Kehoe M, Stoel-Gammon C, Buder E H
University of Washington, Seattle, USA.
J Speech Hear Res. 1995 Apr;38(2):338-50. doi: 10.1044/jshr.3802.338.
This study examined the acoustic correlates of stress in children's productions of familiar words. Previous research has employed experimental words rather than familiar words to examine children's phonetic marking of stress, or has not adequately controlled for phonetic environment. Subjects in this study included 22 children, aged 18-30 months, and 6 adults. Fundamental frequency, duration, and amplitude measures were extracted from stressed and unstressed syllables in two types of comparisons: one that controlled phonetic environment and syllable position (interword) and one that measured the relative effects of stress within the same word (intraword). When the tokens were analyzed on the basis of target stress pattern, results revealed no differences between adults and children in their acoustic marking of stress. Listener judgments showed that approximately 30% of children's two-syllable productions were coded unreliably or were perceived as inaccurately stressed. Overall findings indicate that children control fundamental frequency, amplitude, and duration to derive perceptually identifiable stress contrasts in the majority of their productions but they are not completely adult-like in their marking of stress.
本研究考察了儿童说出熟悉单词时重音的声学相关特征。以往的研究采用实验性单词而非熟悉单词来考察儿童重音的语音标记，或者没有充分控制语音环境。本研究的受试者包括22名年龄在18至30个月的儿童和6名成年人。在两种类型的比较中，从重读音节和非重读音节中提取基频、时长和振幅测量值：一种比较控制了语音环境和音节位置（词间），另一种比较测量了同一单词内重音的相对影响（词内）。当根据目标重音模式对样本进行分析时，结果显示成年人和儿童在重音的声学标记上没有差异。听众判断表明，儿童双音节发音中约30%的编码不可靠或被认为重音不准确。总体研究结果表明，儿童在大多数发音中通过控制基频、振幅和时长来获得可感知的重音对比，但他们在重音标记方面并不完全像成年人那样。
