What this concept is actually saying
The word 有想 (yǒuxiǎng; Skt. *saṃjñin*, "having perception / having the mental factor of *saṃjñā* active") sounds almost like a tautology — of course a sentient being "has perception." The classical sources, however, are doing something far more precise. They are not pointing at the bare fact of being awake. They are pointing at a specific *cognitive operation*: the operation by which the mind takes a raw encounter with an object and, through a three-fold union (the meeting of sense-organ, object, and consciousness), marks it with characteristics, supplies a verbal/mental label, and grasps that label as if it were the thing itself. "Having-perception" means this marking-and-labeling function is *currently active*.
The foundational definition comes from the *Yogācārabhūmi-śāstra* (《瑜伽師地論》, T1579), which places 想 (*saṃjñā*; "perception / the perception mental factor") as one of the five omnipresent mental factors (五遍行心法, *sarvatraga*) that accompany every single act of consciousness. The five are 作意 (*manasikāra*, initial directing of attention), 觸 (*sparśa*, contact), 受 (*vedanā*, feeling tone), 想 (*saṃjñā*, perception), 思 (*cetanā*, volition/intention). The text defines 想 in a single dense line:
「想者，謂由三和合施設所緣，假合而取。」 *Perception is that which, through the three-fold union, constructs the object of attention and grasps it by provisional synthesis.*
Three things are packed into this line:
1. 「三和合」 — the union of sense-organ, sense-object, and sense-consciousness (e.g., eye + form + eye-consciousness). Perception is not a free-floating fantasy; it is the operation triggered the moment these three come together. 2. 「施設所緣」 — *constructing* (施設, *prajñapti*) what will be taken as the object. This is already an active shaping, not a passive reception. The mind is not a camera; it is a camera plus a caption-writer plus an editor. 3. 「假合而取」 — *grasping through provisional synthesis*. The "provisional" (假, *prajñapti*) signals that what is taken is not the thing-in-itself but a fabricated composite. And *grasping* (取, *upādāna*) signals that this operation is the cognitive root of clinging.
The text then describes the function of this perception in another memorable line: it 「於所緣，令心彩畫言說」 — it makes the mind "paint with words" upon its object. That is, perception turns raw encounter into a kind of internal painting-with-language, a marked-up, labeled, glossed picture. This is exactly what 想 means, and exactly what *having* 想 means.
Now, in the *Yogācārabhūmi-śāstra* the term 有想 shows up in four distinct but interlocking contexts. To understand the concept properly, all four need to be held together — they are not four competing definitions but four layers of the same operation.
1. The meditation context: having-perception versus having a specific perception. The text quotes a passage in which a monk (*bhikṣu*) is firmly established in the first *dhyāna* and, through the loathsomeness/breakdown of the eye, forms, and so on, comes to "have no eye-perception," "no form-perception," and so on — yet the text repeatedly adds, 「然有其想」 — "but there is still perception." The point is precise: what is given up is the ordinary characteristic-grasping perception of the eye as pleasant, safe, or self (the 「眼想」); what remains is a *different* perception, namely the perception formed by deliberately contemplating the eye as suffering, as accumulation, as illness (「為苦、為集，或為病等」). The technical phrase is 「不受自相」 — one no longer takes on the object's own characteristic in the usual way, but one has not become perception-less. The text says this range extends to the realm of nothingness (無所有處), and all of it is anāśrava (無漏, undefiled) — i.e., this is the contemplative work of insight (*vipaśyanā*), not ordinary distracted perception. The whole passage is teaching that *having-perception* (有想) is not a binary on/off switch. There are intermediate states: you can give up the perception of an object *as itself* while still maintaining perception of it *as suffering*, *as cause*, *as illness*. Only when one abandons all characteristic-perception does one cross into 無想 (aperception); only in the cessation attainment (滅盡定, nirodha-samāpatti) do *all* perceptions cease arising.
