— Max Kemman (@MaxKemman) September 15, 2015
Early this month I spent two days at ‘Beyond Methods of Mining: a workshop on doing historical research using digital data‘. It was a splendid event and my first outing as a member of the Sussex Humanities Lab! My sincere thanks to the organisers (especially Tessa and Maarten) for both their choice of theme and for their hospitality.
Max’s tweet sums up the conference for me: it was an anxious affair, fraught – at least in the backchannel and during the over coffee conversations I had – with concern around the somewhat structuralist use of language employed in many presentations (including my own). And yet if you judged the conference on the titles of each talk alone (see my notes from the event) you probably wouldn’t spot this. For unlike many DHy events, this one was rather special as it – refreshingly – saw historians first and foremost present historical research into historical phenomena based on their use of data. Which is not to say that the research presented had been undertaken without consideration of theory and method, rather that we agonised about theory and method after getting stuck into the past.
Now if your positivism alarm bells are ringing, you’ll be pleased to hear that when I got home, I stumbled across Structuralist Methods in a Post-Structuralist Humanities, Seth Long‘s musings from afar on our twitter musings ‘in’ the room. Here Seth asks some vital questions, including:
Is it too much of an unnatural vivisection to insert structural, quantitative methods into a post-structuralist humanities?
Thinking through problems of this nature is vital to writing good history, whether or not it happens to be supported by computational work with data. But as Seth picked up on, I’m not ashamed of being positivist first, of beginning by throwing the digital kitchen sink at a problem to see what sticks, and then – through sustained criticism of my own work – uncovering the problems later. Sometimes this process leaves me with nothing. Other times it leaves me with something, only for a collaborator or audience member to pick up on an issue, leaving me again with nothing. But sometimes, after all the checks and balances have been thrown at an initially positivist endeavour, I am left with something. And something is better – to my mind at least – than agonising so much over theory and method first that you end up with nothing.
Code, data, deck, and viz from my talk ‘Acts of being in proxies for prints. People in the British Museum catalogue of Political and Personal Satire, 1770-1830’ is available on GitHub under an admittedly slight confusing array of Creative Commons licences (thanks British Museum for throwing an NC spanner in the works! Though I respect your decision…). The talk builds on research I have blogged about here before, most recently as ‘Code, Control, and the Humanities‘. The Q&A after the talk made me realise that I really need to tease the part of speech data from the descriptions. If you are interested in collaborating to tackle this problem and co-publishing subsequent results, please get in touch.