Happiness as the Only Intrinsic Value

Abstract

As happiness is directly experienced by the individual as valuable, its normative value needs no additional justification. Things like institutions and moral principles may be used to promote happiness directly and indirectly. In time, they may be mistakenly valued for their own sake, while ultimately their values should be based on their contribution to happiness. We are born, brought up, and socially influenced to have certain preferences which are largely consistent with our own happiness. Where they diverge, then apart from the effects on the happiness in the future and of others (hence really no divergence in the longer/wider perspective of happiness), ultimately it is happiness that is really consistent with rationality. Arguments for happiness as the only intrinsic value are made and defended against objections or opposite arguments (including ‘Rather be unhappy Socrates than a happy pig’, Kant's categorical imperatives, Rawls’ maximin, etc.). The apparent acceptability of these opposing positions is due either to our imperfect rationality and/or inadequate account on the effects in the future and on others.

Our argument in the previous chapter for preferring happiness when it diverges from life satisfaction is based on the belief that happiness is ultimately speaking, the only thing of intrinsic value. After arguing for this in Sect. 5.1, we also consider some common objections in Sect. 5.2. The non-welfarist position of Kant’s categorical imperatives and ‘anti-utilitarian’ position of Rawls are criticized in Sects. 5.3 and 5.4 respectively. Further objections are discussed in Appendix B.

5.1 Happiness as the Only Intrinsic Value

A number of reasons may make life satisfaction to diverge from happiness. In the previous chapter, we discuss a case where a person A may have low or even negative (net) happiness but yet have high life satisfaction as she believes that she has made significant contributions towards increasing the happiness of others. Whether this belief is correct or not, it can be argued that only her happiness, not her life satisfaction, should be counted in the ultimate social objective. For simplicity, consider only the extreme cases where her belief is either entirely correct or entirely wrong. The intermediate cases are also taken care of by the combination of the arguments for each of the two pure cases.

If A’s belief is incorrect, she did not actually contribute to raising the happiness of other individuals. Her belief that she did so may increase her own happiness. If so, that increased happiness is already counted in the social objective that takes account of her happiness. If A’s belief is correct, she did contribute to raising the happiness of other individuals. The higher happiness levels of other individuals from her contribution are already counted in a social objective function¹ that takes account of the welfare or happiness levels of all individuals. (‘Welfare’ and ‘happiness’ are used interchangeably as a happiness definition of welfare is adopted.)² Knowing or believing that her contribution has made other individuals happier probably makes A happier. However, A’s happiness level may remain low, though her life satisfaction level may be fairly high. It should be the low happiness level rather than the high life satisfaction level that should count towards social welfare. Why?

Suppose a person B does voluntary social work that contributes to the happiness of say some elderly people. Obviously, the higher happiness levels of these elderly people count towards social welfare. Moreover, if B himself gets happiness from the social work and/or from knowing/believing that he contributed to the happiness of these elderly people, his higher happiness level should also be counted in social welfare. Thus, it is not the fact that the happiness of other individuals that A helps to raise has already been included within social welfare that precludes the counting of the higher life satisfaction of A. It is just that happiness should be counted but life satisfaction (at least the part that is not based on happiness but on say pure contribution to others or what may be called non-affective altruism) not; why?

Happiness, either in the form of pleasure of the flesh like eating delicious food or having sex or in the form of spiritual fulfilment, is what the individual directly enjoys and hence is inherently valuable to herself. Each and every individual wants to have a high level of happiness for its own sake. Happiness is valuable in itself. This is self-evident to anyone. (If fact, this trait of being able to enjoy and suffer is so important for our survival that individuals completely without this capability are extremely rare, if they have ever existed.) Thus, we do not need any philosophical arguments to justify this (that happiness is valuable in itself). However, we do have to justify the point that, ultimately speaking, only happiness is valuable and that all other valuable things derive their values ultimately from their contributions to happiness. This is done below.

First, it may be pointed out that arguments on what is good, valuable or ought to be done, etc. belong to the normative sphere.³ In contrast to the positive sphere where statements/judgments may be either true or false in some objective sense, value judgments can only be persuasive or not. Thus, it is not logically possible for one to prove that only happiness is ultimately or intrinsically valuable or to prove the truth of any other value judgment. One could only try to make the arguments persuasive. However, noting that different people have different views on such normative issues, one does not expect complete agreement.

