Visualizing Impossibility: Thoughts on Lauren Klein

In Lauren Klein’s “The Image of Absence: Archival Silence, Data Visualization, and James Hemings,” we search alongside her for ghosts, silences, and absences in the archive. Over the course of the article, she seeks to illuminate the life and contributions of James Hemings within the Papers of Thomas Jefferson, a digital archive made available through ROTUNDA, University of Virginia Press, and in doing so, discusses the possibilities and pitfalls of data visualization in this process. For Klein, digital technology has the capacity to render visible the invisibilities of archival gaps, and at the same time expose the limits of our knowledge as productive space with which to think.

Recalling last week’s conversation about narrative and database, Klein suggests that archival silences can be produced, in part, by metadata and data structuring decisions (663). This claim dovetails with Lisa Brundage’s suggestion that the most essential word in database theory is the “you,” or human agency responsible for decisions regarding information. In the context of Klein, the locus of “you” as human interacting with or producing an archive becomes a space for determining the nature of archival imbalances, power, and structure—particularly when Klein asks, “How does one account for the power relations at work in the relationships between the enslaved men and women who committed their thoughts to paper, and the group of (mostly white) reformers who edited and published their works?” (664)

This same question of the “you” that must be accounted for appears in the data visualists’ role in rendering information visually, and is part of Klein’s call for a greater theorization of the digital humanities. She states, “the critic’s involvement in the design and implementation—or at the least, the selection and application—of digital tools demands an acknowledgment of his or her critical agency” (668). In Klein’s scholarship, qualifying and elucidating the role of “you” is paramount to understanding the archive, the visualization, and the data collected.

Critique without suggesting an alternative is all too easy, and I admire the way in which Klein posits data visualization as antidote to archival silences and also deeply engages the fraught history of its practice (665). She engages visualization’s vexed history through the figure of Thomas Jefferson himself, who underwent training in early forms of data visualization with William Small at the College of William and Mary. In this section of the article, we gain a sense of how complex it is to engage these forms: can the same tool that Jefferson was so fond of also be a tool for scholars to resurrect the memories and presence of the slaves he owned, centuries later?

Klein also explores the ways in which Jefferson’s note-taking and records use representation in diagrams, charts, and tables to suggest that he was engaged in using data visualization as a “form of subjugation and control—that is, the reduction of persons to objects, and stories to names,” which points at the reductiveness and potential for violence in types of visual display (679). Klein’s portrayal of Jefferson here, as an unthinking white man who recorded Hemings as empirical evidence, to be charted and claimed as thus, is emblematic of the central question of her piece: how can we visualize without appropriation, acknowledge incompleteness, and in a paraphrase of Marcus and Best, let ghosts be ghosts without claiming them for our own purposes or meanings?

Evoking Stephen Ramsay’s idea of “deformance,” or the creative manipulation and interpretation of textual materials, Klein ultimately suggests that rendering Hemings in an act of visual deformance makes legible “possibilities of recognition” that the actual textual content of the Papers of Thomas Jefferson resist, while “expos[ing] the impossibilities of recognition—and of cognition—that remain essential to our understanding of the archive of slavery” in contemporary studies (682).


When confronted with archival ghosts, Klein seems to suggest that the best policy is: illuminate, not explicate. How do you negotiate the difference between these two words, and can you share with us the ways it influences your pedagogy and scholarship?

Is there ever truly a safe way to visualize data, particularly regarding people and especially those who have been silenced, ghosted, or violated, in a way that rhetorically privileges stories and narrative over names and numbers?

To what extent does digital technology provide solutions of access for archival materials, but at the same time reproduce power structures that perpetuate silences? Can digital technology increasingly address this question through innovation, or is this a question of institutional change?

Klein’s argument regarding silences in digital archives seems to address the question of mark-up and encoding, whose granularity is often determined by institutional funding. In a recent conversation, Erin Glass (of Social Paper, an amazing platform for student-centered writing that you should check out!) and I noted that the first invisible document of any archive, institution, or project is often a grant. This document lays out the rationale, timeline, and required resources that shape the development of the project, but it is rarely discussed once secured for an institution, and is often invisible except in gestures towards sponsorship or funding. ROTUNDA is an organization that is part of University of Virginia Press, but whose digitization work is funded through grants. It is likely that decisions of encoding granularity were built into the grant itself and the time requirements of the project.

So, at the roots of the process of creating digital archives, how might we conceive of the entire process–from grant onwards–as a new space to intervene in inclusive, even collaborative, editing processes that produce richer metadata? Does this help address archival silences, or instead offer more opportunities to reproduce them?

5 thoughts on “Visualizing Impossibility: Thoughts on Lauren Klein

  1. Sakina Laksimi Morrow

    Is there ever truly a safe way to visualize data, particularly regarding people and especially those who have been silenced, ghosted, or violated, in a way that rhetorically privileges stories and narrative over names and numbers?

    This provocation reminded me of something I read in Manovich’s piece with regards to data visualization. He states that infovis is inherently reductionist, and that we often must shed many of the dynamic characteristics of a the thing we are visualizing in order to establish a narrative about that thing. As explained more eloquently by Manovich: “Infovis uses graphical primitives such as points, strait lines, curves, and simple geometric shapes to stand in for objects and relations between them – regardless of whether these are people, their social relations, stock prices, income of nations, unemployment statistics, or anything else…However, the price being paid for this power is extreme schematization. We throw away %99 of what is specific about each object to represent only %1- in the hope of revealing patterns across this %1 of objects’ characteristics”. He goes on to explain that this is not the result of modern techniques in infovis but rather the structures of relations of modern science of the 19th century- that both the “natural” world and social world must be understood through their “simple elements” and rules of interaction. In other words, reductionist thinking is inherent to the practices of the sciences (chemistry, physics etc.) and the social sciences (sociology, linguistics etc.) I am interpreting the words “safe” as accurate, fair and/or nuanced when it comes to representing the silenced, ghosted and violated. Does data visualization in this realm reduce the experiences and voices of the marginalized to some basic elements as objectified/determined by reductionist scientific thinking?
    While Manovich highlights some key elements of data visualization that is productive, he also points to the limitations of this medium in providing rich context. One such example is the privileging of the spatial representation of a lot od data visualization. How does this underpinning structure also determine what elements and interactions get told, and how does historical and cultural context fit in within this schema.

