Michael Bublé is sexy and talented. Whether or not you agree with that sentence, it almost certainly activated your medial prefrontal cortex, which is an area of your brain that is activated when you think about the attributes of other people but not about the attributes of inanimate objects such as houses or tools (Mitchell, Heatherton, & Macrae, 2002). Although most of your brain shows diminished activity when you are at rest, this area remains active all the time (Buckner, Andrews-Hanna, & Schacter, 2008). Why does your brain have a specific area that is dedicated to processing information about just one of the millions of objects you might encounter, and why is this area constantly switched on?

Of the millions of objects you might encounter, other human beings are the single most important. Social cognition is the processes by which people come to understand others, and your brain is doing it all day long. Whether you know it or not, your brain is constantly making inferences about others people’s thoughts and feelings, beliefs and desires, abilities and aspirations, intentions, needs, and characters. It bases these inferences on two kinds of information: the categories to which people belong, and the things they do and say.

538

You will recall from the Language and Thought chapter that categorization is the process by which people identify a stimulus as a member of a class of related stimuli. Once we have identified a novel stimulus as a member of a category (“That’s a textbook”), we can then use our knowledge of the category to make educated guesses about the properties of the novel stimulus (“It is probably expensive”) and act accordingly (“I think I will borrow it from the library”).

What we do with textbooks we also do with people. No, not the borrowing part. The educated guessing part. Stereotyping is the process by which people draw inferences about others based on their knowledge of the categories to which others belong. The moment we categorize a person as an adult, a male, a baseball player, and a Russian, we can use our knowledge of those categories to make some educated guesses about him, for example, that he shaves his face but not his legs, that he understands the infield fly rule, and that he knows more about Vladimir Putin than we do. When we offer children candy instead of cigarettes or ask gas station attendants for directions instead of financial advice, we are making inferences about people whom we have never met before based solely on their category membership. As these examples suggest, stereotyping is a very helpful process (Allport, 1954). And yet, ever since the word was coined by the journalist Walter Lippmann in 1936, it has had a distasteful connotation. Why? Because stereotyping is a helpful process that can often produce harmful results, and it does so because stereotypes tend to have four properties: They are inaccurate, overused, self-perpetuating, and automatic.

539

The inferences we draw about individuals are only as accurate as our stereotypes about the categories to which they belong. Although there was no evidence to indicate that Jews were especially materialistic or that African Americans were especially lazy, American university students held such beliefs for most of the last century (Gilbert, 1951; Karlins, Coffman, & Walters, 1969; Katz & Braly, 1933). They were not born holding these beliefs, so how did they acquire them? There are only two ways to acquire a belief about anything: to see for yourself or to take somebody else’s word for it. In fact, most of what we know about the members of human categories is hearsay—stuff we picked up from hearing other people talk. Many of the people who believe stereotypes about Jews or African Americans have never actually met a member of either group, and their beliefs are a result of listening too closely to what others told them. In the process of inheriting the wisdom of our culture, it is inevitable that we also will inherit its ignorance too.

But even direct observation can produce inaccurate stereotypes. For example, research participants in one study were shown a long series of positive and negative behaviours and were told that each behaviour had been performed by a member of one of two groups: Group A or Group B (see FIGURE 13.15). The behaviours were carefully arranged so that each group behaved negatively exactly one third of the time. However, there were more positive than negative behaviours in the series, and there were more members of Group A than of Group B. As such, negative behaviours were rarer than positive behaviours, and Group B members were rarer than Group A members. After seeing the behaviours, participants correctly reported that Group A had behaved negatively one third of the time. However, they incorrectly reported that Group B had behaved negatively more than half the time (Hamilton & Gifford, 1976).

Why did this happen? Bad behaviour was rare and being a member of Group B was rare. Thus, participants were especially likely to notice when the two co-occurred (“Aha! There’s one of those unusual Group B people doing an unusually awful thing again”). These findings help explain why members of majority groups tend to overestimate the number of crimes (which are relatively rare events) committed by members of minority groups (who are relatively rare people, hence the word m-i-n-o-r-i-t-y). The point here is that even when we directly observe people, we can end up with inaccurate beliefs about the groups to which they belong.

