What this concept is saying
相違 (*virodha*, incompatibility or contradiction) is a relation in which two dharmas—two phenomena, properties, functions, mental states, or propositions—conflict so that they cannot stand together in the relevant respect.
That last qualification matters. “Incompatible” does not simply mean “different,” “disagreeable,” or “unlikely to occur together.” Tea and coffee are different, but both can sit on the same table. Affection and disappointment may coexist in a complicated relationship. By contrast, the same patch of water cannot, at the same moment and in the same respect, be both frozen solid and boiling. A reason offered to prove “sound is permanent” is contradictory if it actually proves that sound is impermanent.
Thus 相違 is not one narrowly defined object. It is a pattern that appears on several levels:
1. incompatibility between phenomena; 2. contradiction between a proof and what it is meant to prove; 3. mutual exclusion among simultaneously arising mental states; 4. self-defeat within a philosophical position; 5. and, under Madhyamaka analysis, the mistaken reification of incompatibility itself.
Its opposite is 不相違 (*avirodha*, non-incompatibility or consistency). The test varies with the level under discussion: can the two things coexist, and can the proposed reason function as a valid means of knowledge?
At the level of phenomena and causation, the *Nyāyabindu* (《正理滴论》) distinguishes two forms.
The first is causal or resultant exclusion: when one fully produced result is present, the contrary result is absent. The text’s example is cold touch and hot touch. The point is not that warmth and coolness can never occur in neighboring places, nor that temperature cannot change gradually. It is that the opposed results cannot both be fully present in the same locus, moment, and respect. The emphasis falls on the exclusion found in the result.
The second is existential exclusion: one member is present precisely where the other is absent, as in existence and nonexistence. Here the issue is not a struggle between two forces. “The vase exists here” and “the vase does not exist here,” when said of the same vase, place, time, and sense, cannot both be true. The emphasis falls on incompatibility in status, rather than on one result causally displacing another.
This already protects us from two common mistakes. Incompatibility is neither mere difference nor necessarily active opposition. Nonexistence does not “fight” existence. The two descriptions simply cannot both obtain under identical conditions.
In Buddhist logic, 因明 (*hetuvidyā*, the discipline of reasons and inference), 相違 becomes more precise. The *Nyāyapraveśa* (《因明入正理论》) says:
「不成、不定及与相违，是名似因。」 “Unestablished, inconclusive, and contradictory: these are called pseudo-reasons.”
A 相違因 (*viruddha-hetu*, contradictory reason) is not merely a weak argument. It is a reason that turns against its own thesis: instead of establishing what the speaker intends, it establishes something incompatible with it.
The clearest example is:
- Thesis: “Sound is permanent.” - Reason: “Because it is produced.” - Intended proof: permanence. - What the reason actually supports: impermanence.
Produced things belong with impermanent things. The reason does not merely fail to reach the destination; it points in the opposite direction.
The *Nyāyapraveśa* gives a fourfold classification:
- 法自相相違因: a reason contradictory to the stated property itself; - 法差别相違因: a reason contradictory to a particular implication or qualification of that property; - 有法自相相違因: a reason that undermines the subject itself; - 有法差别相違因: a reason that undermines a particular implication or qualification of the subject.
Here 法 means the property predicated in the thesis, while 有法 means the subject possessing that property. 自相 is the explicitly presented character; 差别 is a particular specification or implication involved in what is being asserted. The fourfold scheme is therefore built on two axes: subject versus predicate, and explicit character versus particular qualification.
The first type is easiest to see: “Sound is permanent because it is produced.” The reason “produced” proves the opposite of the predicate “permanent.”
In the second, the bare words of the predicate may appear intact, but the reason establishes a specification incompatible with what the proponent intended. The classical example concerning eyes and other composite faculties being “for another’s use” belongs to an old debate about whether composites imply an independent user. Its point is logical, not anatomical: the same reason can establish a kind of “use by another” contrary to the particular independent user the proponent meant to establish.
In the third, the offered reason rebounds against the very subject about which the claim is made. It does not merely disprove a property assigned to the subject; consistently applied, it also excludes the proposed subject itself.
In the fourth, it defeats a particular characterization implicit in that subject—for example, establishing “not being an object of the relevant cognition” where the thesis requires the contrary.
The latter three arise from technical Indian debates and cannot safely be reduced to four catchy English slogans. Their shared structure, however, is exact: the alleged proof establishes an implication incompatible with the predicate, the subject, or a specified aspect of either. “The supporting reason contradicts what is to be established” is the heart of all four.
