CANON

There is no such thing as the Buddhist canon. In fact, the concepts of canon and canonicity are especially problematic in Buddhism, given the wide geographical spread and great historical variety of the religion, together with the absence of any central authority. If the term canon is defined loosely as a more or less bounded set of texts accorded preeminent authority and sanctity, then each Buddhist school or tradition to evolve developed its own canon in the process. While agreeing on the centrality of the notion of buddhavacana (word of the Buddha) as capable of leading others to awakening, Buddhists may and do differ over what actually constitutes this buddhavacana.

In view of the perennial possibility of disagreement and misunderstanding, Buddhists formulated explicit guidelines for authenticating religious teachings as true buddhavacana and interpreting them correctly. These guidelines include the four great authorities (mah?pade?a), which directed that teachings were to be accepted as authentic if they were heard from (1) the Buddha himself; (2) a sa?gha of elders; (3) a group of elder monks specializing in the transmission of dharma (i.e., s?tra), vinaya or m?t?k?s (the matrices or mnemonic lists that became the abhidharma); or (4) a single elder specializing therein. But teachings heard from any of these authorities could only be accepted if they conformed with existing scriptural tradition (i.e., with the s?tra and vinaya), and also, according to a variant formulation, if they did not contradict the nature of things (dharmat?). Another set of principles, not subscribed to by all Buddhist groups, held that in receiving and interpreting teachings one should follow the four refuges or reliances (prati?arana), relying on the dharma taught in preference to the person teaching it, the meaning (or spirit) of it rather than the letter, s?tras of definitive or explicit meaning (n?t?rtha) rather than implicit meaning requiring interpretation (ney?rtha), and direct understanding (jñ?na) rather than discursive knowledge (vijñ?na).

Even while emphasizing seniority and tradition, these interpretative principles place a higher premium on content and its realization than they do on form and obedience to it. Truth (dharma) emerges as the primary value, ever the same whether buddhas arise to preach it or not, independent of particular formulations by particular people, so that eventually the statement "All that the Buddha has said is well said" is turned around: Whatever is well said (i.e., true) is the word of the Buddha. Canonicity is therefore defined in functional terms: If a teaching is meaningful, if it is in line with the dharma, and if it tends to eliminate the defilements and lead to liberation, then any product of inspiration (pratibh?na) may be accepted as the word of the Buddha. Under such conditions, innovations inevitably crept in, some of them rejected as not being the true word of the Buddha, but some of them finding acceptance, especially if they accorded in spirit with existing belief. It was in this way that the Mah?y?na s?tras eventually came to be accepted by some Buddhists as buddhavacana, as did the Buddhist tantras after them. Thus Buddhism functioned from early on with what is almost a contradiction in terms, an "open canon," in which commonly accepted principles of authenticity take the place of a rigidly defined and bounded set of texts in a given linguistic form. The latter would have been well-nigh impossible in any case because Buddhism functioned in a situation of regional and linguistic diversity, with Buddhists living in autonomous self-governing communities.

Form, content, and transmission

Agreement in such circumstances was by consensus, despite occasional attempts by kings and emperors to enforce orthodoxy. Several so-called councils (sa?g?ti, group recitations) are supposed to have been held as the fledgling sa?gha tried to maintain unity on what was to be accepted as the true word of the Buddha or the correct interpretation of the rules of discipline. The first council at R?jagrha took place after the death of the Buddha. At this council, the disciple ?nanda recited the s?tras (discourses delivered by the Buddha, or others accorded equivalent authority), and Up?li recited the vinaya (rules of discipline for renunciants). The community accepted their recounting of these two bodies of texts, with only some monks dissenting.