Because all thumbtacks are pretty much alike, our stereotypes about thumbtacks (small, cheap, painful when chewed) are quite useful. We will rarely be mistaken if we generalize from one thumbtack to another. But human categories are so variable that our stereotypes may offer only the vaguest of clues about the individuals who populate those categories. You probably believe that men have greater upper body strength than women do, and this belief is right on average. But the upper body strength of individuals within each of these categories is so varied that you cannot easily predict how much weight a particular person can lift simply by knowing that person’s gender. The inherent variability of human categories makes stereotypes much less useful than they seem.

540

Alas, we do not always recognize this because the mere act of categorizing a stimulus tends to warp our perceptions of that category’s variability. For instance, participants in some studies were shown a series of lines of different lengths (see FIGURE 13.16) (McGarty & Turner, 1992; Tajfel & Wilkes, 1963). For one group of participants, the longest lines were labelled GroupA and the shortest lines were labelled Group B, as they are on the right side of Figure 13.16. For the second group of participants, the lines were shown without these category labels, as they are on the left side of Figure 13.16. Interestingly, those participants who saw the category labels overestimated the similarity of the lines that shared a label and underestimated the similarity of lines that did not.

You have probably experienced this phenomenon yourself. For instance, we all identify colours as members of categories such as blue or green, and this leads us to overestimate the similarity of colours that share a category label and to underestimate the similarity of colours that do not. That is why we see discrete bands of colour when we look at rainbows, which are actually a smooth chromatic continuum (see FIGURE 13.17). When two cities are in the same country (London and Ottawa) people tend to underestimate their distance, but when two cities are in different countries (London and Cincinnati) they tend to overestimate their distance (Burris & Branscombe, 2005). Indeed, people believe that they are more likely to feel an earthquake that happens 370 km away when the earthquake happens in their province rather than a neighbouring province (Mishra & Mishra, 2010).

What is true of colours and cities is true of people as well. The mere act of categorizing people as blacks or whites, Jews or Gentiles, artists or accountants, can cause us to underestimate the variability within those categories (“All artists are wacky”) and to overestimate the variability between them (“Artists are much wackier than accountants”). When we underestimate the variability of a human category, we overestimate how useful our stereotypes can be (Park & Hastie, 1987; Rubin, & Badea, 2012).

541

When we meet a truck driver who likes ballet more than football or a senior citizen who likes Drake more than Bach, why do not we simply abandon our stereotypes of these groups? The answer is that stereotypes tend to be self-perpetuating. Like viruses and parasites, once they take up residence inside us, they resist even our most concerted efforts to eradicate them. Here are three reasons why:

542

If we recognize that stereotypes are inaccurate and self-perpetuating (as you now do), then why do not we just make a firm resolution to stop using them? The answer is that stereotyping happens unconsciously (which means that we do not always know we are doing it) and automatically (which means that we often cannot avoid doing it even when we try) (Banaji & Heiphetz, 2010; Greenwald, McGhee, & Schwartz, 1998; Greenwald & Nosek, 2001).

For example, in one study, participants played a video game in which photographs of black or white men holding either guns or cameras were flashed on the screen for less than 1 second each. Participants earned money by shooting men with guns and lost money by shooting men with cameras. The results showed that participants made two kinds of mistakes: They tended to shoot black men holding cameras and tended not to shoot white men holding guns (Correll et al., 2002). Although the photographs appeared on the screen so quickly that participants did not have enough time to consciously consider their stereotypes, those stereotypes worked unconsciously, causing them to mistake a camera for a gun when it was in the hands of a black man and a gun for a camera when it was in the hands of a white man. Sadly, black participants were just as likely to make this pattern of errors as were white participants. Why did this happen?

Stereotypes comprise all the information about human categories that we have absorbed over the years from friends and uncles, books and blogs, jokes and movies and late-night television. When we see black men holding guns in rap videos, our minds associate these two things, and although we realize that we are watching art and not real life, the association is made and remembered. Later, we cannot decide not to be influenced by that association any more than we can decide not to be influenced by our second-grade teacher or the smell of french fries.