Let us try to see the persuasiveness or even the compellingness of the main point in a number of steps.⁴

Step 1: In an isolated world of no affective sentients, nothing is of any normative significance

Here, an isolated world means a world/universe that is completed isolated from any other existence in the sense not only of having no informational or any other flows between it and others but also of having no any causal connections (including through gravity or any other force) with any other existence. We may conduct our analysis along the line of Einstein’s thought experiments; the question of realism is not a relevant issue. One way to imagine this isolated world is to assume that the real world does not exist. That isolated world is the only existence and hence has no informational or any other causal connections with others. If the whole world has no affective sentients from beginning (if there was a beginning) to end (if there will be an end), it seems clear that nothing is of any normative significance. Here, affective sentients are beings that are capable of enjoying happiness and/or suffering pain/unhappiness. Whether that world is getting warmer or colder, more chaotic or more orderly, etc., nothing will be made better off or worse off. It is of no normative significance.

Step 2: Other things being equal, it is undesirable to inflict pain/unhappiness; it is desirable/valuable to have happiness

If pain/unhappiness is inflicted upon some affective sentient beings without anything desirable, either directly or indirectly, it is clearly undesirable. Similarly, if happiness can be enjoyed without causing anything negative directly or indirectly, it is desirable, as the affective sentient beings (like us humans) can testify to that, at least in principle. This does not rule out the possible desirability of pain/unhappiness that leads to more happiness in the future and/or for others and the possible undesirability of happiness that leads to more unhappiness in the future and/or for others.

Step 3: If something is of normative significance, it must ultimately speaking be due to some effects on the enjoyment of happiness or the suffering of pain/unhappiness.

Comparing Steps 1 and 2 above, it can be seen that, if nothing is of any normative significance in a world of no affective sentients, then in a world with normative significance, the normative significance must be due, directly or indirectly, to the affective feelings (happiness and/or unhappiness) of the affective sentients, since this is the only difference between the two cases.

Consider the situation illustrated in Insert 1 where a host of factors may directly or indirectly affect the affective feelings (happiness and pain) of sentients. These factors and doing things that may affect these factors may thus be important or being of normative significance from Step 2 above. However, if the top box (affective feelings) does not exist right from the beginning and to eternity, then we are in the world of Step 1 and nothing is of any normative significance. Thus, in the world where affective feelings exist and something may be of normative significance, the normative significance must be based on or due to, at least ultimately speaking, on the effects on the normatively important affective feelings (top box in Insert 1).

Consider an analogy illustrated in Insert 2 where the final profits of a firm is shown in the top box. The analogy is not perfect because things other than the final profits may be significant, even just to the firm (or its owners). Thus, to make the analogy hold, we have to either assume that things other than profits are not significant, or hold all other things that may be significant unchanged in the comparison, as indicated in the top right box. For example, if the firm is concerned with the welfare of its employees over and above the effects on its final profits, the welfare of its employees must be among those factors that are held unchanged in our comparison. Apart from these factors that are held unchanged, many variables may affect the final profits of the firm directly or indirectly. Assume that, apart from those factors held constant, the firm is only concerned with its final profits and not, say with its cash flows and sales except for their effects on profits. (These effects are usually on the future period. However, to abstract away from the complication of dynamic illustration, we take all the effects as occurring simultaneously or in the a-temporal framework of Insert 2.) Then, any combined changes (e.g. some decrease in cash holding plus some increase in stock holding) that leave the final profits unchanged must be deemed as equally desirable and hence of no evaluative consequences to the firm. Changes that are of consequence must be those that do affect the final profits.

An exogenous change that increases the sales of the firm is usually of positive consequence to the final profits. Thus, some middle managers, especially those in charge of sales, may think it is highly desirable to advertise to increase sales. However, this endogenous increase has significant costs in the form of advertising outlays. The top management should know that even a highly effective advertising may not be desirable if it is too costly. What is desirable or not has to be judged by the net effects on the final profits. Similarly, for the case of Insert 1, what is normatively desirable or not has to be judged by the final effects on affective feelings, ultimately speaking.

Step 4: Something that is not in itself happiness or unhappiness but that may affect happiness or unhappiness either directly or indirectly may also be of normative significance.

