The last chapter ended on a question: if the felt character of experience belongs to no inner object, what does it belong to?

Two rival pictures have to be cleared away before any answer goes through. One, qualia realism, locates phenomenal character in intrinsic, non-representational qualities of mental states — inner items representation alone cannot capture. The other, the inner-theater picture, makes the immediate object of perceptual awareness a private representation the self watches, with the world inferred from it. The two often get run together; one of this chapter’s tasks is keeping them apart.

The thesis I will defend, against both: phenomenal character consists in representational content of the right kind. Seeing red differs from seeing blue because each experience presents the world differently, not because each carries a different coat of inner color. This is strong representationalism in its identity-claim form — the position Tye and Dretske defend.

---

6.1 The Older Observation: Moore, Brentano, and the Diaphanous Mind

The previous chapter did the demolition. Experience easily becomes a third thing between observer and world, yet attention to that supposed exhibit slides through it and lands on the world, as Harman found when he sent Eloise to introspect her seeing of a tree and she came back with nothing but the tree.¹²

The observation is old. G. E. Moore, in 1903, called the experience-relation diaphanous — see-through — noting that when we try to introspect a sensation of blue, “the other element is as if it were diaphanous.”³ Aquinas had treated the form in the mind as what cognition works through, and Brentano made that outward directedness the mark of the mental. Neither they nor Moore held the identity claim defended here; the family resemblance lies in where they looked to specify the act. What an experience is like turns on what it is of. So this chapter asks what experience actually picks out, once the noun stops pretending to name a thing over and above the world it gives us.

---

6.2 Two Arguments, One Target

Attend to your visual experience of a tomato and you expect to find some inner item: a mental image, a private color-patch, a shimmer of redness behind the eyes. Chapter 4 ran that exercise. Nothing turns up but the tomato.⁴

The observation bears on two distinct claims. The inner-theater picture puts a private representation between perceiver and tomato: a screen behind the eyes, watched by a Self, from which the world gets inferred. Qualia realism need not do that. It can treat phenomenal character as an intrinsic, non-representational property of the perceptual state — Block’s mental paint — while granting that the tomato remains the object of awareness.⁵⁶

Demolishing the theater therefore leaves the identity question open. Strong representationalism says phenomenal character consists in content of the right kind and nothing further; if mental paint does independent explanatory work, the claim fails. Section 6.3 gives the positive case, the inversion cases test it, and Chapter 7 follows the proposed residue into pain and mood.

---

6.3 The Positive Case: From Transparency to Content

The case has three stages. Each needs more than the stage before it supplies.

1. Transparency undermines the supposed inner exhibit. Introspection returns the tomato, wall, music, or headache as presented; it does not disclose a private object or coat of phenomenal paint.
2. The wider evidence supports representational determination. Changes in perception and attention can be measured against a fixed stimulus and compared with reports of changed appearance. These comparisons support an account of the phenomenal contrasts through the world-presenting profile, though no experiment measures that whole profile without theoretical assumptions.
3. Identity requires a further choice. Even explanatory exhaustion leaves a residue-free rival: the profile might ground or constitute phenomenal character rather than character consisting in that profile. The book chooses identity by theoretical economy, and keeps the choice defeasible.

Start with transparency. Michael Tye has spent thirty years refining the thought behind it: introspection provides our primary access to experience from within, yet systematically directs attention to what experience presents. If an inner object or intrinsic paint figured in experience’s basic structure, some access to it would be expected. Its persistent absence counts against both pictures and supports a world-directed account.⁷ It does not prove that every intrinsic feature has been found wanting; a hidden feature remains logically possible.

Tye separates the next move himself. Transparency first creates trouble for qualia realism. He then adds a further plausible hypothesis: the qualities experience presents fix or ground its phenomenal character. Even that hypothesis initially yields only dependence — what he calls weak property representationalism, a supervenience claim. Strong property representationalism makes the additional identity claim. In plain terms, “the feel never changes without the presented qualities changing” still differs from “the feel consists in those qualities.” The chapter must earn the stronger conclusion rather than slide into it.

Nor does transparency decide among every physicalist theory. Higher-order and self-representational accounts explain consciousness by adding a representation of the first-order state, so they too enter the comparative contest.^7b The first-person evidence needs discipline. Dennett’s heterophenomenology takes subjects’ reports and discriminations seriously while refusing to treat their metaphysical interpretation as incorrigible.^7c A report that the tomato appears red constrains the theory. It does not inventory the state’s ultimate constituents.

The second stage asks which account best explains the wider pattern. Take a state already counted as conscious and borne by an identified subject. Its phenomenal character, the book proposes, consists in its complete world-presenting intentional profile: independently fixed content under a perceptual or bodily-affective mode, with determinacy, field structure, and attentional organization.⁸ This profile exceeds a bare object-label. It records whether the tomato appears red or green, clear or blurred, centered or peripheral, seen or imagined, commanding attention or fading behind the mug. Content-fixing history, the test for state consciousness, and the identification of the subject remain separate questions.

The question is how to specify that profile without quietly reading the answer off the feel. Blur gives us a place to start: the printed edge stays put while the eye’s grip on it changes. Before asking how the page looks, we can ask which differences in its lettering the visual system can still resolve. The worked comparison below separates those measurements from the phenomenal report. Even a perfect match between the two would support determination, not yet identity: a grounding theory could explain why they always travel together.

Kind and Stoljar press exactly where this move remains vulnerable. Kind distinguishes weak transparency — we attend to experience by attending to what it presents — from the stronger claim that experience has no intrinsic feature beyond presentation. Stoljar likewise argues that diaphanousness underdetermines representationalism. A qualia realist can grant the tomato exercise and retain an unseen property of the state.⁹ Blur, double vision, and afterimages can even seem to offer indirect access to that property.

Double vision, on this account, changes how many objects seem present or where they seem to lie. That proposal needs the same independent checks as blur; naming a change in content does not establish one.

Kind shows that we can attend to how the world gets presented. She does not thereby isolate a property outside the presenting. But the intentionalist owes more than a new description of each apparent exception. Content must earn accuracy conditions through tracking, producer-consumer organization, and world-involving history; mode, determinacy, field structure, and attention need constraints not supplied by the phenomenal report alone. Otherwise “the complete profile” would mean only whatever made the theory come out true.

The strongest tests run in both directions. Preserve state consciousness, subject, and the independently specified profile while phenomenal character stably changes: that breaks representational determination. Change the complete profile while character demonstrably stays fixed: that challenges the stronger identity, since determination alone can allow different profiles to yield the same character. The tests can guide experiments. Stimulus control, discrimination, generalization, error patterns, and intervention can constrain the profile. Recurrence, flexible reuse, correction, and cross-system influence can constrain state consciousness. Perturbation can help identify the bearer. First-person judgment remains evidence rather than an infallible reading of inner furniture.

Explanatory exhaustion still falls short of identity. A residue-free critic may grant every result above and say that the profile grounds or constitutes phenomenal character. That position adds no paint, and transparency gives no reason to reject it. Section 6.5.2 therefore makes the third-stage argument openly: absent an independently motivated causal, modal, or persistence difference, identity carries one property where grounding carries that property plus an asymmetric dependence relation. Economy favors identity for now.

Chalmers’s pressure remains. Transparency closes neither the epistemic gap nor the modal gap; conceiving a profile without its character may seem to separate them even when no actual case does. Chapter 9 asks what conceivability establishes, and Chapter 11 asks why two concepts of one fact can resist an a priori bridge. This chapter ranks identity as the best explanation of the actual cases without claiming that introspection has settled either question.

