2.1 The Cartesian Tradition

René Descartes wanted to build knowledge on a foundation that could not crack. So he went looking for something he could not doubt, and found that almost everything cracked. The senses deceive: he had been fooled before, and what proof did he have that he wasn’t being fooled right now? Even arithmetic felt shaky. Perhaps some sly demon was tampering with his mind, so that two and two only seemed to make four. One by one he threw the candidates out, until one thing was left standing — the doubting itself. He could not doubt that he was doubting. Cogito ergo sum. I think, therefore I am — though the tag itself belongs to the Discourse and the Principles; the Meditations, where the doubt is staged most starkly, reaches it as “I am, I exist.” Everything else might be an illusion; the thinker having the illusion could not be.

Notice what that proves, and what it doesn’t. It proves there is a thinker. It proves nothing about the world outside. Descartes has planted certainty inside — in the act of thinking, in the mind’s view of its own contents. The world beyond stays a guess. Solid ground lies only within the mind.

The next move proved fateful in the reading of Descartes that became influential: ideas supply what the mind encounters immediately, and the world must be reached through them.¹² Look at a red apple. On this reading, awareness first meets your idea of it; the apple stays one step away, on the far side of the representation. Descartes also permits a different reading, on which an idea names a mode of thinking rather than an item inspected in the apple’s place. That difference matters. The historical subject here is the veil-reading and its influence, not an uncontested verdict that Descartes put a colored picture inside the head.

Descartes noticed further that an idea of a thing need not resemble what causes it. In the Optics he resisted the demand that representation work through resemblance. The word “apple” need not look like an apple to concern one; neural activity need not look like its object either. This insight helps free representation from picturing. It also leaves a question for any account, including ours: what makes a state represent one thing rather than another? The veil-reading adds a separate difficulty. If all you meet directly are your own ideas, how could you step outside them to check their relation to the world?

Descartes did not intend to leave the world a permanent guess. The Meditations attempt to recover knowledge after radical doubt, and a nondeceiving God supplies the guarantee. That recovery faces the famous charge of a circle: can clear and distinct reasoning establish the God whose guarantee secures such reasoning? Descartes had replies; the charge is not a refutation simply because it can be put in a sentence. The relevant pressure here concerns the veil-reading: once private ideas supply the whole starting evidence, the relation to an independent world needs a defense that cannot simply assume the contact placed in doubt.

Berkeley would reject mind-independent material substance while retaining sensible objects and the spirits that perceive them. Hume would press the limits of our rational warrant for an independently existing world. Their answers differ sharply. Both expose the difficulty of beginning with ideas and then asking how those ideas reach beyond themselves.

---

2.2 The Empiricist Renovation

The British Empiricists gave experience a central job: explain how our ideas and knowledge arise. That project need not begin with a private theater. But on the reading that came to dominate their reception, experience supplied ideas within the mind, leaving the world that caused them outside. Each thinker tried to resolve a difficulty inherited from that arrangement.

John Locke (a physician by training, who spent the better part of two decades on his great Essay before letting it out into the world) was read as keeping the core Cartesian picture, and it is that reading which traveled: what we meet directly are our own ideas, not the world. What he changed was where ideas come from. Not from innate ideas or principles, he held, but from experience. Every idea traces back to one of two springs: sensation, what comes in through the senses, or reflection, what the mind notices about its own workings. The newborn mind is a blank slate. Experience does the writing.

For perception, that meant the senses hand us simple ideas, raw bits like color, warmth, hardness, which the mind builds into ideas of whole objects. Sit by a fire and you take in simple ideas of heat, light, and color; your mind assembles them into the idea of the fire. On the veil-reading, the fire itself, the thing in the grate, remains beyond immediate awareness. What you meet is the assembly your mind has run up.

To keep this from sliding into full idealism, Locke split an object’s qualities in two. Primary qualities (shape, size, mass, motion) really belong to the thing, independent of us. Secondary qualities (color, taste, smell, warmth) work differently. They are powers in the object to produce sensations in us, and the sensation a power produces resembles nothing in the object, unlike our idea of its shape, which does.³ So when you see the red of an apple, you are not taking in a color the apple owns the way it owns its shape. You are registering an idea the apple’s real powers stir up in you — a private effect, not a feature of the world.

