Argument

Argument from Perception

Intro

Everyone agrees on the physics. Light leaves an object, enters your eye, gets turned into electrical signals, and those signals get processed in the back of your brain. Nobody disputes any of that.

The argument here is that the physics, however complete, does not add up to seeing, and that this is not a temporary gap waiting on better neuroscience.

Two things are true of perceiving that are not true of any physical process, no matter how complicated.

The first is that perceiving is about something. Your seeing of a tree is about the tree. Trace the whole causal chain and you will find shapes, rates, voltages, and locations. You will never find a step where the process becomes about anything, because aboutness is not on the list of physical properties.

The second is that perceiving can be wrong. The stick in water looks bent and is not. A mistake is a normative failure, a falling short of how something ought to be. And a retina cannot be mistaken. A firing pattern cannot be false. Nerve tissue does not fail to live up to a standard; it just does what it does.

So the naturalist has a bill to pay. Perception has two properties, aboutness and the capacity for error, and the physical vocabulary contains neither. On Christian theism these are not anomalies. A knowing God made knowing creatures, so aboutness is a basic feature of reality rather than something that has to be built out of parts that lack it.

This argument is not a proof. It is a claim about which worldview has to work harder, and by how much.

◎ See this page's graph and discover linked topics →

In full

An abductive argument from the structural features of perceptual experience to the superior explanatory adequacy of theism over naturalism.

Its premises are supplied by the phenomenological description of perception, developed independently of any theological motive by Edmund Husserl and Maurice Merleau-Ponty and set out at Phenomenology of Perception. The description establishes four features: perception is intentional and its intentional object is the mind-transcendent thing rather than an inner representation; it is perspectival while presenting the object rather than the perspective, the structure Husserl analyzed as adumbration; it is normative, admitting of correctness and error; and it is enacted by a lived body whose motor capacities are constitutive of the perceptual field.

The argument then contends that features one and three are not physical properties in any sense a naturalist can accept without helping himself to what he needs to explain, that the two available naturalist responses both fail (indirect realism collapses into skepticism, eliminativism is self-refuting), and that classical theism accounts for the features by locating original intentionality in a knowing God and derivative intentionality in creatures made to know.

The argument's distinctive contribution alongside its siblings is that it targets perception rather than reasoning. Argument from Reason and Argument from the Reliability of Reason concern inference. Argument from Consciousness concerns felt quality. This one concerns the world-directedness of sensory experience, and it carries a payload the others do not: it is the philosophical precondition of appeals to eyewitness testimony, and therefore sits underneath Argument from the Resurrection.

This page is structured as debate prep. Each premise carries a second-order positive case, anticipated objections steel-manned, point-for-point rebuttals, a live-cite kit, and tactical notes.

Argument structure

Form

Abductive, inference to the best explanation, with an embedded dilemma at P4. The argument does not claim that naturalism entails a contradiction, and should not be presented as though it did. It claims that perception exhibits features naturalism has no resources for, that the two ways of accommodating this both carry heavy costs, and that theism carries neither.

Note what is not claimed. Not that perception is infallible. Not that phenomenology proves God. Not that neuroscience is mistaken about the causal chain. The argument grants the entire physiology and asks what has been left out.

P1, Perception is intentional and world-directed

Affirmative case (second-order arguments)

1. The description finds no inner object. Attend as carefully as you like to the experience of seeing a chair. No mental chair-picture turns up standing between you and the chair. What is given is the chair. The representational item is a theoretical posit, introduced to make the causal story terminate somewhere, not a datum anyone has ever encountered.
2. The homunculus regress. If perceiving a chair requires inwardly perceiving a chair-representation, then perceiving that representation requires a further inner act, and so on without end. Making the inner observer smaller, or mechanical, or distributed does not stop the regress. Adding stages to a causal chain never produces, at any stage, awareness of the earlier stage.
3. Horizonal structure has no representational analogue. A cup is given with its hidden back co-intended as really there. That is not an inference performed after the fact, and it is not absent from the experience. A representation, being a bounded item, cannot contain its own beyond. Husserl's analysis of adumbration is the strongest single piece of descriptive evidence here.
4. Action confirms it. Your hand goes to the cup. Not to a location computed from a picture, and not to a picture. The whole organization of embodied behavior presupposes that perception delivers the world, which is Merleau-Ponty's point about motor intentionality.
5. Derived intentionality needs an original. Everything uncontroversially physical that seems to be about something (a photograph, a word, a thermostat reading) is about it only because minds treat it so. Take away all minds and the aboutness goes. If every case were like that, the regress never terminates. See Intentionality.

