Distinguishing between temporary and permanent removal in verbal working memory

We employed the retro-cue paradigm to examine performance improvements resulting from permanent versus temporary removal in verbal working memory. Permanent removal entails discarding a subset of working memory representations marked as definitively irrelevant, while temporary removal involves momentarily setting aside the uncued subset of representations from the attentional focus, preserving accessibility for later refocusing. We observed that both permanent and temporary removal led to marked progressive reductions in reaction time and errors across cue-target intervals (200, 400, 800, and 1,600 ms), reflecting the gradual simplification of the search set following informative cues. Although removal conditions did not differ in accuracy, responses were slower in the temporary removal condition, especially at the longest interval. A key finding was that performance in the temporary removal condition, but not in the permanent removal condition, was modulated by the presentation order of the target's memory set. This order effect was also observed in a nonremoval control condition where double retro-cues marked all presented information as relevant. We suggest that order effects depend on maintaining the integrity of the retrieval framework (all the contextual cues) needed to guide attentional access to specific representations, including those provisionally set aside in the temporary removal condition. We conclude that the primary distinction between permanent and temporary removal processes is that only permanent removal simplifies the retrieval framework by removing unnecessary contextual cues, resulting in a greater reduction in the complexity of the search set compared to temporary removal. (PsycInfo Database Record (c) 2025 APA, all rights reserved).