The One Question Modern Science Has Not Framed, and the Upanishads Structured Two Thousand Years Ago

Ask a neuroscientist why brain activity is accompanied by experience and you will get an honest answer: nobody knows, and no one has a method that would settle it. That is not a gap in the data. It is a gap in the question. David Chalmers named it in 1995: the hard problem of consciousness.

The Mandukya Upanishad does not solve that problem. It was not written to. Twelve short prose passages, the shortest of the principal Upanishads, it starts from the other end. Instead of asking how matter generates awareness, it asks what awareness is like across the three conditions every reader passes through in a day, waking, dreaming and dreamless sleep, and what is constant across them. Its answer, turiya, "the fourth," is presented as the awareness within which the other three appear.

Two traditions, two starting questions. That is the entire claim here, and it is smaller and more durable than what is usually claimed on Hinduism's behalf in this territory. The mainstream position comes first and at full strength; the piece closes with what this argument does not establish.

The problem that has a name but no method

Chalmers published "Facing Up to the Problem of Consciousness" in the Journal of Consciousness Studies in 1995 and split the field in two. On one side sit the easy problems. His list runs: the ability to discriminate, categorize and react to environmental stimuli; the integration of information by a cognitive system; the reportability of mental states; the ability of a system to access its own internal states; the focus of attention; the deliberate control of behaviour; the difference between wakefulness and sleep.

"Easy" is a relative term, he wrote, and getting the details right will probably take a century or two of difficult empirical work. They count as easy because we know what an answer would look like. Specify a computational or neural mechanism that performs the function, and the phenomenon is explained. The form of the explanation is not in doubt, only its contents.

Then the other side. The really hard problem, Chalmers wrote, is the problem of experience. When we think and perceive there is a whir of information processing, and there is also a subjective aspect. He borrows Thomas Nagel's phrase from a 1974 paper: there is something it is like to be a conscious organism.

The hard problem is not "which brain activity accompanies seeing red." That question belongs on the easy list, and is being answered in increasing detail. The hard problem is why any of that activity is accompanied by seeing at all, rather than running in the dark. A perfect functional account would leave the question where it found it.

One further move matters later. Chalmers proposes that a theory of consciousness should take experience itself as a fundamental feature of the world, in his words alongside mass, charge and space-time. His reasoning is that physics has expanded its ontology before when the existing ingredients ran out: nineteenth-century mechanics could not account for electromagnetic processes, so Maxwell and others introduced electromagnetic charge and forces as new fundamental components. He derives none of this from any Indian text and claims no such lineage.

What the mainstream account gets right

The standard position deserves stating at its strongest. The dominant view in cognitive science is that consciousness is something brains do. Matter is the basic stuff, and awareness is a high-level property of particular physical arrangements, the way fluidity is a property of certain arrangements of water molecules rather than a separate ingredient poured in. On this view the hard problem is a confusion that will dissolve as the mechanisms are filled in, much as the old mystery of what separates living matter from dead dissolved once metabolism and heredity were described.

Two research programmes carry most of the weight. Stanislas Dehaene's global neuronal workspace theory, built on Bernard Baars's earlier global workspace model, holds that content becomes conscious when it is broadcast widely across the brain and made available at once to memory, language, planning and motor control, rather than staying local to the process that computed it. It predicts a signature: a late, sudden, widespread "ignition" in prefrontal and parietal regions when a stimulus crosses into reportability.

Giulio Tononi's integrated information theory runs the other way, from phenomenology to physics. It begins with properties Tononi takes every experience to have, asks what a physical system must be like to possess them, and arrives at a measure of how much information a system integrates as a whole beyond the sum of its parts. On IIT consciousness comes in degrees and lives where integration is highest, putting the relevant machinery in posterior cortex rather than the front of the brain.

They make competing, falsifiable predictions, which is why the Cogitate consortium published an adversarial collaboration in Nature in 2025, designed with both camps and preregistered: 256 participants, tested with fMRI, MEG and intracranial EEG. IIT's predicted sustained synchronisation within posterior cortex did not appear. GNWT's predicted ignition at stimulus offset was largely absent too, and prefrontal cortex carried information about some conscious contents and not others. Neither theory came through with its core predictions intact.

Tononi would object to hearing IIT filed under the easy problems, and the objection deserves recording: it is built to address why experience exists, not merely which processes accompany it. Whether it succeeds is contested. What an experiment tests is the neural predictions.

