Post by Farida Vis
As I haven’t posted anything on here for ages, I thought I’d share an abstract that has recently been accepted for the Association of Internet Researchers conference in Salford (#IR13), in the UK (in fact around the corner from where I live). I have been thinking about the topic of what data visualisations might ‘want’ for a while and as the deadline approached, I decided that writing an abstract was a good way to commit to thinking through the idea more critically. So I was delighted when it was accepted and over the next few months I will try to share more developed thoughts with you. One data visualisation I am keen to reflect on is one I have been talking about a lot recently, and was involved in producing with academic colleagues, in collaboration with The Guardian Interactive team: on the ways in which rumours and misinformation spread on Twitter during the UK riots of 2011. Anyway, here is the abstract, comments very welcome of course!
Data visualisations are increasingly becoming ubiquitous within visual culture, specifically online. This year for example sees the first data journalism award, co-sponsored by Google, which includes a category for data visualisation and storytelling. Although this category keeps the visualisation and the telling of stories neatly together, more often than not data visualisations float around the internet rather rudderless and alone. Onto them are projected assumptions that they can ‘speak’ for themselves. We wrongfully make similar assumptions concerning other forms of visual material, most notably photographs.
Taking W.J.T. Mitchell’s ‘What Do Pictures Want’ (2006) thesis as a starting point this paper aims to draw links between the agency Mitchell ascribes to visual material in their subaltern state, specifically pictures and uses this same frame to position the growing visibility of data visualisations in their multiple forms in order to question what they might ‘want’. In asking what pictures might want Mitchell rhetorically questions our relationship with images and fuels and encourages an imaginative engagement with pictures that we do not normally have. He points to the ‘lack’ of images, what we project on to them in acts that are often ventriloquist in nature, and encourages a much clearer understanding of what they can really do and what kind of words they really need. This lack needs to be paid attention to through words. Words that make images speak and give them a place.
In recent years data visualisations like the social graph have become firmly fixed on the visual landscape of social media and beyond. Through such visualisations, drawn up algorithmically through cybermetric devices connections between users are suggested and made visible in ways that can be misleading and on closer inspection may not be what they seem. Work on YouTube networks for example has highlighted that the cybermetric tool can suggest a closeness that on closer, qualitative inspection of the users, is simply not present (van Zoonen et al, 2011). Recently, cultural sociologist Tricia Wang (2012) has made similar points about the misleading nature of such algorithmic visualisations.
Outside such social graphs, in areas like journalism and beyond, data visualisations have flourished and are producing a new aesthetic of (data) journalism. The work of David McCandless, specifically his book and website Information Is Beautiful and his work with The Guardian newspaper in the UK, has taken such data visualisations into an aesthetic realm worth exploring in more detail. Stripped from their raw data bearings these visualisations are aesthetically appealing, but may not be the best visual narrative (or storyteller) in their own right, obfuscating a clear position or indeed suggesting one misleadingly. As Mitchell would suggest, they desire words, a narrative, an interpretation, or as Butler would have it, hard reading.
On the opposite end of the new aesthetic of data visualisation, sit the ‘bad’ graphics, like Wordle, the backlash against which are growing according to Guardian Data Blog Editor, Simon Rogers (2011). ‘World Clouds’, or as senior software architect for The New York Times, Jacob Harris, has deemed them ‘mullets of the internet’ need to apparently go back where they came from. But with the rise of improved software and tools (such Gephi, visual.li and others) producing beautiful visualisations is increasingly possible for those who do not have a big graphic design team at their disposal.
This paper is thus interested in drawing on the process, theory and practice of photography, and in doing so exploring the producers of data visualisations, the tools that create them as well as the images themselves and their ascribed agency. This paper highlights the current moment of increasing ubiquity of such data visualisations and encourages a much clearer understanding of what they can really do and what kind of words they really need. Moreover, it interrogates how we might think through such developments critically by asking a slightly different question to Mitchell’s, namely: what do data visualisations want?
Harris, J. (2011) Word clouds are considered harmful, Nieman Journalism Lab, 13 October 2011. http://www.niemanlab.org/2011/10/word-clouds-considered-harmful/ (accessed 1/3/2012)
McCandless, D. (2010) Information Is Beautiful, London: Collins.
Mitchell, W.J.T (2006) What Do Pictures Want: The Lives and Loves of Images, Chicago: University of Chicago Press.
Rogers, S. (2011) ‘In Defence of Bad Infographics’, The Guardian Data Blog, 17 October 2011. http://www.guardian.co.uk/news/datablog/2011/oct/17/data-visualisation-visualization (accessed 1/3/2012)
Van Zoonen, L., Vis, F., and Mihelj, S., (2011) ‘YouTube interactions between agonism, antagonism and dialogue: Video responses to the anti-Islam film Fitna’, New Media & Society, 13(8): 1284-1300. Abstract here.
Wang, T. (2012) ‘Dancing with Handcuffs: The Geography of Trust in Social Networks’, Lift 12 conference, Geneva, Switzerland, 22-24 February 2012. Video of talk here.