Audiences in an Age of Datafication: Critical Questions for Media Research

Lessons From Audience History

Tensions between media power and audience agency have always been theorized, more or less explicitly, more or less adequately, against the backdrop of wider socio-political transformations (Katz 1980; Livingstone 1993). While an even longer history of audiences can be and has been told (Butsch 2008; Fornäs 2014), the framing of audiences in relation to their times—and especially the assumption of a gullible audience at a time when misuse of media power was widely feared—can be illustrated by what Pooley and Socolow (2013) termed “The Myth of the War of the Worlds Panic.” Conducted at a time of rising political crisis followed by world war, Hadley Cantril (1940) studied the audience reception of what we would now call “fake news” in H.G. Wells/Orson Welles’ radio drama announcing a Martian invasion. But empirical analysis revealed, notwithstanding the widespread if mythical claims of a gullible audience, that the vast majority of the audience exercised critical literacy of one kind or another to check, deconstruct, contextualize, and resist unreasonable media influence. Contestations over what is assumed or learned about audiences can run deep: Herta Herzog’s (1944) early reading of soap opera fans, also from the Columbia School, is another case in point (as discussed by Brunsdon 2000). My first lesson from history is that, like any other text, what audiences say and what is said about them is open to multiple readings; we must, thus, critically attend both to audiences themselves and to what is said about them at the time and subsequently.

Notwithstanding the promise of such early studies, it seems they were out of step with a socio-political climate, which preferred to construe audiences (in both the popular and elite imagination) as gullible, homogeneous, and unthinking. Boosted by major funding for American social science, the quantitative tradition of media effects research instead became dominant, seeking to harness the power of the mass media to serve the propaganda needs of the state. This was followed by an even greater effort to deploy that same power to service commercial interests during the 1950s consumer boom. Yet media effects research struggled on scientific grounds for the reasons that Cantril, Herzog, and others had already identified. Having to resign themselves for the next few decades to the thesis of minimal effects (far from the mythical hypodermic needle or silver bullet of media messaging hoped for by powerful funders), media effects researchers found themselves forced step by reluctant step to complicate their linear models of mass persuasion by adding ever more variables relating to the particularity of audiences, meanings, and context. The second lesson, then, is that audiences are not so gullible as popularly feared, precisely because they are neither homogeneous nor unthinking.

Nonetheless, the postwar rise of the theory of political economy of communication, while rightly critical of media power, and pioneering in tracking the increasingly global transformations of media companies along with the gradual capture of the state by corporate power, has been persistently vulnerable to charges of technological and economic determinisms because of its equally persistent neglect of social, psychological, and cultural processes (Babe 2009). Empirical audiences—plural, located, reflexive—are easily lost in the abstract nouns of political economy theory (market, civil society, population, public opinion, the digital divide). In this tradition, neither the quantitative findings of minimal effects nor the qualitative findings of critical audiences received much of a hearing (with the possible exception of George Gerbner’s ingenious but ultimately flawed effort to amplify minimal effects within the framework of cultivation theory). Rather, a received version (of which more later) of the idea of the commodified audience (“if you’re not paying for it, you are the product”) attributed to Dallas Smythe and, before him, to the Frankfurt School, appears to have legitimated the critical neglect not only of empirical audiences but also of those who researched them. Or as Oscar Gandy ruefully remarks, on reviewing Richard Butsch’s The Citizen Audience (2008),

I actually believed that I had audiences down pretty well from my little corner of critical political economy. For me, audiences were industrial products, packaged for sale (or short-term lease) to folks with something to sell. But Butsch’s finely detailed cultural and political history of audiences in America offers another perspective that I believe is well worth considering.

While such a rethinking is just what an audience researcher hopes for, this instance came several decades after Stuart Hall’s (1973/1980) essay on encoding and decoding triggered what Hall himself described as an “exciting” resurgence of critical interest in audiences internationally.

