FLEAT VI

There have been accumulating studies in L2 lexical processing. However, many of the studies focused on orthographical/phonological aspects, leaving the structure of semantic lexical representation undercultivated. The present study approached this domain in terms of affective/emotional aspect. Despite the long history of separating emotion from cognition, recent studies suggest that they are closely interconnected (e.g. Damasio, 1994 [from neurobiology]; Opitz & Degner, 2012 [from neuropsychology]; Pavlenko, 2008 [from bilingualism]). In the concerning field, there has been an attempt to make a measurable scale of L1 lexical emotionality by assuming multiple dimensions, ‘valence (positive-negative)’ being the most influential of them (ANEW database, Bradley & Lang, 1999). The aim of this study is to make Japanese L2 word lists with valence ratings for future researches in this field, which potentially will contribute to better understanding of L2 lexical processing and acquisition.

The candidate words were selected so that each of them is listed on both of the following English word lists: (a) ANEW Database, and (b) Lexical Familiarity Database of Japanese EFL Learners (Yokokawa, 2006; 2009). As a result, 390 words were selected.

32 Japanese learners of English participated in the experiment (TOEIC score: M = 789.4, SD = 147.5). They were asked to rate the emotional valence of each target word presented on PC screen in four-point rank system.

The statistical analysis revealed that there is (a) strong positive correlation (r = .92, p < .01) between L2 valence ratings and L1 valence ratings from ANEW database, (b) weak positive correlation (r = .33, p < .01) between L2 valence ratings and visual familiarity, (c) no correlation between L2 valence ratings and reaction time but U-shaped relationship was detected between them. The rationale for these results are presented followed by prospects toward the succeeding ERP study utilizing selected words. (298 words)