The researchers varied where the men were when the woman approached them. Half of them were on a scary footbridge that spanned a deep gorge. To cross this bridge people had to stoop over and grasp flimsy hand rails firmly, as the bridge swayed from side to side in the stiff breeze. The other half of the participants had already crossed the bridge and were resting on a park bench. The question was, which group of men were more attracted to the woman: those who encountered her on the bridge or those who were resting on the bench?
This probably seems like a ridiculous question. After all, it was the same woman in both cases, and it was arbitrary whether she approached the men on the bridge or the bench. Or was it? When the woman gave her phone number to the men on the bridge, their heart was beating rapidly, they were a bit short of breath, and they were perspiring. The researchers predicted that these men would be mixed up about exactly why they were physiologically aroused. Surely they recognized to some extent that these symptoms were the result of standing on the flimsy footbridge. Nonetheless, the researchers reasoned, the men might misattribute some of their arousal to attraction to the woman. This is exactly what seems to have happened. Sixty-five percent of the men approached on the bridge called the woman and asked for a date, whereas only 30 percent of the men approached on the bench called and asked for a date. By failing to recognize why they were aroused, people were more attracted to someone than they would otherwise have been.6
Panty Hose, Vacuum Cleaners, and Reasons Why
Maybe these examples of failing to recognize causes of our responses are exceptions rather than the rule. In everyday life, how accurate are people's explanations of why they respond the way they do? And where do these explanations come from?
A number of years ago, Richard Nisbett and I set out to find the answers to these questions with some simple experiments. We placed people in identical situations, save for one or two key features that we varied. We observed how these key features influenced people's judgments or behavior and then asked people to explain why they responded the way they did, to see if they mentioned the features we had varied.
One of our studies was conducted at Meijer's Thrifty Acres, a bargain store just outside Ann Arbor, Michigan. On a busy Saturday morning Nisbett and I placed a sign on a display table that read: "Consumer Evaluation Survey-Which Is the Best Quality?" We then made sure that four pairs of nylon panty hose were arranged neatly on the table and waited for the first passerby to stop and examine them. We were not moonlighting as marketing researchers or working part-time for a panty-hose manufacturer. This was social psychology in action: Would people be able to express accurately all the reasons why they preferred one pair of panty hose to another?
To be able to answer this question, we had to have some idea of what really influenced people's preferences. Here, serendipity was in our favor. In an earlier version of the study, we noticed that people showed a marked preference for items on the right side of the display. We observed this same position effect in the panty-hose study. The panty hose were labeled A, B, C, and D, from left to right. Pair A was preferred by only 12 percent of the participants, pair B by 17 percent, pair C by 31 percent, and pair D by 40 percent, for a statistically significant position effect. We knew that this was a position effect and not that pair D had superior characteristics because in fact all the pairs of panty hose were identicala fact that went unnoticed by almost all our participants.
After people announced their choice, we asked them to explain why they had chosen the pair that they did. People typically pointed to an attribute of their preferred pair, such as its superior knit, sheerness, or elasticity. No one spontaneously mentioned that the position of the panty hose had anything to do with the preference. When we asked people directly whether they thought that the position of the panty hose had influenced their choice, all participants but one looked at us suspiciously and said of course not. The lone exception said that she was taking three psychology courses, had just learned about order effects, and was probably influenced by the position of the panty hose. However, this woman showed little evidence of the position effect-she had chosen pair B.
Nisbett and I soon found ourselves thinking of other ways to test the hypothesis that people do not know reasons for their feelings, judgments, and actions. One night we met in Dick's office to kick around ideas for a new study. Our progress was slow; we couldn't seem to think of any good ideas. After a while it became apparent why (or so we thought): we were distracted by the noise of a custodial worker vacuuming outside the office. Suddenly, we had an inspiration: we had just sat in Dick's office for several minutes frustrated by our lack of progress, failing to recognize that the noise from the vacuum was distracting us. Maybe this was just the kind of situation we were searching for: one in which people would overlook a stimulus (an annoying background noise) that was influencing their judgments.
We tried to "bottle" this experience in the following study. College students watched a documentary film and rated how enjoyable it was. Dick Nisbett, posing as a construction worker, operated a power saw outside the door to the room, beginning about a minute into the film. The noise continued intermittently until I, the experimenter, went to the door and asked the worker to please stop sawing until the film was over. The participants then rated how much they had enjoyed the film and how much the noise had influenced their enjoyment. To see if the noise really did have an effect, we included a control condition in which participants viewed the film without any distracting noise. We hypothesized that the noise would reduce people's enjoyment of the film, but that most people would not realize that the noise was responsible for their negative evaluation (just as we did not realize, at first, that the vacuum was disrupting our meeting).
