A contemporary reading of one of the most precise, and most easily misread, terms in the Yogācāra tradition.
What this concept is actually saying
Most contemporary readers, hearing the word "image," picture something like a daydream, a vision, or a piece of mental imagery one conjures up. The classical term 有分別影像 (savikalpa-pratibimba, "image-with-analysis") wants to take you somewhere quite different. It is a technical term from the *Yogācārabhūmi-śāstra* (《瑜伽師地論》, T1579) describing one specific kind of object used in deep meditation practice — and it sits at a very particular place on the Buddhist map of liberation.
In the *Yogācārabhūmi*'s framework of 遍滿所緣境事 (the four "comprehensive objects of mental engagement"), the meditator's mind is understood to work on four kinds of objects that together cover the entire path:
1. 有分別影像 — the discriminated image (object of vipaśyanā, 毘鉢舍那) 2. 無分別影像 — the non-discriminated image (object of śamatha, 奢摩他) 3. 事邊際性 — the limiting/ultimate nature of phenomena 4. 所作成辦 — the accomplished purpose, i.e., the fruit
The first two are paired. Both of them are 三摩地所行影像 — "images present within samādhi (deep meditative absorption)." Both of them are 所知事同分影像 — "images that are *similar copies* of the thing to be known," generated by the meditator's concentrated resolve (勝解). The difference is what the mind *does* with that image.
- 無分別影像 (avikalpa-pratibimba): once the image is taken up, the mind stops analyzing. It lets the image be, settles into it with 奢摩他 (śamatha, calm abiding), and remains one-pointedly. - 有分別影像 (savikalpa-pratibimba): the mind actively dissects the image. This is the object of 毘鉢舍那 (vipaśyanā, clear insight), and the analysis it performs has a precise classical name-list: 觀察、簡擇、極簡擇、遍尋思、遍伺察 — *observing, discriminating, finely discriminating, thoroughly seeking, thoroughly scrutinizing*.
So 有分別影像 is, literally, "the kind of image that one analyzes." It is not "an image that contains analysis" in any ordinary sense — it is the image-as-analyzed, the image being worked on by insight.
This is why the *Yogācāra* gives the image so many aliases that all hint at the same thing: 三摩地相 (sign of samādhi), 三摩地所行境界 (sphere of samādhi), 三摩地口 (mouth of samādhi), 三摩地門 (gate of samādhi), 作意處 (field of applied thought), 內分別體 (inner discrimination-substance), 光影 (light-shadow). All of these point to the same idea: this image is what the concentrated mind does its insight-work on. It is the working surface of vipaśyanā.
What kind of "thing to be known" is being imaged?
The *Yogācārabhūmi* is unusually thorough in listing what kinds of dharma the meditator generates a 同分影像 for. They include:
- 不淨 (aśubha, impurity-of-the-body contemplation) - 慈愍 (maitrī, loving-kindness) - 緣性緣起 (pratītyasamutpāda, dependent origination) - 界差別 (dhātu-vyavasthāna, distinctions of elements) - 阿那波那念 (ānāpānasmṛti, mindfulness of breathing) - 蘊、界、處 (the aggregates, elements, sense-bases) - 緣起、處非處 (dependent origination, the "proprieties and improprieties") - 下地麁性與上地靜性 (the coarser nature of lower absorptions and the quieter nature of higher ones) - 苦集滅道四諦 (the four noble truths)
Notice what is and isn't here. The list does not include "whatever happens to pop into your head" or "any fantasy image." It includes only the canonical objects of contemplation taught in the Buddhist curriculum of training. The meditator doesn't choose an arbitrary mental picture; the image must be a 同分 (homologous, same-kind) likeness of one of these training objects. It is a meditative analogue, produced by 勝解 (adhimukti, concentrated resolve) within samādhi.
Why this sits where it sits on the path
This is where 有分別影像 becomes easier to grasp: it is the *hinge* between calming and understanding.
