20 April 2017

Heart Sutra Anomaly


It was apparent, even to the late 7th Century commentators Woncheuk and Kuījī, that the Heart Sutra contained quotations from the Pañcaviṃśatisāhasrikā-prajñāpāramitā-sūtra (PPS) (Nattier 1992: 206-7, no. 33). In this essay, I will compare and contrast the various source texts for part of one of the quotations. We will see that a substantial change was introduced in translation of the PPS produced by Kumārajīva's translation team in 404 CE, though we don't know if this was evident in their source text or was an innovation at that time. Jan Nattier (1992: 205, n.26) already noted this in her watershed article on the Chinese origins of the Heart Sutra, but it has received scant attention since. And it raises interesting issues regarding authenticity and the role of modern philologers.

Nattier compared various versions of the quoted passage from PPS with the versions found in the various Heart Sutras. The Chinese Heart Sutra text is nearly identical to the Chinese PPS created by Kumārajīva's translation team ca. 404 CE (i.e., Taishō Sūtra No. 223). The Sanskrit Heart Sutra is, however, different in many ways from the extant Sanskrit PPS manuscripts (one cache from Gilgit, ca 6th Century and another from Nepal, ca. 19th Century). The versions differ in syntax at some points and differ in lexicon at others, but they mostly do not differ in semantics. Where sentence structures and word choices are different, the Heart Sutra still conveys the same message -- except in one case, which I will call "Section 3" in this essay (I've broken the quoted passage into a sequence of sections for my own purposes).

The obvious conclusion is that the Sanskrit Heart Sutra is a paraphrase of the Sanskrit PPS, while the Chinese Heart Sutra is a direct quote from T223. The paraphrasing occurred because the extract went from Buddhist Sanskrit (composed ca. 1st Century CE) through the filter of Middle Chinese (sometime between 404 and 664 CE) and back to something like Classical Sanskrit (before the death of Xuanzang in 664). The meaning was preserved, but many particulars of how that was communicated were changed.

We can see how this might work using Google translator to go between English to Mandarin and back:
  1. Original: Form is only emptiness. Form is not different from emptiness.
  2. Eng→Man: 形式只是空虛。 形式與空虛沒有區別。
  3. Man→Eng: The form is just empty. There is no difference between form and emptiness.
The words mean the same, but they are paraphrased. The effect on a highly inflected language like Sanskrit could be rather more dramatic, because a good deal of grammatical information is lost in converting to Middle Chinese. And if the translator from Chinese into Sanskrit was not familiar with the Prajñāpāramitā idiom in Sanskrit, then this would also amplify the effect.

We have five versions of the quoted passage from different times and places:
  • 《放光般若經》 by Mokṣala (291 CE). T221 8.6a06-6a13.
  • 《摩訶般若波羅蜜經》 by Kumārajīva (404 CE). T223 8.223a13-a24
  • Pañcaviṃśatisāhasrikā-prajñāpāramitā-sūtra. Gilgit Ms. ca 6th C. Facsimile by Karashima et al (2016). 21v-22r.
  • 《大般若波羅蜜多經》 by Xuánzàng. (659-663 CE). T220-ii; Fasc. 401-478 7.13a12ff.
  • Pañcaviṃśatisāhasrikā-prajñāpāramitā-sūtra Nepal Ms. ca 19th C. Edition by Kimura (2012) 1-1: 64
In this essay I will examine two extracts: section 1 and section 3. Section 1 is useful to review since it is the centre piece of Jan Nattier's argument that the Heart Sutra was composed in China and in the Chinese Language. It demonstrates her notion of "back-translation" but, given the alternative reading, also demonstrates the complexities involved.

