Lexical Semantic Change for the Humanities

Language provides a window into societies and cultures of the past and the present. In the past decade, the development of large historical/diachronic text corpora and the popularisation of computational methods in the humanities have allowed scholars to leverage language patterns to answer a large array of research questions in disciplines as varied as linguistics, history, archaeology, literature and media studies. 

Lexical semantic change, a phenomenon describing the evolution of the meanings of lexical units, has long preoccupied historical linguists. The possibility to study this phenomenon with computational methods has far reaching implications in several disciplines since diachronic semantic change can be used to identify conceptual shifts in questions related to many aspects of society such as culture, religion, politics, economics and morality. The recent convergence of Natural Language Processing and Computational Linguistics with a range of humanities disciplines has led questions related to lexical semantic change to gain momentum. 

In this three-part seminar series, we will hear from three timely projects which are pushing the boundaries of Digital Humanities by creating synergies between disciplines to develop cutting-edge methods to investigate lexical semantic change at a large scale. 

The series is co-organised by CAMille (ULB-KBR), the Data Science Lab (VUB-KBR), the Digital Research Lab (UGent – KBR), and LabEL (UCLouvain-KBR).

Video

Semantic change is usually discussed from the angles of onomasiology or semasiology. Thus, one may focus on words representing the same meaning as in child, girl, baby meaning ‘young person’ or focus on meanings attached to the same word, as in girl ‘child, young person’ (c1300) versus girl ‘female young person’. These traditional approaches assume that the meaning lies in data that is conceived of in categorical and linear terms.
In this talk I present a perspective on semantic change in terms of paradigmatic relations across a text. I discuss the findings from the Linguistic DNA research project which analysed concepts in discourse of 55,000 Early Modern English books. I present the most recent theoretical and methodological innovations, which include bottom-up modelling of prosodic meaning.

About the speaker

Justyna Robinson is a Senior Lecturer in English Language and Linguistics at the University of Sussex (UK) and the lead of Concept Analytics Lab. Her work on semantic change is informed by the understanding that meaning making is essentially a socio-cultural and cognitive phenomenon. In her work she experiments with the ways big data and computational approaches can get us closer to understanding how and why meaning changes. More recently her work focusses on concepts. To this end she has co-developed methods for detecting changes in concepts via the Linguistic DNA research project. She is currently working on concept detection in the Proceedings of the Old Bailey and Mass Observation Project archives.

Should you have any further questions please email vfnoryyr.tevobzbag@xoe.or or oerpug.qrfrher@xoe.or.