Acoustic cues to vowel identification: The case of /ɪ i iː/ and /ʊ u uː/ in Saterland Frisian

Abstract

Saterland Frisian is spoken in Saterland in northwest Germany. This language has a complete set of close short tense vowels: /i y u/. Together with the short lax vowels /ɪ ʏ ʊ/ and the long tense vowels /iː yː uː/ they constitute series of phonemes that differ by length and/or tenseness. We investigated which acoustic cues distinguish the sounds within two triplets containing /ɪ i iː/ and /ʊ u uː/ respectively, conducting a traditional reading task in order to obtain 'normal speech' and a listener-directed task in order to obtain 'clear speech'. In the normal speech condition we found for both triplets that short lax and tense vowels were distinguished by F1 and F2. Short and long tense vowels within the /ɪ i iː/ triplet were distinguished by vowel duration and F2. For the /ʊ u uː/ triplet we did not find any cue that distinguishes short and long tense vowels. However, in clear speech, we found that short and long tense vowels in the /ʊ u uː/ triplet are distin­guished by vowel duration. In normal speech and clear speech short lax and long tense vowels are furthermore distinguished by f0 fall size. In clear speech we find for both triplets that short tense and long tense vowels are dis­tinguished by f0 dynamics which was calculated by dividing the sum of the f0 rise size (pitch of f0 peak minus pitch at the beginning of the interval) and the f0 fall size (pitch of f0 peak minus pitch at the end of the interval) by the duration of the interval. In an additional perception experiment we found a strong agreement between the intended pronunciation of the triplet words and their perception with F2 serving as the best predictor.