I think the criteria used to draw the line of demarcation is controversial. The general consensus at Sutta Central is conservative.
1 Like
Conservative in what sense?
I think the Eightfold Path was a mnemonic device to recall the teachings of the Atthakavagga only Right View was No Views. This could be a thread by itself. The first three Noble Truths IMHO was most likely taken from the saying that someone who does not get stuck in the middle of the world, the cessation of the world, and the arising of the world has avoided the seamstress. Getting stuck in the cessation of the world phase is to become attached to it which leads to suffering. This is alluded to in the Atthakavagga and the Pariyanavagga.
The teaching of the Atthakavagga is agnostic with regard to karmic retribution. It does not depend on views. I would say that is a feature, not a bug. In any event, the point is to escape suffering so there is a point. The problem with views is we get so fixated on them. Meditation and a simple understand of how it fits into the practice is what is important.
2 Likes
It is relatively easy and uncontroversial to determine which suttas are in both the Pali Canon and the early Chinese translations. That said, why limit ourselves to that extent. Looking at other evidence reveals a lot. This site is dominated by Theravadins. Whether consciously or unconsciously, they have an interest in designating as much of the Pali Canon as EBT as possible. The academics aren’t crazy. They are open to more evidence.
1 Like
EBTs (such as the principal source, the four Nikayas/Agamas) are just texts , some compiled/edited early, some later. EBTs were gradually developed and expanded. EBTs were not entirely established at once at the first council in complete form (structure) and content. The extant EBTs are sectarian texts. One can seek an understanding of early Buddhist teachings by studying them comparatively.
Very intetesting thoughts. Thanks for sharing them, @Raftafarian !
Thank you too, @thomaslaw . I understand that principle. It’s the discrepancies between the EBTs and the scholarly theories about Earliest/Original Buddhism that made all my questions come up.
1 Like
I think this is a bit strong. @Raftafarian Talks about “no views” Buddhism - but I would argue that this is not confined to the Atthakavagga- In the Vacchagotta suttas for example the Buddha has gone beyond “all convictions”, the similie of the dhamma as a raft implies the transcendence of all views including the “right view” of the path, the 10 unanswered questions are best explained as a transcending of views (since all views are conditional), so I would say there is plenty of continuity apparent IMO between the texts.
On the side of evolution well snp4.11 has a version of dependant origination that has only six linkss an uses lots of non-standard language, I would encourage you to re-read it with this in mind.
The 6 link DO given is:
6DO
conflict
holding dear
desire
comfort/discomfort
contact
name and form
Now DN15 has a 10 link DO that looks like this:
10DO:
death
birth
existance
grasping
craving
feeling
contact
senses
name and form
consciousness
name and form
Where name and form and consciousness lean on each other like an A frame.
Finally we have the “standard” 12 link DO, first appearing in MN and SN for example at SN12.1 that gives
12DO
death
birth
existance
grasping
craving
feeling
contact
senses
name and form
consciousness
choices
ignorance
Now, ONE way of explaining this is to suggest that 6DO gives an example of an early version of DO from before the language was standardised, 10DO represents a time intermediate, and 12DO is the final, definitive version.
It would be logical to infer on this picture that if any of the DO formulations could be put in the mouth of the Buddha it would be 6DO.
This only represents a “problem” if you think that DO is some sort of mechanical folk-physics describing really existing processes that involve separable and substantive parts.
If you take DO to be an example of a principle, a way of understanding and analysing situations then there is no problem at all, the Buddha taught the principle, we have 3 good snapshots of the basic example, once we understand how all 3 are parsimonious with each other and represent the same underlying strategy we have begun to understand the Buddhas fundamental teaching.
The same can be said for many of the other examples you list. I take the sekkha patipada in DN to be an earlier form than the 8fold path, but I see nothing to suggest that they are contradictory or incompatible.
Even if the Nikayas are compiled from a time after the Buddha, and represent not what he said directly but what the monastics where teaching say a century later, why should we think that what was being taught was “wrong”? They where his students after all, if they formalised terms and made mnemonics and elaborated in themes, what’s wrong with that?
Some of the “original Buddhism” scholars are motivated by a desire to claim that there is some radically different “real” Buddhism that is contradicted by Nikaya Buddhism. I don’t think we have any real reason to believe that.
Metta.
2 Likes
I agree that there are suttas consistent with no views found elsewhere in the canon. That said, the Atthakavagga most explicitly and clearly expresses it as a principle and bests attests to its importance early on. I would also say it is honored mostly in the breach elsewhere in the canon.
