Thursday, June 01, 2006

An Argument for Mind

By: Jerome Kagan, Ph.D.

Harvard University emeritus professor Jerome Kagan, Ph.D., describes psychology as “the child of two quarreling parents,” biology and philosophy. He should know, having been on the front-lines of efforts to resolve that quarrel for more than 50 years. In his new memoir, An Argument for Mind, Kagan recounts dramatic changes in psychology’s assumptions and methods over the past half century, set in the context of his own career studying child development and temperament. The relation of brain function to behavior holds a central place in contemporary psychology. But in “Celebrating Mind,” the penultimate chapter of An Argument for Mind from which this excerpt is taken, Kagan explains why he believes that neuroscience can never entirely replace psychology and describes some of the challenges and gifts that each field brings to the other.

Excerpted from An Argument for Mind by Jerome Kagan.
© 2006 Jerome Kagan. Published by Yale University Press in May 2006. Reprinted with permission.

WHAT IS ASSIGNED: IDEAS OR PROCESSES?

When Neil Rudenstine became president of Harvard in 1991 following Derek Bok’s tenure, he recognized that the faculty had become increasingly fragmented. Humanists rarely lunched with chemists, physicists did not attend the colloquia of biologists, and most molecular biologists had no interest in papers written by psychologists. One of Rudenstine’s first acts was to establish informal groups of faculty representing most of the schools of the university—arts and science, divinity, medicine, law, education, business, and public health. The goal of one group, called the Mind/Brain/Behavior Initiative, was to allow neuroscientists, social scientists, educators, humanists, and lawyers to become acquainted with one another’s vocabularies, methods, and premises. Steven Hyman was one of the first directors of the network, and when he left Harvard to become the director of the National Institute of Mental Health, Anne Harrington, a historian of science, and I became codirectors of this consortium. Hyman returned several years later to become Harvard’s provost. 

A primary, but not the only, activity of the consortium consisted of monthly dinner meetings at which someone spoke for about an hour, followed by a lively discussion. This format permitted the varied faculty groups to appreciate the epistemologies of the others. I am not certain, however, whether many minds were changed. Most neuroscientists remained convinced that when the story was complete, the brain would account for the complexities that the humanists and social scientists were probing. I recall an eminent neuroscientist’s reply when I asked, “If you could measure accurately every event going on in my brain at this moment, would you be able to predict that I was about to leap onto the table at which we are sitting and begin to dance?” “Yes,” was his reply, and nothing I said could persuade him otherwise. He was certain that because every action, thought, and feeling has its origin in brain activity, it necessarily followed that all psychological events had to have a distinct instantiation in brain that, once measured, would imply only one inference. Any other belief was irrational. 

“If you could measure accurately every event going on in my brain at this moment, would you be able to predict that I was about to leap onto the table at which we are sitting and begin to dance?”

But it is not certain whether a brain location is the home of a particular experience or the place where a particular process is engaged. Although a face almost always activates a cortical site in the posterior part of the brain, called the fusiform gyrus, a picture of a spider activates this site in adults who are afraid of this animal, and photos of cars activate this site in automobile aficionados. It is not even possible to use activity in the fusiform area to determine whether a person is actually looking at a face or merely imagining it, for this area is activated when individuals are looking not only at a photograph of a face but a photo of the arms and trunk of a man with a homogeneously gray circle that replaces the face. A scientist examining brain activity in the fusiform area could not distinguish between these two states. This cortical area, it appears, is not slavishly reserved for registering faces but is prepared to process any frequently encountered visual event with detailed features requiring close analysis to discriminate between two similar examples from the same class (for example, two faces or two cars). This explains why an autistic eleven-year-old who spent a great deal of time examining a set of cartoon characters, which had facial features, showed greater fusiform activity to the cartoon figures than to human faces. Most cortical areas will be understood best by discovering the processes each was prepared to engage in rather than by trying to determine the contents it was intended to store. My thumb and index finger are prepared to pick up any object with a particular range of size and weight, not just teacups and pencils. 

