By Lucy Phillips, on 21 March 2014
Richard Rogers, Professor of New Media and Chair of Media Studies at the University of Amsterdam, spoke at UCL on 13th of March 2014.
By Lucy Phillips
“If you’re interested in chess, your hero is Batman,” states Rogers, after a quick analysis of Facebook group networks.
On 13 March, Professor Richard Rogers delivered a talk on what he terms Digital Methods, the title of his 2013 book. Rogers is the head of Media Studies at the University of Amsterdam, director of Govcom.org, a group responsible for several info-political tools, and director of the Digital Methods Initiative. His research focuses on web epistemology, the idea that the web is a distinct knowledge platform with its own native methods. His talk explored the use of such innately digital methods in political science research.
How we perceive and interact with the World Wide Web has changed greatly over time. Whilst in the mid-1990s, the web was seen as “cyberspace,” a virtual realm with its own identity and politics, in the early noughties, as Rogers explains, that changed. Social scientists began to unpack the idea of the web as a different realm, recognizing it instead as being part and parcel of society. Around 2007, a point which David Lazer has coined the “computational” turn in social science, there was yet another gear shift: the internet was now seen as a means through which to study society. It became, in effect, a dataset.
Rogers makes the distinction between “digitized methods” – methods that are imported to the internet – and “digital methods,” those that are native. Whilst many of us are perhaps familiar with digitized methods such as online surveys or online investigative reporting, essentially the analogues of existing offline research processes, the use of digital methods such as the analysis of Google searches, web histories, or social networks, remains relatively alien to the traditional researcher.
In the realm of political science, digital methods can be used to reveal the concerns of an electorate, trace the spread of arguments and rumours, or uncover political positions. For example, in 2005, Rogers delivered a paper on the community structure of the American blogosphere.
The outcome was a stark visual of the degree of web separation between bloggers on the political left and those on the right, (inset left). The void in the middle speaks volumes as to the level of offline political polarization in the US, and indeed, may have anticipated it, Rogers argued. In another quirk of digital methods, if the image on the left, or indeed any image online, were a photograph, a simple click would in almost all cases reveal information about the model of the camera. In 2013, Rogers used such embedded photo information to estimate the price of various cameras used to document the 2013 Gezi Park protests in Turkey. This enabled Rogers to distinguish (somewhat crudely) between professional and civilian photography, on the basis that most amateur photographers would not be using top-shelf cameras, and hence make inferences about potential bias in the photographs.
Rogers went on to give several other examples of digital methods. His research has helped map the effectiveness of internet censorship in both China and Iran through the analysis of hyperlinks. Rogers also argued that inferences about the politics of memory can be made based on differences in the reporting of the 1994 Srebrenica Massacre on various Wikipedia language pages, as the details of the event and even whether it is termed a ‘massacre’ at all vary greatly between languages. Beyond the scouring of webpages, the rise of social media lends huge potential to the proliferation of digital methods. Retweeted content on Twitter can be used to generate a summarized story of the evolution of particular political events, whilst the network created by Facebook can be unpacked to reveal trends of preferences. For instance, whilst President Barack Obama’s Facebook followers – those who have liked his page – while away their time watching The Office, Senator John McCain’s have a soft spot for Desperate Housewives.
However, the ease with which the internet propagates new and ’invisible’ networks has a more sinister undertone. Part of Rogers’ research has looked into the web-based counter-jihadist group networks. Such work involved joining extremist groups on Facebook and looking up the names of the most connected people on this “new right.” Whilst the publishing of those individuals’ identities might help deter hate crime by making them publicly known, their circulation raises important ethical questions – do digital researchers need consent to publish public, or semi-public data?
The research potential created by digital methods is exciting. The web is vast, and it is growing. However, if questions over the ethics of covert ethnography trouble the traditional political scientist, questions over access to and the use of online information raise increasingly complex issues for the digital researcher. The use of web data can perhaps be alikened to debates surrounding the creation of a national DNA database – to what extent do the potential benefits, security or otherwise, of the collection and analysis of data outweigh the infringement on privacy? Where do we draw the line? When asked about the issue of ethics, Rogers admitted this question continues to trouble him. It is likely that such questions will play an important role in shaping the future of digital methods.
For more information, or to enroll in Professor Rogers’ 2014 ‘Digital Methods’ Summer School: http://www.digitalmethods.net
Software Resources built by Rogers and his team:
Googlescraper (the Lippmannian device – repurposes Google into a bias-detection machine)
elfriendo (analyses interest compatibility between individuals or groups)
netvizz (allows researchers to export data from Facebook)