2. The mental-factor context: the internal structure of perception itself. Here the text distinguishes two kinds of 想 within the broader category of 有想: - 隨覺想 (*anugata-saṃjñā*, perception accompanied by comprehension) — the perception that goes with articulate understanding, e.g., the perception "human" or "divine" that a verbally fluent person forms. The label comes with its referent clearly understood. - 言說隨眠想 (*vāca-saṃjñā* in latent form, perception embedded in language) — the perception that operates in those who lack articulate speech: infants, animals, even insects. They still mark and label objects, but the labeling is more bodily and pre-verbal, a kind of "dormant speech-perception."
Both are still 有想. The distinction matters because it shows that perception-as-marking is not the same as perception-as-articulation. It runs deeper than language. Even before you can say "this is a cup," the perceptual apparatus has already taken on the cup-shape and frozen it as "cup-ish." This is why the text insists that perception's work is to 彩畫言說 — to "paint in language" — onto objects, and this painting can be articulate or pre-articulate but is never absent in ordinary cognition.
3. The sentient-being classification context. The *Yogācārabhūmi-śāstra* classifies all sentient beings along several axes simultaneously: by locomotion (no feet, two feet, four feet, many feet), by bodily form (with form, without form), and by mind: 有想, 無想, 非想非非想處有情 (beings-with-perception, beings-without-perception, beings in the realm of neither-perception-nor-non-perception). Crucially, the text states: 「有想、無想、非想非非想處有情，依心差別建立。」 This third classification is established on the basis of mental difference — not on bodily form. So "having-perception" is here a way of describing *kinds of minds*, not kinds of bodies. A being with a fine body can have coarse perception; a being without form (in the formless realms) can still have perception. The classification cuts across the body-classification.
4. The heterodox-speculation context. The text also records how certain non-Buddhist teachers, speculating about the self in future lives, posit a self that "has one perception" (一想) — correlating this with the formless realms of infinite space (空無邊處) and infinite consciousness (識無邊處) — versus a self with "various perceptions" (種種想), correlated with the lower realms. They further speak of "narrow perception" (狹小想) and "limitless perception" (無量想). The *Yogācārabhūmi-śāstra* includes these only as wrong views to be refuted (外道計度); they are not endorsed categories within the Buddhist framework.
These four contexts together give the concept its full contour. *Having-perception* is (a) a specific mental operation — the marking-and-labeling of objects that arises with every conscious moment; (b) a state of practice in which one has abandoned ordinary characteristic-grasping of certain objects but not all perception; (c) a way of classifying sentient beings by the kind of cognitive activity their minds sustain; and (d) a category that non-Buddhist thinkers misapply to a metaphysical self.
It is also worth connecting this to the wider Five-Aggregates (*pañca-skandha*) teaching, since 想 is one of the five and the most famous line about them — 「照見五蘊皆空」 in the *Heart Sutra* (《般若波羅蜜多心經》), and the explicit line 「受、想、行、識，亦復如是」 that places 想 alongside 受, 行, 識 in the same pattern — treats 想 as a fabricated, empty aggregate exactly as the *Yogācārabhūmi-śāstra* does. *Having-perception* is what it means for this fabricated aggregate to be currently active and still constructing its object.
Walking it through in contemporary life
The classical text anchors the analysis in meditation, but the operation itself happens on the subway, at the dinner table, in the group chat, at the office. Let me trace a few ordinary scenes and map them carefully onto the classical definition — and pause at each step to mark where modern intuition tends to slide off the track.
Scene 1: The morning commute. You step onto a packed train. The classical operation unfolds in three stages that the *Yogācārabhūmi* already names: - 三和合 is met: eyes encounter a face pressed against your shoulder, ears encounter the tinny podcast leaking from someone's earbuds, body-contact encounters the heat of fifty bodies. - 施設所緣: the perception function starts *constructing* what counts as the object. It does not just register "warm pressure on shoulder." Within a fraction of a second it has marked the encounter as "stranger," "too close," "annoying," and painted a small internal picture: an intrusion. The face is no longer *a face*; it is *an imposition*. - 假合而取: the constructed picture is then *grasped* as if it were the situation. Mood drops; jaw tightens; the commute is now "bad."
