Exercise #6 of the 9 corpse contemplations MN 10, AN 6.29

Discrepancies on the number of bones in exercise #6 from the 9 corpse contemplations.

  • B.Sujato has 12 bones in MN 10 english, pali has 13 bones.
  • what is the EBT perspective on this?
  • AFAIK only AN 6.29 and MN 10 have the 9 corpse contemplations. Can someone confirm?
  • in AN 6.29 english, bodhi has 7 bones, and in MN 10, he has 14 bones.

My guess: 7 bones was probably the original,
and the different Theravada sects started adding different number of bones.
I like 7 also because it reminds one of 7 bojjhanga. It also seems to sufficiently cover
the major areas to feel like you contemplated most of your skeleton.

The main reason I ask this is in my translations I’ve been by default going with
whatever sutta central does. But in this case, I’d like to go with whatever is
most likely to be the earliest EBT perspective. So is it the 7?

MN 10 pali from sutta central
Aṭṭhikāni apaga­ta­samban­dhāni disā vidisā vikkhittāni,

  1. aññena hatthaṭṭhikaṃ
    aññena pādaṭṭhikaṃ
    aññena gop­phakaṭ­ṭhi­kaṃ
    aññena jaṅghaṭṭhikaṃ
    aññena ūruṭṭhikaṃ
    aññena kaṭiṭṭhikaṃ
    aññena phāsukaṭṭhikaṃ
    aññena piṭṭhiṭṭhikaṃ

  2. aññena khandhaṭṭhikaṃ

  3. aññena gīvaṭṭhikaṃ
    aññena hanukaṭṭhikaṃ
    aññena dantaṭṭhikaṃ

  4. aññena sīsakaṭāhaṃ

    MN 10 b.sujato english from sutta central

    1. here a hand-bone,
      there a foot-bone,
      here a shin-bone,
      there a thigh-bone,
      here a hip-bone,
      there a rib-bone,
      here a back-bone,
      there an arm-bone,
      here a neck-bone,
      there a jaw- bone,
      here a tooth,
    2. there the skull—

    AN 6.29 (pali from sutta central)
    1.aññena hatthaṭṭhikaṃ
    aññena pādaṭṭhikaṃ
    aññena jaṅghaṭṭhikaṃ
    aññena ūruṭṭhikaṃ
    aññena kaṭiṭṭhikaṃ
    aññena phāsukaṭṭhikaṃ
    aññena ­piṭṭhi­kaṇ­ṭakaṭ­ṭhi­kaṃ
    aññena khandhaṭṭhikaṃ
    aññena gīvaṭṭhikaṃ
    aññena hanukaṭṭhikaṃ
    aññena dantakaṭṭhikaṃ
    12. aññena sīsakaṭāhaṃ

    english AN 6.29, b.bodhi has 7 bones
    disconnected bones scattered in all directions:
    here a handbone, there a footbone,
    here a shinbone, there a thighbone,
    here a hipbone, there a backbone,
    and there the skull.

    english MN 10, b.bodhi
    here a hand-bone, there a foot-bone,
    here a shin-bone, there a thigh-bone,
    here a hip-bone, there a back-bone,
    here a rib-bone, there a breast-bone,
    here an arm-bone, there a shoulder-bone,
    here a neck-bone, there a jaw-bone,
    here a tooth, there the skull


For AN6.29 the number of bones is seven in the Sri Lankan Buddha Jayanti edition and in the PTS edition, whereas the Chaṭṭha Sangāyana edition and the closely related Mahāsangīti edition (on which Suttactentral is based) have twelve bones.

For MN10 (and DN22, by the way) the number of bones in the Buddha Jayanti edition and PTS edition is again 7, whereas in this case the two other editions have 13. Ajahn Sujato seems to have missed out the ankle bone, the gop­phakaṭ­ṭhi­kaṃ. I don’t know where Bhikkhu Bodhi/Ñāṇamoli got 14 from.

These comtemplations are also mentioned in MN13 and MN 119.

I would agree with you that the shorter list is likely to be earlier and thus more authentic. There is a tendency in the suttas for things to get expanded and/or standardised over time, and this may well be what has happened in the case of the Chaṭṭha Sangāyana/Mahāsangīti version.

