俱舍論 業體·表無表Abhidharmakośabhāṣya · The Nature of Action · Manifest and Nonmanifest

Before Reading This Fascicle

The Chapter on the World took four fascicles to finish its account of the twofold world, of sentient beings and of the receptacle. This fascicle opens with a single sentence: “by what are differences such as these produced?” — the author turns on the heel of one question, and the whole Kośa walks out of “what the world looks like” into “why the world is as it is.” Puguang’s Record, explaining at the head of the chapter why it comes in this order, says that “the reason action is explained next is that the previous chapter explained results and this chapter explains causes; a result does not arise alone but must arise in dependence upon a cause, and as regards the result action is the nearer, and so action is spoken of next” — the previous chapter treated results, this chapter treats causes; a result does not arise all by itself but must rely upon a cause, and action is nearer to the result than the afflictions are, and so action is spoken of first.

The skeleton of this fascicle is really just two words: entity and dispute. First the entity of action is established (volition and what volition brings about; the three actions of body, speech, and mind; the manifest and the nonmanifest), and then the two great disputes are entered — what the entity of bodily manifest action is, and whether the nonmanifest really exists at all. The two disputes together take up more than half of the Chinese text of this fascicle, and are among the densest places of question and answer in the whole treatise. One should be prepared before reading: what is going on here is not the handing down of a verdict but making the reasons of each of the two parties clear. In the midst of the dispute the author himself says the single most important sentence in it — “this and that are alike hard to know, and here I bear no grudge against either.” Both sides are hard to know; toward neither do I bear ill will. That sentence sets the key of the whole fascicle.

There is one more thing to say first. A technical pair of words recurs throughout this fascicle: manifest and nonmanifest. The manifest is that portion of a bodily or verbal act which is displayed outwardly and which others can see; the nonmanifest is that portion which arises along with the act, is not displayed outwardly, and yet is able to continue on. Everything argued in the fascicle — “does it exist?” and “what is it?” — turns over and over upon these two words. Set these two words firmly in place, and the whole fascicle’s questions and answers have something to hang from.

I. By What Are the Differences of the World Produced

〔1.1〕First strike out the item “creator.” First the question: why does the world have so many differences? The treatise answers with the utmost bluntness — “they do not arise from a single lord who first conceived them, but arise from the differences in the actions of sentient beings.” It is not that some sovereign first stirred a thought and thereafter made them; it is only that the actions of sentient beings severally differ. This one sentence occupies a weighty place in the whole treatise: it hands the enormous panorama described throughout the Chapter on the World back, at one stroke, to sentient beings themselves.

The easy error: to read “a single lord” as a general term for “some cause.” It is not. “A single lord” means that class of account which maintains an eternally abiding creator; what the author blocks is “one maker who first conceives,” not “cause” itself — for action is precisely the cause.

In practice: in reading Abhidharma, when you meet “why is it thus?”, the first step is not to ask “who arranged it?” but to ask “whose action drew it?” Make that exchange, and all the classifications and numbers that follow will be entered in the ledger of practice.

〔1.2〕Why the inner body is ugly and the outer goods lovely — two faces of one set of actions. The treatise inserts here an excellent objection: since both alike arise from action, why are saffron and sandalwood, which are outside the body, so lovely, while the form within the body “constantly runs impurity from the nine sores”? The answer is: “because the class of actions of sentient beings is just so.” If what was done was mixed action (wholesome and unwholesome mingled), then the inner body obtained is exactly of this sort; and in order to counteract just this portion, the same set of actions further “obtains outward goods whose color, scent, taste, and touch are most lovely and delightful.” As for the gods, “the hosts of gods do purely pure actions, and so the two things they draw to themselves are both excellent” — the action being pure, inner and outer are both excellent.

First ask: what is this passage proving? It is proving that “being obtained by action” is not one coarse line but a fine-grained ledger: in one and the same person the inner body and the outer goods may be obtained by one and the same set of actions and yet be opposite in quality, simply because the action itself is mixed. The easy error: to read it as a fixed apportionment, “the beautiful to the outside, the ugly to the inside.” It is not — the sentence about the gods is precisely the counter-instance: if the action is pure, inner and outer are both excellent.

In practice: this passage yields a very serviceable self-examination — when you see the unevenness between yourself and your surroundings, think first “my action is mixed,” and not “the world has treated me unfairly.”

〔1.3〕What the entity of action is: volition, and what volition brings about. The author poses himself the question “what is the entity of this action by which they come about,” and answers himself: “it is the volition that arises in the mind, together with what volition brings about,” and adduces a sūtra in proof: “hence the sūtra says: there are two kinds of action, first the action that is volition, second the action of having willed; and the action of having willed refers to what volition brings about.”

First ask: why divide it into two kinds? Because action has two stages, “still in the mind” and “already carried out.” The action that is volition is the stretch within the mind; the action of having willed is the stretch that the mind has brought to completion. The later text distinguishes the two most clearly: what arises at the stage of preliminary effort is the “deliberating volition” — I should do such a thing; what arises once one has deliberated is the “volition that does the deed,” which moves the body and issues speech in accordance with what was deliberated before. The easy error: to read “volition” as ordinary “thinking.” It is one of the mental factors, and its sense is doing; and precisely because it is a doing, it can serve as the entity of action.

In practice: this division into two is the ground floor of all the treatments of action that follow. Whenever, further on, you read of the ten paths of action or the five deeds of immediate retribution and ask “does this constitute an action or not?”, turn back and ask one question: is it in the stretch of the action that is volition, or has it gone on to the action of having willed?

〔1.4〕How the three actions can be established as three — three different measures. The two actions are opened out again into three: body, speech, and mind. The author sets up a very severe objection here: on what ground are three established? From their support, from their own nature, or from what raises them? And he walks each of the three roads himself, and none of them is passable — if from their support, all action depends upon the body, and then there would be only one action; if from their own nature, of the three only speech is of itself a doing, and then speech alone would be action; if from what raises them, all action is raised by mind, and then again there would be only one action.

The answer lies with the Vaibhāṣikas: “the Vaibhāṣika masters say that the three actions are established, in that same order, by the three causes above.” That is to say — the three actions are not measured out with one and the same measure: bodily action is established from its support (all action alike depends upon the body, but bodily action alone takes the body as the name of its support); verbal action is established from its own nature (speech is itself a doing, and its nature as action is just speech); mental action is established from what raises it (mind is that which raises, and the two actions of body and speech are both raised by mind). The author at once draws the conclusion: “the volition that arises in the mind is mental action; and the action that volition brings about divides into the two, bodily action and verbal action, since these are raised by volition.”

The easy error: to suppose that because none of the three roads was passable, the three actions cannot be established. Quite the contrary — the treatise here demonstrates a standing rule of Abhidharma: one and the same set of enumerated dharmas may be established separately by different criteria. Whenever the numbers of a classification fail to agree, the first question is not “who is wrong?” but “which measure was each of these items taken with?”

In practice: the move will be used once more later in this fascicle — the four causes of the four wholesome and four unwholesome (ultimate · intrinsic · associated · raised) run on exactly the same line of thought.

II. Bodily and Verbal Action Both Extend to the Manifest and the Nonmanifest

〔2.1〕One line of stanza, two worlds. “These two, bodily action and verbal action, both alike have the nature of the manifest and the nonmanifest” — ten characters. First ask: why should this one line be set up as half a stanza of its own? Because at one breath it cuts each of the two actions, bodily and verbal, into halves: the half that can be seen (the manifest) and the half that cannot (the nonmanifest). Every dispute in the rest of the fascicle hangs upon this one cut.

The easy error: to understand the nonmanifest as “a thought in the mind.” It is not — for the Sarvāstivādins the nonmanifest is a form-dharma, a form invisible and unresistant, gathered under the sphere of dharmas; it is neither mind nor a mental factor. It is precisely because it is adjudged a form that a challenge such as “the nonmanifest has no characteristic of form” arises (how can what has no characteristic of form count as form?), and precisely for that reason there is the great dispute in the latter half of this fascicle.

In practice: on reaching this point, set a question in your mind at once — “on what ground does a thing invisible, untouchable, and without resistance count as form?” That question is the fuse of the fascicle’s second half.

III. Bodily Manifest Action Is Shape, Not Motion

〔3.1〕The correct position of the Sarvāstivādins: the force of volition raises a bodily shape. The treatise first sets out the account of its own school: “by the force of volition, such and such a shape further arises in the body, and this is called bodily manifest action.” Puguang’s Record marks this sentence as “setting out the Sarvāstivādins and displaying the tenet of his own masters” — telling the reader plainly that this sentence is the author narrating the Sarvāstivādins’ own school, and is not language refuting another.

First ask: since bodily manifest action is a movement, why is its entity said to be “shape” and not “motion”? Because in the Sarvāstivādins’ scheme of dharma-characteristics, what the eye faculty can apprehend is only visible color (blue, yellow, red, white) and shape-form (long, short, square, round, and the rest); and if a movement is to have an entity, it can fall only into the square of shape-form. So the Sarvāstivādins say: as soon as volition sets things going, such and such shapes further arise in the body, and these shapes are bodily manifest action.

The easy error: to imagine “shape” as a process of change in physical form. What the treatise establishes is a shape at each single moment, not a stretch of continuous movement. In practice: this very point is the fuse laid for the refutation of “motion” in the next passage — continuous movement cannot serve as an entity precisely because dharmas have only moments.

〔3.2〕What the school that is refuted says. The treatise records the divergent account: “some other school says: motion is called bodily manifest action, since when the body moves it is by action that it moves.” This school maintains that the true bodily manifest action is “motion” itself — when the body moves, it is action that makes it move, and so motion is action. The three ancient commentaries agree entirely in their attribution here: Puguang’s Record marks it, “‘some other master says’ down to ‘not motion’ explains the second line. Some others, the Sāṃmitīyas, say: there is a separate moving form going from here to there, and that is called bodily manifest action”; Fabao’s Commentary marks it, “refuting the Sāṃmitīyas’ holding that motion is called bodily manifest action. ‘Since when the body moves it is by action that it moves’ is the Sāṃmitīyas’ reckoning. The body has no moments”; and Yuanhui’s Verse Commentary marks it, “‘action; not motion for its entity’: the Sāṃmitīyas say motion is called bodily manifest action, and in order to refute this it says ‘not motion.’”

First ask: how does this school dare to establish “motion”? Because they do not grant that all dharmas cease momentarily — Fabao’s Commentary puts a finger on it in one line: “the body has no moments.” If the body does not cease momentarily, then there is something that moves from this place to that, and that is motion. The easy error: to suppose this a mere quarrel over names. It is not — it is the opposition of two views of time: on one side, dharmas cease as soon as they arise and there is no room at all for movement; on the other, dharmas can persist, and so there is real movement.

In practice: whenever hereafter you meet a dispute over “does such and such really exist,” look first at whether the two sides agree in their view of “the moment”; where they do not, what is being argued is not that thing at all, but time.

〔3.3〕The author’s counterstroke is a single sentence: all conditioned things have moments. The author refutes with the utmost economy: “it is in order to refute this that it says ‘not motion,’ since all conditioned dharmas have moments.” He then defines the moment: “having obtained its own entity, it perishes with nothing intervening; a dharma that has this moment is called ‘having a moment,’ just as a man with a staff is called ‘having a staff’” — having obtained its own entity, it perishes without one thought-instant intervening, and a dharma possessing that nature is called “having a moment,” just as a man holding a staff is called a man having a staff. He presses one step further: “conditioned dharmas, as soon as they obtain their own entity, must from that point, with nothing intervening, perish into nonexistence”; and since “if it arises in this place it ceases in this place,” there is “no opportunity to shift from here to another quarter.”

First ask: what is that simile of the man with a staff saying? It is saying that “having a moment” is a designation carrying an attribute — a dharma carries the attribute “perishing as soon as it obtains its entity,” and so is called having a moment, just as a man carries a staff and so is called a man having a staff. The easy error: to read “moment” as a unit of extremely short time. Here it is not a stretch of time but a property.

In practice: this definition is the sentence most often cited back to throughout the fascicle; the treatments of the continuity of the nonmanifest and of the raising and the concurrent raising in the fascicle’s second half all presuppose it.

