俱舍論 十智·行相·諸門分別Abhidharmakośabhāṣya · The Ten Knowledges · Modes of Apprehension · Distinctions through the Several Gates

Before Reading This Fascicle

This fascicle is the first half of the seventh chapter of the Abhidharmakośa. The previous chapter (the Chapter on the Worthy and the Noble) walked the stairway of the noble stages to its end — the four wholesome roots, the sixteen minds of the path of seeing, the nine grades of the path of cultivation, the four fruits, the six kinds of arhat; that chapter spoke of “who” and of “to where.” This fascicle turns back and asks something more fundamental still: what in the end is that mind upon which one walks this road?

The treatise opens not with a definition but with a question: the previous chapter said at its opening “the several receptivities and the several knowledges,” and further on said “right view and right knowledge”; is there then any dharma that is receptivity and not knowledge? Is there any dharma that is knowledge and not view? — It does not first tell you what knowledge is; it first lets you see clearly where the line is drawn between knowledge and its two neighbors, receptivity and view. This is the typical move of Abhidharma: the sense of a technical term is not given by its definition but by its boundaries.

The ancient masters have a word on the order of this chapter. Puguang, explaining the name of the chapter and its order, says “because it decides and knows with weight it is called knowledge, and because this chapter explains it at length it is called analysis”; and again, “the reason the Chapter on Knowledge is explained next is that the previous chapter explained the result and this chapter explains the cause, and since the cause is the nearer as regards the result, knowledge is made out next” — the Chapter on the Worthy and the Noble speaks of the result, this chapter speaks of the cause that brings that result about, and cause and result are nearest of kin, hence the one follows immediately upon the other. Fabao offers a further gloss, saying that knowledge is “discernment deciding upon what has once been seen, called knowledge; or else, the mind that seeks to see the object having come to rest and the preparatory effort having slackened, it is called knowledge” — the first road speaks from “deciding upon an object already seen,” the second from “the seeking mind having stopped and the preparatory effort having slackened.” Both roads will be used within this fascicle: the first explains why receptivity is not knowledge, the second explains why the two knowledges of exhaustion and non-arising are not view.

As to the skeleton of this fascicle, the three ancient masters adjudge it much alike: first the difference of receptivity, knowledge, and view; second the difference of mark among the ten knowledges; third the modes of apprehension of the ten knowledges; fourth the analysis through the several gates. The first section is a bare hundred and sixty-odd characters, the second twelve hundred odd, the third close to two thousand nine hundred, the fourth over three thousand — the further on, the denser. In reading, it does no harm to keep one line in mind: the first three sections establish what knowledge is, while the fourth asks how these ten knowledges operate at the various stages; the former is a system, the latter a ledger. The numbers in that ledger section are extremely dense and are easily read as a list; but beneath every entry of numbers there is a reason, and that sentence at the close of the fascicle, “all this should be pondered as is fitting,” is the author handing the work of checking the arithmetic back to the reader.

I. Why Receptivity Does Not Count as Knowledge — the Boundaries of Three Technical Terms

〔1.1〕First ask: at which step do receptivity and knowledge part? The treatise’s answer has the weight of a single sentence: the eight receptivities are not knowledge, because “the doubt they are themselves to sever has not yet been severed.” Receptivity is the uninterrupted path, in the act of severing doubt; knowledge is the path of liberation, the doubt being severed. What makes a decision a decision is that the matter to be decided is already settled — and so one and the same noble discernment is called receptivity at the instant of severing, and only when the severing is done is it called knowledge.

Where it is easily mistaken: taking receptivity to be a grade below knowledge, an immature version of it. It is not. Receptivity is the instant in which force is being exerted, knowledge the instant in which the exertion is accomplished; as regards function, receptivity is on the contrary the portion that does the severing. Fabao here cites the Treatise Conforming to Correct Principle in aid, saying “or else, the mind that seeks to see the object having come to rest and the preparatory effort having slackened, it is called knowledge; the receptivities, as they arise, weigh and consider, and their intent and preparatory effort are fierce, and so they are not included in knowledge” — receptivity is not called knowledge precisely because it is too fierce, not because it is too feeble.

Where this lands: in reading this treatise, whenever two technical terms stand next to each other (receptivity and knowledge, the uninterrupted path and the path of liberation), ask first whether they are two time-phases of one and the same thing rather than two things.

〔1.2〕Ask next: why do the knowledge of exhaustion and the knowledge of non-arising not count as view? The treatise says, “because the seeking mind has come to rest and there is no weighing and considering.” The nature of view is to weigh and consider, which is still to seek forward; by the time of the knowledge of exhaustion and the knowledge of non-arising, all that was to be sought has been sought out, and that forward-pressing motion has stopped, and so they are not called view. This is exactly the position pointed at by Fabao’s second gloss, “the mind that seeks to see the object having come to rest.”

Where it is easily mistaken: reading “not view” as “does not see clearly.” It is precisely the reverse — it is not view because the seeing is already finished. Where this lands: in practice, whenever some portion of the mind is said to “come to rest,” first distinguish whether it has stopped from weariness or stopped because the work is done; here in this treatise it is the latter.

〔1.3〕Why is discernment with outflows knowledge one and all, yet only six of it view? The treatise says of discernment with outflows that it is “all included in the nature of knowledge,” and that among it “only six are also of the nature of view, namely the five defiled views and worldly right view, six in all.” There is a nicety here easily read past: the word “knowledge” in this treatise is not the exclusive property of the noble ones. Selecting out with outflows is also called knowledge; only, of what can weigh and consider and so be established as “view,” there are on the side of the outflows only the six, the five defiled views plus worldly right view.

Where it is easily mistaken: supposing that since a defiled view is defiled it ought not to be ranged with right view as “view.” But the criterion for view is the nature of weighing and considering, not wholesomeness or unwholesomeness. Where this lands: this is a measure used throughout the treatise — to judge whether a dharma belongs to a class, find the defining sentence of that class, and do not put wholesome and unwholesome, defiled and pure, in its place. The treatise closes with the sentence “since all of them select out dharmas, they are all included in the nature of discernment,” gathering noble discernment and discernment with outflows alike back into the nature of discernment, and sealing the relation of the three technical terms in eight characters.

II. The Skeleton of the Ten Knowledges — How Two, Three, and Ten Open Out and Gather In

〔2.1〕First ask: why ten, and not nine or eleven? The treatise first lists the names: conventional, dharma, subsequent, suffering, origination, cessation, path, others’ minds, exhaustion, non-arising. Immediately it says that these ten “gather up all knowledge” — not ten picked up at random, but a declaration that these ten exhaust the whole field of knowledge. Then it gathers back: ten gathered into two (with outflows, without outflows), the two opened into three (conventional, dharma, subsequent), the three opened into ten.

Where it is easily mistaken: reading two, three, and ten as three different classifications set side by side. They are not coordinate; they are three degrees of fineness upon one and the same set of dharmas: two is by nature (outflows and no outflows), three is by mark, ten is by difference of function. One and the same set of dharmas takes a different number when a different measure is laid against it — this is the most basic analytical habit of Abhidharma.

Where this lands: hereafter, wherever a formula such as “gathered up there are only two kinds” or “their marks differ into three” occurs, ask first which measure this layer is using, and the numbers will not fall into confusion.

〔2.2〕Why is conventional knowledge called conventional? The reason the treatise gives is “since for the most part it takes conventional objects such as a pot” — note that word “for the most part.” It is not said that it takes only pots and the like as object, but that in the greater part it takes conventional objects, and so the name is established from the greater part. At the same time the treatise says it is “universal,” taking all things conditioned and unconditioned as its object. These two sentences do not conflict: in naming a dharma the treatise commonly establishes the name from the greater part, not from the whole.

Where it is easily mistaken: seeing the word “conventional” and supposing its field to be narrow. In fact its object is the broadest of the three — dharma knowledge and subsequent knowledge are each confined to the desire realm or the upper realms, and it alone takes everything as object. Where this lands: in reading a technical term, distinguish “from what the name is established” from “how broad the object is”; they are two different things.

〔2.3〕How are dharma knowledge and subsequent knowledge divided? The treatise says, “in that order, they take the four truths of the desire realm and of the upper realms as their object.” Dharma knowledge takes the four truths of the desire realm as object, subsequent knowledge the four truths of the two realms of form and the formless. The ground of the division is not the height of the mind but the realm and level of the object taken. This division is to be firmly remembered, for it is the baseline of the long string of ledgers in the latter part of this fascicle (mutual objects, objects taken, and the cultivation at the several stages).

III. The Knowledges of the Four Truths and of Exhaustion and Non-Arising, Divided Out of Dharma and Subsequent

〔3.1〕First ask: are the knowledges of the four truths newly added, or a further cutting of what was already there? The treatise says it plainly: dharma knowledge and subsequent knowledge “are divided by the difference of their objects into the four knowledges of suffering, origination, cessation, and the path” — it is one more cut, made upon the two already established, according to the truth taken as object, and not the establishing of four further new knowledges. So that up to this step the entities are in fact only the two lines, dharma and subsequent, while the names already number six.

Where it is easily mistaken: imagining the ten knowledges as ten independent categories with no bearing upon one another. In fact they include one another layer upon layer, and that ledger of “mutual inclusion” after Root 7 (this knowledge includes one in whole and so many in part) is set there precisely to forestall this misreading.

〔3.2〕How do the two knowledges of exhaustion and non-arising emerge from these six? The treatise says: of these six, “those included in the stage beyond learning and not of the nature of view are called exhaustion and non-arising.” There are two conditions — being at the stage beyond learning, and not being of the nature of view. Exhaustion and non-arising are not a seventh and an eighth new entity; they are names taken on by the six knowledges at a particular stage.

Where this lands: this may be carried over directly to every place in the treatise where a name is established by stage — one and the same dharma acquires different names at different stages, and the entities need not increase.

