What this concept is about
Imagine someone hands you a basket of apples and asks you to "describe what's inside." You can do several quite different things, and they are not the same move:
1. You say, "These are all apples." — you point out what they have in common. This is identifying the 自性 (svabhāva) — the common nature of the class. 2. You say, "These are Fuji, these are Honeycrisp, these are Granny Smith." — you point out differences *within* the class. This is 自性差別 (svabhāva-viśeṣa) — distinguishing members of the same class by their own-natures. 3. You say, "These are sweet, these are tart." — a different axis of classification, perhaps closer to 染淨差別 (pure/impure distinction) or taste-distinction. 4. You say, "These are good for pie, these are good for eating fresh." — pointing to uses, closer to 對治差別 (antidote / functional distinction).
The Yogācāra treatise *《瑜伽師地論》* (Yogācārabhūmi-śāstra) is doing exactly this kind of multi-axis analysis when it examines any class of dharmas. It tells us: any "established meaning" (所成立義, *sādhyārtha*) — anything you can meaningfully establish — has two top-level aspects, 自性 and 差別 (viśeṣa). The 自性 is the basic, common nature: "what is, is established as being; what is not, is established as not-being." Beyond that, there are "immeasurable distinction-doors" (無量差別門) — superior / inferior, permanent / impermanent, with-form / formless, visible / invisible, resistant / non-resistant, afflicted / unafflicted, conditioned / unconditioned, and so on.
Now, 自性差別 is a *specific* narrow form of this general distinction-work. It does not mean "all kinds of distinction." It means: distinctions made *within the same class of dharmas*, based on *differences in their own natures*. It is the move from "I know what this class is" to "I know how the members of this class differ from each other in what they intrinsically are."
This precision is intentional. The Yogācāra takes care to separate 自性差別 from several look-alike terms:
- 品類差別 — distinction by *category-class*, used to establish dharmas as members of different broad kinds. - 自相差別 — distinction in *own-marks* (自相), used to explain 異門 (different expressions/vocabulary). - 俗相差別 — distinction in *conventional-marks* (俗相), used to explain 意趣 (the author's intention/purpose). - 因相差別 — distinction in *cause-marks* (因相), used to explain 殊異 (peculiar / distinctive features). - 所緣差別 — distinction by *object* (所緣). - 因緣相差別 — distinction by *cause and condition* (因緣相).
These are all 「差別」, but each pulls on a different thread. 自性差別 pulls on the thread of *what each member of the class intrinsically is* — its own-texture, its own-fingerprint.
Where does the term actually appear in the source? Three notable places:
First, in establishing the afflictions (煩惱, kleśa). The treatise says that the differentiation of afflictions is built through five aspects (五種相), and the second of these is 自性差別. The other four are: 自性 (own-nature), 染淨差別 (impure/pure distinction), 迷斷差別 (confusion/cut-off distinction), 對治差別 (antidote distinction). The presence of 自性差別 as a separate category tells us something important: when we want to understand the variety of afflictions, we need to see how different afflictions differ in their own nature — not just in their purity, not just in how they are abandoned, not just in their antidote.
Second, in analyzing the Seven Factors of Enlightenment (七覺支, sapta-bodhyaṅga). The text says that through 自性差別 in conjunction with 所緣差別 (object-distinction) and 因緣相差別 (cause-and-condition-distinction), the seven factors have fourteen kinds of distinctions. So here, 自性差別 is *one of three axes* for distinguishing within a class.
Third, in the foundational framework of 所成立義 (established meaning), where 自性 and 差別 are the two top-level moves of any conceptual establishment.
So the structural picture is: when analysing a class of dharmas, the Yogācāra uses a hierarchical precision — first identify what the class is (自性), then distinguish within it. There are several ways to distinguish within a class; 自性差別 is the way that focuses on differences in the own-nature of each member.
Life walkthrough
Let me walk this through with a concrete example that almost everyone has access to modernly: the experience of "stress."
Suppose you say, "I've been stressed lately." A friend asks, "What kind of stress?"
The first move — identifying the class — is when you say: "stress" as a category has a recognizable 自性. It is a felt sense of being overwhelmed, pressed, taxed; the body tightens, the mind races, the field of attention narrows. That's what makes something count as stress in the first place.
The second move — 自性差別 — is when you notice that the *stress* you feel about a deadline at work is qualitatively different from the stress you feel about a relationship conversation, which is different from the stress you feel about money, which is different from the stress you feel about your health, which is different from the stress you feel from scrolling through world news, which is different from the stress you feel about your own future.
