Solitary Mind, Collaborative Mind: Close Reading and Interdisciplinary Research

July 19, 2013, 08:30 | Short Paper, Embassy Regents F

We have discussed in some detail elsewhere, including in a longer paper at this conference, our ongoing development of software that visualizes poems as complex dynamic systems (Abdul-Rahman, et al. 2013a; Abdul-Rahman, et al. 2013b). In this presentation, we will discuss how working within this interdisciplinary context, which forced us to think about how poems work both as large, complex systems and on their most granular levels, has already led us to new insights about poetry, its features, and its operations. Our focus here will be not on the technology itself, nor on the literary perspectives we have been bringing to bear in its design. Rather, we will discuss how working collaboratively with visualization scientists, whose research practices and values could not be more different from our own, has caused us to imagine poems differently, and on how this adjustment in our thinking, which would not otherwise have happened, has in itself, independently of the technology, led us to new insights about poetry. Interdisciplinary dialogues are also revising our ideas about broader subjects like aesthetics and information as well as the specific ways our different fields perceive them acting on each other. For this reason, we argue, collaboration across the boundaries of very different disciplines has its own inherent value, even to a discipline like ours in which writing, reading, and the ‘freedom to explore’ has most traditionally been associated with solitude.

Even as interest in distant reading (Moretti 2007) rises in the digital humanities, close reading remains central to the practices at the heart of literary scholarship. As helpful as it can be in opening new directions of criticism, distant reading — with its reliance on abstract models based on data mining and quantification — removes human readers from the center of the reading process, bringing them in after the ‘reading’ phase is finished to interpret not texts but data. Thus, these approaches fail to support the most important practice in the study of poetry, which is valued precisely because of its experiential richness. In contrast, close readers engage texts directly, intimately and in detail; they trace the finest interactions among such literary features as rhyme and meter, sound, figures, and syntax, noting how even the subtlest movements and operations (a comma, a repeated vowel, etc.) influence a reader’s interpretation(s) and experience(s) of a particular poem. While the words in a given poem in a given version remain the same, different skilled readers will interact with those words in different ways by choosing moment by moment what to engage and what to ignore. Thus, there is no single ‘correct’ solution to any close reading problem, though there may be many incorrect solutions. A reader’s choices are led both by what is happening on the page and by the reader’s preferences and interests. Close reading, as both expression and experience, thus manifests interactions between poems and human minds.

While close reading is sometimes positioned as belonging to the set of practices not amenable to digitization, we have found in our own work and in a review of the work of others that this is not necessarily the case. In fact, there is no shortage of software designed and applied with the intention to aid close reading practice, whether or not it is yet effective in doing so (Chaturvedi 2011; Chaturvedi, et al., 2012; Clement 2012; Plamondon 2006; Ruecker, et al., 2008; Unsworth and Mueller 2009). This abundance indicates, in our view, an urgent desire in the literary community to embrace and explore the power of computation while at the same time prioritizing and protecting the relationship between literature and human readers. As we explain in our longer paper, one distinguishing feature of our project is the strong emphasis we are placing on poetry’s experiential quality, which we locate in its radical multidimensionality — especially its relationship to time. But we are also, through our interdisciplinary discussions, endeavoring to theorize how to position data visualization and computers as potential — and potentially potent—tools to aid close reading.

Our goal of using visualization to heighten poetic experience has been challenging to describe, understand, and pursue for all the members of our team. In fact, we have struggled to integrate into this overarching objective two distinct and not always obviously compatible elements: intensified aesthetic experience and revelation of new information. Early in our research, we looked to music visualizations for possible analogies that might inform our work. But sometimes impassioned discussions have shown us, for instance, that while visualizers like MilkDrop (Geiss 2012) and the ‘visual music’ of multimedia artists Abstract Birds (2012) may enhance aesthetic pleasure by combining visual and aural elements, they do not necessarily lead viewers to new intellectual perceptions about the music. The computer scientists in our group prioritize information clarity and communication through timesaving tools and techniques. Accustomed to collaborating with biologists, engineers, physicists and physicians to visualize phenomena like combustion, the brain and its electrical impulses, magnetic fields, etc., they have emphasized that their work is more than ‘just pretty pictures.’ At the same time, we poets have consistently defended the aesthetic as meaningful: in poet Robert Creeley’s words, "FORM IS […] AN EXTENSION OF CONTENT" (Olson 1966, p. 16, emphasis original). More is at stake than superficial display, we have argued; especially when visualizing aesthetic objects, events, or systems (like poetry), aesthetic choices are necessarily core design considerations. The aesthetics of individual visualizations should somehow reflect the uniqueness of individual poems.

Similarly, our visualization scientist colleagues are used to identifying specific problems to solve, data to isolate, hypotheses to test: explicit aims that help them define programming strategies and evaluation techniques for very complicated software. Our stance on verifiability and accuracy is more ambivalent and ambiguous. Initially we resisted the very idea of literary hypotheses-making and testing, until we saw it as describing the many questions and choices a close reader constantly makes on the fly in her encounter with a poem. We continue to contemplate what ‘poetic data’ might mean, and what its relationship to literary forms and devices like lineation and metaphor might be. Even in our initial focus on poetic sound, the issue of accuracy has proven difficult to treat. How should we mathematically define rhyme? What constitutes the smallest measurable repeating sonic cluster? Of course we want our program to be able to correctly identify single repeating phonemes to show assonance and consonance. But accuracy becomes categorically more uncertain and tenuous as soon as we move beyond single phonemes. Should only full end rhymes qualify as rhymes? What about internal rhymes? Slant rhymes? As we build this software in hopes of its being useful for as many readers as possible, we want to consider not only how we, but how others whose literary perspectives differ from our own, might answer such questions. We want our software to allow individual reader-users to choose how to define some of these parameters in order to best support their unique close readings.

