Fuzzy-Set Social Science (review)

Fuzzy-Set Social Science contains a bid to quantitatively oriented comparative social scientists to heed the insights of qualitative case-study research, without abandoning their distinctive concern for generalization. The contribution of the book lies in the elaboration of a set of logical-analytical techniques to guide the researcher in such a task. More specifically, Ragin extends the principles of Boolean algebra or settheoretic relations to the comparative study of social phenomena.

Ragin's program for diversity-oriented research using fuzzy sets begins from a critique of conventional quantitative comparative analysis and an appreciation of certain features of the case-study approach. Ragin criticizes the tendency of conventional analyses to fit data in a priori homogeneous categories and the additive-linear view of causation that are associated with statistics. In their place, he offers a set of analytical tools designed to explore the diversity of populations and causal complexity. This re-orientation is associated with a rejection of the logical-empiricist focus on the logic of justification and the associated disjunction between theory and empirical analysis. In contrast, Ragin's tools are designed to guide the logic of discovery (rather than the logic of justification) and to further an ongoing dialogue between theory and research. Stated thus, the project resonates with numerous European calls for a reflexive methodology.¹ Whereas Continental scholars tend to focus on the interplay between the researchers' categories and those found in the data, Ragin focuses on the problem of how to operationalize concepts.²

The book is divided into two parts: a discussion of diversity-oriented research and the application of fuzzy sets to that program. Diversity-oriented research focuses on the limited variety of types contained within a given population and the complexity of causal relations. The aim of the research is to produce a conceptual map showing the types of revolutions, protests, welfare states, or other phenomena under investigation and the variety of routes by which they were produced. Social phenomena are defined by a configuration of aspects and features, and cases are analyzed for their membership in those configurations, such that a change in any one aspect changes the classification of the case.

Fuzzy sets provide a means to allow for partial membership in a population or configuration. Not all democracies are equally democratic, just as not all Protestants are equally Protestant. The construction of a fuzzy set involves the quantitative evaluation of degree of membership and the qualitative specification of cut-off points signifying full membership and non-membership (the points beyond which more or less membership ceases to be significant for the outcome). By forcing the researcher to operationalize the concept of membership in a population and to adjust the concepts to the data, fuzzy sets further theory construction as well as empirical analysis and ensure a closer fit between the two. Once constructed, fuzzy sets can be subject to a series of logical operations leading to the simplest possible sentence concerning the causal configurations associated with a particular outcome.

As this brief review indicates, Fuzzy-Set Social Science is explicitly written for social scientists interested in producing parsimonious, logically tight, general sentences concerning types of social phenomena. As Ragin explains, it offers a language that is partly verbal-conceptual and partly mathematical-analytic. This vocabulary may not appeal to most historians, but it does serve to render explicit the range of assumptions contained in specifically situational forms of explanation.³ As such, it should be of interest to anyone concerned with problems of historical explanation.