Locke reached for scientific respectability. Physics describes the world in mass and motion, not in redness and sweetness, and he wanted his account to match. But the bill is steep. If the colors and warmths we live among really belong to us and not to things, then much of what we take ourselves to see (color, texture, the slant of afternoon light on a wooden floor) is something the mind manufactures in response to a world that holds none of it. The world thins out; the inner life swells; Descartes’ gap widens.⁴

George Berkeley spotted a crack in Locke’s split. He came to the problem as a young fellow of Trinity College Dublin, a clergyman set on defending religion against the materialism creeping through the new science, and the crack he found would serve that end. (The bishopric, at Cloyne, came decades later; the Principles is the work of a philosopher barely twenty-five.) If all we ever meet directly are our own ideas, how could we check which ideas resemble the world and which don’t? We would have to step outside our ideas and set them beside the things themselves — the one thing we can never do. So the whole notion of a primary quality “resembling” something mind-independent falls apart on Locke’s own terms. Berkeley’s conclusion was bracing: for sensible objects, esse est percipi, to be is to be perceived. He denied a material substance behind the ideas, not the minds perceiving them. Sensible objects consist of ideas; spirits perceive and act. For Berkeley this meant not despair but rescue: the familiar world remains, upheld by a divine mind.

Berkeley insisted he was denying nothing the plain person believes: the cherry stays as red, as round and as edible as it ever was, and only the philosophers’ inert matter behind it gets struck out. Grant him that and the argument offers a redescription, not an elimination. His appeal to God also gives the sensible world an order independent of any finite person’s wishes. I do not adopt his idealism; this book starts from a mind-independent physical world. But invoking divine order does not refute him, and ridicule cannot settle the metaphysics. His challenge remains: if ideas alone supply our evidence, what justifies the extra material world the realist places behind them?

David Hume pressed hardest on the difference between rational justification and the beliefs we cannot help having. His skeptical arguments question our grounds for continued unperceived objects, necessary causal connections, and a self beneath passing perceptions. Yet he did not expect anyone to stop believing in bodies. Nature, he thought, had not left that to our choice. The difficulty lay in reconciling ordinary confidence with what philosophical reflection could warrant. Hume’s wider work in morals, politics, and history belongs beside that inquiry, not as evidence that a defeated philosopher abandoned it.

Philosophers have a name for what all this builds: the veil of perception.⁵ On the Cartesian-Lockean picture, we never perceive the world directly. We perceive our representations of it: our ideas, our inner images. The world sits behind the veil, causing the show but never stepping onto the stage. And granting the causal story doesn’t help. Yes, the red you see traces back, along a chain of events, to a surface that reflects light a certain way — but you are still stuck with the gap. The surface is one thing. Your experience of its redness is another. How the two ever meet stays a mystery.

A word on those inner images, lest the target look like a cartoon. The careful sense-datum theorists hung no little colored pictures behind the eyes; Russell, Price, and Broad took pains to deny that sense-data are mental images at all.⁶ The thread this book pulls runs under the cartoon: the move that slips an immediate object of awareness (idea, datum, appearance, model) between the perceiver and the world, reached, if at all, only by inference. Whether that object is pictured as a picture is beside the point. The interposition is the target, and every version shares it.

---

2.3 How Neuroscience Inherited the Picture

Modern neuroscience might have dissolved the theater. Some descriptions instead preserve its floor plan. In them, the theater has simply been re-staffed with neurons.

Neuroscience tells a story of hidden perceptual machinery. Light falls on the retina, photoreceptors convert it into electrical signals, those signals travel through the optic nerve to the lateral geniculate nucleus, which relays them to the primary visual cortex, which projects to a cascade of higher visual areas that extract edges, orientations, depths, motions, colors, faces, objects. By the time consciousness enters the picture (if “enters” is even the right word), an enormous amount of processing has already happened, most of it below the level of any awareness. What we experience as the immediate, effortless perception of a scene turns out to be the output of a massively parallel computational enterprise running largely in the dark. The neuroscientist Anil Seth describes the brain as a “prediction machine,” actively generating a “controlled hallucination” of the world rather than passively receiving it.⁷

Trouble enters with the description. When “the brain constructs a model” slides into “the subject perceives that model,” the model’s location inside the brain seems to place awareness there too. But that conclusion belongs to an inherited frame, not to the neural evidence.

