Part II
This book is organized around the “belief–desire model” of action. To understand how people act and interact, we first have to understand how their minds work. This is largely a matter of introspection and folk psychology, refined and corrected by the more systematic studies carried out by psychologists and, increasingly, by behavioral economists. The model is vital not only for explaining behavior, but also for assigning praise, blame, or punishment. Guilt usually presupposes mens rea, intentions and beliefs. Strict liability – guilt assigned merely on the basis of the actual consequences of action – is rare. In fact, sometimes we hold people guilty merely on the basis of intentions even when no consequences follow. Attempted murder is a crime. “Witches,” declared John Donne, “think sometimes that they kill when they do not, and are therefore as culpable as if they did.” “As for witches,” wrote Hobbes, “I think not that their witchcraft is any real power; but yet that they are justly punished, for the false belief they have that they can do such mischief, joined with their purpose to do it if they can.”
The belief–desire model, although indispensable, is fragile. The methods we use to impute mental states to other people do not always yield stable results. If we want to measure the height of a building, it does not matter whether we do it from the roof downward or from the ground upward. In the determination of beliefs and desires, the outcome may depend on such irrelevant factors. Consider, for instance, the idea that people “maximize expected utility” (Chapter 13). To make it precise, we have to assume that they have a clear and stable idea of the value attached to each possible outcome of an action, and of the probability they assign to the occurrence of that outcome. Often that assumption is justified, but sometimes it is not.
Consider first the beliefs of the agent. In eliciting the subjective probabilities an individual attaches to an event, a standard procedure is the following. Beginning with a number p, we ask the person whether he would prefer a lottery in which he gains a certain sum of money with probability p or one in which he gains the same amount if the event in question occurs.1 If he prefers the former, we expose him to a new choice with the probability adjusted downward; if he prefers the latter, we adjust the probability upward. By continuing in this way, we shall ultimately reach a probability p* such that he is indifferent between a lottery in which he gains the money with probability p* and one in which he gains it if the event occurs. We can affirm, then, that the revealed or elicited probability he attaches to the event is p*. In principle, p* should be independent of the initial p: that is, the elicited probability should be independent of the procedure of elicitation. In practice, this is not the case: a higher p induces a higher p*. This finding suggests that, to some extent at least, there is no fact of the matter, no stable mental state that is captured by the procedure.2
Other procedures are even more fragile. Often, scholars impute subjective probabilities to the agents on the assumption that when they know little about the situation they will assign equal probability to each of the possible states of the world. The justification of this procedure is supposed to be the “principle of insufficient reason”: if you have no positive grounds for thinking one state of the world more likely than another, logic forces you to assign equal probability to them. But states of the world can be conceptualized and counted in many ways. Suppose you are pursuing a thief and arrive at a fork in the road where three paths branch off, two going uphill and one downhill. Since you have no reason for thinking it more likely that he followed one path rather than another, the probability that he took the downhill path should, according to the principle, be one-third. But since you also have no reason for thinking he went uphill rather than downhill, the same probability should be one-half. In this case at least, the principle of insufficient reason is too indeterminate to be of any use in constructing or assigning probabilities.
Consider next the elicitation of preferences. In experiments, subjects have been asked whether they would buy various items (computer accessories, wine bottles, and the like) at a dollar figure equal to the last two digits of their social security number. Thereafter, they were asked to state the maximal price they were willing to pay for the product. It turned out that their social security number had a significant impact on what they were willing to pay. For instance, subjects who had social security numbers in the top quintile were willing to pay on average $56 for a cordless computer keyboard, while those in the bottom quintile were only willing to pay $16. Although the procedures were supposed to tap or elicit preexisting preferences, the results show that there was nothing there to elicit, no fact of the matter. The numbers owed more to the anchoring provided by the social security numbers than to any “real” preferences.
There is also evidence that people's trade-offs among values are highly unstable and may owe as much to procedural artifacts as to an underlying mental reality. Trade-offs can be captured either by choice or by matching in experiments. Subjects may be given the choice between saving many lives at a high cost per life saved (A) and saving fewer lives at lower cost (B). Alternatively, they may be asked to indicate the cost per life saved that would make them indifferent between saving the larger number of lives at that cost (option C) and option B. Suppose that a given subject states a cost lower than the cost of A. As the person is indifferent between C and B and may be assumed to prefer C to A (because C saves as many lives at lower cost), she should prefer and hence choose B over A. The overwhelming majority of subjects did in fact choose a cost for C below that of A, and yet two-thirds stated that they would choose A over B. The more important value – saving lives – is more salient in choice than in matching, although logically the two procedures should be equivalent.
There are other reasons why we should not always take statements about beliefs and other mental states at face value. Religious beliefs are especially problematic in this respect. In early seventeenth-century England, a prelate such as Bishop Andrewes could at one and the same time claim that the plague was a punishment that God imposed on sinners and flee London for the countryside. (By contrast, Philip II of Spain was so confident of divine support that he never made contingency plans.) The belief that by virtue of their divine origin the French kings could heal scrofula by touching the sick person was visibly withering by the end of the eighteenth century, when the traditional formula (“The king touches you; God heals you”) was replaced by a subjunctive (“The king touches you; may God heal you”). The eagerness with which the king's court sought out documented proof of successful healings also suggests a belief that was not sure of itself.
For a contemporary example, consider the idea that the behavior of Islamic suicide attackers can be explained, at least in part, by their belief that there is an afterlife to which martyrdom will give them a privileged access. One may ask whether this “belief” is of the same nature as our belief that the sun will rise tomorrow, that is, whether it is used with equal confidence as a premise for action. This is not a matter of certainty versus probability, but of confidence versus lack of it. I may have great confidence in – and be willing to bet on the basis of – a probabilistic belief based on many past occurrences. The belief in the afterlife held by most people is probably not like that.3 Rather, it may be a somewhat shadowy “quasi-belief,” held for its consumption value rather than as a premise for action. If all who claim to believe in the afterlife held the belief with full certainty, or with “confident probability,” we would observe many more martyrs than we actually see. Although some believers may be of this type, and suicide attackers may be recruited disproportionately from this subset, I suspect that for many, religion serves as a consolation once the decision has been made rather than as a premise for decision.4
This is not to deny that faith, or superstitions, can have non-trivial effects. Thus in French Indochina, a person who thought he was pursued by a demon would run across the road just in front of a car, at the risk of being killed, because he thought the car might kill the demon who followed him in the form of his shadow. Yet note that the superstition against staying on the thirteenth floor of a hotel is not strong enough to make the hotel company incorporate an empty thirteenth floor in the building, since customers seem to be happy staying on that floor as long as it is renumbered as the fourteenth. In the Roman Empire, some individuals, including the Emperor Constantine, postponed their baptism until death was at hand, to be able to enjoy earthly pleasures while also entering heaven with a clean slate. Had they really and fully believed in the afterlife and in baptism as a necessary entry ticket, they would have taken account of the fact that death can occur suddenly without any warning signs.
