The Mahākaccāna theory of Theravadin origins

I have earlier proposed the idea that the Suttanipata was compiled in the south, specifically Avanti, among the school established by Mahakaccana. This explains why the Suttanipata is inknown in the northern schools.

It also suggests a further explanation for the origins of the Niddesa. This is a canonical commentary on portions of the Suttanipata, including the portion that was specially favored by Mahakaccana, namely the Atthakavagga, as well as the Parayanavagga which is located in the far south, near Andhra. Like the Suttanipata itself, it is unique to the Theravada.

Many people have commented on the fact that the Niddesa is a canonical commentary of the Suttanipata. But few comment on the text itself, especially its oddly mechanical nature as an explainer of dense poetry.

The Niddesa is, in fact, in the same style as the early Theravadin Abhidhamma texts such as the Vibhanga, the Puggalapannatti, and so on, as well as the Vinaya Vibhangas (which should be considered as Abhivinaya).

This Abhidhamma style is unique to the Theravada. Abhidhamma in other schools are less rigid and mechanical, more discursive and conversational.

Now, Mahakaccana was known for his dense and incisive analytical teachings. It would only be natural that he taught his students this way. And, given that he was also a lover of the Atthakavagga, it would make sense if his idiosyncratic approach was applied to it as well, which would be a plausible explanation for the peculiar style of the Niddesa. The school he established would of course continue to work in this tradition.

What if all the above texts were produced by Mahakaccana’s school in Avanti? That is, the entire Abhidhamma, the Abhidhamma-style books in the Khuddaka, and the Vibhanga material in the Vinaya? This explains both the existence of texts unknown to the north, as well as the style of those texts.

And what if Mahinda, having learned the texts in Avanti, brought them to Sri Lanka, where they formed the foundation of the Mahavihara curriculum?


As I had mentioned earlier, the concept of an “Avanti exception” appears to be unique to the khandakas of the Theravada vinaya (although if someone finds it in Chinese etc please let me know).

This to me would indicate that the compilers of the Pali khandakas were actively engaged in working out what the vinaya meant for them in Avanti, already some distance away from Magadha.

Thank you for sharing more on this topic, Bhante!


Nice, great. If we add up a bunch of points, maybe there’ll be a proper essay in it.


Interesting points, bhante!

Traditionally the Abhidhamma is attributed more to Sāriputta, and the suttas attributed to Sāriputta can, at times, be quite Abhidhamma-like (even the seemingly earlier ones). Mahākaccāna gives detailed analysis, but it has a slightly more organic flavor to me; he will sometimes turn what would be a normal technical term or plain reading into a metaphor (like talking about the abodes of consciousness in SN 22.3).

If Mahākaccāna had originally inspired this type of analysis, and his own students/lineage of students were the ones continuing it in their own way, why would they attribute it to Sāriputta rather than Mahākaccāna? The Nettippakaraṇa is attributed to Mahākaccāna, and the Niddesa to Sāriputta, so seemingly it could have been done.

Just some questions that sprang to mind!


1 Like

My understanding is that the Atthakavagga is the only chapter in the Snp that was translated into Chinese. How do you think that is explained? Also, do we have any reason to believe the other collections weren’t compiled in the south?

@cdpatton, is there reason to believe the atthakavagga Chinese translation is any earlier or later than those of the other Sutta collections?

This explains why the Suttanipata is unknown in the northern schools.

Bhante. Could you kindly provide more detail about the above? Thanks

Connected to my previous question, are you explaining the Abhidhamma is characteristically Sri Lankan? If so, how did Mahayana develop its own Abhidhamma? Does the content of the Mahayana Abhidhamma have any similarities to the Theravada Abhidhamma? Thanks :pray:t2:

The problem I have with the above theory is you seem to have exalted Mahakaccana yet, whilst I understand it occurred hundreds of years later, many Abhidhamma doctrines, particularly concerning the Noble Truths and Dependent Origination, redefine various relevant terms.

Good question! I don’t have any answer, but a more substantial investigation would definitely have to take this into account.

