I have a cross-disciplinary approach, with a strong interest in the ways in which everyday life, individual human beings, organisations, and structures are interrelated with the contemporary, ever more strongly consolidated media ecology. Which actors and phenomena benefit and are disadvantaged? How do digital media structures interrelate with the rest of society?

My main field of research used to be unregulated file sharing on the internet, but as of more lately I have shifted focus towards the newer developments such as platforms, apps, and mobile hardware.

Current research and teaching interests, more broadly:

● System-theoretical, legal, and political-economic perspectives on digital media – in historical context

● History of ideas, more generally (with particular interest in intellectual traditions and thought collectives, and the intersection of technological imaginaries and neoliberalism)

● Sound media and musicking (including, specifically, their systemic conditions)

My teaching and research therefore concern the ways in which increasing digitalisation and mediation interact with everyday life and societal structures. I take an interest in critically studying platforms as structural and mediating institutions, not least in relation to civil society and civic interests, epistemology (the conditions for how knowledge is established), and the data-driven media economy. I am also interested in how the material and the semiotic interact – especially in relation to theories of agency, relationality, decolonisation, and ecological collapse.

Current research

I am currently affiliated to an ongoing research project, led by Stefan Dahlberg (Mid Sweden University), Linguistic Explorations of Society External link, opens in new window. (LES), funded by the Swedish Research Council, where my role is to map the supply side for text data (editorial and user generated) on the Web, in a global context. There is now a complete research report, in which I summarize a number of challenges: The hitchhiker's guide to web-mediated text – Method handbook for quantification of online linguistic data in a country-specific context. It's published through the University of Gothenburg, as LES is originally their project and, also, some data from the Quality of Government (QOG) institute has been used in the report. The PDF itself can be found here External link, opens in new window., as part of the QOG Institute's working papers External link, opens in new window.. I've learned a lot from working together with leading scientists in natural language processing (NLP) and natural language understanding (NLU), like Magnus Sahlgren (AI Sweden External link, opens in new window.).

In machine learning and quantitative social research, large amounts of text as raw data are sometimes required. The report deals with the supply side of such textual data retrieved from the Web, and whether it can be considered representative of larger populations. Such text data is usually either editorially created (news pages on the Web, for example) or user-created (blogs, forums, social media). I make a brief overview of the global market in terms of providers of such text data.

The report is thus a review of the conditions that apply when web-mediated text is used as a data source – in commercial analytics, but also in more serious research. I explore the conditions for representativeness and validity, and give a snapshot of the current state of the supply side (disregarding, however, text corpora offered by universities and national libraries). It's a complex landscape but still rather easily understandable once you get into it.

One of the many things I show is that while editorially created text is relatively unproblematic to obtain, the supply of user-created text data is now irreparably skewed due to platformisation. Since large digital media platforms more or less completely restrict open access, there are nowadays very few places to find user-created text in large quantities, compared to the “open web” of the past. (That is, unless you pay large amounts of money or simply access/scrape platforms without consent.) While the situation is different in different countries, generally, the only large platforms from which one used to be able to retrieve user-generated text “in bulk” were Reddit and Twitter. Now the latter's bulk data is almost as unattainable as that of the platforms owned by Meta (Facebook, Instagram, WhatsApp). Microsoft- owned LinkedIn, and also Snapchat and TikTok are also, in practice, completely closed in terms of such access. Conventional forums on the open Web remain accessible for data harvesting, but these have in many countries become quite irrelevant, since fewer and fewer people use such forums these days.

In other words, good access to user-created text data for academic research purposes is rather poor, even among the commercial providers that I go through. Therefore, it is mainly editorial text data that one would be guaranteed to have access to, which is why we also examine, in the latter half of the report, the popularity of major text-based news sites are in various countries of choice (20 countries), using SimilarWeb traffic data as a proxy for popularity.

The report therefore also serves as a, hopefully, quite useful snapshot of the supply side of text-based news media in several countries, across the globe (including, f.ex., Hungary, Hong Kong, and Russia).

Older research

Together with some researchers within the Competitive Intelligence companies Retriever and M-Brain, I carried out a research project in 2015, exploring the conditions for user-driven communication in social media. We examined the specific factors that play a role in the sharing of political tweets, as well as the type of material from social media that was seized upon by traditional news media – and, not least, the image of social media in the mass media that was created.

That project was part-financed by the Bank of Sweden's Jubilee Fund and the Swedish Internet Fund, and was also part of my larger, three-year Postdoc which was a so-called Flexit position at the large advertising agency Forsman & Bodenfors.

As part of the same Flexit employment, I carried out an interview-based project on the public service mission in a time of increasingly personalized broadcasting and market-ideological pressures to adapt to platform logics and network logics.

In addition to this, I have recently worked with statistical analysis of media content, as well as working on developing methodologies regarding social media data.

In more history-oriented, political theory I could highlight two articles from my file-sharing and piracy studies that attest to my theoretical inclinations towards moral economies in the online/digital context:

(1) The liability of politicalness External link, opens in new window.. Not so many people might be aware that stalwart Swedish internet startups like Spotify and Skype also had their origins in unregulated, unlicensed file-sharing. By contrasting these two ventures with the much more disparaged Pirate Bay (which was, however, extremely popular during its time), we compared different examples of early Swedish internet entrepreneurship that however shared their origins in the same illegitimate or non-legitimised information space. By doing so, we managed to outline the radically divergent strategies with which the entrepreneurs sought to legitimise their ventures and underlying technologies.

(2) Honorability and the Pirate Ethic. External link, opens in new window. Nonlegitimate ventures in internet entrepreneurship did not exclusively pertain to ethically problematic practices and contents. Also more virtuous, honorable ethics could be identified in this information space, unfolding, some might say, by recourse to certain forms of charisma and rhetoric within commons-based meritocracies. I have explored, in the genre of moral economics (E.P. Thompson, Marcel Mauss, John Ruskin), how online pirates act as a kind of illegal entrepreneurs who gain some respect/honor among their peers – while sacrificing their position in official society by being excluded from it. At the same time, there were also significant attempts to impose order/hierarchy in these often otherwise anarchic spaces, and there were also examples of media archives being built in this illegal sphere, of extraordinary quality and integrity.

I'm no longer using Twitter as an outreach platform. Instead you can find me on Mastodon.