Review of Drucker, Graphesis

Review: Johanna Drucker, Graphesis: Visual Forms of Knowledge Production (metaLABprojects; Cambridge, MA: Harvard University Press, 2014).

Johanna Drucker’s Graphesis, at first glance, seems to be a straightforward history of data visualizations. The vast majority of the book is devoted to tracing the histories of various kinds of information visualization, such as tree graphs, maps, bar charts, and the like. Scores of illustrations accompany this discussion, making the book a fine introduction to the history of information visualization. Behind the historical aspect of the book, however, lies the assertion—actually Drucker’s main thesis—that humanists have fundamentally misunderstood what data is and what visualizations can represent.

The book opens with a foreword defining some of Drucker’s key terms (including graphesis, to which I will return shortly). The first three chapters (“Image, Interpretation, and Interface,” “Windows,” and “Interpreting Visualization :: Visualizing Interpretation”) set forth the history of the graphical forms that lie at the heart of modern information visualization, along the way critiquing the way contemporary humanists have used them. The fourth chapter (“Interface and Interpretation”) examines the computer interface as a constructed system that mediates between the user and the computer, rather than directly reflecting the computer’s underlying processes. The final chapter (“Defining Graphic Interpretation”) is a brief guide to humanistic data visualization done well. The book ends with a short afterword predicting the nature of humanistic media in the future.

It has been implied in another review (McLemee 2014) that Drucker does not define the term graphesis in her book. That review is incorrect. Drucker defines graphesis as “the study of the visual production of knowledge” (4). She has also defined it as “knowledge manifest in visual and graphic form.”[i] The best explanation, however—one which includes several enlightening qualifications—comes from Drucker’s 2011 essay “Graphesis: Visual Knowledge Production and Representation”:

Graphesis is defined as the field of knowledge production embodied in visual expressions. This seems straightforward enough. But the range of such expressions is enormous, and defining the principles of a stable symbolic system seems daunting. The term graphical includes specialized writing and notation, codes and symbols. It might also embrace visual art and design. I mean the term to suggest visual expressions that are arrangements of marks or visual forms organized to read on and as a flat surface (in other words, in their literal, visible form, rather than as pictorial illusions). But I intend the term to suggest a more fundamental ground on which to begin to examine the ways all visual expressions work — whether they are forms of writing, pictorial imagery, information graphics, or other images — by virtue of being marks organized on a flat surface. Graphic artifacts present knowledge through the combination of symbolic codes and structured relations of these elements in a flat field. My basic aim is to create a critical framework within which the forms that are generally used for the presentation of information can be understood and read as culturally coded expressions of knowledge with their own epistemological assumptions and historical lineage. A general theory of graphesis addresses the organizing principles of all images for the ways they encode knowledge through visual structures and rhetorics of representation.[ii]

As noted above, the book has two goals, both related to Drucker’s understanding of her term. On one level, Drucker gives a fairly straightforward history of information visualization. On another level, though, Drucker is concerned to critique the contemporary humanist’s infatuation with quantitative information. Her refrain throughout the book is that “data are capta, taken not given, constructed as an interpretation of the phenomenal world, not inherent in it” (128; emphasis original). She argues forcefully and persuasively that contemporary humanists have misunderstood the nature of data qua capta, leading them to “[collapse] the critical distance between the phenomenal world and its interpretation, undoing the concept of interpretation on which humanistic knowledge production is based” (125) and resulting, in some cases, in “an interpretative warp or skew, so that what we see and read is actually a reification of misinformation” (105). This point alone makes the book worth the price of admission for the digital humanist, and constitutes a valuable critique of the way many humanists use data and data visualizations.

