A History of Data Visualization and Graphic Communication

by Michael Friendly and Howard Wainer, Harvard University Press, 2021, 320 pp.
Hardcover, $49.95 ISBN 9780674975231 

Cover: A History of Data Visualization and Graphic Communication in HARDCOVER

Few scientists have done more to write histories of their own discipline than data visualization researchers Michael Friendly and Howard Wainer. The Milestones Project (a multimedia database of visualizations coauthored by Friendly) is one of the most comprehensive historiographical resources on the subject. It collects hundreds of cases in order to create a timeline of visualization “milestones”—that is, the first instantiation of a particular chart type or feature, such as scatterplots or sparklines. Wainer, like Friendly, has written prolifically on the history and practice of graphical methods, and he is the translator of two foundational texts by French cartographer Jacques Bertin, who is widely credited with creating the theoretical foundations of modern visualization. It is fitting, then, that one of the first comprehensive monographs on the history of data visualization comes from practitioners who have made writing this history a crucial component to their scientific contributions.

The authors have taken on the ambitious task of writing whirlwind historical narratives, synthesizing centuries of scientific inquiry within a single chapter. Such is this book’s grand ambition: it chronicles the history of graphic communication since the beginning of human language and reflects decades of expertise from the authors, who were themselves part of a prominent wave of visualization research. Both Friendly and Wainer were graduate students in the 1960s who were mentored by visualization giants John Tukey and Jacques Bertin, and they have since continued their work as statisticians at York University and the Educational Testing Service. 

The book is, therefore, more than a history of data visualization; it is a disciplinary genealogy as recounted by historical actors who participated in its making. The project mirrors the rise of the history of science as an academic discipline, where scientists wrote many of the first monographs themselves. For example, Friendly and Wainer describe four watershed moments in visualization history by writing themselves into the narrative: Wainer’s 1976 Graphic Social Reporting Project from the National Science Foundation, his English translation of The Semiology of Graphics, Edward Tufte’s monographs, and Friendly’s Milestones Project. They also exhaustively document the “firsts” of nearly every statistical concept: the first person to graph data (Michael Florent van Langren, from 1628 to 1644) (30), to invent the idea of “data” as evidence (John Graunt, in 1662) (48–49), or to use a pie chart (William Playfair, in 1789) (101), and so on. While Friendly and Wainer situate the birth of modern graphics within the Scientific Revolution and the Enlightenment, they also take a longer historical view to identify the origins of visual communication.

To manage this feat, the book proceeds chronologically, first setting up the historical origins of data visualization in Neolithic art and Mesopotamian counting systems (chapter 1) and then proceeding toward other key discoveries: the determination of longitude (chapter 2), mapping social statistics with André-Michel Guerry (chapter 3), and statistical developments in the United Kingdom via John Snow’s cholera maps and William Playfair’s line charts (chapters 4–5). The book then crosses from the eighteenth to the nineteenth century, documenting the invention of the scatterplot by John Herschel (chapter 6) and the period from 1860 to 1890, which Friendly and Wainer (echoing Howard Funkhouser) call the “Golden Age of Statistical Graphics” (chapter 7). A key component of this golden age is the changing role of data visualization as a method (and not simply an output) of scientific discovery. The book closes with a discussion of multidimensional visualization (chapter 8), interactive and animated visualizations (chapter 9), and a lyrical ode to “graphs as poetry,” paired with a reimagining of a visual collaboration between W. E. B. Du Bois and C. J. Minard on the Great Migration (chapter 10).

Crucial to their discussion of the golden age, Friendly and Wainer introduce periodization in the history of data visualization (159) by plotting the number of milestones as proof of disciplinary development. The golden age, they argue, is the era where the newest visualizations were published: the more milestones, the more innovation. The focus on the number of new visualization types and on documenting specific individuals as the “inventor of” particular visual concepts underscores the book’s cumulative approach to documenting scientific progress. The book fundamentally asks, How did we get to the visualizations we have today? and answers it with a sparkling, teleological history of a field that canonizes its supposed luminaries, where data visualization researchers such as Tukey and his students become the inheritors of the Golden Age of Statistical Graphics’ visual splendor.

As scholarship in science and technology studies points out (e.g., D’Ignazio and Klein’s Data Feminism), however, this teleological, “Great Men” framework often elides the inequalities and violence embedded in quantification.1 Statistical thinking and its visual translations are also deeply intertwined with eugenic thinking: tools such as significance testing and regression to the mean were developed specifically to illustrate human difference and used to justify violence against marginalized people. Friendly and Wainer’s analysis could be bolstered by identifying and further contextualizing the role of power inherent within data visualization practices, which could inform how practitioners of data visualizations today should think about their use and abuse. Shaded maps of crime are not just historical heat maps; they are Adolphe Quetelet’s and André-Michel Guerry’s contributions to an exploitative system of mass incarceration that disproportionately identifies nonwhite, disabled, and poor people as having a “high propensity” for punishable moral deviance. This is not to disavow the use of heat maps completely. It is, instead, a call toward being critical about both the origins and use of data visualizations both past and present.

Overall, this book provides an indispensable account of how important practitioners of data visualizations write the history of their field. In particular, its detailed digital appendix reproduces many compelling visualizations in full color, making it a valuable resource for academics and hobbyists alike. It is a delight to finally read a monograph from two scholars who have done so much to build this disciplinary history. At the start of A History of Data Visualization and Graphic Communication, Friendly and Wainer cite Aristotle: “If you would understand anything, observe its beginning and its development.” If you would understand data visualization from the perspective of its famed practitioners, observe this book.

Catherine D’Ignazio and Lauren Klein, Data Feminism, MIT Press, 2020.

Crystal Lee, Massachusetts Institute of Technology