The *Yogācārabhūmi* (《瑜伽师地论》) approaches debate through the question of whether a reason and example can serve as 量 (*pramāṇa*, a warranted means of knowledge). It identifies two grounds of incompatibility: the evidence is not decisive, or it is 同所成—just as much in need of proof as the thesis itself. A disputed claim cannot validate another disputed claim merely by being placed after the word “because.” Such a reason or example “cannot be a measure for accomplishing the thesis, and therefore is not called a measure.” A non-contradictory reason is decisive and genuinely distinct from what remains to be established.
This clarifies the difference among three major pseudo-reasons:
- 不成: the reason has not been established where it needs to apply; - 不定: the reason is inconclusive and occurs where both the thesis and its contrary may be found; - 相違: the reason positively supports what conflicts with the thesis.
The *Nyāyabindu* also discusses 相違决定, roughly “decisive counterbalancing contradiction,” in relation to doubt: opposed lines of inference can leave the matter unsettled. It explains why this need not always be established as an additional independent class of faulty reason; in some settings the defect belongs to the thesis or to the conditions under which testimony is being used. The important point is that “two sides each produced an argument” does not imply that both arguments are knowledge. Their terms, scope, and inferential marks still have to be examined.
At the level of mind, 相違 is neither a debate trick nor a metaphysical abstraction. The *Yogācārabhūmi* lists seven reasons why a mind does not arise: absence of an object-condition, absence of attention, non-attainment, incompatibility, severance, cessation, and the fact that another mind has already arisen. 相違 is only one member of this list and must not be confused with all the others.
Its examples are exact. When a mind associated with pleasant feeling is presently occurring, minds associated with painful feeling or neither-painful-nor-pleasant feeling do not arise in that same moment, because they are incompatible. The same mutual exclusion applies among all three kinds of feeling. Likewise, when a mind is bound by 贪 (*rāga*, craving or possessive attachment), a mind associated with 瞋 (*dveṣa*, aversion or hostility) does not arise simultaneously; when hostility is present, craving does not arise simultaneously.
This does not say that a person cannot shift rapidly between attachment and anger. Indeed, attachment frustrated often conditions anger. Nor does it say that an afternoon cannot contain mixed emotions. The claim concerns simultaneous occurrence in the same moment of mind as classified by these incompatible associations. Sequence is not simultaneity; causal intimacy is not identity.
The *Xianyang shengjiao lun* (《显扬圣教论》) employs 相違 to expose positions that undo themselves. If a theory denies the reality or efficacy of function in a way that leaves it simultaneously requiring and denying inclusion, or requiring something to be both useful and useless, its premise generates its own defect. If it denies continuity so completely that every distinction disappears, it conflicts with the causal and differentiated appearances acknowledged in ordinary experience. This is not the crude argument that “whatever common sense sees must be ultimately real.” Conventional experience can refute a view that indiscriminately erases convention, without thereby proving that conventional things possess independent essence.
This leads to the decisive Madhyamaka qualification. The *Mūlamadhyamakakārikā* with its Chinese commentary (《中论》) does not deny that light and darkness are conventionally opposed. It denies that “incompatibility” can be found as a self-existing relation or used by itself to establish an inherently real causal power.
In its critique of the lamp analogy, the text argues:
「若灯与闇相违故，能自照亦照于彼。闇与灯相违故，亦应自蔽蔽彼……是故灯喻非也。」
If the lamp could illuminate itself and another merely because it is incompatible with darkness, symmetry would require darkness, by being incompatible with the lamp, to conceal itself and another. Since the proposed reasoning does not work symmetrically, mere opposition cannot explain the claimed causal operation. The conclusion is not “lamps do not dispel darkness.” It is that a reified relation called incompatibility does not, by its own essence, supply an adequate causal explanation.
The same analysis extends through the *Mūlamadhyamakakārikā*’s examination of identity and difference: going is neither simply identical with the goer nor an independently established thing wholly different from the goer. And in examining conjunction, it says:
「是法不自合，异法亦不合，合者及合时，合法亦皆无。」
A thing does not conjoin with itself; independently different things do not yield a self-established relation of conjunction merely by already being different. Thus neither “opposed things” nor “the opposition between them” can be isolated as intrinsically established entities.