In fact, some research suggests that trying not to use our stereotypes can make matters worse instead of better. Participants in one study were shown a photograph of a tough-looking male “skinhead” and were asked to write an essay describing a typical day in his life. Some of the participants were told that they should not allow their stereotypes about skinheads to influence their essays, and others were given no such instructions. Next, the experimenter brought each participant to a room with eight empty chairs. The first chair had a jacket draped over it, and the experimenter explained that it belonged to a skinhead, who had gone to use the restroom. Where did participants choose to sit? Participants who had earlier been told not to use their stereotypes sat farther away from the jacket than did participants who had been given no instructions (Macrae et al., 1994). As you know from reading the Consciousness chapter, thought suppression is ironic business that often causes us to do the very thing we were trying to avoid doing (Wegner et al., 1987).

Although stereotyping is unconscious and automatic, it is not inevitable (Blair, 2002; Kawakami et al., 2000; Milne & Grafman, 2001; Rudman, Ashmore, & Gary, 2001). For instance, police officers who receive special training before playing the camera or gun video game described earlier do not show the same biases that ordinary people do (Correll et al., 2007). Like ordinary people, they take a few milliseconds longer to decide not to shoot a black man than a white man, indicating that their stereotypes are unconsciously and automatically influencing their thinking. But unlike ordinary people, they do not actually shoot black men more often than white men, indicating that they have learned how to keep those stereotypes from influencing their behaviour. Other studies show that even simple games and exercises can reduce the automatic influence of stereotypes (Phills et al., 2011; Todd et al., 2011).

543

In 1963, Dr. Martin Luther King Jr. gave a speech in which he described his vision for America: “I have a dream that my four children will one day live in a nation where they will not be judged by the colour of their skin but by the content of their character.” Research on stereotyping demonstrates that Dr. King’s concerns are still well justified; we judge others by the colour of their skin—as well as by their gender, nationality, religion, age, and occupation—and in so doing, we sometimes make mistakes. But are we any better at judging people if we could somehow turn off our stereotypes and treat each person as an individual?

Not necessarily. Treating a person as an individual means judging them by their own words and deeds. This is more difficult than it sounds because the relationship between what a person is and what a person says or does is not always straightforward. An honest person may lie to save a friend from embarrassment, and a dishonest person may tell the truth to bolster her credibility. Happy people have some weepy moments, polite people can be rude in traffic, and people who despise us can be flattering when they need a favour. In short, people’s behaviour sometimes tells us about the kinds of people they are, but sometimes it simply tells us about the kinds of situations they happen to be in.

To understand people, we need to know not only what they did but also why they did it. Is the batter who hit the home run a talented slugger, or was the wind blowing in just the right direction at just the right time? Is the politician who gave the pro-life speech really opposed to abortion, or was she just trying to win the conservative vote? When we answer questions such as these, we are making attributions, which are inferences about the causes of people’s behaviours (Epley & Waytz, 2010; Gilbert, 1998). We make situational attributions when we decide that a person’s behaviour was caused by some temporary aspect of the situation in which it happened (“He was lucky that the wind carried the ball into the stands”), and we make dispositional attributions when we decide that a person’s behaviour was caused by a relatively enduring tendency to think, feel, or act in a particular way (“He’s got a great eye and a powerful swing”).

How do we know whether to make a dispositional or a situational attribution? According to the covariation model (Kelley, 1967), we must consider the consistency, consensuality, and distinctiveness of the action. For example, why is the man in the photograph in FIGURE 13.19 wearing a watermelon helmet? Does he have a goofy personality (dispositional attribution) or is he just a regular guy who is on his way to a Saskatchewan football game (situational attribution)? According to the covariation model, you can answer this question by asking whether his behaviour is consistent (does he usually wear this hat?), consensual (are other people wearing this hat?), and distinctive (does he do other goofy things?). If it turns out that he wears this hat every day (high consistency), and if today no one else is wearing a melon hat (low consensus), and if he tends to do other goofy things, such as wear clown shoes and say “honk, honk” to passersby (low distinctiveness), then you should probably make a dispositional attribution (“He’s a certified goofball”). On the other hand, if he rarely wears this hat (low consistency), if today lots of other people are wearing watermelons (high consensus), and if he does not tend to do other goofy things (high distinctiveness), then you should probably make a situational attribution (“He’s a Roughriders fan on game day”). As Figure 13.19 shows, patterns of consistency, consensus, and distinctiveness provide useful information about the cause of a person’s behaviour.