Section 1
Mok.色與空等無異 所以者何?色則是空 空則是色
Kj.舍利弗 色不異空 空不異色* 色即是空 空即是色
Xz.舍利子 色不異空 空不異色 色即是空 空即是色
Gil.na hi śāradvatīputrānyad rūpam anyā śunyatā nānyā śunyatānyad rūpam rūpam eva śunyatā śunyataiva rūpam evaṃ nānyā vedanānyā śunyatā |
Nep.tathā hi śāriputra nānyad rūpam anyā śūnyatā nānyā śūnyatā anyad rūpaṃ rūpam eva śūnyatā śūnyataiva rūpam
*The notes in the Taishō Tripiṭaka say that some editions have 非色異空 非空異色 here. This is important because the two main versions of the Heart Sutra have different versions of this. T251 has the form above, while T250 has this alternate form.
The syntax here is one of the key differences that Nattier noted in 1992 as evidence for her Chinese origins hypothesis. PPS has na anya X anya Y, whereas Heart Sutra has X na pṛthak Y. Note that anya "other" is a pronoun which takes the gender, case, and number of the noun it qualifies. These both mean the same thing; i.e., that X and Y are not different or, in plain English, "X is the same as Y".

There are also two different syntaxes in Chinese: X 不異 Y and 非 X 異 Y (which, in some ways, mirrors the Sanskrit paraphrasing). The Chinese texts all follow very similar conventions:
ChineseEnglishSaṃsṛktam notes
form rūpa
emptiness śūnyatā
not naGeneric pre-verbal negative
without, un-, -lessa-
is not na (asti)Generic negative for sentences, especially identities
different anya/pṛthak
則/即only eva
Comparing the versions, we can see here that Xuanzang largely followed Kumārajīva. All that he changed was the spelling of the name Śāriputra. Kumārajīva uses a phonetic transcription which in Middle Chinese would have been something like sharibut. The final /t/ was pronounced strongly in Old Chinese and may have sounded very like the Central Asian pronunciation of the name in Prakrit (cf. Gāndhārī: Śariputra; Pāli: Sāriputta). Note that final -a is also dropped in modern Hindi with similar effect (compare also Nattier 1992: 216, n.91). Xuanzang kept the shari, but replaced the last syllable with the Chinese character for son, 子, to translate Sanskrit putra.

An oddity is that Gil. has Śāradvatīputra for Śāriputra. This is actually common, but not reflected in the Chinese translations, which all have the latter.

Mokṣala phrases his first sentence differently: 色與空等無異 means "form and emptiness, etc 等, are not different"; while 所以者何?is "And why?". What follows is the same except that Sanskrit eva is conveyed with 則 where Kumārajīva and Xuanzang both use 即.

Section 1 was the entrée, now we move onto the main course.

Section 3
Mok.亦不見生 亦不見滅 亦不見著 亦不見斷 亦不見增 亦不見減 亦不過去當來今現在
Kum.舍利弗 是諸法空相 不生不滅 不垢不淨 不增不減 是空法非過去 非未來 非現在
Xz舍利子 是諸法空相 不生不滅 不染不淨 不增不減 非過去非未來非現在
Gil.yā śāradvatīputra śunyatā na sā utpadyate na nirudhyate | na saṃkliśyate na vyavadāyate | na hīyate na vardhate | nātītā nānāgatā na pratyutpannā
Nep.śūnyatā śāriputra notpadyate na nirudhyate na saṃkliśyate na vyavadāyate na hīyate na vardhate nātītā nānāgatā na pratyutpannā yā ca īdṛśī
These are all saying something similar. Here, something does not arise (不見生, 不生, na sā utpadyate, notpadyate) it does not cease; it's not defiled and not pure; it is not deficient* and does not grow; it is not part, future, or present. The big difference involves what that something is.
* hīyate (passive form of √) is the verb from which the adjective hīna derives. It means "deficient, wanting; excluded; abandoned; etc."
In both the Gilgit and Nepalese PPS texts, it is śūnyatā that is the subject of these sentences; i.e., it is śūnyatā, itself, which does not arise or cease, etc. However, in the Chinese text of Kumārajīva a whole new phrase is inserted which says: "all dharmas are marked with emptiness" (是諸法空相). Xuanzang also has this phrase; he may simply have followed Kumārajīva. As Nattier points out, "In this context, without an explicit subject in the Chinese text, the reader would most naturally conclude that the subject is 'all dharmas'." (1992: 205, n.26). And this is, indeed, how the translator of the Heart Sutra seems to have read the text, translating sarvadharmāḥ śūnyatālakṣaṇā...

