If you have ever caught yourself mid-sentence and thought, "the word I just used is doing more work than it should," the Yogācāra school has been waiting for you. Its "Six Kinds of Discourse" (六種言論, *liù zhǒng yán lùn*) is one of the most precise linguistic-epistemic tools ever developed: a six-part diagnostic that asks of any concept, "Does this term only function because language is propping it up?" If the answer is yes, you have found what the tradition calls a *jiǎ yǒu* (假有, "provisional existence") - real enough to function, but lacking any fixed, independent essence.
The classical statement, from the *Yogācārabhūmi Śāstra* (《瑜伽師地論》, T1579, ch. 36), is uncompromising: "When, in relation to a thing, these six kinds of discourse can be initiated, know that thing to be provisionally existent" (隨於諸物發起如是六種言論，當知此物皆是假有). The same six appear in the *Prakaraṇārya Śāstra* (《顯揚聖教論》, T1602), where they are deployed specifically to demonstrate that the "factors not associated with mind" (心不相應行, *citta-asampṛta-saṃskāra*) - such as *becoming*, *duration*, *ageing*, and *impermanence* treated as abstract categories - lack self-nature (自性, *svabhāva*).
Before walking through the six, a critical clarification: 假有 is not "fake," "illusory," or "doesn't exist." A car is 假有. A marriage is 假有. A career is 假有. A nation's economy is 假有. These things exist, they have effects, they crash and flourish. What they lack is *svabhāva* - inherent, partless, independent self-nature. The word *provisional* is closer, but the cleanest translation is "conventionally real but not inherently existent." This is the middle path the Yogācāra school insists on: not nihilism (nothing exists) and not eternalism (everything has a fixed essence).
What This Concept Is Saying
The diagnostic works in reverse: if a thing can be talked about in any of these six ways, it is provisional. The reasoning is structural. Anything with genuine self-nature would not need language to "hold it together" - it would simply be what it is, irreducible. But everything that requires these linguistic operations is dependent on conditions, on agreement, on parts, or on change. The six operations are how language betrays the absence of self-nature.
A word on the term *yán lùn* (言論, "discourse") itself. This is not a synonym for *yán shuō* (言說, "utterance" or "speech") in the broader sense. It refers specifically to the type of verbal operation that a given object or concept makes possible. To say a thing "gives rise to" 屬主相應言論 means: when you try to talk about this thing, you cannot do so without performing the operation of attaching it to a master. The diagnostic is about the linguistic operation the thing demands, not about the content of what is said.
Now, the six.
1. 屬主相應言論 - Discourse Requiring a Master
The first kind is discourse that only becomes intelligible when attached to a "master" (主, *zhǔ*) or "owner" - not a metaphysical owner, but a base. The texts' canonical example: when we say "the arising of form" (色之生), we cannot just say "arising" alone. We must specify whose arising. "Arising" cannot stand alone; it requires a host: form, feeling, perception, formation, or consciousness. Strip away the host, and the term dissolves into ambiguity.
This identifies a class of concepts that are "parasitic" on something else. They cannot stand alone; they need a host. The *Prakaraṇārya Śāstra* applies this test to abstract categories: when you say "becoming" (生), "duration" (住), "ageing" (異), or "impermanence" (無常), you must always ask whose becoming, whose duration. The very fact that the term cannot be used without specifying a host is the diagnostic signal.
The structural point: these terms are not "things" in the world; they are relational designations that depend on a base. They mark aspects or processes of something else. This is the linguistic fingerprint of *jiǎ yǒu*.
2. 遠離此彼言論 - Discourse That Avoids Both "This from This" and "That from That"
The second is the most subtle. It has three sub-cases:
- 以此顯此 (expressing *this* through *this*): e.g., "the hardness of earth" (地之堅). The quality is intrinsic to the substrate. The texts note this *can* apply to real existents (通實假, "pervading both real and provisional"). - 以彼顯彼 (expressing *that* through *that*): e.g., "the consciousness of the eye" (眼之識). The relationship is between two genuinely distinct things. This too can apply to real existents. - 非以此顯此、非以彼顯彼 (neither *this* through *this* nor *that* through *that*): e.g., "the door of the house" (宅之門), "the wall of the building" (舍之壁), "ten of a hundred" (百之十), "three of ten" (十之三).
