In 1994, at a conference in Tucson, a young Australian philosopher stood up and gave the mystery a name. David Chalmers distinguished the “easy problems” of consciousness - explaining attention, reportability, the integration of information - from the hard problem: why is there something it is like to be anything at all? Why does the firing of neurons come with an inner light? The room understood immediately that he had said something important, and three decades later the problem stands exactly where he left it. No mechanism has been proposed that even begins to close the gap. Integrated information theory assigns consciousness a number; global workspace theory assigns it an architecture; neither explains why the number or the architecture should glow.
The framing assumes the problem was found. I want to argue it was made. The hard problem is a manufacturing defect - an artifact of a particular intellectual production line that was set up in seventeenth-century Europe and has been running ever since. Run the analysis of experience down a different line, one that was operating on the Gangetic plain more than two and a half millennia ago, and the hard problem never appears. Something else appears in its place - a research program, with methods, instruments, results, and a fierce internal culture of verification. The West did not discover the hard problem of consciousness. It discovered the cost of its own starting point.
How the Bill Was Run Up
The story has been told well by others, most recently by Adam Frank, Marcelo Gleiser, and Evan Thompson in The Blind Spot, and it is worth compressing here because the compression reveals the shape. Galileo’s great move was an act of surgery: he divided the qualities of the world into those that could be mathematized - extension, motion, number - and those that could not - color, sound, warmth, taste. The first he assigned to nature. The second he assigned to the observer, declaring that they “reside in consciousness” and would vanish if the living creature were removed. Physics would deal with the real; the observer would hold the remainder.
It was a brilliant accounting trick, and it worked. By excluding the experiencer from the domain of inquiry, the new science gained a purchase on nature that no civilization had achieved through that route before. But the excluded remainder did not vanish. It sat on the books, unexamined, accumulating interest. Descartes formalized the exclusion into a metaphysics - res extensa for the world, res cogitans for the leftover - and then the res cogitans itself was slowly forgotten, dismantled, or redescribed in the vocabulary of the other side. By the twentieth century the inheritors of this tradition were attempting to explain the observer using only the categories that had been defined, at the origin of their method, by the observer’s removal.
That is the hard problem. It is the bill arriving, three centuries late, for Galileo’s surgery. You cannot carve the witness out of your ontology and then buy it back with the currency of objects, because the currency was minted by the carving. Chalmers’s zombie thought experiment - a being physically identical to you with no inner experience - feels conceivable precisely because the physical description was constructed, from the first day, to be complete without experience. The conceivability of the zombie measures the depth of the original exclusion. Nothing more.
The hard problem is what a civilization sees when it looks for consciousness through an instrument designed to exclude consciousness from view.
This is why the problem does not yield to more neuroscience. Every new imaging study, every refined correlate, adds detail to the object-side description - and the object-side description was never going to contain the subject, by construction. The frustration of the field is the frustration of a man searching for his spectacles with increasing rigor while wearing them.
The Analysis Run the Other Way
Now run the production line in reverse. Begin where the Upanishadic rishis began: with the one datum that cannot be doubted, removed, or explained away, because every act of doubting, removing, and explaining presupposes it. Awareness itself. In this analysis, consciousness is the given and the object-world is the derived category - the thing whose status requires investigation. The question “how does matter give rise to experience?” never arises, because matter was never installed as the foundation. The question that arises instead is more interesting: what is the structure of awareness, and what is its relation to the contents that appear within it?
The Mandukya Upanishad - twelve verses, the shortest of the principal Upanishads, and in the tradition’s own estimation sufficient on its own for liberation - performs this analysis with a precision that has no Western parallel before the twentieth century, and arguably none after. It takes the totality of experience and exhausts it across three states: jagrat, the waking state, where awareness illuminates gross outer objects; svapna, the dream state, where it illuminates subtle inner objects; and sushupti, deep sleep, where it illuminates the absence of objects.
