Egoism vs. Altruism: A Philosophical Analysis of Self-Directed and Other-Directed Desires, Summaries of Nutrition

Download Egoism vs. Altruism: A Philosophical Analysis of Self-Directed and Other-Directed Desires and more Summaries Nutrition in PDF only on Docsity! Chapter 7 Psychological Egoism Elliott Sober Psycliological egoism is a theory about motivation that claims that all of our ulti- mate desires are self-directed. Whenever we want others to do well (or ill), we have these other-directed desires only instrumentally; we care about others only because we think that the welfare of others will have ramifications for our own welfare. As stated, egoism is a descriptive, not a normative, claim. It aims to cliaracterize what motivates human beings in fact; the theory does not say whether it is good or bad tliat people are so motivated. Egoism has exerted a powerful influence in the social sciences and has made large inroads in the thinking of ordinary people. Economists typically think of human beings as being moved by "rational self-interest," where this excludes any irreduc- ible concern for die welfare of others. And ordinary folks often claim tliat people help others only because this makes them feel good about themselves, or because they seek the approval of third parties. It is easy to invent egoistic explanations for even the most harrowing acts of self- sacrifice. The soldier in a foxhole who throws himself on a grenade to save tlie lives of his conirades is a fixture in tlie literature on egoism. How could this act be a product of self-interest if the soldier lu~ows tliat it will end his life? Tlie egoist may answer that the soldier realizes in an instant that he would rather die than suffer the guilt feelings that would haunt him if he saved himself and allowed his friends to perish. Tlie soldier prefers to die and then have no sensations at all rather than live and suffer tlie torments of die damned. This reply may sound forced, but it remains to be seen what grounds we have for regarding it as false. Tlie criticisms that have been leveled against psycliological egoism can be divided into three categories. First, there is the claim that it is not a genuine theory at all. Second, there is the allegation that it is a theory that is refuted by wliat we observe in human beliavior. Third, there is the idea tliat, although egoism is a theory that is consistent with wliat we observe, there are other, extra-evidential considerations that suggest tliat it should be rejected in favor of an alternative theory, motiva- tional pluralism, according to which human beings have both egoistic and altruistic Elliott Sober ultimate desires. All three types of criticism will be considered in what follows, but first we need to state the theory more carefully, 1 Clarifying Egoism When egoism claims that all our ultimate desires are self-directed, what do "ulti- mate" and "self-directed" mean? There are some tllings that we want for their own sakes; other tliings we want only because we think they will get us something else. Tlie familiar means/end relation tliat links one desire to another also allows desires to be chained together - Sarah may want to drive her car because she wants to get to die bakery, she may want to go to the bakery because she wants to buy bread, etc. The crucial relation that we need to define is this: S wants m solely as a means to acquiring e if and only if S wants m, S wants e, and S wants m only because she believes that obtaining m will help her obtain e. An ultimate desire is simply a desire that someone has for reasons that go beyond its ability to contribute instrumentally to die attainment of something else. Consider pain. The most obvious reason that people want to avoid pain is simply that they dislike experiencing it. Avoiding pain is one of our ultimate goals. However, many people realize tliat being in pain reduces their ability to concentrate, so they may sometimes take an aspirin in part because they want to remove a source of distrac- tion. This shows that the things we want as ends in themselves we also may want for instrumental reasons. When psychological egoism seeks to explain why one person helped another, it isn't enough to show that one of die reasons for helping was self-benefit; this is quite consistent with there being another, purely altruistic, reason that the individual had for helping. Symmetrically, to refute egoism, one need not cite examples of lielping in which only other-directed motives play a role. If people sometimes help for both egoistic and altruistic ultimate reasons, then psychological egoism is false. Egoism and altruism both require die distinction between self-directed and other- directed desires. This distinction is to be understood in terms of a desire's propositional content. If Adam wants the apple, this is elliptical for saying that Adam wants it to be the case that he has the apple. Tllis desire is purely self-directed, since its propositional content mentions Adam, but no other agent; I assume that Adam does not regard the apple as an agent. In contrast, when Eve wants Adam to have the apple, this desire is purely other-directed; its propositional