The phenomenologist replies: you see the fire, not a model. The model theorist can answer that a transparent model presents itself as the fire. Both descriptions fit. Talk of models alone therefore cannot settle what awareness reaches.⁸

A charitable reading remains true. The brain processes information, predicts, resolves ambiguities, and recovers three-dimensional structure from retinal input. None of that shows that awareness lands on the processing rather than the world. Machinery may lie in the brain while its object does not.

The needed distinction separates the state that represents from what it presents. The Seeing and the Seen earns that tool after we inspect experience. For now: that the brain builds the seeing does not put the seen inside the skull.

Five claims crowd together in the inherited picture, and the rest of the book needs them kept apart.

1. Perception depends on causal mediation through the senses and brain. It does.
2. Neural systems form internal states that represent, predict, and control. They may.
3. We perceive the world only by first perceiving one of those states—an idea, datum, image, or model. That is the veil of perception.
4. The processing culminates at a central place where its results are presented for consciousness. That is the Cartesian Theater in Dennett’s strict sense.
5. An observing self witnesses the presentation, perhaps accompanied by intrinsic private qualities disclosed only to it. Those are further claims again.

The first two do not entail the last three. Internal representation supplies machinery, not an audience; distributed neural work need never converge on a screen. This chapter targets the interposed object and the centralized presentation to a witness. Later chapters ask separately whether phenomenal character contains intrinsic mental paint and how an integrated subject emerges from the distributed organization. A theory can lose either dispute without making neural representation itself Cartesian.

---

2.4 The Same Move in Memory, and in Machine “Memory”

Once you can see the inner-theater move in perception, you start to find it everywhere. Memory makes the cleanest second case.

Ask someone to picture the kitchen they grew up in. Most people, asked to do this, report a small private screening — a faded reel running somewhere behind the eyes, colors a little washed out, the kitchen viewed from a vantage that no longer exists. The picture seems irresistible. It also gets the metaphysics almost exactly backwards.

Keep two things apart that the exercise runs together. Summoning imagery is not the same act as remembering: you can call up a vivid kitchen you never stood in, and you can remember a great deal about the one you did without any picture arriving at all. The inner reel, strictly, is a story about imagery. What makes the case worth having is that the same move gets made about memory itself, and comes to grief in the same way.

The theater-reading supplies a ready account: the mind contains a chamber, perception leaves traces in it, and some inner agency calls those traces back and watches them. That picture can feel true. When you summon the kitchen, it may seem to be a thing summoned, an inner item produced for inspection.

But ask what it would actually mean for the kitchen to sit in the skull, and the picture starts to wobble. The faded colors are not pigments in the head; the vantage point is not a camera position in the cortex. Whatever the right account turns out to be, remembering looks far more like representing an absent scene (the kitchen, the one that no longer figures in your daily traffic) than like screening an inner souvenir of it.⁹ We see the world; we remember the world; the verb changes and the object does not. Remembering needs the same repair perception needs, and it waits on the same later chapter.

Notice how far the cinema picture reaches. Contemporary discourse around artificial intelligence has quietly inherited it, and Part Four will develop what follows from that. The point available here is narrower, and it is enough: if we first reduce human memory to the storage and retrieval of inner items, a database wins by definition. The cinema picture has packed the answer into the comparison.

---

2.5 Five Questions, One Picture

Once the history is visible, return to the questions that make the inner theater feel unavoidable. They no longer arrive as neutral data. They arrive inside a picture with a history, and each carries an answer that springs up so readily it scarcely seems like a theory at all.