This gives 相違 its place in the path of liberation. At the practical level, discernment must distinguish wholesome from unwholesome, valid inference from deceptive inference, and mental factors that can or cannot occupy the same moment. Buddhism does not ask us to dissolve every distinction into fog. But deeper wisdom also sees that opposing terms arise dependently: “success” depends on a standard of failure, “insider” on outsider, “pure” on impure, and “enemy” on a network of memories, interests, labels, and conditions. The distinction can function without possessing an independent essence.
Ethical and meditative discernment therefore operate between two errors. One error is to ignore incompatibilities: to imagine that clinging can be cultivated without suffering, or that a contradictory reason can still produce knowledge. The other is to solidify oppositions: to turn “this view is mistaken” into “this person is essentially my enemy,” or “anger is present” into “I am an angry self.” Relative incompatibility guides conduct; emptiness prevents that guidance from hardening into hatred and identity.
A walkthrough in everyday life
Imagine an evening commute. You are standing in a crowded train, looking at your phone after a difficult day. A message from your partner reads: “We need to talk when you get home.”
First, distinguish what is merely different from what is genuinely incompatible. The message, the train noise, bodily fatigue, anxiety, and affection for your partner can all be present within the same episode. They are not automatically mutually exclusive. “Many things are happening” is not yet 相違.
Then a thought forms: “This message proves they are planning to leave me.” Here we enter the field of inference.
- Thesis: “My partner intends to end the relationship.” - Proposed reason: “They wrote, ‘We need to talk.’” - Evidence actually available: an ambiguous sentence.
This is not necessarily a contradictory reason. More likely it is 不定, inconclusive: the same message could precede bad news, a practical discussion, a confession of distress, or a decision about childcare. Calling every bad inference “contradictory” would itself blur the classical categories.
Now suppose you add: “They are definitely leaving because they have been making plans for our holiday next month.” Unless other facts change its meaning, that reason tends to support continued involvement and therefore conflicts with the thesis it is being asked to prove. The attempted support turns against the conclusion: this approaches 相違因.
Next comes the mental sequence. The message is interpreted as a threat. Fear or painful feeling arises. A moment later you begin rereading old affectionate messages, and craving for reassurance becomes prominent: “Tell me right now that we are safe.” Then the reply is delayed, and attachment gives way to hostility: “They always do this to me.”
Popular speech calls this “feeling everything at once.” The *Yogācārabhūmi* invites a finer inspection. Across a short interval, painful feeling, grasping, and hostility may alternate so quickly that memory bundles them into one emotional block. But its claim about 相違 concerns which classified states occupy one moment together. It does not deny rapid succession, causal transition, bodily complexity, or the ordinary usefulness of speaking about mixed emotions.
The same precision applies to the three feeling tones. Pleasant, painful, and neither-painful-nor-pleasant feeling are not simply “happiness,” “sadness,” and “boredom” as life stories. They are hedonic tones associated with moments of experience. A pleasant memory and a painful bodily sensation may seem jointly present in ordinary description; classical analysis asks more narrowly about the feeling associated with the particular moment of consciousness under analysis. One should not force a coarse report of several bodily and mental processes into a simplistic one-word label.
Now you open a social-media feed to escape the discomfort. The recommendation system quickly supplies videos about “five signs your partner is emotionally withdrawing.” You reason: “The algorithm showed me this because it fits my situation; therefore it confirms my fear.”
The algorithm’s recommendation is not independent evidence. It is partly conditioned by what you paused on, searched for, and watched. Treating its output as proof can make the example 同所成—dependent on the very assumption it is supposed to establish. The feed may amplify a hypothesis without validating it.
This is a useful comparison with AI systems as well. A fluent answer can repeat the premise embedded in a prompt and then be cited as confirmation of that premise. Fluency is not 量. The output may be unestablished, inconclusive, or contrary to reliable evidence. This analogy does not identify Buddhist epistemology with computer science, and contemporary research does not “prove” Buddhist accounts of mind. It simply makes an old warning vivid: evidence that inherits the uncertainty of the claim cannot magically certify it.
You arrive home ready to argue. Your internal position is: “I want an honest conversation, but nothing you say could change my conclusion.” The position defeats itself. If no possible answer could count, the proposed “conversation” is not functioning as an inquiry. This resembles the self-undermining use of 相違 in the *Xianyang shengjiao lun*: the declared function is openness to information, while the operative rule excludes every possible correction.
At this point, practice does not require pretending that every interpretation is compatible with every other. Your partner may indeed be leaving; some explanations may be ruled out by evidence. The practice is to ask what kind of conflict is actually present:
- Are two states unable to coexist in the same respect, or are they merely different? - Is my reason established? - Is it decisive, or merely compatible with several conclusions? - Does it support my claim, or does it turn against it? - Am I describing a succession of mental states as if they were one simultaneous block? - Is my position open to the kind of evidence it claims to seek?