544

As sensible as this seems, research suggests that people do not always use this information as they should. The correspondence bias is the tendency to make a dispositional attribution when we should instead make a situational attribution (Gilbert & Malone, 1995; Jones & Harris, 1967; Ross, 1977). This bias is so common and so basic that it is often called the fundamental attribution error. For example, volunteers in one experiment played a trivia game in which one participant acted as the quizmaster and made up a list of unusual questions, another participant acted as the contestant and tried to answer those questions, and a third participant acted as the observer and simply watched the game. The quizmasters tended to ask tricky questions based on their own idiosyncratic knowledge, and contestants were generally unable to answer them. After watching the game, the observers were asked to decide how knowledgeable the quizmaster and the contestant were. Although the quizmasters had asked good questions and the contestants had given bad answers, it should have been clear to the observers that all this asking and answering was a product of the roles they had been assigned to play and that the contestant would have asked equally good questions and the quizmaster would have given equally bad answers had their roles been reversed. And yet observers tended to rate the quizmaster as more knowledgeable than the contestant (Ross, Amabile, & Steinmetz, 1977) and were more likely to choose the quizmaster as their own partner in an upcoming game (Quattrone, 1982). Even when we know that a successful athlete had a home field advantage or that a successful entrepreneur had family connections, we tend to attribute their success to talent and tenacity.

What causes the correspondence bias? First, the situational causes of behaviour are often invisible (Ichheiser, 1949). For example, professors tend to assume that fawning students really do admire them in spite of the strong incentive for students to kiss up to those who control their grades. The problem is that professors can literally see students laughing at witless jokes and applauding after boring lectures, but they cannot see “control over grades.” Situations are not as tangible or visible as behaviours, so it is all too easy to ignore them (Taylor & Fiske, 1978). Second, situational attributions tend to be more complex than dispositional attributions and require more time and attention. When participants in one study were asked to make attributions while performing a mentally taxing task (namely, keeping a seven-digit number in mind), they had no difficulty making dispositional attributions, but they found it quite difficult to make situational attributions (Gilbert, Pelham, & Krull, 1988; Winter & Uleman, 1984). In short, information about situations is hard to get and hard to use, so we tend to believe that other peoples’ actions are caused by their dispositions even when there is a perfectly reasonable situational explanation.

545

The correspondence bias is stronger in some cultures than others (Choi, Nisbett, & Norenzayan, 1999), among some people than others (D’Agostino & Fincher-Kiefer, 1992; Li et al., 2012), and under some circumstances than others. For example, we seem to be more prone to correspondence bias when judging other people’s behaviour than when judging our own. The actor-observer effect is the tendency to make situational attributions for our own behaviours while making dispositional attributions for the identical behaviour of others (Jones & Nisbett, 1972). When university students are asked to explain why they and their friends chose their majors, they tend to explain their own choices in terms of situations (“I chose economics because my parents told me I have to support myself as soon as I am done with university”) and their friends’ choices in terms of dispositions (“Leah chose economics because she’s materialistic”) (Nisbett et al., 1973). The actor–observer effect occurs because people typically have more information about the situations that caused their own behaviour than about the situations that caused other people’s behaviour. We can remember getting the please-major-in-something-practical lecture from our parents, but we were not at Leah’s house to see her get the same lecture. As observers, we are naturally focused on another person’s behaviour, but as actors, we are quite literally focused on the situations in which our behaviour occurs. Indeed, when people are shown videotapes of their conversations that allow them to see themselves from their partner’s point of view, they tend to make dispositional attributions for their own behaviour and situational attributions for their partner’s (Storms, 1973; Taylor & Fiske, 1975).

546