Dharma · AI Companion

不成

asiddhi

不成(asiddhi),因明术语,指因体或宗之有法、能别等未能于立敌共许而成立之似过。AI-generated

Type PhenomenaDifficulty IntermediateTruth-level span outside the four-fold two-truths frameworkInitial AI estimate · evolves with use

A Contemporary ReadingAI-generated


What This Concept Is About

The One-Sentence Core

不成 (asiddhi, "non-establishment") is a category of fallacy in Buddhist logic (因明, *hetu-vidyā*) that names the moment a debate, a proof, or a piece of reasoning collapses — not because its conclusion is wrong, but because the shared ground on which any reasoning could stand has not been laid down between the two parties. The reason (因, *hetu*), the subject of the proposition (有法, *dharmin*), the predicate, or the locus on which the reason depends — none of these have been mutually acknowledged (*不極成*) by the proponent (立者, *pratijñā-kāraka*) and the opponent (敵者, *pṛthagjana* / counter-thesis holder).

This is the most fundamental failure mode in reasoning: not a failure of *conclusion*, but a failure of *foundation*.

Where It Sits in the Whole Path to Liberation

In the Tibetan Buddhist curriculum, 因明 is not itself a "liberating step" the way 止 (śamatha) or 觀 (vipaśyanā) are. It is something more like epistemological hygiene — the rigorous training that makes the mind a reliable instrument *before* one attempts the far more delicate work of seeing emptiness (空, *śūnyatā*) and dependent origination (緣起, *pratītyasamutpāda*).

不成 is the diagnosis for the *first* disease a valid cognition (量, *pramāṇa*) can suffer. Without being able to spot non-establishment, a practitioner cannot tell the difference between reasoning that leads to insight and reasoning that *looks* like insight but rests on air. The whole edifice of the pramāṇa tradition — the three characteristics of a valid reason (三相, *trairūpya*), the proper relation of example (喻, *dṛṣṭānta*) to proposition (宗, *pratijñā*) — is built on the assumption that the foundation is shared. 不成 is what you get when that assumption is broken. In a path that depends on the reliable operation of valid cognition for its highest fruit, learning to recognize 不成 is not optional. It is the difference between a ladder and a painted ladder.

The Classical Taxonomy — Two Texts, One Core, Different Edges

The two foundational Indian texts for this concept — 《正理滴論》(*Nyāyapraveśa*, attributed to Dharmakīrti / 法稱) and 《因明入正理論》(*Nyāyapraveśa* by Śaṅkarasvāmin / 商羯羅主) — diagnose the same disease but cut it along slightly different lines. A serious reading needs both.

From 《正理滴論》: Three Forms Inside the Reason

This text lists three specific kinds of 不成 within the *reason itself* (似因之不成), and adds a structural claim about the three characteristics of any valid reason:

**1. 猶豫不成 (*samśayā-siddhi*, "uncertain non-establishment") — The reason's own identity is undecided. The classical example: someone tries to prove "there is fire on that mountain" using a column of mist-or-smoke** (疑惑之霧氣等) as the reason. From a distance, you genuinely cannot tell whether what you see is smoke (煙) or vapor (蒸氣、霧氣). The reason — "smoke, therefore fire" — works only if you have smoke. If the evidence is ambiguous between smoke and mist, the reason itself *oscillates*, and an oscillating reason establishes nothing. This is not a problem with the *conclusion* (fire might genuinely be there) — it's a problem with the *warrant* (we cannot commit to calling it smoke).

2. 所依不成 ("non-establishment of the locus/basis") — The place where the reason is supposed to inhere cannot be pinned down. Classical example: someone tries to prove "there is a peacock in this mountain range" using the peacock's call (孔雀鳴) as the reason. The problem is not the call itself (peacocks do call) and not the conclusion (there may well be a peacock). The problem is that *we do not know where the call is coming from*. Is it from this mountain? From the next? From a hidden valley? Until the locus is fixed, the reason — "a call heard here, therefore a peacock here" — has no foothold.

