Establishment versus Newcomers, Critical versus Administrative?: Sketching the structure of the Swedish field of media and communication studies

The purpose here is to the study the contemporary structure of the Swedish field of MCS. To do so, I draw on French, or Bourdieusian, prosopography and MCA (Broady, 2002; Rossier, 2019). While most Bourdieusian studies rely on self-reports, the prosopographic approach combines data gathered from public sources such as libraries and archives with self-reports (such as information on social trajectories of agents.). Bourdieu's (1984/1988) Homo Academicus and Sapiro's (2002) study of the French literary field are examples of the approach. The obvious advantage is the focus upon “objective” properties and not relying solely on self-reports. This study draws on the full population of agents in the field of Swedish MCS (N=254) as found in public registers: university websites, Diva, Libris, Google Scholar, and other archives. As such, the material used primarily consists of digital, so-called “found data” (Bruhn Jensen, 2010), which is why the present study is only a “half” prosopography. Ideally, the study would combine the prosopographic, archival data with a survey and interviews capturing agents’ social origins and various position-takings (Broady, 2002). In Homo Academicus, Bourdieu emphasised the difficulty for a social scientist in accessing such data, since a survey on family histories, salaries, and position-takings of one's colleagues could be interpreted as creating a “sectarian and inquisitional police file” (Bourdieu, 1984/1988: 39). The kind of science of science undertaken here is challenging also because it involves the attempt to “exoticize the domestic” (Bourdieu, 1984/1988: xi) by turning scientific instruments onto science itself and, in effect, onto oneself. In the spirit of transparency and reflexive sociology (Bourdieu & Wacquant, 1992), I have therefore indicated my own position in the field in the results section (Figure 2).

Data was collected on virtually all agents in the Swedish field of MCS in February 2020. The study includes all Swedish media and communication institutions with more than one person holding a PhD. The data includes staff at the following academic positions: PhD student, post-doctoral researcher/researcher, senior lecturer/assistant professor, associate professor/docent, and professor. Teaching (e.g., “amanuens” and “universitetsadjunkt”) and administrative staff were not included in the study.

Nine variables were used to sketch the structure of the field of Swedish MCS. The study set out to capture the positions of agents in the field and, in effect, to study the field's hierarchies. To do so, the study design relied in part on Demeter's (2018) application of Bourdieu's forms of capital (Bourdieu, 1986) to the field of MCS, and in part on Craig's (1999; 2015) and Waisbord's (2019) mapping of the communication theoretical terrain of the field and arguments on hyper-specialisation.

Seven out of the nine variables capture various field-specific capitals – both symbolic and material – that can be acquired or accumulated; these can be classified as “found data” (Bruhn Jensen, 2010). University websites were used to disclose 1) academic position (having or not having a PhD), and 2) university affiliation (all Swedish institutions conducting research and teaching in MCS with more than one staff member holding a PhD). The latter variable was divided into four categories; old universities, old/pioneering universities, new universities, and university colleges. While the PhD degree captures an academic/cultural capital, the institutional affiliation taps into symbolic capital, since the prestige of institutions – both their positions in the field of MCS (such as the field's pioneering institutions in Stockholm and Gothenburg) and in the broader educational field (the old universities in Lund and Uppsala) – may “trickle down” onto their employees (Bourdieu, 1984/1988). Yearly reports from Riksbankens Jubileumsfond [The Bank of Sweden Tercentenary Foundation] and an archive provided by The Swedish Research Council – two of the biggest and most prestigious research foundations in Sweden – were used to disclose accepted research project proposals during the last 18 years (variables 3 and 4). By virtue of granting their recipients both salaries and research time, and internal recognition, these variables constitute both field-specific economic capital and symbolic capital (Demeter, 2018). The public databases Diva and Libris, Google Scholar, and university websites were used in complimentary fashion in order to count the number of published journal articles (variable 5) and monographs (not including research reports) (variable 6). Google Scholar was used to access agent's H-indexes (variable 7). These three variables are additional parts of a field-specific symbolic capital (Demeter, 2018).

In order to also account for agents’ positions in the communication theoretical terrain (which helped shed light on the onto-epistemological divisions and the dynamics of the autonomy/heteronomy of various sections of the field), two “made” variables (cf. Bruhn Jensen, 2010) were created. Samples of research output and self-presentations on university websites (when available) were analysed. In the first variable – “Sub-field” – each agent was categorised in one of the following sub-fields: 1) journalism studies, 2) cultural studies, 3) media history, 4) strategic communication/organisational communication, 5) political communication, and 6) a blend of the above or an interdisciplinary approach (which captured those transcending the borders of two or more sub-fields). While this constitutes a rough categorisation, it will suffice to understand the broader divisions in the relatively small field of Swedish MCS. Using a more detailed categorisation, for instance the over twenty sub-divisions of the International Communication Association would 1) not be statistically feasible, since the twenty modalities would skew the contribution to the total inertia of the model, and 2) not reflect the actual terrain of the Swedish field of MCS (where, for instance, health communication and visual communication are marginalised).

