An investigation of behavioural and electrophysiological markers of integration in learning of novel names for novel concepts
The Complementary Learning Systems model of word learning proposes that newly learned words that have been integrated into semantic memory can interact with familiar words during lexical selection. The study reported here is the first to examine how behavioural markers of integration map onto electrophysiological markers of integration in a version of the semantic priming paradigm that is assumed to rely primarily on automatic semantic processing. 71 young healthy adults learned novel names for two sets of novel concepts, one set on each of two consecutive days. Learning was followed by a continuous primed lexical decision task with EEG measures.The behavioural data was analysed with Bayesian Linear Mixed Effects models, while, for the electrophysiological data, two types of analyses were conducted: Bayesian Distributional Regression models were used to analyse mean amplitude in two pre-defined spatiotemporal windows (N400 and LPC), whereas the Mass Univariate analysis was run to explore other time points and regions. We found evidence against priming effects in either spatiotemporal window or in the behavioural data. Nonetheless, there was evidence for differential processing of the novel names depending on the length of the consolidation period (0-day vs 1-day). We take these findings to indicate that neither the 0-day nor the 1-day consolidation period was sufficient for integration in our study and that, 24 hours after exposure to novel words, the system still relies on episodic memory to distinguish between these novel words (learned 24h ago), those learned more recently and those that haven’t been seen before.