Does anyone have any scholarly resources (traditional or contemporary sources) which detail when suttas were first organized into nikāyas, saṃyuttas, vaggas, nipātas, and so on? Or do we not know?
4 Likes
AFAIK, the tradition has it that from the very beginning… From the First Council, at least.
Scholars think that all that came into existence gradually, and we can’t say exactly when.
There is an inscription from Sri Lanka that dates to the 2nd century BCE, and it mentions reciters of the different Nikāyas. So, perhaps, Nikāyas/Āgamas are the oldest classification.
For the rest, you may consult Norman, von Hinüber, Frasch, Allon, Gethin, Anālayo, etc. Always the same good ol’ names… If you want, I can give you the exact references.
3 Likes
So it must have been fairly early!
Some whole vaggas exist the same or nearly the same between the Agamas and Nikayas so there must have been some common ancestor there at that level of organization as well, even if that wasn’t quite as fixed as the split into “Long, Middle, Connected, Numerical”
3 Likes
Thank you.
Sure. I’ll take them, if you have them. Thanks!
1 Like
Really? I can’t say that I’ve ever really looked deeply into it beyond noticing great variation between the SA and SN, and certainly between the DA and DN.
1 Like
The following discussions may be relevant to your question:
1 Like
Sure, I wouldn’t have said that, hadn’t I had them.
Later today or sometime tomorrow.
2 Likes
Yeah! It’s just a figure of speech, like, “Smoke 'em if you got 'em!”
Thanks. Looking forward to it.
1 Like
I’ve written a couple essays here on the forum comparing the structures of Agamas. MA/MN has the least in common, while DA/DN has fair amount of commonality that’s been reorganized in different ways. SA/SN has the most in common in terms of organization, but the divisions have been jumbled up, obviously, both in SN and SA.
4 Likes
According to Yin Shun, MA/MN and DA/DN were compiled at the same time, but DA/DN slightly later:
《原始佛教聖典之集成》The Formation of Early Buddhist Texts：
「「中阿含」與「長阿含」的集成，是同時的。但二部所類集的各種經典，「中阿含」要早一些。理由是：約文段說，「中阿含」文段短，「長阿含」文段長；大體上，契經是由簡短而逐漸長廣起來的。約內容說，「中阿含」以教內的比丘為中心，分別、抉擇、整理、評判（外道），從佛法的多樣性中，現出完整的體系。「長阿含」是對外——婆羅門、外道的，將「中阿含」的內容，更有體系，更完成的，透過一般的天神信仰，而表現出佛陀的崇高，佛法的究竟。」(CBETA 2025.R3, Y35, no. 33, p. 732a7-11)
「「長阿含」與「中阿含」，是各有特色的：「中阿含」重於僧伽，「長阿含」重於社會。「中阿含」是法義的闡明，「長阿含」是宗教的適應。「中阿含」是「修多羅」勝義的延續，「長阿含」是〈八眾相應〉——「祇夜」隨順世俗的發揚。在「中阿含」——法義分別的確定過程中，部分佛弟子，更有條理的綜合當時的宗教思想，承受佛教界所完成的修道次第，而結集傳出的，是「長阿含」。所以，「中阿含」與「長阿含」，可說同一時代集成的；而「長阿含」多少要遲一些。「中阿含」代表那個時代的前期，「長阿含」是中後。」(CBETA 2025.R3, Y35, no. 33, p. 747a2-7)
2 Likes
For those who can’t read the Chinese, Yinshun’s argument is that the Madhyama Agama contains shorter texts compared to the Dirgha Agama. (Though, actually, it has a couple very long sutras in Chinese.) Also, the Madhyama Agama spends more of its time systematizing the Dharma teachings rather than being concerned with societal issues and relations with non-Buddhists. He reckons that DA and MA were put together during the same era but that MA represent an earlier part of that era that DA.
2 Likes
I have to say, I often have trouble figuring out what you mean.
In this case, I understood that the phrase “give references” somehow prompted you to respond with “take them” in which “take” is an opposite of literally understood “give”. What I don’t understand, is how (and why) the perfectly idiomatic (in academic English) “give (someone) a reference”
This is an example of a footnote.
