Chapter 1: What kind of book are these two hundred fascicles?¶

The Mahāvibhāṣā ranks among the largest works in the Chinese Buddhist canon. It has two hundred fascicles, 1.54 million characters in classical Chinese, and more than 2.3 million in the modern Chinese translation. Faced with a book of this size, our first question concerns its bulk. What makes it so long? This is no rhetorical question. Its answer will establish how this entire essay reads the work. The book is long because it carries out every multiplication its method calls for. It is not complex because there are only a few multipliers. This chapter will turn that statement into a conclusion that can be checked.

First, what does it claim to be? The very first question in fascicle 1 asks who composed the treatise. The answer is the Buddha, the Blessed One. The reason given is that the nature of all knowable dharmas is so profound and subtle that no one other than the omniscient Buddha could fully awaken to it and disclose it. As for who speaks the questions and answers in the book, some say that Śāriputra asks and the Buddha answers.¹ This opening gives the work's entire understanding of itself. The Buddha has already stated the whole truth. The masters' task is disclosure and organization, not discovery. The formulas explaining why the treatise was composed fit this understanding. When the work introduces a topic, it often begins by asking, "Why compose this discussion?" Two stock answers recur. One is "To refute other schools and make our own position clear." Together with variants using "the correct position" and "the correct principle," it occurs more than a hundred times. The other is "That sūtra is the foundation on which this treatise rests; what it did not state must now be stated." The identical wording for that foundation claim appears fifteen times; the family of expressions using "the foundation on which it rests" appears seventy-five times.¹ Its first use, in fascicle 8, concerns the Pond Simile Sūtra. The sūtra briefly distinguishes twenty formulations of the view of a real personality but does not say how many are views of self and how many are views of what belongs to self. "That sūtra is the foundation on which this treatise rests. What it did not state we now wish to state; hence this treatise is composed."¹ Together, these two formulas constitute the work's own charter. Refute others and complete the sūtras. It understands itself to be conducting debate and also to be saying on the sūtras' behalf what they left unsaid. To measure it by the standards of argument is simply to take it at its word. This is hardly an excessive demand.

A few formulas and one root treatise. The exposition of the Jñānaprasthāna supplies the text's framework. Kātyāyanīputra's root treatise sets out topics one by one, and the Vibhāṣā evaluates them one by one, taking up all the neighboring questions along the way. It explains the Jñānaprasthāna while ranging widely. Besides the Jñānaprasthāna, six other treatises are collectively called the "six feet." For this chapter it is enough to say that the text being explained is the root treatise, while the six feet form the same school's inherited resources. A few formulas expand the book's bulk. Their counts suffice to show what each does. Question and answer is the basic form, and "Why compose this discussion?" is a regular opening question. "Furthermore," with 3,625 occurrences, adds parallel answers to the same question. These answers are not ranked and do not replace one another. "Some say," with 3,313 occurrences, opens the door to another position. "The same applies," with 1,626 occurrences, is the book's real multiplication operator. The same structure runs again under another heading. Each run adds another multiple to the bulk. Its densest concentration is in fascicle 92, with fifty occurrences. That fascicle works through how many underlying tendencies are abandoned and how many fetters exhausted when full understanding is attained concerning consciousness that takes the eye faculty as its object. Then the stamp goes down across the list. "The same applies to the ear, nose, tongue, and body faculties"; "the same applies to the male faculty and the faculties of pain and grief"; "the same applies to the mind faculty, the faculties of pleasure, joy, and equanimity, the five beginning with faith, and the three uncontaminated faculties."¹ One case is calculated, and an entire list is stamped. This is the most common kind of page in the two hundred fascicles. "Why is this?" occurs 785 times, asking for reasons. "And so on, as explained at length" occurs 696 times, making omissions explicit, as though the page bore an instruction to expand the formula here. "This requires four alternatives" occurs 312 times. A two-by-two grid is drawn first, and dharmas are then placed in its cells. Fascicle 5 asks, "Can someone possess the supreme worldly dharmas without possessing the acquisition of disjunction from them?" The answer calls for four alternatives. One possesses the former but not the latter, the latter but not the former, both, or neither. Each cell is assigned a kind of person.¹ Multiply the grid by the doctrinal sets, and another section results. "The evaluation states," with 289 occurrences, brings the various positions to a close. Chapters 2 and 3 will address it directly. The quick reference table in Chapter 0 lists these operators. This chapter adds one point. The book has almost no others.

A comparison of equal lengths separates two sources of bulk. A work can be long because it contains many ideas or because its formats keep multiplying. Repetition lets us distinguish the two. This book compares equal lengths of two modern Chinese translations by the same translator, measure by measure.^{2 3} For twelve-character windows, the repetition rate is 15.9 percent in the Vibhāṣā and 1.8 percent in the Abhidharmakośa. For twenty-character windows, it is 7.4 percent against 0.4 percent. The general-purpose compression ratios are 0.270 and 0.346. Yet the Vibhāṣā uses fewer distinct Chinese characters, 1,984 against 2,109. At equal lengths, its density of repeated twelve-character strings is about nine times that of the Kośa, and its density of repeated twenty-character strings about eighteen times. The classical Chinese points in the same direction. Its two hundred fascicles contain 1.54 million characters, with 2,872 distinct characters and an 8.4 percent repetition rate for twenty-character windows. The Kośa scarcely says any sentence twice. An outline and an extensive commentary ought to differ in genre, but genre cannot explain a difference of ninefold and eighteenfold. Only one explanation remains. The formats are multiplying. The number of ideas cannot account for this bulk. The doctrinal sets, multiplied out in full, can.

The multiplication is deliberate. There are three pieces of evidence. "And so on, as explained at length" is an explicit instruction to expand, not a vague ellipsis. Where multiplication is called for, the treatise knows that it is multiplying. "This requires four alternatives" draws a grid before filling it with dharmas. The grid is applied to doctrinal sets, with as many grids as there are sets. The strongest evidence is the charter itself. "That sūtra is the foundation on which this treatise rests; what it did not state must now be stated." The treatise supplies what the sūtra left incomplete. Since the sūtras always leave something unstated, the treatise can never be finished. A book that defines itself as a procedure for completing what is missing may seem out of control when it reaches two hundred fascicles. In fact, this is the normal result of its charter's operation.

The Vaibhāṣika identity is written into the attribution. The line beneath the opening title reads "Composed by five hundred great arhats and others."¹ That line compresses the council legend into an authorial attribution. According to tradition, five hundred arhats compiled the treatise together in King Kaniṣka's time. The records of that legend and the assessment of its historicity belong to Chapter 2 of Essay 3 in Part I. This essay does not retell the history. It takes one point only. This attribution sets the tone of the book's verdicts. "The evaluation states" is a collective adjudication, not the opinion of an individual master. The four Chinese characters 毗婆沙师, "Vaibhāṣika," thus become an identity. Whoever holds this treatise's orthodox positions is a Vaibhāṣika. A treatise creating a school's own name for itself is a rare case in the literature of the Buddhist schools. Chapter 3 will ask about the relation between adjudication and argument. Its roots lie here. A verdict spoken in the tone of a resolution calls for a different kind of support from an argument.

On to Chapter 2. The source of the bulk is clear. A few formats multiply across an immense array of doctrinal sets. The next chapter examines how arguments actually work beneath this format. Two rules now govern our reading. When repetition appears, first identify the operator at work. When a verdict appears, first ask what supports it. The next eight chapters need both rules.

References¶