A good reason for doubting that a particular segment of knowledge is always stored in the same place in all persons is that a carburetor will activate different sites as a function of exposure to this object: in auto mechanics who have replaced thousands of carburetors over their careers and in adults whose carburetors failed when they were driving in a blizzard. To ask where a specific piece of knowledge is represented is similar to asking where a state of hunger is located. A feeling of hunger is the product of activity in many places in the body and brain, and these places may vary across individuals. Adults lying in a brain scanner watching the same movie segment displayed different patterns of brain activity in the frontal lobe because the scenes provoked different thoughts. 

To ask where a specific piece of knowledge is represented is similar to asking where a state of hunger is located. A feeling of hunger is the product of activity in many places in the body and brain, and these places may vary across individuals. Adults lying in a brain scanner watching the same movie segment displayed different patterns of brain activity in the frontal lobe because the scenes provoked different thoughts.

Nonetheless, some scientists continue to believe that particular ideas are assigned fixed places in the brain. I was surprised by a report in which one person lying in a brain scanner was given the role of investor while another, a stranger lying in a separate scanner, had the role of trustee. The investor could invest any part of twenty dollars with the trustee, who then decided how much of the profit to share with his investor partner. The patterns of brain activity, across many pairs of investors and trustees, revealed that after a number of exchanges, the trustees who began to return more money to their partners showed increased neuronal activity in a brain structure called the caudate that is usually involved in motor activity. The scientists concluded that the neurons in this site evaluated “trustworthiness toward another.” The likely flaw in this bold inference is that the caudate is usually active when a person is evaluating incoming information in order to make a motor response indicating a decision. That is exactly what the trustees and partners were doing in the experiment. Because the trustees were more likely to feel ambivalent over being exceptionally generous than over being selfish, we would expect the caudate to be more active under the former mental state. The brain generates a unique waveform whenever a person is instructed to hit a key when he or she sees a particular target appear on a screen. The activity in the caudate is probably reflecting the trustee’s state of indecision over which movement to make (be generous or be selfish) and does not represent “trustworthiness.” 

The neurons of the motor cortex are activated when a basketball player prepares to shoot a basket in the closing seconds of a game in which his or her team is behind by one point. But no one would claim that the “hope of winning the game” is represented in this site. There is an important difference in meaning between “required for” and “represents.” A psychological state, the judgment of pleasure following a good meal, for example, requires particular brain structures, but it is not represented by them because the psychological state is an evaluation that can be mediated by different brain states. Therefore, the mental state need not occur every time these structures are activated and might occur when they are not activated. 

The hope of finding a place in the brain that is the essence of a thought or feeling is as futile as trying to discover the most essential basis for the taste of a piece of milk chocolate. Remove the milk, the sugar, or the chocolate, and the usual sensory experience vanishes. The mind likes the notion of essences, but like the tooth fairy, nature failed to provide examples of this pleasing idea.

The hope of finding a place in the brain that is the essence of a thought or feeling is as futile as trying to discover the most essential basis for the taste of a piece of milk chocolate. Remove the milk, the sugar, or the chocolate, and the usual sensory experience vanishes. The mind likes the notion of essences, but like the tooth fairy, nature failed to provide examples of this pleasing idea. 

Those who hope to find the place in the brain that represents the material foundation for a psychological event forget that many physical phenomena are the product of a relation between structures and are not “in” any of them. The ocean tides and the heat of a stone at high noon under a bright sun are examples. Thoughts, feelings, and action plans are the products of relations among collections of neurons in different places and, like the contagion of a crowd, are not locatable in a particular spot in the brain. 

