Abstract
INTRODUCTION: This study investigates sentence comprehension in Mandarin-English bilinguals, focusing on whether the widely reported, yet contested, subject-relative processing advantage extends to bilingual speakers. We evaluate which theoretical accounts, based on syntactic structure and canonicity, best explain cross-linguistic patterns of sentence processing. METHODS: Using a sentence-picture matching task, we examined the comprehension of canonical (e.g., actives) and non-canonical (e.g., passives) sentence structures in English and Mandarin for bilingual speakers of varying ages and Mandarin and English proficiency levels across two separate studies (n = 18 and n = 35). RESULTS: In English, bilingual participants exhibited a robust canonical sentence advantage across studies, with better comprehension of subject-relative over object-relative sentences and active over passive sentences, mirroring monolingual processing patterns. However, in Mandarin, comprehension patterns were less robust and more variable. While subject-relative and object-relative comprehension did not significantly differ at the group level, passive vs. active sentences consistently posed greater difficulty and increased performance variability across both studies, particularly among lower-performing individuals. DISCUSSION: These results suggest that sentence comprehension is shaped by language-specific constraints rather than a universal subject-relative advantage. Findings align with unified theoretical accounts that incorporate canonicity-based and structural factors, including word order, syntactic structure, and experience-, usage-, and frequency-based influences. Our results highlight the complex interplay between the aforementioned factors that differ across languages, with implications for both theoretical linguistics and clinical applications.