Cooper W E, Eady S J, Mueller P R
J Acoust Soc Am. 1985 Jun;77(6):2142-56. doi: 10.1121/1.392372.
A series of experiments was conducted to examine the influence of contrastive stress on intonational patterns of duration and fundamental voice frequency (F0). In a preliminary study, intonational patterns in spontaneous speech were compared with the patterns in oral reading of the transcribed spontaneous material. In both circumstances, a general declination of F0 peaks throughout the course of a sentence was not regularly observed, contrary to earlier findings with oral reading of isolated sentences. Two experiments were then conducted to examine the possibility that the general absence of declination observed here might be due to the influence of sentence focus, following previous results from Swedish showing that declination was not observed consistently in isolated sentences until after the site of focus. In experiment 1, we found no evidence of this restriction on declination in the oral reading of English declaratives. Declination was observed both before and after the focused item in a sentence. Furthermore, our analyses revealed that focus (as manifested by contrastive stress) is generally accompanied by an increase in duration on the focused word and by a sharp drop in F0 following the focused item. In experiment 2, these effects were replicated and extended using longer sentences. In both experiments, the influence of focus was strictly localized to the site of focus in the case of duration, but not F0. The results for both duration and F0 indicate that the influence of focus assignment on these attributes varies systematically as a function of the focus location in the sentence. Finally, the combined elongation effects for focus and utterance-final lengthening were less than would be predicted by an additive model, suggesting the operation of an expandability constraint on segmental lengthening in speech production.
进行了一系列实验，以检验对比重音对时长和基频（F0）语调模式的影响。在一项初步研究中，将自发言语中的语调模式与转录的自发材料的口头朗读模式进行了比较。在这两种情况下，与早期对孤立句子口头朗读的研究结果相反，在句子的整个过程中，F0峰值的普遍下降并未被规律地观察到。随后进行了两项实验，以检验此处观察到的普遍缺乏下降现象可能是由于句子焦点的影响这一可能性，此前瑞典的研究结果表明，在焦点位置出现之前，孤立句子中并未始终观察到下降现象。在实验1中，我们没有发现英语陈述句口头朗读中存在这种下降限制的证据。在句子中焦点项目之前和之后都观察到了下降现象。此外，我们的分析表明，焦点（以对比重音表现）通常伴随着焦点词时长的增加以及焦点项目之后F0的急剧下降。在实验2中，使用更长的句子重复并扩展了这些效应。在这两项实验中，就时长而言，焦点的影响严格局限于焦点位置，但F0并非如此。时长和F0的结果均表明，焦点分配对这些属性的影响会随着句子中焦点位置的变化而系统地变化。最后，焦点和句末延长的综合延长效应小于加法模型预测的结果，这表明在言语产生过程中，对片段延长存在扩展性限制。
