This year Goldsmiths is hosting the International Visual Sociology Association (IVSA) conference and in the autumn my colleague Nina Wakeford and myself are starting the first MA Visual Sociology (you can still apply, if interested!). A good time then for visual sociology, one would think. Why then does this blog post speak of visual sociology in the past tense? Why not ask: What is visual sociology? (to follow up Noortje Marres’ “what is digital sociology” post?).
Short answer: Because we need to get rid of the “visual” as a denominator of the subdiscipline. By starting an MA in visual sociology I am guilty in perpetuating this unfortunate situation, but, I hope, we may also produce a new generation of students, who can then leave the visual behind.
As in the case of women’s studies, the denominator does not simply announce a subfield that stands next to any other subfield. There is not visual sociology and textual sociology. There is not men’s studies and women’s studies. Women’s studies emerged because disciplines such as history and sociology were men’s studies without saying so in the title. For the same reason, we have sociology, and a special subfield called visual sociology, because for whatever reason, sociology is assumed to be a purely textual discipline (the same situation applies to visual anthropology). “Visual” is considered to be strange, not really sociology, not really scientific, or it is simply forgotten.
In many other disciplines the same situation does not apply. There is no visual astronomy or biology or chemistry. A non-visual astronomy simply does not exist. What would astronomers do without producing visual traces of other stars, planets and galaxies?
But the reason we now have a subdiscipline called visual sociology has repercussions for both sociology proper and the subdiscipline itself. As I explain in more detail in a forthcoming article in the British Journal of Sociology, the underlying problem is that sociology believes in an unequal media-determinism. Let me explain.
Ideally, as is the case in biology or astronomy, sociologists would use whatever media they want – such as writing, audio taping, drawing, or photography – for whatever research question they need to answer. There would be discussions about whether some media are suited to answer certain questions, but these discussions would center on the particular questions and projects. We would ask: what is the best method to research what cooks do? Making drawings of their daily routines? Interviewing them? We would ask: What is the best method to analyse recent changes of capitalist production? Photographing work situations in factories? Creating maps of circulation of goods based on trade statistics?
Whether a certain research project is good or bad, whether it is considered esoteric or exoteric, would not be decided by its use of certain media. In such a world there would be no need for a visual sociology. Unfortunately, this is not the case in sociology.
Implicitly and explicitly (particularly in the discourses of visual sociologists themselves) certain media are ascribed certain intrinsic qualities. Writing is considered to be objective and unproblematic. Whereas photography is variously considered to be problematic and basically good for producing emotional reactions, and for popularizing research results. Both in the view of visual sociology and its detractors, photographs allow to show society more immediately, and to produce immediate reactions. This impression is certainly furthered by a tradition of visual sociology to photograph the marginalized, the poor, the rural, or colourful examples of multiculturalism and the worlds that are entrenched by modernity. Photographs, it appears, allow a more direct access to social “reality”, which is conceived in terms or marginalized humans, as opposed to abstract quantitative data or interview transcripts.
From there it is a small step to think visual sociology is more “art” than “science”, a separation that ascribes art immediate emotionality and science detached observation. Apart from the fact that many contemporary artists would feel insulted by such ascriptions, it mistakes the effects of certain practices for features of its media.
Again, compare this with astronomy or biology. We can learn a lot from studies of scientific representation, and one thing is that scientific images are not means of popularization. Rather, they are very difficult to understand and don’t make a lot of sense without the necessary theoretical, scientific and interpretive apparatus. Open the pages of a random issue of Nature, look at an image, and try to ascertain whether this is more art than science, whether these images speak emotionally and immediately. This is not to deny that certain scientific images are aesthetically thrilling, and indeed are promoted by scientists as “art”. And indeed, the aesthetic qualities matter greatly for how they are understood in scientific discourse. But even so, they are not understood to be problematic or non-scientific “art” within these disciplines. They are data, in the same sense as written answers to questionnaires are data.
The problem then, is not so much that certain media produce certain effects, but that certain disciplines use certain media in particular ways and ascribe them particular effects. Sociologists (and Anthropologists) are rather unique because they have asymmetrical theories about different kinds of media. Sociologists know well that a text by Margaret Archer is rather different from an ethnographic study by Les Back and this is again different from say investigative reporting, even though they are all composed from the same 26 letters of the alphabet printed on paper. Sociologists inherently accept that texts as media are neutral with respect to their effects, be they emotional and “subjective” or detached and “objective”.
