to grasp, then it must be rejected. It would not be a moral theory, but an evasion of moral theory - an attempt to get around the need for a real understanding of right and wrong.

But these criticisms of utilitarianism only apply to what one might call Benthamite utilitarianism. As originally systematized by Bentham, utilitarianism included no theory of ends. It weighed pleasure and pain -- satisfaction and dissatisfaction -- however people felt them, with no discernment of whether the ends to be satisfied made sense or whether they were ignorant or perverse. A utilitarianism that has nothing to say about ends truly is capable of condoning anything. In the absence of any constraints on what is put into the maximization problem (the ends to be pursued), no constraints are implied on what can come out. John Stuart Mill however, did include a theory of ends in his version of utilitarianism, and this changes everything. Anything we can say about what there is to value in the world or about what matters more than what will imply restrictions on what can possibly be a part of a maximizing pursuit of ends. Further, these restrictions will turn on the reasons we can cite as to why certain things are wrong. If a restriction is violated, that means something important is not being treated in accordance with the value we can see to put on it. Utilitarianism still comes down to feelings in Mill's scheme, to the data of individual experience, but they are knowledgeable feelings that comprehend the worth of what is at stake and what a full accounting of that worth requires.

Of course a theory of ends will mostly find gray areas, where firm conclusions about what matters more than what are not available. But we don't need to be able to establish a lot of priorities to establish some very important things. Seeing the waste of human potential in slavery, for instance, is easy. What is needed is to bring the reasons involved into a systematic accounting of value, and this is what utilitarianism tries to orchestrate. Bentham analyzed one half of the problem. He posited an objective function -- the greatest happiness principle -- and that freed him to turn to the theory of means, or the optimal pursuit of ends. Mill also analyzed the other half of the problem. In addition to following the implications of reason for how best to pursue given ends (instrumental reason) he tried to array the implications of reason for following and marshalling evidence of value and arriving at ends (moral reason). My project here is to show how a complete utilitarian analysis of value, combining both the theory of means and the theory of ends, is able to deduce certain indefeasible rights. In particular, it can be used to deduce Mill's principle of liberty (that the only civilized grounds for forcing anyone to do or forbear is to prevent harm to others) and to deduce a much needed extension to Mill's principle, the right not to be enslaved (which has implications for economic liberty rights more generally).

Section II introduces Mill's theory of ends and scopes out some first conclusions that can be drawn from it. The crucial step here is to recognize Mill's criterion for distinguishing higher and lower ends as a basic principle of reason. Then the different incarnations of this criterion in Mill's philosophy can be recognized, allowing the disparate discussions where Mill developed his theory of ends to be pieced together. This is necessary because, after introducing his distinction between higher and lower ends in the essay Utilitarianism, Mill never mentioned higher or lower ends again, making it seem that he had abandoned the theory of ends, when actually he had addressed it very insightfully in several places but always in different terms and from very different vantage points. Mill succeeded in using the principle of reason embodied in his criterion of higher ends to establish the rough outlines for a viable theory of ends -- one that is able to make progress and refine itself, getting a leg up on what should ultimately be a throughly competent analysis of value. For this reason I have taken to calling this principle of reason "the fundamental principle of moral reason" and I think it is very important that it be recognized as such.

I should point out here that it is not my purpose to give an account of Mill's philosophy. The purpose is to develop the theory of ends in the most effective way possible. Mill provides guidance, and at times the exposition becomes exegetical simply because there is little here that Mill cannot be given credit for, but we must not to be limited to Mill's formulations. In particular, a modern predilection for seeing things in terms of rationality pays dividends, and at some other points Mill's insights need not only to be sharpened but corrected and augmented as well.

Section III briefly points out how Mill's arguments for his principle of liberty in On Liberty can be seen to come directly out of his theory of ends and how they can be made watertight with just a few additions. Most of the argument stands directly on the requirements of reason for how to follow and marshall evidence of value. To complete the argument for Mill's principle, these conlcusions about how to make progress in discovering value only need to be complimented by one undemanding conclusion about what values such progress will discover: that they will be not contain anti-social sentiments like envy. That is, it is the concept of higher ends does most of the work, and only a small amount needs to be said about content to secure Mill's principle.

One striking thing about about On Liberty is that, although Mill claimed that he was appealing to the principle of utility, none of his main arguments for his principle of liberty invoke the sum of utilities criterion. Rather, as section III will survey, all of Mill's arguments that are necessary to the case for liberty derive entirely from the theory of ends. That is, Mill effectively argued that every person who abides by the requirements of reason for locating value and arriving at ends will find that what he comes to value is optimally served by abiding by the principle of liberty, regardless of how anyone else behaves. Enforcement of Mill's principle is therefore consistent with utilitarianism -- if a course is maximizing for each individual then it is trivially true that it must maximize the sum of utilities -- but the case for Mill's principle does not depend on utilitarianism. It has an independent foundation, and of a particularly strong form, following for each individual as a direct consequence of abiding by the requirements of right reason in regards to the discovery and pursuit of value. People will want to abide by Mill's principle of liberty simply as a consequence of full rationality, or individual rationality in their choice of ends as well as means. So long as people are thinking straight they will embrace Mill's principle, regardless of whether they embrace utilitarianism as a public or private morality.

Unfortunately, Mill's principle of liberty only addresses cases where there are what Mill called "direct" interests on both sides of the equation (if direct interest are not at stake there is no "harm" in Mill's usage). This makes it only a first step in articulating the full ideal of liberty, and very inadequate if taken as the full ideal. There are many cases where direct interests are in conflict where we would want to assert that one interest must take priority as a matter of right. Slavery provides a central example of the type. Would be slaves and would be slave-owners both have direct interests in whether slavery is allowed. With direct interests on both sides of the equation, it becomes much harder to come to firm conclusions about how people will want to behave based only on uncontroversial deductions from the theory of ends. In particular, it may not be possible to deduce that every person who follows the principles of moral reason to a competent discovery of higher ends will want to free any slaves he might own. We may trust that a full enough discovery of what there is to value in the world will ultimately lead to this conclusion, but to arrive at it as a deduction is a very demanding objective. The problem becomes managable again, however, if we add to the theory of ends some scheme for adjudicating or combining different people's possibly conflicting direct interests. Utilitarianism fits this bill, and it is not to hard to show that even if it turns out that a fully rational slave-owner could prefer the retention of slavery, it is certain that this contribution to the attainment of higher ends is necessarily more than offset by the slaves' loss of attainment of higher ends, so that Millian utilitarianism will necessarily reject slavery. This argument is nailed down in section IV.

It is ironic that utilitarianism, which has long been branded incompatible with rights, actually turns out to be what is needed to extend the argument for Mill's principle of liberty so that it will cover a right not to be enslaved. The only thing to remember in section IV is that, unlike an argument based entirely on the theory of ends, an argument that uses the sum of utilities criterion (or the sum of attanment of higher ends criterion) does not at present have a secure foundation. Mill's attempt to provide an argument for "the greatest happiness principle" -- his "proof" of utilitarianism -- ends up reducing to an argument that everything valuable can be shoehorned into the concept of happiness, a dubious contention in its own right and certainly not to the point.[1] It is beyond the scope of this paper to go into, but I believe that a sound foundation for utilitarianism is available (for Millian utilitarianism that is, using the sum of progress in discovering and attaining Millian higher ends as a criterion for ranking states of affairs)[2]. For now I just note that this piece of the puzzle still needs to be put in place and proceed with the utilitarian argument for a right not to be enslaved. Ultimately, a complete utilitarian analysis of liberty rights should yield the full ideal of liberty. The case for a right not to be enslaved is a key part of this analysis, laying the foundation for addressing economic liberty rights in general.

In the argument against slavery there is a tie in between this paper and R. M. Hare's paper in this volume. Hare's paper considers the method of example building in utilitarian criticism and discusses one possible slavery justifying scenario. But Hare is hampered in countering the issues his example raises because his utilitarianism is essentially Benthamite, invoking no theory of ends. Without a theory of ends, Hare is able address those considerations that arise from the theory of means, but he cannot be unequivocal in his rejection of slavery because he cannot reject possible justifications for slavery based on suppositions about ends. In the last half of section IV I show how little needs to be called upon from the theory of ends to erase equivocation about Hare's example, then I go on to pose some further examples that show the theory of ends cannot be the source of any utilitarian justifications for slavery. Since slavery is also rejected by all considerations arising from the theory of means, the result is an indefeasible utilitarian right not to be enslaved.


II The theory of ends

J. S. Mill distinguished two sorts of reason. One was deduction from premises, or syllogistic reason, and the other was the "Philosophy of Evidence", or the implications of reason for how to proceed in arriving at premises and for judging the tenability of premises, both by examining the direct evidence for premises, and by testing their implications.[3] When the subject is art instead of science, Mill recognized these same two roles of reason, the only difference being that in the case of art, the "premises" are the ends to be pursued. Once ends are in place, deductive reasoning can help figure out how best to pursue those ends, but reason also has implications for how to proceed in arriving at and evaluating ends, just as it does for how to proceed in arriving at and evaluating scientific premises. These implications of reason for what one comes to value -- what we might call moral reason -- Mill placed under the heading of "Teleology, or the Doctrine of Ends."[4] This doctrine would include, not just what reason has to say about how to follow and marshal evidence, but also the evidence itself: all the evidence so far accumulated about what has value as an end. Mill then gathered this business of locating ends together with the business of pursuing ends and called it "the Art of Life".[5]

Mill's criterion for distinguishing higher and lower ends fits into this framework. What can reason tell us about how to follow and martial evidence about the value of ends? Mill's concept of higher ends asserts a preference for informed choice. "Higher" is identified with the choices a person makes as he gains more experience and understanding about what there is to value in the world.[6] There are exceptions to the rationality of a preference for informed choice, but with few qualifications it is a valid principle with wide ranging implications. Rationality does not require that information never be thrown away, because information is costly to process and retain. But in the implicit cost-benefit analyses of the mind, all information about what there is to value must itself be valued positively, and only be thrown away if the cost of processing it is greater than the expected value to be gained. The principle of not uneconomically throwing away information is familiar regarding the rational or maximizing pursuit of given ends. Knowledge about how to pursue ends enhances the pursuit of ends. When the subject is information or understanding about what there is to value, the logic is the same. The pursuit of value depends as much on information about the worth of ends as on information about how ends can be pursued. There will always be gray areas. The task of moral reason is to identify the clear violations of proper valuation, and figure out the implications of allowing or eliminating these violations.

