Bhikkhu Bodhi's "Aggregates and Clinging Aggregates"

A classic, thanks for bringing this up! I wanted to check this before, but didn’t have a copy, so now I do.

It’s been ages since I read this, so I’m going to go back over it using the greater range of sources we have available today.

Firstly, to summarize the main doctrinal point, Ven Bodhi’s argument establishes that an arahant’s experience is still within the realm of suffering. They have escaped suffering in the sense of being beyond rebirth, but in the present life their experience is still dukkha. This is certainly correct, and can be confirmed on multiple grounds.

Also, let me just point out the elegant and meaningful structure of the essay. He begins by formulating the problem in terms of the four noble truths, draws out certain implications of that, examines those implications in the light of various sources, and draws it back with a return to the four noble truths at the end. When I see that, I think, “Huh, if only all my essays were like that!”

But let’s look at how the article establishes its case.

The main source text is SN 22.48. In his translation of this in Connected Discourses, nearly thirty years after the original article, Ven Bodhi includes the following note, which both handily summarizes his earlier argument, and confirms that he still agrees with his earlier interpretation:

This sutta is quoted and discussed at Vism 477-78 (Ppn 14:214-15), in relation to the difference between the aggregates and the aggregates subject to clinging. The key terms distinguishing the pañc’ upādānakkhandhā from the pañcakkhandhā are sāsava upādāniya, “with taints and subject to clinging.” The pañc’ upādānakkhandhā are included within the pañcakkhandhā, for all members of the former set must also be members of the latter set. However, the fact that a distinction is drawn between them implies that there are khandha which are anāsava anupādāniya, “untainted and not subject to clinging.” On first consideration it would seem that the “bare aggregates” are those of the arahant, who has eliminated the āsava and upādāna. However, in the Abhidhamma all rūpa is classified as sāsava and upādāniya, and so too the resultant (vipāka) and functional (kiriya) mental aggregates of the arahant (see Dhs §§1103, 1219). The only aggregates classed as anāsava and anupādāniya are the four mental aggregates occurring on the cognitive occasions of the four supramundane paths and fruits (see Dhs §§1104, 1220). The reason for this is that sāsava and upādāniya do not mean “accompanied by taints and by clinging,” but “capable of being taken as the objects of the taints and of clinging,” and the arahant’s mundane aggregates can be taken as objects of the taints and clinging by others (see As 347). For a detailed study of this problem, see Bodhi, “Aggregates and Clinging Aggregates.”

One detail, and this is from a highly unreliable memory of something over a decade ago, I think Ven Bodhi said he would change one aspect of this essay, namely the equation of the meditation mentioned as a “pleasant abiding in the present life” on page 94 with the arahant’s fruition. This seems to be indirectly confirmed in his notes in CDB. In his note on this sutta (SN 22.122) he doesn’t comment directly on this detail, but refers to a note elsewhere, note 332 on SN 17.30, where he says:

Spk: The pleasant dwellings in this very life (diṭṭhadhammasukhavihārā) are the pleasant dwellings in fruition attainment. For when a meritorious arahant receives conjee, sweets, etc., he must give thanks to those who come, teach them the Dhamma, answer questions, etc., and thus he does not get a chance to sit down and enter fruition attainment. Spk’s identification of the “pleasant dwellings” with fruition attainment is certainly too narrow. The term usually means the jhānas, as at II 278,10-11

Now, to check the relevant parallels, it seems that SN 22.48 has just one Chinese parallel at SA 55. This seems to be closely parallel to the Pali in most respects. However, it adds an extra detail at the end. Whereas in the Pali text the only difference between the “aggregates” and the “grasping aggregates” is that the latter are “liable to defilement and grasping”, the Chinese text adds the following:

And what is the grasping aggregate? Whatever form is with defilement and grasping, and that form, in the past, present, or future, gives rise to greed, hate, and delusion, and so on as above to the many kinds of mental affliction.

Perhaps @vimalanyani or another Chinese speaker can improve this. But it seems to me as if the 彼 here indicates a relative clause of the very common “yo … so …” type, indicating that the extended phrase is part of the same sentence, not a separate addition. It also seems to me that the text is abbreviating and referring to a longer list of afflictions. This isn’t necessarily significant, but it does read to me like this is an organic part of the text, and there is at least a possibility that the Pali has suffered loss.

