Agency, social relations, and order: Media sociology’s shift into the digital

1 Introduction

Media and communication are among the most fundamental concepts in the social sciences, in which it is a firmly established position that technology-based communication media are the prerequisite for the emergence of more complex forms of society such as advanced civilizations, modern nations, or today’s late modern societies (Thompson, 1995). Communication is considered a principle means for the construction of social reality. Peter Berger and Thomas Luckmann (1966, p. 172), for example, spoke of a “conversational apparatus”, through which humans continuously maintain, modify, and reconstruct reality, while system theory bases its conception of society on the communicative character of social systems (Luhmann, 2012). In a broad sense, media sociology can best be understood as a field of the social sciences that deals with the role played by the technologically-based mediation of communication in the construction of the social world (Silverstone, 2005).

Until the end of the twentieth century, media sociology was for many scholars synonymous with an investigation of mass media’s implications for society (Katz, 2009; Wahl-Jorgensen, 2013), but much has changed in recent years. If we talk about media today, legacy mass media such as newspapers or television still have their place, yet, for more and more people media are more frequently comprised of various digital platforms such as Instagram, Facebook, or Netflix. Even when we talk about newspapers and television, these no longer take the form of legacy mass media but have taken on new digital arrangements and have, therefore, fundamentally changed in a variety of ways.

Just as the phenomena of media sociology have leaned towards the digital, so too has media sociology shifted in the approaches it takes. This, in turn, requires innovative theoretical strategies to help make sense of a rapid, fluid media environment characterized by manifold media. It is media sociology’s shift into the digital that I would like to deal with in this article. My intention here is not to rely on too many previous studies but, rather, to establish some of the fundamental connections which, in my view, are characteristic of media sociology’s re-orientation. The main argument is that media sociology has developed from a sociology of mass communication to a sociology of a deeply mediatized world. There are three more concrete changes that we can connect to this shift: first, a rethinking of agency, second, a redefinition of social relations, and third, a rediscovery of order in light of the digital.

Taking such a comprehensive view is certainly risky because it can lead to overlooking the particularities and specificities of individual approaches. This is especially the case for the field of communication, of which media sociology is a part. As Silvio Waisbord (2019) highlighted, we can capture communication as a “post-discipline”, which has no unified theory or methodology and is, therefore, situated “beyond” the paradigmatic idea of a traditional discipline. Nevertheless, we should avoid taking these remarks to mean that we should approach communication with an “attitude of arbitrariness”. By contrast, he argues for something that he calls – referring back to C. Wright Mills (1970) – a “communication imagination” that looks for “connections between different strands of research”, aiming “to link scholarship with public life” (Waisbord, 2019, p. 150). This article can be understood as an attempt to develop a ‘media-sociological imagination’ of the digital. My concern is to identify some common threads running through recent media sociology for the purpose of developing a more general understanding of where media sociology is headed.

However, a preliminary remark on the terminological nature of media sociology itself is necessary. We can define media sociology as “the study of media processes and phenomena anchored in classic and contemporary sociological questions and methods” (Waisbord, 2014, p. 7). Media sociology is, therefore, a certain ‘style of thinking’, regardless of whether or not the person in question is a scholar of media and communication studies or sociology. Over and above the different strands of media sociology, they all share a focus on the analysis of ‘social patterns’. As Rodney Benson puts it, it is about the “patterned character of human action and thus [to] create categories that group together various patterns” (Benson, 2014, p. 26). Certainly, such an elaboration of patterns is always associated with generalizations which – compared to empirical data – in some contexts can generate simplifications. Its advantage, however, is its ability to guide us in achieving an understanding of structural relations in the social world and society and, in the specific case of media sociology, in regard to questions of media and communication.

A broad approach such as the one I intend to take in this article is certainly always limited: It is not possible to give due credit to every publication in the field. By necessity, one can only move through a section of the media sociology canon. This can only be dealt with by making explicit the position from which the article is written; in the foreground will be placed all those publications that orient themselves around a new foundation of media sociology, understood in the broadest sense as a form of “practice theory” (Reckwitz, 2002) and “relational sociology” (Fuhse, 2020). Other approaches, such as systems theory, are thus relegated to the background. The turn to practice theory and relational sociology is, however, not accidental. The implicit argument is that the turn to these approaches is connected to the increasing relevance of digital media, which have challenged media sociology to think differently about media, social practice, and society (see Couldry, 2012; Pentzold, 2015).