〔3.3〕What do they take as object when they first arise? The treatise says, “these two, when they first arise, are only of the subsequent class of suffering and of origination,” the reason being “because they contemplate the aggregates of the summit of existence as their object with the six modes of apprehension that take suffering and origination for their object.” Immediately the treatise poses a question of its own: is the object of the vajra-like concentration the same as here? Answer: on the side that takes suffering and origination for object the reason is the same, and on the side that takes cessation and the path for object it is not. This one question and answer draws a clear line between the knowledge of exhaustion and the vajra-like concentration — the two are neighbors in stage while their objects are not wholly the same, which is where confusion comes most easily.

IV. The Threefold Limit upon the Knowledge of Others’ Minds

〔4.1〕First ask: on what ground is it said that the knowledge of others’ minds “cannot” know certain minds? The treatise says, “this knowledge has fixed marks with respect to its object” — a fixed mark is a hard limit; it is not an insufficiency of capacity but a limit set by its very nature. The limits are of two kinds: it does not know a superior mind, and it does not know past and future minds.

〔4.2〕The three directions of a superior mind. The treatise divides them into three aspects, level, faculty, and stage: as to level, the knowledge of a lower level does not know the mind of a higher; as to faculty, the dull faculty of one resolute through faith and of one whose liberation depends on circumstance does not know the mind of the sharp faculty of one attained through view and of one whose liberation does not depend on circumstance; as to stage, of the non-returner, the fruit of worthiness of a śrāvaka, the pratyekabuddha, and the greatly awakened one, the knowledge of each earlier stage does not know the mind of one at each later and superior stage.

Where it is easily mistaken: supposing this to say that “one at a lower stage cannot see one at a higher at all.” The treatise sets the limit very precisely: what is not known is that mind, not that person. Where this lands: this has a real bearing on practice — of the state of one at a high stage, what an ordinary being or a noble one at a lower stage can know is exceedingly limited; this is not modesty of speech but a limit set in the nature of the thing.

〔4.3〕Why are past and future minds not known? The treatise says it takes “only the apprehending minds and the rest within another’s continuum in the present for its object.” The three qualifiers are every one of them at work: present, another’s continuum, apprehending mind. The knowledge of others’ minds operates only upon the threefold intersection of the present moment, another person, and mental dharmas.

〔4.4〕The classes of dharma and subsequent do not know each other. The treatise says the knowledge of others’ minds included in dharma knowledge does not know the subsequent class, and that included in subsequent knowledge does not know the dharma class, the reason being “since dharma knowledge and subsequent knowledge have for their object the whole of the desire realm and of the upper realms respectively as counteragent” — each one’s range of counteraction is whole and non-overlapping, and so they have no bearing upon each other. This carries the earlier line, that dharma and subsequent are divided by their objects, straight through to the knowledge of others’ minds.

〔4.5〕How far the path of seeing can be known. The treatise here gives a very fine ledger: the knowledge of others’ minds is absent within the path of seeing, since the path of seeing “contemplates the principle of the truths as a whole and turns exceedingly fast”; but the mind of the path of seeing may be taken as object by another’s knowledge of others’ minds. Hence a śrāvaka cultivates the preparatory practice beforehand, and when the practice for the portion of dharma knowledge is complete he knows the mind of the first two thought-instants of the path of seeing; if he seeks further to know the mind of the portion of subsequent knowledge, he cultivates a further preparatory practice, and by the time it is complete the other has already reached the sixteenth mind — and though he then knows this mind, it no longer counts as knowing the path of seeing. The rhinoceros-like pratyekabuddha likewise knows the first two thought-instants, and cultivating further knows the eighth mind, that of the subsequent knowledge of origination, “because this is by a preparatory practice of the lower grade alone”; and the treatise records also a divergent account, “some say: he knows the first two and the fifteenth mind.” As for the World-Honored One, “without any preparatory practice he can know the whole of another’s path of seeing.”

Where it is easily mistaken: reading this passage as a ranking of the three vehicles by superiority. What it in fact states is a matter of method: the depth of knowing depends upon the grade and the speed of the preparatory practice, and a buddha needs no preparatory practice at all. Where this lands: wherever this treatise sets the three vehicles side by side and lists numbers, look first at which capacity is being compared, and do not read it loosely as high and low.

V. The Knowledge of Exhaustion and the Knowledge of Non-Arising Differ by One Sentence

〔5.1〕First ask: where does the difference between the two lie? The treatise cites directly the words of the root treatise (the Treatise on the Development of Wisdom): the knowledge of exhaustion is to know of oneself “I have known suffering, I have severed origination, I have realized cessation, I have cultivated the path”; the knowledge of non-arising is to know of oneself “I have known suffering, and need not know it again,” down to “I have cultivated the path, and need not cultivate it again.” The difference lies only in the sentence appended afterward — the knowledge of exhaustion says “it is done,” the knowledge of non-arising says “it need not be done again.”

Where it is easily mistaken: supposing the knowledge of non-arising to be a new contemplation set up after the knowledge of exhaustion. What they contemplate is one and the same thing; they differ in what they settle about the future: the one reckons up the past, the other seals off what is to come. Where this lands: the two confirmations of realization — that the thing is finished, and that this thing will never come round again — are two distinct layers.

〔5.2〕How can a knowledge without outflows give rise to a knowing such as “I have already known”? This is an objection the treatise poses to itself, and the answer has two houses. The treatise masters of Kaśmīra say: that knowing is made within the subsequently obtained knowledge after emerging from contemplation, and so there is no fault; and by the difference between the two subsequently obtained knowledges the difference between the two knowledges within the prior contemplation is displayed in reverse. There is another account: the knowledge without outflows itself also makes such a knowing, and the sūtra and treatise speaking of “seeing” is merely a convenience of the run of the words, or else because it illumines and so turns with respect to the principle of the truths; and it cites the root treatise’s “and the knowledges too are view” as its warrant. Both accounts are recorded.

Where this lands: this objection is in fact one instance of a general rule — how can a knowing that carries the word “I” stand together with the contemplation of no-self that is without outflows? The treatise’s handling is not to dissolve the contradiction but to divide the stages (during right contemplation / upon emerging from it).

〔5.3〕The ledger of the mutual inclusion of the ten knowledges. The treatise thereupon sets out a table of mutual inclusion: conventional knowledge includes one in whole and one in part; dharma and subsequent each include one in whole and seven in part; suffering, origination, and cessation each include one in whole and four in part; the knowledge of the path includes one in whole and five in part; the knowledge of others’ minds includes one in whole and four in part; exhaustion and non-arising each include one in whole and six in part. The use of this ledger is to prevent the ten knowledges from being read as ten mutually exclusive cells — the ten names overlap one another, and the entities fall short of ten.

VI. Seven Reasons for Opening Two into Ten

〔6.1〕First ask: since the ten knowledges overlap one another, why establish ten at all? The treatise answers with seven grounds: first, by own nature conventional knowledge is established (having a knowledge that is not of the ultimate for its own nature); second, by counteragent dharma knowledge and subsequent knowledge are established (able to counteract the desire realm and the upper realms in whole); third, by mode of apprehension the knowledges of suffering and of origination are established (the objects of these two having no difference in entity, so that what is divided lies only in the mode of apprehension); fourth, by mode of apprehension together with object the knowledges of cessation and of the path are established (mode of apprehension and object both differing); fifth, by preparatory practice the knowledge of others’ minds is established (without it one does not know another’s mind and mental factors); sixth, by the task being done the knowledge of exhaustion is established (arising first in the body in which the task is done); seventh, by the cause being complete the knowledge of non-arising is established (arising with the whole noble path as its cause).

〔6.2〕The ordering of these seven has its point. Note the contrast between the third and the fourth: the objects of the knowledges of suffering and of origination are alike with outflows and have no difference in entity, and so they can be divided only by mode of apprehension; whereas the knowledges of cessation and of the path differ in mode of apprehension and in object alike, and so they require a ground of their own. In one and the same set of dharmas the reasons for separate establishment need not be of one order — the treatise refuses to cover ten with a single reason, and this is exactly the strictness of Abhidharma in the establishing of names.

Where it is easily mistaken: reading the seven grounds as seven coordinate criteria of classification. They are in fact seven several adjudications, each aimed at one knowledge (or one pair). Where this lands: hereafter, meeting the formula “by such a ground such a knowledge is established,” read it by pairing item with item, and do not take it for a single uniform formula.

〔6.3〕The item on the knowledge of others’ minds is particularly worth looking at. The treatise says: originally the preparatory practice is cultivated in order to know another’s mind, and when it is complete the mental factors are known along with it, “yet the name of the knowledge of others’ minds is established with reference to the preparatory practice.” The name stays with the original motive and does not change with the function afterward enlarged. This is the same manner of naming as conventional knowledge’s being named from the greater part, “taking for the most part conventional objects such as a pot.”

VII. Why the Dharma-Knowledges of Cessation and of the Path Can Counteract the Upper Realms as Well

〔7.1〕First ask: does dharma knowledge not counteract the desire realm only? The treatise here opens an exception: the dharma knowledge of cessation and the dharma knowledge of the path included in the path of cultivation “can counteract as well what is abandoned through cultivation in the upper realms.” Two sentences give the reason: “since the cessation and the path of the desire realm are superior to those of the upper realms,” and “having removed one’s own enemy one can remove another’s as well.”

〔7.2〕Why are the cessation and the path of the desire realm on the contrary superior? Here one stumbles easily. The common intuition is that the upper realms are superior to the lower; but the truth of cessation and the truth of the path are not dharmas belonging to a realm and level, and what the cessation and the path of the desire realm sever and counteract is coarser and heavier and their force more evident, and so they are called superior. The treatise closes with a very homely reason — only when one’s own enemy has been removed is there strength to spare for removing another’s.

