The first letters I encoded in the past few days are not really representative of the rest of the corpora, but they forced me to deal with interesting problems concerning the architecture of our index of places. Here’s a sample:
Après la rupture de l’<placeName ref=”#l0004″>Espagne</placeName> et de
l’<placeName ref=”#l0005″>Angleterre</placeName> la guerre paroit se
perpetuer et je n’en vois plus la fin, sur tout, si <lb/>la prediction de Mr.
<persName ref=”#p0008″>Litt</persName> qui a deviné si juste dans les
affaires de la nation, se verifie, que la guerre avoit commencé en Amerique
<lb/>et ne pourroit aussi finir, que là. Le General <persName ref=”#p0009″>Platen</persName> a occupé <placeName ref=”#l0008″>Naumbourg</placeName> et
<placeName ref=”l0009″>Zeitz</placeName>. Sil parvient à chasser les ennemis
de <lb/>l’<unclear reason=”german script, not so good legible, could mean Erzgebirge”>Erthgebürge</unclear><!– wie kann ich das besser machen? –><!– ausserdem schriftwechsel in die dt schrift!!! –>, il reparera en quelque
façon les fautes en <placeName>Pomeranie</placeName>. On dit, qu’en <placeName ref=”#”>Silesie</placeName> tout est en mouvement et on pretend <lb/>même
que le Roi pourroit bien marcher vers la <placeName ref=”#l0010″>Pomeranie
A work in progress, as you can see…
My coworker Alexander Meyer, who is, among many things, in charge of helping us putting up decent TEI and data architecture, first wanted to proceed with an index of places structured in countries>regions>cities>streets as Solenne Coutagne had first suggested (only reversing it so it would be a tree).
The first difficulty I encountered was “Germany”. Germany today, Germany in 1760, Germany in 1840 are different geographical entities. So we have to give up referring to one geographical entity, we have to consider more like the word “Germany”, and let the people using the search engine figure out themselves which Germany is meant in the texts of 1760, 1784, 1806, etc.
The next obvious thing is that we will have to tag those place names both in French and in German to make them searchable. I am now wondering if it would be worth putting the English variant of the place names too.
In those letters written during the Silesia war, as you see, Beausobre’s correspondent gives him information concerning troop movements, and some of the places named there do not really exist any more (they may have been inhabited in the 18th century and are only very small places now, not even villages), or they carry names like “Neustadt”, “Breitenfelde”. There are at least 10, if not 20 places called like that in Germany. Probably historians would know immediately which Neustadt is meant, but I have trouble spotting it. And my problem is, I really don’t care where this or that Neustadt is, and I don’t want to spend all of my encoding time looking for it when I know that there are letters dealing with publications I will be much more eager to retrace. I am looking for a way to isolate all the “unspotted” places and mark them as elements on which I would require feedback from my historically knowledgeable readers, or something like that.
In the end, this WILL be crowdsourcing, I guess… (and about that, see this paper by Hubertus Kohle!)