These are all "stress" — they belong to the same class. But they have *different own-natures*. The deadline-stress has a ticking-clock quality, a forward-looking narrowing, a specific bodily urgency in the chest. The relationship-stress has a more persistent, ruminating texture, less clock-bound, more stuck in the throat, repeatedly re-activating past conversations. The news-related stress has a horizonless, ambient quality — no identifiable endpoint, no clear action to take, a kind of diffuse dread. The health-stress has a hypervigilance about the body, an oscillation between symptom and reassurance. The existential-stress sits behind all of them, has no object, no time-horizon, no action.
Each is stress. Each belongs to the same class. But each has its own inner texture, its own relationship to time, body, and agency. This is exactly what 自性差別 points to: not the category-boundary (what makes stress = stress), but the texture-differences within the category (what makes deadline-stress be what it is, distinct from news-stress).
Why does this matter? Because most contemporary people, when they say "I'm stressed," stop at the category-level. They take the entire field of stress as homogeneous. They apply one coping mechanism — meditation, exercise, distraction, venting, a productivity hack — across all instances. Sometimes this works; more often it doesn't, because they are applying a generic response to phenomena with different own-natures. The Yogācāra framework is saying: in addition to knowing the 自性 of stress, you need to see the 自性差別 — and that seeing will tell you that deadline-stress and existential-news-stress need different relationships to time, different bodily responses, different cognitive interventions. The trained awareness that distinguishes within the class is itself the prerequisite for an appropriate 對治差別 (antidote-distinction).
Let me try a second example: love, or "caring for someone."
The class 自性 might be: a directed, valuing attention toward another, where their flourishing matters to you, where you orient your energy around their good.
Now, the 自性差別 within this class:
- Parental love of a small child: a protectiveness-quality, an asymmetry (the child is not yet a full peer), an anticipatory grief woven into the tenderness. - Romantic love of the early kind: a discovery-quality, an excitement at novelty, a strong cognitive bias toward the other, often blind to their flaws. - Mature long-term partnership: a familiarity-quality, a quiet constancy, a tolerance of imperfection, an intertwining of identity. - Friendship: a chosen-quality, a shared-interest orientation, a lower dependency, more revocable. - Compassion for a stranger: a momentary-quality, a recognition of common vulnerability, no biographical entanglement.
These are all "love" or "care" in the same broad class. But each has a different own-nature — structured differently around time, around identity, around reciprocity, around vulnerability.
Many contemporary relational problems arise because people use one type of love-template as the default for all. You treat your friend like a romantic partner (or vice versa), or you bring parental protectiveness into adult peer relationships, or you demand the security of long-term partnership in the early discovery phase, or you evaluate a stranger's suffering through the lens of biographical intimacy. The recognition of 自性差別 — that within "love" there are different intrinsic textures — is the precondition for relational intelligence. Without it, one ends up either demanding the wrong things or missing what is actually being offered.
A third example, more granular: the experience of "thinking."
The class 自性: cognition, the formation of representations, the working of mind.
The 自性差別: there is the thinking that is *rehearsal* (re-running past scenarios), the thinking that is *planning* (running forward scenarios), the thinking that is *judging* (evaluating present), the thinking that is *wandering* (associative, low-coherence), the thinking that is *analysing* (breaking down a problem), the thinking that is *synthesising* (combining elements), the thinking that is *mentally practicing a response*, and the thinking that is *evading* (using mental activity to avoid feeling).
These are all "thinking" — same class. But each has a different own-nature, a different relationship to time, to affect, to agency. The contemporary habit of treating "thinking" as one undifferentiated activity, and applying one rule (e.g., "just don't think so much" or "try to be more positive") across all these very different sub-classes, is what makes most advice about thinking feel irrelevant.
Modern cognitive science loosely maps onto this when it distinguishes, for instance, between default-mode-network activity and task-positive-network activity, or when observations are made that "rumination" and "reflection" are supported by partly different neural patterns. Neuroscience also acknowledges that what counts as a discrete "type" of cognition is partly constructed by the categories researchers bring to the data — there is no settled taxonomy. So the mapping is suggestive, not equivalent. The Buddhist framework of 自性差別 is doing something more pragmatic and more granular: it is a tool for *self-recognition in real time*, not a taxonomy of brain regions or a metaphysical classification of mental events.
The one move you should *not* make from this walkthrough is to convert these texture-differences into fixed essences. The deadline-stress you feel today may tomorrow have a different shape. The texture-differences are recurring patterns in your experience, not eternal forms. We will return to this point in the misreadings section.
