moralobjectivity.net: home page           'A Theory of Moral Objectivity' contents page

A Theory of Moral Objectivity (Section 3k - Utilitarianism)

By Robert M. Ellis, originally written as a Ph.D. thesis 'A Buddhist theory of moral objectivity' in 2001. This html version copyright 2008.

A hard-copy paperback book of this thesis is now available from lulu.com and also on Amazon, price UK£25 (or equivalent). This relatively high cost is necessary because it is A4 size and has 487 pages (296,000 words). This print version includes an index.

Support independent publishing: Buy this book on Lulu.

A downloadable pdf version of this thesis is available from the British Library at http://ethos.bl.uk (you will need to search the original title 'A Buddhist theory of moral objectivity', and register with the ethos site, but registration is open and the download pdf is free for researchers). Alternatively you can download a pdf for a small cost from lulu.com.

Join discussion or ask questions on any aspect of the thesis on the new phpbb discussion board

 

3k)      Utilitarianism

 

The final type of view I shall consider under the heading of eternalism is perhaps the most influential after Christianity, though that I consider it a type of eternalism may initially occasion some surprise given its strong links with the empiricist tradition. Utilitarianism has been advocated by a string of philosophers from Helvetius in the eighteenth century to the present, though it attains its classical, and most clearly eternalist, form in the hands of Mill and Sidgwick. In general I shall attempt to separate this classical form of Utilitarianism, with its appeal either to dogmatic empiricism or to intuitionism to support its ethical absolutism, from some earlier and later forms (such as that of Hume) which attempt to limit their ethics to a purely descriptive form and appeal only to social convention as the basis of ethics. Though the label is purely a stipulative one for convenience, I shall refer to this sort of ethical theory as empiricist ethics rather than as utilitarianism. This form of ethics, though nihilist rather than eternalist, shares many features with the utilitarianism I shall discuss here, and in this way some points relevant to my treatment of it in the next chapter will be anticipated.

 

i)                    A capitalist form of eternalism

 

The origins of utilitarianism are to be found in the development of capitalism, a connection which also explains its eternalist nature. I have already discussed the relationship between Christianity and the beginnings of capitalism[1], including some mention of the role played in the development of capitalism by the Christian (particularly Protestant) tendency towards an alienated rationalism. This rationalism involved the continual deferral of gratification, which is central to the spirit of capitalism as Weber portrays it. Whether this alienation of desires is motivated by moral or by financial goals, it needs to be accompanied by faith in a guarantee of security which ensures that the sacrifices made in deferring gratification will result in some later fulfilment of desire. As I have already mentioned, this guarantee can be believed in either at the absolute level of God or at the worldly level of an economic system, and it becomes the foundational assurance of a system of book-keeping which can be both moral and financial.

 

But the continual development of rationalism promoted by capitalism also ultimately undermined belief in God, creating a search for a substitute form of cosmic justice, which I have been tracking in each of the forms of transcendental humanism that I have been discussing from Kant onwards. For whilst capitalism relentlessly provided practical grounds for alienation driven by economic necessity, without God this system lacked a full rationalisation with a complete guarantee of moral as well as financial security. Kant, Hegel, Marx and Schopenhauer all provided rationalisations for this alienation in different ways, but none of them completely succeeded in offering a non-theistic moral ideology compatible with capitalism: instead, the appeals respectively to autonomy, monism, communism and asceticism merely offered partial accommodations to capitalism which tended to flourish only insofar as capitalism produced conditions conducive to those ideals. Perhaps of these four Kant came closest to producing a truly capitalist eternalism, stressing the belief in autonomous rationality required for those who engage in a market of any kind: but Kant’s anti-consequentialism makes it difficult for his ethics to be applied directly in the justification of capitalist enterprise, and unappealing to those who dwell, not in examination of the rational consistency of their motives, but in the future satisfaction of present desires. Perhaps Kant’s ethics suited those who sat, like Kant himself, very much on the edge of the sphere of capitalist development, but not those in the midst of it.

 

To fill this vacuum more completely, then, emerges utilitarianism (the beginnings of which in fact pre-date all four of the above-mentioned philosophers). Its most important early developments occurred in England, the cradle of the industrial revolution, during the eighteenth century, and began with the descriptivist ethics of Hume. These ethics do not by themselves provide a capitalist eternalism, since they express the scepticism produced as an incidental by-product of capitalism rather than providing the security which it requires for its full rationalisation[2]. It required Adam Smith to draw a prescriptive natural law ethic out of the very mechanism of capitalism itself to provide this rationalisation, allowing empiricist ethics to achieve a prescriptive status, and providing the moral security that was required from the very same source as the financial security that already existed.

 

At the core of this rationalisation was the invisible hand famously described by Adam Smith.

As every individual…endeavours as much as he can both to employ his capital in the support of domestic industry, and so to direct that industry that its produce may be of the greatest value; every individual necessarily labours to render the annual revenue of the society as great as he can. He generally, indeed, neither intends to promote the public interest, nor knows how much he is promoting it….he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention.[3]

 

The invisible hand is obviously intended as a metaphor for an economic process solely produced by human activity, but its relationship to the cosmic justice tradition of eternalism should be obvious. It seems that in this belief Smith was influenced directly by Stoicism as well as by Christianity[4]. It posited a necessary relationship of requital between the individual act of investment and a result which is not merely individual but collective. In this way it provided the beginnings of a universal moral justification for capitalist activity, though Smith himself maintained an ethics which was firmly descriptive and rejected prescriptive utilitarianism because it did not describe how people really reason about ethics. Although Smith’s approach was thus strictly empirical and claimed to be dealing with facts rather than values, it is the extension of his theory to a status of universality in defiance of the problems of induction, and the subsequent political use of that theory, that clearly incorporates prescriptive values into it. For Smith and his successors in economic thought, the gap between the theory of the inevitable social benefits of capitalism in the future and its shortcomings in the present could only be bridged through recourse to the dogmatic assumption of the invisible hand.

 

This belief about the results of capitalistic activity laid an important foundation for the belief in progress through human activity which is found in classical utilitarianism, and effectively provides its version of cosmic justice. Adam Smith and John Stuart Mill are connected particularly by their preparedness to assume the existence of natural grounds for transcendental humanism on the grounds of empirical enquiry (a preparedness shared also by Marx, who was influenced by Smith). For Mill, as for Smith, the motives of individual actions are irrelevant to the consequences they will create, but unlike Smith, he wanted to apply a prescriptive moral framework to the rational direction of those motives. Selfishness might create an ideal society, but only so long as it was rationally directed. Mill seemed to believe that all human problems could be overcome if only a utilitarian prescription was followed, there being some real and natural link between an individual’s rational estimation of consequences and the consequences that would actually ensue. He expresses faith that, with the aid of science, poverty and disease can be eradicated, and “All the grand sources, in short, of human suffering are in a great degree, many of them almost entirely, conquerable by human care and effort”[5].