The text specifies here that the grasping aggregates give rise to suffering. Now, this does not directly contradict Ven Bodhi’s reading. However his basic argument is that the arahant’s aggregates may give rise to grasping in another person. This has always seemed like slender reading to me. And it seems even more so here. Surely, the overarching theme of the Buddha’s teaching is how grasping gives rise to suffering for oneself.

Ven Bodhi’s argument finds support in the Abhidhamma and commentaries, and indeed one of the purposes of the article seems to be to establish that the Theravadin tradition is, in this respect, a correct reading of the suttas. If we leave this to one side, though, we are left with a number of what I think are rather weak links in the chain of argument.

The argument rests on the premise that there is some kind of meaningful distinction drawn between the “aggregates” and the “grasping aggregates”. In the original essay, Ven Bodhi is careful to point out that these are “not contrasted”, while in the note above he refers to the “difference between” the two terms.

But it is not really clear to me that there must be any real difference. One might define “man” as “homo sapiens” and “mortal man” as “homo sapiens subject to death”, but they are the same thing. The latter definition merely adds more detail.

Rather than denoting different kinds of thing, perhaps the point of the distinction between “aggregates” and “grasping aggregates” is to show them from different teaching perspectives. The former is a simple presentation of “what is” whereas the second explains how “what is” gives rise to suffering.

The strongest textual support for the argument that an arahant possesses the “grasping aggregates” is SN 22.122, where there is a discussion between venerables Sāriputta and Koṭṭhita. This supports the idea that the arahant has the “grasping aggregates”. In my translation:

“But Reverend Sāriputta, what things should a perfected one meditate on?”
“Reverend Koṭṭhita, a perfected one should meditate on the five grasping aggregates as impermanent, as suffering, as diseased, as an abscess, as a dart, as misery, as an affliction, as alien, as falling apart, as empty, as not-self. A perfected one has nothing more to do, and nothing that needs improvement. Still, these things, when developed and cultivated, lead to blissful meditation in the present life, and also to mindfulness and situational awareness.”

The problem, though, is that there is nothing to connect this sutta with the primary source text SN 22.48/SA 55, which says nothing about the arahant.

Another weak point here is that this sutta does not appear to have any parallels. Normally this would not be necessarily significant, but in a case where a text appears to be supporting a sectarian doctrine, its absence from the other collections is noteworthy. It is quite possible, however, that similar statements are found elsewhere.

Moving on, I come to the weakest link in the chain. As indicated in the summary above, and stated in the original essay on page 95, the aggregate of form is said in the Abhidhamma to have no “bare” version. Thus the basic text explicitly states that there are five “bare” aggregates, but the Abhidhamma/Bodhi reading requires that there be only four. Given that so much of the argument rests on slender inferences, such a clear and direct contradiction with the primary source text is untenable.

The point of this classification is that while the arahant has no attachment to their own aggregates, others may be attached to them. Someone might see an arahant’s body and have desire for it, for example. Someone can even read an arahant’s mind and conceive a desire for their mental states. The only exception to this, in the Abhidhamma reading, is when an arahant is sitting in the absorption on fruition. In such a state, their mind is inaccessible to anyone (except possible a Buddha or another arahant) so cannot give rise to grasping.

This is an obscure edge case, and it is not referred to in the sutta. And as I indicated earlier, Ven Bodhi seems to have some reservations about bringing in such a state here.

In conclusion, I think the Abhidhamma/Bodhi reading rests on overly slim inference while directly contradicting the source text.

Rather, it seems to me simpler to see the source text as not pointing to different sets of things, but two different perspectives on the same thing. The “bare” aggregates show “what is”, while the “grasping aggregates” show how grasping to “what is” leads to suffering.

This kind of analysis is not uncommon in the suttas. Compare, for example, SN 12.20, which draws a distinction between “dependent origination” and “dependently originated phenomena”. The point here is not that these are distinct things, but that they are two aspects or ways of looking at the same process.

Thus the question of whether the “bare” aggregates refer to an arahant does not even come up: that’s not what the Sutta is about.

This reminds us of an important principle in reading the suttas. When it comes to central teachings like the aggregates, we should avoid reading special or innovative doctrines into marginal, occasional, or obscure cases. The main teachings were stated again and again in different ways, and it is clear that this is a central mode of the texts as we have them. Unless the reading forces itself on us, we should prefer a reading that adds as little as possible to the content of the suttas elsewhere. This is an aspect of what I call the “principle of least meaning”.

While it makes for impressive exegesis and fascinating connections, when we chase slender meanings into obscure texts, alas, we all too often arrive at a dead end.