Sorting Things Out: Classification and its Consequences. Geoffrey C. Bowker and Susan Leigh Star. Cambridge, MA: The MIT Press; 1999; pp. 377; $29.95 (ISBN 0-262-02461-6.)
Librarians build classifications to sort things out. Linguists use it to study how people name things. Cognitive psychologists have measured how concepts fit in categories. These three, formerly disparate areas, collide in modern information systems that attempt, for example, to organize web pages in categories that the general public will find useful. It is the accomplishment of this book to recognize classification itself as an object of study, as a vehicle for ethnography and as a public object of hidden sociological, linguistic, cultural and enterprise assumptions. This book is destined to become a primary source for “users studies” and “knowledge architecture” in the field of information science.
This book establishes a broad area of study by presenting several detailed examinations of classifications of disease, patients, nursing work and race. The authors establish numerous avenues of typical critique that future scholars will follow. Examples are the origins of categories, change in category definitions, the social meaning of categories, category definition as an artifact of technology and so on. There are a number of extraordinary examples of the enormous human impacts of classification systems such as South African citizens who have been tossed back and forth between being classified as Whites, then Coloreds, then back again. That classification is embedded in our political, social, linguistic and cognitive lives there can be no doubt. The effect of their sample critiques and examples is to illustrate how values, policies and modes of practice become embedded in large information systems and become expressed in classification systems. It is the virtue of the book to force us to recognize classifications and their effects that we may be only vaguely aware of as background.
The survey treatment , however, leads to multiple examples that pile on and do not advance the argument. From the first sentence and through to the last, the authors establish that classifications are human artifacts, and few readers would disagree. The research challenge is to take the next step and establish certain classification characteristics are associated with certain human, linguistic or institutional characteristics.
For example, to describe the Nursing Intervention Classification as an elegant ethnographic tool does not help us take the next step of strategizing how it may be used ethnographically. On the other hand, the authors do give the reader several useful guidelines for the analysis of any large infrastructural system (p. 159), three main areas of challenge in crafting a classification scheme (p. 231), and the several key properties of an infrastructure (p. 237). These are specific suggestions that will engender empirical research and comparative ethnographic studies.
The audience for this book would appear to be the library and information science community, folks who will happily receive a text announcing the new and important context for classification studies. This book is, however, strangely disconnected from a large body of empircal, cognitive research in classification and categorization that already exists. Consider these three quotations from Knowledge, Concepts and Categories, edited by Koen Lamberts and David Shanks (and ironically also published by The MIT Press). Each is found in the introductory paragraph:
"What a person learns about a new category is greatly influenced by and dependent on what this person knows about other, related categories." (Evan Heit, "Knowledge and Concept Learning", p. 7)
"The cognitive system does not treat each new object or occurrence as distinct from and unrelated to what it has seen before: it classifies new objects in terms of concepts which group the new object together with others which have previously been encountered." (Ulrike Hahn and Nick Chater, "Concepts and Similarity", p. 43)
"Even informal observation of everyday categorization reveals that many objects fit into a number of categories." (Gregory L. Murphy and Mary E. Lassaline, "Hierarchical Structure in Concepts and the Basic Level of Categorization", p. 93)
How strange that the editors of the MIT press didn't alert Bowker and Star about the book they had already published by Lamberts and Shanks. And imagine how much more rich would be the analysis of Sorting Things Out would have been if it had benefitted from the empirical research that already exists.
Terrence A. Brooks
School of Library and Information Science
University of Washington
Lamberts, K. and Shanks, D., eds. (1997). Knowledge, Concepts and Categories. Cambridge, MA: The MIT Press.