History of the Pali Canon (older version)
- For criticism see Criticism of History of the Pali Canon (older version)
- 1 Oral tradition
- 2 In writing
- 3 In print
- 4 In translation
- 5 The digital age
- 6 The future
- 7 References
Theravada tradition holds that the Canon was written down from oral tradition in Ceylon in the reign of king Vattagamani. The exact dates of early Sinhalese rulers remain a matter of debate among historians, but they agree that he ruled some time in the last century BC. They also seem to agree with the tradition as regards the writing down of the Canon, though some adopt cautious wording. Where they disagree with it, and each other, is on the evolution of the Canon in the centuries from the Buddha's time (which most recent specialists date around the 5th century BC) to then.
On the relation between the Canon and the original teaching of the Buddha, three different approaches among scholars have been identified. The first of these argues that at least large parts of the first four nikayas show such coherence they must represent in substance the work of a single mind, that of the Buddha himself.
The second argues that there is very little hard evidence so very little can be known.
The third concentrates on detailed examination of particular points. Scholars following this approach reach a variety of different conclusions. Some hold that the Suttanipata is the oldest part of the Canon.
See  for a list of writings by scholars on this and related topics.
As noted above, the Canon was written down in the last century BC. However, the climate in Theravada countries is not conducive to the survival of manuscripts, so those available to modern scholars are much more recent. As the exception that proves the rule, a 2-page fragment from the 8th or 9th century has been found in Nepal. As there is no record of Theravada there before modern times, it is assumed that a monk fleeing the Muslim invasion of India around 1200 must have taken it there. (Similarly, in the 13th century there is a sudden wave of quotations from North Indian Buddhist texts in literature written in Ceylon.) Apart from this, and a few brief quotations in inscriptions, there is nothing before the 15th century, and little before the 18th. Thus the surviving manuscripts are the result of an unknown, but substantial, number of successive copyings. As a result, the evolution of the Canon in this period is still somewhat obscure, though of course nothing like as much so as in the oral period.
Thus some scholars adopt a similarly sceptical attitude to that mentioned above, holding that we know nothing for sure about the relation between the Canon as written down and that existing nowadays. Others claim that it has changed little since then. Others again claim that certain books were added later: the Parivara, Niddesa and Patisambhidamagga are mentioned by various scholars.
One complicating factor is the fact that Pali has no alphabet of its own. Everyone (including us) writes it in their own. The transmission of the Canon between countries therefore involved transcription from one alphabet to another, with extra opportunities for error. Partly as a result of persecution by Western invaders from around 1500 on, parts of the Canon were lost in Ceylon and had to be reintroduced from Southeast Asia.
One peculiar "manuscript" should be mentioned here. Between 1861 and 1868, the entire Canon was inscribed on 729 marble slabs, each about five feet by three. This text was formally approved by the Fifth Council in 1871, and the slabs are on display in the Kuthodaw Pagoda in Mandalay, where tourist guides call them "The World's Largest Book".
A number of significant discrepancies between different editions of the Canon, or between the Canon as we have it now and what the commentators seem to have had before them, or between what different commentators seem to have had before them, imply that there have been more changes in the Canon than mere copying errors. Below, some of these and other interesting points are discussed.
The contents of the Canon
The introductions to the commentaries on the three pitakas (mainly written in the 5th century) give various lists of books in the Canon, which also vary somewhat between editions (see Arrangement of the Pali Canon). Scholars tend to conclude that these lists represent different stages in the evolution of the Canon. The subcommentary on one of the commentaries, probably written in the 10th century, explains the apparent differences in some lists by saying that books not mentioned were in fact counted as parts of other books. At least by about 1800, some authorities in Burma were using the same method to include in the Canon books never before mentioned as such. The head of the Burmese sangha at that time included at least the Netti and Petakopadesa. This may also explain the absence of the Netti, its commentary and its subcommentary from a list of books donated to a Burmese monastery in 1442.
