In this paper, I draw together W. E. B. Du Bois and corporate bureaucrats to compare the graphical representation of race across three distinct racial epochs: the Progressive, Civil Rights, and post-1980s neoliberal era. I illustrate how, through visual and rhetorical strategies, corporate bureaucrats extend a Du Boisian legacy in constructing popular knowledge of race and racism. I show how they do this by making whiteness visible through data visualizations and rhetorically bundling them to liberal American values of equal opportunity. In examining them as epistemic and semiotic objects, I argue that graphical representations of race compel the enactment of meaningful strategies seen to challenge racial inequalities in the workplace. Yet, insofar as these are employed to equate racism with the absence of equal opportunity in capitalist firms, I argue, they also mask whiteness and reproduce systemic racism. The graphical representation of race, in effect, reveals how the practices of knowledge production and processes of signification are entangled in everyday corporate bureaucracies. Thus, I suggest that we need to reject analytical binaries that pose a bounded distinction between “business” and “social justice” to extend research into the cultural production and productive enactment of racial materiality.
Du Bois; race; racism; data visualization; corporations; equal opportunity
At the turn of the 20th century, W. E. B. Du Bois (1899; 1900) innovatively illustrated demographic data, displaying the socio-economic condition of African Americans in the United States to challenge popular conceptions of racial inferiority. Since then, scholars of color and allies have employed data visualizations to expose the operations of systemic racism in the United States. For example, researchers have written about the impacts of environmental, policing, economic, and housing policies on communities of color by representing race in graphs, maps, and other images (e.g. Benjamin 2019; Bond 2017; Gilmore 2007; Noble 2018; Pulido 2000). While we assume—and to some extent know—that these data visualization practices are effective for creating popular understandings of the complex social, economic, and political phenomena that is racism, we know little of how and why this is the case. We also know less of the cultural role these visualizations play in spaces that are not typically considered activist, such as the corporation.
This paper explores how data visualization is currently “reimagined as a form of accountability and even protest in the age of Black Lives Matter” (Battle-Baptiste and Rusert 2018, 22). Diversity professionals—most of whom are people of color and women and are human resource professionals, diversity officers, business scholars, and management consultants—produce graphs to advocate for increased racial and gender representation among corporate leadership. Since the 1960s, these graphs have been critical components of affirmative action reporting. Using ethnographic and historical data, I draw together W. E. B. Du Bois’s and corporate bureaucrats’ graphical illustration of racial differences in the workplace, comparing these practices across three distinct racial epochs: the Progressive, Civil Rights, and post-1980s neoliberal era. I draw on two years of ethnographic research conducted among diversity professional networks in the United States from 2014 to 2016 and in 2020 at a non-profit organization, American D&I, and at over 50 events, including three national and international diversity conferences. I also draw on archival research, analyzing one company’s affirmative action reports: those of the Polaroid Corporation. I inquire into how diversity “is done” in everyday bureaucratic processes and historically as racial projects (Ahmed and Swan 2006; Omi and Winant 2014).
In examining how graphical representations of race are both epistemic and semiotic objects, I argue that graphical representations of race enable corporate bureaucrats to challenge racial inequalities in the workplace. Yet, insofar as they equate racism with the absence of equal opportunity, they also mask whiteness. By tracing their visual and rhetorical strategies, I illustrate how corporate bureaucrats extend a Du Boisian legacy in constructing popular understandings of race and racism. First, I show how they make some aspects of whiteness—a social, political, and economic phenomenon—visible, which enables them to depict racial hierarchies in the workplace. Second, in rhetorically bundling specific semiotic qualities of graphs with liberal American values of equal opportunity I show how they create new meaning over racial inequality and construct popular understandings of racism. Third, insofar as these graphs are tied to specific cultural values and authorities, I explain these newly meaningful graphical representations of race compel the enactment of institutional practices regarded to address racial inequalities in the workplace.