European media sociology can be seen to support this position, which constitutes the primary starting point for my argument. Fervently discussed in the field of social theory over the past few years (Bhambra, 2007), Western theory formation is confronted with the question of the extent to which it needs to ‘de-westernize’ (for media and communication research: Curran and Park, 2000; Gunaratne, 2010; Ray, 2012). The necessity to question the cultural and social ‘blind spots’ of theory-building, of course, also affects media sociology. This brings me, however, to a second argument for adopting a practice theory approach: The turn to practice theory can be partly understood as a shift away from a modern conceptual apparatus of sociology to a more open brand of theorizing. Many practice theory practitioners have consistently endeavored to develop a broader, globally oriented perspective that is sensitive to questions of “de-westernization” and “post colonialism” (i. e., Couldry and Mejias, 2019, pp. 83–112).

Based on these preliminary remarks, I would like to argue along five steps. First, I will accentuate some of media sociology’s implicit premises to make clear the extent to which the spread of digital media is connected with the need for a reorientation of media sociological thinking. I will then discuss three core aspects of this reorientation in more detail: agency, social relations, and order. This finally leads me to a conclusion in which I further reflect on the ‘media-sociological imagination’.

2 Media sociology: From a sociology of mass communication to a sociology of a deeply mediatized world

As already mentioned, media sociology has its origins in the sociology of mass communication (Pooley and Katz, 2008). Mass communication can be defined as “the practice and product of providing leisure entertainment and information to an unknown audience by means of corporately financed, industrially produced, state-regulated, high technology, privately consumed commodities in the modern print, screen, audio and broadcast media” (Hartley, 1994, p. 173). Media sociology was concerned with a critical analysis of mass communication’s social patterns, whether in regard to their production processes (for example, media organizations and newsrooms), their content (ideologies, stereotypes, and so on) or their appropriation (for example, the relations of media use to class and situational context). Mass media have typically been defined as a ‘field’, a ‘system’ or a ‘sphere’ that has an influence over the various domains of society.^[1] What is striking is that even during the time when the “common sense category” (Hartley, 1994, p. 172) of mass media dominated media sociology, its orientation was anything but uncontroversial. Todd Gitlin (1978), for example, criticized the narrow empiricism of classic media sociology as exemplified in the work of Paul Lazarsfeld. Based on his own ethnographic research on media’s “domestication” (Berker, Hartmann, Punie, and Ward, 2006; Morley and Silverstone, 1991), Roger Silverstone (2005) made a case for placing the concept of ‘mediation’ at the center of media sociology – whereby there were clear parallels to the Latin American approach toward media sociology (Martín-Barbero, 1993). In particular, ethnographic research on media appropriation suggests that concepts of “mass media and “(Curran and Gurevitch, 1991) need to be countered with different perspectives that place greater emphasis on the everyday embeddedness of media use, its contradictoriness as well as practices of everyday resistance (i. e., Drotner, 1994; Moores, 1993). A certain ‘messiness’ of media appropriation became clear here, an approach that, in turn, corresponds with non-Western ways of theorizing (Murphy and Kraidy, 2003). This involved addressing issues of inequality and power with respect to ethnicity, class, and gender, from a local as well as a global perspective.

Without wanting to level out the diversity of this discussion on media sociology, it can be stated that there was a unifying element between the different positions which manifested in understanding mass media as a sphere of society that was in a certain sense ‘separate’ – as a ‘field’, a ‘system’, or a ‘sphere’. But this is precisely what is being called into question with the advent of digital media. Digital media are defined here as all media whose basic technologies of production, distribution and use are based on ‘software algorithms’. Historically, the term digital media was first used to describe the (then) ‘new’ media of the internet, from webpages to platform media (Lister, Kelly, Dovey, Giddings, and Grant, 2009, pp. 13–44). Increasingly, however, in the “hybrid media system” (Chadwick, 2017) the legacy media of mass communication are reconfigured into the digital; television, for example, is not only watched in front of the television set (which nowadays is also a kind of computer within a large monitor) but also through various other end devices and their supporting apps.

The crucial point is, therefore, that with digitalization and its associated media transformations – which, above all, can be seen in the software-based character of almost all of today’s media, ‘media’ are no longer simply “what people consume when not working or sleeping; they are interwoven into social life, making ‘mediation’ integral to everyday life” (Waisbord, 2014, p. 6). When Sonia Livingstone interrogated “the mediation of everything” (Livingstone, 2009, p. 1), she characterized a shift which resulted in the idea that media can no longer be considered a separate “domain of society” (Lunt and Livingstone, 2016, p. 3) that influences others. With the emergence of ‘new’ digital media and the digitalization of ‘old’ legacy mass media, a completely different approach is required if media sociology is going to suitably analyze societal patterns of mediated communication and their role in the mediated construction of reality, as well as questions of inequality and power.

