1
The hearth of darkness
Living within occult infrastructures
Stephen C. Slota, Aubrey Slaughter, and Geoffrey C. Bowker
Infrastructure is the story of what happens when the “real story” is taking place. Behind the spectacles of permanent technocultural revolution, effervescent personal expression, and the vertiginous proliferation of new modes of expression lies the operation of physical, computing, organizational forms of action (dispositifs). Often, the most visible products of our culture reflect and represent substantive changes in infrastructure and infrastructural capacity – behind the scenes of film production are attendant infrastructures of chemical reactions and assorted associated refinements for film development, broadly available and reliable electricity, all the industry and science that went into the production of electric lighting (not to mention the substantial complexity of audio recording) (Sterne, 2006). These are prior to the broader cultural infrastructure that supports and responds to particular forms of expression. Silent films, for example, presume a significant visual as well as written skill set for the audience (cf. Deleuze, 1986). The early history of US national broadcast television – especially the content controls enforced by the Federal Communications Commission (FCC) – speaks to the notion of resource availability and societal responsibility (Aufderheide, 1991). Similarly, the apportionment and use of radio spectra depend to some extent on both the technology that is used to broadcast and broader societal assumptions as to the best use of that quite limited spectra (Aufderheide, 1999).
In recent times we are experiencing an unusual proliferation of infrastructural goods. Where previously infrastructure that spanned broad distance required significant investment and management of physical space and material, extant communications infrastructure provides a ground in which new infrastructure might be more easily built and integrated. The infrastructure of the Internet, for example, was largely built upon the existing infrastructure of telephone communication, which itself was partly built on the telegraph, which followed and was built alongside roadways and canal infrastructure (cf. Castells and Blackwell, 1998; Edwards et al., 2007; Edwards, 2010). New infrastructures emerged out of the generic, formal http: protocol, defined largely by their role relative to particular activities – often called “platforms” (Gillespie, 2010). Infrastructure, as defined by Star and Ruhleder (1996), is embedded and transparent; it exists (metaphorically) within or underneath other social, technological, and built worlds and does not need to be reconsidered at the moment of each task it enables. Infrastructure is learned as a part of membership and linked with the conventions of practice therein and embodies some set of standards. It is built over the top of an installed base, becoming visible upon breakdown, and is of a scale or scope that exceeds a single “site” – however that might be conceived (Star & Ruhleder, 1996). This definition of infrastructure allows us to consider the complex plurality of social, organizational, and physical infrastructures that together inform and support our day-to-day activities – something that expansively is referred to as knowledge infrastructures (Borgman et al., 2013). It is important to remember, here, that infrastructure occupies not just a material place but also a social, political and organizational one – hence the emphasis on knowledge infrastructures – and can in fact be almost totally immaterial (Karasti et al., 2016; Borgman et al., 2013). It is difficult to argue, for example, that the TCP/IP routing and addressing protocols are not infrastructural to Internet communication, but it is equally difficult to understand those objects according to their material properties alone (DeNardis, 2012). One of the major methodologies for social scientists interested in infrastructure studies is the notion of “infrastructural inversion” (Bowker et al., 2009), where the supportive technologies, standards, and material are intentionally and specifically foregrounded in order to explore their effect on work, expression, or policy.
The core reason for making this move is the postulate that infrastructure matters. Infrastructure is not a neutral background enabling an infinite set of activities; infrastructure holds values, permits certain kinds of human and nonhuman relations while blocking others, and shapes the very ways in which we think about the world. This is evident in Veyne’s (2005) proposition that it is impossible to trace the development of the concept of democracy over time because “democracy” changes fundamentally with new infrastructural developments. Meeting in an agora or town square to determine matters of concern is fundamentally different from holding discussions through print media or following a 24-hour news cycle on electronic media (Boczkowski, 2005). Or as Richard John (2009) has pointed out, we could not have the American state without the cheap circulation of newspapers nationally through the infrastructure of the post office, permitting the engagement of a national debate among residents of the otherwise remote states. Infrastructural inversion has been applied to work in health care (Jensen, 2008), water management in Thailand and the role of rice production (Morita, 2017), sociotechnical analyses of Wi-Fi (Mackenzie, 2005), as a generative resource in the digital humanities (Kaltenbrunner, 2015), and in studies of policy and development (Pelizza, 2016; Suarez-Villa, 1997; Zick, 2013; Korn & Voida, 2015; Hetherington & Campbell, 2014). It, however, seems most comfortably applied to the area of science studies for its revelations on knowledge production practices (Mayernik et al., 2013; Georgiadou et al., 2009; Lee et al., 2006). Throughout this writing, we “invert” the notion of infrastructure itself to consider it as a relational quantity – something that exists as infrastructure only in relation to particular activities, modes of work, expression, and performance.
We examine how systems and technologies infrastructural to new media such as social media platforms, recommender systems, and entertainment apps serve to inform certain kinds of performance of the self, and we examine broadly the concept of infrastructure as it relates to digital media and communication. Infrastructural systems are deeply embedded in a wide ecology of social interactions, political realities, and assumptions among its users with respect to affordances (Gibson, 1979). As we explore the impact and consideration of the place of knowledge infrastructure in our daily interactions with media, we are talking about effects and assumptions infrastructural to our current interactions and are describing a relational, changing system impacted by a wide array of social, political, and technological factors. In the following sections, we discuss the specific example of the End User License Agreement (EULA), exploring how our basic interactions, legal rights, and assumptions of value are coded into an occluded, often-unread, potentially unenforceable document. Infrastructural to our interactions with service providers, the law, and by extension to the communities enabled and propagated within our services, the place of the EULA is nevertheless in flux – what is reality one day might be history a few days later. The deployment and relationality of infrastructure are subject to relatively rapid changes – as soon as it must be reconsidered at the moment of work there is a substantive difference from the initial infrastructural relationship.
