Michael Franke

Michael Franke is a Professor in the Department of Linguistics at the University of Tübingen. His research focuses on experimental pragmatics, cognitive modeling, and evolutionary approaches to language use, reasoning, and decision-making processes. He employs probabilistic modeling techniques to investigate higher-order cognition and the relationship between empirical data, abstract theory, and mathematical modeling.

His primary research interests include experimental pragmatics, cognitive and evolutionary modeling, probabilistic reasoning, Bayesian statistics, and computational linguistics. Franke's work bridges theoretical linguistics with experimental cognitive science, using both human behavioral data and computational modeling approaches. He has made significant contributions to understanding scalar implicatures, pragmatic reasoning, and the application of game theory to linguistic phenomena.

Recent publications reveal a growing focus on the intersection of human cognition and language models, particularly examining how large language models handle pragmatic phenomena like scalar implicatures, quantifier interpretation, and question answering. His work increasingly integrates neural and symbolic approaches to model pragmatic reasoning processes.

Professor Franke has developed several important research tools including magpie (a framework for web-based experiments), faintr (an R package for interpretation of categorical predictors in Bayesian regression), and R_eproducible_LaTeX (for writing reproducible research papers). He has also created educational resources including web-books on probabilistic pragmatics, Bayesian regression, and understanding large language models.

In his teaching, Franke offers courses on logic, data analysis, and Bayesian regression. His educational materials demonstrate his commitment to making complex statistical and computational methods accessible to students in linguistics and cognitive science.