You are here: Home :: Area CULI :: Linguistics and literature

Debowski, Lukasz

Zipf's law against the text size: a half-rational model

In this article, we consider Zipf-Mandelbrot's law as applied to texts in natural lan-guages. We present a simple model of dependence of the law on the text size, which is featured by variable power-law tail and constant ratio of the most frequent words. As a result we derive several closed formulas, which accord with empirical data qualitatively and partially quanti-tatively. For example, there appears to be a minimal length of literary texts equal to ? 159 word tokens for English.

Glottometrics, RAM-Verlag

Volume: 4, 09/2002
Pages: 49-60

Show full article (external site)

Show all available items of this journal

 

Area CULI