Mind, Matter, and Meaning

Chapter 4: The Argument from Illusion

The argument that drove careful people behind the veil — met at full strength, before we earn the right to take it apart.

What this chapter does. The previous chapter called the inner theater an artifact — built, furnished, inherited, and propped up by habit. But an artifact can still have a load-bearing wall, and the inner theater has one: an argument old enough to predate Descartes and good enough to have kept honest philosophers loyal to the veil long after its historical supports were knocked away. Here you meet that argument at full strength — and, deliberately, you do not yet get to watch it fall. The clean answer turns on a distinction a few chapters off, and reaching for it now would rob you of seeing it land in your own case. So the harder, more honest thing comes first: the best case for the veil, stated as powerfully as its defenders ever made it, left standing just long enough for you to feel its grip.

It builds in stages:

  1. The argument from illusion, at full strength (§I). The stick looks bent in water and isn’t; so what you were directly aware of, the reasoning runs, can’t have been the stick — it must have been something inner that really was bent-looking. Generalized, this makes every perception an encounter with an inner item.
  2. The argument from hallucination, which is worse (§II). When you hallucinate, there’s no external object at all, yet something seems present to awareness. Since a hallucination can be indistinguishable from the real thing, the same inner item must be present in both — so even veridical perception delivers the inner, not the world.
  3. The two pressures compound (§II). Science says the world can’t reach consciousness directly (too much processing intervenes); phenomenology says the same (the inner character of experience is consistent with no world at all). The wall looks solid from both sides.
  4. There are escape routes — but they aren’t yet decisive (§III). Austin thought the whole thing ran on a grammatical trick; the disjunctivists deny its key premise. Both are right to be suspicious. Neither becomes a knockout until we have a distinction this chapter can only point at.
  5. What the argument would cost us (§IV). A hallucinated clock that happens to read the right time shows what’s at stake: if the veil is real, veridical perception is just lucky hallucination — true by coincidence, never genuinely in touch with the world. That price is the clue to where the answer has to come from.

For now, the wall.

§I. The Argument from Illusion

No argument entrenches the inner-theater picture more powerfully than the argument from illusion. It dates to antiquity — versions appear in the ancient skeptics — and has never quite gone away, because at first encounter it looks airtight.

It goes like this. In cases of perceptual illusion, experience presents the world other than it is. A straight stick, half-submerged, looks bent. Now ask the awkward question: when the stick looks bent, what are you directly aware of? Not a bent stick in the world — the stick is straight; you can pull it out and check. So there is something bent that you are aware of, and it isn’t the stick. Call it what you like — a sense-datum, an appearance, an inner image. It is bent, you are aware of it, and it is not the physical stick. Therefore the immediate object of your awareness, in this case, is something inner.

So far this only concerns the case that goes wrong. The argument aims higher: to generalize. A veridical perception of a straight stick and an illusory perception of a “bent” one can shade into each other continuously — lift the stick slowly from the water and watch the bend ease out. There is no point at which the kind of thing you are aware of seems to switch from an inner item to an outer one. Consistency, the argument says, demands that we treat the two cases alike: if the bent-looking case involves awareness of an inner item, so does the straight case. The conclusion lands hard. In every perception, veridical or not, the immediate object of awareness is internal. The world always stands at one remove, behind the representation, on the far side of the glass.

This is the engine. Nearly every version of the inner theater, from Locke’s ideas to the twentieth century’s sense-data, runs on it.1

§II. The Argument from Hallucination

A sharper version of the same argument presses from the limit case, where the world drops out entirely.

In illusion there is at least a real stick being misperceived. In hallucination there is nothing. A patient with Charles Bonnet syndrome sees detailed figures — children, animals, faces — that have no counterpart in the room. Someone given a large enough dose of a serotonergic compound may watch the walls breathe and the colors crawl. More mundanely, a recently widowed person sees the late spouse in the kitchen doorway, briefly and clearly, with no figure in the room to do the work. In each case the experience is, at the moment of having it, subjectively indistinguishable from a genuine perception. The hallucinating patient does not feel himself generating an inner image. He feels himself seeing.

The argument from hallucination presses one simple question. If a hallucination and a genuine perception can be, from the inside, entirely indistinguishable — if the felt character of the two matches exactly — then what you are directly aware of in both cases must be the same kind of thing. And since in the hallucination there is plainly no external object to be aware of, what you are aware of in the genuine case cannot be the external object either. It, too, must be something inner.2

This is harder to dodge than the argument from illusion, and for a precise reason. You cannot answer it by pointing to the real object that is merely being misrepresented, because in hallucination there is no real object to point to. The inner item remains the only candidate for what awareness lands on — and once it is the object in the hallucinating case, the indistinguishability drags it into the veridical case alongside.

Now watch the two pressures compound. Trace the causal story and the world cannot reach consciousness directly — too much intervening processing stands between the stimulus and the experience, exactly as the previous chapter described. Examine the phenomenology and the world cannot reach consciousness directly — the inner character of experience squares with the total absence of an outer world. Science presses from one side, phenomenology from the other, and the wall seems solid. A great many careful people have concluded it was.