Infrastructure encodes values, influences behavior
Critical reflection on the design of information systems and other artifacts shows that humans embody their values and morality, often unconsciously, in the things that they create (Winner, 1980; Latour, 1992; Nissenbaum, 1998). These values may be intentionally designed into the physical state of the artifact or system (Flanagan et al., 2008; Friedman et al., 2002) or be observed as resulting from a myriad of social factors (Pinch & Bijker, 1984). These values can produce bias (Friedman & Nissenbaum, 1996) or otherwise be seen to have and carry politics of their own (Introna & Nissenbaum, 2000). Successful infrastructures serve those with a variety of values but may prioritize certain values in their design (Knobel & Bowker, 2011). For example, mobile technology that automatically reports your location through GPS to your friends and family values connectedness and intimacy above privacy. Although these value propositions are evident in the objects themselves, often they are the result of unconscious assumptions on the part of the designer, making it quite difficult to avoid their potential negative impacts on the quality of life (Introna & Nissenbaum, 2000).
Engagement with media in different platforms has significant and unconscious effects on user behavior (Valkenburg & Peter, 2013; Greenfield, 2014; Johnson et al., 1997). Similarly, with the use of hidden A/B testing, where users are unknowingly presented with one of two different versions of a site or service in order to test the effect of that design choice on use, the potential for emotional and social manipulation by those managing platforms (Kramer et al., 2014) is increasingly problematic – the medium itself is designed to disguise or otherwise impede awareness of the variety of ways in which behavior is being manipulated, information and inference about an individual are collected and put to use (Schüll, 2014), and certain interactions are trivialized while others are highlighted (Metcalf & Crawford, 2016). In design literature, such work is often characterized as a “nudging technology,” which is defined by Thaler and Sunstein as “any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives” (2009, 6). A nudge, in the sense of Thaler and Sunstein, is not a mandate – it is the careful curation of choices presented to users of a system through their notifications, menu layout, and other design aspects of the overall information architecture of that system (cf. Obama, 2015). Nudging is designed to act aggregately over a population, rather than deterministically over a single user (Spiegler, 2015; Marteau et al., 2011).
Nudging technologies play upon the human machine and seek to create certain conformities of behavior according to unconscious responses. This is performance of the human as machine in a way that can only be done in a specific media regime – requiring both the tranching of behaviors sufficient to predict responses and the willingness to engage with a system, device, or environment on “its own terms.” That is to say, the system performs a mode of engagement and interaction that works simultaneously to encourage particular performances of the individual as machine, the individual as a collective uniqueness, the individual as self-representation, and the individual in the production of narratives of the self as presented through media conforming to that system.
This is transhumanism in the mode of De Chardin (1964), or Kurzweil (2000), and Butler in Erewohn (1974) – exceeding the performance of the self through technologies of representation, sharing, and re-presentation of that self: “Firstly, the power of invention, so rapidly intensified at the present time by the rationalised recoil of all the forces of research that it is already possible to speak of a forward leap of evolution” (De Chardin, 1964, 305). Increasingly, we self-curate, creating an archive of the self-reflective not just of a self-product but also of the system in which we are performing ourselves. This is integrally a new mode of being human – one where the global and local are brought to the same level through communication technology; where marginalized populations are able to constitute themselves in ways that would not have otherwise been possible (Cormack & Hourigan, 2007); and where the formation, dissolution, and reconstitution of communities are far more rapid than before (Alonso & Oiarzabal, 2010).
What we need to understand people’s politics is to see what they don’t consider to be political, what is a “natural” form of governance, whether that is democracy, capitalism, or nature-red-in-tooth-and-claw (Schmitter & Karl, 1991). Even before we were “nudged” by technology and intentional design, social cues, interactions, and reactions served a similar role in unconsciously modifying our behavior. We do not consider our unconscious responses to be a part of infrastructural systems; we prefer to believe that we interact with systems agentially (Luhmann, 2000). But people’s reactions and their unconscious, reflexive, immanent responses are perhaps much more relevant to the builder of infrastructural systems than anything else (Tosa, 2010). These are perhaps the most vital elements of infrastructure because they cannot be easily changed by either users or designers, but still invisibly inflect all forms of human-infrastructural relations. In many ways, the unconscious, reactive human is itself an infrastructure of relevance to system designers, particularly in the field of human−computer interaction (Zafar et al., 2017; Van House, 2011; Karashima & Ishibashi, 2007). Now, we characterize the influence of the ‘machine’ on the human in terms of the characteristic ability of infrastructure to ‘fade into the background’ of our daily lives in terms of the occult. This is not (just) the occult in the sense of the magical, the unexplained, or the ineffable but also in the sense of that which is hidden, unseen, or blocked from view. In various ways, we find infrastructural relations both fading from view in daily practice and as actively occluded.
The invisible actor: infrastructure as occult, occulting, and occulted
T...