Neural Mechanisms Of Learning New English Words In Reading Contexts By Advanced EFL Learners: Interface Of Implicit And Explicit Learning

For English as a foreign language (EFL) advanced learners, enriching andbroadening their English mental lexicon in reading contexts is the mostly common wayin their daily life. Although researchers claim the significance of learning by reading,little is known about the neural mechanisms involved.By employing a combined method of behavioral priming paradigm andEvent-related Potentials technique, the present study is intended to answer twoquestions:1) How do advanced EFL learners’ neural responses (N400and LPC) changethroughout the ten presentations of new words during reading?2) After tenpresentations, would new words show semantic priming effects in both the implicit andexplicit memory testing tasks? Participants were required to complete a reading task(training) and thereafter two semantic priming tasks (testing), namely a lexical decisiontask and a recognition task. A free recall task was conducted in the end. Participants’online electroencephalogram was recorded throughout the training and testing tasks.The behavioral data including response time and accuracy were also recorded.The results of the reading task showed that while Real words elicited relativelysteady N400s throughout all ten presentations, New words elicited a larger N400thanReal words over the first four presentations and tended to be similar in N400relative toReal words afterwards. New words also elicited a more positive ERP in a later timewindow (LPC) than Real words, but this enhancing positivity remained throughout allten presentations. The response time in the lexical decision task revealed a main effectof target word condition that participants responded significantly more quickly to Realword targets than to nonword targets. More importantly, a significant relatedness effectwas also found in both Real and New word prime conditions: semantically relatedprimes facilitated responses to target words. Average accuracy in the recognition taskwas74%, whereas it was41%in the free recall task. For the ERP results of the lexicaldecision task, it was found that nonword targets elicited significantly larger N400s thanReal word targets. However, while a significant N400effect of semantic relatednesswas observed in Real prime condition in that unrelated targets elicited a larger N400than related target words, this tendency was reversed in New prime condition. In therecognition task, a larger N400was elicited by unrelated targets relative to relatedtargets in both Real and New prime conditions. This pattern of results suggests that L2vocabulary learning in reading contextsfollows gradually evolving processes indexed by the N400and LPC components. Whilelearners gradually integrate new word meanings into their memory reflected by areducing N400, the link of new word meanings to semantic networks keeps beingstrengthened as can be seen in the LPC. In addition, new words exhibit explicitrepresentations both in the lexical decision and recognition tasks, suggesting thatsemantic knowledge acquired even in implicit reading contexts has been stored inexplicit memory as tested immediately after learning, and implicit representations mayrequire a consolidation process or extended exposure.