For Hall and his colleagues at the Birmingham Centre for Contemporary Cultural Studies, the desire was to know more about the everyday lives of people living in media culture, including their interests and experiences as audiences, without taking these as given. So researchers went out and talked to people, confounding those who preferred staying in their office studying texts or imagining audiences to the quotidian world of unpredictable living rooms, endless cups of tea, and knotty research ethics procedures (Livingstone 1998). The payoff was considerable, however. Considerably extending the insights of early qualitative research, a transformative picture emerged of an active audience, an interpretative audience, far from simply subject to the causal influences of powerful others. Not only were media effects minimal, but texts also proved to be polysemic, necessarily open to audience interpretation, even resistance. Technologies were found to be domesticated in unexpected ways, with unintended consequences. Audiences were found to be diverse in ways unanticipated by producers and marketers because of their diverse culture, history, community, and politics. Among many outcomes, one was a critique of the implied audience still endemic within critical theories of dominant media power. My third lesson, then, is that it is time to end the binary formulation that pits media power against audience power, instead recognizing that the circulation of meanings includes not only encoding but also decoding and, today, audience encoding too. This should not be read as a celebration of individualism but, rather, as recognition of the structurally unequal yet semiotically open processes of the circulation of culture (du Gay et al. 1997).

The Mediatization of Society

Over recent decades, I had thought media studies had reached a quiet (or, perhaps, tired) settlement that recognizes both the theoretical co-constitution of the concepts of media and audience as well as a commitment to exploring their empirical and contingent interdependencies. But today, the nature of media power is shifting substantially, along with deeper geopolitical changes, leaving critical scholarship scrambling to keep up. In the rush, fragile settlements are easily undone. I illustrate this first by reflecting on a debate close to home for European media studies and revealing of our disciplinary assumptions. The theory of mediatization—“an historical process whereby communication media become in some respect more ‘important’ in expanding areas of life and society [and . . .] institutionalized technologies of communication expand in extension and power” (Fornäs 2014, 484)—has captured the imagination of many notable scholars as a way to conceptualize the changing role of media in history (Couldry and Hepp 2017; Hjarvard 2008; Lundby 2014; Lunt and Livingstone 2016). But it already supports Katz’s case that as attention is drawn to the analysis of media power, interest in audiences is eclipsed.

The theory of mediatization builds on earlier ideas of mediation. Again, the wider socio-political context is relevant—in the last decades of the twentieth century, we saw the end of the postwar settlement, the rise of lifestyle politics, reflexive modernity, and the risk society, and above all—globalization. Questions of power were being newly contested, newly theorized—as more dispersed yet more extensive, as more diverse in form yet more convergent in effect. For many scholars, the sense of new possibilities was expressed by the theory of mediation—“the fundamentally, but fundamentally uneven, dialectical process in which institutionalized media are involved in the general circulation of symbols in social life” (Silverstone 2005, 189). Audience researchers began to research the experiences of people who live in media (Deuze 2012), in a media world (Bird 2003), and hear their stories (Thumim 2012). The established focus on audiences sitting in front of TV sets at home, in discrete times and places, gave way to a recognition of audiences being everywhere and nowhere (Bird 2003), reconnected to society through their roles as citizens and consumers, elites and migrants, homeworkers and activists, and more. Audience as object became audiencing as process (Fiske 1992), transcending binaries of public/private by refocusing on people’s mediated participation in society, whether through fandom, protest, or other forms of public connection (Couldry et al. 2010; Livingstone 2013).

A range of consequences followed, as audience researchers embraced the globalization of our topic, centering ourselves less in the West and more around the globe, opening up new questions about identity, positionality, and voice. We decentered both media and audience by contextualizing processes of mediation in a wider landscape (Couldry 2012). We reconnected the audience within the larger circuit of culture, including with production (Mayer 2016). We became more reflexive about how audience knowledge was used, offering less neutral description open to all-comers and more support for emancipation. So when I said, a few years ago, that everything was mediated (Livingstone 2009), I meant to position “the media” historically among other societal mediations—money, language, goods, or urban planning (Silverstone 1994). It was not a claim about the growing concentration of power in the hands of a few global conglomerates, though such institutions must remain in view, or about the exploitation of audiences, although this is clearly a concern. Rather, it was an invitation to attend to the conditions of meaning-making, to amplify audiences’ voices in the interests of social justice, and to imagine with them alternative futures.