content men- tions another person, Adam, but not Eve herself. Egoism claims that all of our ultimate desires are self-directed; altruism, that some are other-directed. Tlie fact that Eve has an other- directed desire is not enough to refute egoism; one must ask why Eve wants Adam to have the apple. Psycl~ological Egoism 2 Is Egoism Empirically Testable? One standard philosophical objection to egoism is that it is not a testable hypoth- esis. As the example of the soldier in the foxhole suggests, it seems that egoism can accommodate any behavior whatever. Whether people are nasty or nice to each other, the theory can explain why. This claim about the flexibility of egoism is then linked to a Popperian criterion concerning what it takes for a statement to be scien- tific, with the conclusion drawn that egoism is not a genuine scientific theory at all. It is, despite appearances, empirically vacuous. This argument is flawed in two ways. The first pertains to its sanguine confidence that no observation could ever disconfirm egoism. The fact that the theory can accommodate the soldier in tine foxhole and other behaviors that have been consid- ered by p1Iilosophers hardly suffices to justify this global claim. As it happens, the experimental work in social psychology on altruism and egoism shows that the relevant observational evidence extends beyond the existence of instances of help- ing behavior (Batson 1991; Schroeder ct al. 1995). In addition, the Duhemian point that theories are testable only in conjunction with background assumptions should lead us to draw back from the charge of untestability. If two theories make the same predictions against one background framework, they may make different predictions against another. How do we know that new background theories will never be developed that allow egoism to be put to the test? The charge of untestability presupposes that we have an omniscient grasp of the future of science. The second defect in this argument is that it neglects to notice that the charge of untestability is a two-edged sword. The argument is advanced as a reason for reject- ing egoism. What, then, are we to accept as a positive account of motivation? Pre- sumably, motivational pluralism is supposed to be the acceptable alternative. However, this cannot be where the argument leads. If egoism is untestable, then so is motivational pluralism. As flexible as egoism is in its ability to accommodate observations, pluralism is more flexible still. After all, pluralism deploys all the vari- ables that egoism invokes, and then some. The two theories are related to each other in the same way that "y = f(x)" and "y = g(x,w)" are related. The reason egoism appears to be untestable is that it is an ism. It does not provide specific explanations for behaviors, but merely indicates the kind of explanation that all behaviors will have. This is why it is possible for egoism to be retained even when specific egoistic explanations are found wanting. Why did George donate all that money to charity? A defender of egoism might suggest that George did so because he wanted to improve his business contacts by impressing others. However, sup- pose one then learns that George donated the money anonymously. This refutes the specific egoistic explanation just described, but it isn't hard to invent another. George made the donation because it made him feel good and because he knew that if he did not, he would experience pangs of guilt. The pattern here is typical - hedonism is the position to which egoists standardly retreat. If external benefits don't suffice to explain, one invokes internal, psycl~ological benefits instead. Elliott Sober That egoism is a claim about a type of explanation, and therefore is distinct from the specific explanations that are of the type required, is a pattern that arises in many debates about isms. Consider adaptationism in evolutionary biology. Adaptationists emphasize the importance of natural selection in explaining the observed traits of organisms. Because this ism, by itself, does not provide a specific explanation for any trait, it remains possible for a biologist to continue to be an adaptationist even after a specific adaptationist explanation is found wanting. Why did wings evolve in insects? The hypothesis that wings evolved as an adaptation for flying is thrown in doubt by the fact that very small wing buds provide no lift whatever; although 5 percent of an eye can still function as a light sensor, 5 percent of a wing does noth- ing to get an organism off the ground. However, wing buds are found in some flightless insect species; they function as thermo-regulators. This suggests an alter- native adaptationist hypothesis - that insect wings started to evolve because they initially promoted thermo-regulation and then continued to evolve because they then facilitated flight. And if this hypothesis is challenged, the adaptationist can cast about for a third alternative. It is no good to reject adaptationism because it has this sort of flexibility; the alternative ism, evolutionary pluralism, claims that natural selection is one among several important causes of evolution. As flexible as adaptationism is, pluralism is more flexible still. 