加州大学Edward F. Chang团队的最新研究提出了人类颞叶中共享的和特定语言的语音处理。2025年11月19日出版的《自然》杂志发表了这项成果。

在这项研究中，该研究团队记录了不同语言背景的成年参与者在听母语和不熟悉的外语演讲时的高密度皮层活动。研究小组发现，无论语言经验如何，母语和外语都能在颞上回（STG）中引发相似的皮质反应，这些反应与基本语音特征（如元音和辅音）的共同声学处理有关。

然而，只有在母语听力过程中，小组才观察到STG中对单词边界、词频和语言特定音序列统计的神经编码增强。在一个独立的双语参与者队列中，单词和序列级别的信息编码出现在同一个体和相同的STG神经群中。这些结果表明，经验依赖的语言加工涉及STG中共享的共音和特定语言的序列和词级信息的动态整合。

据介绍，所有的口语都是由人类声道产生的，声道限定了有限的可能的语音。然而，尽管存在这种限制，世界上的7000种口语中存在着令人难以置信的多样性，每一种语言都是通过在特定语言环境中广泛的听力经验来学习的。目前还不清楚大脑中哪些语言处理因素依赖于日常语言经验，哪些不依赖。

附：英文原文

Title: Shared and language-specific phonological processing in the human temporal lobe

Author: Bhaya-Grossman, Ilina, Leonard, Matthew K., Zhang, Yizhen, Gwilliams, Laura, Johnson, Keith, Lu, Junfeng, Chang, Edward F.

Issue&Volume: 2025-11-19

Abstract: All spoken languages are produced by the human vocal tract, which defines the limited set of possible speech sounds. Despite this constraint, however, there exists incredible diversity in the world’s 7,000 spoken languages, each of which is learned through extensive experience hearing speech in language-specific contexts1. It remains unknown which elements of speech processing in the brain depend on daily language experience and which do not. In this study, we recorded high-density cortical activity from adult participants with diverse language backgrounds as they listened to speech in their native language and an unfamiliar foreign language. We found that, regardless of language experience, both native and foreign languages elicited similar cortical responses in the superior temporal gyrus (STG), associated with shared acoustic–phonetic processing of foundational speech sound features2,3, such as vowels and consonants. However, only during native language listening did we observe enhanced neural encoding in the STG for word boundaries, word frequency and language-specific sound sequence statistics. In a separate cohort of bilingual participants, this encoding of word- and sequence-level information appeared for both familiar languages in the same individual and in the same STG neural populations. These results indicate that experience-dependent language processing involves dynamic integration of both shared acoustic–phonetic and language-specific sequence- and word-level information in the STG.

DOI: 10.1038/s41586-025-09748-8

Source: https://www.nature.com/articles/s41586-025-09748-8