It is the third sub-case - and only the third - that exclusively marks a thing as provisional. "The door of the house" is neither intrinsic to the house (a house is not, in itself, a door - not "this through this") nor a separate, externally related entity (it is not just "the door" and the house just happens to be near it - not "that through that"). The door is constituted only in the relationship to the house. "Ten of a hundred" is even clearer: there is no standalone "ten-of-a-hundred" thing; the term is pure relational designation. Same for "three of ten."
These are part-whole, member-collection, or ratio-based designations. They are constituted by relationship. The term has no content independent of the relation it expresses. The texts call this "一向於假相處轉" - "turning exclusively at provisional phenomena."
3. 眾共施設言論 - Discourse Established by Common Convention
The third kind is the most socially embedded. Common names are not arbitrary; they are established (施設, *shī shè*) by collective agreement, and the agreement is grounded in six perceptual/conceptual aspects:
- 事相狀 (the aspect of the object itself): what consciousness apprehends - 所識相狀 (the aspect of what is to be known): what mental attention brings to consciousness - 淨妙等相狀 (aspects of pleasant/unpleasant, beautiful/unbeautiful): what tactile contact apprehends - 饒益等相狀 (aspects of benefit/harm): what feeling apprehends - 言說相狀 (aspects of language and conceptuality): what perception apprehends - 邪行等相狀 (aspects of perverse or correct conduct): what volition apprehends
The structural point: ordinary language is *grounded*, but in a complex, multi-layered way. The word "fire" is not random - it is established because of the convergence of perceptual contact (heat, light), feeling (danger, comfort), perception (the visual pattern of flames), and volition (the impulse to flee or to cook). The naming is not arbitrary; it is anchored in shared experience.
But - and this is the diagnostic edge - precisely because the name is collectively established, and not based on any single inherent feature, the name is provisional. Different perceptual communities, different historical periods, different cultures will have different concepts of "fire" (or "energy" or "intelligence" or "harm"). The name captures a real convergence, but a convergence nonetheless, and convergences can diverge.
4. 眾法聚集言論 - Discourse Built on Aggregation
The fourth kind is the most intuitive, and arguably the most important for everyday self-understanding. It is discourse that creates a singular, substantial-sounding name from many gathered dharmas (法, *dharmas*).
The internal example: from the five aggregates (色受想行識, the *skandhas* - form, feeling, perception, formation, consciousness) gathered together, we build the discourse of "I" (我), "self," "person." The *I* doesn't exist before the aggregates. The aggregates are gathered, and the name is constructed. The "I" appears to be a singular, fixed entity, but it is in fact a name applied to a constantly shifting configuration.
The external examples are even more vivid: from color (色), smell (香), taste (味), and touch (觸) - i.e., the sensory data we encounter - gathered together, we build the names "house" (宅), "building" (舍), "pot" (瓶), "clothing" (衣), "vehicle" (車乘), "army" (軍), "forest" (林), "tree" (樹). None of these is a thing discovered in the world. Each is a designation given to a configuration of parts.
The radical implication, which the tradition returns to again and again: "I" is a name. "House" is a name. "Army" is a name. These are not discoveries of pre-existing entities but constructions from aggregates. The name gives the appearance of singular, independent existence to what is actually a bundle.
5. 不遍一切言論 - Discourse That Does Not Apply Universally
The fifth kind is discourse that "turns with" (隨轉, *suí zhuǎn*) a thing in one context but "returns" (退還, *tuì huán*) - i.e., fails to apply - in another. The canonical example: the term "house" (舍宅) applies to actual houses but does not apply to villages (村), market towns (聚落), way-stations (亭邏), or countries (國). The term "forest" (林) applies to forests but does not apply to individual tree roots (根), trunks (莖), branches (枝), leaves (葉), flowers (華), or fruits (果).
The diagnostic: a name has a specific, limited scope. It is "stuck" to certain configurations and cannot be extended to related-but-different configurations. The very fact that the name has conditional applicability is evidence that the name is a label rather than a real universal.