That third clause should be read twice, because it carries the whole argument. Deep sleep is a data point. Western philosophy of mind treats dreamless sleep as the off-state, the null condition where consciousness simply is not - which is exactly what the object-first frame requires it to be. The Upanishadic analysis notices something the off-state reading cannot accommodate: we wake from deep sleep and report on it. “I slept well; I knew nothing.” That report is a memory claim, and memory requires registration. Something was present in the dreamless state to register the peace and the absence, or there would be nothing to report and no one to report it. Awareness persisted when its objects did not - which means awareness cannot be a product of its objects, including that subtle object we call the mind. The Mandukya names the fourth, turiya, which is no fourth state at all but the witness of the three, the constant across waking, dream, and sleep, “not that which cognizes the internal, not that which cognizes the external… unseen, beyond usage, beyond grasp.”
From this analysis falls a discipline of discrimination that the darshanas formalized: drg-drsya-viveka, the discernment of the seer from the seen. Whatever can be witnessed is an object of awareness and therefore not awareness. Thoughts can be witnessed. Emotions can be witnessed. The sense of being a self - the ahankara, the I-maker - can be witnessed, arising in the morning, dissolving in sleep. Each of these the analysis strips away, applying the same test, until only the witness remains: sakshi, that which cannot be made an object because every objectification occurs within it.
Here the Sankhya darshana adds the distinction that exposes how blunt the Western category of “mind” really is. In Sankhya’s accounting, manas, buddhi, and ahankara - the sensory mind, the discriminating intellect, the I-maker - all fall on the side of prakriti. They are subtle matter: instruments, evolutes, machinery. Purusha, consciousness proper, stands apart from all of them, the unmoving light in which the machinery is seen. The entire Western debate, from Descartes through Chalmers, conflates the instrument with the light - bundles thought, intellect, selfhood, and awareness into the single undifferentiated term “mind” or “consciousness” and then asks how the bundle relates to the brain. Sankhya would say the question is malformed. Most of the bundle is brain, or at least matter in the same sense the brain is. The residue - the pure witnessing - is what the analysis was for, and it is exactly the residue the Western bundle conceals.
An Empiricism of the Inner
The standard dismissal arrives on schedule: this is mysticism, introspection, the private testimony of ancient enthusiasts. The dismissal mistakes the genre. What the Indic traditions built around this analysis is best described as an empiricism of the inner - observation under controlled conditions, with trained instruments, replicated across observers, schools, centuries, and rival traditions that had every incentive to falsify each other’s claims.
The instrument is attention; the training is yoga. Patanjali’s Yoga Sutras read less like scripture than like a laboratory manual: a taxonomy of mental modifications (vrittis), a graded protocol for stilling them (dharana, dhyana, samadhi - the last three limbs grouped as samyama, the applied method), a catalogue of errors and artifacts (the antarayas, the obstacles, listed with the dry thoroughness of a troubleshooting appendix), and a standing warning against mistaking intermediate phenomena (the siddhis) for the result. The deep-sleep argument of the Mandukya is an inference; samadhi is the direct test - awareness held steady while the contents are progressively withdrawn, the witness examined in conditions of controlled object-poverty, as a physicist isolates a particle by removing everything else from the chamber.
And the replication record is the part the dismissal cannot survive. The Buddhist jhana literature maps the same territory with a different vocabulary and a sharply different metaphysical conclusion - anatta against atman - which is precisely what gives the convergence its evidential weight. These were rival research programs, contesting each other’s interpretations for centuries in formal public debate, while agreeing to a remarkable degree on the phenomenology itself: the staged withdrawal of sense-contact, the persistence of lucidity without content, the dissolution and reconstitution of the I-sense. The Jain tradition ran its own program toward kevala jnana on its own metaphysical chassis. When three independent traditions, armed with incompatible theories and motivated to refute one another, return convergent observational reports across two millennia, the reports have a claim on the word “data.”
