As we humans evolved, we developed neocortical structures in the brain that allow for high-level thought processes: consciousness, self-awareness, language, logic, and rationality. Yet we also have older brain structures, such as the limbic system, that we share with birds and reptiles. The result of having a hybrid brain is that social cognition is governed by two systems of thinking: a rational, and controlled way of thinking—the cognitive system; and an unconscious, intuitive, and automatic way of thinking—the experiential system (Epstein, 1994; Kahneman, 2011; Sloman, 1996). Depending on the individual and the circumstances, a person’s thought and action can be produced primarily by one or the other system. The rise and fall of facilitated communication as a treatment for autism provides an example.

85

In the fall of 1991, Mark and Laura Storch were informed that their 14-year-old daughter, Jenny, had accused her father of repeated sexual abuse that her mother had ignored. Their daughter was promptly removed from their home while her parents spent the next 10 months fighting the charges, which turned out to be false (Berger, New York Times, February 1994). Her stunned parents were not only shocked by the specific allegations; they were also dumbfounded because their daughter was severely autistic and had little ability to communicate with others verbally! With no ability to share whatever thoughts she had, how had her teachers and aides tapped into Jenny’s inner world? Jenny had apparently told of a history of abuse by using a technique known as facilitated communication, which allows individuals with severe forms of autism to spell out their internal thoughts with the help of an assistant. The assistant, called the facilitator, steadies the autistic person’s arm to allow the individual to hunt and peck at letter keys. When first introduced in the United States in the early 1990s, facilitated communication seemed a revolutionary way to unlock the inner world of loved ones who could not otherwise communicate their thoughts.

Facilitated communication quickly aroused skepticism, however, as children such as Jenny began sharing horrific stories of sexual abuse (Gorman, 1999). When the scientific community investigated the technique, study after study suggested that the thoughts being typed out were not those of the autistic child, but rather were those of the facilitator. In one experiment, two autistic middle schoolers were shown pictures of common objects and asked to type out what they saw (Vázquez, 1994). The experimenter could not see the pictures on the cards, and in half the trials, the facilitator was also prevented from seeing the cards. But in the other half of trials, the facilitator could see what was shown to the child. When their facilitator knew what the card depicted, both children typed out correct answers on all 10 of the trials. However, when the picture was shown only to the children and not to their facilitator, one child was unable to identify any of the pictures correctly, and the other got only 2 out of 10 correct. On the basis of this kind of evidence, the American Psychological Association passed a resolution in 1994 denouncing the validity of facilitated communication.

The rise and fall of this controversial technique was fraught with heartache and dashed hopes. But it also illuminated something rather interesting about human psychology. In practically all of the cases where communicated messages were deemed written by the facilitator and not the child, the facilitators adamantly and fervently believed that they had not and could not have constructed the thoughts that had been typed out on paper. But the research clearly suggests that they had played an integral role. Although their conscious, cognitive system produced the belief that they were merely helping their pupil control his or her muscles, their unconscious, experiential system was likely guiding their pupil’s finger toward each letter to spell out meaningful words, phrases, and ideas.

The core idea that thinking is governed by two systems of thought that operate relatively independently of one another forms the basis for a number of theories that you’ll encounter in this textbook. These theories are often referred to as dual process theories because they posit two ways of processing information. They have been developed to explain wide-ranging phenomena, from the attitudes we hold to the inferences we make about why other people act the way they do.

86

To appreciate the gist of these theories, let’s analyze what you are doing right now. You made a conscious intention to read your social psych textbook, and you’re following through with it, pushing distracting thoughts about other matters out of your mind. You are able to consciously think about the concepts and ideas that you’re reading about, and perhaps you are going further to elaborate on that information— that is, think it over, critique it, and compare it with your prior knowledge and experience. In each case, you are using the cognitive system to consciously direct your attention, guide your behavior, and make deliberate decisions.

At the same time that your cognitive system is busy with rational thinking, your experiential system operates in the background, controlling your more automatic thoughts and behaviors. You might read a sentence about a lazy black dog yawning and lying down, and might find yourself yawning involuntarily, even though you don’t feel the slightest bit tired. Or your favorite song might come on and, before you are consciously aware of it, you find yourself in a better mood. It is because these two systems can operate independently of each other that well-intentioned facilitators could guide their pupils’ hands without being aware of doing so, but it’s also why your unconscious mind can interpret your environment at the same time that your conscious focus is on your textbook.

