Windows into the Soul or the Clouded Glass of Surveillance Studies

Conceptualizing the New Surveillance

Windows into the Soul is billed as Marx’s magnum opus, the crowning achievement of his long and accomplished scholarly career. Marx began publishing scholarly research in the late 1960s with influential work social movements and collective behavior (Marx, 1967, 1970; Marx and McAdam, 1994; Marx and Wood, 1975). From here, he completed a path breaking study on undercover policing (Marx, 1974, 1980, 1988), which led to his broader focus on surveillance in the last three decades (Marx, 1998, 2002, 2003, 2005, 2009, 2013). Windows into the Soul picks up where Marx’s last single-authored monograph, Undercover, left off. Marx (1988) concluded his now-classic examination of police surveillance by defining a new surveillance or the then-emergent forms of impersonal and technical monitoring that targets personal information, rather than the physical body (pp. 206–234). Writing at the cusp of the internet age, Marx’s argument was prescient. Undercover became a foundational text of what is now called surveillance studies.

Windows into the Soul endeavors to provide a comprehensive framework to analyze the new surveillance. The first two sections of the book constructs an exhaustive typology of new surveillance as process defined by the contexts and structures of surveillance, the means of surveillance, the goals of surveillance, and the types of data collected. Each of these moments of surveillance breaks down into various subcomponents. Regarding means of surveillance, for example, Marx asks us to consider the comprehensiveness of surveillance, who collects the data, and where and with what technology. He is particularly forceful in challenging what other surveillance scholars have criticized as the tendency to present surveillance with “an undercurrent of technocentric gloom … often combined with a theme of progression, manifest in recurrent suggestions of surveillance undergoing greater expansion, intensification and penetration” (Ball and Haggerty, 2002: 133).

To this end, Marx criticizes what he sees as the uncritical application of Foucault to questions of surveillance, which leads scholars to “collapse or produce the more general process or activity of surveillance to just one context—organizational—and to one goal, which is control” (p. 64). Instead, Marx offers a dozen alternative goals for collecting personal information, ranging from compliance to discovery to profit to self-knowledge. As a related point, Marx contends that “understanding and evaluating surveillance require attention to the setting” (p. 86). For Marx, norms follow from contexts, hence “surveillance is neither good nor bad but context and comportment make it so” (p. 284). From here, he identifies “four major context for surveillance … contracts, care, and coercion, and a residual free-range form of surveillance apart from an organization, group, or role where information is simply available with no need for a mechanism or motivation to reveal it” (p. 10).

The first section is by far the strongest portion of the book. Marx offers a comprehensive middle range theory that dissects surveillance as a process that differentially affects individuals and shapes situations and organizations. Theoretically, he draws on the symbolic interactionism of Goffman and, more implicitly, political process theory. His conceptualization is thorough. His sections on identity and layers of personal information (pp. 90–108), the moves and countermoves that shape the surveillance encounter (pp. 142–148,168–172), and techno-fallacies (pp. 267–275) stand out. While one can quibble whether some of his distinctions are duplications or even subtle gradations of type, not fundamental differences in kind (Guzik, 2017; Regan, 2017), Marx’s thorough effort to parse the intricacies of different surveillance processes is admirable and, often, suggestive. Unfortunately, Marx never uses his proposed conceptualization to approach any empirical problem. He leaves this task to others and only “offer[s] a systematic way for grounding (and comparing) ethical and policy judgments about particular tactics and practices” (p. 11, emphasis added).

Speculative Fiction

Rather than deploying his own framework in a research project, Marx turns to speculative fiction. The third part “offers stories as case studies. These involve surveillance used by an employer, parent, voyeur and government” (p. 173). For Marx, these stories “might be true even if it could not be empirically accurate. While the events in these chapters did not occur together at the imaginary times and places described, they could happen. They may be fiction, but they are not quite science fiction” (p. 175). He claims “The complexity of the situation made me do it” (p. 176). These “stories as case studies” make it difficult to evaluate Marx’s theoretical propositions. They confuse theory with empirical research and collapse both into satire.

