Twelve-membered speech of the Buddha: Preliminary note

From the evidence of the philosopher and commentator Buddhaghosa, the speech of the Buddha, such as it was presented in Ceylon in the 5^th century of our era, was the object of seven different classifications. They are listed in the Samantapāsādikā, p. 16; the Sumaṅgalavilāsinī, p. 15; and the Atthasālinī, p. 18):

Tad etaṃ sabbam pi Buddhavacanaṃ rasavasena ekavidhaṃ, dhammavinayavasena duvidhaṃ, paṭhamamajjhimapac chimavasena tividhaṃ, tathā piṭakavasena, nikāyavasena pañcavidhaṃ, aṅgavasena navavidhaṃ, dhammakkhandavasena caturāsītisahassavidhan ti veditabbaṃ.

“It should be known that the Buddha’s speech is single in its taste, twofold because of the Dharma and the discipline, threefold because of the initial, intermediate and final (words of the Buddha), also threefold because of the baskets, fivefold because of the collections, ninefold because of the members (aṅga) and finally 84,000–fold because of the articles of the Dharma.”

The earliest texts mention a classification of the scriptures into members or aṅgas. These aṅgas are not literary genres but simply composition types in respect to form (e.g., prose or verse) or content (e.g., sermons, predictions, stories, conversations, commentaries, etc).

The major drawback of this classification is that, far from being mutually exclusive, the aṅgas overlap one another. Thus a sūtra is also a geya if it contains verse, a gāthā if it is expressed in stanzas, an udāna if it includes exclamations, an ityuktaka if it begins or ends with certain stereotyped formulas, a jātaka if it tells about previous lifetimes, a vyākaraṇa if it contains explanations or predictions, etc.

The Pāli sources list nine members: the Sanskrit-Chinese sources usually mention twelve; the classification into nine members is called Navāṅga, the classification into twelve members, Dvādaśāṅga.

A. Hirakawa has dedicated a masterful study to the Dvādaśāṅga in his work Shoki daijō no Kenyū (Study on the early Mahāyāna), 1968, p. 721–753. Previously he had condensed his ideas into an article entitled The Rise of Mahāyāna Buddhism in Memoirs of the Research Department of the Tōyō Bunko, no. 22 (1963), p. 61–65.

In the canonical sources, whether they are nine or twelve in number, the aṅgas are set out without any explanation. They are supposed to include the entirety of the Buddhist scriptures, but they could also be applied to any other literature, sacred as well as profane. This type of classification is not necessarily Buddhist in origin and could be derived from literary concepts widespread in the early centuries of Buddhism. This would explain the hesitation that commentators will always feel when they have to define any aṅga in particular.