Comparison of perception-production vowel spaces for speakers of Standard Modern Greek and two regional dialects

This study compared the perception-production vowel spaces for speakers of Standard Modern Greek and two regional dialects. In experiment 1, participants produced the Greek vowels and chose vowel best exemplars (prototypes) in a natural sentence spoken in the participants' dialect. In experiment 2, the speakers who had made the recordings for experiment 1 chose themselves vowel prototypes. Cross-dialectal differences were found in both perception and production. Across dialects and experiments, participants' perceptual space was exaggerated compared to the acoustic one. Because participants' perceptual space in experiment 2 was calibrated to the participants own voice, perception and production data are directly comparable.