Chapter 6: The refutation of self

The first eight chapters organize and criticize the Sarvāstivāda system. The book could end there. Vasubandhu begins another fascicle to address directly the thing whose name the entire work has avoided saying aloud, the self. The chapter refuting self is a different kind of text in form. Its title does not stand at the beginning of a fascicle, and it has no root verses of its own. The alternation of verse and exposition in the first eight chapters gives way to prose alone. Whether it is appended or included in the main work is handled differently in the two Chinese translations. Xuanzang's translation begins it midway through fascicle 29 and continues it as a second part in fascicle 30. Paramārtha's translation devotes fascicle 22 to it, placing the chapter title at the beginning. Fabao assigns it, together with the final four verses, to the concluding section, saying that it was added when the verses were explained and does not belong to the root verses.1 It is the work's point of closure, and the literary form itself makes that clear. Weight is measured by position, not length. The chapter occupies only one-twentieth of the book, yet offers the fullest statement of nonself in Buddhist literature.

The opponent and the stakes. The main opponent is the Vātsīputrīya school. The four characters 補特伽羅, "person" (pudgala), occur nearly seventy times in fascicles 29 and 30 together. The three characters 不可說, "inexpressible," occur sixteen times.2 The Vātsīputrīyas posit an inexpressible person, neither identical with nor separate from the five aggregates, neither permanent nor impermanent. Press the question to its limit, and the answer is "inexpressible." The position seeks to preserve both ends, a bearer of rebirth and memory without violating the doctrine of nonself. It is the most honest struggle within the Buddhism of the schools, because it openly makes a question of the need that other schools hide in acquisition, continuum, and seed.

Vasubandhu's questioning always lands in one place. The debates are rich in detail, but the method is almost cruelly uniform. "Inexpressible" either says something or says nothing. If the person can be known, say whether it is identical with or different from the aggregates. If identity or difference cannot be stated, what object does the consciousness that knows the person take? If the object is also inexpressible, the phrase "knowing the person" loses its content. The classifications of two kinds of truth and three kinds of designation are tested cell by cell. In each cell, the Vātsīputrīyas must choose between being able to say something and being unable to say it. Either choice loses. Chapter 3 of Essay 13 established the basis in the five aggregates and nonself. Chapter 7 of Essay 14 examined how the Southern tradition holds that line. The refutation of self supplies the fullest set of arguments along it. Its new arguments are precisely the Sautrāntika tools established in the preceding chapters. Differences in the continuum's capacities can do everything a bearer does. The bearer therefore has nothing to do.

Two non-Buddhist fronts. At its end, the chapter turns to the selves of the Vaiśeṣikas and Sāṃkhyas. These debates are different in kind. The Vātsīputrīya person is a compromise within Buddhism. The Vaiśeṣika and Sāṃkhya selves are explicitly posited metaphysical entities, agents and experiencers. Vasubandhu's refutations of the non-Buddhist positions take less effort because the positions are clear. They make determinate claims about identity and difference, permanence and annihilation, and each claim can be refuted in turn. It is always the inexpressible that proves troublesome. Seen together, the three debates show the complete form of nonself. Externally it refutes entities; internally it rejects compromise. It then assigns every function of bearing dharmas to a continuum without a bearer.

This chapter's position and the question it leaves. Placing the refutation of self at the end announces that the reckonings in the first eight chapters all lead here. After the three times, acquisition, the nonmanifest, and the unconditioned have been struck from the list one by one, the final cancellation is the general request behind them all. Give experience an owner. Vasubandhu's answer is no, and his argument reaches the point where no owner can be given. The remaining question is here too. A continuum without a bearer must carry memory, responsibility, and liberation alone. The seed theory is pushed to its limit. How that limit pushes Sautrāntika toward Yogācāra is the starting point of Essay 20.

On to Chapter 7. Vasubandhu has finished speaking. Sarvāstivāda's strongest defender read this book and wrote eighty fascicles in response. The next chapter examines that defense.

References

References

  1. The formats of the different versions. Xuanzang's separate one-fascicle Root Verses of the Abhidharmakośa (T1560) lists only the six hundred verses of eight chapters. The ninth chapter has no verses. The separate verse text thus independently confirms that the refutation of self has no root verses of its own. Paramārtha's Abhidharmakośa-bhāṣya (T1559) calls it "Chapter 9: Refuting the assertion of self," gives it all of fascicle 22, and places its title at the beginning. The entire fascicle has one occurrence of 偈曰 ("The verse states") and zero of 釋曰 ("The explanation states"), sharply differing from the alternation of verse and explanation in the first twenty-one fascicles. The adjacent fascicle 21 has 125 instances of 偈曰 and 126 of 釋曰. This shows that the older translation's version of this chapter likewise carries no root verses. The three verses lamenting the state of the world, placed at the end of fascicle 29 in Xuanzang's translation, appear at the beginning of this chapter in the older translation: "The Master, the world's eye, has closed his eyes; those who realized and taught the teaching gradually die and scatter ..." Fabao's Commentary on the Kośa, fascicle 1, assigns it to the concluding section: "The final four verses and the chapter refuting self form the concluding section of the exposition of the verses. The first three and final four were added when the verses were explained; they are not root verses." He thus treats this chapter and the final four verses alike as additions made when the prose exposition was composed. No Sanskrit text was available for this point, so it could not be checked. ↩

  2. Abhidharmakośa Śāstra, thirty fascicles, composed by Vasubandhu, translated by Xuanzang, Taishō vol. 29 (T1558). The chapter refuting self runs from the end of fascicle 29 through fascicle 30. 補特伽羅 (person) occurs 49 times in fascicle 29 and 20 in fascicle 30; 不可說 (inexpressible), 11 and 5 times respectively, according to a complete search of the classical Chinese text in the CBETA edition. The two points that the title is not at the beginning of a fascicle and that the chapter has no root verses of its own have been checked within the scope of the modern Chinese translation. The chapter has zero occurrences of the formulas 頌曰 ("The verse states") and 論曰 ("The exposition states") in its own verse-and-exposition structure. Its quoted verses are introduced by "Therefore there is a verse that says" or "Another verse says." They are citations, not root verses. The opening in fascicle 29 establishes a grid through the distinction between real and nominal: "The Vātsīputrīyas, however, hold that a person exists whose substance is neither identical with nor different from the aggregates. This should be examined. Is it real or nominal? How do the characteristics of real and nominal existence differ? A separately existing thing is the characteristic of real existence, as with material form, sound, and so forth. Mere aggregation is the characteristic of nominal existence, as with milk, curds, and so forth." This is the same criterion for real and nominal existence as in Chapter 1 of Essay 15. The grid of truth and designation, the passages asking what object the consciousness that knows a self takes, and the two sections on Vaiśeṣika and Sāṃkhya all occur within these two fascicles. Each passage had been identified before drafting. The modern Chinese translation appears in this site's Kośa section, translated from the CBETA edition. ↩