This week, I’ve moved on to editing a new translation of DĀ 9 The Gathered Sangha (Saṅgīti). This sutra is remarkable in comparison to its Pali parallel mainly in that it shows much less expansion, both in the opening narrative telling us how Śāriputra came to recite the mātṛka of teachings and in the mātṛka recitation itself.
Noticing this, I decided that it would make a good example to illustrate historical evidence of Buddhist texts expanding over time, since the expansion is exaggerated compared to most EBTs. While I don’t have access to a Sanskrit edition of the Saṅgīti Sutta that might exist hidden away somewhere in the academic literature, I can draw on two other editions in Chinese.
The four editions I’ll use for this essay series are:
- Dharmaguptaka (ca. 412 CE): DĀ 9 - Buddhayaśas & Chu Fonian’s translation
- Sarvâstivāda (ca. 660 CE): T 1536 - Xuanzang’s translation of the Saṅgītiparyāya
- ?? (ca. 10t-11h c. CE): T 12 - Dānapāla’s translation from an unknown tradition
- Theravada (ca. ?): DN 33 - Pali original
I’ve placed these editions in historical order, assuming that the most expanded edition of the sutta (DN 33) is likely the latest version, though we have no historical evidence of when it reached its present state (to my knowledge).
We should note, however, that historical age doesn’t in itself tell us how expanded a given edition is. The latest Chinese translation by Dānapāla has the smallest mātṛka list of the four, being just a little smaller than DĀ 9’s list. Simpler versions of this sutta still existed as late as the Song dynasty in China.
To start the comparison, we can get a basic idea of just how much difference there is between these editions by counting the number of items found in the mātṛkas they recite:
Section | DĀ 9 | T1536 | T 12 | DN 33 |
---|---|---|---|---|
Ones | 2 | 2 | 1 | 2 |
Twos | 12 | 27 | 1 | 33 |
Threes | 37 | 50 | 34 | 60 |
Fours | 36 | 49 | 39 | 50 |
Fives | 15 | 24 | 14 | 26 |
Sixes | 14 | 24 | 15 | 22 |
Sevens | 7 | 12 | 6 | 14 |
Eights | 4 | 10 | 4 | 11 |
Nines | 1 | 2 | 1 | 6 |
Tens | 1 | 2 | 1 | 6 |
Total | 129 | 202 | 116 | 230 |
Thus, the Sarvâstivāda and Theravada editions are both quite large compared to the Dharmaguptaka and the Song translation. Especially notable is that the Theravada edition is the largest of the group, making it difficult to imagine that it represents an early version of the text. Rather, like the Chinese Sarvâstivāda version that can be extracted from its Abhidharma commentary, DN 33 represents a fully developed edition likely dating from the late Abhidharma period of Buddhism.
Why would there be so much difference in these editions of an oral tradition text? Most likely, the issue arises because the Saṅgīti Sutta lacked a controlling structure that prevents material from being inserted. As long as new lists contain the right number of items, each section of its mātṛka could grow or be trimmed down without much ado (as long as the tradition in question doesn’t compare its version to that of others).
By comparison, the Dasuttara Sutta uses a structure that makes adding more material difficult. Each section of its mātṛka has ten questions to which each list is an answer. If someone were to add an additional list to a question, it would be incongruent to the rest of the sutta. Similarly, if someone were to add an eleventh question to one of the mātṛka sections, it would be an obvious corruption. So, the differences we see in the parallels of the Dasuttara Sutta are matters of list replacements or additional elaboration in the lists themselves. On the whole, the parallels match each other quite closely and serve as a testament to the value of repetitive structures. Not only did they aid in memorization, they also preserved the texts by making changes easier to detect.
That said, there is a sutra among the Dasuttara Sutta’s parallels that represents a significant change. DĀ 11 cuts the Dasuttara Sutta in half by truncating each section of the mātṛka to the first four and last question. Besides this reduction of the material, DĀ 11 matches the Dasuttara Sutta quite closely, making it fairly clear that either DĀ 11 is an abbreviation or that the Dasuttara Sutta is an expansion of a smaller sutta. The abbreviation theory seems more reasonable because DĀ 11 is the only extant evidence of a “Dasuttara By Half” Sutta.
Before looking at the contents of the mātṛkas that we find in the Saṅgīti Sutta’s parallels, though, I’d like to first take a look at the narratives that introduce them. The stories themselves differ in size and elaborateness and offer evidence of a conscious expansion of the sutta by the sectarian traditions.
After a comparative analysis of these stories, I’ll then lay out the sections of the parallel mātṛkas in more detail.