Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Multilevel Pre-activation of Lexical Features During Prediction-Driven Naming; an ERP study
Poster C6 in Poster Session C, Wednesday, October 25, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
This poster is part of the Sandbox Series.
Agnes Gao1, Tim Trammel1, Matthew Traxler1, Tamara Swaab1; 1University of California Davis
Successful lexical predictions can pre-activate the meaning and various features of the upcoming word (Gao et al., 2022). However, the underlying process of prediction is unknown. The current ERP study uses a novel self-generated predictive naming task to decipher the content and the precise timing of prediction and how it influences the processing of features of the upcoming word. Participants (N=30, in progress) read 480 sets of prime target word pairs that are semantically related (e.g., circus - CLOWN) or unrelated (e.g., trim - CLOWN). There are also 125 filler sets of word-pseudoword pairs (e.g., cartoon - CRECKED). Participants were asked to verbally name one word that came into mind immediately after seeing the prime word and before seeing the target word. Based on the relation between named words and target words, we compared the N400 across four conditions: match named related to target (e.g., seeing “circus” as the prime, naming and seeing “CLOWN” as the target), mismatch named related to target (e.g., seeing “circus” as the prime, naming “elephant”, and seeing “CLOWN” as the target), mismatch named unrelated to target (e.g., seeing “circus” as the prime, naming “elephant”, and seeing “TABLE” as the target), no response. Our hypothesis is that when the same words were named predictively, participants would have pre-activated the word and its semantic features and form-level features, making processing of the word effortless. We found that: 1) the mean N400 amplitude was the most positive when the same target word was named, suggesting a main effect of predictive naming accuracy; 2) the mean amplitude of the N400 was significant more negative when a related alternate word to the target was named (p<.001), suggesting an effect of semantic priming; 3) the mean amplitude of the N400 reduced even more for mismatch named unrelated trials and for the no response trials (p<.001). Moreover, we examined the pre-activation of semantic features (e.g., semantic concreteness), form-level feature (e.g., orthographic / phonological neighborhood size; word length). So far, we found a significant effect of semantic concreteness (i.e., abstract words eliciting less negative N400 responses than concrete words since there is less semantic information to be processed) for both match named related trials (p<.05) and mismatch named unrelated trials (p<.05). However, there was no difference in the mean amplitude of N400 between abstract vs. concrete words for the mismatch named related trials. This suggests that semantic priming facilitated lexical processing, but there is no evidence that concreteness has been pre-activated by predictively naming the same word. For the next step, we will analyze the form-level features to see whether correct prediction can pre-activate these features. If so, we would not see any N400 difference between word with large vs. small orthographic/phonological neighborhood sizes, or between long vs. short words. Finally, we plan to decode the time period after the naming response but prior to the target onset to examine the exact timing and content of pre-activation.
Topic Areas: Language Production, Meaning: Lexical Semantics