4: Discussion

Considering that these findings relate to utterances of a live performance and not to utterances under experimental conditions, the results strongly indicate that vowel discrimination in CV or CVC context can be maintained up to a frequency range of F0 of 800–900Hz, confirming the third position mentioned in the introduction.

Several questions arise from such a statement. Above all, possible vowel discrimination is found for F0 of sounds exceeding statistical F1 of all vowels except /a/. For /u/, F0 of high-pitched sounds can even match statistical F2. Concerning vowel production, it is questionable whether a possible raised larynx position alone can explain such a phenomenon. Concerning vowel perception, vowel discrimination obviously cannot rely directly on F1-F2. Concerning the vowel context, discrimination of vowel sounds on high pitches produced in isolation and in logatomes should be investigated in order to assess the in-fluence of transitions and meaning. Concerning the investigation of artistic singing and speaking styles, vowel identification proves to be style-specific. Above all, for “Western” Opera singing, no corresponding F0 level with comparable vowel discrimination is reported.

Further studies on vowel production and perception at different F0 and in different contexts of the vowels are needed to clarify the relationship between sounds, F0, and vowel qualities. Such future investigations concerns both artistic as well as everyday utterances and should address documentations of spontaneous everyday utterances, live performances as well as systematic vocalizations under experimental conditions.