Wermke Kathleen, Ruan Yufang, Feng Yun, Dobnig Daniela, Stephan Sophia, Wermke Peter, Ma Li, Chang Hongyu, Liu Youyi, Hesse Volker, Shu Hua
Center for Pre-speech Development and Developmental Disorders, University Clinics, University of Wuerzburg, Germany.
National Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China.
J Voice. 2017 Mar;31(2):255.e25-255.e30. doi: 10.1016/j.jvoice.2016.06.009. Epub 2016 Jul 7.
This study examined whether prenatal exposure to either a tonal or a nontonal maternal language affects fundamental frequency (fo) properties in neonatal crying.
This is a population prospective study.
A total of 102 neonates within the first week of life served as the participants.
Spontaneously uttered cries (N = 6480) by Chinese (tonal language group) and German neonates (nontonal group) were quantitatively analyzed. For each cry utterance, mean fo and four characteristic variation measures (fo range, fo fluctuation, pitch sigma, and pitch sigma fluctuation) were calculated, averaged for individual neonates, and compared between groups.
A multiple analysis of variance highlighted a significant multivariate effect for language group: Wilks λ = .76, F(6, 95) = 4.96, P < .0001, η = .24. Subsequent univariate analyses revealed significant group differences for fo variation measures, with values higher in the tonal language group. The mean fo did not differ between groups.
Data regarding fo variation in infant cries have been suggested as providing critical insight into the maturity of neurophysiological vocal control. Our findings, alongside with auditive perception studies, further underscore the assumption of an early shaping effect of maternal speech, particularly fo-based features, on cry features of newborns. Further studies are needed to reexamine this observation and to assess its potential diagnostic relevance.
本研究探讨产前接触有声调或无声调的母语是否会影响新生儿哭声的基频（fo）特性。
这是一项群体前瞻性研究。
共有102名出生后第一周内的新生儿作为参与者。
对中国新生儿（有声调语言组）和德国新生儿（无声调语言组）自发发出的哭声（N = 6480）进行定量分析。对于每一声啼哭，计算平均基频和四个特征变化量度（基频范围、基频波动、音高标准差和音高标准差波动），对每个新生儿求平均值，并在两组之间进行比较。
多因素方差分析突出显示了语言组的显著多变量效应：威尔克斯λ = 0.76，F(6, 95) = 4.96，P < 0.0001，η = 0.24。随后的单变量分析显示，基频变化量度存在显著的组间差异，有声调语言组的值更高。两组之间的平均基频没有差异。
有关婴儿哭声中基频变化的数据被认为能为神经生理发声控制的成熟度提供关键见解。我们的研究结果与听觉感知研究一起，进一步强调了母语语音，特别是基于基频的特征，对新生儿哭声特征具有早期塑造作用的假设。需要进一步的研究来重新审视这一观察结果，并评估其潜在的诊断相关性。
