Researching Transforming Communications in Times of Deep Mediatization: A Figurational Approach

The purpose of this chapter is to outline the concept ‘figurational approach’ as one possible way of researching transforming communications. To do this, the authors reflect on our changing media environment, which is understood as being marked by deep mediatization. Second, they develop the argument that mediatization research is right to emphasize the domain specificity of (deep) mediatization. However, there is a need to sharpen the idea of social domain. With this in mind, the chapter demonstrates how it can be a help to investigate transforming communications by analyzing changing ‘communicative figurations’. Finally, in the conclusion, some remarks are made about what this means for practical empirical research.

However, nowadays the situation has become much more complicated. With a huge variety of different media and their spread across very different domains of society, it no longer seems appropriate to conceptualize any certain kind of media content or medium as the 'driving force' that is changing society. We have to accept that any possible influence of media as contents and technologies arises out of 'cross-media' (Bjur et al. 2014) and 'polymedia' (Madianou and Miller 2013) situations, or to be more specific, a 'media manifold' (Couldry and Hepp 2016: 11, 53). By 'manifold', we refer not just to the plurality of today's media channels and interfaces, but their interlinked nature and the many-dimensional order that results from this and encompasses our whole media environment. In addition, we have to consider that what we call 'media' has been changed fundamentally by digitalization and a related datafication. nowadays, more and more media-'new' as well as 'old'-are becoming digital. This means not just that they rely on a digital infrastructure which is closely related to the internet. At the same time and much more far-reaching, it also means that all media are tending to be based on software, which means algorithms become part of our media-related sense-making. Media, nowadays, are no longer simply means of communication but at the same time and additionally are means of collecting data about us as their users in real time.
With all these changes, the question of 'transforming communications'-that is, how media change communication and by that our social construction of reality-has not lost any of its relevance. In fact, the question has become even more important as media-related influences enter different societal domains. Mediatization research argues that we can notice a 'domain specificity' of mediatization (Hjarvard 2013: 4;nieminen 2014: 64;Lunt and Livingstone 2016: 1), while remaining rather vague about what 'social domain' precisely means and how we can theorize it properly. The challenge at this point is to clarify how we can conceptualize the very different domains of society so that we are able to undertake comparative research on (deep) mediatization across them.
This chapter aims to outline one possible approach to reattempting the research of transforming communications. We closely relate this to what we call a 'figurational approach'. This is a perspective that moves the figurations of human actors into the foreground and at the same time takes into account how far these figurations are entangled with media as contents and technologies, which on a deeper level refers both to media organizations and infrastructures. Through stepwise and comparative empirical research on media-related changes in certain communicative figurations, we can gain a bigger picture that shows the more complex processes of societal transformations.
In this introductory chapter, we want to outline this figurational perspective and in so doing develop an approach for empirically investigating transforming communications. To do this, we will first reflect on our changing media environment, which we understand as marked by deep mediatization. Second, we will argue that mediatization research is right to emphasize the domain specificity of (deep) mediatization. However, we need to sharpen the idea of social domain. On this basis, we want to argue how far it is helpful to investigate transforming communications by analyzing changing 'communicative figurations'. Finally, in the conclusion we will make some remarks about what this means for practical empirical research.

the changIng MedIa envIronMent In tIMes of deep MedIatIzatIon
The idea of mediatization is a particularly helpful starting point to describe how changes in the media environment are part of an overall 'meta process' (Krotz 2007: 256). 1 This is related to other meta processes of change: mainly individualization, globalization and commercialization. Mediatization is a long-term and non-linear process traceable back at least to the beginning of various modernities (Thompson 1995;Meyen 2009;Hjarvard 2013;Esser and Strömbäck 2014;Lundby 2014b). In essence, the term mediatization captures on the one hand the increasing spread of technologically based media in society; and on the other hand how different social domains are being more and more shaped by these media. As we have already emphasized, this process has fundamentally intensified over the last decade. To approach this, we wish to use the term deep mediatization. By calling the contemporary mediatization deep, we want to indicate that with the recent wave of digitalization, mediatization has entered a new stage 2 : it is no longer expedient to grasp the social impact of 'media' merely as the influence of a distinct domain (i.e. journalism) which is separate from other domains of the social world (Livingstone 2009: 2-4). no matter which domain of society we consider, its formation is in one way or another related to the technologically based media of communication, which are all becoming digital. Deep mediatization is by no means homogeneous or linear. It is highly complicated, contradictory and a conflict-driven process. nevertheless, in the Western hemisphere, deep mediatization takes place across societies as a whole. Yet even when we strive to escape from this all-encompassing contemporary mediatization-for example, individuals who refuse to use certain (digital) media in an attempt at 'coping' (Schimank 2011: 459-462) with being reachable at all times of the day and night, or organizations that introduce email-free holidays-such behaviour merely constitutes what we can call temporary 'oases of demediatization', in loose reference to Hartmut Rosa (2013: 87). In this sense, popular self-help literature on 'mindfulness'-the practice of bringing one's attention to things occurring in the present moment, beyond any mediated communication-is less about any durable containment of mediatization: it is rather an expression that deep mediatization includes spaces of self-reflection and controlled escape in order to remain manageable for us as human beings.
