Affective Footprints

By Trent Wintermeier

This essay follows a low-frequency sound looming around exposed communities in a relatively small area of southeast Austin, Texas. This sound is coming from the cooling equipment of four data center facilities. It’s artificial and metallic, the sound, but also disorienting and jarring—immediately recognizable as a low hum that is definitely other-than-human and, thus, machine generated. As a culmination of intensities pulsating, bouncing, shifting through large tubes and ducts separating the interior and exterior of these facilities, the sound is most accurately described as “low-frequency.” This means that it travels far and wide, and it does so quickly; it moves with purpose, strategically taking up space. It hurdles walls, sneaks around earplugs, and refuses to abide by other barriers that function to control access to seemingly private locations, like sensory extremities and the backyards of homes. It functions with determination and intent, meaning that the sound can seep deep into the resonant things which it surrounds. The whirring of fans creates this resonant ecology. Since data requires cool temperatures—especially in the hot state of Texas—the constant cooling is productive of such a persistent, stagnant sound in this part of the city. 

Photograph of CyrusOne AUS2 cooling system (photo by author)

Dissonant hums become the material and vibratory tissue that form bonds with individuals in close enough proximity to hear and feel their impact. Just beyond the confines of these data center service yards, which are home to sometimes hundreds of fans stacked on top and around each other, a perimeter of seemingly disparate things are brought into relation through this sound. Depending on the exact frequency being emitted, this community includes the bark of trees, the drywall and stucco of homes, porous particles of dirt, sand, and rock, and even the skin of humans and animals. The four data centers create a low-frequency cacophony of rhetorical forces that constitute community assembly. Through simply spatial location, exposure foregrounds how entities crack open together in their material and affective persuadability. In short, being around the sounds of these data centers makes a sense of community. Not in a blanket of impact but through independent resonant encounters. As Nina Sun Eidsheim explains, while “the human body overall has a low-frequency resonance… it does not vibrate as a single mass” (171). This community vibrates separately-together. At once-alone. Discrete-connected. This is where assembly is fashioned. 

Sensing sound becomes the precondition for expressing relationships, and this is what brings together such a southeast Austin community. These immanent and annoyingly reminded ties between entities and the sound of data centers is afforded by the close relationship between residential neighborhoods and the service yards of these facilities. Neighborhoods like Colorado Crossing—this is the residential site of exposure in which the data center sounds live. Homes; endless streets of them. Backyards and balconies facing down, challenging the fan blades cutting through the suffocating Austin heat. Residents are less than a few hundred feet from the hundreds of fans, making the materials of their homes and bodies no match for the vibratory pressure emitted from cooling innumerable terabytes of data. The perpetrator that this project is concerned with is CyrusOne (AUS 2), but the other three data centers are: CyrusOne AUS3, Data Foundry AUS11, and Digital Realty Austin 1. They are all spread across less than a mile of property called “MetCenter” that connects the aforementioned neighborhoods and communities. 

It does not cease: loom and exposure are static, steady, undeviating; the sound is happening now, as these words unfold and recognition becomes encountered and evident. This rhetorical force is never just, in any sense of the word. The correct side of any moral economy is subdued, there is no passing moment, and in no way is this almost anything because it is definite. This exposure to the low-frequency nature of data center sounds is a rhetorical force that constitutes relations and forges connections. This argument urges recognition of two related things, the topics of this essay: first, how did these southeast Austin residents arrive at this site of exposure? This question is interested in revealing the rhetorics underlying the spatial proximities responsible for the sound of data centers to impact and develop along with the residents of Colorado Crossing. And second, what other rhetorical forces are at play when sound and exposure are not the only constitutive elements of auditory relations? What other symbolic actants are operating—circulating, issuing, manipulating—within the tendencies of this situation? 

Encountering Data, Vibrations, Space

It has been nearly a year since I encountered the sounds coming from these data centers, in summer 2023, which during a visit around the property I began to understand the extremity of what was happening in southeast Austin. Extremity: indeed, the conditions are extreme, the sound is strong, violent, powerful. These are conditions experienced through extremities: the resonant nature of my hands and feet. My limbs became an essential part of how I approached my own limits of audibility—the breaking point of sound being “too loud” or “too disruptive.” The sounds of these data centers are right at that edge, immediately present affects in the tips of fingers and toes. To echo Steve Goodman: “if affect describes the ability of one entity to change another from a distance, then here the mode of affection will be understood as vibrational” (71). These data center sounds manifest as intense vibrational affects, which is why they are rhetorically charged in their relationship to other bodies (human, environment, animal).