Mill himself never explicitly stated his principle of higher ends as a principle of rationality.[7] Actually, he never elaborated ot it at all, and never even invoked it by name. Nevertheless, he did proceed in the logical way. In Considerations on Representative Government Mill observed that the concept of the public interest could be broken down into two departments: progress and order. These are really two expressions of the same principle, the principle of valuing everything there is to value. We are to husband the progress we have already made in discovering and securing valuable objects (order) and we are to follow the principles of husbandry in making further progress. Being expressions of the same principle, either could be subsumed under the other. Mill chose to subsume order under progress and let progress stand as his term for the public interest.[8] The concept of the public interest is not the only thing that can be broken down this way. Note that the principle of not wasting information (the principle behind Mill's concept of higher ends) can also be exhaustively broken down into the two departments of progress and order, or progress and conservation. To not throw away information one must first conserve (or optimally account) everything that one has already learned about the value of things in the world. That is, one must bother to account one's information about what there is to value, wherever it is worth it to do so, as best as one can tell from what one knows about where value lies. Second, one must value new information about the worth of ends and means, alerting the eye and the minds eye to pursue indications of value, again, wherever the evidence seems to indicate that this attention will be worth the opportunity cost.

Typically, Mill's discussion of progress and conservation as two halves of the public interest made no mention of progress and conservation as two halves of his criterion of higher ends. Nevertheless, the connection did not elude him. This can be seen in On Liberty where Mill states that utility is to be calculated in terms of "utility in the largest sense, grounded on the permanent interests of man as a progressive being."[9] Here Mill is stating his criterion of higher ends -- his criterion for measuring utility -- in terms of progress and conservation (substituting "permanence" for "conservation," but I take the meaning to be the same: permanent interests are those things that a full accounting of value will conserve). It is important to recognize that Mill did make this connection between higher ends and progress and conservation because it is a key to recognizing where he developed his theory of ends. Mill's most extended contribution to the theory of higher ends appears in his essays on Bentham and Coleridge where he represented these two men as flawed champions of progressivism and conservatism respectively. Both were flawed because each recognized only his own fount of value and spurned the other, thus violating the underlying principle of accounting all information and all discoveries of value. Coleridge had an eye only for traditional wisdom, while Bentham regarded tradition as nothing but an obstacle to rational analysis. What is needed is for these two different sources of insight, and indeed for all available sources of insight, to be "completed and corrected by one another."[10]

In this admonition we can see the sense in which Mill's criterion of higher ends really is "the fundamental principle of moral reason." Broken down into progress and conservation, it becomes clear that Mill's criterion of higher ends is a bootstrapping mechanism, or a scavenging mechanism. The principle of husbanding all understanding and information will gather up any other principle of right reason that is discovered and require that it be abided by. Every bit of understanding, every discovery of the value of things, all must be conserved or accounted henceforth. This is the fundamental principle of moral reason because it is the principle that husbands the rest. To the extent that one follows this principle -- as anyone can only do imperfectly -- there is a ratchet effect. We can learn better but we never ignore, discount, regress. Adherence to the fundamental principle of moral reason creates an engine of progress, learning ever better what there is to value and how to pursue it.

True, a little knowledge can sometimes be a dangerous thing, as when we learn how to make atom bombs before we learn how to live in peace, or learn how to wipe out diseases before we learn how to control population ourselves, but even here, the reason that the "progress" mankind has actually made is of such a dubious nature is because of our manifest failure to abide by the principle of conservation, as we often completely ignore the value of what we plow under, returning civilizations, forests and species to organic dust in exchange oftentimes for what is worthless in comparison. The Panglossian view of the old social Darwinists, declaring that whatever happens necessarily constitutes progress, is patently egregious. Flattening everything just to rule the field does not prove moral superiority, it proves moral imbecility, by gross violation of the principle of conservation. Unfortunately, many moderns have reacted to the social Darwinist view by going as far wrong in the other direction, somehow thinking it necessary to deny that there can be any tenable concept or criterion of progress at all. Different values are seen simply different cultural data. Using one culture's values to criticize another's is seen just as an exercise in power and oppression, as if no one ever makes any real discoveries about what there is to value that anyone else can learn from.

Mill, in contrast to right and left, offers a nice breath of sanity. All of our discoveries of value are partial truths, and to guard against the hegemony of any partial truth, it is the job of progress and conservation to see our partial truths "completed and corrected by one another." One of Mill's unanswerable principles (which the fundamental principle of moral reason requires us henceforth to acknowledge) is that "[n]obody's synthesis can be more complete than his analysis."[11] Every culture cultivates particular discoveries of value, and remains ignorant of many others. Any culture that makes assertions of value or priority beyond the scope of its learning is prone to be in error. The fact that one culture's assessments about value or priority can be at odds with another's does not mean that there can be no overarching conclusions about value. It means that one or both of the conflicting cultures is overreaching what it has the grounds to conclude, and with this overreaching corrected, there should be nothing to keep the learning of both from being collected together.[12]


So this is a propitious start. It is not every day one happens across the fundamental principle of moral reason. The next question is how to prosecute this principle. Again, Mill had a telling strategy. If higher ends correspond to greater knowledge and understanding of what there is to value, then people either make progress in arriving at higher ends by husbanding information or they fail to make progress by wasting information. Thus one key to developing a theory of ends is to locate the important insights into how people typically waste information, then note the consequences. Mill hit this nail on the head by focusing on the concept of fixation. Mill's associationist model of psychology made him regard people as creatures of habit (a view that is thoroughly supported by modern research, though Mill's occasional limitation of cognition to habit and association does not stand up).[13] A person who develops habits of reason will be a sponge for all information about what there is to value and how to pursue it. Alternatively, one may fixate on habitual concerns and only attend to information and opportunities pertinent to these.

In On Representative Government, Mill stated the case as follows:

Everyone has two kinds of interests - interests which he cares for and interests which he does not care for. Everybody has selfish and unselfish interests, and a selfish man has cultivated the habit of caring for the former, and not caring for the latter. Everyone has present and distant inter ests, and the improvident man is he who cares for the present in terests and does not care for the dis tant. It matters little that on any correct calculation the latter may be more considerable if the habits of his mind lead him to fix his thoughts and wishes solely on the former.[14]

A "correct calculation" would be one based on informed choice, where all con sid era tions were weighed as a knowledge of them would call for them to be weighed. When Mill wrote of "interests [a person] does not care for", he was not referring to mat ters of taste - that some things are not pleasurable to some people - he was referring to things that a person could value but re mains igno rant of or shuns consideration of. Thus the selfish man and the improvident man are examples of people mired in lower ends, cut off by fixation on habit ual gratifications from an ex panding awareness of what there is to value. As usual, Mill made no note of how the existence of "interests [one] does not care for" manifests uninformed choice and hence examples his concept of lower ends, but the logic is inescapable, and the mechanism of fixation becomes a psychological key to analyzing the difference between higher and lower ends.

Mill went on to offer the further example of fixation in the person of a wife beater,. His prognosis was fully pessimistic:

He would be happier if he were the kind of person who could [live in love and kindness with his wife and children], but he is not, and it is probably too late for him to become, that kind of person. Being what he is, the gratification of his love of domineering and the indulgence of his ferocious temper are to his perceptions a greater good to himself than he would be capable of deriving from the pleasure and affection of those dependent on him. He has no pleasure in their pleasure, and does not care for their affection.[15]

In Mill's view, people are their habits of mind. This does not mean people are not changeable, because habits can bring about changes both in oneself and one's circumstances, but there is a strong case to be made for a central tendency to either fall deeper into habits of fixation or to follow habits of reason to the world of higher ends. A person who develops habits of reason is secured in these habits both by the benefits of following habits of reason (in terms of improved discovery and pursuit of value) and by the fact of reason (the fact that reason makes sense, as when it descries the benefits of informed choice, so that the person who hears what reason has to say will find it compelling). On the other hand, a person who does not develop habits of reason will become blind to reason. As Mill observed: "The mental and moral, like the muscular, powers are improved only by being used."[16] Habitual interests are presented directly to the imagination by past and present experience while concern for what one has not yet learned the value of is speculative. When the speculative powers atrophy from disuse, habitual gratifications are all that is left.

From here we can make some general observations about the character of higher and lower ends. Most obviously, higher ends are going to include all kinds of discoveries about what there is to care about beyond oneself, while lower ends, since they correspond to a failure to care for and seek the worth of what is unknown, are prone to be limited to the sphere of personal concerns that come to impinge whether one seeks them out or not. By pursuing this a little further we can secure a tidbit that will be necessary seal one arm of the argument for liberty, namely the classification of envy and the anti-social pleasures as lower ends. Note that if people do not follow habits of reason to an expanding awareness of what there is to value, they will be stuck with a dearth of ends. Thorstein Veblen's Theory of the Leisure Class can be taken as a study of how a society of people with little grasp of ultimate ends will develop. Absent any developed sense of what is valuable in itself, people can only compare themselves to others, causing values of "invidious comparison" to take over. The rich compete in such activities as "conspicuous consumption" and the poor emulate on their own level.[17] Thus we see how the anti-social pleasures can arise as lower ends, spawned by the failure to follow habits of reason to a discovery of what there is to value.

The categorization of anti-social sentiments as the exclusive province of lower ends has a small but significant role to play in the argument for liberty. Showing how violations of the requirements of moral reason spawn anti-social sentiments can increase our confidence that we can make this categorization (we know the anti-social sentiments exist, so they must come from somewhere) but the crucial element will be to show how higher ends clearly renounce the anti-social sentiments. The proof of the argument for liberty is in terms of what people who follow the requirements of moral to a discovery of higher ends will embrace, so higher ends are our ultimate criterion. Still it is worth looking closely at how violations of full rationality spawn anti-social sentiments, because that will be the juncture where obedience to the requirements of reason for how to follow evidence of value reject the anti-social sentiments.

The shortfall from full rationality to be noted here is how Mill's concept of fixation can be modeled as a very explicit kind of partial rationality, cohering very closely with Mill's critique of fixation on partial truths. Fixating on what one has already learned to value and shunning other consid erations can be an instrumentally rational strategy for maximally pur suing what one has already learned to value. By dis abling the consideration of any direction that threatens to shift behav ior away from the pursuit of established values, the gratification of established val ues is protected. Of course, if enter taining other consid erations than the habitual ones did lead one to make a different choice, the implication would be that, when choice is in formed, one would rather make a different choice. Fixating on ends con stitutes an instrumentally rational pur suit of the fixated values (i.e. a maximizing pursuit of the fixated values) but at the expense of the rational pur suit of value in toto. We might call it "instrumental reason gone amok." Bernard Shaw, in his excellent essay on Ibsenism, scoffed at the the idea of moral reason and claimed that reason is always biased in this way -- that it always sets itself the goal, not of fully accounting all that it can see to account, but of making the most defensible case for one's predetermined conclusions. He liked to say that reason "cuts the sleeve to fit the arm," i.e. that it arrives at the conclusions that a person, according to her desires, sets out to arrive at.[18]

Certainly reason is often biased in this way. Irrational fixation caused by partial reason even fits many clinically observed cognitive mechanisms. Leon Festinger's Theory of Cognitive Dissonance, which has had great empirical success, is a the ory of how people restructure their beliefs to fit their behavior.[19] This is a clear example of reason "cutting the sleeve to fit the arm", or arranging its reasons to arrive at a predetermined conclusion. Other psychological phenomena -- the various forms of denial and rationalization -- also example biased reason. But while reason may indeed always have to be instrumental to some object, there is no reason that it needs to "run amok" by fixating on some particular set of values or some intermediate object. Reason can just as well be in strumental with respect to value in toto, and if a person is rational he will want to achieve the maximum of what he can care about, not just the maximum of what he has already learned to care about. The point is the same one that utilitarianism tries to clarify with its basic structure: that ends and means must be clearly separated, and each addressed logically in turn. Trying to serve one set of ends, not by deducing the means appropriate to it, but by disabling consideration of other ends, is a most obvious violation of rationality. Technically it shows up in the chucking of information, and we learn the lesson that the proper accounting of considerations requires keeping ends and means straight. (A lesson which the principle of conservation -- the fundamental principle of moral reason -- then urges us to remember henceforth, so that by this little effort we can avoid wasting large efforts being optimizing in pursuit of errant and potentially very perverse objectives.)