A meditator who has stabilized attention (奢摩他品 — the 無分別影像 side) now needs to develop 慧 (prajñā, discriminating wisdom). To do that, they need something to analyze. They cannot analyze the "outside world" — that would be ordinary thinking. They cannot analyze nothing — that would be blankness. So they generate, in deep absorption, a stable mental image that *is like* impermanence, *is like* the four noble truths, *is like* dependent origination, and on that image they apply vipaśyanā.
Hence the *Yogācāra*'s careful pairing: 有分別影像 ↔ 毘鉢舍那品, 無分別影像 ↔ 奢摩他品. Same kind of object, different mode of engagement. The path's two wings, śamatha and vipaśyanā, take off from the same runway.
The crucial 唯識 (consciousness-only) qualification
Then comes the move that protects the meditator from a very subtle trap.
The *Yogācārabhūmi* says: "毘鉢舍那三摩地所行影像與心無異，唯是識故" — *"The image present in vipaśyanā-samādhi is not different from mind, because it is consciousness-only."*
This is doing real philosophical work. A naive meditator might think: "There is the image, and there is me analyzing the image." But the text insists: the image you are analyzing is itself mind. What you take to be "the object being examined" is not a separate thing outside consciousness; it is how consciousness, in concentrated form, presents itself to itself. As the *Yogācārabhūmi* puts it, "如依清淨鏡面與本質為緣而似有影像" — like an image that *appears* when a polished mirror and a substance come together as conditions. The "image" and the "mirror" are not two separate substances; the image arises at the meeting.
This is why the meditator is eventually pointed past the image itself: 止觀和合 (the integration of śamatha and vipaśyanā) is reached "於通達三摩地所行影像唯是其識" — *by realizing that the image in samādhi is nothing other than one's own mind* — and then, "或通達後復思惟如性" — *having realized that, one further contemplates 如性 (tathatā, suchness)*.
In other words: 有分別影像 is a skillful raft, not the shore. It is the working surface of insight, and once insight has done its work, the meditator sees through it — not by abandoning it, but by recognizing it as mind, and through it, contemplating suchness.
A note on the term "有分別"
This must be flagged, because it is the most common source of confusion in modern expositions.
In ordinary Buddhist usage, "有分別" (savikalpa) can mean any kind of conceptual, name-and-form-inflected consciousness. But in this technical context, it does not mean that. The *Yogācārabhūmi* is explicit that "有分別" here is limited to the specific activity of 毘鉢舍那 on the samādhi-image: the operations of 觀察、簡擇、極簡擇、遍尋思、遍伺察. There is even a separate seven-fold classification of "分別所緣" (categories of discriminating engagement) — 有相、無相、任運、尋求、伺察、染污、不染污 — which is the broader, looser usage. The 有分別影像 term does not refer to that broader usage.
If you map it onto everyday English, "有分別" here is not "discriminating" in the sense of prejudice or category-thinking. It is closer to "analytically worked" — the image that has been turned over and examined by insight.
Life walkthrough: what 有分別影像 looks like off the cushion
The hardest part for contemporary readers is that this concept is presented in the language of formal samādhi. But the underlying structure — *a stable mental model of something important, examined carefully under calm attention* — is something almost everyone has experienced. Let me walk through some familiar scenes.
Scene 1: The therapist's office
You sit across from a good therapist. You bring up a recurring conflict with your father. The therapist doesn't immediately give advice. They first help you form a stable mental picture of the pattern: "So when he says X, you feel Y, and then you do Z, and then he responds with W…" Slowly, an *image* of the pattern emerges in your mind — not a hallucination, but a clear, considered representation. You can almost see it.
Once that picture is stable, the analytical work begins. The therapist asks: *what is the function of this pattern? what does it protect you from? is it still accurate?* You are now actively examining the image. You are not just sitting with it. You are turning it over, distinguishing its parts, scrutinizing.