**3. 有法不成 (*dharmin-asiddhi*, "non-establishment of the subject")** — The very thing you are trying to prove something *about* does not exist for your opponent. Classical example: a Vaiśeṣika (勝論派) philosopher tries to prove "I am everywhere" using "because there is a quality (功德) that reaches to all places" as the reason. The reason is built on the assumption that there *is* an "I" (ātman) — a self — that could be everywhere. But a Buddhist opponent (敵論者如佛家) does not concede that any such "I" exists. So the very subject of the debate is not jointly accepted (*不極成*) by both parties. The reason is perfectly logical *given* the subject, but the subject is not on the table to begin with.

A structural claim the text also makes: each of the three characteristics (三相) of a valid reason — that the reason is present in the subject (*pakṣa-dharmatā*), that it is present in things homogeneous with the subject (*sapakṣa-dharmatā*), and that it is absent from things heterogeneous with the subject (*vipakṣa-dharmatā*) — carries its own *two* possible faults: 不成 (the characteristic fails to be established) and 猶豫 (uncertainty about whether it is established). Uncertainty can thus be both a *type* of 不成 and a *cause* of 不成 — a subtle point this text makes explicit.

From 《因明入正理論》: Two Layers — Proposition *and* Reason

This text treats 不成 in two distinct locations, which is itself an important teaching: the disease can infect either the proposition you're trying to establish, or the reason you're using to establish it.

Layer One — In the Proposition Itself (似宗過之「不極成」諸相):

Among the five faults of a faulty proposition, three are directly about 不極成 — non-mutual-establishment:

- 能別不極成 (predicate not mutually established): When you try to predicate something of the subject, but the *predicate* (the thing you're saying about the subject) is not accepted by your opponent. Example: a Buddhist debater says to a Sāṃkhya opponent, "Sound is perishable" (聲滅壞). The Buddhist accepts this, but the Sāṃkhya does not accept that sound perishes. The whole predicate fails to land.

- 所別不極成 (subject not mutually established): The mirror case. The Sāṃkhya says to a Buddhist, "I am thought" (我是思). The Sāṃkhya accepts this, but the Buddhist rejects the existence of any permanent "I" to begin with. The subject doesn't get a foothold.

- 俱不極成 (both not mutually established): Both ends fail at once. A Vaiśeṣika says to a Buddhist, "I am the cause of conjunction" (我以為和合因緣). Neither the "I" as subject nor "cause of conjunction" as predicate is accepted by the Buddhist opponent.

This text adds a fourth, paradoxical category — 相符極成 — where the proposition is *so well* agreed upon by both sides that it becomes useless as a *thesis*. If both parties already accept it, you haven't established anything; you've merely stated the obvious. The text names the disease precisely: 「不容成故」 — it does not *fail* to be established, but it *cannot serve* as something to be established. This is a brilliant diagnostic: the failure of the proposition can be either that the ground is missing *or* that there is no ground *needed* because everyone already stands on it.

Layer Two — In the Reason Itself (似因之四種不成):

- 兩俱不成 (non-establishment for both parties): Neither you nor your opponent accepts the reason as applying to the subject. Classical example: trying to prove "sound is impermanent" using "because it is visible" (眼所見性故) as the reason. Neither party accepts that sound is visible. The reason is shared non-existence, not shared existence — which is worse than no reason at all.

- 隨一不成 (non-establishment for one party): One side accepts the reason, the other doesn't. Example: trying to prove something to a Sāṃkhya philosopher using "because it is produced" (所作性故) as the reason. The Buddhist proponent accepts that produced things are impermanent, but the Sāṃkhya opponent denies that sound (which is eternal in their system) is *produced*. The reason holds in your world but not in theirs.

- 猶豫不成 (uncertainty non-establishment): Same logic as in the 《正理滴論》 account — the reason itself is undecided. Example given: trying to prove "there is fire produced by the conjunction of the great elements" using ambiguous perceptual evidence about whether such fire is present.

- 所依不成 (non-establishment of the basis): Same logic — the locus is uncertain. Example given: trying to prove "space (虚空) is real" using "because it is the substratum of qualities (德所依故)" as the reason, addressed to a school (無空論) that denies space exists at all. If the opponent doesn't even grant that space exists, then *whatever* you say about space's qualities has no place to land.

The Subtle Difference Between the Two Texts

The texts aren't contradictory; they're like two clinicians examining the same patient from different angles:

- 《正理滴論》 locates "subject non-establishment" (有法不成) *within* the faults of the reason, because for that text, if the subject is contested, the whole reason-subject-predicates triangle collapses from within. - 《因明入正理論》 keeps it in a separate category — faults of the proposition (似宗過) — because for that text, the proposition and the reason are two distinct moves, and either can fail on its own.