 

The “invisible hand” of Mill’s utilitarianism is thus the hand which (not just in economic terms, but more generally) ensures that human investment in a rational universalist framework would result in a rationally-ordered world in which happiness was maximised. It is the invisible hand of consequentialism. To assume that there is any such link requires the foundation of a dogmatic empiricism, which Mill indeed offers in the shape of his doctrine of naturalism. As I shall discuss in the next subsection, it is debatable how far Mill himself appeals to empiricism as the foundation of his ethics, but in any case, even beyond this explicit appeal, some empiricist assumptions must be made in any universalist consequentialism (I shall be discussing these assumptions in more detail later in the section). It is this at least implicit appeal to dogmatic empiricism (resembling Marx’s) which takes the place of the appeals to other metaphysical sources of knowledge found in the earlier eternalist tradition, and thus enables Mill to genuinely claim that his ethical system is compatible with Christianity[6]. Mill merely offers a new form of natural law which, like the previous forms, is claimed to be derived from human experience but nevertheless assumes cosmic justice as the basis of its analysis.

 

In later forms of utilitarianism the moral significance of consequences is no longer understood to lie purely in the consequences themselves, but in the intentions of the agent[7]. This involves a move from the purely rational assessment of actions from an apparently abstract standpoint (“God’s-eye view”) to a recognition that consequential estimations are embodied and come from a particular standpoint in time and space. But a distinction is still made between the  rationality of the agent’s position and the particular motives which impel the action, and it is still assumed that a reconciliation can be made between the agent’s subjective desires and an objectivity required by reason through the use of purely rational mechanisms. The invisible hand, providing an impersonal rational prescription based on supposed objective causality entirely independent of the desires of the person making the judgement (except insofar as these themselves constitute supposedly objective facts), still operates here and supports utilitarian consequentialism and universalism. And even if modern utilitarians are not much prone to the claim that their doctrine is compatible with Christianity, they still make use of the eternalist technique of referring apparent errors and inconsistencies in their doctrine to their finite grasp of what is absolutely required  (i.e. to errors in utilitarian calculation) rather than to the base assumption that utilitarian calculation is capable of the universality they claim of it. As I shall argue, such an approach requires an assumed basis of knowledge either in dogmatic empiricism or in intuitionism, even if this is not always acknowledged.

 

My designation of utilitarianism as capitalist eternalism should not be taken to imply the claim that its uses are always capitalist, or that all uses of consequential calculation (such as those made in cases of capitalist investment) are necessarily alienated. Rather, utilitarianism as an ethical foundationalism provides a justificatory position for capitalist activity and its associated alienated rationalism, and thus creates a general correlation between them. Whilst the alienated psychology of the spirit of capitalism necessarily requires a dualist philosophy to express the structure of its beliefs[8], this philosophy could alternatively be another eternalist one, or (in the absence of any universal justification at all) a nihilist one. Moreover, many features of utilitarianism are not necessarily associated with ethical foundationalism at all and thus can alternatively be used in a non-dogmatic fashion. Consequentialist calculation, for example, can take place in a sphere where beliefs about consequential links are adopted provisionally rather than dogmatically, but in this case the consequentialism itself becomes provisional, a tool at the service of non-dualism, rather than being itself a foundational assumption of value. Such secondary use of consequentialism, moreover, does not adopt utilitarian assumptions about the definability of universally desirable goals.

 

All of these points will be given more substantiation in the subsections that follow. I shall begin with the epistemological questions and move onto the ways in which I shall allege the  dualism of utilitarianism impedes its success as an ethical system.

 

ii)                  The epistemological foundations

 

As I have already mentioned, there seem to be two kinds of foundational appeal which are explicitly or implicitly used in the attempt to support the universality of utilitarianism: these are empiricism and intuitionism. Sometimes these two are barely distinguishable, as empiricism amounts to a dogmatic appeal to our intuition that the senses offer unvarnished truth. In John Stuart Mill, as I shall argue, empiricism appears to have intuitionism hiding behind it, particularly when the explicit grounds he gives for his ethics are considered.

 

John Stuart Mill’s empiricism is given a systematic exposition in his System of Logic[9]. It offers a number of arguments which attempt to support a scientific naturalism which claims that scientific propositions can represent truths of nature. He achieves this firstly by making a distinction between “real” and “merely verbal” propositions, similar to that between synthetic and analytic[10]: all truth must thus be inductive in nature. General names, although they have a meaning independent of their relationship to nature, can nevertheless, as part of real propositions, be used to denote kinds which are either genuinely natural or merely conventional: Mill argues that the difference can be determined purely by the fruitfulness of the natural kind distinction in enabling the discovery of inexhaustible further distinctions which are not implied by the original distinction[11]. Natural kinds consist in “uniform co-existences which do not…derive from causal uniformities”[12].

 

Mill thus tries to combine nominalism with naturalism, and provides the first indications of the dogmatic basis of his empiricism in the realism about natural kinds which occurs in his analysis of language. The distinction between genuinely natural and merely constructed kind-terms is reminiscent of the phantasia kataleptike of the Stoics: in both cases an incrementality in the relationship between natural kind terms and their fruitful applicability (in the shape of theories predicting, with relative success, further features of a given kind on the basis of existing resemblances) has been made into a false dichotomy between the absolutely real and the merely assumed.

 

Mill does try to avoid representationalism by adopting nominalism, thus beginning an empiricist tradition which attempts to make knowledge absolute by dividing it from meaning. He does not succeed in this because, at least in the case of supposed real natural kinds, a new representationalism emerges, to replace that of essentialism, in the represented scientific agreement to base the meaning of a term on its empirical evidence. What is now “represented” and is taken to make the term meaningful is not the essential nature of the object, but the fact that such an essential nature is conventionally assumed by people within a certain group who use it. This kind of representation is no less independent of the subject and its purposes than is one based on essentialism.

