俱舍論 義疏通講 · 冊5Abhidharmakośabhāṣya · Exposition Throughout · Vol. 5

Before Reading This Fascicle

The first sentence of this fascicle is the plainest hand-over in the whole treatise — “it was said before that the differences of the world all arise from action; action grows only when the latent tendencies sustain it; parted from the latent tendencies, action has no power to call forth a coming existence.” The chapter before spent three whole fascicles on action, and at the very end did not close upon action, but pointed its finger at another thing: action does not sprout of itself, and only when the latent tendencies moisten it can it grow. So the chapter on the latent tendencies does not raise a new topic beside the chapter on action; it supplies the half that the chapter on action left unfinished.

The first cut made by the three ancient masters in this chapter falls with a rare evenness — Puguang’s Record marks “within this chapter, first the entity of the confusions is explained, second the extinction of the confusions”; Yuanhui’s Verse Commentary marks “within this chapter it divides into two: first explaining the entity of the confusions, second explaining the extinction of the confusions”; and Fabao’s Commentary, though it changes the wording to “first explaining the entity and function of the confusions, second explaining the differences of their abandonment,” still adjudges the same twofold division. This whole fascicle falls within the half “explaining the entity of the confusions,” and moreover at the head of that half: it walks through the five sections that analyze the root confusions (increasing the count · what is abandoned through seeing and cultivation · the five views · the four inversions · the seven and the nine prides), then walks the first five of the distinctions through the several gates (all-pervading · taking objects with and without outflows · the two kinds of increase by attachment · the two ethical qualities · what is and is not a root), appends at the end a section on the four gates of determination, and the fascicle closes.

There is one thing to be said in advance. This fascicle reads as though one were doing arithmetic — six, seven, ten, ninety-eight; four inversions, seven prides, nine classes of pride, eleven all-pervading, six that take the dharmas without outflows as object, three unwholesome roots, three indeterminate roots, four determinations — the numbers grow so many as to daunt one. But what bears the load is not the numbers; it is the measuring rod that is exchanged each time the number changes. Six becomes seven by exchanging the rod of greed; seven becomes ten by exchanging the rod of view; ten becomes ninety-eight by exchanging three rods at once (mode · category · realm). The skeleton of the whole fascicle is this one sentence: one and the same set of afflictions takes a new number with each new rod; before asking which number is right, ask which rod this is.

I. Why the Latent Tendencies Can Be the Root of the Several Existences — the Ten Functions of the Afflictions

First the question: action is already what can call forth a result; why must it further be said that “action grows only when the latent tendencies sustain it”? The treatise’s answer is nowhere else than in the sentence “because the afflictions, once arisen, are able to perform ten kinds of function” — the moment an affliction arises, it is at the same instant doing ten things.

These ten are worth looking at one by one, for they are not a spread of synonyms but ten different faces of function. “Strengthening the root” is the fortifying of the affliction’s own foundation; “establishing the continuity” makes it run on thought after thought unbroken; “tilling its own field” works the ground of mind into a field fit for its own growth; “drawing forth the homogeneous” draws out a sequel of its own kind; “setting going action and existence” is the item we know as ‘starting action off’; “gathering its own equipment” draws in the conditions that furnish it; “bewildering the object” makes the mind unable to see its object clearly; “leading the stream of consciousness” drags consciousness along in the direction it wants to go; “overstepping the wholesome class” is the stepping across of that portion which is wholesome; “the sense of binding widely” is a broad binding, and to it the treatise adds a clause of its own — “because it makes one unable to overstep one’s own realm and level.”

The easy error: to read the ten functions as ten adjectives. They are not. Of the ten, only the fifth speaks directly of action; the other nine all speak of how an affliction lays in its own materials. Only so does the treatise dare to say that action parted from it “has no power to call forth a coming existence” — it is not that action cannot call forth a result, but that action in a field without the latent tendencies simply cannot grow.

In practice: while reading this chapter, whenever the doubt arises ‘what is so important about this affliction?’, go back and count which of the ten functions it occupies, and its weight becomes plain at once. Puguang cites the Treatise Conforming to Correct Principle item by item as the gloss on these ten, his first item reading “first, strengthening the root, making it firm, because what counteracts it is far off” — from which one sees that the gloss on these ten is no late invention: the ancient masters already had an explanation of each.

A further note on two particles. The stanza says “that is, greed, hatred, and also pride,” with the one word ‘also’; and it says “ignorance, view, and doubt,” with the one word ‘and.’ The author himself picks out these two words and explains them: ‘also’ is meant to display that “pride and the rest likewise increase upon their objects by the force of greed” — the other afflictions such as pride also grow their strength upon objects by the force of greed; ‘and’ is meant to display that “the entities of the six are each different” — these six are not six names for one thing. First the question: why should two particles need explaining? Because without the explanation the reader would take the six latent tendencies for six coordinate items and miss the linkage of force among them; or would swing the other way and take them for the several names of one thing. One word props up the one side, and the two words together draw the boundary straight.

II. Why Six Is Opened Out into Seven — Where the Name “Craving for Existence” Is Fixed

First the question: the sūtra says seven latent tendencies and the treatise says six; which is right? The treatise’s way of answering is not to choose one but to explain the ground of the opening and gathering — “within the six latent tendencies spoken of before, greed is divided into two, and so the sūtra says seven.” Six and seven are the same set of things; the difference lies only in whether greed is taken apart.

Once it is taken apart, the portion belonging to the upper two realms is called “craving for existence.” Why is this not also called greed for desire? The treatise gives two reasons, and they are reasons on two levels. The first is stated from the fact — “because that greed for the most part turns in dependence upon the inner gate”: sentient beings of the upper two realms mostly give rise to greed for absorption, and all greed for absorption turns toward the inner gate and does not take the five outer objects of desire for its object. The second is stated from the intent — “in order to block the notion of liberation”: some mistake birth in the upper two realms for liberation, and to set up the word ‘existence’ (the existence of the three existences) is precisely to display that what is taken there as object is “not true liberation.” Puguang divides the two reasons expressly here, saying “the second line, ‘because it turns in dependence upon the inner gate,’ is why it is not called greed for desire; the third line, ‘in order to block the notion of liberation,’ is why it is called craving for existence” — one line explaining why the old name is not used, one line explaining why a new name must be.

The easy error: to suppose that “craving for existence” is another kind of affliction. It is not; it is just greed, only that the realm in which it stands and the gate upon which it takes its object are different. The treatise then adds a further layer — the sentient beings there “deeply give rise to savoring attachment toward meditative attainment and toward the body upon which they depend, and so it is said that they savor and cling to their own entity alone; they do not savor and cling to an object, since they have left greed for desire”: what they are greedy for is the entity of absorption, not an outer object, for greed for outer objects was left behind long ago.

In practice: this section demonstrates a standing practice of this treatise in establishing its terms — the dividing and joining of names follows the object and the mode, not the entity. The entity is only one, greed; because object and mode part into two roads, the name parts into two. Hereafter, when the question arises ‘why does one and the same dharma have two names?’, ask it first along this line.

III. 〔A Passage of Prose Commentary Set Apart〕Are the Latent Tendencies Associated with Mind, or Not Associated with It

This is the first great dispute of the fascicle, and the densest run of question and answer in the whole chapter. First the question: how is this dispute raised? The starting point is so small it hardly seems a question at all — how the four characters ‘the latent tendency of greed for desire’ are to be read. The treatise says, “‘the latent tendency of greed for desire’ — by which sense is it to be explained? Is the entity of greed for desire itself the latent tendency? Or is the sense ‘the latent tendency belonging to greed for desire’?” Read the first way, greed for desire and the latent tendency are one and the same thing; read the second, the latent tendency is another thing belonging to greed for desire. Puguang marks these two readings very clearly — “a twofold alternative pressed to a decision. Is the entity of greed for desire itself the latent tendency, on an appositional-determinative reading? Or is it the sense ‘the latent tendency of greed for desire,’ on a dependent-determinative reading?” — the one appositional-determinative, the other dependent-determinative.

The author’s answer is “both alike have a fault” — both readings have a fault, and so an objection is brought against each side.

The objection to the first reading cites a sūtra: the sūtra says that a man may for a while give rise to the entanglement of greed for desire and yet know as it really is the means of going forth, and “thereby he is able rightly to expel the entanglement of greed for desire, together with the cutting off of the latent tendency.” Since the sūtra says ‘the entanglement is cut off, and the latent tendency too is cut off,’ entanglement and latent tendency are plainly two things; and to say that the entity of greed for desire is itself the latent tendency contradicts the sūtra.

The objection to the second reading cites the treatise: the treatise says “the latent tendency of greed for desire is associated with three faculties.” Since it is associated with faculties, it is an associated dharma; and to say that the latent tendency is another thing belonging to greed for desire makes it a formation not associated with mind, which contradicts the Abhidharma.

The position of the Vaibhāṣikas is to take the first reading — “the entities of greed for desire and the rest are just the latent tendencies”; a latent tendency is the affliction itself, an associated dharma. The objector at once presses, “does this not contradict the sūtra?”, and the Vaibhāṣikas explain the sūtra through by two roads: on one road, the sūtra’s words “together with the latent tendency” mean “together with the accompanying bondage” — what is cut off is that portion together with the bondage accompanying it; on the other road, the sūtra speaks of the latent tendency conventionally upon the ‘acquisition,’ “as one sets up the notion of suffering and the rest in fire and the rest,” fire not being suffering, and yet touching fire is called suffering, which is a conventional way of speaking from what it produces. At the end they come back to their own tenet — “the Abhidharma speaks in accordance with the real mark, calling the afflictions themselves latent tendencies.”

The objector presses again, “by what principle is it proved that they are definitely associated?”, and the Vaibhāṣikas bring three functions in evidence: the latent tendencies can defile and vex the mind, can cover and obstruct the mind, and can go against and injure the wholesome. Their inference is very fine — if the latent tendencies were dissociated formations, and dissociated formations are “constantly present,” then wholesome dharmas ought never to have a time of arising; but since wholesome dharmas do still arise, the latent tendencies are plainly not dissociated. The objector takes the proof apart in one sentence — “none of this is proof” — his ground being that one who does not grant that the latent tendencies are associated will not in the first place grant that those three functions are performed by the latent tendencies. The easy error: to suppose that the objector is denying that the latent tendencies can defile the mind. He is not; what he denies is the premise that ‘this work must be performed by an associated dharma.’ Whenever one party in the treatise says “none of this is proof,” look first at whether he is taking apart the conclusion or the premise.

Here the author lays down a verdict — “yet what the Sautrāntika masters say is the best.” Puguang expressly marks it with four words, “the author appraises and takes the Sautrāntikas.” This sentence matters: in many disputed places this treatise records both parties and passes no verdict, whereas here there is a plain taking of sides. The Sautrāntika account is: a latent tendency is neither associated with mind nor dissociated from mind, because “there is no separate thing” — it simply is not another thing. An affliction in the stage of sleep is called a latent tendency, in the stage of waking an entanglement. What is sleep? “A seed following along without being active.” What is waking? “The afflictions arisen and entangling the mind.” What is the seed of an affliction? “A capability that is a differentiation upon its own entity, arising from an affliction and able to produce an affliction.”

The key to this road lies in two similes. The seed of memory: recollection arises from having once known, and can produce a later recollection; this is only a differentiation of capability, and no one would say that the seed of memory is a separate real thing. A sprout and the rest: a sprout arises from a former fruit and can produce a later fruit, and this too is only a differentiation of capability. Whereupon the Sautrāntikas ask in return: if the seed of memory is a mere capability, why must the seed of an affliction have a further dissociated real thing set up for it? “Because the causal ground of a difference cannot be found” — no reason can be found to make the two unlike.

The last round is the objection from the Sūtra of the Six Sixes. The Sarvāstivāda side cites the sūtra saying “upon a pleasant feeling there is a latent tendency of greed”: if a latent tendency were merely a seed, how could the sūtra say that upon a pleasant feeling it ‘is’ there? The Sautrāntikas explain the sūtra through in one sentence — “the sūtra says only that there is one; it does not say that there is a latent tendency at that very moment”; the sūtra says there is, it does not say that at that time it is active. Asked again, “at what time, then, is there one?”, the answer is “at the time of its sleeping; or else the notion of a latent tendency is conventionally set up upon the cause”: either it is there while the seed lies dormant, or the name of the result, ‘latent tendency,’ is fixed upon the cause, which is greed.

In practice: what this whole passage really teaches the reader is a distinction that can be carried away — ‘there is’ and ‘is active’ are two different things. When a sūtra says that a certain affliction is present somewhere, that does not mean that the affliction is at work there. This distinction will be used once more below in the section ‘in the noble one they are and yet do not arise,’ and again in the gate of increase by attachment. The author closes just as briskly — “let the incidental discussion rest here for now; the main discussion should be taken up again.”

IV. Why Six Is Opened Out into Ten — the Difference in the Modes of View

First the question: since six has already been opened into seven, why open it once more into ten? Because the rod has been changed. What was taken apart before was greed; what is taken apart this time is view — “within the six latent tendencies the modes of view differ and make five; the remaining five are not views; adding the counts together makes ten in all.”

Of the ten, five are of the nature of view: the view of a real person, the extreme view, wrong view, adherence to views, adherence to observances and vows; and five are not of the nature of view: greed, hatred, pride, ignorance, doubt. The easy error: to suppose that ten adds four to the six. It does not. The one item ‘view’ within the six is analyzed into five because the modes in which it operates differ in five ways, while the other five items are left untouched, and five plus five makes exactly ten. The relation of ten to six is therefore of the same type as the relation of seven to six — both are the analysis of the inside of one item, not an addition from outside.

In practice: whenever a new number is met with in this chapter, do one thing first — subtract it from the number before, and see out of which item the surplus has been divided. From six to seven, from seven to ten, from ten to ninety-eight, all three subtractions in this fascicle hold good.

V. Why Ten Is Opened Out into Ninety-Eight — the Three Layers of Mode, Category, and Realm

First the question: why does this one leap all the way to ninety-eight? Because this time three rods are used at once — “the six latent tendencies differ by mode, category, and realm, and so become ninety-eight.” Mode is the difference in the modes of view of the section before (six analyzed into ten); category is “the five categories, what is abandoned by seeing the four truths and what is abandoned by cultivation”; realm is “the three realms of desire, form, and formlessness.”

The treatise gives the reckoning very concretely. The five categories of the desire realm multiplied into the ten latent tendencies give thirty-six — “from what is abandoned by seeing the truth of suffering down to what is abandoned by cultivation there are, in order, ten, seven, seven, eight, four.” These five numbers are not set down at random; each has what it lacks: what is abandoned by seeing suffering has all ten; what is abandoned by seeing origination and by seeing cessation have seven each, “lacking the view of a real person, the extreme view, and adherence to observances and vows”; what is abandoned by seeing the path has eight, “lacking the view of a real person and the extreme view”; what is abandoned by cultivation has four, “lacking the views and doubt.” The upper two realms “each lack hatred in the five categories,” and so have thirty-one each. Thirty-six plus thirty-one plus thirty-one is exactly ninety-eight.

From this table one can read off, in reverse, the distribution of each affliction, and the treatise gives it item by item: the view of a real person is in the one category, what is abandoned by seeing suffering, and the extreme view likewise; adherence to observances and vows extends to two categories, what is abandoned by seeing suffering and by seeing the path; wrong view extends to four categories, and adherence to views and doubt likewise; greed and the other three each extend to all five. The easy error: to misread ‘a given view is in one category only’ as ‘a given view takes only one truth as its object.’ The criterion the treatise gives is another sentence — “if what it takes as object is what is abandoned by seeing this truth, it is named as abandoned by seeing this truth; the rest are named as abandoned by cultivation”: the category of abandonment is fixed by which category’s dharmas it takes as object.

At the end the treatise gives the ninety-eight one more cut: “eighty-eight are abandoned by seeing, being injured by receptivity; ten latent tendencies are abandoned by cultivation, being injured by cognition.” In practice: this section is the skeleton of the whole chapter. Almost every gate that follows (all-pervading, taking the dharmas without outflows as object, increase by attachment, the two ethical qualities) will come back to this table of five categories and three realms to draw its numbers, and it is best held firmly in mind while reading — and if it will not be held firmly, no matter: it is enough to hold firmly the fact of ‘the three rods, mode · category · realm,’ for the table can always be derived again.

VI. Are What Is Abandoned by Seeing and by Cultivation Rigid — the Summit of Existence and the Other Eight Levels

First the question: the section before has just established that ‘eighty-eight are abandoned by seeing and ten by cultivation,’ and the treatise at once asks itself “is it decidedly so?” and answers itself, “it is not.” Why loosen what has just been set up? Because the former table divides by the dharma, while this section divides by the person — one and the same affliction may belong to different modes of abandoning according as a noble one abandons it or an ordinary being abandons it.

The treatise divides into three layers. First: a latent tendency injured by receptivity, if it is at the level of the summit of existence, is “abandoned by seeing alone, because only the receptivity to the subsequent cognition can abandon it” — at the level of the summit of existence only the receptivity to the subsequent cognition can abandon it, and so there is no second road. Second: those gathered into the other eight levels “extend to abandonment by seeing and by cultivation” — “when a noble one abandons them it is by seeing and not by cultivation,” since he abandons by the receptivities to the cognition of dharma and to the subsequent cognition; “when an ordinary being abandons them it is by cultivation and not by seeing,” since he can only abandon by “repeated cultivation of conventional cognition.” Third: those injured by cognition, gathered into all levels, are “abandoned by cultivation alone.”

The easy error: to read ‘extending to abandonment by seeing and by cultivation’ as one person having two roads to choose between. It is not so — what extends is this class of affliction, not any one person. The noble one walks only the road of seeing, the ordinary being only the road of cultivation; only taken together are they called ‘extending.’

In practice: this section demonstrates an important way of reading this treatise — the classification of a dharma may be stated relative to the person who abandons it. Hereafter, when two classifications collide, first ask the one sentence, ‘are these two classifications stated with respect to different persons?’ At the end the treatise records a divergent account as well: some other masters say that the seers among the non-Buddhists cannot suppress or abandon the afflictions abandoned by seeing, and cite the Sūtra of the Great Analysis of the Several Actions and the Brahmā Net Sūtra in evidence; whereas the Vaibhāṣikas explain the sense of that sūtra otherwise, saying it is “a momentary falling back at the time when the view arises, as with Devadatta.” Both accounts are recorded together.

VII. The Entities of the Five Views — Five Different Ways of Going Wrong

First the question: if the five views are all views, wherein lies their difference? The treatise answers by giving one by one the mark of what each holds, and the five marks cannot be exchanged for one another.

The view of a real person holds to “self and what belongs to self.” The gloss on this name is given at particular length in this fascicle — “because it perishes it is called sat; the aggregation is kāya, that is, the sense of impermanent conjoined aggregates”; sat is perishing, kāya is aggregation, and together they are ‘aggregates conjoined and impermanent.’ The author goes on to point out the intent of setting up the name: “this name is set up in order to block the notion of eternity and of unity, since that notion must come first before there is any holding to a self” — first there is the notion of ‘eternal’ and ‘one,’ and only upon that is a self held to; to put the word ‘perishing’ into the name is precisely to fend that notion off. This passage has also another master’s gloss (“because it exists it is called sat”), and the two glosses stand together, their adjudication belonging to the section of Further Resonances.

The extreme view is “just upon the self and what belongs to self that are held to, the holding to annihilation and the holding to eternity” — note that its object is nothing else than the self and what belongs to self held to by the view before it; hence the extreme view necessarily follows behind the view of self, a point that will be used again below when all-pervading is treated.

Wrong view is “to give rise to a view denying outright, within the really existent truths of suffering and the rest.” Here the treatise poses itself a question: all false views are inverted, why is this one alone called wrong? The answer is excellent — “because its fault is the graver. As one says ‘stinking ghee’ — where evil holds and evil and the rest, this alone is diminishment, while the rest are superimposition.” All ghee has a smell, and only the most stinking is called stinking ghee; among the five views, this one alone is pure diminishment (calling what is, not), while the other four are still superimposition.

Adherence to views is “to call the inferior superior,” and ‘inferior’ has a definition — “what has outflows is called inferior, since it is what the noble path abandons.” Adherence to observances and vows is “the view that holds what is neither cause nor path to be cause and path,” and the treatise gives three classes of instance: reckoning Maheśvara, the Lord of Creatures, and the rest to be the cause of the world; reckoning perverse practices such as throwing oneself into water or fire to be the cause of birth in heaven; and reckoning the mere keeping of observances and vows, together with the cognitions associated with it, to be the path of liberation.

The easy error: to run adherence to views and adherence to observances and vows together. What each holds runs in a different direction — adherence to views raises the inferior into the superior, adherence to observances and vows takes what is neither cause nor path for cause and path; the one errs about worth, the other about the road. In practice: under both of the latter two views the treatise adds the clause “in strictness the name ought to be set up as adherence to views and the rest” and “in strictness the name ought to be set up as adherence to observances and vows and the rest,” showing that the current names are abbreviations with the word ‘and the rest’ dropped. Whenever this treatise says ‘in strictness the name ought to be so-and-so,’ it is reminding the reader: the current term is an abbreviation; do not use the abbreviation to bind down its range.

VIII. Why Holding What Is Not a Cause to Be a Cause Is Abandoned by Seeing Suffering Alone

First the question: adherence to observances and vows holds what is not a cause to be a cause, and ‘cause’ is precisely what the truth of origination treats; why then is it not abandoned by seeing origination? The author raises this question himself, and the answer lies in one sentence about the order of cause and effect — “one must first deliberate that its entity is eternal, one, a self, and an agent, and only then does the holding to a cause arise.”

That is to say, a man who reckons Maheśvara to be the cause of the world does not first have an error about ‘cause’; he first has an error about ‘eternal’ and ‘self’: he first settles it that the thing is eternal, is one, is a self, is an agent, and only then takes it for a cause. This being so, “as soon as suffering is seen, the holding to eternity and the holding to self with respect to Maheśvara and the rest are forever cut off without remainder, and so the holding to a cause produced by them is cut off as well” — at the seeing of suffering the root is pulled up, and the branches fall of themselves.

The easy error: to suppose that every error about causes is assigned to abandonment by seeing origination. It is not so. When this treatise adjudges the category of abandonment, what it looks at is where that error arises from, not what topic it discusses.

The treatise then unfolds a long string of interrogations and comes to no conclusion. First an objection: then those who hold that perverse practices such as throwing oneself into water or fire are the cause of birth in heaven, and that keeping observances and vows suffices for purity, have not first reckoned eternity and self — must they not then fail to be abandoned by seeing suffering? Next it cites what the treatise says of the ox-observance, the deer-observance, and the dog-observance, adjudging them “adherence to observances and vows, abandoned by seeing suffering.” Then it presses itself, “on what ground is this abandoned by seeing suffering?”, and to answer ‘because it is bewildered about the truth of suffering’ would “have the fault of overreach, since all afflictions taking what has outflows as object are bewildered about suffering.” It presses further as to when adherence to observances and vows counts as abandoned by seeing the path, why the wrong view and the doubt that take the path as object cannot amount to a holding to purity, and why a wrong view and the rest taking as object what is abandoned by seeing origination and cessation, when held to be the cause of purity, should not belong to the abandonment of those. At the end it leaves only one sentence — “therefore the sense here held should be further examined.”

In practice: what is most precious in this passage is not the answer but the author’s willingness to leave a question that has not yet been closed lying on the page just as it is. In reading Abhidharma one must be able to recognize passages of this kind: whatever closes with words of the class ‘should be further examined’ is a place the treatise itself marks as unsettled, and there is no need to supply a conclusion for it.

IX. The Entity of the Four Inversions — Why Only a Portion of Three Views Is Taken

First the question: the inversions are four (eternity · pleasure · purity · self) and the views are five; how is the entity of the four inversions fixed? The treatise answers, “the entity of the four inversions is set up from three views”: within the extreme view only the view of eternity is taken as the inversion of eternity; within adherence to views those that reckon pleasure and purity are taken as the inversions of pleasure and of purity; within the view of a real person only the view of self is taken as the inversion of self.

What matters is the words ‘only taken.’ It is not that the three views are inversions entire, but that from each of the three a portion is taken. The treatise then explains why the rest do not enter — three causes must be complete, “because they are inverted in every part, because they are of the nature of deliberative judgment, because they are false superimposition.” Adherence to observances and vows does not enter, because “it is not inverted in every part, since a small portion of what it takes as object is pure” (the keeping of observances and vows has itself some small portion of purity to take as object); the view of annihilation and wrong view do not enter, because “they are not false superimposition, turning by the gate of nonexistence” (they are diminishment, not superimposition); the remaining afflictions do not enter, because “they cannot deliberate and judge, not being of the nature of view.” The three causes are like three sieves, and what is sifted out is exactly that portion of those three views.

The treatise also handles an apparent conflict between sūtra and treatise. The sūtra says “to reckon the impermanent as eternal — there is the inversion of perception, of mind, and of view,” so that perception and mind too are called inversions; the treatise says that view alone is inversion. The author explains it through thus — “perception and mind, following upon view, are likewise given the name of inversion, since they are associated with view and share its mode of apprehension”: perception and mind take the name in the train of view. Someone asks further why feeling and the rest are not so spoken of, and the answer is “because they are not well established in the world” — the inversion of mind and the inversion of perception are usages everyone recognizes, and feeling and the rest are not.

The easy error: to read ‘taking the name in the train of view’ as making perception and mind also the entity of inversion. They are not. This is exactly the point where this section parts from the account of the twelve inversions held by another school, for which see Further Resonances. In practice: the standing practice of this treatise for handling ‘the sūtra says thus while the treatise says otherwise’ is just this move — it does not alter the text of the sūtra, it alters the reading: it grants the name the sūtra uses, and explains that the name is set up in the train of something else.

X. A Stanza of Ānanda’s — Has the Inversion of Perception Been Cut Off or Not

First the question: since it was said above that the inversions of perception and of mind are cut off in the train of view, and that the stream-enterer has cut them, why then did the Venerable Ānanda say to the Venerable Sovereign in Eloquence, “because there is a disordered inversion of perception, your mind is scorched; having gone far from that perception, greed is stilled and mind becomes pure”? This stanza is plainly addressed to one who has already entered the station of the noble, and yet it says that he still has a ‘disordered inversion of perception.’

The treatise does not here force an explanation, but cites the account of one master by way of explaining it through — “some other master further says: the eight inversions of perception and of mind are not entirely cut off by one in training. Eight such entanglements are forever cut off only by knowing the noble truths as they really are, and apart from this there is no other means of forever cutting them; therefore what is said here does not contradict that sūtra.” That is: one at the station of training has not entirely cut off the eight inversions of perception and mind; they are forever cut off only upon knowing the noble truths as they really are, and besides this there is no other means of forever cutting them. Sūtra and treatise thus each receive their place.

