No abstract

No abstract

Accentuation results in faster recognition ofwords expressing new (focal) information. To find out whether accentuation speeds up the comprehension of words expressing given information as well, the presence or absence of accents was varied independently for these categories in three experiments. Degree of Givenness was varied across experiments. Listeners verified spoken descriptions of pictures. Accentuation was found to interact with the Givenmew variable: Given information was verified faster when the word expressing it was unaccented, New information was verified faster when the word expressing it was accented. These findings suggest that listeners do not simply give more attention to accented words, but rather process accented and unaccented words in different ways. It is hypothesized that the presence of an accent leads the listener to give primary attention to the acoustic/phonetic properties of the word and to construct an interpretation from the bottom up, and that the absence of an accent on a word leads him to map it onto the limited set of discourse entities which are currently activated, with less attention for the acoustic/phonetic properties.

No abstract

A series of experiments was conducted to determine (1) the accuracy with which vowel segment durations in spoken sentences can be represented in auditory sensory storage and (2) the extent to which phoneme boundaries in the identification of phonetalc vowel length in Dutch are affected by syntactic and/or auditory-phonetic context. In a preliminary production test it was found that durations of both long and short Dutch vowel phoneroes in monosyllabic words embedded in sentences are systematically .. affected by word positions in the sentences. In an initial perceptual experiment, phoneme boundaries and slopes of identification curves were measured for 12 listeners in five different test utterances in a binary forced choice identification test. Perceptual accuracy of vowel duration perception as determined from the slopes of the identification curves corresponds on the average to a just-noticeable difference (JND) of about 5 ms with a test segment duration of about 90 ms. Phoneme boundary values are systematically affected by context in ways predictable from syntactic structure and the auditory-phonetic environment.In a second perceptual experiment it is shown that a major effect on phoneme boundary can be brought about by perceived properties of the test utterance following the monosyllable containing the test segment. In a third perceptual experiment it is shown that the difference between phoneme boundaries in utterance final syllable and in embedded syllable is related to the presence or absence of a perceived speech pause following that syllable. The results of these experiments are interpreted in terms of a simple decision model with a noisy auditory representation of embedded vowel duration, lasting a few hundredths of milliseconds, and a noiseless internal criterion for vowel length identification which is systematically affected by the auditory-phonetic environment.