The rest of the book spends the claim with that status intact. Later verdicts about a historyless duplicate or an artificial system remain conditional on the identity defended here; they do not return as independent proof. A reader can reject wide phenomenal content while accepting much of the argument about perception and meaning. The disagreement then remains exactly where it belongs: at this hinge.

6.3.1 The Paint’s Home Turf: Blur, Afterimage, Attention, Imagery

Tye reports Block’s challenge through a cinema screen: seeing a blurred picture clearly can differ from seeing a sharp picture through unfocused eyes. Can a theory of content capture the difference?¹⁰ The everyday version needs only a page and reading glasses. Without the correction, nearby lettering can go soft while you know perfectly well that the print has not changed. You reach for your glasses rather than for a sharper book.

That knowledge does not settle what your visual state presents. You can know the Müller-Lyer lines have equal lengths while they keep looking unequal. So the intentionalist cannot infer “no visual error” merely from your correct belief about the print. The proposal to test is more specific: in this case, blur consists in a loss of determinacy about the contours, rather than an added smear or a positive presentation of fuzzy ink.

Consider a controlled comparison, not a report of a completed experiment. Keep the same viewer, page, distance, lighting, and fixation point; vary the optical correction. First establish, with correction in place, which small shifts of a letter’s edge the viewer can distinguish. Then repeat with the altered optics, varying edge position and spacing across trials rather than asking whether anything feels blurred. Include trials where the ink itself grades gradually into the paper, so a genuine diffuse boundary has its own measurable profile. Randomize positions and responses to check that guessing or a preference for one answer does not supply the effect.

This gives “less determinate” some content before the phenomenal verdict arrives. Suppose the viewer can still locate the letter but cannot distinguish nearby placements of its edges. The state then supports a coarser spatial claim. Its proposed target remains the printed contour: moving that contour changes the responses, and restoring focus restores the finer discrimination. The relevant error would put the contour outside the region presented. Silence about its exact position need introduce no error. Visual mode remains fixed by the ongoing optical input and its role in guiding inspection, not by calling the result “visual.” Fixation constrains the field, while performance on the same location task helps constrain attention. None of this proves that every aspect of attention stayed fixed, or that discrimination exhausts experience. It prevents those terms from doing whatever we please.

Now ask how the page looks. Suppose reports of softness track the independently measured loss of edge detail and both reverse when focus returns. The report contributes the phenomenal side of the comparison; it did not define the loss we measured. The intentionalist has explained a contrast through the visual system’s changed presentation of the page. But if a diffuse-ink condition and an altered-optics condition match on these measures while looking different, that is a problem to investigate, not permission to announce an unmeasured difference in “mode.” The complete-profile claim would survive only if some further difference earned independent support.

Block’s physicalist can accept the whole pattern. The changed optics alter a neural feature whose intrinsic character constitutes the blur; its representational effects explain the diminished discrimination. No second screen or spectator enters that reply. A still leaner nonidentity rival can grant the intentionalist’s entire profile and say the feel obtains in virtue of it. Neither rival must predict a mismatch. The comparison therefore supports a constrained explanation of the contrast and weakens the original claim that content has nothing to say about blur. It does not uniquely select identity.

You may reasonably think I have explained what changes without yet explaining why it feels different. My answer identifies the altered presentation with the altered feel; it does not promise a further event after the edges lose their precision. That answer earns a preference when the same account explains the accuracy, errors, discriminations, and phenomenal contrast without a separately specified qualitative mechanism. Necessary grounding can match that success. Its proposed direction of dependence, rather than an imaginary canvas, supplies the remaining disagreement taken up in Section 6.5.2.

Afterimages and phosphenes bring the same question into cases without a matching colored object. An afterimage can appear to drift across a wall; a phosphene presents a spark without a light source there. If you take the patch for something on the wall, the account is straightforward: the visual state attributes a color at a location where nothing instantiates it. The representing, not some private patch, is illusory.

The harder case begins once you know perfectly well that you have an afterimage. Boghossian and Velleman put the point exactly: the patch appears there without appearing actually to exist there.¹¹ The identity claim should not answer by pretending you have made the simple mistake you just disavowed. Tye’s better reply distinguishes levels. Your nonconceptual visual state still presents something red and filmy before the background; your higher-level judgment represents that very presentation as unreal. The contents conflict, much as the Müller-Lyer lines can keep looking unequal after you have measured them. Recognition corrects the judgment without repainting the lower-level seeing. Redness and apparent location therefore remain in the visual content, while the subject’s assessment correctly denies any red thing there. No visual-field object has been discovered — only a fallible presentation and a judgment that catches it failing. Remove the object and you do not strand the paint; you get a representation with empty hands, plus a perceiver who knows they are empty. The patch itself will fade before the argument does.

Attention supplies an actual experimental comparison. Marisa Carrasco and her colleagues studied the question with two patches of tilted stripes, briefly shown on opposite sides of a fixation point.^10a A preceding dot drew attention to one location without predicting either patch’s contrast or tilt. Observers reported the tilt of whichever patch looked higher in contrast. Cueing changed which physical contrasts they judged equal; within part of the tested range, it also improved their accuracy in judging tilt. The experiment thus separated a measurable stimulus property, an intervention on attention, an appearance judgment, and performance with a right or wrong answer.

The controls matter. Lengthening the interval after the cue removed the appearance effect. Asking which patch looked lower in contrast reversed the choice, rather than leaving observers mechanically choosing the cued patch. These results count against a simple preference for whatever received the cue. They do not settle every possible decision-bias account. Better tilt discrimination constrains processing without by itself proving that contrast looked stronger; the appearance judgment supplies evidence of that further change. If attention really makes unchanged contrast look higher, the account must accept a small misrepresentation. If attention changes only the decision, this experiment supplies no phenomenal contrast for identity to explain. Either way, “attention changed” cannot serve as a blank check on content.

The imagined tomato differs along another axis. The image comes gappy, unstable, thin where the percept runs dense (try counting the seeds in your imagined tomato). It arrives under an imaginative mode, summoned and dismissed at pleasure. Perception answers to the world whether you like it or not. Both states can influence belief, but under different warrants and through different causal routes. Same object, perhaps; nowhere near the same profile. Across blur, afterimage, attention, and imagery, the proposed paint has yet to explain a stable phenomenal contrast the independently specified profile misses.

---

6.4 Three Objections

A family of classical objections presses on the identity claim, each naming a feature of experience that content supposedly cannot capture. Three belong here — the cases that grant the experience a worldly object and dispute only whether its content fixes its felt character. I take them in order, meeting each here and sending the fuller apparatus to the back matter where a case earns it.

