Word Recognition and Semantic Processing by Japanese English Learners

Abstract

An important aspect of developing fluent L2 reading skills is the development of lexical retrieval skills which enables words to be recognized and matched to lexical entries. Previous studies on the L2-English word processing abilities of advanced-proficiency Japanese students at an English-medium university have shown that, while these L2 learners may exhibit robust facilitation effects for words preceded by phonologically or orthographically related primes, there was no such facilitation effect for semantically related prime words, and in fact, in lower-proficiency readers, semantic priming produced delay effects. That Japanese learners of English fail to exhibit facilitation effects from semantically related primes despite high levels of L2 proficiency suggests that they do not automatically consult semantic information during word identification as would be expected. By contrast, L2 readers of English with L1s written in the same alphabetic script as English have demonstrated semantic priming facilitation in a number of studies. This discrepancy in word processing behaviors between Japanese learners of English and Western-L1-background learners of English indicates the possibility of script-specific L1 reading traits negatively impacting L2 literacy acquisition.

The following study was designed to explore how Japanese L2-English learners process semantic relationships between words and whether the delay effect found in previous research was merely a result of task interference or if it is indeed evidence of a script-specific word decoding effect. Unlike in prior studies wherein a masked priming lexical decision task paradigm was used, in this study, 32 Japanese L1 students at Akita International University undertook a semantic categorization task (again using a masked priming paradigm), thereby making the task maximally favorable to the use of semantic features as the foundation for a suitable word identification strategy. The results demonstrate that even in the context of a task designed to maximize semantic engagement, the subjects nevertheless failed to produce any facilitation effects for semantically related priming; however, they did exhibit a significant delay for semantic distractor items. This strongly suggests that Japanese learners of English are largely incapable of accessing semantic information from rapidly displayed (i.e., >50 ms) prime words and as such impervious to semantic facilitation effects in masked priming conditions.

While the significant results from the semantic distractor items demonstrate that the readers could successfully consult semantic information in the reading task, this disparity between semantic awareness in reading primes and targets seems to suggest that semantic awareness must be consciously activated and is not passively available (at least not in the under 80 ms time course which masked priming requires) as it is for most readers from alphabetic L1 backgrounds.