This challenge for media sociology posed by the changing media environment coincides with the same challenge for other areas of sociology, which also argue that with digitalization, social science theory needs to be further developed to remain in touch with related transformations. From a systems theory perspective, for example, Armin Nassehi (2019) argued in his proposal for a theory of a digital society that digitalization should be seen against the background of the totality of the simultaneous complexity and pattern-like character of today’s societies which then becomes “observable because many of these patterns are, for the first time, made visible by digital data”. Similarly, Noortje Marres argued for a digital sociology because “digital societies” are not only marked by “far more complex interactions between social life and knowledge” but, in addition, “data analysis is put forward as a way of acting on social problems” (Marres, 2017, p. 9) – that is, as a way of ‘recognizing’ and ‘handling’ social patterns.

If the sociology of digital media is to take these ideas further, it is helpful to refer to mediatization research, as Sonia Livingstone does in her reflection on the “meditation of everything” quoted above (Livingstone, 2009, pp. 6–7; Lunt and Livingstone, 2016). At its core, mediatization is a “sensitizing concept” (Jensen, 2013, p. 206, with reference to Blumer, 1954) that draws attention to the interrelations between changes in media and communications, on the one hand, and changes in culture and society, on the other (Couldry and Hepp, 2013). Mediatization research is thus concerned with empirically identifying patterns in these interrelations and gradually arriving at more general theories on the role mediated communication plays in processes of social and cultural change.^[2]

Initial studies into mediatization were more in line with media sociology’s ‘legacy’, describing media as “partly independent institution[s] in society” (Hjarvard, 2013, p. 43), which have an influence on other social spheres through their “media logic” (Altheide and Snow, 1979; Asp, 1990; Mazzoleni, 2008). With the progression of digitalization, more attention began to be paid to the “media-saturated” (Lundby, 2014b, p. 3) character of societies, and mediatization research began addressing the new characteristics of mediatization through digital media and their infrastructures (i. e., Finnemann, 2014; Miller, 2019).^[3] A key argument is that with the swelling ‘wave’ of digitalization, which saw fundamental qualitative changes in our media environment as more and more media base themselves on ‘software algorithms’, a new stage of mediatization has emerged, one that is sometimes referred to as “deep mediatization” (Hepp, 2020a). Deep mediatization attempts to describe the ways in which digital media and their infrastructures are so comprehensively entangled with our everyday practices that we can no longer grasp the various social domains of contemporary society in their distinctiveness as apart from digital media. It no longer makes sense to understand media as an ‘independent’, discrete social domain.

Certainly, other terms are possible. However, regardless which term is preferred, the key task for media sociology is to develop appropriate concepts for the description of advancing media saturation and its related transformations. If we follow the discussion of a media sociology oriented towards practice theory and relational sociology, media sociology faces three specific challenges: first, a rethinking of agency; second, a redefinition of social relations; and third, a rediscovery of order in the digital realm. From an intersectional perspective, all three concern the classic questions of social inequality that media sociology has always been interested in, especially questions of inequality and power and their relation to ethnicity, class, and gender.

3 Rethinking agency: Media practice, entanglement, and communicative robots

As mentioned above, media sociology has focused a lot of its attention on how the everyday spread of digital media was accompanied by a shift towards practice theory (Pentzold, 2015; Reckwitz, 2002; Schatzki, Knorr Cetina, and von Savigny, 2001). It tasked itself with turning practice theory toward the analysis of routine, everyday activities, and the ongoing processes of constructing reality that exists within and passes through them. In today’s deeply mediatized societies, the contrast between practices of specific media use (for example, ‘watching television’) and other social practices (‘cooking’, for instance) becomes less distinct as an increasing number of our social practices also refer to the media (for example, ‘cooking with the help of YouTube tutorials’). As Nick Couldry put it: “If we live in a ‘media-saturated’ world, then it is reasonable to expect that how that world is carved up into recognizable practices may no longer correspond to categorizations formed in a ‘pre-saturation’ world (where audiencing could be assumed to be a discrete activity)” (Couldry, 2004, p. 121).

This is associated with a fundamental shift in media analysis: Re-accentuating older concepts such as medium theory (McLuhan and Lapham, 1994; Meyrowitz, 2019), it is not anymore media content as an ‘entity’ which has a direct or indirect effect on people. More relevant are the ways in which the various practices of today’s social reality are interwoven with digital media technologies. This is typically described as “entanglement”, which finds its roots in the discussion on “sociomateriality” (Jarke, 2014; Scott and Orlikowski, 2014, p. 873). ‘Entanglement’ emphasizes that everyday practices are closely interwoven with material media technologies and that there is “a lack of an independent, self-contained existence” (Barad, 2007, p. ix). This is not merely a transitional situation, but one that is fundamental to a deeply mediatized social world. Many practices constitutively refer to digital media and cannot be investigated beyond them. An analysis of the practice of dating among young people, for example, is not possible without considering their interdependence with various messenger services and other platforms. It is precisely with such an entanglement that the classic media sociological questions of inequality – of ethnicity, class, gender, etc. – are challenged in a new way. Inequalities that were not initially conceptualized in terms of media must now be reconsidered in relation to digital media and their infrastructures.^[4]

