On Translations of 'Khandha' and Inconsistency

[Vaddha]

One of the most important technical terms in Buddhism is the Pāḷi 'khandha, Skt. ‘skandha.’ The translation many in the English speaking world are familiar with is ‘aggregate,’ and this is the translation currently in use by Ven. @sujato here on SuttaCentral. I would like to discuss this a bit, mainly pointing out an inconsistency I see in the translation of related technical terms.

This came to mind when reflecting on the following stock definition in the suttas:

And what is rebirth?
The rebirth, inception, conception, reincarnation, manifestation of the aggregates, and acquisition of the sense fields of the various sentient beings in the various orders of sentient beings.
Yā tesaṁ tesaṁ sattānaṁ tamhi tamhi sattanikāye jāti sañjāti okkanti abhinibbatti khandhānaṁ pātubhāvo āyatanānaṁ paṭilābho

Notice anything?

‘Āyatana’ is translated as ‘sense field.’ But ‘āyatana’ does not mean ‘sense field,’ it simply means ‘field’ or ‘domain.’ ‘Khandha,’ on the other hand, is translated as ‘aggregate.’

Both of these terms are equivalent in that they are (1) both technical terms used in Buddhism to refer to models for dividing up experience, and (2) both ambiguous as stand-alone words without prior context to understand what they refer to within Buddhism.

The favoritism to translate ‘āyatana’ as “sense-X” while leaving ‘khandha’ general and un-qualified is not unique. Ven. Bodhi and Ven. Suddhāso have both done the same. ‘Khandha’ is left translated as a meaningless stand-alone term, but ‘āyatana’ is provided with standard supplementary translation material. Ven. Ānandajoti has the following translation:

khandhānaṁ pātubhāvo, āyatanānaṁ paṭilābho.
the manifestation of the components (of mind and bodily-form), the acquisition of the sense-spheres.

This is a bit better, in that both terms are qualified. But notice here there is the impression given of literal accuracy with ‘sense-spheres’ in that this is given with no parentheses, right next to ‘components (of mind and bodily-form)’ which does give the supplementary material in parenthesis, implying it is added. Still, then, there is a bias to translate ‘āyatana’ as though it needs and entails qualification, whereas ‘khandha’ as though it needs to be stand-alone and any qualification is somehow less literal and secondary than with ‘āyatana.’

I’m not proposing a particular translation solution — that is up to each translator. But I am calling for some consistency, and possibly for some re-consideration of “aggregates” as the translation for “khandhā.” Hoping Ven. @sujato would consider this for SuttaCentral

I personally would advocate not for reducing ‘āyatana’ to “fields” but rather with supplementing (and/or replacing) “aggregates” to fit with the standard with the former, as translators clearly have believed to be important for the sake of clarity. Translating “āyatana” just as “fields” is a perfect demonstration of how non-sensical “aggregates” on its own is. In this case, the text itself is ambiguous, so following the convention of the text would be fine; but at least using a more readable and accessible word like has been done with ‘āyatana’ would be nice.

Much mettā for all the translators out there and the work you do!

[sujato]

It’s an interesting point, thanks for raising it. We have discussed translation of khandha before, I looked for a better rendering but failed.

As a rule I would tend to refrain from adding extra explanation unless it is genuinely unclear in context. For example, where Bodhi has “spiritual faculties” I just use “faculties” because I think the context makes it clear.

Now, let us consider going both ways: a minimal translation:

the manifestation of the aggregates, the acquisition of the fields.

Okay, to me it feels unclear; “fields” (or other renderings like “spheres”) feels like not enough.

Now a maximal translation:

the manifestation of the aggregates (of mind and body), the acquisition of the sense-fields.

Notice that even Anandajoti puts (of mind and bodily-form) in brackets; he wants to make clear it is extraneous.

I’m never really comfortable with using “mind and body”, too many memories of Descartes. What else might we use? From the suttas, we might say:

the manifestation of the five aggregates, the acquisition of the sense-fields.

This doesn’t really say what they are, but it does specify what we’re talking about. or perhaps:

the manifestation of the aggregates of conditioned phenomena, the acquisition of the sense-fields.