Yet even this account of a san ? gha relatively united as to what the Buddha had taught may oversimplify history. Later councils (at Vai?al?, P??aliputra, etc.) were occasions for more serious disagreements, which led to the formation of the different nik?yas by sects or schools each recognizing the validity of its own ordination lineage only. In India it appears that each nik?ya came to transmit its own set of sacred texts, initially dividing them into s?tra and vinaya. In some schools, the s?tra and vinaya were supplemented from about the second century b.c.e. onwards by the abhidharma, an even more variable set of texts (seven for the Sarv?stiv?dins, a different seven for the Therav?dins, and so on), which systematized the teachings in terms of the particular categories they fell under. Some schools rejected this third category, but for most the notion of the canon as consisting of the three baskets (tripi?aka) of s?tra, vinaya, and abhidharma became standard. The tripi?aka of one school, as far as scholars know, was never the same as that of the next, although the loss of the literature of most schools makes it difficult to be certain about the extent of difference. Nevertheless, there are certain commonalities. For example, the s?tra-pi?akas were divided into sections (nik?yas, ?gamas) according to such criteria as length, subject, or numerical category (there was also a miscellaneous category, for texts that did not fit any of these). The vinayas were divided into rules for men and rules for women, these being ordered according to the seriousness of the offense, with other sections devoted to particular aspects of community life (ordination, official acts, property, etc.). The resulting collections of texts, which are referred to as canons, were therefore quite varied, extensive, and structurally complex.

One of the primary functions of the Buddhist order was to preserve and transmit all this literature, at first orally, then in writing, from generation to generation, even though Buddhists have always had a keen sense of the fragility of this enterprise. They believe that this effort is bound to fail in the end, due to human weakness, so that the work of a buddha will need to be done over and over again. Different groups of renunciants took responsibility for the transmission of different sections of their school's canon, committing them to memory, although occasionally people with prodigious mental powers mastered the whole canon. One consequence of this "division of labor" is that the same text can occur in two or more different places in a given canon. Oral transmission also led to extreme redundancy and repetition, the same formulas and blocks of text recurring in many different contexts.

From about the first century b.c.e. onward the texts began to be committed to writing, on palm leaf, birch bark, and other materials. This was only partially successful in preserving the texts for posterity, and most have been lost. The only canon to survive in its entirely tirety is that of the Theravadins, written in the P?li language. It shows that some schools kept their scriptures in ancient tongues, but in the extant fragments of other schools' canons it is apparent that a continuous process of Sanskritization was under way. The use of various Indian languages is another sign of the absence of any central authority. In one sense all Buddhist scriptures, even those in P?li, are translations; it is not known what language (s) the Buddha himself spoke, but he is supposed to have sanctioned his followers' use of their own dialects for transmitting his teachings. The Buddhist canon is thus thoroughly multilingual. Parts of the canons of many Indian schools are extant in Chinese or Tibetan translation, as well as in Sanskrit fragments displaying different degrees of regularization from earlier Prakritic or Middle Indic dialect forms to classical Sanskrit. Thus the vinayas of six schools have survived, as well as parts of the s?tra-pi?akas of the Sarv?stiv?dins, the Dharmaguptakas, and the Mah?s?mghikas. Abhidharma texts from various schools, in particular the Therav?dins and the Sarv?stiv?dins, also survive. But while manuscripts continue to be found, the greater part of the Indian Buddhist canons has no doubt vanished forever. Buddhist teachings, which emphasize the inevitability of transformation and loss, have themselves succumbed to it.

Even when it was fully extant, it is unlikely that many Buddhists ever knew their canon in its entirety, as a Muslim might know the Qur'an or a Christian the Bible. The Buddhist scriptures are simply too extensive, so that most members of the order would have been familiar with and used only a small number of them, a functional partial canon as opposed to an ideal complete one. Scholars also believe that Buddhists belonging to different mainstream or ?r?vakay?na schools would have accepted much of what the other groups transmitted as canonical, agreeing on the broad principles, and differing only on particular points of doctrine, and, more importantly, on points of monastic discipline. Some of the most heated disputes in the history of the order were over the vinaya. With the advent of the Mah?y?na, with its prodigious outpouring of new scriptures, the scope for disagreement increased, and the bounds of the Buddhist canon became less distinct. The Mah?y?na canon was even more open than the mainstream one, and followers of that path are in most cases unlikely to have known more than a tiny fraction of the literature it generated. The same is true of tantric Buddhism, with its many

classes of tantras, ritual and soteriological texts, which outnumbered even the Mah?y?na s?tras.