1. Start with the redness of the tomato. Does it belong out there on the fruit, or does the brain add it? The answer that springs to mind: the brain adds it. Out in the world, on the modern scientific view, we find only invisible wavelengths reflecting off colorless surfaces; the warm red you actually see must come from somewhere inside the brain or mind, or so it seems.
2. Push a straight stick halfway into water and it looks bent. You know better — pull it out and the stick stands as straight as ever. So what is the bent thing you plainly see? The reflex says you see a bent something; and since the stick itself stands straight, that bent something cannot belong to the stick. It can only show up as an appearance, standing in for the stick, and the bent appearance belongs to you, in here, or so it seems.
3. When Shakespeare’s Macbeth, in the grip of hallucination, reaches for a dagger that does not exist, what does he reach toward? Something hangs in the air for him (dagger-shaped, edged, seen) and yet the air holds nothing. The dagger must be an item in his mind that he sees, or so it seems.
4. When you and a friend both point at the tomato and say “red,” do you even share the same red experience?¹⁰ The natural thought: no one could ever really know, since each of us has access only to our own case. Where you see red, I might see green — and since we would both go on calling ripe tomatoes red, no test would catch it. So the experience turns private: an inner screening with an audience of one, so it seems.
5. And when a machine answers you in fluent, feeling sentences, does anyone live behind the words? The reflex that takes the question seriously says you cannot tell from outside. Either some inner light burns, some flicker of what it is like to be the machine, or nothing whatever inhabits the inside, and the fluent words amount to a very good trick. Whatever settles the matter lies behind the words, out of view, or so it seems.

These formulations divide into two families. Vanished color, bent shape, and the dagger in the air invite the same inference: not in the world, therefore in here. The private spectrum and the machine’s hidden light instead borrow the theater’s imagery. That imagery cannot serve as evidence for the picture it already assumes. The underlying questions survive without it, however. Could two experiences differ in feel despite matching in other respects? What would make an artificial system conscious? Neither question requires a screen or spectator, and the later chapters must answer both without treating the questioner as a concealed Cartesian.

Put all five together and you get a single picture so familiar most of us never notice we hold it. According to that picture, behind the eyes lives an inner theater. The real you sits in the audience. When you see the tomato on the counter (round, waxy, that red that seems to glow from inside), you see only your inner representation of it. The fruit itself you only infer.

2.6 The Picture’s Credentials

Do not wave the inner theater away. The picture has earned a wide following. Neuroscience tells us, correctly, that the brain builds models: the redness you see emerges at the end of a long process from photon to retina to cortex, and the process bends to fooling, lesioning, drugs, rewiring. Close your eyes and the tomato disappears; open them and it returns. Seeing looks separable from the seen — something that could in principle run without it. And above all, the picture preserves the one phenomenon any truthful description has to keep on the table: that there is something it is like to see red, a felt quality present to you and to no one else, as private as a held breath. Put those together and the inner theater stops looking like a theory. It looks like what taking science and your own experience seriously forces you to accept.

Follow the picture further and the troubles begin. If you only ever perceive your inner image, how do you ever know the tomato exists? And where does the redness live if not on the tomato, not in the photons, and not in the firing neurons? Does it live in some third place belonging to neither world nor brain?

Pour a cup of coffee and take the first sip of the morning. Something happens there that a physicist could describe at enormous length. Volatile compounds reach receptors in the nose; dissolved acids and bitter alkaloids meet the tongue; heat sensors report; signal after signal runs inward and fans across the cortex, where memory and expectation arrive to help. Every step of that account may be true. One day it may even be finished. Read it back to yourself anyway, and something seems to have gone missing from it. It says nothing about the warmth, or the faint bitterness, or the way the roast announces itself before you have decided anything at all. Why should any of that processing feel like anything from the inside?

The question has a name. So does the paper that gave it one: David Chalmers published “Facing Up to the Problem of Consciousness” in 1995,¹¹ and thirty years on, the question has not gone away. He gives the name easy problems to the ordinary business of explaining what a brain does: how it discriminates, integrates, attends, reports. Easy not because anyone can solve them, but because we know what solving them would look like. The hard problem names what remains. Grant every easy problem its answer. Grant the entire functional story, start to finish. It still seems to make sense to ask why the lights are on: why all that processing does not simply run in the dark.

This chapter will not answer any of it. Chapters 9 and 11 take the question up in earnest.

2.7 Indirect Realism and the Invisible Experience

That picture also goes by a name: indirect realism. The theory can be defended, but the picture it gives cannot digest one small thing. You see the tomato, and yet you cannot see or point to anything you might call your experience of seeing it. Odd words, invisible experience, to set side by side — and yours seems to be one. Hold onto the oddness, unresolved for now. Chapter 4 turns it into an argument.