Probably the most unproblematic doctrinally, but also clumsy.

I feel like there are really two issues here. One is that in the Buddhist community we have come to accept the term “aggregates” for better or for worse. The other is that, unlike with “sense”, there is no really compelling way of expanding it. But maybe someone can suggest one?

[Sasha_A]

@Vaddha

Notice anything?

‘Āyatana’ is translated as ‘sense field.’ But ‘āyatana’ does not mean ‘sense field,’ it simply means ‘field’ or ‘domain.’ ‘Khandha,’ on the other hand, is translated as ‘aggregate.’

Both of these terms are equivalent in that they are (1) both technical terms used in Buddhism to refer to models for dividing up experience, and (2) both ambiguous as stand-alone words without prior context to understand what they refer to within Buddhism.

If these words are indeed to be taken as technical terms with special meanings, and not just words, then ‘āyatana’ and ‘khandha’ are the names of these terms, and as the names of the special terms they do not need to be translated at all, nor expanded in brackets.

This is because any translation of the name of the term here, especially with the expansion in brackets, would be the translator inserting his own interpretation of the meaning of the term, rather than simply translating the original text as it is. If the translator is concerned about the reader’s ability to understand the meaning of the term, the appropriate footnotes can be provided with links to other suttas where the meaning of the term is explained. The same goes for the translator’s personal opinion on the meaning of the term - its place is in the footnotes, not in the translation itself, especially disguised as the original content of the translated text.

[sood]

@sujato

But maybe someone can suggest one?

Ven @Sunyo has aspects of existence in his translations, do you find that compelling, Bhante? That also suggests a meaningful way to expand aggregates with maybe - aggregates of existence.

[yeshe.tenley]

If the khandas are just the mind and body and that is how the Teacher understood them, then I don’t see a problem of just saying:

Maximally:

the manifestation of the mind and body, the acquisition of the sense fields.

If we are at all unsure that this is what the Teacher intended, then I’d prefer the minimal solution:

Minimally:

the manifestation of the aggregates, the acquisition of the fields.

I kinda prefer the ambiguity and to let people decide for themselves what the Teacher intended, but I think even ‘mind and body’ and ‘sense fields’ are ambiguous terms so…

One way or another, don’t let Descartes throw you Venerable!

[Vaddha]

@sujato

translation of khandha before, I looked for a better rendering but failed.

I’ve read up on some of those discussions. I definitely see the difficulty. I think ‘group,’ ‘grouping,’ or ‘heap’ are not bad for khandha. I think that even in the discourses this is why the standard term is in fact ‘upādānakkhandha’ — it qualifies what groups or heaps we are talking about.

@sujato

I would tend to refrain from adding extra explanation unless it is genuinely unclear in context.

Right, and I think it’s fair to produce a translation that is no less clear than the source text. Even in Pāli, the phrasing is ambiguous and unclear without familiarity with Buddhist categories.

@sujato

even Anandajoti puts (of mind and bodily-form) in brackets; he wants to make clear it is extraneous.

This is what I was pointing out in my post though: it’s a classic case of giving a mirage of false precision with brackets or parentheses. “Sense” is also extraneous, but why is it not in brackets?

I think that if translators decide to add extraneous explanation to ‘āyatana’ with no brackets or parentheses, there is no good reason the same is not being done with ‘khandha.’ Looking at the pattern, it looks like it’s just a case of someone doing this in the past and then all subsequent translators following suit as though it were default and standard.

@sujato

the manifestation of the five aggregates

I think this is quite good, and it’s certainly implied in the Pāli text that these are the five khandhā. I personally find this the most satisfying solution: there is a set of groups/categories, and when we add a qualifier like ‘the five,’ we delimit that to be a specific grouping of five groups. So it provides context, draws out the immediately implicit connotations of the source text, and it is not clunky or awkwardly idiosyncratic.

I think wherever the term ‘khandhā’ appears where it does not literally mean “[any] groups” but rather specifically is short-hand for the five groups important to Buddhism, it would be nice for the translation to say so. At least I personally will be taking that away from this.

