Vincent J. van Heuven

Spectral balance as an acoustic correlate of linguistic stress

Although intensity has been reported as a reliable acoustical correlate of stress, it is generally considered a weak cue in the perception of linguistic stress. In natural speech stressed syllables are produced with more vocal effort. It is known that, if a speaker produces more vocal effort, higher frequencies increase more than lower frequencies. In this study, the effects of lexical stress on intensity are examined in the abstraction from the confounding accent variation. A production study was carried out in which ten speakers produced Dutch lexical and reiterant disyllabic minimal stress pairs spoken with and without an accent in a fixed carrier sentence. Duration, overall intensity, formant frequencies, and spectral levels in four contiguous frequency bands were measured. Results revealed that intensity differences as a function of stress are mainly located above 0.5 kHz, i.e., a change in spectral balance emphasizing higher frequencies for stressed vowels. Furthermore, we showed that the intensity differences in the higher regions are caused by an increase in physiological effort rather than by shifting formant frequencies due to stress. The potential of each acoustic correlate of stress to differentiate between initial- and final-stressed words was examined by linear discriminant analysis. Duration proved the most reliable correlate of stress. Overall intensity and vowel quality are the poorest cues. Spectral balance, however, turned out to be a reliable cue, close in strength to duration.

Spectral balance as a cue in the perception of linguistic stress

In this study, the claim that intensity, as an acoustic operationalization of loudness, is a weak cue in the perception of linguistic stress is reconsidered. This claim is based on perception experiments in which loudness was varied in a naive way: All parts of the spectrum were amplified uniformly, i.e., loudness was implemented as intensity or gain. In an earlier study it was found that if a speaker produces stressed syllables in natural speech, higher frequencies increase more than lower frequencies. Varying loudness in this way would therefore be more realistic, and should bring its true cue value to the surface. Results of a perception experiment bear out that realistic intensity level manipulations (i.e., concentrated in the higher frequency bands) provide stronger stress cues than uniformly distributed intensity differences, and are close in strength to duration differences.

Acoustic correlates of linguistic stress and accent in Dutch and American English

In the literature the same acoustic correlates of stress and accent have been established for Dutch and English, i.e. F0 movement, duration, intensity and vowel quality. A.M.C. Sluijter and V.J. Van Heuven (1996) showed that F0 movement and overall intensity in Dutch differentiate only between accented and non-accented syllables, rather than between stressed and unstressed. The most reliable acoustic correlates of stress were duration and high frequency emphasis. Vowel quality differed significantly only in lexical items, but was only a weak correlate in reiterant speech copies. We reconsider the acoustical correlates of stress and accent in American English (AE) and compare the results with the Dutch results. We offer an analysis of the discriminating strength of the parameters in an attempt to optimally distinguish initial and final stressed tokens by machine, using LDA.

Effects of Focus Distribution, Pitch Accent and Lexical Stress on the Temporal Organization of Syllables in Dutch

In recent developments in phonological theory two independent representations for prosodic prominence are needed in languages such as Dutch and English. A nonculminative auto-segmental structure with

The nature of hemispheric specialization for linguistic and emotional prosodic perception: A meta-analysis of the lesion literature

It is unclear whether there is hemispheric specialization for prosodic perception and, if so, what the nature of this hemispheric asymmetry is. Using the lesion-approach, many studies have attempted to test whether there is hemispheric specialization for emotional and linguistic prosodic perception by examining the impact of left vs. right hemispheric damage on prosodic perception task performance. However, so far no consensus has been reached. In an attempt to find a consistent pattern of lateralization for prosodic perception, a meta-analysis was performed on 38 lesion studies (including 450 left hemisphere damaged patients, 534 right hemisphere damaged patients and 491 controls) of prosodic perception. It was found that both left and right hemispheric damage compromise emotional and linguistic prosodic perception task performance. Furthermore, right hemispheric damage degraded emotional prosodic perception more than left hemispheric damage (trimmed g=-0.37, 95% CI [-0.66; -0.09], N=620 patients). It is concluded that prosodic perception is under bihemispheric control with relative specialization of the right hemisphere for emotional prosodic perception.