The two systems have different ways of organizing information. The cognitive system uses a system of rules to fit ideas into logical patterns. Much as your intuitive understanding of English grammar tells you that something is wrong with the statement “Store Jane to the goes,” your cognitive system uses a type of grammar to detect when ideas fit and don’t fit. In this way, it can think critically, plan behavior, and make deliberate decisions. By contrast, the experiential system is guided by automatic associations among stimuli, concepts, and behaviors that have been repeatedly associated through our personal life history of learning. Because of the vast array of information in the world, our brains have evolved to learn from repetition. After repeatedly observing that dark clouds in the sky are often followed by rain, we automatically form an association between clouds and rain. The experiential system organizes these associations into an elaborate network of knowledge.

Because the experiential system stores a large collection of well-learned associations, it can be used to make rapid, “good enough” judgments and decisions at times when using the cognitive system would be too slow and effortful (Epstein, 1990). These mental short cuts, or rules of thumb, are called heuristics. One simple heuristic that people utilize automatically is that “more is better.”

The marble-choice scenario helps us to see the intuitive appeal of heuristics, but it raises an important question: Do heuristics influence judgment when the real-world stakes are high? The answer is “yes.” For example, imagine that a deadly disease is threatening a small town of 600 people, and public health officials are considering two different treatment plans. If Treatment A is adopted, 200 lives will be saved. If Treatment B is adopted, there’s a 1/3 probability that all 600 people will be saved, and a 2/3 probability that no one will be saved. Which would you choose? If you are like most participants asked this question by the Nobel Prize winner Daniel Kahneman and his collaborator, Amos Tversky (Tversky & Kahneman, 1981), you would probably choose Treatment A. But now consider the following version of the same problem: If Treatment A is adopted, 400 people will die. If Treatment B is adopted, there is a 1/3 probability that nobody will die and a 2/3 probability that all 600 people will die. Would you now prefer Treatment B? If you take a close look at these two framings of the issue, they are statistically identical choices. But people’s preferences change dramatically when they are cued to think about what would be lost compared with what might be gained. When the question is framed the second way, in terms of lives lost, the vast majority of people prefer to take the chance with Treatment B where there is some chance of avoiding any loss of life. Our experiential minds are more readily swayed by thinking about what we might lose than by thinking about what we might gain. It takes a much closer and more rational consideration of the odds to realize that these choices are the same.

87

Implicit and Explicit Attitudes

Attitudes are emotional reactions to people, objects, and ideas. If we have two systems for thinking, does that mean we have two ways of evaluating something as good or bad? The answer is “yes” according to dual process theories of attitudes (Gawronski & Bodenhausen, 2006; Nosek, 2007). According to these theories, implicit attitudes are based on automatic associations that make up the experiential system. Some automatic associations can be passed on genetically through evolution (such as an automatic fear response to snakes; Öhman & Mineka, 2003), but most are learned from our culture (such as a negative attitude toward eating pork or fried ants). By contrast, explicit attitudes are often reported consciously through the cognitive system.

Because we have no direct conscious access to our experiential system, measuring people’s implicit attitudes requires a bit of cleverness. One popular task developed by Tony Greenwald and colleagues is the implicit association test (Greenwald et al., 1998). We’ll be describing this task in more detail in chapter 8, where we look more closely at how attitudes are formed and change. We’ll also return to it in chapter 10 because the study of implicit attitudes has been extremely important in our understanding of prejudice. For now, the point to remember is that this task measures the degree to which a person mentally associates two concepts (e.g., “flowers” and “pleasant”), essentially by measuring how quickly she or he can lump together examples of Concept 1 (rose, petunia, tulip) alongside examples of Concept 2 (happy, lucky, freedom). If you are like the average person (and not an entomologist), you’d probably be quicker to throw these flower and pleasant words in the same mental file folder than to group the same pleasant words with insect names such as flea, locust, and maggot. It’s this difference in speed that tells us something about your implicit attitude toward flowers relative to insects, which may or may not be the same as what you would report explicitly on a questionnaire.

88

If the cognitive and experiential systems can both produce attitudes, and if these two systems operate independently of one another, does that mean that the same person can have different attitudes toward the same thing? The answer, again, is “yes.” To illustrate, when volunteers in one study (Nosek, 2005) were asked whether they prefer dogs or cats, what they consciously said—that is, their explicit attitude— was that they prefer dogs. But their responses on a reaction-time measure revealed that, at an implicit level, they associated cats with good more than dogs with good (perhaps because cats seldom have a bad reputation as dangerous animals). People’s explicit attitudes toward dogs and cats were correlated positively with their implicit attitudes, but only moderately so, suggesting that implicit and explicit attitudes can coexist at different levels of consciousness. Not only can they coexist, they can kick in under certain circumstances to influence how we act. Your explicit attitude might dictate which kind of pet you choose to adopt from the local humane society (a very conscious decision), but it’s your implicit attitude that probably accounts for the automatic startle response you might have if you encounter a German Shepherd, rather than a tabby cat, in a dark alley.

