Crunching Data Like an Anthropologist: Rethinking the Divide Between Data Science and Interpretive Social Science.

Talk and workshop with David Moats, 12th November 2019.

There have been many calls for better collaboration between fields like anthropology and sociology and computer science and data science (Neff et al., 2017), though, encounters between these fields remain in many ways fraught. This is despite the fact that there are many proposals for hybrid practices (Jensen 2013, Marres 2017, Venturini and Latour 2010, Blok and Pedersen, 2016) or even a shared lineage between them (Seaver 2015, Munk and Jensen, 2015). Yet, problems and misunderstandings persist due to the assumed roles and inherited divisions of labour between social scientists and more technical researchers and everyday boundary work which performs divisions between “qualitative” and “quantitative,” or “positivist” and “interpretivist”.
What happens if we interrogate these tensions, not as historical fictions or disciplinary givens but as a practical, situated achievement? This talk discusses a series of workshops organised at Linköping University in which anthropologists and other self-identified ‘qualitative’ researchers were encouraged to engage with various digital tools (network graphs, scrapers, text analysis, and Digital Methods etc.). These interactions offer us a glimpse of how so called “interpretivist” researchers might approach data analysis in different ways, if given the chance. But they also raise the possibility that some of the barriers to collaboration might come from these researcher’s own taken-for-granted assumptions about research.

Workshop: ‚Sound‘ Ethics

This workshop will explore the ethics and politics of representing data through an unlikely vehicle: the representation of data as sound (sonification). Data visualizations (like network graphs and sanky diagrams) are ubiquitous in certain areas of social science and studies of online media, but sonification is becoming increasingly common in the natural sciences and computing (Supper 2015). It is often argued that auditory representations of data have different properties than visual ones: that they are more immersive and trigger the emotions.
However, these sorts of claims often draw on purely cognitive explanations or taken for granted ideas of what our ears/brains are capable of. They also tend to reproduce a “hierarchy of the senses” in which vision is rational and other senses are more subjective.
What if we think of these differences as at least partially socially and culturally produced?
What can our fumbling attempts to represent data as sound tell us about how we represent data visually?