Areal biases in stop paradigms

The well-known hypothesis that stop paradigms tend to display gaps at the voiced velar and voiceless bilabial positions was tested. Data from UPSID suggest that this tendency represents a rough average across the world's languages, but that it misses some strong areal biases. For example, African languages do not avoid voiced velar stops, and they are responsible for most of the underrepresentation of the voiceless bilabial stops. Thus, it can not be concluded that velars and bilabials constitute universally underrepresented members of the respective voiced and voiceless stop series. Although this pattern is to be expected from proposed production and perception constraints, it is largely overridden by areal biases.