The easy error: to read these twenty characters as a root stanza of this treatise. They are not — they are a stanza the treatise quotes, and the preceding text has a plain warrant, “if that is so, why then did the Venerable Ānanda say to the Venerable Sovereign in Eloquence.” Fabao’s Commentary marks it here as “the author cites Ānanda’s stanza as an objection,” and Yuanhui’s Verse Commentary marks “hence a stanza is cited that says”; both ancient masters adjudge it a quoted stanza.

In practice: in reading the stanzas of this treatise, look first at whether there is a warrant before it of the class ‘as a stanza of the sūtra says,’ ‘he said to so-and-so.’ What has a warrant is a quoted stanza; only what has no warrant and follows upon ‘the stanza says’ is a root stanza — a distinction that looks trifling and yet directly decides how many stanzas a chapter has.

XI. The Seven Prides and the Nine Classes of Pride — Seven Postures of Exalting Oneself

First the question: pride is merely holding oneself high; why must it be divided into seven? Because the objects met with differ and the postures struck differ. The treatise first gives the general mark — “what makes the mind exalt itself is given the general name of pride; because the modes in which it operates differ, it is divided into seven” — and then gives the particular marks one by one.

Pride and excessive pride make one pair: toward the inferior and toward the equal, to call oneself superior and to call oneself equal, is pride; toward the equal and toward the superior, to call oneself superior and to call oneself equal, is excessive pride. Pride beyond excessive pride goes one step further — “toward the superior to call oneself superior is pride beyond excessive pride.” The pride of self is “upon the five aggregates of appropriation to hold to self and what belongs to self, making the mind exalt itself.” The pride of superior attainment is “to call oneself to have attained, among excellent qualities not yet attained.” The pride of slight inferiority is the most winding — “toward one superior by much, to call oneself but slightly inferior,” plainly far behind and yet saying one is only a little behind. Wrong pride is “to call oneself possessed of quality where one has no quality.”

The nine classes of pride flow out from three of the seven prides (pride · excessive pride · the pride of slight inferiority), “if they arise in dependence upon view, the order in which they operate differing,” making three sets of three. The treatise gives the three matchings: the first three are, in order, excessive pride, pride, the pride of slight inferiority; the middle three are, in order, the pride of slight inferiority, pride, excessive pride; the last three are, in order, pride, excessive pride, the pride of slight inferiority.

A question the treatise poses itself shows great fineness: the pride of slight inferiority still has a place of exalting to speak of (within the superior one still sets oneself as only slightly inferior, and so has still something to lean on); what then is the place of exalting in ‘the pride that there is none inferior to me’? The answer runs, “that is, toward such a host of superior sentient beings as one loves and delights in, though one knows one’s own person to be utterly inferior, one still holds oneself in esteem” — knowing plainly that one falls short and yet still esteeming oneself, because one is of a kind with the host one reveres. The easy error: to suppose that pride must boast. This instance proves otherwise: to set oneself as inferior and yet still esteem oneself is pride all the same.

At the end the treatise records the divergent gloss of two treatises: “thus far the gloss follows the Treatise on the Development of Wisdom”; and if one follows the Treatise on Categories of Dharmas, then the pride ‘I am superior’ flows out from three prides — pride, excessive pride, and pride beyond excessive pride — “because the objects observed differ as inferior, equal, and superior.” Both glosses are recorded together. In practice: the crux of pride lies not in what is said but in whether the mind has ‘exalted’ itself. This rod is far finer than ‘whether one has boasted.’

XII. In the Noble One They Are and Yet Do Not Arise — One Exception among What Is Abandoned by Cultivation

First the question: pride extends to abandonment by seeing and by cultivation; will then that portion of it abandoned by cultivation be active in a noble one before it has been cut off? The treatise answers “this is not settled,” and gives an instance anyone can follow — the entanglement of killing is abandoned by cultivation, and yet “the noble ones are necessarily not active in it.”

The whole teaching value of this section lies in that ‘why.’ The reason the stanza gives is “because they are increased by view and doubt”: these afflictions, though abandoned by cultivation, are increased by view and doubt in person; and since at the path of seeing the noble one has already cut off view and doubt, “because the backbone of view and doubt has been broken, the noble one cannot give rise to them” — the spine that propped them up is snapped, and they cannot stand up. The treatise matches them item by item: the classes of pride and the pride of self are increased by the view of a real person, the entanglements of killing and the rest are increased by wrong view, craving for nonexistence is increased by the view of annihilation, one portion of craving for existence is increased by the view of eternity, and unwholesome regret is increased by doubt.

Two terms may be learned in passing. “Craving for nonexistence” is not ‘the absence of craving’; it is a greedy seeking after this very fact that ‘the three realms are impermanent’ — “that is, the three realms are impermanent, and to seek greedily upon this is called craving for nonexistence.” “One portion of craving for existence” is “to wish to become the great nāga king Airāvaṇa and the like,” to wish that one may in a coming life become some excellent kind of sentient being.

The easy error: to read ‘the noble one has them and yet they do not arise’ as ‘the noble one has cut them off.’ It is exactly the opposite — what the treatise says is not cut off and yet not arising. He has them, and they cannot get up. In practice: this is the second redemption of the distinction made in section three (that ‘there is’ and ‘is active’ are two different things). The sentence usable in practice is: to cut off an affliction one need not always counteract it directly; cut off what feeds it, and it cannot rise of itself.

XIII. Gate One · All-Pervading and Not All-Pervading — the Eleven That Can Take In All Five Categories

From “among the ninety-eight latent tendencies, how many are all-pervading and how many not all-pervading?” onward, this fascicle turns into the distinctions through the several gates, of which it covers the first five.

First the question: what is meant by all-pervading? What the treatise gives is not one definition but three conditions — “taking as object, throughout, the dharmas of the five categories of its own realm and level” (able to take all as object), “as a latent tendency being a cause and producing throughout the defiled dharmas of the five categories” (able to produce throughout), and “upon these three senses the name all-pervading is set up.” Those that meet these three senses are only eleven: “the views and doubt abandoned by seeing suffering and origination alone, and the independent ignorance associated with them,” which counted out are “seven views, two doubts, two ignorances.”

One question the treatise puts to itself shows the finest sense of measure: this taking of the five categories as object — is it gradual or all at once? If gradual, the other latent tendencies can also take them gradually, and then everyone is all-pervading; if all at once, who is able at one time to reckon all the dharmas of the desire realm as superior, as pure, as the cause of the world? The treatise’s way of answering is to narrow the claim — “it is not said that it takes as object at once everything of its own realm and level; it is said that it has the power to take the five categories as object at once”: not that it exhausts all dharmas in one moment, but that it has the capacity to reach the five categories at one time. The easy error: to read ‘throughout’ as ‘the whole of.’ The throughout here is said with respect to category, not with respect to dharmas.

Nine can take an upper realm as object. Of the eleven, setting aside the view of self and the extreme view, the other nine can further take an upper realm and an upper level as object. The treatise says, “the word ‘upper’ properly makes plain the upper realm and the upper level, and at the same time displays that there is no latent tendency taking a lower one as object” — the one word ‘upper’ carries along with it the point that ‘there is no latent tendency taking a lower one as object.’ And it adds a restriction: “although these nine can take both their own and an upper level as object, in principle there is no taking of one’s own and an upper level as object at once.” As for why the two views of self and of extremes cannot take an upper level as object, the reason is clean — “because he does not hold that to be his self or what belongs to his self, and because the extreme view must arise from the view of self”: the view of self takes as object the aggregates affiliated to oneself, and the extreme view follows upon the view of self, so neither can go up. If it be asked what kind of view it is to reckon Mahābrahmā a sentient being or eternal, “the Abhidharma masters say: these two are not views, but are gathered into wrong cognition,” and as for why, the treatise says plainly “taking our own tenet as the measure, it is so stated.”

The entity of the all-pervading is not the latent tendencies alone. The treatise says one must take “together with the dharmas that accompany them,” yet must exclude their ‘acquisition,’ “since it is not of one result with them.” At the end it appends a fourfold alternative: are all all-pervading latent tendencies all-pervading causes? The first alternative is the all-pervading latent tendencies of the future; the second is the dharmas coexistent with them in the past and present; “the third and fourth should be analyzed in accordance with the principle.” In practice: this gate of all-pervading is the foundation of the gates that follow — the gate of increase by attachment will use it, and the sense of taking an upper realm as object will use it too, so it is worth spending a little more effort upon in reading.

XIV. Gate Two · Which Afflictions Can Take the Dharmas without Outflows as Object

First the question: can afflictions take nirvāṇa and the noble path as their object? The treatise answers that they can, but only six — “the wrong view and doubt abandoned by seeing cessation and the path alone, and the independent ignorance associated with them, three each making six.” The rest take the dharmas with outflows as object, and “this is established of itself by extension.”

Taking cessation as object and taking the path as object differ greatly in range. Those that take cessation as object “each take the cessation of their own level as object”: the three afflictions of the desire realm take only the cessation through discernment of the formations of the desire realm as object, and so on up to the three of the summit of existence taking only the cessation of the summit of existence; the reason is “because the cessations are not causes and results with respect to one another” — the cessation of this level cannot be a cause of the cessation of that level, and so it cannot reach across. Those that take the path as object are far wider, “taking six levels or nine levels as object”: the three afflictions of the desire realm take as object only the path of the class of the cognition of dharma in six levels, and the three each in the upper eight levels take as object throughout the path of the class of the subsequent cognition in nine levels; the reason is “because they are of one class” and “because the paths of the several levels are causes of one another” — the paths are homogeneous causes of one another, and so they can reach across.

The treatise then handles a difficulty of symmetry: since the class of the cognition of dharma can counteract form and formlessness, why is it not the object of the three in each of those eight levels? The answer has two layers — “because the classes of the cognition of dharma concerning suffering and concerning origination are not their counteragent” (it is not a universal counteragent) and “because it cannot counteract what is abandoned there by seeing” (it is not a counteragent in the whole part). “These two beginnings being absent, it is not their object.”

As for why greed, hatred, pride, and the two adherences do not take the dharmas without outflows as object, the treatise gives five reasons, each different from the others, and this is the best teaching material in the section: for greed, “because what it takes as object should be relinquished,” whereas what is without outflows should not be relinquished (if it took what is without outflows as object it would be like the desire for wholesome dharmas, and no fault would follow); hatred takes enmity and harm as object, and “cessation and the path are not enmity”; pride takes what is coarse and agitated as object, and “cessation and the path are quiescent”; adherence to observances and vows holds what is not pure to be the cause of purity, and “cessation and the path are truly pure”; adherence to views holds what is not superior to be superior, and “cessation and the path are truly superior.”

The easy error: to take these five as five statements of one reason. They are not — each of the five fastens upon its own mode of apprehension. In practice: from this a general rule can be inferred in reverse: whether an affliction can take a given object depends on whether that object has the handle the affliction needs. Upon the two truths of cessation and the path there is no enmity, no coarse agitation, no defect that can be held as pure or as superior, and so these four afflictions have nowhere to lay hold; only wrong view and doubt need no handle — denying outright and hesitating can be done to anything.

XV. Gate Three · Increase by Attachment through the Object and through Association

First the question: under what conditions does a latent tendency count as ‘increasing by attachment’ upon a given dharma? The treatise divides into two roads: increase by attachment through the object, and increase by attachment through association.

The road of the object is first given in general: an all-pervading latent tendency “increases by attachment through the object upon the dharmas of all five categories of its own level throughout,” and one not all-pervading “only upon its own category.” Then it is given in particular — the six afflictions that take the dharmas without outflows as object and the nine that take an upper realm as object have, “upon their object-fields, no sense of increase by attachment.”

The two similes of this section are excellent and repay reading word for word. Why is there increase by attachment? “That is, if there is a dharma appropriated as its own by the view of self and the craving within this level, then there may be … the principle of increase by attachment through the object, as when a garment is damp the dust follows and stays upon it” — only when the garment is damp can the dust cling to it, and ‘appropriated as its own’ is that layer of dampness. Why is there no increase by attachment? The dharmas without outflows and those of an upper level are not appropriated as its own by the view of self and the craving of a lower level, and are moreover contrary to the lower affliction that takes them as object, “as upon a burning stone the foot does not stay” — the stone scalds the foot, and it cannot stand there. The one says it clings, the other says it does not, and the two similes make an exact pair. The treatise also records alongside another master’s account, which glosses ‘latent tendency’ by ‘following in accord,’ and gives as its simile “as when one with the wind-sickness takes a dry and astringent medicine, the sickness is not what is increased by attachment upon the medicine.”

The treatise also clears the ground, in passing, of one easily confused movement of mind: “to dwell in a mind of the lower level and seek the upper level and the like is the desire for wholesome dharmas, and is not called a latent tendency” — to aspire to an upper level is not an affliction.

The road of association is concise: “whatever latent tendency it be, upon the dharmas associated with it, by reason of association it increases by attachment upon them.” And the treatise recalls one presupposition — “all statements of increase by attachment are said with reference to what has not yet been cut off, and so the head of the first stanza marks the words ‘not yet cut off’”: every statement of increase by attachment is bounded by ‘not yet cut off.’ At the end it sets a question and answer: is there a latent tendency that takes neither the dharmas without outflows nor an upper realm as object, and yet has increase by attachment through association only and not through the object? “There is: the all-pervading latent tendencies that take an upper level as object.”

The easy error: to read increase by attachment as ‘taking as object.’ The two differ — what can be reached as object may still not be increased upon. In practice: what this gate teaches is the condition for the growth of afflictions, and the condition is only one — the object must be something appropriated as its own by the holding to self and by craving. What has never been taken for ‘mine’ an affliction may take as object and still grow no fat upon.

XVI. Gate Four · Which Are Unwholesome and Which Are Indeterminate

First the question: ought not afflictions all to be unwholesome? The treatise answers that it is not so, and first gives a general adjudication of the upper two realms — “all latent tendencies of the form and formless realms are of indeterminate quality alone.” The reason is a chain of causal inference: “because if a defiled dharma were unwholesome it would have a painful maturation; and a painful maturational result the upper two realms do not have, since the cause that vexes and afflicts another is definitely absent there.” The upper two realms have no painful maturational result, and no cause of vexing and afflicting others, and so the afflictions there can only be indeterminate.

Within the desire realm too there are two exceptions — the view of self, the extreme view, and the delusion associated with them. The treatise gives three reasons. First, “because these do not go against giving and the rest, for it is in order that ‘I’ may obtain happiness hereafter that one now diligently cultivates giving, discipline, and the rest”: it is for the sake of ‘self’ obtaining happiness hereafter that one now cultivates giving and discipline, so that the view of self runs together with giving and discipline without conflict. Second, the extreme view of annihilation “can accord with liberation,” and the treatise cites the World-Honored One’s words in evidence — “among the several courses of view of the several non-Buddhists, this view is the best, namely that ‘I am not and what belongs to me is not.’” Third, “because these two views are bewildered about one’s own affair alone, and do not seek to vex and injure other sentient beings”: they are only bewildered about one’s own affair, and do not intend to harm others.

The treatise at once poses itself an objection: if the reason be so, then greedily seeking the pleasures of heaven and giving rise to the pride of self likewise do not harm others — should they too not be indeterminate? This objection is raised and not concluded, and there is then recorded alongside an account of the preceptors of former times — “an innate view of self is of indeterminate quality, as the view of self is active in beasts and the like; if it arises from discrimination it is of unwholesome quality,” dividing by the innate and the discriminated. The remaining latent tendencies affiliated to the desire realm “are contrary to the above and are all of unwholesome quality.”

The easy error: to read ‘indeterminate’ as ‘harmless.’ It is not. Indeterminate speaks of whether a painful maturational result is called forth, not of whether it obstructs the path. In practice: this gate gives a very serviceable rod — in adjudging the good or ill of a movement of mind, ask first whether it calls forth a painful maturation and whether it vexes and afflicts others, rather than asking first whether it looks like an affliction.

XVII. Gate Five, Part One · The Three Unwholesome Roots

First the question: unwholesome afflictions are so many; why are only three called ‘roots’? The treatise’s answering sentence is very short — “all greed and hatred affiliated to the desire realm alone, and unwholesome delusion, are gathered into the unwholesome roots” — and it says, “an affliction whose quality is unwholesome alone is set up as an unwholesome root, being the root of unwholesome dharmas.” To be a root, two conditions are needed: the entity itself must be unwholesome alone (not extending to the indeterminate), and it must be able to be the root of the other unwholesome dharmas.

The treatise also adds an instance of its abbreviating manner: “the rest are not so; that the remaining afflictions are not unwholesome roots is established by extension of the sense already given, and so the stanza does not state it” — that the rest are not roots is plain by extension from the text above, and so it is not stated again in the stanza.

The easy error: to suppose that ‘root’ means the gravest affliction. It does not; root is said with respect to ‘being able to produce,’ not with respect to ‘being the coarsest and heaviest.’ Pride too is grave, and yet is no root. Puguang here cites the five senses given by the Mahāvibhāṣā to make plain why these three alone are set up as roots — “they extend to the five categories, are present throughout the six consciousnesses, are of the nature of latent tendency, can set going coarse and evil bodily action and verbal action, and make the firm and forceful preparatory effort that cuts off the wholesome roots” — only where the five senses are complete is a root set up.

In practice: from the word ‘root’ one may carry away a reminder for practice — to counteract an affliction one must find its root, and the mark of a root is ‘other afflictions grow out of it’, not ‘it is what makes one most uncomfortable.’

XVIII. Gate Five, Part Two · The Three Indeterminate Roots and the Four of the Masters of Other Regions

First the question: does the indeterminate too have roots? The Vaibhāṣika masters of the land of Kaśmīra say that it has three — “the three, indeterminate craving, delusion, and discernment; and even the portion produced by maturation is gathered into the indeterminate roots.”

The treatise asks itself why doubt and pride do not enter, and the answering words are full of picture — “because doubt turns toward two courses and pride turns by exalting.” Doubt swings back and forth between two roads, “being of a wavering nature, and so should not be set up as a root”; pride “turns upon its object with the mark of exalting, being unlike the dharma of a root.” And the treatise gives the image on which this rests: “for what is to be a root, it must abide firmly and turn downward, as the world commonly understands” — a root must strike downward and must stand firm; the one wavers and the other tilts upward, and neither looks like a root.

The masters of other regions, however, set up four — indeterminate craving, view, pride, and delusion. Their reason is stated from the actual condition of those who cultivate absorption: “because the several foolish men who cultivate the absorptions of the upper realms depend on nothing beyond the three of craving, view, and pride; and these three all turn by the force of ignorance.” Craving, view, and pride are what ordinary beings who cultivate the upper absorptions lean upon, and these three all turn in dependence upon ignorance, making four roots in all. The treatise also glosses the name ‘indeterminate’ — “within the name indeterminate the wholesome and the evil are both blocked out.”

The easy error: to suppose that in three and four someone has miscounted. It is not so. The two houses take different angles: the one takes the mark of a root, ‘whether it can abide firmly and turn downward’; the other takes ‘what those who cultivate the upper absorptions actually lean upon.’ Both parties are recorded without adjudication, and this treatise passes no verdict here.

In practice: what most repays turning over in this section on the indeterminate roots is that it grants that movements of mind that are indeterminate also have roots — those inconspicuous movements that call forth no painful result have their own machinery of growth all the same. In practice one cannot watch only what is conspicuously unwholesome.

XIX. 〔A Passage of Prose Commentary Set Apart〕The Four Gates of Determination — How to Answer a Question

The treatise turns into this section from the question of ‘the fourteen indeterminate matters,’ and first draws the boundary: the indeterminate of that sūtra “sets up the name indeterminate only with respect to a question that should be set aside,” and is not the same thing as the indeterminate of the three ethical qualities in the gate before. First the question: since the name is the same and the sense different, why gather it in here? Because at the sight of the two words ‘indeterminate’ the reader is bound to doubt, and the treatise simply finishes the matter off.

The four determinations are: categorical determination, determination by analysis, determination by counter-question, determination by setting aside, each matched to one question — death, birth, superior and inferior, and whether self and aggregates are one or different. The four instances the treatise gives are clean enough to be committed to memory. Asked “will all sentient beings die?”, categorical determination: all will die. Asked “will all who have died be born again?”, determination by analysis: “those who still have afflictions will be born again, and not the rest.” Asked “is a human superior or inferior?”, determination by counter-question: “compared with what?” — compared with what? Compared with gods, a human is inferior; compared with the lower destinies, a human is superior. Asked “are the aggregates and a sentient being one or different?”, determination by setting aside, the reason being “because a sentient being is not really existent, the natures of oneness and difference are not established,” with a simile — “as with the whiteness or blackness of the son of a barren woman”: of what does not exist, black and white cannot be spoken.

Why is setting aside called a determination? The treatise answers itself, “because in determining his question one says that this ought not to be determined” — to utter the words ‘this ought not to be determined’ is itself a determination.

The latter three questions each carry a round of an outsider’s objection and the author’s explaining through, and these three rounds repay looking at one by one, for they demonstrate three different manners of argument. First, the objector says the second question may also be determined categorically (answering straight out, ‘not all will be born again’); the author explains through: what the questioner asked was ‘all,’ and a general answer, “though it makes him know a general statement, still leaves him unresolved.” Second, the objector says the third question may also be determined categorically (a human is both superior and inferior, “because what he is set against differs, as consciousness is result and cause”); the author explains through: the questioner puts a categorical question and your answer is not categorical, so it can only amount to determination by analysis; “but here one should press him as to what his question sets it against, and hence this is called determination by counter-question.” Third, the objector says that since the fourth question determines nothing at all, how can it be called a determination; the author explains through: “yet what he asks should in principle be set aside. Having said that it should be set aside, how is that not called a determination?”

The Abhidharma masters further explain the four determinations at length according to the four questions, and their passage on determination by analysis is the longest string of layered pressing in the whole fascicle — dharmas are past, future, and present, and he takes the past; the past has form up to consciousness, and he takes form; form is wholesome, evil, and indeterminate, and he takes the wholesome; the wholesome has seven, and he takes abstention from killing; abstention from killing is set going by the three wholesome roots, and he takes what is set going by non-greed; what is set going by non-greed has the manifest and the nonmanifest — six layers of division all the way down. Set against it, determination by counter-question divides not a single step: “one should not analyze for him, but only bring him at last to stand silent.” The watershed between the two lies not in the question but in the mind that asks — the treatise says it with the utmost plainness: “because the intent of the question differs as straight or fawning, and the determination differs as analyzed or unanalyzed.” To one who comes to ask with a straight mind one analyzes; to one who comes with a fawning mind one only counter-questions.

The treatise also poses itself an objection: within determination by analysis and determination by counter-question “there is no question at all, only a request to be told,” nor is there “any determination, but only the counter-question ‘which one do you wish told?’” — how then do these count as question and determination? The answer: to request, saying ‘tell me the path,’ is that not asking about the path? and to determine his question by counter-questioning, is that not determining the path?

At the end it analyzes the marks of the four determinations once more according to the sūtra, and the four instances in the sūtra differ from those given above by the Abhidharma masters — categorical determination instances “are all the formations impermanent?”; determination by analysis instances “having done actions performed with deliberate intent, what result is to be experienced?”; determination by counter-question instances “are a man’s perception and his self one or different?” (to which one should counter-question, “in dependence upon which self do you ask thus?”, and if he says in dependence upon the coarse self, then “it is determined that it differs from perception”); and determination by setting aside enumerates the fourteen matters one by one: the four alternatives of the world’s being eternal or not eternal, the four alternatives of the world’s having an end or not, the four alternatives of the Tathāgata’s existing or not existing after death, and the two alternatives of the life-principle being the body or being other than the body.

The easy error: to read determination by setting aside as the Buddha’s not knowing or not being willing to say. The reason the treatise gives is one and the same from beginning to end — the presupposition of what is asked is not established (“because a sentient being is not really existent, the natures of oneness and difference are not established”), and the question itself is a bad one. In practice: this section is in truth a treatise on the art of answering. What it teaches is: before answering, adjudge the type of the question — whether the question is well formed, whether the asker is sincere, whether what is asked needs to be divided into cases. The division of the four determinations comes before any answer whatever.

Further Resonances: The Mirror of the Three Commentaries

The above is the sense of the treatise text of this fascicle itself. Below a further section is set up, gathering 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 another way of reading the same passage — recorded together without adjudication, each restored to its own place.

1. Where the Three Commentaries Diverge

One commentary altogether absent. The commentaries on this fascicle stand in a particular case: Shentai’s Commentary is a fragment, its seven surviving fascicles belonging to the chapters on the elements, on the faculties, and on the worthy and the noble, the chapter on the latent tendencies being wholly lacking; and so this fascicle has no word-by-word gloss of Shentai’s to hang upon. Under mechanical rechecking, the marker sentences of this chapter yield a uniform zero result in Shentai. Only the general conspectus in his first fascicle retains the one phrase ‘the chapter on the latent tendencies, sixty-nine lines,’ and that is a table of the stanza counts of the whole treatise, not a gloss on the chapter on the latent tendencies, and cannot be made to serve as evidence of outlining. The lack is marked as it stands, and no text from another fascicle is put in its place.

The chapter outline: a rare evenness, and one independent divergence. Puguang and Yuanhui adjudge the chapter-level division as twofold in the same way (explaining the entity of the confusions / explaining the extinction of the confusions), and even anchor the head of the second main division to the same sentence of the treatise, ‘now this should be examined’ — the first time in this collection that all three houses agree at the chapter level. Fabao stands apart: he changes ‘the entity of the confusions’ to ‘the entity and function of the confusions’ and ‘the extinction of the confusions’ to ‘the differences of their abandonment,’ and for the subdivision of the chapter on the entity and function of the confusions he does not use the division ‘the root confusions / the distinctions through the several gates,’ but instead a fourfold ‘explaining the latent tendencies / miscellaneous explanation of the confusions / analysis through the gates of sense / a separate explanation of the five hindrances.’ The names are the same and the width of the outline levels differs. From this a rule for reading the commentaries follows: Fabao’s chapter-internal numbering forms a system of its own, and when his ‘the nth’ is cited it must be tied to his house and never made interchangeable with Puguang’s or Yuanhui’s. Recorded without adjudication.

The gloss on the chapter name: each house takes one sense. Puguang takes “following along after sentient beings is called ‘following’; a mode of apprehension that is subtle is called ‘slumber,’” with sleep for a simile — “as when a man sleeps, his mode of apprehension is hard to know”; Fabao takes “a subtle mode of apprehension, binding along and increasing by attachment”; Yuanhui takes “following along after sentient beings, because it increases their dimness and stagnation.” The three houses point at the same thing, and the face each picks out is different: one stresses that it is hard to know, one that it binds along and increases by attachment, one that it increases dimness and stagnation. Recorded without adjudication.

