The objection
CLAIM-412 defines interpretation without experience, so that the identity is not circular. A3 named the cost: “Blindsight illustrates that visual information can guide discrimination and behavior in the apparent absence of corresponding conscious visual experience”, and “much of perception, motor control, and predictive processing involves sophisticated information integration without clearly corresponding conscious episodes.” A1 had already pressed it: “A distributed computational system could produce an impressively coherent interpretation without there being anything it is like to be that system.”
The objection does not need a count. It does not say that every nonphenomenal definition admits unconscious processing: a definition restricted to particular architectures or operations might not, and whether most instances of such a class are unconscious would be an empirical question with no settled way to count instances (events, mechanisms, episodes). Nor does it say that blindsight is shown to be unconscious. Report and behaviour do not settle phenomenality (QUESTION-048), so blindsight is apparently unconscious processing. The obstacle is that the broad definition admits processing that is not independently established as experiential, not that every unconscious interpretation has been identified.
So the thesis has two options. It can take interpretation unrestricted and count these cases as experience, which A1 said “would need to be embraced as a substantive implication rather than allowed to slip in unnoticed”. Or it can restrict interpretation, and then it owes the restriction. A3: the challenge “doesn’t undermine the constitutive thesis, but it forces a specification of which forms of interpretive organization qualify.”
Why it stands
None of the candidates the exchange offered is yet a criterion:
- Global availability. The workspace view (A3). A3 added that it “isn’t necessarily sufficient for phenomenal experience”. The record’s workspace and no-report readings (QUESTION-048) are about access.
- Integrated causal structure. IIT (A3). The COGITATE result (LIT-973) challenged predictions of both IIT and the workspace view.
- Higher-order or recursive modelling (A3, LIT-1362), with the risk of making experience depend on reflection.
- Perspectival organization. A7’s own candidate. A7 conceded that “the existence of an operative frame does not yet prove experience. A corporation could possess such a frame.”
- Endogenous valuation (A7 §12), with the thermostat as its warning: “The presence of a regulatory objective alone cannot be enough.”
A7’s formal statement of the gap: “No necessary-and-sufficient criterion yet distinguishes experiential interpretive organization from nonexperiential information processing.” It called this “not a minor omission”.
What it does not say
- It does not say that the thesis is false. Production views face the same cases. They say which processes produce experience, as the thesis must say which ones constitute it.
- It does not say that unconscious interpretation is a different kind of process. It may differ only in organization, which is what the thesis needs.
- It does not take sides on where to draw the line, or on whether the line is sharp. A7 left a graded account open.
Note of 2026-10-22: the candidates tried since
U8–U11 of the same exchange tried four more conditions (QUESTION-254, note of 2026-10-22): reciprocal integration, interventional sufficiency, self-maintenance and cross-context answerability. None meets this claim’s defeater. Reciprocity also produces self-confirming capture (CLAIM-330). Interventional sufficiency describes a regime adequately but “would not automatically be a judgment about self-governance” (TERM-067). Self-maintenance was withdrawn as necessary (TERM-072). Answerability is untested. A8 stated the objection in this claim’s terms, as “the unconscious cognition objection”: if unconscious neural systems “instantiate all of it, we either need to accept more consciousness than ordinarily attributed or acknowledge that our criterion is insufficient. An appeal to reporting would merely recreate the access-consciousness problem.” A11’s synthesis keeps the gap open: “Which organizational properties distinguish those perspectives from nonexperiential interpretive systems remains an open theoretical question.”
Note of 2026-10-22: the readings
The works the exchange cited, read on 2026-10-22, support the objection. None meets its defeater.
- Whyte et al. (NOTE-1062) say active inference covers “both conscious and unconscious” behaviour. Their criterion is tied to report.
- Gurnee et al. (NOTE-1069) find interpretation in a language model outside its reportable workspace, and draw the blindsight parallel themselves.
- Barrett et al. (NOTE-1067) extend categorization to single-cell organisms, and leave animal experience as “whatever that means for the animals in question”.
- Representationalism concedes the same open sufficiency condition (NOTE-1068). The higher-order criterion (NOTE-1066) fails on necessity (animals) and on sufficiency, and its evidence runs through report. Graziano’s attention schema (THEORY-873) is a candidate for the missing condition, but it is representational and stated through report.
- Lyre (THEORY-880) separates neural correlates of content from a general correlate of consciousness. That concedes the gap from the structuralist side.