3 Butler's Stone As noted above, even though hedonism is a special version of egoism, hedonistic explanations are often what egoists invoke when a nonhedonistic explanation is found wanting. If George didn't donate money to charity to make business con- tacts, perhaps he did so for the warm glow of satisfaction that the donation pro- vided. For this reason, arguments that attempt to refute hedonism have a special location in the dialectical landscape. Although refuting hedonism is not sufficient to refute egoism, it would make an important contribution to that larger enterprise. Many philosophers have thought that Joseph Butler (1692-1752) refuted he- donism once and for all (Broad 1965; Feinberg 1984; Nagel 1970) in the following passage: That all particular appetites and passions are towards external things themselves, distinct from the pleasure arisinj from them, is manifested from hence; that there could not be this pleasure, were it not for that prior suitableness between the object and the pas- sion: there could be no enjoyment or delight from one thing more than another, from eating food more than from swallowing a stone, if there were not an affection or appetite to one thing more than another. (Butler 1965 [1726]: 227) I'll call this argument Butler's stone. Although Butler does not explicitly say in this passage that hedonism is false, let us construe the argument with this as its conclu- sion: Psychological Egoism 1 People sometimes experience pleasure. 2 When people experience pleasure, this is because they had a desire for some external thing, and that desire was satisfied. Hedonism is false. I don't propose to challenge the first premiss. However, I think the second premiss is false and that the conclusion does not follow from the premisses. The second pren~iss is over-stated; although some pleasures are die result of a desire's being satisfied, others are not (Broad 1965: 66). One can enjoy die smell of violets without having formed the desire to smell a flower, or something sweet. Since desires are propositional attitude, forming a desire is a cognitive achievement. Pleas- ure and pain, on the other hand, are sometimes cognitively mediated, but sometimes they are not. Notice that this defect in the argument can be repaired; Butler does not need to say that desire satisfaction is the one and only road to pleasure. The transition from premisses to conclusion is where the argument really goes wrong. Consider the causal chain from a desire (the desire for food, say), to an action (eating), to a result - pleasure. Because the pleasure traces back to an aiite- cedently existing desire, it will be false that the resulting pleasure caused the desire (on the assumption that cause must precede effect). However, this does not settle how two desires - the desire for food and the desire for pleasure - are related. In particular, it leaves entirely open what caused the desire for food. Hedonism says that people desire food because they want pleasure (and think that food will bring them pleasure). Butler's stone concludes that this causal claim is false, but for no good reason. The crucial mistake in the argument comes from confusing two quite different items - the pleasure that results from a desire's being satisfied and the desire for pleasure. Even if the occurrence of pleasure presupposed that the agent desired something besides pleasure, nothing follows about die relationship between the desire forpleasure and the desire for something else (Sober 1992; Stewart 1992; Sober and Wilson 1998). Hedonism does not deny that people desire external tllings; rather, the theory tries to explain why that is so. It is curious that tlis argument has been interpreted so widely as refuting hedon- ism. At the end of the sermon in which the stone passage occurs, Butler says this: "Let it be allowed, though virtue or moral rectitude does indeed consist in affection to and pursuit of what is right and good, as such; yet, that when we sit down in a cool hour, we can neither justify to ourselves this or any other pursuit, till we are convinced that it will be for our happiness, or at least not contrary to it" (Butler 1965 [1726]: 240). And if we return to the language of the stone argument itself, we see tliat Butler is making a claim about the content of "particular appetites and passions." Read narrowly, the argument says merely that if people desire pleasure, their desires do not fall under tliat rubric; the argument does not say that people never desire pleasure nor does it say that the desire for pleasure is never ultimate. Did Butler fail to refute hedonism in the stone argument because he wasn't even trying to do so? Elliott Sober Figure 7.1 Decision Plug Reached In 4 4 Choose to plug in: time --+ Choose not to plug in: T Decision T Don't Reached Plug In The four letters in these two time lines represent how pleasant your experiences will be during different temporal periods, depending on what you decide. If you choose to plug into the machine, you'll liave an immense level of bliss (b) after you plug in. This will dwarf the amount of pleasure you'll experience in die same period