Dharma · AI Companion

因三相

hetu-trairūpya

因三相指正因所必须具足之三条件:于所立法具备、同品定有、异品遍无。AI-generated

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

A Contemporary ReadingAI-generated

What This Concept Is Talking About

We make inferences every day. Seeing a grey sky, you infer "it's probably going to rain"; seeing your boss frown, you infer "today's meeting might involve getting reprimanded"; seeing a news item reposted by countless influencers, you infer "this matter must be reliable." Most of the time, these inferences seem acceptable, but occasionally they face reality checks—and those moments of failure are rarely because the fact itself was wrong, but because there was a problem in the inferential process: the "reason" you brought forward simply was not qualified to serve as a reason.

Buddhist epistemology—因明學 (hetu-vidyā, "Buddhist logic and epistemology")—spent centuries trying to clarify precisely when a reason qualifies as adequate. The *Nyāyabindu* (《正理滴論》, a work by 法稱 [Dharmakīrti] and the foundational anchor of pramāṇa studies in Tibetan Buddhism) provides the answer: a "reason" (因 [hetu, "reason"]—that which we bring forward as evidence) capable of supporting a correct inference (比量 [anumāna, "inferential cognition"]) must simultaneously satisfy three conditions. These are the three aspects of the reason—因三相 (hetu-trairūpya, "three aspects of the reason").

First, let us lay out the basic structure of an inference. A complete inferential argument is supported by three components: the conclusion one wishes to establish is called the proposition (宗 [pakṣa, "proposition"]), the reason brought forward to support the proposition is called the reason (*hetu*), and there is the example (喻 [dṛṣṭānta, "example"]). The three aspects of the reason specifically audit the "reason"—they do not audit whether the proposition is correct (that belongs to another set of examinations), nor do they audit whether the example is good (the *Nyāyabindu* subsumes the example directly into the body of the reason without establishing separate conditions for it, a point where it differs slightly from Chinese Buddhist logic). They focus solely on that middle component: the "reason."

What Are the Three Aspects?

The First Aspect: Must be present in the subject of inference. The "subject of inference" (所推論比度之事 / 有法 [pakṣa / dharmin, "subject of inference"]) is the specific object regarding which you wish to prove a conclusion. The first aspect requires that the reason you bring forward must actually be present in this object. If you say, "This is a tree, because it possesses the features of an agarwood tree," the first aspect checks: does this object indeed possess the features of an agarwood tree? If yes, it passes. This is the minimum threshold: your reason must at least be present at the scene.

The Second Aspect: Must be necessarily present in similar instances. A "similar instance" (同品 [sapakṣa, "similar instance"]) refers to any instance that conforms to the predicate of the proposition—that is to say, anything belonging to "the class you wish to prove." The second aspect requires that across all these similar instances, this "reason" must be present. "Necessarily" is key—not "some," not "most," but "invariably." The features of an agarwood tree must appear in every tree for "having agarwood tree features" to reliably infer "being a tree."

The Third Aspect: Must be completely absent in dissimilar instances. A "dissimilar instance" (異品 [vipakṣa, "dissimilar instance"]) refers to any instance that does not conform to the predicate of the proposition—that is to say, anything that does not belong to "the class you wish to prove." The third aspect requires that across all these dissimilar instances, this "reason" must be absolutely absent. In everything that is not a tree, agarwood tree features must never occur.

When all three aspects are fully present, your "reason" is logically bound tight to the "proposition": wherever the predicate of the proposition is present, the reason must be present; wherever the predicate is absent, the reason must be absent. Only then is it a qualified "reason," termed a valid reason (正因 [samyak-hetu, "valid reason"]).

Why Must All Three Aspects Be Present?