Similarly, people may experience or claim to experience “quasi-emotions” that differ from genuine emotions in that they have no implications for action. Some people who claim to be indignant over third-world poverty and yet never reach for their wallet may enjoy their indignation as a consumption good, because it makes them think well of themselves. (I return to such “warm-glow” phenomena in Chapter 5.) Similarly, the visible enjoyment of many who claimed to feel grief (or “quasi-grief”) after the death of Princess Diana was inconsistent with the horrible feeling of genuine grief. The appropriate term for their feelings is, I believe, “sentimentality.” (The German Schwärmerei is even more fitting.) Oscar Wilde defined a sentimentalist as “one who desires to have the luxury of an emotion without paying for it.” Whether the payment takes the form of donations to Oxfam or the form of suffering, we can tell from its absence that we are not dealing with the real thing.5
A related issue is the immense power of autosuggestion. Once we know that an X is supposed to be a Y, we claim and believe that it is obviously a Y. The world's greatest experts on Vermeer were taken in by (what now seem to be) obvious forgeries by van Meegeren. Proust refers to the “aptitude which enables you to discover the intentions of a symphonic piece when you have read the program, and the resemblances of a child when you know their kin.” Later, I cite how Emma in Jane Austen's novel persuades herself that she is in love. A European jazz fan completely changed his high appreciation of Jack Teagarden upon learning that he was not black. If we are well disposed toward a writer, we may read deep meanings into what an impartial reader would consider trivial remarks. We project our expectations on the world and then claim that the world confirms and justifies our beliefs.
The expectations can have real effects. Thus when the same wine is served with different price labels, the pleasure centers in the brain are more highly activated in subjects who think they are drinking an expensive wine. When subjects are made to believe, falsely, that they are drinking an alcoholic beverage, they behave (up to a point) as if they were drunk. Placebo effects are the best-known examples of this phenomenon. (Even genuine painkillers are less effective when given covertly.) Placebo effects can be completely or partially reversed by opioid antagonists, suggesting that the placebo works by generating endogenous opioids in the body, similar to what produces the “runner's high.” The runner, though, has to work for his pleasure.
The upshot of these remarks is that we should be wary of thinking of beliefs, desires, preferences, emotions, and the like as stable and enduring entities on a par with apples and planets. Later chapters will provide many instances of the coarse-grained, elusive, unstable, or context-dependent nature of mental states. I shall also, however, make statements that may seem to exemplify the very kind of pseudo-precision or make-believe rigor I have been warning against. In some cases, I do this to explain the internal workings of a model, without vouching for its realism. In other cases, I present a precise idea as a specification of a more general one that I believe to be valid. For example, I do not think people update their beliefs by Bayesian reasoning by the precise mechanism I describe in Chapter 13. I do believe, however, that if twelve jurors are confronted with the same evidence, those who enter the courtroom with a stronger belief in the defendant's guilt will on average be more likely to convict. Also, jurors will shift their belief toward guilt when presented with a strong piece of evidence supporting it. (By contrast, inconsistently with Bayesian reasoning they may shift their belief away from guilt when later presented by weaker evidence also supporting it.) These qualitative statements often suffice for explanation, but not for prediction. Similarly, although I do not think that people discount the future by a precise quantitative function, we can often explain their behavior by assuming that they attach more importance to immediate than to remote rewards. In some cases, we might also need to invoke the general shape of the discounting function to explain why they sometimes change their minds, without specifying the value of the parameters. The trick – more a craft than a science – is to know how much detail to provide for a given task. When great detail is required, it is a warning sign that the task may be unfeasible.
Some final comments are called for with regard to unconscious mental states and mental operations. In this book I shall repeatedly refer to the unconscious workings of the mind. Dissonance reduction (Chapter 9), wishful thinking (Chapter 7), and transmutation of motives (Chapter 9), for instance, are caused by unconscious mechanisms. We may not understand well how they operate, but I find it impossible to deny that they exist. Many have also argued for the existence of unconscious mental states. Although these mechanisms and states cannot be accessed directly, they may, like dark matter and dark energy in the universe, be identified by their effects. In principle, one might also access them through neuroimaging.
The existence of unconscious mental mechanisms is relatively uncontroversial. The existence of unconscious mental states is a less straightforward issue. Self-deception, unlike wishful thinking, presupposes that there are unconscious beliefs. On this question, the jury is still out (see Chapter 7). Freud thought that we all have a number of unconscious and unavowable desires, as instantiated, he claimed, by the Oedipus complex. These are speculative and largely unproven ideas. The evidence for unconscious emotions and prejudices is stronger.
To the extent that unconscious mental states have causal efficacy, it should be possible to identify them by their effects. If a denial of a statement is disproportionately strong, for instance, we might infer that it is one that the person in question really, although unconsciously, believes: “The lady doth protest too much, methinks.” There is a story (which I have been unable to track down) told about Sigmund Freud, who was invited to meet a prominent person, Dr. X, in the international Jewish movement. During their conversation, Dr. X asked him, “Tell me, Dr. Freud, who in your opinion is the most important Jewish personality in the world today?” Freud answered politely, “Why, I think that must be yourself, Dr. X.” When Dr. X replied, “No, no,” Freud asked, “Wouldn't ‘No’ have been enough?” Double negation can be equivalent to affirmation. The Proustian character of Legrandin (Chapter 9) offers a fine-grained illustration of this effect.
One can also identify unconscious prejudices by their effects. In an Implicit Association Test, experiments subjects were first asked to classify rapidly (by tapping their left or right knee) each of a list of names into those that are most often considered black (such as Malik and Lashonda) and those that are most often seen as white (such as Tiffany and Peter). Next they were asked to classify rapidly each of a list of words as pleasant in meaning (such as “love” and “baby”) or unpleasant (such as “war” and “vomit”). Next, they classified a randomly ordered list that included all of the black names, white names, pleasant words, and unpleasant words. First they were asked to tap their left knee for any black name or unpleasant-meaning word and their right knee for any white name or pleasant-meaning word. Second, the instructions were changed. They were asked to tap their left knee for white names and unpleasant words and their right knee for black names and pleasant words. It took about twice as long to respond to the second task, even though objectively the tasks were of equal difficulty. In theory, one might use such tests in a court of law, to determine whether the “disparate impact” of a racially neutral law was due to “disparate treatment” based on an unconscious prejudice.
Unconscious emotions can often be identified by observers who infer their existence from the characteristic physiological or behavioral expressions. Most of us have heard and many of us uttered the angry statement “I am not angry.” Envy can manifest itself in a sharpness of tone and a tendency to adopt a derogatory slant that are obvious to observers but not to the subject. In Le rouge et le noir Mme de Rênal discovers her feelings for Julien Sorel only when she suspects that he might be in love with her chambermaid, one emotion (jealousy) thus revealing the presence of another (love).
Self-deception (see Chapter 7) is more problematic in this respect. Suppose I form and then repress the belief that my wife is having an affair with my best friend. Although unconscious, the belief that they are lovers might still guide my actions, for example, by preventing me from going to the part of town where my friend lives and where I might risk seeing my wife visiting him. This may sound like a plausible story, but to my knowledge there is no evidence that unconscious beliefs have causal efficacy. Many arguments for the existence of self-deception rely on (1) the exposure of the person to evidence strongly suggesting a belief he or she would not want to be true and (2) the fact that the agent professes and acts on a different and more palatable belief. To obtain direct evidence for the unconscious persistence of the unpalatable belief one would need to show (3) that it, too, is capable of guiding action, as in the hypothetical example just given. To repeat, I do not know of any demonstration to this effect.
My hunch is that the phenomenon does not exist. It is a pleasant conceit to imagine that my unconscious beliefs could be the handmaiden of my conscious ones, by steering me away from evidence that might undermine them, but it is no more than an unsupported just-so story. Along equally speculative lines, one could imagine that “the unconscious” is capable of inducing indirect strategies (one step backward, two steps forward), for instance, by making a child hurt herself to get the attention of her parents. These suggestions make the unconscious too similar to the conscious mind, by making it capable of having representations of the future and of other people's actions and intentions. Mental states of which we are unaware may cause spontaneous actions such as answering, “No, no” instead of simply “No,” but I do not know of any evidence that they can also cause instrumentally rational behavior.