〔7.3〕Why does it not work the other way? The treatise says “hence subsequent knowledge has nothing that can counteract the desire realm.” The cessation and the path of the upper realms are not superior to those of the desire realm, and so subsequent knowledge cannot counteract the lower realm in return. This item is asymmetrical — counteraction can run only from the superior to the inferior and not the reverse. Where this lands: wherever this treatise speaks of counteracting, ask first which is superior and which inferior; once superior and inferior are fixed, the direction is fixed.

VIII. How the Modes of Apprehension Are Distributed among the Ten Knowledges

〔8.1〕First ask: what is a mode of apprehension? This section gives first the distribution, while the entity is not explained at length until Root 13. The treatise says dharma knowledge and subsequent knowledge each possess the sixteen modes of apprehension, the impermanent, suffering, and the rest; conventional knowledge “has these and others besides,” since it can take the individual and common marks of all dharmas as its object; and the knowledges of the four truths each have four modes of apprehension taking their own truth as object.

〔8.2〕The two portions of the knowledge of others’ minds. The knowledge of others’ minds without outflows has only the four modes of apprehension taking the truth of the path as object, since it belongs from the first within the knowledge of the path; the one with outflows takes as its object the individual mark of the minds and mental factors it takes as object, and “as the object is an individual mark, so also is the mode of apprehension,” and so it is not within the sixteen. This sentence is the key to the section: the object is an individual mark, and the mode of apprehension follows suit — mode of apprehension and object advance and retreat together.

〔8.3〕A single thought takes only one entity as object. The treatise says these two knowledges of others’ minds at any time take in a single thought only one entity as object: taking a mind as object they do not take mental factors, taking feeling as object they do not take perception and the rest. This item immediately draws forth the long exchange that follows.

〔8.4〕The fourteen of exhaustion and non-arising. The treatise says exhaustion and non-arising, setting aside emptiness and non-self, each possess the remaining fourteen modes of apprehension; the reason is that these two knowledges, “though included in the ultimate, touch upon the conventional,” and upon emerging from contemplation one must say “my births are exhausted, the holy life is established, what was to be done is done, I shall not undergo further existence” — and since “I” must be said, they cannot run together with the two modes of apprehension, emptiness and non-self. This is the neatest piece of dovetailing in the whole fascicle: a single sentence of self-declaration turns back and delimits the modes of apprehension of a knowledge.

IX. The Exchange over the Mind with Greed — How a Technical Term Is Forced into a Definition

〔9.1〕First ask: if a single thought takes only one entity as object, what of the sūtra’s “he knows as it truly is a mind with greed”? The treatise answers with great dispatch: that is not to take the two, greed and mind, at once, “as one does not at the same time take a garment and the dirt upon it.” There is dirt upon the garment, but in looking at the garment one is not at the same time looking at the dirt.

〔9.2〕The two senses of a mind with greed. The treatise divides them into two layers: first, associated with greed; second, bound by greed. A mind associated with greed possesses both senses in full, while the remaining minds with outflows are only bound by greed. There is another account, holding that only the first sense is taken.

〔9.3〕How the author forces the conclusion out. This passage is the typical move of the treatise and is worth following step by step. If it be said that “whatever is not associated with greed is a mind free of greed,” then a mind associated with the remaining afflictions ought also to be called free of greed; press further, and there will appear a mind that is neither the counteragent of greed nor of defiled nature, and one would then have to grant that it is neither a mind with greed nor a mind free of greed — press a technical term until it can be hung upon neither side, and that definition is declared a failure. And so the treatise says, “therefore one ought to grant the account of the other master, that what is bound by greed is called a mind with greed.”

Where it is easily mistaken: supposing the case here settled. It is not — further on the author himself turns back and overturns the sense of “bound by greed” (see section XI). This treatise commonly adopts one account first in order to establish an argument, and then in a later stretch of the text refutes it itself; in reading, one must read to the end.

X. The Gathered and the Scattered, the Sunken and the Restless, the Small and the Great — the Glosses on Eight Pairs of Minds

〔10.1〕First ask: who fixed the names of these eight pairs of minds? The treatise records two houses. The Vaibhāṣika masters say: the gathered mind is the wholesome mind, since with respect to its object it does not run off and scatter; the scattered mind is the defiled mind, since it is associated with agitation. The masters of the West say: that which is associated with sleep is called the gathered mind, and the remaining defiled minds are called the scattered mind.

The treatise thereupon adjudges the latter account “not in accord with principle,” on two grounds: first, that on that telling a defiled mind, if associated with sleep, would extend at once to both gathered and scattered (gathered because of the sleep, scattered because of the defilement); second, that it would do violence to what the root treatise says, “to know as it truly is the gathered mind is possessed in full of four knowledges.”

〔10.2〕Who is doing the refuting? This question must be asked clearly. Puguang here marks it “the Vaibhāṣika masters refute,” and explains the identity of the masters of the West thus: “the masters of the West are the masters of the country of Gandhāra; the Mahāvibhāṣā calls them the masters abroad”; Fabao marks it “the Sarvāstivādins refute the explanation of the masters of the West; first refuting by principle, second refuting by contradiction of the root treatise.” ⇒ Both commentaries assign this refutation to the Vaibhāṣika masters (the Sarvāstivādins), and Fabao further divides the refutation into the two layers of refuting by principle and refuting by contradiction of the root treatise. This agrees with the run of the treatise’s own text — this passage is at bottom a contention between two houses within the Sarvāstivāda, which the author records.

〔10.3〕The like explanation of the remaining seven pairs. The sunken mind and the roused mind are matched with indolence and right diligence; the small mind and the great mind with what men of little and of much pure quality delight to practice. The treatise opens a further road as well, speaking of small and great from five aspects, root, price, retinue, concomitant turning, and force: the defiled mind has few roots (associated with two at most), the wholesome many (constantly associated with three); the defiled is of low price (coming to be without effort), the wholesome of high price (coming to be by a great store of provisions); the defiled has a small retinue (having no future cultivation), the wholesome a large one (having future cultivation); the defiled has little concomitant turning (only three aggregates), the wholesome much (extending to four aggregates); the defiled is of small force (the wholesome roots it severs will certainly be joined on again), the wholesome of great force (receptivity certainly severs the latent tendencies forever). Restless and unrestless, stilled and unstilled, concentrated and unconcentrated, cultivated and uncultivated, liberated and unliberated are to be inferred in the same way.

Where this lands: of these five aspects the most worth savoring is the one on “retinue” — the wholesome mind has a large retinue because it has future cultivation; when one wholesome mind becomes present, what it sets in motion is a whole string of the future. The full three thousand characters on “the cultivation of the knowledges at the several stages” in the latter half of this fascicle are the detailed ledger of just this sentence.

XI. The Author Refutes Himself: Those Eight Names Are Still Not Made Clear

〔11.1〕First ask: since each pair has been matched up, what is still lacking? The treatise says, “an explanation such as this neither accords with the sūtra nor is able to make out the several senses of the phrases” — two counts of the charge. Not according with the sūtra: the sūtra says that being inwardly gathered is “when the mind runs together with torpor and sleep, or when what is inwardly associated has calm abiding without insight,” and being outwardly scattered is “when the mind roams among the objects of the five excellent desires, scattering along and streaming along, or when what is inwardly associated has insight without calm abiding”; the sūtra has its own gloss, differing from the glosses of both foregoing houses.

〔11.2〕One round at close quarters. Objection: was it not said above that a defiled mind together with sleep would incur the fault of one mind extending to both gathered and scattered? Answer: it was said, but it does not accord with principle, since the sūtra does not grant that a defiled mind together with sleep is the scattered mind. Objection: was it not also said that this contradicts the root treatise? Answer — and this sentence shows the sense of proportion best — “better to contradict the words of a treatise than to contradict what a sūtra says.” When sūtra and treatise conflict, take the sūtra. This is a general rule of the whole treatise, said here to the full in six characters.

〔11.3〕What it means to be unable to make out the several senses. The treatise says: on that explanation one cannot make out the differing marks of each of the eight phrases, the scattered and the rest and the gathered and the rest. On the author’s own explanation, though they are alike defiled minds, the differences among their faults are displayed; and though they are alike wholesome minds, the differences among their merits are displayed, “and so, following eight senses, eight names are separately established.” Eight names are needed because eight differences are to be displayed; if all were gathered under the two heads of defiled and wholesome, the eight names would be so many superfluous names.

〔11.4〕An objection from the sunken and the restless. The treatise raises one more objection: if the sunken mind were the restless mind, the sūtra ought not to say that cultivating pliancy, concentration, and equanimity when the mind is sunken is untimely cultivation, and cultivating discernment of dharmas, diligence, and joy when the mind is restless is untimely cultivation. The other side saves itself with “this refers to cultivation in the sense of attending to it and wishing to cultivate it, not cultivating it as present”; and pleads for itself further, “as regards their being constantly associated I say their entity is one; this is speaking after my own sense, and who can bar it?” The author’s last sentence does not argue much: “nevertheless the sense of this sūtra is truly not so.”

Where this lands: what is worth watching in this whole round is not who wins but the ranking of the criteria — wherever two houses contend and the sūtra has an express text, the sūtra text is the final court; only where the sūtra has no express text and a treatise alone does, does the turn of reasoned proof come round.

〔11.5〕Turning back to refute “bound by greed.” Only now does the treatise gather in the thread laid down in section IX. If “bound by greed” means that the acquisition of greed follows along with it, then the mind without outflows of one in training must also be called a mind with greed; if it means being taken as object by greed, then the mind with outflows of one beyond learning must also be so called; if it be not granted that that mind can be taken as object by greed, then on what does that mind come to be with outflows; and if it be said to be taken as object by an affliction of the common mark, then it ought rather to be called a mind with delusion. And the most telling stroke: the knowledge of others’ minds does not take as object the acquisition of greed, nor can it be said to take as object “the greed that takes the mind as object” — on what, then, does the knowledge of others’ minds know that another has a mind with greed? Hence, “it is not by being bound by greed that a mind is called a mind with greed.”

