Doculect

This is not the whole explanation why these Mayan doculects have an exceptionally high token frequency, but it explains part of the deviation.

However, for the present study these texts are considered to be specific doculects, which might, or might not, be the 'same' language as described in other resources (e.g. grammars or spontaneous text collections).

The most extreme case is illustrated by the doculect of McDonald and Wurm (1979), in which their consultant was being interviewed for materials in Garlali, but he preferred to speak Punthamara, so that is what he gave them.