As for why the latent tendencies are explained next after the chapter on action, the two commentaries each give their own reason. Puguang says, “because action, as a cause calling forth results, cannot arise alone but must depend on the condition of confusion, and confusion is the remoter with respect to the result, the latent tendencies are spoken of afterward” — confusion is farther from the result, and so is ranged later; Fabao says, “because action is particular and confusion general, action is explained first and the latent tendencies glossed after” — action is particular and confusion general, and so the particular comes before the general. One and the same order, two different reasons, recorded without adjudication.

The two glosses on the view of a real person: three degrees of fineness in outlining. The treatise itself plainly has two glosses here — “because it perishes it is called sat” and “because it exists it is called sat.” Puguang divides into two sections and ties each to its house, marking the former “the Sautrāntika masters’ gloss” and adjudging it appositional-determinative, and marking the latter “second, the Vaibhāṣika masters’ gloss”; Fabao likewise divides into two, marking “the author cites the Sautrāntika gloss” and “the author cites the Sarvāstivāda gloss,” and supplies besides the Treatise Conforming to Correct Principle’s counter-question to the former — “if so, what use is there in marking it with the word sat? The word kāya alone suffices to block eternity, and so the name ought to be set up as the view of kāya only. No dharma that is eternal can be aggregated; what use is it to mark the word ‘perishing’ upon the person?” — a counter-question nowhere cited by Puguang; Yuanhui makes no division of sections and writes them out one after another. ⇒ One and the same passage cut three ways: into two sections, into two sections with a supplement, and written out continuously. Recorded without adjudication.

The Vaibhāṣikas’ not granting the account of twelve inversions: Fabao is the most schematic. Puguang here glosses sentence by sentence, and first ties the school of the divergent account of twelve inversions to the Masters of Distinction (“following the Mahāvibhāṣā and the Treatise Conforming to Correct Principle, they are the Masters of Distinction”), and explains on their behalf the division of eight abandoned by seeing alone and four extending to abandonment by seeing and by cultivation. Fabao, on the same passage, picks out a threefold outline — “from ‘the Vaibhāṣika masters do not grant this sense’ below there are three: first counter-objecting to refute the other school, second explaining in accordance so as to establish his own tenet, third properly explaining through the other school’s objection” — pointing out at one stroke the three stages of the Vaibhāṣikas’ answer. The doctrinal content of the two commentaries agrees entirely; only the fineness of the outlining differs. Recorded without adjudication.

What office each of the stanza’s four lines performs: a stroke that is Yuanhui’s alone. Yuanhui, glossing the one stanza on the four inversions, says “the first two lines set out the entity, the third line establishes and rejects, the fourth line explains the sūtra through” — the first and second lines set out the entity of the four inversions, the third states why these three views alone are taken (establishing and rejecting), and the fourth explains through the inversions of perception and of mind spoken of in the sūtra. This angle neither Puguang nor Fabao makes explicit, and it is of great use to the reader: in a stanza of twenty characters, each of the four lines has its own office, and to understand the offices is to understand the structure. This is an original stroke of outlining by a commentator, and does not enter the main text of the treatise.

The manner of setting out the positions of the three schools: a difference of form with attributions entirely agreeing. Puguang, in the dispute over the association of the latent tendencies, marks the school at every sentence, tying each question and each answer to its house; Fabao instead sets out at the head of the passage one general table of the three schools — “the Sautrāntika tenet: what is active is called entanglement, the seed is called latent tendency. The Mahāsāṃghikas and the rest: the entity of a latent tendency is a dissociated formation, that is, the latent tendency belonging to greed for desire. The Sarvāstivādins: greed for desire is itself the latent tendency” — a form far more schematic than Puguang’s. What matters is this: the attributions of the two commentaries agree entirely, without a single divergence; what differs is only the form of setting them out.

One attribution that must be shown as such. The party that objects in this fascicle to the account that the latent tendencies are associated is not marked in the treatise itself with the name of its school; Puguang, however, marks it plainly as “the Mahāsāṃghikas and the rest” (as in “‘does this not contradict the sūtra?’ — the Mahāsāṃghikas and the rest press” and “‘by what principle is it proved that they are definitely associated?’ — the Mahāsāṃghikas and the rest reproach”). ⇒ This ascription is what a commentator ties, not the plain text of the treatise, and in citing it one must show whose tying it is. Recorded without adjudication.

The nine classes of pride: no divergence at all among the three houses. Puguang, Fabao, and Yuanhui agree item by item on the listing of the nine classes of pride and on the framework of their flowing out from three of the prides, and Fabao besides cites the Treatise Conforming to Correct Principle glossing the word ‘class’ — “class is the sense of category, that is, the differentiations of pride.” Here the three houses have nothing to contend over, and this is recorded as it stands.

In the attribution of the passage on the all-pervading there is one further divergence of angle. Puguang marks the school at every sentence of the four rounds of exchange over whether craving and pride too are all-pervading — the Sautrāntikas object, the Vaibhāṣikas reproach in return, the Sautrāntikas answer, the Vaibhāṣikas state their own tenet; whereas Saṅghabhadra, citing one of those sentences, calls it outright the Sūtra-master’s own interrogation (see below). One and the same sentence, and the two houses tie it to different speakers, recorded without adjudication.

2. Saṅghabhadra’s Refutations and Supplementary Explanations

Saṅghabhadra’s two treatises do not name this chapter ‘the chapter analyzing the latent tendencies’ but ‘the chapter adjudicating the latent tendencies,’ the one as the fifth and the other as the sixth. Checking item by item with the marker of quotation he habitually uses, within the range of this fascicle the Treatise Conforming to Correct Principle quotes Vasubandhu’s sentences in nine places, while the Treatise Revealing the Doctrine, within the same range, has not one word-for-word quotation — and this is an honest negative proposition, which can prove only one thing: that the form of the Treatise Revealing the Doctrine is the more concise and that it does not counter-refute sentence by sentence in this passage; one cannot infer from it that that treatise has no divergent sense.

The place where the quotation is fiercest is precisely that sentence in which this treatise appraises and takes the Sautrāntikas. The author says, “the entity of a latent tendency is neither associated with mind nor dissociated from it, because there is no separate thing; an affliction in the stage of sleep is called a latent tendency”; Saṅghabhadra quotes it word for word and adjudges, “examining that gloss closely, it is not so in principle,” pointing straight at where it contradicts itself — since he himself grants that apart from the afflictions a latent tendency is no other thing, and since an undefiled dharma cannot be the entity of an affliction, then “it cannot be said to be the entity of a sleeping latent tendency, and so there is no least thing named a sleeping latent tendency.” ⇒ The account this treatise holds to be the best, Saṅghabhadra denies in its very gloss. One meaning and two schools, recorded together, with no adjudication of which prevails.

The passage on the Sūtra of the Six Sixes is likewise. The author explains the sūtra through with “the sūtra says only that there is one; it does not say that there is a latent tendency at that very moment”; Saṅghabhadra quotes and refutes — “this gloss is unreasonable, since it is unreasonable in principle that at that moment the latent tendency should be spoken of with the word of present existence” — and goes on to press him item by item with the objection that the entity of a latent tendency would then have to be unknowable. Recorded without adjudication.

In one place what Saṅghabhadra appraises is not the doctrine but the composition. On the ground for setting up the name ‘craving for existence,’ the author speaks first of turning by the inner gate and afterwards of savoring attachment toward one’s own entity; Saṅghabhadra appraises, “what difference in sense has this gloss from the former? … examining it, the Sūtra-master’s gloss differs in sense not at all from the former, but only frames floating words to make it seem a little different” — he holds this to be one sense stated twice over. ⇒ This belongs to an appraisal of composition and not an opposition of doctrine, and the reader should keep it apart from the two items before. Recorded without adjudication.

One place of the most exact word-for-word correspondence. The author asks and answers himself, “what is the mark here of being abandoned by seeing suffering? And so on to, what is the mark of being abandoned by cultivation? If what it takes as object is what is abandoned by seeing this, it is named as abandoned by seeing this”; Saṅghabhadra quotes it word for word and refutes, “this is unreasonable … since the all-pervading latent tendencies take the five categories as object, the latent tendencies abandoned by seeing suffering and origination ought also to extend to being abandoned by seeing origination and suffering and the rest,” and sets up in its place his own tenet, establishing the mark of abandonment by receptivity. In this sentence the two treatises correspond character for character, and it is the passage in the debate of this fascicle most convenient for reading side by side. Recorded without adjudication.

One place of ‘the same conclusion by a different reason.’ On why adherence to observances and vows is abandoned by seeing suffering, the author objects that ‘to say it is because one is bewildered about the truth of suffering has the fault of overreach’; Saṅghabhadra says that this objection is itself not established — “since not all afflictions taking what has outflows as object take the suffering that is result for their object. How then is there any fault of overreach?” — and sets up besides the account that ‘holding what is not the path to be the path is of two classes’ so as to establish the same conclusion well. ⇒ The two houses guard the same conclusion, and what is contested is the reason for reaching it. This type occurs once only in this fascicle and is worth remembering.

One counter-question on the view of a real person. Saṅghabhadra asks in return upon the gloss ‘because it perishes it is called sat’ — “if so, what use is there in marking it with the word sat?” — if it is only to block eternity, the single word kāya would suffice; why mark a further word ‘perishing’? This objection appears also in the Treatise Conforming to Correct Principle passage cited by Fabao, from which one sees that the ancient masters knew of this objection long ago and did not take it to settle right and wrong. Recorded without adjudication.

One place where the treatise does not refute and Saṅghabhadra alone does. The treatise itself records the preceptors of former times’ account that ‘an innate view of self is of indeterminate quality’ and passes no verdict; Saṅghabhadra adjudges it head-on — “this is unreasonable, since to say that what cannot discriminate is gathered into view and is abandoned by the path of seeing is unreasonable in principle” — and establishes the innate view of self as abandoned by cultivation. ⇒ This is the one instance in this fascicle of ‘the treatise citing an account from one side only, the refutation appearing in Saṅghabhadra alone’; in the eight places before, the treatise already has some exchange of both parties. Recorded without adjudication.

One divergence of attribution. In the passage on the all-pervading, the sentence ‘on what ground is it a view when the other afflictions take that as object, while these too take it as object and yet are not views?’ Puguang marks as the Vaibhāṣikas’ reproach in return against the objector, while Saṅghabhadra calls it outright the Sūtra-master’s own interrogation. One and the same sentence, and the two houses tie it to different speakers; recorded without adjudication.

3. The Divergences of Paramārtha’s Earlier Translation

Paramārtha’s Expository Treatise on the Treasury forms a system of its own in the passage corresponding to this fascicle. The first thing to know is that the very name of this chapter differs in his version — it reads ‘the chapter analyzing the confusions’ and not ‘the chapter analyzing the latent tendencies,’ one and the same Sanskrit taken in different renderings (one taking ‘confusion,’ the other ‘latent tendency’). To search Paramārtha’s version with Xuanzang’s words is never to find the chapter name; a comparison across versions can only take the structure of the scene. The two versions also differ in the number of fascicles: Xuanzang three and Paramārtha two, and not one of the fascicle boundaries of the two coincides.

The rendering-names form a system of their own. Xuanzang’s ‘greed’ Paramārtha renders ‘desire for desire’; ‘craving for existence’ as ‘desire for existence’; ‘entanglement’ as ‘the upper mind’; ‘without outflows’ as ‘without flow’; ‘independent ignorance’ as ‘solitary ignorance’; ‘the extreme view’ as ‘the view of extremes’; ‘adherence to observances and vows’ as ‘adherence by holding to precepts’; ‘the unwholesome roots’ as ‘the evil roots’; ‘other regions’ as ‘the outside masters’; ‘determination by counter-question’ as ‘determination by asking back’; ‘determination by setting aside’ as ‘setting-aside determination’; ‘pride beyond excessive pride’ as ‘pride beyond excess’; ‘the pride of slight inferiority’ as ‘the lower pride.’ In form, Xuanzang writes ‘the stanza says / the treatise says,’ the whole stanza coming out first; Paramārtha writes ‘the verse says / the explanation says,’ glossing line by line as embedded, so that one stanza is often broken up and cited in several places.

Two instances that best show the parting of angle. The first is the setting up of the name ‘craving for existence’: Xuanzang has “in order to block the notion of liberation,” Paramārtha “in order to remove that notion of nirvāṇa” — one takes liberation and the other nirvāṇa, pointing at one thing and taking a different name for it; and Paramārtha here brings out besides a transcription, ‘samāpatti,’ where Xuanzang writes with a rendering by sense and the transcription does not appear on the face of his text. The second is the reckoning of the four inversions: Xuanzang has “from three views,” Paramārtha ‘the four inversions arise from two and a half views’ — Xuanzang taking the three views as a rough enumeration, Paramārtha reckoning precisely to the half (of the three views, some taken whole and some in part, making exactly two and a half). The doctrinal content the two versions point at is entirely the same, and what differs is only the manner of reckoning. Recorded without adjudication.

Two places where the face of the text is doubtful and the prose commentary resolves it of itself. First, in listing the names of the six latent tendencies Paramārtha writes ‘view, mind, doubt,’ one character ‘mind’ more than Xuanzang’s ‘view and doubt,’ and its exact sense is unclear. Second, the face of Paramārtha’s stanza on the all-pervading reads ‘the confusions of seeing suffering, origination, cessation,’ one character ‘cessation’ more, which read on its face would seem not to agree with Xuanzang’s ‘abandoned by seeing suffering and origination’; yet his prose commentary immediately states that it is still bounded by suffering and origination, and agrees entirely with Xuanzang in doctrine. ⇒ In both places no explanation is forced, and the fact alone is recorded.

One place of showing and hiding a name. For the party that objects to the account that the latent tendencies are associated, Paramārtha’s version in that passage does not bring out the name of the school; yet when it comes to citing the sūtra in evidence for the four determinations, Paramārtha writes out plainly ‘the Venerable Mahāsāṃghika’ (Xuanzang writing ‘the Mahāsāṃghikas,’ the one a rendering by sense, the other a full transcription). ⇒ From this one sees that Paramārtha does not avoid the name of this school throughout the chapter; it is only that in that passage he did not mark it specially. The fact alone is recorded, and no judgment is made on the translator’s intent in his stead.

One place of the most exact correspondence. In the enumeration of the fourteen indeterminate matters, the order and the content of the two versions agree entirely, and it is the place in the closing passage of this fascicle most convenient for word-for-word parallel reading; only that Xuanzang writes ‘after the Tathāgata’s death’ and Paramārtha ‘the Tathāgata as other than death,’ the angle of the gloss differing slightly while what is pointed at is the same. There is besides the passage glossing the ten functions, where the opening and gathering of the two versions are almost the same, Paramārtha for the most part heading each with a doubled word (firmly strengthening · settledly establishing · repeatedly tilling · producing and arising), the order of the ten functions corresponding one to one; and it is the passage in this fascicle where the two translations come closest.

4. Coordinates for Parallel Reading

What this fascicle sets up is a table of ‘how the afflictions are numbered.’ The move it demonstrates again and again — one and the same set of dharmas takes a new number with each new rod — will be used again in every remaining chapter of this treatise; and in this fascicle it is used most densely.

Looking forward, this fascicle has done only half a piece of work. The entity of the confusions being adjudged, how the confusions bind, their increase by attachment, and their sequence, and further the abandonment and extinction of the confusions, all lie in the two fascicles that follow and in the next main division; and the four gates of determination at the end of this fascicle are an appended discussion inserted after what is and is not a root, the author himself saying that the indeterminate of that sūtra “sets up the name indeterminate only with respect to a question that should be set aside,” and is not the same thing as the indeterminate of the present gate.

Looking back, this fascicle everywhere puts to use the results of the chapters before. The establishing of the three realms and the division of the five categories into seeing and cultivation are coordinates laid down all along the way by the chapters on the elements, on the faculties, and on the world; while the sentence that opens the fascicle, “it was said before that the differences of the world all arise from action,” follows word for word upon the end of the chapter on action. ⇒ The chapter on the latent tendencies does not raise a new topic; it supplies the chapter on action with its other half — action is the seed, and the latent tendencies are the moisture by which the seed is enabled to grow.

Two further things this fascicle leaves the reader as loose threads. First, the author expressly appraises and takes the Sautrāntikas in the dispute over the association of the latent tendencies, and it is at this very place that Saṅghabhadra’s counter-refutation is fiercest — the manner of composition recorded in the Life of Vasubandhu, ‘to set up the Sarvāstivāda sense and, wherever there is a bias, to refute it by the Sautrāntika sense,’ can here be seen for what it concretely looks like, and so can the sort of response it drew. Second, in the passage on what abandons adherence to observances and vows the author leaves the one sentence “therefore the sense here held should be further examined” and draws no conclusion; nor did the ancient masters close it on his behalf. Where the treatise itself marks a place as unsettled, the reader need not close the case for it.

Before Reading This Fascicle

This fascicle opens from an extremely short sentence — 「諸有情類於此事中隨眠隨增，名繫此事」 tr. “when for any class of sentient beings a latent tendency increases by attachment upon this thing, that is called binding this thing.” There is no transitional formula, no laying out of what the previous fascicle had done; the author settles a definition in one sentence and immediately asks: among the three times, past, present, and future, which latent tendencies can bind which things? The three ancient masters all bring the knife of the outline down at the same place here: Puguang marks 「此下第六明惑」 tr. “below, the sixth, explaining the afflictions,” Fabao marks 「已下大文第六明世縛事通局」 tr. “from here below, the sixth of the main text, explaining the general and particular extent of the binding of things across the times,” Yuanhui marks 「從此第六，明惑能繫」 tr. “from here the sixth, explaining that the afflictions can bind” — three different wordings for the division-name, and one and the same stretch pointed at; and each of the three divides his own fascicles, whose boundaries again coincide one by one with Xuanzang’s. This cut was not forced upon the text by later hands.

What matters is the manner of the question. Since it asks “what things can a past latent tendency bind,” it has already presupposed that the past has things to be bound. And so one question forces out one dispute: are the dharmas of past and future really existent or not? The treatise here inserts an incidental discussion, taking in the Sarvāstivādins’ two proofs from teaching and principle for their tenet, the four masters’ divergent accounts within the school of how the three times are established, and the author’s whole refutation line by line, the prose running unbroken to more than fifteen hundred characters, and then closing it with the one sentence 「法性甚深非尋思境」 tr. “the nature of dharmas is profound and is not the object-field that pondering can reach.” Having closed it, the author says 「傍論已了，今應思擇」 tr. “the incidental discussion is finished; now it should be pondered,” and turns back to the main theme as if nothing had happened.

So one should come to this fascicle prepared for this: its first third is one of the greatest disputes between the schools in the whole treatise, and its latter two-thirds one of the finest tables of numbers in the whole treatise (sixteen dharmas and sixteen consciousnesses; forty-one items and fifty-two items; twenty-nine, twenty-eight, thirty-six, and fifteen). Dispute and number stand together in one fascicle, and this is no accident of arrangement — what the dispute disputes is “whether there is,” what the numbers set out is “how many there are”; the former is unsettled, and only the latter can be counted.

One more thing must be said first. Within this fascicle the author records the two parties and does not conclude for either. He records the Sarvāstivādins’ establishing of their tenet in full, and he writes the Sautrāntikas’ refutation with a free hand; and at the end he lets the Sarvāstivādins say one sentence, “profound and not the object-field of pondering,” and the party breaks up. This is not vagueness; it is the form: one sense with two schools, recorded together without adjudication. In reading, one need not divide victory from defeat on behalf of the ancients; first see clearly where each side has laid its weight.

I. The General Rule of Binding — First Dividing the Afflictions of Individual and of Common Characteristic

〔1.1〕Why two classes of affliction must be divided the moment the mouth is opened. First ask: what is asked is “which latent tendencies can bind which things”; why does the answer not set out from the three times but first divide the afflictions into two classes? Because the rule of binding across the three times is not the same for the two classes, and without dividing them first there is no way to state it. The treatise establishes at once: 「一者自相，謂貪、瞋、慢」 tr. “the first is of individual characteristic, namely greed, hatred, and pride,” and 「二者共相，謂見、疑、癡」 tr. “the second is of common characteristic, namely view, doubt, and delusion.” The criterion of the division is given immediately after — an affliction of individual characteristic is so named because 「以貪瞋慢是自相惑，非諸有情定遍起故」 tr. “because greed, hatred, and pride are afflictions of individual characteristic, and it is not that all sentient beings must arouse them universally”; an affliction of common characteristic is so named because 「由此三種是共相惑，一切有情俱遍縛故」 tr. “because these three are afflictions of common characteristic, and all sentient beings are alike universally bound by them.”

Clearing a misreading: taking “individual characteristic” to mean the individual characteristic of the object taken and “common characteristic” the common characteristic of the object taken. The criterion lies not in the object but in whether it arises — greed, hatred, and pride depend upon each sentient being’s own particular object for their arising, while view, doubt, and delusion are had by every ordinary being without exception. Yuanhui explains this most cleanly and briefly, his meaning being that what arises taking a particular dharma as object is named of individual characteristic and what arises taking many dharmas as object is named of common characteristic; and reading the two accounts together one sees that this pair of technical terms is passable at both ends: from the side of arousing, it is universal or not universal; from the side of the object taken, it is particular or many.

In practice: every adjudication of increase by attachment below, the chain of the order of the ten latent tendencies, and even the number of items under the outflows, torrents, yokes, and appropriations all hang upon this pair of terms. Let the first section be blurred over, and every square of figures that follows will fall through.

〔1.2〕Why the binding across the three times is divided so finely. First ask: the two classes being divided, how is the rule of binding to be set out? The treatise’s ordering is this: of greed, hatred, and pride in past and present, only those “that have already arisen” are reckoned as binding — because an affliction of individual characteristic does not arise universally, and where it has not arisen there is no binding to speak of. Coming to the future, those three associated with the mental consciousness bind all three times throughout; while the greed and hatred associated with the five consciousnesses in the future divide into two roads — those that may arise bind the one future time alone, and those that will not arise bind on the contrary all three times throughout. As for view, doubt, and ignorance, past or future, so long as they are unabandoned they bind all three times throughout.

Clearing a misreading: on seeing that “those that will not arise bind on the contrary all three times” one feels it upside down — do those that will never arise bind more widely? In fact not: the portion that may arise is going to fall hereafter upon some one determinate moment, and the range of its binding is thereby held within that moment; the portion settled not to arise has no place to fall, and binds universally simply as an unabandoned potency. This is the first appearance of the sense that “a latent tendency is a potency and not an activity.”

In practice: what really has to be counteracted in practice is often not the mind that aroused greed on that one occasion, but that portion which has never once come into activity and yet has never been abandoned. This passage of the treatise writes just that fact into the rule.

II. The Establishing of Real Existence in the Three Times — Two Proofs from Teaching, Two from Principle

〔2.1〕How this dispute was forced out. First ask: the fascicle was speaking peacefully of the latent tendencies; why must it suddenly argue whether the three times exist? The treatise explains itself: 「應辯諸事過去未來，為實有無方可說繫」 tr. “one must make out whether things past and future are really existent or nonexistent, and only then may one speak of binding” — the things bound extending to past and future, one must first ask whether the things of past and future exist or not. And either side is difficult: if they are really existent, then all conditioned dharmas exist at all times, and would they not become permanent? And if they are really nonexistent, then the three matters of what binds, what is bound, and release from bondage have nowhere to come down. The Vaibhāṣika masters here decide for real existence and themselves resolve the difficulty of “becoming permanent”: 「毘婆沙師定立實有，然彼諸行不名為常」 tr. “the Vaibhāṣika masters decidedly establish real existence, yet those conditioned dharmas are not named permanent,” because conditioned dharmas are conjoined with the conditioned marks.

Clearing a misreading: taking this passage for a digression off the theme. It is not a digression; it is the ground floor of the main argument of this fascicle — what the later passage on “abandoned and yet not released from bondage” handles is precisely “the thing is abandoned and the binding still stands,” and that word “stands” cannot come down without the sense of the three times to hold it up. In practice: whenever one meets the two characters “incidental discussion” in this treatise, look at what it is incidental to; an incidental discussion is often precisely the hidden pile on which the main argument is able to stand.

〔2.2〕The four reasons must be seen in two layers. First ask: how many reasons are hidden in the ten characters of the stanza, 「三世有由說，二有境果故」 tr. “the three times exist because it is said so, and by the two of having an object-field and having a result”? Answer: four, in two layers. The first layer is the proof from teaching — first, 「由契經中世尊說故」 tr. “because the World-Honored One said so within the sūtras” (the Buddha teaching men to strive to cultivate disgust with and abandonment of past form and to strive to sever and to seek future form, from which it may be known that past and future form exist); second, 「又具二緣識方生故」 tr. “and again because a consciousness arises only when complete in two conditions” (the sūtra saying that a consciousness arises from two conditions, so that if past and future were not existent the consciousness that takes past and future as object would lack one condition). The second layer is the proof from principle — third, that a consciousness when it arises must have its object-field (without an object-field no consciousness arises; and since the consciousness that remembers the past does arise, its object-field ought to exist); fourth, 「又已謝業有當果故」 tr. “and again because an action already perished has a coming result” (if the entity of a past action were wholly nonexistent, the coming results of wholesome and unwholesome would have nowhere to come from).

Clearing a misreading: taking the four as four items standing abreast. The text of the treatise divides them very plainly: 「已依聖教證去來有，當依正理證有去來」 tr. “having proved by the noble teaching that past and future exist, we shall now prove by correct principle that past and future exist” — the first two by teaching, the latter two by principle, and the latter two are precisely the former two turned into reasoning and stated again. In practice: whenever one reads a passage establishing a tenet, count clearly how many proofs are from teaching and how many from principle, and then in the refutation set them against each other one by one, and there will be no confusion.

〔2.3〕“The school that says all exists” is a school-name and also a criterion. First ask: is the stanza’s closing 「說三世有故，許說一切有」 tr. “because they say the three times exist, they are allowed to be the school that says all exists” a conclusion or a definition? A definition. The text of the treatise gives the criterion sentence by sentence — 「若自謂是說一切有宗，決定應許實有去來世」 tr. “if one calls oneself a follower of the school that says all exists, one must decidedly grant that the times of past and future are really existent”; and conversely, if one grants only the present and the past actions that have not yet given their result, saying that there is no future and no past action that has already given its result, then 「彼可許為分別說部，非此部攝」 tr. “such a one may be allowed to be of the Vibhajyavādins, and is not gathered into this school.”

