5.1 The Tool on the Table

Lay a walking map on the table and find the hill on your route. Crowded contour lines show a steep climb ahead, though the paper lies flat. You can smooth a crease in the map without making the walk any easier. Three things differ here: the ink and paper, how they depict the terrain, and the hill itself. The map can get the slope wrong; no second hill has to appear on the table to make that error possible.

The last chapter gave us transparency: looking for the seeing ordinarily returns our attention to what seems seen, rather than to a second colored surface.¹ The map helps name a distinction that observation needs. The physical vehicle does the representing; the content specifies how things are presented; an object, when there is one, supplies what the state concerns. We will use that distinction against two moves: treating the brain state as something seen, and demanding an inner bent thing because a straight stick looks bent.

The analogy has a limit. Looking at a map does not amount to seeing its hill. It shows only that a vehicle here can concern something elsewhere without sharing its properties. Direct perception needs more: the object must help cause the very presentation of it under the right world-involving conditions. The map earns the distinction, not a theory of sight.

5.2 The Grammar That Hides the Distinction

Open a dictionary to experience and you find, near the top, a harmless epistemic sense: direct observation of or participation in events as a basis of knowledge.² The definition posits no private object. Grammar nevertheless makes one remarkably easy to imagine. The events sit on one side, the observer on the other, and the noun offers us a third term: experience, ready to become the medium where the two meet.

The noun does not summon a ghost. It leaves a chair out for one.

Watch how the noun behaves. Experience takes adjectives, sits as a subject, gets had and undergone and described. Some adjectives properly qualify the episode: a recollection can be vivid, a headache painful, an afternoon rich. Trouble begins when the character of seeing gets pictured as an exhibit with its own colors and grain. There is the tree; there is Eloise looking at the tree; and now we imagine Eloise’s experience of the tree as a third item that a sufficiently careful science might one day inspect.

Grammar can arrive with the stage already set. Once experience has been pictured as a third item, the famous problems nearly write themselves. How do the events out there get translated into properties in here? Why do the inner properties feel the way they do rather than some other way? How could anyone but the haver ever know about them? The noun did not cause three centuries of debate; it made the proposed scenery feel familiar before the arguments began.

None of this denies that experience has character. Looking at a sunset differs unmistakably from listening to a fugue: the redness, the late gold, the slow cooling of the light all belong in the account. The inner-exhibit picture prejudges where to look for that character. It sends us inward, to a third item whose own qualities supposedly make the difference. Chapter 4 found no such item announcing itself to attention. The positive explanation still lies ahead.

5.3 Vehicle, Content, and Object

Apply the map’s distinction to seeing an apple on the counter.

The vehicle is the physical state doing the work: here, a pattern of neural activity. The content specifies how the apple appears: red, round, and resting on the counter. These are its accuracy conditions, what would have to hold for the presentation to get things right. The object is the apple itself, when one exists and stands in the right relation to the seeing.

The content and object cannot trade places in this list. The same apple can look red or, under misleading light, brown; the object stays while its presentation changes. A hallucination may present a red apple where none exists. Content remains possible without a matching object, and neither case requires the neural vehicle to turn red.

The trouble starts when we let the vehicle’s address determine where its object must be.

Call the mistake the container fallacy: the vehicle of perception is a state of the brain, the brain is inside the skull, therefore what perception delivers must also be inside the skull: an inner item, not the outer world.

That therefore does not follow. The map lies on the table; the hill does not. Likewise, a neural state in your visual cortex can concern the apple on the counter without putting the apple in your cortex. Where representing happens does not by itself tell us where to find what it concerns.

Dress that fallacy in a modern lab coat and it fools serious people. We do not really perceive the world, the argument goes; we perceive a representation in the brain, because the proximate cause of the experience is a brain state. Fred Dretske caught the slip.³ He had a working rule (explain the mind in terms continuous with the rest of nature) and a taste for the homely example that catches a grand mistake red-handed. To say we read ink, because ink is the proximate cause of our reading, confuses the vehicle of the written words with what the words are about. The ink causes the reading; what you read brings news from somewhere else. Causal proximity never adds up to perceptual aboutness. The chain of causes runs through the brain; what the perception is of runs out to the world. Mistake the one for the other and you have built yourself a screen.