With respect to these arguments, the concept of deep mediatization is neither an attempt at a closed theory nor a limited theoretical approach. There are various traditions of mediatization research, and such a range is needed because of the complexity of the field. 3 However, across these different traditions, we can at a first level understand mediatization as a 'sensitising concept' (Jensen 2013: 213-217;Strömbäck and Esser 2014: 4;Lunt and Livingstone 2016: 464); that is, a concept that 'gives the user a general sense of reference and guidance in approaching empirical instances' and that 'merely suggests directions along which to look' (Blumer 1954: 7). This means to look at the overall spread of different media and the related changes in various social domains (Schulz 2014: 58-62). Using the term deep mediatization makes us 'sensitive' to how far mediatization nowadays progresses into what has been called 'mediatized worlds' (Hepp and Krotz 2014: 6) and a 'mediatized way of life' (Vorderer et al. 2015: 259).
At a second level, and departing from this, we need further concepts and approaches to describe in detail how the transformation that we relate to the term mediatization actually takes place. While we have a rough estimate of the processes and practices that constitute deep mediatization, we still lack thorough empirical investigations.
Reflecting this specificity of different phenomena of media-related changes and their particularities, it is nevertheless striking that they are all confronted with certain trends that characterize the change of the present media environment. If we understand the 'media environment' as the entire body of available media at any given time in society (Livingstone 2001: 307;Hasebrink and Hölig 2014: 16;Jensen and Helles 2015: 292), we can initially distinguish at least five such trends: first, a differentiation of a vast number of technologically based media of communication; second, an increasing connectivity of and through these media, which offers the possibility to individually and collectively 'link' across space and time; third, a rising omnipresence of media that creates the possibility to connect permanently and everywhere; fourth, a rapid pace of innovation, the emergence of 'new' media and services in ever-shorter periods of time; and fifth, a datafication, which is the representation of social life in computerized data via media devices and their underlying software and infrastructure.
none of these trends is to be seen as a separate individual media phenomenon; rather, they are all closely linked with each other, and altogether they are characteristic of the present changes in our media environment (Bjur et al. 2014: 15). We have to be aware that these trends are not 'linear'. It is also uncertain whether these trends will continue or whether other trends will emerge. In addition, they are highly contradictory in themselves. However, altogether they are manifestations of deep mediatization, and distinguishing between such trends provides us with a first understanding of the media-related changes in which we are involved.
The trend for differentiation in the media means that the number of media and their functionalities have increased over recent decades. While in the beginning there was a discussion concerning whether digitalization might result in the dominance of the computer as the sole 'metamedium' (Kay and Goldberg 1977;Höflich 2003), 4 it turned out that the result of digitalization was rather the arrival of a variety of very different media, which at the present stage are becoming more and more digital and increasingly based on software (see Manovich 2013). The differentiation of media gives rise to a variety of contradictory impacts. While digital media might support self-paced learning for young people and adults (Wolf 2015;Wolf and Wudarski in this volume), the same media can be used to build up authoritarian relationships in religious organizations (Radde-Antweiler 2015; Radde-Antweiler and Grünenthal and Gogolok in this volume). Reflecting both these aspects, across the variety of possible consequences we can assume that differentiation might result in an optionality (Rusch 2006) of ways of use. These can be related to processes of individualization (Hasebrink 1999) and, following from this, contingency within and across social domains and related questions of inequality and power. This can have various further influences on the segmentations, exclusions and divides articulated in a specific social domain (van Deursen and Helsper 2015; nieminen 2016). For example, an increasing number of media as contents and technologies might weaken (as our preliminary research shows) the binding power of communicative practices within communities (Marszolek and Robel 2016), and the variability of possible contacts might increase (Friemel 2013). This may be especially discussed in relation to internet-based contact platforms, which are understood as supporting 'weak ties' instead of 'strong' relations within the direct living environment (Wittel 2008;Rainie and Wellman 2012: 131-134).