of time if you decide not to plug in and to lead a normal life instead; b > %. If this were tlie only consideration involved, the. hedonist would liave to predict that peo- ple will choose to plug into the machine. How can hedonism explain the fact that many people make the opposite decision? The hedonist's strategy is to look at earlier events. Ifyou decided to plug into the machine, how would you feel before you were actually connected? Presumably, you would experience a great deal of anxiety (a). You'd realize that you were about to stop leading a real life. You will never again see the people you love; all of your projects and plans are about to be terminated. It is clear that you would have less pleasure during this period of time than you would if you rejected the option of plugging into the machine and continued witli your real life instead; a < n^. If hedonists are to explain why people choose not to plug into the experience niacl-line, and are to do this by considering just tlie pleasure and pain diat subjects expect to come their way after they decide what to do, die claim must be that a + b < nl + R. Since b is far greater than n2, tlus inequality will be true only if a is far far smaller than nl. That is, hedonists seem compelled to argue that people reject the option of plugging in because the amount of pain they would experience between deciding to plug in and actually being connected to die machine is 8z)antic - so large that it dwarfs tlie pleasure they'd experience after they are connected. This suggestion is not plausible. The period of time between deciding to plug in and actually doing so can be made very brief, compared witli tlie long stretch of years you'll spend attached to the machine and enjoying a maximally pleasurable Psychological Egoism ensemble of experiences. I grant that people who decide to connect to the machine will experience sadness and anxiety during the brief interval between deciding to plug in and actually plugging in. But the idea that this negative experience swamps all subsequent pleasures just isn't credible. To see why, let's consider a second thought experiment, suggested to me by William Talbott. Suppose you were offered $5,000 if you went through 10 seconds of a certain experience. The experience is believing that you had just decided to spend the rest of your life plugged into an experience machine. After your 10 sec- ond jolt of this experience, you will return to your normal life and will realize that you just had a "nightmare"; you then will receive the money as promised. I expect that many people would clioose the 10 seconds just described because it will earn them $5,000. This shows that hedonism is mistaken if it claims that tlie experience of believing you will be plugged into an experience inacli~ie for tlie rest of your life is so liorrible that no one would ever clioose a life tliat included it. The hedonist still has not been able to explain wliy many people would clioose a normal life over a life plugged into the experience machine. Tlie reason is that a hedonistic calculation seems to lead inevitably to die conclusion that a + b > nl + :. Does this mean that die hedonist must concede defeat? I think that the hedonist has a way out. Quite apart from die amount of pleasure and pain that accrues to subjects after they decide what to do, there is the level of pleasure and pain arising in tlie deliberation process itself. Tlie hedonist can maintain that deciding to plug into die machine is so aversive that people almost always make the other choice. When people deliberate about the alternatives, they feel bad when they think of the life they'll lead if tliey plug into tlie macliine; they feel much better when tliey consider the life they'll lead in tlie real world if they decline to plug in. The idea of life attached to the machine is painful, even though such a life would be quite pleasurable; die idea of real life is pleasurable, even though real life often includes pain. This hedonistic explanation of wliy people refuse to plug in exploits die dis- tinction that Scl~liclz (1939) drew between die pleasant idea of a state and die idea of a pleasant state. To see what is involved in this suggestion, let's consider in more detail what goes through people's minds as they deliberate. They realize that plugging in will mean abandoning the projects and attachments tliey hold dear; plugging into the ina- chine resembles suicide in terms of the utter separation it effects with tlie real world. Tlie difference is that suicide means an end to conscio~~sness, whereas the experi- ence machine delivers (literally) escapist pleasures. Hedonism is not betraying its own principles when it claims tliat many people would feel great contempt for the idea of plugging in and would regard the temptation to do so as loathsome. People who decline the chance to plug in are repelled by tlie idea of narcissistic escape and find pleasure in the idea of choosing a real life. One virtue of this liedonistic explanation is tliat it explains tlie results obtained in both the tliouglit experiments described. It explains why people often decline to plug into the experience machine for the