When Yinshun reorganized the Chinese Saṃyukta Āgama and saw that it fit the headings he found in the Yogâcārabhūmi, he also noticed that it could be divided into three divisions with distinct formats. Those three divisions appear to match the sutra formats of sutra, geya, and vyākaraṇa as defined by Asaṅga in his commentary on SĀ that’s contained in the Yogâcārabhūmi.
The first three aṅgas are defined by Asaṅga in this way:
- Sutras discuss Dharma topics generally to an assembly of monks.
- Geyas are sutras that have concluding or summary verses at or near the end.
- Vyākaraṇas are sutras in which the Buddha or a disciple gives an explanation, such as what happened after a disciple passed away or when someone comes with a question.
Yinshun noticed that Vargas I-IV of SĀ are mostly Sutras, Varga V contains Geyas, and Vargas VI-VII are vyākaraṇas, generally speaking. They all have outliers, but the majority can be put into those categories to make three aṅga divisions.
He came to the conclusion, then, that Asaṅga was a reliable source about the way SĀ was compiled in northern India. From this, Yinshun theorized that the sutra aṅgas, at least the first three, might have been the way the sutras had been originally compiled in the earliest version of the canon. As the sutra canon grew, the Saṃyukta Āgama took shape. As SĀ continued to grow, texts were split off into the other Āgamas (Dīrgha, Madhyama, and Ekôttarika), but then the tradition stopped at these four.
The clearest parallel of these three divisions in SN is the Geya division. It’s a bit complicated, but if you look at the chart I created at the beginning of this thread, you can see that Varga IV, “Eight Assemblies,” matches the saṃyuttas of SN’s Sagātha Vagga. The only outlier is the Bhikkhu saṃyukta, which is located in the Nidāna Vagga. However, when we look at the contents of that saṃyutta, we see that many of its suttas are geya sutras; that is, they have concluding verses. So, it stands to reason that it may have originally been in the Sagātha Vagga, but was moved and a couple prose-only suttas were added to the beginning of it.
Vargas I-IV have corresponding vaggas in the Theravada Samyutta Nikaya (Vagga II, III, IV, and V), but the parallels to the samyuktas in Vargas VI-VII are scattered throughout those same vaggas. So, if they originally had been two separate divisions in the Theravada tradition, they were combined at some point and somewhat haphazardly.
The basic takeaway is that the Theravada Saṃyutta Nikāya seems to have been jumbled in a random order after the Sagātha Vagga, perhaps in a similar way that the Chinese SĀ had been disordered after its translation. This appears to have been a problem that cropped up at some point in Buddhist history, and that would explain the relatively random disorder of the suttas as they’ve come down to us, if it’s true. It’s a little hard to believe they were randomized from the start, though it isn’t impossible, I guess.