ISSN 0439-755X
CN 11-1911/B

心理学报 ›› 2021, Vol. 53 ›› Issue (1): 15-25.doi: 10.3724/SP.J.1041.2021.00015

• 研究报告 • 上一篇    下一篇


雷震1, 毕蓉2, 莫李澄2, 于文汶2, 张丹丹1,2()   

  1. 1西南财经大学中国行为经济与行为金融研究中心, 成都 611130
    2深圳大学心理学院, 深圳 518060
  • 收稿日期:2020-05-03 出版日期:2021-01-25 发布日期:2020-11-24
  • 通讯作者: 张丹丹
  • 基金资助:
    * 国家自然科学基金(31970980);深圳市基础研究自由探索项目(JCYJ20180305124305294);深港脑科学创新研究院(2019SHIBS0003);教育部人文社会科学研究规划基金项目支持(16XJA790004)

The brain mechanism of explicit and implicit processing of emotional prosodies: An fNIRS study

LEI Zhen1, BI Rong2, MO Licheng2, YU Wenwen2, ZHANG Dandan1,2()   

  1. 1China Center for Behavioral Economics and Finance & School of Economics, Southwestern University of Finance and Economics, Chengdu 611130, China
    2College of Psychology, Shenzhen University, Shenzhen 518060, China
  • Received:2020-05-03 Online:2021-01-25 Published:2020-11-24
  • Contact: ZHANG Dandan


准确识别言语中的情绪韵律信息对社会交往非常重要。本研究采用功能近红外成像技术, 探索外显和内隐情绪加工条件下愤怒、恐惧、快乐三种情绪韵律加工过程中的大脑皮层神经活动。结果表明, 对愤怒、恐惧、快乐韵律进行特异性加工的脑区分别为左侧额极/眶额叶、右侧缘上回、左侧额下回, 其中右侧缘上回脑区同时受到情绪和任务的调控。此外, 右侧颞中回、颞下回和颞极在情绪外显任务中的激活明显强于内隐任务。本研究的结果部分支持了情绪韵律的层次模型, 也对该模型的第三层次, 即“额区对语音情绪信息的精细加工需要外显性情绪加工任务参与”提出了质疑。

关键词: 情绪, 语音韵律, 颞上回, 眶额叶, 额下回, 缘上回


Emotional expressions of others embedded in speech prosodies are important for social interactions. Affective prosody refers to a way to express and convey emotions through the dynamic changes of various acoustic cues such as pitch, intensity, stress, and intonation in speech, without relying on vocabulary and grammatical structure. Previous studies have shown that STC, IFG, OFC, and other cerebral cortex and subcortical structures are involved in emotional prosody processing, and gradually formed a hierarchical model. However, existing studies on the neural mechanism of emotional prosody processing mostly focus on the difference between non-neutral emotional prosody and neutral prosody, while the comparison between various non-neutral emotional prosody is less investigated. Besides, the differences involved in brain regions of emotional prosody processing under explicit and implicit tasks are still not clear. Furthermore, it is necessary to further accumulate experimental evidence based on noise-free brain imaging technology, such as the noise-free features of fNIRS are especially suitable for speech processing research.
This study used functional near-infrared spectroscopy to investigate how speech prosodies of different emotional categories are processed in the cortex under different task conditions. A group of 25 college students participated in this study with a 3 (emotion: anger vs. fearful vs. happy) by 2 (task focus: explicit vs. implicit) within-participant factorial design. We manipulated task focus by adopting two different tasks, with emotional discrimination task as explicit condition and sex discrimination task as implicit condition. Ten phonological materials for each of anger, fearful, and happy prosody were selected from the Chinese Speech Emotion Database and consisted of the corresponding emotional prosodies and neutral prosodies. The emotional explicit task was to count the emotional and neutral sentences contained in each 10-second speech, and the emotional implicit task was to count the sentences played by two women in each 10-second speech. A multi-channel fNIRS system was used to record brain activity in a continuous waveform. According to existing literature, the brain regions observed in this study are the bilateral frontal and temporal lobes. Therefore, we used 13 emitters and 15 detectors to form 37 effective observation channels.
We first adopted NirSpark-2442 software to preprocess the data, and then conducted general linear model analyses to calculate the cortical activation related to the task. The results showed that the brain activation was significantly higher when anger was contrasted to fearful and happy prosody in left frontal pole / orbitofrontal cortex, and when happy was contrasted to fearful and anger prosody in left inferior frontal gyrus, and when fearful was contrasted to anger and happy prosody in right supramarginal gyrus. Importantly, there was an interaction between emotion and task. In the explicit task, cortex activity in the right supramarginal gyrus was more sensitive to fearful than to anger and happy prosodies. But no similar results were found under anger and happy prosody. In addition, the brain activation in temporopolar, superior temporal gyrus, and middle temporal gyrus with the explicit task was greater than that in the implicit task.
The present study demonstrated the specific brain regions for processing angry, fearful and happy prosody were left frontal pole/orbitofrontal cortex, right supramarginal gyrus, and left inferior frontal gyrus respectively, and the important role of right superior temporal gyrus and right supramarginal gyrus in emotional explicit task. These findings partially support the hierarchical model of emotional prosody and question the third level of the model.

Key words: emotion, phonetic prosody, superior temporal gyrus, orbitofrontal cortex, inferior frontal gyrus, supramarginal gyrus