11 August 2017

Fixing the Broken Heart Sutra.

In this essay I do several things. I show that Conze bungled the editing of the Section VI. of the Heart Sutra (Section VII in my nomenclature) and how to fix the received text. I review research by Huifeng which shows that the original Sanskrit translator also bungled, and go into detail on how to fix his mistakes. I incidentally show that Red Pine is also a bungler. By doing the basic philology that should have been done a century ago, I explain the grammar of this passage of the Heart Sutra and how to better construct it in Sanskrit. I give a revised Sanskrit text that is far more consistent with existing Chinese texts. I once again ponder the dire state of Buddhist philology. 

One of the many linguistic puzzles of the Heart Sutra is the construction nāstitvād that one finds in Conze's Section VI. I have been worrying away at this problem for five years and just now had a breakthrough, but it requires some context. The text in Conze's revised, 1967 edition and 1975 translation of the Heart Sutra tells us 
Tasmāc Chāriputra aprāptitvād bodhisattvo  prajñāpāramitām āśritya viharaty acittāvaraṇaḥ. Cittāvaraṇa-nāstitvād atrasto viparyāsa-atikrānto nishṭhā-nirvāṇa-prāptaḥ. 
Therefore, O Śāriputra, it is because of his non-attainmentness that a bodhisattva, through having relied on the perfection of wisdom, dwells without thought-coverings. In the absence of thought-coverings he has not been made to tremble, he has overcome what can be upset, and in the end attains to Nirvāṇa (Buddhist Wisdom Books).
Almost everything about this translation is wrong. It is one of the most egregious examples of Buddhist Hybrid English in the entire canon of Buddhist translations. The English is execrable and incomprehensible. "Non-attainmentness" is a rather ugly neologism. And what on earth is a "thought-covering" (and why is it hyphenated)? Nor have other translators ever really got to grip with this passage. Those who purport to translate it from Sanskrit have failed to see a very basic mistake in Conze's Sanskrit (incidentally repeated by Vaidya in his edition). Red Pine simply dismisses the rules of Sanskrit grammar at this point with hand-waving and produces a translation that fits his preconceptions but has nothing to do with the underlying text (See p.137 of his book). 

As with previous essays of this kind I will make frequent use of the Chinese. In a previous essay, I worked through Huifeng's treatment of the Sanskrit acittāvaraṇaḥ in light of the Chinese text. I accepted that he was right to reinterpret the passage, but disagreed on the definition of the key verb. He wanted 罣 to mean "hang" and I insisted on the more basic meaning of "stick".

The Chinese passage in T2521, which most East Asians take to be the Heart Sutra, reads (though note that the punctuation is modern):
以無所得故,菩提薩埵依 般若波羅蜜多故,心無罣礙;無罣礙故,無有恐怖,遠離顛倒夢想 ,究竟涅槃。
And this means more or less the same. I'll work through the differences as we go.  The first difference is to notice that the Chinese text has no "therefore Śāriputra" here.

Next, Huifeng has shown that aprāptitvād is an incorrect translation of 以無所得故.  In fact, Kumārajīva most often uses these characters to represent anupalambhayogena. And Huifeng makes a very good case for taking this word with the previous section and as ending a sentence. So if we make these corrections we get a working text that looks like this

Tasmācchāriputra śūnyatāyāṃ na rūpaṃ… na prāptir nābhsamayo anupalambhayogena ॥  
Bodhisatvo prajñāpāramitām āśritya viharaty acittāvaraṇaḥ. Cittāvaraṇa-nāstitvād atrasto viparyāsa-atikrānto nishṭhā-nirvāṇa-prāptaḥ. 

Now, bodhisatvo does not just start a new sentence, it starts a next paragraph or section. Focusing now on this new paragraph, since we have eliminated the possibility that 以無所得故 means acittavaraṇaḥ, it is extremely unlikely that 無罣礙故 can mean acittavaraṇaḥ nāstitvād (and anyway this is a very weird construction). Moreover, there is no verb like viharati "dwell" anywhere in Chinese.

