Maybe clever advertisers will figure out a way of getting subliminal ads to work. Even if they do, however, the effects of their ads are unlikely to be as powerful as everyday ads presented at conscious levels. Despite people's blase attitude toward ads that they see on television, hear on the radio, and see in the print media, these ads can shape their behavior in powerful ways. Perhaps the best evidence for this comes from studies that use split cable market tests. Advertisers, working in conjunction with cable television companies and grocery stores, show different versions of commercials to randomly selected groups of cable subscribers. The subscribers agree to use a special identification card when they shop, allowing the grocery stores to keep track of exactly what they buy. The advertisers can thus tell whether people who see a particular commercial are in fact more likely to buy the advertised product. The answer is that they often are.'
People fear subliminal advertisements (which have no effect) more than everyday advertising (which often has powerful effects) because they worry that they will be influenced without knowing it. But ironically, everyday advertisements are more likely to influence us without our fully recognizing that we are being influenced. It is not as if we go to the drugstore and think, "Should I buy the house brand or Advil? Well, if Advil is good enough for Nolan Ryan, it's good enough for me .. Instead, we might find a name brand more comforting or familiar and not realize why we feel that way. So we shell out the extra cash for something that is no different from the house brand. Nor does a teenager say, "I think I'll start smoking because I want to be like the Marlboro man I saw on a billboard." Instead, adolescents learn to associate smoking with independence and rebellion, with little recognition that it was advertising that helped create this association. Even when we consciously see and hear something such as an advertisement, we can be unaware of the way in which it influences us.
I do not mean to portray people as automatons, marching mindlessly to the commands of Madison Avenue. The failure to recognize the power of advertising makes us more susceptible to it, though, because we are likely to lower our guard while watching commercials or fail to avoid them altogether. Consequently, we can be influenced in unwanted ways without being aware that we are being influenced. Nancy Brekke and I termed this "mental contamination," because our minds can unknowingly become "polluted" with information we would rather not have influence us.'
Given that many studies find that advertising often influences people in unwanted ways, we might entertain the hypothesis that it has the same effect on us. There is lots of good psychological science out there, and by considering it carefully we might gain insight into our own minds. We can then make more informed decisions, such as whether we should worry more about the word "SEX" in ice cubes or everyday TV ads for painkillers. We would also know which button to push on the remote control, if we ever are given the choice between watching subliminal ads versus regular ads.
ARE YOU RACIST?
By some measures, racial prejudice has decreased dramatically over the past few decades in the United States. As recently as 1945, many states and localities had laws that denied African Americans basic freedoms such as whom they could marry, where they could live, and where they could send their children to school. These laws began to change, notably with the 1954 Supreme Court decision to ban segregation in schools and with the 1964 federal Civil Rights Act. Opinions voiced by Americans in polls have improved over the same period. In 1942 only 2 percent of southerners and 40 percent of northerners believed that blacks and whites should attend the same schools, whereas by 1970 these percentages had increased to 40 percent and 83 percent, respectively. In a 1997 Gallup poll, 93 percent of whites said they would vote for a qualified black candidate for president, compared with 35 percent in 1958. Sixtyone percent said they approved of interracial marriage, compared with 4 percent in 1958.
Though encouraging, these figures belie the fact that racial prejudice persists in the United States and elsewhere throughout the world. In 1989 researchers conducted a sobering study to see if there was still racial discrimination in housing in the United States. In twenty locations throughout the country, accomplices of the researchers met with real estate agents to inquire about buying or renting homes and apartments. The accomplices presented themselves as similarly as possible except for their race; some were white, some were black, and some were Hispanic. In a discouragingly large number of cases, the real estate agents discriminated against their minority clients. They presented them with fewer options than their white clients and were less likely to follow up the meeting with phone calls. The amount of discrimination the minorities encountered was about the same as found in a similar study conducted twelve years earlier, suggesting that there has been little or no reduction in housing discrimination over that period.;
It is not hard to find other signs of continuing prejudice. Hate crimes are all too common, such as the brutal 1998 murder of James Byrd Jr. in Jasper, Texas, who was chained to a pickup truck and dragged, simply because he was black. In 1999 four white police officers shot Amadou Diallo forty-one times when he reached for his wallet, mistaking him for a suspect in a rape case. Many believe that the fact that Diallo was black played a role in the officers' readiness to pull the trigger. Although such tragic cases may be rare, blacks continue to experience many forms of discrimination. Approximately half of the African Americans sampled in a 1997 Gallup poll reported that they had experienced discrimination on the basis of their race in the past thirty days, such as while shopping, eating out, or at work.
