14 April 2017

Further Problems with the Heart Sutra: acittāvaraṇa

In this and the previous essay I am summarising and discussing Huifeng's article on terminology in the Heart Sutra. To reiterate, in this essay, we are assuming that the Sanskrit Heart Sutra is a translation from a Chinese text (though not one of the surviving versions). So we're investigating how a translator might understand the Chinese text, and were subjecting the received Sanskrit text to a critical examination in the light of comparative philology, especially within the Prajñā-pāramitā literature, facilitated by electronic texts and word searches. The fundamental (radical) question we are asking is, "Is the received Sanskrit text an accurate translation of the Chinese Heart Sutra?" But this also hints at an unasked question, "Which Chinese Heart Sutra is the Sanskrit a translation of?" The latter question will have to wait for another day

We now move onto the phrases 心無罣礙 and 無罣礙故 which were translated into Sanskrit (in Conze's edition) as acittāvaraṇa and cittavaraṇa-nāstitvād. Conze is again in poor form translating the latter as "without thought coverings" and "in the absence of thought coverings".

This section of the text was apparently composed to accompany the quotation which finished as the end of the last section. So we do not expect to find exact equivalents in the Sanskrit Prajñāpāramitā literature, but we might wish that the Sanskrit translation of the Chinese was at least consistent with that literature. Although one might equally argue that the requirement is still to be true to early medieval China. 

The easiest part of this passage is 心. We expect citta in Sanskrit and we find citta in the Sanskrit Heart Sutra. So far so good. 

The translation of 心無罣礙 as acittāvaraṇa was a phrase that Conze had a great deal of trouble with. In his edition he notes some manuscripts have acittālambana and he describes how one might have become the other through a series of scribal errors of the type that are so common in the Nepalese manuscripts (e.g., mistaking v and b). However, Huifeng finds that neither of these possibilities—āvaraṇa or ālambhana—are plausible readings of 罣礙 (2014: 98-9). For example, he cites what seems to be an important passage from Kumārajīva's Large text:
Then Śakra, Lord of the Gods, said to Subhūti: Whatever Subhūti has stated is only for the sake of emptiness, without being hung-obstructed [sic] (無罣礙). Just as an arrow shot up into empty space is not obstructed (無礙), so too is Subhūti's Dharma teaching not obstructed (無礙). (92) 
Now, the awkward phase "hung-obstructed" is not very convincing and I find that I cannot go along with Huifeng in either his interpretation of 罣 as "hang" in this context. The basic meanings of the two verbs here are according to Kroll (2015):
  • 罣: catch fish; enmesh, ensnare, entangle.
  • 礙: impede, hamper, hinder; obstruct, block off.
If we are describing the way an arrow shot into space travels as "無礙" then it seems straight forward to say it is "not impeded". If we are looking to compliment this with 罣 then "hung" is not an obvious choice, whereas some kind of entanglement is. The arrow is not impeded or entangled in anything - therefore it continues on its way. 

As Huifeng points out this is only reinforced by the Sanskrit equivalent of the phrase 無罣礙 in this simile which is: na kvacit sajjati. The form sajjati is unusual. The Sanskrit verb is √sañj, which means "adhere to, be attached to, cling to". Other Sanskritists I consulted all took the verb sajjati to mean "cling", on face value. And we expect a passive form, sajyate. In Pāli this is sajjati. So this looks like an example of a Prakrit form being used in Sanskrit. That it occurs in a Buddhist text makes this quite likely, as Buddhists mixed up forms from Prakrit and Sanskrit very often.

One of the things about this verb is that it governs the locative - the place of clinging or attachment is given in the locative case. Hence kvacit "anywhere", a locative adverbial pronoun.

