量

What This Concept Is Saying

The Chinese word *liang* (量, “valid measure,” “standard,” “quantity,” or “magnitude”) carries two distinct meanings in the materials before us. They are related by the broad idea of “that by which something is determined,” but they must not be collapsed into one doctrine.

In the *Yogācārabhūmi Śāstra* (《瑜伽师地论》), 量 concerns the grounds on which a thesis can rightly be established. Here it is close to “valid cognition,” “warrant,” or “authoritative means of knowing.” In the cited passages of the *Abhidharmakośabhāṣya* (《阿毘达磨俱舍论》), however, 量 chiefly means quantity, extent, proportion, or measurement. The latter passages do not present a theory of perception, inference, and authoritative teaching; those categories must not be imported into them merely because the same Chinese character appears in both works.

This distinction matters beyond philology. The first meaning asks, “What warrants this judgment?” The second asks, “How much, how large, how long, or in what proportion?” One tests cognition; the other specifies magnitude.

In the *Yogācārabhūmi*, 量 belongs to the structure of reasoned discussion, together with the thesis (*zong*, 宗), reason (*yin*, 因), example (*yu*, 喻), similar cases, and dissimilar cases. A purported ground is not automatically valid just because it sounds persuasive or comes with supporting language. If it conflicts with what is to be established, fails to determine the matter, or merely repeats the conclusion in another form, it cannot establish the thesis and is “not a measure” (*bu ming liang*, 不名量). A ground counts as a “correct measure” (*zhengliang*, 正量) when it does not conflict with the thesis, is determinate, differs from the proposition it supports, and can genuinely establish that proposition.

That last condition is easily overlooked. “This teaching is true because it is a true teaching” adds no independent support. “I am right because I feel certain” confuses psychological force with cognitive warrant. Certainty may accompany knowledge, prejudice, fear, or intoxication; it is not by itself a measure of truth.

The *Yogācārabhūmi* discusses three principal kinds of warrant.

1. *Xianliang* (现量, direct or present cognition)

This is cognition of what is presently known rather than a conclusion reached by discursive comparison. The treatise uses several related expressions: “worldly direct cognition” (世间现量), “purified direct cognition” (清净现量), and “purified direct cognition not shared with the world” (不共世间清净现量).

The last expression points to supramundane wisdom that knows its field as it is: the conditioned as conditioned, the unconditioned as unconditioned, the surpassable as surpassable, and the unsurpassable as unsurpassable. It therefore should not be reduced to ordinary eyesight or to the slogan “seeing is believing.” The category reaches from worldly immediacy to a mode of knowing transformed by the path.

Nor does direct cognition mean that every interpretation arising during a sensory event is infallible. “I hear a raised voice” and “she raised her voice because she despises me” do not have the same status. The first is closer to what is presently given; the second already incorporates memory, expectation, attribution, and inference. Much unnecessary suffering begins when these layers are fused.

2. *Biliang* (比量, inferential cognition)

Inference is accompanied by reflective discrimination and considers what has been or ought to be examined. The treatise gives five forms, each of which must retain its own structure.

- *Xiang biliang* (相比量, inference from a sign or characteristic) knows something through a sign and its relation to what is signified: seeing a banner and knowing that a carriage is present, or seeing smoke and inferring fire. The same pattern may infer qualities such as faith, dispassion, the conduct of a bodhisattva, or the qualities associated with a tathāgata from relevant manifestations. This does not mean that appearances never deceive; it means that an inference depends on a defensible relation between sign and signified.

- *Ti biliang* (体比量, inference from an observed entity or part) proceeds from a presently seen body, instance, or portion to an unseen entity or its remaining portion. It may also reason from the present to the past or future, and from something near to something distant. Seeing one exposed section of a structure and inferring its continuation is a simple contemporary example, though the strength of the conclusion depends on how reliably the observed part represents the whole.

- *Ye biliang* (业比量, inference from activity or function) infers the bearer or basis of an activity from what it does. Movement at a distance may distinguish a person from a tree stump; tracks, manner of movement, or sounds may indicate an elephant, snake, horse, or lion. The treatise also gives a more technical pattern: from seeing, hearing, smelling, tasting, touching, and cognizing, one infers respectively the eye, ear, nose, tongue, body, and mind as their bases.

- *Fa biliang* (法比量, inference through connected dharmas) moves from one associated or contiguous principle to another: from impermanence to suffering, from suffering to emptiness and non-self; from the conditioned to arising, abiding, alteration, and cessation; and from the unconditioned to the absence of those conditioned phases. This is not merely an association of words. Each transition expresses a contemplated relation that must be understood and verified within the relevant doctrinal framework.

