What This Concept Is Saying
Imagine you are about to cross a river. Someone points to a bridge and says, "Cross here." You step on it and the planks give way — not because the bridge was poorly painted, not because it lacked the *form* of a bridge, but because, despite every appearance, it could not do what a bridge is supposed to do. In Buddhist logic (因明, *pramāṇa* — the classical Indian-Buddhist discipline of valid cognition and reasoning), this kind of failure has a precise name: 似因 (hetvābhāsa), literally "the appearance of a reason." A *hetvābhāsa* is a "reason" (因, *hetu*) that has the shape and surface of an argument but cannot, in fact, establish the thesis (宗, *dharmin*) it claims to support. It looks like reasoning; it functions as something else.
To see why this matters, we have to step back into the architecture of a Buddhist syllogism. In the tradition most clearly codified by Dignāga and Dharmakīrti and preserved in texts like Śaṅkarasvāmin's《因明入正理論》(*Nyāyapraveśa*, "Introduction to Valid Reasoning") and the Tibetan-transmitted《正理滴論》(*Pramāṇavārttikā*-derived drops), a valid inference has three limbs: 宗 (the thesis, "Sound is impermanent"), 因 (the reason, "because it is produced"), and 喻 (the example, "as in a pot, which is produced and impermanent"). The 因 must satisfy what is called the 三相 (*trairūpya*, the "three characteristics"):
1. 遍是宗法性 — The reason must be established as a property of the subject of the thesis (the "subject-property" character). 2. 同品定有性 — The reason must be present in at least some things similar to the predicate (the "similar-positive" character). 3. 異品遍無性 — The reason must be entirely absent from things dissimilar to the predicate (the "dissimilar-negative" character).
A *real* reason holds all three. A 似因 is precisely a reason that fails one, two, or all three of them. The《因明入正理論》sums it up in one line: 「不成、不定及與相違，是名似因」 — "Unestablished, indeterminate, and contradictory: these are called the pseudo-reason." And the text warns that 「以此為因所生之智不能正解」 — the cognition produced by such a "reason" cannot correctly understand its object.
The seriousness of this is captured in the very name. "似" means *appears to be*. The 因明 tradition is not interested in polishing; it is interested in whether the cognitive work gets done. A 似因 is therefore not a stylistic flaw or a minor irregularity. It is a failure of *function* dressed up in the costume of *form*. And the tradition is meticulous about naming exactly *how* it fails, because the type of failure matters for how one should repair the argument.
The Three Master Failures
The classical sources divide 似因 into three umbrella categories, each with sub-types that have stayed largely stable across the Chinese and Tibetan transmissions.
一、不成因 (Asiddha) — "The Reason That Is Not Established"
The first characteristic fails: the reason cannot actually be shown to belong to the subject of the thesis. The classical texts list four to five sub-types:
- 兩俱不成 (ubhaya-asiddha) — Neither the proponent nor the opponent accepts that the reason holds of the subject. Canonical example: trying to prove "sound is impermanent" by the reason "because it is *visible-to-the-eye* (眼所見性)." Sound is not visible. Both sides reject this, so the reason never even gets off the ground. - 隨一不成 (ekatar-asiddha) — Only one of the two parties accepts it. For example, arguing "sound is impermanent because it is *produced* (所作性)" to a *Sabda-darshana* (聲顯論) opponent who holds sound to be eternal. They do not accept that sound is "produced," so for them the reason is unestablished even if you personally accept it. - 猶豫不成 (saṃśaya-asiddha) — The factual basis of the reason is itself in doubt. The classical case is being uncertain about whether a particular fire is the result of "great-element conjunction" (大種和合) and yet using that fact as a reason. If the foundation is shaky, the building cannot stand. - 所依不成 (āśraya-asiddha) — The subject that the reason is supposed to attach to is not accepted by the opponent. Example: arguing to a "denier of space" that space exists, using the reason "because it is a basis of qualities (德所依故)." If the opponent does not even accept that *space* is a real subject, your reason has nothing to attach to. - 有法不成 (dharmin-asiddha) — Added in the《正理滴論》tradition: the subject itself is not established. Argue "the self is pervasive because it has pervasive qualities (遍在功德)" — if the very existence of the self is in dispute, the subject (有法) doesn't get established, and no reason can save the argument.
The unifying logic is simple: before a reason can do anything, it must first *be* — that is, it must be established as genuinely belonging to whatever subject it is supposed to be illuminating. When that fails, nothing downstream can succeed.