Not a single aspect can be missing, because each guards against a distinct type of error:

- Only the first aspect is met: The reason happens to be present at the scene, but has no structural connection to the predicate of the proposition; it is mere coincidence. "My neighbor became rich, so the stock market will rise"—the neighbor becoming rich indeed occurred (first aspect passes), but it has no bound connection to the stock market (second and third aspects fail). - Only the first and second aspects are met: The reason appears at the scene and accompanies the predicate of the proposition, but it also accompanies things that are not the predicate—it is not an "exclusive mark" of the predicate, and cannot infer the conclusion we want. "It barks, so it is a dog"—things that bark are not limited to dogs (dissimilar instances also have it); though the second aspect passes, the third aspect fails. - Only the first and third aspects are met: The reason does not appear in non-predicates, but it does not necessarily appear in all instances of the predicate either—it is a rare phenomenon, insufficient to support a universal inference. "One day I saw a unicorn, so unicorns exist"—it is not seen in the non-unicorn world either (third aspect passes), but that one encounter in the first aspect does not mean all unicorns will be seen by people (second aspect fails).

Three Subclassifications of Dissimilar Instances

The *Nyāyabindu* further divides dissimilar instances into three categories: first, dissimilar instances outside the category—those completely outside the scope of "that domain" (for example, if proving that an object has "mammalian" traits, a stone lies outside the category); second, dissimilar instances directly contrary to the predicate—those within the domain, but in the opposite direction of the predicate (if proving an object is a mammal, a reptile falls into this category); third, dissimilar instances merely lacking the predicate—those within the domain, but simply lacking the property of the predicate (if proving an object is a mammal, an amphibian falls into this category).

The common point of these three categories is "lacking the predicate of the proposition," so all fall under the governance of the third aspect. But keeping this subclassification in mind helps us identify that when an inference fails, it is often because we only thought of one type of dissimilar instance and overlooked the other two.

Three Legitimate Reasons (Three Valid Reasons)

The *Nyāyabindu* further points out that reasons permitted by the three aspects take only three forms.

1. Essential Reason (自性因 [svabhāva-hetu, "essential reason"]): Inference based on the item's own nature. Example: "This is a tree, because it is an agarwood tree." You observe the specific agarwood nature of this item (bark, resin, aroma, etc.), and from this infer that it belongs to the broader class of "tree." The core of an essential reason is that the reason itself is a specific instantiation of the predicate—it is not an "external mark" of the predicate, but the manifestation form of the predicate in a specific context. Elsewhere it is also termed a "reason based on specific characteristics" (自相因 [svalakṣaṇa-hetu]).

2. Effect Reason (果因 [kārya-hetu, "effect reason"]): Inference reasoning backward from an effect to a cause. Example: "There is fire over there, because there is smoke." Smoke is the effect of fire; seeing smoke, you infer the presence of fire. The validity of this type of inference relies on a stable causal chain between the reason (smoke) and the predicate (fire). Also translated as "effect-nature reason" (果性因).

3. Reason Based on Non-Apprehension / Non-Obtainment (未緣到因/不可得因 [anupalabdhi-hetu, "reason based on non-apprehension"]): Inferring non-existence from non-perception. Example: "There is no jar here, because if there were one, it should be clearly apprehended given that all cognitive conditions are complete, yet it is not apprehended." I do not see a jar; if a jar were truly here, I should see it when all cognitive conditions are present; but I do not see it; therefore, there is no jar here. The special feature of this reason is that what it establishes is not "presence," but "absence." Reason based on non-apprehension (Tibetan: *mi dmigs pa'i rtags*) is frequently translated as "reason based on non-obtainment" (不可得因) in other translations—both refer to the same thing.

The three types of reasons can be subsumed under two categories: affirmative reasons (立物因 [vidhi-hetu], i.e., essential reasons and effect reasons, inferring existence from existence) and negative reasons (否定因 [pratiṣedha-hetu], i.e., reasons based on non-apprehension, inferring non-existence from non-existence). The former "establishes" the existence of something, while the latter "establishes" its absence. This distinction is crucial—many people habitually think only of affirmative reasons, failing to realize that negation can also be systematically "established."