But the media moved on fast. And they are moving us on too. With growing economic significance, new forms of networking, and ever-more powerful players, the media’s own story of its importance is changing, as is ours. Recognizing that “traditional social institutions (family, church, labor union, political party, etc.) have been stripped of much of their traditional purpose by the impact of mass-produced communications” (Smythe 1981, 253; see also Hjarvard 2008), mediatization theory conceives of all parts of society being reconfigured according to “media logics.” This is to go beyond the claim for an observed and contingent increase in “mediation” to argue that “the media” are becoming more important, and more influential, in ever more complex ways. Rather than Silverstone’s “fundamentally uneven, dialectic process” of mediation, it suggests more of a takeover. Moral panics have often gained potency at times of changing media power, reinscribing a generally pejorative view of the audience as passivized and homogeneous, and simultaneously naturalizing that view so that the empirical research that might check or contest it seems unnecessary. Seemingly, and perhaps surprisingly, this potency extends to the academic domain.

In accounts of the mediatization of societal domains or fields (say, of politics, education, family, sport, and so on; see Lundby 2014), the lived experience of audiences is largely invisible, partly because audience research favors “voices from below,” while mediatization theory examines the workings of power “from above” (Lundby 2016). It is also because mediatization theory tells a story that spans decades, if not centuries (Lunt and Livingstone 2016), this posing a particular problem of evidence since compared with other parts of the circuit of culture, “the social and cultural aspects of mass media reception are literally disappearing before our eyes and ears” (Jensen 1993, 20–21; see also Mihelj and Bourdon 2015). The result is a double exclusion, not just of audiences, but especially of the experiences of the nonelite—young and old, poor or foreign, or simply ordinary—who rarely make it into the historical record.

We are left with a contradiction—the media are becoming more and more important across ever more fields of society, and yet people’s engagement with such media is, seemingly, unimportant. In the field of political communication, Witschge (2014) counters by pointing variously to audiences’ diverse interpretations of political media texts; their agency in acting on their interpretations as voters and citizens in a mediated world; the social and civic consequences of interactions among audience members; the difficulty of persuading audiences to think in particular ways; the aggregate effect of audiences’ choice, search, selection, and commenting strategies; and the activities of audiences in creating and sharing mainstream, alternative, or resistant content. Related arguments can and should be developed for other fields, perhaps following Schrøder’s (2017) call for a theorization of audience logics (or dynamics) to mirror the media logics of mediatization and thereby to recognize the mutual shaping of media and audiences over time.

The Datafication of Audiences

By contrast with debates over the political economy of communication or the theory of mediatization, the debate over datafication (Lupton and Williamson 2017; Mayer-Schönberger and Cukier 2013; van Dijck 2013) far exceeds the boundaries of media studies (notwithstanding the effort to make it ours by dubbing it “deep mediatization”; Couldry and Hepp 2017). Perhaps its very multidisciplinarity contributes to today’s heady climate in which cautious calls to gather evidence about people’s lives are easily missed in the urgent rush to describe our coming predicament. As boyd and Crawford (2012, 666) wrote at the start of the excitement over “big data,” we must not lose sight of “why people do things, write things, or make things . . . in the sheer volume of numbers,” not least because “bigger data are not always better data” (p. 667), being often partial, biased, or decontextualized. And yet it seems that in accounts of the datafication of society, attention to empirical audiences is easily displaced by a fascination with the data traces they leave, deliberately or inadvertently, in the digital record. Academic attention has turned to the analysis of the algorithmic manipulations of audience’s digital traces that increasingly allow everything people do to be tracked, as their data are bought and sold above their heads and below their radar (Qiu 2018; Ytre-Arne and Das 2018).

In one sense, such data make audiences newly visible—consider the current fashion for visualizations of the “twittersphere” or aggregate “comments” or other digital traces, along with lively discussion of the algorithmic insight to be obtained from the “big data” that fuels digital networks. This also recognizes how people’s actions as audiences are mixed in with all other actions recorded on digital networks, again demonstrating that audience analysis cannot be sequestered from the rest of people’s lives. But this enhanced visibility obscures more than it reveals. In data visualizations of audiences’ (or users’) activities, much of importance is stripped away. Away with the socio-cultural, displaced by individual “behaviors.” Away with context, meaning, interpretation, for it is the hidden patterns beneath awareness that matter. Away with audiences’ motivations, commitments, and concerns—for if data reveal what people “really” do on and through digital media, why talk to them anymore? In response to some of the rhetoric, Couldry and Kallinikos (2017, 153) caution that “on social media, users for practical purposes are not real persons but abstract operations enacted through the aggregation of singular data-points.” At present, the distance between real and data selves is often great—witness the academic uses of big data that fail even to distinguish men or women, adults or children.

