Prediction clearly changes what we perceive
The strongest evidence is about perception, not consciousness. In a challenging visual discrimination task, people performed better when a target's timing was predictable, either because it fell on the beat of a preceding rhythm or because it matched a previously learned interval [4]. Both kinds of temporal expectation improved accuracy compared with an irregular stream, and both produced similar alignment of occipital brain activity to the expected moment [4]. That is a real perceptual benefit from prediction, but it concerns how well a stimulus is processed, not whether the person is conscious.
Subcortical circuits add another route by which prediction can influence perception. The basal ganglia receive information from many levels of the cortex and from higher-order thalamus, placing them where they can shape sensory gating, selective attention, predictive processing, and the representation of when events occur [2]. This is evidence that predictive circuitry can influence perceptual processing and conscious states, but it describes a contributing system rather than a test of whether prediction is required for consciousness [2].
Why a perceptual effect is not yet a theory of consciousness
A perceptual improvement and a conscious experience are different things to measure. The temporal prediction study deliberately collected both objective accuracy and subjective visibility reports, a distinction the authors note has often been overlooked in work on temporal attention [4]. Even there, the finding is that prediction improves performance and shifts brain phase alignment; it does not show that prediction is what makes an experience conscious.
The theoretical analysis makes the gap explicit. Predictive processing portrays the brain as a hierarchical system that continually generates and updates predictions about the world, and it has the ambition to explain consciousness as well as perception [3]. Examining bistable perception, hallucinations, and blindsight, that analysis concludes predictive processing may offer necessary conditions for consciousness, but whether those conditions are also sufficient remains unresolved [3]. It also argues that the theory must clarify the phenomenal, felt aspect of experience before it can explain consciousness fully [3].
What would it take to show prediction is necessary or sufficient?
Necessity would mean consciousness cannot occur without prediction; sufficiency would mean prediction alone is enough. The evidence available does not isolate prediction in that way. The basal ganglia review links those circuits to perceptual processing and conscious states through an integrative relationship with cortex and thalamus, which is consistent with prediction mattering without showing it is the deciding factor [2].
Work on inattentive perception points to a different ingredient. Because the brain's energy supply cannot support continuous processing of everything in view, processing is limited mainly to change, and the authors propose that stable phenomenal content is filled in over time through integration, with qualia treated as initial conditions that cannot be derived from present data [5]. That account emphasizes temporal integration and the felt character of experience rather than prediction as the defining feature. It is a hypothesis, not a demonstration, but it illustrates why prediction shaping perception leaves the central question open.
A testable model shows how far the evidence reaches
One proposal makes prediction concrete by tying the shortest interval a person can distinguish to three consciousness parameters, with the overall threshold set by the larger of a modulated term and a fixed physical floor [1]. It uses this three-channel model to explain why acute stress barely improves real-time resolution while sharply expanding retrospective time, and it lays out sub-predictions for meditation, stress, flow, sleep deprivation, psychedelics, and body temperature, each with measurement indexes and falsification standards [1]. This is a serious attempt to make claims about consciousness testable. Even so, it addresses how conscious time experience is structured, not whether prediction is necessary or sufficient for consciousness, and its predictions remain to be confirmed.
Sources used in this answer
Prediction P16: Consciousness-State-Dependent Temporal Perception Threshold – A Three-Channel Model, the Floor Hypothesis, and Falsifiable Sub-Prediction
Proposes a three-channel model in which the minimum time-discrimination threshold depends on three consciousness parameters, with eight falsifiable sub-predictions covering meditation, stress, flow, sleep deprivation, psychedelics, and body temperature.
Contributions of Basal Ganglia Circuits to Perception, Attention, and Consciousness
Reviews anatomical, functional, and clinical evidence that basal ganglia circuits, through integrated cortex–basal ganglia–thalamus interactions, can influence perceptual processing and conscious states via sensory gating, predictive processing, attention, and temporal representation.
Between hallucination and reality: Predictive processing and the problem of consciousness
Concludes that predictive processing may provide necessary conditions for consciousness, but whether these conditions are sufficient remains unresolved, and the theory must clarify the phenomenal aspect of experience.
Unique Effect of Entrainment on Perception? Context-Specific Temporal Prediction Mechanisms in Multiple Aspects of Perception
Finds that rhythm-based and interval-based temporal expectations improve discrimination performance and subjective visibility to a similar degree, with similar occipital delta phase alignment, questioning a unique perceptual benefit of rhythmic entrainment.
Inattentive Perception, Time, and the Incomprehensibility of Consciousness
Hypothesizes that because neuronal processing is limited to change, stable phenomenal content is filled in through temporal integration, with qualia acting as initial conditions not derivable from present data.