Know then thyself, presume not God to scan;

The proper study of mankind is Man.

—Alexander Pope (English poet, 1688–1744), An Essay on Man

Although our cultural and social environments shape the types of knowledge we have about the self, the self-concept is also influenced by our actual personality and physical attributes. Studies of identical twins reared apart indicate that some of these attributes are likely to be influenced by a person’s genetic makeup (e.g., Plomin et al., 1990). Some people like to be the life of the party; other people like to blend into the background. Some people are gifted athletes; others are gifted in math.

How do people discover which attributes define who they are and what sets them apart from others? Next we consider three ways that people learn about themselves over the course of their everyday social interactions: the appraisals they get from others, their social comparisons, and their self-perceptions. We’ll see that, like any means of acquiring knowledge, these processes are imperfect. Our social cognitions can lead us down a path of mistakes. That means that although it may appear that we know ourselves very well, we nevertheless make errors in how we think about and experience our traits, qualities, and emotions.

Two 20th-century sociologists, Charles Cooley (1902) and George Herbert Mead (1934), examined the social origin of the self-concept from a perspective known as symbolic interactionism. The central idea is that people use their understanding of how significant people in their lives view them as the primary basis for knowing and evaluating themselves. Cooley coined the term looking glass self to refer to the idea that significant people in a person’s life reflect back to her (much like a looking glass, or mirror) who she is by how they behave toward her. People come to know themselves first by observing how others view them, or others’ appraisals of them, and then incorporating those appraisals into their self-concept.

People use others’ appraisals not only to know their attributes but also to evaluate themselves and their actions as good or bad. For example, a person might feel bad about the pile of laundry on the floor because she imagines Mom’s voice saying, “You are such a slob!” Because people internalize others’ appraisals, they evaluate themselves as if those other people were in their heads, observing them act. Cooley also pointed out that a person’s self-concept is more likely to develop and change in response to the appraisals of people who are close or admired than in response to the appraisals of strangers.

Research supports many of Mead’s and Cooley’s insights. In fact, Baldwin and colleagues (1990) hypothesized that even unconscious reminders of approval and disapproval from significant others would influence self-evaluations. In one study, social psychology graduate students evaluated their own research ideas after being subliminally primed with a picture of a highly respected faculty member scowling with disapproval or, in the control condition, a picture of a less prominent figure with an approving expression. In a second study, the researchers first had Catholic participants read a description of a sexual dream, then subliminally presented them with the scowling face of either the pope or an unfamiliar other, also with a disapproving expression, and finally had them rate themselves on dimensions such as morality, intelligence, and talent. Both groups of participants rated themselves less positively (graduate students saw their ideas as less important and original; Catholics viewed themselves as less competent and worthy) after being exposed to the disapproving face of a significant authority figure, but not following exposure to the disapproving face of an unfamiliar person, even though these faces were presented to participants below the level of their conscious awareness. Even at an unconscious level, people carry with them the knowledge of how significant others view them, and they use those appraisals to judge themselves.

159

Mead went beyond Cooley’s emphasis on appraisals of particular significant others to propose that people also internalize an image of a generalized other, a mental representation of how people, on average, appraise the self. This generalized other becomes the internal audience or perspective by which people view and judge themselves. A study consistent with this idea showed that overweight people are significantly less happy if they live in a society that stigmatizes obesity and values thinness than if they live in a society in which obesity is common and accepted (Pinhey et al., 1997). These findings suggest that people can think poorly about themselves because they take the perspective of a generalized other who views them negatively.

Although this research suggests that the self-concept is highly influenced by others’ direct or indirect feedback, other studies show that people’s views of themselves are sometimes very different from the views that others hold of them. Does this mean that Cooley’s and Mead’s analyses are wrong? Not really. Remember that Cooley said that people first assess how others view them, and then they change their self-concept to bring it in line with those internalized views. But we shouldn’t assume that people accurately judge what others think about them—in fact, they often misread the appraisals others are reflecting back onto them (Carlson et al., 2011). What matters, though, is that people base their self-views on how they think others view them. That is, reflected appraisals—what we think other people think about us—do play a significant role in shaping our self-concept, as Cooley and Mead proposed, but these reflected appraisals can be very different from the actual appraisals that people have of us (Ichiyama, 1993; Shrauger & Schoeneman, 1979).

This point is readily apparent when we consider people’s perceptions of their physical attractiveness. If you ask Person A how physically attractive she or he is, and then you ask Person A’s romantic partners and close friends to rate Person A’s attractiveness, you will find that those ratings do not match up entirely. Indeed, the average correlation between the ratings is only about .24 (Feingold, 1988). That means that some people overrate their attractiveness; other people underrate it. In both cases, people’s perceptions of themselves are derived from factors other than the appraisals they receive from others.