5.4 The Argument from Illusion Comes Apart

Return to the straight stick in the glass of water.

Recall the argument that put you there, because its respectability carries the point.⁴ A straight stick, half-immersed in a glass of water, looks bent. Nobody disputes that much; you can run it tonight at the sink. The argument continues: the stick itself remains straight. Pull it out and check; run your finger along it. And yet, the argument presses, you are aware of something bent. Bentness stands present to your mind. Since the public stick lacks it, some other object must bear the quality: an inner item, a sense-datum. Once that item gets its foot in the door, the continuous series from distorted to accurate perception spreads it across every case, and the world retreats behind a screen. That road runs from a stick in a glass to the veil of ideas, and for three centuries it has looked like a one-way street.

The argument turns on one step: from the stick looks bent to you are aware of something that is bent. It converts a presented quality into a bearer; the continuous series spreads the result. The container fallacy instead transfers the vehicle’s location to what we see. One reifying family, two doors.

Of course bentness is presented. Sight presents the stick as bent and, in that respect, misrepresents the straight stick.⁵ Bentness enters the accuracy conditions: the presentation would get that feature right if the stick had the shape it seems to have. Nothing yet requires the vehicle, or a second object, to bear the shape.

Here the careful defender of inner objects digs in at his strongest. He grants the grammar. He grants that “looks bent” can smuggle. And he presses anyway: but I am genuinely presented with bentness; bentness stands there before my mind; presentation requires something that has it.

This is the best version of the argument. A state can represent bentness with nothing bent anywhere, inner or outer, just as a sentence can concern a dragon without a dragon crouching behind the words. Representation needs no instance of its content to do its work.

The comparison leaves one debt unpaid. Reading about a dragon does not feel like having a dragon sensibly present, while the bend at the sink does feel given in vision. The defender can grant the point about sentences and press that difference. Fair enough. But the Phenomenal Principle now shows its hand: it no longer reports a neutral fact about what appears; it offers a theory of what sensory presence must involve. Until the bearer theory wins, the bent object does not follow.

Hunt on for an inner thing, itself bent, carrying the quality the way a sense-datum supposedly carries it, and you have made content into one more object seen. A straight stick, presented inaccurately; no second bent thing follows, in the glass or behind your eyes. And where on the smooth slope from distorted to accurate perception does awareness switch from an inner object to an outer one? Nowhere. The distorted end never earned an inner object in the first place.

Nobody saw the linguistic slide more clearly, or said it more enjoyably, than J. L. Austin. He delivered the lectures later reconstructed as Sense and Sensibilia at Oxford; G. J. Warnock assembled the book from Austin’s notes after his death.⁶ It takes the argument from illusion apart joint by joint, and two of its cuts help here.

1. Austin first exposes a grammatical sleight: the slide from “x looks bent” to “there is something that is bent which I sense.” “Looks bent” no more requires a bent something than “looks likely” requires a likely something, or “looks like rain” requires a thing made of resemblance. The grammar of looks invites the inner object; it does not deliver it.
2. The second cut is one I had underestimated until I read him properly. The argument from illusion, Austin points out, runs together two things English keeps carefully apart: illusion and delusion. In an illusion, a real public arrangement looks other than it is: two equal lines look unequal, or a stage performer appears headless. In a delusion, by contrast, something wholly unreal may seem present: Austin’s patient sees pink rats. The immersed stick belongs on the public-object side of that distinction, though Austin thought the case too familiar to count as a proper illusion without stretching the word. One real, straight stick looks bent through water. Nothing in that description conjures a private replacement.

This cut blocks the quick route from public misperception to a private object. A careful sense-datum theorist can still grant the stick throughout and demand a bearer for the presented bentness. That slower route returns us to the Phenomenal Principle, where the representational alternative has shifted the burden but not yet explained sensory presence. Austin clears the verbal undergrowth. He does not make the remaining argument disappear by taxonomy.

