I know only too well of all the corpora and features that we have still not fully realized (or even not yet begun to work on) for the edition, although they were initially conceived as primary goals. I can foresee, for instance, that we will probably not have enough project time to encode the correspondence of the Friday Society, an exchange of letters between five protagonists, sometimes one-to-one letters, sometime group-to-one or one-to-group. I have been wanting to edit this correspondence for ten years now. For seven years, I was blocked by the difficulty of choosing between presenting it in a chronological or in a person-defined order. Then came the digital edition and with it the unequaled opportunity not to choose, but to offer it all. A dream come true, really. Except that it is another dream that is coming true in the end.
For two years now, we have been more and more working on author-based corpora, which was precisely what I wanted to avoid. My idea was (and still is) that you don’t get to embrace the reality of relationships and networks through an approach based on personality cult. So as we began to focus on authors (Tieck, Boeckh, Chamisso, to name the bigger ones), I thought I had to accept the grieving over never being able to seize networks. I consoled myself thinking that I would not have been able to define said networks in a way that would have impacted my scholarly environment anyway. I said that we did not have enough entries to visualize anything. Crappy arguments, really.
I am not saying that defining relationships is an easy thing. But at the very least, it is something that does not as much resist formalization as I first dreaded. We implemented something of that sort. At first it was something fairly intuitionally conceived and consisted only in capturing family relationships – what we later brought a theoretical step further with the concept of genealogy (meaning by that not only the family, but also the intellectual generations). We encode, so far as they are known and both persons appear in our edition, parent-child, spouse, lover, sibling relationships – the example of August Boeckh shows how long the list can be for one single person. This example also lists in which function Boeckh took an active part in major institutions. This other list is realized on the basis of a tedious work on what we call the group index, for which we need precise entries, e.g. to know exactly, for instance, which professors were in a faculty or another in a given academic year, who directed the seminar for scholarly schools in which year, etc. In order to retrace intellectual genealogies, we want to add student-teacher relationships on the same model as we have parent-children relationships. When this will be done, August Boeckh will appear as “student of Friedrich August Wolf” and all of the students from the philological seminar listed in our database – there are a few of them – will be listed after a phrase like “Teacher of: Immanuel Hermann Fichte, etc.”. In the end, all the things that it takes me pages to explain in my papers – that I.H. Fichte was the son of J.G. Fichte, attended his father’s classes, but was also greatly influenced by Boeckh – would be contained in two lines of a simple biographical entry.
But understanding this information supposes a great deal of background knowledge. It all comes down to being able to “read” the entries, e.g. to know what, in this very short information, matters – and what matters less. And so, after admitting that, to be honest, we have by far reached the critical mass of information needed to approach the corpus not just qualitatively, but also quantitatively, I also have to admit that defining a visualization strategy for this relationship information might, after all, actually help structuring, hierarchising, filtering it – being able to do something with it. Of course, a simple query already allows me to state that everyone seemed to have read Rousseau’s Nouvelle Hélöise between 1761 and 1860. But confirming well known hypotheses is one thing. Finding new ones is another. After running after intellectual networks for ten years, it seems to be time to nail them down properly.