The seminal point is that each brain profile permits more than one inference about a psychological state. To reduce the number of possible inferences, scientists must know something of the individual’s previous history. But that history, whether years of education or talking with one’s spouse an hour earlier, cannot be described with the words appropriate for brain activity. Conversely, the same psychological event can be the product of different brain states. If asked to name a picture of a maple tree, I would give the same correct answer whether I had just awakened, had exercised for two hours, had drunk a bottle of wine, or had been administered a mild anesthetic. Each of these conditions would have altered my brain in a nontrivial way. I would also have given the correct answer a year later if I had developed Parkinson’s disease during the interval. If a psychological measurement can remain unchanged despite alterations in brain, it follows that measurements of brain alone cannot, in principle, be proxies for a psychological phenomenon. 

Conversely, the same psychological event can be the product of different brain states. If asked to name a picture of a maple tree, I would give the same correct answer whether I had just awakened, had exercised for two hours, had drunk a bottle of wine, or had been administered a mild anesthetic. Each of these conditions would have altered my brain in a nontrivial way.

A paper in the official journal of Britain’s Royal Society argued that scientists one day will be able to state that a particular brain state precedes each freely willed decision or action. They suggested that if my brain could be measured with the powerful machines of the future when I was trying to decide between soup or salad, there would be a moment when my brain was in a particular state that would predict my choice of salad. I suggest that this hope cannot, in principle, be realized because there cannot be a single brain state across all individuals that precedes the selection of salad over soup. More seriously, my brain state the moment before I decided on salad will be different at varying times of day, in different restaurants, and in a state of fatigue versus intoxication. If future neuroscientists were able to measure my brain across a hundred occasions when I selected “salad,” they would record a hundred different brain states, but the choice of salad would have remained the same. Thus, the assumption of a deterministic relation between a brain state and a psychological decision is overly optimistic. 

Fortunately, scientific work supports this judgment. The leech’s nervous system is considerably simpler than our own and far easier to study. A trio of California neuroscientists stimulated varied groups of neurons to see if they could predict whether the subsequent motor pattern would be a swimming or a crawling movement if the leech were in its natural state. They could not make this simple prediction 100 percent of the time because the response following stimulation of the neurons depended on the prior state of the nervous system and unpredictable “noise” in the neuronal matrix. There is an inherent unpredictability at the level of the neuron that resembles the uncertainty of a photon in quantum mechanics. If scientists with complete control of a simple nervous system cannot predict with perfect accuracy which one of two acts will occur, it is highly unlikely—I am tempted to write impossible—that any future scientist with the most elaborate equipment will be able to predict whether I will order soup or salad. 

The predictability of a psychological reaction must always be a probability estimate. Remember Erwin Schrödinger’s cat, enclosed in a box containing an apparatus that, upon detecting a photon, releases a toxin that will kill the cat. An observer cannot know with certainty, following the release of the photon, whether the cat is dead or alive. He or she must open the box to answer that question. Moreover, in the quantum world the act of observation alters the state that existed before the observation was made. Imagine a man wanting to know the postures, facial expressions, comments, and moods of a group of people talking in a room behind a thick door in the seconds before he entered the room. The moment he opens the door he has disturbed the properties he wished to know. Hence, it is impossible, in principle, to be certain of the psychological or biological state of a person before these states are measured because the act of measurement alters those states. I am afraid that the observers monitoring my brain will have to wait until I tell the waiter whether my preference is for soup or salad on any given evening. Free will resides in that narrow corridor of unpredictability that nature built to frustrate humans who entertained the illusion that one day they might know it all. 

Free will resides in that narrow corridor of unpredictability that nature built to frustrate humans who entertained the illusion that one day they might know it all.