On one side of me: I face a data center service yard with too many fans to count, so many tubes that it must be nearly impossible to find where they start or end. My notes have an illustration of a body with points of impact marked in pencil: knees, toes, hands, face; these are the sites affected. I’m collecting data, literally, it’s being sprayed all over my body in sonic form. The excesses of sustaining information, written on to me and into my notebook. Clearly, “the presence of noise pervades. It is agent,” as Marina Peterson states (3), and this rings true in my responses: hear, feel, write. What is also true, as she continues to illustrate, is that “noise is atmospheric. Palpable in its sensation” (4). The space between myself and the cooling towers and fans becomes rhetorically charged, and this is true for what’s behind me, too: a few hundred feet away are the backyards of Colorado Crossing residents. Standing there, I’d say the space between the CyrusOne data center and the neighborhood is only around four-hundred feet away. 

This encounter is one of rhetorical force, in that my auditory sensibilities and spatial surroundings become oriented when in the presence of sound. Its low-frequency loom collects affective potentials during expansion, distribution, and contraction. While I’m not suggesting that my experience equates to, or even merely resembles, the lived accounts of Colorado Crossing residents, the sense of affect, vibration, and data function similarly. Reflecting on this moment of standing between the data center and neighborhood, what I most viscerally recall is how I felt. This is the “stickiness” (Ahmed 29; Gregg and Seigworth 1) of affect, in that my own potential and openness is pulled together once again. And this underlies community relations as an assembly, a group parsed together through shared resonant exposure. I think this is what Jeff Grabill may be suggesting when he says that “the work of rhetoric is to assemble a group or a public around a matter of concern” (258): shared attunements are part of what constitutes a community. The rhetorical force of being affected can initiate this work and begin to center certain relational goals.

Arriving to Rhetorical Exposure

I’ll take a few steps back, because the sounds coming from CyrusOne’s cooling equipment have been a nuisance far prior to my visit to the site. CyrusOne is located on the “largest master-planned business park” in Austin, MetCenter (MetCenter “About”). It’s not hard to miss, the 550-acre property—it’s just off of the intersection of State Highway 71 and US Highway 183. Daily, thousands of cars peel by the big “M” standing guard outside of the entrance to the property. And directly across that highway is a chain-link fence separating the roadway from the property of Austin-Bergstrom International Airport; again, another frequented site, but this time by 747s carrying transient travelers, most of whom only step foot in southeast Austin to catch an Uber downtown. MetCenter is surrounded by this loud traffic: car horns, vehicle tires, plane engines. All of which are high-decibel and generally “loud,” serving to mask the sonic presence of lower frequencies. 

MetCenter has been built with these considerations in mind. That is, its proximity to the airport and major highways is one of the reasons why this location in southeast Austin was chosen by developers. In their market study document, MetCenter’s proximity to multiple highways is highlighted as part of the “attractive location” of the property, as it offers “regional mobility,” including access to public transportation, alongside potential airport expansion for international corporations (“Market Study” 5). Southeast Austin provides an apt geographic location for MetCenter in that it serves to connect—to stick together, affectively, through sound—other parts of the city and country where people want to live. Commuting made easy. Local economy ruined. The proximity that MetCenter articulates to various economic and community elements of the city functions rhetorically to position them as a simple, easy, and low-stakes business park that appeals to large international corporations looking to operate in Austin. However, MetCenter leaves behind a massive affective, ecological, and infrastructural footprint for nearby residents. 

As we also know, MetCenter is not only nearby to highways and airports, but also to local residential neighborhoods which border a significant portion of the property. This border is porous, only real in city plans and as security drives by to monitor its certainty. This border, like our skin, walls, and fences, is partial and it is so only to a few existents, including sound. The property’s spatial relationship to neighborhoods is not only outlined but determined by the City of Austin’s Imagine Austin Comprehensive Plan: Vibrant. Livable. Connected. In 2012, the City of Austin developed this plan to “turn around the negatives of the city,” which includes developing certain locations around the city—including southeast Austin—to “energetically leverage [the city and its residents’] strengths as [they] grow” (103). Of course, this plan includes territory that would be developed into MetCenter, which is the largest area in Austin classified as a “Job Center.” The issue is, the purpose of this territory is to “accommodate those businesses not well-suited for residential or environmentally-sensitive areas” (107, added emphasis). “Not well-suited,” undoubtedly a rhetorically charged substitution for “ill-suited,” characterizes the relationship between local communities and data centers.

To develop this in terms of my first question: how did these southeast Austin residents arrive at this site of exposure? Their exposure to the sounds coming from data center cooling is constructed by the economic, infrastructural, and especially the residential entanglements of the MetCenter property that have already been determined by the Imagine Austin plan. Such a classification of MetCenter’s property amplifies the affective tensions between the neighborhood zones and data center property lines. Communities form through this relationship that is determined by being “not well-suited” but, nonetheless, extremely near—living in Colorado Crossing. 

