I. What This Concept Is About
1. Its Place in the Overall Teaching
The very name 论所依 (vādaniśraya, "the basis of argumentation") itself contains a cipher: the two characters 所依 mean "the ground one relies upon." It is not a loose reference to "the foundation of scholarship," nor is it the general sense of "the basis of a rationale" — it is a distinct structural item within the discipline of argumentation (论议法门): it designates, within a structured debate, exposition, or teaching, what exactly you are trying to establish, and what you use to establish it.
In the 《瑜伽师地论》 (Yogācārabhūmi-śāstra, "Discourse on the Stages of Yogic Practice"), the discipline of argumentation is divided into seven headings: the nature of argumentation (论体性), the locus of argumentation (论处所), the basis of argumentation (论所依), the adornments of argumentation (论庄严), defeat in argumentation (论堕负), escape from argumentation (论出离), and the many works of argumentation (论多所作法). Vādaniśraya is the third item. The preceding "nature of argumentation" concerns what kind of speech counts as argumentation; the "locus" concerns where it takes place (the setting and state of mind); only with "the basis of argumentation" do we enter the content level: what, exactly, your argument establishes, and how.
The 《显扬圣教论》 (Prakaraṇāryavācā-śāstra, "Discourse Setting Forth the Holy Teaching") expands the same system into eight headings, singling out "evidence" (论据) as its own item, but 论所依 still occupies the same position and the same scope — the emphasis is merely adjusted: the Yogācāra system keeps "evidence" within "the basis," while the *Prakaraṇāryavācā* system gives it independent standing. This detail is worth remembering, because it is easy to confuse when reading different translations.
2. The Overall Architecture: "Two to Be Established, Eight That Establish" — Ten in Total
The entire content of vādaniśraya is a structure with a very clear number: two kinds of what is to be established (所成立义) + eight kinds of what establishes (能成立法) = ten total.
This number is not arbitrary; it corresponds to the inner skeleton of any argumentative activity: first you decide what is to be established (所成立义), then you assemble what you use to establish it (能成立法). Only when the two work together can an argument be actualized.
3. The Two Kinds of "What Is to Be Established" — What You Are Trying to Establish
The first group is 所成立义 (sādhya-artha, "the meaning to be established") — the goal of the argument. It is divided into two kinds:
1. Self-nature (自性, svabhāva): "what exists, establish as existing; what does not exist, establish as not existing" — establish the matter of whether a given dharma is or is not. For example: is nirvāṇa existent or not? Is tathāgatagarbha established as existent or as nonexistent? This level handles the fundamental determination of "existence or nonexistence."
2. Distinctions to be established (差别所成, viśeṣa-sādhya): establish the various characteristic distinctions of a dharma. This is where the classics truly unfold their vast space: the examples given in the 《显扬圣教论》 include: having a superior / being supreme (有上、无上); permanent / impermanent (常、无常); having form / formless (有色、无色); visible / invisible (有见、无见); with resistance / without resistance (有对、无对); with outflows / without outflows (有漏、无漏); conditioned / unconditioned (有为、无为)... The text explicitly states that such categories of distinguishing marks are immeasurable.
This reminds us: "establishing" in Buddhist argumentation is never so thin as "proving a proposition true." It includes the determination of essence (self-nature), and also the building up of immeasurable distinctions of quality, state, attribute, level, and direction. The layered "therefore it is permanent... therefore it is impermanent..." passages so common in the scriptures are precisely the typical operation of 差别所成.
Special attention is due to a crucial qualification in the 《显扬》:
「为欲生成他信解故，非为生成诸法之自相。」
"For the sake of producing faith and understanding in others, not for producing the own-nature of dharmas." The own-nature of dharmas is already accomplished, already established in its own dharma-nature — argumentation does not "create" or "alter" the real appearance of dharmas; its function is merely to cause others to gain faith and understanding in already-accomplished dharmas. This sentence draws the boundary of vādaniśraya very clearly: argumentation has real dialogical power, but it is not the power of a creator-god.