This discrepancy is not one of paraphrasing or selecting different synonyms as it is in other cases. The new phrase completely changes the meaning of this sentence, though the (intransitive) actions are the same, the subject "undertaking" the actions is different.

Another major difference here, of which Nattier says "most striking of all", becomes apparent when we look at the Sanskrit Heart Sutra passage:
iha śāriputra sarvadharmāḥ śūnyatālakṣaṇā anutpannā aniruddhā amalā avimalā anūnā aparipūrṇāḥ
Nattier notes (172 and notes) that where the Sanskrit PPS has singular verbal forms, consistent with śūnyatā being the grammatical subject, the Heart Sutra has nominal forms in the plural. She reminds us that while plurals may be marked in Chinese, they frequently are not; and that any given character may function as noun, adjective, or verb depending on context.
Gil.Kj.Heart Sutra
na sā utpadyate 不生 anutpannā
na nirudhyate 不滅 aniruddhā
na saṃkliśyate 不垢 amalā
na vyavadāyate 不淨 avimalā
na hīyate 不增 anūnā
na vardhate 不減 aparipūrṇāḥ
"In each case the Chinese is a perfectly good rendition of the terminology contained in the Sanskrit Large Sutra, while the Sanskrit Heart Sutra, in turn, represents a perfectly good rendition of the Chinese" (Nattier 1992: 172).
Repairing this artefact of translation would be relatively easy, were it not for the phrase 是諸法空相. However, the Chinese Heart Sutra contains this phrase because it is in the source text. Admittedly, the terms have been altered from verbal to nominal forms, and we could fix this; the matter of the extra phrase is more difficult because we only have a small number of Sanskrit texts and it is not found in either, despite the antiquity and relative fidelity of the Gilgit ms.

