1. Cross-Language Neighborhood Effects in Bilinguals: An Electrophysiological Investigation Krysta Chauncey 1, Katherine J. Midgley 1,2, Jonathan Grainger 2, Walter van Heuven 3, Phillip J. Holcomb 1 Tufts University, Medford, MA 1 ; LPC-CNRS, Université de Provence 2 ; F.C. Donders Centre, Nijmegen, NL 3 Figure 6 – French Words, English Neighborhood Size Introduction In bilingual speakers, are lexical items from different languages stored jointly or separately? Holcomb, Grainger, & O'Rourke (2002) found that greater orthographic neighborhood density produced an increase in N400 amplitude. van Heuven, Dijkstra, & Grainger (1998) found that larger cross-linguistic neighborhood size slowed response times, but since this study was behaviorally conducted, the data is functionally opaque. The current investigation reproduced this effect cross-linguistically, which implies a joint or entwined storage of multiple lexica in bilinguals. Figure 7 – English Words, French Neighborhood Size Figure 3 – Trial Sequence Results The ERP data revealed a latency-shift of a possible N400 in response to L1 (French) words when compared to L2 (English) words. Also found was a within-language effect dependent on cross-language neighborhood size— larger cross-language neighborhood size produced a more pronounced negative-going waveform (N400), although the contrast was stronger for French (i.e., L1) neighborhood size during the processing of English (L2) words than for English (L2) neighborhood size during the processing of French (L1) words.

2. Methods Subjects 20 participants (13 females, average age = 23) right-handed French native speakers normal or corrected-to-normal vision Stimuli common words in English and French 74 critical items, 30 body parts blocked by language, order counterbalanced French stimuli classified by relevant neighborhood size calculated for English (>4, <2) English stimuli classified by relevant neighborhood size calculated for French Stimuli semantic categorization judgment to body parts This research was supported by NIH Grants HD25889 and HD043251, and by the CNRS (France). Figure 1 – Electrode Montage References Holcomb, PJ; Grainger, J; O'Rourke, T (2002). An electrophysiological study of the effects of orthographic neighborhood size on printed word perception. Journal of Cognitive Neuroscience, 14, 938-50. van Heuven, Walter J. B; Dijkstra, Ton; Grainger, Jonathan (1998). Orthographic neighborhood effects in bilingual word recognition. Journal of Memory & Language, 39, 458-483. Conclusions Increased N400 latency to L2 (English) words confirms the cognitive costs of L2 processing Increased amplitude to L2 (English) words when the L1 (French) neighborhood size is large (+4) indicates that words from different languages are contained in some kind of joint or entwined storage. This finding adds support to an interpretation of cross- language neighborhood effects in terms of an integrated lexical network for bilinguals. These results extend and confirm the behavioral data reported by van Heuven et al. (1998) finding increasing reaction times (and therefore processing loads) to L2 words with larger L1 neighborhood size. CAT SAT COT CAP Figure 2 – Orthographic Neighbors Figure 4 – English Words v. French Words Figure 5 – English Words, French Neighborhood Size