As it happened, we were completely wrong. The students who watched the film in the presence of the noise enjoyed the film as much as the students who saw the film without the noise. In fact they enjoyed the film slightly more. When we asked participants how much the noise had influenced their ratings, however, they had the same hypothesis as we did: most people reported that the noise had lowered their enjoyment of the film. Even though our initial hypothesis was wrong, we managed to find a case in which people reported that a stimulus had influenced their judgments, when in fact it had not-more evidence of people's lack of insight into the causes of their everyday responses.
Why Do People Misunderstand the Causes of Their Responses?
On the basis of studies such as these, Dick Nisbett and I published an article arguing that people often make inaccurate reports about the causes of their responses because there is "little or no introspective access to higher order cognitive processes." If you wonder how we could make such a sweeping statement on the basis of the panty hose and power-saw studies, you are not alone. A number of critics responsed to our article, arguing that our claims were far too extreme. In our defense, the conclusions were based on more than the demonstration studies we did; we surveyed several large literatures that were consistent with our conclusions about lack of awareness and inaccurate causal reports, including many studies like Dutton and Aron's "love on the bridge" experiment. Nonetheless, our argument did not go unchallenged.
Perhaps the most controversial part of our article was the claim that people have limited introspective access to their mental processes. Any sentient human being knows that an extreme version of this argument is false. The fact that people make errors about the causes of their responses does not mean that their inner worlds are a black box. I can bring to mind a great deal of information that is inaccessible to anyone but me. Unless you can read nay mind, there is no way you could know that a specific memory just came to mind, namely an incident in high school in which I dropped my bag lunch out a third-floor window, narrowly missing a gym teacher who happened to walk around a corner at just the wrong time. Isn't this a case of my having privileged, "introspective access to higher order cognitive processes"?
Ah, Nisbett and I argued, it is true that people have privileged a
ccess to a great deal of information about themselves, such as the content of their current thoughts and memories and the object of their attention. But these are mental contents, not mental processes. The real action in the mind is mental processing that produces feelings, judgments, and behaviors. Although we often have access to the results of these processessuch as my memory of the lunch-dropping incident-we do not have access to the mental processes that produced them. I don't really know, for example, why that particular memory came to mind, just as the participants in the panty-hose study did not know exactly why they preferred pair D over A. Maybe I just saw someone who looked like the gym teacher, heard a song that was popular at the time, or saw something fly by my office window that looked suspiciously like a peanut butter and jelly sandwich. Who knows.
As some of our critics pointed out, however, the distinction between mental content and process is not very tenable. Suppose I heard a song on the radio, which reminded me of the lunch-dropping incident, which reminded me that the teacher I almost hit was also the wrestling coach, which reminded me of professional wrestler Hulk Hogan, which reminded me of Minnesota governor Jesse Ventura. Is each step in this chain of associations a mental content, or is the entire chain a mental process that led from hearing the song to the image of Jesse Ventura?
A better distinction, I believe, is the by now familiar one between the adaptive unconscious and the conscious self. The Nisbett and Wilson argument can be reworked as follows:
• Many human judgments, emotions, thoughts, and behaviors are produced by the adaptive unconscious.
• Because people do not have conscious access to the adaptive unconscious, their conscious selves confabulate reasons for why they responded the way they did, just as P. S., Mr. Thompson, and Estabrooks' hypnotized subject did.
In other words, to the extent that people's responses are caused by the adaptive unconscious, they do not have privileged access to the causes and must infer them, just as Nisbett and I argued. But to the extent that people's responses are caused by the conscious self, they have privileged access to the actual causes of these responses; in short, the Nisbett and Wilson argument was wrong about such cases.
THE CONSCIOUS CAUSALITY QUESTION
But to what extent are human responses the products of the adaptive unconscious versus conscious thoughts? It is clear that the adaptive unconscious is responsible for a good deal of our behavior, and in these instances the reasons for our responses are impossible to access directly. But people also possess a conscious self that directs behavior, at least at times.
Suppose, for example, we observe a customer in a fast-food restaurant ask for a chicken sandwich, and we ask her why she ordered what she did. She would probably say something like, "Well, I usually order the burger, fries, and shake, but I felt more like a chicken sandwich and unsweetened iced tea today. They taste good and are a little healthier." These are precisely the thoughts she was thinking before she asked for the sandwich and thus were responsible for what she ordered-a clear case of conscious causality.
Or is it? Suppose that earlier in the day the fast-food customer encountered someone who was quite obese, which primed issues of weight and self-image, which made her more likely to order food with less fat and calories than the burgers, fries, and shake. The customer was aware of part of the reason she ordered what she did-her conscious thoughts preceding her action-but unaware of what triggered these thoughts. This example illustrates that the question of conscious causality is a very difficult one to answer. There may be relatively few cases in which a response is the pure product of only the adaptive unconscious or only conscious thoughts.