Buddhist canons outside India

The complexity of this picture increased still further when Buddhism spread beyond the greater Indian cultural area. Although the P?li canon of the Therav?dins eventually established itself as the standard in Sri Lanka and Southeast Asia, in Central Asia and China different schools coexisted, and the Mah?y?na orientation was dominant. The Chinese translated scriptures belonging to these different schools and to this new movement with great zeal, the result being that the Chinese Buddhist canon took on a rather different shape. At first the work of bibliographers and cataloguers, later the product of imperial decree, authorized and funded by the court, the Chinese Buddhist canon (Dazangjing, literally "Great Storehouse Scripture") was a far more comprehensive collection. It eventually included Chinese translations of texts from the tripitakas of different Indian schools and of huge quantities of the Mah?y?na s?tras and Buddhist tantras produced in India from approximately the first century c.e. onward, as well as commentaries and treatises, texts written in China, biographies of monks and nuns, lexicographical works, and even the catalogues of Buddhist scriptures themselves. The sheer number and diversity of texts made the use of the tripartite structure of the tripi?aka unfeasible. What is more, the Chinese retained different translations of the same text, often produced many centuries apart, affording modern scholars an excellent view of how texts and translation techniques developed over time.

Thus the Chinese Buddhist canon, which became the standard in Korea and Japan as well, is vast. It has appeared in numerous editions, many of them made with imperial patronage, although the one most often consulted by scholars today is the Taish? shinsh? daiz?ky? (New Edition of the Buddhist Canon Made during the Taish? Reign), published in Japan from 1924 to 1934 in one hundred volumes, each of which runs to about a thousand pages (eighty-five volumes of texts containing 2,920 works, twelve volumes of iconography, and three of catalogues). Yet, immense as it is, the Taish? is not the only edition; many others have survived as well, and thus "the Chinese Buddhist canon" is itself an abstraction of many highly variable collections. This proliferation of editions was in part due to state involvement, as each successive set of rulers sought to legitimate themselves politically as patrons of religion, or aspired for reasons of piety to the merit that the propagation of the buddhadharma generates.

These ideological considerations were instrumental in stimulating the invention and spread of printing technologies in East Asia, long before they were known in the West. Thus the world's oldest printed works are Buddhist texts, and from the tenth century onward the earlier manuscript copies of the Chinese Buddhist canon were replaced by printed editions, first using carved wooden blocks, then movable metal type. The production of these editions required resources that in those days only states could muster, although in recent times wealthy religious and commercial organizations have also become involved.

The same is true of Tibet, where in the fourteenth century the efforts of cataloguers trying to make sense of the sheer diversity of Buddhist texts combined with the interests of political authorities, intent on their own kind of order, to produce the first of many editions of the Tibetan canon, the Old Snar thang. Unlike the Chinese, the Tibetans were generally disinclined to preserve multiple translations of the same text, but their canon (upon which the Mongolian canon is also based) is equally vast. It has two major divisions, the Bka' 'gyur (the Word Translated; i.e., buddhavacana) and the Bstan 'gyur (the Teachings Translated; i.e., commentaries and other treatises). The Bka' 'gyur includes the three subdivisions of vinaya (that of the M?lasarv?stiv?da school), s?tra (predominantly Mah?y?na s?tras, in their various categories), and tantra (also arranged in various classes). The Bstan 'gyur also reflects these categories. The arrangement of all these texts differs according to edition, and sometimes one edition carries works not found in another.