I think if translators adopt the convention, like with ‘āyatana,’ of translating ‘khandhā’ as ‘five X,’ it also makes translations other than ‘aggregate’ more in reach. Because ‘five heaps’ or ‘five groups’ can only be read as a specific set of things that the reader should fill in from elsewhere.

Thank you, bhante!

[Vaddha]

@sood

aspects of existence in his translations

Not who you were addressing, but just to chime in:

I think ‘aspects of existence’ is nice. I like that one can read it and get a good sense of the actual meaning, while it also preserves the ambiguity of the Pāli in that it doesn’t explain what the ‘aspects’ are in the translation.

What I find less ideal is that ‘khandha’ is explicitly explained to refer to the fact that these are groupings of every instance of the five categories of phenomena. So the term ‘group,’ ‘heap,’ ‘grouping,’ or ‘category’ is actually significant in the understanding and explanation of the term in the suttas. Here is an example of what I mean:

”Sir, what is the scope of the term ‘heap’ as applied to the five heaps?”
“Kittāvatā pana, bhante, khandhānaṁ khandhādhivacanaṁ hotī”ti?
”Any kind of form at all—past, future, or present; internal or external; coarse or subtle; inferior or superior; far or near: this is called the heap of form.”
“Yaṁ kiñci, bhikkhu, rūpaṁ—atītānāgatapaccuppannaṁ ajjhattaṁ vā bahiddhā vā, oḷārikaṁ vā sukhumaṁ vā, hīnaṁ vā paṇītaṁ vā, yaṁ dūre santike vā—ayaṁ rūpakkhandho.

You couldn’t do this with ‘aspect,’ because ‘aspect’ isn’t a translation of the word ‘khandha.’ So it nullifies a core implication of the original term and the discussions of this term within the discourses.

That said, every translation has it’s pros and cons, and it’s good to have variety like this.

[yeshe.tenley]

But ‘five’ isn’t used in the source text so I don’t see why this should be added. Why is ‘five’ a matching specification to ‘sense’ when applied to the fields? If you’re going to add matching specifics that are not in the source text, then I think it would make sense for those specifics to be related somehow.

Maybe the solution is to just add ‘sense’ in brackets giving:

the manifestation of the aggregates (the mind and body), the acquisition of the fields (the sense organs).

?

The mind and body are commonly understood english words that are used outside of Buddhism so relating the aggregates to them is useful in describing what the Teacher might have been saying in Pali and how it translates to common english words. The same is true for the sense organs. These are english words understood widely outside of Buddhist context although ‘sense fields’ is not. Google it to confirm. So either relate these in brackets or not in brackets consistently if that is what the Teacher intended or leave it fully ambiguous I guess with just ‘aggregates’ and ‘fields.’

[sood]

@Vaddha

I think ‘aspects of existence’ is nice. I like that one can read it and get a good sense of the actual meaning, while it also preserves the ambiguity of the Pāli in that it doesn’t explain what the ‘aspects’ are in the translation.

Indeed, it seems to strike a nice balance.

@Vaddha

What I find less ideal is that ‘khandha’ is explicitly explained to refer to the fact that these are groupings of every instance of the five categories of phenomena. So the term ‘group,’ ‘heap,’ ‘grouping,’ or ‘category’ is actually significant in the understanding and explanation of the term in the suttas.

Yes, I think aggregates also carry this sense for me since I think of several things aggregating together into one category with aggregates, so I found aggregates of existence to have a nice ring to it, along with the balance as above. I find “groups of existence”, “heaps of existence”, “categories of existence” also to be equally good in this regard and interchangeable.

[yeshe.tenley]

@Vaddha

I think ‘aspects of existence’ is nice.

Is bhava or existence mentioned in the Pali here? Why would this be a good translation if not?

[Vaddha]

@yeshe.tenley

The mind and body

The thing is that ‘mind and body’ is missing part of the definition of the khandhā. “External form” doesn’t mean your body or an external body, it also means the external water element for example. Moreover, separating the aggregates into rūpa vs. the other four is a later Buddhist idea that does not mesh with how the early discourses envision things. When the suttas divide the aggregates, they separate viññāna from the other four. Just as they also have ‘nāmarūpa’ juxtaposed to ‘viññāna.

