GENERALIZATION IN COMPARATIVE AND HISTORICAL SOCIAL SCIENCE 59issue. Using criteria that presuppose a modernist goal, they propound a largely, often entirely,
qualitative approach comparing the development, during specified periods of their histories, of a
limited number of cases framed in macro-societal terms. Comparative historical analysts do not
reject a statistical approach outright, but they do reject the belief that statistical techniques, and
the large number of cases they require, are necessary prerequisites for the rigorous construction
of modernist knowledge. Finally, and crucially, they also reject the belief that a single, refined
theoretical scheme is necessary, or even helpful, to the pursuit of this goal.
In support of this stance, comparative historical analysts have sought to specify just how their
approach might achieve the rigor that others contend necessitates statistical analysis or a refined
theoretical scheme. Attention focused first on the claim that comparative historical analysts could—
by careful selection of which societies, and which periods of those societies’ histories, to compare
—set up inductive analyses that approximate the logic of the methods of agreement and differ-
ence as laid out in John Stuart Mill’s account of experimental inquiry. More recently, however,
proponents of the approach have come to recognize that this claim has serious problems. The
volume edited by Mahoney and Rueschemeyer (2003a) testifies to a more recent proliferation of
appeals to other techniques—such as Boolean algebra/QCA,^14 fuzzy set logic, Bayesian analysis,
pattern matching, process tracing, and causal narrative—as potential sources of rigor. While shar-
ing a concern to identify techniques to play this role, comparative historical analysts currently
diverge in the details of their responses such that it is problematic to identify the approach today
with any single technique or set of techniques.
Commonalities among comparative historical analysts can be better located in their use of two
sets of conceptual contrasts, which together construct a common identity for their approach out of
a shared sense of what it is not. The rejection of single, refined theoretical schemes offers a
revealing entry point here. This rejection served initially to position comparative historical analy-
sis in contrast to Parsonian functionalism, which offered not only a contending approach to com-
parative social science more generally, but one that also used qualitative, macro-societal,
comparative history. Though that tradition is now quite defunct, comparative historical analysts
have continued to emphasize this rejection, recently turning to it as a basis on which to contrast
their approach with the rising rational choice tradition. While thus emphasizing their rejection of
approaches that rely on a single, refined theoretical scheme, comparative historical analysts also
insist that they are not atheoretical. They conceptualize their approach as occupying a sensible
middle ground between two extremes: overambitious single theories on the one hand, and narrow
empiricism on the other. This perceived middle ground involves “a pluralistic approach to theory
that allows specific research questions and actual historical patterns to help shape the selection of
appropriate analytic frameworks” (Mahoney and Rueschemeyer 2003b, 21). Although we might
debate whether this center holds—in the sense that a philosophically satisfactory explication of it
can be given—such a debate would be beside the point. The common identity of comparative
historical analysts resides not in any extended account and defense of theoretical pluralism, but in
a shared presupposition as to that stance’s integrity and desirability.
A second set of conceptual contrasts cuts across much of the same terrain as the first, but runs
along a somewhat different axis. Rational choice theory is again assigned to a role once filled by
Parsonian functionalism: here, the role of an ambitious universalizing theory claiming to apply at
all times and places. In conceptualizing this role, comparative historical analysts again pair it off
with an opposite extreme and proceed to see their own approach as occupying a sensible middle
ground. The opposite extreme—assigned to interpretive/area studies scholars—is portrayed as a
particularizing approach uninterested in, or even hostile to, generalization in the modernist vein.
The perceived middle ground here is then seen to lie in the pursuit of “mid-range” modernist