§III. Where the Argument Can Be Pressed — But Not Yet Beaten

It would be dishonest to pretend no one has fought back. The argument has been resisted, and resisted by formidable people. The trouble is that the resistance, stated here and now, does not yet reach all the way to a verdict — and I would rather you saw why than be handed a refutation you have no reason to trust.

J. L. Austin — the wartime intelligence officer who came back to Oxford and turned ordinary-language philosophy into something close to a precision instrument — thought the whole argument ran on a grammatical sleight of hand.3 The slide, he noticed, happens between two sentences that sound nearly identical in English: the stick looks bent, and I am aware of something bent. The first is a remark about how a perception portrays its object. The second is an existential claim about an inner thing that has the property of being bent. They are not the same kind of statement, and treating the first as if it licensed the second is what conjures the inner item into being. That is a real and deep complaint. But notice what it would take to make it stick: an account of how a perception can “portray its object as bent” without there being any bent thing — inner or outer — doing the portraying. Austin gestures at it. He does not, in a way that compels, deliver it.

The disjunctivists press a different seam.4 The argument from hallucination assumes that because a hallucination and a perception are indistinguishable from the inside, they must be the same kind of mental state. Deny that, and the inference never starts. Perhaps genuine perception is one kind of thing — a relation in which the world itself figures — and hallucination is another kind entirely, and the two merely happen to be hard to tell apart, the way a convincing forgery is hard to tell from the original without being the same kind of object. This is an elegant move. But on its face it looks like changing the subject: the whole force of the argument was that indistinguishable states should be treated alike, and the disjunctivist’s reply is, essentially, no they shouldn’t — which needs an independent account of what makes perception the world-involving kind of state it’s claimed to be.

Both lines are on the trail of something true, and neither is yet a defeat. What they are missing — what we are missing, standing here — is a single distinction that would let us say, without hand-waving, how an experience can be wholly directed at the world, can get the world wrong, and can still never contain a private bent-stick for us to bump into. That distinction exists. Naming it, and turning it into the key that springs this entire lock, falls to a chapter still ahead of us — The Seeing and the Seen — and there is a reason not to rush it. The argument from illusion does not just need to be answered; it needs to be answered in a way you see in your own case, or the answer won’t hold against three centuries of pull. Seeing it requires looking first.

§IV. What the Argument Would Cost Us

Before we leave the wall standing, be clear-eyed about what that costs — because the cost is the clue.

Borrow a small case from epistemology. A man hallucinates a clock on the wall of his kitchen. There is no clock there; the space is empty. But the hallucination is vivid and orderly, and at the moment he looks, the hallucinated clock reads 3:42. As it happens, the actual time is 3:42. He forms the belief that it is 3:42. The belief is true. He feels every bit as certain of it as he would with a real clock in front of him. Does he know what time it is?

The answer is no, and the reason matters.5 His belief holds true, but the truth comes by accident. The hallucinated clock responds to nothing outside his head. Move the real clock, stop it, set it to noon — the hallucinated clock ticks on along whatever schedule the hallucination dictates. The match between his belief and the world holds, this moment, by coincidence, and would dissolve in any nearby case. He has the right answer for none of the right reasons. From the inside, he cannot tell this state from the one he would be in if a real clock hung on the wall and his eyes were doing what eyes do. From the inside, knowing and luckily-believing-truly-by-hallucination look indistinguishable. From the outside, they could not be more different. One state involves the world. The other does not.

Here is the part the inner theater cannot absorb. If what you are immediately aware of in perception is always an inner item — something playing on the screen behind your eyes — then perception just is the hallucinated clock, every time. It would track the world only when the inner show happened to align with it. Veridical perception would be lucky hallucination, true by coincidence, never genuinely in touch with anything. The world would enter, at most, as the distant cause of a private display that might or might not match it.

That consequence ought to feel intolerable, and the fact that it does is not nothing. It marks the first faint pull of the answer. Whatever perception turns out to be, it does not feel like a lucky inner match, and the suspicion that it isn’t one — that seeing, when it works, involves the world rather than merely correlating with an inner stand-in for it — is exactly the thread the next chapters pick up.

§V. Standing at the Wall

So here we are, stuck in an uncomfortable place on purpose.

The inner theater is an artifact — the previous chapter showed that. But it has a genuine argument, not just a history, and the argument runs better than its reputation suggests. State it at full strength and it presses from two directions at once, science and phenomenology, and the obvious replies turn out to need something they don’t yet have. The honest report, at this stage of the book: we have not refuted the case for the veil. We have only located it — we can see where it must be vulnerable, and we can see what it would cost us if it weren’t.

The way out does not begin with another argument. It begins with looking — with turning attention onto experience itself and reporting, plainly, what is there to be found. That sounds almost too simple to matter. It is the most important move in the book. The next chapter runs the exercise, and what it turns up is the crack in the wall that everything after widens into daylight.

The stick, by the way, was never bent. You knew that the whole time. The work ahead is to understand why knowing it was never enough — and what it takes to finally believe it.

