Welcome to COALA!
Computational Corpus Annotation for Quantitative Analysis of Latin Lexical Semantics
How do words change meaning over time, and how can we measure it?
Language is constantly evolving, yet the meanings of words are not directly observable. In historical languages like Latin, understanding meaning has long relied on dictionaries and small-scale qualitative studies. While invaluable, these approaches cannot fully capture the complexity of semantic variation across centuries, genres, and contexts.
COALA brings together computational methods and historical linguistic expertise to make large-scale, quantitative analysis of word meaning possible for the first time in Latin.
The development of an innovative corpus annotation system built on advances in word sense disambiguation will allow us to study word meanings at scale and over two millennia.
COALA opens new directions across multiple fields:
- Corpus Linguistics
Developing scalable methods for consistent semantic annotation in historical texts - Computational Semantics
Advancing state-of-the-art techniques for analyzing meaning in context - Historical and Latin Linguistics
Exploring how meanings evolve across genres, registers, and time periods
Our main research questions are:
- How does polysemy vary across different types of texts?
- How do words within the same lexical field evolve together?
- When do semantic innovations emerge and how do they spread?
- Is Latin truly a “fossilised” language, or does it show ongoing semantic change?
Why Latin?
Latin, with its central role in Europe’s cultural and intellectual heritage, offers a uniquely rich testing ground, with over 2,000 years of continuous textual history, extensive digital corpora and research tools. This makes Latin the ideal language to pioneer quantitative historical semantics.
Want to know more?
For collaborations or inquiries, please get in touch with the PI, Barbara McGillivray.
The COALA project was successfully evaluated by the ERC and funded by UKRI (project reference UKRI947).
