Earth Observation and Signal Territories: Studying U.S. Broadcast Infrastructure through Historical Network Maps, Google Earth, and Fieldwork

This article engages with three different modes of Earth observation—historical network maps, Google Earth interfaces, and fieldwork—to develop the concept of “signal territories” and elucidate a critical approach for studying U.S. broadcast infrastructure. This approach: 1) highlights physical infrastructures—technological hardware and processes in dispersed geographic locations—as important sites for historical and critical analysis in media and communication studies; 2) explores multiple modes of infrastructure representation—ranging from cartography to phenomenology, from hand-drawn maps to digital interfaces, from circuit diagrams to site visits; and 3) foregrounds the biotechnical aspects and resource requirements of broadcast infrastructures, probing their dynamic operations and complex materialisms. Engaging with what Richard Maxwell and Toby Miller call a “materialist ecology” of media, the article explores what is at stake in understanding media infrastructures from up close and afar.

Cet article examine trois manières différentes d’observer la Terre (les cartes de réseau historiques, les interfaces Google Earth et les enquêtes sur le terrain) dans le but de développer le concept de « territoires délimités par un signal » et d’élucider une approche critique pour étudier l’infrastructure de la radiodiffusion américaine. Cette approche : 1) souligne les infrastructures physiques (les matériels et processus technologiques dans des lieux géographiques dispersés) comme étant des sites importants pour l’analyse historique et critique en études des médias et de la communication; 2) explore plusieurs sortes de représentation des infrastructures—de la cartographie à la phénoménologie, des cartes géographiques dessinées à la main aux interfaces numériques, de schémas de circuits aux visites sur place; et 3) met en avant les aspects biotechniques et les besoins en ressources des infrastructures de radiodiffusion, examinant leurs opérations dynamiques et leurs matérialismes complexes. Cet article a recours à ce que Richard Maxwell et Toby Miller appellent une « écologie matérialiste » des médias afin d’explorer ce qui est en jeu dans la compréhension à petite et à grande échelle des infrastructures médiatiques.

The Earth is criss-crossed by transoceanic cables, satellite Earth stations, broadcast networks, mobile telephone towers, and Internet server farms. During the past several decades, communication historians and scholars have conducted key studies of such systems. Collectively, this path-breaking scholarship has examined the political and economic strategies that undergird national and international communication networks, the cultural impacts of their emergence, and the imperializing dimensions of their use. This work has articulated the rise of communication networks with the administrative manoeuvres of nation-states and multinational corporations; processes of modernization, urbanization, and globalization; and various stages and forms of capitalism (Carey, 1989; Castells, 1996; Innis, 1951; Mattelart, 2000; Price, 2002; Schiller, 1969). Crucially, this scholarship has also delineated the relations between communication networks, geopolitics, and regulation.

What is often missing from this work, however, is a detailed investigation of the multiple ways that these infrastructures have historically become intelligible to citizen-consumers and intersect with cultures of everyday life as well as a sense of the contrasts across industrialized and developing regions, rich and poor areas, urban and rural settings, and literate and oral cultures. I want to begin, then, by examining several U.S. network maps from the late nineteenth and early twentieth centuries. These maps not only provide a platform for thinking about the ways infrastructures have historically been rendered through cartographic practice, but also serve as telling precursors to Google Earth’s FCCInfo layer in the way they foreground the scattered locations of infrastructures as well as issues of ownership, regulation, and technological literacy. Although there are many historical studies of infrastructures such as electrical grids (Hughes, 1993), railroad tracks (Schivelbusch, 1987), telephony (Fischer, 1994), radio networks (Hilmes, 2011), and global telecommunication systems (Headrick, 1981, 1988, 1991; Hugill, 1999; Winseck & Pike, 2007), there are fewer studies of the historical mapping of such systems (Challis & Rush, 2009; Dodge & Kitchen, 2004; Edney, 1999).

Network maps from the late nineteenth and early twentieth centuries approximated the vectors of transportation, commerce, and communication and developed a symbolic language that transformed massive, physically decentralized technological systems into icons that could fit within a visual frame. Early telegraphy, railroad, and radio maps were often hand-drawn and/or painted on paper. Mapmakers inscribed the locations of interlinked nodes (whether train, telegraph, or radio stations) upon national or regional maps of sovereign territories, encouraging the conceptualization of networked infrastructures as a set of distinct and isolated yet connected parts. The maps typically used simple arrangements of dots or circles (connoting a fixed site of activity) connected by lines (connoting dynamic movements or flows). Though railroad, telegraphy, and radio infrastructures were new in the nineteenth and early twentieth centuries, the cartographic template used to represent them arguably dates back centuries to celestial charts, tree of life diagrams (Harley & Woodward, 1987), or medieval diagrams of pilgrimage routes (Connolly, 2009). Nevertheless, network maps were significant in the nineteenth and early twentieth centuries as they not only documented infrastructural developments, but also gave them a concise visual form. Network infrastructures were dispersed across vast distances and operated at speeds and scales that were impossible to photographically record or perceive in their entirety (Kern, 2003; Schivelbush, 1987; Warf, 2008). Network cartographers compressed and abstracted these processes, enabling viewers to comprehend the spatial organization of new technologies of transportation and communication by implanting them within familiar cartographic spaces. In this sense, historical network maps were both crucially instructive and highly reductive: they offered static and remote overhead projections of massive, dynamically evolving systems.