If I surreptitiously put a tasteless poison in your coffee, it may have no effect on your enjoyment of that cup of coffee. However, if it makes you seriously sick the next day, it is obviously a bad thing for me to do that, at least if you deserve no punishment and no other benefits comes from this. More indirectly, telling a certain lie may in itself cause no or little unhappiness or may even save some embarrassment, but it may have the indirect undesirable effects of reducing marginally the degree of trust between people and even the degree of observance of other basic moral principles in general, and hence may eventually has more negative effects on happiness and is undesirable as a result.

Step 5: Normative (including moral) principles may be fostered to promote happiness and/or reduce sufferings or to promote things that may be indirectly conducive to happiness.

Due to the reliance of the human species on sociability (including cooperation in hunting) for survival, we have in-born (naturally selected or God-made) abilities like the capability to learn languages, the instinct for moral sentiments and outrages, and even the gene for altruism.⁵ However, as the human species also has a long period of childhood for learning, we also rely much on nurture/education, including learning to observe certain principles such as honesty and refraining from littering. This learning takes place at home, in schools, and through social contacts. Most of us benefits greatly from the observance of these principles by most people, largely speaking. Thus, the peer and social pressures against non-observance are big.

Step 6: The adherence of certain outdated normative principles may cause great sufferings.

As circumstances change, certain previously sensible normative principles may no longer be conducive to happiness and may even cause great sufferings. Just a single example suffices to convince. At least in ancient China (for at least a millennium from the Southern Song Dynasty to the recent Qing Dynasty), it was regarded as immoral for a woman to marry twice, even after the early death of her husband (while a man must remarry to have offspring and could even have more than one wives simultaneously). The long tradition to continue adhering to this moral principle cause great sufferings and was the theme of much realistic novels. Very slowly but eventually, this principle was given up over the early decades of the 1900′s.

Step 7: Just like people may adhere to inappropriate normative principles, people, including moral philosophers, may inappropriately believe that certain things are valuable independent of and over and above their contributions to happiness.

Influenced both by our inborn inclinations (including moral intuitions for equality and justice) and our upbringings, many people (including learned moral philosophers) may mistakenly believe in the normative significance of things other than happiness and independent of their contributions to happiness. This is a mistake (in the normative sense, being inappropriate, unpersuasive, or even wrong, rather than being false) since it violates Step 3. This may need some elaboration.

Step 3 says that if something is of normative significance, it must ultimately speaking be due to some effects on the enjoyment of happiness or the suffering of pain/unhappiness. These include the effects on future happiness/unhappiness and the effects on the happiness/unhappiness of others, including other individuals and possibly other affective sentients. When we examine closely normative principles/arguments supposedly based on considerations independent of happiness, we can always find that either they are really related, directly or indirectly, to effects on future happiness or effects on the happiness of others, or that the principles are not acceptable.⁶

Our discussion above is analogous to the three levels of ‘chan’ (禅), a Buddhist teaching on understanding the real world. At the first level, when you see a mountain (or river), it looks like a mountain (river). At the second level, when you see a mountain (river), it does not look like a mountain (river). At the third and highest level, when you see a mountain (river), it still looks like a mountain (river). Here, recognizing the intrinsic importance of happiness is the first level. Recognizing that things other than happiness, like knowledge, freedom, justice, morality, etc. are also important is the second level. Recognizing that why these things are important should ultimately be due to their contributions to happiness is the third and highest level.

It may be thought that what is of ultimate intrinsic values may be multi-factors or multi-dimensional and not just confined to happiness. However, if one (or the society, or any decision maker) has multiple final objectives, one will be in difficulties when these objectives are in conflict with each other. For example, most people including I agree that both freedom of speech and non-racism are good principles that we wish to observe, other things being equal. However, once a radio person made some racist remark and was fired. Here, the two principles are in conflict. Should one always observe one that is ranked higher in importance? How to judge? It is not obvious that non-racism is more important than freedom of speech, or vice versa. My view that these principles are important only because of their contributions to happiness allows us a simple way (in principle, though we may still have difficulties in getting the relevant information) to make decision in the presence of such a conflict of two or more desirable principles. We should roughly estimate the effects on happiness, taking all relevant effects into account. In the example above, perhaps if the remark involves only very slight hint of racism without too much bad effects, perhaps the anchor person should only be warned instead of fired. Whenever we have a conflict of two or more desirable principles/rules, we need some higher criterion to guide our choice. Overall net welfare is the highest criterion that trumps all other principles, as Little puts it (see the interview of Ian Little by Pattanaik and Salles 2005, p. 364, 368).

