Search

This article emerged from the work of the Carnegie Commission on Preventing Deadly Conflict, Washington, DC. The commission addressed several fundamental questions: (1) What are the problems posed by deadly conflict, and why is outside help often necessary to deal with these problems? (2) How can disputes be resolved peaceably? (3) Which strategies work best? (4) Who can do what to implement these preventive strategies? Borrowing from the model of preventive medicine, the commission detailed a repertoire of the most promising political, economic, military, and social tools and strategies that can be mobilized by the international community to assist vulnerable societies in the development of sustainable and equitable arrangements for managing diversity and resolving disputes peacefully. From a comparative examination of intransigent and destructive intergroup conflicts, the commission found that the failure to prevent conflict is most often not a failure of foreknowledge or capacity but of political will. Effective political leadership is often the critical variable for successful prevention. This article seeks to illustrate how the social and behavioral sciences may be usefully applied to the problems encountered by leaders when confronted by the challenges of preventing deadly conflict.

Mel Sabshin and I worked closely together during the 1950s, and his influence on me-as on so many others-has persisted ever since. I have never ceased to be deeply impressed by his intellectual curiosity, his strong sense of social responsibility, his integrative capacity across disciplines, and his constructive problem-solving orientation. One of his pioneering and enduring interests over several decades has been in the study of normality. Within that frame-work, he has devoted special attention to adolescent development. In this article I pick up on this strand of his interest. During much of my career, and especially during the past 14 years at the Carnegie Corporation, I have sought ways to build the knowledge base on adolescent development and put that knowledge to use in preventing lifelong casualties.

Twenty males participated in a 2-month study examining the relationship between 8 a.m. plasma testosterone levels and orgasmic frequency. Within subjects, higher levels of testosterone are associated with periods of sexual activity. Over subjects, however, the direction of the relationship is reversed. Mean testosterone levels were higher for sexually less active individuals.

A long-acting preparation of gonadotropin-releasing hormone (GnRH) was administered to three adult male chimpanzees living in a socially integrated group. Hormonal changes and behavioral responses were monitored for 8 days. Blood serum levels of luteinizing hormone peaked at about 8 hr and returned to baseline at 48 hr. Total serum testosterone reached the highest levels of 140% above baseline at 24 hr and returned to baseline after 48 hr. Out of a wide variety of behaviors observed in detail, only one cluster of three related behaviors showed even a suggestion of response: The two measures of activity, lone travel (p < 0.05) and distance traveled (p < 0.1), decreased on the day of hormone administration followed by an increase on the 2 subsequent days, and idle time (p < 0.1) revealed the reciprocal pattern. In view of various reports of behavioral effects of GnRH in other species, it is particularly noteworthly that parameters of sexual and agonistic behavior of male chimpanzees definitely did not change.

Gonadotropin-releasing hormone in doses up to 500 microgram was administered to 12 healthy male volunteers. Luteinizing hormone and testosterone levels increased subsequent to GnRH administration. No immediate effects of GnRH on mood and behavior were noted, though an increase in alertness, a decrease in anxiety and fatigue, and an increased speed of performance on automatized motor tasks were noted several hours after GnRH administration.

The object of the study was to assess the lability of testosterone levels in plasma of normal human males over a long period of time and to search for periodicities in changing levels. Blood samples obtained from 20 healthy young men every second day for 2 months were assayed for total testosterone concentration by radioligand saturation analysis with late-pregnancy plasma. The flucturations of plasma testosterone levels over the total time span were substantial for most individuals; the coefficients of variation ranged from 14 to 42% (median 21%). The presence of periodic functions in these fluctuations was tested by 4 different, relatively independent methods. Close agreement among at least 3 analytic methods was found for 12 out of the 20 subjects. These 12 subjects had cycles of plasma testosterone levels with periods ranging between 8-30 days, with a cluster of periods around 20-22 days. The majority of such cycles were significant at least at the 5% level. The mean amplitudes of these cycles ranged from 9 to 28% of the subjects' mean testosterone levels (average 17%).

In this era of rapid, far-reaching transformation, our way of life is in many respects a novelty for our species. Opportunities arising from profoundly enhanced capabilities in science and technology are felt in every sphere of life from health to communication, yet each advance has side effects that take time to appear. Grave institutional inadequacies are manifested in the prevalence of totalitarian governments, proliferation of devastating weapons, failure of educational institutions to prepare most people for the modern world, failure to use what we know to prevent damage to a large proportion of the world's children; and the weakness of international institutions to deal with global interdependence in the face of persistent ethnocentrism and prejudice. The American scientific community can usefully become more deeply engaged with these great problems.

The association between mood and plasma testosterone was investigated. Every second day for 2 months, a Multiple Affect Adjective Checklist was filled out by 20 young men. Blood samples taken at the same time were analyzed for testosterone concentration. Relationships between hostility, anxiety, and depression and plasma testosterone levels were tested both within and across individuals. Intrasubject correlation coefficients between affects and hormone were fairly evenly distributed between positive and negative values, some significant on either side. Intersubject correlation coefficients were all positivie. Between depression and testosterone, the correlation was significant only at the 10% level; between the other two affects and testosterone, the correlations were not significant.