4. The Eight Kinds of "What Establishes" — What You Use to Establish
The second group is 能成立法 (sādhana-dharma, "the means of establishment") — the toolbox for establishing the thesis. There are eight kinds:
1. Establishing the thesis (立宗, pratijñā): first display in advance the thesis you "cherish and delight in." The "thesis" (宗) is the core proposition you are defending and seeking to prove in this argument. The source of the thesis may be: what is contained in the teachings (论教所摄, derived from scriptures and treatises), what one establishes through one's own argument (自辩所立), or what one has heard from others (从他所闻). And the purpose of establishing the thesis is not merely "to persuade the other party" — it may include: establishing one's own school, revealing the faults of another school, humbling the other's arrogance, crushing the other's contempt, or setting forth the argument out of compassion (悲愍).
2. Discriminating the reason (辩因, hetu): relying on what is directly seen (现见之事) to disclose a decisive principle, causing others to accept the thesis. The "reason" (因) is the genuine ground that supports the thesis. The 《显扬》 deliberately uses the compound 辩因 ("discriminating the reason") to stress that this is a process of distinguishing and clarifying the principle, not casually citing a cause.
3. Drawing an example (引喻, udāharaṇa): drawing on dharmas that are familiar through worldly habit, commonly accepted, and easy to understand, as similes. For instance, "impermanence is like a mass of foam (无常犹如聚沫)." Examples matter because some principles cannot be directly verified by experience, but can be made intelligible through scenarios everyone knows.
4. Similar class (同类, sajātīya): list instances of the same class as the principle to be established, reinforcing the thesis.
5. Dissimilar class (异类, vijātīya): list instances of the opposite class, highlighting the specificity of the thesis.
6. Direct perception (现量, pratyakṣa): direct, immediately present cognition — the immediate grasp of the senses.
7. Inference (比量, anumāna): inferring the unknown from the known — the core of rational inference.
8. Authoritative teaching (至教, āgama): citing what the noble ones have said and what is recorded in the scriptures as grounds for proof (the 《瑜伽》 calls it 正教, "correct teaching").
The small differences among the three translations are worth noting: the 《瑜伽》 says 正教 (zhèngjiào, "correct teaching"), the 《显扬》 says 至教 (zhìjiào, "authoritative teaching"); the 《瑜伽》 says simply 因 ("reason"), the 《显扬》 specifically uses 辩因 ("discriminating the reason"); the 《瑜伽》 says 论堕负 ("falling into defeat"), the 《显扬》 abbreviates it to 论负 ("defeat in argumentation"). These are merely wording differences of the same system in different treatises — the concepts themselves are identical.
5. The Order: Why Thesis First, Then Reason, Then Example
The three most central means of establishment — thesis, reason, example — have a strict order that cannot be scrambled:
- First establish the thesis (宗): put forward the core proposition you intend to defend, giving the whole argument a target; - Next discriminate the reason (因): disclose the principle on the basis of what is directly seen, without departing from the thesis; - Then draw the example (喻): use familiar scenarios to illustrate the directly seen matter on which you rely.
This order has deep significance: the thesis is "what you claim," the reason is "what justifies your claim," and the example is "what makes others able to understand." In sequence, the three are "proposition → reason → expression." If any link is missing, the argument collapses.
6. Contradiction, Non-Contradiction, Valid Cognition, Invalid Cognition — Judging Whether the Thesis Really "Stands"
Listing all eight means of establishment does not mean every argument needs all eight. Rather, after presenting the reason and the example, they must undergo a final assessment — an assessment based on two pairs of concepts: "contradiction" (相违) and "non-contradiction" (不相违), and "valid cognition" (量) and "invalid cognition" (非量).
Contradiction vs. non-contradiction: - Contradiction has two cases: indeterminacy (不决定, the reason itself is ambiguous), or sameness with what is to be established (与所成相同, what you use to prove the thesis is itself the very thing to be proved — circularity); - Non-contradiction likewise has two cases: determinacy (决定, the reason is itself conclusive), or difference from what is to be established (与所成有别, the reason is something else, something independent).
