Towards a Theory of Power and Political Instability in Latin America

o NNE MAY consider James B. Conant has observed, an attempt. . . lower the degree of empiricism or extend the range of theory. . . . Almost all significant work of scientists today, I believe, comes under the heading of attempts reduce the degree of empiricism. 1 Despite the conception of science and theory prevalent in the natural sciences and reflected in the formulation of Conant, the discipline of political science has yet reject decisively the notion of theory as an aggregation of metaphysical speculations unrelated experience and practice. Academically, in fact, the field of political theory, within the discipline of political science, traditionally has defined its role as recording, with varied degrees of interpretation, the history of metaphysical speculations in the area of politics and the state. 2 But theorizing, even about politics, as Lasswell and Kaplan correctly assert, not be confused with metaphysical speculation in terms of abstractions hopelessly removed from empirical observation and control. 3 Within a scientific framework, consequently, a theory of power and instability in Latin America must attempt discharge at least four functions. (1) An adequate theory must serve to reduce the degree of empiricism; that is, it must place in a broad and meaningful context the apparently disconnected manifestations of instability in Latin America. (2) It must offer generalizations regarding the exercise of power and the prevalence of instability in Latin America which are compatible with the available evidence. (3) It must offer a guide and provide a framework for relevant research. A scientific theory neither anticipates every future event nor solves every practical problem. A theory with scientific pretensions, however, does provide directions for the investigation and solution of subsequent problems. (4) A theory of Latin American instability, if it is