Recurrent processing theory
Explanation
Victor Lamme, a Dutch neuroscientist, has defended since the 2000s a theory of visual consciousness centred on recurrent processing: what distinguishes conscious from non-conscious representations is not their location (prefrontal vs. visual) but their type of processing. Conscious representations are those in which information has flowed back from higher to lower areas in a recurrent circuit.
Perceptual processing has two phases. First, an initial feedforward wave: sensory information rapidly ascends the cortical hierarchy (V1 → V2 → V4 → IT), processing increasingly complex attributes. This phase is very fast (~150 ms) and can generate representations, but these are not conscious. Second, a recurrent phase: information flows back from higher to lower areas, completing loops. This phase, according to Lamme, is what generates consciousness.
The theory rests on diverse neuroscientific evidence. Studies with magnetoencephalography (MEG) and invasive recordings show that visually conscious stimuli generate recurrent activity, while masked stimuli (non-conscious) only generate the initial feedforward wave. Patients with lesions that hamper recurrence have specific deficits of visual consciousness.
A consequence of the theory is the dissociation between phenomenal consciousness and access consciousness. Lamme holds that local recurrent processing in visual areas is sufficient for P-consciousness (there is visual experience), even if information does not reach prefrontal areas for verbal report (A-consciousness). This would support Block's idea of "phenomenal overflow".
The theory has implications for animal consciousness. If what matters is recurrent processing in sensory regions and not prefrontal activation, then many animals with sensory cortices similar to ours would have genuine visual consciousness, even if they do not have prefrontal processing as developed. This supports intuitions about extended animal consciousness.
Recurrent processing competes with rival theories (GNW, IIT, HOT) in explaining the neural correlates of consciousness. Each theory emphasizes a different aspect: GNW global availability, IIT mathematical integration, HOT meta-representation, recurrence local recurrent dynamics. Ongoing adversarial experiments attempt to discriminate among them.
Strengths
- Distinguishes phenomenal and access empirically.
- Supports no-report experiments.
- Specific neural mechanism (local recurrence).
- Coherent with IIT regarding the posterior hot zone.
Main critiques
- Difficulty studying experience without some report.
- The P-A distinction has been questioned.
- Insufficiently specific about which recurrence generates consciousness.
- Mixed evidence from adversarial collaborations.