二、不定因 (Anaikāntika) — "The Reason That Does Not Decide"
The first characteristic may be fine — the reason is genuinely present in the subject — but the second or third fails: the reason does not reliably track the predicate. It is present in *both* things that share the predicate and things that do not, or only in some on one side and not the other. It cannot decide.
Six classical sub-types:
- 共 (sādhāraṇa) — The reason is common to both the similar and dissimilar classes. Example: arguing "sound is eternal because it is *perceivable/measurable* (所量性)." Perceivability applies to eternal things and impermanent things alike, so it proves neither. - 不共 (asādhāraṇa) — The reason appears in *neither* the similar nor the dissimilar class — it is unique to the subject in such a way that no example can be drawn. Argue "sound is eternal because it is *audible* (所聞性)" — audibility is special to sound, so one cannot construct a same-side or opposite-side example to anchor it. - 同品一分轉異品遍轉 — Present in some similar things, entirely present in dissimilar things. (And its mirror image, 異品一分轉同品遍轉.) These arise when the distribution of the reason across similar/dissimilar cases is uneven in ways that allow a counter-argument. - 俱品一分轉 — Present in part of the similar class and part of the dissimilar class, but never fully covering either. - 相違決定 (viruddhāvyabhicāra) — Two opposing arguments are both fully valid and mutually exclusive, leaving the matter undecidable. One side argues sound is impermanent because it is produced; another argues sound is eternal because it is audible. Each is internally coherent; they just point opposite ways. The texts call this also "the reason of doubt," because the mind cannot settle.
What unifies these is the failure of 異品遍無性: the reason does not reliably *exclude* the opposite of the predicate. If the reason can also be found among things that lack the predicate, it loses its discriminative force. It becomes a *wandering* reason rather than a *settling* one.
三、相違因 (Viruddha) — "The Reason That Proves the Opposite"
This is the most dramatic failure. Not only does the reason fail to establish the thesis; it actively *undermines* it by establishing the opposite. The classical text gives four sub-types:
- 法自相相違因 — The reason is found only in things that are *dissimilar* to the predicate. Argue "sound is eternal because it is produced (所作性)." But producedness exists only in impermanent things, so the reason actually proves sound is impermanent, the very opposite of what was claimed. - 法差別相違因 — The reason establishes something that contradicts not the predicate itself but its differentiating properties or connotations. - 有法自相相違因 — The reason, while attempting to negate the predicate's real-character, quality, or action, ends up also negating the very subject of the thesis. - 有法差別相違因 — The reason establishes something that contradicts the differentiating properties of the subject.
The pattern across these is captured in the formula 「亦能成立」相違內容 ("it also establishes the contrary content") 「俱決定故」 ("and both are determined") — meaning the reason doesn't merely fail; it *succeeds* in establishing something that is decisively contrary to the thesis.
Where 似因 Sits in the Larger Architecture
似因 does not stand alone. It belongs to a family of "faults of what is meant to establish" (能立過). Three limbs can fail: the 宗 (thesis) can fail — that's 似立宗 (paksha-ābhāsa), the thesis that only appears to be a thesis. The 因 (reason) can fail — that's our 似因. The 喻 (example) can fail — that's 似喻 (dṛṣṭānta-ābhāsa), the example that misleads. And there is a related category of failure at the level of refutation: someone who correctly spots a real 似因 is performing 能破 (duṣaṇa, "true refutation"), but someone who wrongly declares a sound argument to be a 似因 commits 似能破 (duṣaṇābhāsa, "pseudo-refutation"). The《因明入正理論》makes this explicit: to correctly show that a reason is unestablished, indeterminate, or contradictory is to genuinely refute; to claim these failures when they are not present is itself a pseudo-refutation.
This is an important point that the classical texts keep returning to. The categories are not arbitrary labels for "bad arguments." They are diagnostic instruments. And like any diagnostic instrument, they too can be misapplied — which is why the tradition also has a name for misapplied critique.
A Walk-Through in Contemporary Life
Let us take the classical machinery and walk it through a few ordinary scenes — a workplace decision, an online argument, a moment of self-justification, a conversation with an AI. We'll go slowly, mapping each classical sub-type onto something recognizable, and pausing at the points where contemporary minds most often map the categories *wrong*.
Scene One: The Product Meeting
A team is arguing about whether to delay a software launch. Maya says, "We should delay, because users are complaining." Kenji pushes back: "Users always complain — that doesn't prove anything."