The *Prakaraṇārya Śāstra* sharpens the point: "house" does not apply to villages; "jar" (甕) does not apply to other vessels; "army" (軍) does not apply to individual men and women; "forest" does not apply to the parts of trees. The name fits the configuration, not the constituents and not the larger context. This boundedness is itself the diagnostic.
6. 非常言論 - Discourse of the Impermanent
The sixth kind involves names that change as conditions change. The texts identify four sub-types:
- 破壞 (destruction): when a pot breaks, the name "pot" (瓶) is abandoned (捨) and the name "shards" (瓦) arises (生). - 不破壞 (without destruction): when powders and ingredients are mixed to form pills, the original names are abandoned and the new name "pill" (丸散) arises. Nothing was destroyed, yet the naming changed. - 加行 (crafting/effort): when a gold bar (金段) is fashioned into rings and bracelets (環釧), the name "gold bar" is abandoned and the names "rings and bracelets" arise. The material is the same; the name has shifted with the configuration. - 轉變 (transformation): when food is digested and becomes excrement, the name "food" (飲食) is abandoned and the name "excrement" (便穢) arises. The substance has been transformed; the designation has followed.
The diagnostic: a thing with self-nature would not need its name to change. If the thing were genuinely independent and fixed, then when the name changes, something would be wrong. But in fact, names change constantly as configurations change. The thing never stays the same; the name never stays the same; the conditions never stay the same. The same stream can elicit different names at different times - which means none of the names captures a fixed essence.
Life Investigation
Let me walk each of the six through scenarios a contemporary person would recognize, strictly mapping the classical definition, and noting where people commonly map things wrong.
屬主相應言論 in daily life
Consider the word "stress" (压力). When you say "I'm stressed," can you just say "stress exists"? No. You must specify *whose* stress - work stress, relationship stress, financial stress, health anxiety. The word "stress" is a relational term; it cannot stand on its own. It needs a host (you, your body, your situation).
The diagnostic implication: "stress" is not a thing floating around in the world that then enters you. It is a designation that requires a base. If you talk about "stress" in the abstract - as if it were a substance that moves around - you have already made a category error, because the very form of the word betrays its parasitic nature.
This applies to: "growth" (谁的成长?), "anxiety" (谁的焦虑?), "productivity" (谁的生产力?), "fulfillment" (谁的满足?), "burnout" (谁的倦怠?), "momentum" (谁的势能?). All parasitic. All requiring a master.
The most common mapping error: treating these as if they were free-floating entities that you "have" or "catch." "I have stress" treats stress as a thing. The classical diagnostic would say: the word cannot exist without the "I" as host. The "having" is itself the linguistic operation that betrays the provisional nature.
遠離此彼言論 in daily life
Consider "my company's stock price" (公司股价).
The stock price is not intrinsic to the company (not "this through this" - a company is not, in itself, its stock price; the price is a numerical designation of market valuation). Nor is it an entirely different entity that just happens to be near the company (not "that through that" - the price is constituted in the relationship between the company and the market).
The stock price is relational: it exists only in the conjunction "company + market + moment in time." It is part-whole, member-organization, or relational designation.
Contemporary examples that fit this same diagnostic: "my partner's expectations" (伴侣的期待), "the team's morale" (团队士气), "the algorithm's bias" (算法的偏见), "the news cycle's momentum" (舆论周期), "the country's GDP" (国家的GDP). Each of these is neither intrinsic to its host nor separate from it. Each is constituted by the relationship. "Morale" is 假有: it functions, it affects outcomes, but it has no self-nature independent of the configuration of people, history, and context.
The most common mapping error: treating the relational term as if it were an intrinsic property. "I am an anxious person" treats anxiety as intrinsic (此顯此 - the anxiety is part of who I am in the way hardness is part of earth). "I am not my anxiety" treats it as entirely separate (彼顯彼 - anxiety is something other than me). The Yogācāra diagnostic would say: anxiety is neither intrinsic nor separate; it is a relational emergence constituted in the conjunction of person and situation. The recognition can dissolve the binary.
眾共施設言論 in daily life
Consider the word "home" (家).