Unfortunately, although he gave an English translation which conveyed the correct reading of the Chinese, Huifeng didn't give a Sanskrit reading (and declined to do so when I pestered him in person). With the help of some colleagues I came up with an alternative Sanskrit reading based on the hints in Huifeng's article:
yo bodhisatvaḥ  prajñāpāramitām āśritya asya cittam na kvacit sajjati ।
The mind of the bodhisatva who relies on perfection of wisdom, does not get stuck anywhere.
The phrase "does not get stuck" could also be "does not get attached"  But how to deal with Cittāvaraṇa-nāstitvād atrasto viparyāsa-atikrānto nishṭhā-nirvāṇa-prāptaḥ? It is at this point that Red Pine simply abandons grammar and simply makes up a translation to suit his purposes. But the fact is that there is nothing very difficult here, except that Conze has once again led us astray.

In his 1948 edition, the last word in the sentence is nishṭhā-nirvāṇaḥ. Let us start here. Nirvāṇa is an adjective and thus properly takes the gender of the noun it describes. Buddhists often use it as a noun, and when they do so it is invariably neuter. The eagle eyed reader will note that here nirvāṇaḥ has a masculine nominative singular ending. This can mean only one thing (and there is nothing "vague" here, contra Mr Pine)  which is that nishṭhā-nirvāṇa is a bahuvrīhi or adjectival compound. And it is describing a noun in the masculine nominative singular case. In Conze's Sanskrit, there is no such noun in this sentence. However, there is one in the previous sentence, i.e. bodhisatvaḥ (and note that -aḥ followed by p is unchanged, so bodhisatvo is wrong in Conze).

This tells us that Conze has blundered (again). Here we have just one sentence. It also tells us that we don't need prāpta. Prāpta (the passive past participle of prāpṇoti "to attain") was added to provide a verbal derivative because, just as in English, a Sanskrit sentence has to have a verb or a verbal derivative acting as a verb. By dividing one sentence, with one main verb, into two sentences, Conze has created an ungrammatical entity. Ancient scribes added prāptaḥ or in one case the verb prāpṇoti, but it was never needed. It also makes Red Pines decision to take atikranto as the verb look silly (not to mention that this leaves viparyāsa undeclined which is also not allowed).

The rules of grammar in any language are not simple. But the person who composed the Sanskrit followed those rules and it is the subsequent editors, translators, and commentators who have been at fault. Conze was an expert in Sanskrit so I cannot imagine why he went off piste in this case. Other experts, some of whom I hold in the highest esteem, have also failed to notice which rules were being applied.