There will be limits to how far the human being is capable of adopting habits of reason. The theory of cognitive dissonance, after all, is a theory about our subconscious mechanisms. And there are many other typical ways, not touched on here, that people fail to rationally account and follow information. The requirements of probabilistic reason are a particularly ripe source of violation.[20] But this just says that a theory of ends is likely to be fruitful. Lower ends are going to be common and there are going to be many points where we can see how higher and lower ends will go in different directions, providing guidance for remediation.

Lets look at that guidance. What value is being misaccounted by fixation and what different direction will a full accounting of value take here? Habitual concerns are in contrast to speculative ones. Instead of reaching out, they are close to home. They are the things that seek one out whether or not one seeks them out. They are personal and at the limit self-centered, selfish. Full accounting of value, on the other hand, is a sponge for whatever worth is ever discovered in any thing, conserving it henceforth wherever it is at stake (that is, fully accounting it, and acting for it when that is the course of most value). Then you just have to account how smart people are. We have open ended faculties of intelligence, looking out for what is worth spending a life on, this life that we have no choice but to spend. Our question is how to spend it dear. And with our eyes thus open, we see things to value and act for everywhere. Of course it helps that there is in general not a conflict between self-love and love for what else there is to love, but on the contrary, being productive of value is in general the best way to prosper, and so our practical circumstances do not punish us personally for loving what there is to love outside of ourselves. That is why, when there is a conflict between self love and what else there is to value, people so often are glad for the chance to make sacrifices. Acting for most value is their mainspring. There is no hurdle for them to turn from securing their self interest to spending it because their purpose in securing life has all along been so that they can spend it, and do things with their lives.

What then can we say about anti-social sentiments? If a person abides by the requirements of moral reason for following evidence of value, and fully accounts all discoveries of value henceforth, will he take pleasure in the undeserved misfortunes of others? Can't happen. As a person discovers the range of things to care about, he comes to look upon other people, making similar progress, as compatriots. Even people stuck in lower ends become would be compatriots and one can only wish that they too should make progress in discovering and pursuing value. Competition and cooperation only stir this pot, so long as there is not injustice. But hatred of injustice is a social sentiment, not an anti-social one. Where higher ends do not leave the anti-social sentiments behind they will leave them overruled, and all can expect that people with higher ends will value each other's progress in discovering and attaining value. The yardstick of invidious comparison may never entirely leave us, but that is only because, in addition to our open ended faculties for discovering value in the world, and our instincts to value each other, there are a million ways for reason to shortcircuit, for it to become partial reason, and descend into lower ends. We cannot be completely rid of anti-social sentiments, but they are purged from higher ends.

That is all that is needed for our our argument for liberty (actually, it is overkill), but having come this far with Mill, we should have a look at what he said about the content of higher ends. His approach is wonderful. What will progress in the discovery of what there is to value reveal? Mill simply offered up his own discoveries of what there is to value. If we consider that people who fail to make progress in the discovery of what there is to value will on average secure less value and be less happy, we can interpret Mill's comments about the sources of happiness and unhappiness as his positive discoveries about the content of higher and lower ends. They can hardly be improved upon:

 To those who have neither public nor private affections, the excitements of life are much curtailed, and in any case dwindle in value as the time approaches when all selfish interests must be terminated by death; while those who leave after them objects of personal affection, and especially those who have also cultivated a fellow-feeling with the collective interests of mankind, retain as lively an interest in life on the eve of death as in the vigor of youth and health. Next to selfishness, the principal cause which makes life unsatisfactory is want of mental cultivation. A cultivated mind (I do not mean that of a philosopher, but any mind to which the fountains of knowledge have been opened, and which has been taught, in any tolerable degree, to exercise its faculties) finds sources of inexhaustible interest in all that surrounds it; in the objects of nature, the achievements of art, the imaginations of poetry, the incidents of history, the ways of mankind, past and present, and their prospects in the future.[21]

 Mill's patently correct observation is that the world is full of things to value. Our open ended faculties grasp worth and interest everywhere, and since the world outside oneself is so much larger than the sphere of one's private concerns, much of what one will find to care about will lie beyond one's private concerns, if one makes any progress at all in discovering what there is to value.

This observational method may seem like a prosaic way to pursue the lofty subject of moral reason, but it makes sense once one understands what role reason is necessarily constrained to play in the theory of ends. In Mill's "philosophy of evidence" about what there is to value, the evidence itself is all observation. Reason comes in to make sure that this evidence is properly gleaned and accounted. All of our discoveries about what there is to value in the world, or about what matters more than what, are subject to improvement and revision as we learn better what there is to value. What we can say about higher ends at any given time is simply the total of what we have so far been able to observe. We have no other grounds for saying anything about what has value. In particular, reason alone cannot tell us anything.

That is the substance of Hume's law: "Moral Distinctions not deriv'd from Reason."[22] Intuitively, Hume's idea was that no sequence of "is" syllogisms can ever lead to an "ought" conclusion, and this is unquestionably true. Fact and value are fundamentally separate notions. Mill's criterion of higher ends finds the one role for reason in arriving at ends that does not run afoul of Hume's law. Rather than attempt to derive values from reason alone, Mill's distinction between higher and lower ends sets up rules that the process of value formation must obey. The basic rules for the rational accounting and pursuing of value must not be violated. This means first and foremost that all information is to be rationally accounted and deployed. There is to be no accounting some discoveries of value while distaining others, no "instrumental reason run amok" or any of the other typical ways that people throw away information. Within these guidelines, all of one's evaluative faculties are to come into play, whether or not they employ reason. Maybe it is possible to explain ra tionally what makes a Mozart piano concerto valuable, maybe it is not. What is cer tain is that we do not have to be able to explain its value rationally in order to value it. All rationality says is that what value one discovers in it must be conserved -- that it must henceforth be accounted wherever it is at stake -- and so must all other discoveries of value must be accounted wherever one's clues about where value lies indicate they are worth accounting. Regarding the content of higher ends, this is the only form that moral reason can take.[23]

Following Mill's lead then, progress in discovering what there is to value in the world leads to the discovery of many things to value and care about beyond one's own prosperity. In particular, one comes to see the worth of other people's lives, not just to one's own prosperity, but as something one care's about beyond oneself. We see what there is to love in people, and we see how they too have their eyes open to what there is to value in the world, making us all compatriots. For the purposes of the argument for liberty, we use this steamshovel to crack one small nut: categorizing the anti-social pleasures exclusively as lower end.


[Note, added 5/98 for rawls.org web site: It may seem that I have brought the theory of ends an awfully long way here to only do so little with it. Actually, I have brought it far enough here to secure not just the argument for liberty, but the argument for utilitarianism. More precisely, what can be secured from here is the argument that anyone who properly husbands evidence of value will end up embracing utilitarianism, certainly as their concept of the public good -- of what there is to care about outside of themselves -- and ultimately as a personal criterion of how they want to live their lives. It works this way: the next conclusion of the theory of ends is that people who husband evidence of value will embrace mutuality, and the embrace of mutuality can then be seen to imply an embrace of Millian utlitarianism. (See Getting Rawlsian Moral Theory Right.) Thus the argument against slavery (and for economic liberty rights) in section IV should be understood to be fully grounded.]


III The argument for Mill's principle of liberty

The introduction to any monograph of Mill's On Liberty will outline Mill's sequence of arguments for his principle of liberty. All I want to do here is briefly indicate how Mill's main arguments can be seen to derive from his theory of ends and add what is necessary to make these arguments watertight. The conclusions arrived at here will also be invoked in making the case for a right not to be enslaved.

Mill's principle of liberty states: "That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others", where "harm" in Mill's usage is limited to harm to a person's "direct" or liberty and security interests.[24] Thus in effect Mill's principle only asserts that when one person's "indirect" or vicarious interests conflict with another person's direct interests, the indirect interests at stake are never to be given precedence as grounds for compulsion. The argument for this principle will proceed by sorting out the possible direct and indirect interests people could have in abiding by or rejecting the principle of liberty then tabulating them. The goal is to, through analysis, dispense with those hypothetical situations where to maintain the principle of liberty some vast quantity of one kind of interest would have to be overweighed by an infinitessimal quantity of some other interest, pushing arguments for priority beyond where they can reasonably be established. By tabulating what is really at stake some hypothetically conflicting interests can be determined to, on the contrary, be necessarily in alignment while others can be limited such that priorities can be clearly established.

Mill built his case for liberty by arguing first for liberty of thought and discussion and then extending the same arguments to liberty of action.[25] We can begin in the same way. With freedom of thought and discussion, people typically have a direct interest in being able to think and communicate freely, because information and information processing are critical to the discovery and pursuit of all sorts of value, including direct interests. On the other hand, interests in regulating thought and speech typically are indirect or vicarious interests in what other people are thinking or communicating. Thus freedom of thought and speech is a handy arena for sorting out direct and indirect interests. The interests at stake in this area will be seen to all be in alignment, leaving only the case of prioritization between direct and indirect interests to make the case for liberty rights in this area.