Anticipated objections

1. "Representationalism is the scientific consensus, and phenomenology is armchair introspection." Vision science works with internal representations successfully every day. Introspective reports are notoriously unreliable, and change blindness and inattentional blindness show that people are badly wrong about their own visual experience.
2. "Illusion and hallucination refute direct realism." In a hallucination there is no object, and yet the experience is subjectively identical to veridical perception. If the two are indistinguishable from the inside, the common factor must be an internal state, and that internal state is what you are aware of in both cases. This is the argument from illusion, and it is the strongest objection on the page.
3. "Aboutness is just reliable causal covariation." A state means whatever it reliably indicates. Smoke means fire. There is no mystery, only correlation plus a bit of information theory.
4. "You are describing folk psychology, not reality." How things seem is not evidence about how they are. Physics routinely overturns experience: the table seems solid and is mostly empty space.

Rebuttals

1. The consensus objection conflates two claims. Vision science establishes that the brain has internal states carrying information, and nobody disputes that. It does not establish that those states are what you are aware of, which is a philosophical thesis added on top. Change blindness cuts the other way as well: it shows people are wrong about the richness of their visual field, which is a claim about the field's contents, not evidence that they were looking at a picture. And the objection cannot be pressed too hard without self-harm, since a scientist reading an instrument must be perceiving the instrument rather than an unverifiable inner surrogate for it.
2. The argument from illusion equivocates on "indistinguishable." That two states cannot be told apart from the inside does not entail they share a constituent. Disjunctivism, developed by J. M. Hinton, Paul Snowdon and John McDowell, is the standard reply: veridical perception and hallucination may be different in kind while introspectively indiscriminable, exactly as a genuine banknote and a perfect forgery are different in kind while visually identical. The inference from "cannot tell apart" to "same common object" is precisely the step in dispute, and asserting it begs the question. Concede fully that hallucinations occur and that they matter. What does not follow is that ordinary seeing is a hallucination that happens to be caused correctly.
3. Covariation cannot deliver error, which is what makes it fail. If a state's content is fixed by what it reliably indicates, then a state that fires in the wrong circumstances is not misrepresenting; it is correctly indicating a wider disjunction. The account cannot make room for falsity, and falsity is the thing most needing explanation. This is the disjunction problem, and it is conceded as unsolved by most who work on it. Teleosemantics repairs the timing but inherits the same indeterminacy from selection history.
4. The folk-psychology objection saws the branch. "How things seem is not evidence about how things are" is itself a claim about how things are, arrived at by someone reading instruments, seeing data, and reasoning. It uses perceptual reliability to conclude that perception is not what it seems. The table example also proves less than it appears: the table's solidity is not an illusion but a real macroscopic property, and physics explains it rather than debunking it.

Live-cite kit

• Scholarly: Merleau-Ponty, Phenomenology of Perception (1945), on the failure of the constancy hypothesis. "We must recognize the indeterminate as a positive phenomenon."
• Scholarly: Husserl, Ideas I (1913), on adumbration, that a spatial thing is necessarily given through profiles and necessarily given as more than any profile.
• Scholarly: Thomas Reid, Inquiry into the Human Mind (1764), naming the veil of ideas and the skeptical trajectory it opens.
• Aphorism: "You did not see a picture of the door. You walked through the door."

Tactical notes

• Do not begin with the technical vocabulary. Begin with the door. Everyone grants they walked through the actual door, and the concession is already fatal to the strong representational picture.
• Concede hallucination early and completely. Fighting it looks evasive and it is not necessary. The concession costs nothing once disjunctivism is on the table.
• Do not overreach into infallibilism. The moment you sound like you are claiming perception never errs, you have handed over the bent stick and lost the room.