Both readings are coherent. Both texts are needed to see the full shape of the disease.

Why This Concept Is Not Just "Being Wrong"

不成 is the easiest fallacy to mis-identify because it looks like *other* failures. It is not:

- **相違 (*viruddha*, "contradictory")**: when the reason actively opposes what it's supposed to prove (e.g., using "because it is permanent" to prove impermanence). 相違 names an *opposition* in the reasoning. 不成 names a *missing ground*. - **不定 (*anaikāntika*, "indeterminate")**: when the reason could prove too many things (e.g., "sound is impermanent because it is knowable" — but pots are also knowable and are impermanent for unrelated reasons). 不定 names an *over-reach* in the reasoning. 不成 names a *missing base*.

The classical image that captures all three: a building. 相違 is a load-bearing wall placed backwards (it actively works against the structure). 不定 is a column that holds up too many unrelated roofs (it spreads itself thin). 不成 is no foundation at all — the building is in the air, and we're arguing about what kind of roof to put on.


Life Walkthrough: Walking Through This Concept in Real Scenes

The classical examples (smoke-or-mist from a distance, a peacock's call from an unknown place, trying to prove "the self exists" to a Buddhist, trying to prove "sound is visible") can feel remote. But every category maps precisely onto situations we are in constantly. Let me walk through several.

Scene 1: The Couple Arguing About "Quality Time"

The argument: "You never spend quality time with me." / "I am with you every evening — I just need to decompress by scrolling after work."

What is actually happening, in 因明 terms: The two parties are not disagreeing about whether they spend time together. They are failing at 俱不極成 — both subject and predicate are unestablished between them. What is "quality time"? The proponent (the partner who feels neglected) defines it as engaged, present, undistracted attention. The opponent (the partner who scrolls) defines it as co-located time with low cognitive demand. Neither party has made the other party's definition *the* shared ground. So when the reason "you never spend quality time with me" is offered, the predicate ("quality time") has nowhere to stand. The argument looks like a disagreement about quantity or behavior; it is actually a non-establishment of what the debate is *about*.

How to fix it using the classical framework: Step back from "you're wrong about quantity" and address the missing ground: "I need us to agree on what we mean by 'quality' before we can argue about whether there's enough of it." This is precisely the move of *establishing the basis* before proceeding to the proof.

Where people commonly map this incorrectly: Most couples treat this as a 相違 — "you say X, I say not-X, we're opposed." It's not. You're not yet opposed. You're both standing on different ground, arguing across a gap. The conflict isn't between your positions; it's between your *definitions*.

Scene 2: The Work Strategy Meeting That Goes Nowhere

The argument: Two executives argue about whether to invest in a new product line. One says "the data shows consumer demand." The other says "the data shows market saturation."

What is actually happening: Look at the structure. Each side has its *own* data set, its *own* definition of "demand" (units shipped? survey willingness? search trends? retention curves?), and its *own* time horizon. The reasoning — "we should invest *because* demand exists" — has a reason that 猶豫 is operating on: is the evidence actually evidence of demand, or is it evidence of something else? This is the smoke-or-mist example, exactly. Each side is pointing at a column of vapor and insisting it's smoke (or insisting it's mist). The data is uncommitted between interpretations; therefore the reason is uncommitted; therefore no conclusion can be established.

Classical fix: Before arguing about the conclusion, you have to do the work of establishing what the data is, in a way both sides accept. If you cannot do that — if "demand" remains ambiguous — the right move is to admit that no valid reasoning is possible yet, and to gather more disambiguating evidence. The 《正理滴論》 point is exact: an uncertain reason establishes nothing, and that is not a problem to be argued through but a problem to be *investigated out of*.

Common mis-mapping: Most strategy meetings treat this as "we need to weigh the arguments" (相違 framing) or "we need more data" (vague intuition). The 因明 diagnosis is sharper: the *specific* fault is 猶豫不成, and the *specific* remedy is to commit, by joint investigation, to what the data actually says before invoking it as a reason.

Scene 3: The Online Debate About "Is AI Conscious?"