That second stage is precisely 有想 in the mental-factor sense: the marking-with-characteristics that turns an encounter into a labeled scene. The third stage is where it links forward to the next mental factor 思 (*cetanā*, volition): the labeled picture now steers what you do — you shift your weight, you frown, you rehearse a complaint in your head.
Now compare that with the meditation-context meaning of the term. Suppose, still on the same train, you have been practicing for some time. You notice the shoulder-pressure arises. The ordinary characteristic-grasping perception — "intrusion," "too close," "I am being invaded" — is given up. The text says this is 「無有眼想……然有其想」 — you have no "eye-perception" in the ordinary sense, yet there is still perception: you contemplate the contact as suffering (為苦), as arising from causes (為集), as something sickly (為病). That is not numbness. That is a *replacement* of one perception with another — and the replacement is exactly what the text calls 有想 at this more refined level. It is perception functioning, but no longer on the object's "own characteristic" (自相); it is functioning on the mark "suffering" instead. The classical line 「不受自相」 means: you are no longer letting the object paint its ordinary picture onto your mind. You are, in a controlled way, painting a different picture. *Having-perception* does not stop; it is *redirected*.
Scene 2: Late-night scrolling. Pick up the phone. The screen lights up. Within the first ten seconds, the perception function has already executed 施設所緣 on the thumbnail — "interesting," "this person again," "I don't need to see this," "oh, a recipe." Each tap is downstream of a label. The genius of the feed is that it optimizes precisely for this *saṃjñā* operation: every thumbnail is a designed 彩畫言說 — a "painting-with-language" condensed into a single image-and-caption pair that hijacks the labeling function and turns it into a click.
This is where the 言說隨眠想 distinction becomes relevant. Even before you read the caption, your perception has already done an implicit, pre-verbal marking: the thumbnail "is" whatever category it triggers. That is the 言說隨眠想 of the adult scroller — the verbal labeling is dormant in the perception, awaiting only the slightest cue to fire. Infants and animals operate in pure 隨眠 mode all the time; adult humans usually operate with a thin, fast verbal overlay on top of it. Both are 有想.
Scene 3: The argument with a partner. Partner says something. The classical operation is identical to the train scene, only now with much more at stake. - 三和合: ears meet the partner's words; memory supplies prior context; consciousness arises. - 施設所緣: perception marks the words. Crucially, it marks them as an attack, or as a complaint, or as a misunderstanding — almost never neutrally. The marking is almost instantaneous and almost invisible to the speaker. - 假合而取: the mark is grasped as the *meaning* of the words. The conversation is now "about" what the marking said, not about what was said.
Now the *Yogācārabhūmi*'s meditation insight becomes available in a very practical form. The text says: you can give up the eye-perception (the ordinary characteristic-grasping, "he is attacking me") while retaining perception-as-it-is — perceiving the words as 苦 (suffering-toned), as arising from 集 (causes — his tiredness, your own defensiveness), as something 病 (ill — sick because untreated). You do not have to stop perceiving. You stop *letting the object paint its usual picture*. The 「然有其想」 is the perception that sees the situation through the lens of the four noble truths (or the three marks, or whatever contemplative frame you inhabit), rather than through the lens of "I am being hurt / I am being blamed."
This is a particularly important place where modern readers mis-map the concept. The reflexive assumption is: *if I am still having perceptions, I am still caught*. The classical text says: there is a mode of 有想 in which perception functions but the ordinary self-protective painting-with-language has been replaced by a contemplative one. That is not bondage; that is practice.
Scene 4: The baby's face. An infant cannot yet speak. But as the *Yogācārabhūmi* notes, 言說隨眠想 is already present. The infant sees the mother's face and its perception function is fully active: it marks the face, supplies a label (in pre-verbal, bodily form: *familiar, safe, food-source*), and grasps the label. The infant smiles. This is 有想 in the 言說隨眠想 sense — perception embedded in language-that-is-not-yet-speech. It is also the most basic form of cognitive life: marking is already happening before articulation is.