Thereupon the author fixes it himself: “being associated with greed it is called a mind with greed; being unassociated with greed it is called free of greed and the rest.” Immediately he explains through an objection from a sūtra: the sūtra says a mind free of greed, hatred, and delusion no longer falls into the three existences, and he answers “this is said with reference to the acquisition of freedom”; and he explains through the earlier objection as well: if a mind associated with the remaining afflictions be granted in this sense to be called free of greed, there is no contradiction, only it is not said to be so, since these are to be assigned to the side of having hatred and having delusion.

〔11.6〕Closing the incidental discussion. The treatise closes with six characters, “let the incidental discussion rest, and let us state our own position,” and returns to the marks of the knowledge of others’ minds: it does not take the object of another’s mind, nor the apprehending mode of apprehension of another’s mind; it knows only that he has a defiled mind and the rest, and does not know the form to which his mind is attached, nor the mode of apprehension by which he apprehends. Were it otherwise, the knowledge of others’ minds would have to take form as object, and there would further be the fault of its taking itself as object. At the last it gives a list of fixed marks — it takes only the individual mark of one real dharma among minds and mental factors that are affiliated to the desire realm or the form realm or unaffiliated, within another’s continuum, in the present, of the same class; it is not associated with emptiness and signlessness, is not included in exhaustion and non-arising, and is not within the path of seeing or an uninterrupted path; and the remainder, not being barred, is admitted as is fitting.

XII. Whether Modes of Apprehension without Outflows Can Exceed the Sixteen

〔12.1〕First ask: what is at issue? The treatise masters of the country of Kaśmīra say: there is no mode of apprehension without outflows that exceeds these sixteen. The masters abroad say: there are further modes of apprehension without outflows that exceed the sixteen.

〔12.2〕On what do the masters abroad rely? On one sentence of the root treatise (the Treatise on the Development of Wisdom). The root treatise asks, “can there be an unaffiliated mind that discerns dharmas affiliated to the desire realm?”, answers that it can discern them, and lists the eight modes of apprehension of the impermanent, suffering, emptiness, non-self, cause, origination, arising, and condition, appending at the end the phrase “there being such a case and there being such a fact, it discerns them as drawn forth in accordance with principle.” The masters abroad accordingly take these two closing phrases to be modes of apprehension existing apart from the sixteen.

〔12.3〕How does the author check it? The author does not contend on grounds of principle but checks against the usage of the text. If that passage were meant only to display that “making the eight modes of apprehension is a case that exists and a fact that exists,” then “in the remaining places too this phrase ought to be said” — yet the same treatise, where it speaks of a mind abandoned through seeing discerning dharmas affiliated to the desire realm, lists twenty-one items, self, what belongs to self, annihilation, permanence, down to greed, hatred, pride, and delusion, and does not append the phrases “there is such a case” and the rest. Where one and the same treatise has a formula of the same type here and lacks it there, it is evident that these two phrases are not names of modes of apprehension. And so the treatise judges, “such an explanation is not so, since the remaining places do not say it,” and “since this phrase is absent, the explanation does not accord with principle.”

Where it is easily mistaken: reading this passage as the author adjudging the masters abroad wholly wrong. What the treatise denies is only this one textual proof, not a declaration that there is absolutely no further mode outside the sixteen; while what the stanza establishes with “the pure do not exceed the sixteen” is the tenet of Kaśmīra. The statements of both parties are recorded together, and the choice between them belongs to the adjudication of doctrine.

Where this lands: this passage gives a method of great use — before citing a sentence as proof, look whether the sentences of the same type in the same book say the same thing. A solitary instance proves nothing.

XIII. The Names and the Entities of the Sixteen Modes of Apprehension

〔13.1〕First ask: how many entities lie beneath the sixteen names? The treatise first records a divergent account: some other master says the names are sixteen but the real entities only seven — of those taking the truth of suffering as object, name and entity are four alike, and of those taking the remaining three truths as object, the names are four apiece but the entities one apiece, four plus three making seven. Then it gives the correct account: “one who speaks as we do holds that in real entity they are sixteen as well.”

〔13.2〕How is it proved that the entities are in fact sixteen? The treatise makes no abstract argument; it turns instead to glossing mode by mode — giving three rounds of glosses in one breath. The first round: because it exists awaiting conditions, impermanent; because it has the nature of oppression, suffering; because it runs counter to the view of what belongs to self, empty; because it runs counter to the view of self, non-self; by the principle of a seed, cause; by the principle of an equal manifesting, origination; by the principle of a continuum, arising; by the principle of bringing to completion, condition (with the simile of the lump of clay, the wheel, the cord, and water coming together to make a pot); because all the aggregates are exhausted, cessation; because the three fires have come to rest, stilled; because there are no manifold troubles, excellent; because one departs from the manifold calamities, separation; by the sense of a way through, the path; because it accords with correct principle, conformity; because it goes rightly forward, progress; because it can transcend forever, issuing forth.

The second round exchanges another set of reasons: because it is not final, impermanent; because it is like bearing a heavy burden, suffering; because inwardly it is apart from an agent or person, empty; because it is not under one’s own mastery, non-self; by the sense of drawing forth, cause; by the sense of coming forth, origination; by the sense of nourishing and bringing to birth, arising; by the sense of acting as a support, condition; because it is not joined on again and the continuum is severed, cessation; because it is apart from the three marks of the conditioned, stilled; because it is the wholesome of the ultimate, excellent; because it is utterly secure, separation; because it counteracts the wrong path, the path; because it counteracts what is not in accordance, conformity; because it enters upon the palace of nirvāṇa, progress; because it relinquishes all the existences, issuing forth.

For the third round the author says of himself, “since the glosses of old are thus not of one gate only, let a further explanation be made as each may please,” and gives yet another set.

〔13.3〕The three rounds of glossing are themselves the argument. This is to be seen clearly: if there were only seven entities beneath the sixteen names, it would be impossible to give three sets of glosses each internally coherent and none repeating the others. That one can gloss them three times over, one name to one sense each time, is precisely what shows that each of the sixteen names has its own entity. The treatise does not refute by declaration; it refutes by density.

Where it is easily mistaken: taking the three rounds of glosses for a heap of synonymous repetition and skipping over them. On the contrary, they are the evidence of the section itself.

〔13.4〕The four desires and four similes of the four modes of the truth of origination. At the truth of origination the treatise opens a further passage: since the stages differ, the four desires differ — grasping the general self of the present and arousing the general desire for one’s own being; grasping the general self of the future and arousing the general desire for further existence; grasping a particular self of the future and arousing the particular desire for further existence; grasping the self joined on to birth and arousing the desire at the time of that joining (or grasping the self that does actions and arousing the desire at the time of doing them). The four stand to suffering, in order, as initial cause, equal gathering, particular condition, and near arising, matched with the four similes of the seed, the sprout, the field, and the blossom. The simile of the field the treatise states most finely — by the force of field, water, dung, and the rest, the flavor, potency, ripeness, and quality of the fruit each arise severally.

Following the sūtra’s modes of craving, the treatise sets out eighteen more: five differences in grasping the general self of the present, five in grasping the general self of the future, four in grasping a particular self of the future, four in grasping the self joined on to birth.

Where this lands: this passage restores that abstract truth, “origination,” to four concrete tenses of the grasping of self — the present self, the future self, the future self in another form, the self now being joined on to birth. The cause of suffering is not an abstract desire; it is the grasping of self in four tenses.

〔13.5〕Modes of apprehension have discernment for their entity, and the breadth and narrowness of the three gates. The treatise says a mode of apprehension has discernment for its entity. Objection: then discernment itself has no mode of apprehension, since discernment is not associated with discernment. The treatise answers: it ought to be said that every difference of class in the taking of an object by mind and mental factors is called a mode of apprehension; discernment and the remaining minds and mental factors all have an object and so are all that which apprehends; and all existing dharmas are that which is apprehended. Hence the three gates differ in breadth and narrowness — discernment extends to all three, mode of apprehension, that which apprehends, and that which is apprehended; the remaining minds and mental factors extend only to that which apprehends and that which is apprehended; and the remaining existing dharmas are that which is apprehended alone.

Where it is easily mistaken: reading mode of apprehension, that which apprehends, and that which is apprehended as three coordinate classes. They are three circles of differing compass, and one and the same discernment falls within all three at once, while an ordinary existing dharma falls only within the outermost. Where this lands: this is the treatise’s standard model for handling the threefold division of entity, function, and object, and may be applied directly wherever a like threefold division is met with hereafter.

XIV. Nature, Level, and Body — Three Gates Spoken of Together

〔14.1〕First ask: how are the three natures divided? The treatise says conventional knowledge extends to the three natures, wholesome, unwholesome, and indeterminate, while the remaining nine knowledges are wholesome alone. The reason need not be stated expressly — the nine knowledges are all included in the noble path or have the noble path for their cause, and conventional knowledge alone extends to the ordinary.

〔14.2〕The stairway of levels depended upon. Conventional knowledge extends in its dependence from the desire realm up to the summit of existence (all nine levels); the knowledge of others’ minds depends only upon the four fundamental meditative absorptions; dharma knowledge depends upon these four together with access concentration and the intermediate absorption, six levels in all; the rest depend upon these six together with the three lower formless levels, nine in all.

Where it is easily mistaken: reading the “only four” of the knowledge of others’ minds as its being the weakest. In fact this makes it the most particular about its level — it can arise only upon a fundamental meditative absorption, and neither access concentration nor the intermediate absorption will do. More limits do not necessarily mean inferior; often they mean that the conditions it depends upon are the finer.

〔14.3〕The body depended upon. The knowledge of others’ minds can become present in dependence upon a body of the desire realm or of the form realm alike; dharma knowledge arises only in dependence upon a body of the desire realm; the remaining eight knowledges extend in their dependence to bodies of the three realms. Dharma knowledge depends only upon a body of the desire realm because it takes the four truths of the desire realm as its object, and the place where it becomes present must be the desire realm. Where this lands: in reading this treatise, “level” and “body” must be told apart — the level is the level of absorption depended upon, the body is the place of birth in which one stands when it becomes present, and the two frequently do not coincide.

