How much do categories for spatial relations vary across languages, and can we predict the best example of a category from its members?

Abstract

Is cognition the same across speakers of di erent languages? Best examples of color cate- gories turned out to be fundamental to this debate. Past research, that used best examples, has been conducted to argue for the universalist as well as the relativist view. The current research compared di erent models, which predicted best examples of spatial categories in four di erent languages. Two di erent representational spaces were found and the models were adjusted to work for both of these spaces. Subsequently, the performance of the models was compared. The results demonstrated that the quality of the models predictions was very inconsistent through all languages. By using the Pearson's and the Spearman's correlation coe cient, re- sults could be obtained that suggest that the use of language speci c representational spaces may be considered. Future research could use these language speci c representational spaces.