A particular component of digital media is that it turns media practices (all practices entangled with media) into data practices (practices of producing and processing digital data) (Bolin and Schwarz, 2015). This is due to the fact that each time digital media are used, so-called “digital traces” (Merzeau, 2009, p. 2) are left behind, traces that are then processed by the corresponding software systems and accumulated into what Ruppert refers to as a “data double” (Ruppert, 2011, p. 220, quoting Haggerty and Ericson, 2000), a digital representation of an individual. What is occurring here is an additional layer of entanglement with complex software systems whereby human practices unfold a further, technologically mediated dimension of the construction of reality. It is not only about what one does while realizing a particular practice but also about the data that are generated and processed as a consequence of each practice that then become a large part of our datafied constructions of reality. Online shopping, for example, is not only about acquiring a certain product, but through the automated analysis of the data, it is about the construction of a “computed sociality” (Alaimo and Kallinikos, 2017, p. 185), as data footprints reconfigure social interaction and reorganize users into larger, commercially relevant social objects. This is what is happening, for example, when consumers’ tastes are assumed with the purpose of advertising other products following an online purchase (Turow, 2011).

This entanglement of human practice with digital media and their infrastructures raises new questions of agency. In approaches to media sociology that addressed legacy mass media, the attribution of agency was always quite simple: It was media producers (as individuals or as organizations) and audiences with whom agency was associated (see, for example, the chapters in Curran, 2014). The media were mainly thought of as a means of mediation for the purpose of communication. But it is clear that the role digital media play is already much more far-reaching if we consider the procedures of data processing which take place in parallel with their communicative uses. Even more far-reaching is the fact that digital media themselves produce ‘communicative outcomes’. One can see this on a very simple level in the form of the processes through which the algorithmic selection of a platform’s message feed is presented to the user (Bucher, 2018, pp. 41–65). However, the agential potential of digital media is much more comprehensive, as various “communicative robots” (Hepp, 2020a, p. 4) have begun to enter the public imagination and our domestic spaces: artificial companions such as Amazon’s Alexa and Apple’s Siri (Wilmott, Fraser, and Lammes, 2018), social bots like Facebook and Twitter bots (Gehl and Bakardjieva, 2016), and work bots that can be found in ‘robot journalism’ practices where software is used to automatically produce journalistic texts (Caswell and Dörr, 2018; Clerwall, 2014).

The development of these systems has stimulated many media sociology scholars to fundamentally rethink agency. An important point of reference here is Actor Network Theory (ANT), which early on pointed out that ‘things’ are able to adopt the role of actants when coming together with human actors (Latour, 1991). The idea was originally explained using simple material objects such as speed bumps in the street that force people to drive slowly. With today’s digital media, however, any attempt to rethink agency has dug even deeper. The question here is no longer as simple as whether and how people can, for example, delegate their ability to act with technology; rather, it is a question of how to adequately describe algorithm-based communication media when they become communicators themselves while also becoming closely entangled with human practices (Esposito, 2017; Guzman, 2018). ‘Hybrid agency’ approaches (e. g., Fink and Weyer, 2014; Hanson, 2009) are useful here: As diverse as the various theorizations are, they meet at the point where the decisive question is not what agency the ‘machine’ – for instance, “automated media” (Napoli, 2014, p. 340) or “communicative robots” (Hepp, 2020b) – intrinsically possesses. Rather, it is a question of what new types of agency emerge when people act together with these kinds of systems. In this sense, a hybrid agency is always a kind of supra-individual agency that is produced in tandem by both humans and machines. The discussion on appropriate theoretical models and descriptive methods is constantly expanding and is correspondingly open (Gunkel, 2020, pp. 256–280). It is clear, however, that older concepts of agency that only refer to human practice and delegation will most likely fail to successfully address new forms of autonomous media (Andrejevic, 2020).

4 Redefining social relations: Networks, assemblages, and figurations

With the development of digital media and the supporting infrastructure of the internet, discussion arose in media sociology as to how social relations can be adequately described in the context of transforming communications. Besides older concepts such as field, system, and domain, three descriptive concepts came to the fore: networks, assemblage, and figurations. Despite all their differences, the discussion about all three concepts shows that there exists, from a media sociology point of view, a need for a new descriptive language to adequately explain the complexities of social relations in times of deep mediatization.