Why contemporaries need it
The contemporary worldview has a strong bias toward generalisation. This is partly an inheritance of industrial-age efficiency, partly a function of how marketing and algorithms work (they love flat categories), partly a feature of how we acquire vocabulary as children (we learn "dog" before "terrier"). We name the class, we move on. The Yogācāra's lesson is that this is the *first* move, not the last.
The specific contemporary harms of skipping the 自性差別 step:
1. Mis-applied coping. As above: one coping mechanism for many phenomenologically distinct phenomena. The same meditation, the same productivity system, the same boundary-setting, the same venting-to-friend — applied across situations whose own-natures differ. Sometimes the "antidote" itself becomes a new affliction. A common contemporary example: a person applies "mindfulness" (a tool appropriate to one sub-class of mental experience) to grief, which is phenomenologically different and may need a different kind of attention.
2. Moral flattening. When we say someone is "lazy," "anxious," "selfish," "creative," "smart" — we treat the class as the whole story. The 自性差別 of "laziness," for instance, includes: exhaustion-laziness (the body is depleted), decision-fatigue-laziness (the mind is depleted), avoidance-laziness (the situation is threatening), depressed-laziness (the motivational system is impaired), comfortable-laziness (no threat, no need). Each is "lazy" in some sense, but each has a different own-nature, and each needs a different response. Calling them all "laziness" and applying one moral judgment produces both injustice and failed interventions.
3. Therapy and self-help ceilings. Much of contemporary self-help languishes because it operates at the category level. "Manage your stress." "Set boundaries." "Communicate better." These are correct at the 自性 level but too coarse at the 自性差別 level. The person who has trouble applying them is not necessarily deficient in willpower; they are often confronting a situation whose own-nature doesn't fit the template.
4. AI and the flattening of categories. When we ask an AI assistant or a search algorithm a question, we are often handed back a category-level answer. The technology flattens because it is trained on category-level data and rewards summary-level responses. The human's edge, increasingly, is in the discrimination of cases within the class — in seeing, "this *particular* stress, this *particular* relationship, this *particular* failure of mine, has its own inner texture, and the category-level answer is missing it." This is precisely the move of 自性差別. (Worth noting: contemporary AI systems are also beginning to be trained on case-level nuance, so the gap is not absolute; but the human capacity to *care* about the texture-difference, the interested attention that 自性差別 requires, remains distinctive.)
5. Democratic and political discourse. When political discourse operates at the category level ("the left," "the right," "the urban," "the rural"), it loses the 自性差別 within those classes. The "rural voter" and the "small business owner" and the "factory worker" and the "farmer" are in the same class by external geography but have different own-natures. The political thinker who can do 自性差別 will see further than the one who only operates at the category level.
6. Self-knowledge. The Yogācāra pairs 自性差別 with 所緣差別 (object-distinction) and 因緣相差別 (cause-and-condition-distinction) in the seven-factors analysis. The full trio — nature, object, cause — is the complete inner anatomy of any class. Most contemporary self-analysis is strong on 所緣 (what is the thing about?) and 因緣 (what caused it?) but weak on 自性差別 (what is its own nature, as distinct from its neighbors in the class?). Reintroducing that third dimension is part of what the tradition is offering.
This is why the tool of 自性差別 is genuinely useful for contemporaries: it is a discipline of staying *finer-grained than your categories*, of noticing the texture-difference within the class, of resisting the pull toward the flat category.
Common misreadings and clarifications
Misreading 1: 「自性差別」 means "the difference between self and other" or "subjective vs. objective distinction."
This is a tempting English translation because "self-nature difference" sounds like "self vs. not-self." But it is wrong. The 自性 here is not "self" in the sense of ego or subject. It is the *own-nature* of a class of dharmas — the inner nature that the members of the class share. 自性差別 is *within* that class, distinguishing among the members of the same class. It is not about the self-vs-other distinction; that would be a different teaching (e.g., the analysis of 人我分別 — person/not-person discrimination). Conflating the two collapses a precise tool into a vague philosophical mood.
Misreading 2: 「自性差別」 is just 「差別」, and the 「自性」 prefix is decorative.
