2007/01/24

Ladd, D. R., Mennen, I., & Schepman, A. (2000). Phonological conditioning of peak alignment in rising pitch accents in Dutch. Journal of Acoustical Society of America, 107(5), Pt 1, 2685–2696.

Presentation: Sarah
Summary: Renee

This paper reports two experiments on the alignment of rising prenuclear pitch accents in Dutch. The first experiment required subjects to read prepared sentences, in which two types of test syllables (long/short vowel) were inserted. Results from Experiment 1 showed that rises accompanying phonologically long vowels tend to be longer than those accompanying short vowels. However, there was no significant relation between phonological vowel length and the amount of F0 change. The results found in Experiment 1 consistently replicate Caspers and van Heuven’s (1994) findings, except that the latter found a strong effect of phonological vowel length. There are two explanations for this finding. (1) The durational explanation is that the rise takes a certain amount of time, and the short vowel is too short for the rise to be fully realized on the vowel itself. (2) The structural explanation is that tonal targets are anchored to specific places in structure. The accent peak may be anchored to the end of the syllable, which is equivalent to the end of the vowel if the vowel is long, but is during the consonant if the vowel is short. The second experiment thus aimed to decide between these two explanations. The general approach in Experiment 2 was similar as that in Experiment 1, except that the vowels of the test syllables were restricted to [i:] and [I]. Results of Experiment 2 showed that for both long and short vowels the F0 peak was aligned after the offset of the vowel, as expected by the durational explanation. However, as expected by the structural explanation, the peak was aligned earlier following [i:] than following [I]. The data of Experiment 2 thus partially supports both accounts.