Related to this argument, Drucker calls for humanistic data visualizations to incorporate visual representations of “interpreted phenomena” like “point of view, position, the place from which and agenda according to which parameterization occurs” (133). As an example, “one might imagine skittish points on an unstable grid to display the degrees of anxiety around a particular event or task, for instance, or points that grow hot or cold depending on the other elements that approach them” (134). Drucker’s point is well taken: an overreliance on quantitative visualizations masks the ambiguities of the world the visualizations purport to represent. I wonder, though, exactly how much ambiguity can be included in a visualization before it ceases to be an effective tool of communication, instead becoming just an interesting-looking picture. A better approach, I think, would be to adopt the model of the infographic: emphasizing the legibility of the visualization proper, while at the same time contextualizing and nuancing it through captions and other kinds of peritexts.[iii] This approach allows visualizations to do what we expect them to do—communicate quantitative information in a format more readily understandable to us than a list or table—while not overburdening them with qualifications, which, in our context, are easier to understand when expressed in words.[iv]

The final chapter of the book is most useful for the humanist who is convinced by Drucker’s argument. In this chapter, she gives examples of what she sees as properly nuanced humanistic visualizations and describes some of the traits that make them useful for humanistic research. My only complaint about this chapter is that it feels far too short; after 170-some pages of apophasis, this (12-page) chapter seems like an afterthought. One can only hope that Drucker will write a follow-up, giving humanists guidelines to follow as they put together visualizations.

In all, Graphesis makes a strong, convincing case that humanists need to re-examine and, in many cases, re-formulate, their relationships with data. As such, it serves as a much-needed corrective for contemporary humanistic research.



[i] Drucker 2001: 145.

[ii] Drucker 2011: 3. Drucker seems to have coined the term independently from Marie-Rose Logan, who, in the editorial foreword to a 1975 edition of Yale French Studies, came up with it

to balance the term “grapheme” (the written object) used with different implications by Derrida as well as by Wimsatt [i.e., in Derrida’s and Wimsatt’s articles in the same volume]. “A believable neologism,” “graphesis” stands as an operatory process through which “writing” actualizes itself in a (written) text. The attempt is thus not to formulate a notion which would demand a “defense and illustration” but rather to grasp the nodal point of the articulation of a text. “Graphesis” de-limits the locus where the question of writing is raised, whether on the so-called creative, philosophical, or critical level. “Graphesis” de-scribes the action of writing as it actualizes itself within the text independently of the notion of intentionality. Paralleling “deconstruction” as an operatory process, it includes the deconstructive gesture. As an abstract notion, it cannot be validated without the grapheme (for they are the recto and verso of the same concept), but, through its dynamics, graphesis would preserve the grapheme from theoretical dessication (Logan 1975: 12-13).

[iii] I am indebted to Cairo 2012 for my views on data visualizations and infographics.

Drucker acknowledges that constructivist visualization can decrease a visualization’s intelligibility, but argues that it is a sacrifice worth making. She writes, “Creating bar charts with ambiguity and degrees of uncertainty or other variables in them might cause champions of legibility and transparency some unease, but the shift away from standard metrics to metrics that express interpretation is an essential move for humanists and/or constructivists across disciplines” (130).

[iv] I should add that I am not arguing for a total excision of nuance from humanistic visualizations, but rather a middle road between, say, the (unnuanced) Excel pivot table and the (hypernuanced) unstable grid that Drucker describes.



Alberto Cairo, The Functional Art: An Introduction to Information Graphics and Visualization (Berkeley, CA: New Riders, 2012).

Johanna Drucker, “Digital Ontologies: The Ideality of Form in/and Code Storage: Or: Can Graphesis Challenge Mathesis?,” Leonardo 34 (2001): 141-145.

Johanna Drucker, “Graphesis: Visual Knowledge Production and Representation,” 2011(?), accessed November 18, 2014,

Marie-Rose Logan, “Graphesis…,” Yale French Studies 52 (1975): 4-15.

Scott McLemee, “The Eyes Have It” (review of Johanna Drucker, Graphesis: Visual Forms of Knowledge Production), Inside Higher Ed, September 3, 2014, accessed November 18, 2014,


1 Comment

Filed under Uncategorized

One response to “Review of Drucker, Graphesis

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s