The Visual Display of Quantitative Information by Edward R. Tufte

Abstract

TECHNOLOGY AND CULTURE Book Reviews 175 and decentralized local planning to provide essential social and eco­ nomic support for individuals, families, and communities. The early sections of chapter 4 (“Redefining Public and Private Spheres”) contain some of Frankel’s most specific comments about technology and culture and display typical strengths and weaknesses of his book. He praises Toffler for paying attention to phenomena often ignored by the Old Left, especially “electronic cottages”: homebased businesses spawned by computer and telecommunications rev­ olutions. He cautions, however, that the new technologies and non­ factory work environments will not necessarily lead to better lives for workers. On the contrary, isolated home workers may lose a sense of solidarity, a shared feeling which for Frankel is still a relevant heritage of the Old Left. Occupational gender discrimination could also be perpetuated, since lower-status work in word and data processing are already becoming woman’s work in office and home industries. To overcome the potential loss ofsolidarity and to prevent discrimination, state institutions will have to provide support systems (e.g., child care) and enforce egalitarian regulations. Frankel’s discussion of the electronic cottage demonstrates how he can constructively synthesize Old Left and postindustrial utopian ideas. Nonetheless, his examination is too brief (pp. 149—54) and quite vague. Had he presented even one example (Kerista, a San Francisco com­ mune, would have been an excellent choice), he could have moved beyond introductory generalities to more intricate analyses of rela­ tionships between home computer technologies and gender/family structures. Perhaps this criticism is unfair. Frankel does not strive for intricate analyses. Instead he hopes to raise numerous questions about nu­ merous topics. By doing so he has succeeded in compiling a catalog of challenges that should encourage Old Leftists and recent propo­ nents of alternative technologies and life-styles to learn from each other and to articulate useful and imaginative critiques of the present and models of better worlds. Kenneth M. Roemer Dr. Roemer, professor of English at the University of Texas at Arlington, is the author of The Obsolete Necessity (1976) and Build Your Own Utopia (1981), and the editor of America as Utopia (1981) and Approaches to Teaching Momaday's Way to Rainy Mountain (1988). The Visual Display of Quantitative Information. By Edward R. Tufte. Cheshire, Conn.: Graphics Press (Box 430 06410), 1983. Pp. 197; illustrations, index. $34.00. Historians who watch their numbers should read this book. It is a concise history of the use of graphs to display information and an 176 Book Reviews TECHNOLOGY AND CULTURE incisive discussion of effective graphical methods. The graphs in­ cluded are well chosen, and their incorporation into the text is exemplary. Edward Tufte begins by describing common techniques—data maps, like Edmund Halley’s 17th-century representation of the trade winds; time series, like J. H. Lambert’s plot of seasonal changes in ground temperature; bar graphs, like William Playfair’s chart of Scotland’s imports and exports; and relational graphics, like Playfair’s display of tax rates and population. I wish he bad suggested how Boulton and Watt’s development of the indicator diagram fits into this context. The book then focuses on graphics that lie or distort, misrepresenting the size of a quantity, comparing quantities (like dollar values) without using standardized units, or giving only a few data points. Tufte has harsh words for illustrators, untrained in data analysis and convinced that statistics are boring, who use graphics mainly to catch the eye of readers whom they assume are unsophisticated. The second half of the book advances a theory of data graphics. Tufte begins from the principle that graphs should, above all, show data. Ink devoted to other purposes—to an underlying grid, to deco­ rative patterns, and even to creating symmetry (as in a bar graph) should be eliminated, if possible. The resulting graphs look unfa­ miliar, but are generally easy to interpret. Tufte’s book merits the attention of historians not only because he presents useful principles clearly and concisely but because he illus­ trates them so well. Of special interest are his examples of graphs showing several variables simultaneously. For instance, Charles J. Minard showed in one graphic the path...