Social scientists have critiqued diversity management as a field of business expertise that co-opted social justice movements to employ race and gender as sources of capitalist value, govern already-marginalized individuals, and reproduce systemic inequities while appearing “to do good.” (Ahmed 2012; Berrey 2015; Gordon 1995; Melamed 2006; Mohanty 2003). They highlight that diversity professionals tie diversity to business value by arguing that diverse talent helps organizations identify new markets, innovate products, and after the 2008 economic recession, anticipate political and economic crises (e.g., El-Erian 2016; Thomas 1991). Yet, I contend that the graphical representation of race reveals how the practices of knowledge production and processes of signification are entangled in everyday corporate bureaucracies. As such, scholarship that employs analytical binaries of “social justice” and “business” as separately bounded objects are limited for examining how race and racism are constructed, operationalized, and mobilized across time and space by everyday mundane materiality, such as graphs.
Taking a critical perspective, I further argue that employing graphical representations of race in the corporation inadvertently reproduces power-laden relations that are embedded in capitalist firms. I draw attention to one major purpose of these data visualizations, which is to diversify corporate executive leadership. I show how diversity practices, in particular, are driven by cultural assumptions that if biases were eliminated from capitalist labor markets, people of color could achieve compensation and prestige that are normative to the highest-ranking occupational roles in the corporation. Thus, graphical representations of race employed in corporate cultural processes, I suggest naturalize capitalist market ideologies and systemic racial hierarchies embedded in the structures of corporate firms. Therefore, in attempts to inform the practice of using data visualization to address systemic racism in and beyond the workplace, I draw on scholarship on feminist data visualization, namely on Catherine D’Ignazio and Lauren F. Klein (2016), who beyond asking us to reject analytical binaries, also stressed that we should address the needs of collectives (not just individuals) and embrace a diversity (or plurality) of strategies for creating systemic change.
Graphs are a type of data visualization. As a pragmatic means to communicate information, they make expertise public by displaying large amounts of data in a concise, clear, easy and quick-to-read manner. For this reason, they are ubiquitous in scientific literature and in everyday business practice (Tilling 1975). They are what Bruno Latour and Steven Woolgar (1979) termed as “inscriptions,” which represent material objects in a way that can be used by others. Like maps, graphs have “properties of being mobile, but also immutable, presentable, readable, and combinable with one another” (Latour 1990, 26). Immutable mobiles, Bruno Latour (ibid.) wrote, retain a culturally coherent form and content and, hence, are able to circulate across time and distance. Data visualizations also are “epistemic things,” objects with a force that help drive knowledge production (Rheinberger 1997). As social scientists of scientific imaging and visualization have shown, data visualizations are central to the production of scientific knowledge and do more than just represent, they achieve things in the world (Burri and Dumit 2008; Coopmans et al. 2014; Latour 1987; Latour and Woolgar 1979; Lynch and Woolgar 1990).
Social scientists have shown that data visualizations make the invisible visible and that this is cultural work (Aman and Knorr-Cetina 1988; Knorr-Cetina and Amann 1990; Lynch 1985a). Much of this literature has focused on how laboratory scientists produce visuals of biological, chemical, and physical phenomena that cannot be seen with the naked eye (Calvillo 2019; Cambrosio et al. 2005; Hankins 2000; Kaiser 2009; MacKenzie 2013; Vertesi 2014). Scholars have also shown that visibility is entangled with rhetoric and argumentation as visualizations convey ideas that can challenge popular beliefs (Hentschel 2014; Tufte 1997). For example, Alberto Cambrosio, Daniel Jacobi, and Peter Keating (2005) wrote that Linus Pauling used a series of images to convince academic skeptics that the fit between two molecules—“molecular complementariness”—is the basis of life by illustrating the “lock and key” mechanisms for antibody-antigen reactions. Similarly, Klaus Hentschel (2014) explained that 18th century astronomer, William Herschel, used graphs to plot thermometric heat along a visible spectrum that ranged from violet, indigo, blue, green, yellow, orange, and red to invisible lights. Difficult to explain, his visuals depicted the progression of light from visible to invisible spectrums, ultimately convincing his audience of the existence of infrared light. Little attention, however, has been placed on how data visualizations help make social, economic, and political phenomena visible in ways that help construct meaningful and useful concepts.