In our program, then, visualizations are not authoritative arbiters to reduce poetic complexity, give definitive answers, or merely save time; rather, they are aesthetic agents, using visual constructs and perception to reveal poetic features, patterns, and qualities readers might not otherwise have noticed by prompting them toward fresh experiences, insights, and questions. Thus, the visualizations do not replace close reading; they suggest rich avenues for initial and subsequent explorations, cuing the reader into which operations of the poem may merit further investigation. This orientation has also encouraged us to consider how the human-computer interaction developing through this software may prove more nuanced than the word ‘tool’ superficially suggests. It has provoked us to ask whether and in what sense the computer generating these visualizations might fruitfully be considered a fellow literary entity and even collaborator in the close reading process—questions we have differed on, as we will explain in our presentation.

Given our usual scholarly practices — based in the essentially experiential and qualitative nature of poetry, not to mention in the space of determined solitude where poetry is usually experienced—we were not natural targets for a project in visualizing poetry or even for a project that is highly collaborative. Of course we recognize that collaboration can generate programs and poems that otherwise would not have been possible. And this interdisciplinary project is indeed accomplishing those results. But it is also challenging us to examine fundamental issues, including our expectations and understanding of collaboration itself. We were brought into the project by our collaborators’ commitment to respond to our existing values and practices. Among the surprises for us has been the usefulness of collaboration itself to the practices and priorities we already embrace. At the same time, when one of the computer scientists in our group recently invoked Gregory Bateson’s notion of information as ‘the difference which makes a difference,’ (1972, p. 453, emphasis original) we recognized it as an apt description of our interdisciplinary research: exploring our various differences is not only leading us to new territory, it is helping us see familiar ground anew.


This work was supported by a Digging Into Data Challenge grant: in the US, by the National Endowment for the Humanities; and in the UK by the Arts and Humanities Research Council, Economic and Social Research Council, and JISC. We would also like to acknowledge and thank our collaborators: Alfie Abdul-Rahman, Min Chen, Christopher Johnson, Eamonn Maguire, Miriah Meyer, and Martin Wynne.


Abdul-Rahman, A., K. Coles, J. Lein, and M. Wynne (2013a). Freedom and Flow: A New Approach to Visualizing Poetry. Paper to be presented at Digital Humanities 2013, University of Nebraska-Lincoln. July 2013.
Abdul-Rahman, A., J. Lein, K. Coles, E. Maguire, M. Meyer, M. Wynne, A. E. Trefethen, C. Johnson, and M. Chen (2013). Rule-based Visual Mappings — with a Case Study on Poetry Visualization. Paper conditionally accepted to EuroVis, Leipzig, Germany.
Abstract Birds. (2012). http://www.abstractbirds.com (accessed 19 February 2013).
Bateson, G. (1972). Form, Substance, and Difference. In Steps to an Ecology of Mind. New York: Ballantine.
Chaturvedi, M. (2011). Visualization of TEI Encoded Texts in Support of Close Reading. M.S. thesis. Ohio: Miami University of Ohio. http://etd.ohiolink.edu/sendpdf.cgi/Chaturvedi%20Manish.pdf?miami1323623830 (accessed 6 September 2012).
Chaturvedi, M., G. Gannod, L. Mandell, H. Armstrong, and E. Hodgson (2012). Myopia: A Visualization Tool in Support of Close Reading. Digital Humanities 2012. held July 2012 at University of Hamburg.http://lecture2go.uni-hamburg.de/konferenzen/-/k/13930(accessed 6 September 2012).
Clement, T. (2012). Methodologies in the Digital Humanities for Analyzing Aural Patterns in Texts. Proceedings of the 2012 iConference. Toronto, Canada, February 2012. ACM Digital Library: http://dl.acm.org/citation.cfm?id=2132213 (accessed 6 September 2012).
Geiss, R. (2012). MilkDrop. http://www.geisswerks.com/milkdrop/ (accessed 19 February 2013).
Moretti, F. (2007). Maps, Graphs, and Trees: Abstract Models for Literary History. London: Verso.
Plamondon, M. (2006). Virtual Verse Analysis: Analysing Patterns in Poetry. Literary and Linguistic Computing. 21(Suppl), 127-141. http://llc.oxfordjournals.org (accessed 3 December 2012).
Ruecker, S., M. Radzikowska, P. Michura, C. Fiorentino, and T. Clement. (2008). Visualizing Repetition in Text. CHWP A.46 http://projects.chass.utoronto.ca/chwp/CHC2007/Ruecker_etal/Ruecker_etal.htm (accessed 6 September 2012).
Unsworth, J., and M. Mueller. (2009). The MONK Project Final Report. http://www.monkproject.org/MONKProjectFinalReport.pdf (accessed 6 September 2012).