The argument from illusion therefore no longer forces the veil of ideas. “The stick looks bent” does not by itself place a bent object before awareness, and the continuous series cannot spread an object that never entered at the distorted end. The inner bent thing required an additional theory of presence, quietly entered as a premise.

5.5 And the Theater Goes With It

The familiar theater makes the other mistake: it turns the vehicle into an object. Behind the eyes stands a private screen, with a Self in the dark watching the show. It can feel forced on us by sheer biology: the brain sits in the dark, signals arrive, surely it must be working from an inner image.

But a neural representation need not be an image anyone inspects. Neural states can register features, guide attention, and shape action without another viewing of those states. The container fallacy adds that viewing; the neural facts do not. Just as mapping the hill does not erect a hill on the table, processing its slope does not erect a slope inside the skull. We still need to explain how that processing lets us see. Adding a screen would give us another thing whose seeing needs explaining.

5.6 Disjunctivism, and Why the Book Declines It

Two camps agree that no veil stands between us and the world, and disagree about why. This book and the disjunctivists both send the inner object packing; they part over what perception shares across its good and bad cases. The sense-datum theorist and the representationalist both posit a positive common factor across that divide, then quarrel over what it involves — an inner object for one, representational content for the other. Which side is right decides whether the identity claim ahead has the materials it needs.

Disjunctivism plants its flag at the case Austin taught us to hold apart from illusion: the hallucination, the tomato that seems present when no tomato sits anywhere to be seen. It denies that genuine perception and a matching hallucination share one fundamental experiential nature.⁷ M. G. F. Martin built its modern case. Really seeing the tomato amounts to a relation that includes the tomato, a state you could not occupy if the tomato were not there; the matching hallucination counts as a different kind of state altogether, sharing indiscriminability without sharing that basic kind. Either you stand related to a real object, or you occupy some other state that merely seems the same: a disjunction, with no neutral experiential core left for the sense-datum theorist to make inner. No screen, no veil; the relation reaches all the way out.

Disjunctivism makes for an elegant escape, and I respect it. It pushes realism about perception to its most uncompromising form, building the tomato itself into the constitution of the perception. This book takes another road, locating the world in what experience presents rather than in what constitutes it. Against the sense-datum theory we march together: two parties refusing to let the bent stick smuggle an inner object onto the scene.

Hallucination does not force either camp to admit an inner object. On the common-content account, a state can present a tomato where none exists; its accuracy conditions go unmet. No private tomato replaces the missing public one. That offers an alternative to the bearer theory, though sensory presence still needs explaining.

The disjunctivist has a complaint waiting in return. If one complete content-state can occur in both perception and matching hallucination, the tomato starts to look external to the experience’s fundamental nature — the veil rebuilt one level up. As stated the charge does not stick: content never becomes an object you perceive instead of the tomato. Whether a common content nevertheless leaves the current world too far outside the nature of seeing remains a second question, and the world-involving account ahead must answer it.

Why decline that road? A forged banknote helps locate the dispute. I may spend it because I take it for money. A reply in Martin’s spirit says that its being indistinguishable from a genuine note explains that response; the two notes need not share a fundamental nature. Nor does a common belief about them prove a common kind of experience. The question concerns what best explains the apparent scene that prompts the belief, not just the belief afterward.⁸

I favor common content because it describes that scene positively: a tomato presented as red, round, and within reach, whether or not the tomato is there. It lets one account explain the detailed appearance while the right causal contact distinguishes seeing from hallucinating. Martin instead gives genuine perception the basic role and explains the bad case through its relation of indiscriminability to the good one. His account preserves the object’s place within seeing more directly; ours promises a fuller common explanation of how things seem. That promise must earn its keep in the cases ahead. The banknote cannot decide between them.

So we let the disjunctivist go his own way, with a tip of the hat. He reaches the open air by another route. We keep the common factor and accept the bill: the positive theory now has to earn it.

5.7 The Seeing and the Seen

Right now you have something in view: this page, the wall behind it, the light in the room. The seeing does the presenting; its content specifies how the room appears. A vehicle works inside the head. The room need not move there, and content supplies no further exhibit standing between you and it.