rest of their lives; it also explains wliy people offered $5,000 often agree to have ten seconds of die experience of believ- Elliott Sober ing that they have just decided to plug into tlie machine for the rest of their lives. In both cases, deliberation is guided, not so much by beliefs about which actions will bring future pleasure, but by the pleasure and pain that accompany certain tliouglits d w i n j the deLibe~ation p~ocess its68 The problem of the experience machine resembles the problem of the soldier in the foxhole, discussed earlier. How can hedonism explain tliis act of suicidal self- sacrifice, if tlie soldier believes that he will not experience anything after lie dies? The hedonist can suggest that there is a self-directed benefit tliat accrues before the act of self-sacrifice is performed. It is no violation of hedonism to maintain that tlie soldier decides to sacrifice his life because that decision is less painful than die deci- sion to let his friends die. The problem of suicidal self-sacrifice and tlie problem posed by the experience machine can be addressed in the same way. 6 Burden of Proof Philosopl~ers sometimes maintain that a common sense idea should be regarded as innocent until proven guilty. That is, if a question is raised about whether some common sense proposition is true, and no argument can be produced that justifies or refutes it, then tlie sensible thing to do is to keep on believing die proposition. Put differently, the idea is that the burden of proof lies with those who challenge comnion sense. This general attitude sometimes surfaces in discussion of egoism and altruism. Tlie claim is advanced that tlie egoism hypothesis goes contrary to common sense. Tlie common sense picture of human motivation is said to be pluralistic - people care about tliemselves, but also care about others, not just as means, but as ends in themselves. Tlie conclusion is then drawn that if philosophical and scientific argu- mentation for and against egoism is indecisive, then we should reject egoism and continue to accept pluralism. One objection to this proposed tie-breaker is tliat it is far from obvious that "common sense" is on the side of motivational pluralism rather than egoism. What is comnion sense? Isn't it just wliat people commonly believe? If so, it is arguable that egoism has made large inroads; it now seems to be a view that is endorsed by large numbers of people. Philosophers need to be careful not to confuse common sense with wliat they themselves happen to find obvious. As far as I know, no em- pirical survey has determined whether a pluralistic theory of motivation is more popular than psycliological egoism. Regardless of what people commonly believe about psycliological egoism and motivational pluralism, I reject die idea tliat conformity with common sense is a tie- breaker in this debate. It does not liave tliis status in physics or biology, and I see no reason why it should do so when tlie question happens to be philosophical or psy- cliological in character. In fact, it is arguable that our intuitions in this domain are especially prone to error. People liave a picture of their own motives and the mo- Psycl~ological Egoism poison distinction and other properties found in the environment. Finally, we may add to our list die idea that there can be pluralistic solutions to a design problem. In addition to the monistic solution of having a protein detector and the monistic solution of having a color detector, an organism might deploy both a protein detec- tor and a color detector. Given this multitude of possibilities, how might one predict which of them will evolve? Three principles are relevant here - availability, reliability, and efficiency (Sober 1994; Sober and Wilson 1998). Natural selection acts only on the range of variation that exists ancestrally. A protein detector might be a good thing for the organism to have, but if that device was never present as an ancestral variant, then natural selection cannot cause that trait to evolve. So the first sort of information we'd like to have concerns which proximate mechanisms were available ancestrally. Let's suppose for the sake of argument that both a protein detector and a color detector are available ancestrally. Which of them is more likely to evolve? Here we need to address the issue of reliability. Which device does the more reliable job of indicating which particles in the environment are good to eat? Without further information, not much can be said. A color detector may have any degree of reli- ability, and die same is true of a protein detector. There is no a priori reason why the direct strategy should be more or less reliable than die indirect strategy. How- ever, there is a special circumstance in which they will differ. It is illustrated by figure 7.2. The double arrows indicate correlation; gaining nutrition is correlated with an organism's fitness, and a particle's being red rather than green is correlated with its nutritional content (figure 7.2). In the diagram, there is no arrow from fitness to color except the one that passes through nutrition. This means that an organism's fitness is correlated with the color of