For the second variable – “Communication theoretical approach” – each agents’ main communication theoretical approach was classified according to Craig's (1999; 2015) seven traditions in communication theory. For instance, agents with a sociopsychological communication theoretical approach frame communication as a problem of influence (which is usually to be measured with quantitative methods), whereas a semiotic approach focuses upon meaning and signs (almost exclusively by studying media texts with qualitative methods) (Craig, 1999; 2015). All but one tradition (the cybernetic tradition, which in Sweden is confined to disciplines such as Information Systems and Computer Science) were represented as a main outlook. Some agents can (and do), of course, criss-cross between various theoretical views on communication. This study accounts, however, only for the main tradition that an agent relies upon. Taken together, the variables measuring sub-field and communication theoretical approach allow – in part because they are proxies for research methods used and objects of study, and autonomy/heteronomy – closing in on whether historical disputes (between what has been described as administrative and critical research) still divide the field.

“Sex” (male/female) and the original “institutional affiliation” and “academic position” variables were projected as supplementary, passive points in the space. Since MCA is sensitive to outliers (categories holding less than 5% of the observations) the institutions at Halmstad University, Linköping University, Linnaeus University, and Gävle University – all with relatively small media and communication departments – were combined into one category. Table A1 (Appendix) presents a full disclosure of the variables used and their data sources.

MCA is an exploratory statistical technique that extracts structures from a number of qualitative (nominal and ordinal) variables. These structures, or dimensions, are visualised in a two-dimensional space where both individual agents and the categories of the variables are given coordinates depending on their relation to the overarching dimensions. Distances between individuals or variable categories indicate social distances (Hjellbrekke, 2018). Bourdieu favoured MCA because it corresponded with the open-ended, relational, multivariate thinking of field theory (Bourdieu & Wacquant, 1992; Duval, 2018; Lebaron, 2009). In order to sketch the structure of the Swedish field of MCS, the nine variables presented above were included in an MCA.

MCA creates imperfect representations of fields (Duval, 2018). The statistical representation of the field of MCS presented here should be viewed for what it is – an impressionistic snapshot. This study is concerned with the field as it was crystallised in a set of digital properties in February 2020. The study is thus confined to the struggles leading up to the early spring of 2020. It goes without saying that fields change (oftentimes slowly, but in small fields sometimes more quickly – for instance if an institution is able to recruit many senior staff members from other institutions, or if the only established senior colleague at a small department leaves).

Additional limitations include the fact that not all field-specific resources were accounted for (e.g., Demeter, 2018), particularly certain economic (salaries) and social capitals (Who knows and works with who? Who has been chair of which conference division and member of which editorial board? etc.). Book chapters and conference presentations, as well as international visits, were left out – focus was put on monographs and journal articles. No account was taken in regards to which publisher agents have published with. The social origins and trajectories of the agents were not included for reasons stated above. Lastly, although most variables were recorded “objective properties”, human fallacy in terms of coding and errors in databases or websites may have impacted the results.

Turning to the first research question – What are the main divisions in the Swedish field of MCS? – Figure 1 shows two of the extracted structures, or dimensions, that the MCA produced from the nine active variables presented in the previous section. The second dimension resembled the first (the factorial plane of dimension one and two retains the “horseshoe effect”, which indicates uni-dimensionality [Hjellbrekke, 2018: 96]). As such, focus is here put on dimensions one and three, which together explain 72 per cent of the variation among the variables (see Appendix for details on eigenvalues and contributions of variables and modalities/categories to the dimensions).

Figure 1

The vertical dimension explains 62 per cent of the variation among the variables and is, accordingly, by far the most important one. The VR-grant and the number of published journal articles and monographs are the most important active variables in this dimension (Table A4, Appendix). The dimension represents a division where consecrated agents (those holding a PhD) are pitted against PhD students; where beneficiaries of research grants stand against those without external research funding; and where the publication elite stand against those who have published less or nothing. The vertical dimension describes, in other words, a hierarchy separating an incumbency (top) from challengers and newcomers (bottom), and the volumes of field-specific capital (in academic/cultural, symbolic, and economic forms). Such a separation is found in most fields (Fligstein & McAdam, 2012). Doing research, writing monographs and (successful) research applications, publishing in journals, and writing promotional applications take both effort and time. As such, this dimension emphasises the fact that fields tend to be time-economies (Bourdieu, 1984/1988). Those in the upper echelons of the field possess the social resources and scientific authority required to “influence the way in which [other] agents allocate their time” (Bourdieu, 1984/1988: 98). Compared to assistant professors/senior lecturers – the incumbency (primarily professors with research grants and many publications) tend to have more time to engage in the activities required to reach their positions in the first place, and they manage other agents’ time (not least by directing research at their institutions and by supervising PhD students).

