?
A Quantitative Study of Russian Colour Terms Buryj and Koričnevyj in the Google Books Ngram Corpus
We report a microdiachronic investigation of distributional semantics of two competing Russian colour terms (CTs) for 'brown', buryj (12 th cent.) and koričnevyj (17 th cent.), while using Russian subcorpus of Google Books Ngram. By conducting time-series analysis (1800-2009) of bigrams containing either of these terms, we estimated frequency of occurrences of the two "Russian browns" and explored changes in the extent of the terms' combinability with nouns signifying objects (N=259). Results provide evidence that in total frequency of use, koričnevyj overtook buryj at the beginning of 1920s and unequivocally prevails from the beginning of 1960s. Furthermore, the perplexity index indicates significant increase in the scope of objects whose denotations collocate with koričnevyj. This is complemented by the observed increase of the Jensen-Shannon divergence between frequency distributions of buryj and koričnevyj, with both phenomena being particularly manifested from 1960s. The obtained estimates of distributional semantics corroborate the status koričnevyj as the basic CT for 'brown' in modern Russian. The present diachronic corpus analysis provides novel insights into linguistic evolution of an emergent basic CT-by revealing the process of it gradually supplanting an old term with a similar colour meaning, the timescale of the new term's increase in usage, and significant expansion in its distributional semantics.