This is also wrong. 差別 (viśeṣa) by itself is the umbrella term for any kind of distinction — the 上/下, 常/無常, 有色/無色 list. 自性差別 is a *specific* kind of distinction: the kind that operates within the same class, based on differences in the own-nature of the members. The Yogācāra explicitly distinguishes it from other kinds of distinction (品類差別, 自相差別, 俗相差別, 因相差別, 所緣差別, 因緣相差別). These are all 「差別」 but each is a different tool. The 「自性」 in 自性差別 is doing real work: it specifies *which* distinction you are doing. To drop the prefix is to lose the very precision the term was built for.
Misreading 3: 「自性差別」 is an essentialist claim — each thing has a fixed essence, and we just need to read it off.
This is the most philosophically charged misreading, and it deserves careful handling. The term 自性 literally can mean "self-nature" or "own-being," which in some Buddhist contexts (especially Madhyamaka) is the precise thing to be *deconstructed* — 「諸法無自性」 ("all dharmas are without own-nature"). So how can Yogācāra use 自性差別 without contradicting that?
The answer is in the framework the source itself provides: the Yogācāra is operating at the level of 所成立義 (established meaning) — the level of meaningful conceptual analysis. It is not making metaphysical claims about essences. It is giving tools for *navigating conceptual analysis with precision*. When it says "establish the 自性 of a class," it means: identify what makes this class a class. When it says "establish the 自性差別 of a class," it means: identify how the members of this class differ from each other in their identifiable nature. These are pragmatic-conceptual moves, useful for the practitioner who wants to see and respond well, not metaphysical assertions about fixed essences.
The deeper Mahāyāna insight — that even these conceptual establishments are themselves without fixed essence — is compatible with the Yogācāra's project; it operates at a different level. The Yogācāra itself recognises that conceptual establishments are constructed conventions. So when you use 自性差別 in your own practice, you are not committing to a metaphysics of essence. You are training a sharper eye for the texture-differences within a class. This is a skill, not a metaphysical thesis.
Misreading 4: Once I've identified the 自性差別, I have the final truth about the class.
No. The framework is recursive. Within any sub-class you identify, you can re-ask: what is the 自性 of this sub-class? What are the 自性差別 within it? And so on. The seven-factors analysis gives fourteen sub-distinctions within the seven factors, which can themselves be further analysed. The endpoint is not a final list; it is a habit of finer and finer seeing. Also: the texture-differences you identify are conditioned, contingent, and change with context. The move of 自性差別 is ongoing, not conclusive.
Misreading 5: 「自性差別」 is a uniquely Yogācāra invention, foreign to earlier Buddhism.
This is a scholarly observation, not a truth claim on my part. The term in this technical form is characteristic of the Yogācāra corpus (T1579). But the underlying move — distinguishing within a class based on the inner nature of the members — is present in earlier strata. The Buddha's analysis of 「煩惱」 into greed, hatred, and delusion (or further into the various classical lists of 98 or 108 specific defilements) is implicitly doing 自性差別. The difference is that the Yogācāra makes the move explicit, names it as a tool, and locates it within a battery of other discriminative tools. So the term is new; the practice is ancient.
Where this concept sits in the path
自性差別 is not a destination. It is one of several fine-grained tools that the Yogācāra places in the practitioner's hand. It is used in service of two larger aims: first, the precise understanding of phenomena, which itself contributes to 遍知 (full knowing) and the cutting of 疑惑 (doubt); second, the matching of an appropriate response — 對治 — to the actual nature of the situation.
A practitioner who has only the category-level understanding (自性 but not 自性差別) will be confused by the variety of phenomena and apply the wrong responses. They will live in a flat world of broad categories and generic remedies, and the persistent sense that "the advice doesn't quite fit" will be the diagnostic symptom of missing this tool.
The whole point of the framework is to support 更精準的看見 (more precise seeing) and 更精準的回應 (more precise response). In the classical framework, 自性差別 is bracketed by other tools: on one side, 自性 (what makes the class a class); on the other, 對治差別 (what differentiates the antidotes). The trio — 自性 → 自性差別 → 對治差別 — is a complete arc: *identify, discriminate, respond*.
The contemporary practitioner who wants to navigate skillfully — with their own mind, with their relationships, with their work, with the complexity of the world — is essentially being asked to do this same arc. The classical term is unusual; the practice is universal. The Yogācāra's contribution here is not the practice itself, but the explicit naming of the practice as a distinct move, separate from the broader move of 「差別」 and separate from the adjacent moves of 「品類」, 「自相」, 「俗相」, 「因相」, 「所緣」, and 「因緣相」. That naming forces a clarity of introspection that is otherwise easy to lose. The very act of asking, "have I noticed the 自性差別 within this class, or only the class itself?" — is, in a small way, the practice coming alive.