Clearing a misreading: taking “the school that says all exists” to mean “all dharmas exist.” It does not — what it says is “all three times exist,” the stress falling upon time and not upon the number of dharmas. In practice: when the author closes the passage later he says 「若欲善說一切有者，應如契經所說而說」 tr. “if one would state ‘all exists’ well, one ought to state it as the sūtra states it,” and adduces the sūtra to explain “all exists” by the twelve sense fields or the three times — which is precisely to return this criterion upon its owner. One school’s name, explained by each of the two parties in its own way: the point at issue in this fascicle in fact begins from the school’s very name.

III. The Four Masters Who Establish the Three Times — Class, Mark, Position, and Relation

〔3.1〕What the four masters are disputing. First ask: since all four belong to the Sarvāstivādins, what is left for them to dispute? What they dispute is this: the entity being the same in all three times, upon what are the differences of “time” to be established? Dharmatrāta says upon “class,” Ghoṣa upon “mark,” Vasumitra upon “position,” Buddhadeva upon “relation.” Each of the four has his simile: Dharmatrāta’s is 「如破金器作餘物時，形雖有殊而體無異」 tr. “as when a vessel of gold is broken up and made into some other vessel, the shape being distinct while the entity is not distinct”; Ghoṣa’s is 「如人正染一妻室時，於餘姬媵不名離染」 tr. “as when a man is at the moment enamored of one wife, of his remaining consorts it is not said that he has left off desire”; Vasumitra’s is 「如運一籌，置一名一、置百名百、置千名千」 tr. “as with the moving of a single counting-rod: set at one it is called one, set at a hundred it is called a hundred, set at a thousand it is called a thousand”; Buddhadeva’s is 「如一女人名母名女」 tr. “as one and the same woman is named mother and named daughter.”

Clearing a misreading: reading the four similes as four turns of rhetoric. Each of the four fastens upon one mechanism — the gold vessel fastens upon “the entity unaltered while the class to which it belongs alters”; the wife upon “the marks being all possessed together while the activity is partial”; the counting-rod upon “the thing being one while the position differs and so the name differs”; the mother and daughter upon “the name established by what it is related to.” In practice: hereafter, whenever one meets a set of divergent accounts standing abreast, first find out each master’s mechanism; the mechanisms found, the author’s refutation below has something to come down upon.

〔3.2〕The names of the four masters are themselves a lesson. Puguang glosses each of these four names, and this is the most learned passage in the fascicle. He glosses Dharmatrāta’s name thus: 「能救正法、或正法救彼、或以正法救人，故名法救」 tr. “he can defend the right dharma, or the right dharma defends him, or he defends men with the right dharma, and so he is named Dharma-defended”; he glosses Ghoṣa’s name thus: 「梵云懼沙，舊云瞿沙，訛也」 tr. “in Sanskrit Jusha; the old rendering Qusha is corrupt”; on Vasumitra he carries also the account of how he got his name, saying first 「此即第三師說。世是天名，與天遂友故名世友」 tr. “this is the third master’s account. ‘World’ is the name of a god, and because he came to be friends with the god he is named World-Friend,” and then analyzing his Sanskrit name, 「梵名筏蘇密呾囉，筏蘇名世，密呾囉名友。舊云和須密，訛也」 tr. “the Sanskrit name is Fasumidaluo; fasu means world and midaluo means friend; the old rendering Hexumi is corrupt”; and he glosses Buddhadeva thus: 「能覺悟天，故名覺天。梵云勃陀提婆，勃陀名覺，提婆名天。舊云佛陀提婆，訛也」 tr. “he can awaken the gods, and so is named Awakening-God; in Sanskrit Botuotipo, botuo meaning awakening and tipo meaning god; the old rendering Fotuotipo is corrupt.”

First ask: what use are these glosses? They have two uses. One is that in reading the earlier translation one will not mistake the men — Paramārtha’s earlier translation gives this passage as Damoduoluo, Qusha, Poxumiduoluo, and Fotuotipo, all by transcription; and one who does not know that the renderings of sound and of sense are one and the same man will read four masters as eight. The other is to know that renderings have a history — Puguang marks each of the three old transcriptions “corrupt,” and that is his own glossing verdict. Clearing a misreading: taking Puguang’s word “corrupt” for an adjudication of the treatise. The treatise here only records the four masters’ accounts, and touches not a word upon renderings of names. In practice: whenever one sees an ancient master set down the two words “this is corrupt,” first ask who is speaking; within this fascicle, it is Puguang speaking.

〔3.3〕How the author assesses and refutes the first three and takes the third. First ask: since he takes the third, why refute the first, the second, and the fourth, and not in order? Because what he refutes is three forms of one and the same malady. He refutes the first master thus: 「第一執法有轉變故，應置數論外道朋中」 tr. “the first holds that dharmas undergo transformation, and ought to be placed within the party of the Sāṃkhya non-Buddhists” — the entity unaltered while the class alters is precisely the Sāṃkhya doctrine of “transformation.” He refutes the second thus: 「第二所立世相雜亂，三世皆有三世相故」 tr. “the marks of time established by the second are jumbled, since the dharmas of all three times have the marks of all three times” — the three marks being possessed together, the three times cannot be told apart. He refutes the fourth thus: 「第四所立前後相待，一世法中應有三世」 tr. “the relation of earlier and later established by the fourth would make there be three times within the dharmas of one time” — earlier and later standing in mutual relation within one time, the earlier and later moments of the past become past and future again, and the three times have no fixed standard. He takes the third alone, on the ground that 「彼謂諸法作用未有名為未來，有作用時名為現在，作用已滅名為過去，非體有殊」 tr. “his meaning is that when the function of dharmas is not yet present they are called future, at the time when there is function they are called present, and when the function has ceased they are called past, the entity not being distinct.”

Clearing a misreading: taking “the assessment that takes Vasumitra” for Puguang’s or Yuanhui’s adjudication. It is not. Puguang expressly marks here 「此即評取第三世友」 tr. “this is the assessment that takes the third, Vasumitra,” which is precisely to point out that the assessment comes from the author himself; and Yuanhui, narrating the same passage, likewise opens with “the author assesses, saying.” In practice: the most vital question in reading a commentary is “is this sentence the treatise’s or the commentary’s?” And this passage is just where all three commentaries with one voice tell you: this one is the treatise’s.

IV. Five Preliminary Interrogations — Pressing “Function” into the Corner

〔4.1〕Who is asking whom in these five exchanges. First ask: the third master having been taken as best, why is he at once objected to? Because the author’s own standpoint was never with the Sarvāstivādins. In the five exchanges that follow, the three ancient masters mark them consistently: the first exchange, 「彼應復說，若去來世體亦實有，應名現在，何謂去來」 tr. “he ought further to say: if the entities of the times of past and future are also really existent, they ought to be called present; what is meant by past and future?” is the Sautrāntikas’ objection — Puguang marks it 「經部難。去、來實有，應名現在，何謂去、來」 tr. “the Sautrāntikas object: if past and future are really existent they ought to be called present; what is meant by past and future?”; the second, 「豈不前言約作用立」 tr. “was it not said above that it is established upon function?” is the Sarvāstivādins’ answer — Puguang marks it 「說一切有部答。指同前說」 tr. “the school that says all exists answers, pointing to what was said above”; the third, “the present eye and the other faculties, gathered under the participating, what function have they?” is the Sautrāntikas objecting again; the fourth, 「彼豈不能取果與果」 tr. “can it not take a result and give a result?” is the Sarvāstivādins’ answer; the fifth, 「有半作用，世相應雜」 tr. “there is half a function, and the marks of the times would be jumbled,” is the Sautrāntikas objecting again.

Clearing a misreading: reading this passage as the author’s rhetorical self-questioning and self-answering. There really are two parties, and the three commentaries mark the schools identically, not one round diverging. In practice: hereafter, on meeting a long chain of question and answer, first fix the speakers by the ancient masters’ outlines and then read the content; otherwise one is certain in the end to fit the cap to the wrong head.

〔4.2〕The weight of the objection from “half a function.” First ask: why can the fifth exchange close these five rounds? Because it turns the Sarvāstivādins’ own criterion back upon them. They say that “at the time when there is function it is called present,” and they also say that the participating eye, though it does not see form, can take a result and give a result — then since the past homogeneous cause and the rest can give a result, they too have function, and they too ought to be called present. And thus upon one and the same dharma the two functions of “taking a result” and “giving a result” are the one present and the other absent, which makes 「有半作用，世相應雜」 tr. “half a function, and the marks of the times jumbled.”

Clearing a misreading: taking this for a game with words. It is not — what it points out is that the criterion itself is not self-sufficient: if time is fixed by “function,” then function must first be cut clean, whereas taking and giving a result belong from the outset to two different times. In practice: whenever a criterion is set up, first ask whether it can adjudicate itself cleanly; this is the most serviceable single move the fascicle teaches its reader. The author lays down his pen here himself — “having briefly probed, next comes the refutation at length” — from which it is plain that all the above was still only a warming-up.

V. The Root Stanza’s Refutation at Length Line by Line — What Obstructs the Function, What Time the Function Is, No Distinction, Time Collapses, Unarisen and Ceased

〔5.1〕The four lines of the stanza are a table of contents for the refutation. First ask: how are the twenty characters of the root stanza, 「何礙用云何？無異世便壞，有誰未生滅？此法性甚深」 tr. “what obstructs the function, and what time is the function? If there is no distinction, time collapses; who then is unarisen or ceased? The nature of this dharma is profound,” to be read? Yuanhui here gives the least laborious reading, his general adjudication being 「前三句經部破，第四句有宗答」 tr. “the first three lines are the Sautrāntikas refuting, the fourth line the answer of the school that holds existence.” That is to say: the first three lines are three objections and the fourth is the Sarvāstivādins’ answer; and among the three objections the first line in fact contains two — “what obstructs the function” and “what time is the function.” Puguang’s marking of the schools line by line agrees with this, only that he scatters the general adjudication beneath the prose line by line.

Clearing a misreading: reading the fourth line, “the nature of this dharma is profound,” as the author’s conclusion. Yuanhui expressly marks 「此法性甚深者，薩婆多救也」 tr. “as for ‘the nature of this dharma is profound,’ that is the Sarvāstivādins defending” — it is the Sarvāstivādins’ defense and not the author’s verdict. In practice: that the words of two parties should stand within one stanza is no isolated case in this treatise; in reading a stanza, to ask first “whose is this line?” is often more vital than glossing it word by word.

〔5.2〕The first objection: what is it that blocks the function. First ask: where does the objection cut in? It cuts in at the seam between “existing constantly” and “now arising and now not.” The treatise runs: if the own-entity of a dharma exists constantly, it ought at all times to be able to arouse its function; what force then makes the function aroused by this entity of a dharma be now present and now absent? Thereupon the Sarvāstivādins set up a defense — that the several conditions have not come together; and the Sautrāntikas at once reject it: this defense is unreasonable, since on the Sarvāstivādins’ own account the conditions themselves exist constantly.

Clearing a misreading: taking this for an interrogation from common sense. It is not — what it uses is precisely the Sarvāstivādins’ own “existing constantly.” If they answer “the conditions have not come together,” they must grant that the conditions have a time of coming together and a time of not; and the conditions existing constantly, whence the two times? In practice: the most forceful material for refuting another school is always that school’s own premise.

〔5.3〕The second objection: to which time does the function itself belong. First ask: why does this objection follow immediately upon the last? Because the last forced out “the function is now present and now absent,” and this one presses after that “now.” The treatise runs: how is this function itself to be said to be past, future, or present? Can a further function be established upon a function (which would be without end)? And if the function belongs to none of the three times and is yet said to exist, it becomes an unconditioned dharma and ought to be constant and incapable of being nonexistent — and then it ought no longer to be said that “the function has ceased” or that “the function is not yet present.” Here the Sarvāstivādins defend, saying 「若許作用異法體者，可有此失，然無有異，故不應言有此過失」 tr. “if it were granted that the function is other than the entity of the dharma, there might be this fault; but there is no distinction, and so it ought not to be said that there is this fault” — Puguang marks this sentence as the Sarvāstivādins’ defense, and Fabao likewise marks it a Sarvāstivādin defense and points out that it is precisely the explanation of the two characters “not distinct” in the stanza.

Clearing a misreading: reading “not distinct” as the author’s words. It is the Sarvāstivādins’ defense, and it is also the heading marked upon the first half of the stanza’s second line. In practice: the characters of a stanza are sometimes the words of the refuting party and sometimes of the defending party, and only the commentary’s outline can tell them apart — which is just where the ancient masters’ outlines cannot be dispensed with.

〔5.4〕The third objection: since there is no distinction, the three times collapse. First ask: the Sarvāstivādins say “the function is just the entity”; why is that to their disadvantage? Because the entity existing constantly, the function too ought to exist constantly, and then there is no moment that can be called past or future. The treatise runs, “if that is so, the sense of time you have established must collapse … how could there ever be a time called past or future?” — Puguang marks this as the Sautrāntikas’ refutation and expressly points out that what it explains is the three characters “time collapses” in the stanza. The Sarvāstivādins defend again: how can it not be made out? Not yet arisen is called future, arisen and not yet ceased is called present, already ceased is called past. The Sautrāntikas thereupon answer with the third line of the stanza and press to the bottom by questioning and answering themselves: since the entity truly exists constantly, who is not yet arisen? Who is already ceased? — What was lacking before is “it was not yet”; what was lacking afterward is “it is already gone.” And so they conclude, still in the Sautrāntikas’ own voice, 「故不許法本無今有、有已還無」 tr. “therefore he does not grant that a dharma was not before and is now, and having been returns to nonexistence,” and the sense of the three times cannot be made out on any account.

Clearing a misreading: taking this round for a mere tangle of technical terms. What it asks is in fact something utterly plain — if the two words “arising” and “ceasing” are not allowed to have “was not before” and “is not afterward,” what meaning is left in them? In practice: this question may be carried over and used directly upon oneself — if one says that the nature of mind is originally complete and the entity of a dharma exists constantly, then what is it that the word “cultivation” alters? The treatise does not answer for its reader; it only whets the question to its keenest.

VI. The Quoted Stanza and the Sautrāntikas’ Comprehensive Answer — Having Once Been, Being to Be

〔6.1〕That stanza is not a stanza of this treatise. First ask: why must the four lines 「許法體恒有，而說性非常，性體復無別，此真自在作」 tr. “to grant that the entity of a dharma exists constantly and yet to say that its nature is not permanent, while nature and entity are further not distinct — this is truly the handiwork of Īśvara” be separately marked as not a root stanza? Because there is a warrant in the preceding text — the treatise itself gives 「依如是義故，有頌言」 tr. “following out this sense, there is a stanza that says,” from which it is plain that the author adduces it to help his refutation and that it is not one of the stanzas within this treatise’s own reckoning. Fabao marks it here 「引頌破也」 tr. “a stanza adduced in refutation.” The sense of the stanza is: having granted that the entity exists constantly, and yet saying that its nature is not permanent, while nature and entity are again not distinct — that comes to “let it be permanent when we want it permanent and impermanent when we want it impermanent,” which is no different from the Īśvara the non-Buddhists reckon, who acts at his own will.

Clearing a misreading: counting it into the sixty-nine stanzas of this chapter. In practice: the prose of this treatise often adduces the stanzas of others, and the way to judge them rests not upon doctrine but upon the warrant in the preceding text — whenever one sees such phrases as “as a stanza says,” “hence there is a stanza that says,” “a further stanza is spoken,” one must stop and ask first whose that stanza is.

〔6.2〕The Sautrāntikas take the word “exists” apart into two layers. First ask: the Sautrāntikas having refuted the real existence of past and future, why do they yet say “we too say that the times of past and future exist”? Because they have taken the word “exists” apart: 「謂過去世曾有名有，未來當有，有果因故」 tr. “that is, the past time is named existent as having once existed, and the future as being to exist, there being a cause able to produce the result” — the existence of the past is “it once was,” the existence of the future is “it will be, and there is presently a cause able to produce it”; to say they exist upon these two natures is not to say that past and future are really existent in entity as the present is. Their warrant is 「有聲通顯有無法故」 tr. “because the word ‘exists’ can point to both existent and nonexistent dharmas”: the world says “there is a lamp that was not there before,” and that word “is” does not make the nonexistent a real thing.

Clearing a misreading: taking the Sautrāntikas to be playing with the surface of words. They are not — what they are handling is why the Buddha said that past and future exist. The answer they give is: it was in order to block the view that denies cause and result that he spoke of existence upon the sense of having-once-been and being-to-be. In practice: when in reading the sūtras one meets the word “exists,” ask first whether it is the existence of an entity or the existence of a narration; this fascicle raises that distinction into a general rule for interpreting the sūtras.

〔6.3〕The proofs are each returned in turn. First ask: how do the Sautrāntikas return the four reasons of the establishing? First, the sūtra of the staff-and-topknot non-Buddhist says that “an action, though past and exhausted and perished, is nevertheless existent” — the Sautrāntikas answer: that is said covertly of the “capability of giving a result” within the present continuum drawn forth by the action; unless so explained, a past action becomes really existent in the present, and what past is left? Second, the Sūtra of Ultimate Emptiness says 「以薄伽梵於勝義空契經中說：眼根生位無所從來，眼根滅時無所造集」 tr. “for the Bhagavat said in the Sūtra of Ultimate Emptiness: at the position where the eye faculty arises it comes from nowhere, and when the eye faculty ceases it is made and gathered nowhere,” and says expressly that it was not before and is now, and having been returns to nonexistence — if the eye faculty of past and future were really existent, the sūtra ought not to have spoken so. Third, the proof from a consciousness arising from two conditions: the Sautrāntikas ask back, in “the mental faculty and dharmas as conditions,” does that “dharma” act as a producing condition in the way the mental faculty does, or only as the object-field taken as object? If as a producing condition, how can a dharma a hundred thousand aeons hence, or even the nature of nirvāṇa, produce today’s consciousness? And if only as the object-field taken as object, then past and future as “having once been and being to be” are equally objects taken. Fourth, the proof from an action having a coming result: the Sautrāntikas say expressly 「非經部師作如是說，即過去業能生當果」 tr. “it is not that the Sautrāntika masters say that that very past action can produce the coming result,” their tenet being that the transformation and differentiation of the continuum drawn forth by the action make the coming result arise.

Clearing a misreading: reading the fourth item as the Sautrāntikas denying action and result. Quite the contrary — what the Sautrāntikas deny is that “an action already ceased directly produces the result,” and what they establish is the transformation of the continuum. In practice: among these four returns, the third and the fourth afterward became the old root upon which the whole Yogācāra learning established seeds and the continuum, and it is worth remembering that their prototype lies in this passage.

〔6.4〕The move of the “thirteenth sense field.” First ask: why does the treatise suddenly bring forward a “thirteenth sense field” outside the twelve? This is the Sautrāntikas finding an extreme instance for their own claim that “a consciousness may take the nonexistent as object.” If the other party says that “what is wholly without entity cannot be taken as object,” then when someone plainly penetrates that “there is no thirteenth sense field,” what does this consciousness that takes an object take? And if it be said that what it takes is that name, that is to grant that the name is nonexistent and yet is taken. The treatise concludes therefrom 「故識通緣有非有境」 tr. “therefore consciousness extends to taking as object both an existent and a nonexistent object-field.”

Clearing a misreading: reading it as sophistry. It is in fact pushing the point at issue to the place where no further retreat is possible — whoever is able to think “nonexistent” has thereby demonstrated that a consciousness can take the nonexistent as object. In practice: this move is quoted again and again in later treatises, and once it is well learned one knows where to lay hold when one meets any claim of the type “the nonexistent cannot be taken as object.”

〔6.5〕How the author closes this passage. First ask: after so long a refutation, who is at last adjudged the winner? No one. The treatise first records the Sautrāntikas’ closing — that if one would state “all exists” well one ought to state it as the sūtra does, all that exists being only the twelve sense fields or only the three times; then records the Sarvāstivādins’ response to the difficulty raised at the outset — that though past and future are nonexistent, the producing cause, the latent tendency, is still present, and so one speaks of past and future afflictions that bind and past and future things that are bound, the latent tendency abandoned being named release from bondage; and last records the Vaibhāṣika masters’ closing: past and future are really existent as the present is, and wherever within this something cannot be explained through, one who cherishes himself ought to know that 「法性甚深非尋思境」 tr. “the nature of dharmas is profound and is not the object-field that pondering can reach,” for how can one deny a thing as nonexistent merely because one cannot explain it? And then one sentence, 「傍論已了，今應思擇」 tr. “the incidental discussion is finished; now it should be pondered,” turns back to the main argument.

Clearing a misreading: reading the two characters “profound” as the author’s own closing, or as a concession of his. Yuanhui expressly marks it a Sarvāstivādin defense; and Saṅghabhadra’s Treatise Conforming to Correct Principle, reading this place, reacts with the greatest heat, calling the author’s sentence outright “a slanderous saying craftily set up” — from which it may be seen that within the lineage this sentence is taken as having been fastened upon the Sarvāstivādins. In practice: it is the form of this treatise to record the two parties and leave the last sentence to the other side, and the reader must see for himself to whom that last sentence belongs. To carry on the posture of his closing without carrying his conclusion for him — this is the measure most to be kept in reading this fascicle.

VII. Abandoned and Yet Not Released from Bondage

〔7.1〕The boundary between two words. First ask: are “abandoned” and “released from bondage” not one and the same thing? They are not. The treatise sets up the question: when things are abandoned are they released from bondage, and when things are released have they been abandoned? Answer: what is released has decidedly been abandoned, and yet there is what is abandoned and not released. Puguang here divides the bounds of the two words very clearly, saying that “already abandoned” is stated with reference to having been abandoned and “released from bondage” with reference to being freed from the binding — the one looking at whether the thing itself has been abandoned, the other at whether there is still an affliction binding it.

Clearing a misreading: taking “abandoned” for the terminus. The treatise here tells you: the thing taken as object being abandoned, if the affliction that takes it as object still stands, that thing is bound as before. In practice: this is an extremely serviceable item on the ledger of practice — that the object of some portion of affliction no longer arises does not mean that this portion is liberated; one must still see whether some other affliction still takes it as object.

〔7.2〕Two instances. First ask: which cases are “abandoned and yet not released from bondage”? The treatise gives two positions. The first is on the path of seeing: at the time when the knowledge of suffering has arisen and the knowledge of origin has not, the things abandoned through the seeing of suffering have been abandoned, and yet if the all-pervading latent tendencies abandoned through the seeing of origin are not yet forever abandoned, those that can take them as object bind them as before. The second is on the path of cultivation: among the nine grades the earlier grades have been abandoned, and the latent tendencies of the unabandoned grades, if they can take them as object, bind them as before. The treatise closes with 「斷非離繫，如是應知」 tr. “abandoned and yet not released from bondage, it ought to be understood thus.”

Clearing a misreading: taking the all-pervading latent tendency for some further special affliction. It is just that family of the afflictions of common characteristic spoken of above — because they can take all five categories as object, the things of this category being abandoned, the afflictions of that category can still come and bind them. In practice: this section puts to use the division of the afflictions of individual and of common characteristic made in the first section; the reason the afflictions of common characteristic had to be singled out is precisely the ambush laid for this place.

VIII. The Sixteen Dharmas as the Object-Fields of the Sixteen Consciousnesses

〔8.1〕Why the author first makes a table. First ask: what was asked was “how many latent tendencies increase by attachment upon a given thing”; why speak first of consciousnesses and their objects? The treatise explains itself: to answer thing by thing separately would cost a great many words, and so 「是故應造略毘婆沙」 tr. “therefore one ought to compose a brief Vibhāṣā,” 「而能越渡大大問流」 tr. “and one is able to cross the very great flood of questions.” The device is to reduce the dharmas in brief to sixteen kinds — 「略為十六種，即三界、五部及無漏法」 tr. “in brief they are sixteen kinds, namely the three realms, the five categories, and the dharmas without outflows” — the consciousnesses that can take them as object being likewise sixteen in number; and once one knows which dharma is the object-field of which consciousness, the question of increase by attachment becomes easy.

Clearing a misreading: taking the “brief Vibhāṣā” for laziness. Quite the contrary — it turns an endless run of particular questions into one table that can be consulted. In practice: on meeting a great table in Abhidharma, first find where its “brevity” lies; the brevity of this fascicle lies in the number sixteen.

〔8.2〕How the three numbers five, eight, and ten arise. First ask: five in the desire realm, eight in the form realm, ten in the formless — what is the rule of this staircase? The rule is one only: what can take them as object must be of the same realm or of a higher one, plus one consciousness without outflows. The dharmas of the three categories abandoned through the seeing of suffering, through the seeing of origin, and through cultivation, if affiliated to the desire realm, are each the object of five consciousnesses — the three categories of their own realm, plus one category of the form realm abandoned through cultivation, with the consciousness without outflows as the fifth. Those of the form realm are each the object of eight — the three categories of their own and of the lower realm together, plus one category of the formless realm abandoned through cultivation, with the consciousness without outflows as the eighth. Those of the formless realm are each the object of ten — the three categories of each of the three realms, nine in all, with the consciousness without outflows as the tenth. As for the dharmas abandoned through the seeing of cessation and of the path, to each is added one consciousness that takes its own category as object, and so the desire realm becomes six, the form realm nine, and the formless eleven. The dharmas without outflows are the object-fields of ten consciousnesses — the last three categories of each of the three realms, nine in all, with the consciousness without outflows as the tenth.

Clearing a misreading: reading “three each” as three realms each. The desire realm’s “three of its own realm” means the three categories of its own realm abandoned through the seeing of suffering, through the seeing of origin, and through cultivation, and not the three realms. In practice: the way to check the reckoning lies in the text of the treatise itself — the three categories of each realm take the three categories of its own and lower realms as their floor, one category being added realm by realm as one goes up, and then the one without outflows; multiply that rule through once again and five, eight, and ten fasten of themselves. Whenever one reads the numbers of this fascicle, ask first “is this a category of its own realm or of another realm,” and the numbers will not fall into confusion.

IX. The Summary Stanza and the Instance of the Faculty of Pleasure

〔9.1〕Why this stanza must be specially marked. First ask: are the two lines and forty characters beginning 「見苦集脩斷，欲色無色繫，應知如次第，五八十識緣」 tr. “what is abandoned through the seeing of suffering, of origin, and through cultivation, affiliated to the desire, the form, and the formless realms, should be known to be the object of five, of eight, and of ten consciousnesses respectively” a stanza of this treatise? No. Its warrant in the preceding text is 「為攝前義，復說頌言」 tr. “in order to gather up the preceding sense, a further stanza is spoken” — it gathers the prose just finished back into a stanza, for ease of memorizing and reciting. Puguang marks it 「說頌重攝。總結」 tr. “a stanza is spoken to gather it again; a general conclusion,” and Fabao marks it 「此兩行頌重述前義」 tr. “these two lines of stanza restate the preceding sense.”

