Visual Selection Is Spatially Constrained During Working Memory Consolidation

Working memory consolidation (WMC) is known to consume central attentional resources, impairing performance on subsequent attention-demanding tasks. However, the specific nature of this impairment, particularly its effect on the initial stage of visual selection, remains unclear. The current study tested three competing hypotheses: No-suppression, Pure-temporal-suppression, and Spatial-suppression hypotheses. We used a sequential target-matching task where the inter-stimulus interval (ISI) between a target item and a subsequent probe item was varied to modulate the state of consolidation. Experiment 1 revealed that a spatial cue before the probe item failed to guide attention during consolidation (10-ms ISI) but did so effectively after consolidation was complete (1,000-ms ISI), thus ruling out the No-suppression hypothesis. In contrast, Experiment 2 showed that when a larger target item encompassed the cue's location, the cueing effect was restored during consolidation (0 and 150-ms ISIs), a finding that argues against the Pure-temporal-suppression hypothesis. Taken together, these findings support the Spatial-suppression hypothesis, revealing that visual selection during WMC is governed by an attentional window whose size is flexibly tied to the consolidated representation. This work clarifies the nature of the attentional bottleneck in WMC, reframing it as a spatial phenomenon.