So to translate it in terms of ‘four mental + one bodily’ is an oversimplification at best, and a misrepresentation of how the discourses present a model of the mind at worst.

As for how ‘five’ is an equivalent, it’s because when the suttas say ‘khandhā’ there are several things this can refer to. One is the ‘tayo khandhā’ or ‘sīla/samādhi/paññā.’ But when it says “the manifestation of the khandhā,” it doesn’t mean the manifestation of sīla/samādhi/paññā; it means the manifestation of the five khandhā. The same is true with ‘āyatana,’ which could mean the formless āyatanas (ākāsānañcāytana, etc.). But here it specifically means the six āyatanas (in Pāli sa.lāyatana), which are the domains of sensory experience. The native Pāli way of distinguishing the various types of khandhā is through the word ‘pañca,’ or ‘five,’ and this is precisely what is implied in the passage I cited in my original post.

[yeshe.tenley]

@Vaddha

The thing is that ‘mind and body’ is missing part of the definition of the khandhā.

Right, but in this sutta I think that everyone agrees that it is the individual that is reborn and the parts of the individual (or aggregates) are the mind and body.

In other words, I think everyone agrees that the sutta is not implying that the entire phenomenal world is being reborn aka external form and external mind etc. On the other hand, if this isn’t universally agreed, then ‘mind and body’ shouldn’t be used and probably better to just use ‘aggregates’ and leave the ambiguity.

[Adutiya]

@Sasha_A

If these words are indeed to be taken as technical terms with special meanings, and not just words, then ‘āyatana’ and ‘khandha’ are the names of these terms, and as the names of the special terms they do not need to be translated at all

I’ve thought the same thing about several Pali words, particularly words with difficult to capture nuances such as dukkha. I’ve tended to do deep dives to get a sense of a word’s meaning in different contexts so that I have a mental grasp of what is intended without a (sometimes misleading) translation. However, in a discussion years ago Ven. Sujato pointed out that the task of a translator is to translate satisfactorily and not leave words untranslated. I can understand this wisdom, particularly in Buddhism. Investigation is an awakening factor!

[sujato]

@sood

Ven @Sunyo has aspects of existence in his translations, do you find that compelling, Bhante?

Not really, “of existence” makes me itch. And a khandha is not an “aspect”.

@Vaddha

“Sense” is also extraneous

I think the point is that it is not extraneous: it is what āyatana means in this context. In fact you could just drop “fields” and it works fine. The only reason for saying “fields” is to keep a connection with other uses of āyatana, but that is not really necessary. So a minimal translation would better be:

the manifestation of the aggregates, the acquisition of the senses.

The problem is that in English we have an ordinary word “senses” that covers pretty much the same meaning as āyatana, but we don’t have anything that means what khandha means in this case.

@Sasha_A

would be the translator inserting his own interpretation of the meaning of the term, rather than simply translating the original text as it is

When the translator does their job of rendering the meaning into English, sometimes you might find it challenging because it conflicts with your personal interpretation. To which I say, you’re welcome.

[Sunyo]

@sood

Ven @Sunyo has aspects of existence in his translations

I’m not translating the whole canon, though, which gives a bit more liberty, because a specific rendering doesn’t have to work in every passage. My translations, generally, have different aims, one of which is to provide another perspective on standard (ingrained?) terms. Of which ‘aggregates’ is indeed one.

‘Aggregates’ is meaningless by itself but, ideally, gets its meaning from the larger context. With that in mind, it isn’t particularly problematic. Still, even in modern English the term just reminds me of concrete.

[Jayarava]

The big problem with Pāli khandha and rūpakhandha is that they are not defined in Pāli or anywhere else. This is noted in the two big studies of the khandha doctrine published by (the late) Sue Hamilton and Tillman Vetter. Even the Theravādins have been left guessing what was meant. And this gives rise to anomalies and inconsistencies and to inconsistent translations. “Aggregate” has to be the least accurate and most misleading translation of any Buddhist technical term.

Sue Hamilton came right out and said it. The received khandha doctrine is incoherent. Many of us have looked for better ways to understand the words and the associated doctrine. I’m largely guided by Hamilton, though I now think she underplayed the role of the cessation of sensory experience.