Clearing a misreading: taking it for another’s stanza adduced by the author, or counting it into the chapter’s stanza-count. It is neither adduced from elsewhere nor does it enter the count of the root stanzas — it is the summary of this passage’s own prose. In practice: within this treatise, wherever “a further stanza is spoken” and its content answers line by line to what precedes, it is mostly of this kind; the judgment still rests upon the warrant in the preceding text and not upon guesses from doctrine. This stanza is the first instance of such a mark seen in this collection, and its pattern is worth remembering.

〔9.2〕How the single instance of the faculty of pleasure is held up. First ask: the sixteen dharmas expounded, why does the treatise close by holding up the faculty of pleasure alone? The treatise says so itself: 「故我於此略示方隅」 tr. “and so I here indicate only one corner” — holding up one corner so that the reader may turn back the other three of himself. The instance held up advances in three layers. The first layer asks how many latent tendencies increase by attachment upon the faculty of pleasure: 「應觀樂根總有七種」 tr. “it ought to be observed that the faculty of pleasure has seven kinds in all,” namely one of the desire realm abandoned through cultivation, five categories of the form realm, and the one without outflows as the seventh; and since 「一切無漏非諸隨眠之所隨增」 tr. “all that is without outflows is not such as the latent tendencies increase by attachment upon,” there is increase by attachment only upon the first six. The second layer asks how many latent tendencies increase by attachment upon the consciousness that takes the faculty of pleasure as object: this consciousness has twelve in all, namely four categories of the desire realm (excepting that abandoned through the seeing of cessation), five of the form realm, two of the formless, and the one without outflows as the twelfth. The third layer asks about what takes “the consciousness that takes the faculty of pleasure as object” as its object: fourteen in all, the two categories of the formless realm abandoned through the seeing of suffering and of origin being added to the preceding twelve.

Clearing a misreading: reading the three layers as one question asked three times over. What the three layers ask about moves outward step by step — first the faculty, next the consciousness that takes the faculty as object, next the consciousness that takes that consciousness as object; and with each step outward the things that can take them as object become more numerous, because the all-pervading latent tendencies can climb up layer by layer. In practice: these three layers are the only reckoning taught hand to hand in this fascicle; once it is worked through, the author’s “reasoning from this one corner, the rest ought to be pondered out” is no empty phrase.

X. The Two Kinds of Mind Possessed of Latent Tendencies

〔10.1〕“Possessed of latent tendencies” and “increase by attachment” are two different matters. First ask: when a mind is called “a mind possessed of latent tendencies,” does the latent tendency necessarily increase by attachment upon it? The treatise says expressly that this is not settled. A mind possessed of latent tendencies divides in all into two — 「有隨眠心總有二種，有染、無染心差別故」 tr. “a mind possessed of latent tendencies has two kinds in all, because there is a difference between the defiled and the undefiled mind.” The defiled mind extends to both senses: the latent tendencies associated with it and those that take it as object, if unabandoned, increase by attachment; and if that associated portion has been abandoned they do not increase by attachment, and yet it is still said to be possessed of latent tendencies, because they are constantly associated with it. The undefiled mind, on the other hand, 「若無染者，唯局隨增」 tr. “if it is undefiled, it is confined to increase by attachment alone” — because the latent tendencies that take it as object are decidedly unabandoned, this name being established upon increase by attachment alone.

Clearing a misreading: confounding “possessed of latent tendencies” with “increase by attachment.” Puguang under this division separates the two very finely, pointing out that “increase by attachment” has of itself the two senses of increase through association and increase through the object taken, while what “possessed of latent tendencies” takes is the sense of accompaniment — accompanying does not necessarily mean increasing. In practice: this item is the place in the whole fascicle where the technical terms are most apt to slip out of the hand, and it is also the premise of every later question of “how many latent tendencies increase by attachment”; failing to tell these two senses apart, the table of sixteen dharmas above has been reckoned for nothing.

XI. The Ten Latent Tendencies Arise in Due Order

〔11.1〕The starting point of this chain. First ask: do the ten latent tendencies arise in an order? The treatise says they do, and gives the reason link by link: first, through ignorance one fails to understand the truths and is unwilling to observe suffering and so on to the path; failing to understand, there is next drawn forth doubt, one having heard the accounts of two roads and harboring hesitation; from hesitation is drawn forth wrong view, wrong hearing and wrong reflection producing a wrong determination that denies the truth of suffering and so on to the truth of the path; from denying the principle of the truths is drawn forth the view of self, one denying the principle of suffering within the aggregates of appropriation and determinedly grasping “this is I”; from the view of self is drawn forth the extreme view, one falling by the grasping of self into the two extremes of annihilation and permanence; from the extreme view is drawn forth the grasping of rules and observances, one grasping at one extreme and reckoning this grasping to be able to purify; from the grasping of rules and observances is drawn forth the grasping of views, one having reckoned it able to purify and so grasping it as most excellent; from the grasping of views is drawn forth greed, one being deeply attached in feeling to one’s own view; from greed is drawn forth pride, one having been deeply attached and so relying upon it, arousing self-exaltation and disdaining others; from pride is drawn forth hatred, one being unable to bear in feeling what others arouse against one’s own view and hating and detesting it.

Clearing a misreading: reading this chain as a psychological account of cause and effect. It is in fact a chain fastened upon “view” — the treatise marks it at the end: when greed and the rest abandoned through the seeing of the truths arise, they take the “view” within one’s own continuum as their object-field. That is to say, these ten links from beginning to end are telling how a man comes to fall in love with his own opinion, and then to exalt himself for it and to hate men for it.

〔11.2〕Which kind of ignorance stands at the start. First ask: ignorance is of several kinds; which is this? The text of the treatise gives only 「先由無明於諦不了」 tr. “first, through ignorance one fails to understand the truths,” without qualification; Puguang here adds two characters, giving “first, through independent ignorance, one fails to understand the truths” — limiting it to independent ignorance and not to the general sense of defiled ignorance associated with other afflictions.

Clearing a misreading: reading the two characters Puguang adds as belonging to the treatise’s own text. The treatise here truly does not have those two characters. In practice: this is just the sort of thing most to be watched for in reading a commentary — a limitation the commentator adds in order to make the sense go through is often full of insight, and yet may not be taken for a direct statement of the treatise. Both layers are recorded together here: the treatise gives “ignorance,” Puguang gives “independent ignorance.”

〔11.3〕Another account, and arising out of order. First ask: is this chain the only road? The treatise leaves itself two openings. First, at the link of hatred it records another account alongside — another master says that hatred and detestation arise at the very juncture of taking up or setting aside one’s own view, without waiting until after pride. Second, at the end the treatise says expressly: the above is stated with reference to arising in due order, while for those that arise out of order there is nothing fixed about before and after. In practice: the treatise gives an order, not a law; wherever one sees the words “with reference to,” the author is himself marking the range within which his account applies.

XII. The Three Causes and Conditions of the Arising of an Affliction

〔12.1〕Three conditions of which not one may be lacking. First ask: what does an affliction need in order to arise? The treatise takes the entanglement of desire-greed for its instance and lists three matters: first, that the latent tendency of desire-greed is unabandoned and not yet fully understood; second, that the object-field conformable to desire-greed is present before one; third, that the improper attention taking it as object arises; 「由此力故便起欲貪」 tr. “by these forces desire-greed thereupon arises.” The three in order are just 「即因境界加行三力」 tr. “the three forces of cause, of the object-field, and of preparatory effort” — the latent tendency being the force of cause, the object-field’s being present the force of the object-field, and improper attention the force of preparatory effort.

Clearing a misreading: taking the three for three successive steps. They are three conditions complete at one and the same time, and not three stages. In practice: this is the passage in the whole fascicle most directly usable. Among the three forces, the force of cause (the latent tendency) is not to be abandoned in a day, and the force of the object-field is often not in one’s own hands; only the force of preparatory effort — that one thought of improper attention — can be turned at the present moment. The place where practice takes hold is just this link.

〔12.2〕The exception and the general sense. First ask: must every arising of an affliction be complete in the three conditions? The treatise adds a sentence of its own: this is stated with reference to being complete in causes and conditions, and there are also those that arise by the force of the object-field alone, as for instance an arhat of the lineage liable to falling back. Puguang here hangs the three causes and conditions upon the general Abhidharma frame, explaining 「因謂六因、緣謂四緣」 tr. “cause means the six causes, condition means the four conditions” — joining the three forces of this passage back to the great frame of the six causes and the four conditions.

Clearing a misreading: reading the six causes and four conditions as established by the treatise here. The text here speaks only of three forces; the six causes and four conditions are a frame Puguang adduces in order to make the whole treatise hang together. In practice: both layers are recorded together — the treatise says three forces, Puguang carries it through by the six causes and four conditions; the reader gains both the practical use of the passage and a sense of its position within the whole treatise.

XIII. Outflows, Torrents, Yokes, and Appropriations — the Four Gates and Their Number of Items

〔13.1〕First look at the names of the four gates, then at the numbers. First ask: are the four gates of outflow and the rest four different kinds of affliction? They are not; they are four accounts of one and the same set of afflictions. The sūtras call the latent tendencies together with their entanglements outflows, torrents, yokes, and appropriations: the outflows are three (of desire, of existence, of ignorance), the torrents four, the yokes four, and the appropriations four (of desire, of views, of rules and observances, of assertions of self). Their entity runs through them all — 「應知四軛與瀑流同，四取應知體同四軛」 tr. “it should be known that the four yokes are the same as the torrents, and the four appropriations should be known to be the same in entity as the four yokes.”

Clearing a misreading: memorizing the four gates as four batches of things. They are one heap of afflictions regrouped under four senses of the term; and the reason the groupings differ lies wholly in the opening and gathering of the two items “view” and “ignorance.” In practice: the trick of holding this passage is not to memorize numbers but to hold two rules of opening and gathering — view is gathered in among the outflows and opened out among the torrents and yokes; ignorance is gathered in among the yokes, while among the appropriations view is opened and ignorance gathered in.

〔13.2〕The number of items one by one. For the outflows, 「四十一物總名欲漏」 tr. “forty-one items are named in sum the outflow of desire,” namely the thirty-one root afflictions of the desire realm with the ten entanglements; 「五十二物總名有漏」 tr. “fifty-two items are named in sum the outflow of existence,” namely twenty-six root afflictions in each of the two upper realms; and the fifteen ignorances of the three realms separately as the outflow of ignorance. For the torrents: 「如是已顯二十九物名欲瀑流」 tr. “thus it is displayed that twenty-nine items are named the torrent of desire” (greed, hatred, and pride five each, doubt four, entanglements ten), 「二十八物名有瀑流」 tr. “twenty-eight items are named the torrent of existence” (greed and pride ten each, doubt eight), 「三十六物名見瀑流」 tr. “thirty-six items are named the torrent of views” (twelve views in each of the three realms), 「十五物名無明瀑流」 tr. “fifteen items are named the torrent of ignorance.” For the appropriations: 「三十四物總名欲取」 tr. “thirty-four items are named in sum the appropriation of desire” (the yoke of desire with the ignorance of the desire realm), 「三十八物總名我語取」 tr. “thirty-eight items are named in sum the appropriation of assertions of self” (the yoke of existence with the ignorance of the two upper realms), 「餘三十物總名見取，所除六物名戒禁取」 tr. “the remaining thirty items are named in sum the appropriation of views, and the six items excepted are named the appropriation of rules and observances.”

Clearing a misreading: taking the forty-one of the outflow of desire and the twenty-nine of the torrent of desire for a contradiction. They do not contradict — the outflow of desire contains the afflictions other than delusion together with the entanglements and does not open out view, while the torrent of desire draws view out, and so is short by twelve views and comes to exactly twenty-nine. In practice: every group of figures can be checked by the rules of opening and gathering; wherever the reckoning will not come out, it is most often because one has overlooked the phrase “delusion excepted” or “view drawn out.”

〔13.3〕What the several occurrences of “on what ground” ask about is the reason for the grouping. First ask: why does the treatise set down several questions of “on what ground” in a row? Because the number of items is easy to list while the reason for the grouping is hard to make plain. On what ground are the latent tendencies of the two upper realms joined into one outflow of existence — 「同無記性於內門轉，依定地生，由三義同故合為一」 tr. “alike indeterminate in nature, turning by the inner gate, and arising in dependence upon the levels of absorption: by these three senses being the same, they are joined into one.” On what ground is ignorance separately established as one outflow — 「無明能為諸有本故」 tr. “because ignorance is able to be the root of all existences.” On what ground is view not made a name of its own among the outflows — because an outflow takes the sense of “causing to abide,” while the nature of view is fierce and sharp and does not conform to abiding. On what ground must the grasping of rules and observances be established separately — 「由此獨為聖道怨故，雙誑在家出家眾故」 tr. “because it alone is the enemy of the noble path, and because it deludes both assemblies, the householders and those gone forth”: the householders are deluded by it into reckoning self-starvation and the rest to be the road to heaven, and those gone forth into reckoning the abandoning of agreeable object-fields to be the road to purity. On what ground is ignorance not established as an appropriation — 「謂不了相說名無明，彼非能取，非猛利故」 tr. “that is, not understanding the marks is what is called ignorance; it is not that which appropriates, its nature not being fierce and sharp.”

Clearing a misreading: taking these questions for supplementary explanation. They are the very bone of this passage — the numbers are the result and the reasons for the grouping are the cause. In practice: the item on the grasping of rules and observances is especially to be remembered. The two delusions it reproaches — one in austerity, one in aversion to objects — are still, for the most part, the two squares within which practitioners today go wrong.

XIV. The Glosses on the Names “Latent Tendency” and the Rest

〔14.1〕The four senses of “latent tendency.” First ask: the entity and the numbers having been fully expounded, why gloss the names as well? Because within the name lies the manner of its working. The treatise gives four senses: first, subtle — at the time when a root affliction is present, 「行相難知故名微細」 tr. “because its mode of apprehension is hard to be aware of it is called subtle”; second, twofold increase by attachment — 「能於所緣及所相應增惛滯故」 tr. “because it is able to increase torpor and clogging upon what is taken as object and upon what it is associated with,” that is, increasing torpor at the two places of the object-field taken and the mind it is associated with; third, following after — it can draw forth “acquisition” and constantly follow after a sentient being, and can be made to arise even without any doing; fourth, binding along — even if one spends strength to block it, it still appears again and again.

Clearing a misreading: reading “subtle” as small in quantity. What it says is that its mode of apprehension is hard to be aware of — not weak in force, but hard to detect. In practice: these four senses taken together are just that portion in practice which is hardest to be rid of: undetectable, unceasing in its increase, not to be shaken off, not to be pressed down. Recognizing these four words, one knows why the latent tendencies must be abandoned and not merely suppressed.

〔14.2〕The two layers of gloss on the names of outflow and the rest. First ask: how many layers of gloss does the treatise give for the four names outflow, torrent, yoke, and appropriation? Two. The first glosses them by their function — it can 「稽留有情久住生死」 tr. “detain sentient beings so that they long abide in birth and death,” or make them turn round within birth and death from the heaven at the summit of existence down to the Unremitting Hell; because within the continuum it leaks out harm without end from the six sores it is named outflow; 「極漂善品故名瀑流」 tr. “because it utterly sweeps away the wholesome, it is called torrent”; 「和合有情故名為軛」 tr. “because it joins sentient beings together, it is called yoke”; 「能為依執故名為取」 tr. “because it can act as support and appropriation, it is called appropriation.” The second is the better explanation the author gives of himself — 「諸境界中流注相續泄過不絕故名為漏」 tr. “because within the several object-fields it flows on in a continuous stream and leaks out harm without ceasing, it is called outflow” — and he adduces the sūtra’s simile of drawing a boat against the current in proof: to draw a boat upstream is hard even with great strength expended, while to let it go downstream is not hard even with strength given up; and the arising of a wholesome mind and of a defiled mind is likewise. Following the sense of this sūtra, the afflictions’ not ceasing within the object-fields is named outflow; 「若勢增上說名瀑流」 tr. “if its potency preponderates it is called torrent,” surging and flooding and sweeping and dashing and hard to withstand; what at the time of activity is not utterly preponderant is named yoke, merely joining sentient beings together with sufferings of every kind, or named yoke because it is active again and again; and 「執欲等故說名為取」 tr. “because it appropriates desire and the rest, it is called appropriation.”

Clearing a misreading: reading the second layer as a denial of the first. The treatise itself marks it “one who would explain it well” — it is a closer explanation and not a refutation of what precedes. What the two explanations differ in is where they look: the former looks to the result (long abiding in birth and death), the latter to the mark (the difference between not ceasing and preponderating). In practice: that simile of drawing a boat against the current is a solid object handed to the reader at the fascicle’s end — from disputing whether the three times exist to opening and gathering the number of items, the fascicle comes down at last upon one sentence that anyone may verify: going with the current costs no strength, going against it costs strength, and practice is the going against.

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 take in four matters: the commentaries’ adjudications and their divergences from one another, 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 another way of reading one and the same text. Recorded together without adjudication, each returned to its own place.

1. Where the Three Commentaries Diverge

One commentary wholly absent. The commentaries on this fascicle stand in a special case: Shentai’s commentary is a fragmentary text, of which the surviving seven fascicles belong to the Chapter on the Elements, the Chapter on the Faculties, and the Chapter on the Worthy and the Noble, while the Chapter on the Latent Tendencies is wholly absent; and so this fascicle has no line-by-line explanation of Shentai’s to hang upon. Although his Fascicle One’s general conspectus carries a remark on the number of lines of this chapter, that is a conspectus and not a commentary, and cannot be made to serve, nor may the words of other fascicles be substituted. It is marked absent as it stands.

One and the same stretch, and two patterns of division. All three adjudge Unit 2 to belong wholly to the sixth main division of this chapter, and in this they agree completely; the number of divisions, however, runs in two patterns. Puguang divides beneath the main division into two, “first, explaining binding with reference to the times; second, explaining binding with reference to abandonment”; Yuanhui runs with him and adds in each the one character “can,” giving 「一約世明能繫、二約斷明能繫」 tr. “first, explaining the power to bind with reference to the times; second, explaining the power to bind with reference to abandonment”; while Fabao alone opens four divisions — 「一世縛通局、二證三世有、三建立三世、四破三世有」 tr. “first, the general and particular extent of binding across the times; second, proving that the three times exist; third, establishing the three times; fourth, refuting the existence of the three times” — his latter three answering exactly to the three passages of the establishing of the tenet, the four masters, and the refutation at length. ⇒ The names agree while the number of divisions differs, Fabao alone being fine. The sentence-forms of the three commentaries’ division-names likewise each form a family: Puguang is accustomed to “below, the Nth,” Fabao to “the Nth of the main text,” Yuanhui to “from here the Nth.” Recorded together without adjudication.

On the four masters, Puguang is full and Fabao is neat. Puguang sets out for each of the four masters the gloss of his name and the account of how he got it, most fully for Vasumitra, carrying also the tale that his parents named him “friend of the god” in order to keep spirits from harming him — a piece of lore that among the three commentaries only Puguang has. Fabao uses for all four the uniform sentence-form 「此述第三計也」 tr. “this sets forth the third reckoning,” his doctrinal attributions wholly the same as Puguang’s while his wording forms a manner of its own. Yuanhui again strikes out a line of his own: at his general outline sentence he first sets the four terms side by side, 「一類異、二相異、三位異、四待異」 tr. “first, difference of class; second, difference of mark; third, difference of position; fourth, difference of relation,” a summary peculiar among the three to Yuanhui — Puguang and Fabao both explain the masters one by one as the text runs, and do not list the four terms at the general outline sentence. The differences of fullness and of sentence-form among the three let one see the three characters proper to the commentary as a form. Recorded together without adjudication.

That one assessment of Yuanhui’s occurs twice in his whole book. In this fascicle Yuanhui narrates the author’s assessment and refutation of the four masters as 「應置數論外道朋中，以數論執法有轉變故也」 tr. “it ought to be placed within the party of the Sāṃkhya non-Buddhists, since the Sāṃkhyas hold that dharmas undergo transformation,” continuing below with Vasumitra’s establishing the times best; and one and the same passage occurs exactly twice in the whole of Yuanhui’s book, the two being nearly word for word the same and differing only in the presence or absence of one character “school.” This is a repetition within Yuanhui’s own writing, recorded as a fact, and it does not affect the force of the passage here.

One difference in the density of attribution. Within the passage of refutation at length, Puguang and Fabao handle the sentence “if it be said that the several conditions have not come together, this defense is unreasonable, since they are granted to exist constantly” differently: Puguang folds it into the continuation of the Sautrāntikas’ objection and sets up no separate label for the answering party, while Fabao expressly marks 「牒救非也」 tr. “citing the defense and rejecting it” — citing the Sarvāstivādins’ defense first and then having the Sautrāntikas reject it, dividing one layer more than Puguang. This is the one difference between the two commentaries in the density of attribution at one and the same sentence, recorded as it stands, with no verdict as to which is right.

And at the places that matter most, the three commentaries agree completely. Through the thirteen rounds of question and answer in the passage of refutation at length, all three mark the refuting party “the Sautrāntikas” (Puguang giving the Sautrāntikas object, the Sautrāntikas object again, the Sautrāntikas refute; Fabao the Sautrāntikas object; Yuanhui the Sautrāntikas refute), and the defending party uniformly “the school that says all exists,” “the Sarvāstivādins,” or “the Sarvāstivādins” by another name — different names for the same thing, and compared round by round not one round diverges. Yuanhui moreover passes at the head of the stanza a general adjudication, 「前三句經部破，第四句有宗答」 tr. “the first three lines are the Sautrāntikas refuting, the fourth line the answer of the school that holds existence,” and marks the last line separately, 「此法性甚深者，薩婆多救也」 tr. “as for ‘the nature of this dharma is profound,’ that is the Sarvāstivādins defending.” Such convergence is not common among the long passages of debate in this treatise, and is worth recording.

Puguang marks the boundary of his own division. For the passage of glosses Puguang marks 「此下第二釋名」 tr. “below, the second, explaining the names,” and states of himself what it does not include — 「纏後別明，故此不問」 tr. “the entanglements are explained separately later, and so are not asked after here,” that is, the gloss on “entanglement” falls not in this fascicle but after it. This remark, the boundary of this treatise’s fascicle, and the boundary of Paramārtha’s earlier translation corroborate one another on three sides and fall exactly at one and the same place. That the boundary of an ancient master’s outline should coincide with the boundary of a translation is the hardest set of corroborations for the carving of this fascicle.

2. Saṅghabhadra’s Refutations and Supplements

Saṅghabhadra’s two treatises name this chapter not “Analysis of the Latent Tendencies” but “Adjudication of the Latent Tendencies”; and the stretch answering to this fascicle begins in the Treatise Conforming to Correct Principle at the sentence 「如是略述三世有無理趣正邪有差別已」 tr. “having thus briefly set forth that in the reasoning about the existence or nonexistence of the three times there is a difference of right and wrong.” Saṅghabhadra calls the author “the Sūtra-master,” and within the range of this fascicle that designation occurs twenty-two times, every one of them citing the author’s own words word for word and then refuting them.

The fierceness of the wording is rare in the whole treatise. Upon the Sautrāntikas’ comprehensive answer, “we too say that the times of past and future exist,” Saṅghabhadra cites it and says 「經主於中欲以強力逼令非了」 tr. “the Sūtra-master would herein by main force press it into unintelligibility”; upon the objection from a consciousness arising from two conditions, he rejects it with 「經主此言乖於論道」 tr. “these words of the Sūtra-master are at variance with the way of treatise-reasoning”; upon the sentence “if one would state ‘all exists’ well,” he marks outright 「經主此中作如是謗」 tr. “here the Sūtra-master utters such a slander”; and upon the Vaibhāṣika masters’ closing “the nature of dharmas is profound” he further adjudges 「經主於此詭設謗言」 tr. “here the Sūtra-master craftily sets up a slanderous saying.” At the end of the whole passage Saṅghabhadra concludes for himself, 「正對經主決擇過未，盡彼所能」 tr. “meeting the Sūtra-master directly and adjudicating past and future, to the utmost of his power.”

Here there is one thing worth stopping over. When Saṅghabhadra says that the author “craftily sets up a slanderous saying,” his meaning is that the sentence “the nature of dharmas is profound” is not a closing phrase the Vaibhāṣika masters could have uttered of themselves, but one the author fastened upon them. One and the same sentence: within this treatise it is the Sarvāstivādins’ defense, in Yuanhui too it is marked a Sarvāstivādin defense, while in Saṅghabhadra’s eyes it is the author’s own contrivance. The three sides see it differently, and they are recorded together without adjudication — but the reader may know from this that the attribution of that last sentence of the fascicle is itself a point at issue with a history behind it.

Saṅghabhadra himself also points out the character of that root stanza. Before the passage of refutation at length Saṅghabhadra gives 「然彼經主於此義中，迷執情深復廣興難」 tr. “yet that Sūtra-master, in this matter, being deep in confused attachment, further raises objections at length,” and below cites directly the whole stanza “what obstructs the function” as the text objected with. ⇒ The defining verdict of the debate within the lineage and the three commentaries’ attribution “the Sautrāntikas refute” stand as obverse and reverse: the one side says this is the Sautrāntikas’ refutation, the other that this is the Sūtra-master’s raising of objections at length, and what is pointed at is one and the same stretch.

The passage of glosses was not let pass either. The better explanation at the end of this treatise Saṅghabhadra likewise cites, giving 「經主此中復作是說：若善釋者應作是言」 tr. “here the Sūtra-master says again: one who would explain it well ought to say thus” — from which it is plain that even upon the gloss of the word “outflow” the lineage has something to say.

The two treatises divide the labor differently. Within the passage answering to this fascicle, the Treatise Revealing the Doctrine reproduces the four masters’ divergent accounts almost word for word, as 「尊者法救作如是說：由類不同，三世有異」 tr. “the Venerable Dharmatrāta says thus: it is by difference of class that the three times differ” and 「尊者妙音作如是說：由相有別，三世有異」 tr. “the Venerable Ghoṣa says thus: it is by there being a difference of mark that the three times differ,” nearly identical with this treatise’s own sentences (Ghoṣa’s “there being a difference of mark” differing by one character from this treatise’s “difference of mark”); and then it closes directly with 「是故三世實有義成，諸有智人應隨信學」 tr. “therefore the sense that the three times are really existent is made out, and all who have wisdom ought to follow it in faith and study it,” without one character in between touching the author’s long interrogation from “what obstructs the function” onward, and without one citation of “the Sūtra-master.” ⇒ The whole passage of refutation is absent from the Treatise Revealing the Doctrine. This is a fact of form and not an omission — the Treatise Conforming to Correct Principle is charged with debating line by line, the Treatise Revealing the Doctrine only with restating positively the definitions of its own school. One and the same author, and the two treatises divide the labor differently; marked absent as it stands, recorded without adjudication.