In addition to making new infrastructures legible, historical maps gave new meaning to and tested concepts such as sovereignty, national community, and locale as they inscribed new routes and connections within and between the boundaries of sovereign national territories. The network map was a highly contradictory discursive space. On the one hand, it projected a visual rhetoric of technological integration and unification, inscribing the trajectories of emerging systems upon maps of sovereign national boundaries, which were, in some cases, still being contested. On the other hand, as the network map transposed a series of new lines upon already existing national and state boundaries, it mobilized a visual logic of transection, division, and reorganization. The network map at once played a role in discursively securing or reinforcing sovereign boundaries by showcasing the telecommunication and transportation infrastructures that tethered distant sites together and challenged this integration by delineating an array of new boundaries and modes of organization. In this way, network maps articulated what James Hay (2012) has described as a liberal democratic regime of communication, one that celebrated the new freedoms of mobility afforded by modern transportation and communication technologies while reserving the right to use those very technologies to patrol, secure, and defend its sovereign boundaries (also see Chun, 2008).

To provide a clearer sense of the function of these historical network maps, I will briefly discuss U.S. telegraphy, railroad, and radio maps from the late nineteenth and early twentieth centuries. These maps can be understood as historical precursors to the more contemporary Google Earth–FCCInfo interface since they employed similar cartographic conventions. This 1853 map by Chas. B. Barr of Pittsburgh (Figure 2) shows the locations of telegraph stations in the United States, the Canadas, and Nova Scotia.

The telegraph system is represented here as a series of black lines connected to black dots that stand in for transmission routes and telegraph stations. These lines and dots are layered onto a map of eastern and southern territories in North America to provide a sense of where telegraphic communication could or could not have happened on the continent during the middle of the nineteenth century. For instance, the map shows limited access in the Canadas and in southern parts of the U.S. and in doing so reveals the infrastructure to be limited and partial in its reach—as a differential infrastructure rather than a total one. Even as the map foregrounds telegraphy infrastructure’s partiality, it unveils expansionist ambitions, for remarks in the map’s legend indicate that 27,000 miles of U.S. telegraph line was in operation in 1853, with another 10,000 miles under construction. Other remarks highlight the planned expansion of the system across the American South and into Mexico and Cuba. Finally, details about ownership are conveyed: the legend indicates private party Morse owned 17,500 miles of the existing telegraph line and the U.S. House of Representatives owned the rest. Long before Google Earth, then, network infrastructures were layered onto territorial maps to reveal systems owned and operated by multiple corporate and government stakeholders.

A railroad map from 1876 by G. W. and C. B. Colton & Company (Figure 3) uses similar cartographic conventions to map the routes of the Southern Pacific Railroad Company. Southern Pacific’s two main routes are layered onto a regional map of the Western U.S. and are colour-coded as blue and red lines. The routes of two other companies—Central Pacific and Texas and Pacific—appear in brown and yellow, respectively, conveying the multiple competing railroad companies in the American West. Remarks in the legend indicate the map was presented to the Judiciary Committee in connection with Southern Pacific’s application for rights to pass through federal lands under Congressional Acts passed in 1866 and 1871. Thus as the map reveals Southern Pacific’s routes, it also puts into visual discourse questions about land ownership, access, and regulation that are vital aspects of infrastructure development and expansion. Like the telegraphy map, this map foregrounds the limited reach and partiality of this emergent infrastructure, as most areas are untouched by the tracks. It shows Southern Pacific’s tracks extending from south to north through the agricultural corridor of California’s central valley, with interstate connections into Oregon and Nevada via the Central line and into Arizona via Texas and Pacific. As lines are inscribed on a map of the Western U.S. to signify tracks, they re-write existing state boundaries, emphasizing the railroad’s capacity to support interstate movement and commerce.

