Perceptual pitch differences between high vs. low vowels with identical F0 have been reported in the literature (Fowler 1997, Stoll 1984). However, the reasons for this perceptual pitch shifts are controversial: Fowler (1997) claimed a compensatory mechanism for intrinsic F0 differences to avoid disturbance of prosodic parsing. Stoll (1984) argued that a psychoacoustic pitch shift accounts for the pitch differences. In German a specific phenomenon is found: Tense and lax vowels differ in their tongue height. Interestingly, tense vowels tend to have the same fundamental frequency as their lax counterparts, which is in contradiction to common explanation theories for intrinsic F0. An explanation could be provided by Hoole (2004) who found higher CT activity for lax vowels in comparison to tense vowels, speaking for an active compensation mechanism for F0.
To test whether German listeners judge an laryngeal activity or the F0 a vowel pitch comparison experiment was conducted for three different German vowels showing a similar phonological vowel height, but differing in their tense/lax distinction and roundedness. To test whether pitch perception was influenced by language inventory the same perception experiment was conducted crosslinguistically with Catalan listeners in Spain (who do not have a tense/lax distinction nor front rounded vowels).
The differences in tenseness were not significant in either language family, therefore pitch does not seem to be perceptually exploited by listeners to enhance vowel differentiation for the tense-lax distinction, at least for the conditions in the given experiment. The difference in roundedness showed significant pitch differences in both languages. This can be explained by an influence of 'sibilant pitch' (Traunmüller 1987) on the general pitch judgements of the listeners. Musical education significantly influenced sensitivity to pitch differences for German listeners, the values are in accordance to results found in literature (Rauscher and Hinton 2003). Surprisingly, Catalan listeners show an insensitivity to even large F0 differences and judge the pitch differences in vowels mainly based on the categorical vowel identity. More surprisingly, this insensitivity in pitch discrimination for Catalan listeners was not dependent on musical education.
It is not clear what mechanism causes the existing different pitch sensitivity in the two languages. The design of a new pitch discrimination experiment will be presented which will examine the underlying mechanisms and clarify the causes for the differences in pitch perception, i.e. testing the possible existence of different pitch perception mechanism for speech and complex musical tones.