There are in fact three bahuvrīhi compounds in a row, all of which describe the bodhisatva:
  1. atrastaḥ “one who is without fear”  
  2. viparyāsa-atikrāntaḥ “one who has overcome delusions”
  3. niṣṭhānirvāṇaḥ “one who has extinction as his end”
 So our working text now reads
yo bodhisatvaḥ  prajñāpāramitām āśritya asya cittam na kvacit sajjati  cittāvaraṇa-nāstitvād so atrasto viparyāsa-atikrānto nishṭhā-nirvāṇaḥ
A quick comparison with the punctuation of the Chinese text T251 tells us that the editors of Taishō were on the same wavelength (though inconsistently, T250 they side with Conze and break the passage into two sentences!). But we can refine this further. I said above "it is extremely unlikely that 無罣礙故 can mean acittavaraṇaḥ nāstitvād". Based on Huifeng's research the Chinese characters most like mean asaṅgatvād "because of being without attachment". And this gives us
yo bodhisatvaḥ prajñāpāramitām āśritya asya cittam na kvacit sajjati asaṅgatvād so atrasto viparyāsātikrānto nishṭhānirvāṇaḥ |
However, when we look again at the Chinese we notice some further issues with these adjectival compounds. Firstly there are four of them in Chinese, and secondly the two key texts T250 and T2521 are different.
T250: 無有恐怖,離一切顛倒夢想苦惱,究竟涅槃。
T251: 無有恐怖,遠離顛倒夢想,究竟涅槃。
Well take this a step at a time, and it is well worth taking our time because the first part of this provides us with a nice little insight. From the Sanskrit, we are expecting an adjective which means "he is not afraid". What we get is "non- 無 existent 有 terror 恐怖". Typically we read this as "being without fear" to match the Sanskrit. But let's look again at the passage with the previous word added and without the modern punctuation.
This is what the Sanskrit translator saw in the 7th Century. He had to figure out where the breaks come. We can see where the modern editors have put the breaks, but what if the translator became a little confused at this point and bracketed out the wrong characters and thought he saw something like this:
無 [罣礙故無有] 恐怖 遠離 顛倒 夢想
Now, this is an unlikely reading, but I'll tell you why I am highlighting it.  If we take the bracketed characters 罣礙故無有 the phrase 無 [罣礙故無有] 恐怖 says something like "not—because of the non-existence of mental obstructions—afraid". So is 無有 here the source of nāstitvād? There is no other explanation I can think of and no other explanation has ever been offered, to the best of my knowledge. Although the Chinese text is mangled in the process, we know it is not the first time the translator has mangled the text in the process of translating it into Sanskrit. But we also know that the translator does understand that 無 corresponds to the Sanskrit prefix a- or to the negative particle na. Which is to say that he knows better than to use nāstitvād when a- or na would do, and thus nāstitvād is in need of some explanation. If there is another explanation, I'd love to hear it.

Then we have some pairs of characters:
  • 恐怖 "afraid" 
  • 遠離 "goes beyond" = Sanskrit atikranto
  • 顛倒 "delusion = Skt viparyāsa
  • 夢想 "dream thoughts" i.e. illusions = Skt māyā.

In Chines, the order of 遠離 "goes beyond"  and 顛倒 "delusion suggests an active verb. And on this basis it would be possible to quibble with the construction of the Sanskrit. But I propose to leave the basic structure intact here. The translator has also mushed 顛倒 and 夢想 together into the familiar Sanskrit word viparyāsa, when he might have included the concept of māyā. Adding māyā here would be interesting in light of my observation about the relation of śūnyatā and māyā (soon to be published, but preliminary notes blogged) Now we want viparyāsātikrānta to be a descriptive compound and we would read this as "the bodhisatva has overcome delusions." I think we could follow the Chinese more precisely by having viparyāsmāyātikrānta. 

Note that T250 and T251 differ slightly here (added spaces for comparative purposes)
T251: 遠離        顛倒夢想,
T250:     離一切顛倒夢想苦惱.
The extra bits are 一切 meaning "all" (literally, "a single cut") and 苦惱 meaning "misery and trouble" where 苦 is the character most often used for Sanskrit duḥkha. In T250, 離 means "depart, go away"; while in T251, 遠離 has the same meaning, but with greater emphasis. A similar distinction is found in Sanskrit between atikranta and saṃatikranta for example.

The final piece of the puzzle, then, is how to translate 究竟涅槃. As I have explained in another previous essay, niṣṭhānirvāṇa was a poor choice. Kumārajīva uses these characters for a nirvāṇaparyavasānam. In our text it is being used an adjective of bodhisatvaḥ and must be given in the masculine nominative singular: nirvāṇaparyavasānaḥ. This gives us a final text:
yo bodhisatvaḥ prajñāpāramitām āśritya asya cittam na kvacit sajjati asaṅgatvād so atrasto viparyāsamāyātikrānto nirvāṇaparyavasānaḥ |
Of course there are always difference ways to translate passages from one language to another. One only has to look at English translations of the Heart Sutra which continue to multiply and diverge from each other. With the Heart Sutra each exegete appears to be trying to put their own unique stamp on the text rather than all of us working towards a common understanding and a single standard translation. In other words, in the usual Buddhist critique, translating the Heart Sutra is more often an occasion for egoism than for transcending self. Rather ironic really.  