Three Failure Modes (Three Fallacies of Reason)

The absence of any single aspect causes an inference to collapse. The *Nyāyabindu* summarizes these into three fundamental fallacies of reason (似因過 [hetv-ābhāsa, "fallacies of the reason"]):

- Unestablished (不成 [asiddha, "unestablished reason"]): A problem with the first aspect—the reason fails to hold firm at the scene. For example, using "it is a warm-blooded animal" to prove that this creature is a mammal, when it is actually cold-blooded. The reason is not genuinely present. - Contradictory (相違 [viruddha, "contradictory reason"]): The reason points in the direction opposite to the predicate. For example, wanting to prove "this road is safe," but bringing forward the reason "no one travels this road at night"—which implies danger instead. - Uncertain (不定 [anaikāntika, "uncertain reason"]): The reason appears in both similar and dissimilar instances, failing to differentiate them. For example, using "it can fly" to prove "it is a bird"—bats can fly but are not birds, while ostriches are birds but cannot fly. The third aspect breaks down.

The treatise further points out: each aspect has two faults, namely being "unestablished" (the reason cannot hold at all) or "doubtful" (whether the reason holds is uncertain). Three aspects multiplied by two faults yield six specific errors; but these six, in the final analysis, do not escape the three fundamental categories above.

Governing Both "Thinking Clearly for Oneself" and "Explaining to Others"

The three aspects of the reason are not techniques used only when composing treatises; they run through both inference for oneself (自利比量 [svārthānumāna, "inference for oneself"]) and inference for others (利他比量 [parārthānumāna, "inference for others," i.e., convincing others]). The former generates correct cognition because the reason possesses the three aspects, and its validity in outcome is equivalent to direct perception (現量 [pratyakṣa, "direct perception"]); the latter requires expressing the inference in the format of a "three-aspect reason," divided into two modes of statement: possessing similar properties (具同品法 [sādharmya-vat]) and possessing dissimilar properties (具異品法 [vaidharmya-vat])—focusing respectively on enumerating similar instances and excluding dissimilar instances—yet both contain the complete requirements of the three aspects. Together, these two formats constitute the complete syllogistic form of "pervasion in presence and total exclusion in absence" (隨因遍行遍遣)—enumerating similar instances on one hand and excluding dissimilar instances on the other, leaving the listener no room for refutation.

It is worth mentioning that the *Nyāyabindu* explicitly does not establish the "example" as an independent branch of proof—the function of the example is already contained within the second and third aspects; giving an example is nothing more than providing evidence for necessary presence in similar instances and complete absence in dissimilar instances. This is a clear difference between Tibetan pramāṇa traditions and the Chinese tradition's *Nyāyapraveśa* (《因明入正理論》, authored by 商羯羅主 [Śaṅkarasvāmin], translated by 玄奘 [Xuanzang]). In the Chinese tradition, the example stands independently as one branch of proof (the three-part syllogism: proposition, reason, example—三支作法), whereas the *Nyāyabindu* subsumes it into the body of the reason itself.

In Chinese Buddhist logic, these three aspects are summarized more succinctly: "pervasion as the property of the subject, necessary presence in similar instances, complete absence in dissimilar instances" (遍是宗法性、同品定有性、異品遍無性, from *Nyāyapraveśa*). Both expressions articulate the exact same thing: the first aspect is "pervasion as the property of the subject" (遍是宗法性 [pakṣadharmatva]), while the second and third aspects are "necessary presence in similar instances" (同品定有性 [sapakṣe sattvam]) and "complete absence in dissimilar instances" (異品遍無性 [vipakṣe 'sattvam]). Readers familiar with this Chinese phrasing will find that the *Nyāyabindu* simply unpacks it in detail, offering more systematic classifications such as the three valid reasons and three fallacies of reason.


Walkthroughs in Daily Life

Let us pick a few inferences we make every day and run them through the security check of the three aspects one by one.

Scenario 1: She didn't reply to my message today, so she is mad at me

Deconstruct it: - Proposition (what you want to establish): She has negative emotions toward me. - Reason (the cited evidence): She didn't reply to my message today.

Run through the three aspects:

1. First Aspect—Did she really fail to reply? Open WeChat: unread, unreplied. Passes. 2. Second Aspect—Does everyone who "has negative emotions toward me" fail to "reply to messages"? No. Sometimes angry people send a long text to argue; sometimes a non-reply is simply because they are busy, sleeping, or in a meeting. Fails. 3. Third Aspect—Does everyone who "has no negative emotions toward me" refrain from "failing to reply to messages"? No. Many people don't reply simply because they forgot, didn't see it, or temporarily don't feel like talking. Fails.

