In Köln today, having received an introduction to the Hellespont Project, a joint enterprise of the Deutsches Archäologisches Institut and the Universität Köln to produce a digital annotated version of the Pentekontaetia section of Thucydides. An experience, I must admit, that I found somewhat akin to having one’s brain removed, slapped around a few times and reinserted in a slightly different position. That is no reflection on my hosts, who were astonishingly generous with their time, but on my lamentable ignorance of all this digital humanities stuff. At times like this, one can’t help lapsing into Rumsfeldisch: I knew that I didn’t know much about certain things, but had no idea of all the things about which I don’t have the faintest hint of a clue.
This has, therefore, been a day in which I’ve lurched between the sort of awestruck wonder you’d expect if you put a particularly ill-informed Gaulish peasant farmer in the centre of Rome (Wow, I never knew you could have buildings that big! Wow, I never knew there was a webpage that showed you the location of ancient cities on Google Maps!), the sort of blank incomprehension you’d expect if you tried to introduce him to basic concepts in Roman law (locatio/conductio? TEI 2, CIDOC CRM and CTS-URN? Huh?) and the desperate and rather silly attempts he’d make at trying to relate this new world to his own experience (So that’s like the chief’s hut taking up the whole of that hill? I used to do webpages, you know, so I recognise those angly bracket things…). I really can only admire the patience of Agnes Thomas and her colleagues in talking me through all this, and hope that it starts to come together and make more sense over time.
The reason for the visit – besides a general interest in all things Thucydides – is of course the plan that I’ve mentioned on this blog before to turn some of the results of the Reception of Thucydides project into a digital resource aimed primarily at school and college students interested in ideas of citizenship, democracy and the like. That’s a rather different audience from the Hellespont project, which is intended primarily for researchers (it’s based on the Greek text, for a start), but it seemed prima facie that there might be scope for some sort of productive collaboration, and at the very least it would give me more of an idea of what’s involved in such a resource. Well, I still think that there are some obvious connections to be made (not to mention the fact that I’m going to find it extremely useful to be able to draw on their material in due course), but above all I am now uncomfortably aware of quite what I’m proposing to take on, and how steep a learning curve this is going to be.
The idea of an annotated text, marked up in xml, was pretty easy to grasp; it’s more or less what I had in mind already, though I’d been thinking of links to other webpages, or simply to those floaty comment things that appear when you hover the cursor in the right place [n.b. I am making no claim whatsoever to technical expertise here], rather than to a database. I didn’t have a clue that there are already clear protocols regarding how such a text should be annotated, to make it as open and useable as possible, nor that there is clearly a powerful set of expectations in this area (the revelatory moment came in the middle of a Skype conversation today with Reinhard Förtsch in Berlin, when he complained of someone who had analysed the personal connections within a corpus of letters that this work had not improved the text itself at all, since it involved a separate, personal database; merely analysing a text using computer tools does not a proper digital humanities person make…). Suddenly, the stakes feel a lot higher; if I want to do this properly, and not have proper digital humanities people making derogatory comments behind my back, then merely sticking a few comments into a digital text isn’t going to make the grade.
And then we came to the question of what gets annotated, and in what form (which is where CIDOC comes in, and threatens to give me a headache). Place names, fine (automatic links to gazeteer and maps); names of people, fine (automatic link to the Arachne database of the DAI, and I need to check whether the English version is any good, because the German one is really useul). Events (and activities, which are similar but more human-focused); okay, I can see some philosophical issues with defining and limiting events, let alone the implication that you should then identify a single actor and a single cause (or at any rate that was my initial impression), but on reflection I can also see that thinking through a specific line in Thucydides so that one is in a position to state its contents in a clear propositional format (Actor (Athenians) give rise to (provoke) Event (Peloponnesian War)) could at least identify more precisely exactly what we ought to be debating – to take just one example that Agnes showed me, why is it that so many commentators on Thucydides conflate the various events reported in I.105-6 into a single lump (not necessarily characterised in the same way)? How far does this reflect Thucydides’ own presentation of them, and how far their preconceived notion of the events’ significance?
If we can’t easily define in this manner exactly what Thucydides is saying in a particular sentence, that probably tells us something as well. However, at this point I do start to feel slightly nervous – and I don’t think it’s solely because all this was reminding me strongly of xml indexing, which I still dislike intensely. Is it possible to encode uncertainty and ambiguity, where that’s (arguably) apparent in the text, or is it necessary for the purposes of relating the sentence to the database that some kind of decision is made, so that the action is defined as one thing or the other (or one cause rather than another is identified), and the alternative or alternatives are effectively erased, even though we may imagine that the normal reader of Thucydides would remain conscious of the ambiguity and uncertainty (and was expected to by the author)? What may be lost, therefore, in the process of enriching the text through annotation? The simple answer to this is that I am far too ignorant of exactly what’s going on to be able to comment – maybe this is no more a problem than it is with a commentary deciding to present some issues as problematic and pass over others – but I still worry slightly…
It also raised questions for me as to how far this approach, so clearly suited to conventional historical questions about places, people and events, and equally suited to conventional philological questions about words, syntax and the like, would work for more interpretative approaches, where a key concept may not be identifiable with any specific term or sentence or even section of the text. This is comparable to my regular wish to have a single index entry for a sequence of pages, which is of course possible (albeit a bit messy) in xml, but perhaps not ideally suited to the construction of a database. How, also, is this approach going to work for the study of reception? Here I can see at least a partial way forward: something along the lines of Textual Object (Thuc. II.45) is referred to by (cited by) Textual Object (Leo Strauss) – I should stress that I’m making up these phrases (the real ones are scarier). This doesn’t really engage with the problem of how far many receivers of Thucydides are responding to an idea of Thucydides rather than his text – but then an annotated text isn’t the right place to engage with that anyway.
There are times, I must admit, when I wonder if I shouldn’t just admit that I’m old and past-it, and stop trying to bridge the digital divide, as it becomes evident that it’s even bigger than I thought – dammit, I don’t even have a smartphone. I’m reminded of what the Penguin Guide to Jazz on CD says of the jazz guitarist Pat Metheny. Metheny is noted for what I can only describe as tastefully smooth guitar widdling, as he plays technically complex and sophisticated but utterly bland guitar lines over a tasteful, bland keyboard background (this is not, as you may have gathered, my sort of music). However, despite or because of the fact that this style of music has made him large sums of money (at least by jazz standards), every so often Metheny takes a break to play out-there free jazz with giants of the avant garde like Ornette Coleman and Derek Bailey, producing what most people think is a truly horrible, atonal noise (obviously these are the only Metheny records I like). The Penguin Guide thus describes him as the skinny boy at school who keeps putting himself in the roughest, toughest playing situations so that no one will think him a wimp, even if he gets creamed time after time.
It’s not that Metheny repudiates his more mainstream music (much as I might wish him to), but he doesn’t want to play things safe all the time, is clearly fascinated by the challenge of the avant-garde, and senses a kinship even if no one else can see it. There’s a definite measure of arrogance there, or at least of absolute self-confidence, but also of courage. Trying to get to grips with the world of digital humanities a good five or ten years too late carries, perhaps, a similar risk of looking very foolish and missing the point completely, but the alternative is to remain oblivious to one of the cutting edges of the subject. Somehow, therefore, I need to find the time and the money to develop the Thinking Through Thucydides resource – or, to be more exact, the annotated text that will underpin the resource – properly, meeting the expected standards while remaining true to my original goals.
It’s not that it’s necessarilly done well, it’s remarkable that it’s done at all…