Indian Illusion Arguments and Perception Share a Problem, Not a Solution

The Indian epistemologists arrive with a question that is less decorative than it sounds. What, exactly, lets a cognition count as knowledge?

Not which doctrine has the prettiest metaphysical weather. Not which sage has the most persuasive eyebrows.

What gets through.

Their name for a means of valid knowledge is pramāṇa. Perception is one candidate, pratyakṣa. Inference is another, anumāna. Testimony, comparison, postulation, and non-cognition make up the larger family recognized somewhere in the tradition.

The counting is not clerical. It is doctrine with a calculator.

Cārvāka admits one: perception alone, because inference never entirely escapes doubt. Dignāga and Dharmakīrti, like Vaiśeṣika, admit two. Sāṃkhya takes three. Nyāya takes four. Prabhākara’s Mīmāṃsā takes five. Bhāṭṭa Mīmāṃsā and Advaita take six.

A person who says “the Indian view” here has arrived at the debate without bringing a chair.

Nyāya even gives the disagreement its furniture.

The Nyāya Sūtras begin with sixteen categories, and several of them concern the work and failure of argument: honest inquiry, competitive wrangling, destructive refutation, fallacy, equivocation, futile reply, and the particular manner in which a debater loses.

Philosophy has rarely made losing so administratively specific.

Its inference has five members: thesis, reason, familiar example, application, conclusion.

There is fire on the mountain. There is smoke. Smoke accompanies fire, as in a kitchen. The mountain has smoke. Therefore there is fire.

The point is not merely to reach the fire. It is to leave fingerprints on the route.

Error belongs in that route.

Mīmāṃsā’s svataḥ prāmāṇya says a cognition is true by default unless a specific defeating condition appears. It reverses the burden that asks every appearance to arrive carrying a tiny passport.

A later correction does not erase the earlier episode. It identifies what defeated its claim.

This is where perceptual science walks in, wearing a lab coat and carrying an object that does not quite exist.

Look at a Kanizsa triangle. Three incomplete circles are arranged so their missing wedges point toward the same empty centre. A bright solid triangle appears, with edges that are not in the image.

Measurement finds no corresponding change in luminance or colour along those edges. Perception continues to submit the triangle anyway, with the serene confidence of an office that has misplaced the paperwork.

The figures demonstrate closure: visual organization completes an incomplete outline into a bounded figure.

The relevant effect appears early in visual cortex, not only in some later stage where a little inner critic might be thought to add the triangle by hand.

Bottom-up signal and top-down organization are not politely queued at separate doors. Ordinary perception combines incoming features with context, expectation, and grouping.

At which point the brochure practically writes itself.

Indian philosophers knew that an appearance may need correction. Perceptual science can build the correction into a display. Therefore a Kanizsa triangle confirms an Indian school.

No.

A triangle with imaginary edges does not decide whether testimony is an independent pramāṇa, whether an absence is known through anupalabdhi, or whether liberation is Nyāya’s cessation of suffering, Sāṃkhya-Yoga’s kaivalya, or Advaita’s jīvanmukti.

A sensory mechanism does not settle a theory of valid cognition, and it certainly does not settle a theory of liberation. That would be an ambitious result for three mutilated circles.

The sciences do not even receive one clean mechanism for every illusion.

Müller-Lyer’s lines appear unequal despite being equal. Richard Gregory explains the effect as misapplied size-constancy scaling. A rival centroid account appeals to the overall luminance profile.

Cross-cultural work found the size of the misperception varying widely between societies, and later studies question whether carpentered environments explain the whole range.

The effect persists. The explanation keeps arguing.

And the correction does not dismiss the appearance.

Knowing the lines are equal does not make them look equal. Measuring them settles what one is entitled to claim. It does not retract what is still being seen.

That gap, between the corrected judgment and the persisting appearance, is the whole subject.

That is the pressure shared by the two vocabularies.

Neither gets to dismiss error as a dirty exception to an otherwise immaculate faculty.

Mīmāṃsā needs defeating conditions because cognitions can be corrected. Illusion research needs the false percept because it exposes which cues a constancy mechanism depended on.

In both cases, successful seeing is not defined by never being vulnerable to error. Error helps reveal what success had been relying on.

This is also why the laboratory cannot be promoted into a metaphysical referee.

Gibson’s ecological account argues that the optic array provides structured information for direct perception. Marr instead treats vision as a computational recovery of a three-dimensional world from an ambiguous two-dimensional projection. Predictive coding gives a third answer.

These are rivals, not three stickers on the same microscope.

It is speculation, not a discovery from either tradition, to say that a Kanizsa triangle does exactly the epistemic work of a classical Indian stock case.

The comparison earns only a structural resemblance. An experience presents something. A correction changes the claim one is entitled to make. The experience has already revealed that seeing was more than a retinal delivery.

The deeper continuity is that error is a test of perception, not its scandal.

One tradition asks which instruments can lead beyond confusion and toward liberation. The other manipulates cues until a normal visual competence betrays its dependencies.

Neither gives the other its answer. Both refuse the fantasy that knowledge begins only after appearances have stopped being troublesome.

The world offers a line, a room, three broken circles. Seeing makes a claim anyway.