This second phase — the examining — is structurally the move from 無分別影像 to 有分別影像. The first is "I have the picture and I sit with it"; the second is "I have the picture and I now analyze, discriminate, and search it." Notice: the picture itself is the same. What changes is the mode of engagement.
A common modern error is to confuse this with "being analytical all the time." It is not. In the *Yogācāra* account, the image must first be *stable*, produced under conditions of deep concentration. The contemporary parallel: you cannot usefully analyze an unstable mental model of yourself. You need a *kind* of stillness (śamatha) for the image to be workable, otherwise your analysis is just spinning in anxious thought. The 有分別影像 assumes a base of 奢摩他. It is *calm attention applied to a clear mental object with discriminating insight.*
Scene 2: A programmer debugging a complex system
A senior engineer is called in to debug a system that is intermittently failing. First, she forms a model of the system in her head — based on the documentation, prior conversations, the error logs. This model is *not the system itself*. It is a 所知事同分影像 — a same-kind likeness of the thing being investigated, produced by her concentrated resolve (the *adhimukti* of someone who has studied the codebase deeply).
Now: there are two modes she might enter. She might just observe the system's behavior calmly without analysis — watching, accepting what comes, not interfering. That is the 無分別影像 mode: she has the model, and she holds it still.
Or she might systematically test hypotheses against the model: "If the bug is in module A, then under condition X we'd expect Y; but we see Z, so…" She is now in 有分別影像 mode. She is 觀察 (observing), 簡擇 (discriminating), 遍尋思 (thoroughly seeking), 遍伺察 (thoroughly scrutinizing). The object — the mental model — is the same; what differs is whether the mind *settles into* it or *turns it over*.
The 唯識 insight the *Yogācāra* wants to protect here: a debugger can mistakenly believe the model *is* the system, or that the system exists "over there" while the model is "in her head." The classical correction is that the model — being mind — is not different from the mind doing the analysis. This is not a denial that there's a real system out there; it is a denial that the model-being-analyzed is a separate substance from the analyzer. The image and the mind analyzing it are both mind.
Scene 3: Trying to understand your own emotional reaction
You get a text from a friend that's terse. You feel a pang of anxiety. Now: most people, at this point, either (a) immediately start analyzing the text ("What did they mean? Did I do something wrong?") in a way that's actually anxious rumination, or (b) try to suppress the anxiety entirely.
The classical structure suggests a third possibility. First, let the mental image of the situation stabilize. Feel the pang. Don't push it away, don't narrate it. Just notice its quality, its location, its texture. Once it is stable and clearly seen, you are in the vicinity of 三摩地所行影像 — a same-kind likeness of "the felt situation," arising within your settled attention.
Then, *if appropriate*, you begin discriminating insight-work on that image. Not the endless rumination of "what did they mean" — that is *ordinary* 有分別, the seven-fold 分別所緣 that the *Yogācāra* explicitly distinguishes from the technical term. The technical 有分別影像 work is something like: "What is the *structure* of this reaction? What is its function? Is it accurate? Does it arise from a permanent self or from conditions? What is its 苦 (suffering) component?" You are now 觀察, 簡擇, 極簡擇 — observing, discriminating, finely discriminating.
The crucial guardrail: this analytical work happens on a stable image, not in the middle of emotional chaos. The 勝解 (concentrated resolve) and 三摩呬多地 (the absorption-ground) of the classical text correspond, in the modern reading, to a state where the meditator is *collected enough* to do this work properly. Doing insight-work on an unstable image is, classically, a non-starter.
Scene 4: Learning any difficult conceptual domain
A philosophy student is trying to grasp dependent origination (緣起). She reads the texts, listens to lectures, discusses with teachers. Slowly, a *clear mental image* of the doctrine forms — not yet the doctrine itself, but a same-kind likeness of it, a 所知事同分影像.
Now she can do two things with it. She can simply let it sit in her awareness without analysis — a kind of meditative holding — which is the 無分別影像 mode. Or she can actively analyze it: how does this condition relate to that one? What are the subtle distinctions between internal and external conditions? How does this part of the doctrine cohere or fail to cohere with that part? This analytical engagement with the mental image is the 有分別影像 mode.