6.4.1 Mary’s Room

Frank Jackson’s Mary’s Room, the most famous objection of the family, now has a chapter of its own.¹² Here I record only how the identity claim bears on it. Mary knows every physical fact about color vision from inside a black-and-white room, then walks out and sees a tomato. She learns something — but what she gains is a new way of representing a property she already knew, the perceptual mode of presentation that one acquires only by undergoing the state.¹³ Whether that concession costs physicalism anything is Chapter 8’s question, and Chapter 8 answers it with more care than a summary can.¹⁴

6.4.2 The Inverted Spectrum

The family’s oldest member comes next: the inverted spectrum. Chapter 2 filed a folk version of this case among the puzzles the theater generates — a private inner display, swapped behind the scenes, unstatable once the theater falls. The version that survives the demolition needs no inner items; it states its challenge in the identity claim’s own vocabulary. Suppose your color experience runs systematically inverted relative to mine (you see red where I see green, and the reverse), yet we were both raised to call the tomato “red,” so nothing in what we say or do betrays the difference. If that is coherent, phenomenal character cannot consist entirely in representational content, since our experiences would represent the same properties while feeling different. You have probably asked it yourself, as a child if not since (do you see my red?), and the pull deserves respect: the difference, if real, seems locked away where no test could reach it. The reply, in plain language: the color solid (the three-dimensional arrangement of the colors by hue, brightness, and saturation) runs lopsided, so a full swap that preserved every similarity and every brightness relation has no room to hide — the geometry betrays it.^14a

6.4.3 Inverted Earth

The inverted spectrum imagined two people inside one world. Inverted Earth moves one of them to another world, and it is the harder case. The scenario comes from Block, who built it as the version representationalism cannot absorb.¹⁵

Picture it. One night, without your consent, two things are done to you at once. Fine color-inverting lenses are slipped behind your eyes, and you are carried to Inverted Earth, a planet exactly like ours except in its colors. The sky there reflects what we would call yellow. Ripe tomatoes throw back the reflectance of grass. The locals, raised on an inverted vocabulary, call that sky “blue” and those tomatoes “red.” The two inversions cancel. You wake and notice nothing wrong. The sky overhead looks the way the sky has always looked to you, you call it “blue” as your new neighbors do, and your word now picks out the local sky’s yellow-reflectance, since that is the property your states have started tracking in the community you now belong to. Let years pass. The externalism this book runs everywhere else says your color experience comes to represent the local reflectances: your unchanged experience of the sky now carries inverted content. The felt character held still; the representational content turned over. If the story is coherent, the identity claim is false, for it gives us two states alike in character and opposite in content.

The relational structure that answered the inverted spectrum cannot help here. There the reply ran: genuine inversion would have to flip the entire color solid, and that might not be coherent. On Inverted Earth the whole solid travels intact with your phenomenology; nothing internal to your color experience has moved. What has moved is the worldly content — precisely the thing the identity claim says the character consists in.

Block presses the case home through memory. You remember the childhood sky and notice no difference between that blue and this one. So you seem to hold the very thing the identity claim denies: a felt character that sat still while its content moved.

A structural representationalist can accept that description without reviving mental paint. She can hold that the complete synchronic profile, including color-space relations, discriminations, expectations, attention, memory dispositions, and rational control, fixes how the sky looks, while causal history fixes the state’s wider reference. On that account the present organization continues to present the same structural color even after the wide address changes. Transparency cannot decide between this view and phenomenal externalism. Both reject an inner colored item, and both send attention outward; they disagree over whether historically fixed reference belongs to phenomenal character.

The book therefore cannot use Inverted Earth as independent evidence for its identity claim. Its reply remains conditional on the wide-content theory defended across this chapter and Chapter 15. On that account, the external re-tuning that moves perception also moves what memory presents: the memory keeps its old object (the image stays of that earthly sky), while the color it represents the sky as having goes local. Character and content then move together too slowly for the subject to detect, because the standards of comparison move with them. The structural rival denies that the wide address enters character at all. No further introspective gauge settles the difference, and no mental paint follows merely from taking that rival side.

The book still chooses the wide account, for the abductive reasons stated in Section 6.3. One independently specified content relation fixes accuracy, error, rational role, and phenomenal contrast. The structural rival, meanwhile, owes a principled boundary: the point at which wide content stops contributing to the presentation. The preference remains defeasible. Its cost is exactly what it appears to be: phenomenal character changes without detection during the long re-anchoring. That verdict is a consequence of the identity thesis, not evidence for it. First-person access gives you the present appearance; it does not issue a timestamped certificate of metaphysical sameness. Section 11.3 prices the consequence for acquaintance. The remaining underdetermination is genuine: transparency alone cannot choose between the wide and structural accounts.

Two further objections press here, and they are the hardest the identity claim faces: bodily sensations like pain, and moods, experiences that seem to carry felt character while being about nothing at all. They earn a chapter of their own. Chapter 7, The Hard Cases, shows the toothache turning out to represent the body and the morning’s dread the world entire — each, on inspection, about something after all.

---

6.5 What the Thesis Commits Us To

Strong representationalism does not say consciousness amounts to nothing or that felt life comes as an illusion. It says the felt dimension consists in the complete world-presenting profile of a conscious state. The redness you see remains real: your visual system presents the tomato’s surface as having a property it has, or, in error, one it lacks. No private coat of color need accompany that presentation.

It also commits the book to intentionalist direct realism. When you see the tomato, awareness reaches the tomato immediately; no representation stands in its place. Content interposes no object; it states how the world stands, and in successful perception the current object satisfies it through the right causal-historical connection. Searle supplies one influential route to that compatibility, making the perceived object part of the experience’s causally self-reflexive conditions of satisfaction. The book takes the compatibility lesson without taking over his precise machinery.¹⁶

That commitment needs more than the slogan that content is transparent. Three things must stay apart. By vehicle, I mean the neural episode that does the representing, produced through the causal chain from light at the retina to activity along the visual pathways. The content under a perceptual mode specifies how that state presents the world and what would make the presentation accurate or mistaken. Successful seeing takes the tomato itself as its intentional object, the thing the experience presents; a pure hallucination has no current mind-independent object in that role. These three enter one account, but they do not change places. Neither vehicle nor content becomes something seen merely because the theory needs it to explain seeing.¹⁷

Hallucination tests the distinction. Suppose a hallucination and a perception share the same complete content and perceptual mode. One conclusion follows: content alone does not suffice for seeing, since it can occur unsatisfied. A second conclusion does not follow: that when a tomato stands before you, you first perceive the content and then infer the tomato from it. In the good case, the tomato causes and satisfies the presentation under the right world-involving relation; it is what the experience is of. In the bad case, the same presentation runs unsatisfied. Shared content explains the matched character and rational pull. Satisfaction and current coupling explain why only one episode counts as seeing. No private item enters as the object of awareness or as a premise from which the subject reasons outward.

Relationalists press the hardest version of this. Bill Brewer and M. G. F. Martin both argue that the mind-independent object partly constitutes the experience, so that nothing content-like could have occurred unchanged without it.¹⁸ Charles Travis presses from another side, arguing that experience issues no verdict determinate enough for the world to satisfy, and John Campbell that acquaintance with an object must come first, before any content about it. The reply is that common content explains the matched character of perception and hallucination, while satisfaction and current coupling explain why only one of them counts as seeing.

Perceptual content can remain nonconceptual and more or less determinate; the theory does not make experience whisper a sentence. Its accuracy conditions come from independently measurable work. Producer-consumer organization tracks something; learning corrects it; some variations alter discrimination and control; and one current object, not another, stands in the right world-involving relation to the state. A subject need not formulate those conditions any more than a cat needs the concept trajectory to misjudge a jump. The causal-historical coupling fixes the token object; it does not borrow a demonstrative concept from the perceiver. Chapter 15 owes the full account. If matched history and control leave no stable correction pattern that distinguishes the candidate contents, concede Travis’s silence at that grain. If demonstrative reference survives systematic breaks in current object-coupling, Campbell’s priority claim gains ground.

Block’s physicalist paint proposes a different bargain. It proposes an intrinsic physical property of the vehicle as the feel itself, not as a second screen and not merely as another cause of report or action. Content can explain accuracy and rational direction while paint constitutes phenomenal character; calling that arrangement duplication would assume the identity now at issue. The present advantage for identity is narrower. Across the cases examined here, changes measured independently in content, mode, determinacy, field structure, and attention track the phenomenal contrasts. No vehicle property identified independently of the profile predicts a contrast the profile misses. Until one does, paint adds a second property family and a coordination relation without improving the account. A neural intervention that changed character while the complete profile stayed fixed would supply the evidence Block needs.