Valid cognition vs. invalid cognition: - What is contradictory → cannot accomplish the thesis → cannot be valid cognition, and is not called "valid" (量); - What is non-contradictory → can accomplish the thesis → can serve as valid cognition, and is therefore called "valid".
This assessment tells us: possessing a means of establishment is not the same as its being effective. You may offer ten reasons; if all ten are "contradictory," then your argument is only formally reasoning — in substance, nothing has been established. Genuine establishment requires non-contradictory valid cognition — this distinction is the most easily overlooked, yet most crucial, piece of the vādaniśraya system.
II. A Walkthrough in Everyday Life
The theoretical "two to be established, eight that establish" sounds classical, but its structure actually maps directly onto our everyday practices of reasoning. Below, I walk through this architecture step by step, using several situations familiar to contemporary people.
Scenario 1: Deciding at Home Whether to Transfer the Child's School
Suppose you and your partner are discussing whether to transfer your child from the current school to another.
Step 1: Identify the "What Is to Be Established"
- Determination at the level of self-nature: Is "transferring" actually a real option? This seems trivial, but in fact — is "transferring" possible given our current living conditions, household registration, and policies? This is first clarifying the baseline of "what exists, establish as existing; what does not exist, establish as not existing." If there is simply no school available, the self-nature level fails first. - Determination at the level of distinctions: Even if transfer is possible, there follow immeasurable distinctions — quality of academic resources, commuting distance, the child's social circle, mental health, extracurricular activities, family financial burden, ties with grandparents... each item is a matter to be handled under "distinctions to be established."
Family meetings get stuck in endless argument not because the conclusion is wrong, but because the self-nature level has not reached consensus before jumping to the distinction level — people start comparing school rankings before even aligning on "is this choice actually open to us?"
Step 2: Unfold According to the "Means of Establishment"
- Establishing the thesis: "We maintain: for the child's overall development, we should transfer this semester." (Put the core proposition on the table first.) - Discriminating the reason: "Because the current school can no longer meet the child's needs in area X, and the new school has clear resources in area X." (Disclose the principle on the basis of what is directly seen.) - Drawing an example: "It's like changing doctors — not that the current doctor is bad, but this condition requires more specialized treatment." - Similar class: list cases of other families who transferred successfully (named or anonymous, but real). - Dissimilar class: list counterexamples of transfers that failed or produced adjustment problems (to avoid confirmation bias). - Direct perception: visit the new school in person, interview parents of current students, consult public data (this is immediate and verifiable). - Inference: infer the child's likely reaction from past patterns of adapting to new environments. - Authoritative teaching: child psychology research reports, Ministry of Education statistics, academic literature — these are "what external authorities teach."
Step 3: Assess Contradiction / Non-Contradiction
If the reason offered is "my colleague said that school is good" (this is "heard from others," but it is indeterminate, because the colleague's situation does not necessarily equal yours), or "the child will be happier at the new school" (this is sameness with what is to be established — "happier" is itself the very thing this argument is trying to prove) — these are all contradictory, and cannot serve as valid cognition.
What can genuinely serve as valid cognition are reasons independent of the conclusion and verifiable: for example, empirical data showing that the new school offers a particular gifted program, or an objective assessment of the child's current academic performance.
Scenario 2: Pushing a Proposal in a Work Meeting
You are a product manager, pushing for a new feature to launch at a meeting.
- Self-nature: Can this feature technically be done? This is not a trivial question — many meetings get stuck here because someone assumed "yes" when in fact it's "no." - Distinctions: launch timing, version, accompanying copy, A/B testing scope, budget, risk mitigation — each is a distinction. - Establishing the thesis: "Launching this feature this cycle will increase retention by 5%." - Discriminating the reason: "Based on the last 30 days of data, users who use similar features retain at a significantly higher rate than those who don't." - Drawing an example: "Like the small tool we launched last quarter, which saw a similar retention boost after launch." - Similar / dissimilar class: in the industry, similar features have both success stories and failures — both sides must be listed. - Direct perception: user interview recordings, prototype test data. - Inference: extrapolate from small-scale tests to full launch performance. - Authoritative teaching: industry white papers, academic studies.