In particular, there is no evidence that the unconscious is capable of making intertemporal trade-offs, as some economists have claimed.6 According to one argument, workers form motivated beliefs about job safety if the benefit of holding the belief exceeds the cost. If the psychological benefit of suppressing one's fear exceeds the cost due to increased chances of accident, the worker will believe the activity to be safe. According to another argument, people may form exaggerated (but motivated) beliefs about the dangers of addiction. If I want to quit using drugs but find that my beliefs about their dangerous effects are insufficiently dissuasive, I may adopt the belief that they are more dangerous than I currently believe they are, since this belief would motivate me to suffer the withdrawal pains. Everything we know about addiction suggests the opposite, however: addicts persuade themselves that the drug is less dangerous than they have reason to think it is. More generally, there is no evidence that the unconscious can weigh the present benefits of false beliefs against the future costs of holding them, or the present costs of false beliefs against the future benefits of holding them.
Bibliographical note
Evidence for anchoring of probability assessments is given in A. Tversky and D. Kahneman, “Judgment under uncertainty: heuristics and biases,” Science 185 (1974), 1124–31. Evidence for the anchoring of preferences is given in D. Ariely, G. Loewenstein, and D. Prelec, “Coherent arbitrariness,” Quarterly Journal of Economics 118 (2003), 73–105. The reference to Bishop Andrewes is from A. Nicolson, God's Secretaries (New York: HarperCollins, 2003), and that to the royal healing from M. Bloch, Les rois thaumaturges (Paris: Armand Colin, 1961). The examples about the relation between beliefs and behavior in Indochina and in the Roman Empire are taken from P. Veyne, L'empire gréco-romain (Paris: Seuil 2005), pp. 531–8. A good discussion of sentimentality is M. Tanner, “Sentimentality,” Proceedings of the Aristotelian Society n.s. 77 (1976–7), 127–47. The evidence on the impact of price on pleasure from drinking wine is in H. Plassmann et al., “Marketing actions can modulate neural representations of experienced pleasantness,” Proceedings of the National Academy of Sciences 105 (2008), 1050–4. A large-scale (internet) experiment on unconscious prejudices is reported in B. Nosek, M. Banaji, and A. Greenwald, “Harvesting implicit group attitudes and beliefs from a demonstration website,” Group Dynamics 6 (2002), 101–15. For the argument that Freud made the unconscious too similar to the conscious mind, see L. Naccache, Le nouvel inconscient (Paris: Odile Jacob, 2006). Arguments by economists about intertemporal trade-offs in the unconscious are found in G. Akerlof and W. Dickens, “The economic consequences of cognitive dissonance,” American Economic Review 72 (1982), 307–19, and in G. Winston, “Addiction and backsliding,” Journal of Economic Behavior and Organization 1 (1980), 295–324.
1 We must assume that the event in question is one that, if it occurred, would not affect him personally, such as the discovery of life on other planets. If the event is the victory of his favorite sports team, he might bet money that it will lose so that regardless of what happens he will have something to be pleased about.
2 This statement is probably too strong. Manipulating the procedure might elicit any probability assignment between 50 percent and 80 percent but none outside that range. In that case, we would be justified in asserting that the subject believes that the event is more likely to occur than not but is not certain that it will. This assessment is far more coarse-grained, however, than what is needed in standard models of decision making.
3 In one of his many ironic comments on the fragility of faith, Gibbon notes that the “bishops of Tour and Milan pronounced, without hesitation, the eternal damnation of heretics; but they were surprised, and shocked, by the bloody image of their temporal death.”
4 The idea of religion as the “opium of the people” also suggests that it is a consumption good rather than a premise for action. It could, however, be a premise for inaction.
5 A related phenomenon occurs when people take a third-person perspective on themselves. A proverb says, “Virtue does not know itself.” Also, one cannot coherently assert one's own naivety, since the very idea presupposes lack of self-consciousness. And as Nero Wolfe put it, in one of Rex Stout's novels featuring him, “To assert dignity is to forfeit it.”
6 As we shall see in Chapter 9, the founder of the theory of cognitive dissonance also made this claim.
4
This chapter and the two following ones will be devoted to varieties of motivation. In the present chapter, the discussion is fairly general. In the following, I focus on two specific issues, selfishness versus altruism and temporal shortsightedness versus farsightedness. These two issues complement each other to some extent, the latter being as it were the intrapersonal and intertemporal version of the former, interpersonal contrast. More importantly, they are also substantially related, in the sense that farsightedness can mimic altruism.
The set of human motivations is a pie that can be sliced any number of ways. Although none of them can claim canonical status, there are four approaches that I have found useful. The first proposes a continuum of motivations, the second and the third both offer a trichotomy, and the fourth a simple dichotomy. The classifications are both somewhat similar and interestingly different, allowing us to illuminate the same behavior from different angles. Following the discussion of these typologies, I offer some further comments on motivations.
From visceral to rational
On September 11, 2001, some people jumped to their death from the World Trade Center because of the overwhelming heat. “This should not be really thought of as a choice,” said Louis Garcia, New York City's chief fire marshal. “If you put people at a window and introduce that kind of heat, there's a good chance most people would feel compelled to jump.” There was no real alternative. Subjectively, this may also be the experience of those who drink seawater when freshwater is unavailable. They may know that drinking even a little seawater starts you down a dangerous road: the more you drink, the thirstier you get. Yet the temptation may, for some, seem irresistible. The craving for addictive substances may also be experienced in this way. An eighteenth-century writer, Benjamin Rush, offered a dramatic illustration: “When strongly urged, by one of his friends, to leave off drinking [a habitual drunkard] said, ‘Were a keg of rum in one corner of a room, and were a cannon constantly discharging balls between me and it, I could not refrain from passing before that cannon, in order to get at the rum.’” Sexual desire may also be so overwhelming as to silence more prudential concerns.
Some emotions may also be so strong as to crowd out all other considerations. The feeling of shame, for instance, can be unbearably painful, as shown by the 1996 suicide of an American navy admiral who was about to be exposed as not entitled to some of the medals he was wearing, or by the six suicides in 1997 among Frenchmen who were exposed as consumers of pedophiliac material. Anger, too, may be overwhelmingly strong, as when Zinedine Zidane on July 9, 2006, in the last minutes of the World Cup soccer final, head-butted an Italian opponent to retaliate against a provocation, under the eyes of seventy thousand people in the stadium and an estimated one billion TV viewers worldwide. Had he paused for a fraction of a second to reflect, he would have realized that the action might cost the defeat of his team and the ruin of his reputation.
Except perhaps for the urge to jump from the World Trade Center, it is doubtful whether any of these desires was literally irresistible, in the way a boulder rolling down a hillside might be irresistible to a person trying to stop it in its course. (An urge to fall asleep may be irresistible, but falling asleep is not an action; that is why attempts to do so are self-defeating.) Addicts are somewhat sensible to costs: they consume less when prices go up.1 People in lifeboats sometimes can prevent each other from drinking seawater. Sexual temptation and the urge to kill oneself in shame are certainly resistible. Because of their intensity, these visceral cravings nevertheless stand at one extreme of the spectrum of human motivations. They have the potential, not always realized, for blocking deliberation, trade-offs, and even choice.
At the other extreme, we have the paradigm of the rational agent who is unperturbed by visceral factors, including emotion. He acts only after having carefully – but no more carefully than is warranted under the circumstances – weighed the consequences of each available option against one another. A rational general, chief executive officer, or doctor is concerned merely with finding the best means to realize an objective goal such as winning the war, maximizing profit, or saving a life. The visceral roots of the desires do not enter into the equation.