IV. Cessation Awaits No Cause — the Hardest Inference in the Fascicle

〔4.1〕The pivot sentence of the demonstration. That school retreats a step to save itself: “if all conditioned dharmas have moments, then the sense that they do not shift to another quarter can indeed be established” — first prove that “conditioned dharmas all have moments,” and then I will grant your conclusion. The proof the author gives is a single sentence: “that is, the cessation of conditioned dharmas awaits no cause. Why so? What awaits a cause is a result; cessation is a nonexistence, and a nonexistence is not a result, and so it awaits no cause.”

First ask: what is this sentence inferring? The inference is very short: what needs a cause is a “result”; and “cessation” is a kind of “nonexistence,” and a nonexistence is not a result; therefore cessation needs no cause. And since cessation needs no cause, there is nothing whatever to hold it back — and so it ceases as soon as it arises.

The author adds a further reverse inference: “if the first did not cease, the later ought to be the same, since the later and the first are equal in the nature of existing. And since there is a perishing later, one knows that the earlier was ceasing.” If the first thought-instant did not cease, the one after ought likewise not to cease, since the two are exactly alike in the single point of “existing”; and since later on there truly is a perishing, one can see that all along it was ceasing.

The easy error: to read “cessation is a nonexistence and a nonexistence is not a result” as “after cessation there is nothing at all.” What it says is that the event of ceasing is not itself a newly arisen thing, and so cause and effect cannot be laid over it. In practice: this is one of the very few stretches of pure inference in the whole treatise, and it must be followed sentence by sentence and not by the feel of the language.

〔4.2〕The difficulty from direct perception: when firewood meets fire, is it not plainly extinguished? That school brings out direct perception: “is it not plainly seen in the world that firewood and the rest come to cessation and nonexistence through conjunction with fire?” — and adds a formidable line: no means of knowledge surpasses direct perception, and so one cannot say that all cessation awaits no cause.

The author questions back: “how do you know that firewood and the rest cease through conjunction with fire?” The other answers: because after the firewood is conjoined with fire it is no longer seen. Whereupon the author brings out the key distinction of the whole passage: not seeing admits of two explanations — is it because the fire extinguished it, or because the firewood before ceased of itself and what follows does not arise, so that, there being nothing there, it is not seen? The text offers a ready pair of instances: “as with the conjunction of wind with a lamp flame and of a hand with a bell’s sound” — wind blows on the lamp, a hand stops the bell, and the flame and the sound no longer continue; yet nobody would say that the wind and the hand “extinguished” them; they merely no longer arise.

First ask: why is this distinction fatal? Because the other party’s direct perception can establish only “it is not seen afterward,” and cannot establish “the fire made it cease”; and “not seen” accommodates both explanations. Hence the author says: “therefore this sense must be made out by inference” — this matter can be settled only by reasoning and not by the eyes.

The easy error: to suppose the author is denying direct perception. He is not — he grants that you did indeed see that phenomenon, and only points out that your interpretation was not given you by direct perception. In practice: this is a discriminating method you can carry with you — tell apart “what I saw” and “my interpretation of what I saw.”

〔4.3〕Three alternative accounts, each refuted by the author. The opponent’s camp offers three further accounts, all of which the treatise records and refutes.

The first: “some hold that with a thought and a sound the earlier is the cause and the later ceases” — a thought and a sound being extinguished by the thought-instant that follows. The author refutes: “this too is unreasonable, since the two cannot be simultaneous,” and instances doubt and knowledge, pain and pleasure, greed and hatred: two dharmas whose own characteristics are opposed cannot be simultaneous; the earlier having ceased, the later then arises — so where is there anything to extinguish it? And there is a further pursuit that is very neat: “and by whom is the very last thought and sound extinguished?” The last thought-instant has no successor; by what should it cease?

The second and third: “some hold that the cessation of a lamp flame has the nonexistence of abiding for its cause” (taking the mark of abiding as the cause of cessation) and “some hold that when a flame ceases it is by the power of dharma and non-dharma” (taking the power of dharma and non-dharma as the cause of cessation). Puguang’s Record marks the provenance of the latter two very clearly: of the first, “this is the reckoning of the Elders and the Sāṃmitīyas. ‘Abiding’ means the mark of abiding”; of the second, “this is a divergent reckoning of the Vaiśeṣikas. Dharma and non-dharma are gathered under the category of quality”; and the item of thought and sound is likewise marked, “this is a divergent master of the Vaiśeṣikas.” The author’s refutation is “both are unreasonable, since there is nothing that is not a cause” — for by such a reckoning one might at will contrive such a cause upon any conditioned dharma whatever, and then anyone may say anything and there is no settled standard; and once that is granted, “this being so, the original dispute ought to be laid to rest, since it is granted that without awaiting any other cause all have moments” — the dispute ends of itself, for both sides must then grant momentariness without an external cause.

The easy error: to take the three accounts for trivial side-matters. They are in fact three answers to one and the same question: does cessation require an external cause or not? In practice: hereafter in this fascicle, whenever you meet “does such a dharma require something else to accomplish it?”, these three refutations can serve as your template.

〔4.4〕The three grades of cooking-change: forcing the “productive cause” and the “extinguishing cause” to be one and the same. The author strikes heavily once more: if the cessation of firewood has conjunction with fire for its cause, then in the process of gradual alteration under cooking there are the three grades, low, middle, and high; at the very moment when the middle and high grades of cooking arise, the low and middle grades are ceasing. One and the same conjunction with fire produces on one side and extinguishes on the other — “the entity of the productive cause would become the extinguishing cause.” And the author blocks the line of retreat as well: even granting that where the flames differ one might contrive to distinguish two causes, then “where ash, snow, vinegar, sun, water, and earth are conjoined and can make cooking-change arise in firewood and the rest, how will you contrive to reckon the productive and the extinguishing causes as different?” Ash, snow, vinegar, sun, water, and earth can all bring about cooking-change; is each of them to be fitted out with an extinguishing cause of its own?

The opponent’s last question is very concrete: “if that is so, when it is plainly seen that water boiled dwindles away, what does the conjunction of fire accomplish in it?” Water boiled to dryness — surely the fire did something. The author’s answer is superb: the conjunction with fire strengthens the fire element, and the strengthening of the fire element “makes the mass of water arise ever more slightly in each successive stage, until at the slightest it no longer continues” — fire attends only to “making what arises afterward less each time”; it has never “extinguished” anything.

First ask: why does this answer count as a victory? Because it gives “fire has a function” a complete place, while dispensing with the role of “extinguishing cause.” The easy error: to read it as “fire does nothing.” What the author says is that what fire performs is the work of producing (making the later instants ever slighter), not the work of extinguishing.

In practice: and so the whole passage closes — “therefore there is no cause that makes dharmas cease; dharmas cease of themselves, since it is their nature to perish,” “since they have moments, motion is definitely nonexistent,” and at last back to the topic: “since by this principle motion is definitely nonexistent, that bodily manifest action is shape is established in principle.” As for why people feel that something is moving, the author accounts for that too: like the flame of burning grass spreading along, it is merely “a conceit of superiority about traveling” — an illusion that there is a “going.”

V. Whether Shape-Form Really Exists — Eight Rounds between the Sautrāntikas and the Sarvāstivādins

〔5.1〕The author changes sides. The previous engagement over, the author turns about at once: “yet the Sautrāntikas say: shape is not really existent; that is, where visible color arises abundantly on one side of a mass, just therein a long form is conventionally established . . .” — long, short, square, and round are no more than four arrangements of visible color arising abundantly on one side, sparsely on one side, abundantly on all four sides, or pervading throughout, each given a name. The treatise offers an instance everyone has experienced: a torch swung swiftly in one direction is called “long”; whirled about in a circle, it is called “round” — “therefore shape has no real form-entity of a further kind.”

The three commentaries adjudge this passage identically: Puguang’s Record marks, “‘yet the Sautrāntikas say’ down to ‘a form-entity of a further kind’: from here the author sets out the Sautrāntika sense and refutes the Sarvāstivādins”; Fabao’s Commentary marks, “‘yet the Sautrāntikas say’ down to ‘to be understood as applicable’: the author sets out the Sautrāntika sense and refutes the Sarvāstivādins”; Yuanhui’s Verse Commentary marks, “this is the Sautrāntika school declaring its refutation of the Sarvāstivāda school’s really existent shape-form. Hence the treatise says: yet the Sautrāntika masters hold shape not to be really existent.”

First ask: what is likeliest to leave a reader dizzy here? The change of identity. In the previous engagement the author stood with the Sarvāstivādins to refute the Sāṃmitīyas; in this one he stands with the Sautrāntikas to refute the Sarvāstivādins. The easy error: to suppose the author contradicts himself. He does not — this is precisely the form of the Kośa, and the Life of Master Vasubandhu records his manner of composing it as “establishing the sense of the Sarvāstivādins and, wherever there was a partial position, refuting it with the sense of the Sautrāntikas”: taking the Sarvāstivādins as the ground, and where a thing will not hold, refuting it with Sautrāntika doctrine.

In practice: whenever hereafter you come to a stretch of dispute, ask first, “on which side is the author standing in this stretch?”; the three commentaries almost always mark it expressly, and are the most reliable of signposts.

〔5.2〕The difficulty of apprehension by two faculties — if shape really exists, it will be apprehended by two faculties. The Sautrāntikas’ first difficulty is a very clean reduction to absurdity: “if it be said that there really exists a further kind of shape-form, then one form ought to be apprehended by two faculties” — if long, short, square, and round have an entity besides, then on one and the same mass of form the eye sees its length and the body can also feel out its length, and one and the same form is apprehended by two faculties, the eye and the body; whereas by the Abhidharma scheme of characteristics, “in principle no form-sphere is apprehended by two faculties.” Puguang’s Record explains the line of the reasoning sentence by sentence, saying, “this explains the stanza’s eighth line, ‘since it would be apprehended by two faculties.’ The Sautrāntika masters say: your standing tenet . . .” and so on, pointing out that the Sautrāntikas are striking back with the Sarvāstivādins’ own standing tenet.

First ask: how do the Sarvāstivādins save themselves? They answer: what is felt is in fact only touch, not shape; it is only because touch and shape are in one mass and inseparable from each other that on apprehending touch one can recollect shape — just as seeing the color of fire one recollects its warmth, or smelling a flower’s scent one recollects its color. Puguang’s Record marks this as “‘touch and shape’ down to ‘can recollect the flower’s color’: the Sarvāstivādins save their position.”

The easy error: to read “recollect” as vague association. The Sarvāstivādins’ rescue has a strict condition — “the two dharmas here are definitely inseparable, and so by apprehending one the other can be recollected”: the two dharmas must be definitely inseparable. That very condition is where the next difficulty applies its leverage.

〔5.3〕The Sautrāntikas strike back and the Sarvāstivādins save themselves again. The Sautrāntikas fasten at once on that condition: “there is no touch definitely inseparable from shape; how then can apprehending touch recollect shape with certainty?” — touch and shape are not definitely inseparable (one and the same touch may go with different shapes); on what ground do you recollect shape with certainty from touch? Puguang’s Record marks this as “down to ‘can recollect shape’: the Sautrāntikas object.” The Sautrāntikas press one layer further: if the two are not definitely in one mass and yet shape can still be recollected from touch, then visible color too ought to be recollected with certainty from touch; or else shape-form too ought to be as indefinite as visible color, and then on apprehending touch one ought not to know shape — “but in fact it is not so.” And they add an instance: many shapes are seen upon brocade; are there then many real shapes in one place?

The Sautrāntikas change tack again and strike from the side of atoms: “again, whatever is a real form with resistance must have real atoms of a further kind of its own. Yet there is no atom called long and the rest” — every real form with resistance ought to have atoms of its own class, yet there has never been an atom called “long”; and so long and short can only be “conventionally established within the differing marks of many things arrayed in such and such a way.” Puguang’s Record marks it, “‘again, whatever there is’ down to ‘conventionally establishing long and the rest’: the Sautrāntikas object again.”