A third and final example brings us into the era of broadcasting. A 1937 map of the National Broadcasting Company’s radio network (Figure 4) uses uniquely patterned black lines to indicate the locations of NBC’s red and blue radio network facilities as well as “supplementary legs” that could be used with either network. The lines representing network transmission routes are layered onto an abstract map of the U.S. that is shaded in orange. Like the telegraphy map, the lines connect to dots that represent stations scattered across vast distances. While this map showcases NBC’s national expansion, it offers less geographic detail than either of the other maps discussed, signalling a turn toward greater abstraction. Since this map features the infrastructure of one privately owned radio network—NBC—and none of its competitors, it functions as a document of strategic corporate planning and/or promotion as it delineates and showcases the network’s extensions and enterprises across the country. At the same time, however, like the other maps discussed earlier, this map implicitly reveals the partiality and unevenness of NBC’s infrastructure, since vast swaths of the country remained beyond the network’s reach. Although areas along the east and west coasts, in the upper Midwest, and in the southern U.S. were interlinked via NBC facilities in 1937, those in the West were less trafficked by the network’s signals. What the map does not represent are the localized patterns of coverage surrounding each affiliate station, which vary from site to site and are contingent upon the position, design, licensing, and strength of the transmitter.¹

By briefly discussing these three maps, it is possible to grasp some of the historical conventions used for mapping infrastructures of communication and transportation. Long before the emergence of Google Earth, cartographers designed graphics and icons to signify network nodes and routes, devised practices for layering graphically rendered networks onto territorial maps, and used colour-coding and other visual techniques to differentiate systems with unique owners or routes. These earlier maps circulated information about the location, ownership, expansion, and regulation (or lack thereof) of telegraphy, railroads, and radio and generated discursive spaces for deliberating and assessing the design and political, economic, and cultural effects of infrastructural developments. Historical maps of network infrastructures are instructive in several other ways as well. First, by inscribing emergent networks of telecommunication and transportation within cartography, these maps helped to evolve a visual language for representing national network systems. While today such systems can be observed with one satellite’s pass, in the nineteenth century massive, dispersed systems that stretched across an entire continent in some cases were drawn by one person’s hand. This fusion of an extensive, material infrastructure with an individual’s perceptual and creative capacities speaks to the incredible challenge of infrastructural representation. While telegraphy and railroads helped to augur new perceptual modes and aesthetics articulated with “the annihilation of space and time” (Carey, 1989; Innis, 1951; Kern, 2003), the map or diagram became the only visual discourse that could represent the vastness of these new national systems in a single frame.

Second, and relatedly, historical maps provided a discursive space for articulating infrastructural specifications. As telecommunication and radio networks expanded and grew more complex, a standardized visual language for representing their internal circuitry emerged so that engineers could consistently convey the structure, function, and dynamism of different network nodes and operations. As Edward Jones-Imhotep (2008) observes, “[d]rawings were not just sites where concepts and ideas were worked out—they were instruments through which materials and objects gained meaning and acted back upon the drawings themselves” (p. 449). He continues, “[d]rawings were the places where electronics were first (imaginatively) assembled, erased, operated, and reconfigured, and therefore a key site for contest over the meaning of the devices that went into them” (p. 450).

By the mid-1950s networks were no longer only projected onto territorial maps, they were also increasingly splayed open, so to speak, and given their own semi-cartographic practice as electronic engineers—who also had to be trained as creators and readers of such diagrams—invented a language for delineating their internal dynamics. This meant that physical objects and processes were translated into standardized sets of symbols and icons that were designed to enable quick understandings of the structure and dynamism of these systems. Like the historical network map, the circuit diagram privileged abstraction over realism, resonating with Joe Milutis’ (2006) observation that “[a] network, in a sense, is a pure abstraction, no longer determined by topography. It is a system of etherealized signs” (p. 153).

Third, and finally, historical network maps signalled their own proprietary character, which was encoded through the inscription of the cartographer’s name in the map’s visual field. As cartographers visualized emerging infrastructures owned by competing parties, they were at the same time enacting their own property claims in an economy of signs—staking out intellectual property. While in the nineteenth century individuals or small companies produced network maps, by the late twentieth century the mapping of telecommunication and broadcasting systems had become a full-fledged industry. Today large private firms such as Choisser, Geospatial Broadcasting Network, Analytic Graphics, Inc., and Telegeography are contracted by media and telecommunication corporations and regulatory agencies to produce local, national, and transnational maps of systems around the globe. Such maps are not only used to specify network locations and configurations, but are also instrumental in corporate planning and strategizing (Dodge & Kitchen, 2004) and, as a result, have become increasingly proprietary and inaccessible to most people. As the privatization and abstraction of network mapping intensifies, it arguably becomes more challenging for citizens to gain access to information about the locations, ownership, operation, and regulation of infrastructures, which is problematic given that citizens subsidize, use, and are surrounded by infrastructures in everyday life. This is one of the reasons that Google Earth is a provocative space for thinking about infrastructural intelligibility: it has the potential to extend public access to and knowledge of network cartographies.

