My Dissertation Defense Talk
On April 1st, 2022, I defended my dissertation, “Atmospheric Media: Computation and the Environmental Imagination,” and became a doctor. Below is the text of the talk I gave at the top of my defense, which explores the project’s argument, context, and next steps.
Friends, family, esteemed colleagues: I am proud to welcome you to the defense of my doctoral dissertation, “Atmospheric Media: Computation and the Environmental Imagination.” Thanking all those who have supported me over the three years of its writing would take all day, but I would like to extend a special thanks to my committee, in particular my director Matthew Kirschenbaum, whose support and guidance made this project possible. Over the next twenty minutes, I’ll discuss the dissertation’s contours and major arguments, explore its chapters, situate its work within its allied fields of digital media studies and the environmental humanities, and offer some thoughts on its future development. This is a project whose urgency, I believe, has only magnified during the years I have spent on it, as questions of the air and its mediation have moved from the background to the forefront of our daily lives. So let’s dive in.
I’ll begin with a definitional question: what are atmospheric media? Atmospheric media are techniques and technologies for the rationalization of air. You have no doubt encountered many today, even in this very room. The weather app on your phone providing the day’s forecast, the air conditioning keeping you comfortable in your home or vehicle, and even the masks you are wearing on your faces now to control the distribution of viral particles are all examples of atmospheric media. These are media of quantifying, understanding, and managing the air. The air, to paraphrase German philosopher Peter Sloterdijk, has “become a source of concern” to contemporary culture. When I began working on this dissertation in 2018, my driving concern was environmental. Rising levels of atmospheric carbon, increases in severe weather events, and aerial pollution more generally prompt concomitant strategies of atmospheric tracking, regulation, and modeling through a variety of media technologies. In 2022, thanks to the covid-19 pandemic, those concerns have deepened and multiplied, reaching the air inside our classrooms, even the air inside our bodies.
Broadly, my dissertation positions these atmospheric media as fundamental conduits for the cultural work of managing the air, and in turn, for managing these diverse environmental concerns. It does so in four chapters, each exploring a specific technique of atmospheric mediation. The first, “Imaginary Forecasts, or the Arts of Atmospheric Media,” explores the technique of forecasting through a close reading of three works of contemporary creative media art engaged with the science of meteorology. The second, “Air-conditioning the Internet: Temperature, Security, and the Myth of Sustainability,” turns its attention to the technique of conditioning through an analysis of the role air conditioning plays in keeping data centers, and thus the modern internet, operational. The third, “Collective Tissue: Media Toxicity and Respiratory Temporality,” combines a media archaeology of the toxicity of 3D printing with a close reading of science fiction author Ted Chiang’s 2008 short story “Exhalation” to explore the technique of respiring and the biochemical ramifications of atmospheric media. And the fourth, “Machine Reading for Atmosphere: Modeling, Affect, and Weather Records,” concludes the dissertation with the technique of modeling, using digital humanities tools of natural language processing to explore the place of affect and emotion in official governmental records of severe weather. While my dissertation involves a fair amount of technical detail—even, at times, my own computer code—its focus is squarely on cultural questions. How do scientists, artists, and laypeople alike turn to atmospheric media to develop social and aesthetic relationships to the air? And how do those relationships set horizons of possibility for addressing the global problem of climate change?
My answers to these questions are in many ways located in my dissertation’s subtitle: “Computation and the Environmental Imagination.” While atmospheric media precede the invention of the digital computer, I argue that what binds them together is a shared computational logic. Take, for example, the atmospheric medium of the weather diary. Weather diaries are among our oldest meteorological media, attested as far back as Babylonian civilization. In weather diaries, observers record atmospheric phenomena at regular intervals, often using a constrained set of descriptors. They are prototypes for the detailed records scientists and governments now keep of the weather, and indeed some of the earliest applications of digital computing in the 1940s and 1950s were in the service of recording and forecasting the weather. In atmospheric media, I argue, computation operates as a foundational logic networking many different kinds of atmospheres, from the sky above to the air inside us. Computation smoothes over atmospheric difference with the standardization of data. Computation then inflects what I call “the environmental imagination,” a phrase I adapt from environmental humanist Lawrence Buell. The environmental imagination names the cultural and technical processes through which humans mediate our understanding of what “the environment” is and what available relationships to it are. In many ways, my dissertation is a portrait of what I view as a dominant environmental imaginary in contemporary Western techno-culture: that of imagining the air as a computer, one that we can program as a way out of the crisis of climate change.