However, many people have objection to my welfare-supremacy view. To consider all such arguments would require a monograph in itself. Here, let us just consider some examples that can be discussed and answered together in the next Section.⁷

5.2 Answering Some Objections

One is the multi-century old argument that it is better to be the unhappy Socrates than a happy pig. A more modern version of this or a similar argument is the so-called pleasure or experience-machine argument (Nozick 1974); most people prefer their less happy current situations than to be hooked up to pleasure machines that will give them much more machine-induced pleasurable feelings. Another argument is on the common dis-preference for happiness based on falsehood. An example is a happy woman whose husband is disloyal to her. These and similar arguments are apparently very persuasive, making modern welfarists and utilitarians largely preference (or attitudinal) utilitarians rather than the classical hedonistic ones. (Welfarism maximizes social welfare as a function only of individual welfare; utilitarianism goes for a social welfare function that is just the unweighted, or equally weighted, sum of individual happiness. Classical utilitarianism maximizes the unweighted sum of individual happiness; preference utilitarianism maximizes the unweighted sum of individual preferences or utilities.) The principle of happiness as the only appropriate ultimate objective is defended below against these arguments.⁸

Consider Mr. C. He believes that, in the presence of uncertainty, the appropriate thing to do is to maximize the expected welfare. (Welfare is used interchangeably with net happiness. For simplicity, consider only choices that do not affect the welfare of others.) Suppose you put C in the privacy of a hotel room with an attractive, young, and willing lady. C can choose to go to bed with her or not to. C knows that the former choice involves a small but not negligible risk of contracting AIDS. He also calculates that the expected welfare of this choice is negative. Nevertheless, he agrees that, provided the lady is beautiful enough and the risk not too high (though high enough to reduce his expected welfare), he will choose to go (or fail to abstain going) to bed with her. This choice of C, though irrational (at least from the welfare point of view), is far from atypical. Rather, I am confident that it applies to at least 80% of adult males, the present writer included. Men are genetically programmed to want to make love to attractive (usually implying healthy) women in their reproductive ages who have not yet conceived (simultaneously explaining why slimness in waist and young girls are attractive), since this helps them to pass on their genes.

After the evolution of consciousness to help making choices (like fight or flight) on the spot by sizing up the situation, evolution (or God) makes sure that consciousness-guided choices are consistent with fitness (for survival and reproduction) by also endowing conscious species with affective feelings. So, activities consistent with fitness (like eating nutritious food when hungry and having sex with reproductive members of the opposite sex) are rewarded with pleasures and fitness-reducing activities (like injuries to the body) are penalized with pain. Thus, fitness-consistent choices are usually also welfare-maximizing choices. However, since the ultimate decisive factor is fitness, the coincidence is not 100% (Ng 1995). In particular, programming the organism to be excessively (from the viewpoint of welfare maximization) in fear of death or to be excessively inclined to mate may be fitness-maximizing. This explains why a man like Mr. C above will likely choose to have sex with the attractive girl even if he knows that this reduces his expected welfare. This example suggests that the choice or preference of a person may not be a perfect guide to what should ultimately be valuable to her.

Should our ultimate objective be happiness or should it be our preference? Preference may diverge from happiness or welfare for three reasons: a concern for the welfare of others (possibly including animals), ignorance, and irrationality (or imperfect rationality), as discussed in Chap. 10.1007/978-981-33-4972-8_2 above. These three factors are exhaustive as ‘irrational preference’ is defined to be the preference against one’s own welfare due neither to ignorance nor a concern for the welfare of others. Obviously, if the divergence is due to ignorance, happiness should prevail over preference. If due to a concern for the welfare of others, a distinction should be made between the individual and the society. It is admirable for an individual to sacrifice her own happiness for the welfare of others. However, for the society, the social objective should take account of the welfare of all individuals. (For simplicity, we ignore animal welfare here; on which see the final Chap. 10.1007/978-981-33-4972-8_15.) If the divergence is due to irrational preference, it is also clear that happiness should prevail, since preferences based on irrationality are similar to those based on ignorance. For the case of this divergence due to the genetically programmed tendency to mate, it may be pointed out that our (i.e. persons like you and me) welfare is the affective feelings that we enjoy over our life time, not that of our genes that ‘aim’ at fitness. We should aim at happiness, not fitness. (The maximization of long-term welfare requires sufficiently high fitness though.) We are the feeling persons, not the unfeeling genes. Thus, Mr. C, in his calm and reflective moment, may agree that it is in his interest to resist going to bed with the attractive lady to avoid contracting AIDS. However, due to biology, few men can resist successfully in that hotel room. However, for the ultimate social objective, we should go with his reflective moment rather the moment he was tempted by the attractive lady in the hotel room; we should go with our feeling persons, not with the unfeeling genes.