1 Like

Who needs ankles anyway? All they do is join your foot onto your leg!

I wonder if there’s anything in the chinese or sanskritic texts on this/

1 Like

i was just thinking of checking the agamas also…
in b.anaalayo’s satipatthana perspectives book:
for madhyama agama has:
… a skeleton without flesh and blood, held together by the sinews …… disconnected bones scattered in the main and intermediate directions, here a hand bone, elsewhere a foot bone, elsewhere a shin bone, elsewhere a thigh bone, elsewhere a hip bone, elsewhere a back bone, and elsewhere a skull …… bones bleached white, the colour of shells …… bones heaped up, more than a year old …

for ekott:
… a corpse whose joints have come apart, [with its parts] scattered in different places, here a hand bone, there a foot bone, or a shin bone, or the pelvis, or the coccyx, or an arm bone, or a shoulder bone, or ribs, or the spine, or the neck bones, or the skull. Again one uses the contemplation that one’s own body is not different from that: “I will not escape from this condition. My body will also be destroyed.”

another interesting difference, where as MN 10 has 9 stages:
The parallel versions differ in the way they divide the actual stages of decay. The Madhyama-āgama description presents five, the Ekottarika-āgamaversion eight, and the Majjhima-nikāyadiscourse nine such stages. Nevertheless, the main import of the practice appears to be the same, in that contemplation proceeds from a dead and bloated body to a body that is eaten by various animals.

in AN 4.13(?) or 14, the padhana sutta sutta that gives a recommended practice for each of the 4 right efforts, the fourth one, maintaining kusula, has 5 contemplations: atthika sanna, pulavaka sanna, vinilaka sanna, viccidhaka sanna, uddhamataka sanna. I think this differs in number between BJT and CST as well. I think BJT maybe has 6, CST has 5 if I recall correctly.

I was just thinking the 5 stages of MA maybe matches the 5 of AN 4.13. Perhaps what happened during the Buddha’s time was whenever he gave a talk on the subject he just started naming bones and stages of decay extemporaneously without regard for standardizing on a set number of parts or stages, and as a result the large discrepancies everywhere in exact number.

thanks bhante.
I didn’t know about MN 13. I was aware of DN 22 and MN 119, as I consider them both just derivatives of MN 10.

Is there any plan for sutta central to more explicitly show some of these relationships between similar suttas, and their estimated chronology from EBT perspective?

after reading “history of mindfulness” a few years ago, it really changed my life and how i look at the suttas.

i understand you’re ordained theravada monks and can’t simply just remove MN 10 and DN 22 from suttacentral and replace it with MN 10 mula. But how about the Tv ahidhamma vibhanga version of MN 10? It’s pretty close to MN10-mula, and it’s theravada. at least the vibhanga version could be listed as the first parallel right? currently it’s not even listed as a parallel(just checked).

i know b.anandajoti has a translation of that vibhanga version. i’ll volunteer to make a concise version that elides text and matches lengthwise more closely with MN 10, and MN 10-mula. (b.anandajoti’s version has every repetition expanded out, and it’s massive )

Although this part of the Vibhaṅga is called the Suttanta-bhājanīya (“analysis according to the suttas”), the content does not quote the suttas directly. Rather it quotes selected passages, and as such they are not really sutta parallels. Moreover, even the selected passages are often not verbatim quotes. If you were to include this sort of thing, you are really opening up a pandora’s box, since you would then have to include any pericope from the early texts which is found in any sort of Buddhist literature. This would be an enormous task.

We do plan to gradually enrich our data, and showing chronological connections might be part of that. At the moment our developers are hard at work putting into place the technology to make such things possible.

Indeed it should be listed as parallel. Originally our parallels covered the 4 nikayas only. We are working on extending that, and should have several thousand more parallels in the next few months.

We have it already! He spent ages trying to work out exactly how the repetitions were to be expanded, and as far as I know, this work is unique in achieving this. Unreadable, but unique!

If you want to make elided or Mula texts or whatever, great: that’s what Discourse is for. The main SC is only for the canons, but feel free to add such materials here.

