Perhaps the AAI taps people's chronic level of attachment that has become the signature of the adaptive unconscious, whereas self-report questionnaires tap people's conscious beliefs about their attachment relationships. But how can this be? Can we really have such disconnected systems that disagree on something as basic as an internal model of attachment relationships? The answer may be that we can, not only in the area of attachment but in other basic areas of personality as well.
DUAL MOTIVES AND GOALS
If we were to make a list of the goals that are most important in life, surely the desire for close relationships, success in life (e.g., a career), and power would make most people's short list. There is a long tradition in personality psychology of studying these three motives; indeed, psychologists such as H. A. Murray and David McClelland have argued that people's level of needs for affiliation, achievement, and power are major components of human personality.
There is growing evidence that these motives are an important part of the personality of the adaptive unconscious. Murray and McClelland assumed that these basic motives are not necessarily conscious and must therefore be measured indirectly. They advocated the use of the Thematic Apperception Test (TAT), in which people make up stories about a set of standard pictures, and these stories are then coded for how much of a need for affiliation, power, or achievement people expressed.
Other researchers have developed explicit, self-report questionnaires of motives, with the assumption that people are aware of their motives and can freely report them. A controversy has ensued over which measure of motivation is the most valid: the TAT or self-report questionnaires. The answer, I suggest, is that both are valid measures but tap different levels of motivation, one that resides in the adaptive unconscious and the other that is part of people's conscious explanatory system.
David McClelland and his colleagues made this argument in an influential review of the literature. First, they noted that the self-report questionnaires and the TAT do not correlate with each other. If Sarah reports on a questionnaire that she has a high need for affiliation, we know virtually nothing about the level of this need that she will express, non consciously, on the TAT. Second, they argued that both techniques are valid measures of motivation, but of different types. The TAT assesses implicit motives, whereas explicit, self-report measures assess self-attributed motives.
Implicit motives are needs that people acquire in childhood that have become automatic and nonconscious. Self-attributed motives are people's conscious theories about their needs that may often differ from their nonconscious needs. McClelland reports a study, for example, that measured people's need for affiliation with both the TAT and a selfreport questionnaire. People's affiliation needs, as assessed by the TAT, predicted whether they were talking with another person when they were beeped at random intervals over several days, whereas a self-report measure of affiliation did not. Affiliation needs as assessed with the selfreport measure were a better predictor of more deliberative behavioral responses, such as people's choices of which types of behaviors they would prefer to do alone or with others (e.g., visit a museum). The picture McClelland paints is of two independent systems that operate in parallel and influence different types of behaviors. In our terms, the adaptive unconscious and the conscious explanatory system each has its own set of needs and motives that influence different types of behaviors.
This separation between nonconscious and conscious motives may be very similar to the separation we encountered between nonconscious and conscious attachment styles. It is also characteristic of several other kinds of motives, such as dependency needs (people's desire to associate and interact with other people). A number of tests of dependency have been developed, some of which are explicit, self-report questionnaires and some of which are implicit, projective instruments. The two types of instruments are only moderately correlated and tend to predict different kinds of behavior. Further, women reliably score higher on explicit, conscious measures of dependency, whereas men tend to score higher on nonconscious measures. Indirect measures of dependency appear to tap nonconscious motives, whereas self-report questionnaires tap conscious, self-attributed motives.'"
DO WE SEE OURSELVES AS OTHERS SEE US?
If there are two sides to people's personality-a nonconscious and a conscious one, each producing unique behavior-then it is interesting to consider how other people get to know us. People could form impressions from our automatic, uncontrolled actions that reflect our implicit motives and traits (e.g., our implicit need for affiliation), or they could form impressions from our controlled, deliberative actions that reflect our explicit motives. It seems likely that people attend at least in part to behaviors that emanate from the adaptive unconscious (e.g., "Jim says that he's shy, but he's often the life of the party"). If so, other people might know us better than we know ourselves. As a character in Richard Russo's novel Straight Man said, "The truth is, we never know for sure about ourselves ... only after we've done a thing do we know what we'll do ... Which is why we have spouses and children and parents and colleagues and friends, because someone has to know us better than we know ourselves. 1120
There is some evidence that supports this startling conclusion. First, the correspondence between people's ratings of their own personality and other people's ratings of their personality is not very high. It depends somewhat on the trait; for example, people tend to agree with others about how extraverted they are, but on most other personality traits the level of agreement is modest (correlations in the range of .40). Thus, Suzie's judgment of how agreeable and conscientious she is correlates only modestly with how agreeable and conscientious her friends think she is.
Furthermore, other people agree more among themselves about what another person is like than they agree with that person's own ratings. Jane, Bob, Sam, and Denisha are likely to agree more with each other about how agreeable and conscientious Suzie is than they are to agree with Suzie.