3. The Divergences of Paramārtha’s Earlier Translation

The first thing to know is that Paramārtha’s earlier translation gives this chapter not as “latent tendency” but as “affliction” — the greatest single divergence of rendering in the whole text. To search Paramārtha with Xuanzang’s two characters for “latent tendency” is to search forever in vain; a comparison across the two texts can only be taken by the structure of the scene. The opening sentence of this fascicle stands in Paramārtha as 「若人隨眠惑於境界中生起」 tr. “if a man’s latent affliction arises within an object-field,” agreeing wholly in scene with the question with which this treatise opens, except that this treatise takes “thing” for what is bound and Paramārtha “object-field.”

The names of the four masters are all by transcription. Paramārtha gives 「大德達磨多羅，分別有異故安立三世」 tr. “the Bhadanta Dharmatrāta: because the analysis differs, the three times are established,” 「大德瞿沙說：由相別異故立三世」 tr. “the Bhadanta Qusha says: because the marks are separately distinct, the three times are established,” 「大德婆須蜜多羅說：由位別異故立三世」 tr. “the Bhadanta Poxumiduoluo says: because the positions are separately distinct, the three times are established,” and 「大德佛陀提婆說：由異異故立三世」 tr. “the Bhadanta Fotuotipo says: because of difference upon difference, the three times are established” — the four answering one by one to those Xuanzang lists, while the style of address is “Bhadanta” and not “the Venerable,” and the names are wholly by transcription. Puguang’s glosses on the names are precisely the bridge between this set of renderings by sound and by sense.

For one of the four the whole system of rendering differs. For Buddhadeva, Xuanzang gives “relation” (relation as regarding), while Paramārtha does not use the character for relation at all and gives “difference upon difference” (earlier and later mutually different). What is meant doctrinally is one and the same, while the characters chosen for the rendering are wholly different — the most conspicuous systematic divergence in this fascicle. Recorded together without adjudication.

Divergences of rendering in a sūtra title and a verb. The Sūtra of Ultimate Emptiness stands in Paramārtha as the Sūtra of Real Emptiness, its citation reading 「眼根若生無所從來，若滅無所增集」 tr. “if the eye faculty arises it comes from nowhere, if it ceases it is increased and gathered nowhere,” differing by one character from Xuanzang’s “made and gathered nowhere”; the core sense of “coming from nowhere” and of “was not before and is now, having been returns to nonexistence” corresponds line by line closely in the two texts, and this is the passage among this fascicle’s adduced proofs where the two translations stand nearest each other.

One name of a non-Buddhist reproached differs. In the proof from teaching concerning the staff-and-topknot non-Buddhist, Paramārtha names, within the same line of doctrinal reasoning, a class of non-Buddhist different from Xuanzang’s; and re-checked by a fault-tolerant method, Xuanzang’s two characters for “staff-and-topknot” yield zero throughout Paramārtha’s text — the scene is certainly there, and only the name of the non-Buddhist reproached differs between the two texts. This is not a failure of searching but a real divergence of the translations, recorded without adjudication.

Divergences of format and of fascicle arrangement. Xuanzang gives “the stanza says / the treatise says,” the whole stanza coming first and the prose after; Paramārtha gives “the verse says / the explanation says,” embedding his explanation line by line, so that one stanza is often broken into several passages for citation, the passage on real existence in the three times being especially fragmented. The fascicle arrangement does not align either — Paramārtha gives this chapter in two fascicles only, Xuanzang in three; and the stretch answering to this fascicle straddles a boundary between two of Paramārtha’s own fascicles. The division into fascicles is a matter of the translation bureau and not evidence of doctrinal sections; the opening and closing of fascicles do not align, and the scenes align throughout.

4. Coordinates for Parallel Reading

The problem disputed in this fascicle — if the two times of past and future are not really existent, how are cause and result unbroken — did not come to an end with this fascicle. The Yogācāra lineage took it up head-on, its treatise saying expressly that since past and future are not really existent, to say they are not permanent may be granted while to say they are not annihilated is hard — and setting up the answering line, that if past and future were really existent it might be granted that they are not annihilated; and it resolves the two extremes of annihilation and permanence by a consciousness that turns constantly like a stream. ⇒ Upon one and the same difficulty, this fascicle records the two parties and closes with “profound,” while the Yogācāra lineage sets up another entity of continuity to bear it. The entity established differs, and the difficulty met is entirely the same. This is the straightest line this fascicle draws forward.

The entity of the latent tendency too admits of parallel reading. This fascicle establishes its name by the character “following” — subtle, twofold increase by attachment, following after, binding along, all four senses being stated upon “accompanying.” The Yogācārabhūmi lineage establishes its entity upon seeds, its text saying that among the seeds belonging to one’s own being, what is gathered under the class of the afflictions is named coarseness and also named latent tendency, while what is gathered under the class of maturation and under the remaining indeterminate classes is named coarseness alone and not named latent tendency. ⇒ The one establishes the name by the manner of its working, the other by the entity it depends upon; both lineages grant that the latent tendency is a portion lying dormant and constantly following, while the entity to which it is fastened differs. Where the numbers agree and the frames differ, they may be read side by side; where the frames differ and the problem is the same, they may be read side by side all the more.

As for that simile of drawing a boat against the current at the end of this fascicle, it belongs to no lineage’s point at issue. Whether the three times exist may be disputed, and the opening and gathering of the number of items may be disputed; but that going with the current is easy and going against it hard is a thing neither party will dispute. That the whole fascicle closes here is not without meaning.

Before Reading This Fascicle

This fascicle is the close of the chapter on the afflictions in the Abhidharmakośa, and it is also the fascicle in the whole treatise that most resembles a ledger book. The two foregoing fascicles finished the entity, the number, the realm-affiliation, and the increase by attachment of the latent tendencies; this one opens its mouth and adds no new dharma at all — it merely changes the names: fetter, bond, latent tendency, secondary affliction, entanglement, and grime of affliction besides, six names turning back to look once more over the same set of afflictions. The three ancient commentaries let the knife of their outlines fall at one and the same place here: Puguang’s Record marks “below, the second, explaining the six gates of fetter and the rest,” Yuanhui’s Verse Commentary marks “from here the second, explaining the six gates of fetter and the rest,” and Fabao’s Commentary, citing the treatise’s own words, marks “it gathers up the four foregoing gates and produces the gate that follows” — three houses dividing their fascicles independently and the boundary agreeing throughout.

Why change the names? Because each name measures one function of the afflictions: a fetter ties, a bond binds fast, a latent tendency lies dormant, a secondary affliction follows along, an entanglement enwraps in the present moment, a grime defiles. The same set of things changes its number when the rule changes, and this is the move this treatise loves best; if in reading this fascicle one only remembers the numbers and never asks “which rule was this number taken with,” what one is left with at the end is a heap of names and marks impossible to keep hold of.

There is one more thing to say first. Concealed within this fascicle is a chapter-level main division of the outline — with the four words “now this should be considered” (the page anchor falling in the middle of the fascicle) the whole chapter turns from “explaining the entity of the afflictions” into “explaining the cessation of the afflictions.” This cut does not fall at the fascicle boundary; it falls in mid-fascicle. The three ancient commentaries all mark the second of the main text here at once: Puguang’s Record gives “below, within this chapter, the second of the main text, explaining the cessation of the afflictions,” Fabao’s Commentary gives “from here below, the second of the main text, explaining abandonment,” and Yuanhui’s Verse Commentary gives “from here the second of the main text, explaining the cessation of the afflictions.” This fascicle is therefore in fact two halves: the first half is still the general account book of the afflictions, and the whole second half is the method of settling accounts — the four causes of abandoning the afflictions, the four kinds of counteragent, abandonment from the side of the object, the four kinds of distance, and at the last the six-tiered mechanism of the nine full comprehensions closing the whole chapter. A separate large section will be set up below when that point is reached; take note of the turn made there.

I. Six Names: No New Dharma Added, Only a Change of Rule

〔1.1〕Why five more must be spoken of after the outflows and the torrents. First the question: the outflows, the torrents, the yokes, and the appropriations were just finished in the foregoing fascicle, and the names of the afflictions are already not few; why set up the five of fetter and the rest besides? The treatise’s answer is a single sentence — “it is just these afflictions that, because the senses of fetter, bond, latent tendency, secondary affliction, and entanglement differ, are further spoken of as five kinds.” The key lies in the three words “the senses differ”: it is not that dharmas have been added, but that senses have been added. What “outflow” takes is the leaking out, what “torrent” takes is the drowning and sweeping away, what “fetter” takes is the tying and binding, what “entanglement” takes is the enwrapping of what is presently active — the names are a functional profile of the afflictions and not an inventory of them.

The easy error: to take the six gates as six classes of affliction, and then to ask “how many fetters are there, how many entanglements, and what do they come to when added?” They cannot be added. The six gates are six fresh stocktakings of one and the same set of accounts, each with a different angle of questioning. In practice: when you read on and find nine fetters, ten entanglements, six grimes, do not hurry to add them up; ask first by which function this gate is gathering them in.

II. The Nine Fetters: Nine Cords from Craving and Ill Will to Envy and Avarice

〔2.1〕The list of the nine fetters and the definition of the fetter of craving. First the question: which nine are the nine fetters? The treatise lists them in good order — “first the fetter of craving, second the fetter of ill will, third the fetter of pride, fourth the fetter of ignorance, fifth the fetter of views, sixth the fetter of grasping, seventh the fetter of doubt, eighth the fetter of envy, ninth the fetter of avarice.” Among them there is one defining sentence for the fetter of craving that matters most: “the fetter of craving here means greed in the three realms.” Notice that it does not say “greed in the desire realm” — greed in all three realms counts as the fetter of craving, and so greed for form and greed for the formless are cords no less, not excused because they are subtle. As for the three of pride, ignorance, and doubt, the treatise says “for the rest, their marks should be adjudged as is fitting,” handing the definitions back to the places already adjudged in the foregoing fascicles rather than repeating them.

The easy error: to read “craving” as human affection. The craving here is another name for greed, and what it measures is adhesion; the subtler the greed for a state of absorption, the less it looks like an affliction, and yet it enters the fetters all the same. In practice: the very first of the nine fetters blocks off the line of retreat that says “what I feel is a pure liking” — greed in all three realms goes into the ledger without exception.

〔2.2〕A very fine exchange: the dharma bound by the fetter of craving alone and not by the fetter of views. The treatise inserts here an extremely fine posed question: is there a dharma that is associated with views and yet is bound only by the fetter of craving and not by the fetter of views, while the latent tendency of views still increases by attachment upon it? Answer: there is. It is, at that stage where the cognition of origination has already arisen and the cognition of cessation has not yet arisen, the dharmas associated with the two graspings that are abandoned through seeing cessation and through seeing the path.

First ask: why that stage in particular? Because the all-pervading fetter of views (that which can take another category for its object) has already been forever abandoned at the abandoning through seeing origination, and so the fetter of views cannot bind them; while the fetter of views that is not all-pervading can be neither their object nor associated with them, both roads being cut. But the fetter of craving takes an object universally, and can still bind them as its object. As for the latent tendency of views, that is their own latent tendency of the two graspings increasing by attachment upon them, which is a different matter from whether the fetter of views binds them.

The easy error: to take “being bound by a certain fetter” and “a certain latent tendency increasing by attachment” for one and the same thing. It is exactly these two that the treatise here means to pry apart — binding is spoken of from the side of the fetter’s power to bind, increase by attachment from the side of a latent tendency growing upon its object, and the one may be present where the other is not. In practice: when in reading Abhidharma you meet a fine exchange of this kind wedged into the middle of a list, do not skip it; such passages are very often demonstrating exactly how the criterion just established is to be used.

III. The Fetters of Views and of Grasping: An Equal Count of Things and the Division of What Grasps from What Is Grasped

〔3.1〕Why the three views and the two graspings must be split into two fetters. First the question: since the fetter of views gathers in the views, why not gather the two graspings in as well and set up one fetter only? The treatise answers: “the three views and the two graspings are equal in things and equal in grasping” — two layers of reason. The first layer is “equal in things”: the three views have eighteen things and the two graspings likewise eighteen, the count of entities being exactly equal; things are well gathered only where one side is more and the other less, and an equal number is precisely a reason for setting them up apart. The second layer is “equal in grasping”: the three views are the side of what is grasped, the two graspings the side of what grasps, and since what grasps and what is grasped differ, they are “set up as two fetters.”

The easy error: to suppose that “equal in things” means the two are of the same nature and therefore ought to be gathered. It is exactly the reverse — the treatise here uses the equality of the count to prove that they can each stand independently as a fetter without imbalance, and then uses the difference of what grasps from what is grasped to prove that they must be divided. In practice: hereafter, when you meet a dispute in this treatise over “opening out” and “gathering together,” look first for the two rules — the count of entities and the function; almost every joining and dividing in this treatise is settled by these two.

IV. Envy and Avarice Enter the Fetters: Two Conditions, and the Trouble Brought by the Ten Entanglements

〔4.1〕Why, of the ten entanglements, only envy and avarice are raised into the nine fetters. First the question: the entanglements being so many, why are only the two of envy and avarice promoted to fetters? The treatise answers, “because these two alone are unwholesome and arise of themselves” — two conditions must be present at once: being unwholesome alone (not extending to the indeterminate), and being able to arise of themselves (able to become active on their own without leaning upon a root affliction abandoned through seeing). “These two alone have both senses present; the rest are not so.”

〔4.2〕The author takes his own reason apart on the spot. This passage shows the character of the treatise better than any other. Having established the reason, the treatise at once points out its limit itself: “if the entanglements were eight alone this explanation would pass, but if one grants that there are ten entanglements this explanation is unreasonable” — since rage and concealment equally possess those two senses. The Vaibhāṣika school granting ten entanglements, this explanation is no longer sufficient, and so a second reason is established instead: “one should say that the faults of envy and avarice are especially grave, in that these two occur again and again”; and three further sets of reasons are added — that they can be causes of lowliness and of poverty (envy drawing lowliness, avarice drawing poverty), that they display universally the two classes of secondary affliction, the dejected and the glad (envy belonging to the dejected, avarice to the glad), and that they harass the two divisions of those gone forth and those at home (turned about again into the three pairs of gods and asuras, humans and gods, another’s party and one’s own).

The easy error: to take the string that follows as a “supplement.” It is not a supplement but a replacement — the earlier reason has already been adjudged unreasonable by the author himself under the premise of ten entanglements. In practice: in reading a treatise one must see clearly the premise under which a reason holds; one and the same conclusion does not rest on the same reason under a scheme of eight entanglements and under a scheme of ten. This passage is also a specimen of the treatise’s form: establishing one’s own account, refuting it oneself, and establishing another instead, all completed within one and the same passage.

V. The Five Fetters of the Lower Part: Two Keep You In, Three Drag You Back

〔5.1〕Which five they are, and what it is they favor. First the question: what does “the lower part” of “favoring the lower part” refer to? The treatise answers, “these five favor and benefit the realm that is the lower part” — “that is, the desire realm alone obtains the name of the lower part”; and the five are the view of a real person, the grasping at rules and observances, doubt, greed for the objects of desire, and ill will. The division of labor is quite clear: the latter two (greed for the objects of desire, ill will) make a man “unable to pass beyond the desire realm,” while the former three (the view of a real person, the grasping at rules and observances, doubt) make it that “even if he were able to pass beyond, by the former three he comes down again.” The simile the treatise gives is “like the jailer who keeps guard and the patrolman who goes his rounds” — the jailer shuts you in, the patrolman catches back the one who has got out, the one keeping guard and the other pursuing.

〔5.2〕Another reading: the lower part means sentient beings of the lower sort. The treatise records alongside “some other master says”: the lower part refers not to a realm but to two senses, “the lower sentient beings, that is, the ordinary beings,” and “the lower realm, that is, the desire realm”; the former three can obstruct one from passing beyond the class of ordinary beings (abandon them and one becomes a noble one), the latter two can make one not pass beyond the lower realm. ⇒ One and the same name has two ways of being cut, and the treatise records both without adjudicating.

The easy error: to suppose that one of the two accounts must be right. The form of this treatise is that where the sūtras have no express statement both parties are recorded together. In practice: the simile is very serviceable — when your practice is stuck, first distinguish whether you are being shut in by “greed for the objects of desire and ill will,” or whether you had got out and were caught back by “the view of a real person, the grasping at rules and observances, and doubt”; the two ways of being stuck are counteracted differently.

VI. The Stream-Enterer Abandons Three Fetters: Gate and Root, Two Ways of Gathering

〔6.1〕Six are plainly abandoned; why does the sūtra speak only of three. First the question: the afflictions abandoned by a stream-enterer are in truth six (besides the three fetters there are the extreme view, the grasping at views, and wrong view); why does the sūtra speak only of “abandoning three fetters”? The treatise answers with entire honesty: “in principle one ought to say that six afflictions are abandoned; it is because of the gathering of gate and of root that it speaks only of abandoning three” — speaking of three is not that fewer are abandoned, but raising the three that gather in as representatives of the six.

The gathering has two layers. One is “gate”: among the six there are three kinds of class — that extending to one category alone, that extending to two, that extending to four; speaking of abandoning these three kinds gathers in those three gates. The other is “root”: among the six there are three that turn along with another — “the extreme view turns along with the view of a real person, the grasping at views turns along with the grasping at rules and observances, wrong view turns along with doubt”; speaking of abandoning these three kinds gathers in those three roots. Hence “in saying that three are abandoned it has already said that six are abandoned.”

〔6.2〕A third explanation: the three obstructions to going to another region. The treatise records another explanation besides, taking the simile of a traveler: whoever would go to another region has three obstructions — not wishing to set out, losing the right road (because he goes by a wrong road), and doubting the right road; and one who goes toward liberation has three similar obstructions: by the view of a real person he dreads liberation and will not set out, by the grasping at rules and observances he clings to a wrong path and loses the right road, and by doubt he is deeply irresolute about the path.

The easy error: to read “abandoning three” as a shrunken list. In practice: this passage gives a very useful self-examination — when you halt upon the path, is it because you have no wish to walk at all, because you have taken the wrong road, or because you are walking and yet do not believe this is the road? Each of the three obstructions answers to one fetter.

VII. The Five Fetters of the Upper Part: The Subtler, the Harder to Pass; Five Cords Keeping One in the Upper Two Realms

〔7.1〕The symmetry of the upper five with the lower five. First the question: there being fetters favoring the lower part, why set up fetters favoring the upper part besides? The treatise answers: “if these five kinds are not yet abandoned, they can make sentient beings unable to pass beyond the upper realms. Because they favor and benefit the upper realms, they are called the fetters favoring the upper part” — the five being greed for form, greed for the formless, restlessness, pride, and ignorance. Their symmetry with the lower ones lies in this: the lower keep one in the desire realm, the upper keep one in the upper two realms; and the names are alike taken from “favoring and benefiting” — whichever part they favor gives them their name.

The easy error: to suppose that the afflictions of the upper realms, being indeterminate, are no obstruction. The afflictions of the upper two realms are indeed not unwholesome, and yet they bind all the same — the quality of good and evil and the matter of binding or not binding are two different things. In practice: restlessness and pride within a state of absorption are the hardest of all to notice, because they show no coarse mark; that the treatise sets them beside the two greeds for form and the formless is precisely a reminder that the stiller the place, the finer the cord.

VIII. The Three Bonds: Why There Are Only Three Bonds

〔8.1〕The entity of the three bonds and the reason for establishing three. First the question: the afflictions being so many, why are only three bonds established? The treatise first gives the entity — “first the bond of greed, that is, all greed; second the bond of hatred, that is, all hatred; third the bond of delusion, that is, all delusion” — and then the reason: “because it is in accordance with the three feelings that bonds are said to be three”: toward pleasant feeling the bond of greed increases by attachment (both kinds of increase by attachment, as object and as associated), toward painful feeling hatred, toward neutral feeling delusion.

〔8.2〕One misunderstanding that has to be cleared away. The treatise blocks it beforehand itself: “although toward neutral feeling there are greed and hatred as well, they are not as delusion is” — it is not that there are no greed and hatred toward neutral feeling, but that greed and hatred are not so fitting toward neutral feeling as delusion is. And it adds a limiting sentence: “with regard to one’s own continuum, and taking the three feelings, pleasant and the rest, as the object of the bonds, this settled statement is made” — this correspondence is a settled statement made with regard to one’s own continuum and with the three feelings as object, and is not a claim that greed arises only upon pleasant feeling.

The easy error: to read the three feelings and the three poisons as an iron rule of one-to-one correspondence. The treatise has said expressly that this is only a settled statement made with regard to “the feeling that is the object.” In practice: this set is the most serviceable index for contemplative practice — where there is pleasure look for greed, where there is pain look for hatred, where there is neither look for delusion; the third is the most easily missed, because neutral feeling gives no sting, and that is exactly the hotbed of delusion.

IX. The Latent Tendencies Dealt with in One Line: The Treatise’s Method of Abridgment

〔9.1〕Why this gate has only five characters. The stanza reads simply “the latent tendencies have been spoken of above,” and the prose adds only one sentence: “the latent tendencies are six, or seven or ten, or ninety-eight, as has been said above.” First ask: why, of the six gates, is this one alone so summary? Because the latent tendencies are the very subject of the two foregoing fascicles, where the four tiers of opening and gathering — six, seven, ten, ninety-eight — have been adjudged in the fullest detail; here only one slot needs to be filled on the roster of the six gates.

The easy error: to suppose that this gate has nothing to discuss and to skip it. It in fact points to something: the skeleton of the six gates was fixed beforehand, and the latent tendencies are merely one slot placed into that skeleton; in other words, the treatise does not set up a gate because it has something to say, but because the skeleton ought to have this slot. In practice: when in reading this treatise you meet an extremely short section, it is usually not that there is little content, but that it has been finished elsewhere; what to do then is to go back and fill in that place, not to pass it over.

X. Secondary Afflictions: One General Name, One Particular Name

〔10.1〕Two uses of one and the same name. First the question: do the root afflictions themselves count as secondary afflictions? The treatise’s answer has two layers. The layer of the general name: “these afflictions too are called secondary afflictions, since they all follow along with mind and do the work of harassing it” — the root afflictions too are called secondary afflictions, the sense taken being “following along with mind and harassing it.” The layer of the particular name: there is besides a set of defiled mental factors different from the root afflictions, gathered under the aggregate of formations, which “are also called secondary afflictions because they arise along with the afflictions, but are not called afflictions, since they are not roots” — the sense taken being “arising along with the afflictions.”

The easy error: to run the two layers into one, whereupon one cannot make sense of “why greed too is called a secondary affliction and rage too is called a secondary affliction.” One and the same name often has a general and a particular use in this treatise, and the criterion is the place from which the sense is taken: the one follows along with mind, the other follows along with an affliction. In practice: the ten entanglements and six grimes below all belong to the layer of the particular name; when you meet the words “secondary affliction,” ask first whether the general or the particular sense is being taken here.

XI. The Ten Entanglements: Two Schools on Eight and Ten, and the Glossing of Each Name

〔11.1〕Two schools on the number. First the question: are the entanglements eight or ten? The treatise records the two accounts together — “yet the Treatise on Categories of Dharmas says there are eight entanglements, and the Vaibhāṣika school says there are ten entanglements,” the ten being the eight plus rage and concealment. The treatise does not adjudge which is right; but below, in treating the reason for setting up envy and avarice as fetters, it has already taken the premise of “ten entanglements” in earnest (see section IV), from which one may see that within this treatise both schools’ numbers have to be able to run.

〔11.2〕Each name glossed, and every one given a verifiable mark. Envy is “that which makes the mind unglad at the flourishing affairs of others”; avarice is “that which, contrary to the giving of goods, of dharma, and of skill, makes the mind cling and grudge”; regret is just the regret over the evil deed done; sleep is “that whose nature is to make the mind murky and slight, having no force to sustain the body”; rage is “that which, apart from hatred and harmfulness, makes the mind flare up toward the sentient and the insentient”; concealment is “the hiding of one’s own sin.” Among them regret and sleep carry one important limitation — “the two entanglements of regret and sleep take only the defiled”: wholesome remorse and wholesome sleep do not enter this count.

The easy error: to read sleep as physiological slumber and regret as all reflection. The treatise has already cut them in half with the words “taking only the defiled.” In practice: glossing names is not memorizing names and marks; it is being given a set of marks that can be checked against oneself on the spot — being unglad at another’s flourishing, a mind that clings and grudges, a mind murky and slight, a mind that flares up, the hiding of a sin: all five marks can be verified upon the mind here and now.

XII. The Four Homogeneous Outflows of the Entanglements: Three Divergent Accounts of Concealment

〔12.1〕From which root affliction each entanglement flows out. First the question: what does “homogeneous outflow” mean? It means issuing forth as a continuation of the same class — a certain secondary affliction is the residual force of the same class as a certain root affliction. The treatise’s assignments are: non-conscience, avarice, and restlessness belong to the homogeneous outflow of greed; shamelessness, sleep, and torpor to the homogeneous outflow of ignorance; envy and rage to the homogeneous outflow of hatred; regret to the homogeneous outflow of doubt.

〔12.2〕Concealment alone has three accounts. The treatise here records three houses together: “some say concealment is the homogeneous outflow of greed. Some say it is the homogeneous outflow of ignorance. Some say it is the homogeneous outflow of both,” and with the words “knowing and not knowing, each in its own order” it points out what the three accounts fasten upon: if the man who hides his sin knows in his heart that he has sinned and hides it deliberately, it inclines to greed; if he covers it over without knowing or noticing, it inclines to ignorance; if there is something of both, then it belongs to the homogeneous outflow of both.

The easy error: to suppose that one of the three accounts must be chosen. The treatise here does not even hand down a verdict; it merely gives a rule for reading the three accounts (knowing / not knowing). In practice: what makes concealment hard to classify is precisely that hiding a sin may arise either from cherishing one’s name and gain (greed) or from having no knowledge of oneself (delusion) — and this item, of all of them, comes closest to daily use.

XIII. The Six Grimes of Affliction: Another Set, Coarser in Their Defilement

〔13.1〕The names and marks of the six grimes. First the question: where does a grime differ from an entanglement? The name of grime is established in “these six kinds arise from the afflictions, and being coarse in the mark of their defilement are called grimes of affliction” — what it takes is defilement that is coarse and manifest. The six are vexation, harmfulness, resentment, flattery, deception, and self-intoxication, each with its definition: vexation is “stubbornly holding fast to sinful matters, and by this not accepting reasonable admonition and instruction”; harmfulness is “being able to press hard upon others, and by this able to perform beating and reviling and the like”; resentment is “upon the matter that was the object of rage, pondering again and again, binding up enmity and not letting go”; flattery is “crookedness of mind, and by this being unable to display oneself as one truly is”; deception is “deluding others”; self-intoxication has been explained above.

