DN is the earliest Nikaya

If we accept the traditional narrative that a large body of remembered discourses were later categorized and compiled in to various nikayas based on size and theme, why would one nikaya be older than another?


I think attempting to place individual Nikaya/Agama collections in an order of development is difficult. They may not have been created at the same time, but there may not have been much delay between them, either. The trouble stems from not knowing when exactly sutras were added or removed from a given collection. The conversion of heretic sutras in DN/DA or very long texts like the Parinibbana Sutta may have been later additions to an earlier collection that changed its overall character. The Chinese DA has an even larger cosmological text added to the end of it that’s twice the size of the Parinibbana Sutta (and DA’s version of the Parinibbana Sutta is itself somewhat larger).

Yinshun’s argument that SN and SA descend from the earliest collection makes sense logically because they look like the “leftovers” after larger sutras were moved into separate collections as the canon grew in size. He assumes that there was a single collection at the start, and then new collections were created as it grew and became unwieldy in size. This makes sense, but the history of how things happen is often quite surprising and complex when it’s recorded, and we’ve been left in the dark about the events in early Buddhism. So, we are stuck rationalizing what has come down to us.

It seems more meaningful and profitable to study the contents of all four collections and attempt to discern early and late strata of texts and material inside them. It’s still fairly speculative, though. It’s easy to read a few sutras, form a theory on anecdotal evidence, and then realize that a wider reading of material doesn’t support it. I’ve lost count of how many times this has happened to me over the past couple years.


Thank you for your insights, Charles.

I can’t imagine how we would ever be able to arrive a determination about an urtext or Q source in early buddhist texts. And how would that information inform our practice?

But perhaps with machine learning and data analysis something close to it is possible.


Thank you for sharing your thoughts, @josephzizys! It’s always great to hear different opinions. Like you alluded to, it does seem like the persuasiveness of your thesis depends on what assumptions you make. There is a lot I’ll have to intellectually digest. One thing that comes to mind right now:

Interestingly, the same is true with the Anguttara Nikaya….very little about the five aggregates there, even in the book of 5s.

IMHO the lack of aggregates in the DN/AN is consistent with the “each Nikaya had a different audience” thesis. In this thesis, the AN is more geared to beginner/lay practice, while the DN is geared towards Brahmins. If the aggregates were a more advanced monastic practice, that would explain their relative absence in the DN/AN.

Are there other reasons for thinking the 5 aggregates were an “advanced monastic” teaching? I think so. For example, the second cardinal discourse, which introduced the 5 aggregate formulation, allegedly took the five monks (who had already achieved stream entry) to Arahantship. Likewise, in MN 143, Anāthapiṇḍika laments that the teaching that Sariputta just gave him are rarely taught to lay people. Part of Sariputta’s teaching in that sutta involves the aggregates.


Nikayas as in later tradition where Buddha wasn’t there to approve. While earliest part of suttanipata has commentary which was oral probably which was ignored and replaced by the Nikayas. If nikayas was oral tradition why it wasn’t in the early Oral commentary Of the earliest parts of suttanipata?

Wikipedia has info on the earliest part of suttanipata

I think that ordering suttas and recognizable pericopes by doctrinal elements used or presupposed is the way to go. Archaic language factors in as well. Larger collections may have been developed over long periods of time so some suttas may be earlier and others later. It may be that one collection was started first, but finished last.

I use the absence of the five aggregates in both the Chapter of Eights and MN119 to argue that they are both early texts and from there draw some conclusions about the jhanas in my discussion. I think that we have to be careful when we prooftext to only do it between suttas and perocopes that use and presuppose the same doctrinal elements.

Ordering may take several iterations. Ordering the doctrinal elements based upon simplicity and dependence on others may have to be revised based on whether or not they are consistent with what is found in the suttas. In other words, there may be some chicken and egg issues to work out.


@Gabriel i agree with what your saying, and I think that DN has some
Of the most ornate and therefore probably latest verse forms in the canon, but as I allude to in the essay I think the roots of the material are generally the oldest, even if the text wasn’t closed until later - jhana is older than sati, gradual path is older than 8fold path, 10 DO is older than 12 DO, narrative settings are older than mechanical formula combinations etc.