I’ll say a little bit now about each of the chapters themselves and their argumentative progression. The first, “Imaginary Forecasts,” opens the dissertation with close readings of three pieces of contemporary media arts that take the formal languages of meteorology as points of departure. They are: poet and programmer J.R. Carpenter’s interactive web project This is a Picture of Wind, book artist and digital humanist Johanna Drucker’s experimental chapbook Subjective Meteorology, and textual artist Allison Parrish’s now-defunct Twitter bot The Ephemerides. While these works vary in form and genre, they all have investments in electronic literature. I argue that these works' investment in computational techniques, such as randomness and procedural generation, reflects attempts to aestheticize the practice of meteorology. These works position forecasting as a speculative enterprise, conjuring imaginary weather diaries for days that never happened, pataphysical models for mapping emotional weather, and poems written by space satellites, in doing so imagining alternative models for capturing and processing weather data.
The second chapter, “Air-conditioning the Internet,” moves from seeing atmospheres to manipulating them. Its subject is air conditioning, specifically the air conditioning that maintains thermal equilibrium in data centers. Through a case study of the multinational data center company Equinix, whose DC11 data center in Ashburn, VA I toured in the spring of 2019 (thanks to my committee member Jason Farman), the chapter argues that HVAC is a fundamental technology without which the internet would collapse under its own heat. Moreover, the chapter links data center HVAC practices such as wide-area networks and aisle containment to anxieties about the ongoing sustainability of cloud computing under climate change. Data center companies need air conditioning to maintain the illusion of thermal control, even as data centers themselves contribute to environmental pollution more generally. Conditioning, therefore, becomes a broader conceptual framework for thinking about how atmospheric media construe models of the atmosphere and then work to bring actual air in line with those models.
The third chapter, “Collective Tissue,” moves from global infrastructures to the vulnerable human body. Its interest is pollution and the environmental costs of atmospheric media as revealed through the breath. It’s methodologically a bit more experimental than the previous chapters, combining a media archaeology of 3D printing with a close reading of Ted Chiang’s 2008 science fiction short story, “Exhalation,” which concerns a species of humanoid robots powered by air as they face the revelation of their mortality. My goal in this chapter is to argue that atmospheric media introduce new temporal relationships to the experience of media toxicity, what I call “respiratory temporality.” Respiratory temporality is marked by cycles and rhythms, as toxins move from media to lodge in the body in unknown ways. I’m also interested in this chapter in exploring different ways that atmospheric media engage the sensation of respiration; there’s a fun section, for instance, about the smells of 3D printing, and the capacity for smell to function as an analytic in media studies.
Finally, the fourth chapter, “Machine Reading for Atmosphere,” returns to more meteorological territory, although again with an unusual methodological spin. It begins from a question inspired by environmental media studies scholar Sara Grossman, who asks how official meteorological records can capture the affective and emotional experience of severe weather. To answer this question, I undertake a case study of the Storm Events Database, a record of severe weather in the United States from 1950 to the present day maintained by the National Oceanic and Atmospheric Administration. Among its many data, the Storm Events Database records written narratives for all its severe weather events, ranging from brief sentences to chapter-length documents. To read the Database, I use a technology called sentiment analysis, which seeks to quantify the presence of affect in a given text. The chapter becomes a parallel study of the Database, which I argue deliberately effaces affect from its records, and sentiment analysis itself, which I offer as a flawed but generative tool for media studies scholars interested in questions of machine learning and conceptual modeling.
