Using Contextual Constraints in Bilingual Lexical Access: Evidence from Eye-Movements and Stimulated Recall

Abstract

This study combines eye-tracking measures and stimulated recall interview data to investigate how late Spanish–English bilinguals process interlingual homographs in sentence contexts that differ in semantic constraint, and how they report managing parallel activation in contexts characterized by low semantic constraint. To examine the influence of sentence-level semantic constraint on bilingual lexical processing, 19 late bilingual speakers of L1 Spanish and L2 English read sentences containing English–Spanish interlingual homographs embedded in high- and low-semantic-constraint contexts while their eye movements were recorded. A subset of participants took part in stimulated recall interviews prompted by their own eye-movement recordings from low-constraint sentences, targeting moments of increased processing difficulty. Eye-tracking data were analyzed using early and late eye-movement measures indexing initial lexical access and later integrative processing. Linear mixed-effects models were used to assess the effect of contextual constraint on reading behavior. Stimulated recall data were analyzed qualitatively using thematic content analysis to identify participants’ reported cognitive activities during ambiguity resolution. Results showed that high-semantic-constraint contexts facilitated faster and more efficient processing of interlingual homographs, whereas low-constraint contexts elicited longer reading times and increased rereading, indicating greater processing difficulty. Stimulated recalls revealed participants’ awareness of parallel activation, their use of contextual cues, and monitoring and rereading behaviors during ambiguity resolution in low-constraint contexts, complementing the eye-tracking findings. By integrating eye-tracking with stimulated recall interviews, this study links online processing patterns with readers’ reported cognitive activities, providing a more comprehensive account of how bilinguals manage cross-language ambiguity during sentence reading. The findings have implications for models of bilingual lexical processing and suggest that increased sensitivity to semantic context may support more efficient comprehension of lexically ambiguous input in L2 learning.