I’m not sure this is a strong point here. The language in verse tends to preserve more archaic forms. We do know that suttas which quote the atthakavagga of course come afterwards, and the tutthaba metre is an older one, but a lot of suttas have that metre or siloka which is also older. Check out KR Norman’s essay here on the atthakavagga: https://scdd.sfo2.cdn.digitaloceanspaces.com/uploads/original/2X/2/265995a8b2d9b994fea6a9b85ad51dc83467afd5.pdf

To make this claim larger I’d like to see more suttas from each of the developmental steps you are claiming. For example can you find a large number of suttas with just the short section, a large number with just the short plus middle, etc. How do we also determine the temporal progression here between the patimokkha and the short section in DN? Maybe the patimokkha came first, and then the rules in the ‘short section’ are borrowed from it.

Why does a less technical formulation mean an earlier strata? What if the Buddha was very technical? In some sense the list of the 8fold path is simpler than the gradual training. The gradual training could also be viewed as a way of ‘fleshing out the details’ of the 8fold path suggesting the 8fold path comes first. It’s too easy for me to just flip the argument in this way.

This needs to be explained in more detail. Right now I have no clue why the 10 link version is more easily criticized by annihilationism.

As suggested by others, the missingness of the 5 aggregates could be attributed to other aspects of DN as a collection.

I’m not sure what you mean by recapitulating the later canon here.

Thanks for putting this together, it’s interesting to see a different nikaya besides the samyutta being suggested as the earliest.


@stephen I don’t accept the traditional narrative that the 4 Nikayas as we have them now where recited immediately after the death of the Buddha by Amanda, I think there is internal evidence that they developed over time, drawing from a pre existing literature, and that evolution of form and doctrine is evident in the texts, with simpler forms like 10 DO more often found in DN.

1 Like

@cdpatton I totally agree that my theory is speculative! I also have to say I was extremely hesitant to post anything because as I say in the OP I really haven’t got time to do my arguments justice - I’m working full time, have 3 school age children, I barely manage to fit in my daily meditation! As I was marshalling these ideas I came to the conclusion that at the very least I should read DN MN and SN again, and probably a whole bunch of secondary literature - my understanding of verse forms is woefully naive and second hand, and as you point out I have made no effort to engage with the Agama material! Nevertheless! I still felt I. The end it was worth sharing my thoughts and getting the feedback of the wonderful people on this forum, and I regret nothing!

I think speculative theories are fine if we don’t become too attached to them, and can even act as something of a protection against too easily swallowing and becoming attached to the traditional narrative.


@TheSynergist thanks for your reply, and that is a fascinating observation about AN, which is the N of the first 4N I am by far the least familiar with having only recently acquired Bodhi’s translation. In terms of 5A, my speculative feeling is that it is a later development than DO, sort of a modified version of DO that came to be used more in an effort to combat personalists, I actually think that today it almost suffers from its own success and that people end up thinking something like “well “I” may not really exist but at least the 5 aggregates exist!”.

I should probably say that on my reading all the Nikayas material presents a totally consistent philosophy and approach to awakening, and I see no contradiction between 10DO and 12DO and the 5A or between the jhana formula and the Satipatthana formula and so on, I think it tends to be essentializing and reifiying readings of the suttas that leads to worries about consistency, I also feel like it’s quite plausible that the Buddha themselves taught different groups using different metaphors or formulas based on the local intellectual mileu with some groups like Brahmins responding better to one focus and others to others.

Finally with regard to the Ur texts, my take is that it is far more likely that in the pre-Nikayas period there was a whole world of sutta and geyya, that is the short formulas and the short verses, remembered by a wide variety of followers, and that all 4N drew from this “background” of literature rather than starting as SN and then branching out. Of course all these ideas are necessarily speculative.

1 Like

The first chapter entitled “THE GIST: THE HIDDEN STRUCTURE OF THE DHAMMA” of Venerable Sujato’s History of Mindfulness book basically comes to the conclusion that the Samyutta Nikaya/Samyutka Agama is the earliest “collection”.

I finished reading it a couple of days ago. It’s a very dense and detailed read but in the end I was able to absorb much useful information even though English is not my primary language.

There’s another book called “The Authenticity of Early Buddhist Texts” but I haven’t found time to read that one.

Just posting this here I case you haven’t read it otherwise just ignore.

With Metta. :grin:

1 Like

@SilaSamadhi8 I did read @sujato s wonderful book! But I guess I did not find his argument that SN is the earliest collection very convincing- having said that I would love to hear from others what those arguments, including Yinshuns arguments for the priority of SN are in simple terms - I gave my list of 10 for DN, an equivalent list for SN might facilitate discussion.

