The neural correlates of abstract rule learning: An event-related potential study

Abstract rule learning is a fundamental aspect of human cognition, and is essential for language acquisition. However, despite its importance, the neural mechanisms underlying abstract rule learning are still largely unclear. In this study, we investigated the neural correlates of abstract rule learning by recording auditory event-related potentials (ERPs). Participants were first presented with artificial three-syllable sequences containing ABA or ABB abstract rules for learning. They were then tested on sequences of novel syllables following the ABA or ABB abstract rules, half of which were inconsistent with the rule previously learned. Grand-averaged ERPs revealed significant decreases in positivity at 200-260 ms in response to consistent sequences during the earlier session of the test phase, and increased negativity at around 400 ms in response to inconsistent sequences in the later session. The potentials exhibited a left anterior-dominant distribution. The appearance of the N400-like negativity in the later session suggests that temporal ERP changes occurred with the abstract rule learning process, and that the N400-like negativity is associated with the acquisition of abstract rules. Abstract rule learning is a fundamental aspect of human cognition, comprising the ability to recognize abstract, hierarch- ical regularities from limited exemplars and generalize abstract relations to new instances that share no surface features in common with the exemplars presented initially. Abstract rule learning plays a significant role in many different types of