Scene 5: The person at the computer trying to focus. You sit down to write a report. Within three minutes, the perception function has already labeled the page as "boring," "hard," "I don't know where to start." Each label is a 施設所緣 — a small painting-with-language — that then hands the steering wheel to 思 (volition), and volition reaches for the phone. The classical diagnosis is not that you lack willpower; it is that the *perception function has already done its painting, and the painting is now what volition is acting on*. Practice, in this frame, is not first a matter of forcing volition to obey; it is a matter of pausing the perceptual painting — or replacing it with a different painting ("this is suffering; this arises from causes; this is sick") — so that volition has a different picture to act on.
The place most people mis-map the concept here is to think 有想 = 想法 (thinking in the sense of inner speech, ideas, opinions). It is not. 想法 in the casual sense is closer to the mental factor 思 (cetanā, volition-with-thought) or to 行蕴 (saṃskāra, formations) broadly. 有想 is the *prior*, much quieter, much more pervasive operation of marking that conditions all your thinking but is not itself "a thought." This is why the *Yogācārabhūmi* bothers to define it as 施設所緣 and 假合而取 rather than as "having ideas."
Why contemporary people need this concept
We live inside an apparatus that is engineered, from the feed to the office dashboard to the smart speaker, to exploit the operation the *Yogācārabhūmi* calls 想. The platform's job is to 彩畫言說 — to paint with language — onto every object of attention, faster and more persuasively than you can. By the time volition arrives to act, the painting has already been done for you. Every modern piece of attention-economy engineering — thumbnails, captions, push notifications, micro-targeting — is a refinement of the very marking function that classical Buddhism isolates as one of the five omnipresent mental factors and subjects to its most careful analysis. The concept 有想 gives you a name for what is being engineered *against* you.
It also gives you a *non-binary*, *non-romanticized* picture of contemplative progress. The meditator's progress is not a steady march from "having perception" to "having no perception." The classical text shows a much more textured movement: one gives up the ordinary perception *of this object*, while keeping a refined perception *of the same object as suffering*; then one gives up the perception of *that* object, while still perceiving the situation *as sick*; only at the far end does 一切諸想皆不生起 (all perceptions cease arising) — and even that is the very rare, very specific context of the 滅盡定 (cessation attainment). For modern meditators, this matters enormously. It means the goal is not to blank out, not to "stop thinking," not to become a smooth stone. The goal is to redirect the perceptual painting from the object's own characteristic (自相) to the marks of the three marks / four truths, and eventually to step out of the painting entirely — but not before, and not as a permanent achievement at every moment.
A third reason: the concept dignifies the perception function rather than dismissing it. The *Yogācārabhūmi* does not say 想 is the enemy. It says 想 is omnipresent, ordinary, and the basis on which language, conceptual thought, and the entire human world of meaning is constructed. Without 想, there is no human cognition. The contemplative task is not to annihilate it; it is to see through the 假合 — the *provisional* synthesis — that it performs. The world's pictures are useful; they are also fabricated. Holding both at once is what the practice of 有想, understood properly, is for.
A fourth, more subtle reason: the concept's inclusion of 言說隨眠想 — perception embedded in dormant speech — gives a classical handle on something very contemporary, namely that much of human suffering arises from pre-verbal labeling. The infant has it; the adult has it; the trauma survivor has it; the algorithmic consumer has it. The label arrives before the thought, and by the time thought tries to intervene, the label is already load-bearing. A framework that takes 言說隨眠想 seriously lets you see that what needs interrupting is not necessarily your *ideas* but the marking operation that runs underneath your ideas, marking and labeling fast and silently, before the conscious mind shows up.
Common misreadings and clarifications
Misreading 1: 有想 = 想法 (having thoughts / ideas). This is the most common and most damaging mis-translation. It collapses a specific mental factor into a vague colloquial notion. 想 in the Buddhist technical sense is the *marking* function — the operation that supplies characteristics, builds a picture, and hands the picture to volition. "想法" in modern Chinese tends to mean something more like *opinions* or *inner speech*, which is closer to 行蕴 (saṃskāra) or to 思 (cetanā) working together. If you treat 有想 as "having thoughts," you will conclude that meditation aims at *thoughtlessness*, and you will spend years chasing a state of blank mind. The classical text is not pointing at that. It is pointing at the much quieter, much more pervasive operation that runs even when no explicit thought is occurring.
