Neural correlates of encoding in novel word learning

One of the central observations in the recognition memory literature is that neural processes occurring during encoding of stimuli are predictive of their later recognition and recall. That is, compared to items that are later forgotten, encoding of correctly recognised items has been associated with greater amplitude between 400ms and 800ms post stimulus onset across parietal sites (the ERP difference-due-to-memory effect), increased EEG power in the theta and gamma frequency bands and decreased EEG power in the alpha and beta bands, and increased theta-gamma phase-amplitude coupling. Importantly, theories of encoding developed based on these findings imply that these effects should be domain-general. In this pre-registered study, we tested this assumption by exploring neural correlates of successful encoding in learning of novel names for novel concepts. Following the previous studies, we used three different measures of neural activity, ERPs, time-frequency representations of power, and phase amplitude-coupling; however, we could not reject the null hypothesis for either of these measures. We provide three possible interpretations of our findings, and our main conclusion is that the existing theories of encoding may be underspecified and that properly-powered pre-registered studies are needed to further constrain these theories.