With free dom of thought and discus sion, the central issue is the value of in formed choice and the question of what pro motes it. Mill phrased the argu ment in two complementary parts:

[T]he peculiar evil of silencing the expression of an opinion is that it is robbing the human race, posterity as well as the existing generation - those who dissent from the opin ion, still more than those who hold it. If the opinion is right, they are deprived of the op portunity of exchanging error for truth; if wrong, they lose, what is almost as great a bene fit, the clearer perception and livelier impres sion of truth produced by its collision with error.[26]

As Mill noted, "these two hypotheses" -- that "[w]e can never be sure that the opinion we are endeavoring to stifle is a false opinion; and if we were sure, stifling it would still be an evil still" -- are as compelling from the point of view of the individual who would think of stifling an opinion as from the point of view of society. Thus in general, a person's own freedom of thought and discussion will be a boon to her progress in discovering what there is to value and how to pursue it, which is a direct interest. Further, the more progress other people make the more she can learn from them, giving her a direct interest also in the progress of others and hence in their freedom of thought and discussion as well as her own. A person can also have direct interests in deterring another person's freedom of speech. The person who shouts "fire" in a crowded theater is harming other people's direct interests. This and libel and other such considerations set limits to the liberty of thought and discussion that it will be in a person's direct interests to favor.[27] But except in these special circumstances, a person's direct interests in liberty of thought and discussion will necessarily be on the side of liberty.

Turning to the tabulation of indirect interests in liberty of thought and speech, one category of indirect interests would be the positive or benign interests that people take in the welfare of others. Since the rational conception of a person's welfare is his progress in discovering and pursuing higher ends, people who follow habits of reason to a discovery of higher ends will find that their benign interests in the welfare of others takes the form of an interest in what promotes other people's own progress in discovering what there is to value and how to pursue it. Thus the question becomes whether paternalism -- infringing people's liberty in order to protect them from themselves -- is a benefit or a harm to their progress.

Mill gave several impassioned arguments that paternalism is destructive of progress. One is especially interesting here because in it Mill seems to be struggling to articulate the rationality of informed choice. In answering whether there could be opinions that are so useful regard less of their truth that they should be protected from refutation, Mill wrote: "The truth of an opinion is part of its utility. If we would know whether or not it is desirable that a proposition should be believed, is it possible to ex clude the consideration of whether or not it is true?"[28] If he has difficulty articulating the rationality of informed choice, he certainly is clear on the ramifications:

No one can be a great thinker who does not recognize that as a thinker it is his first duty to follow his in tellect to whatever conclusions it may lead. Truth gains more even by the er rors of one who, with due study and preparation, thinks for himself than by the true opin ions of those who only hold them because they do not suffer themselves to think. Not that it is solely, or chiefly, to form great thinkers that freedom of thinking is required. On the contrary, it is as much and even more indis pensable to en able average hu man beings to at tain the mental stature which they are ca pable of. There have been, and may again be, great individual thinkers in a general atmosphere of mental slavery. But there never has been, nor ever will be, in that atmosphere an intellectually active people.[29]

The paternalistic argument against liberty is rejected on its own terms. Trying to protect people from error harms them, enfee bles them, fun damentally stunts the development of their faculties, on which a person's progress entirely depends. If every form of influence short of force is insufficient to convince someone where value really lies, then he must learn the hard way, by experience. Using force can only keep him from learning and hence keep him from making progress in the discovery and pursuit of higher ends.

We have then that a person's direct interests and his interests in the welfare of others both call for the protection of Mill's principle in the realm of thought and discussion.[30] When turning to liberty of action, all the same arguments about liberty and progress apply, because liberty of action is as neces sary as liberty of thought to discovering what is worth pursuing. When we are talking about actions rather than thoughts there are more opportunities for direct interests to be infringed, but this does not raise any new issues since Mill's principle only protects those activities that do not harm the direct interests of others. Some objections do take on new force however. Now stopping someone from follow ing error can mean stopping him from injur ing himself - saving him not just from contemplating folly but from living folly. But the reply is the same. If he can not learn from ad vice, the only way he can learn is from experi ence, and if what one values is progress in the discovery and satisfaction of higher ends, both for oneself and for others, this learning is what matters. Freedom and responsibility are necessary conditions for the growth of the individual upon which all progress rests.[31] Thus in general a person's direct interests and his interests in the welfare of others both call for the protection of Mill's principle of liberty.

This leaves only a few other categories of indirect interest to consider to make the argument for Mill's principle watertight. First there are any perverse or anti-social interests one might take in the welfare of others, taking pleasure in the misfortune of others rather than welcoming their progress in discovering and pursuing value. Are there any higher interests of these sorts that could weigh against the liberty of others enough to offset one's direct interest in learning from other people's "experiments in living"? The theory of ends says "no". As seen earlier, one of the first conclusions that a theory of ends is able to come to is that anti-social interests can be pretty clearly relegated to the category of lower ends. Fully rational people will find no reason here to part with Mill' s principle.

That leaves just one more tricky case: the indirect interests one can take, not in the welfare of others, but in the ways other people's actions indirectly affect one's own welfare. Experience shows that people can get extremely exercised just knowing that other people are doing what are imagined to be repulsive things. But the scope of such interests can be narrowed considerably by noticing something which Mill did not: namely, how Mill's principle of liberty gives rise to a right to privacy. To be internally consistent, Mill's principle must allow certain sorts of activities to be forced out of the public sphere, and only require that they be protected in private. This possibility of forcing activities out of the public sphere means that protection of Mill's principle will impose much less on the indirect interests of the majority than it otherwise would, leaving much less scope for those indirect interests to provide a grounds for wanting to reject the principle of liberty. Once the scope for opposition to Mill's principle is narrowed in this way, it becomes possible to relegate the remaining grounds for opposition to irrationality and lower ends.

The internal inconsistency in Mill's statement of his principle arises as follows. Consider the case of an act that does not harm anyone's direct interests, but which no one has a direct interest in being allowed to perform. Mill's principle says both that it is not justifiable to hinder the act, since harm to the direct interests of others is not as stake, and that no one should be hindered from hindering the act, since that also does not harm anyone's direct interests. This is inconsistent, and to become consistent Mill's principle must be neutral when there are only indirect interests on both sides, just as it has nothing to say when there are direct interests at stake on both sides.

There is an important class of activities that do not harm anyone's direct interests, but which no one has a direct interest in. Consider the example of public nudity. Public nudity does not harm anyone's direct interests, yet if someone wants to go nude in public we can be pretty sure that his interest is in what other people think. (If his interest is in publicity per se, then by definition his interest is in what other people think.) An interest in what other people think is an indirect, or other-regarding, interest. Thus under a consistent Millian principle of liberty, the majority would be free to decide whether people must dress in public.[32] The close connection between acting in public and being interested in what other people think means there are many actions that people will only have an indirect interest in engaging in public.[33] On the other hand, so long as an activity does not harm the direct interests of others, Mill's principle requires that scope be provided somewhere for engaging in the activity. Thus Mill's principle will protect a sphere of privacy: so long as direct harms to others are not involved, people can do what ever they want in private.

Without trying to flesh out the boundaries of what could be forced into a Millian private sphere, lets look at the self interests a rational person can take in the private behavior of others when no one's direct interests are being harmed. Any personal affliction one feels at such private behavior is self induced, since one does not have to contend with any direct stimulus even to think about the subject. To subject oneself to this affliction is just irrationality. Now it is true that human beings possess many primitive faculties that are not rational, even though they may be functional and approximate rationality in many circumstances, and this fact needs to be taken into account. Just because an affliction is irrational does not necessarily mean that it can be completely discounted. But here the object must be to manage irrational concerns, so as to minimize the irrationality. As with a child who has decided to throw a tantrum, the answer is not to give irrationality its head but to force it to come to grips with rationality. If someone is prone to self-inflicted distress at the thought of homosexuality, for instance, the way to turn this into an obsession is to let it get its hands on power. In contrast, a person who is making progress with higher ends and accepts the goal of trying to manage his irrationality will not find in his irrational afflictions a reason to want to throw anyone in jail or otherwise attack anyone else's life.[34]

That leaves only one's indirect interests in those activities of others that are guaranteed access to the public sphere by Mill's principle, meaning activities that people have a direct interest in being allowed to engage the activities in public and which do not harm anyone else's direct interests. To conclude that a fully rational person would approve of each instance where Mill's principle protects this class of public behaviors it is necessary to refer to the content of higher ends, not just the concept, but a very first observation is all that needs be invoked: that progress in the discovery of what there is to value in the world will lead to at least a modicum of generosity of spirit. To want to restrict activity here more than Mill's principle allows means putting one's own indirect interests above the direct interests of others, and not for their own good -- since it has already been determined that other people's own good is best served by Millian liberty -- and this a person with even a modicum of generosity will not want to do.[35] Since the activities still in question infringe no one else's direct interests, those engaging in them must not be harassing or forcing attentions on others.[36] They are are simply minding their own business -- their direct interests -- and they are already keeping private whatever the majority asks them to that they do not have direct interests in engaging in public. To go beyond this and also want to prohibit what these people have a direct interest in is simply not to value other people's lives. It comes to about the same thing as anti-social sentiment, which we have seen can be fairly clearly relegated to the category of lower ends. There is much to value about other people's lives and much to hope for in their progress. People who make progress in discovering what there is to value will attain the modicum of generosity of spirit necessary to not want to put their own indirect interests over the direct interests of others, within the circumscribed arena of possible conflict. QED.[37]


IV Utilitarianism and slavery

Mill's principle of liberty is just a beginning. It is clearly inadequate as a full ideal of liberty because it secures no rights when there is any conflict of direct interests at stake. It cannot even distinguish between the direct interest that a person has in possessing her own purse and the direct interest that a thief has snatching it from her. It handles a very important grey area, but takes for granted the black and white. An important task then will be to extend the argument for liberty and try to scope out what direct interests should be given priority over others as a matter of right. A right not be be enslaved provides a central example of such a priority, one that goes to the heart of many of the issues that conflicts of direct interests can raise. But the presence of direct interests on both sides will require putting the analysis on a fundamentally different footing that the argument for Mill's principle. It was tricky enough just trying to nail down the argument that a rational person's indirect interests in the behavior of others would favor other people's liberty. With direct interests on both sides it becomes much harder to say anything about how people will want to behave simply as a result of being rational about ends as well as means. If individual rationality is inconclusive, the only other hope for securing a right not to be enslaved is to invoke a scheme for tallying or adjudicating direct interests that purports to be right, not from the point of view of the individual, but from the point of view of society, or morality.

The scheme I take up here is utilitarianism, embodying Bentham's distributional dictum: "all to count as one, none as more than one." This will provide an important test for the claims of utilitarianism to be a proper moral guide. As noted at the outset, a main criticism of utilitarianism has been its seeming incompatibility with indefeasible rights. The argument for Mill's principle is compatible with utilitarianism -- if each person's higher ends are optimally served when the liberty of others is protected, then it is trivially true that the sum of attainment of higher ends is achieved by protecting liberty -- but utilitarianism only chimes in its compatibility after the argument for liberty is complete. In the case of slavery, utilitarianism could show itself, not just compatible with indefeasible rights, but necessary to their establishment. I cannot go into it here, but I believe that Millian utilitarianism can be deduced from first principles to be the ultimately correct moral standard, and that a right secured as indefeasibly necessary according to a utilitarian analysis will be a genuine right.[38] For now it is enough to show that utilitarianism (whatever foundation one can or cannot give it) yields an indefeasible right not to be enslaved.

