Finally, just as astronomers are developing more powerful tools to peek into the distant universe, so are psychologists developing better instruments to measure people's internal states. True, we do not yet have an Inner Self Detector, but increasingly sophisticated techniques are being developed, such as measures of the neurological correlates of emotion and affect.
THE THEORY PROBLEM
Suppose we were given the job of designing the optimal human being (in our spare time). Should we endow humans with feelings and emotions? If so, should we make these feelings conscious or unconscious? It would seem pretty odd to say, "Okay, human, you can have feelings, but some times you are not going to be aware of them" What function could this possibly serve?
Such a functional approach can be dangerous, because it is easy to fall into the trap of assuming that every feature of the human mind serves a useful purpose. Nonetheless, the fact that it is easy to tell a story about why conscious feelings are adaptive, and difficult to tell a story about why unconscious feelings are adaptive, seems to favor the incorrigibility argument.
There are two solutions to the theory problem, one old and one new. The old solution is psychoanalytic theory, which argues that the reason feelings can be unconscious is repression. The newer solution is our friend the adaptive unconscious, which might produce feelings independently of people's conscious constructions of their feelings.
Psychoanalysis and repressed feelings. According to Freud, feelings can be kept out of awareness because they are anxiety-provoking, such as failing to recognize sexual attraction toward one's parents. The most dramatic case of repressed feelings is reaction formation, whereby unconscious desires are disguised as their opposite. Erotic attraction toward a member of the same sex, for example, might be so threatening to people that they unconsciously transform their desire into homophobia.'
The psychoanalytic view of repressed feelings has proved difficult to test in a rigorous way. Not only would researchers have to demonstrate that people have a feeling of which they are unaware-which, as we have seen, is no easy matter-they would also have to show that the reason people are unaware of the feeling is that they have repressed it. A number of writers have reviewed the evidence for repression and found it wanting.'
One recent study, however, is quite suggestive. This study examined the psychoanalytic idea that people who are extremely homophobic may be repressing homosexual urges; that is, that their dread of homosexuality may be a means of disguising sexual attraction toward members of the same sex. The researchers recruited male college students who had scored low or high on a questionnaire measure of homophobia. They asked the men to watch sexually explicit videos while they were attached to a device that measured how much of an erection they had. Now, you might be wondering how the researchers did this, and how they got men to agree to do it. In answer to the first question, they used a device called a plethysmograph, a rubber ring placed around the penis that measures changes in its circumference. The plethysmograph is quite sensitive to changes in penis size and has been used extensively as a measure of male sexual arousal. In answer to the second question, the men watched the videos alone in a room and were allowed to attach the plethysmograph themselves; there was no mad scientist demanding that the men drop their trousers in public.
All the videos the men watched depicted consensual sexual behavior between two adults. One film showed heterosexual sex between a man and a woman, one showed lesbian sex between two women, and one showed homosexual sex between two men. The two groups of men showed similar levels of arousal (as measured by the plethysmograph) to the heterosexual and lesbian videos. Consistent with the psychoanalytic hypothesis of reaction formation, however, the homophobic men showed significantly greater increases in penile erection to the male homosexual film than did nonhomophobic men-even though the homophobic men claimed that they were no more aroused by the homosexual film than nonhomophobic men said they were.
This study does not provide airtight proof that the homophobic men had a feeling (sexual attraction toward other men) of which they were unaware. As the authors of the study note, there is some evidence that anxiety can increase sexual arousal; thus the greater increases in erections may have been the result of anxiety in the homophobic men, not sexual attraction. Though not definitive, the study is at least consistent with the psychoanalytic idea of reaction formation, whereby conscious feelings (homophobia) served the purpose of disguising unconscious ones (homosexual attraction).'
Emotions are functional; but do they need to be conscious? Even if we assume that repression is alive and well, the theory problem would not be completely solved. An advocate of incorrigibility might respond, "I am willing to concede that in rare, neurotic cases, people can keep a painful feeling out of consciousness. This is by far the exception, however. In the vast majority of cases, people are fully aware of their feelings, evaluations, and emotions. In fact it is highly functional for people to be aware of their own feelings. Imagine if we had no idea whether we were attracted to or repulsed by a new acquaintance. Not a good way to ensure procreation." This argument suggests that not only is there no compelling theory to explain why people would be unaware of their own feelings (save for unusual cases of repression), but there is a compelling reason why it is to people's advantage to be aware of their own feelings.
The view that emotions serve important functions is an old one with many supporters. Charles Darwin, for example, pointed to the social, communicative functions of emotions. The expression of disgust signals other members of the same species to avoid a certain food; the expression of fear signals our compatriots that danger is near. Emotions may also further the survival of the individuals who have them. An animal that is angry or afraid reacts in ways that make it appear more dangerous to its foes; a cat, for example, displays bared teeth, an arched back, and raised body hair. Fear makes people flee from dangers, and pain teaches them not to touch hot stoves.'
