Musical Expertise Reshapes Cross-Domain Semantic Integration: ERP Evidence from Language and Music Processing

Background/Objectives: Both language and music are capable of encoding and communicating semantic concepts, suggesting a potential overlap in neurocognitive mechanisms. Moreover, music training not only enhances domain-specific musical processing but also facilitates cross-domain language processing. However, existing research has predominantly focused on Indo-European languages, with limited evidence from paratactic languages such as Mandarin Chinese. In addition, the impact of variations in musical expertise on these shared processing mechanisms remains unclear, leaving a critical gap in our understanding of the shared neural bases for semantic processing in language and music. This event-related potential (ERP) study investigated whether Chinese sentences and musical chord sequences share semantic processing mechanisms and how musical expertise modulates these mechanisms. Methods: This study recruited 46 college students (22 musicians and 24 non-musicians). Participants read Chinese sentences presented word-by-word visually, while chord sequences were delivered auditorily, with each word temporally aligned to one chord. Sentences included semantically acceptable or unacceptable classifier-noun pairs and chord sequences ended with in-key or out-of-key chords. Participants were instructed to focus on reading sentences while ignoring the concurrent music. ERP signals were recorded, and time-locked to final words to capture neural dynamics during semantic integration. Results: The behavioral results showed that musicians were influenced by musical regularity when reading (acceptable: F(1, 44) = 25.70, p < 0.001, η² = 0.38; unacceptable: F(1, 44) = 11.45, p = 0.002, η² = 0.21), but such effect was absent in non-musicians (ps > 0.05). ERP results showed that musical semantic processing had a substantial impact on both P200 (F(1, 44) = 9.95, p = 0.003, η² = 0.18), N400 (musicians: F(1, 44) = 15.80, p < 0.001, η² = 0.26; non-musicians: F(1, 44) = 4.34, p = 0.043, η² = 0.09), and P600 (musicians: F(1, 44) = 5.55, p = 0.023, η² = 0.11; non-musicians: F(1, 44) = 8.68, p = 0.005, η² = 0.17) components. Furthermore, musical expertise exerted modulatory effects during later stages, as evidenced by divergent N400 and P600 latency patterns between musicians and non-musicians. Specifically, ERP amplitudes exhibited opposing trends: musicians showed an enhanced N400 and diminished P600, while non-musicians displayed a weaker N400 and stronger P600. Conclusions: Our findings provide novel evidence that Mandarin Chinese and chord sequences engage partially overlapping neural mechanisms for semantic processing both in the early (P200) and the late (N400 and P600) stages. Crucially, this study is the first to demonstrate that musical expertise may gradually reorganize these shared mechanisms, enabling two initially independent but functionally analogous semantic mechanisms into a domain-general processing system. These insights deepen our understanding of the neurocognitive mechanisms underlying linguistic and musical semantic processing and highlight how expertise shapes the neural architecture of cross-domain mechanisms.