Well, the Parayana is more closely associated with the south, so perhaps it wasn’t circulated in the north? Or perhaps it’s just an accident.

As for the other chapters, the standard theory is that they did not exist independently but were compiled for the Suttanipata itself; I am suggesting that that happened in Avanti.

Well, all the texts that I have identified (the Suttanipata, Niddesa, Abhidhamma, and Vinaya Vibhangas) don’t exist in the north, although there are obviously parallels. Every school had a Vinaya Vibhanga, but it is not the same as the pali one.

As to style, all these texts share the same manner of very strict, almost dictionary-like glossary, rather than a discussion about the text.

No, I am saying the Pali Abhidhamma is characteristically Avantian, and was brought from there to Sri lanka. There are other Abhidhammas in early schools (hint: you can find them on SuttaCentral)

I don’t know much about it. There are similarities, but I’ve not studied it in any detail.

I’m not “exalting” anyone. I’m positing that we can trace the origin of certain texts to a region that was originally missionized by Mahakaccana.

I said nothing about time frames, except for the involvement of Mahinda, which is a well-established fact.


Forgive the lack of citation, but Asanga quoted a/some passages or suttas in the Pārāyanavagga in the Yogacarabhumi, and I think a couple Sanskrit fragments of certain sections exist in the north as well a handful of times. There may be some mentions of it in the Āgamas as well if the suttas that mention it have parallels—I’m not sure. The Milindapañha also mentions the Sutta Nipāta, and is a characteristically Northern text, though I’ve heard some say that that could be explained as being potentially added later in the Pāli translation. There are some stories about the brahmin Bāvarī in Chinese iirc as well. It’s not clear, but it seems this collection was just somehow lost maybe, or was not as important to the people in the north?


1 Like

I’ve not devoted much study of it. It’s one of the earliest Chinese translations, during the early 3rd century AD. It was clearly quite popular among northern Buddhists because it’s quoted frequently even by Mahayanists, who liked many of its passages. I first became aware of it when I was translating the Dazhidulun, Kumarajiva’s giant commentary on the Prajnaparamita Sutra. It’s also quoted by Asanga and Vasubhandhu. If was one of the earliest EBTs, it sure got around in later eras.

One thing I can see easily is that the Chinese version of the Atthakavagga has almost the exact same sutra order. There’s just a little divergence in the second half of the sixteen. Which would suggest it isn’t too distant from the Theravada version in time, really. There’s commentary included in the Chinese version, but I’ve not devoted any time to comparing it to Pali. I believe there’s a first attempt at an English translation that exists. If memory serves, it was by someone more familiar with Pali, so they naturally have that bias of assuming divergences are translations errors.

This is not really true. It just depends on which Abhidharma texts one chooses to read. Most people read Vasubandhu’s Kosa or Asanga’s Yogacarabhumi, which is the very end of the northern Abhidharma tradition. The early texts look like Theravada Abhidhamma.


I don’t think this is so, bhante, unless one is making an apples and oranges comparison of, say, a canonical Pali Abhidhamma text and a late Sarvāstivāda commentarial treatise.

When we compare like with like, the succinct style of the Pali Vibhanga is exactly like that [*] of the Dharmaskandha in the Sarvāstivāda Abhidharmapitaka (in fact there are several entire chapters that are identical in both works), while the conversational style of Vasubandhu’s auto-commentary to the Abhidharmakośa is easily matched by that of the Sammohavinodanī, Buddhaghosa’s Vibhanga commentary (though Vasubandhu’s work is far richer in historical data, since Buddhaghosa only occasionally names the persons or schools whose views he criticizes, while Vasubandhu always does).


[*]. Should say, “corresponds to…”


Oh okay, I’ll have to look these up.

To be clear, I’m strictly comparing the canonical Abhidhammas here.

You’re going to have to give references for this!

Here are the chapters on DO in the Vibhanga and the Dharmaskandha (1st link only). They’re almost completely different.

Sure, other Abhidhammas include the strict matika + analysis style of the Pali, but they include a lot else besides. There are actual discussions and quotations, much more similar to, say, the Visuddhimagga. I don’t think there’s anything remotely similar to the Yamaka and Patthana, nor is there any treatment of verse comparable to the Niddesa.