XV. Inclusion in the Foundations of Mindfulness — a Ledger in Three Sentences

The treatise says the knowledge of cessation is comprised within the foundation of mindfulness of dharmas, the knowledge of others’ minds is included in the latter three foundations of mindfulness, and the remaining eight knowledges extend to the four.

First ask: why is the knowledge of cessation only in the foundation of mindfulness of dharmas? Because it takes cessation through discernment alone as its object, and cessation through discernment is an unconditioned dharma, which only the foundation of mindfulness of dharmas can include. Why is the knowledge of others’ minds only in the latter three? Because what it takes as object is another’s mind and mental factors, and it does not take form as object, so that the foundation of mindfulness of the body is excluded.

Where it is easily mistaken: supposing the foundations of mindfulness to be four gates of practice with no bearing on the classification of the knowledges. What is used here is the “inclusion of objects” side of the foundations of mindfulness — one and the same set of four foundations of mindfulness can serve both as gates of practice and as four cells for classifying objects; this section uses the latter. Where this lands: when a familiar numerical category turns up in an unfamiliar context, ask first which of its sides is being used here.

XVI. How Many Knowledges Each of the Ten Takes as Object

The ledger the treatise gives is this: dharma knowledge takes nine as object (setting aside subsequent knowledge), subsequent knowledge takes nine (setting aside dharma knowledge), the knowledge of the path takes nine (setting aside conventional knowledge, since it is not included in the path); the knowledges of suffering and of origination take two apiece (conventional and others’ minds); conventional, others’ minds, exhaustion, and non-arising take ten apiece; the knowledge of cessation takes no knowledge as object, taking cessation through discernment alone as its object.

First ask: why does dharma knowledge not take subsequent knowledge as object? This carries the earlier line, that dharma and subsequent do not know each other, through to the taking of objects — their whole counteraction of realm and level does not overlap. Why does the knowledge of the path set aside conventional knowledge? Because conventional knowledge is not included in the truth of the path, while the knowledge of the path takes only the truth of the path as object. Why do the knowledges of suffering and of origination take only two? Because they take only what is with outflows as object, and among the ten knowledges only conventional knowledge and the knowledge of others’ minds with outflows are with outflows.

Where it is easily mistaken: memorizing this string of numbers by rote. In fact every one of the numbers can be derived from the three lines of the truth taken as object, the realm, and with or without outflows. Where this lands: meeting a table of numbers in this treatise, look first for its line of derivation and the numbers will remember themselves; if you cannot remember them, it is the line you have not found.

XVII. The Objects of the Ten Knowledges, Ten Dharmas in All

〔17.1〕First ask: how are the ten dharmas counted out? The treatise says the conditioned dharmas divide into eight — the four classes of what is affiliated to the three realms and the conditioned without outflows, each divided into the associated and the unassociated, four times two being eight; and the unconditioned divides into two, the wholesome unconditioned and the indeterminate unconditioned. Together they make ten.

〔17.2〕The object of each knowledge. Conventional knowledge takes ten; dharma knowledge takes five (the two of the desire realm, the two of the outflowless path, and the wholesome unconditioned); subsequent knowledge takes seven (the six of the form and formless realms and of the outflowless path, plus the wholesome unconditioned); the knowledges of suffering and of origination take six apiece, those affiliated to the three realms; the knowledge of cessation takes one (the wholesome unconditioned); the knowledge of the path takes two (the outflowless path); the knowledge of others’ minds takes three (the associated dharmas of the desire realm, the form realm, and the outflowless); exhaustion and non-arising take nine apiece (the eight conditioned plus the wholesome unconditioned).

Where it is easily mistaken: supposing that because the knowledge of cessation takes only one it is the narrowest and so the poorest. The one dharma that the knowledge of cessation takes as its object is nirvāṇa itself. Fewness of object does not mean lightness of sense — this is a proportion the treatise keeps throughout.

〔17.3〕Whether a single thought can take all dharmas as object. The treatise poses this question to itself and answers that it cannot. Objection: does the knowledge that contemplates non-self not know that all dharmas are non-self? Answer: that too cannot take all dharmas as object. This one question and answer leads straight into the next section.

XVIII. Why the Contemplation of Non-Self Cannot Reach Itself

〔18.1〕First ask: which piece is missing? The treatise says “still setting aside its own class” — its own class being its own entity, the dharmas associated with it, and the dharmas coexistent with it. The reason is in three sentences: object and possessor of an object are distinct; they have the same object; and they are neighboring and near — hence they are not what this knowledge takes as object.

Where it is easily mistaken: supposing this to say that the contemplation of non-self has a hole in it. It is not a hole but a structure — the portion that contemplates cannot at the same time be what is contemplated, just as a knife does not cut itself. Where this lands: this may be applied directly to every claim of “taking all dharmas as object without exception”; ask first whether it sets aside its own class.

〔18.2〕Why is it only formed through hearing and through reflection? The treatise says this knowledge is included in the desire realm and the form realm alone and is formed through hearing and through reflection, not through cultivation; the reason is “because discernment formed through cultivation takes a single level severally as its object, and were it otherwise one would become free of defilement all at once” — discernment formed through cultivation takes one level severally as its object, and if this knowledge, which takes everything as object in general, were formed through cultivation, then one would at a single stroke become free of the defilement of all the levels, which plainly is not so.

Where this lands: this is a very handsome form of reductive proof — suppose it were formed through cultivation, derive a result that nobody will grant, and so adjudge that it is not. This adjudication later became one of the disputes within the lineage; see the section of Further Resonances.

XIX. How Many Knowledges Are Accomplished, from the Ordinary Being to One beyond Learning

The treatise’s ledger: at the stage of the ordinary being and at the first thought-instant of the path of seeing, one knowledge is definitely accomplished (conventional); at the second thought-instant three (with dharma and suffering added); at the fourth, sixth, tenth, and fourteenth thought-instants, subsequent, origination, cessation, and the path are added in order; at stages where no addition has been reached the number is as before, and so at the stage of the path of cultivation seven are likewise definitely accomplished; at any of these stages, if one is already free of desire one adds one (others’ minds), excepting only the ordinary being born in the formless realm; one whose liberation depends on circumstance definitely accomplishes nine (with the knowledge of exhaustion added), and one whose liberation does not depend on circumstance definitely accomplishes ten (with non-arising added).

First ask: why is the addition made at just the second, fourth, sixth, tenth, and fourteenth thought-instants? For this one must go back and check against the table of the sixteen minds of the path of seeing in the Chapter on the Worthy and the Noble — the place where the dharma knowledge and the subsequent knowledge of each truth first arise is the place where a new knowledge is first obtained. This section is the knowledge-side portrait of the table of the sixteen minds in the Chapter on the Worthy and the Noble, and the two tables must be read together.

Where it is easily mistaken: reading “definitely accomplishes” as “becomes present.” To accomplish is to have obtained and not lost; to become present is to be using it at the moment. That an ordinary being definitely accomplishes conventional knowledge does not mean that he is using it at every moment. Where this lands: wherever this treatise speaks of accomplishing, it is always with reference to “obtaining,” which is a different thing from actual occurrence.

XX. Cultivation at the Stage of the Path of Seeing — the Conventional Knowledge at the Border of Direct Realization

〔20.1〕First ask: what is cultivated in the path of seeing? The treatise says that at the stage of the path of seeing, whatever receptivity or knowledge arises, one cultivates in the future what is of that same class, and cultivates in full the modes of apprehension and foundations of mindfulness of that same truth. Why only what is of the same class? Two reasons: “because this class-nature has never been obtained before,” and “because its counteragent and its object are both fixed.”

〔20.2〕What the conventional knowledge at the border of direct realization is. Only at the three subsequent knowledges of suffering, origination, and cessation can one cultivate in addition the future conventional knowledge at the border of direct realization, “since it can be cultivated in addition only at the hindmost border of the direct realization of each single truth, this name is established for it” — the name is its position: at the hindmost border of the direct realization of one truth.

〔20.3〕Why is it not cultivated at the subsequent knowledge of the path? The treatise gives two reasons: conventional knowledge has never performed direct realization in terms of entities with respect to the path; and there certainly is no one who can perform thorough direct realization in terms of entities with respect to the path — suffering, origination, and cessation admit of thorough knowing, thorough severing, and thorough realizing, while the path cannot be thoroughly cultivated, since the class-natures of the path are too many. The treatise records also a divergent account (that this is a retinue of the path of seeing while being included in the path of cultivation, and so cannot be cultivated) and does not adopt it, judging, “the principle is not fully established and should not be taken as proof.”

〔20.4〕The most precarious step: it is a dharma that will not arise. The treatise says this conventional knowledge “is a dharma that will not arise, since at no time is there room for it to become present.” The objection follows at once: since it does not arise, how is it still said to be cultivated? Answer: because it was not obtained before and is only now obtained. Objection again: since it cannot arise, upon what does the sense of acquisition rest? — an exceedingly sharp objection, since “acquisition” is ordinarily tied to being able to arise. The treatise cites the ancient master’s explanation in answer: by the force of the noble path this conventional knowledge is cultivated, and after emerging from contemplation a superior conventional knowledge taking the truths as its object becomes present; because the support of its arising has been obtained, it is called obtaining it, “as obtaining a gold ore is called obtaining gold.”

This simile is the pivot of the section: to obtain the ore is not to have gold in hand, but the ore is that upon which the gold depends, and so to speak of obtaining gold is not empty. In the same way, though the conventional knowledge cultivated in the path of seeing does not become present, the support of its becoming present has been obtained.

