Guest Lecture: Dr. Dominic Schmitz (HHU Düsseldorf) — ‘Gender-(un)fair language and the mental lexicon: Computational Insights’

Abstract: Gendered word-forms differ not only in their social connotations but also in how they are mentally represented and processed. This talk explores how computational models 
of the mental lexicon can shed light on these differences. I will introduce approaches 
from distributional semantics, which represent word meanings as vectors in a semantic 
space (cf. Harris, 1954), and from discriminative learning, which models how form
meaning mappings are learned and processed in the mental lexicon (Baayen et al., 2019). 
Together, these methods allow us to simulate and analyse how gendered word-forms – 
such as specific and generic masculines or gender-star forms in German – differ in their 
semantic representations and processing. Through examples from recent studies, I will 
illustrate how computational modelling helps us understand biases in language not as 
comprehension errors, but as reflections of the linguistic input that shapes our mental 
lexicon.