The book therefore takes identity as the leaner live theory. One independently constrained profile covers transparency, error, hallucination’s positive character, rational role, and the measured changes that track determinacy and attention. A rival would gain ground if the same representational profile occurred with a stable phenomenal difference, or if phenomenal character varied independently of content, mode, determinacy, field structure, and attention. The cases examined here supply neither result. Identity leads, defeasibly, because it explains the present evidence with fewer independent commitments — not because transparency has yielded a theorem.

The book pairs this thesis with a separately defended realism about color. The tomato is red, and its redness belongs to it — a type of surface spectral reflectance, genuinely there to be reached. Strong representationalism alone does not tell us which properties colors are. Error theory, dispositionalism, primitivism, and perceptual variation all challenge the reflectance account.^18a On the combination defended here, the felt redness consists in presenting the surface as carrying a reflectance-type it really carries. One world, color included.

One debt comes due here. The directness just defended runs entirely through aboutness, and this chapter leaves that aboutness a primitive, exactly where its ally Searle leaves it. Chapter 15 develops a naturalistic answer: selection, learning, tracking, producer-consumer organization, and continuing world-involving use can make a state answerable to what it concerns, whether the mechanism evolved or was engineered. The state can misrepresent because the function or use that fixes its standard can fail. This supplies a general model with worked cases, not a completed account of every pain and mood.¹⁹

6.5.1 The Consciousness Profile

The phrase of the right kind now has to earn its keep. Brains carry many representations that nobody experiences. Early vision registers edges and motion; priming alters a later judgment; bodily regulation tracks temperature, glucose, and pressure. Sophisticated influence on behavior cannot by itself mark consciousness, since unconscious processing can discriminate, predict, and control. If every representation counted as felt, the identity claim would spread consciousness wherever information flowed. That would erase the distinction the theory was introduced to explain.

The phrase conceals four questions unless they are separated.

1. Content fixing: what does the state present, and what makes that presentation accurate or mistaken? World-involving selection, learning, tracking, and producer-consumer history belong here. Chapters 12–16 take up that debt.
2. Character determination: among conscious states, what makes this one look red rather than green, blurry rather than sharp, painful rather than neutral? Content under a perceptual or bodily-affective mode, with its determinacy, field structure, and attentional organization, supplies the proposed profile.
3. State consciousness: why does this content figure in experience rather than remain unconscious processing? The house proposal is a distributed access-and-control profile, specified below. Influence on behavior alone does not satisfy it.
4. Subject individuation: for which persisting system is the content poised and integrated? A subject cannot enter as a proprietor added to the machinery. It must be identified through the organization across which perception, memory, expectation, belief, desire, planning, correction, and action hang together. Chapter 17 develops that organization; Chapter 24 tests the criterion where candidate systems are nested or distributed.

The identity relation adds no fifth ingredient. It says that, within a content-bearing state already counted as conscious and borne by an identified subject, phenomenal character consists in the complete world-presenting intentional profile; no further paint waits to be attached. The remaining empirical questions ask which architectures fix content, which states meet the consciousness profile, and which persisting control organization bears them.

Failures at one level need not spread to the others. A conscious state outside the proposed distributed profile would refute that account of state consciousness while leaving open what fixes its character. A mistaken subject boundary would force a better account of the bearer, not new mental paint. Conversely, a matched-profile phenomenal contrast would strike the identity claim even if the book had correctly identified both conscious states and both subjects. Keeping the questions apart prevents one success from being spent twice.

Here is the positive proposal. A content-bearing state is conscious for a subject when it participates, as that content, in a recurrently sustained and selectively available pattern of influence across the subject’s integrated economy. Under suitable conditions it can guide attention, working memory, inference, planning, report where report is possible, and flexible action; feedback from those activities can preserve, revise, or displace it. No one output is necessary. Breadth, flexible reuse, recurrence, and two-way correction do the discriminating work together. A subliminal prime that bends one response, or a regulator that adjusts one variable, lacks the profile even though it affects behavior.

The proposal replaces a doorway with a causal pattern. A state can acquire greater or lesser reach, stability, and control; borderline cases should therefore be expected. Global broadcast, local recurrence, higher-order monitoring, and predictive organization become hypotheses about how physical systems realize parts of the profile, not rival definitions among which the book refuses to choose. Dennett’s Multiple Drafts model supplies the right architectural lesson: contents undergo distributed revision and some gain what he later called “fame in the brain” by influencing memory, report, reasoning, and action, with no final presentation to a witness.^19a His denial of intrinsic qualia does not amount to a denial that pain or experience occurs. The book shares that refusal of mental paint while defending its own identity claim about the complete world-presenting profile.

Tye’s 1995 framework, PANIC — Poised, Abstract, Nonconceptual, Intentional Content — gives conscious content four conditions.²⁰

1. Poised content stands available for use in forming beliefs or desires.
2. Abstract content concerns property types rather than particular tokens.
3. Nonconceptual content carries a finer grain than the concepts the subject commands.
4. Intentional content reaches toward properties of the world or body.

The last three conditions specify the form and target of perceptual presentation. Poise specifies its place in the subject’s economy. It requires no current attention, verbal report, or actual judgment. A headache can shape the day before its sufferer names it. A crowded scene can look richer than the few items a subject manages to report. Poise instead names the standing availability at the center of the distributed profile. PANIC helped isolate that feature; it does not by itself fix the content’s worldly address, identify the persisting subject, or specify the recurrence and cross-system correction that make availability more than a latent disposition.

That profile rules out two easy counterfeits. A thermoregulatory mechanism can register the coffee’s heat and adjust a hand without presenting the heat within a wider cognitive economy. A conceptual thought can enter reasoning while lacking the fine-grained perceptual mode of feeling warmth. A subject, here, is not whatever already has experience. It is the relatively maximal persisting control organization within which content remains available across memory, expectation, belief, desire, correction, planning, and action. Competing goals get settled there. Altering the content alters later system-level control. Chapter 17 develops that nonphenomenal test and Chapter 24 applies it to nested and distributed systems.

Two pressures test that proposal: Ned Block’s overflow argument, which reports a visual field richer than anything a subject can access in time to report it, and the live contest among global-workspace, recurrent-processing, and higher-order accounts over how the profile gets realized. Overflow defeats any theory that equates consciousness with actual report; it does not defeat counterfactual availability across a wider economy. The house proposal remains defeasible. Find a stable conscious state that lacks the distributed profile, or an unconscious state that matches it in content-sensitive breadth, recurrence, flexible control, and correction, and the state-consciousness account fails even if the identity claim survives.

6.5.2 The Identity Adopted

On the cumulative case, the book adopts the identity claim and the distributed profile as defeasible accounts of phenomenal character and state consciousness. Take a conscious state borne by an identified subject. Its phenomenal character consists in the state’s complete world-presenting profile: independently fixed content under a perceptual or bodily-affective mode, with determinacy, field structure, and attentional organization. Content-fixing history supplies the address; distributed access and control mark the domain; integrated persistence identifies the bearer. These claims constrain the identity without becoming extra ingredients in phenomenal character.²¹

The claim goes beyond correlation, supervenience, grounding, or constitution. Like the identity of water with H₂O, it would hold necessarily if true while remaining knowable only a posteriori. The analogy gives the logical form, not evidence that this instance must succeed.^21a

The strongest remaining rival grants the entire explanatory result without adding mental paint: the profile explains every phenomenal contrast, but grounds or constitutes phenomenal character rather than exhausting it.^21b Parsimony against inner objects cannot answer that view. The empirical evidence underdetermines the last step.