P2, Perception is normative

Affirmative case (second-order arguments)

1. Misperception is ordinary. Not exotic, not pathological. The stick looks bent. The tower looks round. The dress looks blue. Perception is routinely and unremarkably wrong, which entails that there is something for it to be wrong about and a way it ought to have gone.
2. Physical processes cannot fail. A retina does not err. A cortical firing pattern is not false. Physical description is exhausted by what happens; it contains no room for what should have happened. Norms are not physical properties.
3. Veridicality is the concept the whole field runs on. Vision science itself is organized around the distinction between accurate and inaccurate perception, and cannot state its own results without it. The normative notion is presupposed by the discipline that is supposed to eliminate it.
4. The same structure appears in reasoning. Validity is a norm on inference exactly as veridicality is a norm on perception. This is the argumentative core of Argument from Reason, and the two arguments reinforce each other precisely because they are about different faculties.

Anticipated objections

1. "Norms come from biological function." An eye is for seeing, and something is malfunctioning when it fails to do what it was selected for. Teleology is naturalized by evolutionary function, and no God is needed.
2. "Norms are conventional." Correctness is what a community counts as correct. This is a social fact, and social facts are physical facts about people.
3. "This is just a language quirk." We say perception is mistaken, but that is loose talk. Strictly there are only states and their causes.

Rebuttals

1. The biological-function reply is the strongest and still falls short, in two ways. First, selection explains why organisms with a trait outbred those without it; it does not by itself introduce a standard the individual case can fail to meet. Recasting "malfunction" as "statistically atypical relative to selection history" changes the subject from a norm to a frequency, and a frequency cannot be violated. Second, the account is indeterminate: selection favored the ancestral organism's surviving and reproducing, and indefinitely many content-assignments are consistent with that, so it cannot fix what the state is supposed to represent. Grant the reply its real force, that biological teleology is a genuine and useful notion, and collect the larger concession attached: the naturalist has now committed to real ends in nature, which is a heavier metaphysical bill than he came to pay. See Argument from Conscience for the same move in ethics.
2. The conventionalist reply cannot be run about perception without absurdity. Whether the stick is really bent is not settled by a vote. And conventions are themselves normative, so the account is circular: it explains norms by appeal to what a community ought to be counted as having agreed.
3. The language-quirk reply is eliminativism in miniature, and inherits its self-refutation. The claim "strictly there are only states and their causes" is offered as correct rather than incorrect, which is a normative status, asserted in the act of denying that any such status exists. See Self-refutation.

Live-cite kit

• Scholarly: Wilfrid Sellars on the logical space of reasons as distinct from the space of causes, in "Empiricism and the Philosophy of Mind" (1956).
• Aphorism: "A retina has never once been wrong. You have been wrong all week."

Tactical notes

• Run P2 second, never first. P1 gets the concession, P2 collects on it.
• If the opponent reaches for biological function, do not treat it as a bad answer. It is the best answer available. Grant it and press the two costs.

P3, Aboutness and error are absent from the physical vocabulary

Affirmative case (second-order arguments)

1. Enumerate and check. Mass, charge, spin, position, momentum, rate, energy, structure. Aboutness is not on the list, and it is not a composite of items on the list.
2. Derived intentionality is parasitic. Every physical item that appears to be about something is about it because minds treat it so. See Intentionality. The regress requires a terminus.
3. Indeterminacy. Even granting a correlation-based account, the physical facts do not fix a unique content. Fodor's disjunction problem, Quine's inscrutability of reference and Kripke's rule-following puzzle converge on this from three directions. James Ross pressed it into a direct argument that determinate formal thinking cannot be a purely physical process.
4. The gap is principled, not provisional. This is not a claim that neuroscience is young. It is a claim about what kind of properties the physical vocabulary contains. More detail about the same kind of properties does not produce a property of a different kind.

Anticipated objections

1. "This is a god-of-the-gaps argument." You are naming something unexplained and inserting God. That pattern has failed repeatedly.
2. "Emergence." Novel properties arise from complex arrangements all the time. Wetness is not a property of a water molecule. Aboutness could emerge from neural complexity the same way.
3. "You have simply defined the physical narrowly." If aboutness turns out to be real and natural, it is physical by definition, and the argument evaporates.

