Source
Seth & Bayne (2022), LIT-982, the section “Evaluating theories of consciousness” and Box 3 (“The measurement problem”), as read in NOTE-849.
What was actually shown
An argument, in two steps.
The report channel. IIT and reentry theories predict that posterior cortical activity can support experience “without contribution from anterior areas”. But “we lack reliable methods to verify such claims, since verification relies on subjective report (or at least, executive control in some form), which in turn requires anterior cortical activity”. The evidence channel carries the very activity whose necessity is in dispute. The review locates the front-versus-back debate here: “at its heart is a disagreement about the relationship between consciousness and cognitive access”.
The bypass. No-report markers (PCI, optokinetic nystagmus, bifurcation dynamics) avoid overt report, but “any proposed indicator … must be validated”. Validation by introspection inherits its unreliability and is unavailable for infants, patients and animals; validation by theory “seems unlikely” to give intersubjective validation “because theories of consciousness are themselves contentious”. Of the proposals for a third route, “none is uncontroversial”.
What could have failed: the argument would fail if some validated measure of conscious content existed that used neither route; the review surveys the candidates and finds none accepted.
What this does not say
- Not that posterior theories are wrong or untestable for ever. The limit is “at present”; the review expects methodological advances (for example, optogenetic tests of IIT’s inactive-neuron prediction).
- Not that no-report studies show nothing. They show diminished prefrontal engagement when report is removed; the limit is on what that can establish about experience, which is THEORY-094’s point from another side.
- Not a proof. It is a two-author review’s argument; it has not been tested against a reply.
- Not specific to cortex in its second half. The validation problem applies to any system, including animals, organoids and AI.
Connections
- QUESTION-048 asks what report, no-report, deficit and architectural evidence can establish about experience as opposed to access. This answers part of it: report evidence cannot separate the two on the front-versus-back question, because report brings in the anterior activity in dispute, and non-report markers cannot yet be validated as tracking experience. It adds nothing on deficits.
- THEORY-094 says no-report designs still carry post-perceptual cognition that tracks the percept. That is a limit on the no-report channel; this is a limit on the report channel and on validating bypasses.
- THEORY-023 makes the same evidential point for AI. Box 3 states it for any creature without report.
- COGITATE (THEORY-245) is the adversarial collaboration the review points to; its own concession that it may track stimulus processing is an instance of the validation problem.