A close look at the functional argument, however, raises a question that has seldom been asked: Does an emotion have to be conscious to be functional? Most theorists have said yes, assuming that events occur in this order: People encounter something in their environment that is dangerous, such as a ferocious bear. The perception of the bear triggers an emotion, namely fear. The conscious experience of this emotion causes the person to act in an adaptive way, such as running in the opposite direction.
As reasonable as this sequence may seem, it is not the only possible explanation of emotional reactions. One problem is that emotions are often slow to develop and occur after people have taken steps to deal with dangerous events. Consider something that happened to me several years ago, when I was driving a rental car during a thunderstorm. Unbeknownst to me, the car had severely worn, bald tires. When I went under a highway bridge and drove from the dry pavement back onto the rain-slicked highway, the tires lost their grip, and the rear of the car fishtailed dangerously from left to right. For a few tense seconds, I fought to regain control of the car and avoid slamming into the guardrail. Fortunately I came out of the skid without incident and continued on my way.
The interesting thing is the point at which I experienced a conscious emotion. According to the standard, functional view, the perception that I was in danger triggered fear, which caused me to take action to regain control of the car. In fact, I did not experience any emotion as I felt the car go into a skid and began tapping the brakes furiously. It was only after the car stopped fishtailing and I was no longer in danger that I experienced the "whoosh" of emotion. ("Oh my God, I could have been killed!") How could my fear have been a signal to act in a lifesaving manner, when it did not occur until after the danger had been averted?
Examples like this were familiar to William James, who proposed a different sequence of events from the standard, evolutionary explanation of emotions. James argued that the perception of environmental events triggers bodily responses, which then trigger conscious emotions; "we feel sorry because we cry, angry beca
use we strike, afraid because we tremble." In his famous example, we do not meet a bear and run because we are afraid; we meet a bear, run, and then experience a post-hoc fear that played no causal role in our fleeing-much like the "whoosh" of fear I experienced after regaining control of the rental car.
James's theory triggered a debate on the relationship between bodily responses and emotions that continues to this day. For our purposes, the issue is whether the conscious experience of emotion is necessary for adaptive responses to environmental threats. James's theory suggests that it may not be, and thus turns the entire issue of the function of emotions on its head. Maybe conscious emotions serve no function at all, but are a by-product of nonconscious cognitive processes that size up the environment and trigger adaptive behaviors-like heat that is released as a by-product of a chemical reaction, but does not cause the reaction.'
A similar argument could be applied to the social function of emotions. It might well be adaptive for a cat to rear its back and hiss when it encounters Rex, the Doberman next door, but perhaps it can do so without the conscious experience of fear. The cat might perceive danger (Rex slipped off of his chain again) and react appropriately, without any conscious experience at all.
But if it is not a conscious emotion that triggers adaptive behaviors, what does? How does the perception of a bear lead to fleeing, without any intervening emotional response? One reason James's theory of emotion was so controversial is that it did not explain how the perception of an environmental event could lead directly to behavioral responses to that event. One possibility is that emotions and feelings do precede adaptive behaviors, but that people are not always aware of these emotions and feelings.
The Adaptive Unconscious Feels
From the nonconscious mental processes we have considered so far, it is a small leap to argue that the adaptive unconscious can have its own beliefs and feelings-not because these beliefs and feelings are so threatening that the forces of repression keep them hidden, but because the adaptive unconscious operates independently of consciousness.
Almost by definition, emotions are states that inundate consciousness. They are often accompanied by bodily changes that are hard to ignore, such as increased heart rate and shortness of breath. How could such a state exist outside of awareness? How could we have a feeling and not feel it? The answer, I suggest, is that we need to adjust our definition of feelings, to allow for the possibility that people can have them without knowing it.
THE NONCONSCIOUS EARLY WARNING SYSTEM
One example of such nonconscious feelings is a danger-detection system documented by Joseph LeDoux. Evolution has endowed mammals (e.g., humans and rats) with two pathways in the brain that process information from the environment differently, dubbed by LeDoux the low road and the high road of emotion. Both roads start at the same place, namely at the point at which information from the environment reaches the sensory receptors and from there travels to the sensory thalamus. The roads also end up at the same place-the amygdala, an almond-shaped region of the forebrain (amygdala means "almond" in Greek) long believed to be involved in the control of emotional responses. The amygdala has neural pathways to the areas of the brain that control heart rate, blood pressure, and other autonomic nervous system responses associated with emotion.
The two roads, however, get to the amygdala via different routes. The low road consists of neural pathways that go directly from the sensory thalamus to the amygdala, allowing information to reach it very quickly, but with only minimal processing of the information. The high road goes first to the cortex, the area of the brain responsible for information processing and thinking, and then to the amygdala. The high road is slower but allows for a more detailed analysis of the information in the cortex.