Finally, suppose the conversation reveals that the message concerned a parent’s health, not the relationship. Relief arises, followed by embarrassment. You may now construct another essence: “I am irrational.” Madhyamaka analysis asks you not to replace one rigid opposition with another. The mistaken inference occurred through fatigue, ambiguity, memory, habitual fear, and algorithmic reinforcement. It was real enough to cause suffering and should be understood; yet no independently existing “irrational self” can be extracted from that dependent sequence.
This is the whole arc in miniature. Logical discipline corrects the inference. Mindful observation distinguishes simultaneous incompatibility from rapid succession. Ethical restraint prevents fear from becoming harmful speech. Insight into dependent arising keeps the correction from becoming a new identity. Example leads into doctrine; doctrine returns as freedom within the example.
Why contemporary people need it
Contemporary life rewards collapsed distinctions. Platforms favor declarations such as “If you care about justice, you must accept every claim made in its name,” or “If you criticize this policy, you oppose the people it intends to help.” These formulations manufacture incompatibility where none has been demonstrated. Two commitments may be in tension without being logically contradictory; two people may disagree without becoming existential enemies.
At the same time, contemporary culture often denies genuine incompatibilities when acknowledging them would be inconvenient. We tell ourselves that compulsive consumption and contentment can be cultivated by the same movement of grasping; that constant interruption and collected attention can be maximized simultaneously; that a workplace welcomes dissent while punishing every instance of it; or that one can demand total control in an intimate relationship while preserving the other person’s freedom.
The teaching of 相違 supplies three forms of sobriety.
First is epistemic sobriety: a reason must do more than resemble support. When a claim is repeated by many accounts drawing from the same source, repetition does not create independent confirmation. When an AI answer restates a user’s assumption, confidence of tone does not establish truth. When a scientific finding is invoked, one must still ask about population, method, replication, interpretation, and scope. Scientific disciplines themselves remain contested and developing; they cannot simply be recruited as a seal certifying Buddhist doctrine.
Second is psychological sobriety: the apparent solidity of “my mood” may conceal a rapidly conditioned series. Craving, irritation, pleasure, pain, and neutrality arise and cease under conditions. Seeing their sequence weakens identification. Instead of “I am furious and therefore must act,” one can recognize: painful contact has occurred; grasping sought reassurance; frustration conditioned hostility; hostility is now present. This is not an excuse. It is a more workable map.
Certain wholesome and unwholesome orientations are also practically opposed. A moment of clear non-harming is not the same mental orientation as a moment intending injury. Cultivation matters because what is repeatedly supported becomes more available, while incompatible tendencies are deprived of occupation and reinforcement. Yet Buddhism does not teach liberation by simply suppressing one half of every pair. Hatred of hatred remains entangled with hatred. The path replaces ignorance through understanding and relinquishes craving by seeing its conditions and cost.
Third is ontological sobriety: functional opposition does not imply self-existing enemies. Fire and coolness, truth and falsity, generosity and possessiveness, liberation and bondage can be distinguished conventionally. But each designation depends on conditions, criteria, relations, and conceptual imputation. Madhyamaka’s denial of intrinsic nature does not make contradiction disappear; it prevents contradiction from being transformed into a metaphysics of permanently opposed substances.
This is especially important in moral and political life. Wisdom can say, “This act is harmful and must be stopped,” without concluding, “This person possesses harm as an eternal essence.” Compassion without discernment becomes permissiveness; discernment without emptiness becomes condemnation. The liberating middle is capable of firm distinctions without reified hatred.
On the path, 相違 therefore serves progressively different purposes:
- In hearing and study, it prevents us from accepting self-contradictory teachings or reasons. - In reflection, it helps separate established evidence, inconclusive evidence, and counterevidence. - In ethical discipline, it reveals when our professed values and operative intentions defeat one another. - In meditation, it sharpens observation of which mental events coexist and which occur only in rapid sequence. - In insight, it exposes the dependent construction of every opposition, including bondage and release.
Even the final pair must be handled carefully. Saṃsāra and nirvāṇa are not two eternal territories locked in cosmic combat. Bondage is dependently arisen misapprehension and appropriation; liberation is not another substance but the exhaustion of ignorance and grasping. Conventional distinctions guide the path, while emptiness prevents the path, the practitioner, and the goal from being appropriated as fixed selves.
