Perceptual awareness as an information-distribution problem
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper |
|---|---|
| Key points | Argues that perceptual awareness is better understood as an information-distribution problem, in which stimulus information is organized across objective decisions and subjective reports. Proposes an information-theoretic framework using mutual information, conditional mutual information, and co-information to characterize these relationships at the trial level, distinguishing content-specific awareness, perceptual clarity, and confidence. Introduces Subjective Access Co-information, CoI(K;A;C), to quantify the net redundancy-synergy balance between awareness and confidence in relation to decision correctness. |
Abstract
Why do some sensory signals enter conscious experience while others influence behaviour without being consciously seen, heard, or felt? Research on perceptual awareness typically combines objective decisions with reports of awareness, visual clarity, confidence, or some subset of these measures, yet these outcomes are usually analysed separately. We argue that perceptual awareness can instead be approached as an information-distribution problem: how is stimulus information organised across objective decisions and subjective reports? We develop a theoretical information-theoretic framework based on mutual information, conditional mutual information, and co-information to characterise these relationships at the trial level. Crucially, the interpretation of these quantities depends on what the subjective report measures, allowing the framework to distinguish content-specific awareness, perceptual clarity, and confidence rather than treating them as interchangeable. The framework also specifies how informational relationships can be tracked across stimulus strength, used to formulate conscious–unconscious dissociations, and linked to metacognitive sensitivity. When awareness and confidence are both collected, we further introduce Subjective Access Co-information, CoI(K;A;C), which quantifies the net redundancy–synergy balance in how the two subjective reports jointly relate to decision correctness. By complementing established psychophysical and metacognitive measures with an explicit description of behavioural information structure, this framework provides new experimentally testable targets for theories and neuroscience of consciousness.