Why do mammals have these two emotional pathways? One possibility is that the low road evolved first in organisms that did not have a sophisticated cortex. Once the cortex expanded, perhaps it took over the role of emotional processing and superseded the more primitive low road. In LeDoux's words, the low road may be "the brain's version of an appendix" that no longer has any function. LeDoux rejects this view, however, arguing that the low and high roads work in tandem in a quite adaptive manner. The low road operates as an early warning system that quickly alerts people to signs of danger, whereas the high road analyzes information more slowly and thoroughly, allowing people to make more informed judgments about the environment.
To use one of LeDoux's examples, suppose you are walking in the woods and suddenly see a long, snakelike object lying in the middle of the path. You stop instantly and think "Snake!" as your heart begins to beat rapidly. You then realize that the shape is not a snake after all but a fallen branch from a hickory tree, and you go on your way.
What happened, according to LeDoux, is that the image of the stick was sent directly from the sensory thalamus to the amygdala, with a crude analysis that said, "Snake ahead!" This "low road" processing caused you to stop abruptly. Meanwhile, the image was also sent to the cortex, where it was analyzed in more detail, revealing that the object had bark and knotholes. This "high road" processing overrode the initial, low-road response, recognizing that it was a false alarm. The early warning system (the low road) errs on the side of seeing danger ahead; as LeI)oux puts it, "The cost of treating a stick as a snake is less, in the long run, than the cost of treating a snake as a stick." High-road processing serves to put our fears to rest (at least much of the time), saying, "Hey, calm down, snakes don't have knotholes and bark"
The low road of emotional processing operates outside of conscious awareness. We see the stick and freeze, without any conscious feeling or thought. Does this prove, however, that people have nonconscious emotions, or simply mental processes of which they are unaware? This seems to be largely a semantic issue. If what we mean by fear is how we experience it consciously, with its concomitant shortness of breath and the feeling that our hearts are traveling to our throats, then it is very difficult to have these feelings and not be aware of them. But if we mean "Does the person have a nonconscious evaluation that something dangerous lurks ahead?" then the answer seems to be yes: people believe that something is scary and act accordingly. This seems pretty close to saying that people experience an evaluation or emotion of which they are unaware. LeI)oux endorses this latter point of view, arguing that "The brain states and bodily responses are the fundamental facts of an emotion, and the conscious feelings are the frills that have added icing to the emotional cake."'
LeDoux has amassed an impressive amount of evidence consistent with his low-road/high-road picture of emotional processing. As a theory of nonconscious feelings, however, it is limited in three ways. First, all the research has concerned a single emotion, fear. It makes sense to endow humans with an early warning system that makes them freeze at the slightest sign of danger. But what about other emotions and feelings? Can they exist nonconsciously as well? Second, the dichotomy of rudimentary low-road processing and complex high-road processing may not be the full story. I believe it is useful to distinguish between different kinds of "high road" processing, namely processing by the adaptive unconscious and the conscious system.
Third, the theory does not allow for the simultaneous existence of different feelings, one conscious and one nonconscious. In the Topper example, Blake and Kate believed they loved their pony when at some level they also hated him (or so my argument goes). As compelling as LeDoux's early warning system model is, it cannot account for examples such as these. Once the high road has time to analyze the situation it overrules the low road, saying that "sometimes a stick is just a stick." In contrast, the adaptive unconscious may evaluate the environment in one way, while people believe (consciously) that they feel differently.
LOVING AND HATING TOPPER
Why did Blake and Kate both love and hate Topper when they were children? It is possible that this is an example of psychoanalytic repression. Admitting to themselves that they hated a pet they were supposed to love may have
raised anxieties about parental approval, for example, triggering the mechanism of repression. But although such an explanation is possible, there may be a simpler one for this type of unacknowledged feeling.
The adaptive unconscious is an active evaluator of its environment, and when a pony bites us and steps on our feet, it infers that the pony is mean and evaluates it negatively. However, people also have an active, conscious self that simultaneously forms inferences and evaluations. Often the conscious system gets it right. We notice that we have been avoiding Topper and that we are apprehensive in his presence, and infer correctly that we can't stand him.
Sometimes, though, the conscious system gets it wrong. One way this can happen is that people fail to notice that a feeling has changed until their attention is drawn to it. Over a century ago, William Carpenter argued for the existence of such "unnoticed" feelings, such as "the growing up of a powerful attachment between individuals of opposite sexes, without either being aware of the fact." Carpenter noted that "The existence of a mutual attachment, indeed, is often recognised by a bystander ... before either of the parties has made the discovery ... the Cerebral state manifests itself in action, although no distinct consciousness of that state has been attained, chiefly because, the whole attention being attracted by the present enjoyment, there is little disposition to introspection."0
This example meets our "strong suspicion" criterion of a nonconscious feeling, in that people act as if they have a feeling of which they are unaware, observers believe they have the unacknowledged feeling, and the people themselves later acknowledge that they had the feeling (assuming that Carpenter's lovers come to recognize their mutual attraction). Surely, though, the lack of awareness of such strong feelings is temporary. Once people take the time to introspect, they recognize their attraction for another person. In Carpenter's words, the feeling "suddenly bursts forth, like a smouldering fire, into full flame.""
Strangers to Ourselves Page 14