 

Mill goes on to also argue that causal propositions can be conclusively demonstrated in some instances. He does this by arguing that we can analyse a given causal process into the simple constituent causes for each simple effect, and then, by experimenting on each simple cause and going through a process of elimination, isolate the sufficient cause of the simple effect in each case[13]. In this Mill makes at least four unjustified assumptions: of a universal law of causation, that simple causes can be atomistically determined, that the relevant causes can be conclusively isolated in relation to a given phenomenon, and that the process of elimination can be conclusive.

 

Alan Ryan argues that the assumption of a universal law of causation is of a different, and much more deductive, order from Mill’s insistence otherwise that only particulars can be deduced from particulars, this level of assumption even being one that could reasonably be described as Platonic[14]. It is certainly this level of assumption which seems to be operative in his later assertion that the law of causality operates uniformly on human actions. As the basis of his consequentialism, however, he also needed to assert that the relationships of particular events in human experience could be reliably predicted universally. The assertion of determinism could amount merely to an abstract prediction of future discoveries in the social and cognitive sciences, but the use of  consequentialism requires the possibility of particular and present knowledge of the type that Mill’s causal atomism tries unsuccessfully to establish.

 

Though Mill’s approaches to the justification of empiricism have long been superseded by more recent philosophers, the project which he represents of the possible empirical justification of the universal prescriptivity of utilitarianism appears to have been abandoned[15]. It is questionable even whether Mill himself really believed that his empiricism supported his ethics, since he seemed to make a distinction between the certainty about causal laws required as a minor premise of ethical judgement and the utilitarian principle of the summum bonum that is required as a major premise. If the System of Logic makes it fairly clear that the former was taken as proven, his statements about the latter in Utilitarianism are much more equivocal.

Questions of ultimate ends are not amenable to direct proof. Whatever can be proved to be good, must be so by being shown to be a means to something admitted to be good without proof….[but]…. There is a larger meaning of the word proof, in which this question is as amenable to it as any other of the disputed questions of philosophy. The subject is within the cognisance of the rational faculty; and neither does that faculty deal with it solely in the way of intuition. Considerations may be presented capable of determining the intellect either to give or withhold its assent to the doctrine; and this is equivalent to proof.[16]

 

As Ryan points out, the appeal here is not so much to a notion of proof as to one of rationality, and it is this rationality which connects Mill’s epistemology to his ethics[17]. The resemblance between the rationality of the two spheres, for Ryan, lies in the way in which a general principle (causality in epistemology, utilitarianism in ethics) is assumed as a rational basis. Ryan argues convincingly that those later thinkers, such as G.E. Moore, who accused Mill of the “naturalistic fallacy” of confusing facts and values missed the fact that Mill also clearly believed in this distinction[18]. The universality of Mill’s utilitarianism seems thus not directly based on empiricism so much as an analogy to it, though the possibility of the analogy as a way of making the foundations of ethics intelligible nevertheless depends upon the prior empiricist claims. Mill’s “equivalent to proof” here is rather reminiscent of Schopenhauer’s “metaphorically or figuratively”: in each case it seems clear that the author both wants to assert a universal and prescriptive ethics (which, practically speaking, both do, since neither allows their hesitancy to influence the nature of the prescriptions that follow) and maintain the factual integrity of their appeal to science. In each case the claim to knowledge in the scientific sphere forms a bulwark of assumptions without which universality of ethics could not be claimed, and without which there would be no appropriate object to make the analogy to, even if the exact nature of the relationship between the two compared entities in the analogy is more arguable.

 

Ryan’s argument, if correct, also has the interesting implication of showing the universality of Mill’s utilitarianism to be just as intuitionist as that of later utilitarians, whilst his assumptions about consequentialism remain just as empiricist in their hands. Mill’s “considerations” as offered in Utilitarianism involve an appeal to happiness alone, without any convincing argument as to why it should be universalised. We are offered only a non sequitur:

No reason can be given why the general happiness is desirable, except that each person, so far as he believes it to be attainable, desires his own happiness. This, however, being a fact, we have not only all the proof that the case admits of, but all which it is possible to require, that happiness is a good: that each person’s happiness is a good to that person, and the general happiness, therefore, a good to the aggregate of all persons.[19]

 

Why should each person’s desire for their own happiness imply the value of the aggregate good? Mill clearly does not succeed in avoiding an intuitive assumption here, despite his expressed desire to do so. Perhaps, as I shall be arguing throughout, it is possible to bridge this gap by questioning its basic dualist premises, but without any such questioning, the mere intuitive assertion of universality can only be alienating to an ego when applied in any sphere where it does not already include an idea of “general” good within its sphere of identification. Even when the ego succeeds in identifying with the idea of general happiness, much more than this argument is required to enable it to actually identify with the happiness of other persons.

 

Sidgwick’s approach to this problem is perhaps rather more honest. He asserts the principles of utilitarianism on intuitionist grounds, but then also recognises the practical irrelevance, even counter-productiveness, of any such proof.

There are certain absolute practical principles, the truth of which, when they are implicitly stated, is manifest; but they are of too abstract a nature, and too universal in their scope, to enable us to ascertain by immediate application of them what we ought to do in any particular case; particular duties have still to be determined by some other method.[20]

 

Sidgwick thus turns to “the Morality of Common Sense” as the practical basis of ethics, with the argument that it has a symbiotic relationship with utilitarianism: utilitarianism provides a basis for criticism and amendment of common sense morality and an overall justification for it, whilst common sense morality provides a practicable basis of action with which people already identify[21].  Sidgwick’s honesty consists in the recognition of the role that the conventional plays in ethics, without the abandonment of all claims to universality. But in effect he reproduces a familiar pattern in eternalism, that of a radical discontinuity between an inaccessible absolute and an accessible but unjustified conventional ethics. By appealing to utilitarianism to justify common sense morality Sidgwick makes some large assumptions about the utility of conventional ethics which suggest that his motive, like that of Kant and other eternalists, was ultimately to provide a rationalisation for conventional ethics rather than a genuinely independent ethical standpoint capable of criticising social practices. This reflects the difficulties in distinguishing philosophical intuition from a common sense version without appeal to some further basis of judgement. Mill, although subject to the same difficulty, at least tried to create the foundations of a genuine critical position.

 

After Sidgwick and Moore (who held a similar position), utilitarian thinkers seem little inclined to seek any real epistemological grounds for their ethical position, going little further than Mill’s non sequitur and thus offering no way of persuading anyone from egoism. In this they generally reflect scientistic meta-ethical positions which I shall be discussing in the next chapter[22]. Utilitarianism has thus preserved its universalist form but, oddly, no universalist grounds. It has effectively thus restricted its operation to spheres in which universalist, or at least aggregative, motives exist already.