  2. Sara Vogel, PhD. (she/her)

    When Klein writes about illumination over demystification, she poses as an example the way her arc diagram alludes to the presence of James Hemings, a historical figure silenced at many points in the compilation of the archive. She argues that her arc diagram, even as it “exposes the grid of control of slavery that confined him to social and corporeal death,” illuminates the broader community in which he played an important role, without suggesting that we can use the archive — itself structured by power relations — to get at some deeper, hidden truth about him which we cannot ever know. Her argument works very well with her own example, but I wonder what other kinds of data visualizations might successfully apply her logic about illumination over demystification.

    Her argument makes me go back with fresh eyes to a data viz project that had a real impact for me, the “Million Dollar Block” project ( At first glance, the project seems to illuminate another type of ghost in our databases — prisoners — by calling attention to the addresses where they lived before they were locked up, and plotting those addresses on a map of New York City. Some cultural criticism is attempted by this project — knowing something about city neighborhoods, I can read in these maps the racism inherent in the system, and the argument about how these millions of dollars spent locking people up might be better spent on other forms of community development. Even if we aren’t expecting demystification and deep understanding of those silenced by the database, I don’t see much of an attempt by this data viz to restore agency to those depicted either. The people are represented by little dots labeled with how much they cost the city and state to incarcerate. The deep red on the black background also feels a little manipulative. I’m not saying visualizations like these aren’t powerful. They made a splash when they were published and exhibited at MoMA, and perhaps, were part of the conversation when prison reform policies (like the “close to home” initiative”) were proposed. But I think everyone’s right in saying that there are no truly safe ways to visualize data. Just like many other forms of quantitative data analysis, it can be done without ever interacting with the people behind the data point. Of course with historical research, ghosts are ghosts, but with more contemporary examples, I feel like data visualization can’t be the only research method employed for depiction of those violated, silenced, ghost-ed, if social justice and agency are the goals.

  3. Alexis Larsson

    Thank you, MCK, and Jojo and Robert. I’m also interested in this notion of safety in information. I think of my students coming into ENG110 classes who write that academic writing is all about conveying facts or conveying “subjective opinion,” and the categories are neatly separated and clear to them. Asking students to explore questions of how rhetorical decisions, including selection and representation, shape information and discourse while being shaped by these in contexts, becomes a way of examining essentialist views of language as a transparent medium for conveying information. Unfortunately, when it comes to data and (especially in reference to Sample and Few) visual representation, I experience a mental automatic shutdown, but I think the general concept still holds. There is an idea that numbers, and especially numbers on graphs, have an essence of truth that does not change when “presented” in different ways. Few seems to be challenging the straightforwardness in his comparison of ways of visualizing and pairing data. To be aware of the way numeric/visually represented data seem neutral, thus authoritative, is to lose rhetorical innocence. And to be aware that our rhetorical decisions in shaping access to this data have consequences that include, in the case of archival research, appropriating the stories of others in potentially questionable ways (I’m curious to know other people’s thoughts on archival work and ethics, especially since I have almost zero experience here), to become aware of rhetorical agency is also to feel culpable. And to make matters worse, that rhetorical agency doesn’t equal intentionality–our decisions have consequences in what gains in/visibility, and in the way we engage with others about data, but we usually can’t decide on the perfect way to affect the subject, that is, perform the most responsible and appropriate rhetorical acts.

  4. Robert Robinson

    Wow! I’m seconding Jojo’s “Thanks” and adding a “That was amazing…”

    I would like to respond to the trickiness of finding a “safe way to visualize data.” This reminds me of our discussion in the lab a couple weeks ago. I often feel as though the visualizations themselves serve as a type of narrative, but like we returned to multiple times courtesy of Lisa Brundage–and as you noted in your analysis–the “you” on both ends is essential. Sticking with the “you” in the authorial position, I believe safety is a nearly impossible reality. By sheer nature of the contexts that surround our existence–including the historical moment, cultural backgrounds, learning experiences, and more–we produce non-neutral texts, which I believe are inherently unsafe. Much like your comment last week about critical texts, our very inclusion of items in a database alludes to our inner values. From the position of the reader, we must consider the exact same contextual spaces and the transactions that take place when a reader picks up a text (Sorry, undergrad flashback– Being mindful of the interactions of author and reader and our human limitations, I would venture to say that even if we create narratives from numbers and attempt to provide voice to the silenced, we still run the risk of silence or exclusion in another area. I guess I believe that there is no way to create a safe visualization or narrative, maybe just a less unsafe one?

  5. Jojo Karlin (she/her/hers)

    Thanks for the illumination and provocation, MCK.
    The idea of incorporating “the grant into the continuity of the project seems logical, but also problematic –logical because it is certainly a stage of the conceptualization of the work, but problematic because its authors are extended to include the system of selection and preference belonging to grant giving bodies. Addressing the potential omissions and biases coming from these sources is certainly important, but the reading of the social side of these texts seems to me more a reading of the institutions themselves than of the work. I just wonder if it puts the project into the role of authority (in a corporations are people sort of way) when really the selection switches hands between grant-givers and grant-receivers.

Comments are closed.