When Google Earth first emerged in 2005, geographers suggested that the virtual globe would have the potential to catalyze greater public involvement in mapping and celebrated it as bolstering practices of participatory or volunteer geography that had begun with geographic information systems (Elwood, 2009; Goodchild, 2007). At the same time, however, since Google Earth relies on an international user community to produce content known as “layers”—geo-referenced datasets accompanied by graphics and icons formatted as kmz files that can be viewed in the platform—it also relies upon the “free labour” of Internet users—the uncompensated expenditures of human energy/time/attention, which, Tiziana Terranova  (2004) argues, have become a defining condition of network culture. Although some have lauded Google Earth’s participatory structures, others have critiqued the platform’s implicit disregard for national sovereignty, highlighting the opposition of countries such as Australia, India, Morocco, and South Korea to Google’s release of high-resolution geospatial imagery of their sovereign territories, including sites deemed critical to national security (Kumar, 2010; Parks, 2012). Still others have discussed Google Earth as an Apollonian eye in the sky or as part of a neo-Baroque aesthetic. As Leon Gurevitch (2013) observes elsewhere in this issue, “[t]he digital globe interface communicates with renewed vigour the centrality of the human agent, elevating the scopic powers of that agent through an interface that deploys a ‘code of spectacular vision’ inherited from Renaissance perspective” (p. 338).

Rather than approach Google Earth in broad brushstrokes, I focus on one layer called FCCInfo. Using a series of screen captures, I explore how this layer structures and communicates information about contemporary U.S. broadcast infrastructure. Furthermore, I assess whether FCCInfo has the potential to extend technological literacies and materialist understandings of signal distribution practices beyond a community of experts. FCCInfo was created by the private consulting firm Cavell, Mertz, & Associates and was made available as a free download in 2009. Based in Manassas, Virginia, Cavell, Mertz, & Associates’ (2012) mission is “to make complex technical and related regulatory issues understandable and manageable” (n.p.), and the company took on this project for the FCC to promote its own mapping and consulting services.

The FCCInfo layer generates an interactive cartographic display of all registered and licensed radio and TV stations, antennas, towers, and microwave broadcast facilities in the United States. Colour-coded balloons are used to represent different parts of the broadcast system. AM radio stations are represented in purple balloons, FM radio stations in green, TV stations in red, and registered towers in orange. Microwave links are represented in light blue, lavender, and pink. The layer also provides pop-up windows offering detailed information about each structure and site, including the type of structure, height above ground level, height above mean sea level, the channel, and the owner of the structure. Through the pop-up window the user can click on a hyperlink to an FCC website with even more data about the particular structure and site and its licensing history. The user can also zoom in on specific network nodes or facilities and acquire aerial and, in some cases, street-view perspectives of broadcast infrastructure sites. Never before have U.S. citizens, or members of the international community, for that matter, had such direct and immediate access to information about U.S. broadcast infrastructure, in both well-known urban media hubs and smaller, rural media markets.

To explore the FCCInfo layer in greater detail, I discuss the representation of broadcast infrastructure sites in my hometown media market of Santa Barbara, California. Using Google Earth’s navigation features, I zoomed from a national perspective to a local overview (Figure 5) of this relatively small media market of 80,000 situated on the central coast of California, sandwiched between the denser signal territories of Los Angeles to the south and San Francisco to the north. Santa Barbara has two main transmission facilities located on the mountains above the city, which are in or adjacent to the Los Padres National Forest. One is located to the northeast and one to the southeast. Both of these facilities transmit signals to radio and TV stations in Santa Barbara and to other nearby towns (such as Solvang, Santa Ynez, and Ventura) via microwave links.

By zooming in on one of these facilities, it is possible to obtain a more detailed understanding of the signal traffic in this market as icons draw attention to different structures and transmissions. A close analysis of the facility on Gibraltar Road (Figure 6) reveals there are 20 FM radio transmissions, 12 TV transmissions, eight microwave and three registered towers. Clicking on any of the balloons hyperlinks the user to FCCInfo.com, where further data about the licensee and the licensed structure can be found. For instance, by clicking on KJEE, we find a commercial FM radio station based in Montecito, California that has been operating since 1994. The station also has a microwave link that connects to its parent facility on West Carrillo Street in Santa Barbara.

