Morphology, the study of word formation, is a crucial aspect of language comprehension and production. Languages employ various strategies for word formation, including inflection, derivation, and compounding. Chinese, an isolating language, relies heavily on compounding, while English is a weakly inflectional language. Morphological processing significantly impacts word comprehension, literacy development, and second language acquisition. Previous research using electrophysiological techniques (ERPs) and neuroimaging (MEG, fMRI) has shown that morphological processing involves distinct neural networks, particularly the left frontal and temporal cortices. The left frontal gyrus (LFG) seems to be a language-general region for morphological processing, while the left temporal regions might play a more language-specific role. However, research on the neural correlates of morphological processing in bilingual brains is limited, particularly regarding the relationship between L1 and L2 processing strategies. While some studies have compared L2 processing to that of native speakers, few have directly contrasted L1 and L2 processing within bilingual individuals. This study aimed to investigate the shared and distinct neural correlates of L1 and L2 morphological processing in Chinese-English bilinguals using a visual morphological priming paradigm focusing on derivational morphology, which exists in both languages. The researchers hypothesized that the observed temporal-spatial characteristics would reflect processing strategies rather than linguistic differences.

Existing research indicates a dissociation between semantic and morphological processing in the brain. Studies using ERPs have identified a left anterior negativity (LAN) effect around 450–500 ms, associated with morphological parsing difficulty. Earlier ERP components, such as P/N250, are linked to automatic, form-based morphological decomposition. Neuroimaging studies using MEG and fMRI have implicated the left frontal and temporal cortices in morphological processing. The left frontal gyrus appears to be a language-general area for morphology, while the left temporal regions might be more language-specific. Studies on bilingual morphological processing have yielded mixed results, with some suggesting language-specific patterns and others indicating shared neural resources between L1 and L2. These discrepancies might be due to differences in language typology, participant proficiency, and the morphological structures under investigation. The unified competition model proposes that bilinguals use common cognitive resources for shared language structures, primarily employing L1 resources for L2 processing. However, the extent to which L1 and L2 processing strategies are similar or different in bilingual brains remains an open question, especially concerning shared morphological structures. Previous work with bilingual children suggested both language-general and language-specific neural resources for L1 and L2 morphological processing, with inconsistencies attributed to the different morphological structures used (Chinese compounds and English derivations) and potential proficiency differences.

Thirty native Mandarin-Chinese speakers (15 males; mean age 22.2 years) participated. All were right-handed, had normal or corrected-to-normal vision, and had started learning English as a foreign language at a mean age of 7.4 years. Their English proficiency was assessed using the LexTALE test, revealing a mean score indicating an intermediate level. The experiment used a visual morphological priming lexical decision task with a stimulus-onset asynchrony (SOA) of 200 ms between prime and target. Participants judged whether target words (60 Chinese derivational words and 60 English words with suffixes) were people-related. Control conditions used semantic priming (Chinese compound words and semantically related English word pairs). EEG and fNIRS data were simultaneously recorded. EEG data were pre-processed using EEGLAB, including re-referencing, filtering, epoch extraction, and artifact removal. fNIRS data were pre-processed using nirsLAB, including filtering, artifact removal, and hemodynamic response calculation. Statistical analyses included linear mixed-effects models for behavioral data and repeated-measures ANOVAs for fNIRS and ERP data.

Behavioral results showed a significant language effect (Chinese items processed faster and more accurately) and a significant priming type effect (morphological priming faster than semantic priming). A marginal interaction between language and priming type suggested that the morphological priming effect was primarily driven by L1 Chinese. fNIRS results revealed a neural dissociation between morphological and semantic priming effects. Morphological priming elicited greater activation in the left DLPFC, while semantic priming showed greater activation in the left STG and MTG. Pairwise comparisons showed that this DLPFC effect was driven by L2 English, and L1 Chinese showed greater activation in the frontopolar area compared to L2 English. ERP results showed an ELAN effect (80–150 ms), P250 effect (200–250 ms), and LAN effect (300–500 ms). The ELAN effect was more pronounced for L1 Chinese, while the P250 and LAN effects were present for both L1 and L2 but did not show significant cross-language differences. Analyses of difference waves (morphological priming minus semantic priming) showed marginal differences between L1 and L2 in ELAN.

The findings support a unified competition model, suggesting that bilinguals primarily employ L1 neural resources for L2 morphological processing, especially for shared morphological structures. The left prefrontal cortex appears to be a language-general region for morphological processing in bilinguals, while the left temporal cortex is involved in the differentiation between morphological and semantic processing. The stronger activation in the left prefrontal cortex for L1 Chinese compared to L2 English suggests that this region is related to native language proficiency and morphological competence. The ELAN results indicate similar early structural processing strategies for L1 and L2, differing only in degree, while shared P250 and LAN effects point to shared later structural processing. The absence of a strong morphological priming effect in English reaction times might indicate a difference in the sublexical processing strategies employed for L1 and L2.

This study provides evidence for both shared and distinct neural correlates of L1 and L2 morphological processing in bilingual adults. The left prefrontal cortex appears to be a crucial hub for bilingual morphological processing, with L1 showing stronger activation. Early structural processing (ELAN) is shared, while later processing shows both shared and language-specific components. The results support the unified competition model, suggesting a reliance on L1 resources for L2 morphological processing. Future research should include monolingual controls and manipulate language proficiency to better understand the neural correlates of language nativeness and the influence of proficiency on L1/L2 processing differences.

The study’s cross-sectional design limits inferences about developmental trajectories. The relatively homogenous sample (intermediate English proficiency) might limit the generalizability of the findings to bilinguals with different proficiency levels. The reliance on derivational morphology might not fully capture the complexities of morphological processing in both languages. The use of a visual paradigm might not generalize to auditory processing. Finally, while EEG and fNIRS data were collected simultaneously, the integration of these data was limited in the current analysis.