Both the second and third aspects fail—this is a classic uncertain reason (*anaikāntika*). The reason appears in both similar and dissimilar instances, failing to help us distinguish between "having negative emotions" and "not having negative emotions."

This does not mean you cannot infer she is angry; rather, the reason you brought forward cannot support this inference. To make the inference stand, you need a sturdier signal—such as her consistent behavioral patterns when angry in the past, changes in her tone when speaking recently, whether her responses to others have also changed, and so on. These signals are much closer to being necessarily present in similar instances and completely absent in dissimilar instances.

The most common mapping error here: The vast majority of people only silently pass the first aspect in their minds—"She indeed didn't reply to me"—and then jump straight to the conclusion. The second and third aspects are the two checkpoints that truly determine whether an inference holds, yet they are the ones we skip most easily in daily life.

Scenario 2: This stock has been rising recently, so it will keep rising

- Proposition: It will continue to rise in the future. - Reason: It has been rising continuously recently.

Run through the three aspects:

1. First Aspect—Has it indeed been rising recently? Passes. 2. Second Aspect—Do all stocks that "will continue to rise in the future" show that they "have been rising continuously recently"? No; many stocks that will rise were previously falling and have just begun to rebound. Fails. 3. Third Aspect—Do all stocks that "will not rise in the future" show that they "have not been rising continuously recently"? No; many stocks that have hit their peak will plummet right after. Fails.

Again, an uncertain reason. Worse still, there are actually numerous real counter-examples here (stocks with the largest past gains often perform worst later—counter-evidence known as momentum reversal in behavioral finance). Thus, this inference is not merely uncertain; it is one of the most classic cognitive traps in the market.

The most common mapping error here: Equating "what happened in the past" with "what will happen in the future"—misreading "temporal correlation" as "logical necessity." The second and third aspects are designed precisely to block this move; they force us to ask "do all instances of this class act this way?" and "do all non-instances of this class refrain from acting this way?" Once asked, most time-series inferences collapse immediately.

Scenario 3: There is no kangaroo in this room because I don't see one

This time, let us use an example of a reason based on non-apprehension (*anupalabdhi-hetu*) to see how it differs from the previous two.

- Proposition: There is no kangaroo in this room. - Reason: Non-apprehension (non-perception) of a kangaroo.

Run through the three aspects:

1. First Aspect—Did I indeed not see/hear/touch a kangaroo? Passes. 2. Second Aspect—In all rooms that "have no kangaroo," is a kangaroo "not apprehended"? This holds true only when cognitive conditions are complete—there is sufficient light, the kangaroo is indeed within the reach of your vision/hearing/touch, you are not distracted, and you are not looking in the dark wearing sunglasses. That is to say, "non-apprehension" can support "absence" only under a clean premise. Passes (when conditions are complete). 3. Third Aspect—In all rooms that "have a kangaroo," will you "apprehend a kangaroo"? Similarly requires complete conditions. If a kangaroo is behind your back, in another room, or if you are asleep—you will not apprehend it. Passes (when conditions are complete).

The premise for this inference to hold is "if it were present, it should be clearly apprehended"—a specific qualification added in the original text of the *Nyāyabindu* to set a prerequisite for the second and third aspects. When conditions are incomplete, non-apprehension cannot infer non-existence.

The most common mapping error here: Equating "I haven't seen it" with "it doesn't exist." Police investigations, medical diagnoses, and investigative journalism all rely on reasons based on non-apprehension; but only when cognitive conditions are complete can it support "absence." In criminal investigations, "no third-party fingerprints found at the scene" must be paired with "if a third party were present, fingerprints should have been left" to infer "no third party"; whereas in the dark of night, or if the scene has been cleaned, or if evidence was deliberately concealed—this inference fails. The moment when reasons based on non-apprehension are most prone to misuse is precisely when conditions are incomplete yet we pretend they are complete.