The horizontal dimension explains 10 per cent of the variation among the variables and can be approached as a contemporary variant of what has been described as the historical “tug-of-war” between “Nowakian behavioural research” and “critical left-wing opinion” (Hyvönen et al., 2018: 93; Hyvönen, 2015), or (more recently) between the field's “normal science” and “culturally oriented” media research (Andersson & Fornäs, 2010; Bolin, 2010; cf. Kuhn, 1962). This dimension also taps into the historical divisions between critical and administrative research (Lazarsfeld, 1941).

Two variables are particularly important to this dimension: the agents’ communication theoretical traditions and their sub-fields (See table A4, Appendix). To the right in the map we find agents in a contemporary variant of administrative research – that rely on American sociopsychological traditions (cf. Craig, 1999) and that tend to focus on media effects, influence, and public opinion (often with statistical methods). These agents tend to be active within the sub-fields of strategic/organisational communication and political communication. Projecting all the institutions as passive points in the space, we find that the Strategic Communication Department at Campus Helsingborg (Lund University) and the MCS institutions at Mid Sweden University and the University of Gothenburg are drawn to this pole. Towards the other pole we find a “European” MCS consisting of agents conducting (mostly qualitative) media research or cultural studies from critical and semiotic views on communication. MCS at Södertörn University (not the separate journalism department at the same institution) and Jönköping University stand out in this respect. Karlstad University and Örebro University are the institutions most balanced on both dimensions.

The answer to the question of whether agents in different poles along the second, horizontal dimension continue to live in “entire[ly] different worlds” (Hyvönen et al., 2018: 91) requires some nuance. By virtue of agents “playing the same game” and competing for the same symbolic and material resources, they do live in the same world – but in different parts of it. The distance between factions of the field can be discussed in terms of non-communication between different institutions and that agents are active in different sub-fields, publish in sub-field–specific journals, attend different conferences, and address media and communication problems in different, sometimes conflicting, ways. These oppositions are illustrated by Craig's ideal typical position-takings of researchers in various communication theoretical traditions: for a critical scholar, sociopsychological approaches “reflects ideologies of individualism [and] instrumentalism”, whereas critical theory, in the eyes of the sociopsychological researcher, “confuses facts and values” (Craig, 1999: 134). A semiotician argues that a sociopsychological researcher's so-called media effects are in fact “internal properties of sign systems”, whereas a sociopsychological researcher holds that semiotics “fails to explain the factors that influence the production [and] interpretation of messages” (Craig, 1999: 134).

To some extent, the horizontal dimension also describes a tension between heteronomous and autonomous poles where research is either supposed to be valuable to stakeholders outside the field, or if recognition is mainly sought from peers, inside the (sub-)field (cf. Bourdieu, 1984/1988). One may argue that it is agents in the right side of the space that are richest in external prestige – researchers in strategic and organisational communication are often dealing with “practical” problems that matter to companies, organisations, and the media industry. In some cases, there is an epistemological overlap in that these agents adopt the problems posed by the media industry. One may think of a political communication scholar studying the effects of news consumption on political knowledge who is regularly invited as an expert in the media. Or, the strategic communication researcher who is a consultant on the side.

Taken together, elements of the historical opposition between administrative and critical media research (Lazarsfeld, 1941) of the global field of MCS is reproduced in the Swedish context. I note, however, that this division accounts for 10 per cent of the variation in the space, which means that differences between agents are much better explained by their placement along the incumbency-challenger axis. It should also be noted that while certain strands of administrative research may hold more external prestige because of epistemological overlaps with stakeholders outside the field, they do not necessarily have less internal prestige than agents tending towards critical strands. Historically, American-inspired mass communication research constituted the field's incumbency (Hyvönen et al., 2018). This balance may, however, have tilted.

A key internal debate on the field has revolved around the status of culturally oriented media research – whether it is marginalised (Andersson & Fornäs, 2010) or not (Bolin, 2010). In fact, the culturally oriented and critical faction of the field is more represented among the Swedish Research Council's and Riksbankens Jubilieumsfond's beneficiaries, and over half of the field's professors rely on either critical, semiotic, or sociocultural perspectives. The picture painted here thus aligns with Bolin's argument that “culturally oriented media research is far from marginalized” (2010: 6). Westerståhl's A- and B-leagues of media research is no longer an adequate classification. Södertörn University, for instance – which Bolin (2010) himself used as an example – is found, along with the critical view on media and communication, towards the top-left of the space, where parts the field's incumbency is located.