Why is this word used? Trace it back: there is the visual pattern of dwellings (识所取), the intentions of safety, rest, or retreat (作意), the tactile comfort of familiar furniture (触所取), the feeling of belonging or alienation (受所取), the language we use to describe it ("where I come from," "my place") (想所取), and the conduct associated with it (the way we behave when "at home" vs. "away") (思所取). The name "home" is anchored in a convergence of all six aspects.
But - and this is the diagnostic edge - because the name is collectively established, it is provisional. Different cultures have radically different concepts of "home." Different life stages shift what "home" means. Different emotional states (the home you grew up in vs. the home you fled vs. the home you long for) can produce incompatible uses of the same word.
The diagnostic test: ask why this name is used. Trace it back to the perceptual, affective, and volitional conditions that gave rise to the consensus. If you can trace it back to conditions rather than to a fixed essence, you have identified a 假有.
This applies to: "weekend," "childhood," "love," "success," "freedom," "professional," "adult." All collectively established, all provisional, all functioning.
The most common mapping error: forgetting that the name is collectively established, and so treating it as if the consensus were natural or necessary. "Of course home means X" - the "of course" is the move that conceals the convention. The diagnostic interrupts the "of course."
眾法聚集言論 in daily life
Consider "my morning commute" (早上的通勤).
"Morning commute" is not a single thing. It is an aggregate of: waking up, getting dressed, making coffee, checking the phone, leaving the apartment, walking to transit, riding the train, transferring lines, walking from the station, entering the office. None of these alone is "the morning commute." Together, they get the name. Strip away the aggregate, and there is no "commute" left - just events.
The contemporary relevance: we constantly mistake aggregates for entities. "The economy" is an aggregate of millions of transactions, but we talk about "the economy" as if it had its own behavior. "The market" is an aggregate of buyers and sellers, but we say "the market crashed" as if the market were a thing that crashed. "The team" is an aggregate of people, but we say "the team decided" as if the team had its own mind. "The algorithm" is an aggregate of code, data, and computations, but we say "the algorithm recommended" as if the algorithm had its own will.
Yogācāra's diagnostic: every time you find yourself using a singular noun for what is actually a plural event or a configuration of parts, ask whether that noun is "gathered" or "found." If gathered, it's 假有. It has no self-nature, even if it has all the functions of a real thing.
The most paradigmatic example, and the one the texts single out, is "I" (我). I am a name applied to an aggregate of form, feeling, perception, formation, and consciousness. The "I" appears to be the most singular, most unified thing I know. The texts say: it is a name. It is a name that gathers a configuration.
The most common mapping error: assuming the aggregate has a center. "I" feels like a unified self with a center; the diagnostic says this is the name's work, not the aggregate's nature. The aggregate has no center; the name gives it the appearance of one.
不遍一切言論 in daily life
Consider the word "professional" (专业的).
"Professional" applies in some contexts (work, expertise, conduct) but does not apply in others (personal life, hobbies, casual interactions). If you call someone a "professional eater," you have stretched the term; in its normal scope, "professional" returns (退還) to designated domains. The term has a defined scope, and using it outside that scope is felt as a misuse.
The diagnostic: every time you use a name, ask where it applies and where it does not. If the scope is limited, the name is conventional. The term "mom" applies to a relationship, not to a profession; if you call your mother "mom" at her workplace, the term is out of place. The name has returned.
The contemporary relevance: this is how we police the boundaries of identity-terms. "Artist" applies in some contexts but not others. "Friend" applies in some contexts but not others. "Boss" applies in some contexts but not others. "Patient" applies in some contexts but not others. The fact that the name has a defined scope is itself evidence that it is a label rather than a self-existing thing.
This connects to contemporary debates about "performative" or "contextual" identities - the recognition that identity-terms have contexts of applicability, and that the context is part of what the term means. The Yogācāra insight is older but more precise: the term has a scope because the term is a designation, not because the underlying reality has a fixed boundary.
The most common mapping error: extending the name beyond its scope and treating the extension as natural. "I am always a professional" - the "always" extends "professional" beyond the contexts where it has its proper applicability, and the diagnostic would flag this as the name doing the work, not the reality.
非常言論 in daily life
Consider the person you were ten years ago.