Twelve passages that argue from your own night

The Mandukya has no narrative, no named speakers and no dialogue. Gaudapada wrote a verse commentary on it, the Mandukya Karika, the earliest systematic exposition of Advaita Vedanta that survives. Dating is contested. Patrick Olivelle judges that the Prashna and the Mandukya, the two late prose Upanishads, cannot be much older than the beginning of the common era. Hajime Nakamura and Richard King place the Mandukya in the first or second centuries CE. R. D. Ranade and others group it with the last of the ancient principal Upanishads, several centuries earlier.

Verse 1 puts everything under the syllable AUM, past, present, future and whatever lies beyond time. Verse 2 makes the identification the rest depends on, that this Atman is Brahman, and states that it has four quarters.

Then the text does something a reader can check against last night. Verse 3 gives the waking quarter, Vaishvanara, awareness turned outward on gross objects. Verse 4 gives the dream quarter, Taijasa, awareness turned inward on subtle objects, the same apparatus running on internally generated material. Verse 5 gives deep sleep, the state in which the sleeper neither desires any object nor sees any dream, where awareness is called Prajna, undifferentiated, a mass of consciousness, full of bliss. Verse 6 calls that same Prajna the lord of all, the knower of all, the source and the end of beings.

Verse 7 makes the turn, almost entirely by negation.

"Turiya is not that which is conscious of the inner (subjective) world, nor that which is conscious of the outer (objective) world, nor that which is conscious of both... It is unperceived, unrelated, incomprehensible, uninferable, unthinkable and indescribable... This is Atman and this has to be realized."

Mandukya Upanishad, verse 7, in Swami Nikhilananda's translation.

Verses 8 to 12 map the four quarters onto the syllable itself: A, U, M, and the silence that carries no letter. The fourth is amatra, without measure and without parts.

The method deserves attention more than the conclusion. The raw material of verses 3 to 5 is available to any reader tonight. No revelation is needed to grant that you wake, that you dream, and that some portion of the night passes without dream. Whether the conclusion follows is a separate matter, disputed inside India for centuries. Nyaya, Samkhya, Mimamsa and the Buddhist schools all attacked the Advaita reading, often on the point that follows.

A note on sources. The Mandukya's twelve passages are transmitted alongside Gaudapada's Karika, whose first chapter interleaves them with his verses, and Karika numbering differs between printed editions, so a reference like "Karika 1.12" can point to different lines in different books. Gaudapada's arguments are described here in words rather than by number.

Dreamless sleep is where the two accounts genuinely part

Here the resemblance stops being vague. The standard Western treatment of deep dreamless sleep is that consciousness goes out. No experience, no subject, a gap. You do not report the gap because there was nothing to report and nobody present to report it. Chalmers's list puts "the difference between wakefulness and sleep" among the easy problems.

Vedanta says something else, for a stated reason. In the Brihadaranyaka, Yajnavalkya takes King Janaka through waking, then dream, then the state beyond dream, and reaches this:

"That it does not see in that state is because, although seeing then, it does not see; for the vision of the witness can never be lost, because it is immortal. But there is not that second thing separate from it which it can see."

Brihadaranyaka Upanishad 4.3.23, in Swami Madhavananda's translation.

That is an argument rather than an assertion. The claim is that absence of objects is being mistaken for absence of awareness. What goes missing in deep sleep, on this reading, is the subject-object structure, not the awareness. The evidence offered is the ordinary report on waking: people say "I slept well," a judgement about a stretch of time rather than a report of a blank.

The Mandukya then does something a purely devotional text would not. It declines to call deep sleep the destination. Gaudapada separates Prajna from turiya on the ground that deep sleep still carries ignorance in seed form, a cause not yet sprouted, while turiya carries neither ignorance nor error. If awareness in dreamless sleep were the goal, everyone would reach it nightly.

The obvious objection is a good one. Saying "I slept well" may be a retrospective inference from feeling rested, or a memory judgement about elapsed time, with no experience behind it. Advaita has replies, turning on the difference between remembering an object and the continuity of the witness. Nobody has settled this from an armchair or a laboratory.

Where Western philosophy moved, and how far

Galen Strawson's 2006 Journal of Consciousness Studies paper "Realistic Monism: Why Physicalism Entails Panpsychism" is worth reading precisely because its argument is a physicalist one. Strawson holds that experience is undeniably real, denies that anything real can emerge from constituents wholly lacking the relevant character, and concludes that experientiality must be present in the basic constituents. Philip Goff's Galileo's Error (2019) makes the case for general readers: physics describes what matter does, its behaviour and relations, and by construction says nothing about what matter is, so consciousness could be what fills that silence.