The grounding rival can make a stronger argument than “it still seems different.” The optical and cognitive organization explains why the page looks blurred, she says; the blur does not explain why that organization obtains. Her theory preserves that explanatory direction while granting necessary covariation and adding no independent mental paint. Grounding need not come with different possible-world or persistence conditions. Its purpose is precisely to distinguish what depends on what even where both must occur together.

The book’s reply separates causal explanation from metaphysical priority. Changed optics can explain the onset of one state, and a detailed account of its operation can explain a contrast, even if the state has two descriptions. The causal direction from removing the glasses to altered presentation does not establish a second direction, from that presentation to a distinct phenomenal fact. Nor does our ability to understand one description first establish that what it describes comes first in nature. Grounding could earn that further direction through an independently supported account of what phenomenal character requires. The blur and attention evidence has not supplied one.

The preference for identity therefore concerns explanatory commitments, not a census of invisible objects. Both accounts carry the same demanding empirical profile and the same unexplained necessity; grounding additionally asks us to accept a metaphysical direction that the present comparisons do not distinguish from an order of explanation. I prefer to identify the feel with the presentation that already accounts for its contrasts. A matched-profile phenomenal difference would defeat representational determination; an established modal separation would defeat identity; an independently motivated grounding asymmetry would remove this economy advantage. The book chooses the leaner account while leaving the eraser on the desk.

---

6.6 The Thesis and the Hard Problem

Transparency closes neither the epistemic gap nor the modal gap. Chalmers’s question — why should physical activity feel like anything at all? — survives the disappearance of an inner screen, and Chapter 11 meets it directly.²² The present chapter has defended a conditional claim: phenomenal character consists in the representational profile of the right kind. It has not made that identity obvious from the armchair or ruled out every conceivable separation.

If the identity holds, however, asking why the profile additionally produces a feel double-counts one event. The remaining difficulty concerns our concepts and access: neural organization described from outside and lived experience undergone from within disclose the same process without yielding an a priori bridge. The necessity therefore carries a real intellectual cost. It does not require a second property or law in nature.

The older picture makes that cost look like a production problem. Physical processing occupies one column, private experience another, and the philosopher asks how the first could generate the second. The identity claim refuses the second entry: once the relevant physical activity takes the form of poised, world-directed presentation, phenomenal character consists in that activity’s complete profile. What remains hard concerns why nature contains such organized activity, how brains realize it, where its conscious range begins, and why our phenomenal and physical concepts meet so reluctantly. None of those questions disappears. They no longer require an ontological bridge between two properties already counted separately.

The same separation disciplines the machine question. We must ask what a system’s states present, whether any meet the distributed state-consciousness profile, which persisting organization bears them, and what phenomenal character consists in for any conscious state thereby identified. Nothing in those questions demands carbon or a mammalian silhouette.

Part Four inherits the dependency rather than escaping it. Its conclusions about content and understanding rest on history, use, and integration; its consciousness corollary rests here. A stable phenomenal difference between fully matched conscious profiles would defeat the identity claim and that corollary together. Programhood supplies no substitute argument.

---

Chapter Summary

The claim. Once a state counts as conscious and an identified subject bears it, phenomenal character consists in its complete world-presenting profile. This profile includes independently fixed content under a perceptual or bodily-affective mode, with determinacy, field structure, and attention. No inner paint remains beyond it.

Where the argument stands. Blur and attention show how stimulus, processing, and appearance can be compared without defining every variable by the feel. They support the representational account but do not choose uniquely between identity and necessary grounding; the preference for identity adds a defeasible judgment about explanation. Inverted Earth separately tests whether historically wide reference enters phenomenal character: rejecting that extension need not restore mental paint or defeat every representational account.

What’s left open. Causal history and producer-consumer organization can fix content where a world-involving history privileges a correctness condition, though its exact grain may remain underdetermined. The distributed access-and-control profile marks state consciousness; competing theories propose its physical realization. Integrated persistence identifies the subject, and only then does identity specify character. A change in feel between matched conscious profiles would defeat the identity; a matched conscious/unconscious pair would defeat the state-consciousness account. Borderlines and bearer boundaries remain questions for science rather than hidden additions to content.

The hand-off. Pain and mood seem to offer mental paint its safest refuge. The next chapter asks what they present. Its answer will decide whether that refuge has walls.

