Variability and path dependence of the lexicon in use in multilingual contexts

• Applicant: Dr. Anna Shadrova
• Subject Area: Applied Linguistics, Computational Linguistics ; General and Comparative Linguistics, Experimental Linguistics, Typology, Non-European Languages
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 555347674

This project aims to describe effects of local path dependence and other variability-inducing phenomena in the lexicon in use in task-based corpora and to relate those to observations of structural aspects of the lexicon from cognitive disciplines. It is designed against the backdrop of a massive lexical variability found across task-based elicitations of homogeneous L1-cohorts that cannot be plausibly attributed to external factors such as register, posing a challenge to the existing methodological and analytical approaches. Much of this variability appears to be attributable instead to effects of local cognitive interaction and path dependences, i.e. realizations that depend on previous realizations within the same production. These observations systematically converge with cognitive models of associative structures of the mental lexicon, thus promising generalizable insights into structural developments of the mental lexicon in multilingual contexts and its role in language use in general. This includes well-described phenomena such as priming or alignment in dialogue, which have not been widely considered as explanatory factors for the variability of lexical corpus data to date. Open questions include the role and influence of different elicitations (a typology of task effects), the degree of individual preferences vs. generalized cognitive patterns, and the influence of the network properties of the mental lexicon on those procedural phenomena. Moreover, some observations suggest a dialectical interaction between aspects of those network properties and discourse realizations, i.e. a dependence of discourse content and thematic progression on structural aspects of the mental lexicon, which in turn is influenced in its structure by effects of multilingualism. This extends the project to a sociopolicital dimension concerning the role of multlingualism and successful navigation of educational settings in an increasingly diverse society. The project aims to systematically model these phenomena in corpus data and relate them to cognitive and psycholinguistic models of the structure of the lexicon via a triangulation with experimental and imaging data from the same speakers. The primary goals of the project are a) to gain an understanding of the extent of inter- and intra-individual variability in the use of the lexicon and the role of lexical path dependences; 2) to deduce a model of the developing L2 lexicon through a convergence of corpus and experimental/imaging data; and 3) to differentiate between effects of acquisition dynamics and general multilingual cognition therein. Data will be collected longitudinally from mono- and bilingual German L1 college students and newly immigrated L2-German college students.

• DFG Programme: Emmy Noether Independent Junior Research Groups