2.8 Why the Picture Persists

The inner-theater picture has been under serious philosophical pressure for a long time. And yet it persists — not just in philosophy departments, but in popular neuroscience, in AI discourse, in the way people talk about their own minds at dinner tables. Why?

Part of the answer is grammatical. “I have an experience” and “I feel a pain” invite an inner object one possesses. But “I see red” need not describe an encounter with inner redness; it may describe a perceptual relation to a surface.

Part of the answer is scientific prestige. When neuroscience says the brain processes signals and constructs models, and when this story appears in authoritative textbooks and popular books by distinguished researchers, the inner-theater picture gets wrapped in the authority of empirical science. It becomes harder to question because it seems to have been confirmed rather than merely assumed.

The picture also takes the right things seriously before going wrong in the explanation. It is a good first guess that has overstayed its welcome. The plain sense of direct contact with the world remains the incumbent; a theory that replaces it with a veil carries the burden of proof.

---

2.9 The Inner Theater, in Place

The inner theater does not stand on history and habit alone. Its strongest argument begins with the bent stick: if it looks bent and is not, perhaps what you directly perceive cannot be the stick. The next chapter tests that inference.

---

Chapter Summary

The claim. The inner theater records no discovery. It grew out of a historically contingent philosophical inheritance: put certainty inside and place the world outside. It then acquired furniture from empiricism and borrowed authority from neuroscience.

Where the argument stands. An influential reading of Descartes and Locke placed ideas between perceiver and world; Berkeley and Hume pressed the difficulties of that starting point in different ways. Modern neural facts show causal mediation, model-building, and reconstruction, not that awareness stops at an inner object. Grammar and scientific prestige help explain the picture’s appeal without proving it true.

What’s left open. Genealogy removes inevitability; it does not refute the theater. Illusion, hallucination, privacy, and the hard problem still deserve direct tests.

The hand-off. The first test begins with the bent stick. If appearance can mislead, must awareness terminate at a private image rather than the object itself?

