Strangers to Ourselves Page 8
There is increasing evidence that positive and negative information is processed in different parts of the brain, though the extent to which these different brain regions map onto conscious versus nonconscious processing is unclear. There is at least the possibility that the adaptive unconscious has evolved to be a sentry for negative events in our envi- ronments.19
Is the Adaptive Unconscious Smart or Dumb?
So which part of the mind is smarter, anyway? This question has been posed by several researchers, notably the social psychologist Anthony Greenwald. Greenwald concluded that unconscious cognition is a rather primitive system that can analyze information in only limited ways. He suggested that modern research has revealed a very different kind of unconscious from the Freudian unconscious, one that is considerably less clever.
Greenwald focused mostly on research that presents words to people at speeds too fast to be perceived consciously. Several studies have found that such subliminally presented words can influence people's responses to some extent. For example, Draine and Greenwald presented people with words on a computer (e.g., "evil," "peace") and asked them to make very quick judgments of whether they were good or bad in meaning. Unbeknownst to participants, these words were preceded by very fast presentations of "priming" words that were also good or bad in meaning. The prime words were flashed so quickly that people did not see them consciously. Nonetheless, they influenced people's responses to the second, target words. When the prime word was opposite in valence to the target word-for example, when "peace" was preceded by a subliminal presentation of "murder"-people were more likely to make a mistake and judge "peace" as bad. When the prime word was the same valence as the target-for example, when "peace" was preceded by a subliminal presentation of "sunset"-people made very few mistakes in judging "peace" as good. Most psychologists view this as evidence that people unconsciously saw the subliminal word and processed its mean ing, which either interfered with or helped their judgment of the second word.20
Greenwald notes, however, that the unconscious mind's ability to recognize and process subliminally presented words is limited. There is no evidence, for example, that it can perceive the meaning of a two-word sequence that is different from the meaning of each individual word. Consider the words "enemy loses," which have a positive meaning when read as a unit, but a negative meaning when each word is considered individually. When two-word sequences such as this are flashed subliminally, people extract the meaning of the individual words (negative, in the example above), not the meaning of the unit. Hence, the unconscious mind may have limited cognitive abilities.
This conclusion is at odds, however, with much of what we have just reviewed-for example, research showing that the nonconscious mind is superior to the conscious mind in detecting covariations in the environment. It is no surprise, perhaps, that our minds can make limited judgments of information that it saw for only a few hundredths of a second. What is more surprising is that it can detect any meaning from a word that is flashed so quickly. In fact, a point that is often overlooked is that the unconscious mind is doing a superior job to the conscious mind on these tasks. Even if it is making only udgments of subliminally flashed words, it is still doing better than the conscious mind, which has no idea that it saw anything at all. On these tasks, the unconscious mind is a lot smarter than the conscious interpreter.
What about when people have more time to examine and process incoming information? As we have seen, the nonconscious mind still outperforms the conscious self on at least some tasks, such as covariation detection. One study found, for example, that people could learn a complicated rule in which the presentation of a stimulus on one trial depended on what had been presented seven trials earlier, even though they could not consciously remember what had been presented that long ago.21
To be sure, the adaptive unconscious can be rigid and inflexible, clinging to preconceptions and stereotypes even when they are disconfirmed, in contrast to the more flexible conscious mind. There is no single answer to the question of how smart or dumb each system is-it depends on what you ask them to do. The adaptive unconscious is smarter than the conscious mind in some ways (e.g., detecting covariation), but less smart in other ways. The bottom line is that it is different, and whether we assign the labels "smart" or "dumb" to these differences is arbitrary. A more useful approach is to map out the differences and try to understand the functions of the two systems. The adaptive unconscious is an older system designed to scan the environment quickly and detect patterns, especially ones that might pose a danger to the organism. It learns patterns easily but does not unlearn them very well; it is a fairly rigid, inflexible inferencemaker. It develops early and continues to guide behavior into adulthood.
Rather than playing the role of CEO, the conscious self develops more slowly and never catches up in some respects, such as in the area of pattern detection. But it provides a check-and-balance to the speed and efficiency of nonconscious learning, allowing people to think about and plan more thoughtfully about the future.
It is tempting to view the tandem of nonconscious and conscious thinking as an extremely well-designed system that operates optimally. But this would be a mistake. First, there was no grand design. In real engineering, old designs can be completely thrown out and new ones started from scratch. The Wright brothers, for example, did not take a horse buggy and stick some wings on it to make a flying machine; they were able to begin afresh and build every part of their plane with the final goal (to fly) in mind. By contrast, natural selection operates on the current state of an organism, such that new systems evolve out of old ones. It is not as if someone sat down in advance and drew up the blueprints for the grand design of the human mind. Evolution works with what it has.