Without an inner-object veil, the low-grade skeptical worry (what if the screen and the world disagree?) loses its grip. A second worry remains: whether common content leaves the current object outside the fundamental nature of seeing.

Naming the confusion blocks two arguments for an inner exhibit; it supplies no complete positive account. Seeing the apple’s red still has a definite character. A serious view—I held a version longer than I like to admit—grants this chapter and insists on an intrinsic feel beyond anything represented.⁹ It does not rebuild the screen, and the next chapter owes it a careful answer.

If the felt character of your experience belongs to no inner exhibit, then what in the world does it consist in? Hold the question. Meanwhile, you can take the stick out of the water.

---

Chapter Summary

The claim. One reifying family recurs across three centuries of trouble. The container fallacy turns an internal vehicle into an object of perception; the argument from illusion turns presented content into a further bearer. Both install a second object where representing needs none.

Where the argument stands. The vehicle’s location does not place what we see inside the skull, and a presented bend need not belong to an inner bearer. Content specifies how things appear; an object, when present, supplies what the state concerns. Austin blocks the quick slide from public error to something conjured, leaving the Phenomenal Principle to defend its theory of sensory presence.

What’s left open. What gives content sensory presence, how world-involving contact distinguishes seeing from hallucinating, and whether a common factor creates an interface all remain for the positive account. The qualia realist who still posits an intrinsic feel (Block’s mental paint) gets a separate answer there as well.

The hand-off. Disjunctivism reaches the same anti-inner-object result by denying a common fundamental experiential nature. This book keeps common content for its explanatory work and must now show that content forms no new veil. The next chapter begins paying that bill.