His chapter on parenting, for example, revolves around “a satirical statement from PISHI (Parents Insist on Surveillance Help, Inc.)—a fictitious social movement dedicated to protecting children through the use of technology” (p. 201). In Marx’s construction, this imagined social movement shifts attention “to the relatively more personal, diffuse, informal, dutiful, and caring surveillance of dependent children in the private setting of the home” (p. 200, original emphasis). To claim that the family is “a surveillance context” comparable with conventionally understood sites of surveillance such as the state or workplace, Marx asserts that the “family is the ultimate, total institution” (p. 211). Unlike prisons or work camps, however, the family is not “a place of residence and work where a large number of like-situated individuals, cut off from the wider society for an appreciable period of time, together lead an enclosed, formally administered round of life” (Goffman, 1961: xiii). Households, in sharp contrast to total institutions, are woven into overlapping networks of sociality: civic and faith organizations, the workplace and market, and many administrative arms of the state.

No doubt, there are practices structuring families that can usefully be labeled “surveillance.” There are even surveillance practices in which kinship relations become important components of surveillance networks. But it is unclear whether the family is the appropriate unit of analysis to understand these examples. Indeed, many of the surveillance systems that Marx rolls into his fictional case study such as home drug tests are linked to external forces that impinge upon the relative autonomy of households: the increasing marketization of social relations and the growing reach of criminal justice systems. Moore and Haggerty’s (2001) study of home drug tests, for example, links home drug testing to “neoliberal trends towards mobilizing private entities like the family to engage in regulatory practices that were previously concerns of the state” (p. 377). As Silverstein (2001) showed with probation, the softening of surveillance often takes the form of state-family partnerships: “The family … become cajoled by the state to participate in control and supervision of troublesome populations that historically were solely a state responsibility” (p. 417). In their study of RFID monitoring of children, similarly, Ema and Fujigaki (2011) find that parent’s support such surveillance took place in wider context informed by “top-down governmental projects [that] promoted child monitoring systems in a series of social experiments on a ubiquitous network society” (p. 134). While parents did join teachers in providing the project some bottom-up legitimacy, it would be an overstatement to posit the family as the privledged site of surveillance. In this case, as with many others, kinship relations become integrated into surveillance networks but they do not comprise the organizational center of these systems. In general, these “stories as case studies” are unpersuasive and, in the case of Marx’s imagined “family-as-total-institution,” directly contradict findings from the relevant literature.

Overreaches like these are a predictable consequence of Marx’s curious methods. He justifies his choice of fiction as a form of ideal type. He writes:

Another type of fiction well known to the social scientist is the ideal type, as suggested by Max Weber … This ideal type makes greater claim as to its reality, even if in its pure form it cannot be literally found. It is a synthetic mental construct emerging from the empirical, yet going beyond it to distill the central features of the phenomena in its purest form. (p. 175)

Here, Marx conflates a heuristic with a fiction. In one of the foundational texts of sociological methods, Weber (1949) defines ideal types as “a mental construct for the scrutiny and systematic characterization of individual concrete patterns which are significant in their uniqueness, such as Christianity, capitalism, etc.” (p. 100). In this regard, Marx is correct to label and contrast “new” and “traditional” surveillance as ideal types (p. 54). They are “mental constructs” that enable Marx to scrutinize the unique character of “traditional” surveillance that tracks physical bodies and “new” surveillance of personal identifying information. However, Weber (1949: 102) is clear that this “procedure gives rise to no methodological doubts so long as we clearly keep in mind that ideal-typical developmental constructs and history are to be sharply distinguished from one another”. He warned that the “danger of this procedure which in itself is entirely legitimate lies in the fact that historical knowledge appears as a servant of theory instead of the opposite role. It is a great temptation for the theorist to regard this relationship either as the normal one or, far worse, to mix theory with history and indeed confuse them with each other” (Weber, 194: 102, original emphasis). Marx clearly did not heed this warning.