That person is the same person (continuity) and also not the same person (everything has changed - body cells have turned over, beliefs have shifted, relationships are different, memories are reframed, habits are replaced). The name "you" is a continuous designation applied to a constantly changing aggregate. The name has not changed, but the configuration has - and the name would have to change if the changes were more dramatic (the friend who became a stranger, the home that became a rental, the job that became a career).
The diagnostic: when the configuration changes enough, the name may need to change. The friend who was a "stranger" is now a "friend" - the same person, different name. The hobby that was a "pastime" became a "career" - the same activity, different name. The relationship that was "dating" became "engaged" - same connection, different name.
Contemporary example: the person you call your "ex" was once your "partner." The place you call "home" was once a "rental." The job you call "my career" was once "an internship." None of these names is fixed; all shift with conditions. The very fact that the same underlying stream can elicit different names at different times shows that none of the names is capturing a self-nature.
The most common mapping error: treating the name as a fixed label that the configuration must match. "I am still the same person I always was" - the "still" and "always" are the moves that conceal the change. The diagnostic interrupts this by asking: if you are still the same, why does the name now need new qualifications? ("I've changed, but I'm still me at my core.") The "at my core" is the move that tries to rescue self-nature; the diagnostic would say: the name is the *only* continuity, and a name is a convention.
Why Contemporaries Need This
The Six Kinds of Discourse addresses several specifically contemporary concerns that earlier centuries did not have to face with the same intensity.
1. The inflation of language
We live in an era of unprecedented linguistic production. Every year, thousands of new terms enter circulation: "AI alignment," "deepfake," "creator economy," "climate anxiety," "quiet quitting," "doomscrolling," "polyamory," "neurodivergence," "burnout," "wellness." Most of these are useful; some are hype; almost all are provisional. The Six Kinds of Discourse gives you a diagnostic: when you encounter a new term, ask which of the six it fits. If it requires a host (属主), it's a relational term about a base. If it is an aggregate (聚集), it is a configuration of parts. If it does not apply universally (不遍), it has a limited scope. If it changes with conditions (非常), it is impermanent. This is not a debunking - it doesn't say the term is "fake" - but it does say the term is provisional, and that recognition changes how you relate to it. The relief is not nihilism; it is precision.
2. The AI and algorithm question
When an AI system "decides" something or "recommends" something, what is happening? Contemporary discussions of AI agency often oscillate between two poles: anthropomorphism (treating the AI as a self-nature agent with beliefs and intentions) and denial (treating it as nothing, just code). The Six Kinds of Discourse offers a more precise middle: the AI's "decision" is 假有. It is a designation applied to an aggregate of weights, data, computations, and outputs. The designation is useful; the designation is conventional; the aggregate has no self-nature; the aggregate has no inherent agency. But to call the aggregate a "decision-maker" is a collectively established, scope-limited, changeable designation - which is exactly what 假有 means.
This is also why AI alignment debates often get stuck. The people arguing about whether AI is "really" conscious or "really" an agent are asking a self-nature question. The Yogācāra diagnostic would redirect: the term "agent" or "intelligence" is a collectively established label with limited scope that shifts with conditions. What matters is not whether AI has self-nature but how the designation functions and what it produces. Whether to call an AI an "agent" is a question about which of the six kinds of discourse the term invokes, and what scope the convention should have.
The same diagnostic applies to "the market," "the algorithm," "the network," "the platform," "the cloud" - all aggregates, all 假有, all functioning. The point is not to deny their effects but to clarify their nature.
3. The pressure of self-nature in identity
Contemporary life is saturated with self-nature discourse about identity. "I am an introvert." "I am a creative." "I am a failure." "I am successful." "I am a good person." "I am a bad person." Each of these treats the designation as a fixed essence - an inherent property of the self. The Six Kinds of Discourse, especially 眾法聚集 and 非常, provides a quiet counter-pressure: you are an aggregate, your name for yourself is a convention, your identity-terms change with conditions. This is not a self-help book telling you to "be whoever you want" - that is its own kind of inflation. It is a precise diagnosis: the "I" is 假有, and the relief of this recognition is not nihilism but the end of the burden of defending a fixed self. The recognition of *jiǎ yǒu* does not destroy the "I"; it stops the "I" from having to be a fortress.
