Appendix One: History

On the history of the Udāna

Passages in the Canon mention udānas in a list of nine genres in which the Buddha’s teachings and events in his life were memorized during his lifetime. There has been some speculation as to whether the Udāna we currently have is in any way related to the udānas mentioned in the list. The general consensus is that most of the exclamations at the end of our current udānas might possibly date from the Buddha’s time, but that the stories are obviously a much later invention. There is, however, no proof for either position.

The question of how to prove through textual analysis whether the exclamations actually came from the Buddha is essentially uninteresting: There is no historical evidence to prove or disprove that anything in the Canon came from the Buddha; the only way to test the value of what the Canon contains is to put its teachings to the test.

It is interesting, however, to examine the arguments for assigning a late date to the stories in the Udānas, for when we examine these arguments we find that they teach us more about the assumptions of the people who present them than about the Udāna itself.

The arguments fall into two main classes: those based on the form of the text, and those on the content.

The first formal argument for the lateness of the stories is based on the fact that a fraction of the exclamations occur elsewhere in the major poetry anthologies of the Canon without any connection to the stories in the Udāna. I have noted some of these parallels in the notes. Here they are as a list:

1:9 (last half) = Dhp 393 (last half)

2:3 = Dhp 131–132

3:4 = Thag 14:1 (PTS v. 651); Thag 17:2 (PTS v. 1000)

3:6 (first part) = Sn 3:4 (PTS v. 469) (first part)

4:3 = Dhp 42

4:4 = near equivalent, Thag 2:36 (PTS vv. 191—192)

4:6 = Dhp 185

4:7 = Thag 1:68

4:9 = Sn 3:12 (PTS v. 746)

5:5 = Thag 6:13 (PTS v. 447)

6:3 (last half) = Thag 2:30 (PTS v. 180 — last half)

6:7 (first half) = Sn 1:1 (PTS v. 7 — last half)

7:4 (first part) = Thag 4:8 (PTS v. 297)

In addition to these parallels in the Pali Canon, variants of all the exclamations also appear, again without stories, in the Udānavarga, a compilation of verses in Buddhist Hybrid Sanskrit made by a Sarvāstivādin monk, Dharmatrāta, somewhere between 75 B.C.E and 200 C.E. From these facts the argument maintains that because these verses occur without a story in some places, but with a story in the Udāna, the stories must be later additions.

The problem with this argument is that it ignores the possibility that the compilers of the other texts might have had a different purpose than the compilers of the Udāna. They could have simply been interested in creating anthologies of verses shorn of any narrative frameworks. What’s ironic is that this latter point is sometimes used to “prove” the lateness of the Dhammapada as a collection: Some of the Dhammapada verses occur associated with stories in other parts of the Pali Canon, and scholars have argued that the compilers of the Dhammapada simply lifted the verses from those passages and dropped the stories.

Because this argument can be used either way–to prove that collections of verses with stories are later than collections of verses without stories, and that collections of verses without stories are later than collections of verses with stories–it doesn’t really prove either position.

A second set of arguments for the lateness of the Udāna is based on the fact that thirteen of the udānas–story and verse–also occur elsewhere in the Canon:

1:1 – Mv.I.1.1—3 (different details in the story)

1:2 – Mv.I.1.4—5 (different details in the story)

1:3 – Mv.I.1.6.7

1:4 – Mv.I.2

2:1 – Mv.I.3

2:10 – Cv.VII.1.5—6

4:5 – Mv.X.3—4 (different details in the story)

5:5 – Cv.IX.1, AN 8:20 (no verse)

5:6 – Mv.V.13.1—10 (different details in the story; the verse is also different in the PTS but not in the Thai, Sri Lankan, or Burmese editions)

5:8 – Cv.VII.3.17

6:1DN 16, SN 51:10, AN 8:77

8:5DN 16 (includes an intervening incident not narrated in 8:5)

8:6 – Mv.VI.28, DN 16

What’s notable about these parallels is that seven of the thirteen occur in the two longest biographical accounts in the Canon: the account of the events leading from the Buddha’s awakening to his gaining his two foremost disciples in Mv.I; and the account of his last year in DN 16. From this fact, scholars have argued that these udānas were simply lifted from these longer accounts. Because DN 16 is regarded as a late document, this would mean that the compilation of the Udāna was even later. Sometimes this argument is bolstered with another one: that originally the monks compiled a continuous narrative of the events in the Buddha’s life, but for some reason the narrative was chopped up into the bits and pieces we now find in the early canons, whereas the narrative as a whole was forgotten or lost.

This latter argument, of course, is purely conjectural, based largely on the assumption that the early monks would have had a modern Western desire for a complete biography of their teacher. And it begs the question, why would the monks have thrown away a perfectly good continuous narrative if they had had one?

As for the preceding argument, it is belied by two facts. The first is that the longer narratives containing parallels to the udānas lack a sense of flow. If anything, they read as if they were stitched together from preexisting materials, the udānas being among them. Second, one of the udānas with a parallel in DN 168:5–is composed in a style called campū, in which the narrative is told in both prose and verse. This style is rarely used in the Canon. If it was originally part of DN 16 before being chopped off into an udāna, we would reasonably expect that the rest of DN 16 would also be composed in this style. But it isn’t. The events in 8:5, together with an intervening incident not included in 8:5, are the only parts of DN 16 narrated in the campū style. This suggests–even though it doesn’t prove–that 8:5 was composed separately before it was included in DN 16.

Thus this second set of formal arguments proves nothing about the relative earliness or lateness of the Udāna.

