Effects of Vowel Length and Syllable Structure on Segment Duration in Dutch

Three experiments investigated timing patterns in Dutch mono- and bisyllabic words contrasting
in vowel length.

• In Experiment 1, duration of the postvocalic stop consonant in CV(:)C words did not vary as a function of preceding vowel length.
• Experiment 2 extended this finding tointervocalic stops in bisyllabic CV(:)C~n words. In Dutch, CV:C~n words contain a long vowel in syllable-final position while CVC~n words contain a short vowel followed by an ambisyllabic consonant.

Results from Experiment 2 indicated that the duration of the intervocalic consonant is not affected by the quantity of the preceding vowel or its differential status as a tautosyllabic or ambisyllabic consonant. These results suggest no effect of vowel length on postvocalic consonant duration. However, an additional finding of Experiment 2 was that the duration of second-syllable [~n] is inversely affected by the length of the vowel in the first syllable.

Finally, Experiment 3 established that in CV(:)CC~n words, both medial consonants and [~n] were longer when preceded by a short vowel in the first syllable. These findings indicate that the presence of a short vowel results in a compensation of approximately 25-30 ms, which is distributed across all segments following that vowel.

It is hypothesized that the postvocalic consonant in Experiments 1 and 2 did not participate in this compensation because the consonant is obligatory following short vowels. Thus, the factor affecting whether or not a postvocalic consonant exhibits compensatory behavior may be not so much ambisyllabicity versus tautosyllabicity but rather its obligatory versus optional status. Implications for models of phonetic and phonological timing are discussed.