Common contradictory traps in assessment: "This feature will definitely be popular" (sameness with the thesis), "that's how it was last time" (indeterminate, because the context may differ) — these sound forceful, but are actually holes in the argument.
Scenario 3: Arguments on Social Media
Arguments on X (formerly Twitter) or Weibo commonly fail precisely because the structure of vādaniśraya is not used at all:
- No thesis is established: people jump straight into details, and neither side knows what the other is trying to establish; - No reason is discriminated: only conclusions are given, no grounds; - No example is drawn: each side's metaphors talk past each other; - Similar and dissimilar classes are entirely absent: people only look at cases that support their own view, ignoring counterexamples; - Direct perception is ignored: "I feel" and "everyone says" substitute for empirical evidence; - Inference skips steps: going straight from A to Z with no chain in between; - Authoritative teaching is wielded as a club: "such-and-such a big account also says so" is one form of "heard from others," but if it is indeterminate (that account may itself lack rigor), it degenerates into contradiction.
After walking through these three scenarios, one can see: vādaniśraya is not a relic of ancient Indian scriptural debate. It is, in fact, an analysis of the inner structure of all serious reasoning activity. Our reasoning today often fails precisely because it neglects certain links in this architecture — especially the need for prior consensus on the self-nature of "what is to be established," and the final assessment of "contradiction / non-contradiction."
Places Where People Most Often Map It Wrong
- Understanding 论所依 as loosely as "the basis of an argument" — in fact it is a specific item with a specific tenfold content, not a general reference. - Narrowing 所成立义 to "the conclusion" — it actually has two layers, self-nature and distinctions, and the distinctions to be established include immeasurable categories. - Understanding 能成立法 as a single layer of "reasons" — the eight kinds each have their own division of labor; example, similar class, dissimilar class, perception, inference, and authoritative teaching each have distinct functions. - Assuming all eight means of establishment must be present to be effective — in fact, one assesses which are contradictory and which are not, and then decides which is "valid cognition." - Understanding 量 as "quantity" or "amount" — 量 here is an epistemological term meaning valid cognitive basis, not a matter of size or number.
III. Why Contemporary People Need It
1. In an Age of Exploding Opinions, Distinguishing "What Stands" from "What Does Not Stand" Is a Basic Survival Skill
The contemporary information environment is the most crowded "arena of argumentation" in human history: social media, comment sections, podcasts, short videos, instant messaging — in every second countless people are establishing theses, discriminating reasons, and drawing examples — but the vast majority do so without self-awareness. Vādaniśraya provides a "map": it lets you see which utterances are structurally sound and which merely look like they are.
The practical value of this map is that it separates "I disagree with you" from "you have not established anything." The former is mere subjective preference; the latter is a structural diagnosis. The most common mistake of contemporary people is to conflate the two — with the result that we spend enormous energy "expressing dissatisfaction" and very little "diagnosing the other's argumentative structure."
2. It Makes "Persuading by Reason" Actually Operable
Many people long for "rational dialogue" but find themselves unable to do it. The problem is often not attitude, but the lack of an operable methodology. What vādaniśraya provides is precisely an operable toolbox:
- Whom do you want to persuade? First write down the "what is to be established" clearly — are you establishing a determination of self-nature, or a particular distinction? - What is your basis? Direct perception (what you have seen with your own eyes), inference (what you deduce), or authoritative teaching (what authorities say)? These three cannot be mixed. - Is your reason contradictory? Circular? Indeterminate? Or genuinely independent of the conclusion?
Once this procedure is internalized, reasoning is no longer a pile of vague phrases like "I feel," "everyone says," "experts say," but a structured, testable process.
3. It Treats a Uniquely Contemporary Disease: the Dysfunction of "Authoritative Teaching"
Ancient Buddhist argumentation listed "authoritative teaching" (至教, or 正教) as one of the eight means of establishment — not because authority is self-evidently valid, but because at certain moments, an already-verified inherited wisdom is an unignorable basis. But the contemporary problem is precisely that "authoritative teaching" is torn between two extremes: severe abuse and severe neglect.