- *Yinguo biliang* (因果比量, causal inference) reasons between cause and effect in either direction. One may infer a likely result from its causes, or infer antecedent causes from an observed result. The cited summary gives the example of inferring attainment of a śramaṇa’s fruit from cultivation of the path, or inferring prior cultivation from the fact that such a fruit has been attained.

These five forms show that Buddhist inference is not one vague act called “thinking.” A sign, a part, a function, a doctrinal relation, and a causal relation do different work. Confusing them produces bad arguments: a mere resemblance is treated as a cause, a visible fragment is mistaken for the whole, or correlation is promoted into necessity.

3. *Zhengjiao liang* (正教量, the warrant of authentic teaching), also called *zhengjiao* (正教, authentic teaching)

This refers to teaching spoken by an omniscient one, or heard from that source and transmitted in accordance with the Dharma. Yet “a text says so” is not the complete criterion. The *Yogācārabhūmi* gives three tests:

- *Bu wei shengyan* (不违圣言): it does not contradict the sacred teaching. It is spoken by the Buddha or noble disciples, transmitted through the tradition, and does not violate right Dharma and right meaning.

- *Neng zhi zaran* (能治杂染): it can counteract defilement. Practised according to the teaching, it is capable of taming greed, hatred, delusion, and their associated afflictions. This is a practical and soteriological test. A doctrine that merely supplies spiritual ammunition for vanity and hostility has not thereby demonstrated its curative power.

- *Bu wei faxiang* (不违法相): it does not violate the characteristics of phenomena. It neither adds existence where the relevant characteristic is absent nor denies what is present; it does not confuse what is determinate with what is indeterminate, difference with non-difference, cause with effect, or literal actuality with provisional designation.

These criteria hold textual fidelity, transformative efficacy, and accuracy about phenomena together. None should be isolated. Fidelity without understanding becomes dogmatism. Claimed spiritual benefit without accuracy can become wishful thinking. Analytical precision without the treatment of greed, hatred, and delusion can become an elegant instrument of self-importance.

The second meaning of 量 is more straightforward but philosophically revealing. In the cited *Abhidharmakośabhāṣya* passages, it denotes amount, dimension, proportion, or a measurable relation. The text speaks of the quantity or extent of Dharma aggregates, of the nose and two other faculties taking an “equal-sized object” (*dengliang jing*, 等量境), and of measurements of the container-world, bodily magnitude (*shenliang*, 身量), and lifespan (*shouliang*, 寿量). Measurement and number sometimes occur side by side.

The treatise also preserves a named hierarchy of units. In its fine-length sequence, each successive unit is seven times the preceding one: seven *jiwei* (极微) make one *wei* (微); the progression then continues through gold-dust, water-dust, rabbit-hair dust, sheep-hair dust, ox-hair dust, dust visible in a sunbeam, nit, louse, barley grain, and finally finger-joint. Above that level, three finger-joints make one finger; twenty-four fingers laid crosswise make one cubit; four cubits make one bow-length, also called a fathom; five hundred bow-lengths make one *julushe* (俱卢舍); and eight *julushe* make one *youshanna* (踰缮那). The road between a village and an *āraṇya*, a secluded wilderness dwelling, is used to indicate approximately the road-distance of one *julushe*.

These old units should not be treated as modern laboratory constants. The text itself emphasizes systems established through names. Measurement makes comparison possible by selecting a unit, naming it, and applying a rule of accumulation. It is useful precisely because it stabilizes a relation; it is not an independently existing essence hidden inside things.

Placed in the path of liberation, the two meanings converge without becoming identical. We require sound warrants because ignorance mis-knows experience; we require awareness of measurement because the mind continually measures “more,” “less,” “enough,” “too late,” “not successful enough,” and “not loved enough.” Valid cognition helps distinguish what is known from what is projected. Reflection on conventional measures shows how much of the world we inhabit is organized through useful designations. Neither exercise is liberation by itself, but both can weaken confusion and prepare the mind for seeing phenomena without distortion or appropriation.

A Walk-Through in Everyday Life

Imagine receiving a short message from your partner during an exhausting commute: “Fine. Do whatever you want.” You immediately think, “They are furious, the relationship is deteriorating, and tonight will be a disaster.” At the same time, your phone reports that you have spent six hours online this week, while an app tells you that your relationship “compatibility score” has declined.

Walk through this slowly.

First, identify what is directly present. There are visible marks on a screen, recognized as words. There may be bodily tension, heat in the face, and a tightening in the abdomen. These are candidates for the layer of present cognition. “My partner is furious,” however, is not simply visible in the pixels. It is already an inference.

