Distribution of complexities in the Vai script

TL;DR

This paper analyzes the complexity distribution in the Vai script, showing that traditional uniformity assumptions fail and that Poisson-based models better fit the observed data.

Contribution

It introduces and validates Poisson and hyper-Poisson models to accurately describe the complexities in the Vai script, challenging previous uniformity hypotheses.

Findings

Uniformity hypothesis fails for Vai script complexities.

Poisson distribution models fit the data well.

Hyper-Poisson models effectively describe connections.

Abstract

In the paper, we analyze the distribution of complexities in the Vai script, an indigenous syllabic writing system from Liberia. It is found that the uniformity hypothesis for complexities fails for this script. The models using Poisson distribution for the number of components and hyper-Poisson distribution for connections provide good fits in the case of the Vai script.