Sorting Things Out: Classification and Its Consequences (review)

Sorting Things Out: Classification and Its Consequences. By Geoffrey C. Bowker and Susan Leigh Star. Cambridge, Mass.: MIT Press, 1999. Pp. xii+377. $29.95.

In Sorting Things Out, Geoffrey Bowker and Susan Leigh Star shed new light on a variety of classification systems and practices. The thrust of this eclectic and wide-ranging study is twofold: (1) classification is a ubiquitous human activity; and (2) the ramifications of classification both as part of the information infrastructure and as a social practice affecting human lives deserve much greater attention than they have previously received. This wide-ranging study then instantiates these claims by closely examining instances of classification schemes, such as the International Classification of Diseases (ICD), the Nursing Intervention Classification (NIC), and the system of racial classification under apartheid in the former South Africa. In these studies the authors focus on both cultural and technical aspects of classification, especially the negotiations and implications affecting such fundamental social constructs as identity, disease, race, and work.

Like Gaul, Sorting Things Out is divided into three parts, and indeed they are conquered in quite different ways. Most of the chapters have been published in earlier versions and, perhaps as a result, they do not mesh particularly well. As a whole, the book is a bumpy ride. The three divisions, each encompassing two or three chapters, are devoted in turn to the construction and use of classifications (especially the ICD); the way in which practices and systems of classifications "torque" biographies, that is, how they change lives; and how classification systems "organize and are organized by work practice" (especially the NIC). Several topics one might expect in a defining work on classification are lacking in this breakout, such as systems of library and archival classification, the encyclopedic tradition, bibliographic indexing, and security classification schemes. On the whole, such omissions represent readers' expectations more than the authors' intentions, which appear to be oriented toward practices that operate within the structures of medical and social classification rather than the intellectual problems of creating formal classifications of information or knowledge. On this matter of the power of classification to shape self-identity, influence work, respond to politics, and in many other ways influence people, Sorting Things Out is unique in its painstaking attention to details and observation of multifaceted causes and effects.

Ultimately, this book delivers two messages. The first is a strong programmatic statement; all aspects of our daily life are influenced by classification practices (or their close cousins, such as standards) and, ipso facto, it is crucial that we strive for a better grasp of the epistemological, political, and normative implications of these practices. The first and last chapters concentrate on this message, supplemented by a rather unconvincing attempt in chapter 9 to wrestle an overarching theory of classification to the ground. A problem for both the manifesto ("To classify is Human," p. 1) and the theory is that it is difficult to define classification generally in a way that pulls the book together; the authors define it as "a spatial, temporal, or spatio-temporal segmentation of the world" with three properties: (1) consistency of classification principles in practice; (2) mutually exclusive categories; and (3) completeness (pp. 10-11). Not only is it difficult (as the authors acknowledge) to find a system that actually meets these requirements, but the eclectic organization of the specific case studies in Sorting Things Out counters the spirit of such apparent rigor. In other words, while the authors write convincingly of the need to foreground classification practices that are not always visible, they are less successful in reducing these practices to a coherent notion of classification itself.

Once or twice in Sorting Things Out Bowker and Star muse on the difficulty of classifying their own book on classification. They locate it at the "crossroads of the sociology of knowledge and technology, history, and information science" (p. 6). The Library of Congress removes this ambiguity and classifies it...