An example of the distinction between visceral and rational motivation is provided by the difference between visceral and prudential fear. Although it is common to refer to fear as an emotion, it may be only a belief–desire complex. When I say, “I fear it is going to rain,” I mean only that I believe it is going to rain and that I wish it were not going to rain. If the “fear” inspires action, as when I take an umbrella to protect me against the rain, it is a paradigm of rational behavior (Chapter 13). None of the characteristic features of the emotions (Chapter 8) is present. Visceral fear, by contrast, may induce action that is not instrumentally rational. It has been calculated, for instance, that 350 Americans who would not otherwise have died lost their lives on the road by avoiding the risk of flying after September 11, 2001. By contrast, it does not seem that the Spanish incurred excess deaths by switching from train to car after the attacks on trains in Madrid on March 11, 2004. It is possible that because of the long run of attacks by Basque Homeland and Freedom (ETA), the population had developed an attitude of prudential rather than of visceral fear toward terror bombings. For them, terrorist attacks may have been just one risk among others, similar to – albeit more dangerous than – the risk of rain.
When Franklin Roosevelt wrote that “the only thing we have to fear is … fear itself – nameless, unreasoning, unjustified terror which paralyzes needed efforts to convert retreat into advance,” he probably had in mind rational fear of visceral fear. When Montaigne wrote “It is fear that I am most afraid of,” adding that “fear banishes all wisdom from the heart,” the context suggests that he referred to the same idea. In such cases, a person who rationally fears that he might at some future time be subject to irrational fear, can take precautions against the tendency, by not exposing himself to situations that might trigger fear or by preventing himself from acting out of fear. An admiral might, for instance, burn his ships to prevent himself (or his sailors) from taking flight in a panic. I discuss such strategies of “imperfect rationality” in Chapter 15.
Between the extremes of the visceral–rational continuum, we find behavior that is partly motivated by visceral factors, yet is also somewhat sensitive to cost–benefit considerations. A man may seek revenge (a visceral desire), yet also bide his time until he can catch his enemy unawares (a prudential concern). If he challenges his enemy to a duel (as required by norms of honor), he may take fencing lessons in secret (a dishonorable but useful practice). If a person is made an offer that is both unfair and advantageous, in the sense that she would be better off taking it than not, she might accept it or reject it, depending on the strength of her interest versus the strength of her resentment (Chapter 19). In more complex cases, one visceral factor might counteract another. The desire for an extramarital sexual affair might be neutralized by guilt feelings. An urge to flee from the scene of battle may be offset or preempted by an urge to fight caused by anger at the enemy, by the fear of being shamed by one's comrades, or by the fear of being shot for desertion.
Interest, reason, and passion
In their analysis of human motivations, the seventeenth-century French moralists made a fruitful distinction among interest, reason, and passion. Interest is the pursuit of personal advantage, be it money, fame, power, or salvation. Even action to help our children counts as the pursuit of interest, since our fate is so closely bound up with theirs. A parent who sends his children to an expensive private school where they can get the best education is not sacrificing his interest but pursuing it. The passions may be taken to include emotions as well as other visceral urges, such as hunger, thirst, and sexual or addictive cravings. The ancients also included states of madness within the same general category because, like emotions, they are involuntary, unbidden, and subversive of rational deliberation. For many purposes, we may also include states of intoxication among the passions. From the point of view of the law, anger, drunkenness, and madness have often been treated as being on a par.
Reason is a more complicated idea. The moralists mostly used it (as I shall use it here) in relation to the desire to promote the public good rather than private ends. Occasionally, they also used it to refer to long-term (prudential) motivations as distinct from short-term (myopic) concerns. Both ideas may be summarized under the heading of impartiality. In designing public policy, one should treat individuals impartially rather than favoring some groups or individuals over others. Individuals, too, may act on this motivation. Parents may sacrifice their interest by sending their children to a public school, because they believe in equality of opportunity. At the same time, policymakers as well as private individuals ought to treat outcomes occurring at successive times in an impartial manner by giving each of them the same weight in current decision making, rather than privileging outcomes in the near future. In fact, some moralists argued, a concern with long-term interest will also tend to promote the public good. At the Federal Convention in Philadelphia, for instance, George Mason argued that
we ought to attend to the rights of every class of people. He had often wondered at the indifference of the superior classes of society to this dictate of humanity & policy; considering that however affluent their circumstances, or elevated their situations, might be, the course of a few years, not only might but certainly would, distribute their posterity throughout the lowest classes of Society. Every selfish motive therefore, every family attachment, ought to recommend such a system of policy as would provide no less carefully for the rights and happiness of the lowest than of the highest orders of Citizens.
Either form of impartiality has degrees. The strength of concern for others tends to vary inversely not only with genealogical distance, but with geographical remoteness. Similarly, even prudent individuals usually give somewhat more weight to the near future than to the more remote, a fact that is only partly explained by their knowledge that they might not live to enjoy the distant future.
In addition to the requirement of impartial ends, reason, in its consequentialist form (see later discussion), may be defined by a rational choice of means. Although the terms “reason” and “rationality” are sometimes used interchangeably, I propose (Chapter 13) a conception of rationality that applies only to means, not to ends. Reason, by contrast, applies to both.
As an example of how behavior may be understood in terms of any of these three motivations, we may cite a 1783 letter from New York chancellor Robert Livingston to Alexander Hamilton in which he comments on the persecution of those who had sided with the British during the wars of independence:
I seriously lament with you, the violent spirit of persecution which prevails here and dread its consequences upon the wealth, commerce & future tranquility of the state. I am the more hurt at it because it appears to me almost unmixed with purer patriotic motives. In some few it is a blind spirit of revenge & resentment, but in more it is the most sordid interest.
The phrases I have italicized correspond to reason, emotion, and interest, respectively. The adjectives are telling: reason is pure, passion is blind, interest is sordid. In Chapter 25 I illustrate the distinction with examples from constitution making.
Id, ego, superego
In his analyses of human motivations, Freud also suggested three basic forms, each of them linked to a separate subsystem of the mind. The three systems are the id, the ego, and the superego, corresponding, respectively, to the pleasure principle, the reality principle, and conscience. The id and the superego represent, respectively, impulses and impulse control, while the ego, “helpless in both directions … defends itself vainly, alike against the instigations of the murderous id and against the reproaches of the punishing conscience.” In a more illuminating statement from the same essay (“The Ego and the Id”), Freud wrote that the ego is “a poor creature owing service to three masters and consequently menaced by three dangers: from the external world, from the libido of the id, and from the severity of the superego.” Yet even this formulation does not capture fully what I think is the useful core of Freud's idea. This is the proposition that as the ego is navigating the external world (the reality principle) it also has to fight a two-front war against the impulses from the id (pleasure principle) and the punitively severe impulse control exercised by the superego (conscience).2
This proposition was novel, profound, and true. What it lacks is a mechanism. Why could not the ego itself exercise whatever impulse control might be needed? Why do morality and conscience so often take the form of rigid rules? Do we need to stipulate the existence of separate and quasi-autonomous mental functions? It took the pioneering work of George Ainslie to provide satisfactory answers to these questions. I discuss his views in Chapter 15. Here I only want to draw attention to the fact that many impulses need to be kept at bay because of the cumulative damage they can do if unchecked.3 On any given occasion, drinking or eating to excess, splurging, or procrastinating (such as failing to do one's homework) need not do much harm to the agent. The damage occurs after repeated excesses (or repeated failures). The focus of impulse control, therefore, must not be the individual occasion, since the person can always say to himself or herself that a new and better life will begin tomorrow. Impulse control must address the fact that the impulse will predictably arise on an indefinite number of occasions. The solution arises from reframing the problem, so that failure to control an impulse on any one occasion is seen as a predictor of failure to control it on all later occasions. “Yes, I can postpone impulse control until tomorrow without incurring important harm or risk, but why should tomorrow be different from today? If I fail now, I shall fail tomorrow as well.” By setting up an internal domino effect and thus raising the stakes, the agent can acquire a motivation to control her impulses that would be lacking if she just took one day at a time. The other side of the coin is that the control must be relentless and, as the Victorian moralists put it, “never suffer a single exception.”