In short, while “ethnography is not the one methodology to rule them all” (Schrock 2017, 705), no more are data analytics. The data analytics industry may believe that “by predicting you are able to anticipate what people will want and shape your business accordingly, thus protecting its future value” (Beer 2018, 473), but it is our job to question such claims. Insofar as “audience research has entered the era of ‘big data’” (Athique 2018, 59), the “whole way of life” (Williams 1958) should not escape our grasp, and multiple socioeconomic and cultural determinations shape audience agency and interpretation, of which the digital interface is but one. Or as du Gay et al. (1997, 84) noted two decades ago, the

excessive focus on production and the economic has the effect of shutting down the analysis of culture . . . One is most unlikely to learn anything from people’s everyday practices if one approaches them with the view that they are unworthy of serious study because they are superficial and inauthentic substitutes for a denied alternative existence.

Enough of the critique, for these debates and developments are yet young and both technology and those who deploy it will only gain in sophistication. Behind the many fearful predictions about datafication, we might identify (at least) three. One concerns the creation of value for platforms from the exploitation of audience (human) labor. One concerns discriminatory judgments by algorithms on behalf of infrastructural institutions (insurance, employment, education, banks, police, etc.). One concerns the public and democracy’s vital reliance on the (flawed or biased) judgment and (insufficient or insufficiently rational) participation of ordinary people.

Of the first, Carah and Angus (2018, 193) worry that “value is created where we enable algorithmic media to train on the data we continuously stream, enabling them to make more fine-grained judgments about us.” Athique (2018, 64) calls this alchemy, since turning audiences’ digital traces from “muck” to “brass” may rest on a flimsy foundation—after all, the value Facebook gains from selling users’ attention to, say, Coca Cola for advertising purposes may or may not benefit Coca Cola’s revenues nor, on the other hand, cost the users more than they would wish to pay. At the same time, this transaction does not exhaust the meaning of viewing Coca Cola advertisements on Facebook. In Ang’s (1990) now-classic critique of the ratings industry in Desperately Seeking the Audience, she argues for the use value of the advertisement not only to those who go out to buy Coca Cola but also to the audience that enjoys laughing at it or deconstructing it or deliberately buying Pepsi because of it. More negatively, the calculation of value accruing to Facebook does not include the cost to society of an audience that learns the world is composed of fun-loving beautiful white people from which one is personally excluded. Beyond questioning the limits of the claim that platforms create value by exploiting audiences, I also wonder, when we trace how, “over the course of ten years, users have negotiated their relationship vis-à-vis platforms through appropriation and protest” (van Dijck 2013, 160), whether the resulting “cat-and-mouse game” is so different from that which Jenkins (2003) described between Star Wars fans and Lucas Enterprises. As Artz (2015) argued, it is vital that the audience-as-commodity (or audience-as-exploited-labor) argument does not “overreach” itself either by reducing audiences to data (see also Athique 2018; Fisher 2015) or by confusing the activities of platform users with the work of the platform in collecting and monetizing those activities (i.e., generating exchange value).

Of the second, I can only urge attention to the whole circuit of culture including regulation. Undoubtedly the audience or user will suffer if algorithms that make discriminatory judgments are deployed by infrastructural institutions. But what matters here is not only the expressive relation between audience and platform (important though this is, justifying audience ethnographers’ call for greater attention to audience voice and audience interests; Lupton and Williamson 2017; Ytre-Arne and Das 2018). Nor can we rely on the heroic actions of citizens, though one sympathizes with Jack Qiu’s (2018, 307) urging that “the future of digital labor, including social media labor and free labor, is up to agentic human actors on both sides of the circuits—to resist top-down control that reduces us into subhumans and to expand our liberty and humanity through networking and innovative interventions.” Rather, it must be for the democratic state and international civil society and governance bodies to act in the public interest, intervening not only in relation to the transparency and accountability of platforms (as there are growing calls for them to do; Mansell 2017) but also in relation to the legitimacy of the decisions taken by society’s infrastructural institutions, which, after all, cannot escape regulation. In other words, while the social justice implications of the dominance of platforms are, undoubtedly, both worrying and urgent, one strategy is to fight for regulation that reduces the burden on audiences’ media literacy and capacity for resilience and resistance by designing a digital environment that treats ordinary people more fairly and equitably.