The argument: Commenter A: "AI is conscious because it responds in emotionally appropriate ways." Commenter B: "AI is not conscious because it has no subjective experience." Commenter C: "Consciousness is just information processing, so of course AI is conscious."

What is actually happening: This is a textbook 有法不成 or 能別不極成 situation, depending on which move you examine. The *subject* — "AI" — might be granted by all parties, but the *predicate* — "conscious" — is not. Or, worse, the *subject* of the discussion ("consciousness as such") is itself contested: behaviorists, functionalists, dualists, and Buddhist no-self theorists each grant a different thing when they say the word.

If the parties cannot first establish what they mean by "consciousness" in a way all parties accept, then *every* subsequent reason — "because it responds emotionally," "because it lacks qualia," "because it processes information" — falls into 不成. Not because the reasons are bad in themselves, but because the predicate they are trying to attach to AI is not mutually established. You are arguing about what color to paint a wall whose existence is contested.

Classical fix: The 《因明入正理論》's "相符極成" warning is also alive here: if everyone *already* agrees on what "consciousness" means (within some narrow technical sense), then arguing whether AI has it is fine. If no one agrees — if the term is doing radically different work for different speakers — then before you can even *have* the debate about AI, you need a meta-debate about the term. And if no agreement is possible on the term, the specific claim "AI is conscious" cannot be established *as that specific claim*. It can perhaps be established as some weaker, rephrased claim using mutually accepted terms. But that's a different argument.

Common mis-mapping: People treat this as "we just need more evidence about AI" (treating it as 不定, a problem of insufficient data). It's not. You could give me all the neural-network weights in the world and the disagreement would persist — because the disagreement is *upstream* of evidence, in the definition of the predicate.

Scene 4: The Family Member Who Says "You're Just Like Your Father"

The argument: "You always do that — it's exactly what your father would do." / "I am nothing like him. I have nothing in common with that man."

What is actually happening: Both parties are making a claim that requires the predicate "alike your father" to be mutually understood. But "alike" is the contested ground. One person means "alike in this specific behavioral pattern." The other means "alike in essence, in character, in fate." The predicate is not mutually established (能別不極成), and the reason being offered — "you do X, just like he did" — is a 隨一不成: it holds for the speaker (who sees the pattern), but is rejected by the hearer (who does not accept either the likeness or its significance).

Classical fix: Separate the proposition. "You did X today" can be argued. "Doing X is a sign of being like your father" can be argued. "Being like your father is bad" can be argued. But *conflating* all three into "you're just like your father" is an 不成 failure: the predicate is too thick, carrying too many contested elements at once. The 因明 remedy is to *thin out* the proposition until both parties can at least accept the thinnest version.

Common mis-mapping: Most people treat this as a 相違 ("I say we're alike, you say we're not"). It is *pre*-相違 — it is the failure of a shared predicate that would *make* a 相違 possible. Until the predicate is established, there is no opposition; there are two separate arguments happening in two separate rooms.

Scene 5: The Doctor and Patient With Different Definitions of "Better"

The argument: Doctor: "Your numbers are improving — you're getting better." Patient: "I don't feel better. I'm not better."

What is actually happening: Two predicates are in play — "better" as biomarker improvement, and "better" as felt well-being. The doctor is operating in the medical-pramāṇa frame. The patient is operating in the experiential-pramāṇa frame. Neither is *wrong* within their own frame. But when the doctor offers the reason "your numbers improved, therefore you are better," the patient experiences this as 能別不極成: the predicate "better" does not mean what the doctor thinks it means to the patient. The patient's 所依不成 (basis non-establishment) is also operative: the "basis" for "better" in the patient's world is not the lab values, so the reason has nowhere to attach.

Classical fix: Before arguing about whether the patient is "better," the parties need to establish what "better" is jointly taken to mean — at least for the purpose of this conversation. If they cannot, they should speak in two separate vocabularies: "your biomarkers have improved" (a proposition the patient might accept as factually true) and "your felt sense of health" (a separate proposition). The 因明 lesson is that one proposition cannot do the work of two.