My own impression is that SN probably does have some of the earliest short suttas and formulaes, but that it formed as a collection somewhat later than DN and MN, perhaps partly out of an urgency to preserve the early material that DN and MN had overlooked - my reasoning is basically my impression that the collection itself has significantly more “mechanical” features that create doctrinal sophistication by combining formulas into more complex versions of ideas that are not evident in DN, this is suggestive to me that as a whole SN had later roots than DN, but as I say that doesn’t mean that SN doesn’t preserve sutta and formula that are every bit as early as DN.

Once again I just want to reiterate that my theory is a speculative one, that I am not a monastic or an academic and so I earnestly mean to enjoy and provoke robust discussion but in no way mean offence to my elders.

1 Like

@Preston thank you for your comprehensive feedback! You and @Gabriel have given me a lot to chew on! I think I will have to spend some time thinking about my responses so I will do that and reply to your post more thoroughly later. Thanks again!

I second the motion! You cannot test a hypothesis until you have formulated it. That is how progress is made. Testing is through analysis of the text and practice.

This is an interesting theory. My gut reaction is that it seems quite plausible. It explains why the Samyutta Nikaya has some “archaic” features that make it seem like the building blocks of the longer nikayas, while also explaining why it has so much inelegant, vague, and/or repetitive fluff. Sometimes this “fluff” consists on entire samyuttas that actually ought to be dealing with important subjects….why, for example, are the samyuttas on the efforts, five powers, and Jhanas such nothingburgers? Under your theory, that the SN was created to salvage old material, the “fluff” could perhaps be seen as attempts to fill in material that has since unfortunately been lost.

I’m curious how the Anguttara Nikaya would fit into your reconstruction. Bhikkhu Bodhi once pointed out that there is surprisingly little overlap between the Samyutta and Anguttara Nikayas, despite the suttas being of comparable length, suggesting that the redactors of these nikayas deliberately avoided sharing too much material (btw, this could be another reason why the 5 aggregates are so much more emphasized in the SN compared to the AN. Ditto with Dependent Origination, and probably some other things as well). Under your theory, perhaps the SN redactors didn’t feel a need to duplicate all that much from the AN since the whole point of SN was to preserve short material that was not preserved elsewhere. But there could be other explanations, of course.

Keep up the good work, @josephzizys!

1 Like

I do, too. If we want to try to work out these issues, we have to go through alternatives and vet them. I was just giving my two cents, having done some of this myself.

1 Like

Just to clarify, we need to distinguish between “earliest collection” and “collection with earliest content”. All the collections clearly have content added over time, SN no less than others. When we argue for it being the earliest collection, we mean that the way it is structured and organized was probably the earliest.

Just a few quick points to consider:

  • other accounts of the First Council do not place DN at the start
  • there is much material in DN that is obviously late. So that needs to be considered.
  • You say there’s no abhidharmic tendencies, this is clearly incorrect: DN 22 for a start, but also the last couple of suttas; Sangiti Sutta was even used as the foundation for an Abhidhamma book!
  • Regarding the 10-factored DO, you might be interested in Rod Bucknell’s argument to this effect. I disagree with it, on the basis that you can’t draw historical conclusions from the mere shortness or longness: such conclusions must be triangulated with a range of supporting conditions. (per our method in The Authenticity of the early Buddhist Texts). So if you were to show on other grounds that DN14 was early, the 10-factors might be held to improve the case. Unfortunately, comparative study with the Chinese version reveals the opposite: it has a lot of differences, and is likely to have been a late development.

@sujato I am all a flutter that one of my dhamma-heros has made comment on my work! thank you Bhante! I am so pleased to have gotten so much constructive feedback so quickley on this essay and am inspired to go back and look at many of these aspects and address them, however I must first find the time and right now i really don’t have it, perhaps one day I will have the opportunity to return to full time study and make a proper fist of this, in the meantime thank you very much for your observations and thank you and everyone especially for suttacentral and discuss and discover, they have truly reinvigorated my passion for the dhamma and my practice! Metta!


Here’s a cheeky thought; perhaps SN and AN originate from DN33 and DN34 respectively, and their contemporaneous development explains the tendency to avoid duplication of content.

1 Like