Well, not really, you can still put limits on it. Our scope is the Tipitaka and parallel literature, which obviously includes the Abhidhamma. There’s actually not a huge amount of parallels in such material: maybe a few hundred. (The Dhammapada, by way of contrast, has maybe ten thousand.) Where appropriate we’ll mark such parallels as “quotes”, etc. But it is, I think, useful to see such conection, or alternatively, the lack of them. Moreover, in some Sanskritic Abhidharma texts (Dharmaskandha) there are some precious examples of sutta parallels in old Sanskrit. So while such material is not our main focus, we do aim to cover it. Currently Ayya Vimala is working with this, extracting the cross-reference data found in the Mahasangiti text.

1 Like

In MA 81, MA 98, and MA 99, I see a consistent list of eight bones:

  1. 足骨
  2. 膊骨
  3. 髀骨
  4. 臗骨
  5. 脊骨
  6. 肩骨
  7. 頸骨
  8. 髑髏骨

In Kashmiri dhyāna manuals… I see seven bones in 達摩多羅禪經, and nine bones in 禪祕要法經. The latter has a lot of material on skeleton visualizations.

I am not sure why the Abhidhamma should be included. It seems to me that this muddies the waters. The Suttacentral “About” page reads: “It focuses on the texts that represent ‘Early Buddhism’, texts preserved not only in the Pali Sutta and Vinaya Piṭakas but also in Chinese and Tibetan translations and in fragmentary remains in Sanskrit and other languages.” Is Suttacentral about Early Buddhist Texts or is it about the Tipitaka? These are quite different things.

When I was in Thailand recently I met one of the main exponents for the Buddhavacana movement and he immediately criticised Suttacentral for including the Abhidhamma. I understand his concern, for this approach may give the impression that Suttacentral considers the Abhidhamma as an EBT. And once you start quoting texts outside the EBTs, why stop with the Abhidhamma and other late Canonical texts? Limiting it to the Tipitaka is in some ways quite arbitrary, since this is just a container and it says little about authorship or even the chronology of texts.

I certainly agree that the exposition of satipaṭṭhāna in the Vibhaṅga is interesting, and that it may well point to a very early version of this sutta. The same, however, could be said about any quote in any chronological strata of Buddhist literature, whether canonical or not. It seems to me that these sorts of arguments, although very interesting and important, should be confined to independent monographs and essays.


We made a decision long ago to include the Abhidhamma. Obviously our focus is the early texts. But, well, the Abhidhamma is there. It seems to me that it is unproductive to try to exclude it.

The Abhidhamma’s alright, I don’t have anything against it. It’s just the way that a generation of Buddhists organized and structured the main teachings. Go and have a read! The problem is not Abhidhamma, it’s fundamentalism.

As it stands, SuttaCentral is one of the few sites on the web that actually hosts Abhidhamma translations. That means that someone who wants to read Abhidhamma will be likely to do it here. If they don’t come here, they’ll go somewhere else. But when they’re on SC, they’ll see, even if they don’t want to, the context. Abhidhamma students often have been schooled in a fundamentalist “Pali-only” background. Seeing the fact that there are texts in other traditions, that the links between these in the Suttas and Vinaya are strong, and in the Abhidhamma are weak, cannot help but have an effect.

Fundamentalism isn’t overcome by fundamentalism in this world (to paraphrase the Buddha!) If we assume that people are too stupid to work things out for themselves, does that make us educators or indoctrinators?

Fundamentalism is overcome by being open and encouraging. Inquiry, information, and education are not something to be feared. The best way to understand how the Abhidhamma texts are developed from the EBTs is to actually read them. Otherwise you’re just accepting one authority over another.

If someone wants to make a canon that is solely early texts, go for it. All our material is on github, and it’s literally a minute’s work to fork our project and start your own. It would be very interesting.

Fix everything that’s wrong! Start by excluding the Abhidhamma of course, but also the Lakkhana Sutta, the Atanatiya Sutta, the Sangiti Sutta, the Dasuttara Sutta, the Isigili, the Mahasamaya, the Girimananda, the Mahasatipatthana, the Mahasudassana, 80% of the Vinaya except for the basic rules, all the books of the Khuddaka except the six early ones—and maybe not even some of them?— the Anupada, the Mahapajapati fable, and so on.

Then start the hard work. Then do the same for the Chinese texts, the Sanskrit and Tibetan. Good luck getting anyone to agree with your choices!