The contemporary trap: students often think they are "meditating on the doctrine" when they are actually in ordinary conceptual analysis, with no stable image, no calm base. The classical term 有分別影像 specifically *requires* the image to be there, and the analysis to be done *on the image*. Otherwise, the analysis is just discursive thought (戲論, prapañca), not vipaśyanā.
The systematic mapping
Let's lay this out cleanly so the structure doesn't get muddled:
| Classical term | Contemporary parallel | |---|---| | 三摩地所行影像 | A stable mental representation arising in settled attention | | 所知事同分影像 | That representation being a *same-kind likeness* of the dharma being trained in (impermanence, the four truths, etc.) | | 勝解 (adhimukti) | The concentrated resolve that produces and sustains the image | | 無分別影像 | The image held in calm, non-analytic attention (śamatha) | | 有分別影像 | The same kind of image, now turned over by vipaśyanā's analytical operations | | 觀察、簡擇、極簡擇、遍尋思、遍伺察 | Observing, discriminating, finely discriminating, thoroughly seeking, thoroughly scrutinizing | | 唯識 | The realization that the image being analyzed *is itself mind* — not a separate object "over there" | | 如性 (suchness) | What is contemplated once the image is seen through as mind |
The most common mapping error is in the second row of the left column: collapsing 有分別影像 into "ordinary analytical thinking." It is not that. It is insight-work specifically on a concentration-born stable image.
Why contemporary people need this concept
Three reasons stand out.
1. We have become a culture that knows how to think, but not how to *examine*
This is a strange paradox of the information age: more analysis has been performed by more people on more topics than at any time in history. And yet there is widespread confusion between discursive analysis and insight.
Discursive analysis operates on abstract propositions, often in language, often under time pressure, often in a state of mild anxiety. It produces opinions, arguments, positions. It does not, by itself, liberate.
The 有分別影像 concept insists on something discursive analysis cannot deliver: a *stable, calm, mind-born image of the thing to be understood*, and then careful, unhurried scrutiny of it. The image is *held* — that is the śamatha component — and then *examined* — that is the vipaśyanā component. The two are not opposed; they are sequenced and integrated.
In an age of hot takes, doomscrolling, and reactive commentary, the discipline of *forming a stable image first, and only then carefully examining it*, is a quiet but powerful corrective. It is the difference between arguing about impermanence and calmly turning over a clear mental representation of impermanence until it is understood from the inside.
2. Modern psychology is rediscovering the importance of "the workbench object"
Contemporary therapeutic practice — particularly cognitive-behavioral, schema, and insight-oriented therapies — has stumbled onto something close to the 有分別影像 idea without naming it. The therapist's job is often to help the client form a stable, considered *representation* of a pattern (the "workbench object"), and then to examine it together. The representation is not the reality — it is a *same-kind likeness* (所知事同分影像). The therapist and client examine it: what is its structure? its function? its accuracy? its cost?
The classical framework is more articulated: the workbench object must be formed in a state of 勝解 (concentrated resolve, which contemporary therapists might call "focused engagement" or "therapeutic presence"), and the examining must be careful and unhurried (the 觀察、簡擇、極簡擇 list is essentially a description of a clinical attention sequence). The *Yogācāra* adds something modern therapy often lacks: the 唯識 guard — reminding the practitioner that the workbench object is itself *made of mind*, and that mistaking it for the thing-in-itself is a subtle reification that the insight-work is meant to dissolve.
3. The 唯識 (consciousness-only) warning is more urgent than ever
Modern life is filled with objects that *feel* like they exist "out there" but are actually projections of mind: social-media profiles of people, AI-generated images, deepfakes, algorithmically curated newsfeeds, the "personality" attributed to brands, the "character" inferred from someone's texting style.