Now I want to situate this project within the work that has inspired and influenced its development. “Atmospheric Media” is a work of environmental media studies. Environmental concerns have long been an interest to media studies—I think of early twentieth century critic Lewis Mumford’s work on mining in relation to technology, for instance, which has been a crucial influence of mine. Over the past two decades, spurred by cross-pollination from the environmental sciences and humanities, media studies as a field has developed what media theorist Nicole Starosielski has called “an environmental consciousness [in] the study of digital systems.” This work has taken many forms which have been touchstones for me in working on this dissertation: Starosielski’s own work on undersea cables, Jussi Parikka’s research on media and geology, Melody Jue’s work on media and seawater, and John Durham Peters' elemental media philosophy, to name but a few. These projects combine an interest in media’s materiality with research into their environmental effects, through heterogeneous methods spanning theory, sociology, media history, and aesthetic critique. My own work on atmosphere began with what I viewed as a gap in this scholarship around aerial materialities, although I have been pleased to see over the past three years more work emerge in this area from scholars such as Yuriko Furuhata and Jean-Thomas Tremblay. But more to the point, I have come to view atmospheres as a necessarily interdisciplinary concept, which requires thinking across media studies, the environmental humanities, literary studies, and the digital humanities. I consider this interdisciplinarity, which returns time and again to the concepts of computation and mediation, as this project’s core hallmark.
It is an interdisciplinarity that reflects my own intellectual career both here at Maryland and beyond. We are, of course, sitting in an English department right now, rather than a media studies department. This is a distinction that I believe both matters relatively little for this project, but also very much. It matters little in that media studies has always been a hybrid field that crosses disciplinary boundaries; English departments, conversely, have long been sites for doing media studies in institutions that do not distinguish the practice with departments of its own. Yet it matters very much in that my dissertation’s archives and methods are marked by my parallel trainings, here in an English department, as a literary critic and as a digital humanist. These are reflected, for instance, in my dissertation’s investment in sustained close readings of individual pieces of textual media, from electronic literature to SEC filings to science fiction stories to narratives in databases. They also emerge in my use of computational methods, which come to the fore in my fourth chapter’s analysis of affect in weather records. As such, this dissertation is both a record of my own intellectual interdisciplinarity, as well as an argument for that interdisciplinarity’s suitability to the study of atmospheres. Atmospheres, after all, are expansive. They touch all manners of living on this planet. It has made sense to me these past three years to write a project that attempts to match that expansiveness.
This project has many roots, but two in particular strike me as crucial for understanding how I began work on it and how it took shape over the years. The first was in my qualifying exams in the fall of 2018, in which I gave a presentation titled, a bit cumbersomely, “The Clock Rate of Digital Studies: Textuality and Historicism in Atmospheric Media.” (Which is funny given that I don’t consider myself a historicist in the slightest.) In it, without realizing, I had begun to trace many of the concerns that became central to the dissertation: an interest in the media of meteorology and forecasting; attempts to think computation in theoretical relationship to the environment; and the importance of media criticism, the close reading of media objects, to my method. (It was also where I coined the phrase “atmospheric media.") The second root was an article I wrote for the journal Amodern right before I began working on my first chapter (which would actually become the dissertation’s third). It was for a special issue on media technique, and concerned grids as a technique for imagining the air as data for the purposes of weather forecasting. Technique would go on to become a central organizing framework of my dissertation. Technique is a tricky concept to pin down in media studies, but I have come to value media theorist Grant Wythoff’s pithy definition that technique are ways of doing things with technologies. Each of the four techniques that organize my chapters—forecasting, conditioning, respiring, and modeling—then name ways that different stakeholders use computation in the ongoing work of managing the air.
As I come to the end of this talk, and to the end of my time as a graduate student here at Maryland, I want to offer some brief thoughts on how I hope to develop this project further into a book, which has admittedly always been its goal. From a publication perspective, I think this is a project with potentially wide appeal beyond exclusively academic audience. Reimagining it for commercial publication would certainly entail shifting its conceptual and methodological focus, but I think that many of its core arguments—namely, that we come to know and shape the air through media, and that the media with which we do so influence that knowledge and shaping—can remain. There is also more opportunity to integrate the profusion of work in environmental media studies and “atmospheric studies” that has emerged over the time of its writing, particularly on questions of race, gender, sexuality, and class. And there is, of course, the question of the covid-19 pandemic, which I made a deliberate decision to sideline in this project, but that I now think warrants explicit integration.
I want to thank again all of you for coming to this defense and sharing with me a celebration of this work. It has no doubt been challenging times to write a dissertation, particularly one almost comical in its continued relevance to our day-to-day lives in the twin atmospheric crises of climate change and the pandemic. I am very much looking forward to having a generative conversation with you all now, and I welcome any and all questions and comments. Thank you.