Presentation

LIN Zhou¹, Charles Perfetti; ¹University of Pittsburgh, ²Learning Research and Development Center, ³Center for the Neural Basis of Cognition

Exploring neurophysiological signatures of novel word learning processes: The role of semantic training and 24-hour consolidation with overnight sleep Lin Zhou and Charles Perfetti Introduction. Word learning processes, in the context of neurocognitive theories of memory (Davis and Gaskell, 2009), involve an initial rapid acquisition process in episodic memory and then a gradual integration process into the existing lexicon in neocortical long-term memory. This gradual integration process from episodic to the neocortical long-term memory can be affected by various factors, such as training methods, re-experiencing, rehearsal or sleep processes. Previous research found that after 24 hours of post-learning consolidation with overnight sleep, but not immediately after learning, newly learned words (trained with both form and meaning) induce N400 amplitudes (Bakker et al., 2015a) and power in the theta and beta bands (Bakker et al., 2015b) comparable to those of real words. This suggests that both ERP responses (e.g., N400) and oscillatory responses can be used as indicators in tracing the lexicalization process of newly learned words. The present study extends these findings by further manipulating training methods (form-only vs. form-meaning training) and examining the lexicalization process of newly learned words with different training methods and different post-learning consolidation process (24 hours vs. 0 hour) with the EEG methods. Methods. Adult participants were taught four sets of new visual words on two consecutive days (day 1 vs. day 2) with the training methods (semantic training vs. form-only training) and 24-hour consolidation with overnight sleep (day 1/remote vs. day 2/recent) factorially manipulated, together with two sets of existing real words (e.g., autumn) with their original meaning (e.g., a season lasting from September to November). On each day, half the new words were trained with both form and meaning (i.e., semantic training) and the other half with meaningless string of asterisks (i.e., form-only training). Thus, on day 2, new words learned on day 1 had experienced 24 hours post-learning consolidation with sleep, whereas those learned on day 2 had not. Immediately after learning on day 2, participants were asked to read the newly learned words, existing real words and unfamiliar nonwords while EEGs were recorded. Hypothesis and predictions. Our behavioral experiment with the same design has shown that semantic training, relative to form-only training, benefits the integration of newly learned words into existing knowledge. We hypothesize that newly learned words in the semantic training condition, compared with those in the form-only condition, are read more real-word like, which can be traced with ERPs (e.g., N400) and oscillatory responses (e.g., theta and beta power). Data with 32 participants have been collected and are being analyzed.

Topic Areas: Reading, Meaning: Lexical Semantics