How can we reconcile the advances that have been made with such stark evidence for lingering bias? To what extent are Americans still prejudiced, and what form does this prejudice take? One possibility is that people are as prejudiced as they ever were, but have learned to hide it better because it has become less culturally acceptable to be openly racist. Although there might be some truth to this, the very fact that cultural norms have changed is a sign of progress. Further, it is not just what people say that has changed. The percentage of people who chose to marry someone of a different race was more than six times higher in 1992 than in 1960. Another possibility is that prejudice has decreased in segments of the American population but persists in a sizable number of people, accounting for the fact that there are still hate crimes, housing discrimination, and bias in the workplace.
A great deal of research in social psychology, however, suggests another possibility: the same person can be both prejudiced and nonprejudiced. A number of researchers have argued that many people abhor prejudice and discrimination and try their best, at a conscious level, to adopt egalitarian attitudes-more so, perhaps, than at any other point in American history. At a more nonconscious, automatic level, however, many of these same people have unknowingly adopted the racist viewpoint that still pervades American culture.
The adaptive unconscious might have learned to respond in prejudiced ways, on the basis of thousands of exposures to racist views in the media or exposure to role models such as one's parents. Some people learn to reject such attitudes at a conscious level, and egalitarian views become a central part of their self-stories. They will act on their conscious, nonprejudiced views when they are monitoring and controlling their behavior, but will act on the more racist disposition of their adaptive unconscious when they are not monitoring or cannot control their actions.
In one study, for example, white college students reported their opinions of an African-American and a white interviewer. The researchers also measured any nonverbal signs of discomfort during the interviews (e.g., the amount of eye contact the students had with each interviewer). The students' opinions were predicted by their conscious beliefs about how prejudiced they were. The less prejudiced people believed themselves to be, the less likely they were to favor the white interviewer over the black interviewer. Their nonverbal reactions, however, told a different story. People's discomfort during the interviews (e.g., the amount of eye contact, how often they blinked) was not related to
their conscious beliefs, but was predicted by a measure of their implicit, automatic prejudice (more on how this was measured in a moment). People who were prejudiced at the automatic level exhibited more negative nonverbal behavior toward the black interviewer, even if they were not at all prejudiced at the conscious level.'
This research might tell us something about ourselves. How can we know if we are prejudiced toward members of various groups, be they African Americans, Hispanics, Asians, whites, women, men, lesbians, gays, or Rotary Club members? Consciously, we might not be prejudiced at all toward these groups, and if it were not for social psychological research on the topic, that would be all there is to it. But on the basis of the research, we might at least entertain the possibility that we have automatic, habitual prejudiced responses toward members of some of these groups of which we are not fully aware.
How can we measure people's level of implicit prejudice, bypassing their conscious beliefs and desires? Most of the techniques rely on computer presentations that time how long it takes people to respond to words and pictures. In one version, people think they are taking part in a study of how well they can do two things at the same time, namely memorizing faces and responding to the meaning of words. A photograph of a face is flashed on a computer screen for about a third of a second, which is quite fast but long enough for people to see consciously. The face is followed almost immediately by an adjective. People are asked to memorize the face and then press one button if the adjective has a positive meaning (e.g., "likable," "wonderful") and another if it has a negative meaning (e.g., "annoying," "disgusting"). The computer times how long it takes them to respond.
It just so happens that some of the pictures that are flashed before the words are of white people and some are of black people. The assumption is that if people are prejudiced at an automatic level, then the race of the face will trigger affective reactions that influence the speed with which they can respond to the words. If people have a negative reaction to a black face, for example, it should be easier to press the "bad" key when a negative word appears, because the negative feelings that are already there will facilitate this response. By the same reasoning, the negative feelings should make people take longer to press the "good" key when a positive word appears, given that the bad feelings are inconsistent with the meaning of the words. The opposite pattern of results should occur when a white face is flashed: because the face triggers positive feelings, people should respond relatively quickly to the good words and slowly to the bad words. On the other hand, if people are not prejudiced, then the race of the face should not influence the speed with which they respond to the words.
The pace of this task is very fast, and people cannot control their responses consciously. There is not enough time for people to say, "Oh, that's a black face; even though I feel a little negatively toward it, I should respond quickly to the positive word that just appeared." Moreover, people do not know that this task has anything to do with their attitudes or stereotypes; they think it is a test of how well they can do two things at once. By observing the speed with which people respond to the words, depending on the race of the face that preceded them, researchers can assess the existence of a pattern of automatic, habitual prejudice.
But can such an artificial task conducted in a psychology laboratory really tap deep-seated feelings toward members of other groups? Well, the proof is in the pudding, namely whether responses on this task predict anything of interest. And indeed they do. The study assessing nonverbal discomfort toward an interviewer used a measure of automatic prejudice much like this one, and other studies have similarly found that responses on the computer task predict how people act toward people of different races. In one study, participants who responded in a prejudiced manner on the computer task were more likely to avoid physical contact with a black student, by placing a pen on the table when it was his turn to use it instead of handing it to him.