The media environment of deep mediatization is characterized by the trends of an intensified connectivity. By connectivity, we primarily mean the interconnectedness of various media owing to their digitalization and the infrastructure of the internet. This is the case for 'old' media such as television and the digital press, but increasingly and with reference to personal communication for 'new' media such as online platforms and mobile phone applications. As a consequence, there is a close relationship between more recent processes of mediatization and globalization (Krotz 2008). A characteristic of contemporary everyday life is our ability to socially connect globally, across various media, if we want to. But, '"connectivity" does not necessarily mean "social connectedness"' (van Dijck 2013: 4). Increasing media connectivity can result in a spatial extension of processes of construction (Wessler and Brüggemann 2012: 119-136;Hepp 2015: 13-18), and through that social domains can extend and their borders become blurred. This might 'disembed' (Giddens 1990: 20) social processes being maintained across large distances. For example, it can become easier to build networks for learning over long distances (Thomas and Brown 2011: 53;Ito et al. 2009: 213), popular cultures can exist transnationally (Buckingham and Kehily 2014) and whole organizations or networks of organizations can be built up across various locations (Breiter 2003;Ribes et al. 2013;Lammers and Jackson 2014: 33-47;Jarke 2015)-all of this held together by technologically based communication. However, we must be careful in assuming any one single set of possible consequences of media's connectivity. In other words, the further consequence of connectivity is very much context dependent.
Besides the increasing differentiation and connectivity, the social, temporal and spatial spread of media relates to their omnipresence. Face-toface meetings, talks or walking and other social situations, which for a long time were not related to media, have nowadays become so in one way or another. These dynamics are especially propelled through the spread of mobile communication technologies (Katz and Aakhus 2002;Ling and Donner 2009;Goggin 2011;Vorderer et al. 2015). It has become possible to be 'always on' (Chen 2011: 63) and 'constantly in touch' (Agar 2003: 22); that is, reachable at any time. This omnipresence of various media can result in an increasing 'acceleration' (Rosa 2013: 41-43) of social processes. We might, for example, expect immediate answers, a quick delivery and a fast response. With reference to this, social domains can be marked by new temporalities, especially with expectations of a new 'immediacy' (Tomlinson 2007: 72-93) of communicative reaction. Arguably, the result of this is a general acceleration of life (Wajcman 2015: 13-35). This can be the case in the sphere of work, and also in our private life. More recent research indicates that the omnipresence of media also stimulates a new appreciation of 'media-free' situations and spaces, in highly institutionalized contexts such as politics (Pritzlaff-Scheele and nullmeier in this volume) as well as in private life (Roitsch 2017). It is, again, worth noting that substantial differences exist between one social domain and another.
A rapid pace of innovation has accompanied recent media developments. This means that the time sequence of more or less fundamental media innovations has-at least in the perception of many media usersshortened over the past few decades (Rosa 2013: 71-74). 5 This pace of innovation might result in a constantly perceived adjustment pressure, a perceived pressure to 'conform' to these changes with a possible breakdown of the ability to adapt. While various innovations surrounding the smartphone and its apps have become widespread, the most recent assumption is that the 'internet of things' and its 'locations awareness' might once again change 'everything' (Greengard 2015: 60). However, we should generally be cautious about any 'rhetoric of the technological sublime' (Morley 2007: 235) related to the present pace of innovation, because complex articulations of segmentation and exclusion are evident as they are reflected in such concepts as divide (norris 2001;van Dijk and Hacker 2005;Zillien 2009;Livingstone and Helsper 2007;Tsatsou 2011). Being able to appropriate and adjust to certain media innovations means to be in a power position, no matter whether this is within the family, a group of friends or in certain organizations, especially when it comes to questions of regulation (Schulz et al. 2011). Even an attitude of openness towards innovations might privilege entire social groups as 'pioneer communities' (Hepp 2016), such as, for example, the Quantified Self movement vis-à-vis other social groups such as excluded homeless people (Koch 2016a). An outcome of all this can be segmentation between different sections of the population (Drgomir and Thompson 2014;Friemel 2016).