The gap between reflected appraisals and actual appraisals is partly due to distortions in the feedback that senders provide to us: People often try to be tactful, softening their feedback to others (e.g., DePaulo & Kashy, 1998). But people also selectively interpret the feedback they are given. In one study, O’Connor and Dyce (1993) interviewed members of bar bands, asking band members to rate the ability of others in the band (actual appraisals of band mates), what feedback they gave to others in the band about their abilities, and how those individuals perceived their own ability. The researchers found that band members were pretty honest with their band mates, just a little more positive in their feedback then their private evaluations of them would suggest. However, each individual seemed to think his band mates appreciated his musical ability less than they actually did and less than the feedback he got from them would suggest. So each band member seemed to let his own insecurities color how he was perceived by others, creating a gap between reflected appraisals and actual appraisals.

160

A second way in which people learn who they are is by comparing themselves with others. Leon Festinger (1954) first described this process in his social comparison theory. He pointed out that people often don’t have an objective way of knowing where they stand on an attribute. Therefore, many, if not most, of their beliefs about themselves are on dimensions that can be assessed only relative to others. Consider, for example, whether you are a fast runner. Well, compared with a four-year-old, you probably are. Festinger called this comparison of the self with those who are worse off a downward comparison. But how does your running speed compare with that of an Olympic runner? Not so well. People engage in upward comparison when they compare themselves with those who are better off. Festinger pointed out that we generally compare ourselves with people to whom we feel similar because those people provide the most informative indication of our traits, skills, and abilities. So if you think you’re fast, you’re probably making that judgment relative to other people similar in age who also play the same sports you do, rather than to four-year-olds or to Olympic runners. What’s more, people are more likely to compare themselves with individuals in their local environment than with large groups, national averages, or other entities that lie outside their familiar day-to-day experience (Zell & Alicke, 2010). Festinger further suggested that people are particularly likely to make these comparisons when they lack objective indicators of how they’re doing and are uncertain of where they stand.

Consider, for example, a job interview situation where one is likely to be uncertain about one’s standing on certain characteristics. One study (Morse & Gergen, 1970) used this context to demonstrate how social comparisons can influence people’s views of themselves. College-student participants came to the lab in response to an advertisement for a data-entry position. On arriving, each student and another person were asked to wait in a room and complete some initial personnel questionnaires. In one condition, the other candidate (who was actually a confederate of the research team) was dressed in a suit and carried a briefcase, giving off an air of competence and dependability. In the other condition, the confederate was dressed in wrinkled clothing, his hair was unkempt, and he carried an unorganized stack of papers. Among the questionnaires participants completed was a measure of self-esteem. According to social comparison theory, downward comparison with the person who seems like a worse candidate for the job than yourself would make you puff up with pride, whereas upward comparison with the more impressive candidate would be more likely to leave you feeling down in the dumps. This is exactly what the researchers found: Participants rated themselves more positively when they sat in the room with Mr. Sloppy compared to when they sat in the room with Mr. Neat.

Just as reflected appraisals sometimes are a poor match to what people really think of themselves, the self-knowledge people gain through social comparison is not always accurate. In fact, people consistently make errors in their use of social comparisons to judge their own attributes. These errors can come from over- or underestimating one’s own attributes or from over- or underestimating the attributes of those with whom one compares oneself.

One particularly well-documented social comparison error is the better than average effect, people’s tendency to rank themselves higher than most people on positive attributes (Alicke, 1985). In two of the many demonstrations of this effect, 42% of engineers thought their work ranked in the top 5% of their peers (Zenger, 1992), and 94% of college professors thought they did above-average work (Cross, 1977). Even people in prison think they are kinder and more moral than the average person (Sedikides et al., 2014). Of course, if you think about it, many of these individuals can’t be accurate. It’s statistically impossible for most people to be above average—instead, the average response should be average! As we will discuss in detail in the next chapter, this and other biases in self-perception are caused in large part by the need to maintain a positive sense of self-worth.

161

But why else do people make this common error? To find out, Dave Dunning and colleagues (2003) reviewed evidence from various studies that examined which people are most likely to overestimate their abilities and knowledge in a variety of domains, including performance in psychology, reading comprehension, grammar, and logic tests, sense of humor, debating ability, hunters’ knowledge of firearms, medical residents’ interviewing skills, and medical lab technicians’ problem-solving ability. In an ironic twist, the researchers found in each case that many of the same people who rated themselves as above average were the worst performers. Even when offered money for being accurate, poor performers did not become more accurate in making these judgments, suggesting that they couldn’t assess themselves accurately even when highly motivated to do so (Ehrlinger et al., 2008).