2 Likes
The only scholar cited that says otherwise is Vetter, who is not an expert, but is a scholar of Madhyamaka dabbling outside his field. In that Wikipedia page, I have, long ago, noted that the actual experts in the early texts have a more positive view as to what they contain, whereas the extreme sceptics almost always come from a background of late Buddhism.
The other cited scholars, who are experts in the field, are saying that the formal compilation of the prose suttas into the nikayas as we have them today is somewhat later than the Aṭṭhakavagga and Pārāyanavagga. Nobody disputes this. But the prose suttas themselves existed long before they were collected into nikayas, and there is no serious evidence that the Aṭṭhakavagga and Pārāyanavagga are earlier than the prose suttas on the whole.
All the collections contain texts of different ages; heck, the Pārāyanavagga itself contains some of the latest passages in the entire canon. The age of the collection tells us nothing about the age of the contents that are collected.
Merely collating arguments, as does Wikipedia, is no way to find the truth. Those who have argued for the supremacy of the Aṭṭhakavagga and Pārāyanavagga have done so on the basis of certain specific reasons, and those reasons are incorrect.
9 Likes
I am not sure I fully agree with this interpretation, the scholars cited for the most part are a bit vague tbh, one gets the impression that some, like Von Hinuber have a sympathy for the idea of doctrinal evolution, some, Like Nakamura and Mizuno are more swayed by the fact that they take SN to be early and that therefore Atthakavagga is earlier still due to the qoutation of it in SN, some like Warder seem to think that there are metrical reasons to think that Atthakavagga is earlier than Therigatha, but are more circumspect with regards to the relation to the prose…
The position they present does not seem to me to be simply that the text of the atthakavagga was fixed earlier than the text of the suttas, if anything they do not go that far! I would love to see a few articles where the secular scholars nailed their flags to the mast, but they all seem to just say “early” without saying definitively that they think:
-
that the texts where all from a similar time and that some where simply fixed in vernacular language earlier or later
-
that there was genuine evolution of doctrine that different strata of the texts represent (Von Hinuber sort of hints that this might be their position)
-
That different geographical contributions might go some way to explaining different language features.
-
I don’t actually have a 4 to hand but I am sure there are other options I haven’t considered.
Unfortunately what the secular scholars actually say, that I could find, amounts to saying that (the core of) the atthakavagga as we have it now appears to use earlier language and metre than much of the rest of the canon, as to what explains that fact very few that I could find where willing to give an opinion.
My IMPRESSION of Norman is that they probably would have a great deal of sympathy to “formal compilation” argument, and my IMPRESSION of Warder is that they would probably have a sympathy for the “significant evolution over time” argument.
My impression of BOTH is that they are far to clever as academics to risk actually putting in print what they think on balance at the end of the day, which after all would be nothing more than informed speculation.
In either case you have a towering figure in the field to lean on, Norman doing a great job of critiquing the revisionists in the paper i link to above, and Warder giving sufficient hints to those who want to go looking for the seams amongst the suttas so to speak.
(Warder for example in Indian Buddhism suggests that the order of the NIkayas where at the time of their compilation taken to be in order of authority, i.e DN is more authoritative than MN which is more authoritative than SN which is more authoritative than AN, which I agree with, they also seem to hint that the grammar of DN is in general more uniform than the subsequent nikayas, and to hint that the core parts of he Atthakavagga and the Parayanavagga are very early.)
All these positions are of course more or less “thumbnail sketches” of more nuanced positions, obviously as you point out there are late verse forms in the parayanavagga, there are late forms in the DN, the Nikayas as we have them now are not in the form that the teachings where organized in the buddhas lifetime, which we know because the Nikayas themselves tell us this.
All that said, I feel like while we can probably definitely rule out the “Extreemists” like Schopen’s claim that “we can know nothing” prior to the common era and like the “Buddha taught the abhidhamma to his dead mother” in 450bce, it is still open to reasonable people to take opposing views on balance as to whether and to what extent we can say that the contents of the Nikayas reflect the actual teachings of the Buddha in their lifetime.
I guess I am a “compatibablist” in that I think on balance it looks like the canon formed over centuries and some ideas are absent from earlier layers and become prominent in later ones and the most likely explanation is that this is because they where new ways of explaining the truth, some of which where possibly invented by the early monastics not by the buddha themselves, I just think that this make little difference at the end of the day as the ideas being explained are compatible and parsimonious, not contradictory or radically different.