The term datafication refers to digitalization: a growing number of media are based on software. As a result, through the use of these media we leave 'digital traces' (Karanasios et al. 2013), data that can be aggregated and processed in automated ways on the basis of algorithms. This is the case across the variety of digital media platforms (van Dijck and Poell 2013), which are also understood as 'social software' (Stegbauer and Jäckel 2007: 7-10). In public discourse, this change of the media environment is mainly discussed with reference to the concept of 'big data' (Mayer-Schönberger and Cukier 2013; for a critique see boyd and Crawford 2012; Lohmeier 2014). This means that the representation of social phenomena by quantified data plays an increasing role in societal self-understanding and self-conception, with the result that technical intermediaries (search engines, platforms, etc.) disguise agency by 'quantification' (Passoth et al. 2014: 281-283;Pasquale 2015: 32-38). On the other hand, there is the hope of new, technologically based forms of transparency that might support participation as it is discussed, for example, with reference to open data and smart cities (Townsend 2013;Koch 2016b: 210, 218). Furthermore, such a datafication can result in a stabilization of sociality, which is 'society made durable' (Latour 1991: 103). At the same time, as the public debate following Edward Snowden's revelations has illustrated (Schulz 2013), new possibilities of surveillance emerge-for governmental agencies (Fuchs 2013;Lyon 2014) as well as for private actors (Andrejevic and Gates 2014; Christensen and Jansson 2014).
As we have already pointed out, one must be cautious about the trends of deep mediatization we outlined above: these are preliminary interpretations on the basis of the general state of media and communication research. Keeping the uncertainty about their directedness and future stability in mind, these trends offer us guidance in respect of how our media environment is changing with the process of deep mediatization.

the doMaIn specIfIcIty of deep MedIatIzatIon
Beyond detailed research results, existing studies on mediatization agree that any process of mediatization is very specific in relation to the social domain under consideration. This term social domain is used by various representatives of mediatization research (amongst others Hjarvard 2013; Ekström et al. 2016;Lunt and Livingstone 2016), while coming close to the everyday understanding of 'spheres' of society. In its widest sense, the term 'social domain' refers to those 'spheres' as being meaningful in everyday practice. The scaling of the different 'meaningful domains' can be very different, reaching from certain social groups or organizations to whole social fields or systems. We can understand this scaling to be a problem of terminological blurriness. However, the main argument being pushed forward in mediatization research is different. By hinting at the domain specificity of mediatization, scholars want to emphasize the variety of mediatization across different spheres of society. Mediatization is not a homogeneous process but very much differs from one area to another. It is a 'domain-specific' phenomenon.
Mediatization research investigates its 'domain specificity' with reference to such different theoretical conceptualizations and scalings. For example, there is a discussion about the mediatization of different 'social fields' in Bourdieu's understanding (Couldry 2012: 144-153), of different (sub-)systems in the sense of Luhmann (Kunelius and Reunanen 2016: 8-12), or different 'social worlds' in the sense of phenomenology and symbolic interactionism (Hepp and Krotz 2014: 6-9). Therefore, using the term 'social domain' within mediatization research does not have the intention of suggesting that these different theoretical conceptualizations are the same. Rather, using the less theoretically loaded term 'domain', it emphasizes one fundamental empirical result across these theoretical conceptualizations within mediatization research: mediatization takes place very differently in different spheres of society.
To empirically operationalize such a domain specificity, the level of society as such seems to be inappropriate and even the level of whole fields of society seems to be too general. One becomes able to operationalize such a kind of research as soon as one moves to what is called the 'meso-level' (cf. Donges and Jarren 2014: 181-182), that is the level of specific kinds of collectivities (groups, communities, etc.) and specific kinds of organizations (enterprises, schools, etc.). We are able to investigate the domain specificity of (deep) mediatization if we move to a level which is called in actors-centred sociology 'supra-individual actors' (Schimank 2010: 327-329), that is a structured constellation of individual actors: collectivities then become concrete as 'collective actors', whose members share certain practices of meaning construction and reciprocal observation. Organizations become 'corporative actors' whose shared practices are based on binding agreements.
