The student who begins his study of international relations today is faced with a discipline that has changed greatly in the last decade. Some would say that it has matured to the point of being a true science. Others would argue that it has fallen into the abyss of data collection, technological quantrophobia and unenlightened description, never to return to the light of theoretical explanation. In part, such debates are healthy signs of growing pains in a young science; but there may be a disheartening aspect to them—one that is not immediately obvious. Let us begin by looking at the two sides in this argument.

When and why people abandon their old institutions, patterns of behavior, and even places of residence in favor of new ones that are associated with sustained economic growth and development is a question of central importance to social science. Increasingly, political scientists have associated these processes of change with the occurrence of disintegration of particular political units, groups, and systems, and with the integration of others. The changing nature of social and cultural ties signals concomitant modifications in what people define as their community, where they place demands, and where they look for authoritative decisions to be made.

The contemporary arms-control community was born in the years 1958–1960, but now—thirteen years later—that community has still failed to confront its fundamental subject matter, arms competitions, in any systematic and rigorous fashion. Elsewhere, I have explored definitions, the ways in which an arms race could be analyzed, and arms-race strategies, outcomes, and dynamics. Here, the cartographic exercise is extended so as to provide a systematic discussion of the fundamental question: Why do states race (or renew a race) in armaments?

Shortly after James Forrestal resigned as Secretary of Defense in late March 1949, the nation was shocked to learn that he was under treatment for a severe mental illness. Within a few months Forrestal committed suicide. This tragic occurrence, coming after Forrestal's highly successful career in government, directly challenged the long-standing mental-health mythology prevalent in Washington. The essence of the myth, as noted by Albert Deutsch at the time, was the belief that “no Very Important Person, under any circumstances, can possibly suffer from a psychosis.” The denial of this possibility in official Washington was of a piece with widely shared beliefs that to suffer a mental illness was a disgrace that automatically and permanently rendered one unfit for public office.

Until recently, our best studies ofpoint of departure the systemicthe developinglevel of analysis—theareas tookna- as tional parties,1 the state institutions,2 the cosmopolitan central elites,3the overarching economic and communications networks4—but the particular assumptions and inevitable shortcomings of this approach led to a call for more attention to subsystem levels of analysis. By now it has become commonplace to note the need for studies of politics at the local level to gain a more intimate understanding of the way in which processes such as political development and political integration take place in the villages and small towns of the Third World's agrarian hinterland.