So the Sanskrit phrase na kvacit sajjati means "it doesn't stick anywhere" (and this time Conze accurately translates "...does not get stuck anywhere"). Elsewhere, saṅga means "attached" and asaṅga means "unattached" or "without attachment" (hence the name of the Yogācāra co-founder). Huifeng's argument for "hung, hanging", is not wrong, but it overlooks the obvious meaning. It is true that √sañj can mean "hang", but when describing the arrow or the mind of the meditating bodhisatva., the basic sense works fine. Also an arrow shot into space doesn't stick anywhere, as opposed to an arrow shot at a target which does. We don't say that an arrow that strikes a target is "hung" on the target. It's sticks into, or is embedded in it.

Some other passages show that Kumārajīva also used the character 礙 for Sanskrit prati√han, "to strike against." This adds a certain semantic richness, but the basic sense of √sañj, "stick", conveys the meaning of the text. There's also an unnoticed play on words here, because the secondary verb in the Sanskrit, translating 依, is āśritya and this has a connotation of "to adhere to, to attach one's self to". So the sentence is saying that because he is attached to the perfection of wisdom, the bodhisatva doesn't stick elsewhere. In Chinese 依 means "to rely on, fall back on; be guided by". 

Huifeng's grammatical commentary also informs us that 故, "alone after a verbal form is usually grammatically equivalent to a Sanskrit ablative form" (81) suggesting that 無罣礙故 corresponds to an ablative form. The received Sanskrit cittāvaraṇa-nāstitvād i.e. (citta-āvaraṇa-na-asti-tvād), is such a form, though the idea of negating a verb by compounding it with nāstitvā 'non-existence-ness' is idiosyncratic at best, and I can find no examples of this form in PPS. Where PPS does use the abstract form of nāsti (as a stand alone term), it opts for nāstitā, not nāstitvā (e.g., Kimura 1-1: 154, 1-2: 17). The form ‑nāstitvād doesn't fit the Prajñāpāramitā idiom. I've not been able to turn up this idiom anywhere, except in the Sanskrit Heart Sutra. We can see what it means, but not why anyone would chose to translate 無 this way.

Translating from Chinese into Sanskrit

So in 心無罣礙 無罣礙故: 心 is citta; 無罣礙 is most likely na [kvacit] sajjati; and 無罣礙故  is asaṅgatvāt. We just need to assemble the elements into a Sanskrit sentence. We also have to keep in mind that the preceding passage describes the bodhisatva as being "engaged in non-perception" (anupalambhayogena) and we need to translate this phrase to fit that setting. We're looking for a phrase that says that the bodhisatva's mind, while engaged in non-perception, does not stick anywhere.

Unfortunately, Huifeng does not commit to a Sanskrit translation at this point, though he does give an English translation that reinterprets the Chinese text.
The bodhisattvas, due to being supported by transcendental knowledge, have minds which do not hang on anything; due to their minds not hanging on anything, they are without fear... (103)
Given my discussion of this above, my rendering of this same passage (trying to avoid Buddhist Hybrid English) would be
Being supported by the perfection of wisdom, the mind of the bodhisattva does not get stuck anywhere; and because it does not stick, they are fearless ...
We can see now why some manuscripts have bodhisatva in the genitive case; they want to indicate that the mind in question is possessed by the bodhisatva. What I take the passage to mean is that through insight style meditations such as the contemplation of the skandhas, and dwelling in emptiness, the bodhisatvas disentangle themselves from the snares of sensory experience and become free (vimokṣa).

Now that we are clear what the text says and what it means, and we know roughly what kind of words the authors of the Prajñāpāramitā literature might have used, we can try to translate 菩提薩埵 依 般若波羅蜜多故 心無罣礙 無罣礙故 into Sanskrit.

So we want a compound that says "unattached mind" or "mind which is unattached" or "mind which is not stuck". We can see why a translator might have adopted a-citta-āvaraṇa, since in Buddhist Sanskrit and in Pāḷi āvaraṇa means "hindrance" or "obstruction" (BHSD). The translator has understood impediment and opted for a common word, but without reference to the Prajñāpāramitā idiom. Still it can mean "without mental hindrances". Note also that the Sanskrit texts interpolates viharati as a main verb and treats 依 as a gerund āśritya. The former is odd, but the latter fits the context.