The first time you cite a source
Give a full reference (called a full footnote in this guide). Author names should be Initial(s). Surname (i.e. C. N. Adichie) and are not reversed (i.e. Adichie, C. N.). Many footnote systems suggest full names should be given rather than initials, however the University of Hull has decided to use initials only as many academic journals do not give full names and this would lead to inconsistency within the system. Page numbers indicating where you found the information you are referring to should be given at the end of the reference. If you want to point to a specific page in a book chapter or journal article reference, you should add a colon after the page range to do this.
Examples of a book and a journal article for single authors are given below.
N. Rogers, The press gang: naval impressment and its opponents in Georgian Britain (Continuum, 2007), 45-47.
K. N. Panikkar, ‘Literature as history of social change’. Social Scientist, 40, 3 (2012) 3-15:4.
Hull Footnotes - Referencing your work - Library at University of Hull
caused to come into being the unidiomatic, strangely sounding (in English) “take references”. You say it is a figure of speech, but I can’t confirm the existence of such a figure of speech in English. So I am a bit perplexed as to how to understand you.
Not that it is of crucial significance, but since we had this little exchange, I wanted to underscore that.
Anyway, let’s get back to The Giving of References:
Mark Allon (2021) writes:
The earliest datable references to bhāṇakas of specific collections of texts are found
in Buddhist inscriptions, of which the earliest date to the 2nd century
BCE. For example, in the Sri Lankan inscriptions dating from the 2nd
century BCE onwards published by Paranavitana (1970) we find
reference to Majjhima-bhāṇaka (majhimabaṇaka, no. 330), Ekottarikabhāṇaka = Aṅguttara-bhāṇaka (ekautirakabaṇaka, no. 407), and Saṃyutta-bhāṇaka (sayutakabaṇaka, no. 666),
84 while in early Indian inscriptions we find references to those who know or are a master of three baskets, the equivalent Sanskrit forms being paiṭakin, traipiṭaka,
traipiṭikā, and traipiṭakopādhyāya.85p. 44
https://www.buddhismuskunde.uni-hamburg.de/pdf/4-publikationen/hamburg-buddhist-studies/17-allon.pdf
then, Ven. Anālayo (2016) says:
The Four Āgamas
Accounts of the first saṅgīti in the Dharmaguptaka, Haimavata (?),
Mahāsāṅghika, Mahīśāsaka, and Mūlasarvāstivāda Vinayas employ
the plural form āgamas in their description of the compilation of the
teachings that according to tradition took place under the leadership
of Mahākassapa soon after the Buddha’s demise.15 The different
Vinayas reporting this event agree that the Buddha’s personal
attendant Ānanda recited the discourses on this occasion.
They further report that the resultant textual material was divided
into groups, presumably to facilitate oral transmission, by collecting
long discourses and discourses of middle length into corresponding
āgamas; and short discourses were further separated into those
assembled according to topic and those assembled according to a
numerical principle (a discourse can discuss one or more items and
can accordingly be allocated to the Ones, the Twos, the Threes etc.).pp. 12-13
https://www.buddhismuskunde.uni-hamburg.de/pdf/5-personen/analayo/agama-anga.pdf
then, the passage from the Mahāvaṃsa about the writing down of the canon in the 1st cent. BCE is discussed by Frasch (2022), from p. 6 of the pdf on:
https://d-nb.info/1280687487/34
then, regarding the segmenting into the saṃyuttas / vaggas / nipātas, etc.,
Norman (1997) writes:
Something similar can be seen in the uddānas (the lists of contents) which are
found in, or at the end of, a number of texts, e.g. in the Dhammapada where the uddāna
gives the names of the vaggas, and in the Thera- and Therī-gāthās, i.e. the verses ascribed
to male and female elders, where we are told how many elders there are and how many
verses as a whole they have recited. Despite their position in our texts, these uddānas
were probably intended originally for use at the beginning of the recitation, and even as
the reciter progressed. Someone reciting the Dhammapada, for example, had a guide to
tell him which vagga came after which, so that he could keep them in order. It must be
pointed out that the system is not foolproof, because the numbers given in the uddānas to
the Thera- and Therī-gāthās do not agree entirely with the numbers as we have them
now,21 and they presumably refer to an earlier recension of the text, where such numbers
were relevant and of value to the reciters. They have been retained in a written recension,
even though they are no longer of any value.Listing things by numbers is a very common mnemonic device in India, and both
the Buddhists and Jains make use of this idea. The Jains have the Sthānāṅga and the
Samavāya, the contents of which are listed numerically, and the Buddhists have
something very similar in the Saṅgīti-suttanta of the Dīgha-nikāya. The name Saṅgīti is
reminiscent of the saṅgītis at which the texts were recited in a joint recitation, and the
name of the sutta suggests that it represents a recitation of doctrinal matters, arranged in a
numerical way, and intended for chanting together, perhaps in an attempt to provide a
summary of the doctrine as a precaution against the confusion among the Jains after the
death of their leader Nāthaputta, which was the occasion for the preaching of the
Saṅgītisutta. Whether the recitation was at one of the great saṅgītis or at some other
chanting is a matter for investigation.