In general, mahayana directly absorb/ borrow Abhidhamma texts from early school, most notably the Sarvastivada.

The Tibetan favors Asanga’s Abhidharmasamuccaya and Yogacarabhumisastra, and Vasubandhu Abhidharmakosa.

In china, one popular abhidharma text is Tattvasiddhi sastra (Chengshilun) by Harivarman.

But those are later be supplemented by text containing specific doctrine of particular school, which can also be classified as Abhidharmic.

The Mahdhyamaka school of thought by Nagarjuna. The most important text is Mulamadhyamakarika.

And the Yogacara (Vijnanavada/ Citramattrin) school. Some of their texts:
大乘百法明门论 Mahāyāna śatadharmā-prakāśamukha śāstra (Hundreds Dharma of Mahayana)

Cheng Weishi Lun (Chinese: 成唯識論; pinyin: Chéng Wéishì Lùn, CWSL, Sanskrit reconstruction: *Vijñapti-mātratā-siddhi, English: The Demonstration of Consciousness-on

The new Mahayana school of thought build their idea on top of the earlier abhidharma philosophy. There are differences and similarities with Theravada Abhidhamma.

1 Like

I was going by what I remembered of the summaries and quotations from the Sarvāstivāda’s seven books in Abhidharma Buddhism to 150 AD, volume VII in the Encyclopedia of Indian Philosophies series, edited by Karl Potter.

In the case of the Dharmaskandha I had in mind Fred Greiner’s summary, which is chapter 3 of the above volume.

Looking at it afresh, I see that my “identical to…” needs changing to “corresponds to…” Sorry about that.

As for the similarity in style, what I particularly had in mind was the account of the twenty-two faculties in the Indriyavibhaṅga and its counterpart in chapter 17 of the Dharmaskandha Unfortunately the latter isn’t given in Greiner’s summary and I can’t now recall where I read a translation of it.


Well, then, it’s not really true, as I said. The same style and format are found in the oldest canonical Sarvastivada texts as well as in the Sariputra Abhidharma. It’s the way all the old Abhidharma texts read. The first division of the Sariputra Abhidharma has the same format as the Vibhanga, the Dhatukaya is very much like the Dhatukatha, and the Vijnanakaya is very similar to the Kathavatthu. The parallels are striking, both the subjects and the format.

The Dharmaskandha isn’t the closest parallel to the Vibhanga, so it’s likely to mislead if one isn’t able to read the other parallels.

There is an English translation of the Dhatukaya by Swati Ganguly. There’s also Frauwallner’s Studies in Abhidharma Literature, which compares all three sets of canonical texts.


indeed, the correspondence is interesting, basically they treat a similar set of topics, but the treatment is mostly different. Not to say there aren’t any similarities, but still.

Okay, we find the Dharmaskandha passage here:

Google translates it as:

For a while, Bhagavan was in Shiluofa and lived in the Garden of Solitude in Siduolin . From time to time, there was a Brahman named Shengwen , who came to visit the place of the Buddha , put his palms together respectfully, and said to the Buddha : " I would like to ask the Venerable Gautama less .[4] Only willing to obey . At that time, the World Honored One told Fan Zhi : " Ask as you please , and I will explain it ." Fan Zhi asked , “How many kinds of roots are there ?” The World Honored One said , " There are twenty-two . What are twenty -two ?

Sounds pretty different to the Pali. It then lists the twenty-two faculties. Let’s look at the definition of the “eye”:

What eye root ? It is said that the eyes and forms have been rightly seen and shared with each other , which is called the root of the eyes . Also , when the eye increases and develops eye consciousness , it is called the root of the eye . Also, the eyes and the color are rightly hindered from reaching the same part , which is called the root of the eyes . Also, the right eye for form and the same division with it are called the eye root . Such as the eyes of the past, future, and present are called eye -roots , also known as known, understood, comprehensible, omniscient , cut off , understood , seen , obtained , felt , realized , and so on . , what is observed , what is observed , what is examined, the decision , the touch , the touch , the certificate, etc. What is the answer? It is said that the pure form created by either hell , or near life , or ghost world , or heaven or human , or middling existence and cultivation . All the names are different words ,[5] Thinking , etc. , giving speeches and sayings are called the famous eye , the famous eye place , the famous eye realm , the famous eye root , the famous view , the famous way , the famous guidance , the famous white , the famous purity , the famous store , the famous door , the famous field , the famous things , Celebrities , famous pools , famous seas , famous sores , famous sore gates , and famous shores . If it is the root of the eye , it is taken from the inside .