Just as we are born with the excessive inclination to mate, we are also genetically programmed with certain traits that tend to increase our fitness even if our happiness may be compromised somewhat, perhaps at the margin. One such trait is our inclination to do things rather than just enjoying existing accomplishments or just enjoy the stimulation of a pleasure machine. The drive to achieve helps us to increase our fitness. This drive is also much reinforced by education and social influences. We find such drives so natural and so important that we do not know that, when such drives conflict with our happiness, they (usually only at the margin, as too low a drive level is bad both for fitness and for welfare) are really bad for our true interest, just like the excessive drive to mate in the case of Mr. C in his hotel room. Thus, while it may be true that most persons will reject the pleasure machine option offering many times the amount of happy feelings, the choice of this option (assuming no external costs on others) is the more rational one, just like the choice of not sleeping with the lady.

Will I choose the pleasure machine option? Still not, because I believe that I can contribute to the welfare of others through my work. If I cannot and if hooking up to the machine will not put anyone in misery, I will in fact gladly choose to hook up!⁹ Similarly, I prefer to be a happy pig rather than be a learned philosopher if my philosophy cannot help others, directly or indirectly, to increase happiness by a lot more. My choice may be the exception. However, as explained above, most people (the present author included) are not perfectly rational due to our genetic programs and our upbringings.

We are similarly programmed and socially influenced to be curious and want to find the truth and so on. Thus, our preference for ‘authentic happiness’ or dis-preference for a happy but deluded life may again be explained by either the effects on the happiness in the future or of others, or by our imperfectly rational preferences. For the ultimate social objective, we should go with the rational objective of happiness rather than the imperfectly rational preferences. In fact, it may be argued that it is a mistake for failing to see that happiness is the only rational ultimate objective and all supposed qualifications to this can be explained by the effects on the happiness in the future or of others (hence really no qualification) or that their apparent acceptability is due to our imperfectly rational preferences programmed by our genes and shaped by our upbringings and social interactions. If moral philosophers can see this fundamental point in ethics, they would probably have no difficulties in accepting happiness as the only right ultimate moral principle, making most of the muddled discussion in moral philosophy unnecessary.

This does not deny the value of secondary virtues like truth, autonomy, accomplishment, justice, etc. However, it is important to keep in mind that all these virtues are ultimately based on the effects on happiness. Failure to do so may end up causing great suffering such as the case of blindly adhering to the moral principle of not permitting women to marry twice in ancient China discussed above. Injustice is the denial of due happiness or the undue imposition of unhappiness.¹⁰ Why certain denial is or is not due denial or undue imposition is ultimately also based on the effects on happiness. However, once secondary virtues/principles are accepted, their violation may be detrimental to happiness not only due to the direct effects but also due to indirect effects, including reducing the general adherence to good moral principles and making those believing in the principles concerned less happy. Viewing happiness as ultimately the only thing of value does not preclude taking all these direct and indirect effects into account.

Griffin (2007, p. 147) asks, ‘What could be the bridging notion that would allow us to compare a short life of supreme moral achievement with a long ordinary life?’ A short life of supreme moral achievement (safeguarding the country in Griffin’s example) may be more valuable, but only because it helps others to achieve more happiness, both directly in safeguarding the country and in setting a good example for virtuous and courageous acts. Accepting happiness as the only ultimately valuable thing allows us (subject to practical difficulties of estimating the quantitative effects) to compare different secondary values and to make the logical choice when different secondary principles are in conflict.¹¹

Most, if not all, objections to happiness as the only ultimate value ignore the effects on others and in the future. For example, consider Hausman’s (2010, p. 336) objection: ‘A crucial problem with the proposal to diminish the time people spend doing things they find unpleasant is that a myopic policy of maximizing current net pleasure is no more likely to maximize net pleasure over a lifetime than is a policy of maximizing weekly profits likely to maximize profits over a decade.’ Obviously, if adequate effects on the future and on others are taken into account, maximizing (net) happiness is not open to such objections.