Taking account of consequences
Finally, motivations may be consequentialist or non-consequentialist, that is, oriented either toward the outcome of action or toward the action itself. Much of economic behavior is purely consequentialist. When people put aside money for their old age or stockbrokers buy and sell shares, they attach no intrinsic value – positive or negative – to these actions themselves; they care only about the outcomes. By contrast, the unconditional pacifist who refuses to do military service even against the most evil enemy takes no account of the consequences of his behavior. What matters for him is that certain actions are unconditionally forbidden, such as taking a human life. It is not that he is unaware of the consequences, as may be the case in emotional action, only that consequences make no difference for what he does.
Public policy may also be adopted on either type of motivation. A policymaker might adopt the principle “Finders keepers” (e.g. in patent legislation), on the assumption that if the person who discovers a new valuable resource is assigned the property right in it, more valuable resources will be discovered. This is a consequentialist argument. A non-consequentialist argument for the same policy might be that the person who discovers a new resource, whether it be a piece of land or a cure for cancer, has a natural right to property in it. For another example, we may consider the speech (XXXI) of Dion Chrysostomos against the practice of the Rhodeans to reuse old bronze statues to honor benefactors of the city: this, he argued, was both to violate the rights of those in whose honor the statues had originally been erected and to discourage potential new benefactors who knew that statues erected in their honor might soon be recycled in favor of someone else. Consequentialist arguments may (seem to) warrant harsh measures toward terrorists even if the steps that are taken violate the non-consequentialist values associated with human rights and civil liberties.4
A special case of non-consequentialist motivation is the principle that I shall refer to variously as everyday Kantianism, the categorical imperative, or magical thinking: do what would be best if all did the same. In one sense this principle is linked to consequences, since the agent does what would bring about the best outcome if everybody else did the same. These are not the consequences of her action, however, but of a hypothetical set of actions by her and others. In a given case, acting on the principle could have disastrous consequences for all if others do not follow suit. In the international arena, unilateral disarmament is an example.
Another case is the following principle of Jewish ethics. Suppose that the enemy is at the door and says, “Give me one among yourselves to be killed and we shall spare all the others; if you refuse, we shall kill you all.” The Talmud requires that in such cases the Jews let themselves all be killed rather than name one to be killed so that others can be saved. If, however, the enemy says, “Give me Peter” under the same conditions, it is acceptable to hand him over. There is not a ban on causing a person to be killed to save others, but on selecting who it shall be. The novel Sophie's Choice presents the same dilemma.
Social norms (Chapter 21) offer a further special case of non-consequentialist behavior, with an important twist. Social norms tell people what to do, such as take revenge for an insult or refrain from eating a kid boiled in its mother's milk, not because there are any desirable results to be brought about, but because the action is mandatory in itself.5 While not taken to bring about any outcome, such actions may be seen as undertaken to prevent an outcome, namely, being blamed by others for not taking them. We may then ask, however, whether the blaming is also undertaken for similar consequentialist reasons. In general, I shall argue, they are not. Moreover, when people are hurt by the actions of others they retaliate even in one-shot interactions under full anonymity, such as may be obtained in experimental settings. Because the interaction is one-shot, they have nothing to gain in later encounters, and because it is anonymous, they need not fear the blame of third parties. I shall return to these experiments in several later chapters.
Even for a professed non-consequentialist, consequences may matter if they are important enough. Consider a principle that many would consider an unconditional one, the ban on torturing small children. In a “ticking bomb” scenario, imagine that a necessary and sufficient condition for preventing the detonation of a nuclear device in central Manhattan is to torture a terrorist's small child in her presence. If the scenario could be made credible, many non-consequentialists might acquiesce in the torture. Others would say that since the conditions in the scenario will never obtain in practice, the absolute ban remains in effect. Still others would ban the torture even if the scenario did occur. My task here is not to argue for one of these conclusions, but to make the empirical observation that in real-life situations stakes are rarely so high and knowledge rarely so certain as to force the non-consequentialist to consider the consequences of his behavior. It is possible that with more at stake or more definite knowledge he would abandon his principles, but since the situation does not arise we cannot tell for certain whether we are simply dealing with a very steep trade-off or with a total refusal to engage in trade-offs.
These four approaches to motivation capture some of the same phenomena. Visceral factors, passions, and the pleasure principle clearly have much in common. The last applies to a wider range of cases, because it involves pain avoidance as well as pleasure seeking. When students procrastinate in doing their homework, it is not necessarily because there is something else they passionately want to do. Often, they are merely taking the path of least resistance. The superego, reason, and non-consequentialist motivations also have some features in common. Although not all systems of morality are rigid and relentless, some are. Kant's moral theory is a notorious instance. (In fact, his moral philosophy may have originated in the private rules he made for himself to control his impulses, such as his maxim of never smoking more than one pipe after breakfast.) At the same time, morality can rise above rigidity, in individuals not subject to ambiguity aversion. The toleration of ambiguity is, in fact, often said to be the hallmark of a healthy ego. By contrast, the relation among rationality, interest, the ego, and consequentialism is more tenuous. It would be absurd to claim that the hallmark of a healthy ego is the rational pursuit of self-interest.
Wanting and wishing
We often think of motivations as taking the form of wanting to bring about some state of affairs. They may also, however, take the form of wishing some state of affairs to obtain, whether or not there is something one can do to bring it about.6 This distinction between wants and wishes is important if we look at the motivational component of emotion (Chapter 8). As both Seneca and Adam Smith noted, emotions can, in fact, be accompanied either by a want to do something or by a wish that something be the case. In anger or wrath, A's urge to take revenge on B cannot be satisfied by C's doing to B what A had planned to do or by B's suffering an accident.7 What matters is not simply the outcome, that B suffer, but that he suffer by A's agency, and that B knows that A is the author of his suffering.8 In the Rhetoric, Aristotle quotes Homer to this effect: Ulysses makes sure that Polyphemus knows who blinded him, “as if Ulysses would not have been avenged unless the Cyclops perceived both by whom and for what he had been blinded.” In Racine's Andromaque, Hermione says that her vengeance on Pyrrhus will be “lost unless he knows, when dying, that it was I who murdered him.”
By contrast, in hatred what matters is that the hated person or group disappear from the face of the Earth, whether this happens by my agency or by someone else's. Similarly, whereas love simply induces a wish for the person I love be happy, gratitude also makes me want the other person's happiness to be due to my agency, and for her to know that it is. In sadism, what matters is to make the other suffer; in malice, it is merely that the other should suffer. Adam Smith observed that a person to whom it would be “agreeable … to hear that a person whom he abhorred and detested was killed by some accident … would reject with horror even the imagination” of causing that event. This is even clearer in envy. Many people who would enjoy seeing a rival's losing his possessions and would do nothing to prevent it from happening if they could, would never take active steps to destroy them, even if it could be done without costs or risks to them.9 A person who would not set her neighbor's house on fire might abstain from calling the fire brigade if she saw it burning.