Now, as for the Vibhanga, it abbreviates the Dhammasangani definition, which I quote here in full:

Yaṁ cakkhu catunnaṁ mahābhūtānaṁ upādāya pasādo attabhāvapariyāpanno anidassano sappaṭigho, yamhi cakkhumhi anidassanamhi sappaṭighamhi rūpaṁ sanidassanaṁ sappaṭighaṁ paṭihaññi vā paṭihaññati vā paṭihaññissati vā paṭihaññe vā, cakkhumpetaṁ cakkhāyatanampetaṁ cakkhudhātupesā cakkhundriyaṁpetaṁ lokopeso dvārāpesā samuddopeso paṇḍaraṁpetaṁ khettampetaṁ vatthuṁpetaṁ nettampetaṁ nayanaṁpetaṁ orimaṁ tīraṁpetaṁ suñño gāmopeso
The eye, that is to say the sentient organ, derived from the Great Phenomena, included in the self-state, nature of the self, invisible and reacting—by which eye, invisible and reacting, one has seen, sees, will, or may see material shape that is visible and reacting—this that is sight, the sphere of eight, the element of vision, the faculty of vision, [this that is] “the world”, “a door”, “an ocean”, “lucent”, “a field”, “[physical] basis”, “a guide”, “guidance”, the “hither shore”, an “empty village”—this is that [material] shape which constitutes the sphere of vision.

It looks like they share some things in common. The list of synonyms at the end is pretty interesting! Still, it looks like the similarities finish with the Pañhāpucchaka, which is about 80% of the text. So 80% is completely different, and 20% has some similarity. And specifically, it is the most “abhidhammic” portions that are different.

Let’s test that theory!

The first of three parts of the Dhatukaya is here:

It begins with a matika outlining various dhamma categories, which don’t obviously correspond to any Sutta grouping, then gives definitions of them, and goes on to expand and explain the definitions.

The Dhatukatha begins like this:

It’s a matika that has nothing to do with Dhamma categories, but rather is a set of logical operations for determining relations between dhamma categories; in other words, it’s an order of magnitude more abstract. The categories themselves are introduced on the next page, and they are the samyutta matika, with nothing in common with the categories of the Dhatukaya.

A couple more pages define the scope of the text, then we get to the meat at Dt 2.1. Here we see that there is no expansion or definition of the terms, rather it applies the logical categories of inclusion and exclusion, etc., to the matika.

I’m going to go ahead and say, “almost nothing in common”. I’m sure a deeper analysis would find more things that relate, but they’re really quite different.

I haven’t studied it in a while, but I do not believe this is the case. The Vijnanakaya includes, as a small part, a discussion of some disputed topics, among them being discussions of the puggala and sarvamasti, both of which are among the first topics in the Kathavatthu. But the discussions are much shorter, and there is nothing of the literally hundreds of further topics in the Kathavatthu.

In any case, the Kathavatthu is not really an example of the specifically Theravada Abhidhamma style, as it consists of quotes and argumentation, rather than the playing out of a matika. The peculiar style, which I still hold to be unique to the Theravada, is best exemplified by the Dhammasangani, the Patthana, and the Yamaka, as well as places like the Dhatukatha.

The same format as the Vibhanga suttantabhajaniya or the abhidhammabhajaniya? I can’t recall that myself, but I’m going to go out on a limb and guess it’s the suttantabhajaniya, and that here too, the specifically Theravadin manner of Abhidhamma is quite distinct.

Note that this shows that the Theravada understood that there is a a uniquely “abhidhammic” style within the Abhidhamma, and it is that that I am referring to.