Do the measures of automatic prejudice predict more important behaviors than eye blinks and pen passing? An intriguing study by Keith Payne suggests that they might. Participants saw a picture of a white or black face on a computer screen, flashed for a fifth of a second. Then a picture of either a hand tool (such as a pair of pliers) or a handgun appeared, and people had half a second to indicate which type of object it was by pressing a button labeled "tool" or "gun." Given how little time people had to respond, they often made errors by pressing the wrong key.
The interesting questions are what kind of mistakes people (who were nonblack college students) made and whether these mistakes were influenced by the race of the face that preceded the object. Payne hypothesized that many people have an automatic association between blacks and violence, which might make them more prone to mistake a tool for a weapon when it was preceded by a black face. This is in fact what happened. People were significantly more likely to press the "weapon" button when they saw a tool preceded by a black face than when they saw a tool preceded by a white face. The extent to which people made this error was not predicted by a standard, paper-and-pencil measure of racial prejudice; it was an automatic association of which people were not fully aware.
This was, of course, only a laboratory study in which people were seated in front of a computer, pressing buttons in response to pictures of faces and objects, not in response to real people. The parallels of the findings to the Amadou Diallo shooting, however, are sobering. When the police officers saw Diallo reach into his pocket for his wallet, they had about the same amount of time as participants in Payne's study to make a critical decision: Did he have a gun? Tragically, they decided he did, when in fact he was unarmed. We will never know if, had Diallo been white, they would have made a different decision. The Payne study, however, suggests that such errors are influenced by the race of the victim.'
It is important to remember that the police officers had to act extremely quickly. It is not as if they stood around and thought, "Well, let's see, he's black, so he is probably armed." They didn't have time to think at all, at least not consciously. In fact the police officers might well hold completely egalitarian and nonracist beliefs at the conscious level, and would not have been influenced by Diallo's race if they had had time to think. A number of studies have found that there can be a dissociation between people's automatic attitudes rooted in the adaptive unconscious and their conscious beliefs. The person who believes that he or she holds completely egalitarian views might have deeper, automatic reactions toward minorities that are quite negative.
Research on automatic prejudice is in its infancy, and we need to discover a lot more about how best to measure it and what it predicts. From the point of view of self-knowledge, though, this research might make us question and, perhaps, monitor better our own beliefs and behavior. In fact we might not have to speculate about whether these findings apply to ourselves as tests of automatic prejudice become more widely available. It is possible to take one version of these tests on the Internet and to receive a score that is, purportedly, an index of your automatic prejudice.' Clearly, a lot more research is needed to understand fully what these tests are measuring. Nonetheless there is something to the idea that people can be nonprejudiced at a conscious level while their adaptive unconscious feels otherwise, and we should at least question whether this is true of us.
Before individual tests of nonconscious prejudice and other states are perfected and made widely available, the question remains how people can gain greater access to their own idiosyncratic feelings and traits, and not just the general tendencies of participants in research studies. Are there other forms of "self-outsight" that can inform us more directly about our own nonconscious yearnings and motives?
Seeing Ourselves through the Eyes of Others
I have a friend, Mike, who insists that he is shy, to the surprise of everyone who knows him. He appears to meet people easily and has always had plenty of friends. When he travels, he invariably strikes up a conversation with his fellow passengers. He is a great storyte
ller and enjoys regaling people at parties with tales of his childhood in New Jersey. He is an engaging college teacher and appears quite comfortable lecturing in front of hundreds of students.
How can Mike possibly think he is shy, when he clearly possesses such great people skills? Maybe Mike experiences anxiety when he is around other people, despite looking so comfortable and relaxed in social settings. Mike's friends can't get inside his skin to see whether he feels nervous and sweaty before each lecture, or whether he has to force himself to be outgoing and gregarious at parties, when he would really rather be home reading a book.
People's friends are, in fact, less likely to see them as shy than people are to see themselves as shy, precisely because people are good at masking the social anxiety they often feel. If you ask Mike, however, he reports-quite honestly-that he does not feel particularly anxious when he teaches or when telling stories at parties, and that he genuinely enjoys being around large groups of people. Why, then, does he claim that he is a shy person?
From what Mike has told me, he was an introverted child. While most of his classmates were running around the playground shouting and yelling, he was likely to be off to the side drawing in the dirt with a stick. He did not have many friends, though he always had one best buddy. He gravitated toward solitary activities such as writing and computer games, avoiding more social ones such as team sports.
Mike outgrew his introversion by the time he was in college. Beginning in high school he had a wide circle of friends and began taking drama classes. It is not uncommon for children to become less introverted as they age; for example, 50 to 60 percent of college students who say they were shy at ages eight to fourteen report that they are no longer shy. This is what seems to have happened to Mike, except for one thing: he never changed his self-theory that he was shy. We have a case of someone who has a self-theory about his personality ("I am shy and introverted") that is at odds with his adaptive unconscious, which has become more extraverted."
Strangers to Ourselves Page 21