Wishful thinking (Chapter 7) is based on wishes rather than on wants. In some cases, the agent refrains from the hard work of making the world conform to his desires and adopts instead the easy path of adopting an appropriate belief about the world. If I desire to be promoted but am reluctant to make an effort, I may rely instead on insignificant signs to persuade myself that a promotion is imminent. In other cases, acting on the world is not an option. I may be unable to cause my love to be requited or my sick child to recover. In such cases, I may either engage in gratifying fantasies or face the facts. A further distinction is between cases in which the fantasies have no further consequences for action and those in which they are used as premises for behavior. I may delude myself into thinking that a woman of my acquaintance harbors a secret passion for me and yet not make any overtures to her, either because I am constrained by morality (or self-interest) or because the deluded belief is entertained mainly for its consumption value. The delusion may also be expressed overtly to its object, as happened when a secretary of John Maynard Keynes's told him that she could not help being aware of his great passion for her. Her life was ruined.
States that are essentially by-products
A factor that complicates the wish-want distinction is that in some cases I can get X by doing A, but only if I do A in order to get Y. If I work hard to explain the neurophysiological basis of emotion and succeed, I may earn a high reputation. If I throw myself into work for a political cause, I may discover at the end of the process that I have also acquired a “character.” If I play the piano well, I may impress others. These indirect benefits are parasitic on the main goal of the activity. If my motivation as a scholar is to earn a reputation, I am less likely to earn one. To enter a political movement solely for the sake of the consciousness-raising or character-building effects on oneself is doomed to fail, or will succeed only by accident. In a passage that has become proverbial, Seneca claimed that “glory follows those who avoid it.” Proust noted that a musician “may sometimes betray [his true vocation] for the sake of glory, but when he seeks glory in this way, he moves further away from it, and only finds it by turning his back on it.” Self-consciousness interferes with the performance. As he also wrote, although “the best way to make oneself sought after is to be hard to find,” he would never give anyone advice to that effect, since “this method of achieving social success works only if one does not adopt it for that purpose.”
Musical glory or social success falls in the category of states that are essentially by-products – states that cannot be realized by actions motivated only by the desire to realize them. These are states that may come about, but not be brought about intentionally by a simple decision. They include the desire to forget, the desire to believe, the desire to desire (such as the desire to overcome sexual impotence), the desire to sleep, the desire to laugh (one cannot tickle oneself), the desire to ignore someone,10 and the desire to overcome stuttering. Attempts to realize these desires are likely to be ineffectual and can even make matters worse. It is a commonplace among moralists and novelists that intentional hedonism is self-defeating,11 and that nothing engraves an experience so deeply in memory as the attempt to forget it. Although we may wish for these states to be realized, we should beware of wanting to realize them.
Many people care about salvation (in the afterlife) and redemption (for wrongs they have done). They may also believe they can achieve these goals by action. To die the death of a martyr in the fight against the infidels may provide the passport to heaven, or so some believe. To fight against the Nazis after having collaborated with them at an earlier stage may redeem the wrongdoing. Yet if these actions are undertaken for the purpose of achieving salvation or redemption, they may fail. In Catholic theology, the intention to buy a place in heaven by voluntary martyrdom would be an instance of the sin of simony. According to the most rigorous mystic doctrine, only work for the salvation of others is permitted. (Could this be a social contract: I pray for your salvation, you pray for mine?) Some Islamic scholars make a similar criticism of suicide attackers who are motivated by the belief that they will get a privileged place in paradise. Montaigne writes that when the Spartans “had to decide which of their men should individually hold the honor of having done best that day, they decided that Aristodemus had the most courageously exposed himself to risk: yet they never awarded him the prize because his valor had been spurred on by his wish to purge himself of the reproach he had incurred in the battle of Thermopylae.” The French press magnate Jean Prouvost, who had collaborated with the German forces during the occupation of France, tried to redeem himself by writing a large check to the resistance when it became clear that the Germans were losing the war. After Liberation, the High Court granted him a non-lieu (a judgment that suspends, annuls, or withdraws a case without bringing it to trial), something the Spartans presumably would not have done.12
The Palo Alto school of psychiatry has emphasized the importance of states that are essentially by-products, and notably how people can tie themselves into mental knots by trying to achieve directly what can only arise indirectly. Injunctions such as “Be spontaneous!” or “Don't be so obedient!” and statements such as “Nice girls don't even think about sex” can paralyze those to whom they are addressed. Stendhal was obsessed by the impossible goal of appearing to be natural. When he wrote that “all I need to be certain of achieving [social] success is to learn to show my indifference,” he ignored the fact that the will to appear indifferent is inconsistent with true indifference. As Montaigne wrote, anticipating what psychologists call the “white bear effect,” “there is nothing which stamps anything so vividly on our memory as the desire not to remember it.”
Push versus pull
Why do people leave one country for another? Why do academics leave one university for another? Why do peasants leave the countryside for the city? Often, answers are classified as “push versus pull.” One may emigrate either because the situation at home is unbearable or because the situation abroad is irresistibly enticing – at least this is a common way of viewing the matter. In many situations, however, it is misleading. Typically, people move because they compare the situation at home and abroad and find that the difference is big enough to justify a move, even taking account of the costs of the move itself.13 Yet it can make sense to distinguish push motives from pull motives, when the former are closer to the visceral end of the continuum and the latter closer to the rational end. People in the grip of strong fear sometimes run away from danger rather than toward safety. The only thought in their mind is to get away, and they do not pause to think whether they might be going from the frying pan into the fire. Depending on the drug, addicts can be motivated either by the pull from euphoria (cocaine) or by the push from dysphoria (heroin). Suicidal behavior, too, may owe more to push than to pull. It is an escape from despair, not a flight to anything.
The operation of social norms (Chapter 21) can also be viewed in terms of push versus pull. The desire to excel in socially approved ways exercises a strong pull on many individuals, whether they strive for glory (being the best) or for honor (winning in a competition or combat). Other individuals are more concerned with avoiding the shame attached to the violation of social norms. In some societies, there is a general norm that says, “Don't stick your neck out.” To excel in anything is to deviate, and deviation is the object of universal disapproval: “Who does he take himself for?” The relative strength of these two motivations varies across and within societies. Classical Athens illustrates the competitive striving for excellence.14 In modern societies, small towns often show the stifling effects of the hostility to excellence. To risk a generalization, overall the push from shame seems to be a more important motivation than the pull toward glory, which is not to say that the latter cannot be powerful.
Motivational conflict
The existence of competing motivations is commonplace:
I need a book so strongly that I am tempted to steal it from the library, but I also want to behave morally.
In the face of a bully I am both afraid and angry: I want to run but also to hit him.
I want all children to have public education, but I also want my child to go a private school to obtain the best education.
I want a candidate who favors legal abortion, but I also want one who favors lower taxes.
I want to smoke, but also to remain healthy.
If I am made an advantageous but unfair offer, “take it or leave it,” I want both to reject it because it is unfair and to accept it because it is advantageous.
I want to donate to charity, but also to promote my own interest.
I am tempted to have an extramarital affair, but I also want to preserve my marriage.
How is the conflict among these motivations resolved? A general answer might go as follows. Where the situation is one of “winner takes all,” so that no (physical) compromise is possible, the strongest motivation wins. If my concern for my child is stronger than my concern for the schooling of children in general, I will send him or her to a private school. If my pro-choice concern is stronger than my tax-cut concern and no candidate favors both positions, I vote for a pro-choice candidate who proposes to raise taxes. If somebody offers me $3 out of a common pool of $10, intending to keep the rest for himself, I accept it. If I am offered only $2, I reject the offer. When compromise is possible, the stronger motivation has a stronger impact than the smaller one. A smoker may cut down his cigarette consumption from thirty to ten cigarettes a day. As a reflection of the strength of my altruism, I may spend 5 percent of my income on charity.15
This answer is not exactly wrong, but it is pretty simplistic, since the idea of “strength of motivation” is more complicated than these quick examples suggest. A motivation may owe its strength to its sheer psychic force; this is the sense in which, for instance, visceral motives are often stronger than what Madison called “the mild voice of reason.” A strong motivation may also, however, be one that the agent endorses strongly because of the high value placed on it in her society. Each society or culture is in fact characterized by a normative hierarchy of motivations (Chapter 9). Other things being equal, a person would rather perform a given action for motive A than for motive B if A ranks higher in the hierarchy. These are metamotivations, needs to be animated by desires of a certain kind.16 Even though weaker in the visceral sense, they may in the end win out over other motivations.