The Sarvāstivādins save themselves again: then let it be said that the atoms of shape-form arrayed in this way are called long and the rest. The Sautrāntikas cite this and refute it: “this is mere partisanship, since it is not thoroughly granted” — it is only a statement partial to one’s own school, and the two parties do not grant it in common; for shape-form has no atoms whose own characteristic is thoroughly granted, as visible color has, and whence then the “assembling and arraying”? Puguang’s Record marks this as “‘if it be said that just by’ down to ‘assembled and arrayed’: the Sautrāntikas cite the Sarvāstivādins’ rescue and refute it.”

The Sarvāstivādins offer a further instance from daily life: “is it not plainly seen that earthen vessels and the like have their visible appearance alike while their shapes differ” — the same color of clay, and yet the vessel-shapes differ (from which it appears that shape has an entity of its own). Puguang’s Record marks it, “‘is it not plainly seen’ down to ‘while their shapes differ’: the Sarvāstivādins save their position.” The Sautrāntikas answer beautifully: is this not exactly what I said — “as with many ants, whose appearance is no different, and yet arrayed in a line or in a ring their shape differs”; every ant in a swarm looks the same, and yet ranged in a line or in a ring the shape differs. Puguang’s Record marks it, “‘has it not already been made out’ down to ‘ought in principle to be so’: the Sautrāntikas refute again.”

The last difficulty is the dark and the distance: looking at a tree stump in the dark or from afar, one makes out only its shape and not its color, from which it appears that shape exists independently of visible color. The Sautrāntikas explain it through: precisely because in the dark and at a distance the color cannot be seen clearly, one gives rise only to the discrimination of long and short; just as looking at a crowd of people from afar, one knows only that it is a marching army and not the several particular marks — “because at times one does not know visible color and shape and knows only a general mass.” Puguang’s Record marks it, “‘knows only a general mass’: the Sautrāntikas explain it through. Because in the dark and at a distance one looks at the mass of visible color.”

First ask: what is to be remembered from these eight rounds? Remember the criterion, not who won: the whole engagement argues one sentence only — are long, short, square, and round “another kind of thing,” or “an arrangement of the same kind of thing”? The easy error: to suppose one ought to pick a side when the reading is done. The treatise passes no sentence here, and the words of both parties stand in the treatise together.

In practice: this passage is one of the finest models of debate in the whole treatise. Whenever hereafter you meet any dispute over “does such and such really exist,” you can measure it with the three tools used here: can it be reduced to absurdity (apprehension by two faculties), does it have atoms of its own (the test of real form), and can it be explained away by differences of array (the test of conventional establishment)?

〔5.4〕What the Sautrāntikas establish as bodily manifest action — a one-sentence disclosure. The author closes with a question: “since both motion and shape have been set aside, what do you of the Sautrāntika school establish as bodily manifest action?” Motion refuted and real shape refuted, what do you establish? The Sautrāntikas answer: “we establish shape as bodily manifest action, but as a convention and not as a real thing.” Puguang’s Record marks it here with the utmost clarity: the question is “since the Sāṃmitīya masters’ motion as bodily manifest action and the Sarvāstivādins’ real shape as bodily manifest action have both been set aside, what do you of the Sautrāntika school establish,” and the answer is “‘we establish shape as bodily manifest action, but as a convention and not as a real thing’: the Sautrāntikas answer.”

First ask: what kind of answer is this? It is “keep the name in use, refuse the entity” — the name shape is retained to serve as bodily manifest action, but only conventionally established. This is the standing method of the Sautrāntikas.

Ask a layer deeper: what then is bodily action? The Sautrāntikas answer: “if action established in dependence upon the body is called bodily action, then the volition that can move the body in its various ways, proceeding by the gate of the body, is called bodily action” — the real action is that volition which can make the body perform its various movements; because it proceeds in dependence upon the gate that is the body, it is called bodily action. Verbal action and mental action likewise. Yuanhui’s Verse Commentary restates it here: “in the Sautrāntika school shape is bodily manifest action, a convention and not a real thing. Shape, though it is the manifest, is not action” — separating “manifest” and “action” with perfect clarity.

The easy error: to take “the three actions have volition for their entity” as the author’s settled position. In this fascicle it is one of two parties recorded; the Vaibhāṣikas at once conclude by reverting to their own school: “the Vaibhāṣika masters say: shape is really existent, and so bodily manifest action has shape-form for its entity, and the entity of verbal manifest action is just vocal sound.” Both accounts stand in the treatise.

In practice: this parting of ways is the fuse for the fascicle’s second half — precisely because the Sautrāntikas gather action up into volition, the nonmanifest will be turned by them into “a continuum of the seeds of volition”; the whole of the next great dispute grows out of here.

〔5.5〕If manifest action is definitely nonexistent, what of the nonmanifest — the Sautrāntikas answer a round for themselves first. At the end of this passage the treatise sets up three further questions, each tighter than the last. The first: since shape is conventional, “if that is so, manifest action is definitely nonexistent”; and manifest action being nonexistent, the nonmanifest of the desire realm ought also to be absent (since the nonmanifest arises in dependence upon the manifest), and “this becomes a great fault.” The Sautrāntikas answer: “such a great fault we have a principle by which to fend off” — from the two spoken of above, bodily manifest and verbal manifest, there arises by a superior volition a certain difference of volition, and this is called the nonmanifest. The second question: what fault is there in this? “This ought rather to be called action that turns along with the mind, since the nonmanifest of absorption turns together with the mind” — since your nonmanifest arises from volition, it ought to turn together with the mind at the same time; how then does it differ from the nonmanifest that turns along with the mind produced by absorption? The Sautrāntikas answer: “there is no such fault, since it is produced by the superior volition that resolves and the superior volition that sets in motion” — it is drawn forth by the resolving volition and the setting-in-motion volition, and is not of that class which turns together with the mind. And they add a conceding line: even granting that there is manifest action, manifest action too must rely upon the force of volition, “since its own nature is dull.”

First ask: why is the second question keen? Because it seizes the Sautrāntikas’ soft place — if you gather the nonmanifest up into volition, it becomes very hard to say that it does not turn along with the mind. The easy error: to read the Sautrāntikas’ answer as equivocation. Their distinction is in fact very clear: “being drawn forth by volition” and “turning together with the mind” are two different things; the former is a sequence of cause and effect, the latter a simultaneous co-arising.

Look again at one sentence within the first question — “if that is so, why does the sūtra say that there are two kinds of action?” The Sautrāntikas take the opening to redefine the action that is volition and the action of having willed: what arises at the stage of preliminary effort is the “deliberating volition,” what arises once one has deliberated is the “volition that does the deed,” and “doing the deed to be done in accordance with what was deliberated before, moving the body and issuing speech, is called the action of having willed.” ⇒ The Sautrāntikas have not thrown away the sūtra text; they have only shifted where the text comes down, from “form” to “volition.” In practice: to judge whether a school is “denying outright,” look at whether it has thrown away the sūtra text; one that has merely shifted where it comes down is not denying outright.

VI. Whether the Nonmanifest Really Exists — the Vaibhāṣikas’ Eight Proofs

〔6.1〕The stanza has only ten characters, and yet it is a table of contents. “Because it says three, and says form without outflows, and says increase, and says not doing, and so on” — it says three kinds of form, says form without outflows, says that (meritorious action) increases, says that (having another do a deed and) not doing it, and so on. First ask: why is the stanza so short? Because it is a list of evidence, not the argument itself; every word of it points to a passage of sūtra. Puguang’s Record, marking the Sautrāntikas’ refutations, numbers them one by one, saying “below, the Sautrāntikas refute the eight foregoing proofs; this is the refutation of the first proof”; and Yuanhui’s Verse Commentary likewise has “the eight proofs above prove the nonmanifest real. From here the Sautrāntikas explain the foregoing through” — all three count eight proofs.

The eight, set out, are these: ① the sūtra says form is of three kinds, and the third has form yet is invisible and unresistant — if it is not the nonmanifest, what is it? ② the sūtra says there is form without outflows — if it is not the nonmanifest, what form can be without outflows? ③ the sūtra says meritorious action increases, and it goes on increasing even when another mind has arisen or when there is no mind — without the nonmanifest, by what does it increase? ④ one dispatches another to do a deed and does not do it oneself, and yet a path of action is accomplished — without the nonmanifest, by what is it accomplished? ⑤ the sūtra says that the dharma sphere is invisible and unresistant and does not say “without form” — from which it appears that there is form within the dharma sphere. ⑥ without the nonmanifest, there being no speech and the rest in absorption, the eight limbs of the noble path fall short. ⑦ without the nonmanifest, after the precepts are received there is no entity of discipline continuing on, and on what ground is he still called a monk? ⑧ the sūtra says that the discipline of abstaining from killing and the rest is called “the discipline that is a dyke,” able to dam and hold back transgression for a long time — “a thing without entity cannot be called a dyke.”

The easy error: to read these eight as eight independent theses. They are in fact eight applications of one and the same method of inference: there is a slot standing empty in the sūtras, and only the nonmanifest will fit it. In practice: whenever hereafter you see an argument that “such a dharma must really exist,” look first whether it is of this type; if it is, there is only one way of refutation — find another thing that fits the slot as well.

〔6.2〕The author’s closing words and the sentence that lays the dispute to rest. The Vaibhāṣikas conclude: “by these proofs it is known that nonmanifest form really exists.” And after the whole train of refutation has ended, the author lays the dispute to rest with his own hand: “let these many disputes rest here for now”; whereupon the Vaibhāṣika masters conclude by reverting to their own school: “the Vaibhāṣika masters say: there is a real thing called nonmanifest form, and this is the tenet of my school.”

First ask: who says these two sentences, and what is the difference? The first is the author’s laying the dispute aside; the second is the Vaibhāṣikas declaring their own position. The treatise closes the passage with this and adds no sentence of judgment — and this is of the utmost importance: it is the standard form of “recording both parties without adjudication.”

The easy error: to read “this is the tenet of my school” as the author’s own confession. It is the Vaibhāṣikas’ confession; Puguang’s Record marks it here, “‘the Vaibhāṣika masters’ down to ‘the tenet of my school’: the Sarvāstivādins conclude by reverting to their own school.” In practice: when you meet the first person in reading a treatise, first make sure who is speaking — within this fascicle the word “I” changes owners at least twice.

VII. The Sautrāntikas Refute the Eight Proofs One by One

〔7.1〕The blanket denial, and the first pattern of refutation: find another thing to fill the slot. The Sautrāntikas’ blanket denial is a single sentence: “the Sautrāntika masters say: these proofs, though many and of all kinds of marvelousness, are nevertheless unreasonable.” Fabao’s Commentary marks it, “on the treatise’s ‘the Sautrāntika masters say’ down to ‘nevertheless unreasonable’: the Sautrāntikas deny the foregoing proofs — that there is a separate real entity of the nonmanifest is unreasonable.”

The method of refuting the first proof is just the “find another thing to fill the slot” spoken of above: the third kind of form, invisible and unresistant, may be the form in the object-field of absorption obtained by cultivating absorption — “the yogins say: when one cultivates the meditative absorptions, the form in the object-field of absorption produced by the force of absorption”; it is invisible because it is not the object-field of the eye faculty, and unresistant because it does not obstruct a place. Yuanhui’s Verse Commentary agrees entirely here, giving “as for the three forms, they are the form in the object-field of absorption and not nonmanifest form.”

The Sarvāstivādins at once turn and challenge: since it is invisible and unresistant, on what ground is it still called “form”? The Sautrāntikas answer very neatly: the explaining through of this difficulty is “the same as with the nonmanifest” — we are under one difficulty together; answer me as you answer yourself.

First ask: does this manner of answering, “the same difficulty,” count as evasion? It does not. It points out that the other party’s difficulty is a double-edged blade: if you can explain why the nonmanifest is form, I shall explain the form in the object-field of absorption with the same explanation. The easy error: to suppose that the Sautrāntikas have granted that the form in the object-field of absorption really exists. They say only that “this slot need not be filled by the nonmanifest alone.”

In practice: to refute an argument of the “only explanation” type, the standard technique is to supply a second explanation — one need not prove it better; it is enough to prove that it holds just as well, and the original argument’s claim of uniqueness lapses.