The above argues in favour of happiness mainly against preference (as the ultimate objective). However, the difference between happiness and life satisfaction and that between happiness and preference are very similar. Thus, our argument in favour of happiness against preference can also be used to argue in favour of happiness against life satisfaction where the two differ.

5.3 Rejecting Kant’s Categorical Imperatives

Our welfarist position that ultimately only welfare or happiness is of intrinsic value is different or even inconsistent with some other moral philosophical positions, especially the Kantian. While this is not a place to provide a full refutation of these alternative positions, I will just briefly explain that welfarism/utilitarianism is not inconsistent with the valid parts in Kant (e.g. universalizability; treating humanity not as mere means); and those parts in Kant not consistent with welfarism/utilitarianism are not acceptable (e.g. disclosing the hiding place of potential victims to a murderer). For simplicity, we ignore the difference between utility and welfare discussed in Chap. 2 above.

If you want to quench your thirst, you have to drink something. Drinking something is then a conditional imperative, being conditional on wanting to quench your thirst. A categorical imperative is not conditional on anything; it is something you must do (morally speaking) unconditionally. Accepting some imperatives or moral principles categorically irrespective of their welfare consequences may certainly be inconsistent with welfarism, as discussed below with reference to Kant.

I have no qualm with Kant’s first formulation of universalizability: ‘Act only according to that maxim whereby you can at the same time will that it should become a universal law.’ (Kant 1785/1993, p. 30; 4:421). Obviously, it is also consistent with welfarism and utilitarianism. Each individual utility or welfare value has the same significance with any anonymous SWF (social welfare function). Since welfarism/utilitarianism does not rule out but rather requires anonymity, there is no inconsistency with universalizability.

Now consider Kant’s second formulation of humanity: ‘Act in such a way that you treat humanity, whether in your own person or in the person of any other, never merely as a means to an end, but always at the same time as an end’ (Kant 1785/1993, p. 36; 4: 429). The consistency of this with welfarism is ambiguous, depending on interpretation.

Kant’s imperative of ‘no mere means for humanity’ may be interpreted in a morally compelling way and as being consistent with welfarism. The central part of treating any person ‘never merely as a means to an end, but always at the same time as an end’ is certainly consistent with welfarism. Welfarism requires that the utility of each and every individual should enter into the SWF (i.e. is taken into account of) positively. This is inconsistent with treating a person as a mere means, not also as an end. If person is a mere means, his utility does not enter the SWF directly but may only affect social welfare by contributing (as a means) to the utility levels of some other individuals that enter the SWF. Welfarism requires the consideration of the utility of each and every person. We may use a pencil as a mere means. We may also use a person as a means to achieve something, but in doing so, we must also consider her feeling. If we use her as a mere means, not as also an end, then we do not need to consider her feeling, just like we do not have to consider the feeling of a pencil.

In fact, for this compelling part of ‘never merely as a means to an end, but always at the same time as an end’, I would go beyond humanity and include all sentients capable of welfare (enjoyment and suffering). Animal welfare should be a part of human morality, as discussed in the final Chap. 15.

Both Kant and many interpreters of Kant went beyond the above compelling part of ‘not merely a means’. Without entering into a full discussion, I will just mention two points quickly.

First, from his categorical imperative, Kant derived some moral principles too absolutely, to the disregard of possible huge welfare losses. One clearly unacceptable example is Kant’s insistence that ‘To be truthful (honest) in all declarations is therefore a sacred unconditional command of reason, and not to be limited by any expediency’ (Kant 1799; last page), even to an intended murderer about the hiding place of the potential victim. Kant made this clearly absolute imperative in his reply to Benjamin Constant’s criticism. Most people, myself clearly included, find this unacceptable.

One reason that led Kant astray is his mistaken all-or-nothing reasoning. This happened in many of his arguments. Just one simple example here. In his argument against stealing, Kant considered the moral proposition that it is permissible to steal. He argued that this proposition would result in a contradiction upon universalization. The concept of stealing itself presupposes the existence of private property rights. However, the universalization of stealing would lead to no property rights, leading to a logical contradiction. This reasoning is based on contrasting only the two opposite extremes of free stealing for all and absolutely no stealing in any circumstances. In the real world, we are always in-between these two non-existent extremes. Most morally upright persons would avoid stealing as far as possible, but may be compelled to steal if that is the only way, say, to avoid starvation of his children to death. Such a morality that allows stealing in some extreme circumstances does not end us up with complete absence of property rights. (For a similar mistake of all-or-nothing comparison by a Nobel laureate ending up with the grave mistake against the sensible taxation of pollution, see Ng 2007.)

