Publications
Chosen filters:
Chosen filters:
Markov models for multi-state language change KU Leuven
Historical linguistics has witnessed an upsurge in quantitative corpus studies. The bulk of these studies involve the use of regression modelling. We point out a number of potential problems with this approach, and offer an alternative. For a multi-state language change, we propose a Markov model in continuous time. The major advantage of this technique, which has been used in medical contexts, is that it is especially geared towards dealing ...
Changing preferences in cultural references KU Leuven
This proof-of-concept paper advocates the use of lasso regularization for tracking changes in our collective cultural memory, as documented in the journal De Gids, issues 1837–1999. Mentions of 264 well-known historical figures are treated as predictors in an aggregate binomial regression for the era of publication, to see which historical figures are more likely to be mentioned in issues in the (long) nineteenth century, as opposed to the ...
De sterke werkwoorden in het Nederlands. Een diachroon, kwantitatief onderzoek. KU Leuven
The goal of this project is to trace the diachronic development of the preterite morphology in Dutch from the early Middle Ages up until present-day Dutch. The primary focus is on the competition between (1) the so-called strong inflection (e.g. vaar - voer ('sail')), which is based on root apophony (Ablaut), and has developed from the Proto-Indo-European perfect (Mailhammer 2007), (2) the so- called weak inflection (e.g. vaar - vaarde), which ...
De kracht van sterke werkwoorden KU Leuven
A corpus-based quantitative analysis of twelve centuries of preterite and past participle morphology in Dutch KU Leuven
Germanic preterite morphology has been the subject of a bewildering number of studies, looking especially at the competition between the so-called strong inflection (operating with ablaut), and the so-called weak inflection (operating with suffixation). In this study over 250,000 observations from twelve centuries of Dutch were analyzed in a generalized linear mixed-effect model gauging the effects of a multitude of language-internal factors, ...
Semantic differences between strong and weak verb forms in Dutch KU Leuven
Dutch, like other Germanic languages, disposes of two strategies to express past tense: the strong inflection (e. g., rijden – reed ‘drive – drove’) and the weak inflection (spelen – speelde ‘play – played’). This distinction is for the most part lexically determined in that each verb occurs in one of the two inflections. Diachronically the system is in flux though, with the resilience of some verbs being mainly driven by frequency. ...