Contrasting L1 profiles through early orthographic and phonological processing of written English

Reading is fundamental to everyday life and can appear automatic and effortless, but this seemingly simple skill involves complex interactions between neural networks within ~200ms of seeing a word. While a wealth of work has investigated when and how orthography and phonology interact during visual word recognition (VWR), much is still unclear about such early native language (L1) processing and significantly more so when considering reading a second language (L2). To investigate this, monolingual English participants (native, alphabetic L1), late bilingual Spanish-English participants (non-native, alphabetic L1), and late bilingual Chinese-English participants (non-native, non-alphabetic L1) read English words, pseudohomophones, and pseudowords in orthographic and phonological lexical decision tasks and real English words in an orthogonal rhyme recognition task. Behavioural performance was evaluated alongside event-related potential (ERP) analysis of occipital, occipitotemporal, and frontal-central activity in the ~100ms post-stimulus timeframe (i.e., P1-O, P1-OT, and N100-FC components, respectively) as well as the ~170ms timeframe at occipitotemporal sites (N170-OT) following documented associations with early orthographic and/or phonological processing. Crucially, interpretations of analyses took the markedly different language profiles into account to examine how L1 might influence L2 processing. Patterns of behaviour and electrophysiology showed similarities but distinguished between groups based on language profiles. Analysis highlighted evidence for early orthography-phonology mapping at occipitotemporal sites, parallel orthographic/phonological processing at frontal-central sites at ~100ms, and a lack of VWR-related occipitotemporal N170 effects across groups. Findings, overall, suggest that language profiles underlie early orthographic and phonological processing and that the three groups have distinct neural strategies for VWR linked to orthographic and phonological aspects of their language profiles. Ultimately, the outcomes support the importance of L1 properties and L1-L2 relationships (i.e., language profiles) in bilingual VWR and the increased consideration of these factors in future development of monolingual and bilingual theories and models of VWR.