All Dharmas and the Mark of Emptiness

Where does the phrase 是諸法空相 come from? The CBETA Lexicon tool shows that the phrase does not occur in any Chinese text before Kumārajīva uses it in T223, his translation of PPS. In the Sanskrit PPS (Kimura) the compound sarvadharma occurs quite often with svabhāvaśunya; e.g.:
  • tathā hi svabhāvaśūnyāḥ sarvadharmāḥ. (PSP_2-3:129)
  • svabhāvaśūnyā hi subhūte sarvadharmāḥ. (PSP_4:55)
  • svabhāvaśūnyā hi kulaputra sarvadharmāḥ. (PSP_4:94)
  • tathā hi bhagavan sarvadharmāḥ śūnyāḥ. (PSP_4:130)
But in the whole text śūnyatālakṣaṇa occurs only twice:
bhagavān āha: śūnyatālakṣaṇā hi devaputrā iyaṃ gambhīrā prajñāpāramitā, ānimittalakṣaṇā hi devaputrā iyaṃ gambhīrā prajñāpāramitā, apraṇihitalakṣaṇā hi devaputrā iyaṃ gambhīrā prajñāpāramitā. (PSP_4:67)
"The bhagavan said, for this profound perfection of wisdom, O small gods (devaputrā), has the mark of emptiness; this profound perfection of wisdom, O small gods, has the mark of signlessness; this profound perfection of wisdom, O small gods, has the mark of desirelessness." (Cf Conze 1975: 351 - the text is almost obscured by imposed subject headings).
Which is a reference to the three vimokṣas and thus to śūnyatā as meditative state, not abstract principle. And:
sarvadharmā hi subhūte viviktā asvabhāvāḥ svabhāvaśūnyāḥ, anena subhūte paryāyeṇa yena lakṣaṇena prajñāpāramitā saṃvidyate tenaiva lakṣaṇena sarvadharmāḥ saṃvidyante yad uta viviktalakṣaṇena śūnyatālakṣaṇena. PSP_5:12
"For, Subhūti, all mental objects are isolated, without essence, empty of essence. In this way, Subhūti, perfection of wisdom is recognised by this mark, that is, by the mark of isolation, the mark of emptiness." (Cf Conze 1975: 441, who seems to translate every other verb as "exists", but here and elsewhere saṃvidyate clearly does not mean "exist", but instead means, "is known, is recognised; is perceived")
This doesn't really help us, because, here, it is prajñāpāramitā that is marked with emptiness (śūnyatālakṣaṇa). Now ,we have a third object to which this condition can apply. In a related passage from the Aṣṭasāhasrikā we find:
śūnyatāsvabhāvā hi subhūte pañca skandhāḥ, asvabhāvatvāt / na ca subhūte śūnyatā lujyate vā pralujyate vā (Aṣṭa XII; Vaidya 126. Cf. Conze 1973: 173)
For, Subhūti, the five skandhas are the essence of emptiness, because they have no essence. And, Subhūti, emptiness cannot break or destruct.
The point is that śūnyatā cannot be broken (√ruj, pra√ruj), which is at least related to the idea that it doesn't arise and pass away.
As an aside, here lujyate is from a PIE root *leug̑- "to break" (the only common English cognate is lugubrious). The dialect of the composers of the Ṛgveda only had r; however, the text was redacted by speakers of a dialect that retained the r/l distinction who reinserted l. The Eastern dialect of Māgadhī developed into l-only (King Asoka referred to himself as lāja rather than rāja); whereas Western dialects tended towards r-only (See Despande p.70ff.). This is interesting because recent evidence has shown that the original Prajñāpāramitā text was composed in a western dialect, namely Gāndhārī (Falk and Karashima).
There is an interesting passage in Aṣṭa XV:
iha subhūte bodhisattvā mahāsattvā anuttarāṃ samyaksaṃbodhim abhisaṃbuddhāḥ santo lokasya ākāśagatikaṃ rupamiti dharmaṃ deśayanti | evaṃ vedanā saṃjñā saṃskārāḥ | evameva subhūte sarvadharmā ākāśagatikā anāgatikā agatikā ākāśasamāḥ | yathā ākāśam anāgatam agatam akṛtam avikṛtam anabhisaṃskṛtam, asthitam asaṃsthitam avyavasthitam, anutpannam aniruddham, evam eva subhūte sarvadharmā anāgatā āgatā ākṛtā avikṛtā anabhisaṃskṛtā asthitā asaṃsthitā avyavasthitā anutpannā aniruddhā ākāśakalpatvādavikalpāḥ | (Aṣṭa 15.2)
Here, Subhūti, the bodhisatvas mahāsatvas, being unexcelled fully-enlightened Buddhas, teach the Dharma that form has the [same] condition of space in the world. So also sensation, apperception, and volition. In the same way, Subhūti, all dharmas have the condition of space, not coming, not going, just like space. Just as space does not come or go, it is not made or unmade or shaped, it does not last, remain, or endure, it does not arise or cease, so also all dharmas do not come or go; they are not made or unmade or shaped, they do not last, remain, or endure, they do not arise or cease, they are not falsely distinguished from these aspects of space.
And the reason this is true is that, "all dharmas are in a state of emptiness" (śūnyatāgatika sarvadharmāḥ).

Another interesting passage is the section which uses the Gāndhārī alphabet (a ra pa ca na...) as an acrostic by which to remember various aspects of emptiness upon which to meditate. In PPS (Kimura 1-2: 85) we find the phrase:
akāro mukhaḥ sarvadharmāṇām ādyanutpannatvāt
The letter 'a' is a door, because of the primordial non-arising of all dharmas.
Here mukha is usually (following Conze) rendered as "door" or "opening", but may also mean "mouth, face, head; chief". What the letter 'a' (a-kāra) is, in practice, a mnemonic, a place holder or a reminder, for a word that begins with that letter, i.e., anutpanna 'unarisen'. More hints about the meditation practice are found elsewhere in the text:
"Moreover, Subhuti, the Bodhisattva, who courses in perfect wisdom, admonishes the Bodhisattvas as follow: 'Sons of good family, may you become skilled in the consummation of the letters! May you become skilled in one letter, in two letters, etc., to: in forty-two letters! May you through these forty-two letters come to a state which has moved away from everything. May you meditate on the 42 letters as contained in one letter, and may you meditate on one single letter as contained in 42 letters!" (PPS VIII 5.3; Conze 1975: 587. For more on this, see my essay The Wisdom Alphabet Meditation on visiblemantra.org.)
Although I don't think there is any direct connection between the Heart Sutra and the Arapacana Alphabet, this does, at least, confirm that the idea of dharmas not arising was also stated in this context. However, as discussed in the essays about form is emptiness, I think this refers to the state of śūnyatā-samādhi, where there is no experience. Buddhists involved in the Prajñāpāramitā texts seem to have come to ontological conclusions on the basis of this experience. By this I mean they adopted the stance that śūnyatā-samādhi was reality, or at least a more fundamental reality than what we normally experience. When you take consciousness and subtract all experience, what you are left with is awareness with no subject or object, no spatial or temporal orientation, and so on. This state is often described as "luminous".