As shown in FIGURE 3.3, some attitudes are pretty similar when assessed implicitly or explicitly (Nosek, 2007). For example, in a study on judging political parties, the correlation of about .75 suggests that people’s reported party preferences on a questionnaire correlate quite strongly with their automatic evaluations of Republicans and Democrats. Other attitudes can be quite distinct, so the preference people say they have for family versus career might be only weakly correlated (about .30) with their implicit attitude for one over the other. Our implicit and explicit attitudes are more likely to align when we feel strongly about the issue in question, have given it a lot of thought, and feel comfortable expressing our attitudes (Nosek, 2007). On the other hand, when we are explicitly undecided about an issue, our implicit attitudes predict our later explicit preferences (Galdi et al., 2008). It seems that our experiential system is a bit of a backseat driver at times, whispering directions when our cognitive system is not sure which way to turn.

Automaticity and Controlled Processes

The example of the German Shepherd in the dark alley illustrates how the experiential system guides simple behaviors such as automatic reactions to the environment. But this system guides behavior in more sophisticated ways, as well. In particular, it can control the behaviors necessary to reach our goals. By building up mental associations through routine interactions with the physical and social environment, we can automatize certain behaviors, meaning that we can perform those behaviors without devoting much conscious attention to what we are doing. It is as though we were on autopilot. Think about how you can brush your teeth, drive home from work, or go through the grocery checkout lane without much thought. Such automatization of behaviors is highly adaptive, because it allows us to accomplish goals while saving our mental energy.

But what happens when we encounter a novel challenge that our automatized routine is not prepared to handle? Imagine that you are brushing your teeth with your electronic toothbrush, just as you’ve done every day for the past few years, and all of a sudden the toothbrush stops working. Your experiential system probably won’t be able to handle this situation, because it does not have a set of well-learned associations about toothbrush malfunction. Fortunately, the cognitive system is designed to override the experiential system in these situations, applying controlled processes of reasoning and decision making to solve unexpected problems and help you reach your goals. However, three conditions must be met in order for the cognitive system to successfully override the experiential system:

89

90

We are aware that controlled processes are necessary either to get the job done or to counteract automatic processes that are not working as they should. For example, you consciously have to remind yourself to stop by the drugstore to pick up a new toothbrush battery because your automatic tendency will be to drive straight home.

We are motivated to exert control over our thoughts and behaviors. You need to care enough about getting your toothbrush fixed to change your usual habit of driving home.

We have the ability to consider our thoughts and actions at a more conscious level, because controlled processes require more mental effort. Sometimes we do not have enough cognitive resources to engage controlled ways of thinking. In these cases, the need for nonspecific closure kicks in, usually leading us to think and act in ways that are familiar and automatic. For example, after a long, exhausting day studying at the library, you may be more likely to fall back on unconscious routines and habits (such as driving straight home) rather than working toward new, consciously chosen goals (such as getting that new toothbrush battery) even if you’re aware that you need to remind yourself and are motivated to do so.

By knowing that these three conditions must be in place for the cognitive system to operate, researchers have a powerful way of testing dual process theories in the laboratory. If the cognitive system’s style of deliberate, effortful thinking requires awareness, motivation, and ability, then when people are put into situations where one or more of these conditions is missing, the cognitive system will lose its control over thinking and behavior. For example, if people are asked to memorize a long series of numbers, they lose the ability to focus attention on difficult decisions. In these situations, the experiential system will take over, because it can operate automatically without awareness, motivation, and ability; hence, people will tend to think and act more on the basis of automatic associations, heuristics, and gut-level attitudes. Throughout this book we will see how researchers have used this reasoning to test dual process theories of diverse phenomena.

Although it is tempting to view the conscious, rational cognitive system as the entire basis of human intelligence, and the experiential unconscious as more primitive, in actuality the unconscious is quite smart in at least five ways (FIGURE 3.4). For one, the basic motives that we said earlier guide social cognition—the needs for accurate knowledge, clear knowledge, and preferred knowledge—are largely unconscious. People rarely seem to be aware that these motivations are influencing their judgments and behavior. Second, during sleep, our cognitive system shuts down, but our unconscious stays busy consolidating memories—that is, organizing and solidifying what we’ve learned and experienced (Diekelmann & Born, 2010). Third, studies of ground-breaking artists and scientists indicate that after these individuals engage in extensive conscious deliberation on some problem or issue, an incubation period in which conscious attention is shifted to more mundane matters tends to precede moments of creative insight, which seem to just pop into consciousness out of nowhere, or more precisely, out of the unconscious (Cattell, 1971; Csikszentmihalyi, 1996; Wallas, 1926).

