Self-organizing Pattern in Multilayer Network for Words and Syllables
This provides a new universal property for linguists studying language evolution, with practical application in analyzing trends like Internet slang.
The authors discovered a universal law in human languages showing that syllables play a complementary role to Zipf's law, with rank-rank frequency distributions for English and Chinese corpora fitting a master curve. They derived an analytic scaling form and applied it to quantify the evolution of Internet slang, demonstrating its utility as a tool for evolutionary linguistics.
One of the ultimate goals for linguists is to find universal properties in human languages. Although words are generally considered as representing arbitrary mapping between linguistic forms and meanings, we propose a new universal law that highlights the equally important role of syllables, which is complementary to Zipf's. By plotting rank-rank frequency distribution of word and syllable for English and Chinese corpora, visible lines appear and can be fit to a master curve. We discover the multi-layer network for words and syllables based on this analysis exhibits the feature of self-organization which relies heavily on the inclusion of syllables and their connections. Analytic form for the scaling structure is derived and used to quantify how Internet slang becomes fashionable, which demonstrates its usefulness as a new tool to evolutionary linguistics.