Data Visualization and Defunct Visual Metaphors
This dossier concerns the relationship between visualization as a mode of knowledge-production and vision as a (visual) metaphor for knowledge. It looks at the history of data visualization and its relationship to epistemology, particularly renderings of statistical data. Key moments in the history of statistics and data visualization are discussed, with a particular focus on so-called political arithmetic and the decline and re-emergence of visualization techniques as an effect of complex quantification methods and computing, respectively. The dossier advances the argument that visual metaphors are fundamentally inadequate for the contemporary forms of data acquisition, analysis, and knowledge-production that are endemic to the informatic mode. The persistence of visual metaphors for surveillance is considered alongside the increased incidence of opaque data capture and mining, practices which do not lend themselves to familiar metaphors or renderings. The dossier concludes with a critical discussion of the use of data visualization techniques in artistic interventions which seek to make this process more legible.
Visualization and Cognition
A discussion of the complex relationship between vision and epistemology may well begin with Plato, as does Martin Jay's influential book on the treatment of vision in twentieth century French theory, but this dossier takes a different path. It instead follows the unique methodology developed by Bruno Latour in his discussion of "Visualization and Cognition," wherein he argues for the careful consideration of representational techniques in the production of scientific facts. His concern, like that of this dossier, is with the way in which inscription practices both make legible their objects of study and how these same inscriptions are then discursively marshaled in support of specific scientific claims. In other words, Latour examines "the transformation of rats and chemicals into paper," (3) as "[n]othing can be said about the rats, but a great deal can be said about the figures [which reductively represent them]" (15). Latour is careful to acknowledge that such claims teeter on the brink of triviality, but responds by insisting that it is upon the basis of representational efficacy and economy that truth claims obtain their status. Inscriptions should not be dismissed as mere visual aids; they provide the necessary grammar for the representational transposition of objects or phenomena into a semiotic system. Whatever form this may take, the meaningful specification and analysis of objects or phenomena must pass through a process of rendering. As Latour notes: "If scientists were looking at nature, at economics, at stars, at organs, they would not see anything […] Scientists start seeing something once they stop looking at nature and look exclusively and obsessively at prints and flat inscriptions. In the debates about perception, what is always forgotten is this simple drift from watching confusing three-dimensional objects, to inspecting two-dimensional images which have been made less confusing" (15). For Latour, science happens in the specific and consistent rendering of perceptual information on a material substrate. The phrase "data visualization" would therefore seem redundant in this context: visualization is the production of data. The visualization of economic trends by way of mathematical or graphical representation is what makes economics, from an ontological perspective, a possible object of study. This is not to deny that economic activity happens, but rather to suggest that any meaningful understanding of economics as a measurable and data-producing activity must always begin with inscription and visualization. Visualization is always already a way of understanding.
Latour extends this insight into discussions of the unique relationship between science and bureaucratic technologies like paperwork and record-keeping. He examines the degree to which visualization techniques achieve popular support and thus, in turn, confer increased legitimacy on the arguments to which they are attached. Visualization, he argues, is key to understanding the mobilization of consensus. And it is here where data visualization may no longer seem like a redundant expression: one form of inscriptions (say, numerical tallies, measurements, formulae, etc.) can effectively feed another (say, a graph). Of course, the reverse is true, too: graphical inscriptions produced by way of mechanical instruments can be the basis upon which numerical data-collection occurs. The rise of statistical thinking, an expression coined by Theodore Porter, is therefore concomitant with the emergence of specific forms of visualization. Visualization brackets statistical reasoning on both sides: the visualization of objects or phenomena by way of specific inscriptions which lend themselves to mathematical reductionism, and which, in turn, for reasons of economy, clarity, or analysis, may be once again rendered graphically. For example, in Christopher Scheiner's Tres Epistolae to the right, which visually charts the change in sunspots over time, visualization produces the data that others would later evaluate mathematically. Below that, William Playfair's well-cited chart plots the historical change in wages, price of wheat, and reign of specific monarchs. Statistics elsewhere obtained (and by way of yet another form of inscription) are here rendered anew in graphic form. Playfair, to whom many of the most familiar forms of contemporary data visualization are still ascribed, understood the unique power of charts thusly:
"The advantage proposed, by this method, is not that of giving a more accurate statement than by figures, but it is to give a more simple and permanent idea of the gradual progress and comparative amounts, at different periods, by presenting to the eye a figure, the proportions of which correspond with the amount of the sums intended to be expressed. As the eye is the best judge of proportion, being able to estimate it with more quickness and accuracy than any other of our organs, it follows, that wherever relative quantities are in question […] this mode of representing it is peculiarly applicable […] In a numerical table there are as many distinct ideas given, and to be remembered, as there are sums, the order and progression, therefore of those sums are also to be recollected by another effort of memory, while this mode unites proportion, progression, and quantity, all under one simple impression of vision, and consequently one act of memory" (ix-x).
Of particular note is the degree to which Playfair privileges vision at the same time that he acknowledges the fundamental commensurability of statistical figures and their derivative graphic representation. Playfair would later describe this advantage as the ability to "speak to the eyes." Notwithstanding its peculiar use of an auditory metaphor (which conceivably refers to the presumed passivity or oral comprehension), Playfair's visualization techniques ushered in a host of transformative practices that buttressed the burgeoning field of so-called political arithmetic. No doubt Playfiar's chart possessed a didactic quality. This chart, in particular, accompanied a text which argued for the apparent improved conditions of worker in recent years. Graphic literary, however, remained a concern. Apparently, one had to learn how to hear with one's eyes. As Michael Friendly points out, Playfair's charts were so novel that he was forced to devote several pages of The Commercial and Political Atlas and Statistical Breviary to explaining how one should go about interpreting a line graph. Graphic vision was an acquired skill. Once acquired, however, it radically increased the fluency with which the graphically literate could engage in comparative thinking. But this was a visual competency once removed from the key transformation in sight: "The implicit project of the construction and appropriation of social reality in terms, perhaps not of mathematics, but nonetheless of calculation and classificatory and systematic analysis was conceptualized in the notion of the 'statistical gaze': de statistische Blick" (Hans Erich Bödeker, 176). The ability to see statistically preceded the statistical mode of inscription which in turn preceded the capacity to interpret the visual rendering of statistics.
Vision as Visual Metaphor
In a brief essay entitled "The All-Seer: God's Eye as Proto-Surveillance," Astrit Schmidt-Burkhardt argues that the semantic multi-dimensionality of vision can be understood by examining the etymological connection between the French voir (vision), savoir (knowledge), and pouvoir (power). Schmidt-Burkhardt suggests that vision never achieves a fixed semantic position within this network of meaning. His genealogy of vision as a visual metaphor speaks to this insight. His history begins on the cusp of antiquity, a period in which the idea that gods had eyes or were indeed themselves eyes had already achieved currency. "Divine symbolism should be seen as a relic of hieroglyphic knowledge that had also Greek and Roman authors," he argues. Although Christianity inherits this symbolism, no recorded instances of the figure of the eye occurs in religious art prior to the the sixteenth century.
Eye of Providence
Data visualizations as the new metaphor?
Downcast Eyes: The Denigration of Vision in Twentieth-Century French