Early Chinese Buddhism - A big picture

I’m looking for recent publications which give us a big picture of Chinese Early Buddhism vs. Pali Early Buddhism.

I’m looking specifically for takes on several questions:

  • How do the Agamas compare with the Nikayas (number of suttas, vaggas, complete parallels)
  • How do the direct parallels compare on a word-level? Are they more-or-less direct equivalents, or are the core concepts and the structure the same but differ in the formulas and details?
  • What are the conclusions regarding the bhanakas and transmitters of the two text traditions: did they have the same source material? which agamas show the most/least similarities? what can be concluded regarding the transmission process of Early Buddhism?

I know, there are many publications on detailed topics about Chinese Buddhism, especially by Analayo, Bingenheimer, Dhammadina, Choong Mun-Keat, and Tse-fu Kuan. What I’m looking for is a big picture roundup with references to these detailed publications - kind of like Salomon’s ‘The Buddhist Literature of Ancient Gandhara’ which so nicely presents a big picture for his field.

Very grateful for references from those who are knowledgeable in Chinese Buddhism.


A part of your questions are addressed in the first part of Bhante Sujato’s A History of Mindfulness. As I understood when reading it, much research has still to be done in that field!


That was my uneducated guess. I would love to have been wrong, though.


Analayo has done some articles for the Encyclopedia of Buddhism, that would probably have what you want.

It’s a good question, we should have a point on SC that can aggregate our parallels data. A bit like our new languages overview, but for parallels;

I’ll give some thought to it.

This varies. Often, especially in doctrinal passages, the correspondences are literally word for word. If you look at something like, say, the jhana formula, you can see that all the translations, early and late, were clearly based on an almost identical Indic original.

Other things differ, especially when it comes to narrative and such, but sometimes also in the doctrinal passages.

In other cases, especially in verse, it is often hard to discern exactly how close the passages are.

I think the working hypothesis these days is that the early Buddhist community shared a largely common basis of texts, but they were never 100% standardized; there would have always been some variations. The various Indic schools drew on those and eventually fixed them in a specific form. the Chinese (and other) versions we have today are derived from various of those fixed versions of the schools (mostly Sarvastivada).

That’s a bit hard to state exactly. The two Dighas seem similar in some ways, yet DA has some texts entirely absent from Pali, and vice-versa. So do we consider the similarities of the actual parallels, or the whole content of the collection?

I believe that in terms of the similarities of the structure of the collection (as opposed to the content), the samyuttas show the most similarity.

A few things. Doctrine was clearly given priority over narrative. The early tradition deeply valued preservation, and made extensive efforts to preserve things as well as they could.


Thanks for your conclusions, bhante!

A hypothesis I had was that the collections we have available rely on certain archives of influential monasteries back then. So for example when texts were shipped to Sri Lanka texts from a certain influential archive were chosen and thus became canonical.

Or that the Chinese texts are a mix of collections found in important monasteries, frequented by bhanakas both for learning and for archiving, and by translators like Faxian…

In that sense the different canons would be a bit random as they would represent specific collections of one or a few monastery archives. And that in total there were probably many more texts of each genre (majjhima, digha, dhammapadas, etc.) floating around but not ‘chosen’ and not reproduced.

As difficult as it is to say today, but do you think this is a plausible version?


This is my impression. The Chinese canon is a selection of texts that happened to be translated. The Agamas were translated partially by different translators from different traditions. It would suggest that there were several different canonical lineages that existed in India in addition to the Theravada’s Pali canon. The “original” canon I think was lost to natural degradation of the local oral traditions as they diverged from each other. Then when writing was adopted, those diverged versions were written down by different sects or local monastic centers. Some have survived to this day, but I think half or even most didn’t survive. That seems like a reasonable conclusion given what we have in other languages.



I just wonder if there were ‘canons’ at all. Of course eventually there were, but it seems to me equally plausible that there were collections - without the claim of “This is what the Buddha said, and nothing more”, or “The text that this other monastery has is surely not Buddha-word, because we have everything he said”. These collections would have been open until a relatively late stage when, for some reason, the need for ‘closing’ and ‘authorizing’ would have come up. Only then I think we could reasonably speak of a ‘canon’.


I see what you mean. I would imagine that was true before the schisms began and the advent of Mahayana sutras, and so forth. We don’t really know much about that period because Indian Buddhism adopted writing somewhat late in the game, as I understand it.

Later on, by the time the complete Agamas were translated to Chinese (between 375-443CE roughly), the Agama/Nikaya literature was in direct competition with the Lotus Sutra, the Prajnaparamita Sutras, late Abhidharma texts, and so forth. Discerning what was taught by the Buddha or not would have become quite important to the curators of these canons, which were supposed to be a record of what he actually taught to his direct disciples. The Theravada probably were contending with this to a lesser extent, but it still would have become an issue with all these new sutras circulating.

To give an example of the confusion: In China, a common view was that the Agamas were a record of the Buddha’s early teaching career. He gradually advanced his pupils through the Prajnaparamita, taught the Lotus Sutra, and he preached the Mahayana Parinirvana Sutra on his death bed. It happened while Ananda wasn’t present. This seemed to make perfect sense of the wide variety of texts they were receiving from Central Asia. The upshot was the Chinese became very sanguine about sectarian philosophies. Hence, Chan ended up the overall “winner” of the sectarian competition in East Asia.



The textual tradition of Jainism supports this hypothesis. In their case there is no proper canon to speak of. Different sects have their own collections and very loosely share a handful of core texts, some comparable to what we call discourses (suttas) others even less structured than that!


Texts weren’t shipped: monks were. The texts existed in their minds. Presumably what became canonical in Sri Lanka was the scriptures as preserved by Mahinda and his students.

Yes, to some degree, but I would not think there are “many” uncollected texts. There are, in fact, various ways of discerning what was included in lost collections: for example, explicit references that occur in existing texts, whether canonical or later. And yes, sometimes there are references to discourses or collections that do not appear to have survived. But this is the minority. On the whole, I think the collections mostly overlapped, with a few extra texts added or missing here and there.