Source

  • Block (1990), Inverted Earth, LIT-634, read in full in NOTE-504: pp. 53–58 (the two contents, long-arm roles), 62–71 (the case, its advantages and the objections), nn. 26–29. Page numbers are the journal’s.
  • Block (2003), Mental Paint, LIT-635, read in full in NOTE-511: pp. 165, 168–171 (water emigrant, swampman) and 183–188 (Inverted Earth against representationism, long-term memory).

What was actually shown

Two contents. Block (1990), LIT-634, separates the intentional content of a colour experience, what it represents, from its qualitative content, what it is like (p. 54). Intentional content is fixed by what the experience is normally caused by and used about, so it is functional in “long-arm” roles that include real things in the world as inputs and outputs (p. 58).

The case. Inverted Earth has the complementary colour of every object and inverted colour vocabulary. A subject kidnapped there and fitted with inverting lenses notices no change. On the first day his intentional contents are still Earth’s, and wrong. “After enough time has passed on Inverted Earth, your embedding in the physical and linguistic environment of Inverted Earth would dominate”, and fifty years on he uses ‘red’ to mean green as the natives do. “Once your intentional contents have inverted, so do your functional states” (p. 64). In the interpersonal version, a twin with a molecularly identical brain, raised there with the lenses, has the same qualitative content looking at his sky and inverted intentional content and long-arm role. So, for a functional state identity theory, one qualitative state would be identical to two distinct functional states, “Q = F₁, Q = F₂ and ~(F₁ = F₂), which contravenes transitivity of identity” (pp. 64–65). The case needs no internal switch, so it escapes the doubt about testimony that sinks the intrapersonal inverted spectrum, and identical brains make sameness of qualia the supervenience default (pp. 67–68).

Against externalist representationism. Block (2003), LIT-635, restates the case against the view that phenomenal character is, or is determined by, representational content individuated externally. When a Twin Earth emigrant’s water content changes, sameness of experience is explained by represented appearances (colour, sheen, flow) that did not change, and the representationist wins that round (pp. 168–170). For colour there is no further represented appearance left over, so what stays the same on Inverted Earth must be a feature of the experience: “Mental paint is what stays the same”. Lycan’s reply, that the traveller’s experience shifts gradually without his knowing, makes ordinary phenomenal memory “inherently defective”, “an error theory” with nothing to recommend it beyond saving the view under attack (pp. 186–188). And if the traveller’s contents never shift, he is inverted relative to the natives, which defeats representationism too (p. 188).

What remains open. Block concedes that the brain-internal functional descriptions of the twins are identical, and that a “two factor” functionalism, long-arm roles for intentional content and short-arm, purely internal roles for qualia, would escape: “if there is some functional characterization of qualia, that will sink the Inverted Earth Argument”. His answer is a burden of proof, to say what those short-arm states are, the qualia in “the ‘program’ of the mind”: “Good luck to him!” (pp. 69–71). His own bet, n. 29, is that “the best bet for a theory of qualia lies in neurophysiology”, and in 2003 he calls himself “a physicalist phenomenist” (p. 182). So the account rules out two families and leaves internal ones standing.

What could have come out otherwise: the externalist could have found an internal appearance content for colour, as for water, or a reason to doubt the traveller’s memory that does not assume the conclusion. Block argues neither is available.

What this does not say

  • It does not refute functionalism about qualia as such. Only its long-arm form. Chalmers’s organisational invariance (LIT-392, LIT-406) is a principle about fine-grained internal organisation, and Block’s twins are internally identical, so the two agree that the twins share qualia (NOTE-504, Bearing). This account must not be cited against it.
  • It does not refute internalist representationism. Inverted Earth holds the brain fixed, so a view on which phenomenal content is representational content that locally supervenes on the brain is untouched by it. Block presses that view with a different case, Erisa, raised among colours that keep changing (LIT-635, pp. 166–168), which is a separate argument, offered as pressure, and not part of this claim.
  • It rests on externalism for intentional content. The shift after fifty years is an externalist intuition (Burge, Putnam), and readers who deny it lose the intrapersonal case; the interpersonal twins are Block’s answer to them (NOTE-504, Limitations).
  • It assumes qualia realism. The argument presupposes that there are intrinsic features of experience to stay fixed. Someone who denies that there are qualia at all is not refuted by it.
  • The empirical latitude is assumed. If human physiology rules out an undetectable inverter, Block’s reply is that a race very like us whose physiology does not would do (p. 64).

Rivals, and why none is declared

Dennett. “Quining Qualia” (LIT-433) argues that from the inside a subject cannot tell a change of qualia from a change of memory-linked reactions, and concludes that “there simply are no qualia at all”. LIT-634 answers it by name: Block grants “some justice” in the memory-malfunction diagnosis of the intrapersonal inverted spectrum, and builds Inverted Earth so that there is no internal disturbance for it to attach to (p. 61, n. 20). The two works rival each other, and the LITs say so. But they disagree about whether there are qualia, which this account assumes, not about what fixes them, and the record holds no theory stating Dennett’s view. So no rivals is declared here. If the record files qualia eliminativism as a theory, it rivals this account’s presupposition rather than its claim.

Metzinger and the self-model. Metzinger (LIT-568, LIT-450) holds that phenomenal content is representational content with specific functional properties, and that it locally supervenes on the brain: a brain in a vat could have it (NOTE-460, NOTE-372). That is an internalist representationism, one of the accounts this claim leaves open. THEORY-064 (the phenomenal self is a transparent self-model) is about what is experienced as a self, not about what fixes phenomenal character, and THEORY-063 (the unity of experience as one functional cluster) does not depend on representationism about character either (NOTE-511, Bearing). Neither is a rival, and neither is touched by Inverted Earth, which holds the brain fixed.

Troubles with Functionalism. LIT-407 is where the case began: Block presents Inverted Earth as a better version of an example there, inverting lenses and switched colour words in a village without standing colour beliefs (LIT-634, nn. 21–22), and its point that psychophysics would come out the same with inverted qualia is what the later case turns into an argument that needs no internal switch. It is a precursor, not a source: this account would stand without it.

Connections

  • THEORY-023 (AI consciousness cannot currently be settled). If phenomenal character were fixed by long-arm role or by externally individuated content, a system’s history of causal contact with the world would bear on what its experience is like. This account says it does not, so the dispute that account describes is between internal accounts, short-arm functional and biological, which is where Block’s 2002 and 2009 papers already place it. No relation is declared.
  • Two-factor content (THEORY-089, filed alongside). Block’s note 26 cites “Functional Role and Truth Conditions” (LIT-625) for the equivalence of long-arm roles to short-arm roles plus reference. On that analysis, what inverts on Inverted Earth is the referential factor, and the short-arm role, the twins’ shared internal factor, does not. This account asks the functionalist to do for qualia what the two-factor theory does for content, and says nobody has.

Note of 2026-10-29: what holds it back

Two premises are argued, not settled: that the traveller’s colour contents shift with long embedding, and that his phenomenal character stays the same. The record holds no externalist reply (Burge’s reply to Block is unread, and none of Harman, Tye, Lycan or Dretske is held), so neither the shift nor the case against Lycan’s gradual-shift reply has been tested against the side it attacks.