Doctoral thesis

This PhD project examines lexical creativity in scientific writing, focusing on metaphoric speech, semantic change, and colexification. Scientific language uniquely combines technical precision with creative expression, often employing metaphors and analogies to explain novel phenomena. This creative process involves semantic change, where word meanings extend or shift, and colexification, where a single word carries multiple related meanings. For example, “radium,” derived from the Latin “radius” (ray), exemplifies how linguistic creativity conveys scientific discovery. The study aims to analyze these linguistic phenomena in scientific texts using computational techniques, leveraging Natural Language Processing (NLP) and machine learning models such as Word2Vec and BERT. The corpus will include multilingual writings by scientists like Alexander von Humboldt and Marie Curie, collected from reputable scientific archives and libraries. Specific objectives include analyzing metaphoric speech for its systematicity and communicative principles, investigating semantic change patterns over time, and studying colexification to uncover semantic nuances.

The methodology includes preprocessing the corpus, implementing metaphor identification algorithms, and using word embeddings to detect colexification. Both quantitative and qualitative analyses will evaluate the frequency, distribution, and impact of these linguistic features, enhancing understanding of lexical creativity and its role in scientific communication.

The research, structured in five phases from literature review to result analysis, aims to bridge linguistic theory and scientific practice, contributing to the fields of lexical semantics, language evolution, and the understanding of scientific discourse.