The human mind is an incredible achievement, perhaps the most amazing in the history of the Earth. This does not mean, however, that it is an optimal or perfectly designed system. Our conscious knowledge of ourselves can be quite limited, to our peril.
Knowing Who We Are
Our greatest illusion is to believe that we are what we think ourselves to be.
-H. F. Anriel, The Private Journal of Henri Frederic Amiel (1889)
We tell ourselves stories in order to live ... We live entirely, especially if we are writers, by the imposition of a narrative line upon disparate images, by the "ideas" which we have learned to freeze the shifting phantasmagoria which is our actual experience.
-Joan Didion, The White Album (1979)
In the play Pygmalion, Henry Higgins succeeds in transforming Eliza from a crude flower girl into a refined and lovely lady-while failing to do anything about his own unsavory personality. Higgins is convinced that he is a gracious, fair-minded, cultured English gentleman with the most honorable of intentions, failing to see that he is coarse, misogynous, controlling, and fussy. After his housekeeper, Mrs. Pearce, chastises him for swearing, using his dressing gown as a napkin, and putting a saucepan of porridge on the clean tablecloth, Higgins is genuinely perplexed. He remarks to his friend Colonel Pickering: "You know, Pickering, that woman has the most extraordinary ideas about me. Here I am, a shy, diffident sort of man. I've never been able to feel really grown-up and tremendous, like other chaps. And yet she's firmly persuaded that I'm an arbitrary overbearing bossy kind of person. I can't account for it."
How can Higgins be so blind to the nature of his own personality? Freudian repression might be the culprit; viewing himself as a refined English gentleman, instead of looking in the mirror and seeing himself as he really is, might allow him to avoid considerable psychic pain.
There may, however, be a simpler explanation. Many of people's chronic dispositions, traits, and temperaments are part of the adaptive unconscious, to which they have no direct access. Consequently, people are forced to construct theories about their own personalities from other sources, such as what they learn from their parents, their culture, and yes, ideas about who they prefer to be. These constructions may be driven less by repressi
on and the desire to avoid anxiety than by the simple need to construct a coherent narrative about ourselves, in the absence of any direct access to our nonconscious personalities. Like Henry Higgins, people often construct narratives that correspond poorly to their nonconscious dispositions and abilities.
This is surprising because one of the main things people want to know about themselves is the core of their personality. "Am I a truly honest person?" "Do I have what it takes to be a successful teacher?" "Am I capable of being a good parent?" This is the self that people want to uncover when they ask, "Who am I?" It is the self that the Greek oracle at Delphi advised people to know and the self to which, according to Shakespeare, we should all be true.
But it makes little sense to talk about a single "self" when we consider that both the adaptive unconscious and the conscious self have regular patterns of responding to the social world. This distinction has largely been overlooked by psychological theories of personality.
The Current State of Personality Psychology
Gordon Allport defined personality as the psychological processes that determine a person's "characteristic behavior and thought"-a definition that is as good today as it was when Allport proposed it.2 Few issues are so fundamental or have received as much attention as those concerning the nature of human personality-and few have been as controversial. The field of personality psychology is made up of a fragmented collection of conflicting approaches that disagree on such basic questions as, is there a single, core self that determines people's behavior? If so, what is it, and how can it be measured?
Consider these brief sketches of the major approaches to human personality. Classic psychoanalytic theory argues that the defining feature of personality is how people deal with their repressed drives, such as sexual and aggressive impulses. The battles, compromises, and truces among the id, ego, and superego define who we are. This is the only major approach to personality that stresses the importance of unconscious forces in shaping who the person is. At the opposite end of the mentalism continuum is behaviorism, which asks why we should look inward at the person when it is behavior we want to predict. Though dwindling in number, there are still behaviorists who focus solely on the external contingencies that determine behavior, rather than internal, psychological constructs.
Midway on the mentalism continuum is the phenomenological approach, which argues that to understand why people do what they do, we must view the world through their eyes, examining each person's unique construals of herself and the meaning she finds in her social world. Many social psychologists have adopted this approach by studying the self-concept, which consists of people's beliefs about who they are. For the most part, researchers have assumed that people are aware of their construals, although this approach has typically skirted questions about consciousness.
In recent years the dominant area of personality research has been the trait approach, which attempts to isolate a small number of basic personality traits that are common to all people. This approach is less concerned with theories about the origins of traits and more with quantitative analyses of the results of personality tests, on which people rate their own or others' personalities. Sophisticated analyses have uncovered five basic traits: extraversion, emotional stability, agreeableness, conscientiousness, and openness to experience. These traits are viewed as the fundamental building blocks of personality that everyone possesses to some degree; the particular constellation of people's standing on these traits defines their core or "true" self. The trait approach has been adopted by behavior geneticists, who study the extent to which human personality traits are heritable (largely by comparing the personalities of identical twins reared in different families). Typically, genetic factors have been found to account for 20-50 percent of the variance in personality traits.'