〔7.2〕The section on form without outflows: three accounts side by side and a short exchange. In refuting the second proof, the Sautrāntikas bring out first the yogins’ explanation: “within the form produced by the force of absorption, that produced in dependence upon an absorption without outflows is said to be without outflows.” Another master gives a divergent explanation — that the form on the body of one beyond training and all outer form are without outflows, since they are not what the outflows depend upon for their increase; Puguang’s Record marks this as “‘another master says’ down to ‘obtain the name without outflows’: setting out a divergent explanation. Some other master of similes says.”

Whereupon there is a short and dense exchange. The Sarvāstivādins object: that being so, why does the sūtra say “as for dharmas with outflows, they are all the eye whatever, and so on at length” (speaking of form as with outflows throughout)? Puguang’s Record marks it, “the Sarvāstivādins object. Why does the sūtra say that dharmas with outflows are the fifteen elements.” The masters of similes answer: that is said from the side of its not being the counteragent of the outflows. The Sarvāstivādins press again: then one and the same thing is both with outflows and without outflows. The masters of similes ask back, “if so, what fault is there?” The Sarvāstivādins answer: the fault of mixture. The masters of similes answer: when it is called with outflows with regard to this principle, it was not with regard to this principle called without outflows, and conversely likewise; each is said with regard to one sense, and where is the mixture?

First ask: where is the technical point of this short passage? In “one thing may receive different names with regard to different senses.” The easy error: to read “mixture” as logical contradiction. It means disorder in the system of names and characteristics; and the masters of similes’ answer is precisely that so long as one marks with regard to which sense a thing is said, the system is not disordered.

In practice: this is Abhidharma’s general solution for “multiple adjudications of a single dharma”; further on, in treating how one dharma extends over several natures and several realms, the treatise resorts to this move again and again.

〔7.3〕The section on the increase of merit: the seeds of volition make their first formal entrance. In refuting the third proof, the Sautrāntikas bring out the explanation of the masters of the norm of former times: meritorious action increases by the force of the very nature of things — what the donor gives is enjoyed by the recipient, the merit gathered and benefited differs, and so afterward, though the donor’s mind take some other thing for its object, “the subtle continuum perfumed by that earlier volition which took the giving for its object goes on gradually transforming and differentiating as it arises, and by this it can in time to come draw forth many fruits.” Puguang’s Record marks it, “below, the refutation of the third proof. As for the increase of merit, it is the explanation of the Sautrāntika masters of the norm of former times.”

This passage is the fullest statement of the Sautrāntika “seed” doctrine in this fascicle, and three words are to be seized in reading it: perfuming (the earlier volition leaves a trace), continuity (the trace is handed on instant by instant), and transformation and differentiation (there is change in the handing on, and so a new fruit can be drawn forth).

The Sarvāstivādins ask back: how should the differences of merit in another (the recipient) produce a transformation in this continuum of mine? The Sautrāntikas answer: this difficulty is “the same as with the nonmanifest” — you cannot make it out either. And they strike back with something heavier: “and how should he, by differences of merit and benefit in another continuum, cause a really existent nonmanifest dharma to arise besides within another continuum?” On my side there is only a continuum undergoing change; on yours a really existent dharma must arise out of nothing — the difficulty is yours, not mine. Puguang’s Record marks this as “‘if it be asked how’ down to ‘a nonmanifest dharma arises’: the Sautrāntikas cite the difficulty and question it back.”

Then comes the question and answer on merit without a support. Question: with no recipient to give to, how does merit increase? The Sautrāntikas answer: also by the volition that repeatedly practices taking that object-field for its object, “so that even in dreams it turns along at all times.” And then the Sautrāntikas object in return to the Sarvāstivādins: “you proponents of the nonmanifest, since in merit without a support you have no manifest action, whence should you have a nonmanifest?” With no bodily or verbal movement, where does your nonmanifest come from?

The author at last cites a sūtra to refute a divergent master who has gone too far (the one who maintains that even merit with a support rests on repeatedly practiced volition alone): the sūtra says that after a monk has received what was given he enters the immeasurable states of mind, and the donor’s merit at that time increases at all times — “but is there necessarily always a superior volition taking that matter for its object?” The donor is not necessarily giving rise to that volition at every moment. And so one must come back to “the subtle continuum perfumed by volition goes on gradually transforming and differentiating as it arises, and this is decidedly the reasonable account.” Puguang’s Record marks this as “‘if that were so, the sūtra says’ down to ‘decidedly the reasonable account’: the author cites a sūtra to refute the divergent master’s account.”

First ask: whom is the author striking here? He is striking a wayward branch within the Sautrāntikas, not the Sarvāstivādins. The easy error: to read this passage as the author changing sides. In practice: in reading passages of dispute, keep track at every moment of “which way is this arrow flying?”; within this fascicle the author looses more than one arrow at his own side.

〔7.4〕The section on having another act: the path of action and that Venerable. In refuting the fourth proof, the Sautrāntikas give the correct account: by the original preparatory effort, when the messenger has done the deed as instructed, “by the very nature of things the subtle continuum on the instructor’s side is made to arise transforming and differentiating”; and “it should be known that just this subtle continuum transforming and differentiating is what is called the path of action” — this being a conventional establishing of the name of the cause upon the result, since it is the result drawn forth by bodily and verbal action. Puguang’s Record marks it, “below, the refutation of the fourth proof.”

The treatise sets out in passing the divergent account of the Venerable: that within the appropriated aggregates one is touched by the sin of killing by giving rise to volition at the three times, “I shall kill,” “I am killing,” and “I have killed.” Puguang’s Record gives his name, “the Venerable means Dharmatrāta,” and Fabao’s Commentary likewise gives, “on the treatise’s ‘yet the Venerable says’ down to ‘I am killing, I have killed’: setting out a divergent reckoning. This is the reckoning of Dharmatrāta” — the two commentaries agree entirely in the attribution. The author’s treatment of it refutes in part and accepts in part: if merely giving rise to volition at the three times completed the path of action, then when one’s mother has in fact not been harmed but one supposed oneself to have harmed her, a deed of immediate retribution would be accomplished, which is wrong; yet if these three volitions arise in a matter of one’s own doing that is not a killing by mistake, “if it is said in this way, it is after all not unreasonable.”

First ask: is this treatment of “refuting one half and accepting the other” common? In this treatise it is very common, and it tries the reader most of all. The easy error: to see “not unreasonable” and take it for the author’s wholesale acceptance. He accepts only the version that has had a qualifying condition added to it.

In practice: when reading ancient treatises and meeting double negatives (“not unreasonable,” “not without sense”), stop and see clearly what has been qualified; the qualifier is where the judgment lies.

〔7.5〕The sentence that sets the key: alike hard to know, and no grudge borne. The Sarvāstivādins here voice the most emotional question in the whole fascicle: “why do you bear a grudge against the nonmanifest in particular and definitely deny it as nonexistent, while granting a perfumed subtle continuum transforming and differentiating?” — both being equally invisible, on what ground do you grant seeds and not the nonmanifest? Puguang’s Record marks this as “‘why against the nonmanifest’ down to ‘transforming and differentiating’: the Sarvāstivādins complain.”

The author’s answer falls into two layers. The first is a posture: “this and that are alike hard to know, and here I bear no grudge against either” — both sides are hard to know, and I have no ill will. Only the second layer is the reason: since you grant that a path of action is of the class of mind, and yet would say that when the matter is brought to completion by bodily preparatory effort a further nonmanifest dharma arises within the instructor’s body apart from mind and body — “a tenet such as this cannot make one glad”; whereas if it be said that it is just this continuum of mind that arises transforming and differentiating by force of that, “a tenet such as this can make one glad,” since by the transforming and differentiating of the continuum of mind and the rest alone a future fruit can be produced.

First ask: what exactly is the author refusing? Not “things that cannot be seen,” but the step of “setting up a new dharma apart from mind and body.” The easy error: to read it as the author’s opposing all invisible dharmas.

In practice: this is the passage in the whole fascicle from which most can be learned — a debate may be as keen as you like while the heart bears no grudge. The author first declares that he bears no ill will, and then presses his reasons to the end; posture and reason are kept apart, and neither is discounted.

〔7.6〕The refutation of the remaining four proofs: form in the dharma sphere, the eight limbs of the path, individual liberation, and the dyke. The refutation of the fifth proof is very short: the reason the dharma sphere is not said to be “without form” is that there is the invisible and unresistant form gathered under the dharma sphere in the object-field of absorption spoken of above, and it need not be the nonmanifest. Puguang’s Record marks it, “‘again, as to the dharma sphere’ down to ‘form gathered under the dharma sphere’: this refutes the fifth proof.”

The refutation of the sixth proof is a complete exchange. The Sautrāntikas first make the Sarvāstivādins speak for themselves: while one is in the path, whence right speech, right action, and right livelihood? Is there truly a raising of words, a doing of deeds, a begging for robes? The Sarvāstivādins answer: there is not; it is that at that time a nonmanifest of a class without outflows is obtained, and after emerging from contemplation one is able by the force of what went before to give rise to the three right ones and not the three wrong — “since the name of the result is conventionally established upon the cause.” The Sautrāntikas take it up at once: “if that is so, why do you not accept this sense?” — the name of the result being likewise established upon the cause, I say that what is obtained in the path is a “resolve and support” and not a nonmanifest, and the eight limbs of the noble path can be established just as well; why will you not accept it? Puguang’s Record marks it, “‘if that is so, how’ down to ‘the eight limbs of the noble path’: the Sautrāntika masters say”; and Fabao’s Commentary marks the same passage with “on the treatise’s ‘let them first say’ down to ‘begging for a support and the rest, or not’: the Sautrāntikas ask the Sarvāstivādins in return” and “on the treatise’s ‘there is not’: the Sarvāstivādins answer,” the two commentaries agreeing in attribution. The treatise sets out besides the account of a divergent Sautrāntika master: that it is merely a matter of saying “the not doing of wrong speech and the rest” and taking that as a limb of the path, with the instance that among the eight worldly dharmas “not gaining robes and food” is not a further real entity; Puguang’s Record marks it, “‘another master says’ down to ‘this too ought to be so’: setting out a divergent Sautrāntika account.”

The refutations of the seventh and eighth proofs are taken together: the discipline of individual liberation is a matter of setting up a limit undertaken by the force of a resolve of intention, so as to hold off the unwholesome actions of body and speech; and to the Sarvāstivādins’ earlier objection that “once a mind of another object has arisen there ought to be no discipline,” the Sautrāntikas explain it through: “since by the force of perfuming, at the time when a fault is about to arise a single act of recollection stops it.” Puguang’s Record marks it, “‘if a mind of another object arises’ down to ‘a recollection stops it’: citing the Sarvāstivādins’ earlier objection and explaining it through.” The discipline that is a dyke runs the same way — one first sets up a vowed limit definitely not to do what is unwholesome, and afterward, by repeated recollection and by conscience and shame being present, restrains and upholds oneself, “and so the sense of a dyke is upheld by the mind.” At the end the author reproaches them in one line: “if transgression of the discipline were held off by a nonmanifest, there ought to be nobody who transgresses through loss of mindfulness” — if a nonmanifest were truly holding the door, how should anyone lose mindfulness and break the precepts? Puguang’s Record marks this as “this refutes the eighth proof.”

First ask: have these eight refutations a common skeleton? They have — wherever the Sarvāstivādins would set up a real dharma, the Sautrāntikas put in a function of the continuum of mind. The easy error: to suppose the Sautrāntikas deny that discipline exists. What they deny is only that “discipline is an entity of the class of form”; of discipline’s function, its continuity, and its holding off of wrongdoing they cancel not one item.

In practice: this whole stretch gives the reader a most serviceable measure — tell apart “does this function exist?” and “does this function require an entity to carry it?” Many disputes in the whole treatise, taken apart to the bottom, are just this one question.

VIII. The Great Elements That Produce the Nonmanifest Differ from Those upon Which the Manifest Depends

〔8.1〕One line of stanza, one reason. “These great elements that produce it differ from those upon which the manifest depends” — the great elements that produce the nonmanifest are not the same set as those upon which manifest action depends. The reason is a single sentence: “the nonmanifest arises from great elements different from those of the manifest. Why so? Because for one combination to have a subtle result and a coarse result is unreasonable.” For one and the same combination of the four great elements to produce at once a subtle result (the nonmanifest) and a coarse one (the manifest) is unreasonable. Puguang’s Record marks the outline of this passage as “‘it was said above that the nonmanifest’ down to ‘or is there a difference’: below, the second, explaining what produces it.”