---

Notes

1. Merriam-Webster’s Collegiate Dictionary, 11th ed. (Springfield, MA: Merriam-Webster, 2003), s.v. “experience” — the same entry the preceding chapter quotes. The grammar I describe (experience as a noun that takes adjectives and gets had and undergone) runs across the entry’s senses and across the corresponding entry in the Oxford English Dictionary (3rd edition). ↩
2. Gilbert Harman, “The Intrinsic Quality of Experience,” Philosophical Perspectives 4 (1990): 31–52, esp. 39. The Eloise-and-the-tree example sits at the heart of Harman’s transparency argument; the paper is short and worth reading whole, especially for its move from the introspective datum to the negative thesis about intrinsic non-representational properties of experience. ↩
3. G. E. Moore, “The Refutation of Idealism,” Mind 12 (1903): 433–453. Moore develops the transparency observation at 446 and calls sensation “diaphanous” at 450. The Aquinas comparison is deliberately limited: Summa theologiae I, q. 85, a. 2 treats the intelligible species as that by which the intellect understands, with the external thing as its primary object; see Robert Pasnau, Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997). That is a family resemblance about cognitive direction, not an anticipation of representational identity. For contemporary scrutiny of transparency, see Amy Kind, “What’s So Transparent About Transparency?,” Philosophical Studies 115 (2003): 225–244, and Daniel Stoljar, “The Argument from Diaphanousness,” in New Essays in the Philosophy of Language and Mind, ed. Maite Ezcurdia, Robert J. Stainton, and Christopher Viger, Canadian Journal of Philosophy, Supplementary Volume 30 (Calgary: University of Calgary Press, 2004), 341–390. ↩
4. The transparency datum is canonically introduced in Harman, “The Intrinsic Quality of Experience”; refined in Michael Tye, Ten Problems of Consciousness (Cambridge, MA: MIT Press, 1995); developed further in Alex Byrne, “Intentionalism Defended,” Philosophical Review 110 (2001): 199–240. For a recent overview, see Tim Crane and Craig French, “The Problem of Perception,” Stanford Encyclopedia of Philosophy, Fall 2021 edition. ↩
5. Ned 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, esp. 170–73. Block’s view differs from classical sense-data theories: mental paint counts as a property of mental states, not as the object of an inner perceptual relation. The distinction matters for Section 6.2’s argument that qualia realism and the inner-theater picture are separable targets. ↩
6. Block, “Mental Paint,” sections 1–3, makes the distinction between qualia-as-properties-of-states and qualia-as-perceived-inner-objects explicit. The point underwrites Section 6.2’s claim that a sophisticated qualia realist can concede Argument Two without conceding Argument One. For the position the book ultimately rejects but acknowledges as the strongest qualia-realist option, see also Galen Strawson, “Real Materialism,” in Real Materialism and Other Essays (Oxford: Clarendon Press, 2008), 19–51. ↩
7. Michael Tye, Consciousness, Color, and Content (Cambridge, MA: MIT Press, 2000), ch. 3; and Tye, “Transparency, Qualia Realism and Representationalism,” Philosophical Studies 170, no. 1 (2014): 39–57, https://doi.org/10.1007/s11098-013-0177-8. The later essay makes the sequence explicit. Transparency creates trouble for qualia realism; the claim that represented qualities fix or ground phenomenal character enters as a “further plausible hypothesis.” Tye distinguishes weak property representationalism, a supervenience thesis, from strong property representationalism, an identity claim. The main text preserves those steps rather than treating transparency as a shortcut to identity. ↩
8. Higher-order and self-representational theories offer physicalist alternatives to the first-order account of state consciousness. Higher-order theories require a suitable representation of oneself as being in a first-order state: David Rosenthal, Consciousness and Mind (Oxford: Clarendon Press, 2005); Hakwan Lau and David Rosenthal, “Empirical Support for Higher-Order Theories of Conscious Awareness,” Trends in Cognitive Sciences 15, no. 8 (2011): 365–373, esp. 365–367, doi:10.1016/j.tics.2011.05.009. The higher-order representation need not itself be conscious, so this requirement introduces neither a spectator nor an automatic regress. Uriah Kriegel’s self-representationalism instead places the world-directed and self-directed aspects within one conscious state: Subjective Consciousness: A Self-Representational Theory (Oxford: Clarendon Press, 2009). The substantive dispute concerns whether such self-representation is necessary and whether changing it can change consciousness while first-order content remains fixed. Lau and Rosenthal explicitly distinguish subjective awareness from objective task performance; matching one discrimination score does not establish matching the complete first-order organization. Misrepresentation supplies a further test: when the higher-order state characterizes its target incorrectly, or lacks a target, the theory must specify which representation determines conscious character. See Ned Block, “The Higher Order Approach to Consciousness Is Defunct,” Analysis 71, no. 3 (2011): 419–431, and Rosenthal’s reply in the same volume. Following the higher-order representation offers a substantive answer rather than a contradiction, but it requires an account of why its content fixes appearance while the first-order state’s content governs the relevant discrimination. The house proposal declines a necessary higher-order condition pending evidence that this additional explanatory role cannot be supplied by the distributed access-and-control profile. Its comparative claim remains defeasible; transparency alone does not decide between these theories. ↩
9. Daniel C. Dennett develops heterophenomenology in Consciousness Explained (Boston: Little, Brown, 1991), ch. 3, esp. 72–98. The method treats a subject’s sincere reports as data from which to construct the world as it appears to that subject, while withholding assent from the subject’s metaphysical interpretation of what the reports disclose. The present chapter adopts that neutrality about inference without importing Dennett’s entire account. His rejection of qualia targets purported additional intrinsic, private, ineffable properties; ch. 12 and the Appendix preserve content-bearing discriminative states and the reality of pain. The book’s own positive thesis identifies phenomenal character with the independently specified world-presenting profile, while treating reports about it as fallible evidence within the same causal order as the state reported. ↩
10. The thesis remains identity rather than correlation or mere covariance, but the inference to it is abductive and defeasible. The identified relatum for phenomenal-character determination is the independently specified world-presenting intentional profile — content under a perceptual or bodily-affective mode, with determinacy, field structure, and attentional organization. Three neighboring questions remain distinct. World-involving producer-consumer history fixes the content and its accuracy conditions; Chapter 15 develops that account. The distributed access-and-control profile determines which content-bearing states fall within the identity claim’s domain; global-workspace, recurrent-processing, higher-order, and other proposals bear on its realization. Integration and persistence identify the system that bears the state; Chapter 17 develops that account. None of these distinctions adds a proprietor, and none permits the phenomenal profile to expand after the fact merely to absorb a difference. ↩
11. The objection that the transparency datum underdetermines the representationalist conclusion — that one can grant the phenomenology and deny the thesis — is pressed by Amy Kind, “What’s So Transparent About Transparency?,” Philosophical Studies 115 (2003): 225–244, esp. 230–33. Kind distinguishes weak from strong transparency and also presses questions about transparency’s scope; the main text isolates the strength distinction relevant to the identity argument. See also Daniel Stoljar, “The Argument from Diaphanousness,” in New Essays in the Philosophy of Language and Mind, ed. Maite Ezcurdia, Robert J. Stainton, and Christopher Viger, Canadian Journal of Philosophy, Supplementary Volume 30 (Calgary: University of Calgary Press, 2004), 341–390. The representational reply to this underdetermination objection is developed by Michael Tye, “Transparency, Qualia Realism and Representationalism,” Philosophical Studies 170, no. 1 (2014): 39–57, esp. secs. 6–7. The book adds a constrained parsimony claim: once content has been fixed without reading it off the feel, a residue compatible with every report yet explanatory of none has no work left to do. ↩
12. Michael Tye, Consciousness, Color, and Content (Cambridge, MA: MIT Press, 2000), ch. 4, “Case 3: Blurred Vision,” 79–83, reports Block’s then-unpublished cinema-screen example at 80 and answers it through representational indeterminacy. The argument distinguishes representing an object’s boundaries as indefinite from failing to specify their precise location. At 82–83 Tye allows phenomenal indistinguishability in some suitably matched cases and develops a watercolor comparison, credited to Frank Jackson, for cases with a phenomenal difference. Block’s general defense of non-representational phenomenal properties appears in “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; the cinema example here is attributed through Tye, not located in that essay. The edge-discrimination comparison in the main text is the book’s proposed test of the determinacy account, not an experiment reported by either author. Discrimination constrains a candidate content attribution; it neither exhausts the conscious profile nor establishes phenomenal sameness. Tim Crane, “Introspection, Intentionality, and the Transparency of Experience,” Philosophical Topics 28, no. 2 (2000): 49–67, sec. 4, offers a different intentionalist treatment: blurry content can remain while mode distinguishes its presentation from the subject’s taking the world to be that way. His mode-based reply should not be identified with Tye’s indeterminacy reply. ↩