Translate Maya's argument into 因明 form. 宗 (thesis): "This launch should be delayed." 因 (reason): "Because users are complaining." 同品: examples of launches that *should* be delayed — they share the predicate "should be delayed." 異品: examples of launches that *should not* be delayed.
Now ask the three-characteristic test. Is the reason established on the subject? Yes — there genuinely are user complaints, both parties know this. So 遍是宗法性 holds. But then: are user complaints present in things similar to "should be delayed"? Yes, plausibly. Are they *entirely absent* from things dissimilar to "should be delayed"? No. Many launches proceed on schedule despite complaints, and some launches are delayed for reasons that have nothing to do with user sentiment (regulatory issues, supply chain, internal bug). The reason does not satisfy 異品遍無性. It also does not clearly satisfy 同品定有性 as an *exclusive* marker — it might be a partial marker only.
This is, structurally, a 共不定因 (sādhāraṇa-anaināntika) — a "common indeterminate reason," or in plain language, "a reason that everyone agrees is true but that doesn't actually decide the issue." It wanders between the two sides.
What does this tell Maya? Not that she is wrong to worry about user complaints, but that her *argument as stated* doesn't bear the load she needs it to bear. To fix the argument, she would have to do work the original syllogism skipped — show that this class of complaint correlates with launches that fail, distinguish the type of complaint from other types, demonstrate why this specific case sits on the "should be delayed" side. The 因明 tradition would say her 因 *appears* to be a reason (it sounds reasonable, both parties acknowledge the underlying fact) but in its current form cannot establish the 宗.
*Where contemporary minds go wrong:* We often confuse "the premise is true" with "the argument is sound." Maya's premise (users are complaining) is true. But truth of the premise is not the same as the premise's doing its job in the inference. 因明 insists on this distinction with great care. A 因 that fails its second or third characteristic cannot be rescued by the truth of its content; the failure is structural, not factual.
Scene Two: The Late-Night Argument
Two friends are arguing about whether a public figure is corrupt. One says, "They are corrupt, because everyone says so." The other says, "No, they aren't, because smart people don't engage in that kind of thing."
Translate. 宗 (Proponent): "X is corrupt." 因: "Because everyone says so." 宗 (Opponent): "X is not corrupt." 因: "Because smart people don't do that."
Each side has a reason; each reason *appears* to support its thesis. But test the opponent's claim. "Smart people don't do that" — is this established on the subject? That depends on whether the opponent accepts that X is in fact smart. If the opponent's own view includes "X is smart," then yes, the reason attaches. But does it satisfy 異品遍無性? Are there no smart corrupt people in history? Many, in fact. So the reason is present in things dissimilar to "is not corrupt" — meaning it cannot decide.
More interesting still: examine the proponent's reason. "Because everyone says so." This is, structurally, very close to a classical 不定因. "Everyone saying so" is a reason that is *common* — it could equally be cited to support the opposite thesis ("X is innocent, because everyone says so," if public sentiment shifted). It is also suspiciously close to 猶豫不成 (saṃśaya-asiddha) — the factual basis (what people actually think, who counts as "everyone") is itself unsettled.
Neither side, in fact, has produced a viable 因. Both are using reasons that look like reasons — they have grammatical form, they cite something in the world, they are the kinds of things one says when defending a position — but neither can carry the load of establishing the thesis. Both are 似因 of one variety or another. The argument is not "X is corrupt / X is not corrupt." It is "we have not yet produced a real reason either way."
*Where contemporary minds go wrong:* We tend to treat the existence of an opposing argument as evidence for the *uncertainty* of the matter. 因明 actually points elsewhere: the existence of two opposing arguments that are *both* internally plausible is itself a diagnostic category — 相違決定, the "counter-determined" case. The texts say this is also called "the reason of doubt" because the mind cannot settle between them. The classical diagnosis is not "we need more opinions"; it is "both reasons are failing to decide, and the failure pattern is recognizable."
Scene Three: Self-Justification at 2 AM
Someone has just done something they regret. In the aftermath, the mind generates a reason: "I did it because I had no other choice." They say this to themselves. They might even believe it.
Translate. 宗: "I had no other choice." 因: "Because the situation was unbearable." Test the three characteristics. Is the reason established on the subject? The situation *was* difficult — this passes. But is "situation was unbearable" present in cases where people *did* have other choices? Almost certainly yes — many people face unbearable situations and still perceive and exercise alternatives. Is it absent in cases where people genuinely had no choice? Often no — some bearable situations still close off options. The reason fails 異品遍無性.