Metta
1 Like
Thank you Bhante, this seems to sum it up.
2 Likes
Yes, I agree completely.
The two vaggas are part of Sutta-nipata. According to Ven. Yinshun, he suggests that the anga 4. Gatha corresponds to certain texts of Sutta-nipata, compiled in the Khuddaka-nikaya rather than being made part of the 4 basic Agamas/Nikayas (p.10, note 34):
Pages 10-11 from The Fundamental Teachings of Early Buddhism Choong Mun-keat 2000-2.pdf (161.9 KB)
1 Like
Fun fact about that: people often cite this detail, but they rarely look at the context. The Atthakavagga is cited in two episodes in the EBTs, once in SN 22.3, and once in the Vinaya, which is repeated at Ud 5.6 (the latter passage adds the specifier “sixteen” to show the number of suttas).
On both occasions it is in the context of Mahākaccayana in Avanti. Yep, that’s right. One specific teacher, on the borders of the old region of Buddhist expansion, is responsible for 100% of the early citations of the Atthakavagga, once directly (SN 22.3) and once through his student Soṇa.
Perhaps we shouldn’t be regarding it as a specially early text at all, but one that represents an early period of Buddhist expansion. Might the poetic style and specific subject matter represent, not temporal, but regional characteristics? Which would make sense, given that the Parayanavagga then stems from further down the Southern Road.
And perhaps our Pali text was maintained and taught to subsequent generations of students in Avanti who followed Mahakaccayana’s lineage, including a monk you may have heard of: Mahinda, the son of Ashoka and founder of Buddhism in Sri Lanka, who famously stayed in Avanti. Mahinda only lived a few teaching generations after Mahakaccayana; a similar gap to that between us and, say, Ledi Sayadaw or Ajahn Mun.
Sure, that sounds reasonable enough. But I am left wondering: have you read the Apadanas? The Buddhavamsa? The Cariyapitaka? I would encourage it!
The striking thing about dating canonical literature is not that there are some differences, but that the gulf between early and late is so vast. Once you read the late texts—dating maybe 200+ years after the Buddha—any differences within the EBTs seem trivial.
The problem with this, as with Wikipedia’s “so long as it is referenced” style, is that it doesn’t take into account the grounds on which conclusions are reached, and the growth of knowledge over time.
Warder was the senior scholar, and he wrote he Pali Metre, the foundational work on the subject. This was based on his Phd work submitted in 1954 and published later in 1967. And it is from Warder that the idea of dating verse based on metre mostly stems. But this was a foundational, general survey, described by him as a “preliminary outline”. He says:
we should hope to be able to arrange our texts in chronological order, using a criterion more objective than any proposed hitherto.
Okay, what then of the Atthakavagga and Parayanavagga? Warder makes a chart of his proposed dating of verses, then says:
The Atthakavagga and Parayanavagga are omitted: according to Oldenburg’s figures the Atthaka tutthuba is 85% upajati-rhythm, which would make it later than S1 (Sagathavagga) (81%), whereas we have suggested that the vatta of these vaggas is extremely old. The problem of these texts remains, but I consider it certain that they cover a very wide range of time. The Atthaka may be much later than the Parayana (it includes the Tuvatakasutta in giti): whilst the Atthaka is almost entirely in tutthuba, the Parayana contains a good many vattas, although these are found mainly in the frame story.
Are you following? I’m not, BTW, no shame! But what I do understand is that Warder’s conclusions about dating the Atthakavagga and Parayanavagga are hedged about with layers of uncertainty due to the incomplete and contradictory nature of the evidence.
Norman then went on to make a series of studies of Pali verse over several decades, editing, translating, and annotating most of the early verses in detail. In effect, he fleshed out Warder’s outline. He concluded of the Sutta Nipata that “dating by metre is not particularly helpful" and criticized some of Warder’s conclusions. Thus even Warder’s partial and cautious conclusions were not always supported by later research.
These two bodies of work—a student’s research in the early 50s, and the lifetime scholarship of the greatest 20th century Indic linguist—are by no means comparable. But where they do agree is that neither of them support any strong conclusions regarding dating of the Atthakavagga and Parayanavagga on metrical grounds.
Again, no-one disputes the fact that these chapters contain much that is among the oldest strata of Pali verse. I’d just like to see the over-interpretation of the Atthakavagga and Parayanavagga reined in a little. Sorry if I’m bashing away at this point for too long! I’m currently writing essays on all these suttas!