 

iii)                Compatibilism

 

I have already argued that Mill’s utilitarianism shows many of the standard features of dualism and eternalism: representationalism (from which linguistic idealism follows), cosmic justice and ethical foundationalism. I also want to argue that although later utilitarians do not appeal to these points as Mill does, in making some of the same assumptions about universality they continue with some of the same eternalist errors. This argument is bound to be complicated by the close relationship of these continuing eternalist assumptions with nihilist ones which will only be discussed in detail in the next chapter, after which the picture should become fuller and clearer. For the moment, however, I want to consider the status of one remaining feature of eternalism – that of freewill – again as it appears in Mill. Again I shall argue that the problems evident in Mill’s utilitarianism are very difficult for his successors to avoid.

 

Mill is a compatibilist in the sense that he believes in the causal determination of actions and in our capacity to influence those determinations. His determinism is described in terms of predictability.

Given the motives which are present to an individual’s mind, and given likewise the character and disposition of the individual, the manner in which he will act might be unerringly inferred;…if we knew the person thoroughly, and knew all the inducements which are acting upon him, we could foretell his conduct with as much certainty as we can predict any physical event.[23]

 

His understanding of freewill, however rests on a distinction between cause and compulsion.

The causes…on which actions depend are never uncontrollable, and any given effect is only necessary provided that the causes tending to produce it are not controlled.[24]

 

Mill is not particularly explicit about exactly what is doing the controlling. He believes that we act in accordance with our characters, and evidently that it is these characters which make our actions predictable, yet also asserts that “we are exactly as capable of making our own characters, if we will, as others are of making it for us”[25]. This power, however, apparently consists only in a sort of veto over volitions that have already arisen, whereby we either allow our characters to proceed in the determined way, or interrupt them and thus change their nature. But Mill does not explain why no volition is required to interrupt the volition (or indeed to let it pass): if we allow a free volition here, then of course it is without sufficient cause and a metaphysical freewill has been allowed.

 

It thus appears that Mill is so committed both to a belief in determinism as predictability and to a belief in freewill that he is prepared to pursue these beliefs to the point of clear self-contradiction. Ryan casts some light on Mill’s motives:

Mill believed the establishment of a science of society to be an essential step toward implementing rational policies of social change. If determinism took away our freedom of choice, so it seemed to him, there could be no place for science as a guide to choice; we should simply be helpless spectators of a future which we knew to be inevitably ours. But if the reality of choice eliminates determinism – and this for Mill was as much as to say that it eliminates science – then it seemed that choice could not be guided by science. So Mill was forced to show that a scientific sociology and psychology could in no way impair our freedom of choice.[26]

 

It is perhaps here that the contradictions of utilitarianism become most clear. Mill had good reasons for believing that the universality of utilitarianism depended on the fact of its being able to appeal to universal scientific laws of causation, yet the assumption of the possibility of rational choice inherent in any utilitarian calculus implied a requirement of freewill. If Mill had compromised on the predictability of human behaviour he would also have had to compromise on the universality of utilitarianism. It is difficult to see how any utilitarian who sincerely believes in universality[27] can avoid this problem, unless he derives the universality from some other source than empiricism: but this seems to mean leaving the sphere of utilitarianism for that of some other eternalist doctrine.

 

The difficulty here is based on the metaphysical assumption that we must “know” either freewill or determinism to be the case. Once it is assumed that one knows one of them to be the case, it often seems, in the eternalist tradition, that the other follows. This is perhaps because of the inevitable tension between cosmic justice and freewill: no eternalist doctrine is thus entirely free of the embarrassments of compatibilism, even if none makes it quite so obvious as Mill. This is because where most eternalist doctrines give freewill a foundational importance, most avoid the recognition that determinism is also required for cosmic justice. Other eternalist doctrines avoid this recognition by the use of some philosophical device: in Christianity this is the distinction between divine determinism and divine foreknowledge, in Kant and Schopenhauer the restriction of freewill to the noumenal realm and determinism to the phenomenal. In Marx the opposite strategy is employed of stressing the determinism of cosmic justice and theoretically denying freewill. But all of these merely avoid facing the contradictory nature of their doctrines, which emerges when the implication of the denied metaphysical assumption is drawn out.

 

Thus in utilitarianism, we perhaps reach the breaking point of eternalism. In response to the development of capitalism, the “invisible hand” of cosmic justice has moved entirely to the sphere of human activity, where it appears entirely comprehensible by human beings. At the same time, the development of science has come to make it increasingly difficult to rationalise belief in an independent source of freewill such as the soul. The metaphysical mystification of the eternalist tradition thus becomes harder and harder to sustain. From this point, the assertion either of freewill over determinism (to such an extent that any objective basis for the relative predictability of experience is denied), or of determinism over freewill (to such an extent that the subject of experience is denied) becomes difficult to avoid. The road forks into these two sorts of nihilism, with only the faint, barely-trodden track of non-dualism between them giving any hint of an alternative route.

 

iv)                Consequentialism and alienation

 

I shall now turn to the fuller substantiation of the claim I made earlier, that utilitarian consequentialism itself requires a dogmatic empiricism. That this is so can be seen more fully if it is considered in the broader context of the possible dualism or non-dualism of all types of consequentialism, and in relation to the psychology of the consequentialist process. In particular it can be linked to the philosophical debate which is already in progress about whether or not utilitarianism is alienating, since I shall claim that a feature of utilitarianism which has already often been identified as a weakness – its impersonal rationalism – is inseparable from its dogmatism. In the course of my discussion the empiricist nature of this dogmatism should also emerge.

 

It may help to begin with some aspects of the modern debate about whether utilitarianism is alienating, to make it clearer how I wish to resolve the problems which form the basis of this debate by changing the contextual assumptions. I shall discuss two representative contributions to this debate, by Bernard Williams arguing against utilitarianism[28] and by Peter Railton defending consequentialism (of a type which could well be utilitarian)[29].

