Abstract

Alcock’s (1972) two-phase theory of war and an earlier test of the theory are critically reviewed. Two pendulum hypotheses linking the onset of war (defined both broadly and narrowly), the formation of alliances, and changes in military expenditures are derived, operationalized, and tested. Contrary to Alcock’s earlier findings, the tests, concentrating on 1900-1965 great power war and alliance behavior, fail to provide any systematic sup-port for the two hypotheses. This does not mean that we can reject the theory, but the evidence indicates that Alcock’s theory does not advance our understanding of great power war and alliance behavior. Explanations of why wars begin are abundant. In fact, their number is matched only by the variety of the variables emphasized in the various approaches to the subject. A list of the variables might easily include alliance structures, polarity patterns and other types of capability dis-tributions, uncertainty, aberrant personalities, merchants of death, military-industrialist complexes, capitalism, imperialism, domestic strife, public moods, arms races, types of political systems, status dis-crepancy, population growth, technological change, cultural/ideolog-ical/religious/racial differences, balances of power, proximity, number of frontiers, territorial irredentism, psychological stress, atavistic instincts, territorial imperatives, misperception and other decision-making pathologies, lust for power, and even Freud’s death-seeking