Why Contemporary People Need This Concept

1. We Live in an Argument-Saturated Ecosystem Without an Argument-Saturated Training

The contemporary information environment — social media, op-eds, podcasts, group chats, comment threads — produces more *attempts at reasoning* per day than most premodern humans encountered in a year. But the training in *how to diagnose the failure of those attempts* has not kept pace. Most people can name a fallacy if it is "ad hominem" or "slippery slope" — but they cannot see the more common, more silent failure: two people arguing with elaborate reasons, where neither party's foundation is shared, and so the entire debate establishes nothing.

不成 is the diagnosis for the silent failure. We need it because we live inside the silent failure constantly.

2. The AI Era Has Made "What Does This Term Mean?" the Central Question

The rise of large language models, recommendation algorithms, and AI assistants has turned *definitional grounding* from a philosopher's hobby into an engineering bottleneck. Every system that uses a term like "helpful," "safe," "relevant," "fair," "accurate," or "harmful" must do some version of the work 《因明入正理論》 describes: establish what the term means to the system and to the user before the system can act on it. When this fails — and it fails often — the failure has the exact shape of 不成: the system and the user are reasoning, but their predicates are not mutually established.

A clear-eyed contemporary reading of 不成 can sharpen how we think about specification, alignment, prompt design, and the experience of "the AI just doesn't understand what I mean." The 因明 framework says: that experience is real, and it has a name, and the fix is to do the foundational work the framework demands — not to keep arguing past each other.

3. It Is the Antidote to a Particularly Modern Vice: Confusing Eloquence for Establishment

A speaker who argues beautifully for a position can make an audience *feel* that the position has been established. In a media ecology optimized for engagement, the felt sense of establishment is often *sold in place of* actual establishment. 不成 is the concept that lets you pause and ask: "Before I respond to the force of this argument, is the foundation actually shared? Is the reason actually a reason in this discussion? Is the predicate something we both mean by the same word?"

The classical example of 兩俱不成 — using "visible" to prove something about sound, when no one accepts that sound is visible — looks absurd on the page. But its contemporary cousins are everywhere: using metrics that no one agrees measure what they claim to measure; using analogies that no one agrees are analogous; using authority that no one agrees is authoritative in this domain.

4. It Respects the Other Person Without Caving to Them

不成 is sometimes mistaken for a kind of relativism — "if we don't share the ground, no reasoning is possible, so anything goes." This is the opposite of what the framework teaches. 不成 says: *reasoning requires shared ground*. That is a *high* standard. It tells you that many things currently passing as reasoning are not reasoning at all. It is a rigorous demand, not a permissive one.

In a time when public discourse routinely confuses "loudly stating a position" with "establishing a position," 不成 is one of the most clarifying concepts available. It restores the distinction between *voicing* and *establishing*.

5. It Is a Quiet Form of the Buddhist Path's Core Commitment

The Buddhist path, at its deepest, is committed to the unreliability of unexamined cognition — to the idea that much of what we take as established is, on inspection, not established at all. 不成 is that commitment, applied at the level of reasoning before it is applied to the level of self and world. To practice spotting 不成 in daily argument is to do a *microcosm* of the larger Buddhist work of seeing that what we treat as solidly given (a self, a world, a permanent ground) is, in fact, not established when looked at carefully.

There is a deep continuity between 因明 and 般若: the pramāṇa tradition trains the mind to see what is and is not properly established at the level of *cognition*, and the prajñā tradition extends the same discipline to the level of *being*. 不成 is the early, technical form of the same instinct.


Common Misreadings and Clarifications

Misreading 1: "不成 just means 'wrong'"

It doesn't. The classical examples make this clear. You can have a conclusion that is *true* — there *might* be fire on that mountain, there *might* be a peacock in that range — and still have the *reason* be 不成 because the warrant (this is smoke, that call is from here) has not been jointly established. 不成 is a diagnosis about the *warrant structure*, not the truth value of the conclusion.

Misreading 2: "不成 is the same as 相違 (contradiction)"

相違 names a *clash within* the reasoning — the reason actively opposes what it's supposed to prove. 不成 names an *absence* beneath the reasoning — there is no shared ground on which any reasoning could stand at all. The classical image: 相違 is two people on the same floor pushing opposite walls; 不成 is two people in different rooms arguing through the wall.