---

Notes

1. The transparency observation this chapter inherits rather than re-runs belongs to Gilbert Harman, “The Intrinsic Quality of Experience,” Philosophical Perspectives 4 (1990): 31–52, and, in its older form, to G. E. Moore, “The Refutation of Idealism,” Mind 12 (1903): 433–453, at 450 (the “diaphanous” character of the sensing). The phenomenological demonstration — attend to the experience and you find only its object — is the business of the preceding chapter (The Transparency of Experience); this chapter assumes it and supplies the diagnosis. ↩
2. Merriam-Webster’s Collegiate Dictionary, 11th ed. (Springfield, MA: Merriam-Webster, 2003), s.v. “experience”: “direct observation of or participation in events as a basis of knowledge.” The definition itself carries no commitment to an inner object. Section 5.2 concerns the ease with which the noun can be reified, not a claim that lexicography created the metaphysics. ↩
3. The point that the proximate cause of an experience is not thereby its object — that to “perceive” a brain state because a brain state proximately causes the experience confuses the vehicle of a representation with what the representation is about — is Dretske’s, and runs through his representationalist work: Knowledge and the Flow of Information (Cambridge, MA: MIT Press, 1981), on the carrier of information versus its informational content, and Naturalizing the Mind (Cambridge, MA: MIT Press, 1995), on displaced perception and the transparency of the representational medium. The “reading ink” formulation in the text — we read the words, not the ink that carries them — illustrates Dretske’s point rather than quoting a fixed passage. ↩
4. The argument from illusion is reconstructed at full strength, and assessed, in A. D. Smith, The Problem of Perception (Cambridge, MA: Harvard University Press, 2002), chs. 1–2 — the most thorough modern treatment, and notably one sympathetic to the argument’s seriousness even where it ultimately resists the sense-datum conclusion. Smith isolates the principle the literature, following Howard Robinson, calls the “phenomenal principle” — that if one is sensibly aware of something as having a quality, then there is something of which one is aware that has that quality (Robinson, Perception [London: Routledge, 1994], 32) — as the argument’s central premise. Intentionalists reject that inference: a state may represent a quality without anything’s instantiating it. For the contemporary intentionalist diagnosis on which this chapter leans, see Tim Crane and Craig French, “The Problem of Perception,” Stanford Encyclopedia of Philosophy (Fall 2021 edition), and Crane, “Is There a Perceptual Relation?,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), 126–146. The dragon case makes the alternative coherent; it does not by itself explain why perceptual content feels present. Chapter 6 takes up that debt. ↩
5. This chapter locates the presented bentness in the content of a representation that misrepresents a real, straight stick. That move presupposes the same general representational structure across accurate and illusory seeing. Metaphysical disjunctivism need not contest that pairing; its characteristic denial concerns veridical perception and a subjectively matching hallucination. Section 5.6 explains why this book keeps a positive common factor across that stronger divide. ↩
6. J. L. Austin, Sense and Sensibilia, reconstructed from the manuscript notes by G. J. Warnock (Oxford: Oxford University Press, 1962). The illusion/delusion distinction and the critique of the “looks F” → “is aware of something F” inference are developed in Lectures III–V. Austin separates illusion from delusion and objects to treating the familiar immersed stick as though it introduced a second bent object. That blocks the quickest route from misleading appearance to a private intermediary. A defender who explicitly adds a Phenomenal Principle or another bearer premise has made a further argument, however; Austin’s taxonomy alone does not settle it. ↩
7. The metaphysical (or “phenomenal”) disjunctivism at issue in Section 5.6 is developed in M. G. F. Martin, “The Transparency of Experience,” Mind & Language 17 (2002): 376–425, https://doi.org/10.1111/1468-0017.00205, and “The Limits of Self-Awareness,” Philosophical Studies 120 (2004): 37–89, as well as “On Being Alienated,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), 354–410; its modern origins lie with J. M. Hinton, “Visual Experiences,” Mind 76 (1967): 217–227, and Experiences: An Inquiry into Some Ambiguities (Oxford: Clarendon Press, 1973), and with Paul Snowdon, “Perception, Vision and Causation,” Proceedings of the Aristotelian Society 81 (1980–81): 175–192. The claim relevant here is not that veridical perception and subjectively indistinguishable hallucination share no true description or downstream effect, nor that indiscriminability entails positive phenomenal sameness. It denies a common fundamental experiential nature. Veridical perception receives a positive, object-involving account; Martin characterizes hallucination through its indiscriminability from genuine perception. A separate, epistemological disjunctivism, associated with John McDowell (“Criteria, Defeasibility, and Knowledge,” Proceedings of the British Academy 68 [1982]: 455–479), concerns the reasons perception supplies rather than the metaphysics of phenomenal character. For the current taxonomy, including fundamental-kind, content, phenomenal, and epistemological formulations, see Matthew Soteriou, “The Disjunctive Theory of Perception,” Stanford Encyclopedia of Philosophy, substantive revision July 28, 2026, secs. 2–5. ↩
8. The common-factor pressure follows Tim Crane; see “Is There a Perceptual Relation?,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), 126–146, and (with Craig French) “The Problem of Perception,” Stanford Encyclopedia of Philosophy (Fall 2021 edition). The banknote contrast in the text extends that pressure. It distinguishes a neural event’s efficient causal role from the subject’s reason for belief and action, and both from whatever constitutes the experience itself. A disjunctivist may grant common downstream beliefs or dispositions while denying a common experiential nature. The objection therefore asks whether that concession explains enough; it does not show that the disjunctivist has restored the disputed common factor under another name. ↩
9. The deferral flagged in Section 5.7 is to the qualia realist’s residual claim — that experience carries intrinsic, non-representational qualitative character (Ned Block’s “mental paint”) over and above its representational content. That this chapter removes the inner object does not by itself answer the claim that experience has intrinsic qualities; the two arguments come apart. The answer belongs to the next chapter, which defends the positive identity claim, meets the inverted spectrum, and assembles the case against mental paint — and to the hard-cases chapter after it, where the paint’s favorite refuges, pain and the moods, turn out to be content as well. Block, “Mental Paint,” in Reflections and Replies: Essays on the Philosophy of Tyler Burge, ed. Martin Hahn and Bjørn Ramberg (Cambridge, MA: MIT Press, 2003), 165–200. ↩