First ask: why can the subtle and the coarse not have one source? Because in the Sarvāstivādins’ causal scheme, the results produced by one set of four great elements ought to be of equal grade; for one and the same set of causes to produce two results of vastly different class would throw the marks of cause and effect into disorder. The easy error: to suppose this is saying that “the nonmanifest is higher than the manifest.” It is not — subtle and coarse here are the fineness and grossness of resistance, not a matter of value.

In practice: this item looks trifling, and yet it is the ground floor of the two later lines of stanza, “arising from different great elements” and “not from different great elements”; those two lines are precisely counting how many sets of these great elements there are.

IX. The Nonmanifest of the Desire Realm from the Second Thought-Instant On Depends upon Past Great Elements

〔9.1〕Why past great elements have to be brought in. “In the desire realm the nonmanifest thereafter arises in dependence upon past great elements” — only for the nonmanifest affiliated to the desire realm, from after the first moment on, does it all arise from past great elements. First ask: why is it arranged so? Because the nonmanifest of the desire realm must be carried on by the great elements on which the very first moment of manifest action depended; that set of great elements passes in a single instant, and if the nonmanifest thereafter is still to have a support, it can only depend upon the set that has passed.

The text accounts for the role of the great elements upon the present body: “the great elements of the present body can serve only as a support, and as the raising cause and the concurrently raising cause each in its own order” — the present great elements serve only as a “support,” and serve in order as the raising cause and the concurrently raising cause. And it gives a very vivid simile: “as a wheel runs upon the ground, with the hand and the ground for its support” — the wheel can run, the force of the hand that pushes being one kind of support (like the raising cause) and the force of the ground that bears it another (like the concurrently raising cause).

The easy error: to read “depends upon past great elements” as “the nonmanifest resides in the past.” The nonmanifest itself is still a present or future dharma; only its support is taken from the past. In practice: this item brings in the pair of words “raising / concurrently raising” for the first time — the same pair will be used again in the fascicle’s closing passage on the raising of action by mind, and the two places should be read together.

X. Action with Outflows Depends on the Great Elements of Its Own Level, Action without Outflows Follows the Place of Birth

〔10.1〕Three reasons hold up one line of stanza. “With outflows depends on its own level; without outflows follows the place of birth.” Bodily and verbal action with outflows of the desire realm is produced by desire-realm great elements, and so on up to the fourth meditative absorption, whose action is produced by the great elements of that level — realm and level being of a piece. Bodily and verbal action without outflows has no level of its own, and can only be produced by the great elements of whatever level the practitioner is presently born in.

The treatise gives three reasons: “because dharmas without outflows do not fall within the realms; because there are definitely no great elements that are without outflows; and because it is by the force of their support that what is without outflows arises.” First ask: what does each of the three govern? The first says that what is without outflows does not belong to the three realms, and so has no “own level” to speak of; the second says that no great element is itself without outflows, so that the producer of action without outflows can only be great elements with outflows; the third says that what enables what is without outflows to arise is precisely the force of its support.

The easy error: to suppose that “action without outflows is produced by great elements with outflows” is a contradiction. It is not — the producer and the produced may be of different natures, and this same item will be used again later in the fascicle in treating that the great elements are not abandoned by seeing.

In practice: this is an excellent point of self-examination — whenever you read that “a certain dharma is produced by a certain dharma,” ask separately once about the nature of the producer and the nature of the produced.

XI. The Class of the Nonmanifest and the Manifest: Appropriated, Outflowing, of the Class of Sentient Beings, Unconcentrated and Concentrated

〔11.1〕First pin down a few technical terms. The stanza gives seven or eight adjudications at one breath, and three words must be set up before reading it: appropriated (is it that portion which consciousness appropriates as its own entity and which can give rise to pain and pleasure?), homogeneous (is it produced by a homogeneous cause?), and of the class of sentient beings (does it belong to a sentient being?).

The treatise says: the nonmanifest is unappropriated, “since it has no alterability and resistance”; it is also of homogeneous nature, and the word “also” in the stanza carries a further account — “the word ‘also’ shows that a portion of it is momentary, namely the first without outflows,” the first thought-instant of nonmanifest without outflows being not homogeneous but momentary; the rest are “all of homogeneous nature, that is, produced by a homogeneous cause.” As for the class of sentient beings, “it belongs to sentient beings alone, since it arises in dependence upon the inner body.”

First ask: why must that word “also” be explained specially? Because the stanza is extremely economical, and the addition or removal of one word admits or excludes a whole class of dharmas. Puguang’s Record commonly supplies instances at such places, and Fabao’s Commentary likewise marks the outline of this passage as “on the treatise’s ‘these manifest and nonmanifest, what is their class,’ the two stanzas below are the third main division, explaining the appropriated and the classes.” The easy error: to see “also homogeneous” and take it that all of it is homogeneous; the first thought-instant of what is without outflows is precisely the exception.

In practice: in reading the stanzas one should form a habit — whenever you meet the words “also,” “only,” “all,” or “the rest,” stop and ask what it is taking in and what it is letting go.

〔11.2〕The unconcentrated nonmanifest and the nonmanifest of absorption: the one several, the other shared. The nonmanifest of the desire realm is “homogeneous, appropriated, arising from different great elements”; and as to “different great elements,” the treatise explains itself: “the words ‘arising from different great elements’ show that the seven limbs of body and speech are each produced by separate great elements” — three of body and four of speech, seven limbs in all, each with its own set of four great elements. The nonmanifest produced in absorption is the reverse: the discipline of the meditative absorptions and the discipline without outflows “both depend upon what is nourished by absorption,” being unappropriated and not produced from different great elements; the treatise explains itself: “the words ‘not from different great elements’ show that this nonmanifest has its seven limbs produced from one and the same set of four great elements,” the reason being that “the great elements on which it depends are, like the mind, but one and admit of no difference.”

First ask: why must the unconcentrated levels be several and the level of absorption shared? Because the nonmanifest of the unconcentrated levels arises along with several distinct bodily and verbal movements, and the movements being several, their supports are naturally several; while the nonmanifest of the level of absorption arises along with the mind, and the mind of absorption being one, its support is one as well. The easy error: to read “the seven limbs from one and the same set of four great elements” as the seven limbs merging into one. They do not — the seven limbs are still seven; they merely share one support.

In practice: this pair of contrasts yields a rule that can be extended — the number of supports follows the number of what raises them. One mind, one support; seven movements, seven supports.

〔11.3〕Does manifest action swell the body — a very concrete pursuit. The treatise inserts here a most homely question and answer: when manifest action arises, must the original bodily measure be destroyed? If it must, then the maturational form already cut off could be continued anew, contrary to the Vaibhāṣika tenet; if it need not, how should one place contain two bodily measures? The answer is: there are further newly arisen homogeneous great elements which produce manifest action without destroying the original body. The pursuit is not over: then the place where manifest action arises ought to be larger than before. The treatise’s answer is only eight characters — “there are interstices in the body, and so they can be contained.”

First ask: this exchange looks trifling; why is it kept? Because it is a demonstration of pressing an abstract characteristic down into the concrete: having said that manifest action is produced by newly arisen great elements, one must answer “where is it put?” The easy error: to take “interstices” for anatomy. It is a “place of containing” in the scheme of dharma-characteristics.

In practice: in reading treatises one may often learn this move — push an abstract thesis down to the most concrete scene and ask it once; whether it can answer is a good way of testing whether it is thorough.

XII. The Three Natures and Realm and Level: The Nonmanifest Extends to Two Natures, the Manifest to Three

〔12.1〕Why the nonmanifest has no indeterminate. “The nonmanifest extends only to the wholesome and unwholesome natures and has no indeterminate,” the reason being: “because the force of an indeterminate mind is feeble and inferior and cannot draw forth a strong action and make it arise, such that after the cause has ceased the result may still continue to arise.” First ask: what is this reason saying? The mark of the nonmanifest is that “the cause has ceased and the result still continues” — the movement is past, and that portion is still there. For such continuing force there must be a mind strong enough to set it going; and an indeterminate mind is too weak to set it going.

The easy error: to read “an indeterminate mind is weak” as “the indeterminate is unimportant.” Here it is called weak only as regards this one function of drawing forth a strong action. In practice: this item yields a measure for weighing actions — to see whether an action has continuing force, look first at whether the mind that set it going was strong.

〔12.2〕The manifest and volition extend to all three natures; the unwholesome only in the desire realm; the nonmanifest pervades desire and form. “What is called the rest refers to the manifest and to volition,” and these two extend to the three natures, wholesome, unwholesome, and indeterminate. The unwholesome nonmanifest is only in the desire realm, since in the two higher realms “the unwholesome roots and non-conscience and shamelessness have already been cut off.” The nonmanifest pervades the two realms of desire and form, “since within the formless realm there are no great elements” — the formless realm has no great elements, and the nonmanifest being produced from great elements, there naturally is none. And the treatise adds a general rule: “wherever bodily and verbal action operates, there and there alone is there discipline of body and speech.”

Someone at once poses a difficulty: then one whose body is in the desire or form realm and who has entered a formless absorption ought to have discipline (just as, when a mind without outflows arises, there is a nonmanifest without outflows). The author answers: he will not — “since that does not fall within the realms”; and moreover, if the formless realm had a nonmanifest, there ought to be a nonmanifest not produced from great elements; nor may the great elements of another realm and level be taken as its support; and besides, entering a formless absorption is precisely turning away from and abandoning the forms, “and so within that absorption form cannot be produced, since that absorption suppresses the perception of form.” The Vaibhāṣika masters add another road: discipline arises in order to counteract bad conduct, and bad conduct exists only in the desire realm, while the formless realm has toward the desire realm the four kinds of remoteness (remoteness of support, of mode of activity, of object, and of counteragent), and so within the formless realm there is no nonmanifest form.

First ask: why are so many reasons given here? Because this is a boundary question across realms, and the author blocks all four roads at once — realm, support, object, and counteragent. The easy error: to push “there is no nonmanifest in a formless absorption” into “there is no discipline in the formless realm.” What the treatise says is that the formless realm has no nonmanifest form, not that those there have no virtue.

In practice: whenever you meet an adjudication about the boundary of a realm or level, you may question yourself along these four roads: does it belong to this realm, has it a support, can it reach the object, and does it require a counteragent?

〔12.3〕Manifest form is only in the two levels with sustained thought; the desire realm has no obscured-indeterminate manifest. “Manifest form is only in the two levels that have sustained thought” — the desire realm and the first meditative absorption; above that there is no manifest. “Obscured-indeterminate manifest is definitely absent in the desire realm, and can be said to exist only in the Brahmā world,” and the treatise adduces an old story in proof: “it was once heard that Mahābrahmā uttered words of deceit and flattery” — Mahābrahmā, within his own assembly, dissembling and praising himself in order to evade the challenging questions of the monk Aśvajit.

Someone asks: since the higher levels have no speech, whence a sound sphere? Answer: “there are outer great elements which serve as cause and give rise to sound.” Another master says that the upper three meditative absorptions too have unobscured-indeterminate manifest action, and the author adjudicates, “the former account is the good one.” As for why the desire realm has no obscured-indeterminate manifest, the reason is “because there is no raising mind able to set action going” — only a mind with applied and sustained thought can set the manifest going, and the second absorption and above have no such mind; and further, “a mind that sets the manifest going is only such as is abandoned by cultivation, since the afflictions abandoned by seeing turn toward the inner gate,” while the obscured-indeterminate of the desire realm is abandoned by seeing alone, turning inward and not issuing in bodily or verbal action.

First ask: what is the axis of this whole train of adjudications? It is the mind that sets action going. Whether there is a manifest depends on whether there is a mind able to set it going. The easy error: to misread “the desire realm has no obscured-indeterminate manifest” as “the desire realm has no obscured-indeterminate.” The desire realm has obscured-indeterminate afflictions; only that class of affliction does not set the manifest going.

In practice: here for the first time the fascicle expressly puts forward the role of “the mind that sets action going”; it is the forerunner of the words “what raises” in the next stanza, and the two passages should be read continuously.