By isolating different broadcast systems (radio, TV, microwave), facilities (stations, towers, links), sites (urban, rural), and processes (transmission, repeating, linking), the FCCInfo layer compels a site-specific, object-oriented, and ecological approach to the study of broadcasting. If integrated within media studies courses, the layer could help students to conceptualize and research broadcasting not only in relation to what appears in the frame or through the speaker, but also as property claims, transmission processes, spectrum and resource allocations, maintenance labour, and signal traffic. Approaching a location as a signal territory requires an investigation of the multiple signals being transmitted through the air in a location of any given moment, the technological and natural resources that are required to do so, and the spatio-temporal dimensions of transmission. Signal territories exist across urban and rural, wealthy and poor, local and national, terrestrial and circumterrestrial domains. They are fields shaped by human and non-human actors, perceptible and imperceptible activities. Signal territories cannot be seen with the naked eye; rather, they must be produced discursively or through phenomenological engagements. In general, their presence must be inferred rather than observed. By examining the FCCInfo–Google Earth interface, it is possible to perceive Santa Barbara not only as a city with a particular location, population, and climate, but also as a particular kind of signal territory.

Just as the Google Earth and FCCInfo can be thought about as producing signal territories, it is important to consider the multiple industrial and governmental entities that shape and whose interests are transparently embedded within the interface, including broadcasters, transmission facility owners, satellite and aircraft operators, geospatial imaging firms, cartographers, software developers, Internet service providers, and regulatory agencies. Indeed, the Google Earth–FCCInfo interface interweaves multiple corporate and governmental entities (broadcasters, digital corporations, satellite operators, regulators), technologies (remote sensing, broadcasting, cartography, Internet service), and discursive strategies (Earth observation, market development, and public education) and, in the process, produces maps of broadcast infrastructure in a highly proprietary fashion.

Even though Google Earth offers the basic version of its virtual globe platform as a free download, clearly none of this is free. To access the Google Earth–FCCInfo interface, users must own or have access to a computer and an Internet connection. Once this access is established, the user can navigate public domain data (the locations of all licensed broadcast facilities) gathered by a federal governmental agency (the Federal Communications Commission). This information has been re-structured as a layer by a private third party (Mertz, etc.) and can only be accessed and viewed on a privately owned yet publicly accessible Web-based platform (Google Earth), which is composited from geospatial data that has been acquired by satellites historically subsidized by taxpayers. When this public domain FCC data is accessed and viewed in Google Earth, it becomes the privately owned intellectual property of Google, despite the fact that it is layered upon geospatial imagery and distributed through Internet networks that have historically been funded by taxpayers and consumers. In this way, Google Earth functions (not unlike a Santa Barbara real estate agent) as a property development engine, since data prepared for display using the platform becomes the intellectual property of Google when formatted for and layered into Google Earth. In this sense every layer, which is typically generated through the “free labour” of private individuals, non-profit organizations, government agencies, or companies, is not only an inscription in Google Earth, but also is an investment in and gift to Google, reinforcing Siva Vaidhyanathan’s (2011) provocative contention that the world is undergoing the Googlization of Everything.

While all U.S. broadcast facilities are now fully displayed in Google Earth, interestingly, the array of facilities, data centres, and server farms that makes Google Earth possible is not organized as a layer, hinting at the proprietary and strategic nature of this information—and the risks that such a layer might pose to Google’s security, particularly given the world’s growing dependence on Google’s search engines, clouds, mapping, file sharing, and email services (Hillis, Petit, & Jarrett, 2012). Google Earth is constituted through an expansive assemblage that includes aerial and satellite imaging technologies, transmission facilities, data storage centres, computer processing, software and applications, high-speed Internet networks, lands, spectrum, and myriad forms of labour.

Though Google Earth’s infrastructure has not been organized as a layer, it is possible to view some of Google’s data centres in Google Earth. Overhead views of facilities in Lenoir, North Carolina, and Council Bluffs, Iowa, resemble military installations enclosed by security gates, echoing the “closed world” discourses that characterized visions of networked computing during the Cold War (Edwards, 1997). In 2012 Google opened up some of these sites and released a series of lustrous, colour photographs of its data centres’ interiors.² These photographs enable viewers to peer inside Google’s infrastructure sites in Iowa, Oregon, Georgia, South Carolina, Oklahoma, Belgium, and Finland and vibrantly showcase the blocs of servers, fibre optic cables, LED lights, and colour-coded cooling systems that support platforms like Google Earth. Not unlike the historical map of NBC’s radio network facilities, these photos function first and foremost as publicity shots. They construct Google’s infrastructure as a sleek, perfectly lit, machinic playground, but offer scant detail about its historical development, costs, and/or resource requirements.