4. The relief of precision in relational discourse
The category 遠離此彼言論 is particularly useful in contemporary relational discourse. How often do arguments arise because someone treats a relational term as if it were an intrinsic property? "I am an anxious person" treats anxiety as intrinsic (此顯此). "I am not my anxiety" treats it as entirely separate (彼顯彼). "My anxiety is part of who I am" - intrinsic again. "I need to defeat my anxiety" - separate again. The Yogācāra diagnostic would say: anxiety is neither intrinsic nor separate; it is a relational emergence. The recognition can dissolve the binary that traps people between "I am my anxiety" and "anxiety is alien to me."
The same applies to: "I am my trauma" / "I am not my trauma." "I am my work" / "I am not my work." "I am my body" / "I am not my body." The diagnostic would say: these are relational terms, and the binary is itself the error.
5. A non-religious ethics of precision
For people who do not adopt Buddhist metaphysics wholesale, the Six Kinds of Discourse still functions as an ethics of precision. It trains you to notice when you are mistaking a label for a fact, a name for a thing, a convention for a necessity, an aggregate for an essence. This precision is valuable in itself. It doesn't require you to accept the three natures (三性) or the eight consciousnesses (八识) to be useful. It is a tool for seeing through reification - the cognitive habit of treating provisional designations as if they had self-nature.
Common Misreadings and Clarifications
1. 假有 does not mean "fake" or "non-existent"
The most common Western misreading is to translate 假有 as "illusory" or "doesn't exist." This is a serious error that reverses the diagnostic. A car is 假有. A house is 假有. A person is 假有. A relationship is 假有. These things exist. They function. They have effects. They can crash. What they lack is *svabhāva* - inherent, partless, independent self-nature. The English word "provisional" is closer, but the cleanest translation is "conventionally real but not inherently existent." 假有 is a technical term that protects against both nihilism (nothing exists) and eternalism (everything has a fixed essence). The diagnostic is selective, not universal.
2. This is not nominalism
Western nominalism says that universals or aggregates have no real existence at all - they are "just names." The Six Kinds of Discourse is not nominalism. The texts explicitly say that 施設 (establishment by convention) is grounded in real perceptual, affective, and volitional conditions: 事相状, 净妙等相状, 饶益等相状, 言说相状, 邪行等相状. The naming is anchored in actual experience. The aggregate exists; the convention captures something real. What it does not capture is a self-nature. The Yogācāra position is closer to "conventional realism" - things are real, but they don't have the kind of self-existence we tend to attribute to them. This is also why the texts do not say that aggregates "do not exist"; they say aggregates "lack self-nature" (無自性). The verb matters: "lack self-nature" is not "do not exist."
3. This is not the same as the "六種論" in chapter 53 of the *Yogācārabhūmi Śāstra*
The *Yogācārabhūmi Śāstra* has two distinct classifications, both sometimes called "six kinds of discourse":
- Chapter 53's six kinds of discourse by function: 言論 (general discourse), 尚論 (disputation), 諍論 (conflict), 毀謗論 (denigration), 順正論 (right-discourse), 教導論 (instruction). This is an ethical evaluation of types of speech - which lead to benefit, which lead to harm, and which must be examined case by case. - Chapter 36's six kinds of discourse as diagnostic for 假有: the one treated in this essay - 屬主相應言論, 遠離此彼言論, 眾共施設言論, 眾法聚集言論, 不遍一切言論, 非常言論.
These are different classifications with different purposes. The first is about the ethical quality of speech. The second is about how speech reveals the ontological status of its referent. Confusing them is easy because they share a similar Chinese rendering, but they operate on entirely different axes. A serious reader should be careful to keep them separate. The one-sentence definition given for this concept (the Yogācāra diagnostic for distinguishing provisional from real) refers only to the second set.
4. "Owner" (主) in 屬主 is not a metaphysical substance
When the texts say 屬主相應言論, the term 主 is not positing a metaphysical "self" or "subject" behind the discourse. It is a linguistic diagnostic: the term requires a host to be intelligible. The "host" is just whatever the term is about. There is no separate metaphysical entity that the host points to. The diagnostic is purely about how the term functions in language. This is important because 屬主 is sometimes mistakenly read as evidence for a permanent self (atman) - but the diagnostic is actually the opposite: it shows that certain terms (becoming, duration, ageing) cannot stand alone, which is a sign of their provisional nature, not of a substantive owner.