Much of the utilitarian argument for a right not to be enslaved is Benthamite, analyzing how best to maximize the sum of utilities without bothering to discriminate between higher and lower ends. That is, a large part of the argument stems entirely from the theory of means. It is pure economics. R.M. Hare's utilitarian argument against slavery is of this sort and I will show how Hare's economic argument can be strengthened so that, regarding the issues raised by the theory of means, utilitarianism indefeasibly rejects slavery. But the Benthamite arguments about efficient means leave other issues unattended - questions of whether slavery might be advantageous for pursuing particular values which are posited to merit priority in a utilitarian calculation. Rejecting such perfectionist possibilities requires bringing in the theory of ends and showing that any ends based arguments for slavery are rejected by a utilitarianism that uses progress in the discovery of what there is to value as its criterion of higher ends. Slavery will be rejected by the theory of ends for much the same reasons that violation of Mill's liberty was rejected: because people need freedom and responsibility to progress, so that when concern is for the welfare of others, as it always is in a utilitarian calculation, that concern calls for a high standard of liberty.

The Prima facie case that slavery cannot be optimizing in respect to either ends or means is pretty obvious, since slavery makes such a colossal waste of human resources. In its historic incarnations at least, slavery has allowed little leeway or inducement for slaves to improve themselves and make what contributions they are able. Neither have slaves been allowed much leeway for discovering and pursuing value for themselves. Both of these things -- the waste of a person's worth to society, and the waste of his personal progress -- are injuries to progress. Unless slavery has some important other advantage over such alternatives as a free market in paid labor, this waste will disqualify slavery from possibly being a part of the maximal pursuit of progress. To the extent that the circumstances in which slavery cannot be maximizing are seen to be general, the result is an indefeasible right not to be enslaved.

A market system where people are free to buy and sell labor and commodities at market determined prices has, in its theoretically ideal form, some very desirable qualities. The first fundamental theorem of welfare economics states that if markets are competitive and complete and there are no market failures (most importantly, no externalities, or costs that are not reflected in market prices) then a market system will lead to an efficient allocation of goods and services, in the sense that it is impossible to make anyone better off without making someone else worse off. The second fundamental theorem states that every efficient allocation of goods and services can be attained by a market system given an ap propriate initial distribution of resources.[39] The basic idea is that of Adam Smith's "invisible hand", that when people sell their resources to the highest bidder, resources are automatically directed to their most valued uses. (Of course, the highest bidder could be oneself, if one has a better use for one's own resources than anyone else on the market.) Thus efficient markets achieve at least one component of a utilitarian ideal. Individuals are able to make the most of what they have.

Perfect markets are an unattainable ideal, but notice how the ideal comes about: by a kind of maximization of economic liberty. All mutually beneficial trades and combinations are unimpeded. Approaching the perfect market ideal consists largely in trying to promote this economic liberty or replicate its operation. Of course maximal economic liberty is not the same thing as laissez faire. The natural tendency of economic power, in the absence of any ground rules, is toward concentration and monopoly, reaping advantage by depriving others of economic opportunity. It takes restrictions to maintain competition, but the effect of these restraints is to increase economic liberty. Similarly with other sorts of market failure. Internalization of externalities through the use of taxes is an attempt to achieve what economic liberty would achieve if only it were possible to establish markets in the external goods and bads and let people make mutually beneficial trades. Similarly with public goods. There is no way to tax public good externalities so as to get efficient provision of public goods. (A defining feature of public goods, such as clean air, is that there is no way to make people pay for the individual value they get out of it, so there is no market price on which to levy a tax.) Thus the second best solution is to try to use government to approximate the results that economic liberty would achieve if a market for the public good or bad was possible. The upshot is that utilitarianism may call for a lot of restrictions -- you can't establish a monopoly, you can't build a toxic waste dump wherever you want, you have to pay taxes, and in general, you have to take into account, in the prices you face, the costs that your activities impose on others -- but the effect of these restrictions is to come as close as possible to the unattainable ideal of full economic liberty. The ideal set of utilitarian restrictions on economic activity can be regarded as securing in the economic sphere something like what John Rawls called for in general: "the most extensive basic liberty compatible with a similar liberty for others."[40]

Slavery is prima facie the antithesis of this ideal. To the extent that slaves are not allowed to pursue the opportunities that would be available to free men, they cannot use their private information to direct their personal resources to their most valued uses, and this is a dead weight loss. As Friedrich von Hayek put it "practically every individual has some advantage over all others because he possesses unique information of which beneficial use might be made, but of which use can be made only if the decisions depending on it are left to him or are made with his active cooperation."[41] When this private information cannot be exploited, the value it could have reaped is lost. This hangs a millstone around slavery's neck in a utilitarian calculation, not just vis a vis free labor markets, but vis a vis any liberal alternative. If for some reason markets were not workable -- suppose some deus ex machina from outer space threatened to destroy any country that did not communize its whole economy -- the utilitarian object of government would still be to mimic economic liberty, and afford it as far as possible. At every marginal step, utilitarianism would lean away from slavery, not towards it.

Interestingly, Mill thought that communism might well prove the better system for letting people make the most of their lives. Writing in mid-nineteenth century England, and very alert to the imperfections of capitalism, Mill could plau sibly suggest that collectivism might provide both more occupa tional choice and stronger in ducement to work effort than capitalism. On occupational choice, Mill noted: "The generality of labourers in this and most other countries, have as little choice of occupation or freedom of locomotion, are practically as dependent on fixed rules and on the will of others, as they could be in any system short of actual slavery."[42] On work effort Mill wrote: "If communistic labor might be less vigorous than that of a peasant proprietor, or a workman laboring on his own account, it would probably be more energetic than that of a laborer for hire, who has no personal interest in the matter at all."[43] Mill also noted that joint-stock companies face the same incentive problems (what are today called "principle-agent" problems) that government run industries do, so he was optimistic that government agents could be at least as efficient as hired managers.[44] Thus Mill maintained that "which of the two will be the ultimate form of human society ... will probably depend mainly on ... which of the two systems is consistent with the greatest amount of hu man liberty and spontaneity."[45] We have learned a bit since 1848 about which system is consistent with the greatest amount of human liberty, but the basic point was nailed home by Mill: under either system, efficiency will come from maximizing economic liberty. With the advantage of hindsight it is pretty obvious that the system whose fundamental nature is to curtail liberty will afford less liberty than the system whose fundamental nature is to afford liberty.

With the economic considerations piled against slavery, how can it be argued that slavery might maximize the sum of utilities? To show possibility it is enough to come up with an example, in which one attempts to wed slavery to some great benefits that might offset its liabilities. R.M. Hare's essay in this volume inquires into the business of example building in utilitarian analysis in general and on the subject of slavery in particular. Hare's efforts provide a good jumping off point for considering whether attempts to wed slavery to some great benefits can successfully create an example where utilitarianism could call for slavery. Hare's own philosophical stance is utilitarian, but he does not invoke a concept of higher ends (that is, he is what I have called a Benthamite utilitarian). As noted earlier, with no restrictions on what is put into the utilitarian maximization problem, no restrictions are implied on what can come out. Any conclusions about what will maximize the sum of utilities must await the facts about how the people in a particular society feel pleasure and pain. As Hare puts it: "The utilitarian cannot reason a priori that whatever the facts about the world and human nature, slavery is wrong. He has to show that it is wrong by showing, through a study of history and other factual observation, that slavery does have the effects (namely the production of misery) that make it wrong."[46] But the facts about human nature cannot be just anything. The theory of ends can tell us important things about what the relevant part of human nature will come to value (the relevant part being, for the purposes of a Millian utilitarian analysis, the part which is manifest when people follow habits of reason and make progress in discovering what there is to value in the world).

Hare makes all the key arguments, but because of his philosophical position he is unable to make the small jump necessary to reach an unequivocal rejection of slavery. Recapping Hare's example, Hare imagines two islands of slave plantations, both colonies of a European power that has suddenly lost in war and is unable to keep up its empire. The slave-masters flee and on one island ("Juba") those slaves who had formed the manpower of the island's military take over and retain the slave system, improving its conditions and productivity, so that all share in the prosperity. The other island falls into a Hobbesian war of all against all. The people on this island are miserable and take to boats and try to escape to Juba where they will, in their own view, be much happier as slaves.

Given these choices, utilitarianism clearly seems to favor the set up on the slave-holding island. The attempt to wed the liabilities of slavery to offsetting benefits is carried out by artificially limiting the choices. But Hare recognizes that this is untenable. He asks: "Did the retention of slavery in particular contribute very much to the prosperity of Juba that could not have been achieved by other means?"[47] Or again: "What on earth are the slaves doing that could not be more efficiently done by paid labor?"[48] If the ex-military on Juba could control the situation, presumably they could have set up functioning liberal institutions of law and order and property and commerce. Liberal institutions take less resources to run than a command economy, not more, because they decentralize decision-making. There is no non-artificial way to presume that a slave economy can be efficiently maintained but liberal law and order cannot, unless we are talking about an island of incorrigible criminals. But even here, liberal institutions will do. Those who cannot resist breaking the law could be thrown into work camps and effectively reduced to slavery, leaving those who are willing to abide the laws free to make their contributions and get rewarded for it. If everybody ends up in jail, well, then liberal institutions have not done any worse than the slave state, but they have the potential to do much better.

With the economic arguments stacked against slavery, the only way that slavery could be favored in a utilitarian calculation is if people are so enamored of slavery itself that this preference will more than offset the liabilities of slavery. It is true that in the absence of any conclusions about ends, this cannot be ruled out a priori, but it is certainly very implausible on its face: many more people are crushed by slavery than are empowered by it and, as Frederick Douglass so eloquently recorded, those who are empowered are corrupted and degraded by their impunity. To turn this implausibility into an impossibility it is only necessary to invoke the theory of ends in the most minimal way. The invidious pleasure of valuing slave ownership for its own sake is clearly a lower end that is conditional on not making progress in discovering what there is to value in other human beings. And of course any preference for being a slave shows a clear lack of progress. Thus slavery for its own sake can be seen to make a necessarily negative contribution to the sum of progress in attaining higher ends. Combined with its incompetence as a means, slavery is indefeasibly rejected by a higher ends based utilitarianism.