Marx’s turn to satire is especially disappointing, given Marx’s decades-long work on surveillance. He tells the reader that the “data for this book come from observations, interviews, the academic literature, government reports, periodicals, court records, and popular culture.” He mentions he interviewed “more than four hundred people” and, while he makes anecdotal reference to some interviews, he does not analyze this primary data in any systematic way (p. 7). Marx claims the mantle of “social science,” while, simultaneously asserting that Windows into the Soul “is not the type of social science that systematically tests ideas with freshly plucked quantitative data. Nor is this book a theoretical statement in the sense that it offers a tight system from which one can logically derive propositions to be tested. Nor does it offer a single, hard-driving argument.” Instead, the book is “a soft-driving argument that identifies questions central for explanation, evaluation, and regulation and parses empirical possibilities into categories involving types of behavior and four basic surveillance concepts” (p. 8). Whether the reader finds these satires suggestive or tedious, they do not meet the basic standards of empirical social science. Simply put, Windows into the Soul would be a much more compelling book if Marx had used his accumulated research and knowledge of the surveillance literature to systematically apply his processual and interactive theorization of surveillance to the problems of actually existing “surveillance societies.”

Surveillance as Care or Surveillance and Power?

These loose and speculative methods are not offset by sophisticated theorizing. Marx does not seriously engage with enough of the relevant thinking on surveillance to make Windows into the Soul a work of “grand theory.” In particular, his dismissal of Foucault, the theoretical center of surveillance studies, limits the scope and ambition of the project. While Marx’s quick critique of Foucault resonates with some early readings (see Garland, 1990), today, it is hard to sustain Marx’s contention that “Foucault, and many of those uncritically under his spell, collapse or reduce the more general process or activity of surveillance to justice context—the organizational—and to one goal, which is control, a term often used interchangeably with domination and repression” (p. 64, emphasis added). This criticism ignores Foucault’s (1995) basic contribution: to rethink power as a productive force that shapes and molds subjectivity, “that dissociates power from the body,” and “turns into it ‘aptitude,’ a capacity, which it seeks to increase” (p. 138). Indeed, the operations Foucault famously called “disciplinary power” entail the very goals that Marx posits as alternatives to Foucault’s “control” such as compliance,” “verification,” “documentation,” and “self-knowledge” (p. 65). This poststructural theorization of capillary powers places Marx’s conceptualization of surveillance within an expanded analytic that explodes the binaries of self/society or structure/agency and considers how social relations produce subjectivities. In short, Marx could easily be read as providing the conceptual rigor to link to Foucaultian theory to the empirical study of concrete instances of surveillance.

This potential resonance with Foucault’s argument also becomes clear when Marx makes historical claims that are broadly consonant with Foucault’s work. For example, Marx contends that “The current softening (and feminization?) of surveillance (and its frequent corollary, control) involves, if not a marriage, at least a tense cohabitation of science and technology with the ideals of the modern democratic state that appeared with the French and American revolutions” (p. 116). This is more or less the basic thesis of Discipline and Punish: the messy, bodily violence of pre- and early-modern forms of power by softer and “rational” forms of surveillance, regimentation, and “control,” what Foucault (1995) described as a “subtle, graduated carceral net, with compact institutions, but also separate and diffused methods” (p. 297). In dismissing of Foucault (and implicitly Deleuze), Marx refuses to build on or otherwise engage with perhaps the most dominant theoretical tradition within surveillance studies (for reviews of this work see Caluya, 2010; Lianos, 2003; Murakami-Wood, 2007). This is a lost opportunity to bring greater coherence to surveillance studies by advancing a cumulative theory that recuperates or otherwise builds upon Foucault’s influential work.