In contrast to all these approaches, postmodernists argue that there is no single, coherent personality or self. In today's complex world, the argument goes, people are subjected to a multitude of conflicting influences, making it very difficult to have a single, unified sense of "me." The self may be fluid, changing as our culture, roles, and context change, and attempts to measure and define a core set of traits that people carry with them is meaningless.'
MISCHEL AND THE EMPEROR'S CLOTHES
These major approaches have little in common and make fundamentally different assumptions about the nature of personality. Further, in a review of personality research published in 1968, Walter Mischel found that none of the approaches met the gold standard of personality research very well, namely Allport's criterion of predicting with any certainty what people actually do. An extravert should make friends more easily than an introvert, whereas a conscientious person should meet more deadlines than a person who is not conscientious. Mischel found, however, that the typical correlation between personality traits and behavior was quite modest. This news shook up the field, because it essentially said that the traits personality psychologists were measuring were just slightly better than astrological signs at predicting behavior.
Mischel did not simply point out the problem; he diagnosed the reasons for it. First, he argued that personality researchers had underestimated the extent to which the social situation shapes people's behavior, independently of their personality. To predict whether a person will meet a deadline, for example, knowing something about the situationthe consequences of not meeting it, how much time the person has, how much work remains to he done-may be more useful than knowing the person's score on a measure of conscientiousness. Situational influences can be very powerful, sometimes overwhelming individual differences in personality.'
This argument set off a turf war between personality psychologists, who place their bets on individual differences as the best predictors of behavior, and social psychologists, who place their bets on the nature of the social situation and how people interpret it. This war has often been waged in silly ways, with researchers in the two camps waving correlation coefficients and effect sizes at each other like sticks, arguing that theirs is bigger than the other camp's. Nonetheless, this battle was useful in revealing some important lessons. Personality variables, as traditionally conceived, are not all we need to know to predict human behavior.'
By criticizing trait research and pointing to the importance of the social situation, Mischel has often been portrayed as the Antichrist of personality theory. His second explanation of the low correlations between traits and behavior, however, is sometimes overlooked: personality is a good predictor of people's behavior; it's just that it has been conceptualized poorly. Ironically, it is Mischel and his colleagues who have demonstrated how individual differences can be conceptualized and measured in such a way that they account for impressive amounts of variance of behavior.
Rather than a collection of static traits that we can use to classify people, Mischel argued, personality is better conceived as a set of unique cognitive and affective variables that determine how people construe the situation. People have chronic ways of interpreting and evaluating different situations, and it is these interpretations that influence their behavior. Barbara's cognitive and affective personality system causes her to feel threatened when she suffers academic setbacks, and it is then that she is most likely to act aggressively. Sam's cognitive and affective personality system causes him to feel threatened when he perceives that he is being ignored by significant others, and that is when he is most likely to act aggressively. According to this view it makes little sense to try to classify how aggressive Barbara and Tom are on a single trait dimension; instead, we must understand how each person interprets and understands a social situation and acts accordingly.
It has not been entirely clear how aware people are of the operation of their cognitive and affective systems. In fact very few personality theories, with the exception of psychoanalysis, have said much about the role of conscious versus nonconscious processing. A recent collection of scholarly, cutting-edge articles on personality psychology takes up a full 967 p
ages and is touted by its publisher as "the most comprehensive single volume ever published on the subject." However, the index contains only two page references to consciousness and only six to unconsciousness. There are several more entries on psychoanalysis, but none on the modern, adaptive unconscious (or its synonyms).'
A lot of the confusion about personality and its relation to behavior has resulted from a failure to distinguish between the conscious and nonconscious systems. I believe that Mischel's cognitive and affective personality system is best thought of as part of the adaptive unconscious, whereas other personality theories have focused more on people's conscious construals of themselves.
TWO PERSONALITIES: THE ADAPTIVE UNCONSCIOUS AND THE CONSCIOUS SELF
My central thesis is that human personality resides in two places: in the adaptive unconscious and in conscious construals of the self. The adaptive unconscious meets Allport's definition of personality. It has distinctive, characteristic ways of interpreting the social environment and stable motives that guide people's behavior. These dispositions and motives are measurable with indirect techniques (i.e., not by self-report questionnaires). They are rooted in early childhood, are in part genetically determined, and are not easily changed.
But the conscious self also meets Allport's definition. Because people have no direct access to their nonconscious dispositions and motives, they must construct a conscious self from other sources. The constructed self consists of life stories, possible selves, explicit motives, selftheories, and beliefs about the reasons for one's feelings and behaviors. As Joan Didion says, "We tell ourselves stories in order to live."