Civilization is usually understood as the history of how people get along with each other—ways to foster cooperation while avoiding theft, rape, murder and war.  Governments and laws have evolved to maintain interpersonal harmony.  Each person’s problem of getting along with herself has been less obvious to modern thinkers.  From Jeremy Bentham’s utilitarianism through the behaviorist’s reinforcement to the modern economist’s revealed preference, a person’s only problem has been depicted as how to find rewarding circumstances and avoid punishing ones.  In their eyes, civilization comprises techniques for finding rewards and compacts for sharing them.

However, another task, more puzzling and thus more disturbing, has been recognized since earliest history:  How do we keep from undermining our own wishes?  This problem was so big and seemed so unnatural that it was given a religious name—sin, and the concepts in non-Abrahamic religions that are often translated as sin.  People tend to equate sin with crime, despite the efforts of theologians, but the victims of this kind of crime are usually the criminals themselves.  Of the seven deadly sins of Christianity, four—gluttony (?including addiction), sloth, pride, and envy—are said to damage only the self, and the others—avarice, wroth, and lust—damage the self concomitantly with others.  What was supposed to be wrong with sin was that it offended God or the gods, but how did we come up with that explanation?  Researchers have gone beyond the just-so stories of religion to detect a human readiness to perceive the presence of an Other  (Atran & Norenzayan, 2004), but why should the Other want us not to be lustful, lazy, etc. more than we want this ourselves?

Perhaps people do not know their best interests—that sloth brings poverty, lust brings entanglements, and pride goeth before a fall?  Or, more subtly, that pursuit of the most obvious rewards eventually hardens the heart?  But even ancient cultures knew these things and accordingly promoted dispositions to rise above sin.  Could one knowingly sin?  Socrates said no, but at the same time put his finger on why people did: 

Do not the same magnitudes appear larger to your sight when near, and smaller when at a distance?... Is not [the power of appearance] that deceiving art which makes us wander up and down and take the things at one time of which we repent at another?... Men err in their choice of pleasures and pains, that is, in their choice of good and evil, from defect of... that particular knowledge that is called measuring.  Plato's Protagoras (sections 356-357) in Jowett's translation (1892/1937).  His theories and Aristotle's are thoroughly discussed in Charlton, 1988, pp. 13-59.

This diagnosis suggests a simple cure: education.  But in centuries of trying to teach people rational self-interest, sages have not come close to eliminating the problem.  From early on they have excused their failure by positing a separate entity that demands our obedience—Plato’s wild chariot-horse of passion, the Devil, the id, down to a part of the brain that is wired to think hotly and rapidly (e.g. Metcalfe & Mischel, 1999) {Theories using dual choice mechanisms only roughly correct RCT}.  But such split decision-makers are neither necessary nor adequate. 

          

 Picoeconomics is a motivational science-- a science of how people choose things.  It grew out of a series of experiments in the mid-20th century to show how consequences affect choice.  That approach, known as behaviorism, developed precise ways of measuring motivation.  But it became obsessed with the notion that only external rewards and punishments matter—that any meaning arising inside the choice-maker is superfluous, maybe even illusory. 

Also in mid-century the same attitude developed in a sister science, economics: that personal intentions don’t matter, only preferences revealed by actual choices.  This view is known as ordinalism (e.g. Robbins, 1932/1985).   In both disciplines, the person was an indivisible whole.  How she might resolve conflicting estimates of utility was not the proper subject of science.

As you might expect, a reaction developed, at least in psychology.  This cognitive revolution saw the return of words like goals, suggestions, interpretations, intentions ((Gardner, 1985; Baars, 1986, pp. 4 – 10, 141-196).  But words like motivation, reward, and aversion disappeared.  The whole idea that choice is governed by a basic selection process based on differences in reward came to be viewed as a denial of people’s humanity—reductionism.

Economists have mostly continued to view choice as determined by the bottom line.  Even though they allow for emotions like fears, appetites, and altruistic wishes to be factored into choice, they have not wondered about the factoring process until recently. Mostly they have seen the person as externally motivated and indivisible. 

A new approach, game theory, gave behavioral scientists a way to study conflicts among whole  individuals, and, potentially, among differently motivated interests within one individual.  But you can’t study conflicts among interests without knowing what interests there might be.  Long ago, Freud made models of a person’s conflicting parts—he even called one model “economic”-- but he lacked both the measuring tools of behaviorism and analytical tools of economics.  His hypotheses about internal agents that fight for dominance fell into disuse.

Now behavioral scientists are making another try.  We can discern the conflicting interests within a person if we accept the behaviorists’ notion that reward is the absolute determinant of choice, but abandon their rule not to look inside the person.  I have called this approach picoeconomics (micro-micro-economics—Ainslie, 1986, 1992)—the strategic interaction of successive motivational states within the person.  Our challenge is to make a model with simple transactions and few assumptions, but which is still consistent with our humanity.