In such a culture, the *Yogācāra*'s quiet insistence — "毘鉢舍那三摩地所行影像與心無異, 唯是識故" — is not a metaphysical curiosity. It is a practical inoculation. The images you are forming, examining, and acting on (about yourself, about others, about the world) are mind-made. This is not a claim that "nothing exists outside"; it is a claim that the *images you operate on* are not separate from your mind, and that reifying them — treating them as the things themselves — is the trap the entire practice is meant to undo.
To put it bluntly: a contemporary person who has internalized 有分別影像 properly does not mistake their working models for reality. They use the models (the images) as rafts. They examine them with care. And they remember, in the end, that the raft is not the shore.
Common misreadings and clarifications
Misreading 1: "有分別影像 means a visualized image that contains analysis."
This is the most common error in popular expositions. The Chinese 有分別 here is not modifying 影像 as "an image that is differentiated/discriminated in content." It is naming the mode of engagement: the image *as it is being analytically worked by vipaśyanā*. The image itself is the same kind of mental object whether you hold it still or examine it. What differs is the operation.
The *Yogācāra* is painstaking here: it lists seven other kinds of 分別所緣 (有相、無相、任運、尋求、伺察、染污、不染污) for ordinary "discrimination," and it explicitly states that the 有分別 in 有分別影像 does not refer to those. It refers narrowly to the vipaśyanā operations on the samādhi-image.
Clarification: 有分別影像 = "image under vipaśyanā's scrutiny." Not "image full of mental chatter."
Misreading 2: "This is just Buddhist visualization or imagination."
No. Visualization (e.g., visualizing a Buddha-figure, or imagining a deity) is sometimes treated as a separate practice category in some tantric and Pure Land contexts. It is not what 有分別影像 describes.
The 有分別影像 is constrained in three ways ordinary imagination is not: (a) it must be a 同分 likeness of a canonical training object (one of the listed dharmas: impermanence, the four truths, dependent origination, etc.); (b) it must be generated in 三摩呬多地 (a state of deep meditative absorption), not in ordinary consciousness; (c) it must be worked on by 毘鉢舍那行, the specific analytical operations listed in the text.
Ordinary imagination may also be a 染污分別 — contaminated discrimination — which the *Yogācāra* explicitly distinguishes from the technical 有分別 of vipaśyanā.
Clarification: 有分別影像 is a concentration-grounded, dharma-targeted, insight-analyzed mental representation. It is not visualization, not daydreaming, not creative imagination.
Misreading 3: "The image is separate from the mind that analyzes it."
This is the trap the 唯識 statement is meant to defuse. A meditator might suppose: "I (the analyzer) am looking at the image (the analyzed)." The *Yogācārabhūmi* explicitly denies this: "毘鉢舍那三摩地所行影像與心無異, 唯是識故."
This is not a denial that there is a *functional distinction* between the image and the analyzer — the image is the object, the analyzer is the subject-of-analysis. It is a denial that they are two separate substances. They are both mind. The image is how mind presents itself to mind when vipaśyanā operates on a concentration-grounded object.
The contemporary parallel is real and important. In therapy, a client can become attached to the model of themselves that emerged in session, treating it as "the truth" rather than as a skillful working representation. The 唯識 insight says: the model is *made of your mind*. Mistaking the model for the thing-modeled is the trap. Recognizing it as mind-made and continuing to use it skillfully — that is liberation.
Clarification: The image is not "out there." It is not a separate substance. It is mind examining itself, by means of a mind-formed likeness.
Misreading 4: "You have to be in advanced samādhi to do this practice."
In the classical structure, yes — the *Yogācārabhūmi* places this practice in 三摩呬多地 (the absorption-ground) and as the object of "一向修毘鉢舍那" (continuous vipaśyanā practice). But the *underlying structure* of the practice is not inaccessible to a contemporary reader who has never entered formal samādhi.