Dunning and colleagues offer a simple but interesting explanation for this phenomenon. If you ask people how good they are as writers, most people would rate themselves as better than average. But why would the worst writers do this? Because they lack the knowledge of writing—of grammar, composition, and so on—to realize how bad their writing is! If they had this knowledge, they would probably be better writers! Dunning and colleagues (2003, p. 83) characterize this as a double curse: “The skills needed to produce correct responses are virtually identical to those needed to evaluate the accuracy of one’s responses.”

This finding raises an important question: Can people be trained to assess their weaknesses accurately, so that they know what they need to do to improve? To find out, Kruger and Dunning (1999) had participants take a test of logic. Poor performers greatly overestimated their performance. Then they trained some of these poor performers in how to distinguish correct from incorrect answers and gave them their tests to look over. Their self-ratings now became more accurate. In another ironic twist, they now rated their own logical reasoning ability lower than they did before being trained, even though the training probably strengthened that ability!

Thus, ignorance of ignorance is one source of inaccuracy in evaluating the self in comparison to others. As people get smarter and become aware of their own ignorance, they tend to become more accurate about themselves. As Dunning and colleagues (2003) note, although knowledge of your deficiencies can be humbling, it is often better than remaining blissfully unaware, because you won’t be motivated to take steps to improve until you realize you have deficiencies in skills or knowledge.

According to self-perception theory (Bem, 1965), we often discover who we are in the same way that we form impressions of other people. In chapter 4, we talked about how you might form an impression of your new college roommate by observing her behavior. When her behavior cannot be explained by factors in the situation, you infer that it was caused by some internal disposition she possesses. In the same way, we sometimes form impressions of ourselves by observing our own behavior and making attributions for what we do. If behavior can be explained by factors in the situation, we attribute our behavior to those external factors. However, if there is no salient external factor to account for our behavior, we attribute our behavior to an internal attitude or trait.

162

We are most likely to learn about ourselves through this self-perception process when we find ourselves in new or unusual situations. Imagine that a friend invites you to go spelunking with him. You’ve never been spelunking, but you like the idea of exploring caves and trying something new. You are 30 feet into the cave and navigating a tight corridor when your heart starts racing, your palms start sweating, and you start backing out of the cave. Your friend is perfectly calm, and you had only a single cup of coffee this morning, so what could possibly be causing your behavior? You realize for the first time in your life that you are claustrophobic. This is the sort of self-knowledge that you gain only through self-perception, that is, by finding yourself in a new situation where the best explanation for your behavior is something about who you are and the traits (or in this case, phobias) you possess.

In a less dramatic way, we often find ourselves relying more on self-perception processes when we come to transition points in our life. A freshman who just showed up on campus doesn’t really know what kind of college student she is yet. If asked whether she is a punctual student, she can remember back to her most recent class and think, “Yeah, I did show up on time to chemistry and was actually 20 minutes early to my Spanish class, so maybe I am punctual.” Or she might think about whether she is punctual in other domains, such as her job, and generalize from that. In either case, she can make this judgment about whether she is a punctual college student, but not without first engaging in a self-perception process of consulting recent or more tangentially related examples of her own behavior. By her senior year, enough of this experience should have built up so that now when asked if she is a punctual college student, she gives an immediate “yes,” without having to think about recent times when she demonstrated that trait (Klein et al., 1996).

We don’t use self-perception processes just to determine our personality traits and abilities; we also use them to determine how we are currently feeling. Imagine going for a routine visit to your doctor, who asks, “So how are you feeling today?” To answer that question, you probably rely on the evidence coming from your internal physiological and psychological states. If your head is throbbing, you will likely respond that you aren’t feeling so great. In this situation, your physical states clearly inform who you are (i.e., a person with a headache) in that moment. But our physical states also inform our emotions and attitudes in ways that are less obvious.

Using One’s Feelings to Know the Self

Lift your chin up, and then down, and repeat this motor movement. This is a pretty basic physical act. Taking it on its own, we might not imagine that it would have any power to influence our judgment. But think about when we usually engage the muscles in our head and neck in this way. Often it is when we are signaling our agreement with something. Could this mean that our brain unconsciously uses this same sequence of muscular movements to infer agreement? Research suggests that the answer is yes.