Interest and passion, notably, often show a certain deference to reason.17 As Seneca said, “Reason wishes the decision that it gives to be just; anger wishes to have the decision which it has given seem the just decision.” As there are very many plausible-sounding conceptions of reason, justice, and fairness, it will indeed often be possible to present a decision made in anger as conforming to reason. The trials of collaborators in countries that had been occupied by Germany during World War II were in many cases anchored in a deep desire for revenge. Yet because of their deference to reason, combined with their desire to demarcate themselves from the lawless practices of the occupying regimes, the new leaders presented the severe measures as justice-based rather than emotion-based. A person may have a first-order interest in not donating to charity and a second-order desire for not seeing himself as swayed by interest only. In deference to reason, he may then adopt the philosophy of charity (Chapter 2) that can justify small donations. If others give much he will adopt a utilitarian policy that justifies small donations, and if others give little he will adopt a fairness-based policy that justifies the same behavior.
In these cases, reason has no independent causal role. It only induces an after-the-fact justification for actions already decided on other grounds. The conflict is not resolved, but swept under the carpet. I cite further examples of such rationalization in Chapter 7. In other cases, the search for a reason-based justification may change behavior. If I adopt a fairness-based policy of charity because others give little and they suddenly begin donating much more generously than before, I have to follow suit. The same need for self-esteem that caused me to justify self-interested behavior by impartial considerations in the first place also prevents me from changing my conception of impartiality when it no longer works in my favor. We may imagine that in King Lear both Burgundy and France initially fell in love with Cordelia because of her prospects, but that only the former cared so little about his self-image that he was able to shed the emotion when it no longer coincided with his interest. This is a case of interest paying deference to passion rather than to reason, suggesting that passion, or rather this particular passion, ranks above interest in the normative hierarchy. Other passions, such as envy, might well rank below interest. We might then observe efforts to undertake only such envy-based action as may be plausibly presented as interest-based. Actions that cannot be viewed in that light will not be undertaken.
Cognitive dissonance theory predicts that when one motivation is slightly stronger than another, it will try to recruit allies so that the reasons on one side become decisively stronger, causing any lingering doubts to disappear. The unconscious mind shops around, as it were, for additional arguments in favor of the tentative conclusion reached by the conscious mind. In such cases, “strength” of motivation cannot be taken as given, but should rather, to some extent at least, be seen as a product of the decision-making process itself. Suppose that when buying a car I attach values to differently weighted features (speed, price, comfort, appearance) of each of the alternatives and reach an overall assessment by comparing the weighted sums of the values. I might, for instance, attach overall value of 50 to brand A and of 48 to brand B. Because of the uncomfortable closeness of the comparison, I unconsciously modify the weights so that A becomes a clear winner, with 60 versus 45 as overall value. Before making the purchase, I come across brand C, which with the old weights would have scored 55, but with the new only achieves 50. Had I seen the alternatives in the order C–A–B, I would have chosen C. Because I met them in the order A–B–C, I chose A. Such path-dependence undermines the simple idea that motivational conflicts are resolved according to given motivational strength.
On what I called the simplistic view, the decision whether to steal a book from the library might be represented as follows. On the one side of the balance is the benefit of being able to use the book; on the other side, the cost of guilt feelings. What I end up doing depends only on whether the cost exceeds the benefit, or vice versa. But this cannot be right, for suppose someone offered me a “guilt pill” that would remove any painful feelings of guilt for stealing the book. If guilt entered into my decisions merely as a psychic cost, it would be rational to take the pill, just as it would be rational to take a pill that would prevent me from developing a hangover from a planned drinking binge. I submit, however, that most people would feel just as guilty about taking the pill as they would about stealing the book.18 I am not denying that there cannot be, in some sense, a trade-off between morality and self-interest, only that it cannot be represented in this simplistic way.
Here is a more complex case. I wish that I did not wish that I did not want to eat cream cake. I want to eat cream cake because I like it. I wish that I did not like it, because, as a moderately vain person, I think it is more important to remain slim. But I wish I were less vain. But is that wish activated only when I want to eat cream cake? In the conflict among my desire for cream cake, my desire to be slim, and my desire not to be vain, the first and the last can form an alliance and gang up (or sneak up) on the second. If they catch me unaware, they may succeed, but if I understand that the salience of my desire not to be vain is caused by the desire for cake I may be able to resist them. On another occasion, my desire for short-term gratification and my long-term desire for spontaneity may form an alliance against my medium-term desire for self-control. When more than two motives bear on the choice between two options, the idea of “strength of motivation” may be indeterminate until we know which alliance will be formed.
The seventeenth-century French moralist La Bruyère summarized two forms of motivational conflict as follows: “Nothing is easier for passion than to overcome reason; its greatest triumph is to conquer interest.” We have seen that when passion “overcomes reason,” it may still want to have reason on its side. Although St. Paul said, “For I do not do the good I want, but I do the evil I do not want,” a more common reaction may be to persuade oneself of the goodness or justice of what, in the grip of passion, one wants to do. When passion “conquers interest,” it can do so in one of two ways. The agent may, because of the urgency that is typical of emotion (Chapter 8), not take the time to find out where her interest lies. Alternatively, the force of emotion may be so strong that she knowingly acts against her interest. Such behavior may amount to weakness of will (Chapter 6).
Material and formal preferences
Sometimes, it is useful to represent motivations as preferences. In Chapter 13 and in Chapter 24 I discuss how, respectively, individual and collective choice can be explained in terms of preferences. Here, I propose a distinction between material and formal preferences.
Material preferences are defined over all sorts of tangible and intangible options. In everyday life, we constantly express preferences over consumption goods, such as different varieties of fruit or different car brands, and explain our choices by citing these preferences. One may also prefer betraying one's country to betraying a friend (E. M. Forster), having loved and lost to never having loved at all (Tennyson), living in the desert to living with a contentious and angry woman (Proverbs 21:19), or burning out to fading away (Neil Young). Some of these preferences are wishes, defined over states of affairs, others are wants, defined over actions.
Formal preferences include attitudes toward other people (altruism or selfishness), attitudes toward time (patience or impatience), and attitudes toward risk (risk aversion, risk neutrality, or risk seeking).19 I discuss them in, respectively, Chapter 5, Chapter 6, and Chapter 13. Altruism can be measured, at a first approximation, by how much personal welfare an agent is willing to sacrifice to increase the welfare of another person by one unit. Impatience can be measured by how much an agent is willing to pay for getting an immediate reward rather than suffer a delay. Suppose she is indifferent between receiving $100 – x today or $100 in ten days. The larger is x, the more impatient is the person. If x = 0, she is perfectly patient. (Are there cases in which x < 0?) Risk aversion can be measured by the extent an agent A has to be compensated for risk. Suppose a lottery ticket will yield either $100 or $0 with 50 percent probability each. An agent is willing to pay up to $50 – x for the ticket. If x > 0, the agent is risk-averse, the more risk-averse the larger is x. For x = 0, the agent is risk-neutral. For x < 0, he is risk-seeking.