This is the quietest and most dangerous form of 似因 in everyday life. The mind is highly skilled at producing reasons that *feel* like reasons — they are syntactically complete, emotionally convincing, often containing genuine facts — but that cannot do the cognitive work they are being asked to do. The self-justifying reason of a 2 AM regret is often structurally a 共不定因: it is true that the situation was hard, but the truth of that fact does not establish the thesis of "no other choice."
The classical texts treat this kind of failure with no special gentleness. They do not say "but you were suffering, so it is understandable." They say: *this reason cannot establish this thesis.* Whether you *were* suffering is a separate question from whether your reason works. The diagnosis is structural, not biographical.
*Where contemporary minds go wrong:* The single most common error is to take a 似因's *plausibility* as evidence of its validity. A reason can be plausible — even emotionally true, even touching on real facts — and still be a 似因. 因明's whole project is to separate the *function* of a reason from its *appearance*. The 2 AM self-justification knows this in the abstract; it just isn't interested.
Scene Four: The AI Assistant
You ask a large language model a difficult question. It produces a confident, well-structured answer with supporting evidence. You push back on one of its claims. It produces another confident, well-structured answer with supporting evidence. You push back again. It changes its answer. The supporting evidence in the third answer is, on inspection, fabricated.
What has happened is not a debate. It is the AI producing, in sequence, three different 因-arguments for three different 宗-theses, none of which were grounded in any genuine 遍是宗法性. The "reasons" look like reasons — they have shape, citation, paragraph structure — but they are, by the classical taxonomy, a cascade of 不成因 (asiddha): reasons not established on the subject. Many would also fall into 猶豫不成 — the underlying factual matter was never settled, because the model was generating plausible-sounding rather than verified statements.
The diagnostic power of 似因 here is striking. It tells us, in a vocabulary developed over a thousand years ago, exactly what has gone wrong: the "reason" is not established; the cognition it produces cannot correctly understand its object (「以此為因所生之智不能正解」). One need not use the Sanskrit term to recognize this; one only needs to hold the classical categories steady while looking.
*Where contemporary minds go wrong:* It is tempting to treat any AI failure as a hallucination and leave it at that. 因明's taxonomy is finer-grained. There are reasons that fail because they were never grounded (不成因), reasons that fail because they don't decide (不定因), and reasons that fail because they actively prove the opposite of what was claimed (相違因). An AI that confidently asserts "X is true because Y" when Y has nothing to do with X is, structurally, a 相違因. These are different kinds of failure, with different implications for how one should correct them.
Why Contemporary People Need This
We live inside the most argument-saturated environment human beings have ever inhabited. Every scroll, every feed, every headline, every group chat, every workplace memo, every wellness article is, structurally, an attempt to present a 宗 supported by a 因 (or sometimes only a宗 without any real 因 at all). The sheer volume of this makes the diagnostic skill that 因明 offers not academic but *survival-relevant*.
Three pressures make this especially acute in the present.
First, the appearance of reasons has become industrialized. A century ago, a bad reason usually looked rough. Now bad reasons are professionally dressed. They are formatted, cited, footnoted, presented in confident paragraph form. The cosmetic apparatus of reasoning has become nearly indistinguishable from the function. 因明's insistence that we ask about *function* — does this reason actually establish what it claims to establish? — is exactly the kind of question that this environment makes it easiest to forget.
Second, our own minds are prolific producers of 似因. The 2 AM self-justification, the post-purchase rationalization, the resentment-narrative that selects only confirming evidence — these are not personal failures of character. They are predictable outputs of a cognition that is very good at generating *plausible-looking* reasons and not very reliable at evaluating whether they actually decide the matter. 因明 offers a vocabulary for catching oneself in the act. When you notice yourself producing a "reason" that cannot be tested against counter-cases, you are noticing a 共不定因 in your own stream of thought. That noticing is itself a form of vipassanā — insight into the structure of one's own cognitive activity.
Third, the categories are precision tools. Modern discussions of "logical fallacies" tend to lump things together. 因明 is unusual in its insistence on diagnostic specificity: not "this is a fallacy" but "this is a fallacy of *this kind*, and here is the precise structural reason it fails." The difference matters. A 不成因 needs different repair than a 不定因, which needs different repair than a 相违因. A reason that fails because it is not established (不成) needs a totally different response than a reason that fails because it is too common (共不定) or because it actually proves the opposite (相违). Treating them all as "a bad argument" leaves one unable to know what kind of repair to attempt.