Another important dimension of Google Earth’s infrastructure is the constellation of remote sensing satellites orbiting the Earth. In 2008 AGI Analytics released a Google Earth layer (kmz file) that maps all known satellites in orbit and refreshes their position in real time. While navigating this layer, a user can locate the very satellite(s) that acquired the image data used to constitute the Google Earth interface. In this context, Google Earth functions as a kind of overview mirror that exposes and directs attention to the orbital infrastructure that enables users to observe and navigate the world in this way (Parks, 2013). To acquire a comprehensive understanding of Google Earth’s infrastructure, then, it is necessary to piece together its various pieces and parts, which extend from data centres on the ground to satellites in orbit and from sites in Oregon to others in Finland.

While the FCC gathers data on all broadcast facilities because it licenses and ostensibly regulates them, the Commission’s authority over Google is less pronounced. The FCC has asserted very little regulatory authority over Google as the company has established monopolies over the Internet search engine and the virtual globe. A Google search of the “FCC and Google” turns up a handful of articles about Google having its hand slapped by the FCC after the company was discovered to have gathered private data from households via Wi-Fi during data collection for Street View between 2007 and 2010. Even though Google flagrantly violated privacy laws, the FCC fined the company a mere $25,000 for impeding a federal investigation by not producing information that had been requested (Streitfeld, 2012). As Vaidhyanathan (2011) suggests,

To be sure, Google has risen to dominance in a media environment characterized by massive deregulation that began during the Reagan era and has only intensified in recent years (Holt, 2011). One could read the FCCInfo layer as symptomatic of Google’s symbolic absorption of broadcasting and the diminishing power and authority of the FCC, or as the FCC’s struggle for relevance during an era in which the digital behemoth threatens to transform everything from broadcast facilities to the FCC into intellectual property that is only accessible through Google.

While the Google Earth–FCCInfo interface provokes a consideration of various proprietary and regulatory issues, it also enables users to glimpse the thousands of individual physical installations (not only from above but also often in Street View) that make up the U.S. broadcast system and compels phenomenological investigations. In this section I present findings based on a visit to a transmission facility I first encountered in Google Earth. The results of the site visit are intended to infuse the Earth-observing platform with situated knowledges and affective engagements—to “unfreeze” infrastructure (Star, 1999) using a node-centric approach (King, 2006) that explores the material conditions and affective relations that support and surround a single broadcast infrastructure site. As King (2006) observes, the affective relations that inform and emerge through infrastructure nodes can generally not be represented with the tools used by cartographers. Thus to access these affective relations, it is necessary to experiment with different modes of mapping, Earth observation, and mediation.

Inspired by the Google Earth–FCCInfo interface, I set out to develop what Kember and Zylinska call a “creative mediation”—a form of “media analysis” that is “simultaneously critical and creative” (2012, p. xvii) and that “seeks to promote the invention of different forms of engagement with media” (p. 203, italics in original). In their book Life After New Media, Kember and Zylinska (2012) challenge conventional forms of media analysis and call for “a more dynamic, networked, and engaged mode of working on and with ‘the media,’ in which critique is always already accompanied by the work of participation and invention” (p. 203). Rather than approach media as static objects, they urge us to reflect upon the vital processes of mediation, which we, as critics, are always already implicated within. They argue for “envisaging and enacting playful, experimental, yet rigorous crossdisciplinary interventions and inventions that will be equally at home with critical theory and media practice and that will be prepared to and able to make a difference—academically, institutionally, politically, ethically, and aesthetically” (p. 201).

My creative mediation began with a printed screenshot of the Google Earth–FCCInfo interface (Figure 7), which I used to guide my drive into the hills above Santa Barbara along the curvy Gibraltar road to the transmission site turn off. There I found an open gate with a sign that read NO TRESPASSING! VIOLATORS WILL BE PROSECUTED!! If I drove down the road I would apparently be breaking the law, which was ironic given that the site is owned by an organization called Community Radio, Inc. Since the gate was open and since I am inspired by the work of place hackers like Bradley Garrett (2013), I drove half a mile down the very narrow recently paved road and parked my car so that I could walk up the rest of the driveway to the facility. As I walked up the final stretch of the road, I encountered a series of signs on the fence warning “Radio frequency fields beyond this point may exceed the FCC general public exposure limit.” Other signs indicated CAUTION—HIGH LEVEL RADIO FREQUENCY ENERGY AREA—NO TRESPASSING in English and Spanish. I continued to the end of the road and discovered a second gate (see Figure 8), this one closed and locked, with posted information about the site administrator and more warning signs. I also found a wooden box next to the fence with a white notebook inside used by site workers to sign in and out during their visits. The sheet was empty. Since I could not get in the main locked gate, I started to return to my car. As I was walking I noticed a side fence had been left open (see Figure 9, top left) and decided to venture inside to get a closer glimpse of the medley of towers, antennas, and dishes on site. They appeared in different shapes and sizes and were painted matte gray or white or were aluminum so that they camouflaged in with the area and were difficult to see from a distance, calling forth the kinds of complex plays of visibility and invisibility that Hanna Rose Shell (2012) explores in her cultural history of camouflage, Hide and Seek. A rickety wood scaffolding supported some of the antennas, suggesting that the site’s elevation was not high enough for some transmission needs. I also noticed jumbled debris and unused material strewn about the site, including piping, bricks, and metal parts (see Figure 9).

