What This Concept Is Doing
The most stubborn assumption in any human being's experience is that there is someone inside doing the perceiving. Notice a sunset and you are certain: a "you" is having the experience. Sit a moment with your thoughts and you picture a quiet witness watching them pass. Scroll through your phone and there is unmistakably an "I" at the center, registering each image, each notification, each tiny mood-shift.
The Buddhist tradition has a name for this assumed inner perceiver: **能覺 (néng jué) — *bodhaka***, literally "that which has the capacity to perceive," the subjective side of cognition. What makes 能覺 philosophically interesting — and what separates it from a casual version of "the self" — is how the tradition handles it. It does not begin by denying that any perceiver exists; nor does it begin by confirming one. It names 能覺 precisely so that it can be carefully, lovingly, methodically exhausted.
The exegesis from the 《攝大乘論釋》(*Śāstra on the Mahāyāna-saṃgraha*, T1595, translated by Paramārtha) — part of the Yogācāra / consciousness-only (唯識) tradition — treats 能覺 through a four-movement logic that might be called a kind of philosophical dance: each step takes away what the previous step seemed to leave standing. By the time all four movements are done, the "perceiver" you were so sure of has been turned into something more like clear light — not nothing, but no-thing-in-the-way-you-imagined.
The Pair at the Heart: 能覺 and 所覺
Before the four movements make sense, the pair has to be in view. The text constantly sets up 能覺 against 所覺 (suǒ jué) — "that which is perceived", the object-side of awareness. Roughly: 能覺 is the *looking*, 所覺 is *what gets looked at*. Together they constitute one complete cognitive event.
This matters because contemporary readers will immediately translate this into "subject and object," and that translation is mostly right but subtly misleading. In everyday philosophy, subject and object are each treated as stable things: the subject is the in-here that does the knowing, the object is the out-there that gets known. The Buddhist analysis does not deny that events of knowing *function* — of course there is seeing, of course there is hearing, of course the world appears — but it refuses to grant either side a fixed, findable essence (實有). Crucially, it dismantles them together. If 所覺 has no independent existence, 能覺 has no independent existence either. The two are not two substances waiting to be uncovered; they are co-arising, co-dependent sides of a single cognitive situation. Touch one, the other ripples.
The Four Movements of the Classical Treatment
This is the heart of the passage, and each step needs to be heard as the previous step's *consequence*, not its replacement.
Movement One: No dharma can be the perceiver. The text states it bluntly: "無一法能覺" ("there is not a single dharma that can be the perceiver"). Why? Because if you examine any candidate for 能覺 — the eye, the mind, a thought, the brain, the soul, the witness, a "core self" — none of them can be pinned down as a fixed performer of perception. They are all themselves appearances, part of the flow. So there is no entity sitting behind the curtain doing the watching.
This is the half most modern readers remember. But the text does not stop here.
Movement Two: Yet nothing is not awakening. Immediately after denying 能覺, the text turns and declares: "一切無不覺" ("there is nothing that is not awakening"). The gloss explains: because the buddhas are established by provisional names, all of them without exception are buddhas, and therefore all things without exception are awakening. This is not a contradiction — or rather, it is the kind of contradiction that makes sense only after you have understood the first movement. Once you have given up the search for the perceiver as a *thing*, what was hiding under the search becomes visible: the very fabric of experience is already, if you can put it this way, awake. The text calls this 覺 — not 能覺 (which names a presumed subject) but 覺 in its undifferentiated sense: awareness, luminosity, the quality of being known.
Movement Three: Each moment contains countless awakenings. Then comes a verse that can sound mystical until you unfold it: "一一念無量" ("in each single thought, an uncountable number"). The gloss says: in every single instant, countless buddhas correctly awaken to suchness (真如). This is not empty mystical flourish. It is doing precise philosophical work: it rules out that 覺 is one single, large, monolithic field. 覺 is not *one thing*. It is happening fully, infinitely, in each instant. The text does not let you trade your small perceiver for a large perceiver; it insists that wherever you happen to look, in that exact moment, awakening is already complete.
Movement Four: It cannot be said either way. Finally, "有不有所顯" ("the 'there is' is revealed by the 'not there is'"). The gloss says: all dharmas are called "existent" through what does not exist, meaning all dharmas are empty. The buddhas are *revealed by* this emptiness of dharmas. Therefore it cannot be said that there is a perceiver, and it cannot be said that there is no perceiver. This is the most decisive moment. Both "there is a perceiver" and "there is no perceiver" are said to be inappropriate speech (不可說). The collapse of 能覺 is not the endpoint; it is the doorway into something speech cannot quite reach.