The Netti is included in the Burmese Phayre manuscript of the Canon, dated 1841/2, where it is bound in one volume with the Patisambhidamagga. The Netti, Petakopadesa and Milindapanha are included in the inscriptions of the Canon approved by the Burmese Fifth Council.
Some editions include a series of verses at the end, others omit it.
The 10th of the 152 discourses in the Sinhalese and Thai editions is a shorter version of the Maha Satipatthana Sutta of the Digha Nikaya. In the 6th Council edition, the full-length version is included instead. Even the Burmese edition of the commentary seems not to notice the extra material.
In the Valahaka Vagga of the 4th book, the Sinhalese, Burmese and Thai editions all include a note saying that the 6th discourse is mentioned in the commentary but not found in any of their manuscripts. This suggests a very thin line of transmission for a major part of the Canon, with presumably either only a single, damaged manuscript, or a small number with substantial gaps overlapping here, at some point in time.
According to the early commentaries there are 82 udanas, but the present-day text and Dhammapala's commentary have 80.
According to the early commentaries there are 110 itivuttakas, but the present-day text and Dhammapala's commentary have 112.
In the Maha Supina Jataka the commentary says certain verses are not canonical, yet they appear in the main editions of the Canon.
Some jatakas appear in different orders in different editions.
According to the major commentaries there are 550 jatakas, and some mediaeval sources give a list of them, but only 547 are found in the surviving canonical text and commentary.
This book is divided into 4 parts:
Parts 1 and 2 comprise a single apadana each, and part 4 40.. Part 3, in the 6th Council edition, has 561 apadanas, in the Thai edition 550, and the Pali Text Society edition 547; the commentary covers 561. Dr Cutler has suggested that only 547 actually survived in manuscript transmission, while the rest have been reconstructed in modern times. She also has suggested that the missing ones were deliberately "lost" in order to bring the number down to 547, to match the Jataka.
Dhammapala's commentary on the Theragatha and Therigatha quotes many apadanas, but apparently from a different version from the present-day book. This commentary is some centuries older than that on the Apadana itself, suggesting the book may have changed substantially in the interim. Alternatively, as suggested by Hinüber, Dhammapala may have been following a South Indian recension, over which the Ceylon version prevailed
The text covered by the commentary is noticeably different from that in the Canon we have now. In particular, the commentator seems unaware of the last chapter.
Passages quoted in the commentary on the Jataka differ noticeably from the corresponding passages in the present-day text. Hinüber suggests the Jataka commentary used the Ceylon recension, but Dhammapala's commentary on the South Indian recension ensured it prevailed.
The translator says the text is very corrupt. Parts are in the wrong order, part is missing, a page of another book has got mixed in by mistake, and there are numerous smaller errors.
Variations between editions are sometimes systematic. This is essentially impossible for copying errors. Either someone has gone through "correcting" the text or it has been expanded from a condensed form.
The first complete printed edition of the Canon appeared in Burma in 1900 in 38 volumes. It was based on the Kuthodaw slabs.
This has now been superseded by the edition of the text recited at the Sixth Buddhist Council in Rangoon. The council sat from 1954 to 1956, and each volume was issued once the recitation reached the end of that part of the text. It is in 40 volumes. It includes the Netti, Petakopadesa and Milindapanha.
There are some interesting differences between printings, indicating disputes among Burmese Buddhists. For example, an edition of Samyutta Nikaya, volume 3, published in 1977, has an appendix headed Pāṭhabhedā, "Differences of readings": it lists 3 variant readings for the volume, continuing
[this] triad of readings exists, just this is more appropriate here
In the above case the text was not altered, but in another it was. The council expanded some elisions in the Patthana.  A Burmese edition of Patthana, volume 5, published in 1983 has the following "viññāpanaṃ" (note or notice) at the beginning:
In the 4th and 5th volumes of the Patthana book, some enumeration sections were added by the revising elders at the 6th Council, but we consider them interpolations, and they do not exist both in the 5th Council root [text] and in the Sinhalese, Siamese and English foreign books. Therefore the enumeration sections added by them should be seen as removed here.
