This section discusses how material infrastructures add another layer of methodological complexity to platform research. By infrastructures, we draw attention to those visible and invisible structures that impinge upon and make possible networked communication (DeVoss, Cushman, & Grabill, 2005). More than technical structures and systems, infrastructures influence the relational—ethical—contours of networked life (Beck et al., 2016; Brown, 2015). By emphasizing materiality, we mean to shed light on how digital information is entangled with materialities of all kinds (cables, plastics, and metals, but also land, water, and land-based community practices). Attention to materiality, as Laura Micciche (2014) and many others have recognized, is a feminist project, as “all forms of matter, living and nonliving, are significant to sociocultural, political, as well as biological systems” (p. 491). Here, we position questions of material infrastructure as fundamentally feminist—and indeed technofeminist—questions, as they often are entangled with conditions of environmental harm, histories of colonialism, and legacies of militarism. In keeping with an intersectional approach, we draw on decolonial theory (Haas, 2012; Powell, 2012) and new materialism (Parikka, 2015) as allied lenses to investigate the materialities of Facebook’s New Mexico data center.
Although digital platforms are discursively constructed through interface design, branding, and stated policies and guidelines, they also have an expansive material footprint that stretches across the globe in the form of fiber optic cables, data centers, electrical grids, water cooling facilities, and more. Despite metaphors of clean and connective clouds that sit atop the Earth, digital data are neither immaterial nor particularly clean. The major companies of the platform economy have an immense impact on hyper-localized ecologies, as data centers require vast amounts of energy to power and billions of gallons of water to cool. For example, a Greenpeace report from 2009, “How Dirty is Your Data?,” presented staggering statistics about the energy consumption of data centers, indicating that “the combined electricity demand of the internet/cloud (data centers and telecommunications network) globally is 623bn kWh” (p. 11). Put otherwise, according to these statistics, if all the world’s data centers formed into one country, it would rank as fifth in energy consumption worldwide.
Stacked in rows from floor to ceiling in walled spaces, servers in data centers produce massive amounts of heat as a byproduct of constant computational work and the vast amounts of electricity that rush into one space. In response, data centers often rely on air conditioning and water to cool the servers and keep them online. As the data center industry has matured (and faced public scrutiny about its water and energy waste), many centers consider the physical locations of their storage sites to offset the need for expensive and inefficient cooling solutions. For example, the data center market is growing in the Arctic Circle, because, for data, colder is better, and moving cold Arctic air into data centers is a cost-effective way of keeping them running smoothly. In this regard, data centers fundamentally shape and are shaped by the locations in which they are placed. As Jussi Parikka (2015) described, “Data demand their ecology, one that is not merely a metaphorical technoecology but demonstrates dependence on the climate, the ground, and the energies circulating in the environment” (p. 24). In other words, data centers are deeply entangled with land, water, and other geophysical elements.
The notion that “data demand their ecology” is nicely illustrated in a promotional video from Volvo that describes how the climate conditions in Sweden are apt for data-driven platforms. In the video, shown below, the former mayor of Luleå, Sweden, discussed how the city attracted Facebook to locate its first international data center in Sweden: “Bringing the Facebook data center to Luleå,” he noted, “is changing the way people do things. Before they think winter is just freezing. Now they can understand that cold is a good thing for a lot of things. Data centers need cooling. We use the Arctic cold to get more from less...now the world’s data is flowing through Sweden.” This video showcases how the decisions to place data centers in particular locations are not random or neutral—platforms like Facebook understand the value of choosing ecologically sound locations for their data centers.
It might seem strange, then, that platforms would locate data centers in places in the world that don’t provide the same kind of environmental resources. But climate conditions aren’t the only considerations platforms keep in mind when choosing new data center locations; they also consider the extent to which an area has a pre-established Internet infrastructure and are persuaded even more by tax incentives and other state-sanctioned deals (such a free water rights). It is for this latter reason that Facebook has invested billions of dollars on a data center in Los Lunas, New Mexico, which is currently under construction and slated to be operational in late 2018. Yet, for this project, no such promotional video about data enjoying a certain climate exists.
Tracing the decision to place the data center in Los Lunas reveals a different kind of story about how platforms go about choosing data center locations. News reports indicated the state of New Mexico agreed to give Facebook $30 billion in industrial revenue bonds, giving the platform a 30-year property tax break. Furthermore, the village of Los Lunas agreed to give Facebook free water rights for a year in the amount of 162 billion gallons. And though Governor Susana Martinez heralded the deal as a move to bring long-term economic prosperity for New Mexican residents in the form of jobs, it remains unclear how many New Mexicans will be employed by Facebook, as the main construction company on the project is headquartered in Oregon and data centers typically are staffed by only a few dozen employees.
In addition to questionable economic motives, the data center is being built on colonized lands and will use water pumped from the increasingly dry Rio Grande River. Facebook’s new location is just miles from the Isleta Pueblo, whose peoples have lived along the Rio Grande valley for generations. It is difficult to grasp the deep materiality of the Los Lunas data center, but to grease the wheels of the Facebook platform—to have a smooth user experience where we can connect with friends, like cat pictures, and post messages of protest and critique—data will travel via fiber optics cables to a series of buildings in the arid climate of New Mexico. In route, that data will get tangled up with land, water, and colonial histories. Technofeminist research on and practices with digital platforms can help to engage such an expansive, and admittedly fragmented, understanding of material infrastructure.
As Angela Haas (2012) argued, “for decolonial ideologies to emerge, new rhetorics must be spoken, written, or otherwise delivered into existence” (p. 287). If engaging with an intersectional strand of technofeminism that redresses histories of racism, sexism, colonialism, ableism, and other cultural biases, scholars might begin thinking more critically about how the material infrastructures of data intensive platforms unevenly affect the communities and naturecultures in which they are placed. And we might think of new rhetorical approaches to do this work. To fail to do so entraps us in colonial practices. Malea Powell (2012) reminded us of the epistemological and material violence that ensues when we erase bodies from everyday life. To keep analyses focused on the screen or interface or program or code alone dismisses the bodies often abstracted from most people’s everyday engagements with platforms. In the next section, we focus on the body from a different perspective—noting how networks of support, activism, and care can emerge on commercial platforms.