Nov 2, 9:58am

1. What the fable actually stipulates

The REM sleep behaviour disorder (RBD) world rests on two postulates pulled in opposite directions:

  1. Phenomenal decoupling — the dream narrative is generated autonomously; its content is never input-driven at the grain that matters, so it is "not about anything external to itself."
  2. Civilizational equivalence — the dreamers nonetheless build telescopes, write journals, and develop a science in which the token "Alpha Centauri" behaves, socially and technologically, exactly as ours does.

The strongest naturalistic theory of reference begins by putting pressure on the conjunction of these two. A scientific civilisation is, above all, an error-correcting machine. For your dreamers to do astronomy, their "Alpha Centauri" tokens must stand in counterfactual-supporting dependence on that star: intervene to remove it, and — through telescopes, star charts, textbooks — their tokens change. If the dream generator is genuinely input-blind at that grain, no such channel exists, and the "astronomy" is coincidental theatre that collapses at the first novel contingency. An open-loop controller cannot do engineering in a contingent world. So either the coupling channels exist, or the world is not physically indistinguishable from ours. That dilemma, developed carefully, is the naturalist's whole answer.

2. The theory: etiological-interventionist structural externalism

Construct it in layers, each absorbing a classic objection:

Layer 1 — Informational covariance (Dretske). A state carries information about F iff tokening lawfully covaries with F. Defect: too easy. Cow-tokens caused by horses on dark nights carry information about horses-or-cows-on-dark-nights; no room for error — the disjunction problem.

Layer 2 — Etiological function (Millikan, Neander, Papineau). The token type has the function of correlating with F because past tokens that correlated with F were selected. Function fixes "normal conditions," and misrepresentation is tokening under abnormal ones. This is the only known way to get semantic normativity out of causal facts without residue.

Layer 3 — Consumer systems (Millikan again). Content is not in the producer; it is fixed by what downstream systems do with the state. The "interpretation" is another natural process, not a homunculus — and not a God. Nothing needs to be understood from nowhere for anything to be about anything.

Layer 4 — The interventionist criterion. Genuine reference is difference-making: had the referent been relevantly different, the tokening would have differed, through a channel stable across the relevant range of interventions. This is the crucial move against the fable, because it is a local, third-person, checkable test. Chance covariance — the early stage of the fable — is precisely the absence of intervention-invariant channels. You can probe it from inside the world.

Layer 5 — The structural anchor. At the level of individual objects, concede Quine's inscrutability: proxy functions show that which object "Alpha Centauri" denotes is not a further fact. But structure is determinate: the term latches onto whatever node plays that role in the causal-structural web. This is structural realism transposed into semantics — and it answers the "spooky acausal connection" worry directly. The reference relation between your "Alpha Centauri" thought and the star is not a synchronic, instantaneous link; it is a diachronic causal channel running back through starlight, telescopes, and teachers. Remove the spookiness by removing the synchrony.

Put together: a state refers to F iff (i) tokens of its type were selected for correlating with F, (ii) they are consumed by systems whose proper function is to guide action on the assumption of F, and (iii) the channel sustains intervention-invariant counterfactual dependence on F — or, at the structural level, on F's structural role.

3. Applying it: the fable self-destructs, in either direction

Now run the dilemma. If the mature dreamers are behaviourally indistinguishable from us, conditions (i)–(iii) are satisfied, and the theory's verdict is that their states refer. "It's all a dream" becomes a claim about phenomenal provenance, not about semantic status. This is not a defect of the theory — it is the externalist consequence that content never supervened on phenomenology in the first place. Meanings just ain't in the head (Putnam); they ain't in the phenomenal simulation either. The fable's insistence that dream-reference is not real reference quietly assumes an internalist criterion of genuine aboutness — exactly the "magical" criterion the naturalist must reject. The residual intuition that there is a difference is the intuition of the phenomenal intentionality school (Searle's "intrinsic" intentionality, Horgan and Tienson) — defensible, but it is a rival metaphysics, not a neutral premise.

If instead the dream generator stays input-blind at the counterfactual grain, the theory denies them content — but then the stipulated equivalence fails. Their "science" could not correct errors, their language community could not triangulate (Davidson: shared distal causes of mutual adjustments are what make communication possible between skull-bound simulators), and their civilisation would be experimentally detectable as a brittle open-loop puppet show: perturb the environment and watch their tokens fail to respond. Objective content just is differential responsiveness of the right etiological kind, so the distinguishing test is empirical, not metaphysical.