In 2013 I wrote an essay exploring this and decided that skandha comes from a word that means “branch”. And I interpreted this in the light of Hamilton’s approach to reading Buddhist texts as concerned with sensory experience. I now routinely talk about the skandhas as “the five branches of experience”. I’ve got this past reviewers and editors and into print several times now.

In my most recent publications on the Heart Sutra I explain the five branches of experience like this.

• rūpa - “appearance”: the appearance of something in the sensorium.
• vedanā - “valence”: the +ve and -ve hedonic quality of the appearance.
• saṃjñā - “recognition”: identifying the experience(s) that the appearance engenders.
• saṃskāra - “volition”: an opportunity for karmically significant reaction (i.e. cetanā).
• vijñāna - “discrimination”: discriminating the external object that accounts for the sense experience; or the objectification of experience.

Objectification then leads to proliferation (prapañca).

Note that my definition of rūpa depends on rūpa as an āyatana. In that scheme, rūpa is to the eye as śabda is to the ear. Rūpa is what emerges from the object, crosses the intervening space, impacts on the eye, and causes cakṣuvijñāṇa. So rūpa is not substance (dravya) and not the object per se. And it is not “the body” because of the relationship of spraṣṭāvya to kāya in the same scheme. In modern terms, rūpa refers to reflected light (though this never occurred to the ancients). As rūpaskandha, I take rūpa to be a metonym for the appearance across sensory modes. In nāmarūpa I think rūpa just means “appearance”. I don’t think rūpa ever means “form” except in the sense of how something appears.

There is a good essay on āyatana by Gabriel Ellis: Āyatana, the Buddha’s forgotten teaching. And it’s mercifully brief.

Based on Vedic usage, Ellis describes āyatana in this context as a “centre of experience”; or “the place where experience happens”. “Domain” covers this quite nicely. Given the other main use as a “stage” of meditation, āyatana seems to have similar connotations to English “stage” as both the place/sphere/domain where things happen and a point along a journey.

He concludes: “I understand the six internal saḷāyatanas as ‘sight’, ‘hearing’, ‘smelling’, ‘tasting’, ‘body-cognition, and ‘mind-cognition’.”

[Green]

I think the best model is that there is an sich no meaning in smell,-ear, eye, tactile, taste vinnana.
While they arise they only present sounds, smells, visuals, tactile sensation, tastes to the mind.

For example, a small child does not hear meaningful words. The ear-vinnana only present sounds to her mind. There is allready the ability to notice the different qualities of different sounds but there is no meaning yet.

What provides this meaning? I do not think the sense-vinnana’s. I believe it is the mind. We learn that a certain sound is the letter A. We learn to combine hearing different sounds into a word. Different words into sentences. And gradually we start to live in a world in which sounds are immediately given meaning. But is it not the sense-vinnana’s who do this. There is an intelligence active, mind. Sense vinnana’s are not intelligent.

I also consider mental vinnana’s this way. While ear vinnana’s just present sound to the mind due to which they become aware, the mental vinnna’s only presents arising ideas, plans, emotions, thoughts to the mind in which they become noticed, become aware.

But what meaning is given to those arising mental vinnana’s? These perceptions of thoughts, emotions, plans etc. That is, again something of the mind.
One can see those mental vinnana’s as me, mine, value high, value low, but this is not the mental vinnana doing this.

Anusaya’s are also not in vinnana’s. They are in the mind.

If we purify the mind we also do not purify the 6 vinnana’s.

[Sasha_A]

@sujato

When the translator does their job of rendering the meaning into English, sometimes you might find it challenging because it conflicts with your personal interpretation. To which I say, you’re welcome.

I am sorry, Venerable, but the reader’s views have nothing to do with the problem of the quality of the translation: the problem of excluding an easily avoidable influence of the translator’s views on the resulting translation lies solely with the translator himself, with his qualification, honesty to himself and to his readers about his real intentions behind doing the work of translating.

What I find really challenging here is the obvious problem of distinguishing between discussions of purely technical translation problems and discussions of personal views on the meaning of the translation.

[Vaddha]

@Jayarava

decided that skandha comes from a word that means “branch”.