Network is, first of all, a structural metaphor to describe the relations between human actors within a certain social entity (such as a group or the family) and the relations between these entities. The internet and digital media shaped this concept into an analytical perspective to describe the complexities of today’s social relations, and the ‘network’ was even understood as society’s “new operating system” (Rainie and Wellman, 2012, pp. 3–20): By “incorporating gadgets into their lives”, people changed the “ways they interact with each other” and became “increasingly networked as individuals, rather than embedded in groups” (Rainie and Wellman, 2012, p. 6). Local communities and their “strong ties” matter less; more important are “weaker ties” (Granovetter, 1983) that can be expanded by the internet and other communication infrastructures. As a consequence, society appears to be nothing more than a large, complex network: “Societies – like computer systems – have networked structures that provide opportunities and constraints, rules and procedures” (Rainie and Wellman, 2012, p. 7). For media sociology, network is less a metaphor to describe the digital infrastructure of the internet and more a bridging concept to grasp the change in social relations contextualized by digital media. The main points of discussion were “networked publics” (boyd, 2010), the “networked self” (Cohen, 2012), and the “networked culture” (Jenkins, Ford, and Green, 2013).

Extending this perspective from examining social relations within society to describing the organizational foundation of entire societies, Manuel Castells has argued that the internet has become “an appropriate material support for the diffusion of networked individualism as the dominant form of sociability” (Castells, 2001, p. 131). However, for Castells, the “network society” is necessarily a global one as it is a social structure that “is made around networks activated by microelectronics-based, digitally processed information and communication technologies” (Castells, 2009, p. 24). While national states do not cease to exist, they change in the realm of this global network society by increasingly developing themselves in relation to network structures, either between them as networks of states such as the European Union, or within them by moving to more regionalized and localized networks of structural power.

Much like the notion of network, assemblage operates as a bridging concept that offers the possibility of an analysis of various social entities. As an analytical term it was introduced by Gilles Deleuze and Felix Guattari to describe “complexes of lines” that build a “territoriality” (Deleuze and Guattari, 2004, p. 587). Assemblages in this sense are “wholes” characterized by the relations of exteriority. In a terminology that is closer to social sciences, “social assemblage” refers to a “set of human bodies properly oriented (physically or psychologically) towards each other” (DeLanda, 2006, p. 12). Examples of “assemblages of bodies” are conversations, communities, and hierarchical organizations that govern cities and nation states. The term assemblage refers to, as Jennifer Slack (2012, p. 152) puts it, “the dynamic collection or arrangement of heterogeneous elements (structures, practices, materials, affects, and enunciations) that expresses a character or identity and asserts a territory”. The idea is to describe society as consisting of various, differently scaled assemblages. This concept’s main appeal is that it considers the possibility that heterogeneous elements can be held together without actually forming a coherent whole (Allen, 2011, p. 154; Wiley, S. B. C., Becerra, and Sutko, 2012).

There exists a certain parallel with the concept of network that is particularly striking: Assemblages are understood as being made up of individual parts and are articulated through their relational structures as well as by relations of exteriority. One difference between the idea of assemblages and networks is that these parts can be ‘objects’ and ‘things’, leading to an emphasis placed on assemblages’ material character. At this point we can see a direct reference to Actor Network Theory and its redefinition of agency: Objects must be considered as possessing an agency of their own that unfolds in an assemblage with humans and their actions (Latour, 2007). This “network” of “assembled” humans and non-humans is what ANT is principally interested in (Latour, 2011, p. 799; Luckhurst, 2006, p. 4). Various kinds of media-related assemblages became a focus for media sociology, including the ways in which “discursive coalitions” between discrete communicating actors can be analyzed as “ethno-epistemic assemblages” (Allgaier, 2012, p. 308), the idea that media platforms operate as “assemblages” with “software actors” (Bucher, 2012, p. 480), or the ways in which social participation takes place in “assemblages of media technologies” (Carpentier, 2011, pp. 268, 272).

Another take on the complexities of social relations in times of deep mediatization refers to process sociology and the concept of figurations. Put simply, a figuration is used as a “simple conceptual tool to loosen [the] social constraint to speak and think as if ‘the individual’ and ‘society’ were antagonistic as well as different” (Elias, 1978, p. 130). Instead, the idea is to think of both as coming together as one. Figurations are constituted in ‘processes of interweaving’ (Elias, 1978, p. 130), in which the practices of the people involved are interdependent on and oriented toward each other. With figurations the ‘behavior of many separate people intermeshes to form interwoven structures’ (Elias, 1978, p. 132). A figuration is constituted in the continuously changing pattern of interaction between all those involved, which also indicates that the material objects and technologies are entangled with the practices through which a figuration is articulated.