Rebuttals

1. The gaps charge misidentifies the argument's form. A gaps argument says a mechanism is unknown and posits God as the mechanism. This argument says the category is absent and posits a different ontology, not a different mechanism. It also does not claim that theism supplies a mechanism for intentionality, because theism supplies none. It claims theism has the resources to have intentionality be basic. That said, the charge deserves a real concession: if a naturalist account of determinate, error-permitting, non-derived content were produced, this argument would be substantially weakened. That is what it would take, and it is worth saying plainly which evidence would count against you.
2. Emergence cases are not analogous, and the disanalogy is specific. Wetness is a rearrangement of properties already present: molecular structure, cohesion, viscosity. It is quantitative aggregation with a new description, and it is derivable in principle from the microphysics. Aboutness is not a rearrangement of shapes and rates, because no quantity of shape and rate is a quantity of aboutness. Calling it emergent names the problem rather than solving it. See Access and Phenomenal Consciousness for the same objection in the consciousness debate.
3. The definitional reply concedes the argument's substance while disputing its label. If "physical" is expanded to include the intrinsically world-directed and the normatively assessable, then physicalism no longer excludes what theism asserts, and the dispute is verbal. Naturalists who take this route generally end up at panpsychism or neutral monism, both of which grant that mind is basic. That is the concession the argument was after. See Naturalism and Materialism.

Live-cite kit

• Scholarly: Brentano, Psychology from an Empirical Standpoint (1874), on intentional inexistence as the mark of the mental.
• Scholarly: Searle's Chinese Room (1980), on syntax never sufficing for semantics.
• Scholarly: James Ross, "Immaterial Aspects of Thought" (1992), on the indeterminacy of physical process with respect to formal content.
• Aphorism: "A camera has a better lens than you and has never seen anything."

P4, Both naturalist responses fail

The dilemma

Horn one, indirect realism. Keep the physics and make the object of awareness an internal representation. The cost is the veil of ideas, named by Thomas Reid: awareness never reaches past its own representations, so no check on them is possible, since every check is another representation. This is the road from Descartes through Berkeley to Hume, and it terminates in skepticism about the external world. See Cartesian Skepticism.

Horn two, eliminativism. Deny the phenomena. Treat aboutness, seeming and error as folk-psychological posits due for retirement. The cost is immediate: the position is a belief about the world, offered as true rather than false, which uses at every step exactly what it denies. See Self-refutation.

The middle way, and why it is the real battleground. Most working naturalists take neither horn. They accept the reality of intentionality and pursue a reductive account of it (informational semantics, teleosemantics, functional-role semantics). This is the honest and serious position, and the argument's actual claim about it is modest and should be stated modestly: no such account has yet delivered determinate, error-permitting, non-derived content, and each faces a well-known objection its own proponents concede. That is a report on the state of play, not a proof of impossibility.

P5, Theism accounts for both features

1. Original intentionality has a home. God knows, and divine knowing is not conferred by any interpreter standing behind God. The regress terminates in something that was never going to be built out of syntax.
2. Creatures are made to know. On the imago Dei, human beings are not mechanisms that had to bootstrap semantics out of shapes. They are constituted as knowers by a knower. Aboutness is inherited rather than manufactured. See Imago Dei.
3. Norms have a ground. If a faculty was made for something, there is a standard it can fall short of, and the standard is not a statistical frequency. This is the same structure as the moral argument, transposed from conduct to cognition.
4. Embodiment is not an embarrassment. A tradition confessing creation, incarnation and bodily resurrection has no stake in treating the body as a container the mind is trapped inside, so Merleau-Ponty's finding that perception is bodily is a convergence rather than a difficulty.
5. Intelligibility is expected, not lucky. If the Logos is prior to the cosmos, a knowable world encountered by knowing creatures is what you would predict. See Argument from Intelligibility.

What theism does not do. It supplies no mechanism for intentionality, and claiming otherwise invites a fair charge of evasion. It offers an ontology in which aboutness is basic instead of anomalous. That is a real explanatory advantage and it is not a demonstration.