Hi @Jayarava . Interesting article. I do agree that there is a good case to be made that the five khandhas are describing the what of sense experience whereas the six domains (āyatanas) describe the where. The sense-domains are like setting up the parameters of the chess game and all of the pieces involved, whereas the aggregates are like the specific design and movement of the pieces, in one loose metaphor. More literally, the khandhas show us what the detailed constituents of sensory perception are, such as the vedanā, saññā, and sankhārā happening within the framework of each sensory domain. And there is overlap at rūpa and viññāna. There is explicit discussion of this at MN 28. I don’t agree with everything there or your analysis of every category, but that’s not exactly what this topic is about.

To comment specifically on ‘skandha’ as ‘branch’ and the refutation of ‘heap/group/category,’ I’m not sure why you didn’t cite the reasoning for this from the discourses. You cite the the Punnama Sutta, but you leave out the part that specifically discusses the word ‘khandha’ which does describe it in terms of a collection/category for a general set of phenomena. I linked it above; here it is again:

”Sir, what is the scope of the term ‘khandha’ as applied to the five khandhas?”
”Any kind of form at all—past, future, or present; internal or external; coarse or subtle; inferior or superior; far or near: this is called the khandha of form.”

You also didn’t cite other common uses of the word khandha in the discourses apart from Gombrich’s fire example. But here is an extremely common stock formula:

evametassa kevalassa dukkhakkhandhassa samudayo hoti.
That is how this whole mass/heap of suffering comes to be.

This cannot mean a ‘branch’ as in a specific sub-set or off-shoot piece. And this is the primary usage of the term ‘khandha’ in doctrinally significant Buddhist passages outside the pañcakkhandha, i.e. the ‘dukkhakkhandha’ — ‘heap of suffering.’

We have the ‘tayo khandhā’ as well, the khandha of sīla, of samādhi, and of paññā. These are discussed at least once in the suttas themselves (MN 44), and there too it seems that ‘khandha’ more likely is referring to a category or grouping of various practices. Once we go from the idea of a ‘heap/mass/grouping,’ we get more analytical ‘categories’ and then ‘chapters’ or ‘sections’ as groupings of intellectual material or ‘divisions’ as groups of soldiers, etc. So this is by no means less reasonable a usage of the meaning ‘heap’ than ‘branch’ IMO.

There’s also the use of the word ‘dārukkhandha, which is a big log of wood, not a stick/branch. The idea seems to be that skandha refers to the trunk/torso/shoulder, not just branches or twigs. Of course over time we see the word used in various evolving ways and the etymology seems uncertain, as you write in your blog. But from the Pāli alone it reads more that ‘khandha’ refers to a large grouping or category, as well as a heap/mass, of things.

I would say that this image is also implicit in how the suttas describe the khandhas as a “burden” (bhāra). This is given as an alternative term for discussing them, i.e. the burden, picking up the burden, and laying it down. To call five “heaps” or “masses” a burden has a continuity of imagery, but not ‘branches [of experience].’ This is of course not definitive evidence of anything, just an aside.

Either way, I would say it’s clear that the term ‘branch’ doesn’t encompass ‘any instance of this kind of phenomena at all,’ which is the canonical explanation of the use of khandha; only group/heap/aggregate/etc. does. ‘Branches of experience’ is similar to Ven. @sunyo’s ‘aspects of existence.’ I think it’s an exegetical device which nicely captures one element of how this is a framework for the categories of things in our experience/existence; but I don’t think it’s a particularly accurate translation of the Pāli words and how they integrate into the discussions of them in the canon itself.

All the best!

[Vaddha]

@Sunyo

My translations, generally, have different aims, one of which is to provide another perspective on standard (ingrained?) terms.

And I think you do a nice job of this I appreciate the sheer readability and colloquial style you’ve presented as a supplement.

@Sunyo

Still, even in modern English the term just reminds me of concrete .

Yeah! And as I pointed to above, I think the idea is that these are ‘heaps,’ i.e. burdensome groups that we lug around. ‘Aggregates’ is very dry and impersonal on top of being pure jargon. But it does seem to have been hard-wired into the English speaking Buddhist circles.