As is the case with the Chinese canon, the Tibetan translations preserve much that is lost in Sanskrit. Some of the most prestigious editions (Peking, Sde dge, Snar thang, etc.) have been mass-produced woodblock prints; others have been manuscript productions, written by hand on expensive papers with ink made of precious metals and enclosed between ornate covers studded with jewels. The resources expended on this activity have been enormous, and the results are objects of great beauty. For Tibetans, as for other Buddhists, the sanctity of the canon derives from the sanctity of the liberating truth it contains and of the person who uttered it, and therefore the scriptures too are the focus of worship and veneration. They are not like any other books, but embody a special power, and must therefore be treated with reverence and respect, in a way similar, but not identical, to the way in which Jews approach the Torah, Christians the Bible, and Muslims the Qur'an.

Canon and canonicity are therefore never the same from one religion to the next, even if common themes can be found. Furthermore, the Buddhist canon turns out to be a large family of collections of texts in different languages and from different places, all sharing descent from a common set of forebears—the divergent oral reports of what the Buddha had taught, which were circulating among his disciples at his death some time in the fifth century b.c.e. Not unitary in content or linguistic expression even at the beginning, it is unimaginably diverse in both respects two and a half millennia later, as it continues to grow with editions and translations into English and other modern languages. At the same time, the Buddhist canon is unified by a common concern with setting out the path to salvation. Just as the waters of the ocean, however vast, have the same taste of salt at any point, so too all

the many teachings of the Buddha have a single taste everywhere, that of liberation. And as to the path by which liberation is attained, Buddhists are fond of quoting the verse (Ud?navarga 28.1):

Not doing any evil, accomplishing what is good,
Purifying one's own mind: this is the teaching of the Buddha.

See also:?gama/Nik?ya; Apocrypha; Catalogues of Scriptures; Councils, Buddhist; Languages; Scripture

Bibliography

Davidson, Ronald M. "An Introduction to the Standards of Scriptural Authenticity in Indian Buddhism." In Chinese Buddhist Apocrypha, ed. Robert E. Buswell, Jr. Honolulu: University of Hawaii Press, 1990.

Grönbold, Günter. Der buddhistische Kanon: Eine Bibliographie. Wiesbaden, Germany: Harrassowitz, 1984.

Harrison, Paul. "A Brief History of the Tibetan Bka' 'gyur." In Tibetan Literature: Studies in Genre, ed. Roger Jackson and José Cabezón. New York: Snow Lion, 1996.

Lamotte, Étienne. "La critique d'authenticité dans le bouddhisme." In India Antiqua (1947): 213–222. Leiden, Netherlands: Brill, 1947. English translation by Sara Boin-Webb, "The Assessment of Textual Authenticity in Buddhism." Buddhist Studies Review 1, no. 1 (1983): 4–15.

Lamotte, Étienne. "La critique d'interprétation dans le bouddhisme." Annuaire de l'Institut de Philologie et d'Histoire Orientales et Slaves 9 (1949): 341–361. English translation by Sara Boin-Webb, "The Assessment of Textual Interpretation in Buddhism." In Buddhist Hermeneutics, ed. Donald S. Lopez, Jr. Honolulu: University of Hawaii Press, 1988.

Lancaster, Lewis. "Buddhist Literature: Canonization." In The Encyclopedia of Religion, ed. Mircea Eliade, Vol. 2. New York: Macmillan, 1986.

Norman, K. R. P?li Literature: Including the Canonical Literature in Prakrit and Sanskrit of All the Hlnayana Schools of Buddhism. Wiesbaden, Germany: Harrassowitz, 1983.

Pagel, Ulrich. "Buddhism." In Sacred Writings, ed. Jean Holm. London and New York: Pinter, 1994.

Ray, Reginald. "Buddhism: Sacred Text Written and Realized." In The Holy Book in Comparative Perspective, ed. Frederick M. Denny and Rodney L. Taylor. Columbia: University of South Carolina Press, 1985.

von Hinüber, Oskar. A Handbook of Pali Literature. Berlin: de Gruyter, 1996.

Paul Harrison