 

Williams attacks the impersonal rationalism of utilitarianism by using the well-known example of “Jim and the Indians”, a classic dilemma of the type in which consequentialist and deontological ethical responses conflict. Twenty Indians have been sentenced to die unjustly, but Jim is confronted with a choice between shooting one Indian himself, which will enable the reprieve of the other nineteen, and watching all twenty be killed by another. In this kind of example, understood merely as a rational choice between alternatives, the utilitarian obviously has a strong case for saying that Jim should put his “squeamishness” aside and do what appears wrong for the greater good. Williams argues, though, that this is not obviously the case, because of the ways in which the consequentialism of the utilitarian approach creates a fixed context for Jim’s judgement which ignores the separate agency of others (such as that of the soldier who will kill the twenty Indians if Jim does not kill one) and underestimates the central importance of personal integrity to our ethical motivation (according to which whether he kills one Indian is a matter for Jim’s integrity, and whether he kills twenty a matter for the soldier’s). Williams argues that such psychological considerations are not treated in proportion to their ethical importance for us if they are incorporated into the utilitarian scheme, as many utilitarians have suggested, as they there become only one relatively insignificant factor to be weighed among many heavier points of external fact. This reasoning leads Williams to the view that there is such a marked incompatibility between human ethical propensities and the direct use of utilitarianism as a method of deliberation, and that if it were generally employed, “it could lead to disaster”.  Since the absolute viewpoint requires a better practical mode of deliberation, “utilitarianism’s fate is to usher itself from the scene”[30]. A utilitarianism which “retires to the totally transcendent standpoint” and is never actually used for deliberation, he concludes, has become a complete irrelevance.

 

Williams’ position can be contrasted with Railton’s, which begins by recognising alienation as a problem in the deliberative use of consequentialism, but maintains that it can be solved within the consequentialist framework. The approach that he calls “objective consequentialism” (and which Williams calls “indirect utilitarianism”) – effectively a reversion to Mill’s view that it is the consequence itself, considered quite apart from the agent’s knowledge of it, which is the seat of moral value – can, he claims, encompass the avoidance of alienation through psychological sophistication, recognising points where undue consequential deliberation is alienating and therefore self-defeating. He also maintains that consequentialism would not be refuted in a world where it was never consequentially advantageous to use it in deliberation, and that this would not show it to be self-defeating, but on the contrary show “that objective consequentialism has the virtue of not blurring the distinction between the truth-conditions of an ethical theory and its acceptance-conditions in particular contexts, a distinction philosophers have generally recognised for theories concerning other subject matters”[31].

 

This debate should benefit from being put into the context of my discussion so far. Firstly, its implications are not limited to the question of the acceptability of utilitarianism, because similar examples to Williams’ could be constructed in relation to the impersonal demands on any absolute system of ethics in relation to his personal integrity. If we merely adapt Williams’ example to the discussion of a deontological ethic[32], we might have equal doubts that it was obvious that Jim should not kill the Indian because it was not in accordance with, say, Christian or Kantian ethics to do so, regardless of the consequences. He might have actually felt it was more in accordance with his personal integrity to perform the killing. The problem Williams identifies is thus not just one specific to consequentialism, but one general to ethical foundationalism, and its adoption of absolute standards in relative judgements. In Jim’s specific situation, this means that, despite his ignorance of the actual intentions of others, the further factors that may bear on the situation, and the actual long-term effects of his actions, utilitarianism, like other eternalist philosophies, urges him to reach a decision based purely on the rational deliberation of the limited and distorted evidence available to him.

 

In response to this general difficulty of eternalist ethics, however, Williams’ appeal is only to the way in which we actually make ethical choices and to the distance between utilitarian prescription and this actuality. He does not offer us any alternative ground of prescription, and in this sense does not address the concerns of utilitarians. Elsewhere, Williams suggests that there is no alternative ground of prescription, at least as a definitive philosophical formula[33]. This does not, however, release us from the requirement for clear grounds of judgement in  particular cases that the utilitarian demands[34].

 

In contrast, Railton’s appeal to truth-conditions as recognised “for other subject matters”(appealing only to conventional judgement in those spheres) can be seen as a dogmatic empiricism employed, like that of Mill, in the defence of the eternalist absolute. The implication seems to be that Railton would continue to adhere to absolute consequentialism in a universe where God was its only deliberative practitioner: a situation, in fact, where faith in consequentialism has exactly the same relationship to experience as faith in God. And what would it mean for Railton himself to have such a belief in such a universe? To be consistent, he should submit to having it erased from his mind in order to serve the greater good. Railton’s appeal to an entirely abstract perspective, like other metaphysical perspectives, cannot be conclusively refuted, but its lack of relationship to experience makes it entirely useless as a foundation for ethics.

 

Secondly, investigation is needed into the concept of alienation that is being used in this debate. Williams does not use the term “alienation” at all but suggests its converse in his use of “integrity”. This integrity has some relationship to Williams’ use of “character” elsewhere as a description of a person’s “projects”, i.e. consistent desires which maintain one’s reason for existing[35]: he may mean something similar to the ego, perhaps the relatively integrated ego, here. Williams’ criticisms imply that utilitarianism can create alienation in the sense that it recommends the making of ethical judgements based, not on this integrity, but on purely abstract rational criteria which may require the complete suppression of this integrity: an view consistent with my account of alienation[36].

 

However, there is a danger in following Williams’ approach that the dynamic character of the ego may not be sufficiently taken into account in one’s view of “integrity”. If integrity is not to be a fixed psychological entity it can only be understood as a temporally-changing process: as integration rather than integrity. But what is integrating with what (overcoming the alienation of what from what)? If the answer is to be put in terms of “reason with emotion”, ”conscious with unconscious”, or (in the terms I have been using) ego with alienated psyche, the rational analysis of the deliberative utilitarian is itself part of what needs to be integrated. In other words, the fact that a rational consequentialist case encompassing psychological processes (along Railton’s lines) can be made is a fundamental factor (though not the only factor) involved in any integration. One strength of Williams’ argument is thus that it recognises that this rational analysis alone is insufficient either to motivate, guide or justify ethical judgement, but the corresponding weakness in his approach lies in the danger of falling back into a more or less anti-rational position based largely on convention or intuition, where the challenging role abstract and universalising reason is insufficiently appreciated. I do not mean to imply here that Williams himself (whose overall position is a subtle one) necessarily does this: but I shall give further support in the next chapter[37] to the view that the Aristotelian approach on which he generally relies is at least subtly nihilistic in its definite rejection of the universalist claims of the utilitarian.

