A large-scale analysis of metonymy across languages

Metonymy is regarded as a universally shared cognitive phenomenon; as such, humans are taken to effortlessly produce and comprehend metonymic senses. However, experimental studies on metonymy have been focused on Western societies, and the linguistic data backing up claims of universality has not been large enough to provide conclusive evidence. I introduce a large-scale analysis of metonymy based on a lexical corpus of 20 thousand metonymy instances from 189 languages and 69 genera. Drawing on corpus and statistical analysis, evidence of universality is found at three levels: systematic metonymy in general, particular metonymy patterns, and specific metonymy concepts.