Approaching from such a point of view the 'domain specificity' of (deep) mediatization makes a specific challenge of mediatization research explicit. It is less helpful to understand (mass) media as a domain of their own in the sense of a certain social field or system, and to investigate their influence on other social domains in a way that 'media logic' (Altheide and Snow 1979) would colonize the logics of other domains. For a critique of such an approach see for example (Strömbäck 2008;Esser 2013;Landerer 2013), who all ask for further concretization of 'media logic as a metaphor' (Hjarvard 2017). Beyond such a need to make the idea of media logic more concrete, there is a second problem about theorizing media as a domain of its own that relates closer to the character of present deep mediatization: when digital media permeate the various domains of society (Livingstone 2009: 2f.), it becomes more complicated to see them as a domain of their own.
To substantiate this, we can think about collectivities. To research, for example, the question about how the communicative construction of families transforms with today's media changes, questions arise regarding which representations of the family become communicated by mass media (and if the 'logic' of journalistic coverage has consequences for an everyday construction of the family). In addition, further questions matter: for example, how communicative networking in the family takes place, how unknown or forgotten family members are 'found' by the algorithms of Facebook, how family members construct their family memories by exchanging digital images and so on. Again, we are confronted with the necessity of reflecting the cross-media and technologyrelated character of present communicative constructions of the family (Hasebrink 2014: 232).
Generalizing this, there is a certain paradox. This is that today's media of communication are not a domain on their own. They are a phenomenon across domains. At the same time the characteristics of the transformations that relate to these media are 'domain specific', and we have to have different levels of scaling in mind. This domain specificity becomes especially concrete at the level of supra-individual actors, that is collectivities and organizations.

researchIng transforMIng coMMunIcatIons
Having an understanding of the changing media environment of deep mediatization, and within this the domain specificity of related transformations as outlined so far, it is evident that the possible consequences of a changing media environment can differ depending on context. But how can we research and compare the possible consequences of a changing media environment with reference to very different social domains? Basically, we are confronted with the challenge of firming up the idea of social domains in a conceptual framework. This framework has to be substantiated sufficiently enough to offer a stable design for collaborative empirical research, comparison and theory development; and it has to be flexible enough to reflect the specificity of the social domain under investigation. This is where an actor's point of view is particularly important, as it is related to the figurational approach we want to outline. From such a point of view, two aspects matter above all. First, a changing media environment can develop only if practices change. When it comes to media, these are predominantly practices of communication. Second, such changing practices are not just individual phenomena; they have to be analyzed with respect to the social domains in which humans act. We refer here to the already mentioned concept of communicative figuration.

Communicative Practices and Their Entanglement with Media
In media and communication research, approaches that move agency and social practice into the foreground have a long tradition and can be traced back to the beginnings of sociology. This perspective first peaked in the 1970s across different areas, such as audience research (Teichert 1972;Blumler and Katz 1974;Renckstorf and Wester 2001) or cultural studies (Hall 1973;Morley 1980;de Certeau 2002) that no longer considered media users as 'dopes' but as persons acting reflexively with media, being situated in a wider social and cultural surrounding. On such a basis, it became common to consider people as actors who 'deal' with media (Hasebrink 2003;neumann-Braun 2000;napoli 2010;Bonfadelli and Friemel 2014)-no matter whether they come from the side of media production, media use or various kinds of hybrids (Bruns and Schmidt 2011).
Based on this tradition, we can witness a recent and more focused move in research towards media practice. 6 This has to be seen in the wider context of a practice turn in the social sciences. 7 There are two aspects to be learned from this development: first, to consider every activity as 'embodied', and second, to consider the nexus of practices with 'artefacts, hybrids, and natural objects' (Schatzki 2001: 11).
When it comes to human acting, a practice approach is interested in the 'embodied doing' of an activity as such. This doing is based on 'practical consciousness' (Giddens 1984: xxiii), which is learned in highly contextualized ways. Based on this learning, practices can be realized in a meaningful way without being 'discursively' accessible to the actors; that is, they cannot personally explain their doing, and this is also the case for communication. 8 'Practical consciousness' as embodied capacity is rather understood as know-how, skills, tacit knowledge and dispositions, related to the habitus of a person. Most practices are based on this 'practical knowledge', which has its own potential for situational creativity. Practices are anchored in the body and cannot be described as a mechanical obedience to rules. In this sense, practices of communication-with media but also without-are also embodied and have to be considered in their interrelation to other forms of practice (Bourdieu 1977: 16-22;Reichertz 2009: 118-120).