On the one hand, "Professor So-and-So says," "mainstream media reports," "a famous company does it this way" are wielded as unchallengeable clubs of authority; on the other hand, genuine academic research, historical experience, and cross-cultural wisdom are cast aside as "outdated" or "irrelevant." Vādaniśraya reminds us: authoritative teaching can be valid cognition, but only on the condition that it is non-contradictory — that is, independent of the thesis to be established, testable, and not falling into circularity. If the "authoritative teaching" is itself what the argument is trying to prove (e.g., using the thought of school A to prove a proposition of school A), then it is the "sameness with what is to be established" type of contradiction, and cannot be valid cognition.
The concrete implication for contemporary people: neither worship authority unconditionally nor reject tradition unconditionally — look at what role it actually plays in the structure of this particular argument.
4. A Dialogue with Modern Epistemology and the AI Era
Contemporary discourse talks a great deal about "epistemology": What is knowledge? What is belief? What is a justified inference? Vādaniśraya was already dealing with these questions two thousand years ago — the threefold classification of "direct perception" (现量), "inference" (比量), and "authoritative teaching" (至教) corresponds strikingly to the empirical, inferential, and testimonial sources in contemporary epistemology.
But it must be clearly stated: this correspondence does not mean Buddhism has been scientifically proven, nor that the Buddhist classification is an ancient version of modern epistemology. The contexts, goals, and presuppositions of the two differ — vādaniśraya serves the argumentative activity of "causing others to understand and believe," and is not pure academic epistemology. Contemporary readers may draw inspiration from it, but cannot therefore claim the shallow correspondence that "Buddhism already had modern logic."
As for the advent of AI, it gives this architecture a new field of application: when we want to assess whether an AI output "stands," we can apply this tenfold framework — What thesis has the AI established? What reason does it use? Is its example reasonable? Are its similar and dissimilar classes well chosen? What is its source of direct perception? Is its chain of inference complete? Is its authoritative teaching credible? Is the overall argument contradictory?
This is not to say that AI is a new "Buddha," but rather: this ancient architecture of argumentation remains valid for evaluating any system of reasoning — because its focus has never been "who said it," but "whether what is said actually stands."
IV. Common Misreadings and Clarifications
Misreading 1: "论所依 = Buddhist Logic"
Clarification: vādaniśraya is one structural item within the discipline of argumentation, closely related to, but not identical with, 因明 (hetuvidyā, Buddhist logic). The science of hetuvidyā (as developed in the systems of 陈那 (Dignāga) and 法称 (Dharmakīrti)) later developed far more refined tools — theories of pramāṇa, apoha, and parāvṛtti — beyond vādaniśraya. Vādaniśraya is the earlier, more basic Yogācāra framework of argumentation.
People often conflate the two, but when reading scriptures, "论所依" should be understood as the tenfold architecture of the Yogācāra school, not the specialized terminology of later hetuvidyā.
Misreading 2: "所成立义 Is a Proposition, 能成立法 Is an Argument"
Clarification: this is a slippage of modern logical language. 所成立义 is the object to be accomplished by the activity of argumentation — it can be a proposition (at the level of distinctions to be established), or it can be a fundamental determination of "yes / no" (at the level of self-nature). 能成立法 is the eight tools by which this object is established in the mind of the other — not "argument" in a single sense.
Translating 能成立法 directly as "argument" loses the nuance of the eightfold classification — especially the five items "similar class," "dissimilar class," "direct perception," "inference," and "authoritative teaching," which look like they all belong to "argument," but each has an independent function and cannot be lumped together.
Misreading 3: "立宗 = Stating a Claim, 辩因 = Giving a Reason" — That Simple
Clarification: establishing the thesis must handle more than "stating a claim" — it must specify the source (derived from the teachings? established by one's own argument? heard from others?) and the purpose (establishing one's own school? revealing another's fault? humbling arrogance? crushing contempt? compassion?). Discriminating the reason is also not a vague "giving a reason" — it must discriminate and clarify the principle by means of the six resources of "example, similar class, dissimilar class, direct perception, inference, and authoritative teaching," without departing from the thesis.