A choice may involve both material and formal preferences, if I am given the choice between one orange today and two apples tomorrow or between one apple with certainty and a 50 percent chance of two oranges. We are probably less good at making such comparisons than at comparing options that differ only in the material or only in the formal respect. We are often able to state quite clearly what we prefer when other things are equal, but become confused when they are not (“trade-off aversion”).
Folk psychology tends to assume that formal preferences are character traits that shape the behavior of an individual across the board (see Chapter 12). There is some evidence, however, that formal preferences are domain-specific. Most obviously, the degree of altruism depends on our closeness to the recipient. The ability to defer gratification – an expression of patience – differs when the future good is a sum of money and when it is health-related. Casual observation suggests that rock-climbers do not enjoy risk-taking in everyday life.
Bibliographical note
A theory of visceral motivations is offered by G. Loewenstein, “Out of control: visceral influences on behavior,” Organizational Behavior and Human Decision Processes 65 (1996), 272–92. The estimate of “excess car accidents” after September 11, 2001, is from G. Gigerenzer, “Dread risk, September 11, and fatal traffic accidents,” Psychological Science 15 (2004), 286–7. The lack of similar excess accidents in Spain is documented in A. López-Rousseau, “Avoiding the death risk of avoiding a dread risk: the aftermath of March 11 in Spain,” Psychological Science 16 (2005), 426–8. The trichotomy interest–reason–passion is analyzed in A. Hirschman, The Passions and the Interests (Princeton University Press, 1977); M. White, Philosophy, The Federalist, and the Constitution (Oxford University Press, 1987); and in my Alchemies of the Mind (Cambridge University Press, 1999). George Ainslie's Picoeconomics (Cambridge University Press, 1992) provided the lacking mechanisms for Freud's insights. A classic study of push versus pull is D. Gambetta, Did They Jump or Were They Pushed? (Cambridge University Press, 1983). I take the arguments of Dion Chrysostomos from P. Veyne, L'empire gréco-romain (Paris: Seuil, 2005), p. 217. The principle I cite from Jewish ethics is explored in D. Daube, Collaboration with Tyranny in Rabbinic Law (Oxford University Press, 1965), and D. Daube, Appeasement or Resistance (Berkeley: University of California Press, 1987). The passages from Aristotle and Andromaque are lifted from F. Heider, The Psychology of Interpersonal Relations (Hillsdale, NJ: Lawrence Erlbaum, 1958), p. 265. I develop the idea of states that are essentially by-products in Chapter 2 of Sour Grapes (Cambridge University Press, 1983) and apply it to the question of redemption in “Redemption for wrongdoing,” Journal of Conflict Resolution 50 (2006), 324–38, and to the question of salvation in “Motivations and beliefs in suicide missions,” in D. Gambetta (ed.), Making Sense of Suicide Missions (Oxford University Press, 2005). See also L. Ross and R. Nisbett, The Person and the Situation (Philadelphia: Temple University Press, 1991), pp. 230–2. I discuss the “deference to reason” in trials of collaborators after World War II in Chapter 8 of Closing the Books (Cambridge University Press, 2004). A good introduction to the Palo Alto school of psychiatry is P. Watzlawitz, The Pragmatics of Human Communication (New York: Norton, 2011). Evidence for change in the weights attached to various features of alternative options is found in A. Brownstein, “Biased predecision processing,” Psychological Bulletin 129 (2003), 545–68, and J. Brehm, “Postdecision changes in the desirability of alternatives,” Journal of Abnormal and Social Psychology 52 (1956), 384–9. Evidence for domain-specific patience is found in G. Chapman, “Your money or your health: time preferences and trading money for health,” Medical Decision Making 22 (2002), 410–16.
1 That might also be, however, because their budget does not allow them to consume at the same level (Chapter 10).
2 To combine two of Freud's metaphors, the ego is like a rider on an unruly horse (the id) who is at the same time ridden by an incubus (the superego).
3 There is also a fact of cumulative risk. The chance of unwanted consequences from unprotected sex may be small on any given occasion, but the lifetime risk might be considerable. On any given trip, the chance of being injured in a car accident while not wearing a seatbelt is small, but the lifetime probability is about one in three.
4 The parenthetical “seem to” reflects the possible operation of the “psychology of tyranny” (Chapter 2). A classical dilemma of deterrence is that the hatred it inspires may in the end more than offset the fear it is intended to cause.
5 With regard to the rules of kosher food, of which the ban on eating a kid boiled in its mother's milk is only a historical illustration, it was thought for a while that they were justified on hygienic grounds. Today, as far as I have been able to learn, they are recommended on the grounds that it is good for one to do something that is both difficult and pointless. This idea seems to embody the fallacy of by-products that I discuss later: behavior that is justified merely by its character-building effects will not even have those. I do not know, though, how many of those who follow the rule do it for this reason.
6 Directors of funeral parlors may wish for people to die, but do not actively try to create clients. Although the wish may seem harmless, if morally repulsive, Seneca tells about a successful suit by Demades (fourth century BC) against a mortician on the grounds that he had hoped for great profits and could only have attained that aim the by the deaths of many.
7 Hume says, though, that “the distress which Charles [VII of France] had already suffered, had tended to gratify the duke [of Burgundy]'s revenge.”
8 Experiments show, though, that A will punish B even if A knows that B will not know that he is being punished, although, perhaps surprisingly, the punishment is weaker than when A knows that B will know.
9 Some envious people, to be sure, have no such qualms. They may live in a society where little shame attaches to envy or they may just be shameless.
10 According to La Bruyère, “A women who always has her eye on the same person, or always avoids looking at him, makes us think the same thing about her.”
11 In the final volume of À la recherche du temps perdu Proust, probably reflecting on his own life, wrote that the vain search for happiness can nevertheless lead to insights into the human condition that may “offer a kind of joy.” The search for states that are essentially by-products can bring them about indirectly, as when a child's instructing someone to laugh causes the person to laugh out loud at this preposterous demand.
12 The reason he went free was probably that the resistance needed the money and later found itself obliged to keep the tacit promise of immunity that acceptance of the check implied.
13 This formulation presupposes that the cost of moving enters on a par with the benefits of having moved, as determinants of the overall utility of moving. Yet the costs of moving can also enter as constraints on the decisions. If the cheapest transatlantic fare costs more than the maximal amount a poor Italian peasant can save and borrow, he will remain in Italy no matter how much better he could do for himself in the United States (Chapter 10).
14 Aeschylus, for instance, wrote his plays for performance at a dramatic competition. When the young Sophocles defeated him, he was so chagrined that he left Athens for Sicily.
15 In Chapter 6 I discuss the more puzzling phenomenon of “loser takes all” observed in weakness of will.
16 The idea of metamotivations is unrelated to the concept of metapreferences. An example of the latter would be a person who had two different preference orderings, one for eating over dieting and one for dieting over eating, and a metapreference favoring the latter. Following La Bruyère's insight that “men are very vain, and of all things hate to be thought so,” a metamotivation could amount to a preference for preferring dieting to eating on grounds of health over having the same preference for dieting on grounds of vanity (see Chapter 9).
17 As we shall see in Chapter 14, agents may also show a sometimes excessive deference to rationality.
18 In Chapter 13, I argue that a person who had a short time horizon would, for somewhat similar reasons, refuse to take a “discounting pill” that would make him attach more importance to future consequences of present actions. The general principle illustrated by these two pills is that a rational person would not want to do in two steps what he would not want to do in one step. He might, to be sure, want to do in two steps what he could not do in one.
19 Risk aversion should not be confused with loss-aversion, the tendency for losses to be weighted more heavily than equal-sized gains (Chapter 14).