Here's another complication: in examples like this one, it is not even clear that the conscious thoughts preceding the action played any causal role at all. As noted in Chapter 3, Daniel Wegner and Thalia Wheatley argue that the experience of conscious will is often an illusion akin to the "third variable" problem with correlational data. We experience a thought followed by an action and assume that it was the conscious thought that caused that action. In fact a third variable-a nonconscious intention-might have produced both the conscious thought and the action. Seeing the obese person, for example, might have been the cause of thoughts about healthy food and the ordering of the chicken sandwich. The conscious thoughts may not have caused the behavior, despite the illusion that they did so.
Wegner and Wheatley's provocative theory illustrates that a sense of conscious will cannot be taken as evidence that conscious thoughts really did cause our behavior. The causal role of conscious thought has been vastly overrated; instead, it is often a post-hoc explanation of responses that emanated from the adaptive unconscious.
DO STRANGERS KNOW THE REASONS FOR YOUR RESPONSES AS WELL AS YOU DO?
Where do people's confabulations come from? Suppose that someone asked you to describe the major influences on your daily mood. To the extent that your adaptive unconscious influences your mood, you will not be able to examine these influences directly. Instead, there are four general types of information you can use to create an explanation:
• Shared causal theories. There are many cultural theories about why people respond the way they do, such as "absence makes the heart grow fonder" and "people are in bad moods on Mondays." If people do not have a ready-made theory to explain a particular response, they can often generate one based on their storehouse of cultural knowledge about what makes people tick. ("Why did Jane break up with Tom? The fact that he kept calling her by his ex-girlfriend's name probably had something to do with it.")
• Observations of covariation between one's responses and prior conditions. People can observe their own responses and infer what is causing them. People discover what they are allergic to, for example, not by directly examining their digestive processes but by observing the covariation between eating certain foods (e.g., pecans) and allergic reactions (e.g., breaking out into hives). In the same way, people might deduce that they like movies starring Robert DeNiro, are in bad moods when they get less than seven hours of sleep, and catch colds when they forget to wear a jacket in freezing weather.
• Idiosyncratic theories. People have idiosyncratic theories about the causes of their responses that are not shared by the culture at large, such as the theory that going to large parties often makes them depressed. These theories might result from observations of covariation; for example, Jim might observe that he was depressed after the last few parties he attended. People can also learn idiosyncratic theories from others. A person's spouse, for example, might say, "Honey, I noticed that you were down in the dumps at the Jones's lawn party, the Greenbergs' anniversary party, and Sam's birthday party. What's up with you and large parties?"
• Private knowledge (thoughts, feelings, and memories). Although access to one's own mind is not perfect, people have a wealth of privileged knowledge about their own conscious thoughts, feelings, and memories that they can use to deduce what is causing them to respond the way they do. If Jim is feeling sad and knows that he has been thinking a lot about the time his cat ate his favorite goldfish, he might deduce that it is the memories of Goldie's demise that are making him sad.
Perhaps the most radical part of Nisbett and Wilson's argument is that despite the vast amount of information people have, their explanations about the causes of their responses are no more accurate than the explanations of a complete stranger who lives in the same culture." How can this possibly be true? Can it really be the case that a complete stranger, chosen randomly out of the phone book, will know as much as we do about why we respond the way we do? Surely the vast amount of "inside information" we have about ourselves gives us an advantage. Suppose that we are avid baseball fans and that during baseball season our mood fluctuates with the fortunes of our favorite team. Because a stranger doesn't know whether we are baseball fans, political junkies who watch Crossfire every night, or frequent bidders at auctions on the eBay website, how could this person possibly know as accurately as we do what influences our
moods?
True enough, we do have a lot more information about ourselves than a stranger does. However, this information may not always lead to accurate inferences about the causes of our responses. Of the four kinds of information listed above, the stranger has only the first-shared cultural theories. The fact that we also have covariation information, idiosyncratic theories, and private knowledge, however, can be a hindrance as well as an advantage.
For one thing, some of this privileged knowledge isn't as accurate as it might seem. There is considerable evidence that people are not very skilled at consciously observing the covariation between their responses and its antecedents. Sometimes a covariation is so striking that we can't help but notice it, such as the fact that we broke out in hives immediately after eating pecans for the first time. More commonly there are many antecedents of our responses, and it is difficult to tease apart which ones are the causes. Because of this difficulty, people's beliefs about covariation are often a function of their shared cultural theories, rather than deductions based on accurate observations of their own behavior. There is no evidence, for example, that going outside without a jacket increases the likelihood that people will catch colds, despite the cultural theories to this effect.'
In addition, the vast amount of privileged information people have might make it harder to recognize causes of their behavior that a stranger, relying on cultural theories, would see. Suppose, for example, that a medical student who has just learned about diabetes stands up quickly and gets dizzy. The student might think, "Uh oh, I better have my blood sugar checked; I could be in the early stages of diabetes, which is impeding my blood circulation." A stranger, who knows nothing about what the student has been studying or thinking about, is likely to say, "She got dizzy because she stood up too quickly." The stranger may be right in this case, which would be an example of a person's inside information (the student's knowledge of diabetes) leading to inaccurate causal reports.
Strangers to Ourselves Page 12