What Is Consciousness

Every field that touches this question uses the same word, "consciousness," to point at something slightly different. That's not a minor terminological nuisance, it's the reason arguments about consciousness so often talk past each other. So before I say what I think it is, I want to lay out the major ways it actually gets defined, disregard them one at a time for being incomplete on their own terms, and only then say where I land.

The clinical definition: a dial, not a question

Walk into a hospital and ask a neurologist what consciousness is, and you'll get a practical, two-part answer: consciousness is the combination of wakefulness (also called arousal, are you awake or in coma, stupor, or deep sleep?) and awareness (are you actually taking anything in?)¹. This framework is enormously useful, it's how doctors distinguish a patient in a vegetative state (awake, eyes open, but seemingly no awareness) from one who is locked-in (fully aware, but almost entirely unable to move or respond)². It's a dial you can point a scanner at: measure glucose metabolism for wakefulness, measure responsiveness to commands for awareness, plot the patient on a two-axis grid.

But notice what this framework quietly assumes rather than answers: it takes awareness itself, the fact that anything is being experienced at all, as a given, and just asks how much of it, and how alert, a given brain currently has. It's a definition built for triage, not for the deeper question of why there's anything it is like to be aware in the first place. Useful for medicine. Not what I'm after.

The philosophical definition: what it's like to be something

Philosophy draws a sharper line. The clearest version of it comes from Ned Block's distinction between two things people usually lump together under "consciousness": access consciousness, information that's available to your reasoning, your speech, your decisions, the stuff you can act on and report, and phenomenal consciousness, the raw, subjective, felt quality of an experience, independent of whether you ever report or act on it³. You can, at least in principle, have one without the other: information can be processed and used to guide behavior without any felt quality attached to it (this is roughly what most AI systems do today), and, more controversially, Block argues there might be felt experience that never makes it into the reportable, "accessible" part of your mind at all.

Phenomenal consciousness is the part philosophers actually find mysterious. It overlaps closely with what David Chalmers called the hard problem: the puzzle isn't explaining what the brain does, attention, memory, decision-making are all, at least in principle, explainable by ordinary mechanism, the puzzle is explaining why any of that processing is accompanied by felt experience at all, rather than happening in the dark, with no one home⁴. This gets closer to the actual target. But on its own, it's a description of the mystery, not a resolution of it, it names the gap without telling you what's actually on the other side of it.

The functional / levels-based definition: consciousness as a spectrum of capacity

A third approach, common in both cognitive science and popular writing, treats consciousness as something that comes in degrees, correlated with the sophistication of what a system can do, how much information it integrates, how flexibly it can respond, whether it can model itself, make choices, and so on. Integrated Information Theory, for instance, tries to give this a precise, measurable footing, defining consciousness in terms of how much integrated information a system generates. Global Workspace Theory takes a related but different functional route, treating consciousness as what happens when information gets broadcast widely enough across a cognitive system to be used by many different processes at once.

This approach is genuinely useful, it's utilitarian in exactly the way you'd want for medicine, for comparing species, or eventually for asking whether a given AI system has any of the relevant ingredients. But I think it has a specific failure mode worth naming directly: it quietly redefines the question. It starts by asking "what is consciousness," and ends by answering "what does a system need to be able to do for us to call it more conscious." Those aren't the same question. A thermostat that can integrate more information than a simpler one isn't obviously any closer to having an inner life, it's just a more capable thermostat. Treating "more integration" or "more function" as though it straightforwardly means "more felt experience" smuggles in exactly the assumption that needs defending, not asserting. Useful for utilitarian purposes. Still not what I'm after.

Where I land: awareness itself, separate from its contents

Here's my actual position, and I want to be precise about exactly how far it goes: consciousness, in the sense I care about, is not a level or a capacity. It's awareness itself, the bare fact that there is experiencing going on, as distinct from whatever that awareness happens to be filled with at a given moment. That much I'm settled on. It's the reason I disregarded the clinical and functional definitions above: both of them are answers to "how much" or "what kind of processing," and I'm after "is there anyone home," full stop.

What I am not claiming (and I want to flag this clearly, because it's a different and much bigger question) is that this awareness is actually separable from content. That it can exist on its own, with nothing in it, the way Advaita's account of turiya claims. I find that a genuinely interesting hypothesis. I don't currently have any way to confirm it. I haven't experienced awareness without content myself, not in meditation, not in any memory of deep sleep, and I'm suspicious of any account of consciousness that leans on a claimed experience the author hasn't actually had. So everything below should be read as exploring whether the separation is real, not asserting that it is.

Ordinary life gives us two states built from a combination of awareness and something else: waking and dreaming both consist of awareness plus content (sensations, thoughts, a world, a body, a story of "me" moving through it). The open question is whether, underneath both, there's ever awareness with nothing in it at all, or whether "awareness" turns out to be a concept that only ever shows up already attached to some content, in which case the very idea of separating them may be a conceptual move that doesn't correspond to anything real.

