Mining Queneau and the Encyclopédie (or: back from the break)

When I went on a short holiday in August, I did not expect it would be the beginning of such a long break in my blogging activities here. Well, I have been busy! And busy mostly with things that are related less to text analysis and rather to text encoding. (Of course, a blog post in itself would be in order about how these two are actually related.)

Today, I want to get back into the blogging spirit! So I will simply indicate that I am currently in Leipzig, at the “Frankoromanistentag”, the bi-annual conference of the French Scholars in Germany. Lars Schneider and me are organizing a session on digital philology here, and among the papers we had yesterday and are having today (the full programm is here), there are two which are especially relevant to computational text analysis.

The first paper was the one by Anne-Sophie Bories, which we heard yesterday. She told us about a database she built up in which she recorded descriptive information on the metrical structure of all of Raymond Queneau’s poetry (written over 50 years, and in more or less regular verse and in “vers libre”). She used the database to identify recurring patterns of metric structure in certain poems, and then went on to relate those patterns to specific themes in the verses and poems concerned. Very impressive and completedly original!

The second paper, which we have just been hearing, is the one by Glenn Roe, who has been talking about the various ways the ARTFL project is using data mining strategies to explore the famous Encyclopédie de Diderot et de D’Alembert. One of his topics was the question of the system of the “renvois” (references to other articles): how to use and visualise them, but above all on how to go beyond them in the electronic edition using data mining and machine learning. It turns out, for example, that automatic classification of articles into 360 semantic classes brings out a lot of agreements and some very interesting disagreements with the classification present in the print edition.

Of course, I also took advantage of this opportunity to talk about my stylometric experiments with the Molière-Corneille controversy, but this will be for another blog post. Coming soon!


Christof Schöch

Christof Schöch is a Research Associate at the Department for Literary Computing, Würzburg University, Germany. He leads the junior research group CLiGS (Computational Literary Genre Stylistics). His interests are in French and Spanish Literature, Digital Humanites, and Open Access.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *