Volume 42, Issue 2-3

This study examines subglottal air pressure (Ps) variations in Hindi stop consonants. The stops were produced in initial, medial and final positions and under different stress conditions in nonsense utterances of the type CV, VCV and VC which were embedded in a frame sentence. The results show no systematic correlation between Ps variations during the oral closure of the stops and the voicing and/or aspiration contrasts associated with them. However, aspiration and a momentary drop in Ps immediately after the release of aspirated stops seem to be systematically related. A systematic relationship between a momentary increase in Ps and production of stress is also observed. The variations in Ps associated with the production of stop consonants appear to be reflections of the changes in the glottal and supraglottal impedances. Ps variations related to normal stress appear to be controlled by the phonatory system rather than the respiratory system. On the other hand, for the implementation of heavy or emphatic stress, the phonatory and the respiratory systems appear to function synergistically.

Patterns of linguopalatal contact during the production of intervocalic /t/ in different vowel contexts were examined by means of electropalatography, in order to assess the degree of dependency between tongue tip/blade and tongue body, the relative strength of anticipatory and carryover effects and their temporal extent. Results show that the coupling between tip and body of the tongue is rather loose: constraints imposed on tongue body by consonant production are described qualitatively and quantitatively. Carryover effects are stronger and more systematic than anticipatory effects: both can extend across the consonant. Results also indicate that coarticulatory effects depend on the identity of the vowel held constant and, to a lesser extent, on stress position. Some hypotheses are presented concerning articulatory goals and strategies in /VtV/ productions.

The present paper is a continuation of a previous study, in which several cues in French stop consonant perception were investigated. In a first series of experiments, identification of a stop as fortis versus lenis turned out to be influenced by whether the following consonant was a fricative or a plosive. Taking the results of this first series as a point of departure, the next experiment focussed on the possible masking influence of friction noise on the perception of a preceding stop. This hypothesis, however, was not confirmed by the outcome of the experiment. Two further experiments showed that periodicity in a stop plus stop cluster is assigned to the first or the second stop, depending on its position in the cluster, even when the first stop is not released. The general discussion places the results within a general frame of consonant perception in French.

This paper presents a re-evaluation of the data contained in two articles by W. van Dommelen published in Phonetica 40 (1983) and 42 (1985). It is argued that greater generalization is possible for lenis/fortis perception in French C1C2 stop-obstruent clusters. First of all, they follow single stops in that the extent of voicing in the whole cluster has the strongest cue value and that the release feature of C1 is also important; its absence results in less clearly defined response functions. This means, secondly, that stop-stop and stop-fricative sequences produce the same results when the characteristic plosive aspiration features are absent in the former and merged into the fricative in the latter. Thirdly, a fortis obstruent in second position heightens lenis perception in the preceding stop by auditory contrast, not by its phonological status.

The speech signal contains many linguistic and prosodic cues which determine the perception of speech rate. We investigated which cues could optimally explain rate judgements of short Dutch and Italian utterances. For this, the correlation between an objective measure of speech rate (linguistic syllables per second, phonetic syllables per second, or phonetic segments per second) and perceived speech rate has been evaluated for rate judgements of normal, monotone, and unintelligible utterances. The effect of language background on speech rate perception was also studied, by using Dutch and Italian listeners.

Syntactic theories of intonation assume that intonation is congruent to syntax. It is, however, in the cases where it is not that intonation can restructure an utterance and become informative. In a first stage, we identify the prosodic units which play a part in the organisation of utterances. In a second stage, we examine how intonation contributes to the linear and hierarchical structure of utterances. The functioning of prosodic units in semantic and pragmatic organisation of the sentence allows us to bring to light syntactic constraints. This method seems the best way to evaluate the respective interactions of intonation with syntax and pragmatics. These syntactic constraints constitute formal marks which justify the content of prosodic rules (syntactic, semantic, accentual and rhythmic) which make it possible to predict or explain the intonation of utterances.