XIII. The Four Kinds of Wholesome, the Four Kinds of Unwholesome, and the Two Constant Indeterminates

〔13.1〕Four measures for the one word “wholesome.” The transitional question is put very bluntly: is it by what raises them alone that dharmas are made wholesome and unwholesome? Answer: it is not — “it is by four kinds of cause that they come to be of wholesome nature and the rest: first by the ultimate, second by their own nature, third by association, fourth by what raises them.”

Each of the four has its definition and its simile, and the text gives them with perfect regularity: the ultimate wholesome is true liberation, “since within nirvāṇa there is the utmost security and all sufferings are forever stilled, as when one is free of illness”; the intrinsically wholesome are conscience, shame, and the three wholesome roots, which “need not await association nor anything else to raise them, but are in their nature wholesome, like a good medicine”; the associated wholesome are the mind and mental factors associated with them, likened to “water mixed with medicine”; the wholesome by what raises them are bodily and verbal action and the dissociated formations, likened to “the milk drawn forth by the juice of a good medicine.” The four unwholesome answer to them one by one: the ultimate unwholesome is the dharmas of birth and death, “like a chronic disease”; the intrinsically unwholesome are non-conscience, shamelessness, and the three unwholesome roots, “like a poison”; the associated unwholesome are “like water mixed with poison”; and the unwholesome by what raises them are “like the milk drawn forth by the juice of a poison.” Puguang’s Record marks the outline of this passage as “‘the treatise says’ down to ‘as when free of illness’ explains the first line, showing the ultimate wholesome,” and in the section on the intrinsic marks, “these five, being strong and superior, are called intrinsic.”

First ask: why are the four similes ranged so regularly? Because they are deliberately ranged as a descending chain: freedom from illness (secure of itself) → good medicine (good in itself) → medicinal water (called medicine because medicine is mixed in) → milk drawn forth by the medicine’s juice (sweet because of the medicine). Wholesomeness is “borrowed” down one layer at a time, and the further along, the more it depends upon what precedes. The easy error: to read the four as four parallel kinds of wholesome. They are a chain of transmission, not four squares side by side.

In practice: this chain is the general outline of value judgment in the whole treatise. Whenever hereafter you judge a dharma wholesome or unwholesome, you may ask: is it wholesome of itself, or wholesome by association, or wholesome by what raises it?

〔13.2〕Two places of “should be pondered” — blanks the author leaves himself. The treatise leaves two questions unfinished in this section. The first: “if it be a case of obtainment and the rest raised by a mind of another class, how does it become wholesome?” How should the dissociated formations such as obtainment, raised by a defiled mind, count as wholesome? The author answers only, “this sense should be pondered.” The second: how should the nonmanifest that turns along with the mind in absorption become wholesome, being neither drawn forth by an act of attention while in absorption nor set going by the preparatory effort of an unconcentrated mind — or else the divine eye and divine ear would have to become wholesome in nature? The author answers, “upon a sense such as this one should expend some labor.”

First ask: why does the author not answer? Because these two places touch the boundary of the item “wholesome by what raises them”: the wholesome by what raises them rests upon the mind that can raise, and in these two instances the mind that raises and the dharma raised do not agree in class. The easy error: to read “should be pondered” as the author’s not knowing. It is the formulaic language for marking expressly a place awaiting study, and it is honesty and not vagueness.

In practice: when in reading ancient treatises you see “should be pondered” or “one should expend some labor,” note them down as signposts — they are the hard places the treatise itself has marked, and they are often just where the treatise masters of later ages spent most of their strength.

〔13.3〕If the ultimate unwholesome holds, is not everything with outflows unwholesome? Someone at once seizes the gap: if the dharmas of birth and death are all ultimate unwholesome, then no dharma with outflows is indeterminate or wholesome. The author’s solution is very fine: “if the matter be argued from the ultimate, it is indeed as you say; yet here it is spoken of with regard to maturation” — from the ultimate it is indeed so, but here the speaking is with regard to maturation: “of dharmas with outflows whatever, those that cannot record a maturational result are given the name indeterminate,” and those that can record a lovely maturation are said to be wholesome. As for the ultimate indeterminate, “it refers to the two unconditioned, since empty space and cessation independent of discernment are of indeterminate nature alone and there is no other gate.”

First ask: what kind of solution is this? It is changing the measure: the ultimate is one measure and maturation another, and one and the same set of dharmas may receive different names under the two. The easy error: to suppose the author contradicts himself. He does not; he has merely changed the criterion — and he says expressly that he has changed it.

In practice: this is a repetition of the move of “three actions, three measures” in the first section. Instances of “one set of dharmas, different criteria, different conclusions” are extremely numerous throughout the treatise; see the measure clearly and the numbers will not fall into confusion.

〔13.4〕Why the great elements do not become wholesome and unwholesome by what raises them. The treatise poses a further question: if the force of what raises them can make bodily and verbal action wholesome and unwholesome, then the great elements ought by the same rule to be so. Answer: “because what the mind of the doer originally intends to raise is action and not the four great elements, and so the rule cannot be extended.” What the doer’s mind intends to raise is “action,” not the great elements; the great elements are only produced incidentally and lie outside the mind’s intention.

First ask: what is the key word of this reason? It is what the intention is directed at. The force of what raises a thing dyes only what the mind is directed at. The easy error: to suppose that “raised at the same time” entails “of the same nature.” In practice: this item will be used once more at the end of the fascicle — that the great elements, though raised by a mind abandoned by seeing, are still not abandoned by seeing, rests on this very principle.

XIV. What Raises Action Is of Two Kinds: the Raising and the Concurrent Raising

〔14.1〕Two minds, two positions. “What raises manifest and nonmanifest action is of two kinds, the causal raising and the momentary raising, since the former is prior and serves as cause and the latter is present with that moment. In order the first is called the raising and the second the concurrent raising.” The treatise explains itself: “the causal raising, at the time when the action is about to be done, is able to draw it forth, and so is called the raising”; “the momentary raising, at the time when the action is being done, is inseparable from it, and so is called the concurrent raising.” Yuanhui’s Verse Commentary glosses this most concisely and clearly: “the mind before the action is raised is called the causal raising, since it is prior and serves as cause. The mind together with the action is called the momentary raising, since it is of the same single moment.”

First ask: why must two be distinguished? Because a bodily or verbal action from its setting going to its completion sets at least two minds in motion — the mind that decides to do it (prior) and the mind that accompanies the doing (simultaneous). The treatise at once asks what use the concurrent raising is, and answers most vividly: “though there be a prior cause able to draw it forth, yet if there were no concurrent raising it would be like a dead man and action ought not to be established” — with only the prior decision and no mind at the time of doing, it would be like a corpse twitching, and would not count as action.

Someone poses a difficulty: how then does one who has entered a mindless absorption generate discipline? The author answers: “whoever has a mind, his action arises distinctly” — for one who has a mind, action arises distinctly; the mindless stages are a separate matter and cannot be used to deny the function of the concurrent raising.

The easy error: to read the raising and the concurrent raising as the ordinary continuity of earlier and later thought-instants. They are functional positions: one governs the drawing forth, the other the bringing to completion. In practice: this pair of words is the key tool for judging the wholesome or unwholesome nature of an action, and the whole of the next section rests upon it.

〔14.2〕The fourfold distinction: ranging four minds into two squares. The treatise sets out the four members: “there is what is the raising and not the concurrent raising, namely a mind abandoned by seeing”; what is the concurrent raising and not the raising is the five consciousnesses beginning with the eye; what is both is the mental consciousness abandoned by cultivation in all three natures; “there is what is neither the raising nor the concurrent raising, namely minds without outflows and maturation-born minds.” The reasons too are given clearly: the five consciousnesses, “being without discrimination and arising by the outer gate,” can only be the concurrent raising; the mental consciousness abandoned by cultivation, “having discrimination and arising by the outer gate,” extends to both; and minds without outflows and maturation-born minds, being “only in absorption” and “turning spontaneously without preparatory effort,” are neither.

First ask: what are the two axes of this table? One is whether there is discrimination (whether it can draw forth), the other is whether it turns by the inner gate or the outer (whether it can accompany bodily and verbal movement). The values of the four minds on these two axes range out exactly into the four members. The easy error: to read the five consciousnesses’ “being without discrimination” as their having no cognitive function. The discrimination meant here is the deliberative discrimination that can draw forth bodily and verbal action.

In practice: the fourfold method is the ordering technique most used in Abhidharma. Wherever two criteria cross, you may set out a fourfold table with your own hand, which is far more serviceable than memorizing conclusions.

〔14.3〕If a mind abandoned by seeing sets manifest form going, is that form abandoned by seeing? This section has a short and keen exchange. The Sautrāntika masters ask: if a mind abandoned by seeing sets manifest form going, that form ought to be abandoned by seeing; and if it be granted so, what fault is there? Puguang’s Record marks it, “‘if a mind abandoned by seeing sets manifest form going, the form ought to be abandoned by seeing. If it be granted to be so, what fault is there’: the Sautrāntika masters ask.” The Sarvāstivādins answer: that would “transgress the Abhidharma,” and moreover “knowledge and ignorance are not mutually opposed”; Puguang’s Record marks it, “‘that would transgress’ down to ‘not abandoned by seeing’: the Sarvāstivādins answer.”

The Sarvāstivādins at once strike back: then the great elements too ought to be abandoned by seeing, since they likewise are set going by the force of a mind abandoned by seeing. Puguang’s Record marks it, “‘if so, the great elements’ down to ‘drawn forth by that force’: the Sarvāstivādins object in return. The Sautrāntikas.” The Sautrāntikas answer: there is no such fault — just as the great elements, though raised by a wholesome mind, are neither wholesome nor unwholesome in nature. The Sarvāstivādins settle it at last: it should not be granted, since undefiled dharmas of every class are not opposed to knowledge or ignorance, and so the great elements are definitely neither abandoned by seeing nor of the not-to-be-abandoned.

First ask: what is this passage arguing? It is arguing whether “the nature of the dharma raised follows the mind that raises it.” The easy error: to see this and the earlier “the great elements do not become wholesome and unwholesome by what raises them” as two different matters. They are two applications of one and the same principle.

In practice: the passage closes with “that sūtra speaks so only with regard to the prior causal raising, and so there is no contradiction” — the sūtra’s saying that “from wrong view there arise wrong speech and wrong action” speaks of the causal raising (prior and drawing forth), not of the momentary raising (simultaneous). A debate seemingly so intricate turns out in the end to be only for the sake of finding a place for one sentence of sūtra. This is exactly the character of Abhidharma.

〔14.4〕Whether the raising and the concurrent raising are of one nature — only in the Buddha is it otherwise. Question: are the raising mind and the concurrently raising mind necessarily of one nature? Answer: “this is not settled.” If the raising mind is wholesome, the concurrent raising may extend to all three natures; and so likewise if it is unwholesome or indeterminate. There is only one exception: “only in the Sage are the raising and concurrently raising consciousnesses for the most part of one nature, with a small portion not so” — in the Buddha the raising and the concurrent raising are for the most part of one nature, with only a small portion otherwise: a wholesome raising has a wholesome concurrent raising, an indeterminate raising an indeterminate concurrent raising; and yet “at times there is a wholesome concurrent raising following upon an indeterminate raising, while there has never been an indeterminate concurrent raising following upon a wholesome raising.” The reason is “because the Buddha, the World-Honored One, in expounding the Dharma and the like, has a mind that may increase and has never flagged.”

First ask: why is the Buddha alone the exception? Because the Buddha’s mind does not fall back — it may rise from an indeterminate raising to a wholesome concurrent raising, but will not fall from a wholesome raising to an indeterminate concurrent raising. The easy error: to read “for the most part of one nature” as “necessarily of one nature.” The stanza itself says “a small portion not so.”

In practice: this section brings a set of cold, hard adjudications about mind to rest at last upon a very warm conclusion — the Buddha’s mind goes only upward, never downward. At this point in the fascicle, the line of dharma-characteristics and the line of practice and realization are joined for the first time.