For instance, Wells and Petty (1980) had participants listen to an audio recording that included an editorial advocating tuition increases. Under the guise of testing the durability of the headphones, participants were asked to move their chins up and down or from side to side while listening to the tape. Afterward, participants were asked how much they thought tuition should be. Those who had been nodding their heads the entire time reported tuition numbers that were about 38% higher than those who had been shaking their heads!

Another example comes from work on the facial feedback hypothesis. We become so accustomed to expressing our emotional states through our facial expressions that changes in our facial movements become a signal of the emotion we might be feeling. In the first test of this hypothesis, James Laird (1974) attached electrodes to participants’ faces and asked them to evaluate a series of cartoons. Before showing the participants each cartoon, he gave them instructions to contract their facial muscles or squeeze their eyebrows together in certain ways, such as, “Use your cheek muscles to pull the corners of your lips outward.” Participants were told that the electrodes were measuring the activity of their facial muscles, but in reality Laird was subtly inducing them to make either a smiling expression or a frown. Those induced to make the smiling face rated the cartoons as funnier and reported feeling happier than those induced to frown. You can try this out at home without the fancy electrodes. Try putting a pen or pencil between your teeth as shown in the photo on the left. This activates your zygomaticus major muscles, forcing your lips to draw back as they do when you smile. Now hold the pen or pencil with your lips as shown in the photo below. This activates your corrugator supercilii muscles, which you use when you are frowning. In one study using this technique, participants who (unknowingly) “smiled” rated cartoons as more humorous than did participants in the control condition, whereas participants who held a marker with their lips (the “frowners”) found them less amusing (Strack et al., 1988). Participants seemed unconsciously to infer, “If I am smiling (or frowning), I must be amused (or turned off).”

163

The basic idea of self-perception is that we often lack a strong internal feeling about who we are or how we feel, so we look to our own behavior to make inferences about what we are like. Because we don’t have conscious access to the processes underlying our preferences and actions, we often lack insight into why we do the things we do. As a consequence, our inferences about ourselves often can be wrong.

One of the reasons we can be wrong is that we aren’t very good at accurately judging how situations influence our thoughts and behaviors. We underestimate the effects of some situational factors and overestimate the effects of others. For example, Nisbett and Wilson (1977b) showed that people liked a movie less if it was out of focus part of the time but were unaffected by a loud noise outside the room where the movie was showing. However, when the researchers asked participants about the factors that influenced their enjoyment, the participants thought the focus problems did not affect their liking (underestimating the effect) and that the loud noise did (overestimating the effect).

In much the same way as our memory is not a verbatim record of the past (see chapter 4) but is instead reconstructed online, our conscious experience of ourselves is constructed online. As a result it is susceptible to error and bias. This doesn’t mean our beliefs about why we feel and behave the way we do are always wrong, but rather that they are often, if not always, based on an imperfect inference process that sometimes leads to inaccurate or incomplete understanding. If you just found out a close friend died, and you felt very sad, you would probably be right in inferring that this news made you sad. But so would an objective observer who witnessed you learning of this news but did not have access to your internal feelings. At other times, the cause of negative feelings may not be so obvious. You might erroneously attribute them to particular reasons, such as lack of sleep, when they are really due to something else. In fact, studies comparing daily fluctuations in mood with other things going on in people’s day-to-day lives show that people are not very accurate in their beliefs about the factors that affect their moods (e.g., Stone et al., 1985). For example, one study found that college students thought their moods were affected by amount of sleep the prior night and the weather, but they actually were affected by neither (Wilson et al., 1982).

164

Using the Self to Know One’s Feelings

Self-perception processes can also play an important role in the emotions we feel. Stan Schachter’s (1964) two-factor theory of emotion proposed that people’s level of arousal determines the intensity of the emotion, but the specific type of emotion they experience is determined by the meaning that is assigned to that arousal based on contextual or environmental cues. Thus, from this perspective:

emotion = arousal × cognitive label

One startling implication of this theory is that the same arousal could be attributed to one or another emotion, depending on the self-perception process of interpreting cues present in one’s environment. FIGURE 5.2 shows an example of how the two-step process can elicit different emotions.