From a media-sociological perspective, we can consider each figuration as a communicative one (Hepp, 2020a, p. 3): Figurations are typically articulated through practices of communication that take place across a variety of media. Family members, for example, can be separated in space but connected through multimodal communication such as (mobile) phone calls, e-mail, and sharing practices on digital platforms, all the while maintaining the everyday dynamics of family relationships. Organizations, considered as figurations, are kept together through the use of databases, communication across intranets, printed flyers and other media for internal and external communication. Individuals are involved in these figurations according to the roles and positions they play in their respective actor constellations. In society, figurations are not simply positioned alongside each other, they overlap as a consequence of the fact that individuals are a part of different figurations. They are embedded in meaningful discourses into the role of a single figuration in society at large. Furthermore, there are also “figurations of figurations” (Couldry and Hepp, 2017, p. 73) – figurations that are embedded in other figurations constructing a nesting of figurations through which more complex social relationships are formed. One example are departments that form the organizational structure of a company which, in turn, is part of a larger corporation as a subsidiary. The media-related transformations that journalism (Kramp and Loosen, 2018), media use (Kobbernagel and Schrøder, 2016), or families (Lohmeier and Böhling, 2017) are undergoing can be researched by comparing the transformation of their figurations. The idea is to gain a better understanding of how the transformation of culture and society relates to changes in media and communications.

The concepts of network, assemblage, and figuration differ fundamentally both from their origin and from their theoretical location. Nevertheless, a shared line of theory development can be seen in all three concepts: They try to grasp the complexity of social relationships in times of deep mediatization which have intensified as digital media bring about new “chains of interdependence” (Elias, 1978, p. 68) possible. They focus on the changing interrelations and interactions between humans and the role of digital media within them. Such concepts try to address one of the main challenges of contemporary media sociology, which is to arrive at an appropriate description of the media-related transformations of social relationships and to design comprehensive descriptive languages that offer us a better understanding of all related transformations. Here, again, the discussion has not reached definitive conclusions, although the problems are more clearly articulated.

5 Rediscovering order: Infrastructures, data, and capitalism

We can see the turn of media sociology toward concepts such as network, assemblage, and figuration in a broader context that does not only refer to a description of social relations and the ways in which they change in times of deep mediatization but also to a reflection of changes to the social construction of order. Manuel Castells, for example, raises the question of power in the “network society” (Castells, 2009, pp. 42–47), which for him does not only refer to “networking power” (the power that the actors included in a network have over those who are not included) but also to power based on influence imposed on the rules of networks (“network power”), power exerted through positioning an actor within a network (“networked power”), and power applied to the ability to constitute or connect networks (“network-making power”). The figuration also serves to describe the shifting articulation of social order (Couldry and Hepp, 2017, pp. 190–210). Each figuration is an order characterized by power balances, whereby not only do figurations and their power balances change within the context of deep mediatization, they also enter new relationships as a consequence of the data infrastructures that work across them. Deep mediatization triggers a “re-figuration” of society, a change in its fundamental order (Hepp, 2020a, pp. 115–144; Knoblauch and Löw, 2017, p. 1).

Social order was rediscovered as a topic in media sociology, and deep mediatization provides it with a completely new apex. Over and above various theoretical approaches, we can define ‘social order’ as relatively stable patterns of interdependences between not just individuals, groups, and institutions but also (looked at from a higher dimension) between the numerous types of relations involved in social life that all depend on larger stabilities of resource (Hepp, 2020a, p. 190). With deep mediatization, the establishment of order in and with media has fundamentally changed. The role of digital media in the production of social order is much more far-reaching in that they permit a new “microphysics” (Foucault, 1991, pp. 26–29) of the production of order through their saturation of everyday life. The discussion on these new forms of creating order by means of digital media principally takes place along the concepts of infrastructure and data, cumulating in a critique of contemporary capitalism. Here, too, it is not simply a question of how media as mass media legitimize the economic order or advertise certain products, it is mainly about a globalized transformation of the ‘microphysics’ of social order alongside capitalism.

Infrastructure is a crucial concept because we can only understand the specifics of digital media if we look at the infrastructure of the internet and the data centers on which they are based. Terminologically, infrastructure is a ‘kaleidoscopic’ concept with many implications. This has to do with the fact that any definition of infrastructure depends on the perspective of the observer. Railroad tracks, for example, represent infrastructure for the traveler; for the railroad engineer they are the focus and the result of their work, which is made possible by other kinds of infrastructures. We can conclude, then, that analytically, “infrastructure appears only as a relational property, not as a thing stripped of use” (Star and Ruhleder, 1996, p. 113). We should ask, therefore, ‘when’ something constitutes infrastructure, not ‘what’ it is. Taking this into account and extending the definition by Lisa Parks and Nicole Starosielski (2015, p. 4), digital infrastructures are situated sociotechnical arrangements that are designed and configured to support the distribution of digital signal traffic.