This practice seems to have started only a few years after the council, and some printings omit both the "interpolations" and the note saying the text has been changed. The first electronic transcript of the 6th Council edition seems to have followed such an abbreviated printing, which was the, or a, reason for the publication of the version sponsored by the Supreme Patriarch of Thailand (see below).
The first complete printed edition of any book of the Canon was probably Viggo Fausbøll's edition of the Dhammapada, with Latin translation, published in 1855. He also produced an edition of the Jataka, with the first volume appearing in 1877. The first volume of Oldenberg's edition of the Vinaya appeared in 1879, but the process of producing Western editions of the canonical books got organized with the founding of the Pali Text Society in 1881. It completed its edition of the Canon in 1927 with the appearance of the 2nd volume of the Apadana. The above editions of the Jataka and Vinaya form parts of it, and were reprinted by the PTS when their original publishers declined to reissue them. (The PTS produced its own Dhammapada edition.) Since then a few volumes have been replaced by updated editions. The complete set is currently available in 57 volumes, including index volumes.
An edition of most of the Canon was issued in 1893 for the silver jubilee of King Rama V, in 39 volumes. It is speculated that its incompleteness was simply due to the failure to meet the deadline for the remaining material.
A complete edition appeared in 1925-8, in 45 volumes.
Recently, the Dhamma Society Fund, sponsored by the Supreme Patriarch, has produced a Latin-script transcript of the Sixth Buddhist Council's edition. This includes the Netti, Petakopadesa and Milindapanha, which the 1920s edition does not.
A longer-term project is that of the Bhumibalo Foundation. Generally, they take a particular text and prepare editions and Thai translations of it and its commentary and subcommentary in parallel.
After the Sixth Buddhist Council an Indian edition was prepared, mainly based on the council's text, but omitting the Netti, Petakopadesa and Milindapanha. It was in 41 volumes and appeared from 1957 to 1961. This edition gradually went out of print.
More recently, a full transcript of the council's edition has been produced.
The Khmer edition of the Canon was a parallel-text one, i.e. with Khmer translation on facing pages, in 110 volumes. The work of translation greatly slowed down the publication, which took from 1931 to 1969. The Khmers Rouges then burnt every copy in the country, with only a handful surviving elsewhere. After the Vietnamese "liberated" the country, a Buddhist centre was set up in Phnom Penh. It was unable to obtain a set of the native edition until one was donated by the Catholic Missionary Society. This edition includes the Nettippakarana, Petakopadesa and Milindapanha.
Although all the individual books of the Canon had been published in separate editions, no collected edition of the Canon appeared until after the Sixth Buddhist Council. This edition was a parallel-text one like the Cambodian, and so likewise took decades. The complete edition comprises 58 volumes. It includes the Nettipakarana and Petakopadesa, but not the Milindapanha.
Theravada Buddhism, like Judaism, Islam, Hinduism amd Sikhism, but unlike Mahayana Buddhism and Christianity, stresses the importance of the original scriptural language. Nevertheless, translations into the vernacular languages have existed for a long time. The only outside language into which the whole Canon has been translated is Japanese.
In the West, the first translation of a complete book of the Canon was Fausbøll's Latin version of the Dhammapada in 1855. The first such in English was Childers' Khuddakapatha in 1869. The Pali Text Society undertook publication of translations, including reprints of works originally published by others. Its set of translations currently runs to 43 volumes, the oldest dating from 1895. Between them they represent about 3/4 of the Canon.
The digital age
The 2nd Siamese edition of the Canon was transcribed electronically by BUDSIR (Buddhist Studies Information Retrieval) and made available in 1988 in WORM format, and later on CD-ROM. It went online in 2002, though there currently seem to be problems accessing the site.
The 6th Council's edition was transcribed by the Vipassana Research Institute onto CD-ROM issued in India in 1977. This transcript is also now available online.
