GENERALIZATION IN COMPARATIVE AND HISTORICAL SOCIAL SCIENCE 65tions are, however, quite different. A belief that the contemporary world is marked by an in-
creasingly complex web of global interactions calls for adjustments of approach here also, as I
have earlier suggested with regard to how general “movements” of history are envisioned. But
such shifts would be far from involving the jettisoning of any of the core commitments of an
interpretive social science that employs a comparative and historical stance in pursuit of gen-
eral perspectives.
NOTES
- In characterizing a “genealogical” approach in this fashion, the exemplar I have most in mind is John
Gunnell’s The Descent of Political Theory: The Genealogy of an American Vocation (1993). While recogniz-
ing points of overlap with other proponents of genealogy, such as Foucault (1984), I would emphasize that,
in my understanding, viewing a present approach genealogically does not in itself dictate any specific con-
clusion as to the merits of that approach. One might, in principle, recognize the contingency of an approach,
explore and evaluate alternatives to it, and then choose, at the end of this process, to endorse it. - The phrase “comparative historical analysis” or close equivalents were used before the mid- to late
1970s. It was only then, however, that the phrase became specifically associated with one particular ap-
proach to comparative and historical social science. Previously it was used, when used at all, to refer to any
of several such approaches, or to all such approaches. For examples of earlier usage, see Bendix (1968) and
Flanigan and Fogelman (1971). - For detailed interpretations of most major figures and broad trends within this wave of conversations,
see Bock (1956; 1974), Burrow (1966), and chapter 7 of Collini, Winch, and Burrow (1983). My brief
sketch here is indebted to these scholars while also drawing on my ongoing dissertation work in this area. - For more detail on the dominance and diversity of comparative and historical method among Ameri-
can scholars of politics from the mid-nineteenth up into the early twentieth century, see Farr (2002). - This change within sociology is exemplified and promoted in Park and Burgess’s Introduction to the
Science of Sociology (1921), which was foundational for interwar “Chicago School” sociology. Park and
Burgess see the history of sociology from Comte to their own era in terms of “the transformation of sociol-
ogy from a philosophy of history to a science of society,” the most recent stage of which consists in the new
“period of investigation and research” in which sociologists are “more concerned with social problems than
with social philosophy” (1921, 44). - On the key role of Giddings and his students William Ogburn and F. Stuart Chapin in the rise of an
“objectivist” science of sociology centered on an agenda of statistical analysis, see Bannister (1987). See
also D. Ross (1991, 428–48) on the widening influence of “instrumental positivism” in American sociology
in the first three decades of the twentieth century. - The temporal scope of “modernism,” as I use the term, parallels that of such phrases as “modern art”
or “modernist literature,” rather than the more expansive notions of “modernity” that take the Enlighten-
ment, or even the scientific revolution, as their beginning point. For a wide-ranging exploration of “modern-
ism” in this sense, tracking family resemblances across changes in math and science, through philosophy,
into art and literature, see Everdell (1997). - For his views on Mill, see Giddings (1896, 54, 64–65). In the 1890s, Durkheim made a parallel
departure in his Rules of Sociological Method. Like Giddings, Durkheim moved away from Mill to adopt the
experiment as a logical model for social science. He also specifically took the “method of concomitant
variations” as foundational in this regard (Durkheim 1938 [1895], chapter 6). - Examples of both responses are provided within the tradition of comparative work that continued
throughout the interwar period at Yale. The first response is exemplified by Albert Keller, who completed the
massive synthetic project begun by his mentor William Graham Sumner with the four volumes of the Sci-
ence of Society (Sumner and Keller 1928). The second response is exemplified by George Murdock, who
was Keller’s student and, later, also a faculty member at Yale. Murdock directed the Cross-Cultural Survey
begun in 1937. Conducted under the aegis of Yale’s Institute of Human Relations, the survey was a major
effort to gather, classify, and analyze a worldwide collection of data, which in 1949 became the better-
known Human Relations Area Files project. Murdock’s analyses of the data collected by these projects made
him a pioneer of cross-national statistical research in American social science.