The crucial point here is that the functionality of digital media requires the existence of a globalized digital infrastructure. It is not just the internet itself but also the data centers in which data are stored, which can be accessed in various applications we understand as digital media and which constitute the physical reality of the “cloud” (Durham Peters, 2015; Mosco, 2014). Furthermore, software itself is increasingly gaining its own infrastructure status. Examples include platforms such as Facebook, Google Maps, or the Apple and Google app stores. For many digital media, the use of these platforms is a prerequisite for being able to function as a medium at all. Platforms which are digital media themselves become infrastructures for other digital media (Plantin and Punathambekar, 2019; van Dijck, Poell, and de Waal, 2018).

It is crucial for any reflection on today’s social order that these globalized infrastructures do not simply serve the functioning of these digital media. They form the basis of a comprehensive collection of data across the various networks, assemblages, and figurations in which digital media are used. As mentioned above, with practices entangled with digital media all of us leave behind digital traces that are processed into a ‘data double’. In utopian visions of the digital era, the availability of these data provides an insight into patterns of human practice, enabling an optimization of social order based on endless information streams. At least, this was the hopeful aspiration expressed in the discussion on ‘big data’; one idea being the ability to prevent future crime by means of outreach and police work through online datasets (Mayer-Schönberger and Cukier, 2013). However, as recent research has shown (Elmer, Langlois, and Redden, 2015) – and as Snowden’s NSA leaks revealed – these data are often used by companies for highly personalized advertising (Turow, 2017) and by state agencies for surveillance (Hintz, Dencik, and Wahl-Jorgensen, 2019). Recent studies have shown that knowledge about this kind of surveillance is comparatively widespread but increasingly considered fatalistically by the population at large (Dencik, 2018; Draper and Turow, 2019).

Recent research has argued that the status of digital media, their infrastructures, and the data they produce has made capitalism an even stronger force for order than it already was in modern societies. Two of the more comprehensive approaches in which this consideration becomes tangible are those of ‘surveillance capitalism’ and ‘data colonialism’. Shoshana Zuboff (2019) refers to “surveillance capitalism”, which she connects to value creation around the mass analysis of data. Surveillance capitalism’s starting point is the one-sided use of human experience as a resource and its transformation into behavioral data. Companies initially argued that they used these data simply to improve their products and services, however, the sheer volume of data they have been able to accumulate has led to new products and services, namely those related to behavioral prediction and facial recognition. Assumptions of future developments constructed in this way are, in turn, traded at the level of financial markets. Nick Couldry and Ulises Mejias (2019, pp. 3–35) go one step further in their approach to “data colonialism”, which occurs as a result of the increasing interrelatedness of figurations through digital infrastructures and the processing of data. The reason why they speak of data colonialism is that, similar to colonialism in the regular sense of the term, they identify a pattern of ordering that structures the entire world with ever-increasing intensity: The new relationality of figurations by digital infrastructures leads to the commercial handling of data which is primarily concerned with the external appropriation of data on terms that are partly or wholly beyond the control of the person to whom the data relates. These data are then used by technology companies to generate their added value (while at the same time enabling state surveillance). As more facets of everyday life are monitored by digital media, the result will be the “capitalization of human life without limit” (Couldry and Mejias, 2019, p. 5): Through data, capitalism develops new interrelations with every corner of our lives, and human practice becomes structured along capitalist principles more intensely than ever before.

The new ordering of capitalism is but one example of the transformation initiated by the interrelatedness of networks, assemblages, and figurations through digital infrastructures and data, and we must be cautious to not reduce deep mediatization to these ideas. However, it clearly demonstrates the extent to which the establishment of the social order is changing in the context of deep mediatization processes. On the one hand, the complexity and contradictoriness of existing patterns of everyday practice in today’s late modern societies become visible and negotiable for the first time via digital media and their infrastructures. Ultimately, this is embodied by Armin Nassehi (2019), when he argues that digitalization has provided today’s societies with the technologies required for observing and negotiating their complexity. On the other hand, digital media enable companies, state agencies, and other regulatory institutions to penetrate much deeper into everyday practice than was previously possible. It is perhaps at this point where digital media’s power can be seen.

6 Conclusions: A ‘media-sociological imagination’ for a deeply mediatized social world

This article has primarily been an attempt to sort out from a practice-theoretical and relational perspective the strands of media sociology’s shift into the digital. In doing so, I have isolated three lines of discussion. First, the rethinking of agency; second, rethinking social relations; and third, rediscovering social order through the lens of digital media and infrastructures. In relation to all three points, it can be said that they are also linked to a reassessment of classical media-sociological questions of inequality and power, especially with reference to ethnicity, class and gender. In hybrid figurations of humans and communicative robots, for example, inequality and power relations are still (re)produced, but media and communication technologies have a different status here than in the age of legacy media. A special feature of digital media is that they are not only a means of communication, but also a means of data generation, which always raises the question of how inequalities and power relations are inscribed in those data and how they are processed. What can we conclude in trying to establish a ‘media-sociological imagination’ of a deeply mediatized social world?

