The function of consciousness is a more tractable question and is the one with which I will be most concerned. Before considering how best to obtain self-knowledge, we need to make at least some headway on such questions as whether it makes any difference to know ourselves. Does gaining insight (becoming conscious of previously unknown things about ourselves) change anything? Does the person who has limited insight into the reasons for her actions, for example, behave any differently from the person who has great insight?
A standard analogy is that consciousness is the president in the executive branch of the mind. In this conception, there is a vast network of agencies, aides, cabinet officers, and support staff who work out of view of the president. This is the adaptive unconscious, and a smoothrunning government could not exist without it. There is simply too much for one person to try to do, and a president could not function without his or her many (nonconscious) agencies operating out of view. The president is in charge of this vast network, setting policy, making the major decisions, and intervening when serious problems arise. Clearly, consciousness plays a crucial function in these activities. The adaptive unconscious is subservient to consciousness (the president) and reports to it. At the same time, the president who becomes too out of touch is in trouble. If he or she is ignorant of what is occurring out of sight (lacking in self-insight), then the agencies of the adaptive unconscious may start to make decisions that are contrary to the wishes of the president.
Others have questioned the consciousness-as-chief-executive analogy, arguing that consciousness may not play such a crucial role. At one extreme are philosophers who argue that consciousness does not serve any function at all. This position, dubbed "conscious inessentialism" or "epiphenomenalism," holds that consciousness is an epiphenomenal byproduct of a skilled, nonconscious mind that does all the real work. Consciousness is like the child who "plays" a video game at an arcade without putting any money into it. He moves the controls, unaware that he is seeing a demonstration program that is independent of his actions. The child (consciousness) believes he is controlling the action, when in fact the software in the machine (nonconsciousness) is completely in control.'
The philosopher Daniel Dennett notes that this view equates consciousness more with the press secretary than with the president. The press secretary can observe and report on the workings of the mind but has no role in setting policy and is not privy to many of the decisions made behind the closed doors of the Oval Office. It's an observer, not a player.'
How can this be, you might ask, when it so often feels as though we are consciously controlling our actions? Recent work by Daniel Wegner and Thalia Wheatley suggests an answer: the experience of conscious will is often an illusion akin to the "third variable" problem in correlational data. We often experience a thought followed by an action, and assume it was the thought that caused that action. In fact a third variable, a nonconscious intention, might have produced both the conscious thought and the action. My decision to get up off the couch and get something to eat, for example, feels very much like a consciously willed action, because right before standing up I had the conscious thought "A bowl of cereal with strawberries sure would taste good right now" It is possible, however, that my desire to eat arose nonconsciously and caused both my conscious thought about cereal and my trip to the kitchen. The conscious thought might have been completely epiphenomenal and had no influence on my behavior, just as consciousness appears to be unnecessary in lower species in order for them to seek food and survive. Even humans sometimes behave in seemingly intentional ways in the absence of relevant conscious thoughts, such as when I find myself getting off the couch to get a bowl of cereal without ever consciously thinking about what I am doing or willing myself to do so.'
Wegner and Wheatley acknowledge that conscious will is not always an illusion, just that it can be. The most reasonable position, I believe, is between the extremes of consciousness-as-chief-executive and consciousness-as-epiphenomenal-press-secretary. If consciousness were purely epiphenomenal, then a book on self-insight would not be very satisfying. It might give people a better seat from which to observe the action, but these observations could not change the course or outcome of the game. On the other hand, we have already seen that the adaptive unconscious is quite extensive and includes such higher-order, executive functions as goal-setting. Thus, I think the analogy of consciousnessas-chief-executive or head coach is also misleading. We may have the impression that we, our conscious selves, are in complete control, but that is at least in part an illusion.
The philosopher Owen Flanagan notes that different U.S. presidents have exerted differing amounts of control over governmental policy, and that a more accurate view of the role of consciousness may be consciousness-as-Ronald Reagan. According to many historians, Reagan was more of a figurehead than most presidents and did not exert very much control over the government. In Flanagan's words, "Reagan was the entertaining and eloquent spokesperson for a cadre of smart and hardworking powers (actually layers of powers), some known to outsiders, and some unknown. This is not to deny that Reagan felt as if he were in charge in his role as `The Great Communicator'. . . The point is that one can feel presidential, and indeed be presidential, but still be less in control than it seems from either the inside or outside."'
In other words, we know less than we think we do about our own minds, and exert less control over our own minds than we think. And yet we retain some ability to influence how our minds work. Even if the adaptive unconscious is operating intelligently outside our purview, we can influence the information it uses to make inferences and form goals. One of the purposes of this book is to suggest ways this can be done.
In a memorable Saturday Night Live skit from the 1980s, President Reagan was portrayed as a brilliant, cunning leader whose "Great Communicator" persona was all a shtik. In public, he was the fatherly, slightly bumbling Hollywood actor the voters knew and loved. Behind the scenes, he was a ruthless visionary who could think circles around his aides and negotiate brilliantly with foreign leaders. (In one scene, he gets tough with an Iranian leader over the phone-while speaking Farsi.) The goal of this book is to make us all more like the Ronald Reagan in the skit-an executive who knows and manipulates, at least to some extent, what is going on behind the scenes.
