Abstract
Attentional templates are representations of target features in Working Memory (WM) that guide visual search. While transiently active templates are as efficient as templates held in a sustained fashion, their simultaneous activation generates costs for the sustained template (Berggren, Nako, & Eimer, 2020). Based on models that conceptualize WM as a limited resource, Huynh Cong and Kerzel (2020) proposed that WM resources could serve a protective function in visual search. In a series of behavioral experiments, they demonstrated that the allocation of resources in WM could modulate the sustained template costs. Here, we investigated this question by recording electrophysiological correlates of resource allocation (CDA) and attentional selection (N2pc). Two target colors were cued before onset of the search display. In blocked trials with variable targets, both colors changed on every trial. In blocked trials with fixed targets, both colors remained identical throughout. In the mixed condition, one color was fixed while the other varied from trial to trial. Consistent with sustained template costs, RTs were delayed and N2pcs were attenuated to fixed compared to variable targets in the mixed condition. In contrast, RTs and N2pcs were similar to fixed and variable targets in the blocked conditions. Critically, we observed the exact inverse pattern on the CDA. While transient templates elicited larger CDAs than sustained templates in the blocked conditions, there was no difference between both types of attentional templates in the mixed condition. Follow-up analyses showed that the CDA elicited by the sustained template increased from the blocked to the mixed condition. Therefore, more WM resources were allocated to the sustained template when maintained concurrently with an interfering transient template. While the associated costs were still present, the increase in resources may be necessary to protect the sustained template, allowing it to conserve a precise representation in WM despite interference.