The result: the theory distinguishes the worlds exactly when they are physically distinguishable, and refuses to distinguish them when they aren't. That is the theory's strength, not its weakness. It converts "semantic nihilism" from a metaphysical possibility into an ordinary physical hypothesis — one that is jointly incoherent with physical indistinguishability, because the content-fixing facts (selected channels, consumer systems, counterfactual dependencies) are themselves physical facts. If the physics is identical, the content is identical. There is no further fact.

4. The three candidate grounds

Causal coupling
The information channel; difference-making
Alone, no — and in the fable it is stipulated absent or present by construction, so it just restates the dilemma
Evolutionary function
Normativity, error, normal conditions
No — selection acting on the whole dreamer gives the dream generator the function of tracking the world; a dream generator whose proper function is adaptive guidance is a perceptual organ
Quantum phenomenal binding
A unified subject; a single world-simulation
No — it distinguishes both our world and the RBD world from classical-substrate imitations of either, but not our world from the RBD world

Three remarks on the last two rows. First, evolutionary function is world-indifferent in a way that actively undercuts the fable's premise: selection cannot tell "genuine perception" from "functional perception," and that indifference is precisely its naturalistic credential. By teleosemantic lights, Pearce's dreamers, once their enacted dreams are shaped by selection to track fitness-relevant structure, have perception wearing dream garb.

Second, quantum-dependent phenomenal binding deserves separate treatment, because on Pearce's own terms it is the most interesting candidate. If classical machines are mere "packers" of micro-signals with no phenomenal unity, then binding is what makes a subject of experience possible at all — a single simulation rather than a crowd of micro-experiences. That is a necessary condition for phenomenal intentionality: there must be a unified "someone" whose states could be about things. But binding is not a semantic coupler. It binds the simulation internally; it does not lock the simulation to Alpha Centauri. What does the locking is the causal-teleological channel of Layer 4 — and note that the dreamers' brains bind exactly as ours do, so binding cannot separate the two worlds. However — and this is the peace treaty — if the "perfect structural match" thesis is right, and the phenomenal character of the world-simulation is the intrinsic nature of its physical causal-structural profile, then phenomenal aboutness and causal locking are two aspects of one fact. The felt directedness of experience would be the inside of the very invariance relation the naturalist identified. On Pearce's own monist grounds, the dream/perception contrast then dissolves rather than the naturalist theory failing.

5. The Liar analogy fails

"Semantic anti-realism is true" is not "This sentence is false." The Liar's truth-value flips on its own semantic status; the anti-realist thesis doesn't. Its tokens are fixed in content by the same causal-teleological web it describes — benign hermeneutic circularity, of the same kind that lets a physicalist write "all facts are physical" with a physical brain. Nor does the theory need the God's-eye view Pearce says naturalism is forced to assume. The demand for a comparison from outside all causal webs is generated by one particular picture: that reference requires an anchoring over and above causal-structural invariance. Drop that demand, and each content-fixing fact is local, third-personal, and checkable from within the web. In fact the charge inverts: it is the fable that needs the impossible perspective — the verdict "their tokens never genuinely refer" is only available to someone comparing inner glow with inner glow from nowhere. The naturalist never has to leave the web; the fable's narrator does.

6. Verdict

Could semantic nihilism be physically indistinguishable from meaningful thought? On the layered theory above, no — the hypothesis is incoherent, not merely false. The physical facts that constitute content (selected correlation, consumer architecture, intervention-invariant channels) are part of the physical inventory, so a world physically indistinguishable from ours contains them by definition. Skull-bound minds are not "condemned to talk only to themselves," because talk between private simulations is keyed to common distal nodes — publicness is causal commonality, not shared access — and each simulation is counterfactually tethered to the structure it was sculpted by.

What the fable genuinely teaches, run honestly, is this: either the dream world is not physically indistinguishable from ours (then there is no paradox), or it is semantically identical to ours (then the "solipsism" label loses its sting, and semantic access to the cosmos is exactly what causal locking can give without magic). What survives is not a paradox of naturalised meaning but a residue — the felt difference between dreaming and waking, between seeming to refer and referring. That residue is not semantic; it is the hard problem of consciousness wearing the clothes of the theory of reference. It will be resolved, if at all, where Pearce himself locates it: either phenomenal aboutness is identified with structural causal locking (the monist move, which his own metaphysics invites), or semantic content was never a phenomenal property in the first place — and the "magical" theory was the only thing that ever made the two worlds look different.

Nov 2, 9:58am

This is a shared conversation and may include unverified content that doesn't reflect Mistral AI's views. Login to resume chatting.

Vibe can make mistakes. Check answers. Learn more
Naturalized meaning and reference