We also find this “more by one” principle used in the Aṅguttara-nikāya, which is
listed numerically for the same reason. It is quite clear that this method of enumeration is
purely for artificial purposes, because we find that although the lists of ones and twos and
threes and fours, etc., are genuine groups, when we get to the higher numbers the authors
of the text were obliged to make combinations, e.g. one group of ten is made of the five
fears (bhaya), the four elements of stream-entry (sotāpattiyaṅga), and the ariyan method
(ariya ñāya).
22pp. 51-52
https://drive.google.com/file/d/1kQvxmSH-NDrTXhHgHjYjt1UmdsAix46Q/view
then, Gethin (1992) says that saṃyuttas likely are the continuation and elaboration of the basic mātikās (“matrixes”) of the Teaching(s) from the early oral period (pp. 162, 167):
I hope this helps, and if you need to take more references, which I’m giving out here so readily, I’d be happy to give them all up. 
1 Like
@Sphairos cites
The Composition and Transmission
of Early Buddhist Texts
with Specific Reference to Sutras
by Mark Allon
as saying
And Allon in turn citing:
So the text cited above is;
Paranavitana, S. 1970. Inscriptions of Ceylon Vol.1, Containing Cave
Inscriptions from 3rd century B.C. to 1st Century A.C., and Other
Inscriptions in the Early Brāhmī Script. Colombo: Department of
Archaeology
Now this text is the basis for the claim that these nikaya bhāṇakas may date from as early as
However the cited text itself says;
“Palaeographically, therefore, the inscriptions contained in Section I of this Volume may be taken as ranging in date between the last quarter of the third century B.C. and the middle of the first century A.C.”
(pp xxvi)
The subsequent evidence for the dating of these inscriptions thereafter hinges principally on two things; first the connection of names in the inscriptions to figures connected by the lankan chronicles to the reconstructed periodisation of Ashoka, and the paleo graphic nuances of the carvers hand.
Both these, but especially the paleography, are subject to doubt.
In short, the confidence with which Allon ascribes a period to his evidence is belied by the source from which he draes which presents a much more ambiguous picture.
It is much more faithful to Paranavitana to acknowledge that sans compelling evidence we can get no nearer than “Ranging in date between the last quarter of the third century B.C. and the middle of the first century A.C.”
The other thing to keep in mind is the ambiguity inherent in the terms themselves, by what evidence prior to Buddhaghosa can we claim to know that the reciters where reciting "our"texts?
It seems fairly certain based on the Gandaharan evidence that texts recognisable as “samyuttas” and so on where in circulation by the beginnin of the common era, but proior to that how do we know which texts and when the reciters where reciting?
What I mean is, we can well imagine a circumstance whereby a monk is a speciallist in reciting teachings arranged numerically, or arranged by theme, or is renound for reciting a small number of long texts or a larger number of medium texts and so on and so forth, but how do we know which texts these where?
We can assume that it is very likely that the samyutta reciters from the textual basis for the eventual samyuttas we have now, but how many of them where originally recited in the first generation?, how many where added in the purely oral period? how many where added when literacy began to exert an influence? when did literacy begin to exert an influence? when did complete manuscripts start to be produced? what influence did they have?
We have a reference to manuscript production in the chronicles, but what information does that give us about Gandahra?
