The Folly of Fools (23 page)

BIASED ENCODING AND INTERPRETATION OF INFORMATION

Assuming we do attend to incoming information, we can still do so in a biased way. One experiment invited people to look at a figure that could be either a capital B or the number 13 (or a horse or a seal) and were told the stimulus could be either a letter or a number (or a farm animal or ocean animal). Having been provided differential food reward for the general categories ahead of time, people quickly developed a sharp perceptual bias in the appropriate direction on items presented for only four hundred milliseconds, that is, ones just reaching consciousness. Eye tracking showed that the first look was usually toward the preferred category (about 60 percent). These studies suggest that the impact of motivation on processing information extends to preconscious processing of visual stimuli and thus guides what the visual system presents to conscious awareness. Similar work has now been done using colors.

The point is that our perceptual systems are set up to orient very quickly toward preferred information—in this case, shapes associated with food rewards. This itself has nothing to do with deceit and self-deception—it will often give direct benefits. But the same quick-biasing procedure is available to us when the information is preferred because it boosts our self-esteem, or our ability to fool others. There are few more powerful forces in the service of self-deception than personal fantasies, so when these are aroused, selective attention is expected to be especially intense.

The related effect was shown sixty years ago: hungrier children, when asked to draw a coin, draw coins larger. Instruments to gain satisfaction (money buys food) are more attractive and are perceived as being larger. Recent confirmation shows that a glass appears larger to you when you are thirstier, especially if attention is called to your thirst, and even garden implements appear larger if gardening has subliminally been linked to suggestions that it is fun.

Our initial biases may have surprisingly strong effects. In one experiment, people were preselected for strong attitudes for and against capital punishment. They were then presented with a mixed bag of facts supporting both positions. Instead of leading to group cohesion, this action split the group more sharply. Those who were already against capital punishment now had a new set of arguments at hand, and vice versa. Biased interpretation ran the process. Those in favor of capital punishment accepted pro arguments as sound and rejected anti arguments as unsound. As before, self-affirming thoughts were negatively associated with this behavior—think better of yourself, you practice less self-deception. One important implication is that self-deception is a force that often drives people apart—certainly friends, lovers, neighbors—although under common group aims, such as war, shared self-deceptions are also uniquely powerful in binding people together.

BIASED MEMORY

There are also many processes of memory that can be biased to produce welcome results. We more easily remember positive information about ourselves and either forget the negative or, with time, transmute it to be neutral or even positive. Differential rehearsal, as in telling others, can itself produce the effect, an example of self-deception at the end of the process (the “tooter”) affecting earlier processes. Complementary memory biases may actively work in the same direction. When given a “skills class,” people remember their skills prior to the class as being worse than they rated them at the time, probably to create an illusion of progress. They then later misremember their actual performance after the class as being better than it was, presumably in service of the same delusion. What we are doing here is producing a consistent set of biases in our own favor by a series of biased memories.

Memories are continually distorted in self-serving ways. Men and women both remember having fewer sexual partners, and more sex with each partner, than was actually true. People likewise remember voting in elections they did not and giving to charity when they did not. If they did vote, they remember supporting the winning candidate rather than the one they actually voted for. They remember their children as being more precocious and talented than they were. And so on.

Although people often think of memory as a photo whose sharpness gradually degrades with time, we know that memory is both reconstructive and easily manipulated. That is, people continually re-create their own memories, and it is relatively easy to affect this process in another person. If a police officer asks a witness about a nonexistent red sports car right near an accident, the officer will often learn about the red sports car in subsequent questioning—it can sometimes end up as one of the most vividly remembered details of the accident itself. As mentioned, differential rehearsal of material after the fact can produce reliable biases in memory.

Take another example. Health information can easily be distorted in memory even when it is presented in a clear and memorable fashion. People were given a cholesterol screening and then one, three, and six months later tested for the memory of the result. Respondents usually (89 percent) recalled their risk category accurately and their memory did not decay with time, but more than twice as many people remembered their cholesterol level as lower rather than higher than it actually was. This same kind of memory bias is true of daily experiences in which people recall their good behavior more easily than bad but show no such bias in recalling the behavior of others.

Or we can invent completely fictitious memories. As has been said, “My memory is so good I can remember things that never happened.” One case is memorable in my own life. For many years I told the story of how in 1968 I went deep into the bowels of Harvard’s Widener Library to find a book coauthored by my father in 1948, published by the State Department, which laid out the de-Nazification procedures for all Nazis too unimportant to be hung at Nuremberg. It was a complex system of graded steps. If you were a member of SA, two slaps on the wrist; if SS, you lost your job for five years—that kind of thing. Yet almost none of this is true. No such book exists. Yes, the trip to the bowels took place and a book on Nazis was located with my father as a coauthor and it was published by the State Department. Only it was published in 1943 and is a minor piece on the structure of Nazi organizations in Nazi-occupied territories. Hardly the basis for the reinvention of Germany, but is this not the point of false memory—to improve things, especially appearances? I added nice little touches along the way. I liked to say that I trusted no one, including myself, and thus went to the bowels to see whether this family story was true. But this added to the falsehood, since there really was no “family story” about this minor 1943 work, and is a general feature of false-memory construction—new details are added that support the general argument and then become part of memory.