 

Railton, for his part, makes use of the term “alienation” whilst taking praiseworthy care not to petrify the psychological terms on which its understanding rests:

It seems to me wrong to picture the self as ordinarily divided into cognitive and affective halves, with deliberation and rationality belonging to the first, and sentiments belonging to the second. John’s alienation is not a problem on the boundary of naturally given cognitive and affective selves, but a problem partially constituted by the bifurcation of his psyche into these separate spheres.[38]

 

The only problem with this view is that Railton does not extend its implications consistently to the standpoint of consequentialism itself. He assumes that there is still a “naturally given” neutral standpoint from which to make consequentialist deliberations, despite the absence of such a standpoint in the actual examples he discusses. There is thus, for him, still a wholly consequentialist solution to the problem of alienation, despite the fact that he is prepared to admit that the deliberative use of consequentialism involves the bifurcation of the psyche he mentions.

 

The debate between Williams and Railton here can thus not be resolved at its own level, at least insofar as it is true (as it appears to be) that each of them is actually committed to a metaphysical position which is opposed to the other. Railton is committed to an ethical foundationalism based on supposed scientific truth (or at least an analogy to it, like that of Mill). Williams, on the other hand, whilst not denying the possible value of some consequential calculation, appears to deny the value of seeing ethics in a universal light altogether, primarily on the ground of the psychological effects of doing so. The debate thus provides another example of the fruitless struggle between eternalist and nihilist.

 

However, the debate does point in the direction of the key issue of the role of the agent in consequentialism. Both utilitarianism and Williams’ criticism of it are inadequate because they oversimplify the position of the agent. Utilitarianism has to assume an agent which is both fixed and isolated from other agents around it, making a decision the justification of which can only be considered in rational abstraction from the formative processes of decision-making, the ways in which the psychological propensities of the agent influence real decisions, the confusions of actual states of belief, and the instinctual accommodations we make to others which can not be reduced to the rational elements of a single calculation. The solution, however, is not to make the converse negative assumption that universal prescriptive value cannot be extracted from the whole set of conditions with which the agent engages: for this too over-simplifies the position of the agent, ignoring the respects in which it is actually part of the nature of sustained agency to change in order to meet justifiable demands coming from beyond the limitations of one’s current understanding[39].

 

In order to make a utilitarian calculation which avoids these oversimplifications and incorporates them into an objective demand in the way Railton would like, the mind of the agent needs to be capable of extremely subtle reflection. It needs not merely rational clarity, but awareness of the broad range of relevant factors, awareness of its own psychological states, sympathetic awareness of those of others, and openness to the possibility of new factors which have not yet been considered. Most importantly, it needs to be able to judge when it is appropriate to make a judgement based on its own limited knowledge in order to be able to act speedily, or when to avoid doing so by deferring judgement or deferring to authority.

 

For example, in the case of Jim and the Indians, Jim needs to decide first of all whether he really has enough knowledge, awareness and psychological balance to be justified in making a decision based on an attempt at direct utilitarian calculation. He may, for example, be aware of the fact that he is so overcome by fear in that situation that his judgement is likely to be badly affected. In such circumstances he might decide to fall back on what Sidgwick calls “common sense ethics”, or perhaps to the authority of expert advice[40], even if these can only be applied to the complexities of the situation he faces with the relative crudity of a command not to kill. The second-order justification for doing so, however, will not be the supposed knowledge of any higher rational authority that such a recourse will necessarily lead to the best consequences, only that Jim’s own degree of moral objectivity receives its strongest expression through this method.

 

The advantage of consequentialism over other forms of eternalism is thus that, where the conditions of consequential calculation are themselves optimal, it provides a relatively subtle instrument of moral decision-making. Since the amount of knowledge included in a consequentialist calculation is potentially infinite, the extent of the subtlety of the instrument is only limited by the capacity of the mind or minds performing the calculation. I therefore want to argue that the usefulness of consequentialism as a method varies in a direct ratio to the (psychological) objectivity of the mind(s) performing it. The lower the degree of objectivity, the more inclined the person using consequentialist method will be either to dogmatism or scepticism in their assessment of actions, and hence the more likely they are to choose actions which, based on false beliefs, do not actually lead to the desired ends. Since alienation is also directly associated with dualism either in its dogmatic or sceptical forms (as the ego rejects the unknown beyond its established belief-structures, it also narrows and fortifies its attention on one area of experience), a non-alienated use of consequentialism must thus be a non-dualist use of consequentialism.

 

I thus conclude that consequentialism, as an important aspect of utilitarian thought, is not itself intrinsically dualist (nor intrinsically alienated). It does, however, need to be extracted from the dogmatic assumptions associated with the utilitarian appeal to the absolute in order to be used in a non-dualist fashion, and when this is done it becomes apparent that other methods of making ethical judgements may well reflect the objectivity of the agent better in some circumstances.

 

What this leaves open is the question of how far the desirable consequences employed as ends in utilitarian calculation are also dualist, and what kind of idea of desirable consequences might be compatible with non-dualism. This is a question with many ramifications, which I shall be tackling throughout the rest of the book. It cannot be answered without first considering the whole nature and justification of moral prescription from a non-dualist point of view.

 

In the critical consideration of utilitarianism it may also be thought that I need to discuss whether hedonism provides a justifiable starting-point for the utilitarian analysis of desirable ends. I do not intend to do this, because in practice it appears that utilitarianism, because of its absolutism, is rarely if ever hedonistic in any meaningful sense[41]. If my arguments above are correct , it is the absolute nature of utilitarianism, rather than its supposed hedonism, which provides it with a prescriptive value: so that in making a criticism of hedonism and then applying it to utilitarianism I would merely be repeating Mill’s non sequitur. I will however be discussing hedonism in the next chapter in relation to empiricist ethics.

 

v)                  Utilitarianism and politics

 

Before moving on from the topic of utilitarianism I wish to discuss how the features I have referred to so far can be related to tendencies in political thought. Related to this is also the question of whether the justification of utilitarianism as a political philosophy is different from its justification as a theory of ethical judgement by the individual.

 

In subsection (i) above I discussed the relationship between utilitarianism and the economic belief in the “invisible hand”. This forms one sort of cosmic justice belief of the kind found in all eternalist doctrines, having like other such beliefs a dogmatic basis. It differs from other such beliefs, however, in having a direct political applicability. A philosophy which has as its basis the belief that particular types of individual human action, regardless of motive, can produce an absolutely desirable result will naturally promote those particular types of individual human action, whatever it takes them to be, and whether the promotion is of a direct or an indirect type. The state has a power to produce such activities in a way in which it does not have power to produce Christian faith or Kantian good will, and indeed the responsibility for producing the desirable results through corporate control of the population might (consistently with utilitarianism) even be laid entirely before the state.