This is where people most often map the categories incorrectly: they call the entire first interpretation “direct perception” because it arrived instantly. Speed does not turn inference into direct cognition. An interpretation can be learned so deeply that it appears before deliberate thought, yet it still goes beyond what was immediately given.

Next, examine the inference by its actual kind.

As an inference from a sign, the clipped wording may signify irritation—but only if that wording is reliably connected with irritation in this person and context. The same sign could arise from fatigue, hurry, distraction, pain, or an awkward attempt to avoid conflict. A sign licenses investigation; it does not automatically license certainty.

As inference from a part, one message is a fragment of a longer interaction. You may reasonably use a part to consider the whole, but a fragment is not the whole relationship. Algorithms routinely exploit this mistake: one click is treated as your identity, one pause as your preference, one week as your enduring character. Sometimes such predictions are useful; sometimes the sample is unrepresentative.

As inference from activity or function, you might consider observable patterns: replies have become unusually delayed, plans are repeatedly cancelled, or attempts at conversation are avoided. These activities may indicate an underlying difficulty. But they still do not reveal its nature without further evidence. Behaviour can disclose that something is happening without telling you exactly what it means.

As inference through connected principles, you can notice that this emotional event is conditioned and changing. What changes cannot provide permanent security merely by being grasped. The unease produced by demanding permanence from changing conditions opens contemplation of suffering, emptiness, and non-self. This does not mean, “The relationship is unreal, so I need not care.” It means that neither the message, the fear, nor the threatened self possesses the fixed, independent nature that panic assigns to it.

As causal inference, you may reason prospectively: if you answer with accusation, defensiveness will probably intensify; if you pause, regulate your agitation, and ask what is happening, a more truthful conversation becomes possible. Looking backward, you may infer that sleep loss, unresolved resentment, workplace pressure, or your own earlier words contributed to the present tension. Causal inference should widen responsibility, not manufacture a single villain.

Now apply the three tests of authentic teaching. Does the advice you are about to follow accord with the Dharma rather than merely borrow Buddhist vocabulary? If practised, does it actually reduce greed, hatred, and delusion, or does it make resentment feel holy? Does it respect the characteristics of the situation, neither adding a betrayal that has not been established nor denying pain that is plainly present? Does it distinguish a provisional label—“toxic,” “avoidant,” “bad partner”—from the full reality of a changing person and relationship?

Finally, notice the other meaning of 量. “They took three hours to reply.” “I initiate conversation 70 percent of the time.” “We spend only one evening together each week.” These may be accurate measurements. They can reveal a pattern that vague impressions conceal. But numbers answer questions of amount, not automatically questions of meaning. Three hours can signify indifference, a medical emergency, concentrated work, or nothing of consequence. A compatibility score is a quantity generated by a model; it is not direct access to the moral or emotional truth of a relationship.

The same applies at work. Hours logged, tickets closed, citations received, followers gained, calories counted, and a child’s examination scores are forms of measurement. They can support wise decisions. They become oppressive when a conventional unit is mistaken for a complete measure of a person.

Brain science and machine learning offer helpful comparisons here. Both distinguish raw inputs, learned feature extraction, probabilistic inference, and output decisions. This can remind us that an apparently immediate judgment may contain many hidden stages. But the analogy has limits: scientific models remain contested and incomplete, artificial systems do not thereby possess supramundane wisdom, and correlations in neural or computational processes do not prove Buddhist claims about liberation. The comparison can sharpen a question; it cannot substitute for contemplative realization.

Why Contemporary People Need It

Contemporary life produces a peculiar combination: an excess of information and a shortage of trustworthy orientation. We see images instantly, receive conclusions before context, and inherit measurements chosen by institutions whose purposes may not be ours. 量 offers two disciplines suited to this condition.

The first is epistemic humility with teeth. It does not merely say, “Everything is subjective.” It asks what kind of warrant a claim has. Was it directly known? Inferred from a sign? Generalized from a part? Inferred from an effect? Received from an authority? Does the alleged evidence determine the conclusion, or merely repeat it? This makes humility precise rather than evasive.

The second is freedom from metric enchantment. Measures are indispensable: medicine needs dosage, buildings need dimensions, schedules need duration, and communities need ways to distribute resources. But contemporary systems often convert a useful measure into a goal and then the goal into an identity. Attention becomes screen time, friendship becomes engagement, competence becomes productivity, health becomes a dashboard, and worth becomes rank.

The Buddhist response is not to abandon numbers. The *Abhidharmakośabhāṣya* itself carefully records scalable units. The point is to know measurement as measurement: a named, relational system that serves particular purposes. What can be counted is real enough to function conventionally, but the count does not exhaust what the thing is.