While walking around I acquired a more proximate, physical sense of the site. I could hear the buzz of electrical equipment, feel the burning heat of the mid-afternoon sun baking the pavement, smell the scent of warm chaparral bush mixed with ocean air, and see massive power lines stretching across the hillsides and three large transmission towers further up the mountain. While approaching the facility I was not entirely sure whether I was on private or public land, what the penalties for trespassing would be, and whether there were serious health hazards linked to being close to transmission sites. My wandering of the site led me to think about broadcast infrastructure very differently than when I viewed it in Google Earth. Rather than a massive system that could be seen from above, the infrastructure became something that could be sensed, felt, and mediated by my own body. Mapping the signal territory in this more affective and physical way required recognizing that the site, which looked quite orderly in Google Earth, was actually quite eclectic and messy on the ground (Figure 9), a scattered archaeology of transmission equipment that was more akin to “high-tech ramshackle” or “twenty-first-century hillbilly.” The visit also challenged me to think about the biotechnical dimensions of broadcasting. This infrastructure site is not only made of metals, plastics, rubber, and electromagnetic fields, but also of soil, rock, vegetation, insects, lizards, birds, and animals, organic materials and life forms that cannot be perceived in Google Earth and are not typically thought of as part of a media ecology.

As my fieldwork around the site tuned me in to the biotechnical dimensions of this tiny sliver of U.S. broadcast infrastructure, it also enabled me to think about issues that I would never have considered by examining this site only in Google Earth. Some key aspects of the site were completely invisible in Google Earth. The series of gates and trespassing signs I encountered in geophysical space, for instance, contrasted with unfettered access to the site in digital space. The pieces of hardware represented by colour-coded balloons in Google Earth were camouflaged at the site so that they blended in with the local environs, and since they were fenced off, they were more difficult to see and identify. High-voltage currents and electrical radiation were announced as human health hazards at the site (yet were apparently safe for other wildlife in the vicinity), but were absent in Google Earth. The contingency of this facility’s operations upon other infrastructures (see Figure 10) such as the electrical grid, water sewer system, and fixed telephony was not evident in Google Earth but became clear during the site visit. And, finally, the sound of the broadcast infrastructure—the persistent ambient buzz of electrical activity—which provides an acoustic sense that something dynamic is happening, is altogether missing from the Google Earth interface. While I was there I could not help but imagine the initial installation of equipment at the site, the upgrading of that equipment, and physical manipulation of transmitters and antennas that make broadcasting possible. In this sense, the site visit also evoked a history of labour—electrical engineering and maintenance—that is increasingly being replaced by sensors and computers that can be used to monitor such sites remotely.

Most of all, the site visit led me to consider the importance of fieldwork as a vital method for investigating media technologies. In an era in which the world can be navigated in Google Earth and seen in Street View, accessed through Google’s search engines, or streamed from YouTube’s archives, media researchers seem to spend more time than ever at screen and computer interfaces, and less and less time venturing out in the field, wherever it may be. It is worth thinking carefully about how our critical practices and methods are shaped and informed not only by our social subjectivities, technological literacies, and intellectual training, but also by the geophysical positions we are willing to engage with, occupy, or explore. How often do we move beyond what Mark Andrejevic (2009) calls the digital enclosure or what Nick Couldry (2003) calls the media centre? If we are willing to take seriously the fusion of the biological and the technological, it is important to consider not only how consumer electronics become human prostheses but also how automated facilities on the outskirts of cities that are dug deep into the dirt and surrounded by plants and wildlife—seemingly in the middle of nowhere—are integral to broadcasting in the digital age. In this sense, my project is invested in what Diana Coole and Samantha Frost (2010) refer to as “new materialisms,” which, as they explain, bring “biopolitics, critical geopolitics, and political economy together with geneaologies and phenomenologies of everyday life” (p. 28). By thinking and moving between historical network maps, the Google Earth–FCCInfo interface, and a specific infrastructure site, I began to recognize and investigate the ways in which resources such as lands, water, electricity, heavy metals, and other materials are organized to transmit signals and, in the process, turn the Earth’s surface into signal territories. As Maxwell and Miller (2011) insist, “understanding media requires studying them up, down, and sideways …” (p. 17).