XV. The Nāga Stanza: Is the Buddha Always in Absorption

〔15.1〕One sūtra stanza and one small debate. The treatise records the account of another school: that the Buddhas, the World-Honored Ones, being always in absorption, have minds that are wholesome alone and no indeterminate mind; and it adduces a sūtra stanza in proof — “the nāga when he walks is in absorption; the nāga when he stands is in absorption; the nāga when he sits is in absorption; the nāga when he lies down is in absorption.” Nāga means the great dragon, by which the sūtras designate the Buddha; the four postures of walking, standing, sitting, and lying down are line by line in absorption.

The Vaibhāṣika masters’ explanation has a fine sense of proportion: “this shows the Buddha’s intent — if he takes no delight in an unconcentrated mind, then in the four postures he is able to be always in absorption; yet at the remaining stages it is not that there arise no mind of the modes of deportment, no maturation-born mind, and no mind of penetrative effect” — what the stanza shows is the Buddha’s “intent”: so long as he takes no delight in an unconcentrated mind, in all four postures he can be always in absorption; but at the remaining stages it is not that the mind of the modes of deportment, the maturation-born mind, and the mind of penetrative effect (which are of indeterminate nature) do not arise. Fabao’s Commentary marks it here as “down to ‘the nāga when he lies down is in absorption’: setting out a divergent account.”

First ask: why do the Vaibhāṣikas take it apart in this way? Because if one bit down on the letter and insisted that the Buddha has no indeterminate mind whatever, it would conflict with the dharma-characteristics of “the Buddha has deportment and has the maturation-born.” So they read the stanza as displaying the Buddha’s inclination, and not as an exhaustive adjudication of dharma-characteristics. The easy error: to take this passage as a slighting of the Buddha. Quite the contrary — it holds fast to “the Buddha does not lose absorption” while not letting the system of dharma-characteristics contradict itself.

In practice: this is a good specimen of how the relation of sūtra stanza to treatise doctrine is handled — a sūtra stanza displays the intent, the treatise settles the characteristics; the two need not be mortised word for word.

〔15.2〕The last question: does manifest action become wholesome after the manner of the raising or of the concurrent raising? At the end of the fascicle a challenge is left: does the wholesome nature and the rest of the several manifest actions follow the manner of the raising mind or of the concurrent raising? Each side has its difficulty: if after the manner of the raising mind, the desire realm ought to have obscured-indeterminate manifest action (since the view of self and the extreme views can act as its raising); if after the manner of the concurrent raising, the manifest of individual liberation arising together with an unwholesome or indeterminate mind ought not to be of wholesome nature.

The author’s correct explanation is: “it should be said that manifest action comes to be of wholesome nature and the rest after the manner of the raising mind” — after the manner of the raising mind; yet not after the manner of that raising mind abandoned by seeing, since a raising mind abandoned by cultivation intervenes between them. And so that sūtra should be read thus: “that sūtra speaks only with regard to the causal raising that is separated off by other minds” — the sūtra speaks with regard to a causal raising separated off by other minds; and so, though a mind abandoned by seeing can act as the raising, within the desire realm there is still definitely no obscured-indeterminate manifest action.

First ask: what does this “separating off” mean? It means that between a mind abandoned by seeing and the actual issuing of bodily or verbal action a mind abandoned by cultivation must intervene; and what settles the nature of the action is the latter, not the foremost. The easy error: to read it as meaning that the mind abandoned by seeing takes no part at all. It takes part; only it does not settle the nature of the action.

In practice: the fascicle closes here, and closes upon a most typical Abhidharma move — not cancelling the point at issue, but adding a qualifier to it. With the four words “separated off by other minds” added, both horns dissolve at once. The three disciplines to be argued in the next fascicle walk out from just here.

Further Resonances: The Mirror of the Three Commentaries

The above is the sense of the treatise’s own text in this fascicle. Below, a further section is set up to gather four matters: the adjudications of the three commentaries and their divergences, Saṅghabhadra’s refutations and supplements, the divergences of Paramārtha’s earlier translation, and coordinates for parallel reading. A mirror is not the treatise; it is only looking at the same text by another way of reading. Both are recorded without adjudication, each restored to its own place.

1. Where the Three Commentaries Diverge

One commentary absent. The commentaries on this fascicle are in a special situation: Shentai’s commentary is a fragment, and the seven fascicles extant belong to the Chapter on the Elements, the Chapter on the Faculties, and the Chapter on the Worthy and the Noble; the Chapter on Action is wholly wanting, and so there is no line-by-line commentary of Shentai to hang upon this fascicle. Within the extant fascicles on the Chapter on the Elements, Shentai’s commentary still has two forward references to the Chapter on Action (“in the chapter analyzing action this will be further considered” and “as is said below in the chapter on action”), from which it appears that he did once comment on the Chapter on Action, and only that portion no longer survives. The absence is marked as it stands, and text from other fascicles is not made to stand in for it. Only the phrase in the general conspectus of the whole treatise in Shentai’s first fascicle, “the chapter on action, one hundred thirty-one lines,” is a second independent source for the stanza count of this chapter, and it agrees with the interlinear note under the chapter title in the separate edition of the root stanzas.

Whether the chapter divides into three or into four. Puguang’s Record marks at the head of the chapter, “within this chapter, first the nature of the entity of action, second the explanation of the actions in the sūtras, third miscellaneous explanations of the actions,” and Yuanhui’s Verse Commentary marks, “three gates in all: first the nature of the entity of action, second the explanation of the names of the actions, third miscellaneous explanations of the actions” — both dividing into three. Fabao’s Commentary alone makes it four: “the main text of this chapter divides into four: first the entity of action, second discipline and the rest, third the actions in the sūtras, fourth miscellaneous explanations of the actions.” The difference lies in the section on “discipline and the rest”: Puguang and Yuanhui gather it into the chapter on the nature of the entity of action as its third subsection (the extended explanation of the manifest and nonmanifest), while Fabao sets it up separately as a chapter-level division. The three in fact agree entirely as to the extent covered by this fascicle — all adjudge this fascicle to be the whole of the chapter on “the entity of action”; what differs is only the layer from which the next fascicle is to be hung. Recorded without adjudication.

How action gets its name: one sense or three. Puguang and Yuanhui both explain action by the single sense of “doing” (Yuanhui gives, “doing is called action, and this chapter explains it at length, hence ‘analysis’”); Fabao alone brings out three senses: “by three senses it is called action: first function, second upholding a norm, third distinguishing results.” The one takes the simplicity of the etymology, the other the fullness of sense and function. Recorded without adjudication.

The internal subdivision of this fascicle: two sections or six. Puguang divides the chapter on the nature of the entity of action into two sections only (properly explaining the nature of the entity of action, and analyzing action through the several gates); Fabao has six sections of his own, each hung with its stanza count — “the first four and a half stanzas explain the nature of the entity of action, the second one and a half stanzas explain what produces the great elements” — and thereafter lists the appropriated and the classes, analysis through the gates of sense, the four wholesome and the rest, and the two raisings. Yuanhui agrees with Puguang. ⇒ Fabao’s outline is the finest of the three and takes the stanza as its unit, which makes it the one most convenient for collating against the stanza text. Recorded without adjudication.

The two commentaries differ as to who asks the closing question “what do you establish as bodily manifest action.” On the treatise’s sentence “since both motion and shape have been set aside, what do you of the Sautrāntika school establish as bodily manifest action,” Puguang marks it as a “closing question,” and his explanation takes what is set aside to be the Sāṃmitīyas’ motion and the Sarvāstivādins’ real shape and the one questioned to be the Sautrāntikas — “since the Sāṃmitīya masters’ motion as bodily manifest action and the Sarvāstivādins’ real shape as bodily manifest action have been set aside, what do you of the Sautrāntika school establish” — so that the questioner is the author; Fabao, on the same sentence, marks it as “the Sāṃmitīyas and the Sarvāstivādins put the question,” assigning the questioner to those two schools. The two point at one and the same sentence while differing as to whose the question is. ⇒ One sentence, and the two commentaries give different questioners; recorded without adjudication; the vernacular layer of this fascicle follows Puguang’s outline and takes it as the author’s closing question. Here a further layer must be distinguished: Fabao’s mark “‘since you hold that only’ down to ‘as bodily action’: the Sarvāstivādins ask” hangs upon the next sentence, “what dharma then do you establish as bodily action,” and not upon the closing question above; while Puguang on this next sentence marks only the single word “question” and gives no school name. Each of the two sentences has its own mark, and in reading the commentaries the mark on the lower sentence may not be moved up to stand for the upper — this is the general rule of the commentaries’ sentence-by-sentence attribution, and the place in this fascicle where collating the two commentaries most easily slips out of alignment.

The posture of the author’s laying the dispute to rest: complaint or reproach. On the passage “why do you bear a grudge against the nonmanifest in particular,” Puguang marks it as “the Sarvāstivādins complain,” and Fabao marks it as “here the Sarvāstivādins reproach the Sautrāntikas for partiality of feeling.” The one marks it a plaint, the other an arraignment — Puguang inclining to neutral narration, Fabao to the tone of reproach. One and the same passage, and the two commentaries read a different temperature out of it; recorded without adjudication.

The explanation of the four wholesome: simile and counter-illumination. Puguang sets forth the four wholesome sentence by sentence with similes, his mark being “‘the treatise says’ down to ‘as when free of illness’ explains the first line, showing the ultimate wholesome,” and in the section on the intrinsic, “these five, being strong and superior, are called intrinsic”; Fabao in the section on the intrinsic goes the other way and lists the five senses of the unwholesome roots by way of counter-illumination — “this explains the second, the intrinsically wholesome. Non-conscience and shamelessness have two senses” — and thereafter lists five senses: that the unwholesome roots extend over the five categories, pervade the six consciousnesses, are of the nature of latent tendencies, make a firm and strong preparatory effort when the wholesome roots are cut off, and can set going the coarse and evil actions of body and speech. Yuanhui is the briefest here, preserving only the formulaic gloss “within it the utmost security, all sufferings wholly stilled, as when free of illness.” The three differ in their inclination as to fullness and brevity. Recorded without adjudication.

The gloss on the two raisings: Yuanhui is the fullest. Puguang in this section marks only “the distinction. ‘The treatise says’ down to ‘is called the concurrent raising’ explains the first stanza, and may be known,” and adds no gloss of his own; Yuanhui takes it apart word by word — “‘what raises is of two kinds’ is the heading”; “‘the cause and the moment of that’ is the listing of names”; “the mind before the action is raised is called the causal raising, since it is prior and serves as cause; the mind together with the action is called the momentary raising, since it is of the same single moment.” The most abstract set of words in the whole fascicle is glossed most clearly, of all places, by the commentary of outlines. Recorded without adjudication.

Each commentary’s outline terminology forms a system of its own. Puguang uses “below, the Nth”; Fabao uses “the Nth of the main text”; Yuanhui uses “from here the Nth.” In citing the “Nth” of the three commentaries one must attach it to its own house and not let one stand for another — for one and the same outline boundary the three often give different numbers. This is where reading the three commentaries most easily goes wrong, and it is recorded as it stands.

Fabao once uses “Sūtra-master.” At the end of the section on the raisings Fabao gives “this is because the Sūtra-master did not understand the tenet of my school” — designating Vasubandhu as “the Sūtra-master.” This is the usage of the Treatise Conforming to Correct Principle lineage (Saṅghabhadra throughout calls Vasubandhu the Sūtra-master), and its appearance in a commentary on the Kośa is well worth noticing, showing the influence of that lineage upon Fabao’s composition. Recorded without adjudication.

A one-word divergence in the root stanza as cited by Yuanhui. Yuanhui cites the last line of the root stanza as “in the Sage they are necessarily of one nature; where indeterminate, the concurrent raising may be wholesome,” using the character for “may,” where the base text has the character for “delusion.” This one-word divergence has further collateral evidence: the Taishō collation notes for this treatise record the Ming and the Palace editions as having “may,” and the separate edition of the root stanzas likewise has “may” here. Four lines of collateral evidence run one way, and the base text still has “delusion.” By the form of this collection the base-text reading is not altered; here only its divergence is recorded, the choice between them belonging to doctrinal adjudication.