Secondly, considering some real-world examples where the observance of some categorical imperatives has ended up in dismal situations may illustrate the unacceptability of sticking to some imperatives without regard to their welfare consequences. As already discussed above, in ancient China, for nearly a thousand years since the Southern Song dynasty in the twelfth century, people virtually universally believed in the categorical imperative that ‘one woman should not serve two men/husbands’. This does not just mean that a woman should not marry two husbands at the same time. Rather, it means that, once a woman is married to a man, she should not marry again even after the death of her husband, and even if the death occurs at the wedding night, after the ceremony but before real consummation of the marriage. This imperative or principle was dictated by the sacred unconditional command of chastity, irrespective of its welfare consequences. For many centuries, this imperative led to enormous misery. It was only after many decades of severe criticisms, including by many novelists, that the imperative was gradually given up over the early decades of the twentieth century. The change was partly assisted by the Western thinking of gender equality and women emancipation.

It may be thought that such backward moral principles as ‘one woman should not serve two men/husbands’ were only observed by ignorant people in a backward country in the past; modern people do not commit such a silly mistake. Actually, right today, almost throughout the whole world and for all countries except one, we have a similar sacrosanct moral principle of categorical imperative that has contributed to unnecessary, intense and prolonged suffering of many people. Those trying to help reduce this suffering have been persecuted and jailed. What is this principle? The sanctity of (human) life.

At least in some important aspects, the belief in the sanctity of life serves some very useful purposes. When a person dies, she can no longer enjoy life. Thus, ending a life prematurely is a grave matter. In addition, living persons are also threatened by this possible premature death. The costs of causing deaths are enormous. Emphasizing the sanctity of life helps to prevent or at least reduce wanton disregard to lives and serves some useful purpose. The mistake consists in absolutising it and divorcing it from welfare. This absolute sanctity leads to the imprisonment of doctors helping desperate patients to end their miserable lives earlier. This has led to much unnecessary suffering.

In conclusion, Kant and other deontic arguments have not refuted welfarism which is consistent with their acceptable parts; those parts inconsistent with welfarism are not really acceptable.

5.4 A Critique of Rawls

If all welfare-independent rights are unacceptable as fundamental moral principles, as argued above, Rawls’ (1971) second principle of justice (maximin) is absurd, despite its widespread influence.

Rawls’ first principle requires that each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberty for all. I am prepared to accept this first principle on the following understanding:

• (1) that it is adopted because it promotes the general welfare;

• (2) in circumstances where it is disastrous to the general welfare, it may have to be suspended;

• (3) in deciding what is the “most extensive total system” and what is compatible … for all”, the ultimate criterion is the general welfare.

A sex maniac may be in favour of freedom to rape and claim that this is compatible with everyone's freedom to rape.¹² It may also happen that the sex maniac is the person of the lowest welfare level such that freedom to rape for all will maximize the welfare of the worst off, hence consistent with the spirit of Rawls’ second principle to be discussed below. However, if freedom to rape results in the reduction of the welfare of those raped and scared of being raped by more than (in aggregate) the welfare gains of the rapists (even though the former still have higher welfare levels than the latter group even with freedom to rape), then freedom to rape should be regarded as not compatible with the freedom of not being raped. The “most extensive total system of basic liberties” should then not include the freedom to rape. However, thus interpreted, the first principle is really a device to promote the general welfare. It is not ultimate.

Despite its obvious absurdity, Rawls’ second principle is very popular. For example, Temkin believes that, in “one form or another, many philosophers have come to advocate a maximin principle of justice, and one can see why. There is strong appeal to the view that just as it would be right for a mother to devote most of her effort and resources to her neediest child, so it would be right for society to devote most of its effort and resources to its neediest members” (Temkin 1986 p. 109). In my view, the ethical appealingness of this argument as well as that of the maximin principle of justice itself is not difficult to refute.

