Dynamic competitive mechanism of spatial representations in implicit and explicit numerical processing: Evidence from behavior and ERPs.

Debate has long persisted regarding whether the SNARC effect represents a stable representation in long-term memory (LTM) or a dynamic construction in working memory (WM). Adopting an integrative perspective, this study proposes that the SNARC effect is a dynamic instantiation of LTM templates within WM. By introducing the SPoARC effect as a competitive reference and utilizing event-related potentials (ERPs), we compared the mechanisms underlying implicit (parity judgment) and explicit (magnitude comparison) numerical processing. Behavioral results revealed a significant trade-off between the two effects: SPoARC was relatively stronger in the parity judgment task, whereas SNARC was relatively stronger in the magnitude comparison task. ERP data unveiled the complete neural time course: early N1 and P2 components reflected the pre-allocation of attentional resources and conflict detection, while the critical Late Negative Component (LNC) showed significantly increased amplitude in the explicit task. The LNC pattern suggests that explicit processing recruit additional cognitive resources to activate and maintain spatial-numerical templates from LTM. These findings suggest that the SNARC effect is not a fixed representation but a dynamic product regulated top-down by the central executive system: explicit tasks reinforce this association through gain control, whereas in implicit tasks, resources are prioritized for sequential information.