I can think of a couple of other examples that test this conclusion by making stronger attempts to wed slavery's liabilities to offsetting benefits - examples that probe for a weakness in the theory of ends. To capture the most direct attack, consider how industrializing countries often manifest a backwards sloping supply of labor curve. Workers only work as much as needed to satisfy their habitual needs, so that when wages rise, people work less. Now suppose this was the case in Roman society. In these circumstances it would have been impossible to get the Roman roads and aqueducts and buildings built with free wage labor. The only way to get the job done would have been with slave labor, as the Romans used. Might even a higher ends based utilitarianism justify slavery in this case, if without slavery the Roman empire, with all its high accomplishments, would have been impossible? The answer is "no", because this would be a form of perfectionism, where higher ends are not open ended, but some particular end or set of ends is held to be the one set of particular ends whose attainment is to be maximized. In Millian utilitarianism, higher ends are whatever people discover to be valuable. If workers trade their tools for independent time as soon as they are able, that is great from the point of view of higher ends. "Leisure" affords people the opportunity to devote their resources to discovering and pursuing what is valuable in life. If they want to earn money, that is great too. It suggests that they have found things worth pursuing in life that take money to pursue. Mill's theory of ends rejects any perfectionist standard.

Of course people do get stuck in lower ends, but so long as Mill was making sense when he argued that the best way to foster higher ends is to allow freedom and responsibility, a higher ends based utilitarianism must favor the liberal alternative. This was the key to the argument for Mill's principle of liberty and it is worth reiterating its force. Drawing the line at forcing people to be wise means leaving available every mechanism short of force for passing on learning about where value lies and how to pursue it. People individually and collectively are free to educate, to remonstrate, to reason, persuade or entreat, and if someone cannot learn in any of these ways then he can only learn the hard way, for himself, and this learning is the criterion by which a higher ends based utilitarianism measures progress.

Another example seems even stickier. Suppose Roman society would have found it collectively worthwhile to finance public goods like roads and aqueducts but lacked the modern taxing power to finance them? Might this provide a justification for using slaves in a higher ends based utilitarianism? Again the answer is "no". Markets are just too facile. There will always be some market solution that is easy to implement and will provide a liberal way to bring about the efficient use of resources. (The reason that present day market economies are so often terribly inefficient is primarily because corrupt politicians have vested interests in not using market mechanisms to internalize externalities and otherwise eliminate inefficiency). In the present example, each person could be encumbered with a transferable obligation to work x weeks a year on public projects. A market could then be established where some people would take on other people's obligations for a fee, which would be determined by supply and demand, and this was in fact a common institution in earlier times. The problem for slavery is that the advantages of liberal alternatives are unambiguously superior and virtually always feasible. Thus on the most general principles something as illiberal as slavery could never be a part of the maximizing pursuit of higher ends. Add that liberty is the best food for the development of higher ends and the result is an indefeasible utilitarian right not to be enslaved. There is only one more place I can think of that could conceivably harbor a leak in this argument.

Suppose a labor force large enough to depress wages to a point below subsistence. Of course, entrepreneurs and capitalists will be making more money than ever in this circumstance, because they will only be paying a pittance for wages. Might a benign form of slavery, like the communist system at its best, actually be able to redistribute wealth from the rich to the poor in this case, thereby delivering resources to where they have the greatest impact on happiness and increasing the sum of happiness? Granted, we might not want to call a such a system slavery, if it redistributes wealth towards the slaves rather than stealing the value of their labor from them, but even here there will always be market alternatives that can achieve the same distributional outcomes without curtailing people's liberty. Just set up an equitable tax system. Of course you would have to correct the underlying problem by strictly enforcing obligations of parents to children in order to deter irresponsible childbearing. Otherwise in the long term there would be no solution to the "reserve army of the unemployed" under any sort of regime.

The essence of slavery, to sum up the above examples, is a deprivation of liberty which in itself is extremely noxious both to the efficient pursuit of ends and to the development of higher ends. For these liabilities to be outweighed this essence of slavery must somehow be wedded to some other valued quantity. But this wedding is always artificial. Slavery in itself is nothing but these noxious elements, and liberal institutions are too flexible and too powerful to not be able to separate the noxious elements of slavery from any arbitrarily attached valuable elements. Slavery starts out with a ton of bricks strapped to its back and varies everything else to see how fast it can still manage to go while the competing liberal institutions are free to vary everything. In a utilitarian calculation, slavery can only lose this competition, just as a car can only go slower if you give it four flat tires. For slavery to win, liberal alternatives would have to be artificially disabled, and so on general principles, we can deduce an indefeasible utilitarian right not to be enslaved.


1.  Besides his "proof" of utilitarianism (Utilitarianism, chapter IV, Hackett, 1979/1861), Mill offered other arguments for utilitarianism that are also inadequate. Near the end of Utilitarianism Mill explicitly embraced what today we think of as the defining feature of utilitarianism - Benthams idea that every person's welfare of should be considered interchangable on an even par with everyone else's - but he actually maintained that this followed from the meaning of words! In justifying Bentham's dictum "all to count as one, none as more than one", Mill wrote: "[E]qual amounts of happi ness are equally desirable, whether felt by the same or different persons. This, however, is not a presupposition, not a premise needful to support the principle of utility, but the very prin ciple itself; for what is the principle of utility if it be not that "happiness" and "desirable" are synonymous terms?" (See Utilitarianism, op. cit., pp. 60-61., Chapter V, PP 36 (3rd from last) and the accompanying footnote 4.) Because it is the principle it does not need support? So Mill apparently thought, and so he never bothered to ask "desirable for whom?"

Because Mill's arguments for utilitarianism are inadequate, it is highly significant that his arguments for liberty are consistent with but not dependent on the sum of utilities rule. It is significant both in its own right - if the argument for utilitarianism requires its own assumptions, it is valuable not to have to rely on utilitarianism. And it is significant to an assessment of Mill's merits as a philosoper. I surmise that Mill's refraining from utilitarian arguments in On Liberty stemmed from an understanding of what he had given sound arguments for and what he had not. It seems to me that Mill was an extremely systematic and coherent philosopher, even if his system was often not stated and perhaps not even explicitly grasped by Mill himself. His judgement about what made sense and what was well founded within his philosophy was practically unerring.

2. A sketch of what I believe is the tenable argument for utilitarianism appears in footnote 38, after pertinent discussions in the text, in particular about the theory of ends, have been presented.

3. Mill made clear in An Examination of Sir William Hamilton's Philosophy, (London 1865) that he recognized a role for reason, not just in making deductions, but in marshaling evidence for premises. As Mill put it: "what the Logic of mere consistency cannot do, the Logic of the ascertainment of truth, the Philosophy of Evidence in its larger acceptation, can. It can explain the function of the Ratiocinative process as an instrument of the human intellect in the discovery of truth." (p.457.) See also book six of Mill's System of Logic, available in a separate volume entitled The Logic of The Moral Sciences, Open Court, La Salle Illinois, 1987/1843.

4. ibid. System of Logic, chap. book vi, chapter xii, section 6.

5. ibid.

6. Mill wrote: "Of two pleasures, if there be one to which all or almost all who have experience of both give a decided preference, irrespective of any feeling of moral obligation to prefer it, that is the more desirable pleasure." (Utilitarianism, chapter II, PP 5.) Bentham did urge that pleasures not all be weighed the same. He wanted intensity and duration and other measures of pleasure and pain be taken into account (see his Principles of Morals and Legislation, Prometheus Boks, Buffalo New York, 1988/1871, chapter IV). But Bentham did not go beyond the idea of weighing each person's pleasures and pains however they were felt. In Mill's scheme, lower ends are to be valued as higher ends - as more informed judgement - would weigh them, not as the person whose utility is being tabulated judges them.

7. While he did not describe informed choice as a matter of rationality, Mill did maintain that "no intelligent human being would consent to be a fool, no instructed person would be an ignoramus, no person of feeling and conscience would be selfish and base, even though they should b e persuaded that the fool, the dunce, or the rascal is better satisfied with his lot than they are with theirs." (Utilitarianism, Chapter II, PP 6.) Mill did not consider this preference for knowing to be a matter of rationality because he was too well aware of the weight hanging in the other side of the balance: that "[a] being of higher faculties requires more to make him happy, is capable probably of more acute suffering, and certainly accessible to it at more points, than one of the inferior type" (ibid.). Thus to account for this preference for informed living Mill adverted to "pride" and "dignity".

Mill was absolutely right that the complex choice he used as his example cannot be decided as a simple matter of rationality. Rather, it falls under one of what I have called the "few qualifications" to the rationality of a preference for informed choice. Informed choice can occlude some choices. One can no longer in Mill"s example be a "pig satisfied" and not know what one is missing. The value of this lost opportunity must be weighed against the always positive value of informed choice per se to determine whether informed choice is preferable in the given instance. There is no question that informed choice per se will always have positive value. A rational person would at least want to choose whether to be a pig or a person on the basis of knowing what she was choosing between, even if she would like to have the knowledge genie back in the bottle after choices have been made.If Mill had not bit straight into such a subtle example he might well have stated the preference for informed choice directly in terms of rationality. Certainly it is not that his analysis came up short. Rather, his failure to identify the forest came from the clarity with which he saw the trees and was drawn to describe their complexity before he had clarified the larger simplicity.

8. See On Representative Government (Library of Liberal Arts, Bobbs-Merrill, Indianapolis, 1958/1861) the first several pages of chapter II.

9. On Liberty, op. cit., p. 10 (chapter I, PP 11).

10. From Mill's essay "Bentham", in Utilitarianism and Other Essays, Penguin Books, London, 1987, p. 146. Besides Mill's essays on Bentham and Coleridge, the other place where he gives an extended discussion of partial truths needing to be completed and corrected by each other is in On Liberty, near the end of chapter II, pp. 43-50 (PP 34-39).

11. From Mill's essay "Bentham", op. cit., p. 146.

12. Amongst liberals, Isiah Berlin has been seminal in maintaining that the liberal tolerance depends on moral relativism, rejecting the idea of knowable truth. These strains begin with his wonderful/ awful essay "J.S. Mill and the Ends of Life" (reprinted in Berlin's Four Essays on Liberty, Oxford, 1969) and culminate in The Crooked Timber of Mankind (Vintage Books, New York, 1992) where Berlin offers Roman Valor in war and Christian pacifism as an example of two irreconcilable sets of values. But on the contrary, what Rome and Christianity really example are two singular sets of wisdom that very much need to be "completed and corrected by one another." Indeed, in On Liberty Mill used Christian thought as his own central example of a philosophy that is one sided and needs to be augmented by the wisdom of Greece, Rome, the Koran, and more (op. cit. pp. 46-49, chapter II, PP 8-6 from end).

Berlin's thinking is that that if values can be reconciled, that implies a unitary kingdom of ends, which he associates with fascist and other utopian visions that only recognize certain values and are dismissive or murderous towards others. But this last association does not follow. There is every difference between a theory of value that achieves unity by being exclusive and one that achieves unity by accounting all value. Berlin has always been right in holding with Mill that the diversity of valuable ends is much of what makes liberty so valuable as a means. But he made the mistake (which Mill did not) of thinking that if diversity is the foundation of liberty then irreconcilable diversity (diversity squared?) must make an even stronger foundation. On the contrary, since it is nonsense it provides no foundation.

