Unmasking Unconscious Working Memory using a Bias-Free Face Ensemble Paradigm

Abstract

The extent to which visual working memory (VWM) operates on unconscious input remains a matter of debate, with no definitive evidence confirming or refuting its existence. A central challenge in this inquiry lies in methodological constraints, particularly criterion biases in measuring conscious awareness. To overcome these limitations, we conducted two experiments that combined a bias-free two-interval forced-choice (2IFC) task with an ensemble perception paradigm, minimising content-related criterion confounds. Participants were briefly presented with visual ensembles of emotional faces and were assessed for both perceptual processing and WM maintenance using a delayed response task. Visual awareness and unawareness were objectively quantified using detection sensitivity derived from signal detection theory. Results showed a dissociation between awareness and perceptual sensitivity: participants could not explicitly detect task-relevant ensemble features, yet they discriminated them above chance level, even after a delay and in the presence of distractors, thereby providing evidence for the maintenance of unconscious information in VWM. To address potential detection inefficiencies in the 2IFC design, we conducted a follow-up single-interval task with an added distractor interval to disrupt VWM maintenance. Together, these results corroborate our findings demonstrating that WM can operate on information in the absence of conscious awareness. Ensemble-based paradigms thus provide a promising framework for probing the boundaries of conscious and unconscious processing in visual WM.