13. Marisa Carrasco, Sam Ling, and Sarah Read, “Attention Alters Appearance,” Nature Neuroscience 7 (2004): 308–313, esp. 309–312, doi:10.1038/nn1194, reports attention-induced changes in apparent contrast. The uninformative peripheral cue, contingent orientation task, and points of subjective equality are described at 309; the longer cue interval controls at 310–311; the lower-contrast judgment control at 311–312. Orientation accuracy improved within the dynamic range at both tested spatial frequencies in the low-contrast experiment. In the high-contrast experiment it improved at 4 cycles per degree, while performance at 2 cycles per degree was already at ceiling. The main text therefore distinguishes evidence of improved discrimination from the appearance judgments used to infer changed apparent contrast. Patrick Cavanagh, Sheng He, and James Intriligator, “Attentional Resolution: The Grain and Locus of Visual Awareness,” in Neuronal Basis and Psychological Aspects of Consciousness, ed. C. Taddei-Ferretti and C. Musio (Singapore: World Scientific, 1999), 41–52, esp. 41–43, argues that attention has a coarser spatial grain than early vision and thereby limits which registered detail reaches visual awareness. The interpretation remains contested: an apparent-contrast result can reflect represented contrast, representational determinacy, or post-perceptual decision effects. The intentionalist cannot infer the content from the reported feel alone; the empirical work constrains which independently specified representational variable changed. ↩
14. Paul Boghossian and David Velleman, “Colour as a Secondary Quality,” Mind 98, no. 389 (1989): 81–103, use the recognized afterimage to press a projectivist difficulty: the colored patch appears at a location without appearing to be a colored object located there. Michael Tye answers in Consciousness, Color, and Content (Cambridge, MA: MIT Press, 2000), ch. 4, by distinguishing the nonconceptual visual content, which remains illusory, from the subject’s higher-level conceptual assessment that no such object is present. The two contents conflict, as perceptual presentation and corrective judgment can conflict in the Müller-Lyer illusion. The main text follows this two-level reply rather than treating every recognized afterimage as a flat judgment that a colored object occupies the wall. ↩
15. Frank Jackson, “Epiphenomenal Qualia,” Philosophical Quarterly 32 (1982): 127–136, esp. 130, and “What Mary Didn’t Know,” Journal of Philosophy 83 (1986): 291–295, esp. 291–94. Jackson’s case has generated an enormous secondary literature; the canonical collection is Peter Ludlow, Yujin Nagasawa, and Daniel Stoljar, eds., There’s Something About Mary: Essays on Phenomenal Consciousness and Frank Jackson’s Knowledge Argument (Cambridge, MA: MIT Press, 2004). ↩
16. The reply pressed here — that Mary acquires a new representational route to a property she already knew, underwriting a new perceptual demonstrative thought (“this is what red looks like”) rather than any new fact — is developed in full, and set against its neighboring rivals (including the ability hypothesis), in Chapter 8. For a related anti-dualist route that grants Mary new subjective facts while denying that all facts must be physical, see Tim Crane, “Subjective Facts,” in Real Metaphysics: Essays in Honour of D. H. Mellor, ed. Hallvard Lillehammer and Gonzalo Rodríguez-Pereyra (London: Routledge, 2003), 68–83, esp. secs. 2 and 6. ↩
17. Jackson’s later reassessment appears in “Mind and Illusion,” in Minds and Persons, ed. Anthony O’Hear (Cambridge: Cambridge University Press, 2003), 251–271, esp. 251–52; Chapter 8 walks that change of mind as the chapter’s human spine. ↩
18. The inversion case is Sydney Shoemaker’s, “The Inverted Spectrum,” Journal of Philosophy 79, no. 7 (1982): 357–381. The representationalist reply through asymmetries in color space is developed in Michael Tye, Ten Problems of Consciousness (Cambridge, MA: MIT Press, 1995), ch. 7. The geometry weakens the inference from a describable inversion to a possible undetectable difference; it does not prove inversion impossible. ↩
19. The case originates in Ned Block, “Inverted Earth,” Philosophical Perspectives 4 (1990), esp. 62–64, and is restated and pressed in Block, “Mental Paint” (cited in n. 5 above). Block intends it as the scenario externalist representationalism cannot accommodate: by holding the subject’s phenomenal character fixed (the inverting lenses) while the wide content of his color states inverts (the move to a complementary-reflectance environment, followed by re-calibration), the case pulls phenomenal character and representational content apart — two states identical in felt character, opposite in what they represent. It is the diachronic, cross-world strengthening of the static inverted-spectrum case of Section 6.4.2 (developed in Appendix B.5): where the inverted spectrum can be met by the integrated relational structure of color experience, Inverted Earth leaves that structure untouched and dissociates content from character at the level of world-involving wide content, which is the level the identity claim occupies. ↩
20. John R. Searle, Seeing Things as They Are: A Theory of Perception (Oxford: Oxford University Press, 2015), esp. chs. 1 and 3, develops the diagnosis at book length; see also Searle, “The Philosophy of Perception and the Bad Argument,” in Wirklichkeit oder Konstruktion? Sprachtheoretische und interdisziplinäre Aspekte einer brisanten Alternative, ed. Ekkehard Felder and Andreas Gardt (Berlin: De Gruyter, 2018), 66–76. The Bad Argument is not merely the inference from sometimes noticing how things seem to always noticing only appearances. It combines the possibility of hallucination with a common analysis of veridical and hallucinatory experience, then mistakes the content common to the cases for a common object of awareness. Searle’s own alternative gives perceptual content causally self-reflexive, de re conditions of satisfaction: the experience presents its object directly by requiring that very object to cause that very experience. The book borrows the diagnosis and the compatibility of intentionalism with direct realism, not this machinery. Its abstract content concerns property types; an appropriate current causal relation within a causal-historical organization fixes the token object, a debt Chapter 15 pays. Tim Crane and Craig French classify intentionalism as a direct-realist theory on this causal understanding: direct perceptual representation plus the appropriate non-deviant causal relation to the represented object constitutes seeing, while the content common to perception and hallucination supplies no common object of awareness. See Crane and French, “The Problem of Perception,” Stanford Encyclopedia of Philosophy (Fall 2021 edition), secs. 3.3.3–3.3.6. ↩
21. The terminological tangle is documented in James Genone, “Recent Work on Naïve Realism,” American Philosophical Quarterly 53, no. 1 (2016): 1–25, esp. sec. 2, which separates direct realism (the denial that we perceive the world by perceiving something else first), representationalism (perceptual experience has representational content), and naïve realism (perceived objects partly constitute experience) as three distinct theses too often run together under one label. Direct realism is the genus; representationalism and naïve realism are rival species of it, agreeing that perception is direct and disagreeing about what makes it so. Charles Travis’s prior challenge appears in “The Silence of the Senses,” Mind 113, no. 449 (2004): 57–94, esp. 57–60 and 73–78: perceptual experience, as such, need not select one truth-evaluable face value rather than countless compatible descriptions. For John Campbell’s argument that object-involving acquaintance grounds demonstrative reference, and for its relation to disjunctivism, see Matthew Soteriou, “The Disjunctive Theory of Perception,” Stanford Encyclopedia of Philosophy, substantive revision July 28, 2026, secs. 3.5–3.6, which reconstructs the argument and surveys Campbell’s primary treatments. The main text answers both at the hinge: content must receive independent accuracy conditions and token reference rather than inherit either from the phenomenal report or from an already mastered demonstrative. ↩
22. Bill Brewer, “Perception and Content,” European Journal of Philosophy 14, no. 2 (2006): 165–181, esp. 165–170, rejects the content view he had previously defended because its truth-evaluable generality and possibility of falsity fail, in his judgment, to capture perception as direct conscious access to the mind-independent objects themselves. He develops the positive Object View in Perception and Its Objects (Oxford: Oxford University Press, 2011). The related constitutive reading 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. The disjunctivist defense and the indiscriminability-based treatment of hallucination appear in Martin, “The Limits of Self-Awareness,” Philosophical Studies 120 (2004): 37–89, esp. 38–41 and 67–72, and “On Being Alienated,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), 354–410. The pressure point is hallucination: because the perceptual relation is genuine, both relata must exist, so a veridical perception and a matching hallucination cannot share their fundamental nature. Martin accepts that the cases may be introspectively indiscriminable; he denies that this epistemic match establishes positive phenomenal sameness. The representationalist must therefore defend shared content abductively, through its explanatory work, rather than treat it as common ground. For the distinctions among fundamental-kind, content, phenomenal, and epistemological disjunctivisms, and for relational views that allow more commonality than Martin does, see Soteriou, “The Disjunctive Theory of Perception,” secs. 2–5. ↩
