Chapter 7: Latent Defilements and Their Severing¶

In this system the proper name of the defilements is the latent defilements. Part 15 dealt with their neighbours (possession, unmanifest matter, the life faculty), and defilement itself was the one thing it never took head-on. This chapter supplies the omission, and carries throughout the question of this part: is the ever-finer table of stages for the severing of the defilements an achievement of classification, or a specimen of multiplication?

The dispute over their standing. The great divide should be recorded first. The Sarvāstivāda holds that a latent defilement just is the defilement in its own substance, associated with mind, lying dormant and breaking out being no more than a difference in the force exercised by one and the same dharma. This the Mahāvibhāṣā took over rather than invented. The root treatise, the Jñānaprasthāna, glosses growth at fascicle one as "that latent defilement is associated with this mind and unsevered, and takes this mind as its object," so that association is already in the wording of the root text.¹ Fascicle fifty of the Mahāvibhāṣā uses it as a settled premise, asking how one may speak of the dissociated when every latent defilement is associated with mind and the rest, and answering that there the name of latent defilement is applied to possession.² The other side is not one party. The Vibhajyavādins hold that the latent defilement is the seed of the entanglement, that a latent defilement is in its own nature dissociated from mind while the entanglements are in their own nature associated with it, and that the entanglement is born from the latent defilement; these sentences stand at fascicle sixty and nowhere else in the whole work, and they are not a free-standing position but the premise of their proof that an arhat does not fall back, the latent defilement being severed, no entanglement arises, and with no entanglement there is nothing to fall back from.² The record of the Dārṣṭāntikas runs another way. Fascicle twenty-two reports that they allow a latent defilement no growth upon its object and none upon the dharmas associated with it; fascicle ninety reports that they hold mind and mental factors to arise in succession and not to be associated with one another at all.² The two parties collide at the conclusion and reach it by different roads: the Vibhajyavādins set up a further dissociated entity to house the seed, while the Dārṣṭāntikas abolish the relation of association itself. By fascicle nineteen of the Kośa the Sautrāntikas have the thing said out whole: the substance of a latent defilement is neither associated with mind nor dissociated from it, there being no separate thing there, a defilement in its sleeping state being called latent and in its waking state an entanglement. The same passage turns back on the Vibhajyavādins: if a defilement had besides itself a dissociated latent defilement for its seed, then memory too would need a dissociated dharma to give rise to later memory, and it does not.³ The word "entanglement" occurs seven hundred seventy-five times in the whole work, and the division of labour between latent defilement and entanglement is the ground this dispute is fought on.²

What is at issue. It can be said in one sentence: in what manner is a defilement still there when it is not occurring? The Sarvāstivāda's answer is built on the same plan as its answer wherever persistence is at stake (the dharma-substance exists always, and belonging is managed by possession), and so a latent defilement has to be a real dharma associated with mind. Both of the other roads describe the defilement that is not occurring as a kind of latency, and that is the direct ancestor of the doctrine of seeds: the line Chapter 3 of Part 17 will pick up is already buried here. The Mahāvibhāṣā did not hand this dispute to the evaluation. Fascicles twenty-two and fifty do not carry the word once, and the two evaluations in fascicle sixty belong to other questions. The side of association is held in place by two things instead, fascicle fifty treating it as a premise no longer worth asking about, and fascicle sixty answering the Vibhajyavādins with four terms of denunciation, the fruit of not knowing, the fruit of darkness, the fruit of ignorance, the fruit of not striving.² That runs in the direction generalized in Chapter 3: an account by latency would loosen the ground under the doctrine that the dharma-substance exists always, and whoever defends the dharma-substance wins.

The ninety-eight latent defilements, a standard piece of multiplication. The phrase "the ninety-eight latent defilements" occurs one hundred nine times in the whole work.² The way they are opened out is the machine readers of this part already know. The root defilements are six: greed, hatred, conceit, ignorance, doubt, and view. View opens into five, which makes ten. The ten latent defilements are then divided by the three realms, and by the five sections (severed by the seeing of suffering, by the seeing of origination, by the seeing of cessation, by the seeing of the path, and severed by cultivation), and when the multiplying is done the total is ninety-eight. Chapter 1 said that this book is long because it multiplied out everything there was to multiply; the ninety-eight latent defilements are the standard illustration of that sentence. There is one layer of position and no more (six root defilements against a grid of five sections and three realms), and every other line on the table is a product of the multiplication.

Seeing and cultivation: the density of the table of stages. In the whole work "the path of seeing" occurs nine hundred forty-three times, "the path of cultivation" two hundred sixty-eight, "the unhindered path" six hundred eleven, "the liberated path" six hundred sixty-eight.² The density of these words shows what standing the table of stages has in this book. The severing of the defilements is not one act but a timetable: which section of defilements is severed on which path, how many moments the severing takes, that the unhindered path severs while the liberated path realizes—all of it settled cell by cell. The table is laid out with the greatest precision; what must be asked is what kind of thing it is. If it is a description of the experience of practice, then it answers to experience. If it is legislation derived from the enumeration of dharmas, then it answers only to the cells. This book's judgment is that it is mostly the latter: the subdivision of the stages runs far beyond any report of experience that could be checked, and its regularity is the regularity of the cells.

The three realms and nine grounds: two tables meshing. "The three realms and nine grounds" occurs forty-seven times in the whole work, "the nine grounds" one hundred fifty, "the nine grades" more than two hundred fifty.² The grounds are the spatial axis of the severing, the grades are the intensity axis of the defilements, and the stages are the axis of progress. Multiply the three axes together: each ground has nine grades of defilement, each grade is fitted with nine unhindered and nine liberated paths, and the advance runs ground by ground from the realm of desire up to the peak of existence. This is the most spectacular exhibit of Mahāvibhāṣā multiplication, more spectacular than the list of seventy-five dharmas, because it is three-dimensional. Here the classificatory work reaches its summit; and argument was called for at exactly one point, the dispute over standing at the head of this chapter, which is precisely the point that never went to the evaluation and was closed by assertion and denunciation instead. The answer of this chapter therefore meets up with Chapter 2: the table is a genuine achievement, and the ground the table stands on was adjudicated into place.

It is handed on to Chapter 8. What the table of stages arranges is how a person walks through time; the next chapter looks at what time itself was rewritten into by this system.

References¶