2. Saṅghabhadra’s Refutations and Supplementary Explanations

The chapter name in Saṅghabhadra’s two treatises is not “Analysis of Action” but “Adjudication of Action” — the Treatise Conforming to Correct Principle has “Adjudication of Action, chapter four,” the Treatise Revealing the Doctrine has “Adjudication of Action, chapter five.” To search them with the Kośa’s chapter name is bound to yield a false zero.

The gloss on the moment is expressly denied. Vasubandhu’s definition of the moment (having obtained its entity it perishes with nothing intervening, as a man with a staff is called having a staff) is upheld as settled doctrine by all three commentaries; Saṅghabhadra, however, cites it word for word and refutes it — “yet the Sūtra-master says: what is called a moment? Having obtained its entity it perishes with nothing intervening; a dharma that has this moment is called having a moment, as one has” — and adjudges, “his explanation is unreasonable, since a staff is other than the man and so it cannot be said; the simile is not of like dharma, and there is no separate dharma other than the obtaining of the entity.” His reason is that a staff and a man are two things and so “having a staff” may be said, whereas “perishing with nothing intervening after obtaining its entity” is not a separate dharma other than the dharma-entity, and the simile is not of like dharma. ⇒ This is not a refutation of an argument but an outright denial of the gloss itself. One sense with two schools, recorded together, with no adjudication of who prevails.

Cessation awaiting no cause is likewise met with an express counter-thesis. Saṅghabhadra cites “the Sūtra-master explains it thus: the cessation of conditioned dharmas awaits no cause,” and adjudges, “his explanation is unreasonable; exhaustion is just cessation, and the Buddha said that exhaustion and cessation are a mark of the conditioned” — exhaustion is cessation, and the Buddha said that exhaustion and cessation are a mark of the conditioned; being a “mark,” it is a dharma that is had, and cannot be said to be a mere “nonexistence.” He establishes by scriptural proof that “cessation too is a mark of the conditioned,” meeting head-on Vasubandhu’s “cessation is a nonexistence and a nonexistence is not a result, and so it awaits no cause.” This is the place where the two parties of this fascicle stand most sharply opposed. Recorded without adjudication.

On the dispute over shape-form the two treatises are almost word for word the same. Saṅghabhadra cites “yet the Sūtra-master says: shape is not really existent; that is, where visible color arises abundantly in one quarter” (the Kośa has “on one side,” Saṅghabhadra’s citation has “in one quarter,” a one-word divergence), then cites the difficulty of apprehension by two faculties and adjudges, “this principle is not so, since what is made out differs in mark” — what the eye makes out and what the body makes out differ in mark, and so the fault of one form apprehended by two faculties does not follow. He further cites “the Sūtra-master here says again: whatever two dharmas are definitely inseparable, therefore” and refutes it. The corresponding passage in the Treatise Revealing the Doctrine gives “the supports being separate, that the entities are separate is established in principle,” and so on, almost word for word the same as in the Treatise Conforming to Correct Principle. ⇒ Saṅghabhadra here expressly establishes that “shape-form has atoms of its own” in order to refute the Sautrāntikas — that is, to refute the Sautrāntika position as set out by Vasubandhu. Recorded without adjudication.

A direct challenge to Vasubandhu’s closing words. “What do you of the Sautrāntika school establish as bodily manifest action? Here the Sūtra-master states their school’s position, saying: bodily manifest action is just shape, yet conventional” — Saṅghabhadra cites Vasubandhu’s closing words and goes on to press the ambiguity of the position stated: does “conventional and not real” mean that the shape of long and the rest is conventional and not real, or that “the atoms that make up the long and the rest, called shape,” are conventional and not real? What this question addresses is not the Sautrāntika school itself but Vasubandhu’s own wording. Recorded without adjudication.

The sentence that the nonmanifest is not really existent is gone over word for word. “Yet the Sautrāntikas say: this too is not real, since it is merely a not-doing by force of a limit vowed beforehand” — Saṅghabhadra cites the Kośa’s original sentence word for word and goes on to refute, “all such causes are unreasonable. And a mere not-doing is not to be called nonmanifest action.” The sentence in the two texts agrees character for character, and this is the most exactly matched passage of word-for-word collation in the disputation of this fascicle. Recorded without adjudication.

What is the decisive cause of an action’s becoming wholesome and the rest. The Kośa establishes that “all minds without outflows and maturation-born minds are neither the raising nor the concurrent raising”; Saṅghabhadra cites and comments, “yet the Sūtra-master says: because they turn spontaneously without preparatory effort, the maturational consciousnesses are neither the raising nor the concurrent raising,” and adjudges, “by this the Sūtra-master has the fault of both diminishing and adding, making a cause of what is not a cause,” and at the end establishes his own school’s position, “that an action becomes wholesome and the rest is definitely by the force of the raising and not by the force of the concurrent raising, and this principle is well made out.” Whereas the Vaibhāṣika answer at the end of the Kośa is precisely “it should be said that manifest action comes to be of wholesome nature and the rest after the manner of the raising mind” — the two run in the same direction while the lines of reasoning differ. Recorded without adjudication.

The nāga stanza passage is not a disputation. The two treatises cite the same stanza, word for word the same with no divergent reading; Saṅghabhadra goes on to set it forth with the Vaibhāṣika explanation, agreeing in doctrine with what the Kośa records while a little fuller in wording. ⇒ This passage is a difference of fullness in setting forth the sense of one and the same school, not a disputation, and is recorded as it stands.

The two treatises also differ in fullness between themselves. The Treatise Revealing the Doctrine is far briefer than the Treatise Conforming to Correct Principle in the passages corresponding to this fascicle: a fault-tolerant re-check of the two phrases “motion is called bodily manifest action” and “the nāga when he walks is in absorption” within the corresponding range of the Samayapradīpikā yields a word-for-word zero, and only the one sense of the real existence of shape-form has a condensed correspondence. This is a fact about the form of the two treatises, not a failure of searching.

3. The Divergences of Paramārtha’s Earlier Translation

Paramārtha’s Commentary on the Abhidharmakośa forms a system of its own in the passages corresponding to this fascicle. The first thing to know is the master switch of the renderings: Xuanzang’s manifest / nonmanifest is throughout Paramārtha’s instructive / uninstructive. This pair runs through the whole passage, and everything touching the manifest and nonmanifest changes with it; search Paramārtha’s text with Xuanzang’s words, and in this field you will never find anything.

A basic difference of form. Xuanzang has “the stanza says / the treatise says,” the whole stanza coming out first and the prose commentary after; Paramārtha has “the verse says / the explanation says,” glossing line by line inset, so that one stanza is often broken into two or several places of citation, and his verse lines are not confined to five characters. So Xuanzang’s fourteen stanzas in this fascicle appear in Paramārtha as a number of short verses scattered about. The two also divide their fascicles differently: Xuanzang’s Chapter on Action has six fascicles, Paramārtha’s four; the passage corresponding to this fascicle still falls within one fascicle of Paramārtha, and its boundaries can be fixed only by the structure of the scenes.

The renderings form a system of their own. What is without outflows becomes “without flow”; yogins become “the observing masters of former times”; invisible and unresistant becomes “without display and without obstruction”; great elements become “the four greats”; sentient beings become “the designation of living beings”; homogeneous nature becomes “the flowing result”; unappropriated becomes “not grasped by mind”; nourished becomes “increased”; having sustained thought becomes “having observation”; conscience and shame become “conscience and abashment”; abandoned by cultivation becomes “extinguished by the path of cultivation”; maturation-born becomes “born of retribution.” Of these the one most to be watched is what raises / conditioned arising: Paramārtha renders “what raises” by “conditioned arising,” the same characters as the “conditioned arising” of the twelvefold chain but pointing at something else, and the two must be told apart in reading.

Two instances where the methods of translation most clearly part. The first is the mark of bodily manifest action: Xuanzang has “by the force of volition, such and such a shape further arises in the body, and this is called bodily manifest action,” while Paramārtha’s verse has “it is said that the body has the instructive mark,” and his explanation, “in accordance with intention, the body is of such and such an appearance, and this is called instructive” — Xuanzang taking “bodily shape,” Paramārtha “appearance.” The second is the yogins: Xuanzang has simply yogins, Paramārtha “here the observing masters of former times say,” with the two extra words “of former times” expressly marking them as an account of an earlier age, a qualifier Xuanzang does not have.

A twofold divergence in the nāga stanza. Xuanzang cites the stanza with the four lines of walking, standing, sitting, and lying down, all in absorption; Paramārtha has “the nāga in walking is at peace, the nāga in leaning is at peace, the nāga in lying down is at peace, the nāga in sitting is at peace” — “standing” becoming “leaning,” “in absorption” becoming “at peace,” and the lines of sitting and lying exchanged in order. A fault-tolerant re-check of “the nāga when he stands is in absorption” across the whole of Paramārtha yields zero: this is a real divergence of the two translations, not a failure of searching. Recorded without adjudication.

Fullness does not run all one way. In several passages Paramārtha is the fuller: in the section where the Sautrāntikas establish bodily manifest action, Xuanzang answers with a single sentence while Paramārtha first sets up a question and then brings out the answer. In several passages the two texts agree almost word for word: the simile of the man with a staff for momentary cessation, the inference that the productive cause becomes the extinguishing cause, and the opening sentence of the Sautrāntikas’ blanket denial of the eight proofs — and especially that last sentence, where Xuanzang has “these proofs, though many and of all kinds of marvelousness, are nevertheless unreasonable” and Paramārtha has “these proofs are very many and of all kinds of rarity, yet in truth it is not so,” differing by only a few words, from which the stability of this sentence in the translation halls may be seen. Fullness varies passage by passage, with no settled direction.

4. Coordinates for Parallel Reading

The nonmanifest form established in this fascicle is the central pivot of the “side of the entity of action” in the Abhidharma of the schools; its being taken up by the Yogācāra lineage is one of the most direct lines in the whole treatise.

The Treatise on the Establishment of Consciousness-Only poses on the same topic exactly the same question: are manifest and nonmanifest form not really existent? Its reason for refuting shape-form is that “if shape is said to be really existent, what is its nature? — if it is said to be shape, then it is not really existent, since it can be analyzed”; its reason for refuting motion is that “what arises ceases at once, and so there is no sense of motion” and “the cessation of a conditioned dharma awaits no cause.” ⇒ These are precisely the three tools used by the Sautrāntikas in this fascicle (analyzability, the absence of separate atoms for long and the rest, momentary cessation), not one of them missing. Its final conclusion is that “the manifest being really nonexistent, how should the nonmanifest be real,” and it establishes instead that “the nonmanifest is conventionally established upon the wholesome and unwholesome limits of resolve and intention.” ⇒ The Sautrāntika position that stands in this fascicle as one of two parties recorded is taken up entire in the Yogācāra lineage and becomes its own school’s position. The same set of arguments in two positions — this is the most readily collated place of all (for the word-for-word citations and page anchors, see item seven of this unit’s apparatus).

The entity of the three actions is likewise. In this fascicle the Sautrāntikas answer that “bodily action is the volition that can move the body”; that lineage too gives “it is not denied as nonexistent, but only said not to be form” and “the volition that can move the body is called bodily action, the volition that can issue speech is called verbal action” — the one sense that “the three actions have volition for their entity” is handed on in both systems; what differs is only that this fascicle takes it as one of two parties recorded, while that lineage takes it as its own school’s position.

And one thing further, on the side of enumeration: this fascicle establishes the nonmanifest as invisible and unresistant form, while the prose commentary records the yogins’ separate explanation of the three forms without outflows by the form in the object-field of absorption. The Yogācārabhūmi, in listing the sphere of dharmas, divides the form gathered under the dharma sphere into two — the form gathered under discipline and non-discipline, and the form ranged over by concentration. ⇒ These two forms of the dharma sphere in that lineage correspond one to one with the two matters of this fascicle: the former to the nonmanifest, the latter to the form in the object-field of absorption. ⇒ The coordinates by which one and the same set of dharmas is opened out in the two systems are recorded only, and no settled judgment is passed (for the word-for-word citations and page anchors, see item seven of this unit’s apparatus).

As for the “raising and concurrent raising” with which this fascicle closes, the next fascicle will use them again in treating the obtaining and relinquishing of the three disciplines; that is the business of the next. This fascicle does two things only: it sets up the entity of action, and it records the two great disputes exactly as they stand.