Obviously I needed to define my terms more precisely in the OP. I’m well aware that there are substantial similarities in both content and method between the various Abhidhammas. In fact, one of my longstanding desiderata has been to create a set of Abhidhamma parallels. But the similarities tend to be in those things that are closer to the suttas, as you’d expect.

The Abhidhammas share, in addition, a fairly small core of specifically Abhidhammic ideas (like the, I think, six conditions mentioned in the Sariputrabhidharma vs. the 24 of the Patthana) but the extended treatment is almost completely different. And the other Abhidhammas lack completely (?) the Theravada Abhidhamma method of mathematical creation of entire texts from permutations of a matika.

1 Like

The original question was whether other Abhidharma texts were all conversational in style and not full of laborious stuff, yes? Not that they were identical. But, anyway …

You’re focusing on the subjects rather than the method. Yes, the matrkas are different, but it has the same analytical method of applying a set of questions to a set of subjects. This is the type of format the early Abhidharma texts have. It’s basically like they invent two-dimensional tables before writing, so it’s all read out loud in a laborious fashion. This is also the basic format in the Vibhanga. There’s a set of subjects and then a bunch of standardized questions designed to classify and describe them.

Anyway, it isn’t my theory, scholars like Frauwallner who did the deeper analysis say so, as does Ganguly who translated it for us.


Ven. YinShun also publishes a book on Abhidharma texts 說一切有部為主的論書與論師之研究 :

1 Like

Alright, let’s move on. I wanted to check what Frauwallner said that might be relevant, and thankfully it’s on the internet archive.

He first establishes a common stratum for the Abhidhammas, then asks, where did the differences arise? Which is the question I’m interested in here. He says that the Sarvastivada:

displays such obvious signs of a keen, markedly individual development … that it seems justified to assume that this entire development occurred in one and the same area.

This of course is the missionary area of the north-west.

As for the Pali, he supposes that the books were “written” (sic!) in the “mother country”, specifically where the missions originated from in Vedisa. This is of course on the southern road, one of the stops of the journey of the Parayana, and not far from Avanti. So the geographical idea is the same as mine, but he doesn’t connect this with Mahakaccana.

Now let’s look at what he says about the Vibhanga. I’ll paraphrase from page 45.

The Suttantabhajaniya has the same kind of analysis as the Dharmaskandha and is thus based on an ancient heritage. However this didn’t satisfy the “rampant scholasticism” (!) of the Pali Abhidhamma, thus the Abhidhammabhajaniya and Panhapucchaka were added. These draw from the Dhammasangani and belong to a considerably later period.

So there is a shared substrata of common proto-Abhidhamma. As the schools became localized, Sarvastivada in the northwest, the Pali schools in the south, these were elaborated in a way that is “markedly individual” in the case of the Sarvastivada and characterized by “rampant scholasticism” in the Pali. That “rampant scholasticism” consists of endless variations and repetitions of the core matikas.

Frauwallner comments on this tendency in the Yamaka. He says (paraphrasing again) that it is “deficient in content”, merely contrasting things in an infinite variety of combinations and linkages, every item being followed by its negation. As to the Patthana, he says it does not introduce new ideas, but that the Pali Abhidhamma is inexhaustible in inventing new cases and exceptions. And when something has already been discussed, a new variation is introduced, and the whole thing is repeated once more.

It is these tendencies, which Frauwallner repeatedly identifies as being characteristic of the Pali Abhidhamma, that I am proposing originated among the community established by Mahakaccana in Avanti, just as Frauwallner proposed they originated around Vedisa.

You wouldn’t be able to summarize Yinshun’s ideas about Abhidhamma for us, I suppose? :pray:


But there is a northern Sūtra Nipata, in fragmentary form, found in Turkestan / Xinjiang.

I think the British Library has the Rhinoceros sutra in a Karoshthi manuscript too.


Haha that just invalidates my whole argument!

But wait, no that’s the Atthakavagga. It’s well known that potions of the Suttanipata are found elsewhere, but not the collection, and that’s what I’m talking about.