Seymour Epstein
Warren Kaplan
University of Massachusetts

Intraindividual correlations were computed between stomachaches and
headaches, on the one hand, and emotions and other feeling states, on the
other, for four subjects who kept daily records over 30 days. For all subjects,
anger directed at the self was positively associated with symptoms, and anger
directed at others was either unassociated, or negatively associated, with
symptoms. Different individuals exhibited different relationships between
emotions and symptoms, with some exhibiting emotion-symptom specificity,
while others exhibited only more general relationships. The same emotion was
found to be embedded in different cognitive networks for different in-
dividuals, providing one reason why the same emotion may produce different
psychosomatic effects in different individuals. A cathartic hypothesis received
no support from three subjects and weak support from the fourth. Implica-
tions of the procedure and findings for diagnosis and therapy were discussed.

About all that is widely accepted with respect to the relationship between
psychosomatic symptoms and emotions is that the two are related. There is
little agreement on whether some negative emotions are more implicated in
psychosomatic disorders than others, and on whether there are specific rela-
tionships between particular emotions and particular symptoms. There is
also little agreement about the role that suppression of the expression of an
emotion plays in the production of symptoms. The present study examines
these issues by observing the intrasubject variation of two common

-79-