This is exactly the claim classical Advaita Vedanta makes with its analysis of the three ordinary states of consciousness, waking (jagrat), dreaming (svapna), and dreamless sleep (sushupti), and a proposed fourth, turiya, which isn't a state alongside the other three but the constant background awareness in which all three arise and dissolve, the "screen" the movie plays on rather than any particular scene in the movie⁵. The tradition's own analysis is careful here, and worth taking on its own terms rather than mine: it says deep sleep already gives us a clue, because on waking from dreamless sleep, people reliably report something like "I slept well, I was aware of nothing", which is itself a strange thing to be able to report, since it implies some thread of continuity was present to notice that nothing was noticed. Advaita's claim is that this thread is turiya, present the whole time, just without any content for it to illuminate.

And this is precisely the claim I find genuinely testable, and where I want to be careful rather than take the tradition's word for it: is there actually a difference between ordinary dreamless sleep and a state where awareness is present with no content? If dreamless sleep really is just the total absence of experience, nothing going on, no witness, full stop, then consciousness never actually occurs without content, and the whole "awareness prior to content" idea is a conceptual castle with nothing underneath it. But if some form of awareness genuinely does persist through dreamless sleep, even without content, that's a real, empirical foothold for the claim that awareness and content are separable.

This exact question, is dreamless sleep really dreamless, or just unremembered, turns out to be under live investigation, not just ancient speculation. Contemplative traditions across multiple lineages describe a trained ability often called "witnessing sleep": remaining subtly aware through deep, dreamless sleep, rather than blacking out into it⁶. The philosopher Evan Thompson has done the most careful cross-cultural work connecting this claim to contemporary sleep science, arguing that dreamless sleep may not be the total blackout the standard scientific default assumes, but may in some cases involve what he and others call "minimal phenomenal experience", the simplest possible form of consciousness, awareness with essentially no content⁷. This is not a settled question in sleep science, but it's a live and actively studied one, which matters a great deal for how seriously the underlying claim should be taken.

There's also a specific, useful empirical thread here: EEG studies of long-term meditators have found measurable differences in brain activity during deep, slow-wave sleep compared to non-meditators, including, in one line of research on Transcendental Meditation practitioners, greater power in a specific frequency band during deep sleep that correlated with practitioners' own reports of experiencing "higher states of consciousness" during that sleep⁸. A more recent study of elderly expert meditators similarly found EEG patterns during sleep consistent with partially preserved consciousness, rather than the total absence typically assumed⁹. None of this proves the contemplative claim outright, self-report during sleep is inherently hard to verify, and correlation with an EEG signature isn't the same as confirming subjective presence, but it means "can awareness persist through deep sleep" is no longer purely a philosophical thought experiment. It's an open empirical question with early, real data on both sides, and I plan to go much deeper into this specific research thread in a future article.

The tradition draws one more distinction worth flagging here, because it sharpens the whole question rather than just restating it: Ramana Maharshi, a 20th-century teacher in this same lineage, distinguished sahaja samadhi, a permanently established, natural awareness that persists through waking, dreaming, and deep sleep alike, without needing to withdraw from any of them, from nirvikalpa samadhi, a temporary, induced, eyes-closed absorption that some traditions treat as functionally close to a "yogically activated deep sleep"¹⁰. His point was that if the witnessing awareness only shows up when you shut out the world through effort, it hasn't actually been established as constant, it's just a state you can enter and leave. The stronger, more interesting claim is the constant one. That's a real distinction, not a hair-split, and it's the difference between "I can produce a special state" and "there's something present in every state that I've simply never noticed before."

Why this framing, and not the others

I'm not rejecting the clinical or functional definitions, they're doing real, different work, and I'll use them when the topic calls for it. What I'm rejecting is treating either of them as an answer to the question I actually care about. The clinical definition measures a dial without asking why there's anyone home to read it. The functional definition risks quietly substituting "does more" for "experiences more," which are not obviously the same claim. The philosophical hard-problem framing gets the target right but stops at naming the mystery.

What I'm carrying forward, then, isn't a settled answer but a sharpened question: is awareness ever present without content, even in principle, and if it is, is it something that's always there, prior to and independent of any brain or body that might host it, or is it something that gets switched on, generated by the right kind of system, and absent otherwise?

That's not a rhetorical flourish, it's the actual fork the next article has to deal with. If awareness turns out to be inseparable from content, always co-arising with some system complex enough to produce it, that points toward it being emergent: something the universe builds, under the right conditions, rather than something that was already there. If awareness can genuinely exist without content, if turiya, or something like it, is real, that opens the door to it being non-emergent: already present, everywhere or nowhere in particular, with brains and bodies acting as filters or channels for something that doesn't originate in them at all, rather than as the source of it. Advaita is explicitly a non-emergent account in this sense. Integrated Information Theory, Global Workspace Theory, and standard functionalism are, in different ways, emergent accounts. Panpsychism sits in an odd, interesting middle position I want to get into properly. That's the question the next piece takes on directly, and it's the one I actually don't know the answer to.