Brain imaging: Will involves synergy of choice bundling and moment-to-moment control of attention/emotion]  

There have been a number of psychology experiments on self-control and “willpower,” defined as resisting urges to get immediate rewards.  In a much discussed proposal, practicing this resistance is said to build up the “strength” of a contravening factor as in a muscle. (Baumeister et.al, 2006),  Practice leads to more skill in this exercise.  However, this is not intertemporal bargaining, which hinges on revaluation of the alternatives, and, if successful, reduces the need for other commitment mechanisms-- Self-control takes less effort when you have more confidence in a personal rule (e.g. Dar et.al., 2005).
Temptation experiments during brain imaging have begun to identify the neurophysiological components of self-control, which seems not to be a simple bidding process between SS and LL options.  Activity in the lateral prefrontal cortex, sometimes called an “executive center,” is clearly correlated with LL choice. (Chambers et.al., 2009; Crockett et.al., 2013; Monterosso et.al., 2010).  Furthermore, increased coordination of activity in this center with reward centers is observed to be proportional to LL choice, perhaps from suppression (“downregulation”) of reward center activity  (Hare et.al., 2014).  If these findings had been available when I wrote Breakdown of Will, I might have classified the suppression of a competing option as a short-acting form of emotion control (Section 5.1.3).  Food-tempted subjects who show this effect also use “attention filtering” (Section 5.1.2), which occurs faster and is also governed by the lateral prefrontal cortex (Harris et.al., 2013).  The latency of both kinds of response from the presentation of the options is less than a second, showing that they are part of the decision itself.  Thus self-control seems to be a complex process of 1. evaluating whether to control, which should include intertemporal bargaining and other framing effects, 2. filtering attention, 3. inhibiting appetite, and probably 4. evaluating whether #2 and #3 are working.  Reviewing examples of apparent downregulation of reward centers by other sites in the brain, Anouk Scheres and colleagues suggested that a self-control mechanism should replace both single currency and dual motivation models of choice (2013).  But although there are clearly neuronal steps between evaluation of options and behavioral choice, the evaluation of whether and how much to initiate these steps must still involve competing motivations.

In any case, the framing of evaluation can obviate much effort of will, as shown in another SS/LL experiment:  Subjects were variously reminded or not reminded of payoffs at both delays when they made their choices.  When they were reminded, they chose SS less, showed less activity in reward centers, and showed less downregulating activity in the lateral prefrontal cortex when choosing LL.  A simple reframing of the values of the options had reduced “willpower expenditure” (Magen et.al., 2014).   The bundling of choices in intertemporal bargaining looks like a similar process, but where in the brain such bookkeeping takes place remains a question.

In terms of adaptiveness, imagination is a two-edged sword

Even though most of the reward we seek is virtual, we have given very little thought to its properties.  First of all, how does it relate to the purpose we evolved to accomplish, the production and protection of offspring?  What are the potential implications for the survival of a species whose members can reward themselves by imagination?  Evolution has developed the reward process as a proxy for survival and reproduction, outcomes that are too great (often life vs. death) and too distant to select the behaviors of individual organisms.  To the extent that organisms are engineered to learn choices at all, they are engineered to maximize prospective reward.  “Realism”—the accurate prediction of what choices will have what consequences—is often a good way of procuring reward, but it is just a means to this end, and not the only possible means. 

Imagination is an ability that exists to an apparently limited extent in nonhuman animals.  Young mammals play, but seem to lose this tendency with age.  The likelihood that reward by imagination or play would be selected against in evolution is graphically illustrated by the case where strong reward is physically available ad lib from an animal’s own body.  Both male and female animals stimulate their own sex organs.  The more such reward can compete with the reward of actual sex, the more it fails in its function as a proxy for eventual adaptiveness.  The adaptiveness issue can be discerned in humans in how degrees of imaginativeness are distributed along a continuum ranging from fantasy-prone personalities, who are drawn readily into slightly suggested scenarios—sometimes mistaking them for reality (Wilson & Barber, 1983)—and sensation-seeking personalities, who are bored in the absence of concrete stimulation (Schalling et.al., 1983).

The power of imagination to subvert evolutionary purposes is potentially enormous.  Imagination has obvious adaptive advantages, for instance in foresight and instrumental problem solving.  But imagination has to have the power to reward in order to motivate people to use it.  And conversely, nature can afford to give it this power only if there is some factor that keeps it from running amok—to keep reward from short-circuiting its adaptive purpose as a selective factor for adaptive behaviors.

It is with this problem that hyperbolic discounting first seems to contradict evolution.  The conventional belief is that selective pressure has restricted imagination to the service of adaptive goals.  However, the mammalian reward system evolved when organisms did not have much imagination—when they could not do what Suddendorf & Corballis have called mental time travel (2007)—and thus were wired to learn only from outcomes that occurred soon after their choices.  There was no reason that valuation of prospective reward could not be governed by a hyperbolic curve rather than a “rational,” exponential curve.  The hyperbolic over-valuation of imminent rewards should not matter when animals’ long term interests are served by instincts that make long term preparations such as hoarding, dam building, and migrating rewarding in the short term.  However, with selection for increasing intelligence has come increasing imagination, and with imagination the unhitching of reward from adaptiveness, of short term from long term interests, and, as we shall see, of belief from truth.  

People have learned to mate without reproducing, fight without needing to, and commit themselves to costly hobbies that do not contribute to surviving offspring.  Furthermore, we have learned to rob future welfare for present pleasure, not just with addictive substances but also with socially accepted activities ranging in excitement level from death-defying adventure to simple procrastination (Ainslie, 2009 [McKay]).  Most importantly for the present discussion, we have learned to make occasions for current emotional reward from events that are not currently happening, in the form of memories, fantasies and beliefs.  Intelligence obviously has many adaptive features, but its ongoing evolution must be limited—probably has already been limited—by the availability of constraints on the urge for diverting long term resources to current consumption.  We have discussed some methods that are probably uniquely human both to restrain consumption by willpower and to make the reward for consumption more intense by cultivating appetite.  However, imagination does not depend on external events, and, as we will see, is hard to govern by willpower.  Its rewarding power is limited only by appetite, and methods of intensifying appetite can create puzzling non-physical addictions.