Nunc profana tractemus. Detecting Code-Switching in a Large Corpus of 16th Century Letters

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 2901–2908 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 2901 Nunc profana tractemus. Detecting Code-Switching in a Large Corpus of 16th Century Letters Martin Volk, Lukas Fischer, Patricia Scheurer, Bernard Schroffenegger, Raphael Schwitter, Phillip Str¨obel, Benjamin Suter Department of Computational Linguistics, University of Zurich volk@cl.uzh.ch Abstract This paper is based on a collection of 16th century letters from and to the Zurich reformer Heinrich Bullinger. Around 12,000 letters of this exchange have been preserved, out of which 3100 have been professionally edited, and another 5500 are available as provisional transcriptions. We have investigated code-switching in these 8600 letters, first on the sentence-level and then on the word-level. In this paper we give an overview of the corpus and its language mix (mostly Early New High German and Latin, but also French, Greek, Italian and Hebrew). We report on our experiences with a popular language identifier and present our results when training an alternative identifier on a very small training corpus of only 150 sentences per language. We use the automatically labeled sentences in order to bootstrap a word-based language classifier which works with high accuracy. Our research around the corpus building and annotation involves automatic handwritten text recognition, text normalisation for ENH German, and machine translation from medieval Latin into modern German. Keywords: language identification, historical languages, Latin, Early New High German 1. Introduction “Nunc profana tractemus” (Now let us deal with pro- fane things) writes Joachim Vadian in May 1548 in the middle of his letter to Heinrich Bullinger and switches from Latin to Early New High (ENH) German. In the first part of the letter he had written about the church and the bible in Latin, afterwards in German he deals with politics in France, Poland and Switzerland. Latin was the language of education and science all over Eu- rope at the time. ENH German was used for every-day conversation which often resulted in mixed language exchanges (cf. (Jung, 2016) page 21). Joachim Vadian was an intellectual from St. Gallen and in frequent letter contact with his colleague and friend Heinrich Bullinger in Zurich. The aforementioned let- ter is the 203rd letter in the preserved exchange between the two. They knew each other well. Did they always switch between languages when switching topics? Is there a pattern of code-switching for Vadian, or for oth- ers of the time? We investigate the automatic detection of code- switching in this corpus of 16th century letters. How often do the writers switch between Latin and Ger- man? What other languages are used? How good are off-the-shelf language identifiers in distinguishing be- tween medieval Latin and German when being trained on classical Latin and modern German texts? 2. Corpus Building Heinrich Bullinger (1504-1575)1 was a collaborator and successor of Huldrych Zwingli and an important multiplier for the ideas of the Reformation in Switzer- land and Europe. From his extensive correspondence, 1https://de.wikipedia.org/wiki/ Heinrich_Bullinger some 2000 …

Content truncated to 500 words