Properties of the Adaptive Unconscious versus Consciousness
But what is going on behind the scenes, and how does this differ from conscious processing? It is useful to map out the different functions of these mental systems, which are summarized in the table.
MULTIPLE VERSUS SINGLE SYSTEMS
As already noted it is a bit of a misnomer to speak of the adaptive unconscious, as there are a collection of modules that perform independent functions outside of conscious view. One way we know this is through studies of brain-damaged patients; different areas of the brain seem to be associated with quite different aspects of nonconscious learning and memory. Damage to some areas can impair explicit memory, for example (the ability to form new memories), but leave implicit memory intact (e.g., the ability to learn new motor skills). Strokes can impair language abilities without influencing other cognitive functions. Because the adaptive unconscious is a collection of many independent abilities, some of the properties of the adaptive unconscious I describe may apply to some modules more than to others.
Consciousness, on the other hand, seems to be a single entity. Exactly how to define it, and exactly how it is related to brain functioning, are not known. It is relatively clear, however, that it is a solitary mental system, not a collection of different modules. There may be special cases in which consciousness can split into two or more independent systems, such as multiple personalities (although the exact nature and frequency of multiple-personality syndrome is the topic of much current debate). Most people, however, do not possess more than one conscious self. There is only one president, even if that entity does not have as much power or control as it thinks.
PATTERN DETECTOR VERSUS FACT CHECKER
A number of psychologists have argued that the job of the adaptive unconscious is to detect patterns in the environment as quickly as possible and to signal the person as to whether they are good or bad. Such a system has obvious advantages, but it also comes with a cost: the quicker the analysis, the more error-prone it is likely to be. It would be advantageous to have another, slower system that can provide a more detailed analysis of the environment, catching errors made by the initial, quick analysis. This is the job of conscious processing.
Joseph LeDoux, for example, suggests that humans have a nonconscious "danger detector" that sizes up incoming information before it reaches conscious awareness. If it determines that the information is threatening, it triggers a fear response. Because this nonconscious analysis is very fast it is fairly crude and will sometimes make mistakes. Thus it is good to have a secondary, detailed processing system that can correct these mistakes. Suppose that you are on a hike and suddenly see a long, skinny, brown object in the middle of the path. Your first thought is "snake!" and you stop quickly with a sharp intake of breath. Upon closer analysis, however, you realize that the object is a branch from a small tree, and you go on your way. According to LeDoux, you performed an initial, crude analysis of the stick nonconsciously, followed by a more detailed, conscious analysis. All in all, not a bad combination of systems to have."
THE HERE-AND-NOW VERSUS THE LONG VIEW
Useful though the nonconscious pattern detector is, it is tied to the hereand-now. It reacts quickly to our current environment, skillfully detects patterns, alerts us to any dangers, and sets in motion goal-directed behaviors. What it cannot do is anticipate what will happen tomorrow, next week, or next year, and plan accordingly. Nor can the adaptive unconscious muse about the past and integrate it into a coherent selfnarrative. Among the major functions of consciousness are the abilities to anticipate, mentally simulate, and plan.
An organism that has a concept of the future and past, and is able to reflect on these time periods at will, is in a better position to make effective long-term plans than one that does not-providing a tremendous survival advantage. In some lower organisms, planning for the future is innate: squirrels "know" to store nuts for the winter, and migratory birds "know" when to fly south to warmer weather. Imagine the advantage of having a more flexible mental system that can muse, reflect, ponder, and contemplate alternative futures and connect these scenarios to the past. The practice of agriculture, for example, requires knowledge of the past and thinking about the future; why bother putting seeds in the ground now if we cannot envision what will happen to them over the next few weeks?
The idea that consciousness plans for the future probably does not come as much of a surprise. Those who endorse the consciousness-aschief-executive model would agree that a major function of consciousness is to engage in long-term planning. A good CEO leaves the little stuff to underlings and spends his or her time on the big questions, such as what the long-term goals should be and how to implement them.
Our consciousness-as-Ronald Reagan model, however, portrays longterm planning a little differently. The federal government (the mind) is a vast, interrelated system that operates quite well on a day-to-day basis. The chief executive can look into the future and try to set long-term goals, but might find it difficult to make major changes in policy. Often the best he or she can do is to nudge the vast bureaucracy onto a slightly different course. In fact there is a danger to making major policy changes for which the rest of the mind is unsuited.
Consider Herman, who believes that he is a loner who is happiest when by himself doing his own thing, when in fact he has a strong, nonconscious need for affiliation with other people. Because it is his conscious self-view that plans his future and determines his behavior, Herman avoids large gatherings and parties and chooses a career as a computer consultant so that he can work out of his home. His nonconscious need for affiliation is unfulfilled by these choices, however, leading to unhappiness. Perhaps the best use of consciousness is to put ourselves in situations in which our adaptive unconscious can work smoothly. This is best achieved by recognizing what our nonconscious needs and traits are and planning accordingly.'
But how do we recognize what our nonconscious needs and motives are? That is the million-dollar question. For now, I note simply that the ability to think about and plan for the future endows humans with a tremendous advantage, but can be a two-edged sword. Following our conscious wishes can be problematic if they conflict with the desires of the adaptive unconscious.
AUTOMATIC VERSUS CONTROLLED PROCESSING
It is well known that people can perform many behaviors (e.g., riding a bicycle, driving a car, playing the piano) quickly, effortlessly, and with little conscious attention. Once we have learned such complex motor behaviors, we can perform them better when we are on automatic pilot and are not consciously thinking about what we are doing. The moment I begin to think about what my pinkie and index fingers are doing as I type these words, typos result. There is a term for this in athletics: when a player is "unconscious," she is performing at an optimal level without any awareness of exactly what she is doing. She is in the zone.
Although we do not often conceive of thinking in the same way, it, too, can happen automatically. Just as playing the piano can become automatic, so can habitual ways of processing information about the physical and social world. Indeed, a defining feature of the adaptive unconscious is its ability to operate on automatic pilot. Automatic thinking has five defining features: it is nonconscious, fast, unintentional, uncontrollable, and effortless. As noted by the social psychologist John Bargh, different kinds of automatic thinking meet these criteria to varying degrees; for our purposes we can define automaticity as thinking that satisfies all or most of these criteria.
We have already encountered examples of this type of thinking in Chapter 2-namely, the way in which the adaptive unconscious selects, interprets, and evaluates incoming information. Consider the cocktailparty phenomenon, in which the adaptive unconscious blocks out all the conversations except the one we are in, but at the same time monitors what other people are saying (and alerts us if they say something important, such as our name). This process meets all five of the criteria of automaticity: it occurs quickly, nonconsciously, and without intention, in the sense that our nonconscious filter operates even when we have no intention that it do so. It is uncontrollable, in the sense that we have little say over the operation of the nonconscious filter and could not stop it if we tried. Finally, it operates effortlessly, in the sense that the nonconscious filter takes up little mental energy or resources.
Another example of automatic thinking is the tendency to categorize and stereotype other people. When we meet somebody for the first time, we pigeonhole them according to their race or gender or age very quickly, without even knowing we are doing so. This process of automatic stereotyping is probably innate; we are prewired to fit people into categories. The nature of the pigeonholes, however-the content of our stereotypes-is certainly not innate. No one is born with a specific stereotype about another group, but once we learn these stereotypes, usually from our immediate culture, we are inclined to apply them nonconsciously, unintentionally, uncontrollably, and effortlessly. In contrast, conscious thinking occurs more slowly, with intention (we typically think what we want to think), control (we are better able to influence what we think about), and effort (it is hard to keep our conscious minds on something when we are distracted or preoccupied)."'
THE RIGIDITY OF THE ADAPTIVE UNCONSCIOUS
A disadvantage of a system that processes information quickly and efficiently is that it is slow to respond to new, contradictory information. In fact we often unconsciously bend new information to fit our preconceptions, making it next to impossible to realize that our preconceptions are wrong. An example is my assumption that Phil, the man I met at a PTO meeting, was the pushy, rude fellow I had heard about, when in fact he was not.
What happens when the noncons
cious system quickly detects a violation of a pattern? Does it recognize that the old way of seeing things no longer applies? Suppose, for example, that a business manager notices (at a nonconscious level) that the last two employees she had to fire had degrees from small, liberal-arts colleges and that the last three people she promoted had degrees from large, state universities. It is now jobperformance time, and the manager is evaluating a new batch of employees, some of whom went to small, liberal-arts colleges and some of whom went to state universities. On average, the two groups have performed at the same level, although each did better on some tasks than on others. How will the manager evaluate these employees?
A smart, flexible system would recognize that the previously learned correlation, from a very small sample, does not generalize to this larger sample of employees. And yet once a correlation is learned, the nonconscious system tends to see it where it does not exist, thereby becoming more convinced that the correlation is true. When evaluating the employees who went to small colleges, the manager may focus on and remember the times they did poorly. When evaluating the employees who went to large universities, she is likely to focus on and remember the times they did well, thereby strengthening her belief that the size of a person's alma mater is predictive of job performance-even though it is not.
Even worse, people can unknowingly behave in ways that make their expectations come true, as in Robert Rosenthal and Lenore Jacobson's classic research on the self-fulfilling prophecy. They found that teachers not only view their students in the ways that they expect them to be, but act in ways that make these expectations come true. At the beginning of the school year, they administered a test to all the students in an elementary school and told the teachers that some of the students had scored so well that they were sure to "bloom" academically. In fact this was not necessarily true: the students identified as "bloomers" had been chosen randomly by the researchers. Neither the students nor their parents were told anything about the results of the test. The "bloomers" differed from their peers only in the minds of the teachers.
Strangers to Ourselves Page 6