Phonetic Correlates of Statement versus Question Intonation in Dutch

In recent years, the formal elements of Dutch intonation have been laid down in two comprehensive models (’t Hart, Collier and Cohen, 1990;Gussenhoven & Rietveld, 1992. With these two formal models at our disposal, the stage seems set for further explorations, notably of the relationship between form and function. The present study focuses on acoustic and perceptual correlates of one major functional contrast, viz. the opposition between declarativity (statement) and interrogativity (question), two functions featuring prominently in everyday communication.

Hearing feelings: A quantitative meta-analysis on the neuroimaging literature of emotional prosody perception

With the advent of neuroimaging considerable progress has been made in uncovering the neural network involved in the perception of emotional prosody. However, the exact neuroanatomical underpinnings of the emotional prosody perception process remain unclear. Furthermore, it is unclear what the intrahemispheric basis might be of the relative right-hemispheric specialization for emotional prosody perception that has been found previously in the lesion literature. In an attempt to shed light on these issues, quantitative meta-analyses of the neuroimaging literature were performed to investigate which brain areas are robustly associated with stimulus-driven and task-dependent perception of emotional prosody. Also, lateralization analyses were performed to investigate whether statistically reliable hemispheric specialization across studies can be found in these networks. A bilateral temporofrontal network was found to be implicated in emotional prosody perception, generally supporting previously proposed models of emotional prosody perception. Right-lateralized convergence across studies was found in (early) auditory processing areas, suggesting that the right hemispheric specialization for emotional prosody perception reported previously in the lesion literature might be driven by hemispheric specialization for non-prosody-specific fundamental acoustic dimensions of the speech signal.

Speech rate as a secondary prosodic characteristic of polarity questions in three languages

Questions (almost) universally differ from statements in that the former have some element of high pitch that is absent in the latter. Therefore, the difference in speech melody (intonation) is considered to be the primary prosodic correlate of the contrast. We now pursue the possibility that an other, secondary prosodic correlate may exist that signals the difference between statement and question. We noted in Manado Malay (an Austronesian language) that questions were spoken at a faster rate than the corresponding statements. We then examined speech rate in questions and statements in two Germanic languages, viz. Orkney English and Dutch. In all three languages we find faster speaking rate in questions than in statements, but with different distribution of the phenomenon over the sentence. In Manado Malay, the difference seems restricted to the boundaries of prosodic domains, in Orkney it is evenly spread over the sentence, and in Dutch it is only found in the middle portion of the sentence. Some speculation on possible causes of the rate difference between statements and questions is offered in conclusion.

Stress, tone and discourse prominence in the Curacao dialect of Papiamentu*

This paper investigates the word-prosodic system of the Curacao dialect of Papiamentu. Curacao Papiamentu has both lexically distinctive stress and, inde- pendently, a word-level tone contrast. On the basis of a detailed acoustic investi- gation of this tonal contrast, we propose a privative phonological interpretation of the tone contrast, similar to proposals for the Scandinavian word-accent systems (Riad 1998, to appear). As compared to previous treatments of Curacao Papiamentu word prosody, our hypothesis makes crucial reference to intonation and to tonal underspecification. We also investigate the realisation of primary and secondary stress in Curacao Papiamentu.

Why stress position bias

There is ample evidence in the literature that English and Dutch listeners tend to perceive stress on the word-initial syllable. This bias is most easily seen in the perception of (nonsense) words containing repetitions of identical syllables. In four experiments the possible causes of this bias are investigated. The results show that the bias disappears when (i) words are preceded by a spoken context, when (ii) the voice source is replaced by noise (whisper), or when (iii) the fundamental frequency level of the utterance as a whole is lowered. The data are best explained by assuming that the listener interprets the onset of voicing of an isolated word as a (silent) pitch rise from the bottom of the speakers pitch range.