〔13.2〕Four homogeneous outflows and one proof from a sūtra. Deception and self-intoxication belong to the homogeneous outflow of greed; harmfulness and resentment to the homogeneous outflow of hatred; vexation to the homogeneous outflow of the grasping at views; flattery to the homogeneous outflow of the views. For this last the treatise adds a proof from a sūtra of set purpose — “as it is said, what is crookedness? It is the evil views; hence flattery is definitely the homogeneous outflow of the views.”

The easy error: to confound resentment with rage. The treatise has already divided their work — rage is the flaring up of the present moment, resentment is the aftermath, “pondering again and again, binding up enmity and not letting go”; the one is fire, the other embers under ash. In practice: four of the six grimes are directed at other people (harmfulness, resentment, flattery, deception), and this set is the best test of whether “my afflictions have come down upon somebody else.”

XIV. The Gate of Abandonment for the Entanglements and Grimes: Five Extend to Both Seeing and Cultivation, the Rest Are Abandoned through Cultivation Alone

〔14.1〕Why only five extend to what is abandoned through seeing. First the question: the ten entanglements and six grimes are sixteen in all, and in respect of abandonment they fall into two batches; what is the criterion of the division? The treatise answers: the five of non-conscience, shamelessness, sleep, torpor, and restlessness “extend to what is abandoned through seeing and through cultivation, since these extend to arising in association with the afflictions of both categories” — they can arise in association with the afflictions of both the categories abandoned through seeing and abandoned through cultivation, and so their abandonment is settled by whichever they accompany: “as it is associated with what is abandoned through seeing this truth, so it is said to be abandoned through seeing this truth.” The rest — envy, avarice, regret, rage, concealment, together with the six grimes — “are abandoned through cultivation alone, since they arise of themselves.”

〔14.2〕The definition of “arising of themselves” must be seen clearly. The definition the treatise gives is: “because they are associated only with the ignorance dependent on another that is abandoned through cultivation, they are called arising of themselves” — “of themselves” does not mean that they are strong; it means that they do not lean upon a root affliction abandoned through seeing, and keep company only with the solitary ignorance abandoned through cultivation.

The easy error: to read “arising of themselves” as free and unconstrained, or as powerful. It is a technical term, and what it measures is the relation of dependence. In practice: this item explains something very real — the path of seeing abandons a great many afflictions, and yet not one of envy, avarice, rage, and the other things that trouble one most in daily life is abandoned by it; because they do not hang on the side of what is abandoned through seeing at all, and can only be ground away grade by grade by the path of cultivation.

XV. The Gate of the Three Ethical Qualities: Three of the Desire Realm Extend to Two Qualities

〔15.1〕Which of them are not purely evil. First the question: are the secondary afflictions unwholesome without exception? The treatise answers that they are not — “the three of sleep, torpor, and restlessness affiliated to the desire realm all extend to the two qualities of unwholesome and indeterminate; all the rest are unwholesome alone”; while those of the upper two realms are “all gathered under the indeterminate quality alone.”

The easy error: to read “indeterminate” as harmless. Indeterminate says only that it does not draw a maturational result; it does not say that it is no obstruction to the path. Torpor, restlessness, and self-intoxication in the upper realms are all indeterminate, and yet they remain fetters binding one to the upper realms (as was said in section VII). In practice: this gate should be read together with the earlier item “envy and avarice are unwholesome alone” — it is precisely because envy and avarice are purely unwholesome that they qualify for promotion to fetters, while sleep, torpor, and restlessness extend to two qualities and so cannot be promoted. The two gates corroborate each other, and this is the use of the six gates’ repeated stocktaking.

XVI. The Gate of Realm-Affiliation: A Story of Flattery and Deception Reaching the First Meditative Absorption

〔16.1〕The great Brahmā king deludes the monk Aśvajit. First the question: how do such coarse mental factors as flattery and deception get up as far as the first meditative absorption? The treatise takes an event from a sūtra as its proof: “because the great Brahmā king concealed the facts of his own condition and, displaying an appearance, deluded the monk Aśvajit” — the great Brahmā king hid his real state and put forth an appearance that turned the questioner aside. And the treatise states its own form: “although these two were analyzed above, since the sense is connected they are adjudged over again here.” As for the rest: torpor, restlessness, and self-intoxication extend to the three realms; the remaining eleven are in the desire realm alone.

The easy error: to suppose that no such mental factor could be found in the states of the heavens. The one the treatise adduces is the Brahmā king in the highest position of all. In practice: this story is cited more than once in Abhidharma, and its weight lies here — a high station does not mean that the afflictions have been purified; what keeps flattery and deception in being is that there is still an image of an “I” to be maintained.

XVII. The Gate of the Ground of Consciousness: Which Arise in the Mental Consciousness Alone

〔17.1〕The three batches belonging to the mental ground alone. First the question: can all six consciousnesses give rise to the afflictions? The treatise divides them into two batches: all that is abandoned through seeing, all pride and sleep, and among the secondary afflictions those that arise of themselves — “all these depend upon the mental consciousness, since there is no room for their arising in dependence upon the group of five consciousnesses”; the rest extend to all six consciousnesses — greed, hatred, and ignorance abandoned through cultivation, and the non-conscience, shamelessness, torpor, and restlessness associated with them, and those gathered under the factors of the great defiled ground.

The easy error: to suppose that “arising in dependence upon the five consciousnesses” means that eye, ear, nose, tongue, and body are giving rise to afflictions. The treatise’s criterion is whether or not they can arise apart from discrimination: an affliction abandoned through seeing must depend upon discrimination, and the five consciousnesses do not discriminate, so there is no room for their arising. In practice: this gate explains why the path of seeing can abandon a great batch of afflictions all at once — whatever is propped up by discrimination collapses as soon as the principle is seen through; while the batch that extends to all six consciousnesses must be ground away slowly by the path of cultivation, because they do not live upon reasons.

XVIII. The Feelings Associated with the Root Afflictions: The Glad Mode and the Dejected Mode

〔18.1〕A new rule: the glad mode and the dejected mode. First the question: how does one adjudge which feeling a given affliction is associated with? The treatise crosses two lines — whether the mode of apprehension is glad or dejected, and whether it pervades the six consciousnesses or not. Greed is of the glad mode and pervades the six consciousnesses, and so is associated with joy and pleasure; hatred is of the dejected mode and pervades the six consciousnesses, and so is associated with sorrow and pain; ignorance turns in both the glad and the dejected modes and pervades the six consciousnesses, and so is associated with all the four foregoing feelings; wrong view turns in both modes but belongs to the mental ground alone, and so extends to sorrow and joy; doubt is of the dejected mode and belongs to the mental ground alone, and so is associated with sorrow; the remaining four views and pride are of the glad mode and belong to the mental ground alone, and so are associated with joy.

〔18.2〕Why wrong view can be either glad or dejected. The treatise poses itself a question and answers it: “because they had formerly done, in that order, sinful or meritorious action” — one who had formerly done sin is glad when wrong view arises (supposing there is no retribution), while one who had formerly done merit is dejected when wrong view arises (supposing his merit is void). For doubt there is a separate explanation: “one who harbors irresolution seeks a settled knowing, and so his mind is downcast.”

〔18.3〕Above the particular mark there is one further item, neutral feeling. The particular marks done, the treatise adds one general mark: “all of them are associated with neutral feeling, since the latent tendencies, at the stage where the continuum is about to be abandoned, have their force spent and must abide in neutral feeling.” For the upper grounds, “all of them, as is fitting, are associated with the feelings that arise together with their own ground and their own consciousness.”

The easy error: to suppose that neutral feeling conflicts with the joy, sorrow, pain, and pleasure spoken of above. There is no conflict — the particular mark is spoken of from the side of the mode of apprehension of what is presently active, the general mark from the side of the stage where the force is spent; the two rules are not measuring the same stretch. In practice: this gate is the most serviceable passage of all for contemplation — from “is this moment glad or dejected” you can infer backward which class of affliction is present.

XIX. The Feelings Associated with the Secondary Afflictions: The Four That Are Associated Universally, beyond Sorrow and Joy

〔19.1〕The assignments group by group. The six of envy, regret, rage, vexation, harmfulness, and resentment “are all associated with the faculty of sorrow, since they turn in the dejected mode and belong to the mental ground alone”; avarice is associated with joy, and the treatise adds a reason — “since the mark of avarice is extremely like that of greed”; flattery, deception, sleep, and concealment arise together with both sorrow and joy, because “sometimes one performs flattery and the rest with a glad mind, and sometimes one performs them with a dejected mind”; self-intoxication is associated with joy and pleasure, being associated with pleasure in the third meditative absorption and with joy in the grounds below.

〔19.2〕Why the last four are associated with all five feelings. Non-conscience, shamelessness, torpor, and restlessness “are all four associated universally with the five feelings,” the reason lying in the ground-factors they belong to — the first two are factors of the great unwholesome ground (pervading every unwholesome mind), the latter two factors of the great defiled ground (pervading every defiled mind). Whatever “pervades” naturally pervades in respect of feeling as well.

The easy error: to memorize the assignments by rote. The treatise gives a reason for every one of them (the glad or dejected mode, belonging to the mental ground alone, being gathered under a ground-factor); what is to be learned is the reason, not the table. In practice: the group of flattery and deception is most worth noticing — one and the same flattery may be done with a smile or done with a heavy heart; which shows precisely that a secondary affliction is fixed in its nature not by the emotion but by the mode of apprehension.

XX. The Five Hindrances: In the Desire Realm Alone, Two Entities Gathered into One, Five Because They Obstruct the Aggregates

〔20.1〕How it is proved that the five hindrances are in the desire realm alone. First the question: are the three of torpor, restlessness, and doubt among the five hindrances, like sleep and regret, in the desire realm alone? The treatise answers that they are in the desire realm alone, and takes its evidence from a sūtra — “since the sūtra says that these five kinds are purely and entirely an unwholesome heap”; and “in the form and formless realms there is nothing unwholesome,” so the five hindrances do not extend to the upper realms. This is a way of inferring that fixes the realm from the quality: being purely unwholesome, they cannot go up.

〔20.2〕Food, counteragent, and function being the same, two entities are gathered into one hindrance. First the question: torpor and sleep are two dharmas, restlessness and regret are two dharmas; why is each pair gathered into one hindrance? The treatise gives three words and glosses them one by one — “food is what is eaten, also called provision”; “counteragent is what counteracts, also called non-food”; “function is the activity performed, also called capability.” The food of the hindrance of torpor-and-sleep is five dharmas: murkiness, dissatisfaction, yawning and stretching, unevenness of diet, and dullness and inferiority of mind; its non-food is the ideation of light; and their function is the same in that “both can make the nature of mind sink and grow murky.” The food of the hindrance of restlessness-and-regret is four dharmas: thought of kinsfolk, thought of one’s native land, thought of not dying, and recollection of the various sport, mirth, merrymaking, and being waited upon formerly undergone; its non-food is calm abiding; and their function is the same in that “both can make the mind unquiet.” The three being the same, they are gathered into one.

〔20.3〕The reason for establishing five only: they obstruct the five aggregates. “Because these alone can be an eminent obstruction to the five aggregates” — greed and ill will obstruct the aggregate of discipline, torpor and sleep obstruct the aggregate of discernment, restlessness and regret obstruct the aggregate of concentration; and concentration and discernment being gone, doubt arises with regard to the four truths, and “because of doubt they can make it that even liberation and the knowledge and vision of liberation do not arise.” The five hindrances obstruct all the way down, blocking off the whole of the five aggregates without outflows (discipline, concentration, discernment, liberation, and the knowledge and vision of liberation).

The easy error: to read “food” as food and drink. The food here has the sense of nourishing and sustaining — what is feeding this hindrance. In practice: this is the passage in the whole fascicle that can most directly be taken up and used. If you would abandon torpor and sleep, what the treatise gives is not willpower but an inspection of its food (have you eaten too much, is there something you are unhappy about) and of its non-food (arouse the ideation of light); if you would abandon restlessness and regret, likewise first inspect the four thoughts (longing for home, longing for one’s country, supposing one will not die, savoring the merriment of former days), the non-food being calm abiding. To abandon an affliction, first cut off its board and keep — this is the method this passage leaves behind.

XXI. The Dispute over the Order of the Five Hindrances: One Public Exchange

〔21.1〕The author first refutes the old explanation of his own school. First the question: since greed and ill will obstruct discipline, torpor and sleep discernment, restlessness and regret concentration, then the order in the sūtra (torpor-and-sleep before, restlessness-and-regret after) is in difficulty — for “it is necessarily in dependence upon concentration that discernment arises, and so what obstructs concentration ought likewise to come before what obstructs discernment”; restlessness and regret obstructing concentration, they ought in principle to be ranked before torpor and sleep. This objection is raised by the author himself; and Puguang’s Record marks the refuter plainly here, giving “the author refutes the explanation of the Sarvāstivādin masters given above.”

〔21.2〕The first explanation: exchange the assignments. The treatise adduces “some other master says”: among the five hindrances, torpor-and-sleep and restlessness-and-regret “can obstruct, in that order, the aggregate of concentration and the aggregate of discernment” — torpor-and-sleep obstructing concentration and restlessness-and-regret discernment; the assignments once exchanged, the order in the sūtra falls right. This account has a sūtra in proof besides: “one who cultivates meditative poise dreads torpor and sleep; one who cultivates the discernment of dharmas dreads restlessness and regret.”

〔21.3〕The second explanation: set up the five by five causes instead. The treatise adduces further, “some other, differing account: five are established by five causes,” and unfolds it by question and answer: “what do they say? That is, at the stage of moving about . . .” — first, at the stage of moving about, one takes up toward form and the rest the two marks of the lovable and the detestable; afterward, at the stage of dwelling, there arise the two hindrances of greed for the objects of desire and ill will, and these two obstruct the mind that is about to enter absorption; by this, at the stage of actually entering absorption, calm and insight cannot be rightly practiced, and so torpor-and-sleep and restlessness-and-regret arise, obstructing in that order calm abiding and insight; and by this, at the stage of emerging from absorption, when one considers dharmas, doubt is an obstruction in its turn. ⇒ This explanation leaves the assignments untouched and rearranges the five hindrances upon the time-axis of practice (the stage of moving about, of dwelling, of entering absorption, of emerging from absorption), so that the order falls into place of itself.

The easy error: to read the three passages as three accounts set side by side. They are one continuous line: the objection → the first explanation (exchanging the assignments) → the second explanation (changing the axis). In practice: this passage demonstrates the two legitimate ways of dealing with a difficulty about the order in which a sūtra’s items appear — either change the assignments, or change the axis relied upon; and neither requires altering the text of the sūtra. As to which of the two explanations prevails, the treatise does not adjudge, nor do the ancient masters (see Further Resonances).

XXII. 【Chapter-Level Main Section】From Explaining the Entity of the Afflictions to Explaining Their Cessation

At this point the section of the whole chapter that explains “the entity of the afflictions” is finished. With the four words “now this should be considered” the chapter enters the second of the main text: explaining the cessation of the afflictions. The three ancient commentaries let their knives fall here at once: Puguang’s Record gives “below, within this chapter, the second of the main text, explaining the cessation of the afflictions,” Fabao’s Commentary gives “from here below, the second of the main text, explaining abandonment,” Yuanhui’s Verse Commentary gives “from here the second of the main text, explaining the cessation of the afflictions” — and the six subsections the three of them mark agree as well: the four causes of abandoning the afflictions, the four kinds of counteragent, the place from which the afflictions are abandoned, the four kinds of distance, the cessation obtained in abandoning the afflictions, and the nine full comprehensions.

This cut does not fall at the fascicle boundary but within this fascicle; in reading, one should take a fresh breath here oneself: up to now the counting has been of how many afflictions there are and which slot they belong in; from here on, everything asks one thing — how they came to be gone.

XXIII. The Four Causes of Abandoning the Afflictions: Not Every Affliction Is Abandoned by “Seeing Clearly”

〔23.1〕Look first at how the question is raised. The treatise’s transitional sentence is itself a difficulty: as to the all-pervading afflictions of another realm, and among those abandoned through seeing cessation and seeing the path those that take what has outflows for their object, “at the stage of their abandonment their object is not known, and when their object is known they are not being abandoned” — at the time of abandoning, their object is not yet known; by the time the object is known, they were long since abandoned. Whence it appears that “it is not necessary that they be abandoned through fully comprehending their object.”

〔23.2〕The division of labor among the four causes. The afflictions abandoned through seeing are abandoned by the first three causes: first, abandoned through fully comprehending their object — those abandoned through seeing suffering and seeing origination that take their own realm for their object, and those abandoned through seeing cessation and seeing the path that take what is without outflows for their object (their object being fully comprehended, the affliction is abandoned); second, abandoned through the abandoning of what takes them as object — those abandoned through seeing suffering and seeing origination that take another realm for their object, “since what takes one’s own realm as object can take them as object, and when what takes them as object is abandoned they are abandoned along with it”; third, abandoned through the abandoning of their object — those abandoned through seeing cessation and seeing the path that take what has outflows for their object, “since what takes the outflowless as object can be their field, and when their object is abandoned what takes it as object is abandoned along with it.” The afflictions abandoned through cultivation are abandoned by the fourth cause alone: abandoned because the counteragent arises, “since when the path counteracting this grade arises, the afflictions in this grade are abandoned all at once.”

〔23.3〕One rule of counteragents running the other way. The treatise adds besides: “the afflictions of the highest of the high grades have the path of the lowest of the low grades for their counteragent; and so on down to the afflictions of the lowest of the low grades, which have the path of the highest of the high grades for their counteragent.”

The easy error: to take “abandoned through fully comprehending their object” as the sole model of abandoning the afflictions. This whole section is meant precisely to reduce it to a quarter. In practice: the second and third causes speak of abandonment by association — abandon what takes it as object and the object is abandoned along with it, abandon the object and what takes it as object is abandoned along with it; in practice it is not that every affliction must be dealt with separately once. As for the coarsest afflictions being abandoned by the weakest path and the subtlest waiting for the strongest, this item will be used again and again at the stage of the path of cultivation.

XXIV. The Four Kinds of Counteragent: Four Names, Two Arrangements

〔24.1〕The definitions of the four kinds. The counteragent of abandoning is the path of no interval (the moment of actual abandoning); the counteragent of upholding is “the path after this, since by it the acquisition of this abandonment can be upheld”; the counteragent of distancing is “all the paths after the path of liberation, since by those paths what is abandoned here can be made more distant,” and the treatise records alongside “some other master says: the path of liberation as well” (the path of liberation itself counting too); the counteragent of disgust is “where there is a path that sees the faults of this realm and deeply gives rise to disgust.”

〔24.2〕The second arrangement the author gives. Having listed the four names, the treatise at once states of its own accord: “yet if one would speak well of these counteragents, in principle the order ought to be as follows” — disgust (the path of preparation, arising with suffering and origination for its object), abandoning (the path of no interval, arising with all for its object), upholding (the path of liberation, arising with all for its object), distancing (the path of distinguished progress, arising with all for its object). ⇒ The former is a listing by the names, the latter a listing by the order in which they actually occur in practice.

The easy error: to take the two orders as contradictory. The treatise says it quite clearly itself: the second is “if one would speak well of it” — one and the same set of dharmas, listed once by name and once by fact. In practice: these four answer exactly to the four stages of the path of preparation, the path of no interval, the path of liberation, and the path of distinguished progress; hang the counteragents of the afflictions back upon the four paths, and the map of practice comes right.

XXV. The Afflictions Are Abandoned from the Side of the Object: A Principle Very Easily Misread

〔25.1〕It cannot leave what is associated with it; it can only leave its object. First the question: what, after all, does abandoning an affliction abandon? The treatise answers: “it cannot be made to leave the dharmas associated with it, but it can be made to depart from its object, so that it no longer arises toward that object” — the associated dharmas are simultaneous and coexistent and cannot be pried apart; the only thing that can be pried apart is the affliction and its field.

〔25.2〕How past afflictions are abandoned. The treatise poses itself a difficulty: “abandoning future afflictions” passes well enough (it can be made not to arise toward its field again), but what of past afflictions? And it blocks off one ready answer besides — if “from the side of the object” were taken to mean “abandoned through fully comprehending the object,” that too will not do, “this too being unreasonable, since it is not settled” (the foregoing section having proved that to be only one of four causes). The correct account the treatise gives changes the road: the abandonment of the afflictions in one’s own continuum is “because their acquisition is abandoned” (the acquisition of the affliction being abandoned, the affliction no longer belongs to this body); the abandonment of the afflictions in another’s continuum, and of all form and all undefiled dharmas, is “because all the afflictions in one’s own continuum that take them as object are at last completely abandoned.”

The easy error: to suppose that abandoning an affliction is digging something out of the mind. The treatise’s model is nothing of the kind — what is abandoned is the acquisition and the road to the object, not the affliction annihilated. In practice: this also explains why another’s afflictions are not abandoned by me, nor outer form abandoned by me — all I can abandon is that portion of me which takes them as object.

XXVI. The Four Kinds of Distance: Four Similes and a Passage of Four Rounds

〔26.1〕The four kinds of distance and the four similes. The treatise begins with the two words “it is transmitted” and lists four kinds: distance of mark, as the four great elements (in one mass together and yet differing in mark); distance of counteragent, as keeping and breaking the discipline (in one body together and yet counteracting one another); distance of place, as the eastern and the western seas (in one world together and yet separated by quarter); distance of time, as the past and future ages (resting upon one dharma together and yet separated in time). The structure common to the four similes is distant while in the same place: so “distant” is not a spatial notion but the notion of “not reaching one another.”

〔26.2〕Four rounds of exchange. What follows is the longest passage of question and answer in this fascicle, and Puguang’s Record marks the speaker for each round: “relative to what is distance spoken of” is the author’s question, “relative to the present age” is the Sarvāstivādins’ answer; “what has ceased with nothing intervening . . . how are they called distant” is the author’s objection, “because the natures of the ages differ . . . only then are they called distant” is the Sarvāstivādins’ answer; “if that is so, the present . . . would be called distant” is the author’s further objection. The further objection has three limbs: if the unconditioned dharmas have no activity, how is it that they are on the contrary called near; if it is because the present universally obtains the unconditioned that they are called near, then the past and future ages are equally so, and how much more so empty space, which is unconditioned and has no obtaining at all; and if it is said that the past and future are called distant because they are separated by the present when set against one another, while the present is near to both ages, then the past and the future, being both adjacent to and separated from the present, “have both names, and it should not be said flatly that they are distant.”

〔26.3〕The closing sentence. At the end the treatise gives one explanation: “if one goes by correct principle, one should say that the past and the future are called distant because they are apart from the intrinsic mark of a dharma, the future because it has not yet obtained the intrinsic mark of a dharma, and the past because it has already cast off the intrinsic mark of a dharma” — near and distant being spoken of not from temporal interval but from whether the intrinsic mark of a dharma has been obtained. The word “and the rest” at the end of the stanza the treatise also accounts for: “the word ‘and the rest’ is to show that the instances raised are not exhausted.”

The easy error: to read past the two words “it is transmitted.” When this treatise uses “it is transmitted,” the tone keeps its distance — this is passing on what has been transmitted without taking it as one’s own. In practice: this passage is the best specimen of reasoning in the fascicle; all three objections use one and the same stroke — take the opponent’s criterion and measure something else with it, and see whether it still holds. As for the attribution of this closing sentence and Saṅghabhadra’s counter-refutation, see Further Resonances.

XXVII. The Afflictions Are Not Abandoned Twice, but Release from Bondage Is Obtained Again: Six Occasions

〔27.1〕First distinguish two things. First the question: the path being always in distinguished progress, will the afflictions already abandoned be abandoned a second time? The treatise answers that they will not — “by that path these afflictions are abandoned all at once, and there can be no abandoning of the afflictions again at a later time.” But release from bondage (the unconditioned displayed by abandonment) does have a “repeated obtaining”: release from bondage itself does not improve, “and yet when the path advances there may be the sense of arousing again a superior acquisition of it” — what changes is the acquisition, not the release from bondage.

〔27.2〕The six occasions and the subtraction stage by stage. The six occasions are the arising of the counteracting path (the path of liberation), the obtaining of a fruit (the four of stream-entry, once-returning, non-returning, and arhatship), and the sharpening of the faculties (the turning of the faculties). Below this the treatise performs a subtraction stage by stage: for what is abandoned through seeing the four truths in the desire realm, and through seeing the first three truths in the upper realms, all six occasions are present; for what is abandoned through seeing the truth of the path in the upper realms there are five (the arising of the counteracting path being itself the obtaining of a fruit, not divided into two occasions), and for the first five grades abandoned through cultivation in the desire realm likewise five (the fruit of stream-entry being deducted); for the sixth grade four (the obtaining of the fruit of once-returning falling together with the arising of the counteragent), and for the seventh and eighth grades four (the first two of the four fruits being deducted); for the ninth grade three (the arising of the counteragent being itself the obtaining of the fruit of non-returning), and for what is abandoned through cultivation in the upper two realms, apart from the ninth grade of the summit of existence, likewise three (the first three of the four fruits being deducted); and for the ninth grade of the summit of existence two alone (the arising of the counteragent being itself the obtaining of arhatship).

〔27.3〕The author marks this himself as a provisional statement. At the end he adds a limiting sentence: “thus is it said for the present, as a statement of what may be” — those of sharp faculties do not sharpen the faculties, and those who realize by leaping do not pass through the fruit of stream-entry and the rest; each requires a further subtraction.

The easy error: to memorize this string of numbers by rote. It has in fact only one rule of computation — wherever the arising of the counteracting path and the obtaining of a fruit fall in the same moment, they are merged into one occasion; and wherever that fruit was obtained beforehand, one occasion is deducted. In practice: this section also accounts for one fact of practice: what has been abandoned need not be abandoned again, and yet one and the same “already abandoned” will be certified afresh at every checkpoint that follows, and certified more eminently at each.

XXVIII. The Nine Full Comprehensions, One: How the Nine Are Counted Out

〔28.1〕First distinguish the full comprehension that is cognition from the full comprehension that is abandonment. Before turning to the nine full comprehensions the treatise first sets up the names: “full comprehension is of two kinds: first the full comprehension that is cognition, second the full comprehension that is abandonment” — the full comprehension that is cognition is the cognition without outflows; the full comprehension that is abandonment is “just the abandonments, the name of the cause being established here upon the result,” that is, those very “abandonments,” merely called by the name of the cognition that can fully comprehend. The nine full comprehensions all speak of the latter.