the particles that it eats. There is no a p r i o r i reason that color should be relevant to fitness only by virtue of indicating nutri- tional content. For example, if eating red particles attracted predators more than Figure 7.2 fitness <Ñ nutrition <-> color of organism in particle of particle ^ ^ ^ ^ behavior behavior Elliott Sober eating green ones does, then color would have two sorts of relevance for fitness. However, if nutrition "screens off" fitness from color in the way indicated, we can state the following principle about the reliability of the direct device D and the indirect device I: (D/I) If nutrition and color are less than perfectly correlated, and if D detects nutrition at least as well as I detects color, then D will be more reliable than I. This is the Direct/Indirect Asymmetry Principle. Direct solutions to a design prob- lem aren't always more reliable, but they are more reliable in this circumstance. A second principle about reliability also can be extracted from this diagram. Just as scientists do a better job discriminating between hypotheses if they have more evidence rather than less, so it will be true that organisms make more reliable discriminations if they have two sources of information about what to eat rather than just one: (TBO) If nutrition and color are less than perfectly correlated, and if D and I are each reliable, though fallible, detectors of fitness, then D and I working together will be more reliable than either of them working alone. This is the Two-is-Better-than-One Principle. It requires an assumption - that the two devices do not interfere with each other when they are both present in an organism; they function fairly independently. The D/I Asymmetry and the TBO Principle pertain to the issue of reliability. Let us now turn to the third consideration that is relevant to predicting which proxi- mate mechanism will evolve, namely efficiency. Even if a nutrition detector and a color detector are both available, and even if the nutrition detector is more reliable, it doesn't follow that natural selection will favor the nutrition detector. It may be that a nutrition detector requires more energy to build and maintain than a color detector. Organisms run on energy no less than automobiles do. Efficiency is rel- evant to a trait's overall fitness just as much as its reliability is. With these three considerations in hand, let's return to the problem ofpredicting which motivational mechanism for providing parental care is likely to have evolved in the lineage leading to human beings. The three motivational mechanisms we need to consider correspond to three different rules for selecting a behavior in the light of what one believes: (HED) Provide parental care if and only if doing so will maximize pleasure and minimize pain. (ALT) Provide parental care if and only if doing so will advance the welfare of one's children. (PLUR) Provide parental care if and only if doing so will either maximize pleasure Psychological Egoism and minimize pain, or will advance the welfare of one's children. (ALT) is a relatively direct, and (HED) a relatively indirect, solution to the de- sign problem of getting an organism to take care of its offspring. Just as an organ- ism can find nutrition by detecting color, so it is possible in principle for a hedonistic organism to be built in such a way that it will provide parental care; what is required is that the organism be so constituted that providing parental care is the thing that maximizes its pleasure and minimizes its pain (or that the organism at least believes that this is so). Let's consider how reliable these three mechanisms will be in a certain situation. Suppose that a parent learns that its child is in danger. Imagine that your neighbor tells you that your child has just fallen through the ice on a frozen lake. Figure 7.3 shows how (HED) and (ALT) will do their work. Figure 7.3 child needs help Ñ parent believes child needs help Ñ parent feels anxiety and fear 4. 4. i behavior i behavior The altruistic parent will be moved to action just by virtue of believing that its child needs help. The hedonistic parent will not; rather, what moves the hedonistic par- ent to action is the feelings of anxiety and fear that are caused by the news, or the parent's belief that such negative feelings will continue unless the child's situation is improved. It should be clear from figure 7.3 that the (D/I) Asymmetry Principle applies. In the circumstance specified, (ALT) will be more reliable than (HED). And by the (TBO) Principle, (PLUR) will do better than both. In this example, hedonism comes in last in the three-way competition, at least as far as reliability is concerned. The important thing about this example is that the feelings that the parent has are belief mediated. The only reason the parent feels anxiety and fear is that the parent believes that its child is in trouble. This is true of many of the situations that egoism and hedonism are called upon to explain, but it is not true of all of them. For example, consider the following situation in which pain is a direct effect, and belief a relatively indirect effect, of bodily injury (figure 7.4).