〔20.5〕The Vaibhāṣika masters do not adopt this sense. The treatise thereupon records their divergent adoption: whichever level the path of seeing that becomes present depends upon, one can cultivate in the future what belongs to that level and to the lower levels — when the path of seeing dependent upon access concentration becomes present, one can cultivate in the future the path of seeing of one level and the conventional knowledge of two levels; and so on to the path of seeing dependent upon the fourth meditative absorption becoming present, whereupon one can cultivate in the future the path of seeing of six levels and the conventional knowledge of seven levels. Both accounts are recorded.

〔20.6〕The foundations of mindfulness and the own nature of this conventional knowledge. What is cultivated at the border of suffering and at the border of origination extends to the four foundations of mindfulness, while at the border of cessation it is the foundation of mindfulness of dharmas alone; at whichever truth’s border of direct realization it is cultivated, it takes that truth as its object with the modes of apprehension of that truth, since it is obtained by the force of the path of seeing. It is obtained through preparatory practice alone, and because knowledge prevails in it the name “knowledge” is established for it; spoken of together with its retinue, it has the four aggregates of the desire realm and the five aggregates of the form realm for its own nature.

XXI. Cultivation at the Stage of the Path of Cultivation and at the Stage beyond Learning

〔21.1〕The first thought of the path of cultivation. At the moment of the sixteenth mind, the subsequent knowledge of the path, two knowledges are cultivated in the present; one not yet free of desire cultivates six in the future (dharma, subsequent, and the four truths), and one free of desire cultivates seven (with others’ minds added, conventional knowledge not being cultivated, since it is the counteragent of the summit of existence).

〔21.2〕The difference between severing the desire realm and severing the upper levels. The treatise lists one by one what the uninterrupted path, the path of liberation, the preparatory path, and the path of advance cultivate in the present and in the future in severing the nine grades of the desire realm and in severing the seven levels above, the totals being seven or eight. The essential point lies in two dividing lines: conventional knowledge is not cultivated where the summit of existence is counteracted (what counteracts the summit of existence may not be with outflows), and the knowledge of others’ minds is not cultivated in a path able to sever obstruction (a path that severs obstruction does not admit others’ minds). These two recur again and again and are the bones of this whole ledger.

〔21.3〕The first thought beyond learning. At the ninth path of liberation severing the summit of existence, the subsequent knowledges of suffering and of origination and the knowledge of exhaustion are cultivated in the present as is fitting (since the object is the summit of existence); in the future nine or ten are cultivated as is fitting — one of dull faculty setting aside only non-arising, one of sharp faculty cultivating non-arising as well. So too for the path of advance.

Where it is easily mistaken: reading the difference of dull and sharp faculty as a difference in the height of the fruit realized. What differs here is only the single item of whether the knowledge of non-arising is cultivated; the entity of the fruit is alike arhatship.

XXII. Refining the Faculties, Intermixed Cultivation, Cultivating the Supernormal Knowledges, and the Remaining Stages

〔22.1〕Cultivation in the refining of the faculties. In the uninterrupted paths of refining the faculties by one in training, the dharma knowledges and subsequent knowledges of the four truths are cultivated in the present, and six are cultivated in the future — conventional knowledge not being cultivated (since this resembles the path of seeing), and others’ minds not being cultivated (since it can sever obstruction). For the paths of liberation, one not yet free of desire cultivates six and one already free of desire seven (with others’ minds added); and the treatise records also a divergent account, “another master says: at the stage of the path of liberation conventional knowledge too is cultivated.” For the preparatory paths, seven if not yet free of desire and eight if already free; for the paths of advance, seven if not yet free of desire and eight if already free.

In the uninterrupted paths of refining the faculties by one beyond learning, seven are cultivated in the future (the dharma knowledges and subsequent knowledges of the four truths plus exhaustion, conventional knowledge not being cultivated, as resembling the counteracting of the summit of existence); for the first eight paths of liberation, eight; for the ninth path of liberation, nine; for the last path of liberation, ten; for the preparatory paths, nine in the future; for the paths of advance, nine for the dull and ten for the sharp.

〔22.2〕The intermixed cultivation of the meditative absorptions. For the uninterrupted paths of intermixed cultivation by one in training, seven are cultivated in the future; for the paths of liberation there are only the dharma knowledges and subsequent knowledges of the four truths, with conventional knowledge added at the preparatory stage; for the paths of advance others’ minds is further added, and all cultivate eight in the future. For the uninterrupted paths of intermixed cultivation by one beyond learning, eight are cultivated in the future for the dull and nine for the sharp; for the paths of liberation, nine in the future for the dull and ten for the sharp; and the paths of advance are the same as at the refining of the faculties.

〔22.3〕Cultivating the supernormal knowledges. For the five uninterrupted paths of one in training who cultivates the supernormal knowledges, conventional knowledge is cultivated in the present and seven in the future; for the two paths of liberation of the supernormal knowledge of former abodes and of supernormal powers and for the five preparatory paths, conventional knowledge is cultivated in the present, while for the path of liberation of the supernormal knowledge of others’ minds it is dharma, subsequent, path, conventional, and others’ minds; for all the paths of advance, the knowledges of suffering, origination, and cessation are cultivated in the present along with them as is fitting, and all cultivate eight in the future. For one beyond learning cultivating the supernormal knowledges, eight in the future for the dull and nine for the sharp, nine for the dull and ten for the sharp. The two paths of liberation of the supernormal knowledge of the divine eye and of the divine ear, being of indeterminate nature, are not called cultivation — this sentence shows the sense of proportion best: the definition of cultivation contains wholesomeness within it, and the indeterminate does not enter into cultivation.

〔22.4〕Arousing the remaining merits, and the stages of the ordinary being. When a noble one arouses merits with outflows such as the four immeasurables, one conventional knowledge alone is cultivated in the present; one in training cultivates seven in the future if not yet free of desire and eight if already free; one beyond learning cultivates nine in the future for the dull and ten for the sharp, excepting only the subtlest of subtle minds (which in the future cultivates conventional knowledge alone). When arousing merits without outflows, those included in the meditative absorptions cultivate in the present the dharma knowledges and subsequent knowledges of the four truths, and those included in the formless cultivate in the present the four subsequent knowledges alone. When an ordinary being becomes free of defilement, conventional knowledge is cultivated in the present; for the ninth path of liberation and for the path of advance and the preparatory path for becoming free of defilement that arise in dependence upon the four fundamental meditative absorptions, two are cultivated in the future (with others’ minds added), and for the rest conventional knowledge alone. In cultivating the five supernormal knowledges, for the preparatory paths and two of the paths of liberation conventional knowledge is cultivated in the present, for one path of liberation conventional knowledge and others’ minds, and for the five uninterrupted paths conventional knowledge alone in the present and in the future. In cultivating the remaining merits in dependence upon the fundamental meditative absorptions, conventional knowledge is cultivated in the present and two in the future in every case; only the factors conducive to penetration certainly do not cultivate the knowledge of others’ minds, “since they are the near retinue of the path of seeing”; and in cultivating the remaining merits in dependence upon the absorptions of the remaining levels, there is conventional knowledge alone in the present and in the future.

Where this lands: this whole ledger of three thousand characters is in fact only five or six rules combined again and again — where the summit of existence is counteracted the conventional is not cultivated; a path that severs obstruction does not cultivate others’ minds; one not yet free of desire does not obtain others’ minds; the dull of faculty leave out non-arising; the indeterminate is not called cultivation; the factors conducive to penetration do not cultivate others’ minds. Draw the rules out and the ledger turns from a list into a derivation. The closing sentence of the fascicle, “all this should be pondered as is fitting,” is precisely the demand that the reader work out for himself the cells that were not set out.

XXIII. The Levels of What Is Cultivated in the Future, and Whether “Obtaining” Is Cultivation

〔23.1〕First ask: of how many levels is what is cultivated in the future? The treatise says that when the paths depend upon this level and when this level is obtained, one can cultivate in the future what is with outflows of this level; and that when a noble one is becoming free of this level, when this level is obtained, and when the paths within this level become present, in every case he can cultivate what is without outflows of this level and of the lower levels. The phrase “becoming free of this” extends to two of the four kinds of path.

〔23.2〕The prerogative of the first knowledge of exhaustion. Only when the first knowledge of exhaustion becomes present is the force able to cultivate throughout the merits with outflows of the nine levels — the immeasurable merits, the contemplation of the impure and the rest, “can bind the manifold confusions and sever them without remainder, just as when what binds is severed the vital force of what is bound is severed through.” The treatise adds a simile: at that time his own mind is just now ascending the royal throne, and the acquisitions aroused by all wholesome dharmas come to pay court; “as when a great king ascends the throne and is consecrated by sprinkling of the head, all the lands within his borders come to render tribute.”

This is one of the few flourishes in the fascicle, and it is not mere rhetoric. It states a structural fact: at the very thought-instant in which the confusions of the summit of existence are severed to the end, the acquisitions of all the merits with outflows of the nine levels arise together at one time. Where this lands: attainment in practice is often not an item-by-item accumulation but a whole tract falling into place at once after one crucial severing.

〔23.3〕Two limits. First, “one born in a higher level certainly does not cultivate a lower one.” Second, the phrase “the first knowledge of exhaustion” refers only to becoming free of the summit of existence and to the ninth path of liberation of the five stages of refining the faculties, and is not to be used loosely.

〔23.4〕What was once obtained does not count as cultivation. The treatise defines: only what was not obtained before, is now obtained, and is obtained by effort is what is cultivated; if it was once obtained before and afterward relinquished, then though it be obtained again now it is not what is cultivated, “since it was not realized by any setting forth of toil.” It divides this into two aspects as well: what was not obtained before and is made present by effort can cultivate the future, since its force is superior; what was once obtained and arises again does not cultivate the future, since it does not arise by much effort and its force is inferior.

Where it is easily mistaken: supposing “cultivation” to mean repeated practice. The cultivation of this treatise is here far stricter — it must be a first obtaining, and an obtaining by effort; obtaining again does not count.