A third set of formal arguments is based on the fact that neither the Chinese nor the Tibetan canons contain any text corresponding to the Udāna. Both canons contain versions of Dharmatrāta’s Udānavarga mentioned above, with stories relegated to commentaries on the verses. Because these canons treat the stories as later additions, it has been argued that the Udāna in the Pali Canon was composed after the schools represented in the Chinese and Tibetan canons split off from the Theravāda.

This argument, however, is based on the assumption that these other two canons contain complete accounts of what was available in India at the time they were compiled. However, the Tibetan canon contains very few “Hīnayāna” texts, as its compilers were much more interested in the later vehicles, so the lack of the Udāna in this compilation proves nothing.

Similarly, the collection of “Hīnayāna” texts in the Chinese canon, while more complete than that in the Tibetan, is still fairly haphazard. There is no complete canon from any of the early schools; the different nikāyas (or āgamas as they are called in the Chinese collection) apparently come from a variety of early schools. And the Chinese canon itself was a late attempt, during the early Sung dynasty, to gather whatever texts, through happenstance, had made their way to China and into Chinese translation by the time of the T’ang dynasty and had survived into the Sung.

We do know that some of the texts brought to China during the T’ang are not in the collection. In 645, the pilgrim-monk Hsüan-tsang returned to the Chinese capital with a hoard of more than 675 Buddhist texts, many of them new to China, that he had acquired during a long overland trip to India. The emperor at the time was impressed with Hsüan-tsang’s achievement and provided him with the resources needed to set up an expert board of translators. However, most of the texts were never translated. After Hsüan-tsang’s death and the death of the emperor, the emperor’s successor, who had no interest in Buddhism, disbanded the board of translators and sequestered the texts in the imperial library, where they were eventually lost or destroyed. We know that at least one “Hīnayāna” text was in Hsüan-tsang’s hoard, as there is a partial Itivuttaka among his translations. There may have been other similar texts as well.

Thus the lack of an Udāna in the Chinese canon does not prove that the Udāna as we have it was a late text. It simply may not have attracted the attention of the Chinese or their Central Asia teachers; or it may have been in Hsüan-tsang’s collection but later lost.

So none of these three sets of arguments from the form of the text prove anything about the relative earliness or lateness of the Udāna we now have.

The two main arguments from content are similarly inconclusive.

The first of these arguments is that the stories of the Udāna contain too many supernatural elements to be genuine. This, of course, assumes that people close to the Buddha witnessed no supernatural events surrounding his teaching–an assumption that derives less from any knowledge about what actually happened in the Buddha’s life, and more from a modern discomfort with the supernatural. Stories about people levitating, using clairvoyant powers, etc.–unless presented as fiction–offend modern materialistic sensibilities. Modern scholars like to assume that the Buddha and his early disciples shared these sensibilities, and that religious and supernatural elements could have been added to the texts only after those who had directly known the Buddha had passed away. Yet the Canon consistently shows the Buddha to have been an opponent of materialism (DN 2; AN 3:138; MN 60). Thus there is no reason to assume that he or his direct disciples would have been disinclined to believe or report what, from a materialistic perspective, would count as supernatural powers or events.

As for the actual possibility of such powers, modern science has yet to disprove that they exist. Recent advances in sub-atomic physics and the study of the dynamics of complex non-linear systems show that the physical world is much stranger and less deterministic than the materialistic linear sciences of the eighteenth and nineteenth centuries would have allowed for, and yet a lot of modern thinking outside of the sciences has not caught up with this fact.

At the same time, there is no way that a fair assessment of the powers attainable through meditation could be made by anyone who has yet to master meditation. It would be like a future race of philosophers trying to assess modern scientific discoveries without having mastered the scientific method themselves. Just because there is no room in one’s philosophy for a particular skill doesn’t mean that such skills can’t be acquired. Why make the limitations of one’s imagination the measure of the actual world?

The second argument for the lateness of the Udāna based on its content focuses not on the stories, but on some of the exclamations themselves: those in 8:1, 8:3, and 8:4 describing unbinding as a dimension that can be experienced. The argument is this: Because the consciousness-aggregate ends with the attaining of unbinding, any description of unbinding as a dimension that can be experienced is suspect. Therefore these exclamations must be later additions to the Canon.

This argument is based on the assumption that there can be no consciousness outside of the consciousness-aggregate, inasmuch as the definition of that aggregate concludes with the phrase that it includes, “all consciousness, past, present, and future” (SN 22:59). However, the Buddha elsewhere limits the term “all” to what can be known in conjunction with the six senses (SN 35:23). And there are other passages, aside from these passages in the Udāna, indicating that there can be something known outside of the six senses (DN 11, MN 49, SN 35:117). DN 11 and MN 49, in fact, refer to this awareness as “consciousness without feature” or “consciousness without surface” (viññāṇaṁ anidassanaṁ). Because this consciousness lies outside of the dimension of time–it’s akālika–it is neither momentary nor eternal, and cannot be labeled as past, present, or future, for all such concepts have meaning only within the dimension of time. Thus it lies outside the definition of the consciousness-aggregate, and would not be ended when that aggregate ceases, either in the experience of awakening or after the death of the arahant.

(Another argument that there can be no consciousness outside of the consciousness-aggregate is based on a mistranslation of MN 38. Because the argument is technical, I will omit it here. If you are interested, you can find it in the notes to my translation of that discourse and in Skill in Questions, chapter 5, §72, note 2.)

All of this means that the exclamations in 8:1, 8:3, and 8:4 do not conflict with the rest of the Canon. In fact, Iti 43 also contains the exclamation in 8:3, and MN 144 and SN 35:87 cite the exclamation in 8:4 as a teaching of the Buddha. So there is no reason to dismiss these passages as late.

All of which means that the arguments for the lateness of the Udāna–whether based on form or on content–have yet to provide any compelling reason to regard the Udāna as a late addition to the Canon.