Social History: Problems, Strategies, and Methods (review)

Social History: Problems, Strategies and Methods. By Miles Fairburn (New York, St. Martin's Press, 1999) 325 pp. $59.95 cloth $21.95 paper

New Zealand social historian Fairburn's textbook on research design and argumentation should be widely adopted for graduate classes in history. Drawing on the philosophy of science, Fairburn questions the widespread practices in social history of, for instance, generalizing from isolated instances, unsystematically assessing differences and similarities between and within various groups, and evaluating interpretations on mainly stylistic or ideological grounds, instead of on the basis of the logic and power of their models. Although he persuasively criticizes the cultural relativism of postmodernist and hermeneutic approaches as self-refuting, he is better, and more comfortable, discussing such "soft" stances than he is discussing economics and other "hard" social sciences--which he virtually ignores, despite economists' numerous contributions to social history--or statistical methods--which he distrusts when authors cannot completely explain them in simple terms to skeptical innumerates.

Fairburn's chapters about seven "standard methodological problems" (leaving out certain groups, fragmentary evidence, causality, differences and similarities, socially constructed data, anachronism, and hypothesis testing) and his examinations of five "modes of enquiry" (hermeneutic, intentional, causal, "how possible?" and focused information gathering) are grounded in acute discussions of books by such major social historians as Edward P. Thompson, Fernand Braudel, David Hackett Fischer, Stephan Thernstrom, and Daniel Goldhagen. Graduate students will appreciate these capsule summaries and critiques--helpful on qualifying exams: These synopses could also stimulate discussion in classes, especially if assigned in combination with the original works themselves. Historians pay too little attention to the logical structures of arguments that they read and write. Fairburn's penetrating analyses, if widely emulated, would heighten professional self-consciousness.

The book's best discussions, such as that on fragmentary evidence, not only demonstrate how to assess completed historical works but also offer rules that may help historians analyze primary evidence better. Fairburn rejects reliance on contemporary "experts," who may know only parts of the truth, and appeals to "common sense," which may be culture-bound. Instead, he suggests attacking the problem of fragmentary evidence by concentrating on cases that are inherently biased against a particular hypothesis, comparing similar, well-documented cases, eliminating rival hypotheses, and fully elaborating the theory and auxiliary hypotheses so as to suggest a wide range of evidence that would strengthen or weaken the conclusions. His strictures are anchored in a particularly trenchant analysis of Stone's Crisis of the Aristocracy. ¹

When Fairburn turns from qualitative to quantitative history, however, he is reduced from calm adept to panicky dilettante. Counting, beyond the computation of averages, he asserts, is now "practically taboo" in social history, apparently because during the "mass quantification craze of the 1970s," cliometricians made "unscholarly and slipshod use of sources" and arbitrary, unexplained decisions about the appropriate categories for people, failed to explain their methods, and reached "extravagant" conclusions (148-151). In fact, plenty of sophisticated counting still occurs in social history, though most of it takes place in economics or sociology departments; cliometricians generally agonize over the meaning and validity of sources, how to categorize people, and how to specify models; useful statistical methods are often complicated; and it is inconsistent for one who, like Fairburn, praises "bold conjectures" to denounce them.

J. Morgan Kousser
California Institute of Technology

Note

1. Lawrence Stone, The Crisis of the Aristocracy, 1558-1641 (Oxford, 1965).

...