Master objections to the argument as a whole

1. "Phenomenologists were not theists, so their findings cannot support theism." Correct on the first half, irrelevant on the second. The findings are descriptive claims evaluated on their merits, and they are strengthened by coming from people with no theistic motive. Merleau-Ponty's non-belief is stated openly on his page precisely so no one can spring it as a gotcha.
2. "Direct realism is a minority position in contemporary philosophy of mind." Disjunctivism and naive realism are genuinely minority positions, and this should be conceded plainly rather than obscured. But the argument's core needs only that perception is intentional and normative, which is close to consensus. The stronger direct-realist thesis makes the argument sharper without being load-bearing.
3. "You are ignoring predictive processing." Predictive-processing models are the most interesting current development and deserve engagement rather than dismissal. They are also, on their own terms, a story about the mechanism by which the brain settles on a perceptual outcome. That the brain generates and updates predictions says nothing about how any state comes to be about anything, and the models freely use representational and normative vocabulary (prediction, error, accuracy) that is exactly what is in dispute.
4. "This proves at most a mind, not the Christian God." Granted, and the codex does not run this argument as a standalone case for Christianity. It is one strand in Cumulative Case for Christian Theism. What it uniquely adds is the philosophical precondition for the historical argument, since eyewitness testimony presupposes that perception reaches the world.

Tactical opening / closing

Opening. "You walked in here through a door. Not through a picture of a door. Before we argue about God, I want to know what your view says about that, because I do not think it can say anything."

Closing. "You have granted the whole causal chain, and I never disputed it. Light, retina, optic nerve, cortex. What I asked is where in that chain the process becomes about the tree, and where it becomes capable of being wrong about the tree. The answer is nowhere, because those are not the kind of properties that chain contains. I am not saying you have a gap to fill in. I am saying you have brought the wrong kind of thing to the job. On my view a knower made knowers, and seeing is what knowers do. On yours, seeing is something that had to appear out of ingredients that contained no trace of it."

Connection to Scripture

The biblical writers stake their central historical claim on perception in exactly the world-delivering sense this argument defends, and they are unusually precise about the sense modalities involved.

What was from the beginning, what we have heard, what we have seen with our eyes, what we have looked at and touched with our hands, concerning the Word of Life. (1 John 1:1, NASB95)

Four perceptual verbs in one clause: akouō and horaō in the perfect tense, a completed act whose result stands; theaomai, sustained looking rather than a glance; psēlaphaō, verifying by touch. The two words for a firsthand witness are perceptual rather than juridical, autoptai at Luke 1:2 and epoptai at 2 Peter 1:16.

See My hands and My feet, that it is I Myself; touch Me and see, for a spirit does not have flesh and bones as you see that I have. (Luke 24:39, NASB95)

The invitation is to handle, the modality hardest to hallucinate and easiest to falsify. Thomas is offered the same terms at John 20:27.

The dependency runs in a direction rarely noticed. On a strict veil-of-ideas view, no witness ever perceives anything but their own representations, and apostolic testimony reduces to a report about apostolic mental states, which is precisely what the hallucination hypothesis requires. This argument is therefore load-bearing under Argument from the Resurrection and The Bible Is Just Second-Hand Sources Objection Defeater rather than an ornament beside them.

See also

• Phenomenology of Perception, the concept hub for the descriptive base
• Intentionality, the aboutness structure P1 and P3 turn on
• Argument from Reason, the same structure applied to inference rather than perception
• Argument from the Reliability of Reason, the evolutionary defeater against naturalist cognition
• Argument from Consciousness, the sibling argument from felt quality
• Fitness Beats Truth Argument, which stands in a tension resolved at Phenomenology of Perception
• Argument from Intelligibility, the wider claim that mind is basic
• Modal Argument from Mind, another route from mental properties to non-physicalism
• Argument from the Resurrection, which this argument is a precondition of
• The Bible Is Just Second-Hand Sources Objection Defeater, the testimony case that depends on world-delivering perception
• Cumulative Case for Christian Theism, the frame this strand feeds
• Maurice Merleau-Ponty, source of the embodiment findings
• Edmund Husserl, source of the intentionality and adumbration analyses
• Cartesian Skepticism, the trajectory horn one opens
• Self-refutation, the structure that closes horn two
• Materialism, the position P3 presses against
• Imago Dei, the anthropology P5 rests on

Scripture quotations taken from the New American Standard Bible® (NASB), Copyright © 1960, 1971, 1977, 1995 by The Lockman Foundation. Used by permission. All rights reserved. www.lockman.org