There is also something gently subversive in the tradition's insistence that even *good faith* arguments can be 似因. The diagnosis is not an accusation of bad intent. It is a recognition that the cognitive apparatus — yours, mine, everyone's — produces outputs that look like reasons more reliably than it produces reasons that actually work. 因明 is the technology for noticing the gap.
Common Misreadings and Clarifications
Misreading 1: "似因 means a fallacy in general." *Clarification:* 似因 is specifically a reason (因) that fails its three-characteristic test within a structured inference. It is narrower than "fallacy" in general — it concerns only the reason-limb of an argument — and it is also more specific than "fallacy," because it requires diagnosing *which* of the three characteristics failed. An ad hominem argument might involve a 似因, but it would be a 似因 only insofar as the reason offered fails its structural role; the ad hominem character is a separate diagnostic question.
Misreading 2: "If the conclusion happens to be true, the reason cannot be a 似因." *Clarification:* 因明's judgment is structural, not factual. A reason can be a 似因 even when its content is true, and even when the conclusion it is attached to is true. The question is not "is the conclusion correct?" but "does the reason, by virtue of its three characteristics, establish the conclusion?" A surgeon might perform the correct operation for the wrong reasons; the operation is not retroactively justified. 因明 makes a parallel point about inference. The tradition treats this with some seriousness — the whole point is to disentangle truth from valid establishment.
Misreading 3: "似因 is just an ancient Indian thing." *Clarification:* 因明 is a Buddhist adaptation of broader Indian logical traditions (notably Nyāya), developed most rigorously by Dignāga and Dharmakīrti. It was transmitted into Chinese Buddhism as a discipline for philosophical debate, and independently developed in Tibet as part of the scholastic curriculum. Its survival in multiple cultures, its continued study in monastic universities for over a millennium, and its structural clarity all suggest it is more than a historical curiosity. Contemporary philosophy of argument, while using different vocabulary, often re-discovers distinctions that 因明 had already named. The framework is portable.
Misreading 4: "If I just don't use these fallacies, I'm safe." *Clarification:* The tradition's deeper claim is that the cognitive system *produces* 似因 as a matter of course. Mindfulness (smṛti, 念) and clear examination (prajñā, 慧) are required to catch them. This is why 似因 sits within Buddhist practice, not outside it. The category is not a rule-book for avoiding mistakes but a training tool for noticing the structure of one's own reasoning, including the moments when one's own mind supplies a beautifully shaped reason that, on inspection, cannot carry the weight it has been asked to bear.
Misreading 5: "相违因 is just 'proving the opposite.'" *Clarification:* 相违因 has a specific structural signature: the reason is found *only* in things dissimilar to the predicate, so that the very act of citing the reason establishes the contrary. The four sub-types — 法自相, 法差別, 有法自相, 有法差別 — distinguish whether the contradiction is at the level of the predicate's surface character, the predicate's differentiating properties, the subject's surface character, or the subject's differentiating properties. Each has a distinct repair path. To collapse them into "proving the opposite" is to lose the diagnostic power.
Misreading 6: "If two people argue opposite things, at least one is using a 似因." *Clarification:* Not necessarily. Two people could each be using a valid reason that establishes opposite conclusions, in which case the texts classify this as 相違決定 — itself a category of 不定因, not a generic accusation of fallacy. The point is that the matter is genuinely *undecided* by the available reasons, not that one side is making a structural error. Sometimes the world is such that reason cannot decide it. Recognizing this is part of what the tradition offers.
A final thought. The classical tradition's most poignant line on this topic is the closing remark of the《因明入正理論》: 「以此為因所生之智不能正解」 — the cognition produced by such a "reason" cannot correctly understand its object. It does not say "the person using the 似因 is wrong." It says "the *cognition* is wrong." The diagnostic is gentler and more precise than moral accusation. It points to a structure in the world — a reason that doesn't carry weight — and invites us to notice it, including in our own thinking, with the same equanimity we might bring to noticing a cracked beam in a building. The beam is not evil. The diagnosis is not a judgment. It is simply the recognition that what was asked to bear weight cannot bear it, and that knowing this is the beginning of repair.
This is why 似因, despite being a treatise topic about error, sits within a tradition whose larger aim is the end of error. The meticulous naming of how reasons fail is, paradoxically, an act of hope — the hope that if we can see the structure of failure clearly, we may eventually build structures that hold.