Some EBT-ists, so I hear, accept the Atthakavagga only. That would make things much easier. But for myself, I think all boundaries are to some degree arbitrary, and so I prefer to keep them a little loose and flexible.

This is pretty speculative, but some of the design shifts we are talking about under the hood might make it possible to designate and select multiple overlapping versions of the “early texts”. That is, someone could go through the texts and designate what they thought was early and what was late, someone else could do their own version, and so on. A reader could select to view any of these. Maybe!

1 Like

I agree. The Canonical Abhidhamma appears to me to mostly be lists of relationships and so on that systematize the information scattered somewhat randomly amongst the Suttas. It’s a bit heavy going for me to study in detail, but it’s interesting to see how it develops.

Unfortunately, when people say “Abhidhamma” these days (whether they are lovers or haters) they are mostly thinking of much later developments. I haven’t found references to mind moments and the like in the Canonical texts. But it seems difficult to find many who have seriously studied those canonical texts. Most enthusiasts (and non-enthusiasts) seem to rely on the Visuddhimagga and the Abhidhammatthasangaha, and the latter is well over a millennium later than the suttas.

Any serious study of the suttas necessarily involves some systematization to make sense of the teachings. Whether it’s something relatively simple, like Bhikkhu Bodhi’s thematic lists in In the Buddha’s Words or the AN translation, complex analyses such as the Abhidhammas or modern philosophical approaches (Nanavira, etc), or the historical scholarship approach championed here.

But there are real issues here. Nothing is lost in changing one’s mind, if that becomes necessary. One needs to be careful of the sunk cost fallacy.

If the focus is on early texts, what exactly are the limits? Or is the point not to have any boundaries and just include whatever feels right? Is it not useful to have some sort prior policy?

You say the scope of Suttacentral is the Tipitaka. I just want to reiterate my point that this is an artificial container that does not have much to do with whether a text is early or not. Worse, I understand that you also intent to include the Jataka stories, and these are not even canonical. The only advantage with setting the limit at the Tipitaka is that it is (mostly) clearly defined.

But since we are supposed to deal with the EBTs, why use the Pali Tipitaka as the standard for which texts should be included? All the early schools appear to have had their own Tipitakas, and we know that they often varied quite significantly in content. Shouldn’t all this material be included, at least in principle (since we don’t actually have many of these texts)? And what about the Chinese Tipitaka: shouldn’t we then include it too in its full version? If we did this, we would have a large number of texts that would be canonical in one school but not others. If a text is canonical in one school but not others, should non-canonical parallels in these other schools be included, for comparative purposes?

We know that the Abhidhamma and some texts in the Khuddaka were not considered canonical by certain early schools. I think a good approach for deciding which early texts should be included is to look at what was accepted by all early schools. We have good reasons to believe that they all accepted four Āgamas and a Vinaya Pitaka divided into a section on pātimokkha rules and their explication and a more general Khandhaka type section. So that to me would seem like a natural starting point. And perhaps there is no need to go beyond these texts.

Of course, it is true that these texts too include non-early material. But the point is that there is a fairly clear dividing line between the four Āgamas and the Abhidhamma, whereas the dividing line between early and slightly later suttas is much more blurred. The Dhammasaṅgaṇi, for instance, is very different from anything found in the suttas.

I hope you will forgive me for saying all this. I do not mean to simply be difficult. But I do think a bit more clarity about what Suttacentral is all about would be useful. I thought the main purpose was to give people access to the suttas, but it seems your conception is much broader.


This is one of the problems of being too restrictive:

There appear no parallels for the Atthakavagga, or just about anything in the SNP, for that matter: https://suttacentral.net/snp

Of course, that may be just a matter of the imperfect sets of parallels that exist, but, if taken to an extreme (which is probably not what is intended) this sort of argument risks excluding the Metta, Ratana, and Mangala suttas. Not to mention the Pārāyana vagga suttas that Bhikkhu Nanananda is fond of quoting…

This refers to certain modern individuals and does not relate to the early schools.

It does actually have a parallel in Chinese translation, as does the Pārāyana vagga and the Ratana Sutta.

Thanks for that. I was wondering why a collection that seems so important in the Theravada school would not have parallels. Hopefully that information will be added to SC at some point…

Yes, we’re still lacking parallels for most of the verse collections.