91

Fourth, intuition plays a critical role in good decision making. It was long believed that successful decision making relies on a conscious, systematic, and deliberative process of weighing costs and benefits. In choosing a college, you might have been encouraged to weigh the pros and cons of each school, scrupulously comparing features such as the availability of student aid and the student-to-faculty ratio. A sense of how a campus feels to you when you visit might seem irrelevant, and you might be encouraged to ignore it and focus instead on the facts. But research is beginning to show that unconscious, intuitive processes can steer us toward the best decisions in an automatic way. For example, our unconscious can intuitively sense when information is logically coherent, and it responds with a burst of positive affect (Topolinski & Strack, 2009; Winkielman & Cacioppo, 2001; Winkielman et al., 2007).

In many cases, though, we fail to listen to our unconscious feelings when forming attitudes and making decisions. One reason for this is that we often have difficulty verbalizing—that is, putting into words—why we like or dislike something. In chapter 1 (pp. 11-12) we described a study by Nisbett and Wilson (1977) that made this point by revealing the factors that influenced shoppers’ stocking preferences without their conscious awareness. Because we have so little internal access to what actually determines our emotional reactions, when we are deciding things such as what fruit jam or poster we prefer or even how we feel about a relationship partner or a college, a conscious consideration of what we like or don’t like will lead us to focus on factors that are easy to verbalize. And yet those factors may not reflect our feelings deep down.

In fact, when we think consciously about why we hold an attitude toward something, we often come up with a story that sounds reasonable but that does a poorer job than our gut feelings at predicting later behavior (Wilson et al., 1989). In one study (Wilson & Kraft, 1993), some participants were first asked to analyze the reasons that they felt the way they did about their current romantic relationship, and were then asked to rate their overall satisfaction with the relationship. Another group of participants did not do a reasoned analysis; they just rated their overall satisfaction on the basis of their gut feelings. You might think that the people led to analyze their reasons would figure out how they really felt about the relationship, so that their satisfaction ratings would predict whether their relationship stayed together or not. But the results revealed the exact opposite. It was the people asked to rate their satisfaction based on their gut feelings whose satisfaction ratings predicted whether they were still dating that partner several months later. As for the people asked to think hard about why they felt what they felt, their satisfaction ratings did not predict the outcome of their relationship.

A fifth way that the unconscious is smart is that our unconscious evaluations are essential for good judgment. According to Damasio’s (2001) somatic marker hypothesis, there are certain somatic (i.e., bodily) changes that people experience as an emotion. These somatic changes become automatically associated with the positive or negative contexts for that emotion. When people encounter those contexts again, the somatic changes become a marker or a cue for what will happen next, helping to shape their decisions even without any conscious understanding of what they are doing. We can see this when we compare the decisions made by healthy adults with those made by adults who have suffered damage to areas of the brain responsible for social judgments, particularly the ventromedial sector of the prefrontal cortex.

92

Here’s an example of a typical study. Participants are given a gambling task in which their choice of cards from four different decks can either win or lose them money. Two of the decks are risky; they can give big payouts, but choosing from them repeatedly over the course of the game is a losing strategy. The other two decks give more modest payouts, but the losses are milder as well, and a normal participant eventually learns to stick to these less risky options. Patients with ventromedial damage to the prefrontal cortex, however, don’t learn to avoid the risky decks. Why do these people continue to make high-risk decisions that will lose them money in the end? Part of the reason is that they don’t show any fear that their choices will have negative consequences. Bechara and colleagues (1996) assessed participants’ skin conductance as a measure of arousal just before deciding which deck to choose from. Normal participants showed elevated arousal prior to each pick (FIGURE 3.5). They were anticipating that their choice could be a bad one, and as such, were more likely to learn from their mistakes. Ventromedial patients did not show evidence of this increased arousal, and without that somatic marker to warn them against the riskier decks, they chose from them over and over again as their money dwindled away!

We don’t need to be consciously aware of how our brain is interpreting our emotional associations for those emotions to aid our decision making. In one study (Bechara et al., 1997), 30% of normal participants were unable to explain why they chose cards from one deck more or less than from another. They had no conscious understanding of the patterns that had shaped their decision making over the course of the task, yet they showed the same pattern of improved performance as the participants who had developed a clear hunch that two of the decks were riskier than the others.

93