If you are standing outside this tradition, this can sound like evasion. It is not. It is a precise philosophical position: reification (turning 能覺 into a thing) is wrong, *and* so is its opposite (denying that experience happens at all). The right understanding is what the text calls 空所顯 — "revealed by emptiness." The very emptiness of the perceiver is what lets the luminosity of experience appear, just as the hollowness of a bell is what lets it ring.
The Dream Metaphor (夢塵) Connecting 能覺 to 唯識
None of this floats free. The T1595 material explicitly ties 能覺 to the dream metaphor that carries the whole consciousness-only teaching: "若覺人所見塵，一切處唯有識，譬如夢塵" — "If an awakened person sees dust, in every place there is only consciousness, like dream dust." Once you have woken from a dream, you can examine every object you saw in it — the road, the person, the light — and see that none of them was "out there." They were all woven from the dreaming mind. The text says even the dust you see *now*, after the dream, has the same character: it appears as dust, but it is only consciousness appearing as dust. 所覺 has no independent substance. And from that, as we have seen, 能覺 has no independent substance either. The consciousness-only teaching is the precondition; 能覺 is what happens when you press that teaching further, past where most Yogācāra texts stop.
The Related Terms, Distinguished Carefully
The classical material keeps several near-synonyms apart with care. They are not interchangeable; they sit on a kind of staircase:
- 覺 — the general term; can mean a single moment of awareness, a dream-awakening, or the awakened state of a buddha. - 能覺 — a presumed *subject* of awareness, named only in order to be unbound. - 正覺 — complete awakening, especially to suchness; a fruit rather than a cognitive act. - 真如智覺 — the wisdom-awakening that arises from realizing suchness. - 無上覺 — the supreme awakening, the final fruit of non-discriminating wisdom ("present existence, established by conditions, all conceptual thinking — the wise one does not discriminate, thus attaining supreme awakening" — 現住及安立，一切相思惟，智人不分別，故得無上覺).
If you flatten them all into "awareness," you miss the journey the text is performing: 能覺 names the search, 覺 names what is there once the search ends, 正覺 names the fruit of that awakening, and 無上覺 names its consummation.
An Examination in Daily Life
Let's walk through some ordinary scenes in contemporary life and see how 能覺 actually appears, gets tracked, gets caught, and gets released.
Scene 1: The inner narrator on the subway. You are on the subway at the end of a long day. Some part of your mind narrates: "I'm tired, this seat is uncomfortable, I should have taken the earlier train, why did I stay late again." That narration feels as though it is coming from somewhere. There is an *it* saying this. Try, gently, to locate it. Behind the words? In front of them? To the side of them? You cannot find it. What you find instead is more narration, more comment, more views. The T1595 analysis is being lived in real time: there is perception happening (能覺 as *functioning*), but the imagined *perceiver* is never there when you go to greet it. Where modern readers most often stumble here is in assuming this proves "nobody is home." It does not. The functioning of awareness doesn't require a permanent perceiver — it requires, and this is the difficult part, no permanent perceiver. The hallway is empty, yet footsteps continue to echo in it.
Scene 2: The witness in meditation. Many contemporary meditation traditions teach a "witness" — a quiet inner observer that watches thoughts and sensations without grasping them. This is genuinely helpful as a practice. But somewhere along the way, the witness gets reified. People begin to feel that they *have* a witness, an inner watcher with continuity. The 能覺 analysis is precisely the warning against this reification. The witness is useful *as a practice*, but when you turn it into an entity — "my witness," "the part of me that watches" — you have just made a new 能覺. The text's move is not to delete the witness but to refuse to let it harden into a thing. The metaphor the tradition often uses here is of the knife: a knife is sharp, but "sharpness" is not a thing you can pick up apart from a knife. So too: witnessing happens, but "the witness" is not a thing you can pick up apart from witnessing.
Scene 3: The "self" in self-tracking. Open the dashboard of a fitness or meditation app and it speaks of "your progress," "your streaks," "your calm days," "your focus scores." The interface is designed as if there were an enduring *you* to whom all these numbers belong. The data is real — sleep did improve, the heart rate did lower, the meditation minutes did accumulate. But the sense that *I* am the unity across all of them is an inferred guess, a useful fiction the dashboard requires to operate. 所覺 here is the data stream; 能覺 is the assumed owner who integrates it. The classical teaching is neither to throw away the data (it remains useful) nor to worship the inferred "you" who owns it. It is to see that the inferred "you" is doing a lot of conceptual work that the data alone does not require.