Misreading 3: "不成 is the same as 不定 (indeterminate)"

不定 names an *over-extension* of the reason — it proves too much (the reason "knowable" applies to pots as well as sound, so it cannot distinguish impermanent things from permanent ones). 不成 names an *under-extension* of the foundation — there is no ground to extend from. Different diseases, different remedies.

Misreading 4: "不成 is just about formal debates — it doesn't apply to ordinary life"

The opposite is closer to the truth. Formal debates have at least some procedural pressure toward establishing the foundations. Ordinary life — text messages at 11pm, family meals, group chats, performance reviews — has almost none. This is precisely why the *failures* of 不成 are so much more common in ordinary life, and why the concept is so practically valuable.

Misreading 5: "If there's 不成, the only response is to give up"

No. The 《因明入正理論》 and 《正理滴論》 are not defeatist texts. They are diagnostic. The proper response to diagnosing 不成 is to *do the foundational work that the argument skipped*: clarify the predicate, fix the locus, agree on what evidence is evidence of what, separate overlapping propositions, and *then* re-attempt the proof. 不成 is the beginning of a more rigorous argument, not the end of one.

Misreading 6: "不成 means the opponent is being unreasonable"

This is perhaps the most insidious misreading. The classical examples are scrupulous: 不成 is not a charge that the *opponent* is being unreasonable. It is a charge that the *argument* has not done its foundational work. In many cases (the peacock call example, the smoke-or-mist example), the *proponent* is the one at fault — the argument was set up before the ground was cleared. 不成 is, in this sense, a humbling concept: it tells you, as often as not, that the failure is in *your* setup, not in your opponent's resistance.

Misreading 7: "Modern scientific reasoning has solved 不成"

Partially. Modern science has developed rigorous procedures (operational definitions, controlled experiments, peer review, pre-registration) precisely to address what 不成 names. But science is a *practice*; it can be honored in the breach as easily as any other. Paradigm shifts, replication crises, definitional drift in clinical endpoints, and the replication of "evidence-based" claims that aren't actually mutually established between researchers — all of these are modern, real-world instances of 不成, sometimes hiding inside vocabulary that sounds like science. The 因明 framework can *see* these failures even when the scientific vocabulary cannot name them, because the 因明 framework cuts along a different line — not the line of method, but the line of *what has been jointly granted*.


A Closing Image

The classical examples are vivid and worth carrying with you: the column of vapor at a distance that you cannot commit to as smoke; the peacock's call whose source you cannot locate; the proof of a self that your opponent does not concede exists; the visible quality attributed to sound that no one accepts; the "substratum of qualities" attributed to a space your opponent denies.

These are not exotic cases. They are the structure of the moment, daily, when we argue with reasons that look powerful but rest on air — when the predicate has not been jointly defined, the locus has not been jointly fixed, the evidence has not been jointly accepted as evidence, or the very subject of the debate has not been jointly granted to exist.

不成 is the old, precise, gentle name for that moment. To learn it is to learn how to stop building on air. And in a path that ultimately rests on seeing through all *apparent* establishment to what is *actually* established, that is no small beginning.

Canonical EntryAI-generated

一、定义与核心要义

「不成」为因明学中似因三分及似宗过之重要类目,总谓因之体性、所依或有法等不能于立敌双方成立、不能决定之过。《正理滴论》将似因分为不成、相违、不定三类,列不成为其中之首。「《因明入正理论》」则于似因三类(不成、不定及与相违)外,于似宗过中亦以「不极成」之诸相说明宗义不能成立之病。其核心判准在于「能为立敌共许」——因之成立必须不疑、不缺、所依明确,方为正因;反之即犯不成之过。