Critical theorists and other leftists of various persuasions spurn the concept of progress in similar fashion. They seek to show that valuations have no sure foundation so that all valuation can be attributed power and oppression. Barbara Herrnstein Smith, in the first chapter of her book Contingencies of Value (Harvard University Press, Cambridge, 1988), offers alternative readings of Shakespeare's sonnet 116 as evidence of supposedly incompatible valuations of the poem. (On one plausible reading it is a bad poem. On another it is good. Which is it?) But again, what is really exampled is how different discoveries of value can correct and complete each other. As with Berlin, Herrnstein Smith's intention is perhaps commendable but the result is to deny exactly the grounds on which moral theory can require different societies to take full account of each other's different merits.

13. For a discussion of Mill's associationist psychological model, see chapter 1 of Fred Berger's book Happiness, Justice and Freedom: the Moral and Political Philosophy of John Stuart Mill, University of California Press, Berkeley, 1984.

14 On Representative Government, op. cit. pp. 96-97 (chapter 6, PP 16).

15 Ibid.

16. On Liberty, op. cit. p.56 (chapter III, PP 3).

17. Thorstein Veblen, The Theory of the Leisure Class, MacMillan, 1899.

18. See "The Quintessence of Ibsenism" in Bernard Shaw: Major Critical Essays, Penguin Books, London, 1931, pp. 40-46.

19. Leon Festinger, A Theory of Cognitive Dissonance, Stanford University Press, 1957.

20. In this, Jane Austen made a career out of analyzing the various ways people jump to unwarranted conclusions rather than maintain the more complex structure of probabilistic reasoning. More recently, economists have done quite a bit of research on some simpler violations of probabilistic reason. Economists have found, for instance, that people will go to great expense to avoid taking on new risks yet will not go to much less expense to eliminate much greater status quo risks. See Fatal Tradeoffs: Public and Private Responsibilities for Risk, by W. Kip Viscusi, Oxford University Press, London, 1992. For a brief discussion of violations of probabilistic reason that lie behind religious intolerance, see footnote 30.

21. Utilitarianism, op. cit., pp. 13-14. (chapter II, PP 12).

22. Treatise on Human Nature, book 3, part 1, section 1. See also, An Inquiry Concerning the Principles of Morals, appendix I.

23. In his book John Stuart Mill (Routledge, London, 1989) John Skorupski makes a similar point about Mill's observational method. In defending Mill's "proof" of utilitarianism from G.E. Moore's charge that it commits "the naturalistic fallacy", Skorupski writes: "When Mill argues that happiness is desirable by appealing to the "evidence" of what human beings reflectively desire, he points out that he is not putting forward a deductive proof. So he does not sin against the elementary truth. But Moore's preoccupation with the purity of ethical analysis blinds him to the very simple point Mill makes - and to the inescapability of his way of making it: deliberation about ends can take no other form than appeal to what we discover by reflective analysis to be our categorical ends." (p. 3)

My earlier comments on the nature of lower ends might seem to provide a counter-example to this assertion. There I did not make empirical observations about what there is to value, but instead projected the consequences of violating the principles of progress and conservation in particular ways. This is a form of moral reason, but its subject is non-ideal theory. It can tell us specific things about the content of lower ends -- that they will tend to be selfish for instance -- but the only way to say anything about the content of higher ends is via empirical observations of value. Another role for moral reason is in figuring out how evidence of value fits together and how it can be understood and abstracted. It is here that moral reason comes closest to actually deducing content to higher ends, but still this is the science of how to martial evidence. Evidence is not itself deduced. Finally there is the role of instrumental reason in moral reason, figuring out how to pursue value. Instrumental reason becomes a part of moral reason when value is at stake.

24. On Liberty, op. cit. p. 9 et seq. (Chapter I, PP 9-12.)

25. This pattern holds not just for Mill's argument but for the development of established liberties in England and the American colony. For a thorough study of this development in American constitutionalism, see David A. J. Richards, Toleration and the Constitution, Oxford University Press, Oxford, 1986.

26. On Liberty, op. cit. p. 16 (chapter II, PP 1).

27. The question arises whether the sphere of direct interests in other people's speech or any other sort of activity can really be delimited at all. Mill himself acknowledged the thorny problem of how a person's disregard for his own welfare can harm the direct interests of those who love him by draining their resources by trying to help and their health by worrying. (On Liberty, op. cit. p.78 et seq. (Chapter IV, PP 8 et seq.)) Here Mill adverted to the criterion of "assignable rights", which require a utilitarian analysis to determine. If this were the only solution it would undercut my claim that the argument for Mill's principle can be made without invoking the greatest happiness principle. But there is another solution, right in the identificaton of "indirect interests" with vicarious interests. So long as the harm to ones direct direct interests works through one's indirect interests -- one's vicarious concern for the welfare of others -- one's interest is properly classified as indirect.

Some other difficulties in separating direct and indirect interests cannot be so easily dispensed with, but they can be delimited. Most obviously, economic interests are direct interests and almost anything about a person's progress in discovering what there is to value may affect his economic decisions and affect other people's direct interests in that way. Here there is no alternative but to advert to Mill's concept of assignable rights, which is clearly the correct criterion for settling conflicts of economic interest. This problem can be delimited by noting that the issues of economic liberty where direct interests are hard to delimit are a very different set of issues than Mill's principle is primarily concerned with. We can just note that the economic questions are not settled by Mill's principle, and that that does not keep Mill's principle from being able to settle other questions. Some such accomodation must be expected when attempting to analyze one part of moral theory independent of the whole. So long as the argument that is eventually made for utilitarianism does not depend on the principle of liberty (and it does not) the integrity of the whole is preserved.

28. ibid., p. 21 (chapter II, PP 10).

29. ibid., p. 32-33 (chapter II, PP 20).

30. Historically, the most influential paternalistic arguments were the religious ones, claiming to know God's will and the way to salvation. As this case is no longer of central interest to as many people as it once was, I will treat it here in a footnote.

John Locke noted in his Letter on Toleration that "religion consists in the inward persuasion of the mind". (Library of Liberal Arts, Macmillian, London, 1955/1689, p. 18, PP 9-10.) Anyone who accepts this must be swayed by Mill's argument for religious toleration. "[F]aith is not faith without believing" Locke wrote, and "such is the nature of understanding that it cannot be compelled to the belief of anything by outward force."(ibid.) Thus Locke maintained that one's path to heaven can only be found when "every private man's search and study discovers it unto himself." (ibid. p. 31, PP 38.) Thus the paternalistic argument falls through here too (even before the fallibilism of religious beliefs are taken into account).

It is worth noting that Mill's principle of liberty could also be attributed to Locke. Locke's category of "civil interests", which referred to "life, liberty, health, and indolency of the body; and the possession of outward things, such as money, lands, houses, furniture, and the like" is essentially identical to Mill's "direct interests" (Letter on Toleration,.PP 7). And like Mill, Locke maintained that the only grounds for laws were to protect these civil or direct interests (ibid. PP 9). Mill was explicit that the law was justified only in protecting people's direct interests from infringement by others, not from themselves, but this is implicit in Locke too. (See for instance ibid. PP 37-38 (pages 30-31).

More needs to be said to fully address religious arguments for forcing people to do or forebear. In particular I think it is necessary to criticize the way all of our major religions violate the fundamental principles of moral reason (the principles of reason as they pertain to the discovery and accounting of value) with their violations of probabalistic reason. They ask us to attend only to one possibility: the possibility that what they teach about God and the fundamental questions is true, or that they are in possession of the true word of God. But since the contrary possibilities cannot be ruled out, this violates the requirements of probabalistic reason. Christianity tries to get around this by making a distinction between faith and belief. This avoids the embarrassment of falsely asserting a certainty that one cannot honestly possess, but the moral error is the same. Whether one proceeds on faith or on a dishonest certainty, the result is the same refusal to account relevant possibilities.

When the requirements of reason for how to follow and marshal evidence are abided by, I do not believe that any religious grounds for wanting to violate Mill's principle of liberty remain. If this argument leaves any gray areas, certainly the utilitarian argument against state establishment of religion does not. As Mill noted in his introduction to On Liberty, the mutual benefit of religious toleration is the one seminal lesson in liberty that by the mid nineteenth century had finally been learned.

31. There may be exceptions to the claim that another person's progress is best served by letting him learn by experience, if he cannot learn any other way. For instance, can we really not determine that on a "correct calculation" it makes sense to buckle one's seat-belt? Yet a person who learns the hard way that it makes sense to buckle his seat-belt may well have lost the opportunity to benefit from this knowledge. The paternalistic argument for seat-belt laws seems to withstand the Millian counter-argument. Mill noted that the attempt to exercise control must not "produce other evils, greater than those which it would prevent." (ibid., p. 11, chapter I, PP 11.) But if the punishment for not wearing one's seat-belt is a fine small enough that it will not very often leave a person unable to pay her rent, well, some exceptions like this to Millian liberty are not prima facie untenable. But this is a very rare case. We know that a car driver will be sitting in the exact same place doing the same things whether she has her seat belt on or not. There is no other side of the equation except the couple seconds necessary to buckle the belt. But we cannot make a "correct calculation" that rock climbing, for instance, is too dangerous without putting a value on rock climbing. Then all of Mill's arguments about how we cannot be sure an opinion is right apply again and are persuasive. Progress depends on each person following his own discoveries about where value lies as best as he is able. What exceptions do hold up will have the effect of marginally narrowing the sphere of liberty that is secured by Mill's arguments, but the core of Mill's principle, that people should not be ruled by the mere "likings and dislikings of society", remains untouched (ibid., p. 7, chapter I, PP 7).

One might object to my characterizing Mill's argument for liberty as hinging in any way on whether liberty promotes the welfare of who it is accorded to. After all, Mill was adamant that "[h]is own good, either physical or moral, is not a sufficient warrant" for interfering with a person's liberty. (ibid., p. 9, Chapter I, PP 9.). But this was because restraining a person supposedly for his own good does more harm than good. At least, such is the nature of the arguments that Mill provided.

32. Mill arrived at a similar conclusion, but he got there in an untenable way. He was aware of the need for a distinction between public and private behavior, but he addressed this issue in a way that only exacerbated the inconsistency in his formulation. Mill wrote: "[T]here are many acts which, being directly injurious only to the agents themselves, ought not to be legally interdicted, but which , if done publicly, are a violation of good manners and, coming thus within the category of offenses against others, may rightly be prohibited." (On Liberty, op. cit., p.97, chapter V, PP 7.) He admits that the direct interests of others are not involved, but includes their remaining interests under his concept of "harm" anyway, a clear violation of his own categories.