In constructing meaning over otherwise invisible phenomena, data visualizations also achieve things in the world. Visualizing viruses, for instance, can incite cultural fear and anxiety over infectious diseases, and hence, a desire to combat them (Buiani 2014). Images also stabilize facts, as well as shape and even make human action normative (Burri 2012; Dumit and de Laet 2014; Garnett 2016; Latour 1987; Lynch 1985b). For instance, science and technology scholars have shown that data visualizations drive organizational change and managerial processes (Quattrone, McLean, and Thrift 2013; Thrift 2005). In the study of financial cultures, Daniel Beunza and David Stark (2004) explained that traders turn the trading room into a laboratory, using graphs, colors on screens, computer instruments and their calculations, and their rivals’ “footprints” to calculate whether to buy and sell. Thus, the Dow Industrials and NASDAQ graphs allow stock traders to identify and profit from a “market opportunity” (Beunza and Stark 2004). The role of visual technology in finance is central to how these markets operate; they tame and manage the future (Pryke 2010).
While we know that epistemic objects shape cultural action, studies in the data visualization of race must also account for how race is always-already a meaningful category. Donna Haraway (1997) revealed, for instance, that race is a material-semiotic object of knowledge, whose materiality actively constructs it. For example, craniometric visualizations—which purport to represent data of racial difference acquired through pseudoscientific practices—shaped public, scientific, medical, and legal discourses as centered on conceptions of racial inferiority. Through and beyond Haraway’s material-semiotic analytic, scholars have examined the relationship between the racial materiality of race and systemic racism by drawing attention to the construction and operationalization of race in biomedicine and, in particular, genetics (Rodríguez-Muñiz 2016; Doucet-Battle 2016). Additionally, scholars have shown that social actors do not just subscribe to, but challenge the meanings historically inscribed in material-semiotic objects or racial (semiotic) materiality. Here, I draw attention to Amade M’charek (2013), who showed that racial materiality extends beyond the body, including in criminal report sketches and court documents, and that race is neither fact or fiction, but culturally situated and relationally enacted in practice.
To examine how racial materiality can compel change, I draw on Charles S. Peirce (1958) who wrote that signs and their meaning do not reside in the mind; rather, they are achieved through one’s experience in the world. This is, he wrote, because all signs exist in a triadic relationship between the sign, the object to which the sign refers, and the interpretant, which relates sign and object. Webb Keane (2003) drew productively on Peirce (1958) and anthropologist Nancy Munn (1986) to explain how the analysis of semiotic materiality is productive for understanding culture, as it helps us understand how signs exist in a meaningful and compelling relation to cultural values and practices (a semiotic triad). For instance, Munn (ibid.) originally revealed that qualities or “qualisigns” (a quality that is a sign) such as “lightness” represented in things like the weight of a canoe are valuable because they expand the spatial and temporal reach of other valuable goods, such as bracelets and fish, and that this explains why similar, but different practices that are “light” are also valuable. Keane wrote that signs have causal effects, but not because they determine action; rather because signs “are thoroughly enmeshed with the dynamics of social value and authority” (2003, 415). In examining graphical illustrations of race as epistemic and semiotic objects, I trace how they can compel meaningful institutional changes in intended and unintended ways. In what follows, I show first how by constructing popular understandings of race and racism, W. E. B. Du Bois’s data visualization practices laid the foundation for corporate bureaucrats to advocate for increased workplace racial representation.
For much of his academic trajectory, W. E. B. Du Bois was a prolific Black positivist scholar who believed that unbiased scientific inquiry would dismantle scientific and systemic racism (Allen and Jobson 2016; Hatch 2020; Zuberi 2004). In one of the earliest scholarly uses of statistics and the first social scientific study of race, “The Philadelphia Negro” (1899), Du Bois analyzed data from over 5,000 interviews, providing a comprehensive study of African Americans and Black immigrant income, marriage, literacy, crime rates, religious practices, and employment status in Philadelphia. In the time of Plessy v. Ferguson and the rise of eugenics ideology in the United States and abroad, Du Bois used graphs to create scholarly and public understandings of racism in American society that are still employed today.