5. The six are not mutually exclusive; a single thing can elicit several
A single thing can elicit multiple of these discourse types. "My home" is 屬主 (requires "my" as host), 聚集 (built from materials, memories, and relationships), 不遍 (does not apply in all contexts - my home is not "home" when I am traveling), and 非常 (changes as the home ages, the family changes, the meaning shifts). The diagnostic is not asking "which of the six is it?" but rather "how many of these apply?" The more that apply, the more obviously the thing is 假有. A thing that is simultaneously parasitic, aggregated, scope-limited, and changeable has no self-nature from any angle.
6. This is not just about language
A common misreading is to treat the Six Kinds of Discourse as merely a "linguistic" analysis, a kind of Buddhist version of "ordinary language philosophy." But the texts say the diagnostic reveals something about the things themselves, not just about how we talk about them. The classical statement is: when these six kinds of discourse can be initiated "in relation to a thing" (隨於諸物), that thing is provisionally existent. The point is not to analyze language but to use language's structure to reveal how things exist. The linguistic operation is a window onto ontology. The aggregate "I" really is an aggregate; the linguistic fact that the term "I" is built from gathered dharmas reflects the fact that the referent is a configuration. The diagnostic points to an ontological feature (or rather, the lack of one).
7. 假有 is not the same as 無 (non-existence)
The opposite of 假有 is not 無 (non-existence). The opposite of 假有 is 自性有 (existence with self-nature, or 實有). 假有 is the middle position: real but not inherently existent. In Yogācāra's three-nature framework (三性, *tri-svabhāva*) - 遍計所執性 (imagined nature), 依他起性 (dependent nature), and 圓成實性 (perfected nature) - the 假有 discussed here falls under 依他起性 (dependent nature): real, but dependent on conditions, lacking self-nature. The diagnostic is selective: it identifies which things fall into 依他起性 by showing which things elicit these six kinds of discourse. The whole point is to avoid the binary of "exists / doesn't exist" and to land on the third option: "exists dependently."
8. This is not "postmodern" in the relativist sense
There is a temptation to read the Six Kinds of Discourse as a kind of "linguistic turn" similar to postmodernism's claim that "reality is socially constructed" or that "all is text." But the Yogācāra position is more precise and more discriminating. It is not saying reality is constructed by language. It is saying that certain kinds of things (conventions, aggregates, relational terms, scope-limited designations, changeable names) lack self-nature - and that language reveals this. Other kinds of things (the dharmas that make up the aggregates) have a different status. The diagnostic is selective, not universal. "Everything is 假有" is not the claim; rather, "things that elicit these six kinds of discourse are 假有." The diagnostic discriminates between what lacks self-nature and what does not, rather than dissolving all distinctions.
9. The diagnostic is not a debunking but a liberation
One last clarification, and perhaps the most important. The Six Kinds of Discourse is sometimes read as a kind of ancient Buddhist debunking - "you think X is real, but X is just a convention." This reading misses the point. The recognition that a thing is 假有 is not a debunking; it is a liberation from the burden of having to defend the thing as having self-nature. The "I" is a convention, but the convention is functional, and the relief of seeing through its self-nature is precisely what allows one to act without the need to defend an essence. The conventional name works; the conventional name is provisional; the recognition of its provisional nature does not weaken the name but frees it from the impossible task of being permanent. This is the practical, transformative edge of the diagnostic: it relieves the weight of self-nature claims while leaving the functioning conventional reality intact.
In the end, the Six Kinds of Discourse is a quiet tool. It does not roar. It does not demand belief. It simply asks, of any concept you bring to it: where is its host? what is its relation? how was it collectively established? what is it aggregated from? what is its scope? how does it change? If you can answer all six, you have a 假有 - real enough to use, not so real as to be untouchable, not so fixed as to be unforgeable, and not so solid as to be the end of inquiry. The diagnostic, in the end, is not the end of inquiry but the beginning of a more precise one.