In other words, both establishing the thesis and discriminating the reason are activities with internal structure, not merely "a single sentence."
Misreading 4: "同类 and 异类 Are Just Giving Examples"
Clarification: similar class and dissimilar class do have an "example-giving" component, but their functional positioning is more precise — they are specific operations for determining whether the reason and example are contradictory. Similar class means "everything of the same class as the thesis exhibits this principle"; dissimilar class means "everything of a different class from the thesis does not exhibit this principle." Logically, this corresponds to the positive and negative verifications of induction — not simply grabbing examples at random.
Misreading 5: "现量 = Seeing Is Believing, 比量 = Reasoning" — That Direct
Clarification: direct perception and inference have far more refined classifications in Buddhist epistemology — direct perception includes the five sense cognitions, mental discrimination, and sometimes even meditative direct perception; inference must be assessed by whether the reason is determinate and different from what is to be established, to determine whether it is valid cognition. Reducing direct perception to "seeing" and inference to "reasoning" loses Buddhism's refined requirements for "valid cognition."
Misreading 6: "至教 = Quoting Scripture to Overwhelm Others"
Clarification: authoritative teaching is indeed "citing the noble teaching as proof," but its validity is conditional — it must be non-contradictory (that is, not sameness with what is to be established, and not indeterminate). In other words, quoting scripture is not the whole of "authoritative teaching" — if the scripture quoted is itself the object the argument is trying to prove, that is circularity; if the scripture quoted is ambiguous and cannot decide the matter, that is indeterminacy. Only when the authoritative teaching cited is an independent, determinate basis can it become valid cognition.
The practical reminder for contemporary people: when citing an authority, ask yourself — is the authority I am citing independent of the conclusion I am trying to prove? Can it itself be verified?
Misreading 7: Treating "量" as Quantity or Degree
Clarification: 量 here is an epistemological term, equivalent to "valid cognitive basis" (pramāṇa). It is binary: either it is valid cognition, or it is not — not a matter of degree like "this reason is 80% valid." Whether something is "valid" depends on whether it is non-contradictory, determinate, and different from what is to be established — pass these tests and it is valid; fail them and it is invalid.
Misreading 8: Ignoring the Boundary "For the Sake of Producing Faith and Understanding in Others"
Clarification: this is the most important spiritual qualification of the entire vādaniśraya system. The purpose of argumentation is to cause others to gain faith and understanding in already-accomplished dharmas, not to create or alter the own-nature of dharmas. Contemporary people are most prone to treating argumentation as a "real battlefield" — if you win, you grasp the truth; if you lose, the other side is wrong. But the position of vādaniśraya is: the own-nature of all dharmas is already accomplished; argumentation merely causes this fact to appear in the mind of the other. Winning a debate does not mean grasping reality, and losing does not mean being wrong — what matters is whether the "thesis established" is truly non-contradictory with "valid cognition."
This boundary has a particularly healing effect for contemporary people: we need not treat every conversation as a life-or-death struggle, nor assume we stand on the side of truth merely because we "won" some debate. Truth lies outside argumentation; argumentation is the tool that reveals truth, not truth itself.
Conclusion
Vādaniśraya looks like an obscure item in the discipline of argumentation, but unpacked, it is a precise anatomy of "the activity of reasoning": what you are trying to establish (the two kinds of 所成立义), what you use to establish it (the eight kinds of 能成立法), and whether what you establish really stands (the assessment of contradiction / non-contradiction, valid / invalid cognition).
The reason this architecture retains value across more than two thousand years is that its core question — "what kind of utterance counts as established?" — has never gone out of date. In an age of exploding opinions, of stance-over-everything, of infinite expression and scarce argumentation, revisiting this ancient anatomy can make our dialogue clearer, more responsible, and gentler — because it always reminds us: the purpose of argumentation is not to win, but to let the already-accomplished real abide in the mind of the other.