In a section entitled “The Occupations of Negroes,” Du Bois displayed employment data in tables spanning eight full pages before stating that the only way to “grasp the true meaning of these figures” was “by comparing the distribution of occupations among the Negroes with that of the total population of the city” (ibid., 108). He visualized this data in bar graphs, depicting in ascending and descending order the proportion of African Americans employed in five occupational groups, compared to that of the whole population (see figure 1). Du Bois paired this graph with one that breaks down this data analytically in terms of gender (see figure 2). He ordered the bars in ways that highlighted differences across race and gender: almost no Black people were employed in professional services and almost all Black women were employed in domestic services. In comparing these two graphs, Du Bois kept a similar scale, but extended the bar displaying the proportion of Black women in domestic and professional services to the right, making it appear to fall off the chart. He explained that given that most Black women are in one industry, they experience more competition, and as a result, lower wages. Du Bois held no reservations to explain these three graphs as telling one coherent narrative, “The causes of this peculiar restriction in employment of Negroes are twofold: first, the lack of training and experience among Negroes; second, the prejudice of the whites.” (ibid., 111).
In these two graphs we see visual and rhetorical strategies that Du Bois and his associates would continue to use in the 1900 Paris Exposition: scaling to make the smallest of African American populations in professional fields visible, categorizing, coloring and ordering to highlight differences across race, as well as rhetorically bundling data visualizations with equal opportunity values. Such strategies would help drive forward new understandings of race and racism. For instance, in this case Du Bois’s audience was academic, white, and may have mostly subscribed to popular scientific racism. For him, these graphical representations of race, and more specifically its features, such as differences in the length of the bars, were symbolic of the inequality that he experienced in the world as a Black man (that he would write about his entire life), and representative of the inequality expressed in observations and interviews with Philadelphia African Americans.
Du Bois challenged discourses of Western superiority and possible interpretations of such graphs as evidence of African inferiority by illustrating that Black people were capable of holding positions in professional fields. For instance, with regards to why there were few African Americans in trade unions, he wrote,
In sections where Negro labor in certain trades is competent and considerable, the trade union welcomes them, as in Western Pennsylvania among miners and iron-workers, and in Philadelphia among cigarmakers; but whenever there is a trade where good Negro workmen are comparatively scarce each union steadfastly refuses to admit Negroes, and relies on color prejudice to keep up the barrier (ibid., 128).
He explained to his white audience that the absence of African Americans in manufacturing was a matter of economics and resulting prejudice. A lack of training, he wrote, drives down the cost of labor and fuels conflict across white union workers and Black non-union workers. Du Bois, thus, taught his academic audiences a nuanced understanding of systemic racism, writing “a combination of social problems is far more than a matter of mere addition,—the combination itself is a problem” (ibid., 385).
At the 1900 Paris Exposition, Du Bois foregrounded a colorful graphical display of social analysis, but this time to (re)educate a white popular audience who had attended the world fair to witness the “great achievements” of European civilizations. In a collaboration with students and faculty from Atlanta University, including Booker T. Washington and Thomas J. Calloway, these scholars produced over sixty graphs employing data from a social scientific study conducted in Georgia. In these data visualizations, we see similar strategies as those employed in the Philadelphia study. In “Occupations of Georgia Negroes: Males over 10,” Du Bois’s team used a bar graph to display the number of African Americans employed in various occupations, organizing the bars in descending order from top to bottom. This illustrated that most African Americans were employed in agriculture and labor, industries associated with slave labor in the South (see figure 3). Du Bois curved the largest bar to conform it to the page, making visible that African Americans occupied some of the most prestigious occupations in Georgia.
Similarly, this last employment-focused visual compares Black and white employment in Georgia (see figure 4). This comparison challenged understandings that only Black men worked in this industry, while highlighting that African Americans were capable but obtained fewer jobs in professions requiring higher education. Employing vivid colors, Du Bois and his team made their graphs attractive and used a color code to highlight differences across white and Black groups. Visibility is important, he said, reflecting on the goals of their exhibit: “It was a Massachusetts lawyer who replied to the Patent Office inquiry, ‘I never knew a negro to invent anything but lies;’ and yet here is a record of 350 patents granted to Black men since . . . [African Americans are] here shown to be studying, examining, and thinking of their own progress and prospects” (Du Bois 1900). In addition to these data visualizations, Du Bois and his team exhibited dozens of photographs, over 400 original patents held by African Americans, and more than 200 books written by African American authors.