Von der Emergenz zur Etablierung komplexer Präpositionen, modelliert mithilfe token-basierter semantischer Karten (am Beispiel des Russischen, Spanischen, Deutschen)

Abstract

This dissertation examines the processes by which complex prepositions emerge, stabilize, and become functionally integrated within the linguistic systems of Russian, Spanish, and German. It adopts a token-based semantic mapping framework which provides a model of how recurrent usage patterns—often involving collocational structures—lead to the conventionalization of multi-word prepositional units. Through a combination of corpus-based analysis and cross-linguistic comparison, the study traces the semantic and syntactic shifts that accompany these developments, identifying recurring trajectories of formal consolidation and semantic narrowing. The approach emphasizes usage-driven dynamics and gradual restructuring over time. In addition to traditional qualitative and manual analyses, this study integrates contextualized embeddings generated by BERT (Bidirectional Encoder Representations from Transformers) to perform large-scale semantic clustering of prepositional expressions across historical periods. By modeling the semantic similarity of token distributions, BERT enables the construction of data-driven, usage-based semantic maps that reveal patterns of semantic change, desemanticization, and refunctionalization. The combination of manual annotation and AI-based vector analysis allows for a nuanced interpretation of the semantic reorganization of multi-word prepositions and their integration into linguistic structure. This hybrid methodology offers a replicable and scalable approach to historical semantics, and deepens our understanding of how lexical constructions become entrenched across time and languages. By offering a systematic account of how prepositional expressions evolve and become entrenched in discourse, the study contributes to broader models of language change, structural variability, and linguistic categorization.

Keywords: complex prepositions, semantic maps, usage-based linguistics, corpus analysis, BERT, Russian, Spanish, German, multi-word expressions, language change, syntactic reanalysis, lexical development.