〔28.2〕The reckoning of six plus three. For what is abandoned through seeing the truths, six are established: in the desire realm the abandonment of the first two categories (seeing suffering, seeing origination) together establishes one, and the next two categories each establish one, three in all; for the upper two realms reckoned together, likewise “the first two abandonments one, the two each one,” making three — six in all. For what is abandoned through cultivation, three are established: what is abandoned through cultivation in the desire realm establishes one, namely the full comprehension of the exhaustion of the five fetters of the lower part; what is abandoned through cultivation in the form realm establishes one, namely the full comprehension of the exhaustion of craving for form; what is abandoned through cultivation in the formless realm establishes one, namely the full comprehension of the everlasting exhaustion of all fetters.

〔28.3〕Why what is abandoned through cultivation divides form from formless while what is abandoned through seeing does not. The treatise closes this subsection with a question and answer: “because the counteragents of what is abandoned through cultivation are not the same” — what is abandoned through cultivation has in the form realm and in the formless realm each its own counteracting path, and so they must be set up separately; while for what is abandoned through seeing the counteragent of the upper two realms is the same, and so they are established together.

The easy error: to suppose that the nine were thrown together at random. Nine = six abandoned through seeing + three abandoned through cultivation, and the six and the three each have their own rule of joining and dividing. In practice: note the two occurrences of “because it is established together with the foregoing” — the exhaustion of the lower part is reckoned together with the three abandoned through seeing in the desire realm, and the exhaustion of all fetters likewise together with those before it; full comprehension is a cumulative account, not an itemized one, and this point is the key later on at the losing and obtaining of full comprehensions.

XXIX. The Nine Full Comprehensions, Two: Whose Result Each of Them Is

〔29.1〕Six the result of receptivity, three the result of cognition. The treatise first adjudges receptivity and cognition: six are the result of receptivity, namely the six full comprehensions of what is abandoned through seeing in the three realms; three are the result of cognition, namely the three of the exhaustion of the lower part, the exhaustion of craving for form, and the exhaustion of all fetters, “since these three full comprehensions are the result of the path of cultivation.” The treatise poses itself a difficulty: how can receptivity be called full comprehension (full comprehension being properly the name of a cognition)? The answer is twofold: “because all the receptivities are the retinue of cognition, as the retinue of a king is conventionally given the name of the king”; “or because receptivity and cognition have one and the same result.”

〔29.2〕The results divided by ground. The result of access concentration is all nine (in dependence upon access one can abandon what is abandoned through seeing and through cultivation in the three realms); the result of the fundamental meditative absorptions is five or eight — “as for what is called five, the Vaibhāṣika masters say” that the fundamental grounds can only abandon the afflictions gathered under the form and formless realms, while abandonment in the desire realm is the result of access concentration alone; “as for what is called eight, the Venerable Ghoṣa says” that the fundamental grounds are counteragents of abandonment for the afflictions of the desire realm too, for whoever has first become free of desire and enters upon the seeing of the truths in dependence upon a fundamental ground, in respect of the abandonment of what is abandoned through seeing in the desire realm, “grants that a separate path draws forth the acquisition without outflows,” except only the full comprehension of the exhaustion of the lower part (that one being the result of access concentration alone). The intermediate meditative absorption is to be spoken of as the fundamental ones are. The result of the marginal ground of the formless (the access of the sphere of space) is one alone, namely the exhaustion of craving for form; and the result of the first three fundamental formless grounds is likewise one alone, namely the exhaustion of all fetters.

〔29.3〕The results divided by path. The result of the mundane path is two (the exhaustion of the lower part, the exhaustion of craving for form); the result of the noble path is nine; the result of the cognition of dharma is three, of subsequent cognition two; the result of the class of the cognition of dharma is six, of the class of subsequent cognition five, “since the word ‘class’ gathers in both cognition and receptivity.”

The easy error: to read “five or eight” as hesitation on the treatise’s part. This is two houses set side by side — five is what the Vaibhāṣika masters say, eight what the Venerable Ghoṣa says, and the text of the treatise itself carries the speakers’ labels. In practice: this section is the treatise’s art of “counting one and the same set of dharmas over again from a different axis” at its height — counted once by receptivity and cognition, once by ground, once by path, the nine figures rearranged each time. That they can be rearranged without conflict is precisely the proof that the classification of the nine full comprehensions stands.

XXX. The Nine Full Comprehensions, Three: Why Nine and Not More

〔30.1〕The four conditions. First the question: the abandonment of dharmas with outflows has “many entities and stages”; why are only nine established? The treatise answers that it is by four conditions. The first three conditions establish the six that are the result of receptivity: obtaining the acquisition of release from bondage without outflows, being deficient in the summit of existence, and the removal of the twofold cause — “the abandonments must have these three conditions present to be given the name of full comprehension, and where one is wanting they are not.”

〔30.2〕Verifying these three conditions stage by stage. At the stage of an ordinary being there is the removal of the twofold cause but no acquisition of abandonment without outflows, and no deficiency yet in the summit of existence, so they are not called full comprehension; at the stage of a noble one, from entering upon the seeing of the truths up to before the receptivity of the subsequent kind concerning suffering, there is the acquisition of abandonment without outflows but no deficiency yet in the summit of existence and no removal yet of the twofold cause; at the stage of the cognition of the subsequent kind concerning suffering and the receptivity of the dharma kind concerning origination, there is deficiency in the summit of existence and yet no removal of the twofold cause (the all-pervading causes abandoned through seeing origination not yet being abandoned); and only at the stages of the cognition of dharma and subsequent cognition that follow are the three conditions present in the abandonments there obtained, so that a full comprehension is established at each such stage.

〔30.3〕The fourth condition and a fifth that is absorbed. For the three that are the result of cognition, the fourth condition of “crossing the realm” is added — “what is called crossing the realm is that the afflictions and the rest within this realm have all been wholly departed from.” There is besides an account that establishes “departure from what is bound together with it” as a fifth condition; the treatise’s choice is that this condition “does not differ in function from the removal of the twofold cause and from crossing the realm, and so, although it differs in sense, it is not separately stated.” At the end a necessary distinction is added: every stage of crossing the realm removes the twofold cause, but the removal of the twofold cause is not always a crossing of the realm, and so crossing the realm must be established apart from the removal of the twofold cause.

The easy error: to read the four conditions as a list of four requirements. It is in fact a calculating machine — the treatise takes it and runs every stage through it, and whatever fails to pass has no full comprehension established. In practice: this section demonstrates the most exacting mode of proof in the treatise: first establish the conditions, then take all the stages and try them one by one, eliminating on the spot whatever does not fit; the number nine was not laid down by decree but reckoned out.

XXXI. The Nine Full Comprehensions, Four: Who Possesses How Many

〔31.1〕The staircase at the stage of the path of seeing. An ordinary being “decidedly has no possibility of possessing a full comprehension.” A noble one abiding at the stage of the seeing of the truths: from the beginning up to the receptivity of the dharma kind concerning origination he possesses none; at the cognition of the dharma kind concerning origination and the receptivity of the subsequent kind concerning origination he possesses one; at the cognition of the subsequent kind concerning origination and the receptivity of the dharma kind concerning cessation, two; at the cognition of the dharma kind concerning cessation and the receptivity of the subsequent kind concerning cessation, three; at the cognition of the subsequent kind concerning cessation and the receptivity of the dharma kind concerning the path, four; at the cognition of the dharma kind concerning the path and the receptivity of the subsequent kind concerning the path, five. ⇒ Step by step it mounts, one added at every two minds.

〔31.2〕The increase and decrease at the stage of the path of cultivation and at the stage beyond training. At the stage of the path of cultivation, beginning from the cognition of the subsequent kind concerning the path, up to one not yet wholly free of desire in the desire realm, and one fallen back from freedom from desire, all possess six. One wholly free of desire whose craving for form is not yet exhausted, or one previously free of desire, from the cognition of the subsequent kind concerning the path up to before the arising of the path of the distinguished fruit that exhausts form, possesses one alone (the exhaustion of the lower part); and one who from the exhaustion of craving for form or from the stage beyond training falls back and arouses the entanglements of form likewise possesses one. One who has craving for form, from the everlasting exhaustion of craving for form, or one previously free of form, from the arousing of the path exhausting form, up to before he is wholly free of craving for the formless, possesses two (the exhaustion of the lower part, the exhaustion of craving for form); and one who falls back from the stage beyond training and arouses the entanglements of the formless likewise possesses two. One abiding at the stage beyond training possesses one alone, namely the full comprehension of the everlasting exhaustion of all fetters.

The easy error: to suppose that the number increases all the way. It increases first and is then merged — at the path of seeing the six are added one by one, and at the path of cultivation they are on the contrary gathered into one (the exhaustion of the lower part having swallowed the five before it). In practice: this is exactly the cashing out of that “because it is established together with the foregoing” of section XXVIII; when you read “one beyond training possesses one alone,” do not suppose this a falling back — that one is the one that has taken everything into itself.

XXXII. The Nine Full Comprehensions, Five: Why the Non-Returner and the Arhat Must Be Gathered into One

〔32.1〕Two conditions. First the question: why is it that the fruit of non-returning and the fruit of arhatship alone “gather the abandonments together and establish them as one full comprehension”? The treatise answers that it is because two conditions are present: “first crossing the realm, second obtaining a fruit. Those two stages alone have both conditions present.” — these two stages have both crossed a whole realm (the non-returner crossing the desire realm, the arhat the upper two realms) and are at the same time stages of obtaining a fruit of the ascetic life; only when the two conditions fall together at once is there any qualification to gather them.

The easy error: to suppose that the gathering is for the sake of remembering fewer names. It is not — the gathering is itself a qualification in doctrine, granted only where two conditions are present. In practice: this section interlocks with the earlier item on the condition of crossing the realm (section XXX); wherever this treatise meets a “gathering” or an “opening out,” it must without exception first produce the qualification for gathering, and never gathers for convenience.

XXXIII. The Nine Full Comprehensions, Six: Losing and Obtaining, and the Close of the Whole Chapter

〔33.1〕The four cases of losing. Losing one: falling back from the stage beyond training or from the stage of the exhaustion of craving for form into complete freedom from desire being lost. Losing two: a non-returner who from the stage of the exhaustion of craving for form falls back and arouses the entanglements of the desire realm, and the same person when he attains arhatship. Losing five: one previously free of desire who afterward enters upon the seeing of the truths, at the cognition of the subsequent kind concerning the path, obtaining the exhaustion of the lower part and losing the five before it. Losing six: a noble one not yet free of desire at the time when he obtains freedom from desire.

〔33.2〕The three cases of obtaining, and one empty slot. “‘Obtaining likewise’ means that there is obtaining one, obtaining two, and obtaining six, only obtaining five being excepted.” Obtaining one: obtaining what was not obtained before, and falling back from the stage beyond training and arousing the entanglements of form. Obtaining two: falling back from the stage beyond training and arousing the entanglements of the formless realm. Obtaining six: falling back from non-returning.

〔33.3〕Notice that empty slot. The words “only excepting five” in the stanza’s “obtaining likewise, only excepting five” are the mechanism of the whole passage: there is a losing of five, but no obtaining of five. Because “losing five” occurs at the moment of obtaining the exhaustion of the lower part and swallowing up the five before it, which is a stage of outgo only and no income; there is no stage whatever at which five are obtained at one time.

The easy error: to read losing as regression and obtaining as progress. Losing here has two sources — losing by falling back (arousing the entanglements and falling back) and losing by advancing (obtaining the exhaustion of the lower part and losing the five before it, attaining arhatship and losing two); the latter is a merging of accounts and not a loss. In practice: the whole chapter closes upon the sentence “in adjudging the latent tendencies, the analysis of abandonment is finished” — for the sake of adjudging the latent tendencies, the analysis of “abandonment” has been finished incidentally. Read the fascicle through and then look back at that sentence in section I, “because the senses differ, five kinds are further spoken of,” and you will find that the head and tail of the whole fascicle are one and the same thing: one and the same set of afflictions must have its accounts settled afresh at every change of angle; until, at the last settling, nothing at all is left on the books.

Further Resonances: The Mirror of the Three Commentaries

The above is the sense of this fascicle’s text taken by itself. A separate section is set up below 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 a different way of reading one and the same passage — recorded together without adjudication, each returned to its own place.

1. The Adjudications of the Three Commentaries and Their Divergences

One commentary wholly absent. The commentaries on this fascicle are in a peculiar 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 the Latent Tendencies is wholly wanting, and so this fascicle has no line-by-line commentary of his to hang upon. Although his general conspectus in Fascicle One does contain a catalogue phrase such as “the chapter on the latent tendencies, sixty-nine lines,” that is not commentary and cannot be made to stand for commentary, nor is text from his other fascicles pieced in to supply it. The want is marked as it stands.

The three houses agree wholly in the skeleton of their outlines and differ in character. Their words for opening the section: Puguang’s Record gives “below, the second, explaining the six gates of fetter and the rest,” Yuanhui’s Verse Commentary gives “from here the second, explaining the six gates of fetter and the rest,” the two houses agreeing almost word for word; while Fabao’s Commentary, citing the treatise’s own words, marks “it gathers up the four foregoing gates and produces the gate that follows,” taking the momentum of “gathering what precedes and producing what follows.” In the divisions themselves the three agree — Fabao gives “below, five and a half stanzas, the first explaining the fetters. Within it three kinds: first the nine fetters, second the five fetters of the lower part, third the five fetters of the upper part,” no different from Puguang’s divisions, save that Fabao alone marks the stanza count (five and a half stanzas), which neither Puguang nor Yuanhui does. In their manner of adducing evidence each has his preference: Puguang inclines to cite the Mahāvibhāṣā, Fabao to cite the Treatise Conforming to Correct Principle — as in glossing the name “fetter,” where Fabao cites “the Treatise Conforming to Correct Principle says, ‘it has the power of binding with respect to an object and with respect to arising, and so is called a fetter.’” Recorded together without adjudication.

Two places of “no conclusion”: Puguang does not adjudge where adjudication was due. This is the point most worth noticing in reading the three commentaries on this fascicle. The first is the dispute over the order of the five hindrances (section XXI of the main text): Puguang marks the refuter plainly as “the author refutes the explanation of the Sarvāstivādin masters given above,” marks the words of the Sautrāntikas, marks “setting out the Sautrāntikas’ explanation,” and marks “what do they say” as the question and everything from “that is, at the stage of moving about” on as the Sautrāntikas’ answer — the speakers marked out plainly all the way through, and yet at the end of the passage he hands down no verdict of any kind, and having finished the exposition turns at once to the next subsection. The second is the passage on distance (section XXVI): Puguang marks each round, “the author asks,” “the Sarvāstivādins answer,” “the author objects,” “the Sarvāstivādins answer,” “the author objects further,” and for the last sentence marks “the author sets out the Sautrāntikas’ explanation of the mark of distance” — saying expressly that that sentence is not the author’s own school but the author setting out another’s; and then equally does not adjudge which of the Sautrāntika and Sarvāstivāda explanations is right. ⇒ Puguang’s form here is: mark the speakers through to the end; do not hand down right and wrong on the author’s behalf.

One place with a conclusion: “five or eight” for the result of the fundamental grounds. As against the two places above where he does not adjudge, here Puguang hands down his own judgment first — “although the fundamental meditative absorptions are said to be five by some and eight by others, the account of five is the correct sense” — fixing the Vaibhāṣika masters’ “five” as the correct sense; and when the account of the Ghoṣa line is finished he adds the note “the Venerable Ghoṣa means: the fundamental grounds, set against the desire realm, have a counteragent of abandonment,” marking plainly that “eight” is the Venerable Ghoṣa’s divergent account and not the correct sense. Puguang also cites the Treatise Conforming to Correct Principle, Fascicle Fifty-Six, in support of his judgment. ⇒ One and the same commentary, three passages, three ways: two without adjudication, one with adjudication first; and the boundary between adjudging and not adjudging lies precisely in whether the text of the treatise itself already carries the speakers’ labels (the passage on “five or eight” carrying in the treatise itself “the Vaibhāṣika masters say” and “the Venerable Ghoṣa says,” Puguang therefore dares to adjudge; in the two former passages the treatise says only “some other master says” and “some other, differing account,” and Puguang therefore does not). Recorded together without adjudication.

Puguang’s characters in citing the treatise differ from the base text. Puguang cites the opening sentence of this fascicle, “thus have already been adjudged,” with two characters differing from the base text — the particle for “already” written in a near-identical variant form, and the verb “to adjudge” written with the discriminating-radical graph where the treatise habitually writes the disputation-radical graph. The latter is a difference of usage at the level of whole books between this treatise and the three commentaries, and is recorded as it stands without alteration.

The chapter title at the head of Yuanhui’s fascicle is not in the main text. The head of the main text of this fascicle in Yuanhui gives simply “Verse Commentary on the Root Text of the Abhidharmakośa, the twenty-first,” without the chapter name; the eighteen characters of the chapter name, “Chapter Five, Analysis of the Latent Tendencies, part three,” are a variant of the second witness and are found in that fascicle’s collation note. This is a methodological reminder for reading the commentaries — wherever one meets a chapter title at the head of a fascicle in a commentary, one must check its note separately and may not take it straight off as belonging to the main text.

2. Saṅghabhadra’s Refutations and Supplementary Explanations

Look first at the name: the chapter name in Saṅghabhadra’s two treatises is not “Analysis of the Latent Tendencies.” The Treatise Conforming to Correct Principle and the Treatise Revealing the Doctrine both give this chapter as “Adjudication of the Latent Tendencies.” To search the two treatises with this treatise’s chapter name is bound to yield a zero, while the scenes are plainly there — this is a common trap of cross-text collation, and one can only fix position by the structure of the scenes.

The passage on distance: Puguang has no conclusion, Saṅghabhadra has one. This is the clearest contrasting pair in the fascicle. The closing sentence of section XXVI, “if one goes by correct principle, one should say that the past and the future are called distant because they are apart from the intrinsic mark of a dharma . . . ,” Saṅghabhadra cites word for word and prefaces with “the Sūtra-master here says thus,” and immediately adjudges it: “his account is one-sidedly at variance with correct principle. Whatever has no intrinsic mark is not distance; this, being distance, must have an intrinsic mark, since it is gathered under distance, as the other distances are” — the line of reasoning being that if there were no intrinsic mark it would not be distance at all, and since the past and future are granted to be gathered under distance, it ought not to be said that they are “apart from the intrinsic mark of a dharma.” There is besides an earlier occurrence in the same passage: the self-posed difficulty in the treatise’s prose, “if it is because the present universally obtains the unconditioned that they are called near, then the past and future ages are equally so,” Saṅghabhadra likewise cites as the Sūtra-master’s account and refutes with “by the principle explained above, the parallel does not hold.” ⇒ On one and the same topic, Puguang chooses to hand down no verdict and Saṅghabhadra chooses to adjudge it at variance; the two houses differ in their choice about “conclusion,” and it is only recorded, not adjudged.

The passage on obtaining and losing full comprehensions: one technical correction of a stage-term. The treatise glosses “losing five” as “that is, one previously free of desire who afterward enters upon the seeing of the truths, at the cognition of the subsequent kind concerning the path,” which Saṅghabhadra cites as “as for what is called losing five, the Sūtra-master explains: that is, one previously free of desire, at the stage of the cognition of the subsequent kind concerning the path,” and corrects with “this should be said only of the moment of the receptivity of the subsequent kind concerning the path, since at the cognition of the subsequent kind concerning the path they have already been lost. To speak of obtaining and losing is to speak with regard to what is about to be”; and he adds a distinction — it must be stated that this is for one who enters upon the seeing of the truths in dependence upon access concentration, since for one who enters in dependence upon a fundamental absorption there is no acquisition of release from bondage without outflows in respect of abandonment in the desire realm, and those three full comprehensions were never obtained at all, and “what was not obtained before cannot be said to be lost now.” ⇒ What is at issue in this item is not a doctrinal position but the precision of the stage-term and the distinction of the absorption relied upon. This item recurs identically in the Treatise Revealing the Doctrine.

One honest negative proposition. The two occurrences on distance (the first set above) are wanting in the Treatise Revealing the Doctrine — the corresponding passage there lays out flatly the four kinds of distance and the four similes, with neither the words “Sūtra-master” nor any restatement or criticism of this treatise’s sentences, and passes straight on to the next stanza. This is not a failure of searching but a difference of fullness between the two treatises themselves: the Treatise Conforming to Correct Principle has the refutation and the Treatise Revealing the Doctrine has none. It is recorded as it stands.

3. The Divergences of Paramārtha’s Earlier Translation

His renderings form a system of their own, and one group of them calls for the greatest care. Paramārtha’s Expository Treatise on the Treasury forms a system of its own in the passage corresponding to this fascicle. Among the nine fetters, Xuanzang names them from the intrinsic nature of the affliction (fetter of craving, fetter of ill will), Paramārtha from the mode of the affliction (fetter of conformity, fetter of contrariety); and so too with the three bonds (bond of greed and bond of delusion / bond of desire and bond of ignorance). What most calls for notice is the six grimes of affliction: Xuanzang gives vexation, harmfulness, resentment, flattery, deception, self-intoxication, Paramārtha gives not letting go, oppressive harassing, binding up fault, flattery, deception, and intoxication — the six names differing almost one for one, so that they can be matched only by Paramārtha’s own glosses (not letting go = stubbornly holding fast and not accepting right teaching, close to Xuanzang’s “vexation”; binding up fault = pondering again and again the object of a raging mind, close to Xuanzang’s “resentment”; oppressive harassing = injuring and humiliating another’s intention and performing beating and reviling, close to Xuanzang’s “harmfulness”; intoxication = Xuanzang’s “self-intoxication”). They may not be read across at face value. The personal name differs doubly as well: Xuanzang’s “the monk Aśvajit” Paramārtha gives as “Aśvajit of pure livelihood,” one and the same person and the rendering-characters entirely different.

Differences of form and of opening and gathering. Xuanzang has “the stanza says / the treatise says,” the whole stanza coming out first and the prose after; Paramārtha has “the verse says / the explanation says,” glossing line by line inset, so that one stanza is often broken into several passages. Hence the “four homogeneous outflows of the entanglements,” which Xuanzang writes as one stanza of four limbs continuously, Paramārtha breaks into five passages of verse before its sense is complete; and the four causes of abandoning the afflictions, which Xuanzang writes as one stanza of four lines continuously, Paramārtha lists in two separate places, the three causes for what is abandoned through seeing and the one cause for what is abandoned through cultivation. Conversely, in the passage on the faults of envy and avarice Paramārtha adduces a passage of ancient verse in evidence besides, which Xuanzang does not have. Fullness varies passage by passage, with no fixed direction.

One discrepancy in a number, recorded and left undecided. In the section on the results of receptivity and cognition among the nine full comprehensions, Xuanzang’s stanza gives “of these, six are the result of receptivity,” while Paramārtha’s explanation gives “these are the result of eight receptivities.” The doctrinal joint is the same in the two texts (the six full comprehensions of what is abandoned through seeing being the result of receptivity, the three of what is abandoned through cultivation the result of cognition), and yet there is indeed a discrepancy in the statement of the number. This is the one manifest numerical divergence in this fascicle, and it is recorded without adjudication — Paramārtha has not been altered to follow Xuanzang, nor Xuanzang to follow Paramārtha.

One instance of the two texts distributing matter differently between stanza and prose. In the gate of the secondary afflictions, Paramārtha’s verse has one line more, “it says they are not afflictions”; Xuanzang’s equivalent lies not in the stanza but in the prose (“they are not called afflictions, since they are not roots”). One and the same sense, put by one text into the stanza and by the other into the exposition: this is the structural difference of the two translations in miniature.

4. Coordinates for Parallel Reading

Two restorations in the collation. The base text of this fascicle has two places restored from the Koryŏ canon: the word “poverty” in “again, because these two can be causes of lowliness and poverty” (the Taishō original printing giving “greed”), and the word “instruction” in “and by this not accepting reasonable admonition and instruction” (the Taishō original printing giving “regret”). With the first restored, the interlocking pair — envy drawing lowliness, avarice drawing poverty — holds together; with the second restored, “admonition and instruction” becomes a word at all. Both Taishō original printings are recorded alongside in the apparatus, recorded without adjudication. There are besides several variant readings that touch sense more plainly — the word “after” in “all the paths after the path of liberation” for the counteragent of distancing (one witness giving “those”), the phrase “a separate path draws forth” in the result of the path of seeing in a fundamental ground (the Song, Yuan, and Ming editions inverting its word order, the Palace edition doubling “separate”), and the word “one” in “as for what is called obtaining one” in the obtaining of full comprehensions (the Palace edition giving “two”) — all recorded alongside for reference; the choice among variant readings is a doctrinal adjudication and does not belong in this section.

This slot of the secondary afflictions was afterward recompiled. This fascicle sets up the two senses of the secondary afflictions, general and particular (the general, following along with mind and harassing it; the particular, arising along with the afflictions and not being roots), and the layer of the particular name lists the ten entanglements and six grimes. The Yogācāra lineage, at the same doctrinal position, establishes instead twenty items — rage, resentment, concealment, vexation, envy, avarice, deception, flattery, harmfulness, self-intoxication, non-conscience, shamelessness, restlessness, torpor, lack of confidence, indolence, heedlessness, lost mindfulness, distraction, and incorrect knowing — and gathers them under the three classes of small, medium, and great (those arising each separately being small, those pervading the unwholesome medium, those pervading defiled minds great). ⇒ The one sense of “not a root, yet arising along with mind” is unchanged in the two lineages, while the items listed and the rule of classification have changed: this fascicle classifies by the homogeneous stream from which they arise (the outflow of greed, of hatred, of ignorance, of doubt), the Yogācāra by the extent of what they pervade. One and the same set of mental factors, two ways of indexing, which may be read side by side.

The setting side by side of the six gates: the Yogācāra lineage too has a table of the same type. This fascicle gathers one and the same set of afflictions under the six names of fetter, bond, latent tendency, secondary affliction, and entanglement; the Yogācārabhūmi, in its gate of defilement, has a setting side by side of the same type — nine fetters, three bonds, seven latent tendencies, three secondary afflictions listed in order, and after them the five hindrances as well. The listed names of the five hindrances agree entirely in the two lineages, only the third and fourth items differing in the fullness of their rendering (this fascicle giving torpor-and-sleep and restlessness-and-regret, the other torpor and sleep, restlessness and regret) — and the mechanism of this fascicle’s “two entities gathered into one hindrance” (food, counteragent, and function being the same) may be taken as the coordinate for reading them together.