But who is more "right"? Does Suzie know best how agreeable she is, or do her friends know her better than she knows herself? To try to answer this question, some researchers have looked at who can better predict what a person actually does: the person's ratings of his or her own personality or other people's ratings of his or her personality. If we wanted to predict how nervous Suzie will be when she meets someone new, for example, would we be better off going by her own report of how extraverted and agreeable she is or by her friends' reports? There is some evidence that peer reports (Suzie's friends' ratings) predict people's behavior better than their self-reports (Suzie's own ratings). In one study, for example, college students were worse at predicting how nervous and talkative they would he when chatting with a new acquaintance than were peers who had just met them for the first time.-''
Other studies have found that people are worse at making specific predictions about how they will behave than they are at predicting how other people will behave. When asked whether they would purchase a flower as part of a campus charity drive in the upcoming weeks, students made overly rosy predictions; 83 percent said they would, whereas in fact only 43 percent actually did. When asked how likely it was that other students would purchase a flower, people were more accurate; they predicted that 56 percent would, which was closer to the 43 percent figure. In another study, people predicted that they would donate an average of $2.44 of their earnings in an experiment to charity, whereas other people would donate only $1.83. Once again they were more accurate in their predictions about other people; the actual figure donated was $1.53.
One reason people fail to predict their own behavior very accurately is that they believe that they are "holier than thou" and would be more likely than the average person to perform moral acts of kindness. Another is that people use different kinds of information when predicting their own versus other people's behavior. When predicting other people's actions, we rely mostly on our cumulative experience of how the average person would act, including our hunches about the kinds of
situational constraints people will face ("Probably many people who intended to buy a flower will never walk past one of the people selling them"). When predicting our own actions, we rely more on our "inside information" about our own personalities ("I am a kind person who wants to help others"). This can be a problem for two reasons: relying only on inside information causes people to overlook situational constraints on their actions, such as the possibility that they, too, will fail to pass by someone selling the flowers; second, as we have seen, people's inside information is not the full story about their personalities and might not be completely accurate.22
When it comes to asking who makes better judgments about our personality, however-we ourselves or other people-it might not make much sense to ask who is more accurate. Suzie and her friends might have different views of her personality, but both may be "right" in some sense. Her friends might be keying in more on her adaptive unconscious, as revealed in her behavior-particularly behaviors she is not monitoring and controlling consciously, such as how much Suzie fidgets and plays with her hair when she meets a new acquaintance. Suzie, on the other hand, might be basing her estimate on her general theory about how nervous she is in novel social settings.
Suzie's friends might be more accurate at predicting her future behaviors that are spontaneous and unmonitored, such as how nervous she will appear to be on a first date. Suzie's self-view, however, might be more accurate at predicting her more controlled, deliberate actions, such as whether she decides to accept a blind date. Suzie has a constructed self that may be at odds with her adaptive unconscious but still predicts behaviors that she consciously monitors and controls.
The Constructed Self
What is the nature of the conscious self that exists independently of the personality of the adaptive unconscious? There has been a great deal of research on the self-concept, including how it helps people organize information about themselves, interprets ambiguous information, and guides behavior. This research also examines different functions of the self, its affective implications, and how the self-concept differs across cultures.23
Self theorists have been reticent, however, about discussing the extent to which the self-concept is conscious or nonconscious. It is important to focus on this question, I believe, in order to clarify a number of con fusing findings (such as those reviewed already, in which implicit and explicit measures of personality predict different kinds of behavior). We need to distinguish between those aspects of the self-concept that reside in the adaptive unconscious and those that consist of conscious beliefs about the self."
Dan McAdams has studied an important part of the conscious selfconcept, namely the life stories that people construct about themselves, which he describes as a continuing narrative that people tell about their past, present, and future. The major function of these stories, McAdams argues, is to integrate the many aspects of oneself into a coherent identity that is stable over time but also subject to revision. McAdams' work suggests that an important role of this deliberative system is to link together the many disparate parts of the self into a coherent story.
McAdams argues that life stories do not (and need not) correspond perfectly with external reality. They are people's construals of their lives rather than the fact-based reporting of an objective historian. However, a life story should not be a complete fabrication; people whose life stories bear no relation to their actual lives often end up in mental hospitals. One of McAdams' criteria for what makes a good story is that it beat least somewhat reality-based.2-'
Although life stories constitute a compelling approach to personality, others have questioned the extent to which such stories are important determinants of people's behavior, versus epiphenomenal, after-the-fact accounts of one's actions. The personality psychologist Robert McCrae phrased this question well: "I do not yet know quite what to make of them. Are life stories the unifying themes that guide our life, as the jet stream guides weather systems, or are they mere epiphenomena, moreor-less adequate rationalizations and secondary elaborations that convey the gist of our life history in a form suitable for the occasion?"-'"
McCrae's question gets to the heart of issues we discussed in Chapter 3 concerning the role of consciousness, such as whether it is similar to the child at a video arcade who turns the steering wheel on a racing-car game without putting any money into it, unaware that she is viewing a demonstration program, which is not at all influenced by her conscious intentions and goals-an agent that thinks she is in control of the action but really isn't.