Chapter Summary

Met at full strength, the inner theater’s best argument is better than its reputation: from the bent stick, and more sharply from the hallucination where no object exists, it presses science and phenomenology from both sides toward the conclusion that the immediate object of every perception is inner. The serious escape routes — Austin’s charge of grammatical confusion, the disjunctivist’s denial of a common factor — are each on the trail of something true but not yet decisive without a distinction the book hasn’t drawn. So the chapter leaves the wall standing on purpose, with the hallucinated clock to mark the cost — on the veil picture, veridical perception is merely lucky hallucination — and the way out begins not in another argument but in looking, the next chapter’s exercise.

Notes

  1. The classic modern statement of the argument from illusion, with sense-data as the inner items, is in H. H. Price, Perception (London: Methuen, 1932), and A. J. Ayer, The Foundations of Empirical Knowledge (London: Macmillan, 1940). The generalizing step — from the misperceived case to the claim about all perception — is the argument’s load-bearing and most questionable move; it is reconstructed and criticized in J. L. Austin, Sense and Sensibilia, reconstructed from manuscript notes by G. J. Warnock (Oxford: Oxford University Press, 1962), and in Howard Robinson, Perception (London: Routledge, 1994), who defends a version of it via what he calls the Phenomenal Principle (roughly: if something appears to a subject to have a sensible quality, then there is something of which the subject is aware that does have that quality). Naming the Phenomenal Principle is useful precisely because, once stated explicitly, it is far less obvious than the argument needs it to be. ↩
  2. The argument from hallucination is the modern refinement of the argument from illusion; the two are often run together but are philosophically distinct. Illusion begins with misperception — a real stick seen as bent — and infers a mediating inner item. Hallucination begins with cases where no external object is present at all and infers that the immediate object of awareness must be inner. The hallucination version is the harder challenge for the direct realist because it cannot be deflected by pointing to the real object that is merely being misrepresented; there is no real object. Modern formulations appear in Frank Jackson, Perception: A Representative Theory (Cambridge: Cambridge University Press, 1977), and Robinson, Perception (1994). The intentionalist response — which grants that veridical and hallucinatory experiences can share representational structure while denying that this makes inner objects the targets of awareness — is developed later in the book, in the chapters on transparency and the identity claim; it is deliberately withheld here. ↩
  3. J. L. Austin, Sense and Sensibilia (1962). Austin’s central charge is not merely that the argument selects unrepresentative examples (though he presses that too — perception consists overwhelmingly of cases where things are as they appear), but that it depends on what he treats as a grammatical confusion: the move from “x looks F” to “there is something that is F of which the subject is aware.” For an extension of Austin’s approach into epistemology, see John McDowell, “Criteria, Defeasibility, and Knowledge,” Proceedings of the British Academy 68 (1982): 455–479. The decisive form of the diagnosis — recasting the confusion as a conflation of the vehicle of a representation with its object — is reserved for the later chapter that names that distinction. ↩
  4. Disjunctivism about perception holds that the common factor between a veridical perception and a subjectively indistinguishable hallucination is merely epistemic — the subject cannot tell from the inside which is which — and that this epistemic commonality implies no shared metaphysical nature. The view was given its canonical modern formulation by Paul Snowdon, “Perception, Vision and Causation,” Proceedings of the Aristotelian Society 81 (1981): 175–192, and developed in significantly different ways by John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994), and M. G. F. Martin, “The Transparency of Experience,” Mind and Language 17 (2002): 376–425. Martin’s version is the more radical and has generated a large critical literature; see Alex Byrne and Heather Logue, eds., Disjunctivism: Contemporary Readings (Cambridge, MA: MIT Press, 2009). This book does not finally rest on disjunctivism; it takes the disjunctivist’s core instinct — that two states indistinguishable from the inside can differ profoundly in what they are — and vindicates it by a different route (the representational account), without inheriting Martin’s commitment to perception as a primitive, unanalyzable relation. ↩
  5. The hallucinated-clock case is structurally a variant of the cases Edmund Gettier introduced in “Is Justified True Belief Knowledge?”, Analysis 23 (1963): 121–123, in which a subject holds a justified true belief whose truth is disconnected from the grounds of the justification. The version used here pushes the disconnection further than Gettier’s originals: there is no perceptual contact at all between believer and truthmaker, only an internal state that happens to align with it. The case fails several standard post-Gettier conditions on knowledge — the safety condition (Ernest Sosa, A Virtue Epistemology, Oxford: Clarendon, 2007; Duncan Pritchard, Epistemic Luck, Oxford: Oxford University Press, 2005), on which a belief counts as knowledge only if it could not easily have been false; and the sensitivity condition (Robert Nozick, Philosophical Explanations, Cambridge, MA: Harvard University Press, 1981), on which the belief must vary with the truth of what is believed. The case is used here not to adjudicate among theories of knowledge but to expose the structural parallel between the inner-theater picture of perception and a model of knowledge on which true belief is held for reasons internal to the believer with no constitutive contact with what makes it true. The point is not that perception is a species of knowledge, but that both are world-involving in a way the inner theater cannot accommodate. ↩
← Chapter 3: How the Inner Theater Was Built
Contents
Chapter 5: The Transparency of Experience →