Third, while recent investigations into the audiences’ vulnerability to viral misinformation (Newman 2018) certainly raise concerns, burdening audiences with the power—and the responsibility—to underpin or undermine democracy writ large is disproportionate. Initial anxieties that election hacking on Facebook and other social media tipped the outcome of recent elections because the voting public was newly vulnerable to manipulation and mass persuasion have not stood up to critical examination. More important, it seems, were the deeper forces shaping distrust of elites, disaffection with democratic participation, and processes of economic inequality and cultural exclusion. As Nick Couldry, Tim Markham, and I found in our “public connection project” (2010), not all public connection is significantly mediated, and nor is all mediated experience determining of democratic participation. Why? Because of the layers of societal infrastructure between audiences and the state, just as there are between audiences and commerce or, indeed, among audiences themselves. So while there is much to fear from datafication (and perhaps something to welcome also), we will not understand it critically if we “collapse social classes, productive relations and all of the complex, diverse means of production into one amorphous factory churning out private profit in every human action” (Artz 2015, 312).

Conclusion

Many of our contemporary debates not only have long roots but also risk repeating old problems as hopes and fears about the abuses of media power rise again. In arguing that we should keep firmly in mind what history has taught us about people as audiences, I do not mean to advocate either complacency or a celebration of audience agency. Rather, I have argued for the recognition that all analyses of media power include, implicitly if not explicitly, claims about audiences, meaning that research with audiences to examine these claims must be brought within in the critical project. Furthermore, audiences are necessarily social, embedded in society and history in many more ways than through their relation with the media, so the critical analysis of audiences cannot be satisfied with sporadic inclusion of disembodied, decontextualized observations of behavior or cherry-picked survey percentages but must engage with audiences meaningfully in and across the contexts of their lives. It is vital that media studies bring into focus those many dimensions of society—consider the deep shifts in the global economy and world politics that underpin contemporary problems of corporate power, malign states, regulatory failure, and, indeed, dispossessed publics—if we are to avoid technological determinism or dystopian fatalism. Or, to quote Dallas Smythe (1981, 253),

The mysticism attached to technique (and “technology”) has incorrectly assumed that the medium basically defines the audience. But as a historical analysis of the rise of the mass media will show, the opposite has been true . . . By placing the contradiction between advertisers/media on the one hand and audiences on the other on the level of social relations we are on solid ground and can repudiate the mysticism of the technological trap by which audiences are tied to hardware, software, and technique.

Inquiring systematically into the experience of audiences (people) will not always produce happy answers. But it will help us check and qualify grand claims, and it will remind us of the many potential levers for change, including but also going beyond the technological. Some of this work is already underway, finding that—as for every preceding generation of audience research, people do not always fall obediently into line with the responses presumed of them. For example, Bucher’s (2017, 42) qualitative study of social media uses found that “the lived reality of the Facebook algorithm generates a plethora of ordinary affects which may be distancing as well as enticing, generating resistance as well as appeal.” For Andrejevic (2014, 1685), similarly, the primary response from audiences is “frustration over a sense of powerlessness in the face of increasingly sophisticated and comprehensive forms of data collection and mining.” These and other accounts of frustrated, distrustful, or resistant audiences do not suggest the effective imposition of power by the big platforms and those who harness them, giving encouragement to those who call for alternative approaches that respect audiences and publics (Kennedy and Moss 2015).

Roger Silverstone (1999) concluded Why Study the Media, by saying, “It is all about power, of course. In the end.” Maybe it is, but this claim has always worried me, seeming to efface so much about people’s lives, including their meanings, values, cultures, indeed their humanity. I mentioned this to my colleague Robin Mansell, who recalled how Dallas Smythe, then a member of her PhD committee, asked of her thesis on the political economy of communication, “but Robin, where are the people?” Indeed, including the people in a mediated, perhaps mediatized, increasingly datafied age—that’s the task in front of us.