Cognitive functions, for example speech processing, are distributed asymmetrically in the two hemispheres that mostly have homologous anatomical structures. Dichotic listening is a well-established paradigm to investigate hemispherical lateralization of speech. However, the mixed results of dichotic listening, especially when using tonal languages as stimuli, complicates the investigation of functional lateralization. We hypothesized that the inconsistent results in dichotic listening are due to an interaction in processing a mixture of acoustic and linguistic attributes that are differentially processed over the two hemispheres. In this study, a within-subject dichotic listening paradigm was designed, in which different levels of speech and linguistic information was incrementally included in different conditions that required the same tone identification task. A left ear advantage (LEA), in contrast with the commonly found right ear advantage (REA) in dichotic listening, was observed in the hummed tones condition, where only the slow frequency modulation of tones was included. However, when phonemic and lexical information was added in simple vowel tone conditions, the LEA became unstable. Furthermore, ear preference became balanced when phonological and lexical-semantic attributes were included in the consonant-vowel (CV), pseudo-word, and word conditions. Compared with the existing REA results that use complex vowel word tones, a complete pattern emerged gradually shifting from LEA to REA. These results support the hypothesis that an acoustic analysis of suprasegmental information of tones is preferably processed in the right hemisphere, but is influenced by phonological and lexical semantic processes residing in the left hemisphere. The ear preference in dichotic listening depends on the levels of speech and linguistic analysis and preferentially lateralizes across the different hemispheres. That is, the manifestation of functional lateralization depends on the integration of information across the two hemispheres.

中文翻译：

---

语音的二分音听的横向化受语音内容的影响。

认知功能（例如语音处理）不对称地分布在两个具有相同解剖结构的半球中。二重听是一种成熟的范例，用于研究语音的半球偏侧化。然而，分叉听的混合结果，尤其是当使用音调语言作为刺激时，使功能偏侧化的研究变得复杂。我们假设在二分之一听音中出现不一致的结果是由于在处理在两个半球上进行差异处理的声学和语言属性混合时的交互作用。在这项研究中，设计了一个主题内的二项式听觉范例，其中，在需要相同音调识别任务的不同条件下，逐渐增加了不同级别的语音和语言信息。在嗡嗡声的情况下观察到左耳优势（LEA），与在二项听觉中通常发现的右耳优势（REA）相反，在这种情况下，仅包括音调的慢频率调制。但是，当在简单的元音音调条件下添加语音和词汇信息时，LEA变得不稳定。此外，当语音和词汇语义属性包含在辅音元音（CV），伪单词和单词条件中时，耳朵的偏好变得平衡。与现有的使用复杂元音单词的REA结果相比，从LEA到REA逐渐出现了完整的模式。这些结果支持这样的假说：音调的超节段信息的声学分析最好在右半球中进行，但受驻留在左半球的语音和词汇语义过程的影响。双耳听力的耳朵偏爱取决于语音和语言分析的水平，并且优先偏向于不同半球。也就是说，功能侧向化的表现取决于两个半球之间信息的整合。双耳听力的耳朵偏爱取决于语音和语言分析的水平，并且优先偏向于不同半球。也就是说，功能侧向化的表现取决于两个半球之间信息的整合。双耳听力的耳朵偏爱取决于语音和语言分析的水平，并且优先偏向于不同半球。也就是说，功能侧向化的表现取决于两个半球之间信息的整合。