Schedule of Events | Symposia

Semantic activation hinders suppression from visual working memory

Poster Session B - Sunday, March 30, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Edward Leung¹ (edjoeleung@utexas.edu), Jarrod Lewis-Peacock¹; ¹The University of Texas at Austin

Suppressing the maintenance of an item in working memory reduces subsequent access to its representation. However, the view of working memory as activated long-term memory suggests that its representations are integrated with their context, including semantic connections to related long-term memory information. While these semantic activations can facilitate the retention and retrieval of information from working memory, they may also hinder the suppression of associated information. The current experiment tested this hypothesis. On each trial, participants (N=30) encoded four unique images (e.g. lettuce, cake) from two semantic categories (e.g. vegetable, dessert) in either a grouped (high semantic activation) or an interleaved (low semantic activation) order. They were then cued to either maintain or suppress one of the items for 3 sec, before responding to a three-alternative forced choice memory probe for one of the items (the cued item on 50% of trials). Our results replicate the maintenance suppression effect from previous studies, where reaction time was fastest for maintained items, followed by uncued items, and slowest for suppressed items. Furthermore, there was a significant interaction between semantic grouping and operation on accuracy (p=0.01). While suppression significantly reduced accuracy in the low-semantic activation condition relative to maintenance (-8.61%, p < 0.001, BF = 160), this effect was abolished in the high-semantic activation condition (p = 1, BF = 0.198). These results provide behavioral evidence that the activation of semantic knowledge for an item in working memory may impair the ability to remove that item from mind.

Topic Area: EXECUTIVE PROCESSES: Working memory