Seriously engaging the research and thinking about the power and subjectivity that revolves around Foucault’s work would have challenged Marx to take a more complex view of the relations among social structures, historical processes, and the self. It could have forced a reckoning with the historical specificities of existing power relations in ways that the mainline of Foucaultian surveillance studies largely has not. ¹ At the very least, it would muddle the seemingly clear distinction Marx draws among different “surveillance contexts,” and particularly between “consent” and “coercion”—categories that the massive literature on hegemony, for example, has long conceptualized as dialectically interpenetrated moments in the development of social-formation, not dichotomous contexts (Gramsci, 1971; Morton, 2007; Thomas, 2009). Moreover, a deeper engagement along these lines would also raise questions about “care” as context of surveillance, a notion that is not unique to Marx. Some surveillance studies scholars advance the notion of “surveillance-as-care” as part of a new ethics that will recuperate existing surveillance technologies and practices into a humanism ethos, informed by Christian theology (Stoddart, 2011). They “ask what might happen if surveillance were guided by an ontology of peace rather than of violence, an ethic of care rather than control, an orientation toward forgiveness rather than to suspicion” (Lyon, 2001: 153).

Proponents of this view, however, fail to consider that “surveillance-as-care” already exists and is compatible with more traditional “surveillance contexts” like “control.” Indeed, Neocleous (1996), has theorized 19th century “poor relief” in the UK—a possible instance of “surveillance-as-care”—as a form of “political administration” that subsumed the emergent working class within institutionalized politics, solidifying bourgeois rule and the order of private property. The stabilization of capitalist society in the advanced industrial countries entailed a massive project of “surveillance-as-care” in the form of the health, welfare, and safety initiatives of increasingly ambitious administrative states. Neocleous’ work is not just a relevant finding. It is an alternative agenda for the study of subjects claimed by surveillance studies. Like surveillance studies, this line of thinking developed in response to Foucault but, unlike surveillance studies, which remains predisposed to more presentist cultural studies approaches, Neocleous places Foucault in conversation with Hegel, Marx, and Gramsci to develop a Hegelian history-as-theory of what he called “political administration.” This study, along with related work on police power (Neocleous, 2000) and pacification (Neocleous and Rigakos, 2011; Rigakos, 2016; Wall et al., 2017), represents a different answer to the same question that prompted surveillance studies, one that is more attuned to the historical specificities of capitalist social relations. From this perspective, surveillance (and the related concept of privacy) forms part of the discursive and structural integument that provides coherence to particular social relations: a social order organized around private property and the increasingly acute polarization of wealth, which leaves most of humanity prostrate before massive accumulations of capital and power (Henry, 2013; Neocleous, 2002).

Definitional Problems and Political Confusion

This ahistorical view of power is rooted in a tendency among surveillance studies scholars to bend the definition of surveillance to the point of meaninglessness. Here, Marx is the most extreme example of a common problem. Marx defines the “the most general level” of surveillance “as a regard for or attendance to person or factors presumed to be associated with a person” (p. 15). Hence, Marx is redefining many acts of perception and cognition as “surveillance.” This definition downplays the essential core meaning of surveillance: hierarchical observation and control. The Oxford English Dictionary shows that “surveillance” first appears in English in 1799 and is used throughout the period of French Revolutionary and Napoleonic Wars to connote “the watch or guard kept over person.” After the Second World War, the word came to be used as a descriptive adjective of various types of surveillance equipment. When surveillance studies scholars stretch the meaning of surveillance outside of questions of hierarchy and control it leads to theoretical and political confusion.