This observation also makes for an argument that the field of Swedish MCS is not homologous to the wider educational field in Sweden. In the latter, old universities wield the most symbolic capital and recruit top-performing students from socially privileged backgrounds (Börjesson, 2016; Börjesson & Broady, 2016). In the Swedish field of MCS, however, the universities at Lund and Uppsala do not hold the most field-specific capital. This has generated the situation where the agents with some of the most scientific authority and internal recognition in the field of MCS teach students with low or medium status (in terms of admittance thresholds), and where smaller and less-established MCS departments, such as the institution of Informatics and Media at Uppsala University, teach the most prestigious students (the undergraduate programme in MCS at Uppsala University is the most popular MCS education in Sweden).

This is not an implicit argument about the quality of the educational programmes at the different institutions. It should also be noted that the department at Uppsala University was severely understaffed at the time of data collection. During 2020, the department recruited two senior researchers and one full professor. Had data collection been conducted later during 2020, the position of this institution in the field is likely to have changed.

I turn now to the second research question, regarding the gendered character of the field (cf. Chakravartty et al., 2018): How are gender differences articulated in terms of the distribution of social resources? Projecting sex as a supplementary variable onto the space representing the Swedish field of MCS suggests, somewhat surprisingly, that the hierarchies explored above do not connect well to differences between male and female agents (Figure 1). Exploring this further, Figure 2 reveals the positions of all agents in the field – that is, their positions in relation to the two axes of differentiation explored above. The colouring of Figure 2 shows the sex of the agents. Although males on average have slightly higher positions, the gender categories are found at less than 0.2 deviations from the centre of the space, which indicates a very weak association (Hjellbrekke, 2018). This small difference is explained in that while there are slightly more women (n=133; 52%) than men (n=121; 48%) in the field, males are slightly overrepresented amongst the professors (55%). Among the beneficiaries of Riksbankes Jubileumsfond, the distribution between males and females is 50/50 but males are over-represented amongst beneficiaries of the Swedish Research Council (62% are male). The fact that the hierarchies explored above only marginally affect males and females warrants a closer study of the field's power elite, which brings us to the final research question: What characterises the field's power elite?

Figure 2

The power elite is constituted by the incumbent agents who place highest out of all agents in terms of overall volume of capital on either side of the space (the onto-epistemological dimension). Subsequently, the elites can be divided into three groups: The cultural studies-journalism studies elite consists of Göran Bolin, Monika Djerf-Pierre, and Christian Christensen; André Jansson and Mats Ekström constitute the balanced elite; and finally, the political communication elite is made up of Jesper Strömbäck, Bengt Johansson, and Lars Nord.

The exact positions of the elite incumbents along the vertical axis should be approached with some caution. Since MCA uses ordinal and nominal data, it underestimates some fine-grained differences in terms of, for instance, number of publications. For example, had the analysis been able to account for the exact number of publications (rather than categories holding various ranges of numbers of publications), Jesper Strömbäck would be positioned higher in the space.

Four observations are worth highlighting in regards to the field's elite incumbents. First, all of them are somehow connected to the field's pioneering institutions – JMK (at Stockholm University) or JMG (at the University of Gothenburg). Djerf-Pierre, Strömbäck, Ekström, Johansson, and Christensen are currently affiliated with either of these two institutions. Jansson, Bolin, and Nord are not – but they obtained their PhDs from them. Second, the two poles that make up the onto-epistemological divisions in the field are less diverse amongst the power elite compared to the field as a whole. The incumbents are oriented towards cultural studies (Bolin) and journalism studies (Djerf-Pierre, Christensen) and political communication (Strömbäck, Johansson, Nord) – and two are “balanced” in terms of the onto-epistemological division (Jansson, Ekström).

Third, none of the “old universities” (Lund, Uppsala) are represented in the power elite, which is a case in point regarding the absence of a homology between the wider educational field and the field of MCS discussed above. Fourth, and lastly, seven out of eight are male which suggests that while the wider field is structured by gender inequalities only to a limited extent (in terms of distribution of social resources), the power elite remains a male domain, like in many other fields (cf. Göransson, 2007).

While it is clear that male agents dominate the very top of the space (and potentially the field as such), it should also be noted that the pattern of male overrepresentation holds also if we expand the view and look, for instance, at the agents above 0.1 deviations on axis 1. Here, 14 out of 19 agents are male.