The objection that first-person observation is private and therefore inadmissible proves too much. Every observation is first-person; the third person is a construction built from disciplined first-person reports - a point Husserl saw clearly and Western science quietly relies upon every time it trusts a trained astronomer’s reading over a layman’s. We do not reject the microscopist’s report because lens-work takes years to learn. The trained contemplative is in the same epistemic position: an observer whose instrument requires calibration, whose reports are checked against those of other trained observers, and whose errors are catalogued by the tradition itself with unsparing rigor. The asymmetry in how the two are treated is cultural, an artifact of which civilization’s instruments we were taught to trust.
The West built an empiricism of the outer and left the inner to confession. India built both, and the second is the one the hard problem is waiting for.
Husserl deserves a paragraph of his own, because he came closest. The epoché - the suspension of the natural attitude, the bracketing of the object-world’s claim to independent existence so that the structure of experiencing itself comes into view - is a yogic move performed without the manual. He worked out, alone and in the teeth of his own tradition, the first step of a path whose later stages had been charted in another civilization for over two thousand years. One wishes one could go back and put the Mandukya-karika in his hands - not to score a point for Hindu primacy, but because phenomenology stalled roughly where the manual would have predicted: at the threshold where method must become practice, where the bracketing of the world has to be done with the whole instrument of attention rather than affirmed in prose. His successors inherited the stall. Phenomenology became a literature about experience rather than a discipline of it.
The Muddle We Are Exporting
This would remain a quarrel among philosophers were the stakes not arriving in hardware. The question of machine consciousness is now asked in earnest - in labs, in policy papers, in the welfare deliberations of AI companies - and it is being asked with the unexamined bundle. Is the system conscious? Does it experience? The askers inherit a concept that conflates intelligence, selfhood, reportability, and witnessing, because the tradition they inherited it from never performed the discrimination. So we get integrated-information scores assigned to thermostats, solemn papers on whether a language model’s self-reports are evidence of an inner life, and the spectacle of a civilization that cannot define consciousness in humans preparing to legislate its presence in silicon.
The Sankhyan distinction does real work here. Nearly everything an AI system exhibits - perception, discrimination, memory, even the construction and report of a self-model - falls on the prakriti side of the ledger: instrument, machinery, evolute. The Indic analysis would find nothing surprising in a machine that performs buddhi-functions; the buddhi was always matter. The live question, the only question, is whether there is witnessing - and the analysis simultaneously explains why that question cannot be settled from the outside, since the witness is precisely what never appears among the objects. This is a sharper diagnosis of the verification problem in machine consciousness than anything the object-first frame has produced, and it was available before Alexander reached the Beas.
The fractal holds, as it always does. As with the jiva, so with the samuha. An individual who loses the discrimination between witness and instrument identifies with the machinery - with thought, role, the I-construct - and the tradition calls this the root confusion, the seed of suffering. A civilization that loses the same discrimination identifies with its own productions, mistakes its instruments for its essence, and ends by asking its machinery to certify whether anyone is home. The hard problem at the scale of the person becomes, at the scale of the culture, the strange spiritual vertigo of the present moment: a civilization of unprecedented object-mastery that can no longer say what an experiencer is, scanning cortical tissue for the ghost its founding method evicted.
There are signs of the blind spot being noticed from within. Thompson and his collaborators name it. Panpsychism’s return to philosophical respectability is the object-first frame buckling under the weight of its remainder - if consciousness cannot be derived from matter, smear it into matter and call the smearing a theory. Galen Strawson declares physicalism compatible with experience being fundamental. Each of these is a fumbling toward a position whose full articulation, with method attached, sits in the blind spot itself - in the darshanas the academy still files under religion, the deep-sleep argument still unread by the discipline that needs it most.
The choice in front of philosophy of mind is between inheriting a completed analysis and re-deriving it syllable by syllable under other names. Re-derivation is the current plan: another decade of correlates, another integrated-information decimal, another conference where the zombie shuffles through its proofs, panpsychist credences assigned to electrons with the confidence of men measuring the ghost they evicted by the draft it left. Or the other path: take seriously that somewhere between the Sarasvati’s drying and the Buddha’s death, the witness problem was posed correctly - and that twelve verses on waking, dream, and sleep still contain what three decades of Tucson conferences have been circling without entering. Tat tvam asi was a finding. The method that found it is still available.