Now the honest sizing. In the 2020 PhilPapers survey of professional philosophers, 51.9% accepted or leaned towards physicalism about the mind and 32.1% towards non-physicalism, with panpsychists roughly a quarter of the non-physicalist group. Among working neuroscientists the proportion is smaller. Panpsychism has moved from unpublishable to arguable inside philosophy departments, which is a real change and is not a consensus. Anyone reporting it as the coming scientific view is misleading you.

The differences from Vedanta matter as much as the resemblance. Panpsychism as Strawson and Goff develop it distributes consciousness across many things, generating the combination problem of how micro-experiences add up to a unified human one. Advaita runs the other way. Consciousness is one and undivided, and plurality is what requires explanation. A panpsychist and an Advaitin agree that consciousness is fundamental and then disagree about nearly everything downstream. Cosmopsychist positions, which start from a single cosmic subject rather than many tiny ones, sit closer. Miri Albahari argues for a view she calls perennial idealism, drawing explicitly on Advaita, but that is a small corner of a minority position.

The people already doing this carefully

Francisco Varela, Evan Thompson and Eleanor Rosch published The Embodied Mind with MIT Press in 1991, arguing that cognitive science needed a disciplined first-person method and turning to Buddhist contemplative traditions for one, because Western philosophy had largely abandoned trained introspection.

Thompson's Waking, Dreaming, Being (Columbia University Press, 2015) takes the Mandukya's own itinerary as its structure, moving through waking, dreaming, deep sleep and the states around dying, reading each against sleep neuroscience and against Indian and Western philosophy of mind.

He has been explicit about the source. In a 2015 Open MIND paper written with Jennifer Windt, he sets out the classical Indian dispute directly: Advaita Vedanta and Yoga held that consciousness is present in dreamless sleep, Nyaya held that it is absent, and the Advaita rebuttal of Nyaya puts pressure on the neuroscientific habit of defining consciousness as whatever disappears in dreamless sleep and returns on waking.

Windt, Tore Nielsen and Thompson took the question to a journal the following year, publishing "Does Consciousness Disappear in Dreamless Sleep?" in Trends in Cognitive Sciences. They argue that the assumption of a total blank is oversimplified, and propose a taxonomy of the kinds of experience that may occur in dreamless sleep. This is a research programme, not a result. Their own outstanding questions include which stages of sleep, if any, involve a genuine loss of phenomenal consciousness.

Thompson is worth citing for what else he has written. Why I Am Not a Buddhist (Yale University Press, 2020) attacks what he calls Buddhist exceptionalism: the claim that a contemplative tradition is really a mind science rather than a religion, that science can validate it, that it is wholly rational. Every one of those criticisms transfers to Hindu material unchanged, and saying so is the price of citing him at all.

What none of this establishes

The Upanishads did not solve the hard problem and did not anticipate it. Chalmers's problem is posed inside a framework of physical causal closure and functional explanation that did not exist in India in any century. A text that never assumed matter was fundamental cannot have foreseen the difficulty of extracting experience from matter. Asking a different question is not answering this one early.

No brain study demonstrates the metaphysics. Contemplative neuroscience measures reports and their neural correlates, and some of it is careful work. The strongest thing it can show is that trained subjects produce distinctive reports with distinctive signatures. Whether consciousness is fundamental is not a claim an electrode can adjudicate.

An open problem is not evidence for anything in particular. That science has not explained experience does not raise the probability that Vedanta is correct, any more than an unsolved burglary raises the probability that a particular person did it. This is the error most writing in this area makes.

And the Mandukya's conclusion outruns its own method. Verses 3 to 5 report what anyone can check. Verse 7 asserts what no report could establish, and verse 2's identification of Atman with Brahman is stated rather than argued. Advaita understood this: the tradition holds that scripture, reasoning and direct realisation each have separate work to do, and that reasoning alone does not finish the job.

What survives is narrower than the usual claim and much harder to knock over. One tradition asked how matter produces experience and built an extraordinary science on the parts of that question it could reach, leaving the middle untouched. The other asked what remains when the contents of experience are subtracted, and built a method out of a night's sleep. Neither holds the other's answer. The question Chalmers named in 1995 has not moved in three decades, and the twelve passages of the Mandukya were never addressed to it.