THE NEED FOR BIOLOGY

This critique of the reductionists’ hope should not be construed as an argument for ignoring the contributions of biology to a deeper understanding of psychology. Every psychological phenomenon is ambiguous with respect to its historical origins, and adding biological information often illuminates the meaning of a behavior, thought, or emotion. The evidence from brain activity can provide a more profound appreciation of the behavior. For example, the blink of an eyelid that had been conditioned to close at the sight of a clenched fist is mediated by a circuit different from the one activated when a person blinks voluntarily or when startled by a loud sound. The research on temperament I described in chapter 6 suggests that the biology associated with shy behavior in adolescents who had been high-reactive infants is different from the biology that accompanies equally shy behavior in low-reactives. For example, among the eleven-year-olds described by their mothers as shy, quiet, and worried about the opinions of others, failure, and the future, more adolescents who had been high-reactive as infants showed signs of high cortical arousal, whereas most of the shy, anxious low-reactives did not. And among the fifteen-year-olds who were seriously anxious over social interaction, only the high-reactives had high cortical arousal; the few low-reactives who reported similar worries did not display this feature. The socially anxious adolescents who were high-reactive might profit from therapeutic regimens different from those given to socially anxious adolescents with a different biology. 

The syndrome called autism, characterized by seriously impaired social behavior, consists of different types of children, and measures of brain might allow clinicians to detect the distinct types. For example, only some children diagnosed as autistic, not all, fail to scan the eyes in photographs of human faces, suggesting that they may have a dysfunction of the amygdala, because adults with lesions of the amygdala scan the mouth and nose but do not look at the eyes. This fact implies that there is a small group of autistics who have a very particular abnormality. The argument for gathering biological information is the same that explains why physicians order blood and urine tests for patients complaining of exactly the same symptoms, for example, headache, back pain, or fatigue. The body chemistry revealed in the laboratory tests permits the physician to arrive at a more accurate diagnosis and to prescribe more effective therapy. 

The meaning of a single psychological feature is analogous to the meaning of a verb in a sentence. Many verbs can follow different nouns and be followed by different objects and in so doing assume distinct meanings.

The meaning of a single psychological feature is analogous to the meaning of a verb in a sentence. Many verbs can follow different nouns and be followed by different objects and in so doing assume distinct meanings, as in “The boulder struck the car,” “The clock struck noon,” and “The man struck the spectator.” As in language, the more prevalent a psychological feature, the more ambiguous its meaning; the less common, the more restricted its meaning. An extensive vocabulary is more common among children raised in upper-middle-class homes than in children raised in families where no one has a high school diploma. Thus a large vocabulary in a child from a less educated family is more likely to be due to a special biological talent, and biological measures might reveal that fact. 

A second, equally important reason for gathering biological evidence is that it can lead to rejection of an incorrect explanation of a psychological phenomenon. The discovery that the brains of many autistic patients are abnormal has led to rejection of the popular explanation in 1950 that their mothers were cold and aloof. The popular interpretation of separation fear in eight-month-old infants thirty years ago was that the cry of distress when the mother left the child unexpectedly in an unfamiliar place meant that the child was anxious over the loss of its target of attachment. In plainer words, the infant missed its mother. But, as I noted in chapter 3, the more correct explanation is that brain maturation permits the eight-month old to retrieve the representation of the mother’s former presence and to compare it with the present moment while still being unable to understand the discrepancy between the two. The resulting uncertainty leads to the cry of separation. 

The addition of biological to psychological evidence often illuminates the meaning of a feeling, thought, or action, even though the psychological phenomena will have to be described with a distinct vocabulary.

Many more examples are possible. The important point is that the addition of biological to psychological evidence often illuminates the meaning of a feeling, thought, or action, even though the psychological phenomena will have to be described with a distinct vocabulary.



About Cerebrum

 
Bill Glovin, editor
Carolyn Asbury, Ph.D., consultant

Scientific Advisory Board
Joseph T. Coyle, M.D., Harvard Medical School
Kay Redfield Jamison, Ph.D., The Johns Hopkins University School of Medicine
Pierre J. Magistretti, M.D., Ph.D., University of Lausanne Medical School and Hospital
Robert Malenka, M.D., Ph.D., Stanford University School of Medicine
Bruce S. McEwen, Ph.D., The Rockefeller University
Donald Price, M.D., The Johns Hopkins University School of Medicine

Do you have a comment or question about something you've read in CerebrumContact Cerebrum Now.