Abstract
An earlier article by Hansson (3) studied Swedish verbs in the speech of five grammatically impaired children. Verb Type-Token Ratios (TTRs) were calculated to compare children's lexical diversity for main verbs, and to investigate the relative diversity of state verbs and change verbs. Results showed that for all children except one, change verbs were more varied than state verbs. However, the validity of the TTRs is called into question because the most advanced child in the study had the lowest verb TTR. It is argued here that this anomaly, together with one child being an exception to the finding of greater diversity in change verbs, is an artefact of calculating TTRs from unequal sample sizes.