CREATe researchers Kris Erickson and Martin Kretschmer traveled to Berlin this past week to present a paper with co-author Fabian Homberg, ‘Is Originality Overrated? Measuring the success of original and recombinatory works on Kickstarter’ [Download slides]. We were joined in Berlin by CCPR researcher Inge Sorensen, who presented her paper on the (d)evolution of British public service media’s professional standards and codes of conduct across social networks. This was our first outing at AoIR, and we found it to offer a stimulating set of topics and approaches with considerable relevance to research in CREATe. Below in this blog post, I highlight some of the most interesting presentations we attended at the conference and explore potential implications for future research.
Platforms and algorithms
The two stars of the conference were undoubtedly platform intermediaries and algorithmic governance. These topics were also frequently paired together. For example consider YouTube, which offers a platform for user-generated video while also sorting, recommending and policing content using pattern-matching algorithms and software. One set of problems raised by these practices is political: how should society regulate platforms in order to achieve goals like inclusive citizenship and economic growth? This is a particular challenge for Europe, where the perception is that platforms come from abroad (USA) without being subject to national regulation or democratic oversight. A provocative panel discussion by Tarleton Gillespie, Mike Ananny, Karine Nahon, Angèle Christin, Balazs Bodo, and Solon Barocas considered the question, ‘what is a platform, and why has the category achieved stability as an object for policy making?’. Tarleton Gillespie made the observation that what we take to be monolithic and effective social actors (e.g. Facebook) may in fact be very flimsy constructs made up of a few people and some code.
The widespread use of algorithms by platform intermediaries also raises important issues for research method. When we collect data from online services, we may be introducing unknown biases as a result of hidden recommendation or ordering of content based on geographic location, browsing history or other factors. Direct crawling of content via Application Programming Interface (API) may in certain cases be preferable; however this approach may fail to capture a realistic human browsing experience. Certainly, more care is required when making claims based on behavioural data obtained through online intermediaries.
Trust and legitimacyThese remain perennial issues of concern for Internet communication researchers. As we highlighted in our Introduction on the Sharing Economy, many new services are dependent on trust in strangers, constituting new forms of mutuality. Rating mechanisms are one common but imperfect means for online communities to self-organise, but there are other ways that trust and legitimacy can be established. A presentation by Robert William Gehl explored convergent forms of legitimacy in Dark Web search engines. The Dark Web is commonly associated with illegitimate behaviour, but Gehl demonstrated how nascent search engines achieved three forms of legitimacy: one form focused on sovereign state power, a second form based on professional propriety and a third form derived from authenticity and policed via informal social norms. Another presentation by Nathalie Marechal explored the geopolitical legitimacy of the Tor project, given its complicated links to American Government funding to achieve foreign policy goals.
Overall the conference provoked cross-disciplinary thinking about the role of copyright and intellectual property in the wider Internet governance field. Working in IP, it is easy to focus narrowly on specific rules and exceptions, without necessarily considering how these are embedded within social practices which occur in different contexts and through intermediaries with divergent aims and objectives. Intellectual property remains an important regulatory tool in the digital economy, but digital networks are perhaps more contested and more fluid than we can capture from a purely legal perspective.