I agree that, in most cases, a mother should devote more, and in many cases, most of her effort and resources to her neediest child, but only because this maximizes the welfare of the whole family (ignoring effects on others for simplicity). The most disadvantaged child is usually the neediest one because he/she will suffer most in the absence of extra help. Also, the extra help for the neediest also promotes the good spirit of helping the needy. Thus, the extra care for the most disadvantaged need not be inconsistent with overall welfare maximization. However, the maximin principle requires the mother to go much further.

For simplicity, suppose that the mother is faced with only two alternatives. One is to go away with the most disadvantaged child to live in a mountain resort for certain marginal benefit to the health of the sick child. The other is to stay at home looking after all the five children but still with possibly more care for the most disadvantaged one. Suppose the two welfare profiles for the children are

$$ \begin{array}{*{20}c} {{\text{WP}}_{{\text{A}}} = (10,10,10,10,9)} \\ {{\text{WP}}_{{\text{H}}} = (9000,9000,9000,9000,8)} \\ \end{array} $$

and that the mother is indifferent herself and no one else is affected by the choice. The maximin principle requires choosing WP_A. This, in comparison to the alternative of staying at home, increases the welfare of the worst-off child from 8 to 9 at the costs of a huge reduction in welfare (from 9000 to 10) for every other child. No sane mother in the world would make such an absurd choice.

Note that the welfare profiles WP_A and WP_H above are ultimate outcomes, as must be the case for all discussion of the ultimate ethical principles. Thus, the more equal welfare profile of WP_A would not promote further gain in welfare through, say, a more harmonious family relationship. Such effects, if any, should already have been incorporated into WP_A and WP_H.

Consider the much-cherished principle, “From each according to his ability; to each according to his needs” (which I personally approve, assuming no disincentive effect). Why doesn't it read, “An equal amount of work from each; an equal amount of income to each”? If a weak man is tired by four hours of work, it is better for a stronger man to work longer to relieve him. Similarly, if the worst-off child will not gain much more happiness by extra effort and resources, it is better that these resources be spent on other children.

In the original position, behind the veil of ignorance as to which child I would be born into, I would have not the slightest hesitation in wishing my mother to maximize the welfare of all the children, i.e. the unweighted sum of all children's welfare (the welfare of the mother and that of any other person are being held constant in this comparison). This maximizes mine as well as other children's expected welfare.

It is sometimes argued that a risk-averse person may not want to maximize expected welfare. It is quite rational to be risk-averse with respect to income or any other objective rewards since one may have, with good reasons, diminishing marginal utility/welfare of income. But since utility/welfare is the ultimate objective one is presumably maximizing, it is not rational not to maximize expected utility/welfare, if the relevant utility/welfare profiles already included all relevant effects, including such things as anxiety, excitement, etc. which explain most paradoxes of choices involving risk such as the Allais (Allais and Hagen 1979); see Harsanyi (1976, Part A) and Ng (1984). Secondly, if one chooses to be risk-averse with respect to welfare, it is still impossible to reasonably justify an absolute degree of risk-averseness as implied by the maximin principle.

After the lifting of the veil of ignorance, I would still think that it is right for my mother or anyone's mother to maximize the welfare of all children together whether I were the worst-off child or not. Again, my bias in favour of my own welfare may mean that I would hope that my mother would spend more effort and resources on me somewhat beyond the level justified by unweighted sum maximization. (But I think it is unjust for my mother to follow the partial wish of any child.) However, even then, I would definitely not want my mother to maximin in my favour, implying a zero trade-off on the welfares of my brothers and sisters as long as their welfares remain higher than mine. Thus, given the welfare profiles WP_A and WP_H above, even if I were the worst-off child, I would want my mother to choose WP_H. It may be thought that since my welfare is higher in WP_A, I could not have wanted my mother to choose WP_H. This ignores the differences between welfare and preference due to a concern for the welfares of others as discussed in Chap. 2 above.

From the above, it may be concluded that the maximin principle of justice is not only utterly unacceptable but an ethical principle similar to the one in favour of its adoption seems to require the worst-off group itself not to accept it. Why then is the principle so popular? One possible explanation is that it appeals to the guilt feeling of the better-off. They have admirable sympathy for the worst-off but yet are not prepared and/or find it ineffective to alleviate this by substantial personal contribution to the worst-off (by charity or the like). Paying lip service by advocating the maximin principle of justice is a much more cost-effective way of alleviating their sense of guilt. Of course, this explanation need not apply to all advocates of maximin.

In addition to those considered in this chapter, other arguments against happiness as the only intrinsic value are discussed in Appendix B.