Browse Books

Reviewer: H. Van Dyke Parunak

No activity is more characteristic of humans than that of classifying. The world around us is much richer than our limited minds can accommodate directly. Categories are an abstraction of reality, a tractable model that buffers us from the mass of detail that would otherwise overload our processing capabilities. Because classification is ubiquitous and unavoidable, it is mostly invisible. We manipulate categories as though they were the “real world,” often ignorant of the assumptions and history that they impose on our reason. The spread of digital information processing has forced us to become more aware of categories, because computers are even more dependent on a reduced model of the world than we are. At the same time, as we rely more extensively on computers to extend our own cognition, the hidden assumptions embedded in our classification schemes become more insidious. Once instantiated in an information system, a classification becomes much harder to adjust, and its biases are much more difficult to avoid. Bowker and Star offer a sociological analysis of classification. There are many important issues surrounding the various forms of classification (such as Aristotelian vs. prototypical approaches, or alternative logical formalisms for knowledge representation, or the debate between atomic reductionism and “natural kinds”). They touch only on the first of these, and that only cursorily. Their burden is rather to explore the human side of how classifications grow and are used. Their repeated emphasis is that classification is essentially an ethical activity, embodying the deepest intuitions and values of its practitioners, and capable of transferring power and inflicting pain. Their approach is based on four case studies, examples of classification systems with associated social stories. After a lengthy introduction arguing for the importance of the study and an introductory chapter surveying basic methodological issues, Part I (chapters 2-3) studies the International Classification of Diseases (ICD). A wide variety of public health issues require that health workers in different countries be able to agree on causes of morbidity and mortality, and for the past 100 years an international process has produced ten successive editions of the ICD. This historical depth makes the ICD an ideal subject to examine the political side of classification. Part II focuses on the impact of classifications on the lives of people who are classified, by examining two systems. Chapter 5 shows how various categories applied to patients of tuberculosis impact not only their treatment but also their personal lives, while chapter 6 examines apartheid in South Africa. Part III turns to the impact of classification on work practices, focusing on a fairly recent classification scheme, the Nursing Interventions Classification, and its impact on the profession and practice of nursing. Part IV, “The Theory and Practice of Classifications,” draws together the final insights derived from these cases. One of the most valuable is an emphasis on the importance of borderline cases or “monsters.” These entities do not fit neatly into the boxes drawn by a classification scheme, and force attentive users to recognize the arbitrariness and situatedness of their categories. Ultimately, the book offers a rather bleak picture of human intellectual activity. We cannot function without categories, yet by its account the categories we construct are arbitrary, often violated, and reflect our political and social quirks at least as much as they do the realities of the external world. These sobering insights are valuable, but the case studies selected tend to bias the argument. They all are efforts to classify humans and human activities, a domain in which crosstalk between the classifier and the classified is almost unavoidable. It would be interesting if the authors were to apply their sociological tools to some other classification system, such as the physical elements, or organic chemicals, or plants, or stars. Are we really to believe that the periodic table of the elements embodies important ethical distinctions__?__ No doubt, the sociology of the scientists in each field has some impact on these schemes, but one would hope that more objective criteria might have a freer hand in these domains, and perhaps suggest principles that could be used to discipline the development of classification schemes that impinge more directly on humans and their lives. Online Computing Reviews Service