The Oxford English Dictionary (OED) is
sponsoring a webinar that dives deep into the nature of language and how we create definitions.
Words that occur together can tell us something about meaning in language. Theories about such co-occurrence originated in the 19th century, and have been revised (and occasionally re-invented) into the 21st century. The Linguistic DNA project has designed a new computational linguistic approach to model historical word meanings by identifying and ranking a specific kind of lexical co-occurrence: co-occurring non-adjacent lexical trios in discursive spans of text.
Join Dr Seth Mehl, post-doctoral researcher at the University of Sheffield, as he shows us how this tool can be used to mine texts and identify the co-occurrence of specific lemmas, investigate the implication of these patterns on meaning, and inform the way that definitions are written and presented by the OED. For example, the tool can mine texts and identify paragraphs containing: democracy, history, and Greece, and paragraphs containing: democracy, diversity, and tyranny. A linguist can then investigate whether this distinction reflects two pragmatic meanings associated with democracy: as a virtuous ancient example of popular government, or as a threat to the early modern social order.
This new analytical technique has been applied to Early English Books Online (specifically EEBO-TCP), a collection of over 60,000 works from the 16th and 17th centuries, totalling over 1 billion words.
This looks to be fairly technical but some of my writer friends might be interested in it. The webinar is on Thursday, March 12 at 7:00 a.m. EDT.
No comments:
Post a Comment