In the first experiment to test this idea, Schachter and Singer (1962) gave participants an injection of epinephrine (also known as adrenaline), which causes arousal in the sympathetic nervous system. However, they told participants that the study concerned the effects of a drug that influences memory and that the injection was a dose of the memory-enhancing drug. In the critical conditions of the study, participants were told that the injection would have no side effects. They were then asked to wait for the drug to take effect in a room with a confederate who was either happily shooting balls of paper into a trash can or voicing his anger over what he saw as intrusive questions on a survey he was filling out. Why did Schachter and Singer put participants in the room with different confederates? They wanted to determine if they could alter the participants’ emotions by varying the salience of a label for that arousal. Thus, participants were later asked how they felt during the study. Those who witnessed the happy confederate reported being happy, whereas those who spent time with the angry confederate reported being angry themselves. Other participants, who also were given the injection but were told to expect symptoms of physiological arousal as a side effect, were much less likely to experience these emotions. This is because they already had a cognitive label for their arousal: it was a side effect of the injection. Only participants without a ready cognitive label for the arousal attributed that arousal to an emotion. Equally important, the emotion to which they attributed their arousal was vastly different—either happiness or anger—depending on the cues provided by the confederate.

165

This phenomenon has become known as misattribution of arousal, which occurs when we ascribe arousal resulting from one source (in the case of the first study, an injection) to a different source, and therefore experiencing emotions that we wouldn’t normally feel in response to a stimulus. Although certain emotions do induce very specific and differentiated physiological responses (e.g., Barrett et al., 2007; Reisenzein, 1983), initial arousal states often are subjectively ambiguous; thus, people’s emotions can be greatly influenced by their interpretation of the circumstances of the arousal. This research shows how self-perception processes influence not only how we think about the self but also how the self experiences emotions.

Understanding how people use self-perception to label their emotions can be rather useful. Consider a psychological problem such as insomnia. Storms and Nisbett (1970) asked insomniacs to take a placebo pill described as causing arousal right before they went to bed: a seemingly paradoxical form of treatment for insomnia! The researchers reasoned that insomniacs have trouble falling asleep in part because they experience anxiety about their lives and, specifically, about being able to fall asleep. But if they could be led to misattribute some of their arousal to a pill, they would feel less anxious and consequently would be able to fall asleep faster. This is precisely what happened. Studies have similarly used placebo pills to reduce self-perception of a variety of emotions, including fear, anger, sexual attraction, and guilt, and to assess the role of such emotional reactions in various kinds of behavior (e.g., Dienstbier et al., 1980; Nisbett & Schachter, 1966).

Of course, outside the confines of social psychology labs, rarely are people unwittingly given injections of adrenaline or placebo pills. This led Dolf Zillmann (e.g., Zillmann et al., 1972) to wonder: Under what conditions does misattribution of arousal occur in everyday life? According to his excitation transfer theory, misattribution happens when an individual is physiologically aroused by an initial stimulus and then a short time later encounters a second, potentially emotionally provocative, stimulus. Leftover or residual excitation caused by the first event becomes misattributed or, using Zillmann’s terminology, transferred, to the reaction to the second stimulus, resulting in an intensified emotional response to that second stimulus.

In one study demonstrating this phenomenon (Zillmann et al., 1972), half the participants were asked to exercise by riding a stationary bike vigorously for two and a half minutes. The other participants were seated comfortably at a table and asked to pass thread through discs with holes (not particularly arousing!). In the second part of the study, participants were provoked with insults or not provoked by another participant (in actuality, a confederate). Participants were then given an opportunity to punish the other participant by delivering painful electric shocks. As you see in FIGURE 5.3, the unprovoked participants were not very aggressive, regardless of whether they exercised. Provoked participants, however, were more aggressive if they exercised than if they did not, suggesting that they misattributed arousal as anger in response to the provocation, thus producing anger-motivated aggression. Other studies show that residual arousal, whether from exercise or viewing sexually arousing media, can intensify various other emotion-based responses later on—such as prosocial behavior (Mueller & Donnerstein, 1981), the enjoyment of music (Cantor & Zillmann, 1973), and laughter (Cantor et al., 1974). It is critical to note that these effects occur primarily when participants are not aware of the leftover arousal from the first event. If participants are put in another emotion-provoking context immediately after being aroused, they don’t misattribute their arousal because they can easily connect that arousal to the prior inducing source (“My heart is racing because I was just on the exercise bike!”; Zillmann, 1978). Also, if people have their attention focused inward toward the self, they become more aware of the residual arousal and don’t misattribute their arousal or show intensified emotional reactions to the later situation (Reisenzein & Gattinger, 1982).

166

These findings show how the misattribution of arousal, leading to intensified emotional reactions, can happen in the course of daily events. Here’s just one of many examples of how this process can affect people: Imagine a typical teenage boy getting off an intense roller-coaster ride with his date. If a few minutes later he gazes into his date’s eyes, he is likely to experience stronger attraction toward her because of residual arousal from the roller-coaster ride (Meston & Frohlich, 2003).

167