Cooling and Constituting Relations

Returning to 2023: after I read the Imagine Austin report—and after hearing the sounds of data center cooling through my own feet, ears, face, chest—I was curious about how this rhetorical connection, between the facilities and residents, is configured, sustained, and reproduced. CyrusOne affirms the sounds coming from their cooling equipment in a sustainability report published the prior year (2022). The sustainability report, then, serves not only to acknowledge the ecological footprint of the data center industry—this has been thoroughly reviewed by scholars such as Mél Hogan (2015), Anne Pasek (2019), and Dustin Edwards (2020); Edwards, with Rich Shivener, explains that “data centers are intricately woven with conditions of climate and climate change” (2020). A sustainability report is constructed to “save face,” highlighting not only the funds and energy but the community efforts involved in environmentally-sensitive initiatives and programs. Data center sustainability reports communicate to different public stakeholders how the industry is not only siphoning the limited amount of natural resources left on Earth, like water, but how they are making a positive impact to offset this detriment.

Aptly, a major point of discussion in this report is resilience, how to build resilience in the local areas of these data centers. Because creating the conditions for exposure to sonic pollution, among other types, requires diverting attention of local publics from the negative impacts of data centers. The report, under a section titled “Strategy: Building Networks of Resilience,” does just that; CyrusOne rhetorically embeds individual responsibility within a web of community sustainability. The contours of affect become strategically tweaked to, quite literally, make “sense” of a collective task, a common exposure. Community is positioned as a vibratory mass; we resonate together, we are responsible for each other. They enable the tired metaphor of a “network” that includes community, workforce, and ecology, stating that communities must be “resilient” to the “disruptions” that are part of the relationship between local residents and data center facilities (79-80). Auditory relations become rhetorically charged in reinforcement, toughness, and ability to withstand unexpected damage or harm. 

This is not just rhetoric of resilience, and I recognized this when I walked away from the data center facility with a subdued hum retained in the center of my chest and pulsating out through my arms. Framing resilience to the sonic output of these fans as requisite for community sustainability is nasty work. It’s destructive. Dangerous. Vile. Not only does resilience authorize the geographic proximity of MetCenter as “not well-suited” for nearby residences, but it frames exposure as an obligatory state. It puts into jeopardy individual agency, as CyrusOne retains control of the sensory faculties of people experiencing this sound, even as it never lets up. Momentarily, the affectability of being creaks and cracks open as a response to seemingly minor rhetorical adjustments. What enters in is the network that David Ferdman, CyrusOne President and CEO, constitutes:

“We seek to maintain a resilient workforce by ensuring our employees and contractors have the support they need to adapt to changing business and world conditions, while we nurture new and diverse talent through internships and training programs. Workforce resilience will only be possible in resilient communities, where needs are met and people are ready for disruptions. Ultimately, the success and health of all of our human and technological networks depends on a resilient environment—a world with ample water and habitat to supply the needs of both humans and wildlife” (3, added emphasis). 

To be clear, the report does mention that they get complaints from neighbors about the “noise coming from [their] operations,” and that they “make it a point to engage with neighbors rather than retreating behind the letter of the law” (80). Their sense of community engagement is organizing a quelled community that ignores the sounds of data center cooling practices. These communities are “ready for disruptions,” trained to ignore them in maintenance of a data center status quo: accumulate and wield information. This includes selling data to support large language model innovations, storing the data of clients, and informing the enhancement of algorithmic media and surveillance technologies. Resilience to sonic disruptions allows for these practices to ensue. But when I listened to the sound of CyrusOne, this wasn’t immediately clear. Like some forms of rhetoric, some sounds mask meaning—they operate to sustain certain relational imaginaries and attempt to destroy others. The point I’d like to emphasize is this: using the language of “resilience” to frame the conditions of this community attempts to support a network that emerges within and is produced by the noise of data center cooling. 

Outro: Haunted Openings

In the past year, I’ve found that the sounds coming from data center cooling have stuck with me, almost stuck on me around certain bodily sensory points. I can’t hear these sounds, but I can still feel them: this prior exposure haunts me, manifests in and as me, even in my remove from where the sound still is actively produced. Affect expands the spatial and temporal parameters we give to sonic experience. This is what I mean when I say that this sound looms: it hovers over, around, and through us, it emerges again in the intersection of ontological uncertainty and sensory openness. In other work, Steven Gonzalez Monserrate has said that the “cacophonic hum” of data centers close to residential neighborhoods “haunts nearby residents” (84). And he has called this sound “disturbing,” and says that “no one can” stop it because “it does not sleep” (2022). The rhetorical force of this sound comes from, in part, its ubiquitous presence, an inability to escape it. Simply knowing it’s there is enough to engender our affective pasts, reinvigorating the potential for effects to circulate and signify. Certainly, it does not sleep, and knowing this exposure—feeling it, viscerally—is sufficient proof of being acted upon. 

