Slide Slam E4
Behavioural and electrophysiological markers of integration in novel word learning
Maria Korochkina1,2,3, Lyndsey Nickels1, Audrey Bürki2; 1Macquarie University, 2University of Potsdam, 3International Doctorate for Experimental Approaches to Language and Brain (IDEALAB): Universities of Groningen (The Netherlands), Newcastle (United Kingdom), Potsdam (Germany) & Macquarie University (Australia)
According to the Complementary Learning Systems model, learning involves the formation of new episodic memory representations and their later integration into semantic memory (McClelland et al., 1995; McClelland, 2013). Applied to word learning, this model predicts that only integrated newly learned words can compete with familiar words during lexical selection in tasks that require activation of multiple components of the lexical system (Davis & Gaskell, 2009; McMurray et al., 2016). Semantic priming is one such task, in which participants are presented with pairs of words – primes and targets – and asked to make a judgement to the target (e.g., McNamara, 2005). When the prime is semantically related as opposed to unrelated to the target (e.g., bed-sofa vs. hat-sofa), participants make judgements faster and the amplitude of the N400 ERP component is reduced, while that of the Late Positive Component (LPC) is enhanced. Studies tracking semantic integration of new vocabulary have frequently (although not always) reported the behavioural priming effect, yet it is unclear whether it indexes integration as it could be subserved by either semantic or episodic memory system. This issue can be addressed by examining whether the behavioural effect co-occurs with the modulation of both ERP components or LPC only (e.g., Bakker et al., 2015; Batterink & Neville, 2011; Liu & van Hell, 2020). This is because the N400 component is believed to reflect automatic processes of lexical-semantic retrieval (e.g., Kutas & Federmeier, 2011), while the LPC has been linked to episodic memory retrieval (e.g., Rugg & Curran, 2007). This study used a version of the semantic priming paradigm that taps into automatic semantic processing to investigate this issue. 72 monolingual native speakers of Australian English learned novel names for two sets of novel concepts (20 words per set), one set on each of two consecutive days. EEG was recorded on Day 2 when learning was followed by a primed continuous lexical decision task, with newly trained words as targets, and familiar (semantically related or unrelated) words as primes. Recall of the novel words was also tested. For the EEG data, two types of analyses have been pre-registered: the amplitude averaging approach and the mass univariate analysis (e.g., Groppe, Urbach, & Kutas, 2011a, 2011b). The former approach is used as it is most common in the literature, while the latter will explore other time windows and ROIs for the contrasts of interest. Averaged amplitude and behavioural data will be analysed with linear mixed effects models. Data collection for this study was completed in the second half of May 2021. Data analysis is expected to be completed in July 2021. This study is the first to examine electrophysiology of behavioural effects commonly interpreted as markers of integration with a paradigm that taps into automatic semantic processing. We hypothesise that response speed as well as the N400 and LPC amplitude in response to the newly trained words will be modulated by prime-target relationship and time after exposure (24h vs. 0h), indexing differences in integration.