One can even reverse exactly who is saying what to whom. Gore Vidal remembers an interview with Tom Brokaw on NBC’s morning
Today Show
in which Brokaw began by asking about Vidal’s writings on bisexuality, to which Vidal replied that he was there to talk politics. Brokaw persisted with bisexuality; Vidal stood firm until they concentrated on politics. Yet years later, when Brokaw was asked what his most difficult interview had been, he cited his interview with Vidal. Reason: Vidal kept insisting they talk about bisexuality when all he wanted to discuss was politics. Positions exactly reversed—and, as expected, in the service of self-improvement: Brokaw looks better being interested in politics than in bisexuality.

In arguments with other people, lab work shows that we naturally tend to remember the good arguments on our side and the poor ones on the other, and to forget those that turn out badly for us and good for the other. This bolsters our own side and image, of course, which presumably is its function. Memory distortions are more powerful the more they are motivated to maintain our self-esteem, to excuse failures or bad decisions, and to push into the deeper past causes of current problems. Thus, most people maintain the illusion of improvement, where such mistakes as must be acknowledged can at least be attributed to the failings of an earlier version of oneself.

RATIONALIZATION AND BIASED REPORTING

We reconstruct internal motives and narratives to rationalize otherwise bad or questionable behavior. We can attribute behavior to external contingencies rather than internal, thereby helping defend ourselves. So a general belief that cheating is not bad—or is unintentional or occurs in a world without free will—will all serve to rationalize our cheating, as indeed they do.

Biases show up in unexpected places, even when there are no clear benefits or costs at issue. The classic experiment in this domain was beautifully designed to put people in an awkward situation with one of two escapes. People were offered the chance to sit next to a crippled person or one who was not. Each was watching a television set in front of him or her. Sometimes the two TVs had the same show, sometimes different ones. When it was the same show, people preferentially chose to sit next to the handicapped person, as if demonstrating their lack of bias, but if the two TVs had different shows, people chose to sit away from the crippled person, as if now having a justification (more interesting show) for an otherwise arbitrary choice. Similarly, a meta-analysis of many studies shows that white Americans choose to help black Americans more or less equally (compared to helping whites) but not when they can rationalize less helping on grounds such as distance or risk. Here people are not denying or misremembering their behavior—rather, they are denying the underlying intention and rationalizing it as the product of external forces. This has the advantage of reducing their responsibility for behavior performed.

A belief in determinism can provide a ready excuse for misbehavior, just as can unconsciousness: the “I had no choice” defense. Relatively deterministic views of human behavior may provide some cover for socially malevolent behavior. Experimentally inducing a deterministic view (reading an essay on how genes and environment together determine human behavior) increases cheating on a computer-based task that permits cryptic cheating. What this work shows is that by manipulating a variable that reduces personal responsibility, we easily induce immoral behavior in ourselves (at least as viewed by others).

PREDICTING FUTURE FEELINGS

It is an interesting fact that we show systematic biases in our ability to predict our own future feelings. We make systematic errors in the process, under the general rule that what we are feeling at the present will extend into the future. When imagining a good outcome, we overestimate our future happiness, and vice versa for a bad one. It is as if we assay our current feelings and then project them into the future. We do not imagine that we will “regress to the mean,” that is, return naturally to the average value of happiness. We do not assume we will be less happy in the future than our current state of happiness or happier in the future if we are currently down. Thus, one week after the 2004 US elections, Kerry supporters were less dejected than they thought they would be and Bush supporters, less ecstatic.

There is evidence that we make similar mistakes when trying to predict the feelings of others, whether friends or strangers. We overestimate the effect of an emotional event on their future feelings, much as we do for ourselves. Indeed, our forecasting of them is positively correlated with their own, but neither is very predictive of the future.

The problem is in the interpretation. Some see this as a form of self-deception in which we are unconscious of the degree to which our system of self-deception will readjust our thinking in the future. I doubt this. We project easily into the future because it expresses our current emotional state. Verbal predictions regarding our future mental states may be a relatively recent invention with limited selective effects. The relevant trade-offs are already built into our behavior whatever our verbal predictions.

Certain exceptions to this rule also stand out. I remember “courting” a Nigerian beauty at a very safe distance at a club in Amsterdam for three hours without ever having the courage to approach her. When she left, she threw me a look of withering contempt that burned right into my soul. If a social psychologist had been there to measure my “affective forecasting,” I doubt I would have guessed that twenty-five years later, the memory still sears in my consciousness. I believe I would have predicted that within a year or two the whole evening would have been completely forgotten.

ARE ALL BIASES DUE TO SELF-DECEPTION?

A hallmark of self-deception is bias. Mere computational error is not enough. Such error is often randomly distributed around the truth and shows no particular pattern. Self-deception produces biases, patterns where the data point in one direction—usually that of self-enhancement or self-justification. Are there biases that are real but not driven by self-deception? Of course there are.

Consider the following. Sounds that are coming toward us are perceived as closer and louder than they really are, while the opposite is true for receding sounds. This is a bias and it has a perfectly good explanation. Approaching objects are inherently more dangerous than are receding ones—hence the value of earlier and more acute detection. Perhaps the organism is measuring distances in Darwinian units rather than Newtonian ones. From that viewpoint, there is no bias.

Or consider another example. From the top of a tree, the drop to the ground looks much farther than does the same distance viewed from the ground up. There is no social component to these biases. You are directly saving yourself—not trying to manipulate the opinions of others. Many other errors have similarly innocent explanations. Some are simple optical illusions, holes in our sensory system that produce startling biases under particular conditions. Others are general rules that work well in most situations but fail badly in some.