A Counter-Example: When the same statement passes the three aspects

Lest people mistakenly think that "the three aspects can never be passed," here is a contemporary example that passes:

"It rained today, because there are puddles on the ground, and the time frame is within today." - Proposition: It rained today. - Reason: There are puddles on the ground, and within the time frame of "today."

Run through the three aspects: 1. First Aspect—Are there puddles on the ground today? Observation confirms it, passes. 2. Second Aspect—On all days when "it rained today," are there puddles on the ground today? On conventional road surfaces and under normal weather, generally yes. Passes. 3. Third Aspect—In all cases where "there are puddles on the ground today," is it because "it rained today"? Not necessarily (street sweepers, burst water pipes, floor washing), fails.

Well, this aspect failed too—but the boundary of failure is far narrower than in the previous examples (we can imagine scenarios where "there are puddles on the ground today but it didn't rain," but the categories we can enumerate are few). If we add an auxiliary condition—"and there is no other obvious water source"—the third aspect gets much closer to passing.

This walkthrough demonstrates one thing: The three aspects are not a black-and-white binary switch, but exist on a spectrum of degree. Most of our daily inferences fail slightly at some point among the three aspects; the more important the matter (medical treatment, law, investment, life choices), the more worthwhile it is to spend effort narrowing that failure down to a point where it does not affect the conclusion. When that cannot be done, one ought to admit "the reason I brought forward is not strong enough," rather than shouting a weak reason louder to make it sound strong.


Why Contemporary People Need It

The three aspects of the reason are not logic tricks stored in a museum. They serve four concrete purposes in our present condition.

1. In the information age, we live in "second-hand judgments," where almost all cognitions must be screened

In the information we consume daily, the portion seen with our own eyes (direct perception) is shrinking; the vast majority of conclusions come from "hearing others say," "reading data," or "reading AI summaries." This means we are actually making "inferences" every day—accepting conclusions based on reasons handed to us by others (or algorithms). The three aspects provide a set of tools to at least judge whether the "reasons" handed over by others (or ourselves) can hold up. It cannot guarantee that a conclusion is true, but it guarantees that we won't be fooled by the crudest reasoning.

2. It is the minimum defense line against cognitive biases

Behavioral economics speaks of dozens of biases—anchoring, availability, confirmation bias, survivorship bias—and their shared structure is: the first aspect passes, while the second or third aspect quietly fails. You see your neighbor make money in the stock market (first aspect: made money), so you infer "this stock will rise" (second aspect fails: cases of making money do not equal all cases that rise; third aspect fails: far more people may have lost money). The three aspects offer a universal metacognitive framework, allowing you to take one extra step when biases occur.

3. It shares commonalities with the operation of AI / algorithms—but must not be equated with them

Contemporary machine learning classifiers work precisely by "finding features that are necessarily present in similar instances and completely absent in dissimilar instances"—the training set provides the similar and dissimilar instances, and the selected features are the "reason." When a model fails on certain populations, it is often because the feature it found failed the third aspect on those populations (the feature appeared in groups where it shouldn't, or didn't appear where it should). In this sense, the spirit of the three aspects aligns with algorithms.

However, the fundamental difference between the two must be pointed out. The goal of an algorithm is maximizing predictive accuracy; its "similar instances" and "dissimilar instances" are samples defined by a dataset, which may contain sampling bias, label noise, and hidden covariates. The three aspects care about whether a reason genuinely supports a conclusion; its similar and dissimilar instances are logically defined as "all instances possessing / lacking the predicate"—a universal proposition that is in principle infinite. The former may use a feature that is statistically strongly correlated but logically fragile (correlation ≠ causation), whereas the latter demands structural necessity. AI has no concept of a "reason," only "weights"; the two cannot be equated. Neuroscience is similar—the brain's fast thinking (Kahneman's System 1) almost never performs three-aspect checks; only slow thinking (System 2) can do so. The three aspects are a method designed for slow thinking; its function is not to replace intuition, but to set up an additional checkpoint beyond intuition.