The argument that we should focus on the entanglement of practices with objects is of special interest, because with deep mediatization communicative practice increasingly turns into a media-entangled and therefore object-related practice. Here, practice theory itself puts emphasis on the media as a specific kind of object when it comes to the production of meaning: 'writing, printing and electronic media mould social (here, above all, discursive) practices' (Reckwitz 2002: 253). This is the reason why many communicative practices are 'media practice[s]' (Couldry 2004: 125); that is to say, they are undertaken in relation to media.
Following this line of reasoning, we can understand practices of communication as complex and highly contextualized patterns of doing. Or to put it another way, certain forms of communicative action build up complex practices of communication as they are realized today in the increasingly complex media environment of the media manifold. Communication therefore involves the use of signs that humans learn during their socialization and which, as symbols, are for the most part entirely arbitrary. This means that the meaning of communicative practices depends on social conventions. Practices of communication are fundamental to the human construction of reality: we 'create' the meaning of our social world in multiple processes of communication; we are born into a world in which communication already exists; we learn what is characteristic of this social world (and its society) through the (communicative) process of learning to speak; and when we proceed to act in this social world our practices are always also communicative practices.
This understanding of communication has certain implications for conceptualizing media. Putting aside symbolic generalized media of influence such as 'love' or 'money' (Luhmann 2012: 190-238) and focusing on technologically based communication media of 'second order' (Kubicek 1997: 218-220), we can understand them as means of communication, distinguished by specific technologies and their infrastructures, a system of signs and various institutionalizations and organizations that furnish us with services for communicative practice (Beck 2006: 14). Media of communication 'institutionalize' and 'objectify', that is to say 'materialize' symbol systems and practices (Berger and Luckmann 1967: 49-61;Fornäs 2000;Knoblauch 2013: 300f.;Couldry and Hepp 2016: 15-33). This is how they 'mould' (Reckwitz 2002: 253) communication. With deep mediatization, the challenging question is the 'moulding influence' of a medium in its respective typical constellation with other media. We have to address this constellation on at least three levels. These are, firstly, the level of the entire media environment. As we have already noted above, what we mean by media environment is the entire body of available media at any given time. Secondly, there is the level of the media ensemble. This is the subset of the media in a media environment as it is used in a particular social domain (family, company, etc.) with respect to the available options (Bausinger 1984: 349). Thirdly, there is the level of media repertoire. This is the individuals' selection of the media as they use and appropriate them as part of their everyday practices (Hasebrink and Popp 2006).
With deep mediatization, our practices of communication typically reach across media. When we inform ourselves with reference to a certain topic, we talk with people, we email others, read online articles and possibly books, and we might 'ask' Apple's software assistant Siri to search for information on the internet. Therefore, when it comes to the question of how our social domains are moulded by media, we have to consider such cross-media influences with regard to various types of communication.

Social Domains as Communicative Figurations
Investigating transforming communications from a cross-media and therefore 'non-mediacentric' point of view entails defining the starting point of analysis, 9 via the social entity-the 'social domain'-under consideration. But this is exactly the point where we have to become clearer about what we have, up to this point, loosely called the 'meso level' of social domains.
To theorize this further, the process-sociological approach of norbert Elias (1978) is of great help and importance. 10 Elias identified two problems for any social analysis: the relative autonomy but co-dependence of individuals and society, and the distinction between social change (the fact that each progression of life means variances) and structural transformation (fundamental changes in society). His solution was to argue that structural transformation could be explained in terms of the shifting relation between individuals and society through time. Elias referred to these dynamics as figurations-or as we would put it, as figurations of certain social domains. Figurations are 'networks of individuals' (Elias 1978: 15) or, in more encompassing terms, actors, including collectivities and organizations. These actors constitute, by their interaction, larger social entities. Therefore, figuration is a 'simple conceptual tool' to understand social domains in terms of 'models of processes of interweaving' (Elias 1978: 30, 130).
A development that Elias could hardly reflect, though he had some presentiment of it (Elias 1991: 163), is that today many figurations are made up by the use of media. This is one possible driving force of their transformation: the figurations of individuals, collectivities (families, peer groups, communities, etc.) and organizations (media companies, churches, schools, etc.) change with their media ensembles. In addition, deep mediatization makes new figurations possible, such as online gatherings in chatrooms, on platforms or through apps. But there are even further developments. nowadays, some figurations are entirely built up by media technologies. One example is the 'collectivities of taste' as they become represented by groups of people with the same shopping interests in online stores such as Amazon (Passoth et al. 2014: 282). Other examples are 'networked media collectivities' (Friemel and Bixler in this volume) that are constituted around certain media events and topics.