23. For reflectance realism, see Alex Byrne and David R. Hilbert, “Colors and Reflectances,” in Readings on Color, Volume 1: The Philosophy of Color, ed. Byrne and Hilbert (Cambridge, MA: MIT Press, 1997), 263–288, and “Color Realism and Color Science,” Behavioral and Brain Sciences 26 (2003): 3–21. C. L. Hardin, Color for Philosophers: Unweaving the Rainbow (Indianapolis: Hackett, 1988), develops the error-theoretic and observer-variation challenge; John Campbell, “A Simple View of Colour,” in Reality, Representation, and Projection, ed. John Haldane and Crispin Wright (New York: Oxford University Press, 1993), 257–268, defends a primitivist alternative. ↩
24. The debt Section 6.5 marks is the burden of aboutness: the directness defended there runs entirely through the experience’s content being satisfied only by its object, so the whole apparatus leans on the further claim that this aboutness is a natural fact rather than a spook. Strong representationalism cannot help itself to directness on credit. The contrast is between two ways of treating intentionality. Searle holds that intentionality is intrinsic to certain biological systems and not further reducible: “that property of many mental states and events by which they are directed at or about or of objects and states of affairs in the world” (John R. Searle, Intentionality: An Essay in the Philosophy of Mind [Cambridge: Cambridge University Press, 1983], 1), a property he takes to belong exclusively to organisms with the right causal powers and to resist explanation in more basic terms — he offers no reductive account of what perceptual aboutness consists in, and in the Minds, Brains, and Programs exchange treats the matter as unsettled. This book takes the opposing, naturalizing line: aboutness is a physical relation cashable through selection, learning, tracking, producer-consumer organization, and world-involving history and use. The commitment is introduced in Chapter 1 (intentionality as world-involving, content fixed by a system’s history of engagement rather than by intrinsic inner stuff) and the case for it is made in Chapter 15, which develops the tracking-and-function tradition of Fred Dretske, Knowledge and the Flow of Information (Cambridge, MA: MIT Press, 1981), esp. ch. 7, and Naturalizing the Mind (Cambridge, MA: MIT Press, 1995), esp. ch. 1, and Ruth Garrett Millikan, Language, Thought, and Other Biological Categories (Cambridge, MA: MIT Press, 1984), esp. chs. 2 and 6. The two strands differ in emphasis — Dretske begins from an information-theoretic notion of indication and recruits teleology to fix the content where mere correlation underdetermines it, while Millikan begins from etiological biological function — but both cash representation in terms of what a state has the function of indicating, and Chapter 15 draws on both, with the frog’s fly-detector and the misrepresentation it makes possible as the test case. The point is dialectical, not doctrinal: Searle’s cure for the Bad Argument can be adopted without adopting his primitivism about intentionality, and indeed must be, since a directness that bottoms out in unexplained aboutness has relocated the mystery rather than removed it — which is also where this book’s intentionalism parts from Searle’s, even as it borrows his cure. Chapter 15 also meets the standard objection to any history-based account of content — Davidson’s Swampman (Donald Davidson, “Knowing One’s Own Mind,” Proceedings and Addresses of the American Philosophical Association 60 [1987]: 441–458, esp. 443–44), the molecule-for-molecule duplicate with no past — by accepting absence rather than disguising it as indeterminacy: at the first instant the duplicate lacks any content-relevant history and therefore lacks content and, on this chapter’s identity claim, phenomenal character; consequential engagement can begin the learning history from which both later arise. Chapter 15 develops the general naturalizing model and its worked applications. Evaluative pain content, mood’s diffuse accuracy conditions, and the warrant for experience-directed relief remain substantive limits on its completeness; locating the debt does not discharge those applications. ↩
25. Dennett develops the Multiple Drafts model against the Cartesian Theater in Consciousness Explained (Boston: Little, Brown, 1991), chs. 5–6. The chapter takes the distributed architecture as a positive resource. Dennett’s qualia eliminativism denies additional intrinsic qualitative properties, not the reality of pain or of content-bearing experience; see ch. 12 and the Appendix. The house identity claim remains a distinct positive commitment rather than following from that shared denial. ↩
26. Michael Tye, Ten Problems of Consciousness (Cambridge, MA: MIT Press, 1995), ch. 5, defines poised content as content “poised for use in the formation of beliefs and/or desires.” The book uses PANIC as a historical framework, not as Tye’s unrevised current position: Tye later calls PANIC his “old view (1995)” and rejects its equivalence to his newer externalist account in “Yes, Phenomenal Character Really Is Out There in the World,” Philosophy and Phenomenological Research 91, no. 2 (2015): 483–488, secs. 11–17, https://doi.org/10.1111/phpr.12167. Daniel C. Dennett’s Multiple Drafts model rejects a finishing place where content becomes conscious and instead tracks distributed processes whose effects on memory, report, reasoning, and action give some contents greater “fame”; Consciousness Explained (Boston: Little, Brown, 1991), chs. 5–6. The present synthesis treats that architecture as evidence for a distributed access-and-control profile while rejecting the inference that phenomenal character should be eliminated. The four PANIC conditions distinguish phenomenally conscious states from unconscious representations, conceptual attitudes, and mere sensory registrations, but Poise means availability to a cognitive consumer, not inevitable report or judgment. The Nonconceptual condition invites a Sellarsian objection worth heading off. “Nonconceptual” marks the grain of the content, not a brute sensory item licensing belief from outside the space of reasons; compare Wilfrid Sellars, “Empiricism and the Philosophy of Mind” (1956). Fine-grained content can stand available for conceptual uptake without already being conceptual or automatically consumed. See Michael Tye, “Nonconceptual Content, Richness, and Fineness of Grain,” in Perceptual Experience, ed. Tamar Szabó Gendler and John Hawthorne (Oxford: Oxford University Press, 2006), and contrast John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). ↩
27. Michael Tye states the strong thesis in exactly these terms: “phenomenal character is one and the same as representational content that meets certain further conditions” (Michael Tye, Consciousness, Color, and Content [Cambridge, MA: MIT Press, 2000], ch. 3). “Consists in” names this identity in the main text, not the weaker constitution relation exemplified by Lynne Rudder Baker, Persons and Bodies: A Constitution View (Cambridge: Cambridge University Press, 2000). Naming the distinction does not refute the modal rival. Chapters 9 and 11 test whether conceivability and explanatory difference supply a genuine separation between character and the complete, independently specified intentional profile. ↩
28. For theoretical identification, U. T. Place, “Is Consciousness a Brain Process?,” British Journal of Psychology 47 (1956): 44–50, and J. J. C. Smart, “Sensations and Brain Processes,” Philosophical Review 68 (1959): 141–156. The water/H₂O and heat/molecular-motion analogies serve here as heuristics for the logical form of an a posteriori identity — conceptual difference without ontological difference — not as historical demonstrations that explanatory convergence entails identity. The adoption in the text is abductive and defeasible, per the conditions it posts. ↩
29. Constitution’s paradigm cases carry independently motivated differences between the relata — persistence conditions, modal profile — that support both the nonidentity and the direction of dependence. Lynne Rudder Baker, Persons and Bodies: A Constitution View (Cambridge: Cambridge University Press, 2000), builds the view on differences of exactly this kind; Tye deploys the statue and the clay the same way in Consciousness and Persons: Unity and Identity (Cambridge, MA: MIT Press, 2003). The rival’s most careful contemporary form replaces constitution with grounding, an asymmetric and hyperintensional dependence built to survive agreement in causal and modal profile: see Kit Fine, “Guide to Ground,” in Metaphysical Grounding: Understanding the Structure of Reality, ed. Fabrice Correia and Benjamin Schnieder (Cambridge: Cambridge University Press, 2012), 37–80. The text claims no refutation of grounding-theoretic nonidentity; it observes that no independent motivation for the asymmetry has been supplied in this case, and conditions the identity claim’s adoption on that absence. ↩
30. David Chalmers, “Facing Up to the Problem of Consciousness,” Journal of Consciousness Studies 2 (1995): 200–219, esp. 200–03, and The Conscious Mind: In Search of a Fundamental Theory (New York: Oxford University Press, 1996). The identity claim treats the explanatory demand as epistemic rather than as evidence for an ontological remainder: if a conscious state’s phenomenal character and its complete world-presenting profile are identical, no causal bridge connects two properties in nature, though the conceptual question of how the identity can be understood remains. Chapter 11 engages that distinction directly. ↩