XXIV. The Four Kinds of Cultivation, and the Six of the Masters Abroad

〔24.1〕First ask: since cultivation is not obtaining alone, how many kinds are there? The treatise lists four: cultivation as acquisition, cultivation as practice, cultivation as counteragent, cultivation as expulsion.

〔24.2〕Upon what dharmas are the four established? Cultivation as acquisition and as practice are established upon conditioned wholesome dharmas — for the future there is cultivation as acquisition alone, for the present both acquisition and practice; cultivation as counteragent and as expulsion are established upon dharmas with outflows. Hence wholesome dharmas with outflows possess all four kinds of cultivation in full (being both conditioned wholesome and with outflows); the conditioned without outflows possesses only the former two, and the remaining dharmas with outflows only the latter two.

This passage is extremely brief and its apportioning extremely exact: two measures (conditioned wholesome / with outflows) cross to make four cells, and the four cultivations each take their place. Where this lands: meeting a passage in this treatise that apportions by the crossing of two conditions, draw a two-by-two table and the relations become clear at once.

〔24.3〕The six cultivations of the masters abroad. The masters abroad add cultivation as guarding and cultivation as contemplating upon the four — to guard and protect the several faculties is cultivation as guarding, to contemplate the body is cultivation as contemplating, and they cite a passage of sūtra for each (cultivating the faculties, cultivating the body). The treatise masters of the country of Kaśmīra answer: the two cultivations, guarding and contemplating, are comprised within cultivation as counteragent and cultivation as expulsion, and need not be separately established.

This is the close of the whole fascicle, and it answers exactly to its opening: the opening asks whether receptivity is knowledge, which is to ask whether a name ought to stand on its own; the close asks whether guarding and contemplating ought to be separately established as cultivations, which is still the same question. What one work of Abhidharma does from beginning to end is just this — whether a name deserves to stand on its own is settled by whether it has some portion that cannot be taken away by another. Both accounts are recorded, and the choice between them belongs to the adjudication of doctrine.

Further Resonances: The Mirror of the Three Commentaries

The foregoing is the sense of the treatise text of this fascicle itself. Below, a further section is set up, gathering four things: the adjudications of the three commentaries and their divergences, Saṅghabhadra’s refutations, the divergences of Paramārtha’s earlier translation, and the coordinates for parallel reading. A mirror is not the treatise; it is only a change of reading upon one and the same text. Recorded without adjudication, each returned to its own place.

1. Where the Three Commentaries Diverge

A commentary wholly absent from this chapter. Shentai’s Commentary is a fragmentary text; the seven fascicles now extant belong to the Chapter on the Elements, the Chapter on the Faculties, and the Chapter on the Worthy and the Noble, and the Chapter on Knowledge is wholly wanting, so that this fascicle has no line-by-line exposition of Shentai’s to attach. This is marked as wanting according to the facts, and the text of other fascicles is not made to stand in its place. But one thing is worth setting down: fascicle one of Shentai’s Commentary has a general conspectus of the whole treatise, in which the sentence “the chapter on knowledge, sixty-one lines” remains an independent witness to the stanza count of this chapter — the general conspectus of a fragmentary commentary can bear witness beyond the range of its own fragmentation, which is one method this fascicle yields. As for the line-by-line exposition of this chapter, it remains wanting.

A division into four and a division into seven. Puguang and Yuanhui alike adjudge the gate of “explaining the differences among the knowledges” to divide into four — the difference of receptivity, knowledge, and view; the difference of mark among the ten knowledges; the modes of apprehension of the ten knowledges; and the analysis through the several gates — the two agreeing word for word, Yuanhui here plainly following Puguang’s outline. Fabao, on the other hand, sets up seven — the difference of receptivity, knowledge, and view; the difference of mark among the ten knowledges; the establishing of the ten knowledges; the difference of what dharma knowledge severs; the modes of apprehension of the ten knowledges; the analysis by gates of sense; and the extended explanation of the cultivation of the knowledges. Wherein lies the difference? Fabao makes independent sections of “the establishing of the ten knowledges” (the seven grounds on which the ten are established) and “the difference of what dharma knowledge severs” (the dharma knowledges of cessation and the path counteracting the upper realms as well), which Puguang gathers into “the difference of mark among the ten knowledges”; and he splits Puguang’s “analysis through the several gates” into the two sections of analysis by gates of sense and the extended explanation of the cultivation of the knowledges. ⇒ In one and the same stretch of the treatise, one house sees four joints and another sees seven. Recorded without adjudication.

And from this a rule for reading the commentaries follows. Because the divisions differ, the ordinal numbers differ too: at one and the same passage, “within these ten knowledges, which has which modes of apprehension,” Puguang marks “this is the third, explaining the modes of apprehension of the ten knowledges” and Yuanhui “from here the third of the main text,” while Fabao marks “below, the fifth, explaining the modes of apprehension of the ten knowledges.” ⇒ In citing an ancient master’s “the Nth,” the name of his house must be cited along with it, or one and the same passage will come out with ordinal numbers contradicting one another. This is not that one of the three commentaries is wrong; it is that there are three different knives.

The gloss on the chapter’s name: two glosses and three. Puguang glosses the name of knowledge as “because it decides and knows with weight it is called knowledge,” and gives the reason of the order — “the previous chapter explained the result and this chapter explains the cause, and since the cause is the nearer as regards the result, knowledge is made out next.” Fabao gives two glosses: first, “discernment deciding upon what has once been seen is called knowledge”; second, “or else, the mind that seeks to see the object having come to rest and the preparatory effort having slackened, it is called knowledge.” Yuanhui is briefest, giving only “deciding is called knowledge, and this chapter explains it at length, hence ‘analysis.’” ⇒ On the matter of the chapter’s name, Puguang is full on the order, Fabao full on the glossing, and Yuanhui brief; and that second gloss of Fabao’s is really put to use in the text of this fascicle (in explaining why the two knowledges of exhaustion and non-arising are not view), while neither Puguang nor Fabao expressly uses it there — a layer to be recorded without adjudging which is the better.

Why receptivity is not knowledge: the treatise’s own gloss and Fabao’s citation. The reason the treatise itself gives is “because the doubt it is itself to sever has not yet been severed” — spoken from the severing or not severing of doubt. Fabao here cites the Treatise Conforming to Correct Principle besides: “or else, the mind that seeks to see the object having come to rest and the preparatory effort having slackened, it is called knowledge; the receptivities, as they arise, weigh and consider, and their intent and preparatory effort are fierce, and so they are not included in knowledge” — spoken from the fierceness or slackness of intent and preparatory effort. ⇒ Two glosses standing together by different roads: the treatise speaks from what is severed, the Correct Principle from the force of what apprehends. Fabao cites it in aid of explanation and does not say which of the two prevails. Recorded without adjudication.

The speaker-labels of the passage on the gathered and the scattered mind: the two commentaries agree entirely. This is the one passage of debate in this fascicle in which Puguang, Fabao, and Saṅghabhadra all three have a part, and the assignment of the speakers is therefore clearest here. Puguang marks “the Vaibhāṣika masters refute,” and explains the identity of the masters of the West thus: “the masters of the West are the masters of the country of Gandhāra; the Mahāvibhāṣā calls them the masters abroad”; Fabao marks “the Sarvāstivādins refute the explanation of the masters of the West; first refuting by principle, second refuting by contradiction of the root treatise.” ⇒ What the two commentaries assign agrees entirely: what refutes the masters of the West is the line of the Vaibhāṣika masters (the Sarvāstivādins), and not some further house set up beside them; and Fabao divides the refutation into the two layers of refuting by principle and refuting by contradiction of the root treatise, answering one by one to the two reasons the treatise gives (the fault of extending to both gathered and scattered, and the contradiction of what the root treatise says). This set of labels is the most reliable anchor in this fascicle for “who asks and who answers.”

The passage on what is formed through hearing and reflection: Fabao marks “the difference is one of school.” Under the treatise’s passage “not formed through cultivation … one would become free of defilement all at once,” Fabao says expressly, “this is not the tenet of the Sarvāstivādins; how is it known? The Treatise Conforming to Correct Principle says …,” citing Saṅghabhadra’s refutation word for word, and concluding at the end, “the two accounts differ, the difference being one of school.” ⇒ Fabao here passes no verdict but only points out that the difference of the two accounts issues from a difference of school. This sentence, “the difference being one of school,” is the ancient masters’ own statement of the form of “one meaning and two schools recorded together” — not a rule laid down for them by later men.

What has not been checked out and is marked as wanting according to the facts. For the passage on whether modes of apprehension without outflows exceed the sixteen (the second half of Root 12), the passage on what is formed through hearing and reflection, and the passage on the objection about what the sense of cultivation rests upon, the word-for-word outline labels of Puguang and Fabao for “who establishes and who refutes” have not been checked out in the apparatus of this fascicle, and are marked as awaiting supplement according to the facts, and not filled in by conjecture. ⇒ At these three passages the reader may simply read the assignment from the tone of the treatise text itself; and of the three, the first two have both parties expressly marked in the treatise text (Kaśmīra / the masters abroad; the author’s own adjudication), while the third is an objection the author poses to himself and answers by citing an ancient master (see the following item).

Yuanhui has no chapter-title line. Of the three commentaries, Yuanhui alone has no separate chapter-title line at the head of his fascicle for this chapter, the fascicle title beginning it directly; the five characters “Analysis of Knowledge” yield a zero result within the range of Yuanhui’s Chapter on Knowledge, and the chapter title appears only in the collation note of the B recension. ⇒ This is a difference purely of form, touching no doctrine, yet it has a practical consequence for searching — to search Yuanhui’s Chapter on Knowledge with the string “Analysis of Knowledge” is bound to yield a false zero. Set it down.

2. Saṅghabhadra’s Refutations

Saṅghabhadra’s two treatises name this chapter not “Analysis of Knowledge” but “Making Out Knowledge”; the Treatise Conforming to Correct Principle gives it the same chapter number, the seventh, while the Treatise Revealing the Doctrine gives the eighth — the shift of the latter being due to its setting up a preface chapter at the head of the whole work, which adds one throughout, and not a peculiarity of this chapter. The opening sentences of the two treatises are word for word the same.