One way to say in Sanskrit what we understand the text to say and which doesn't diverge too far from the Chinese source text would be:
[Yo] bodhisatvaḥ prajñāpāramitām āśritya [tasya] cittam na kvacit sajjati. Asaṅgatvāt so 'trasto...* 
That bodhisatva, relying on perfect wisdom, his mind does not stick anywhere. Being unattached, he is unafraid...
Words in square brackets have no Chinese word equivalent but are added to create grammatical Sanskrit.This is necessary because the grammar of Sanskrit is more elaborate than that of Chinese.  Kvacit is implied by Kumārajīva's idiom and a locative is required by the verb. In plain English
Because the bodhisatva adheres to the perfection of wisdom, his mind doesn't get stuck elsewhere. Not being stuck, he is fearless... 
* Thanks to Dhīvan and Dayāmati for help with composing this Sanskrit. It will no doubt be subject to revision.


The conclusions of both part one and two of this essay are similar. Whoever translated the Chinese Heart Sutra into Sanskrit could have done a better job of it. They mistook 以無所得故 for aprāptirvād and mistook 心無罣礙 無罣礙故 for acittavaraṇaḥ | cittāvaraṇa-nāstitvād  when, in fact, these are not very likely translations, in the light of the Sanskrit Prajñāpāramitā traditions. The idiom nāstitvād is particularly nasty. In fact, although we stipulated the Chinese origins hypothesis at the outset, the awkwardness and idiosyncrasy of these two phrases alone make the hypothesis seem plausible.

So ,over and above what we can glean from the manuscript tradition of the Heart Sutra, it seems that the Sanskrit ur-text was botched. The question for scholars and Buddhists is then: Should we correct these errors?

Given that the errors are in the order of 1300 years old and universally attested in the Sanskrit manuscript and epigraphical witnesses, given how familiar they are to Buddhists and scholars alike, perhaps we ought to leave them alone? We can treat the Heart Sutra as a quaint oddity, a somewhat perplexing, but unarguably positive one. I imagine that many practitioners would like to continue to chant the familiar text.

On the other hand, what Huifeng has identified in his article are errors. Since I have identified other errors, I think there is a growing case for retranslating the Heart Sutra from Chinese into Sanskrit that is more consistent with the Prajñapāramitā tradition and the improved understanding of the text that philology affords. Of course, the Indian tradition was originally one that spoke Gāndhārī and reconstructing that vocabulary would be difficult, to say the least. But we can, and I argue that we should, construct a better Sanskrit Heart Sutra. One that makes sense, that eliminates the bizarre contradictions, and is in keeping with the other Sanskrit Prajñāpāramitā texts, at least at the time the sūtra was composed, ca. 7th Century.

To this end we have three Gilgit manuscripts of the PPS, one of which has been published in facsimile and partially transcribed. We know from recent publications that full transcriptions of all of the manuscripts are well underway and that we can expect a critical edition at some point. There is a huge amount of comparative information in Karashima, et al. (2016).



Karashima, Seishi, et al. (2016) Mahāyāna Texts: Prajñāpāramitā Texts (1). Gilgit Manuscripts in the National Archives of India Facsimile Edition Volume II.1. The National Archives of India and The International Research Institute for Advanced Buddhology, Soka University, Tokyo.

Kimura, Takayasu (2006). Pañcaviṃśatisāhasrikā Prajñāpāramitā. 6 vols. Tokyo: Sankibo Busshorin.

Kroll, P. W. (2015). A Student's Dictionary of Classical and Medieval Chinese. Brill.


Note, 18 April 2017

evam ukte āyuṣmān subhūtir bhagavantam etad avocat: kiṃlakṣaṇā (PSP_5:12) bhagavan prajñāpāramitā? bhagavān āha: asaṅgalakṣaṇā subhūte prajñāpāramitā na subhūte prajñāpāramitā lakṣaṇan na prajñāpāramitāyāḥ kiñcil lakṣaṇam.

Related Posts with Thumbnails