Social Media Data and Their Utopian Assumptions

The next speaker at "Compromised Data" is Ingrid Hoofd, whose interest is in how new technologies make certain types of representation possible or impossible. The neoliberalisation of universities, for example, leads to a quantification of research data which generates poor research. This is the violence of numbers: how do we assess the way new media technologies change the face of social sciences research, then?

Social media data mining methodology provides an allegory of the technological apparatuses that use it. This hinges on these technologies' propensity to speed up, and on the associated notion of change. There is a strong emphasis on objectivity, generating more true as well as more questionable coverage of the conditions of the real. Social science via datamining tools is implicated in a push towards an idealised data-driven utopia.

Ingrid cites The Guardian's investigation of the UK Riots in this context: datamining extracts the numbers around activity in social networks, and thereby quantifies the questions around risk management; the representation of the riots does not inform us about the conditions which led to their occurrence, but merely track the data associated with them. The data generate phantom collectives of individuals whose data appear similar, but which do not actually exist in the social. Any use of these data to understand the social follows an idealistic hope for a better future.

The facilitation of collections and communities has been a long-term focus of social policy, and social media data are increasingly used to address these issues; questions of humanity and sociality are increasingly addressed by social science which utilises such data - the hope is that new technologies will enable the social sciences to dig deeper. These new and revealing datasets, the hope is, will enable new groups and online alliances to intervene in society. But this assumes that more sociality is always better; the supposed existence of the social functions as the foundational assumption of the social sciences.

On the other hand, if new media engender a rapid fragmentation of society, if collectivities in general are becoming increasingly computerised through social media, then social media data provide us with an illusion of collectivity; the data image is distanced from the person and their responsibilities. Contemporary philosophy is replaced by the techno-sciences.

If we are implicated in a representational regime that is intertwined with business and industrial structures, The Guardian study of the riots offers a useful example, Ingrid says: it combines social media data in the riots with interviews with a range of participants and stakeholders, generating a larger map of interrelations between the data points. Researchers found correlations between age and gender, unemployment, moral decline, and rioting, for example, and highlighted the value of data in this research.

This dissimilates the disintegration and disappearance of the social in England. It generates a pretence of truth that is highly implicated in social risk management, producing accelerated flows of information while eradicating the context of a decline of democracy in contemporary England - it offers an alibi for technocracy. But there is a twist to this story: the criticism of such processes harks back to a utopia which never existed: scientific data was always compromised, and new approaches to the data only point out these issues in new detail.