But when it comes to images, sociologists suddenly believe that it is the media that produce objectivity and subjectivity effects. A photograph surely must produce very direct emotions, and drawings are out of bounds altogether, unusable for the stern task of social science. And those who champion drawings, do so because they are supposedly more artistic, touchy-feely. But drawings can be as cold, analytic and objective data as are texts. Sociologists would only need to learn how to do this (see for example this project by my colleagues of shared inc.).
Not only is such an unequal media determinism a case of poor epistemology and ontology, it also hinders progress in sociology. Faced with the problem of legitimating their own subdiscipline, visual sociologists become complicit in this logic, by arguing that visual sociology is a compensatory exercise for the analytic coldness of their parent discipline.
How can we escape this trap? First of all, we should stop discussing visual media as if they were inherently problematic. Then we should endorse all kinds of media for our research purposes.
Doing so, there is an imperialist and a post-imperialist route. The imperialist route is exemplified by Howard Becker’s brilliant book on “Telling about Society”. As he writes: “I’m very imperialistic, always wanting to call smart people who do interesting work ‘sociologists’” (p. 151), which allows him to include Georges Perec and David Antin as sociologists. This is certainly fine with me and I am a proud flag bearer of this kind of imperialism.
But I have two doubts: first I am not sure whether people like Perec and Antin would like to be colonialized by sociology (though I am sure, if there is one colony one would like to be part of, it were Beckerland). More problematically, I am not sure this solves our problems with those sociologists that believe in a kind of unequal media-determinism that makes all kinds of non-textual representation smell of subjectivity.
To alleviate these problems, I think we may need to go the post-imperialist route by being more precise about the difference between “telling about society” and “doing research”. Visual sociology from this perspective does not suffer from subjectivity, but from being too often content with “telling” rather than researching.
Let me explain: The impression one gets from many (though certainly not all) visual sociology projects is that they show photographs or videos, as something that is seemingly self-explanatory. There is a double problem here: First, the way how these photographs are taken (and it is typical these are always photographs, and not drawings): They show something we supposedly can all recognize: say a poor woman sitting in the streets, or a carpenter doing his craft. This involves a certain “objective” framing that supposedly makes the object easily recognizable (no filters, no false colours, no extreme angles etc.). Thus they are based on a certain understanding of photography as a medium and a way of using it that ensures “objectivity”. The implication is that if only one does not interfere with the medium, objectivity is secured.
Second, the way these photographs are not interpreted. Very often, we see visual sociology as “exhibitions”, where people are assumed to understand what is seen, precisely because of their framing as per the previous point. Because the object is well recognizable, there is supposedly no need for further manipulation and interpretation. But such a use of visuals is what I would call “documentary”. It is a kind of popularization, in the sense that it shows what the viewer already knows (If the viewer wouldn’t know it, she wouldn’t understand it). To understand that the woman is a poor woman, we already need to know what poor women look like.
But research, as scholars in science studies have shown, is precisely the opposite of the above: it is about making the invisible visible, about creating data of “objects” that do not exist independently of their representation and that only become visible because of their manipulation. Again, look at any random image from biology, such as these and these, or think about what happens to recorded interviews. What you can see in these images is not “nature”, but a highly manipulated object, made to fit the frame, forced under the camera, often in false colours and distorted angles, with writings on the image. These images are not “subjective” because they are manipulated. If anything, the manipulation makes them more objective. This is because they are part of a chain of translations that allows rendering something visible through research that would be invisible by simply looking at a photograph of a fly (in the links above: brain control of muscular movement in drosophila).
Thus, if we would really want to produce visual research, we would need to feel comfortable about taking a route, which is precisely not documentary, but manipulative for the sake of making the invisible visible, the untouchable touchable, the unhearable hearable, or even the invisible touchable and the unhearable readable. We would need to become comfortable turning objects step by step into (visual) representations and accounting for each of these steps. Once we get there, we can hopefully abandon the IVSA and our MA course. In the meantime, we can maybe use the IVSA and our MA course to get there a bit faster.