At the very least this essay has shown, again, how poorly the Buddhist community has been served at times by philologists and traditional exegesis. Here is a short text, barely 250 words or characters, that is supposed to be chanted daily by millions of Buddhists, and we still don't have an accurate text to chant.

I have spent a considerable part of the last five years forensically examining this text in Sanskrit and Chinese, with help from key allies, and have been posting my many notes here as I go in the form of more than 30 essays. My third peer-reviewed publication on the text will be out in November 2017. Not everyone is going to be able to follow the argument in this essay, but I hope some of you will at least try. Experience shows that traditionalists will resist any call to change the Heart Sutra and will probably deny that anything is wrong with the text. But people need to know that the familiar Heart Sutra is deeply flawed and in need of surgery. And that the so-called authorities on the text have never noticed this stark fact.

I confess that I am thoroughly vexed by all this bungling. I am far from the best person to be trying to sort this mess out, but the best people have come and gone and the mess remains. Some of them have made huge contributions to the field and clearly had bigger fish to fry, but the Heart Sutra is not exactly inconsequential in Buddhism. That is has been left in this state by philologists does my head in. If I can understand it, why did they not? I'm not special. Something as simple as noticing that a neuter noun is declined as masculine and therefore must be an adjective is just basic Sanskrit. You learn this in the first month of the first year of studying the language. How does anyone miss this, let alone everyone? I also missed it for years. I have studied this passage before and thought about how to translate it. I suppose I'm now something of an expert on the text (if not an authority) but if you look around the internet there are dozens of people with strong opinions on how to understand the text. It's just that none of them seems to actually read the text in Sanskrit and think about how to parse the sentences. 

In particular, in my circles, Conze and Red Pine are treated as reliable guides, but are in fact often, or even mostly, unreliable. Neither of them deserves their reputation for scholarship. Both are bunglers and who have set back understanding of this text amongst Buddhists by decades. Kazuaki Tanahashi is not much better. Though his book is an improvement on Pine's, it is full of careless errors or just plain ignorance of Sanskrit. The dozens of commentaries are just rote recitations of sectarian gibberish. Mu Soeng frankly seems like a lunatic and his opinions on Sanskrit are laugh-out-loud wrong. The Dalai Lama is just going through the motions, reciting some ancient commentary by rote. D T Suzuki is busy doing his Mr Spock impression. It's all so depressing. How does the most popular text in Buddhism come to be treated so shabbily? It is gold. But not for the reasons that any famous Buddhists think it is. 

People often ask me what book they should read on the Heart Sutra. I get blank looks when I say that, in my opinion, it is better not to read any of the books currently available. They are full of inaccurate information. I mean it. The more you read about the Heart Sutra, the more wrong information you are assimilating that will only have to be unlearned in the end. I hope to remedy that situation soon by publishing a small book which gives an overview of recent research on the text (including mine, of course, but a few other people like Huifeng and Jan Nattier). In the meantime beware of fake Heart Sutra news, I guess. You're better off setting the sūtra  to one side and talking to someone who has some real depth of experience in meditation. Come back to it when you have experienced cessation for yourself. That would help head-off the most egregious wrong views. 


Conze, Edward (1948) Text, Sources, and Bibliography of the Prajñāpāramitā-hṛdaya. Journal of the Royal Asiatic Society, April 80(1-2): 33-51.

Conze, Edward. (1967) The Prajñāpāramitā-Hṛdaya Sūtra in Thirty Years of Buddhist Studies: Selected Essays, Bruno Cassirer, pp. 147-167. Modified version of Conze (1948).

Conze, Edward. (1975) Buddhist Wisdom Books: The Diamond Sutra and the Heart Sutra. George Allen & Unwin. First Ed. 1957.

Huifeng. (2014). ‘Apocryphal Treatment for Conze’s Heart Problems: “Non-attainment”, “Apprehension”, and “Mental Hanging” in the Prajñāpāramitā.’ Journal of the Oxford Centre for Buddhist Studies. 6: 72-105. http://www.ocbs.org/ojs/index.php/jocbs/article/view/75
Related Posts with Thumbnails