While this certainly tells us something interesting and profound about the nature of our minds, I think it is a mistake to turn from epistemology to ontology on this basis. Defining reality on this basis seems, frankly, foolish to me. Reality is almost impossible to understand from a single point of view, which has led to a tendency to solipsism in both Western and Eastern philosophy, even after the power of comparing notes on experience has been demonstrated by scientists.

The solipsistic tendency in "hardcore" Buddhism is pronounced and, perhaps, unavoidable. The experience of no (normal) experience is so vivid and compelling that it must be hard not to use it as an absolute reference point around which we organise our worldview, if we have it, just as ontological dualism seems entirely plausible to those who've had out of body experiences, or God seems to exist for those who've had that kind of experience.


In many cases where the Heart Sutra is problematic, where Conze has made a mistake (Nattier 1992, Attwood 2015) or where the original Sanskrit translator has made a mistake (Huifeng 2014, Attwood 2017), the philologist can see the error and suggest a solution (although some philologers seem reluctant to offer such solutions, I am not). Of course, whether religieux accept such suggestions is another matter. Even errors can be authoritative when they are over 1000 years old.

But in this case there is no obvious resolution. The introduction of the phrase "all dharmas are marked with emptiness" is a discontinuity, because it is not found in any Sanskrit witness, albeit that we have very few Sanskrit witnesses: only a handful of manuscripts in two small caches.

That said, the idea is itself fairly orthodox and in keeping with many statements found elsewhere in the Prajñāpāramitā literature. So it is not wrong in the way that some other parts of the Heart Sutra are wrong. Once again, we see the issues of authority and authenticity are complex with respect to the Heart Sutra. The creator of the text appears to have faithfully copied a passage from Kumārajīva's text, and the Sanskrit translator to have tackled it with some success, even if some of his word choices were not. But where did Kumārajīva get it from? Did Xuanzang also have a source with this phrase, or did he include it because it was in Kumārajīva's text, which he was apparently copying (at least in this passage)?



Conze, Edward. (1973). The Perfection of Wisdom in Eight Thousand Lines and its Verse Summary. San Francisco: City Lights, 2006.

Conze, Edward. (1975). The Large Sutra of Perfect Wisdom. University of California Press.

Deshpande, Madhav M. (1995) 'Vedic Aryans, non-Vedic Aryans, and non-Aryans: Judging the Linguistic Evidence of the Veda', in Erdosy, George. The Indo-Aryans of Ancient South Asia: Language, Material Culture and Ethnicity. Berlin: Walter de Gruyter & Co.

Falk, Harry and Karashima, Seishi. (2012). ‘A First-Century Prajñāpāramitā Manuscript from Gandhāra — parivarta 1 (Texts from the Split Collection 1)’. ARIRIAB 15: 19-61.

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. (2010). Pañcaviṃśatisāhasrikā Prajñāpāramitā. Tokyo: Sankibo Busshorin.

Nattier, Jan (1992). ‘The Heart Sūtra: a Chinese apocryphal text?’ Journal of the International Association of Buddhist Studies, 15 (2) 153-223. Online: http://archiv.ub.uni-heidelberg.de/ojs/index.php/jiabs/article/view/8800/2707
Related Posts with Thumbnails