 

Since the nature of the claim to moral authority has changed in utilitarianism relative to earlier forms of eternalism, however, so has the kind of political view created by that claim. Whilst the dogmatic claims of earlier forms of eternalism were associated with supporting groups (formed by shared allegiance to an authority) and thus tended to be either conservative or radical, those of utilitarianism are primarily associated with empirical evidence, with a political group created around the authority of a particular way of interpreting the evidence. Since empirical evidence, together with interpretative arguments, can be offered for almost any political position, utilitarianism thus has great flexibility in its political application. As Alan Brown concedes “Both laissez-faire capitalism and state-controlled socialism can be given seemingly plausible utilitarian justifications”[42]. Utilitarianism thus serves to unite groups otherwise separated in values who are prepared to agree on interpretations of evidence and work together in what they believe to be the public interest.

 

It is probably misleading to talk merely in terms of the utilitarian influence over Western democracy, since in many respects utilitarianism is Western democracy: at least, the mode of justification of utilitarianism is the dominant mode of justification for the actions of Western democratic government. In almost every case of proposed legislation from every party, the appeal for its justification comes, not directly from its relationship to tradition or to an abstract principle, but from a prediction of beneficial effects upon the general public. Even those political arguments which involve a substantial appeal to some apparently non-utilitarian principle, such as retribution in punishment, are rarely free of  subsidiary utilitarian arguments that the proposed action will also lead to good results, simply because politicians need to persuade as wide a section of people as possible.

 

In gaining such wide currency, however, utilitarianism may also have been diluted to such an extent that its eternalist nature is no longer obvious. Dilution occurs as the agents appealing to utilitarian justifications for their actions are no longer individuals seeking the objective rational support of universalism, but the representatives of groups making compromise deals. Whilst an individual may hold an uncompromised ideal of universality, as soon as this is negotiated against the ideal of another holding different universalist beliefs, the level of discussion becomes reduced to that of concrete group interests. The resulting political decision may still be justified in prescriptive moral terms, but has actually been constructed on the basis of little more than an extended conventional morality which combines the interests of different groups. In this political slide from utilitarianism to empiricist relativism we can again see the discontinuity of eternalism at work.

 

I do not intend to offer detailed examples to justify these claims, since this would take a disproportionate amount of space. However, two brief representative examples may be useful, both of which illustrate an ambiguity of political goals among Western democracies where they have offered universalist arguments of a utilitarian type for actions which could also be interpreted as manifesting corporate self-interest. One example could be the Gulf War of 1990-91, which was taken by different commentators to be motivated either by a requirement to maintain international order or by a threat to oil supplies to the West. Another might be the continuance of the “first-past-the-post” system for general elections in the UK, the advocates of which defend it on broadly utilitarian grounds, whilst they are accused by detractors of mere self-interest. In both of these cases it would probably be an over-simplification of a complex issue to suggest that the universalist reasoning given is obviously unjustifiable, or that the self-interested motivation is the “real” one. In each case, though, the universalist argument attempts to confine the discussion to the sphere of the interpretation of empirical evidence used in an attempt to support predictions of the consequences of different policies,  without considering the more fundamental question of why we may be disposed to interpret evidence in different ways and how these differences might be resolved.

 

Thus the application of utilitarianism to politics appears to support a more subtle version of the eternalist authoritarianism to which it is related: one which relies on the dogmatic interpretation of empirical evidence to provide a basis of political authority, rather than the more direct manifestation of dogmatism. But if this version is more subtle and at least makes use of empirical evidence, is it at least relatively more ethically justifiable? In assessing this it is helpful to apply the distinction I made in the last sub-section between dualist and non-dualist consequentialism. If, as I want to claim, the moral usefulness of consequentialism varies in a direct ratio with the moral objectivity of the person using it, some general conclusions about the usefulness of utilitarian consequentialism in politics can be drawn from considering how far politics is conducive to moral objectivity.

 

The circumstances of a political decision clearly differ from those of a personal one in important respects. Firstly, the degree of information available about the likely consequences of legislation or of an executive decision (at least in a modern Western democracy) is likely to be higher, due to the employment of researchers. Secondly, the decision is likely to be made collectively rather than individually, providing a check on the distortions and prejudices which may be held by one individual but not others. Both these reasons can be advanced as making utilitarianism better-suited as a method of political decision-making than of purely personal decision-making, since they make the estimation of consequences more likely to be correct. I want to argue, however, that this reasoning can only be successfully advanced in favour of the political application of non-dualist consequentialism and not of the dualist consequentialism represented by utilitarianism as I have characterised it.

 

The availability of more information only makes a difference to decision-making insofar as the framework of belief within which the decision-making takes place is capable of modification to take into account the new information. As I commented in chapter 2 in relation to scientific theories, successful new theories are created in response to new information when the old theory no longer explains new data, but a new modification is available which explains both old and new data  and is fruitful in predicting further observations. There seems little doubt that such shifts do occur in politics, or political policy reforms would never occur. However, they do not occur simply in response to new information, since this can be always be dismissed or re-interpreted: they also require sufficient moral objectivity on the part of politicians to modify their view.

 

Neither does the collective nature of political decision-making necessarily produce greater objectivity, since the objectivity of a group is not necessarily greater than that of an individual. As I shall argue in more detail in 6.d, we cannot assume either the supervenience of groups over individuals, or the independence of individual judgement, either of which would involve metaphysical premises. Whilst group decision-making may in some circumstances increase objectivity, then, it will not necessarily do so in all. Where it does so, this will be due to the relative objectivity of the group[43]

 

In the case of both these arguments, then, the claimed advantages of utilitarianism to political decision-making do not apply where utilitarianism is understood as involving the dogmatic assumption of universalist values. Rather, these advantages are those of consequentialism used as a method of decision-making by those with relatively greater moral objectivity. The political process cannot be exempted from the wider moral requirement for moral objectivity, even if when it is present consequentialism often (though contingently) has some clear advantages over other methods of decision-making in that context as better reflecting the objectivity of those involved.

 

l)      Intermediate conclusion

 

Here I reach the end of this lengthy account of the historical manifestations of eternalism in the West. As an account it is obviously limited in scope and detail, but I have tried to include all the most influential manifestations of eternalism. If there are still some glaring omissions (e.g. Epicureanism, Medieval Christian thought, Jewish and Islamic thought, Leibniz, Spinoza, and many modern refinements of Christianity, Kantianism, Hegelianism, Marxism and utilitarianism), I think I have probably provided some resources on the basis of which an explanation of the eternalism of these doctrines could be constructed.