二、主要阐述

关于「不成」之内容,两部论典所论角度与细目略有不同。

《正理滴论》就似因之不成分举数种形态:其一为「犹豫不成」,若对因之性质及所成立之事起疑惑,因体本身犹豫不定,则犯不成过。论中举「成立有火,以疑惑之雾气等为因」为例:远处似烟而实为蒸气、雾气,究竟为烟为雾不能确定,故因体不决定而犯过。 其二为「所依不成」,因所系属的地方不能确定时亦犯此过,如以孔雀鸣为因成立此山间有孔雀,然对鸣处尚未明了,因体所依之处不确定,故亦犯不成过。 其三为「有法不成」,有法(即宗上之主词)本身不成时亦犯此过,如胜论派以「于一切处缘到之功德」为因成立「我于一切处有」,然敌论者如佛家不许有我,则此有法根本不成立,由有法连带而来之过,故名有法不成。 其四则关于异品遍无相之过,若于同品异品皆遍转或于任何一品遍转,则为似因不定过之「共不定」,细分三类:同品遍转异品分转、同品分转异品遍转、俱品一分转。 此外,论中明言三相中之每一相皆各有二过,即「不成」与「犹豫」,若对因之相起疑惑,其结果或为不成,或为犹豫不定。 又言「相违决定」乃犹豫之因,然以其不成为比量推理之对象,于自性因、果因、未缘到因之性相中既然无相违决定,故不另立叙述。

《因明入正理论》之论「不成」,分见于似宗过与似因两处。经文先标似宗过五相——自语相违、能别不极成、所别不极成、俱不极成、相符极成,其中三种与不成直接相关:能别不极成(如佛弟子对数论师立「声灭坏」)、所别不极成(如数论师对佛弟子说「我是思」)、俱不极成(如胜论师对佛弟子立「我以为和合因缘」)。相符极成以立宗反成共许之义而失立宗之用,故经文总结其病曰「不容成故」。 经文又于似因开四种不成:(一) 两俱不成——立者与敌者双方皆不许因之成立,如立「声为无常」以「眼所见性故」为因;(二) 随一不成——立者或敌者一方不许因之成立,如以「所作性故」对声显论立宗;(三) 犹豫不成——因之成立犹豫不决,如于雾等大种和合之性起疑惑时为成「大种和合火有」而有所说;(四) 所依不成——因之所依于敌者不极成,如「虚空实有,德所依故」对无空论。

三、不同部类与译本的表述差异

两部论典于「不成」之名相体系互有异同:

《正理滴论》于似因不成中,列「犹豫不成、所依不成、有法不成」,未明列「两俱不成」「随一不成」。《因明入正理论》则明列四种:「两俱不成、随一不成、犹豫不成、所依不成」。《正理滴论》之「有法不成」实即《因明入正理论》似宗过中「能别不极成、所别不极成、俱不极成」所示之义——两论皆以有法或宗之义未能成立而名过,唯《正理滴论》将之摄入似因不成之中,《因明入正理论》则别立于似宗。

就犹豫与不成之关系,《正理滴论》明确说三相各具「不成」与「犹豫」二过,将二者并列为对治对象;《因明入正理论》则将「犹豫不成」列为似因不成四种之一,两种体系中「犹豫」与「不成」之摄属略异。

此外,《正理滴论》另论及「共不定」之三细类(属于相违或不定过之范畴),《因明入正理论》则未细分此类。

四、与相关概念的关系

「不成」与「相违」「不定」并列为似因三分。《正理滴论》明言相违有二种:一者非未全具因而生之果,有他则自无,如冷触与热触;二者一方排另一方而存之唯实相,如「有」与「无」。 又《正理滴论》以疑惑明不成之理,云:「虽未见到,但凡一切智者皆未言说,此说所遣,亦不成立,盖生起疑惑故」——言说与一切智二者不相违,故虽遣除而不能决定成立,此即以疑惑(犹豫)说明不成之所以成立之因由。 三相各具不成、犹豫二过之说,显示此三者于因明体系中相互交涉:不成为过之根本,犹豫为生起不成之因由,不定、相违又与不成共摄属于似因之范畴。

《因明入正理论》中,似宗之「不极成」诸相与似因之「不成」相通:前者就宗之能别、所别不为敌者所许而立,后者就因之体性不为敌者所许而立,二者皆以「不为敌者所许」为过之核心。

五、代表性原文引用

1.「虽未见到,但凡一切智者皆未言说,此说所遣,亦不成立,盖生起疑惑故。」——此即以疑惑明不成之理,出《正理滴论》。

2.「若对因之相起疑惑,其结果或为不成,或为犹豫不定。」——此明三相各具不成、犹豫二过,出《正理滴论》。

3.「不成、不定及与相违。」——《因明入正理论》标似因三类之语。

4.「两俱不成、随一不成、犹豫不成、所依不成」——四种不成之名目,见《因明入正理论》。

5.「似宗五相……能别不极成、所别不极成、俱不极成、相符极成……总结其病曰『不容成故』。」——出《因明入正理论》。

By SutraAI-generated

正理滴論(譯自西藏文本)B0042 · 藏傳論典部

《正理滴論》對「不成」之論述,可從似因三分與三相二過兩方面把握。

第一,宗、義或所作若不能成立,則犯似因之「不成」過。論中將似因分為三類——不成、相違、不定,此即法稱論師所立之體系。

第二,「不成」有多種形態,論中列舉如下:

其一為「猶豫不成」。若對因之性質及所成立之事起疑惑,因體本身猶豫不定,則犯不成過。論中舉「成立有火,以疑惑之霧氣等為因」為例:遠處似烟而實為蒸氣、霧氣,究竟為烟為霧不能確定,故因體不決定而犯過。

其二為「所依不成」。因所繫屬的地方不能確定時亦犯此過,如以孔雀鳴為因成立此山間有孔雀,然對鳴處尚未明瞭,因體所依之處不確定,故亦犯不成過。

其三為「有法不成」。有法(即宗上之主詞)本身不成時亦犯此過,如勝論派以「於一切處緣到之功德」為因成立「我於一切處有」,然敵論者如佛家不許有我,則此有法根本不成立,由有法連帶而來之過,故名「有法不成」。

其四則關於異品遍無相。若於同品異品皆遍轉或於任何一品遍轉,則為似因不定過之「共不定」,細分為「同品遍轉,異品分轉」、「同品分轉,異品遍轉」、「俱品一分轉」三類。

第三,三相各具二過。論中明確指出,三相中之每一相皆各有二過,即「不成」與「猶豫」。若對因之相起疑惑,其結果或為不成,或為猶豫不定。

第四,論中亦論及「相違」之義,以明其與不成之關係。相違有二種:一者非未全具因而生之果,有他則自無,如冷觸與熱觸;二者一方排另一方而存之唯實相,如「有」與「無」。又言「相違決定」乃猶豫之因,然以其不成為比量推理之對象,於自性因、果因、未緣到因之性相中既然無相違決定,故不另立敘述。

又論中云:「雖未見到,但凡一切智者皆未言說,此說所遣,亦不成立,蓋生起疑惑故」,此即以疑惑明不成之理,言說與一切智二者不相違,故雖遣除而不能決定成立。

因明入正理論T1630 · 論集部

《因明入正理論》論「不成」,分見於似宗過與似因兩處,內涵層次分明。

一、似宗過中之「不極成」諸相

經文先標似宗過五相——自語相違、能別不極成、所別不極成、俱不極成、相符極成。其中三種與「不成」直接相關:能別不極成(如佛弟子對數論師立「聲滅壞」,聲之滅壞非數論師所許)、所別不極成(如數論師對佛弟子說「我是思」,思非佛弟子所許)、俱不極成(如勝論師對佛弟子立「我以為和合因緣」,和合之義雙方皆不許)。相符極成雖名「極成」,然以立宗反成共許之義而失立宗之用,故經文總結其病曰「不容成故」。此即宗義於論敵不能成立之相。

二、似因中之「不成」四種

經文又標似因三類,謂「不成、不定及與相違」,其中「不成」別開四種:(一) 兩俱不成——立者與敵者雙方皆不許因之成立,如立「聲為無常」以「眼所見性故」為因,眼所見性於聲上雙方俱不許故;(二) 隨一不成——立者或敵者一方不許因之成立,如以「所作性故」對聲顯論立宗,聲顯論不許聲是所作故;(三) 猶豫不成——因之成立猶豫不決,如於霧等大種和合之性起疑惑時為成「大種和合火有」而有所說;(四) 所依不成——因之所依於敵者不極成,如「虛空實有,德所依故」對無空論,虛空既不許有,其所依之因亦隨之不成。

三、本經獨特之表述

本經以具體例證而非抽象定義展示「不成」之相,每一類皆配以立者、敵者、宗、因之具體情景,使讀者得以按例辨明。其核心判準為「不極成」與「不容成」——前者指自他共許之義未能成立,後者指所立之宗不容於敵者而失其義。本經所用術語自成體系:不極成、相符極成、兩俱不成、隨一不成、猶豫不成、所依不成等,皆本經所特立之因明名相,宜據本經原稱記錄歸一。

Related