An Optimization Model of Language Complexity

Revista Glottometrics. Noviembre de 2016.

In this paper we develop a theoretical model of global language complexity, based on a constrained optimization approach. We assume that language is a system that chooses different levels of complexity for its different domains (i.e., phonology, morphology, syntax, vocabulary) in order to minimize a global complexity function subject to an expressivity constraint (which also depends on nonlinguistic variables related to geographic, phylogenetic and demographic factors). The model is illustrated with the aid of a dataset based on a short text translated into 50 languages, for which global complexity is measured using a version of Kolmogorov complexity. That dataset is used to run simultaneous-equation regressions, which represent different relationships between language complexity measures.