But surely this position is too extreme. Consistent with the consciousness-as-Ronald Reagan analogy, it is clear that people's conscious beliefs about their traits and motives play a causal role (albeit not as much as they might think). The conscious self system is not completely epiphenomenal; as we have seen, explicit beliefs about attachment and motivation influence some important social behaviors.
A number of theorists, for example, have pointed to the importance of people's conscious constructions of the kind of person they ought to be or might become. In psychoanalytic theory, children are said to develop ego ideals as part of the superego, based on their conceptions of their parents' moral stance, and these ego ideals play an important role in the kinds of decisions people make when faced with moral dilemmas and the kinds of emotions they experience. Social psychologists have also discussed the importance of people's constructions of alternative selves. People have mental constructs of the kind of person they would like to become (e.g., a successful lawyer), the kind of person they feel they ought to become (e.g., a parent), and the kind of person they are afraid of becoming (e.g., a bag lady). Possible selves are conscious embodiments of our hopes and fears about ourselves, and these constructions shape our behavior, at least to some extent.27
The bottom line is that when people describe their own personalities, they are often reporting their conscious theories and constructions that may or may not correspond to the dispositions and motives of their adaptive unconscious.
Origins of the Nonconscious and Conscious Personalities
If people have two "selves"-a nonconscious and a conscious one that are only loosely related-where do these systems come from? There is evidence that some of the dispositions of the adaptive unconscious, such as temperament, have a genetic basis. It is also clear that culture and experience play a role. A hallmark of the adaptive unconscious is auto maticity, whereby information is processed in rapid, nonconscious, involuntary ways. One way a construct can become automatic is through lots of repetition. People are not born with the kinds of "if-then" patterns of construal discussed by Mischel, or the chronically accessible constructs discussed by social psychologists. These constructs, rooted in childhood experiences, become automatic through frequent use.
What kinds of experiences? David McClelland and his colleagues offer the hypothesis that nonconscious motives are rooted in early infancy, whereas conscious, self-attributed motives result from more explicit, parental teachings. To test this idea, McClelland and his colleagues interviewed a sample of adults in their early thirties, measuring both their nonconscious motives (i.e., their responses to TAT pictures) and their conscious, explicit motives (their responses on a self-report questionnaire). The fascinating thing about this study is that the participants' mothers had been interviewed twenty-five years earlier about their childrearing practices, allowing the researchers to test the extent to which people's implicit and explicit motives, as adults, were related to the childrearing practices of their mothers twenty-five years earlier.
There was some evidence that early, prelingual childrearing experiences were correlated with implicit but not explicit motives. For example, the extent to which mothers used scheduled feedings correlated with the implicit but not explicit need for achievement in the adult sample, and the extent to which the mothers were unresponsive to their infants' crying was correlated with the implicit but not explicit need for affiliation. Postlingual childhood experiences were more likely to corre
late with explicit than with implicit motives. For example, the extent to which children were taught not to fight back when provoked was correlated with the explicit but not implicit need for affiliation, and the children of parents who set explicit tasks for them to learn were more likely to have an explicit but not implicit need for achievement."
The nonconscious and conscious selves thus seem to be influenced by one's cultural and social environment, but in different ways. The kinds of early affective experiences that shape a child's adaptive unconscious surely have a cultural basis, given that childrearing practices differ markedly from culture to culture. The conscious theories people develop about themselves also are shaped by the cultural and social environment.
Implications for Self-Insight
To understand better our own nonconscious personality dispositions, we cannot simply remove the veil obscuring our view, for there is no direct view. Instead, we are forced to make educated guesses about our nonconscious dispositions.
But why don't people realize, eventually, that their conscious conceptions are at odds with their nonconscious personalities? Doesn't it seem that over time, people would discover that they are not the person they thought themselves to be? Why didn't Henry Higgins eventually realize that he was not the refined, kindhearted gentleman who abhorred profanity? How can people be so out of touch?
One reason is that people are motivated to have an overly positive view of themselves, and avoid looking too closely at their warts and flaws. There is a good deal of evidence that people see themselves through rose-colored glasses and that, within limits, it is healthy to do so. What is the harm in thinking we are a little more popular and extraverted and kind than we really are?
Another reason is that once people develop a conscious theory about themselves-shaped, perhaps, by explicit parental teaching-it can be difficult to disconfirm. We may be more likely to notice the times we act in accord with our conscious theories than the times we do not. Even if inconsistencies are brought to our attention, we can easily dismiss them as exceptions. When Mrs. Pearce points out to Henry Higgins that just that morning he has uttered swear words "to your boots, to the butter, and to the brown bread," Higgins replies, "Oh that! Mere alliteration, Mrs. Pearce, natural to a poet."29 A crude person who swears is simply not part of his self-narrative, and thus he easily dismisses any evidence to the contrary.
Strangers to Ourselves Page 10