33. Of course there are many activities that people have a direct interest in being allowed to engage in public. In particular, caring what other people think can itself become a direct interest when the interest is mutual. Then it becomes a matter of communication between two people whose decisions, and hence whose direct interests, may depend on the information communicated. Thus while some dress code, for instance, could be admissible (as discussed above) it would have to be minimal, still leaving all necessary scope for personal style.

34. It might be objected that I am being inconsistent here. Earlier I maintained that rationality required not hiding from the truth and here I am asserting that rational people will suppress distasteful thoughts. But there is no inconsistency. One wants to know if a flame can hurt. That does not mean that one should hold ones hand in a flame and keep learning what one already knows. The question is, knowing what you know, what is the rational way to allocate your attention?

35. The defining virtue of utilitarianism is the way it separates for individual tabulation the different considerations and the means appropriate to them, so that people cannot, for instance, continue to believe that they are motivated by other people's good, when the only reasons that actually weigh in the direction they favor are selfish. When people cannot fool themselves about their motivations, it becomes impossible for a person who has made progress in discovering what there is to value in the world to want to constrict other people's lives more than Mill's principle allows.

36. Mill defined what seems to be the lower boundary of direct interests as follows: for a person's liberty to be protected, "he must not make himself a nuisance to other people." (On Liberty, op. cit., p.53, chapter III, PP 1).

37. Some other arguments for and discussions about Mill's principle are relevant to my discussion here. In his 1992 article "Mill's Deliberative Utilitarianism" (Philosophy and Public Affairs, V21, No. 1, pp.67-103) David Brink also interpretted Mill's argument for liberty as an argument from higher ends, though his is an argument from higher ends based utilitarianism, wheras the argument here proceeds entirely from the theory of ends and does not invoke the sum of utilities criterion. Brink's article was important for bringing higher ends into the picture. The problem with his analysis is that, rather than frame the argument for liberty in terms of Mill's concept of higher ends, he frames it in terms of Mill's speculations about the content of higher ends. This is natural, because as soon as Mill introduced his criterion of higher ends, he turned immediately to speculation about the content of higher ends, which he identified with the exercise of the "higher faculties" (Utilitarianism, Hackett Publishing, Indianapolis, 1979/1861pp. 7-10, chapter II, PP 4-6). Mill agreed with the Epicurean theories that "assign to the pleasures of the intellect, of the feelings and imagination, and of the moral sentiments a much higher value as pleasures than to those of mere sensation." (ibid. pp. 7-8, chapter II, PP 4). Mill also claimed that this priority approaches the absolute - that no amount of the lower pleasures are worth giving up the higher pleasures for (ibid.).

From this content to higher ends Brink is able to argue that liberty rights are required by Mill's utilitarianism. Without liberty of thought and discussion and action, people cannot develop their higher faculties and hence are deprived of higher ends, which have a dominant weight in the utilitarian calculus. Thus "these liberties are necessarily, and not just contingently or epistemically, connected with the realization of dominant components of value." "Necessarily connected with" does not quite a prove indefeasible rights, but as far the argument goes it does work, in the sense that the conclusions follow from the premises. But because Brink bases his Millian argument on Mill's speculations about the content of higher ends Brink's premises are philosophically suspect. The premises are valid if the content of higher ends is what Mill supposed, but when people follow habits of reason to a discovery of what there is to value in the world, do they really give absolute (or lexical) priority to the exercise of the higher faculties? To identify higher and lower with complex and simple is a strong perfectionist claim, and is problematic at best. There are some very nice simple pleasures after all, and the case for lexicality does not seem to hold water. Brink's way of framing a Millian argument for liberty is internally consistent, but not externally well supported by the evidence of what there is to value.

Moving from an argument based on the content of higher ends to the concept of higher ends makes the case for lexicality more tenable but ultimately it is necessary to get away from reliance on claims of lexical priority. Focussing on the concept of higher ends makes lexicality more tenable because there is a sense in which higher ends are literally lexically superior to lower ends. According to Mill's concept of higher ends, 'higher ends' just refers to a better knowledge or understanding of what there is to value, and this better understanding is to completely superseed the less complete understanding which it has grown from. Of course we can lose understanding as well as gain it, but when we do our subsequent ends are not to be called 'higher', at least respecting the degraded parts of understanding. When ends are higher, they are all that is to count, and the superseeded lower ends are not to be used at all in adding up value. This does not mean that the things that superseeded lower ends held valuable are no longer valued, only that they are valued as higher ends - as better understanding - has come to value them. So there is a way in which actual lexicality does hold, and this can almost be parlayed into a lexical priority argument for liberty.

Correctly emphasizing the concept rather than the content of higher ends, Will Kymlicka describes Mill as arguing that: "liberty is needed precisely to find out what is valuable in life -- to question, re-examine, and revise our beliefs about value." (Liberalism, Community and Culture, Clarendon Press, Oxford, 1991, p.18.) If liberty is, as Mill believed, a linchpin for the discovery of what is worth pursuing in life, then since this discovery of higher ends is a linchpin to the attainment of higher ends, and the attainment of higher ends is all that matters, a lexical priority for liberty might seem to be indicated. The problem is that the discovery of higher ends is not the only linchpin for the attainment of higher ends. The pursuit of ends is also critical, and other things may be necessary to the discovery of higher ends besides liberty. Amongst necessary conditions, there can be no lexical priority, hence no conclusions can be reached about what is required for the maximal attainment of higher ends without going into the ways that these different considerations pair up.

As section III of this paper tries to demonstrate, analysis of an encompassing classification of considerations shows that they do "pair up" and in the simplest way. All can be seen to be aligned in favor of Mill's principle of liberty, thus eliminating the fiction that the smallest amount of harm to liberty as a key the the discovery of value might be pitted against some large amount of harm to other important considerations. This is the sense of Mill's variety of arguments -- to show how the various considerations all favor liberty, so that arguments for the lexical priority of some concerns over others is not necessary to make a conclusive argument for liberty. At least, this is the way I proceed with Mill's beginnings. John Gray and G.W. Smith, in the introduction to their edited volume of Essays on Mill, describe reliance on lexicality as a general flaw of higher ends based attempts to square utilitarianism with liberty rights. (John Stuart Mill: On Liberty in Focus, Routledge, New York, 1991, p. 16.) They are right to regard claims of lexicality as problematic, but wrong to think that higher ends based arguments must rest on such claims.

38. Having already given a brief introduction to the theory of ends, it is possible at this point to make a quick sketch of the argument for utilitarianism for those who are interested. The argument stands only on the theory of ends and hence provides a foundation for utilitarianism of the same strength and character as the argument for Mill's principle in section III -- namely, that anyone who follows habits of reason and makes basic progress in discovering what there is to value in the world and how to pursue it will embrace utilitarianism as her concept first of the public interest and, with a bit more progress, as her private morality.

The key thing that has to be established is that as people make progress in discovering what there is to value in the world and how to pursue it, they come to embrace the value or principle of mutuality. As progress leads people's interests away just from self interested concerns to an interest in the many things there are to care about in the world around, people come to want to live in a way that contributes to that world rather than diminishes it.Similarly as they come to appreciate the value of particular other people's lives and the potential value of unknown other people's lives, they come to want to live in a way that contributes to rather than injures those they interact with.The great divide here is the concept of mutuality: a willingness to act on terms that all can embrace.

Just such a concept is embodied in the concept of "fairness" that John Rawls parlays into "the original position" in his Theory of Justice (Harvard University Press, Cambridge, 1971). This "original position" sets up a Millian utilitarian criterion for choosing principles of justice. (It sets up an expected utility maximization problem over states of the world where one has an equal chance of holding any position in society. This calls for maximization of the sum of utilities divided by population, or average utility.) Nominally Rawls lists utilitarianism as a principle of justice to be ranked by deliberation in the original position, and he rejects it in favor of his two principles of justice: a priority for liberty, followed by what he calls the "difference principle" (a principle of distributional justice). But what Rawls really does is use Millian utilitarianism (a higher ends based utilitarianism, making substantive judgements about value and priority in the utilitarian calculation) to reject Benthamite utilitarianism and arrive at his two principles of justice. The confusion of terminology is unfortunate but of little consequence once understood. The coincicence of Rawls's "original position" with Millian utilitarianism extends to the priorities, or content of higher ends, that he invokes. Rawls gives lexical priority to "the development of the moral faculties", which is identical to "progress in discovering what there is to value and how to pursue it", which is the measure of higher ends in my account of Mill's philosophy.

Rawls not only establishes the Millian utilitarian criterion for choosing principles of justice. He also goes on to analyze what principles of justice this criterion favors - his two principles of justice. His analysis of what principles of justice Millian utilitarianism calls for is almost entirely correct. I will have to write a paper about it someday.

39. The second theorem also depends crucially on the assumption that preferences are convex. For an account of the fundamental theorems of welfare economics in the relatively simple case of a pure exchange economy, see Hal Varian, Microeconomic Analysis, Norton, New York, chapter 5 section 5. For the simplest case involving production as well as exchange, see Michael Intriligator, Mathematical Optimization and Economic Theory, Prentice-Hall, Englewood Cliffs, N.J., 1971, 258-279.

40. That each person should be afforded an equal right to the most extensive basic liberty compatible with a similar liberty for others is the first of Rawls's two principles of justice. See Theory of Justice, op. cit., p.60.

41. Friedrich von Hayek, "The Use of Knowledge in Society", American Economic Review, 1945, v. 35, pp. 521-3. For an interesting discussion of the implications of private information for an ideal utilitarian scheme, see Partha Dasgupta's essay "Utilitarianism, Information and Rights" in Utilitarianism and Beyond, Sen and Williams eds., Cambridge University Press, Cambridge, 1982.

42. John Stuart Mill, Principles of Political Economy, Books IV and V, Penguin, London, 1970/1848, p. 361 (book II, chapter I, section 3, PP 9).

43. ibid. ,p. 355 (book II, chapter I, section 3, PP 2).

44. ibid., p. 326 (book V, chapter XI, section 11, PP. 1).

45. ibid., p. 360 (book II, chapter I, section 3, PP 9).

46. Editor: This quote comes from the sixth to last PP of Hare's essay "What is Wrong with Slavery". In Philosophy and Public Affairs, 1978. no. 2, it is on page 118. You will have to enter the page reference from the reprint of Hares' essay in this volume yourself.

47. Hare's fourth to last PP. PAPA p. 119. Editor: enter page reference from Hare's article in this volume.

48. Hare's PP 14. PAPA p.111. Editor: enter page reference from this volume.


Tne next item in the Liberty volume of Moral Science is: Reframing our System of Liberty

Rawls for Sheriff Home Page | Rawls for Sheriff | Moral Science | Checklist/Contents | Rate this Page | Submit Reply

Top of Page 

Liberty volume

Date Last Modified: 8/27/99
Copyright Alec Rawls © 1998