This has a direct place on the path. Ignorance does not only consist in lacking information. It includes superimposing what is not there, subtracting what is there, confusing the certain and uncertain, mistaking a designation for an essence, and reversing cause and effect. Those are precisely the kinds of distortion guarded against by valid measure and authentic teaching.

Practice therefore begins before dramatic mystical experience. It begins when one says:

“I directly know this much.”

“I infer this, and here is the relation on which the inference depends.”

“I received this teaching, and I will examine whether it accords with Dharma, treats defilement, and respects the characteristics of phenomena.”

“This number is useful for this purpose, but it is not the whole truth of a life.”

Such discipline restrains impulsive appropriation. It creates room between contact and story, evidence and certainty, description and selfhood. In that room, craving and aversion are less able to pose as knowledge.

Yet 量 is not the final aim. A perfectly argued map is not arrival. Correct inference about impermanence can guide practice, but liberation requires that impermanence, suffering, emptiness, and non-self cease to be merely conclusions one can defend. The distinctive “purified direct cognition not shared with the world” marks this horizon: the path moves from hearing, through disciplined reflection, toward direct, transformative knowing.

Thus reason and teaching are not enemies of realization. They protect the journey from fantasy and point beyond themselves. Their success is not measured by how many arguments one wins, but by whether misapprehension and defilement are actually relinquished.

Common Misreadings and Clarifications

“量 always means *pramāṇa* or valid cognition.”

No. In the cited *Abhidharmakośabhāṣya* passages it means quantity, magnitude, extent, or measurement. Even the *Yogācārabhūmi* uses 量 in ordinary expressions concerning high or broad dimensions, low or narrow dimensions, and the measure of worldly speech. Context determines the sense.

“Because the *Yogācārabhūmi* teaches direct cognition, inference, and authentic teaching, the *Abhidharmakośabhāṣya* passages must teach the same three.”

That is precisely the kind of unwarranted addition the category warns against. The supplied *Abhidharmakośabhāṣya* passages do not establish that classification. A doctrine found in one treatise cannot be inserted into another passage without textual evidence.

“Direct cognition means whatever feels immediate and obvious.”

No. A thought can arise immediately while containing interpretation, memory, and inference. “There is a message on the screen” and “the sender no longer loves me” are not the same cognitive act. Nor should ordinary present cognition be equated with the supramundane, purified cognition described by the treatise.

“Inference is just guessing.”

No. Guessing lacks an adequately established relation. Inference proceeds through a sign, a known part, an activity, an associated principle, or a causal relation. It can be strong or weak depending on whether that relation is reliable and whether alternatives have been considered.

“All five inferences are basically the smoke-and-fire example.”

No. Smoke and fire exemplify inference from a sign. Inferring a whole from a seen part, a bearer from its activity, non-self through a doctrinal relation, and a cause from its effect have distinct structures. Flattening them hides the very discipline the classification supplies.

“Authentic teaching means blind obedience to scripture or teachers.”

No. The *Yogācārabhūmi* gives a threefold criterion: non-contradiction with sacred teaching, efficacy in treating defilements, and non-violation of the characteristics of phenomena. Lineage or quotation alone does not erase the need for right meaning, practice, and accurate discernment.

“If a teaching makes me feel calmer, it must be authentic.”

Not necessarily. Temporary comfort may coexist with denial, attachment, or distorted causality. “Treating defilement” means the actual taming of greed, hatred, delusion, and associated afflictions, together with fidelity to Dharma and phenomena—not simply producing a preferred mood.

“If something is conventionally measured, Buddhism says it is unreal.”

No. Measurements function. Dosages, distances, durations, and proportions have consequences. The point is that units are established through names and relations; they should not be mistaken for self-existing essences or for exhaustive descriptions of what they measure.

“Numbers are objective, so they interpret themselves.”

Numbers can be accurate while the conclusion drawn from them is false. A duration does not reveal a motive; a score does not define a person; a correlation does not by itself establish a cause. Measurement supplies magnitude. Interpretation still requires an appropriate warrant.

“Valid cognition is chiefly a weapon for defeating other people’s views.”

Within debate, 量 does help establish or reject a thesis. Within the path, however, its deeper target is our own ignorance. The crucial false thesis is often the one silently asserted by craving: “This changing thing must secure me,” “this feeling proves my story,” or “this label is the person.”

The wisdom of 量 is therefore double. Learn to ask what truly establishes a claim, and learn to see the limits of every scale by which the world is counted. The first protects the mind from falsehood; the second protects it from mistaking a useful convention for ultimate reality. Together they train a person not merely to possess more information, but to know with greater honesty—and eventually to release the grasping that turns knowledge, authority, and measurement into new forms of bondage.