To clarify, it is useful to return to C. Wright Mills’ (1970) original idea of the “sociological imagination”. Two basic ideas come together here. On the one hand, it is about developing a view of the extent to which individual conditions and experiences always refer to society and its structure. As Anthony Giddens and Philip Sutton (2013, p. 5) put it, a “sociological imagination” means “to be able to break free from the immediacy of […] personal circumstances to set things into a wider context”. Like Norbert Elias and the concept of figurations (Elias, 1978), the idea is to overcome the opposition between the individual and society. On the other hand, with his idea of the “sociological imagination”, Mills also implied that the study of these interrelationships should take place critically and with a mind to the societal problems at hand. So, it is not necessarily about developing an abstract general theory of society but about developing theory that reflects the struggles, conflicts, and problems of present societies.

Both are important points when we ask what a ‘media-sociological imagination’ should look like in times of deep mediatization: We must reflect the changing relationships between the individual and society through digital media, and we must keep in mind the societal issues of the present time. With this in mind, I have formulated three conditions for a media sociology for a deeply mediatized social world.

The first is that, in a best-case scenario, media sociology could become a cross-sectional sociology. This relates to the basic difference between the sociology of mass communication and a sociology of a deeply mediatized world. The first kind of media sociology works in a similar way to other areas of sociology that are typically defined by the social domain they address, such as ‘economic sociology’, for example, which addresses the economy, ‘political sociology’, which addresses politics, or the ‘sociology of religion’, which addresses religion. However, as I have argued in this article, a sociology of a deeply mediatized world ‘overcomes’ such a domain-centered orientation because it attempts to consider the role played by digital media and their infrastructures across all domains of society.

This is certainly a considerable challenge for any work in media sociology since it is a matter of incorporating both the theoretical and empirical knowledge of social science research in very different domains. However, it has emerged as a pressing question, namely, to ask what changes when digital media and their infrastructures become the basis for construction processes across the whole of society. A more “holistic” (Hepp and Loosen, 2020, p. 1) perspective is needed than has been absent from many detailed studies. A media sociology oriented in this way could develop a particular potential especially in a time of increasing “fragmentation and hyper-specialization” (Waisbord, 2019, p. 10) of knowledge in media and communication research: A media sociology that argues and thinks from a cross-sectional perspective can act to ‘bridge’ the different areas of media and communication research.

The second condition requires that the media sociology described here is a sociology of entanglement. As many examples in this article have shown, in times of deep mediatization technologies such as communicative robots cannot simply be considered a counterpart to human practices but only in relation to their entanglement with human practices. If we take this seriously, we begin to develop something we might call a ‘hybrid approach’ to agency: an approach that neither understands media technologies merely as a ‘delegation’ of human agency nor as an agency in its own right. Instead, it seeks to understand how new forms of agency develop in the entanglement of human practice and the latest media technologies. In an era when data processing and AI become a general characteristic of society, this is not just a question of media sociology.

Again, this has a lot to do with the challenges of the present. As processes of automation based on digital media and infrastructures, which also contribute to algorithmic decision-making in many areas of society, permeate evermore societal domains, one of the most pressing social questions is how these technical systems relate to people and how we want to shape this relationship (Pasquale, 2020, pp. 1–19). At best, therefore, media sociology can contribute to dealing with one of the greatest challenges posed by current media-related transformations.

A third condition is that media sociology should become a new critical sociology of technologically-based deep structures. One central point of this article was that the question of social order emerges in a new way within the context of deep mediatization. Aided by globalized digital infrastructures and the possibility of continuous data collection, new questions related to the creation and maintenance of social order arise. Media sociology has, therefore, developed an eye for how power relations are (re-)established through digital media, their infrastructures, and the continuous processing of data. For this scenario, concepts such as ‘surveillance capitalism’ or ‘data colonialism’ work as highly functional metaphors

As social order is increasingly ‘molded’ by the deep embedding of networked communication technologies, media sociology should, at the very least, develop into a critique of these novel technologically-based power relations – a critique that addresses once more some of the most urgent societal questions of the present: How do the possibilities for the individual’s life change when, through digital media and their infrastructures, overarching social structures become even more ‘closely’ interwoven with the practice of everyday life?

Certainly, all three demands represent far-reaching challenges for the development of a ‘media-sociological imagination’ situated in a deeply mediatized social world. But at the same time, they demonstrate the potential of a media sociology that develops a comprehensive view of our current, deeply mediatized society. Developing this potential will continue to be a common task for all those interested in media sociology.