In this article I have explored broadcast infrastructure through three Earth-observing practices—historical network cartography, the virtual globe, and fieldwork. In the process, I have suggested that the mapping of infrastructure can be used to analyze issues of ownership, intellectual property, privatization, regulation, and materialism. Though infrastructures have become part of the built environment and surround us in daily life, publics are socialized to know very little about their development, operations, and resource requirements. By highlighting multiple modes of infrastructure mapping, I hope to encourage infrastructure re-socialization—a technological literacy project that urges publics to notice, document, and ask questions about infrastructure sites and become involved in discussions and deliberations about their funding, design, installation, operation, and use. Most people in the U.S. are familiar with commercial television and radio, but relatively few have ever visited a TV or radio station, attended municipal hearings to comment on cable operator franchises, or followed FCC proceedings on spectrum auctions and allocations. There is a serious disjuncture between the amount of public time and attention dedicated to screened entertainment and the amount of public time and attention dedicated to understanding the facilities and sites that distribute the signals that become entertainment. In public life and critical media studies alike, we need a firmer understanding of where and how media distribution has historically occurred in the U.S. and beyond (Sterne, 1999). This will involve studying media technologies in relation to a wider array of industrial sectors such as real estate, energy, electrical engineering, architecture, design, public health, and security.

This project of infrastructure re-socialization also requires developing a more robust awareness of the temporalities and dynamism of infrastructures. It is important not only to know where infrastructure sites are, what they are made of, and how they function, but also to specify the unique temporalities of their historical emergence and ongoing operation. Such a process could begin, for instance, with a consideration of Google Earth’s own movement through the Internet. It takes about five minutes to download and install the Google Earth application and the FCCInfo layer on a computer using a high-speed Internet connection. It took several decades to develop and deploy the satellites and aircraft used to acquire the image-data that makes up the Google Earth platform. It took several decades to build the broadband Internet infrastructure through which this platform and its layers move. It took a century to construct, expand, and refine the U.S. broadcast system and to survey and map its multiple locations. And it takes seconds to access and navigate these sites in Google Earth. In short, the Google Earth–FCCInfo interface is symptomatic of a host of durational acts that are effaced and folded into the neatly framed experience of world navigation at one’s fingertips.

To generate a deeper understanding of such temporalities, we need critical models or creative mediations that approach Earth-observing platforms as time-based media systems. For instance, it would be helpful to see a time-lapse map of Google Earth downloads around the world, a visualization of user activity in Google Earth during a certain period of time, or a real-time display of all Google Earth updates. In relation to photos of broadcast infrastructure sites we might similarly ask: How long does it take a signal to move from site to site? How long have specific towers and facilities been in place? How long did it take to construct them? What was happening in the site before the station was erected? What is happening at that site now? Since historical network maps, virtual globes, and photographs efface the movements and dynamisms that are inherent parts of infrastructural histories and operations, there is a need for creative mediations of infrastructure that foreground temporal relations.

Finally, rather than approach Earth-observing practices only as representing infrastructure sites and processes, we need to understand these practices as performative acts that have different relations to time, which need to be specified and considered. That is, the drawing of a map of telegraphy stations, the compositing of aerial and satellite images and layering of icons to construct broadcast infrastructure in a digital Earth, or the visiting and photographing of an infrastructure site are all performative acts in time that are aimed not only at “representing” but also at enabling certain capacities and potentials. As such they are what Kember and Zylinska (2012) refer to as vital mediations, wherein “mediation becomes a key trope for understanding and articulating our being in, and becoming with, the technological world, our emergence and ways of interacting with it, as well as the acts and processes of temporality stabilizing the world into media, agents, relations, and networks” (p. xv). The more ways in which infrastructures can be perceived, sensed, felt, and understood as part of life, the more potential there is for publics to reflect upon and intervene in their complex materialities and temporalities. This article closes, then, with a call for further creative mediations of the infrastructures that surround us.

I would like to thank Jon Jablonski, Patrick McCray, and Barney Warf for research suggestions, and Chris Russill for helping to inspire me to write this essay and providing vital feedback. Other versions of this paper were presented at the Society for Cinema and Media Studies conference and the Conditions of Mediation: Phenomenologial Approaches to Media, Technology and Communication ICA pre-conference at Birbeck University of London, and I am grateful for attendees’ questions and comments.