Common misunderstandings and clarifications
“相違 means any difference.”
No. Difference is broader than incompatibility. Blue and round are different properties and can belong to the same object. Incompatibility requires exclusion in a specified locus, time, and respect, or a proof relation that establishes what conflicts with the thesis.
“Two things that alternate are not incompatible.”
They may be incompatible precisely as simultaneous states. Hot and cold conditions can succeed one another; attachment can condition anger; pleasant and painful moments can alternate. Succession does not show simultaneous compatibility.
“People can have mixed feelings, so the *Yogācārabhūmi* must be wrong.”
This compares different levels of description. “Mixed feelings” ordinarily summarizes a temporally extended, multi-object, bodily and mental episode. The treatise is analyzing momentary minds and their associated feeling tones or afflictive bindings. Its claim is not that a human situation must feel psychologically simple.
“Pleasant, painful, and neutral feeling are three opinions about an experience.”
No. 受 (*vedanā*, feeling) here is the experienced hedonic tone, not an intellectual judgment. And 相違 concerns the mutually exclusive association of these feeling classes in the analyzed moment, not a ban on complex reports about an entire afternoon.
“If an argument is false, it is a contradictory reason.”
Not necessarily. A reason may be unestablished, inconclusive, irrelevant, or genuinely contradictory. A 相違因 is distinctive because it establishes, or tends in the relevant inferential system to establish, what conflicts with the intended conclusion. “Sound is permanent because it is produced” is paradigmatic because producedness supports impermanence.
“The four contradictory reasons are just four translations for the same mistake.”
They share one structure but attack different locations: the predicate’s explicit character, a particular qualification of the predicate, the subject’s explicit character, or a particular qualification of the subject. Their classical examples presuppose technical debates; translating the nouns without explaining these axes would conceal rather than clarify the doctrine.
“If two arguments oppose each other, truth lies halfway between them.”
No. Opposed conclusions can arise from unequal evidence, ambiguous terms, different scopes, or parallel defects. The discussion of 相違决定 and doubt warns us to examine both inferences. A compromise is not automatically knowledge.
“Whatever contradicts ordinary experience must be ultimately false.”
The Yogācāra critique of views contrary to 世间现见—what is ordinarily observed—does not make common sense infallible. Conventional observation can expose a theory that erases every causal and practical distinction. Ultimate analysis then asks whether the convention exists intrinsically. Conventional validity and ultimate non-establishment are not competitors on a single flat plane.
“Because Madhyamaka says incompatibility is empty, contradictions do not matter.”
This is the most dangerous misunderstanding. Emptiness means absence of intrinsic, independent nature; it does not mean absence of function. Poison and medicine are not interchangeable because both are empty. A reason that proves the opposite remains defective conventionally. The lamp-and-darkness argument denies that reified incompatibility alone explains causal efficacy; it does not deny that illumination conventionally removes darkness.
“Emptiness reconciles every pair into a mystical unity.”
Not in the sense that truth and falsehood, cruelty and compassion, or valid and invalid inference become indistinguishable. Madhyamaka rejects both intrinsically separate substances and an intrinsically unitary substance beneath them. Dependent arising permits distinctions to function precisely without requiring fixed essences.
“To eliminate anger, I should cultivate craving, since the two are incompatible.”
The *Yogācārabhūmi* uses craving and hostility to illustrate simultaneous exclusion, not to prescribe one affliction as the antidote to the other. Craving can condition hostility when obstructed, and hostility can renew craving for vindication. Buddhist cultivation works toward non-greed, non-hatred, and non-delusion, not strategic oscillation between afflictions.
“相違 teaches us to win debates.”
Its deeper purpose is to stop being deceived—especially by our own reasons. A person can identify every contradiction in an opponent and remain blind to the contradiction between professed compassion and contemptuous conduct. In the liberation path, logical examination becomes ethical and contemplative examination: What does this belief actually support? What state of mind is operating? What suffering does it condition? What identity am I constructing around the opposition?
The mature understanding of 相違 is therefore neither a love of conflict nor a wish to erase it. It is the capacity to distinguish accurately, reason honestly, observe the mind closely, and then release the fantasy that the opposing sides exist by their own power. We learn to say both: “These cannot stand together in this respect,” and “Neither side, nor the relation between them, possesses an independent essence.”
That double vision is not indecision. It is what allows wisdom to be precise without becoming hard, and compassion to remain warm without becoming vague.