4. Its position on the path to liberation: Laying the groundwork for "right view"

From the perspective of the Buddhist path to liberation, the three aspects of the reason belong to the branch of epistemology (pramāṇa)—it is not meditation, nor is it precept-keeping, but the study of "whether our cognition itself is reliable." Its position lies in the "thinking" stage of hearing, contemplating, and cultivating (聞思修 [śruta-cintā-bhāvanā]): acquiring information through hearing the teachings (hearing), organizing information into right view (正見 [samyag-dṛṣṭi, "right view"]) through inferences that conform to the three aspects (contemplating), and only then taking right view as the object to engage in practice that observes reality as it is (cultivating).

If "right view" is built on inferences that cannot stand—such as taking "I feel there is an unchanging self" as a fact needing no proof, or building "samsara is suffering" on a flawed analogy—then no matter how hard one works during contemplative practice, the object to be refuted is not real, and the state to be established will be hard to stabilize. The view of emptiness (空性 [śūnyatā, "emptiness"]) requires rigorous inference as a precursor; otherwise, the "emptiness" perceived may merely be conceptual emptiness rather than ultimate emptiness. The three aspects are not meant to turn everyone into a dialectician, but to ensure that at crucial cognitive junctures, we do not support the heaviest conclusions with the flimsiest logic.

Hence, epistemology is positioned as a "preliminary foundation for the path" in the Tibetan Buddhist curriculum—before studying Madhyamaka (中觀) or Prajñā (般若), one must first sharpen one's cognitive tools. In a sense, its contribution to the path of liberation is implicit: you will not achieve enlightenment simply by learning the three aspects of the reason, but without it, the "path to enlightenment" you study may be laced with logical loopholes that shouldn't be there.


Common Misinterpretations and Clarifications

Misinterpretation 1: Regarding the three aspects as "formal logic"

The three aspects of the reason are argumentation within Buddhism, possessing its own goals (serving the path to liberation and right view) and its own categories (it deals with universal statements like "wherever there is..." and "wherever there is no...", but is not symbolic calculus). It is a distant relative to Aristotelian syllogism, Boolean algebra, or modern symbolic logic—not the same thing. Treating the three aspects as an "Indian version of formal logic" loses its liberation-path significance—it is not an intellectual game, but a foundation laid for seeing clearly "what is real."

Misinterpretation 2: Understanding "similar instances" as "similar examples"

A "similar instance" is not an instance that "looks similar," but any instance that possesses the predicate of the proposition. For example, to prove "this object is flammable," similar instances are all flammable things (paper, wood, alcohol...), not all things that "look like this object" (such as all white objects). Once similar instances are equated with "surface resemblance," the second aspect immediately becomes unverifiable—because "resemblance" is an open concept that can refer to similar color, similar shape, or similar material, any of which would collapse the inference.

Misinterpretation 3: Understanding "necessary presence" as "having some"

"Necessary presence" means invariability—every single similar instance must have it; it is not "most," nor is it "several." Once "having some" is permitted, the second aspect collapses; as long as a single counter-example is found (a similar instance that lacks the reason), the inference fails. When we casually say "I've seen it happen this way a few times," we are precisely mistaking "several observations" for "necessity," failing even the first threshold of the three aspects.

Misinterpretation 4: Understanding "complete absence" as "I haven't thought of a counter-example"

"Complete absence" is exhaustive: every single dissimilar instance must lack it. If even a single dissimilar instance possesses the reason, the inference fails. When we casually say "I haven't heard of any counter-examples," this is empirical non-observation, not logical impossibility. Equating "I haven't seen it" with "it doesn't exist" is the most common abuse of reasons based on non-apprehension—and the *Nyāyabindu*'s special emphasis on "if it were present, it should be clearly apprehended" as a prerequisite is designed precisely to block this abuse.

Misinterpretation 5: Taking "seems right" or "many people say so" as evidence for the three aspects

This is mistaking consensus or intuition for evidence of the three aspects. What the three aspects demand is structural binding—all instances possessing the predicate have the reason, and all instances lacking the predicate lack the reason—not that everyone agrees or feels it to be so. Otherwise, geocentrism also "seemed right" in history and was believed by almost everyone; medieval European medical consensus also "seemed right." Substituting "many people say so" into the place of the three aspects is the most pervasive and hardest-to-spot fallacy of substitution in argumentation.