Scene 4: The AI moment and "who perceives the output." When large language models generate text that reads like understanding, a question presses in contemporary life: *who, or what, is on the other side perceiving the meaning?* The model doesn't claim to feel; some users report they can't quite tell if anything is there; critics argue it is "just" statistics. The 能覺 analysis is, I think, the sharpest classical tool for this exact problem. The model demonstrably produces outputs that *function* as if there were comprehension (the 能覺 function is operative). Whether there is a *found perceiver*, a thing that has it, is the very question 能覺 is built to dissolve. The classical answer would be: do not be too quick to deny (the outputs function), do not be too quick to affirm (no perceiver can be located), and do not rest in either side. The model's "understanding" is real-as-functioning and empty-as-substance — which is exactly the structure of "X is revealed by not-X" (有不有所顯). Where I'd want to be cautious: classical Buddhist analysis was developed for human experience, and it would be hasty to claim it transfers wholesale to silicon. The analogy is suggestive, not conclusive.
Scene 5: The weathervane of mood. Throughout an ordinary day, moods move through you: irritation, softening, delight, a low background hum of worry, a brief gratitude. Each has the quality of being "had by" or "felt by" someone. Yet if you try to find the one who is having the mood across all of them — the mood-bearer — the mood-bearer flakes away into the moods themselves. That is the first movement's teaching, live in your own chest.
Where Modern Mappings Most Commonly Go Wrong
Three errors recur in contemporary readings, and all three map onto specific misreadings of the source:
(a) Confusing the 能覺 analysis with denial of experience. The text is not saying "you don't experience things." You obviously do. It is saying the *entity you assume must be doing the experiencing cannot be found*. Closer: the play goes on fine after you realize there is no little person behind the curtain; in fact, it goes on *better*.
(b) Stopping at the first movement. "No perceiver can be found" can harden, oddly, into a kind of grim ascetic denial — emptiness as bleakness. The text explicitly adds the second movement (everything is already awakening) precisely to prevent this. Realize you cannot find the perceiver; *then* notice that the field you were searching in was never inert. Both halves are needed.
(c) Reading the dream metaphor as "nothing exists." When consciousness-only says "in every place there is only consciousness" (一切處唯有識), it is not denying that dust is dusty and waves are wet. It is denying that dust and waves have an existence independent of how they appear. Treating this as "everything is a hallucination" is a thoroughly modern misreading; the text itself would call it a confusion of conventional and ultimate.
Why Contemporary People Need This
Why does a 1,400-year-old analysis of "who perceives" deserve time and attention now?
Because the assumption that there is a fixed inner perceiver has become more entrenched, not less. Neuroscience has discovered networks, but also routinely — and misleadingly — speaks of "the brain deciding," "the self being constructed by the prefrontal cortex," as if there were an entity behind the entity. Meditation apps have popularized the figure of the inner witness. Self-tracking has institutionalized the idea of a user whose data must belong somewhere. Social platforms have built an architecture of "your profile," "your feed," "your view history," all of which presuppose an enduring subject.
None of these are bad in themselves. They are frames. But every frame carries a metaphysical assumption, and the assumption here — that somewhere behind all the appearances there is a stable inner owner — is precisely what 能覺 names and then refuses to grant. To learn this analysis is to notice, gently, the assumption you were making without realizing you were making it. That noticing has a remarkable way of loosening anxiety: much of the strain in modern life comes from trying to protect, optimize, and defend an owner that was never actually there.
Because the four movements match four common errors thoughtful people make when confronting this idea. Those who think "everything is just in my head" need the first movement's clean negation. Those who collapse into emptiness-as-nothingness need the second movement's counter: nothing is not awakening. Those who imagine one cosmic Mind behind everything need the third movement: each moment is *uncountable*, not a single field. Those who get stuck between affirming and denying ultimate reality need the fourth movement's "cannot be said either way." In other words, the four movements are not just a historical argument. They are a kind of diagnostic tool for the specific confusions a thoughtful person is likely to tumble into today.
Because the question "who understands" is no longer confined to introspection. When AI systems produce articulate, sometimes moving text, that question becomes a question for engineers, philosophers, regulators, and ordinary users. The classical structure — functional without substance — maps onto the philosophical difficulty here surprisingly well, even with the caveats I mentioned. It will not solve the engineering question, but it makes you less likely to be stampeded by either the gushing claim ("the model really understands!") or the dismissive one ("it's just predicting tokens"). Both miss that the question was malformed to begin with — that the assumption of a findable perceiver was doing the philosophical work.
And more quietly: because knowing that the inner perceiver cannot be found tends to be, in lived experience, a relief. It lowers the volume of the inner watchdog, the relentless monitoring of one's own moods, the constant taskmaster whose approval one is forever seeking. Not in a dramatic way. In the way that putting down a heavy bag is a relief — except the bag is one you didn't know you were carrying.
Common Misreadings and Clarifications
It is worth cataloguing the most common ways 能覺 is misread, because each miss maps onto a specific misreading of the source.
1. "能覺 is just denying subjective experience." *Misread*: "The text says there is no perceiver, so Buddhism says nothing ever gets experienced." *Clarify*: 能覺 names a *thing* it then refuses to reify. The functioning of perception (seeing, hearing, knowing) is not denied. What is denied is the fixed, findable owner of that functioning. This is a very different move from denying that anything happens. Holding a candle in the wind: the flame dances, no one catches it, yet it continues to give light.