---

Notes

1. The gentler reading of Cartesian “ideas” — as acts or modes of thinking rather than tertia inspected in the world’s place — was pressed by Descartes’ own contemporaries; Arnauld is the locus classicus, and the Stanford entry on Descartes’ theory of ideas traces the direct-realist line from him forward. For the parallel Locke case see note 4 below. Thomas Reid, Essays on the Intellectual Powers of Man (1785), is the source of the influential attribution of an “ideal theory” to Descartes, Locke, Berkeley, and Hume, and thus one of the reception’s principal transmitters; the nineteenth- and twentieth-century textbook traditions carried it, and the label veil of perception itself dates only to Jonathan Bennett in 1971 (see note 5). Nicolas Malebranche’s vision-in-God theory requires a separate distinction: intelligible ideas exist in God, whereas sensations modify the finite perceiver’s mind. His ideas therefore cannot simply be classified as private inner objects within that mind. See Dialogues on Metaphysics and on Religion (1688), Dialogue I, sec. 10, and Dialogue IV, sec. 10, which distinguish intelligible extension, material bodies, and sensations. A genealogy of the picture does not require Descartes guilty; it requires only that the picture got transmitted, and it plainly did. ↩
2. On the Cartesian representational theory of perception, see especially René Descartes, Meditations on First Philosophy, trans. John Cottingham (Cambridge: Cambridge University Press, 1996 [1641]), Third and Sixth Meditations; and Optics, in Discourse on Method, Optics, Geometry, and Meteorology, trans. Paul J. Olscamp (Indianapolis: Hackett, 2001). Discourse IV of the Optics resists the assumption that the mind must contemplate resembling images transmitted to the brain. Signs and words can stimulate thought without resemblance; even pictorial representation depends on selective resemblance and systematic departures from it. This separates representation from copying rather than establishing that representation fails. The remaining content-fixing question concerns what makes the relevant activity represent one object or property rather than another. The Cartesian-circle objection and Descartes’ distinction between present clear perception and remembered demonstration appear in the Fourth Objections and Replies. See Gary Hatfield, “The Cognitive Faculties,” in The Cambridge History of Seventeenth-Century Philosophy, ed. Daniel Garber and Michael Ayers (Cambridge: Cambridge University Press, 1998), for helpful historical context. ↩
3. John Locke, An Essay Concerning Human Understanding, ed. Peter H. Nidditch (Oxford: Clarendon Press, 1975 [1689]), Book II, ch. viii, esp. secs. 9–10 and 15. Locke’s distinction between primary and secondary qualities has generated an enormous secondary literature. For the purposes of this book, the important point is its role in driving the inner-theater picture: by relocating secondary qualities from objects to perceivers, Locke widened the gulf between the world as physics describes it and the world as it appears. Jonathan Bennett, Locke, Berkeley, Hume: Central Themes (Oxford: Clarendon Press, 1971), remains a useful critical survey of the positions and their internal tensions. ↩
4. The reading that resists treating Locke’s “ideas” as a veil of interposed objects is developed by John W. Yolton, Perceptual Acquaintance from Descartes to Reid (Minneapolis: University of Minnesota Press, 1984), and Perception and Reality: A History from Descartes to Kant (Ithaca: Cornell University Press, 1996). On Yolton’s account Lockean ideas are acts or contents of perceiving — the having of which is the perceiving of the world — rather than tertia perceived in the world’s place. Michael Ayers, Locke, 2 vols. (London: Routledge, 1991), esp. vol. 1, Epistemology, should not be assimilated to that reading without qualification: Ayers resists the crude veil picture but finds in Locke an unresolved ambiguity, and titles a chapter of his own “Ideas as intentional acts and ideas as intentional objects.” The contested state of the exegesis is itself part of this section’s point. The contrary “veil of ideas” reading is the one entrenched in the textbook tradition and in the scientific vocabulary descended from it (see, e.g., Jonathan Bennett, Locke, Berkeley, Hume: Central Themes [Oxford: Clarendon Press, 1971]). That is why the book engages the veil-reading: not as a verdict on Locke exegesis, but as the picture that left the study for the laboratory. Whoever the historical Locke was, the interpretation his Essay set loose is the one this chapter traces. Its private sensory effects will return later, in newer dress, as qualia. ↩
5. The “veil of perception” label entered the literature with Jonathan Bennett, Locke, Berkeley, Hume: Central Themes (Oxford: Clarendon Press, 1971), esp. 69, and is standardly credited to him; its entrenchment as the standard shorthand for the epistemological predicament generated by indirect realism owes much to Howard Robinson, Perception (London: Routledge, 1994). It should be distinguished from the “veil of ideas” discussed in Locke scholarship, which refers to the mediating role of mental representations specifically in the context of Locke’s theory — the broader “veil of perception” captures the general predicament of any view that makes inner representations the immediate objects of awareness. For a direct critique, see Tim Crane, “Is There a Perceptual Relation?”, in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006). ↩
6. That sense-data need not be construed as mental images, and were not so construed by the most careful sense-datum theorists, is explicit in the founding statements of the view. Bertrand Russell takes sense-data to be objects of acquaintance whose status as mental or physical he holds open (The Problems of Philosophy [London: Williams and Norgate, 1912], chs. 1–2), and H. H. Price (Perception [London: Methuen, 1932]) and C. D. Broad (Scientific Thought [London: Kegan Paul, Trench, Trübner, 1923]) both insist that a sensum is not an image inspected in the head. What the chapter targets is therefore not the literal-picture caricature but the structural commitment common to all versions — the positing of an immediate, non-worldly object of awareness from which the external world is reached only by inference. The classic exposure of that structural move (as opposed to the cartoon) is J. L. Austin, Sense and Sensibilia, reconstructed by G. J. Warnock (Oxford: Oxford University Press, 1962), taken up in the next chapter. ↩
