Decade of the Brain
[Proclamation] [Activities] [Publications & Tapes] [Other Links]
[Decade of Brain Logo]

Discovering Our Selves: The Science of Emotion
Executive Summary

Panel: The Affect of Emotions:
Laying the Groundwork in Childhood

"Understanding the Effects of
Temperament, Anxiety, and Guilt"

Jerome Kagan, Ph.D., is Professor of Psychology at Harvard University and Director of Harvard's interdisciplinary Mind, Brain, and Behavior Initiative.

Rats behave as if they have the capacity for being fearful, surprised, or sexually aroused. There is little evidence, however, to suggest that they experience any of the emotions we call pride, sadness, joy, anxiety, guilt, shame, anger, jealousy, disgust, love, or moral outrage.

Neuroscientists use "fear" to explain the empirical relation between two events: for example, rats freeze when they see a light previously associated with electric shock. Psychiatrists, psychologists, and most citizens, on the other hand, use the word "fear" to name a conscious experience of those who dislike driving over high bridges or encountering large spiders. These two uses suggest there are several fear states, each with its own genetics, incentives, physiological patterns, and behavioral profiles. These related states of fear should be distinguished from states of anxiety, in which future aversive events are anticipated.

For two decades I have been involved in studying a large group of children from whom we have learned that different combinations of temperament and life history can produce an excessively fearful or anxious adolescent or adult. Different fears demand different therapeutic interventions.

Some children appear to have inherited a temperament that biases them to develop a tendency--but does not determine them--to become subdued, timid, and avoidant to unfamiliar events during childhood. Some children in this subset go on to develop a profile that closely resembles social phobia in adolescence.

We began when the children--more than 450 healthy, Caucasian, middle-class infants born at term without any medical complications--were four months old. When stimulated, twenty percent of the infants responded with agitated limb movement, muscle spasticity, and occasional arching of the back, accompanied by fretting and crying. We called those infants high reactive; we assume, based on neuroscience data, that they inherited a low threshold of excitability in the amygdala and its projections.

About 35 to 40 percent of the 450 children showed little motor activity and no distress in response to the same stimuli. We believe the low reactive children inherited an amygdala with a high threshold of excitability due to a different neurochemistry in the amygdala.

Both groups of children were observed at various intervals. When they were 4-1/2 years old, about one-third of the infants who had been high reactive at four months were extremely subdued, and a month later, shy when they played with unfamiliar children. By contrast, only 10 percent of the low reactives were both subdued with the examiner and shy with unfamiliar children.

Only 13 percent of those who had been high reactive at four months, however, showed consistently high fear at both 14 and 21 months as well as extreme shyness at 4-1/2 years. By the time they were 4-1/2, most high reactives (87 percent) had become less fearful and began to resemble typical children, but none became consistently extroverted. The temperament of the high reactive infants constrained their development of an extremely extroverted personality rather than causing a fearful, anxious personality in most of these children.

The children who showed high fear at all ages were most likely to develop anxious symptoms by the time they were seven. About half the children with a serious fear of kidnapping or other danger had been high reactive infants; only 10 percent of those who had been low reactive possessed these anxious symptoms.

The high reactive infants who became inhibited children also displayed extreme sensitivity to criticism over violation of standards they believed were appropriate. That is, they seemed as prone to guilt as to fear and anxiety.

Humans are unique among species in the capacity for guilt. With phylogeny, the central nucleus of the amygdala, critical to the acquisition of conditioned fear reactions in animals, became progressively smaller, while the basolateral nucleus and its connections with the prefrontal cortex were enhanced. This fact, together with an appreciation that humans, but not primates, show signs of distress when they violate their moral standards, implies that guilt and shame may be as important as fear in everyday human functioning, as well as in psychopathology. Natural selection favored those with built-in restraint: The human moral sense, which generally keeps us from harming others of our species, is a unique product of evolution maintained through selection because it ensures our survival.

Post-traumatic stress disorder (PTSD) in veterans has been assumed to be a consequence of an extreme fear reaction to the possibility of being killed. This symptom appears more frequently in soldiers who have participated in an atrocity or feel guilty for surviving than in those who came close to death but did not participate in or witness an atrocity or lose a friend. Feelings of guilt and depression are as salient in PTSD as is fear. The future study of human emotions will profit from a careful examination of this complex emotional state, which is qualitatively different from fear but shares a position of influence in both normal and pathological human functioning.


Go to:
If you have questions or comments on the LC/NIMH Decade of the Brain Project, please contact: [email protected].

Library of Congress
Contact Us
February 8, 2022