Audio recording of CyrusOne AUS2 cooling system (recording by author)

In Specters of Marx: The State of the Debt, the Work of Mourning, and the New International, Jacques Derrida asks: “what does it mean to follow a ghost?,” and this seems to be an appropriate question to end with. In this essay I’ve followed low-frequency sounds coming from data center cooling systems which function rhetorically to produce and reproduce community relations through affective resonances. To do this, I’ve moved through different spaces and times to, as best I can, track down what has constructed this sonic exposure and what it means for this construction to be rhetorically packed. This movement is part of following a ghost that is sound, affirming that “sound is not the object, it is the movement, the relation, always different from itself,” as Ian Foreman astutely recognizes (269). This work has been a sounding, in this way reproducing a relationship to the sonic for myself and the Colorado Crossing community. A recognition that movement toward a looming ghost pronounces a version of myself that is already being followed. While, indeed, this sounding has been different from its original self, the affects seem to remain the same. 

To track down this ghost means that haunting doesn’t follow exposure but haunting makes possible exposure as a visitation. In the context of this essay, this means that the Imagine Austin plan and CyrusOne’s sustainability report are rhetorical commitments that make this visitation possible, manifest. Arriving to and constituting exposure, then—as I’ve discussed in the above two sections—is not a question of how space has been determined but how affective relations have been conjured, i.e., called upon to appear. The sounds of data center cooling systems are merely a facilitator for this rhetorical event, this process of making evident an openness to the movement of sound and to being followed. 

Works Cited

Ahmed, Sara. “Happy Objects.” The Affect Theory Reader, edited by Melissa Gregg and Gregory J. Seigworth, 2010.

City of Austin. Imagine Austin Comprehensive Plan: Vibrant. Livable. Connected. 15 June 2012, www.austintexas.gov/sites/default/files/files/Imagine_Austin/IACP_2018.pdf.

CyrusOne. “2022 Sustainability Report.” 2022.

Derrida, Jacques. Specters of Marx: The State of the Debt, the Work of Mourning and the New International. Routledge, 2012.

Eidsheim, Nina Sun. Sensing Sound: Singing and Listening as Vibrational Practice. Duke University Press, 2015.

Foreman, Ian. "Spectral Soundscapes: Exploring Spaces of Remembrance Through Sound." Interference: A Journal of Audio Culture, vol. 4, 2014, pp. 1-11.

Grabill, Jeff. “The Work of Rhetoric in the Commonplaces: An Essay on Rhetorical Methodology.” JAC: A Journal of Composition Theory, vol. 34, no. 1, 2014.

Gregg, Melissa, and Gregory J. Seigworth. "An Inventory of Shimmers." The Affect Theory Reader, 2010, pp. 1-25.

Gonzalez, Steven. Cloud Ecologies: An Environmental Ethnography of Data Centers. Diss. Massachusetts Institute of Technology, 2023.

—. “The Cloud Is Material: On the Environmental Impacts of Computation and Data Storage.” MIT Case Studies in Social and Ethical Responsibilities of Computing, 2022.

Hogan, Mél. "Data Flows and Water Woes: The Utah Data Center." Big Data & Society, vol. 2, no. 2, 2015, https://journals.sagepub.com/doi/pdf/10.1177/2053951715592429.

MetCenter. “Information Relevant to Mission-Critical Facilities.” 2017, https://assets.websitefiles.com/60f5fa7dfdb7ee3be0a56f26/6109783253ed2b432283439b_ Mission-Critical-Facts-1-2017.pdf.

—. “Market Study for Locating in MetCenter.” 27 April 2021.

—. “MetCenter.” www.metcenter.com.

Pasek, Anne. "Managing Carbon and Data Flows: Fungible Forms of Mediation in the Cloud." Culture Machine, vol. 16, 2019.

Peterson, Marina. Atmospheric Noise: The Indefinite Urbanism of Los Angeles. Duke University Press, 2021.

Shivener, Rich, and Dustin Edwards. “The Environmental Unconscious of Digital Composing: Mapping Climate Change Rhetorics in Data Center Ecologies.” Enculturation: A Journal of Rhetoric, Writing, and Culture, 10 November 2020, https://enculturation.net/environmental_unconscious.



Previous
Previous

Secondhand Extraction and/as Mediation

Next
Next

Meditations on Tech-Fueled Visions for Canada’s “Green” Future