The tragic hero

For something I was writing recently, I thought it would be interesting to draw on a bank of verbal data to which I had access.  I had some broad overall intentions, but I didn’t quite know what would emerge from the data.  After putting all the text into a database, I explored and interrogated it first from one angle, and then from another, and then from another; the dependent and independent variables alike were up for grabs; the purpose of the exercise, and its overall framing, morphed as I went along. Eventually, the lengthy process – much lengthier than I had planned it to be – ended when I got the data to tell what I thought was an interesting story.

There’s nothing particularly strange about all this: I felt as if I was acting out a chapter from a standard Science & Technology Studies textbook (even if I make no claims that my methods met the natural scientific ideal). But what interests me in this sort of situation is precisely the way that the messy data can only be presented to the world at the point that they can be made to tell a story. What results is ‘fraudulent’ (Medawar, 1963) in the sense that the facts which it makes real are underpinned, and the choices to make them real are constrained, primarily by the requirements of story-telling.

It’s fascinating to think of any academic article as a short story, and about how this is more and less obviously the case – though still the case – across different disciplines.  A particular discursive universe is set up, with its own finite set of actors and causal relations; various plot devices (such as research questions) are deployed; and so on. But perhaps the plot always has pretty much the same structure. At a panel discussion last year on ‘The Allure of Happy Endings’ (at the LSE Space for Thought Utopias literary festival), Jonathan Gibbs made the interesting suggestion that non-fiction always has an implicitly happy ending. And I see academic ‘stories’ in particular as always tending towards comedy rather than tragedy: at the beginning, a particular problem with (one version of) the world is outlined; by the end, that problem is in some way resolved, but the world has been changed.

So I’m left wondering what it might mean to write an academic paper with an ‘unhappy ending’.  Would that only mean that the paper had been badly written?  Perhaps it could tell the unfortunate tale of some of the other unpolished findings, proto-facts, and half-glimpsed perspectives, which had been suppressed and expunged because they were at odds with the world narrated – because they did not serve the story. I’m not sure: the fact of their inscription might mean that the tale was no longer a tragic one.  But, at least, it might be more honest one.

London, 29 April 2017



Medawar, P. (1963). Is the Scientific Paper a Fraud? The Listener, 70: 377–78.