Misreading 2: 有想 = being conscious / being awake. This is the loose translation that loses the precision. A stone is not 有想; a fetus may or may not be 有想 depending on whether the cognitive apparatus has matured enough; a dog is 有想 (in the 言說隨眠想 form); a being in deep dreamless sleep is not 有想 in the active sense; a being in 滅盡定 is not 有想 in *any* sense. The classical text is using 有想 to refer specifically to the active exercise of the perceptual marking function, not to bare sentience. To translate it as "being conscious" is to erase the very thing the concept exists to isolate.
Misreading 3: giving up "eye-perception" means all perception has stopped. The classical text anticipates and explicitly blocks this misreading. It says the monk 「無有眼想」 but 「然有其想」 — the perception of the eye *as itself* is gone, but the perception of the eye *as suffering / as arising from causes / as sick* is still functioning. The text goes out of its way to say this is 「不受自相」 (not taking on the self-characteristic), not 「一切想不生起」 (no perceptions arising at all). The latter phrase is reserved for the 滅盡定, which is a separate and very rare attainment. Anyone who teaches that abandoning any specific "perception" (e.g., of the body, of pain) means the perception function has shut down is contradicting the text.
Misreading 4: 有想 and 無想 are physical / bodily categories. The *Yogācārabhūmi* is emphatic that 「有想、無想、非想非非想處有情，依心差別建立」 — these categories are established on the basis of mental difference, not bodily difference. So 無想 is not "the body without thought" in any physical sense; it is a mind in which the marking function has been suspended at a particular level. This is also why the text can speak of formless-realm beings as still 有想 — they have no body, but they have active perception of the formless object (infinite space, infinite consciousness, etc.). Misreading 有想 / 無想 as bodily categories collapses the entire framework.
Misreading 5: the "perception of the eye as suffering" is a trick or a trick label. Some modern readers, encountering the line 「於眼作意思惟為苦、為集，或為病等」, hear it as a rhetorical device: just call everything suffering and you are done. This is not what the text is doing. The contemplative re-marking of the eye as suffering / as cause / as sick is a *contemplative perception* built on the analysis of the three marks (*tilakkhaṇa*) and the four noble truths. It is not a slogan; it is the actual practice of seeing the object through a different lens, in which the marking function is being deliberately redirected from 自相 (the object's own pleasant or neutral characteristics) to 共相 (shared marks such as impermanence, suffering, and non-self). The text is careful to note that this re-marking is *still 有想* — it is a refined perception, not the end of perception. To use it as a slogan is to mistake the contemplative act of painting-with-language for the perception that is being painted.
Misreading 6: 有想 = saṃskāra (行). Both are translated loosely as "mental activities" in English, but they are different. 行 (saṃskāra, formations) is the broader category of volitional, habitual, karmic constructions — including volition, decision, intention, and the habitual patterns that propel samsāra. 想 (saṃjñā) is the *narrower*, *more specific* operation of marking. Within the Five Aggregates, 受想行識 are listed separately, so collapsing 想 into 行 is a category error. The *Yogācārabhūmi*'s definition — 「由三和合施設所緣，假合而取」 — is the test: if the operation in question is about *constructing the object*, it is 想; if it is about *volitional shaping* of action and habit, it is 行. Conflating them loses the precision the concept is built to provide.
If you want a single sentence to carry the concept into a working day, it might be this. *Having-perception* is the mind's ongoing, often-invisible operation of marking every encounter with a characteristic and a label, and then treating the label as the encounter itself. The classical teaching is not that this operation should be killed but that it should be seen: seen as a fabricated synthesis rather than as the world's given texture, redirected from the object's own pleasant face to the marks that liberate attention, and ultimately — at the rare and specific threshold of 滅盡定 — let to cease. Everything else, including most of contemplative life, happens *within* 有想, not outside it.