11 Likes
I think this highlights another issue that somehow has been completely ignored in this discussion, namely that the Buddha did not speak Pali. Moreover, a number of renowned scholars in the field like von Hinüber, postulated the existence of a Buddhist MIA koine preceding Pali. This koine was supposed to be understood by speakers of all or most MIA dialects and as such must have been stripped of the most conspicuous dialectal characteristics (like Eastern r>l or North-Western retention of all three sibilants s, ş, ś).
One Bryan Levman conducted a comparison of Pali phonological and grammatival forms with those of Western and North/North-Western Pali edicts. His conclusions that merit at least a more in-depth research are that Pali that we know has ist specific linguistic features because it was first wrote down (or codified) in the North/North-Western are of the Middle Indic civilization, i.e. Gandhara and adjacent regions. The reasons for this are threefold: that the concept of writing was first introduced to Middle Indians in that region, most likely as Persian Aramaic writing > Kharosthi > Brahmi, that it was an area of intensive trade between Persia, India, Afghakistan, etc., and finally because the Northern and North-Western dialects were more prestigious than, say, their Eastern counterparts. However, thisnporcess of Pali-building was not instantaneous and must have taken quite some time.
What that means is that not only dating doctrines by meter is not very reliable as Ven. Sujato and Norman remarked, but also dating doctrines by Pali grammar. More archaic Pali grammar and lexemes may mean an earlier date of translating from the supposed MIA Buddhist koine into Pali as a Kunstsprache for NW Middle India and not being an earlier text per se. It also may mean that these verses were spared grammatical and terminological unification with the rest of the texts in Sutta Nipata and older parts of the prose Nikayas for accidental reasons. E.g., different translators, different line of transition, high esteem of the verses for their æsthetic value, regional variation, regional peculiarities in terminilogy and exposition of the doctrine (e.g. the use of certain Pali terms in the Thai forest tradition), personal preference, etc.
In the same vein, comparing different English and German translations of the same Pali texts may give you completely different doctrinal pictures and different diachronical language strata (although, admittedly, Pali is infinitely closer to the MIA koine than English to Pali). And I believe it is one of the bains of the Pali translations: the translator almost always translates texts through the lense of his personal views and preferences, sometkmes to the detriment of a more exact expression. As Levman shows, thisbis very likely to have been the case for Pali as well, albeit to a smaller extent.
3 Likes
I’m not so confident that we have a really clear picture of this, but the point remains, that the tiny details of linguistic variation may, or may not, have been subject to various transformations before reaching us.
To be fair, the arguments on the primacy of the Atthakavagga are, indeed, well-formed arguments, because they are based on several independent grounds: the language, the metre, the cross-references, the Niddesa, the content. It’s just that some of these are not so strong points as people think, and the conclusions we can safely draw from the remaining points are fairly modest.
3 Likes
Then it would be a big coincidence that he happened to sprak a dialect that in most of its features was almost the direct opposite of the language variaties found in the Eastern area of the Ganges basin. Pali being close to Western or Northwestern variaities of MIA dialects is hardly a disputable fact.
What language the Buddha speaks is unknow. But, the Buddha in Vinaya, Cullavagga (Vin. II, PTS, p. 139 ), advises bhikkhus to use your own language (sakāya niruttiyā) for his teachings.
Well, I would say that for some reason Karpik stays with the argumentstion of a more or less general nature instead of rolling-up hisbsleeves and coming down to the nitty-gritty of particular linguistic forms.
Even a brief review of those will show that Pali overwhelmingly possesses features of the Western / Northwestern dialects (exact location is irrelevant for the current discussion with only vestiges of features of Eastern forms). The Ashokan inscripts are merely 150-200 years away from the Buddha so imagining that the real MIA dialects diverged so far from a dialect not yet having all these features is also a stretch. Besides, there are multiple proofs that the Ashokan pillars used an archaic, solemn language that was already different from the regular vernacular, which is also a sign that Ashokan dialects may represent the approximate state of the language at the Buddha’s time.
There are other hypotheses that were bot tslen knto account:
-
the Buddha spoke an Eestern dialect that was Westernized and froze in its forms and grammar just like Vedic did;
-
the Buddha spoke a lingua franca thatvwas Westernized and given an archaic flsvour afterwards.
So yeah, while all in all expressing some fresh ideas, this paper is not a real deal breaker, unfortunately.
2 Likes
I can recommend Richard Gombrich’s little book “Buddhism and Pali”, for some interesting food for thought on the “Did the Buddha speak Pali?” question.
(Quick answer: -yes, perhaps he did!)
2 Likes