 

In the course of the survey I hope to have shown not only the continuance of the basic features of dualism and of eternalism as they were proposed at the outset, but to have identified some general reasons for the failure of eternalist ethics, to demonstrate which is the main feature of this chapter. I shall now briefly try to draw out some of the ways in which this failure has manifested itself.

 

1. Eternalist ethics have been unconvincing. Their reliance on Dogmatism and construction of a metaphysical absolute means that any sceptical examination reveals no particular reason for adopting them.

 

2. Due to their unconvincing nature eternalist ethics have generally been reliant on group support. Eternalist assumptions could only seem self-evident when social conditioning made them so.

 

3. Due to their reliance on group support, eternalist ethics have been subject to a discontinuity between the ethical absolute supported by their metaphysics and the conventional ethics created by the group. This discontinuity undermines any incremental improvement in ethical dispositions.

 

4. The belief in an ethical absolute, and the fortification of the ego in support of that belief, has created alienation from the rest of the psyche. Such alienation is even compatible with entirely conventional ethical behaviour, but has been associated with fanaticism, literalism, perfectionism and conflict. Even in its more moderate manifestations it is associated with a rationalism which creates an unnecessarily narrow focus for decision-making. This alienation has generally undermined the practice of ethics.

 

5. The reliance on group support, and the tendency to alienation, have created a tendency to authoritarianism in the political application of eternalism. At the extreme this can mean authoritarian conservatism or radicalism, or even in more moderate versions a rationalistic over-attachment to verbal formulations of the universal which influences political decision-making.

 

Whilst these weaknesses have led to a decline, or at least a moderation, of eternalism, in the last few centuries, this is largely only to be replaced by nihilism in certain spheres. Often the sole dominance of eternalism in the medieval world has now been replaced by systems of symbiotic co-existence between eternalism and nihilism:  for example, theism and atheism, absolutism and relativism, utilitarian economics and consumerism. Eternalism thus continues to play an important role in maintaining the conditions for the existence of the nihilistic views that attempt to supplant it, even if this is a role of counter-dependency. I shall be saying more about these symbiotic relationships in relation to the features of nihilism that I shall be exploring in the next chapter. The continuing influence of eternalism, in more and less explicit forms, is thus not to be underestimated.

 

Every aspect of eternalism on which I have commented, however, can be seen positively, as a potential contribution to the Middle Way, as well as negatively. As I have been indicating throughout the historical survey, there are respects in which we need a vision of universalist ethics, we need some provisional ideas of an ethical foundation, of freewill and cosmic justice. In some respects even the alienation associated with eternalism simply marks a necessary stage in the development of the psyche from infantile immediacy to the development of a capacity for abstract thought and egoistic self-consciousness. The problematic feature of eternalism consists only in the assumption that this stage of development is the apex of human achievement and cannot be surpassed. Much more will be said in chapters five and six about the ways in which suitably qualified aspects of eternalist approaches can contribute positively to the Middle Way.

 



[1] See  3.f.x

[2] See 4.c

[3] Smith (1976) p.456

[4] Raphael (1985) p.73

[5] Mill (1962) p.266

[6] ibid. p.273

[7] E.g. in Scarre (1996) p.11: “the world would have been spared a vile dictator if the infant Hitler had been smothered by his nurse, but the killing would still have been wrong, from a utilitarian perspective, unless the nurse had possessed remarkable powers of clairvoyance.”

[8] In accordance with my general hypothesis in 2.c.i

[9] Mill (1970)

[10] ibid. p.74

[11] ibid. p.79-81

[12] Britton (1953) p.151

[13] Mill (1970) p.247-266

[14] Ryan (1987) ch.4

[15] This does not mean that other forms of justification for utilitarianism (or at least empiricist ethics) have been abandoned: see for example Hare, whom I discuss in 4.d.v

[16] Mill (1962) p.254-5

[17] Ryan (1987) p.188-9

[18] ibid. ch.11

[19] Mill (1962) p.288-9

[20] Sidgwick (1962) p.379

[21] ibid. p.460-474

[22] See 4.d

[23] Mill (1970) p.547

[24] ibid. p.549

[25] ibid. p.550 (Mill’s emphasis)

[26] Ryan (1987) p.104

[27] More modern utilitarians have often followed the Humean route of claiming the compatibility of determinism and freedom, and denying anything other than an illusory relationship between our experience of choice and the requirement for an absence of sufficient causality. But this form of compatibilism lets go of the eternalist metaphysics which supports freewill and effectively follows a nihilist route, the grounds of which will be explored in 4.c & d.

[28] Williams (1973)

[29] Railton (1984). Brink (1986) offers a very similar argument.

[30] Williams (1973) p.134

[31] Railton (1984) p.116

[32] Williams himself does something similar to this, in relation to Kantianism, in Williams (1981) p.1-19.

[33] Williams (1993) p.200

[34] See 8.a & b for an account of how non-dualism might fulfil this demand.

[35] Williams (1981) p.5-14

[36] See 3.b.iv

[37] See 4.b.ii-iv

[38] Railton (1984) p.97

[39] The crucial link on which I rely here, between the incremental transcendence of any given state of mere egoism and provisional universality, is one which I shall be developing throughout ch.4 and Part 2.

[40] See 8.a

[41] See 4.a.iii

[42] Brown (1986) p.48

[43] See 6.d.ii for some explanation of how groups may be relatively objective.

 

Return to thesis contents page

Return to moralobjectivity.net home page

 

A Theory of Moral Objectivity: quick links to other sections

Contents

1. introduction

2a. Psychology of belief

2b. Heuristic process

2c. Psychology & philosophy

3ab. Eternalism

3cd. Plato

3e. Stoicism

3f. Christianity

3g. Kant

3h. Hegel

3i. Marx

3j. Schopenhauer

3kl. Utilitarianism

4a. Nihilism

4b. Scepticism & Aristotle

4c. Hume

4d. Analytic Philosophy

4e. Wittgenstein

4f. Pragmatism

4g. Nietzsche

4hi. Existentialists

5. Integration

6. Philosophical Problems

7. Normativity

8. Middle Way Ethics

9. Conclusion

10. Appendix

Bibliography

 

Other books:

A New Buddhist Ethics

The Trouble with Buddhism

 

List of all books and papers

 

Objections & responses

 

Comment on the blog

 

Home page