7. Anil Seth, Being You: A New Science of Consciousness (London: Faber & Faber, 2021), esp. chs. 4–6. Seth’s “controlled hallucination” formulation is a deliberately provocative way of expressing the predictive-processing framework associated with Karl Friston and, in earlier form, Hermann von Helmholtz’s account of perception as “unconscious inference.” The claim is not that ordinary perception is pathological but that the generative process underlying hallucination and veridical perception is the same — a point that feeds directly into the argument from hallucination taken up in the next chapter. For a more cautious philosophical treatment of what the predictive-processing framework does and does not entail about the objects of awareness, see Jakob Hohwy, The Predictive Mind (Oxford: Oxford University Press, 2013). The present book takes no settled position on the predictive-processing framework as neuroscience; it uses it to illustrate how naturally the Cartesian-Lockean picture maps onto the vocabulary of contemporary cognitive science — a mapping that any adequate philosophy of perception must confront. ↩
8. The claim that neuroscientific language inherits the Cartesian picture without acknowledging it is made carefully by Tim Crane, “The Intentional Structure of Consciousness,” in Consciousness: New Philosophical Perspectives, ed. Quentin Smith and Aleksandar Jokic (Oxford: Oxford University Press, 2003), and by Galen Strawson, “Real Direct Realism: Reflections on Perception,” in Phenomenal Qualities: Sense, Perception, and Consciousness, ed. Paul Coates and Sam Coleman (Oxford: Oxford University Press, 2015), 214–253, esp. secs. 1 and 9. Strawson’s point is that the neuroscientific fact that perception is causally mediated by brain processes does not entail that perception is epistemically indirect — that the immediate object of perceptual awareness is a brain state or model rather than the world. Causal mediation and epistemic directness are compatible; conflating them is the deepest mistake the inner-theater picture makes. ↩
9. The diagnostic point — that the phenomenal character of a remembering consists in how it represents an absent scene rather than in the properties of an inner picture — is Michael Tye’s; see Ten Problems of Consciousness (Cambridge, MA: MIT Press, 1995), ch. 1, esp. 4–5, where he observes that “in and of itself there is nothing it is like to remember that September 2 is the date on which I first fell in love.” (The transparency material in Consciousness, Color, and Content [Cambridge, MA: MIT Press, 2000] is ch. 3, not ch. 1; ch. 1 there is the knowledge-argument essay.) The distinction between episodic memory (the recall of a participated event, with a recoverable point of view) and semantic memory (the storehouse of facts, including facts about events never witnessed) traces to Endel Tulving, Elements of Episodic Memory (Oxford: Oxford University Press, 1983), who marks the asymmetry as one of recollective experience — autonoetic in the episodic case, noetic in the semantic — rather than as a flat presence or absence of phenomenal character; Tim Crane, The Mechanical Mind, 2nd ed. (London: Routledge, 2003), treats where the line falls as an empirical question common sense cannot settle. The main text needs only the weaker asymmetry: it is the participated-event case that invites the cinema metaphor. The full representational treatment of memory’s phenomenal character belongs to Part Two; it is flagged here only to mark that the inner-reel picture is a special case of the inner-theater picture and falls to the same correction. ↩
10. The question raised here in plain terms — whether two people who agree in all their color words might nonetheless differ in the experiences those words track — is the inverted-spectrum hypothesis. The thought traces to Locke, who entertained the possibility that “the same Object should produce in several Men’s Minds different Ideas at the same time” (An Essay Concerning Human Understanding, ed. Peter H. Nidditch [Oxford: Clarendon, 1975], II.xxxii.15). Sydney Shoemaker (“The Inverted Spectrum,” Journal of Philosophy 79, no. 7 [1982]: 357–81) and Ned Block (“Inverted Earth,” Philosophical Perspectives 4 [1990]: 53–79) sharpened the thought into the canonical challenge to functionalism: if spectrum inversion is coherent while leaving all functional organization intact, then functional role cannot exhaust phenomenal character. David Chalmers treats inversion as one of his central cases for the explanatory gap (“Facing Up to the Problem of Consciousness,” Journal of Consciousness Studies 2, no. 3 [1995]: 200–219). The strong-representationalist reply, developed in Chapter 6, denies the coherence of the case rather than conceding and then patching it: if phenomenal character consists in representational content of the right kind, then an inversion that preserves every world-tracking relation has not in fact inverted anything. The main text deliberately states the puzzle in pre-theoretical form; the apparatus arrives later. ↩
11. David J. Chalmers, “Facing Up to the Problem of Consciousness,” Journal of Consciousness Studies 2, no. 3 (1995): 200–219, esp. 200–203. Chalmers distinguishes problems susceptible to functional or neural explanation—discrimination, integration, access, report, attention, and behavioral control—from the further question why any such processing comes with experience. The main text states that explanatory demand before later chapters dispute Chalmers’s inference from the apparent gap to an ontological remainder. ↩