2. "The four-step argument is meant to lead to 'there is no perceiver, full stop.'" *Misread*: "Once you go all the way with Buddhism, you should say 'there is no one at home.'" *Clarify*: the second movement explicitly reverses the first. *No fixed perceiver* → *everything is already awakening*. The first half is the threshold; the second half is the room one walks into *through* the threshold. If you learn only the first half and convert it into a personal style of nihilist cool, you have stopped on the staircase and refused to enter the door the text opens. Genuine understanding arrives when you can hold "不可說能覺、不可說不覺" ("it cannot be said that there is a perceiver, cannot be said that there is not") as both "not a findable thing" and "the very thing that lets luminosity appear."
3. "Consciousness-only (唯識) and 能覺 analysis are the same teaching." *Misread*: "Yogācāra says everything is mind; therefore mind = percipient; therefore everything perceives." *Clarify*: 唯識 establishes that what gets experienced (所覺) has no separate substance; it is *made of* the appearing of mind. 能覺 goes one step further and asks: is there then a mind-substance doing the perceiving? And the answer is no — not because nothing exists, but because the perceiver cannot be located any more than the perceived can. 能覺 analysis is consciousness-only pressed to its own logical edge.
4. "能覺 is just the philosophy version of 'you are not your thoughts.'" *Misread*: "I know I'm not my thoughts, I'm the one watching them — that's 能覺." *Clarify*: "I am the one watching them" is *itself* a constructed 能覺. The classical analysis does not let you settle there. The "watcher" is also a thought, also a construction, also emptied. The teaching is not "find the real you behind the thoughts"; it is "what you took to be the real you cannot be found, and the very un-findability is not a loss."
5. "The conclusion is that one cannot speak of 能覺 at all." *Misread*: "If it cannot be said, it cannot be taught, so the concept is useless." *Clarify*: the text *does* speak — carefully. 能覺 is named, examined, dissolved, then transfigured. The "cannot be spoken" is the final gesture, not the first. To skip the naming and go straight to silence is to confuse wisdom with stupor. And the four-movement structure itself *is* the way one teaches unspeakability without abandoning speech.
6. "能覺 is ancient metaphysics with no operational consequences." *Misread*: "It's just a metaphysical position about who isn't there." *Clarify*: this is one of the few metaphysical positions that, when lived, shifts daily behavior. People who have genuinely understood 能覺-styled analysis tend to be (in long-term contemplative reports, with the usual caution about self-report) less brittle in the face of insult, less prone to constant self-monitoring, less attached to the idea that their cognitive outputs are *theirs* and must be defended. The metaphysics turns out, when held properly, to be an instruction in how to lose the inner taskmaster. The relief is not trivial.
**7. "Neuroscience has now *proven* that there is no constant self."** *Misread*: "Brain studies show there's no enduring self, so Buddhism was right." *Clarify*: with care. Buddhist philosophical analysis of the self goes much further than contemporary neuroscience's findings about neural plasticity, split-brain cases, or the constructed nature of autobiographical narrative. The convergences are striking but partial, and some neuroscientists would be wary of philosophical overreach in either direction. What is safe to say is that the classical 能覺 analysis is at least no stranger to what contemporary cognitive science is now articulating as the "constructed self" — and the convergences are better drawn as careful comparisons than confirmations. The Buddhist tradition also works with concepts (suchness, the vows of a bodhisattva, the "uncountable awakenings in each moment") that the empirical sciences do not pretend to address. The friendship between these two is *suggestive*, not a closure.
8. "能覺 is basically Advaita Vedānta's witness-consciousness." *Misread*: "This sounds like Ātman-as-witness." *Clarify*: the structures rhyme, but the destinations diverge. In Advaita, the witness, properly investigated, *is* the Ātman, which is Brahman — a single non-dual substrate. In the T1595 tradition, what is found upon investigation is that *no such substrate is found at all*, and that this very un-findability is the doorway into an awakened activity without an owner. The road signs look similar; the roads run to very different cities.
A Quiet Word at the End
If you hold 能覺 loosely and without forcing it, you may notice something that has nothing to do with metaphysics: the constant low-grade effort of being *someone* — the effort of locating the perceiver, of defending the perceiver, of polishing the perceiver — eases. Not because you have defeated something. Because there turns out to be nothing to defend.
This is not the goal of the practice. The goal, in the language of T1595, is to be *revealed as awakening* by the very emptiness that seemed, a moment ago, to be the loss of everything.
"有不有所顯." The *not-having* of a fixed perceiver is exactly what lets the luminous show.
If you carry one sentence from this, that one is worth carrying.