We have evidence of Brahmi script from the 3rd century bce and perhaps even earlier, we know almost nothing about it’s use on perishable materials in the 400 years before the chronicals mention the productiono of a manuscript tripitaka.
The production of such a manuscript, from scratch, would have been an enormous, multi decade undertaking, as can be seen form the history of the Chinese translation efforts, and we again know just about nothing about it.
I was going to draw more heavily from Paranavitana’s book, which is a marvel, but the work involved would take weeks, so I will leave it as an exercise to the reader to 1. identify the donors in each of the banaka related inscriptions, and 2. compare the paleography in each inscription to the provided chart in the book to see how reliable the periodisation is.
The last thing I would say is that at least in the texts themselves there is almost no evidence at all of text groupings, or even of texts, except for the mention of the " suttaṃ geyyaṃ veyyākaraṇaṃ gāthā udānaṃ itivuttakaṃ jātakaṃ abbhutadhammaṃ vedallaṃ" list (mentions that occur in what appear themselves to be later texts) and the specific texts:
make of that what you will.
1 Like
I’m sorry to hear that. I’m from Brooklyn, though, and we are indeed known for having our own style of idiomatic speech: e.g., waiting “on line” instead of “in line,” using the simple past when others might use the present perfect: “Did you eat yet?” for “Have you eaten yet?” is a famous one. I don’t know where you’re from, but, if you’re not a native speaker of English, it could be confusing I guess. “Take” really just means “receive.”
I have to say, though, that I am flattered to hear you say “often”: do you often read my posts? That’s nice to hear.
(By the way, I pronounce “often” with an audible “t.” Do you?)
Again, my apologies. The most I can suggest is that you watch more movies with characters from New York, maybe? I don’t know. But, please, only the old movies: the new movies with the exaggerated caricatures of New York speech are, quite honestly, offensive!
No, please, feel free. Obviously, it was of some significance to you. I’m glad to be of service. Also, I really enjoy learning about the different varieties of English from the different parts of the world. So it’s a win-win for me!
Thank you for the references, by the way.
And, if there’re more, sure, I’ll take 'em if you got 'em! 
2 Likes
Obviously if only according to Pali tradition, one certainly could not have clear evidence to understand how the texts were first arranged into chapter (vagga), collection of fifty (paññāsaka), connected collection (saṃyutta) and section (nipātaka), composed the Basket of Āgamas (Āgama-piṭaka) which is known by name of Sutta (suttasammata).
However, YinShun’s interpretation of the Sarvāstivāda structure based on Yogācārabhūmi tradition is the reasonable textual evidence to understand how the texts gradually developed, arranged into the four principal Agamas/Nikayas.
It is about the gradual development of the four principal Agamas/Nikayas in Early Buddhism, of which Saṃyukta-kathā (相應教, ldan pa’i gtam) was the foundation, although the extant Agamas/Nikayas are sectarian texts.
It just makes so much sense when the texts are structured like that.
Cf.:
Texts Arranged into nikāyas, saṃyuttas, vaggas, nipātas, and the Like - Q & A - Discuss & Discover
Thank you to all. The response was overwhelming: far more in-depth than I expected. There’s so much information here, I will really have to review these answers piece-by-piece.
(And, thank you for the links as well, @thomaslaw!)
I think we’ll probably close out here, though if anyone else has anything more to add, please, feel free.
1 Like
I consider one of the main reasons why the Pali tradition does not have clear evidence to understand how the texts were first arranged into the four principal Nikayas/Agamas is:
Buddhaghosa “may also have had the purpose of making it appear that the Pali Piṭakas had originated from the first Saṅgha council, and that the Pali language of the texts was identical with Magadhi, the language spoken by the Buddha.”
(p. 902 in Choong Mun-keat “Ācāriya Buddhaghosa and Master Yinshun 印順 on the Three-aṅga Structure of Early Buddhist Texts” in Research on the Saṃyukta-āgama (Dharma Drum Institute of Liberal Arts, Research Series 8; edited by Dhammadinnā), Taiwan: Dharma Drum Corporation, August 2020, pp. 883-932).
Buddhaghosa is generally recognised by Theravāda Buddhists as the most important commentator of the Theravāda school.
1 Like