From a media and communication research point of view, we can consider each figuration as a communicative one: practices of communication are of high importance when it comes to a meaningful construction of the respective figuration. Communicative figurations are (typically cross-media) patterns of interweaving through practices of communication. Members of families as collectivities, for example, are possibly separated in space but connected through multi-modal communication such as (mobile) phone calls, emailing, sharing on digital platforms and so on that keep family relationships alive (Madianou and Miller 2012;Hepp et al. 2015) and allow the construction of family memories (Lohmeier and Pentzold 2014). Or organizations as communicative figurations are kept together with the help of databases and communication across the intranet, as well as printed flyers and other media of internal and external communication. Individuals are involved in such figurations through the roles and positions they have in the respective actor constellations. An approach of media and communication research that starts with figurations, therefore, is able to link perspectives on individuals, collectivities and organizations in a productive way.
Taking such a perspective, there are at least three features that are characteristic of a communicative figuration (see Hepp and Hasebrink 2014: 260-262;Couldry and Hepp 2016: 66f.): • First, a communicative figuration has a certain constellation of actors that can be regarded as its structural basis: a network of individuals who are interrelated and are communicating amongst themselves. • Second, each communicative figuration has dominating frames of relevance that serve to guide its constituting practices. These frames define the 'topic' and therefore the character of a communicative figuration. • Third, we are dealing with specific communicative practices that are interwoven with other social practices. In their composition, these practices typically draw on and are entangled with a media ensemble.
Investigating communicative figurations offers us a cross-media and processual meso-level approach to the construction of what are called 'social domains' in mediatization research as well as their transformation through deep mediatization. Today, we are confronted with various dynamically changing media-related figurations. We gain access to them by researching their actor constellations, frames of relevance and communicative practices, all of which are entangled with a media ensemble. Summing up this understanding of communicative figurations and referring back to the main trends in a changing media environment, we can visualize such an analytical approach as follows (see Fig. 2

.1):
A changing media environment moulds the communicative figurations of social domains-their actor constellations, frames of relevance and communicative practices. As outlined above, with deep mediatization we may expect at present five dominant trends in a changing media environment: a differentiation of media, increasing connectivity through various media, their rising omnipresence, a rapid pace of innovation and datafication of human interaction through media. It depends on the social domain under consideration how strongly these trends shape or Based on the research discussed so far, we can assume a number of possible consequences as a hypothetical set: the optionality, contingency and chances of participation in social domains might increase; social domains' communicative figurations might extend spatially; their borders might blur; there might be an acceleration and increasing immediacy within and across them; a disguise of agency might come about; media technology might stabilize sociality in social domains; social surveillance might take place; or all might result in segmentation, exclusion and division. While these assumed consequences are a starting point for future research, it remains an open question as to which of them is characteristic for which social domain, how these different consequences interfere with each other and even if there might be further consequences we are not aware of at present. In addition, we have to consider the different ways in which social domains relate to these trends in deep mediatization's changing media environment. They can be supportive of such changes, for example by always appropriating the latest media. Alternatively, by rejecting certain media, they can hinder these trends.
For any empirical research, we need to have the dual character of possible consequences in mind. On the one hand, a changing media environment might have 'internal' consequences for a social domainfor example, optionality, disguise of agency or segmentation of figuration might take place. This is the case, for example, when relations in an organization change partly owing to the media that are used for  (Loosen 2014). The same can be said for families in which the segmentation of knowledge transforms when digital media become part of the family's memory construction (Lohmeier and Pentzold 2014). On the other hand, there might be 'external' consequences: figurations also change in their relation to each other. If journalism organizations change, for example, their relationship with audiences transforms: we are confronted with so-called '"blurring boundaries" of journalism' (Loosen 2014: 68). Detailed comparative empirical research into the communicative figurations of different social domains can offer us the chance to make more general statements about transforming communications, focusing on individuals, collectivities and organizations.