First: on the adjudication of the gathered and the scattered mind, Saṅghabhadra reverses the direction. The treatise records the two glosses of the Vaibhāṣika masters and of the masters of the West, and adjudges the explanation of the masters of the West “not in accord with principle.” Saṅghabhadra, calling Vasubandhu directly “the Sūtra-master,” reverses the adjudication: he says that the Sūtra-master holds this not to accord with the sūtra, whereas “the explanation of the Westerner does accord with the sūtra,” and says of Vasubandhu that he “does not examine the sense of the sūtra and praises and blames at random.” ⇒ Three layers are to be seen here clearly: the refutation in the treatise and the refutation of Saṅghabhadra issue alike from the Sarvāstivāda line, and yet their adjudications are opposite — the treatise (following the Vaibhāṣika masters) adjudges the masters of the West contrary to principle, and Saṅghabhadra adjudges them in accord with the sūtra and reproaches Vasubandhu. The fierceness of the wording of this refutation is the greatest in the fascicle. All three parties recorded, without adjudging right and wrong.

Second: the defining of a mode of apprehension. Answering the objection that “if discernment is not associated with discernment then discernment has no mode of apprehension,” the treatise establishes that “every difference of class in the taking of an object by mind and mental factors is called a mode of apprehension.” Saṅghabhadra cites this and refutes it, saying “here the Sūtra-master, leaning upon another school, speaks thus,” and questions back, “one should consider what is called a difference of class in the taking of an object by mind and mental factors.” ⇒ What is at issue is whether the one name “mode of apprehension” may be gathered up under “difference of class in the taking of an object.” Saṅghabhadra also reproaches Vasubandhu for relying upon what is not his own school. Recorded without adjudication.

Third: the adjudication of what is formed through hearing and reflection. The treatise adjudges the conventional knowledge that makes the mode of apprehension of non-self to be formed through hearing and reflection alone and not through cultivation, the reason being that discernment formed through cultivation takes a level severally as its object. Saṅghabhadra cites it word for word and refutes it, saying “to say that what is formed through cultivation takes a level severally as its object alone is not fully established,” and saying further that this knowledge, taking everything as object in general and having only the mode of apprehension of gladness, “has no capacity with respect to becoming free of defilement, and so what he says is exceedingly contrary to principle.” ⇒ This passage is the most concrete point of contention in the fascicle, since it touches a real question: whether that knowledge which takes all dharmas as object in general and contemplates non-self can after all be formed through cultivation and have any power of freeing from defilement. Fabao cites this refutation and marks it “the difference being one of school” (see the preceding section). One meaning and two schools recorded together, without adjudging which prevails.

Fourth: the objection about what the sense of cultivation rests upon — the place where assignment most needs care. The run of the treatise’s text here is that the author poses a question to himself (since it cannot arise, upon what does the sense of acquisition rest) and answers it by citing the explanation of an ancient master; this question and answer lie within the author’s own running text and are not words refuting another. Yet when Saṅghabhadra takes this objection up again in the Correct Principle, he marks it directly as “here the Sūtra-master raises such an objection,” and thereupon likewise adduces the ancient master’s account that “the sense of cultivation can be made to stand.” ⇒ One and the same stretch of text is in the treatise a self-questioning and self-answering, and under Saṅghabhadra’s brush becomes an objection raised by the Sūtra-master. This nicety of assignment is recorded only as a collation of the wording of the two texts, without adjudging whether Saṅghabhadra’s move is warranted. The form of this item also differs from the three foregoing — in those three Saṅghabhadra refutes directly, while here he first cites the objection and then adduces the ancient master, without refuting directly.

An honest negative proposition. The word “Sūtra-master” yields a zero result in the Treatise Revealing the Doctrine within the range of this fascicle, and of the four passages listed above, three (the gathered and scattered mind, the defining of a mode of apprehension, what the sense of cultivation rests upon) have not preserved the content of the debate, only the doctrine of “formed through hearing and reflection” remaining. ⇒ All that this negative proposition can prove is one thing: the Treatise Revealing the Doctrine has not preserved the framework of that debate in this fascicle; one cannot infer from it that it lacks those doctrines.

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, and the first thing to know is that the chapter name itself differs: Xuanzang gives “Analysis of Knowledge,” Paramārtha “Analysis of Discernment” — he does not take the word “knowledge” for the chapter name but styles it directly by “discernment,” while within the chapter the “knowledge” of the ten knowledges is still rendered out passage by passage. The chapter number in both texts is alike the seventh, not shifted. To search Paramārtha’s text with Xuanzang’s chapter name is never to find it.

The renderings form a system of their own. With outflows is rendered “with flow,” foundation of mindfulness “station of mindfulness,” Kaśmīra “the country of Jibin,” the masters abroad “the masters of the western country,” cultivation as guarding and as contemplating “cultivation as warding and cultivation as selecting.” ⇒ None of these sets is peculiar to this chapter; they are the systematic wording of the two translations throughout. Cross-text collation can be taken only by the structure of the scene — who, to whom, what is asked, on what principle, and what answer is got — and one cannot take a word of the one text and search for it in the other.

The two ends, opening and closing, agree entirely in scene. At the opening, Xuanzang gives “the previous chapter said at its beginning the several receptivities and the several knowledges, and afterward said right view and right knowledge,” and Paramārtha “having spoken of the several receptivities and the several knowledges, it speaks of right view and right knowledge”; at the close, Xuanzang gives that the masters abroad say cultivation is of six kinds, adding guarding and contemplating, while the treatise masters of the country of Kaśmīra say these two cultivations are comprised within cultivation as counteragent and cultivation as expulsion, and Paramārtha that the Abhidharma masters of the western country say there are six kinds of cultivation, the fifth cultivation as warding and the sixth cultivation as selecting, while the masters of the country of Jibin say these two cultivations enter into cultivation as counteragent and cultivation as purifying counteraction. ⇒ At the two ends, head and tail, of one fascicle, the two texts answer to each other sentence by sentence, which is the strongest collateral evidence for the boundaries of this unit.

One instance of a small difference in opening out and gathering in. At the general statement of the ten knowledges, Xuanzang first asks “how many kinds of knowledge are there? and what are the differences of their marks?” and then answers “there are ten kinds of knowledge, gathering up all knowledge,” while Paramārtha lifts “gathering up all knowledge” into the question, giving “by how many measures can all knowledge be gathered up? by the ten knowledges.” The order is the same in both (first asking how many, then answering ten, then gathering into two), but the weight of the question and answer falls in the one upon “how many kinds” and in the other upon “gathering up all.” Recorded without adjudication.

The opening out and gathering in of the basis of the simile. In the gloss on conventional knowledge, Xuanzang gives “since for the most part it takes conventional objects such as a pot,” raising the single item of the pot; Paramārtha gives “since this knowledge for the most part takes as object conventional kinds such as pots, garments, men, and women,” raising four items, pot, garment, man, and woman. ⇒ One and the same gloss with the basis of its simile abridged in the one and detailed in the other, with no change of sense; recorded only.

At several places in this fascicle the word-for-word collation of the two texts has not been checked out. The passage on the three limits of level, faculty, and stage of the knowledge of others’ minds; the passage on the three gates of mode of apprehension, that which apprehends, and that which is apprehended; the passage on the gathered, scattered, sunken, and restless minds; the item-by-item collation of the full names of the four cultivations; and the item-by-item collation of the full names of the ten knowledges — for these the corresponding wording of Paramārtha’s text has not been collated sentence by sentence in the apparatus of this fascicle, and is marked as awaiting supplement according to the facts, and not filled in with Xuanzang’s words. ⇒ At these several places the reader may take Xuanzang’s text as sufficient for the present; the fine collation of the two texts awaits later supplement.

4. Coordinates for Parallel Reading

Two direct dovetailings with the Chapter on the Worthy and the Noble. The first is the opening of this fascicle — the treatise cites word for word the “several receptivities and several knowledges, right view and right knowledge” of the Chapter on the Worthy and the Noble and raises its question from them, so that those hundred and sixty characters of the first section are in fact a reflection back upon the receptivities of the path of seeing in that chapter. The second is the conventional knowledge at the border of direct realization — it is cultivated in addition only at the three subsequent knowledges of suffering, origination, and cessation, and these three moments fall upon the table of the sixteen minds of the path of seeing in the Chapter on the Worthy and the Noble; the two tables must be read together, and only then will the section on “how many knowledges are accomplished” in this fascicle turn from a list into a derivation.

One forward juncture with the Chapter on Absorption. The intermixed cultivation of the meditative absorptions appears in this fascicle only within the ledger of “how many knowledges are cultivated,” and the treatise does not explain for what purpose the intermixed cultivation is made; the result of that intermixed cultivation — the cause of the five Pure Abodes — is not opened out until the Chapter on Absorption. ⇒ This fascicle establishes only the number of knowledges cultivated and not its result, which is a plain juncture.

The three rounds of glossing of the sixteen modes of apprehension are a coordinate for reading alongside another lineage. The treatise gives three sets of glosses in a row upon the sixteen modes of the four truths, and says of itself, “since the glosses of old are thus not of one gate only, let a further explanation be made as each may please.” ⇒ One and the same set of sixteen names admits of glossing through several gates, and this form is itself a place for parallel reading: the several houses of later times each have their own explanation of the sixteen modes of apprehension, and in reading one may set side by side the gate of sense each has taken, without needing to adjudge which is the better.

One last loose thread. That sentence in the closing passage of this fascicle, “all this should be pondered as is fitting,” hands back to the reader the cells not set out among the twenty-odd gates. This is not an economy of the brush — it is the form of Abhidharma: the rules have been given in full, and the ledger must be reckoned by oneself. May the reader, after the latter half of this fascicle, truly take up a brush and reckon it once through.