Categorial Responses to Synthetic Vocalic Stimuli By Speakers of Various Languages

This paper reports the results of presenting native speakers of thirteen different languages a set of synthetically produced vocalic stimuli. The speakers were asked to judge the relevance of each such sound to the vowel sounds of the native language and to group those judged to be relevant. The results of this procedure closely resemble vowel phoneme inventories as derived by analysis-of-production techniques but is claimed to be superior to the more common procedure due to its freedom from experimenter bias, its freedom from non-linguistic factors present in speech production, its speed, and its automability.