Theoretically, this expanded definition conflates surveillance with “sociability” or the type of playful interpersonal communication that Simmel (1949) identified as an “emancipating and saving exhilaration” form of social interaction (p. 261). Most importantly, sociability can level hierarchies and temporarily create “an ideal sociological world” where “the pleasure of the individual is always contingent upon the joy of others” (p. 257). At the same time, Simmel does not romanticize interaction. Sociability is fragile. It can easily become “entangled with real life” and become complicated “purposive, objective content” of his more formal economic and political associations (pp. 257, 258). Sociability—and, perhaps more importantly, the failed moments where power differentials overwhelm sociability and impose hierarchy on social interaction (being put in one’s place)—are different processes than surveillance. Conflating the two is unhelpful. For example, surveillance studies define forms of sociability engendered by interactive and social media as “lateral,” “reciprocal,” or “participatory surveillance” (Albrechtslund, 2008; Andrejevic, 2006; Marx, 2012). This conceptualization draws a false equivalency between, say, the sociability of “friends” on Facebook and commercial surveillance that has transformed social media platforms into billion dollar firms (Fuchs, 2011). It sets an expectation that interaction will be tainted by the diffuse surveillance of norms. It works against appreciating new forms of sociability and (collective) agency that are engendered by digitally mediated relations.

Thinking in terms of surveillance also makes it harder to conceive of non-hierarchical relations. Consider the notion of “sousveillance” (Mann et al., 2002), which connotes a reversal of surveillance hierarchies and potential imposition of transparency and accountability from below. While a compelling notion, the political possibilities embedded in this concept are ultimately truncated by the way it is shoehorned into the larger concept of “surveillance.” As a result, “sousveillance” restricts the horizon of our political imagination to defensive measures, implicitly conceived within a liberal discourse of rights. Sousveillance makes demands on power. The same holds for surveillance scholars reflections on symmetry and empowerment. In their attempt to create a more “nuanced” sub-discipline, surveillance studies scholars have sought to move beyond pessimism that is associated with surveillance. Monahan et al. (2002), for example, asked to consider the way surveillance is not only something that dominates those being watched. It also empowers the watchers. They provocatively ask how we could bring “symmetry” to surveillance: “how might traditionally marginalized groups use surveillance to challenge their positions of marginality? Or, even broader, how can surveillance be designed, employed, and regulated to contribute to democratic practices and/or the social good?” (p. 107). Framed exclusively in terms of surveillance, this provocative question leads to a political dead end: a demand for “better” surveillance. Hence, Lyon (2013) called for “democratic surveillance” that is reciprocal, participatory, proportional, and decentralized.

The discussion has become absurd. If democracy is more than a form of government and, instead, connotes a “mode of associated living predicated upon conditions of social equality, along lines of race, class, gender, and other categories of difference,” then it is misguided to ask “What manifestations of surveillance support democracy?” (Monahan, 2010: 92, 101). Surveillance—by definition—only exists within hierarchical relations of domination and subordination. Surveillance can be productive or “caring” (in a paternal manner) but it can never be liberating. In the capitalist context of private property and structured subordination between various laboring populations, “democratizing” “surveillance” would mean transforming global information infrastructures into a commonly held and equally accessed system that transcends and obliterates the idea of “surveillance.” In a true democracy, “surveillance” would be an anachronism. The problem here is not the ethical call for more democratic ways to produce and transmit useful information. The problem is theorizing it on a spectrum that includes disciplinary surveillance of the workplace, the market, social service agencies, and the security apparatus. Rather than making demands for “democratic surveillance,” we should be asking what an administratively legible and collectively managed reproductive infrastructure looks like—and, more importantly, what emergent social relations may prefigure viable alternatives to existing arrangements. Surveillance studies, largely, obscures these important questions.

As evinced in Windows into the Soul, surveillance studies can still be described as the over theorized, non-cumulative, and unduly abstract field that Gary Marx criticized a decade ago. Marx’s magnum opus offers yet another conceptual framework with no systemic effort to use the proposed theorization to analyze data. He dismisses the main line of thinking within the subfield and misses opportunities to explore potential points of shared concerns among seemingly rival interpretations. Most importantly, he takes an ahistorical view of power and surveillance that leads to analytic and political confusion. Surveillance becomes sociability, democracy becomes “better” surveillance, and a utopian vision for non-oppressive relations becomes a dim light on the horizon.