The structure is: *form a stable mental representation of a dharma you want to understand; calm your mind around it; then carefully examine it; recognize the representation itself as mind-made.*
The strength of attention and stability required varies by dharma and by practitioner. Mindfulness of breathing (阿那波那念) is on the classical list and is accessible to beginners. The discipline of *calming first, then examining*, is learnable. The 唯識 insight — *the representation is mind, not a separate object* — is the long-term horizon of the practice, not the entry requirement.
Clarification: The full classical practice requires formal samādhi. The underlying structure is available in graduated forms to serious practitioners.
Misreading 5: "This is the same as 'the mind is all there is.'"
The 唯識 (consciousness-only) language can sound like metaphysical idealism — the claim that "only mind exists." That is not what the *Yogācārabhūmi* is asserting here. The text is making a precise, limited point about the meditative image: *the image that vipaśyanā analyzes is itself mind, not a separate object.*
This is a methodological and experiential claim about how to relate to the working objects of meditation. It is not a global metaphysical thesis. The *Yogācāra* tradition does develop full-blown idealism in later texts (e.g., the *Vijñaptimātratā* treatises), but in this passage, the claim is local to the practice: when you analyze the image, know that the image is mind. The point is to prevent reification of the working object, not to deny the existence of anything outside it.
Clarification: The 唯識 statement here is a guard against reifying the meditative image. It is not a metaphysical claim about the non-existence of the external world.
Misreading 6: "Once you have 有分別影像, you're done."
Quite the opposite. The *Yogācārabhūmi* places 有分別影像 as first of the four 遍滿所緣境事. It is the working surface. After it, the meditator moves to 事邊際性 — the limiting nature of phenomena, the ultimate edge of what can be known about any dharma (in the commentarial tradition, this is often glossed as 空性 / śūnyatā, emptiness, or 如性 / tathatā, suchness). After that, 所作成辦 — the goal accomplished, the fruit of the path.
In other words, 有分別影像 is not the destination. It is the *raft*. The classical image of the raft is unusually apt here: you build the raft (form the stable image), examine it carefully (apply vipaśyanā), see what it is made of (唯識 realization), and then — having used it to cross — let it go. The shore is 事邊際性, the ultimate nature; and the standing-on-the-shore is 所作成辦, the fruit.
Clarification: 有分別影像 is a stage, not a goal. Use it; don't cling to it.
A closing image
Imagine you are a careful craftsperson with a piece of wood. The wood is your dharma — impermanence, say, or the nature of self, or the shape of an old pattern of suffering. Before you can work the wood, you need to see it clearly. So you bring it into good light, hold it still, let its shape become distinct in your awareness. That clearly-seen wood, held still, is the 三摩地所行影像, the image-in-samādhi, the 所知事同分影像.
Then you pick up your tools — observation, discrimination, finer discrimination, thorough seeking, thorough scrutiny. You begin to work the wood, examining its grain, its knots, its tendency to split along certain lines. The wood-under-work is now 有分別影像: the same wood, the same image, but now being analyzed by vipaśyanā.
At some point, an honest craftsperson notices: the wood is not separate from the work. The image I am examining is wood I shaped; the wood I am shaping is shaped by my hands. There is not "wood out there" and "me in here"; there is the joint activity of shaping and being-shaped. This recognition is 唯識 — the image is mind.
And if the craftsperson is wise, she eventually stops mistaking the shaping for the thing. The work was on the wood; the wood was a skillful representation of impermanence; the impermanence itself is not the representation. She lets go of the confusion between raft and shore, and the shore — the actual nature of impermanence as it is — opens to her directly. That is 事邊際性, the limit of what can be known. That is the practice that 有分別影像 exists to make possible.
The image is not the destination. But without the image — without a stable, calm, mind-formed representation of the dharma to be known, examined with careful, discriminating insight — the practice cannot even begin.
有分別影像 is therefore not a curiosity of Yogācāra metaphysics. It is the working surface of liberation. The careful reader's task is not to define it more cleverly, but to use it more carefully: form stable images of what is to be known, calm the mind around them, examine them with the five operations of vipaśyanā, recognize them as mind, and — when the time comes — let them go.