The claim
The owner, at U1: “a necessary condition for grounded experience is sensory apparatus”. A1 accepted the direction and changed the terms. Sensory apparatus “may not be sufficient for grounding”, since a camera attached to a system does not ground it, and grounding “arguably requires an appropriately structured causal relationship between an agent and its environment, potentially including feedback, calibration, and opportunities for corrective action”. A7 added that it is not necessary either, in the interpreter’s own body: “A system might receive reports from another system’s sensors without possessing those sensors itself.”
The claim has three parts.
-
Grounding is a relation. A7 wrote it as (G(I, r, \mathcal C)): an interpretation, its purported referent, and the constraining relationships available. It is also relative to a referent. An envatted brain “may have a coherent experience of a tree that is ungrounded with respect to the tree it apparently perceives, while still being causally grounded in the apparatus that produces its signals” (A7; A2 had made the same point).
-
Its senses come apart. A3 separated current, developmental and functional grounding. A7 kept current and developmental and replaced functional with epistemic:
- current: whether present interpretations are constrained by their purported external referents;
- developmental: whether the organization that makes interpretation possible was acquired through embodied or environmental interaction;
- epistemic: whether interpretations can be evaluated or corrected using evidence “sufficiently independent of the interpretive process being assessed”.
The epistemic sense is CLAIM-342’s. A loop is kept from confirming itself by evidence “whose arrival or scoring does not pass through the interpretation it tests”. A5 drew the consequence: sensory signals ground “only insofar as their arrival, content, or scoring has some independence from the interpretive process being evaluated”. It carried over the caveat QUESTION-121 leaves open: “no sensory input arrives with an interpretation-independent meaning”.
-
Grounding and occurrence come apart. A5: “Ungrounded ≠ Unconstrained”. A dream is poorly grounded in its apparent objects but “strongly constrained by memory, emotional organization, and the dynamics of the sleeping brain” (A7). Hence A7’s pair of claims: “What makes an interpretation occur is not necessarily what makes it accurate”, and “What makes an interpretive organization experiential is not necessarily what makes its contents veridical.” A7’s narrower form, which the claim takes: “current veridical grounding need not be constitutive of the occurrence of an experiential organization at a given time”.
Where it sits in the record
- It rests on CLAIM-342 for the epistemic sense. Without it, “grounded” would mean only “caused by”, and the claim would lose the sense in which grounding can be checked.
- It complements CLAIM-412. That claim says what makes an organization experiential; this one says that grounding is a different question. Neither rests on the other.
- Block’s Inverted Earth (THEORY-095) is the record’s reading of the same split for content. Phenomenal character stays fixed while externally individuated content and long-arm role invert. Putnam’s brains in a vat (LIT-1343) is A2’s version: the constitution of an experience and “the determination of its referential content might come apart”. A7 accepted the consequence: “distinguish the internal organization of experienced content from its external semantic reference”.
- Charles Bonnet hallucinations (A3) and dreams are the natural cases. In both, weakened current constraint leaves rich experience. The record holds no reading of either.
The objection it leaves open
The developmental objection, as the exchange stated it (A3): “the independence of experience from current sensory grounding doesn’t establish its independence from sensorimotor grounding in general”. Every case the claim cites is a system that “acquired their interpretive organization through embodiment” (A7). A6 separated four independence theses: from current veridical input, from current bodily feedback, from past sensorimotor development, and from biological realization. “The available evidence supports these to very different degrees.” The claim asserts only the first. It is silent on the other three. A6 named the contrast that would sharpen the dispute: “current veridical sensory constraint” against “counterfactual sensorimotor organization”. On the brain-bound side, the exchange described Villalobos and Cea (LIT-1350) as arguing that a disembodied brain’s physiological implausibility does not defeat a contrastive, brain-centred explanation. Their paper could not be reached, and that description is unchecked.
What it does not say
- It does not say that a system with no sensory history can have experience. That is A3’s thesis C, which “does not follow” and is not claimed.
- It does not say that experience is invariant to grounding over time. A7: “Altering external constraints can reorganize interpretation over time.”
- It does not say that internal coherence is evidence of grounding. A7’s epistemic thesis says the opposite: “Internal coherence alone is not evidence of external grounding.”
Note of 2026-10-22: what the readings show about the objection
The works the exchange cited for the open objection were read on 2026-10-22. Two of them place it differently from the exchange.
- IWMT does not require current coupling (NOTE-1057, THEORY-876). In Safron et al., imagination is the same world model “decoupled from sense-data”. What IWMT requires is a capacity for action–perception coupling and a history of it, and it allows virtual embodiment. So it agrees with this claim on current grounding and parts from it only on development and capacity, the half the claim leaves open.
- Ward’s objection is organicist, not developmental (NOTE-1065, THEORY-874). The paper never discusses brains in vats, development, current input or grounding. Its enactivism holds that meaning, and so experience, depends on belonging to a living, self-maintaining system. That challenges the claim’s split differently: it ties what makes an organization experiential to what keeps it alive, not to what grounds its contents. The claim’s three senses of grounding do not answer it. That is the question of whether experience requires life, which TERM-072’s self-maintaining subclass leaves open.
- Gładziejewski (NOTE-1056) separates the grounding of a modelled structure from the grounding of its realizer, and treats dreams as the off-line exception in which constraint persists without current input. That supports “ungrounded is not unconstrained”.
- Laukkonen et al. (NOTE-1063) describe lucid dreamless sleep as awareness without current content. That is a case for the occurrence half, though it rests on report.
- Villalobos and Cea and Putnam’s “Brains in a vat” are Unavailable (LIT-1350, LIT-1343). The claim’s use of each rests on the exchange’s description.