Models of Lithuanian Grapheme Frequencies

Abstract

While the study of grapheme counts dates from ancient Greeks, that of their frequencies dates from the ninth century AD. These studies moved beyond the comparison of means after almost a millennium. Because of their use in cryptanalysis and elsewhere grapheme frequencies continued to be studied. Published work in the area has been concentrating on the more widely spoken languages and comparatively little has been done with minority languages. An alphabet for Lithuanian was first developed in the sixteenth century and finalized in the twentieth. Twenty models of its grapheme frequencies were examined here. A relative simple model was found to be the best fit. This model may also be applicable to other languages.