Jean Uhrmacher Institute for Clinical ENT Research, University of Cologne, Cologne, Germany.
Ear Hear. 2011 Jul-Aug;32(4):459-67. doi: 10.1097/AUD.0b013e3182064882.
Sentence stress is a vital attribute of speech since it indicates the importance of specific words within an utterance. Basic acoustic correlates of stress are syllable duration, intensity, and fundamental frequency (F0). Objectives of the study were to determine cochlear implant (CI) users' perception of the acoustic correlates and to uncover which cues are used for stress identification.
Several experiments addressed the discrimination of changes in syllable duration, intensity, and F0 as well as stress identification based on these cues. Moreover, the discrimination of combined cues and identification of stress in conversational speech was examined. Both natural utterances and artificial manipulations of the acoustic cues were used as stimuli.
Discrimination of syllable duration did not differ significantly between CI recipients and a control group of normal-hearing listeners. In contrast, CI users performed significantly worse on tasks of discrimination and stress identification based on F0 as well as on intensity. Results from these measurements were significantly correlated with the ability to identify stress in conversational speech. Discrimination performance for covarying F0 and intensity changes was more strongly correlated to identification performance than was found for discrimination of either F0 or intensity alone. Syllable duration was not related to stress identification in natural utterances.
The outcome emphasizes the importance of both F0 and intensity for CI users' identification of sentence-based stress. Both cues were used separately for stress perception, but combining the cues provided extra benefit for most of the subjects.
句子重音是言语的一个重要属性，因为它可以指示一个句子中特定单词的重要性。重音的基本声学特征包括音节时长、强度和基频（F0）。本研究的目的是确定人工耳蜗使用者对这些声学特征的感知，并揭示他们使用哪些线索来识别重音。
多项实验旨在探讨对音节时长、强度和 F0 变化的辨别能力，以及基于这些线索的重音识别能力。此外，还研究了综合线索的辨别和会话语音中重音的识别。实验使用了自然语音和声学线索的人工处理作为刺激。
与正常听力对照组相比，人工耳蜗使用者在音节时长辨别任务上的表现没有显著差异。然而，他们在基于 F0 和强度的辨别任务以及重音识别任务上的表现明显更差。这些测量结果与他们在会话语音中识别重音的能力显著相关。与单独辨别 F0 或强度相比，辨别 F0 和强度变化的表现与识别表现的相关性更强。在自然语音中，音节时长与重音识别无关。
研究结果强调了 F0 和强度对于人工耳蜗使用者识别基于句子的重音的重要性。这两个线索都被单独用于重音感知，但对于大多数受试者来说，结合使用这些线索会带来额外的好处。