Misinterpretation 6: Believing that passing the three aspects equals "absolute correctness"

Passing the three aspects only shows that this reason logically supports this conclusion—it does not guarantee that the conclusion is true. If our understanding of a basic fact about the world is wrong (for instance, if we mistakenly believed all smoke comes from fire), an inference can pass the three aspects formally while its conclusion remains wrong. The three aspects are a legitimacy test for inference, not a guarantee of truth.

Misinterpretation 7: Believing the three aspects are only useful for "logical debate" and unnecessary in daily speech

The true utility of the three aspects lies precisely in daily life. The more significant a judgment (whom to choose, what job to take, what news to believe, what investment to make, judging whether oneself has fallen into a pattern), the more necessary it is to run through the three aspects first. Daily conversation certainly doesn't require technical jargon at every turn, but running a quiet check in one's mind is the simplest way to prevent major errors.

Misinterpretation 8: Thinking the three aspects only work for affirmative inferences, not negative ones

One of the special features of the *Nyāyabindu* is its explicit handling of negation—reasons based on non-apprehension establish "absence" from "non-perception." Thus, the three aspects are not merely tools for "establishing presence from presence," but also include the legitimacy of "establishing absence from non-perception." In daily life, we frequently need to prove "this thing doesn't exist," "this person doesn't have that intention," "this news is unreliable," or "this therapy is ineffective"—these are contemporary versions of reasons based on non-apprehension. The three aspects apply equally, and often require even greater care than affirmative inferences (because negation slides more easily into emotional denial).

Misinterpretation 9: Regarding the "example" as dispensable rhetoric

In Chinese Buddhist logic, the example stands independently as one branch of proof; in the *Nyāyabindu*, the example is subsumed into the reason itself. Regardless of the treatment, the example is never decoration—it is the bridge that grounds the abstract second and third aspects onto concrete instances. If a three-aspect inference lacks an example (a concrete instance), the listener has no way to verify necessary presence in similar instances and complete absence in dissimilar instances for themselves; at that point, the inferrer is actually demanding blind faith from the other party. The true power of the three aspects can only be fully revealed when paired with concrete examples.

Misinterpretation 10: Believing that "having learned the three aspects" means "one's own reasoning will automatically be rigorous"

This is the same as "having learned grammar doesn't mean one won't make writing errors." The three aspects of the reason are a tool for post-hoc audit—you can use it to test an inference that has already been made, but making it run automatically at the very moment an inference forms requires long-term training. Its value lies not in making every inference we make automatically perfect, but in enabling us to identify where an error lies when we make one—Did it fail the first aspect? The second aspect? The third aspect? Once the location of the error is identified, rectification becomes possible; if unidentified, the error will be repackaged as "I have ample reason" and recur repeatedly.

Canonical EntryAI-generated

因三相(hetu-trairūpya)为藏传论典《正理滴论》所立正因之根本定义。

一、定义与核心要义 因三相指能够成立正确比量之因所必须具足之三个条件。此三相贯通于自利比量与利他比量二者之中,为一切正确比量成立之根本所依。

二、主要阐述 三相之具体内涵为:第一「于所推论比度之事上必须具备」,第二「于同品上定有」,第三「于异品上遍无」。所比度之事即欲了知之各别有法;同品乃概与所成立之法相符之事;异品则非同品,复分为「彼外之异品」、「与彼相违之异品」、「彼上没有之异品」三种。三相之中若缺一相未言及,即名似成立;三相已言而立论者或敌论者不极成或有疑义,亦名似能立。每一相各有二过,谓不成或犹豫;合计三种似因过,谓不成、相违、不定。

论主明确指出:「三相所在之处,唯有三种因,谓未缘到因、自性因、果因」。未缘到因如「于各别地方无瓶,若有则当为缘虑之所明了,但未缘到故」,凡有自性之处,若缘虑之其他条件具备,必然成为现实;自性因如「此是树木,是沉香树故」;果因如「彼处有火,以有烟故」。此三种因概括为立物因与否定因二类。其中自性因于别处亦作自相因之名。

