News from Jena

With the end of my scholarship, I have left Paris now and pursue my work in Jena, where I work as a senior scientist at the Max Planck Institute for the Science of Human History. From April on, I will pursue an ERC-funded research project on Computer-Assisted Language Comparison (CALC) with a duration of five years. I will announce further details soon and the project will also have its specific landing page where the most recent results will be presented.

I also thought, now that I came back to Germany, that it might be time to share some of my general thoughts on language and linguistics in my own mothertongue, and I started a blog titled "Von Wörtern und Bäumen. Historische Sprachwissenschaft nach der quantitativen Wende" at Hypotheses. My goal is to write one post every month, not more and not less, and to cover also topics of more general interest rather than the very technical issues on language comparison which I discuss in my papers.

Last not least, a new paper by myself, Simon Greenhill, and Russell Gray just came out with PLOS ONE, titled "The potential of automatic word comparison for historical linguistics". This paper treats, of course, the problem of cognate detection. We tested the current methods as they are available in LingPy on a new testset and could confirm that the results are really close to 90% when comparing them with the cognates identified by experts. Not a bad score which could still be improved, but it is obvious that in order to further improve cognate detection, we need to shift the paradigm, as the computational methods which are only based on sequence comparison are now reaching their limits.