The Sri Lanka Tripitaka Project transcribed the Buddha Jayanti edition, making it available in 1994. From 1999 it was hosted by the online Journal of Buddhist Ethics, but they declined to carry it over when they moved to a new website. It is not clear where, if anywhere, the files currently are. There are at present at least 4 sites hosting incomplete selections, none including the Puggalapaññatti.
The Dharmakaya Foundation in Thailand produced a CD-ROM transcript of the Pali Text Society's edition in 1996. This is no longer available, but a 2nd edition is being prepared.
More recently, the Dhamma Society Fund, sponsored by the Supreme Patriarch of Thailand, produced a new online transcript of the 6th Council's edition, though this now seems to have been taken down, and a complete printing has been made available in pdfs on a Spanish website.
Summary of the current situation:
- the 6th Council edition is available on the internet in 2 versions, both accessible free of charge and authorizing copying free of charge
- the Thai edition is available on CD-ROM at a price; it is supposed to be available online, but there seem to be problems accessing the site
- one of the files for the Sinhalese edition seems to have been mislaid, and the rest are scattered among 4 sites
Summary of the summary: the 6th Council edition is dominant at present.
Details can be found in Editions of the Pali Canon.
Could more material be added to the Canon? Scholars disagree (maybe Theravadins do). According to Professor von Hinüber, the Khuddakanikaya always remained open for additions, but Professor Collins says the Canon is closed.
- Ruegg & Schmithausen, Earliest Buddhism and Madhyamaka, Brill, Leiden, 1990, pages 1f; Williams, Buddhist Thought, Routledge, 1st ed, 2000, pages 32f/2nd ed, 2012, pages 23f; Anderson, Pain and Its Ending, Curzon, 1999, page 17
- Nakamura, page 46
- Routledge Encyclopedia of Buddhism, 2007, page 584
- Hinüber, page 20
- Hinüber, page 59; Nakamura, page 48
- Hinüber, page 60
- JPTS, volume XXVIII, pages 61f
- The list is given in an appendix to Bode, Pali Literature of Burma, Royal Asiatic Society, London, 1909
- JPTS, 1882, page 61
- Bollée in Pratidanam (Kuiper Festschrift), Mouton, Paris/the Hague, 1968, pages 493-9
- Hinüber, page 55
- JPTS, volume XX, page 1
- JPTS, volume XX, page 36
- JPTS, volume XX, page 36
- JPTS, volume XX, pages 36f
- JPTS, volume XX, pages 23f
- page 61
- Norman, page 93
- Story of Gotama Buddha, Pali Text Society, 1990, page viii
- pages 63f
- PTS edition, page 356
- PTS edition, page 352
- e.g. Duka Patthana, PTS edition, pages 150ff
- Cambridge University Library, call number 834:01.b.34.9, page 438
- ... pāṭhattayaṃ atthi, etad ev' ettha yuttataraṃ
- Conditional Relations, vol II, Pali Text Society, 1981, page x
- Cambridge University Library, call number 834:01.b.34.29
- "Paṭṭhānapotthakassa catutthapañcamabhāgesu keci saṅkhyāvārā chaṭṭhasaṅgītiyaṃ paṭivisodhakattherehi pakkhittā, te pana 'adhikā' 'ti maññāma, pañcamasaṅgītimūle c'eva sīhaḷasyāmaiṅgalisavidesikapotthakesu ca n'atthi. Tasmā tehi pakkhittasaṅkhyāvārā idha apanītā 'ti daṭṭhabbā."
- for most of this section, see the Routledge Encyclopedia of Buddhism, 2007, pages 288-90
- page 76
- JPTS, volume XV, page 91
- Hinüber, Handbook of Pali Literature, de Gruyter, Berlin, 1996
- JPTS: Journal of the Pali Text Society
- Nakamura, Indian Buddhism, Kansai University of Foreign Studies, Hirakata, Japan, 1980; reprinted by Motilal Banarsidass, Delhi, 1987, 1989
- Norman, Pali Literature, Harrassowitz, Wiesbaden, 1983Template:2010 madeTemplate:2011