Wieke Eefting

Cross-language switching in stop consonant perception and production by Dutch speakers of English

Voiceless /p,t,k/ are implemented as aspirated stops in English, but as unaspirated stops in Dutch. We examined identification of a voice onset time (VOT) continuum ranging from /da/ to /ta/ in two language “sets” designed to induce native Dutch subjects to perceive the stimuli as if they were listening to Dutch or English. The effect of language set was highly significant, but the boundary shift was very small (2.1 ms longer in English than in Dutch) for three groups of subjects differing widely in English language proficiency. It nevertheless showed the subjects were aware of acoustic differences distinguishing Dutch and English /t/ and that the procedures were effective in creating differing language sets. Nearly every subject produced a longer mean VOT in English than Dutch /t/. The magnitude of the production shift was significantly greater for proficient than non-proficient subjects. Proficient Dutch speakers of English produced Dutch /t/ with shorter VOT values than non-proficient subjects, suggesting they formed a new category for English /t/. We speculate that the language set effect was small because subjects used their English /t/ category to identify stops in both sets. This was probably due to the fact that the synthetic stimuli, which were modeled on the English /t/-/d/ contrast, differed substantially from their Dutch /t/ category.

The effect of ‘‘information value’’ and ‘‘accentuation’’ on the duration of Dutch words, syllables, and segments

A production study was conducted to investigate the consequences of the features ‘‘information value’’ (‘‘new’’ versus ‘‘old’’ information) and ‘‘accentuation’’ (‘‘+accent’’ versus ‘‘−accent’’) on word durations. A professional speaker read aloud speech fragments in which both features were varied systematically. The results revealed that the factor information value by itself did not have an effect on word duration. Accentuation (which is closely related to information value) caused a difference in word duration of about 25%. Simultaneous variation of both factors in the conditions [new, +accent] versus [old, −accent] caused an average difference in duration of 21%. Measurements of the syllable and segment durations revealed that all segments and syllables in the words contributed to the durational changes that are caused by accentuation, which is in favor of our assumption that the word is a relevant unit of tempo.

Evidence for the Adaptive Nature of Speech on the Phrase Level and Below

This paper reports some production and perception experiments challenging the idea that the rate of articulation of interpause intervals of speech is solely conditioned by internal phonological factors. The results reported demonstrate that, at least for one professional newsreader, rate of articulation in interpause intervals is conditioned by context and that deviations from the intended rate are easily noticeable and affect perceived naturalness negatively. It is also demonstrated that a context-conditioned rate of articulation does not affect interpause intervals uniformly but is unevenly distributed. These results are interpreted in terms of speaker adaptation to the listeners running access to signal-independent information.

Relation between the features “givenness” and “accentedness” and the duration of Dutch words

The aim of this study is to investigate whether the duration of a word in speech production is related with the information value (carrying “new” or “old” information) and with the accentedness (+ or − accent) of the word. Fowler and Housum [C. A. Fowler and J. Housum, J. Mem. Lang. 26, 489–504 (1987)] found that speakers tend to shorten words conveying old information. It has also been found, by others, that accented words have a longer duration than words that do not carry accent. Because the features new/accented and old/unaccented appear to be closely related, it is unclear which of the two features causes the differences in duration as found by Fowler and Housum. In the present study, it is attempted to sort out the effects of accentedness and information value. Data will be presented showing how both factors separately and combined affect word durations.

Accentedness and givenness: Change in word, syllable, or segment duration?

In an earlier paper [W. Eefting, J. Acoust. Soc. Am. Suppl. 1 85, S97 (1989)] the results of a study concerning the relation between the features “givenness” and “accentedness” and the duration of Dutch words were presented. Production data from a highly trained, professional speaker indicated that the duration of one syllable and three syllable words is lengthened when accented. Information value had only an effect on one syllable words: One syllable words containing new information were lengthened as compared to the same words containing old information. It remained unclear whether the differences in word duration affected the temporal structure of the whole word; a local change within the word might be responsible for the differences. In the present study, the durations of the syllables and the segments within the target words are measured. Data will be presented showing that all segments in the one syllable words and all syllables in the three syllable words contribute to the differences in word duration as described above. These findings support our claim that in speech the control of word duration is an active skill and that words are the most important units of speech communication. [Work supported by the SPIN‐ASSP Program.]

Cross‐language switching in Dutch‐English bilinguals

A major issue in L2 research is whether bilinguals develop separate or merged phonetic systems. This study tested whether native speakers of Dutch, a language in which /p,t,k/ are realized with short‐lag VOT values, would label the members of a VOT continuum ranging from /da/ to /ta/ differently when listening in a Dutch as opposed to English set. High‐, mid‐, and low‐proficiency subgroups (ten each) were formed from a group of 50 Dutch university students on the basis of (a) global accent judgments by native British English listeners, (b) self‐rating of ability to pronounce English, and (c) the VOT measured in their production of English /t/. A language specific set was established by having the subject produce and process Dutch or English speech prior to, or during, each half of the experiment. The mean location of the phoneme boundaries was only slightly longer in the English (35.6 ms) than Dutch set (33.5 ms). The effect of the language set was significant, but not the effect of proficiency or set x proficiency, indicating that the extent to which L2 learners adopt different, language‐specific criteria in making phonetic judgments does not depend on language proficiency. However, the Dutch subjects showed a considerably larger shift in producing /t/ in English (60.3 ms) versus Dutch (22.8). English VOT was positively correlated with the global accent ratings, and subjects judged to have the least authentic English accent failed to distinguish /t/ in English and Dutch. [Work supported by NIH grant NS20963.]