三相通于自利比量与利他比量。自利比量由具三相之因生诸认识,于量果之理与现量相同。利他比量则「说为三相因者,谓于因安立果故」,格式分为具同品法与具异品法二种,皆具此二义无有差别。具同品法者如「凡缘虑之所明了物,若未缘到,即为无之名言安立之处」;具异品法者如「若是青色等支分……为瓶,以未如实缘到瓶」。两种格式可联合为「随因遍行遍遣」之量式,以同品法或异品法与三种正因分别结合,构成各种不同之推理论证。

三、不同部类与译本的表述差异 本论以「未缘到因」标示藏文 madmigs pavi rtags,与他译之「不可得因」同指凡为缘虑可明了而未缘到者即是。「自性因」、「果因」亦可见「自相因」、「果性因」之异名。

四、与相关概念的关系 本论不另立相违决定为似因过,以其在果因、自性因、未缘到因之性相中本无,于他处更无相违决定可言。又喻摄入能立之中,谓「喻者虽为能立之支分,但离之别无有体,故未另立其相」。所成立之义与因之相系属分为自性及从生二种,唯有自性因与果因成立事物;未缘到因则就否定而立,谓物若有则不致未缘到彼。因与所立法之关系称「相系属」,而立论者意乐、被现量等四种排除所遣,则论及宗之性相。

五、代表性原文引用 「于所推论比度之事上必须具备,于同品上定有,于异品上遍无」。「三相所在之处,唯有三种因,谓未缘到因、自性因、果因」。「说为三相因者,谓于因安立果故」。「喻者虽为能立之支分,但离之别无有体,故未另立其相」。

By SutraAI-generated

正理滴論(譯自西藏文本)B0042 · 藏傳論典部

本論以「三相」為正因之根本定義,貫通於自利比量與利他比量二者之中,並以此統攝一切正確比量之成立。

一、三相之定義與所立 論主明示具三相之因須具三個條件:「於所推論比度之事上必須具備,於同品上定有,於異品上遍無」。所比度之事即欲了知之各別有法;同品乃概與所成立之法相符之事;異品則非同品,復分為「彼外之異品」、「與彼相違之異品」、「彼上沒有之異品」三種。三相之中若缺一相未言及,即名似成立;三相已言而立論者或敵論者不極成或有疑義,亦名似能立。每一相各有二過,謂不成或猶豫;合計三種似因過,謂不成、相違、不定。

二、三相所在之因唯有三種 論主明確指出:「三相所在之處,唯有三種因,謂未緣到因、自性因、果因」。未緣到因如「於各別地方無瓶,若有則當為緣慮之所明瞭,但未緣到故」,凡有自性之處,若緣慮之其他條件具備,必然成為現實;自性因如「此是樹木,是沉香樹故」;果因如「彼處有火,以有烟故」。此三種因概括為立物因與否定因二類。其中自性因於別處亦作自相因之名。

三、自利與利他比量共用三相 自利比量由具三相之因生諸認識,於量果之理與現量相同。利他比量則「說為三相因者,謂於因安立果故」,格式分為具同品法與具異品法二種,皆具此二義無有差別。具同品法者如「凡緣慮之所明瞭物,若未緣到,即為無之名言安立之處」;具異品法者如「若是青色等支分……為瓶,以未如實緣到瓶」。兩種格式可聯合為「隨因遍行遍遣」之量式,以同品法或異品法與三種正因分別結合,構成各種不同之推理論證。

四、獨特之取捨 論主不另立相違決定為似因過,以其在果因、自性因、未緣到因之性相中本無,於他處更無相違決定可言。又將喻攝入能立之中,謂「喻者雖為能立之支分,但離之別無有體,故未另立其相」。所成立之義與因之相系屬分為自性及從生二種,唯有自性因與果因成立事物;未緣到因則就否定而立,謂物若有則不致未緣到彼。

五、術語特點 本論以「未緣到因」標示藏文 madmigs pavi rtags,與他譯之「不可得因」同指凡為緣慮可明瞭而未緣到者即是無。「自性因」、「果因」亦可見「自相因」、「果性因」之名。因與所立法之關係稱「相系屬」,而立論者意樂、被現量等四種排除所遣,則論及宗之性相。

Related