2.5 coMMunIcatIve fIguratIon as an approach for eMpIrIcal research By investigating communicative figurations, we therefore adopt an open analytical approach that gives us the chance to research the transformation of social domains with deep mediatization. This approach is open to various macro-concepts of society such as 'network society' (Castells 2000), 'media society' (Imhof et al. 2004), 'communication society' (Münch 2002), 'next society' (Baecker 2007) or a 're-assembling of the social' (Latour 2007). Such concepts offer more general considerations of how the social world might transform with the changing media environment, and are therefore an important source for posing empirical questions about media-related changes. Yet, as we are living in the middle of the changes we capture with the term deep mediatization, it might be too early to draw conclusions about any particular communication model of media-related transformations of society. Taking into account all of the above, we still need further detailed comparative research on different social domains before we can make general claims. For this kind of research, communicative figuration constitutes a highly productive 'bridging concept' because of its process perspective on practices and its emphasis on actor constellations. 11 The concept of figurations links a micro-analysis of individual practices with a meso-analysis of certain social domains and thus offers us various possibilities to contextualize this with macro questions about society at the very least (see Ryan 2005: 503). In so doing, it offers an important contribution to the discussion of the 'micro-meso-macro link in communications' (Quandt and Scheufele 2011: 9) that is open to various empirical and theoretical approaches. 12 To link the detailed analysis of specific figurations with macro questions of transformation, it is important to be aware of the fact that figurations of social domains are interrelated in various ways: via their overlapping actor constellations, different figurations can be linked with each other. In addition, figurations of collectivities and organizations can become 'supra-individual actors' (Schimank 2010: 327-342) that are part of the actor constellation of other figurations and thus build 'figurations of figurations' (Couldry and Hepp 2016: 71-78). One example here is constituted by figurations of various organizations acting together in a certain institutional field. Besides that, it is important to take into account that figurations do not simply co-exist side by side, but that they are arranged with each other in a meaningful way. For example, in the majority of Western societies, the family is given some special societal meaning because of recreation and bringing up children; organizations such as schools or adult education centres are constructed with certain responsibilities for educating people; journalism organizations deal with information and entertainment, while as companies they also have the role of generating income and jobs. One could continue with many other examples.
On this basis, it is clear that communicative figurations are hardly 'harmonious' phenomena. In contrast, we have to be aware that certain power relations, inequalities and conflicts characterize many figurations. Therefore, all the criteria which are used to describe social disparitiesclass, race, gender and others (norris 2001;Zillien 2009;Stegbauer 2012;Pollock 2013;Klaus 2015;Maier 2015)-matter when it comes to the analysis of figurations. We even go so far as to argue that a figurational analysis has specific capabilities for analyzing such disparities: the origin of the concept is rooted in analyzing the 'power balances' of actor constellations Scotson 1994 [1965]). 13 Describing communicative figurations with reference to their actor constellations, frames of relevance and communicative practices always imply that we have to be sensitive to all lines of inequalities and conflicts that are inherent in or characteristic to them. When analyzing communicative figurations, we can expect to be confronted with the entirety of social disparities concerning media use and appropriation that have been researched so far, and possibly also new ones too.  (2014). 5. While the assessment of a rapid pace of innovation corresponds with our everyday experience, we must be very careful not to over-emphasize this. Referring to social studies of technology, the challenge is to reflect what actually constitutes an innovation: a so-called 'key innovation' and 'improvement innovation'. Moreover, there are 'recursive innovations' and other complex patterns of innovation processes (see Dosi 1982;Rammert 2007: 28;Häußling 2014: 331-335). Hence, we have to be aware that 'pace of innovation' relates to experiencing an apparent acceleration of minor improvements that are constructed, among others for marketing reasons, as 'ground-breaking'. Examples for this are smartphones or tablets where the latest software only works on the most recent generations. 6. See for this discussion Couldry (2004); Raabe (2008); Postill (2010); Schmidt (2012); Genzel (2015); Pentzold (2015). 7. Compare for this, among others, Giddens (1984); Bourdieu (1992); Schatzki et al. (2001); Reckwitz (2002); Hörning and Reuter (2006); nicolini (2012). 8. However, methodologically we do not share the position that we could not gain access to practices and their meaning via interviews. Depending on the interview strategy, we can in an indirect way gain access to (media related) practices (of communication), for example by asking questions about specific habits and everyday experiences (Klein et al. in this volume). 9. There is a long discussion in media and communication research about the necessity of such a 'non-mediacentric' perspective on transforming communications that does not always consider media as the angle of change (see recently Couldry 2006;Morley 2009;Moores 2012;Krajina et al. 2014). We refer here to the argument that only research which