CONTENDING CONCEPTIONS OF SCIENCE AND POLITICS 29within the European Union and its aspiring member states; regional and genocidal wars in the
Balkans and in Rwanda and Burundi; the end of Apartheid in South Africa; the emergence of a
global civil society through the development of the Internet and the proliferation of nongovern-
mental organizations; the reemergence of terrorism on the global scene; the unprecedented inter-
national assent to a “war” that is not between states or between factions within a state, but pits a
coalition of Western military forces operating under executive order against non-state actors; the
Al-Qaeda network, which moves clandestinely within and across national boundaries of multiple
states; or the role of the U.S. Supreme Court in determining the outcome of the 2000 U.S. presi-
dential election, political science failed to predict or explain these critical events.
Perhaps then it is time to reconsider the limitations of the positivist assumptions that undergird
the dominant research methodologies in political and other social sciences, to challenge overly
simplistic versions of empiricism in order to afford social scientists more sophisticated analytic
tools.^1 According to certain versions of empiricism, a simple and direct relation exists between
knower and known. The senses function as faithful recording devices, placing before the “mind’s
eye” exact replicas of that which exists in the external world, without cultural or linguistic media-
tion. Precisely because observation is understood as exact replication, strategies for the acquisi-
tion of knowledge are said to be “neutral” and “value free.” In this view, scientific investigations
can grasp objective reality, because the subjectivity of individual observers can be controlled
through rigid adherence to neutral procedures in the context of systematic experiments, logical
deductions, and statistical analysis of data.
Empiricist assumptions of this sort have been central to the development of the discipline of
political science and to the scientific study of politics in the twentieth century (Finifter 1983; Greenstein
and Polsby 1975; King, Keohane, and Verba 1994; Seidelman and Harpham 1985; Tanenhaus and
Somit 1967). In the following sections, I will explicate and critique the positivist and Popperian
conceptions of science that have profoundly influenced the recent practice of social science, ad-
vance an alternative conception of science, and explore its implications for political inquiry. Clari-
fying and criticizing the methodological presuppositions of the social sciences provides the grounds
for challenging the myth of methodological neutrality, identifying new areas for investigation con-
cerning the political implications of particular modes of inquiry, thereby fostering theoretical self-
consciousness about the relation of social science to contemporary politics.
CONTENDING CONCEPTIONS OF SCIENCE
Positivism
Within the social sciences, empiricist commitments have generated a number of methodological
techniques to ensure the objectivity of scientific investigations. Chief among these is the dichoto-
mous division of the world into the realms of the empirical and the non-empirical. The empirical
realm, comprising all that can be corroborated by the senses, is circumscribed as the legitimate
sphere of scientific investigation. As a residual category, the non-empirical encompasses every-
thing else—religion, philosophy, ethics, aesthetics, and evaluative discourse in general, as well as
myth, dogma and superstition—and is relegated beyond the sphere of science. Within this frame
of reference, social science, operating within the realm of the observable, restricting its focus to
descriptions, explanations, and predictions that are intersubjectively testable, can achieve objec-
tive knowledge. The specific techniques requisite to the achievement of objective knowledge
have been variously defined by two conceptions of science that have shaped the practice of social
science: positivism and critical rationalism.