Variations in the Heart Sutra in Chinese

Huifeng (慧峰)

It's well known that seven versions of the Heart Sutra are preserved in the Chinese Canon. Of these, two are versions of the short text:

• T 8.250  摩訶般若波羅蜜大明咒經
  Móhēbōrěbōluómì dàmíngzhòu jīng
  (= Mahāprajñāpāramitā-mahāvidyā-sūtra).
• T 8.251 般若波羅蜜多心經
  Bōrěbōluómìduō xīn jīng
  (= Prajñāpāramitā-hṛdaya-sūtra).

T 250 is attributed to the Kuchan bhikṣu Kumārajīva, though only by later catalogues, for the first time in 730 CE, which is four centuries after he died (Nattier 214, fn 71); and T 251 is attributed to Xuánzàng,  again only by later catalogues.

There is a third version of the short text, which is a complete transliteration of the Sanskrit using Chinese characters (T 8.256). It has traditionally been attributed to Xuánzàng,  though more recently it has been attributed to Amoghavajra. The remaining versions are all later translations of the long text of the Heart Sutra and don't concern us here.

Both of the two Chinese short Heart Sutra texts are almost identical to sections of Kumārajīva's translation of T 8.223 摩訶般若經 Móhēbōrě jīng (= Mahāprajñāsūtra; Large Wisdom Sutra - a translation of the Sanskrit Pañcaviṃśatisāhasrika-prajñāpāramitā-sūtra or 25,000 Line Perfection of Wisdom Sutra). It is this identity in Chinese, combined with the differences in idiom and sentence structure in Sanskrit versions of these two texts, that lead to the conclusion that the Heart Sutra was assembled in China, albeit from parts of the Large Wisdom Sutra.

Shi Huifeng (Orsborn 2008) has raised a question about part of Jan Nattier's (1992) argument that the Heart Sutra was assembled in China. Although I think the weight of evidence is against his objection, it will be instructive to work through it and think about the implications.

One of the besetting problems of the religious and academic study of Buddhism is the tendency to see texts as monolithic and fixed. We will write about 'The Heart Sutra' for example, as though it was clear what this meant. If my recent work has showed anything it is that the Heart Sutra is far from a unitary phenomenon. It exists in multiple, often competing, recensions: Tibetan Buddhists almost exclusively use  one of two long texts found in their Canon, whereas in Japan and China the short text is most important. In China the version by Xuánzàng is more or less the only text is use. In Japan they have an old (though not as old as claimed and somewhat corrupt) manuscript in Sanskrit that is influential alongside the Chinese. In the West we tend to prioritise the Sanskrit editions by Conze, though I hope to change this by publicising the errors in his edition. And we have very many English translations each of which takes a slightly different approach.

It is very difficult to know which reading of any particular passage is the correct reading. It was only in 1992 that this most popular text in all of Mahāyāna Buddhism, a text which has received intense scrutiny over centuries, was discovered to have been assembled in China. But even the two Chinese short texts are not original: T 8.251 appears to have been edited to conform to some of Xuánzàng's word coinages, but largely follows Kumārajīva's translation (something which Xuánzàng does not do in other cases) and T 8.250 has additions which are drawn from Kumārajīva's Mahāprajñāpāramitāśastra translation (see below). Both contain a phrase that is not found in any Sanskrit manuscript and which must therefore have been added after the creation of the Sanskrit archetype or been deliberately left out of the Sanskrit (viz 度一切苦厄 "overcame all states of suffering"). The former being the more likely of the two explanations because while Buddhists frequently add to their texts they seldom if ever deliberately leave lines out.

Each manuscript is different and not just because scribes made mistakes. Each manuscript has been subtly altered by an anonymous editor who presumably thought to "improve" the text. Yes, we can find and undo these additions where appropriate, but this assumes that we know better than them! Our main advantage being that we have access to more versions than most historical Buddhists would have had. All these versions were presumably in use by Buddhists at some point and were deemed to be authoritative. Of the two versions of the text that were preserved in the Tibetan Canon, both have obvious errors. But they are the canonical versions in Tibet so in practice pointing out the mistakes makes little difference. Tibetan exegetes deftly work around the  textual difficulties. I looked in more detail at some of these points in my article on the Vajrasattva Mantra (which is mostly transmitted in a corrupt version even where canonical versions are less corrupt). Thus in practice it would make more sense to speak of a Heart Sutra tradition or even a number of traditions, including the traditions of exegesis, rather than a single Heart Sutra text. The text that any group had access to at any given time was the Heart Sutra for that group at that time - and this is true even now. 

In the study of Pāli texts we seldom see discussions of variant readings, and the study of the manuscript tradition is neglected compared to Sanskrit studies. Westerners are only beginning to unlock the riches of the Chinese canonical works, and there are still no English translations of the bulk of the Chinese Tripiṭaka, though the first generation of translations are now being expanded in the BDK Tripitaka Translation Series, under the auspices of the Numata Foundation. Some scholars are starting to make use of Chinese texts and it is becoming more common to see comparisons particularly of Pāli and Chinese translations. But we usually see one Pāli and one Chinese translation compared. We see the critical editions being compared without any attention given to alternate readings. The Pali Text Society critical editions and the Vipassana Research Institute Pāli editions, as well as the CBETA/Taishō Chinese edition, all include footnotes about variant readings in other editions or manuscripts. Also note that Silk found considerable variation in the Tibetan Canonical versions of the Heart Sutra based on comparing fourteen exemplars of the various editions of the Tibetan Kanjur.

It may well be that these are of no consequence in the majority of cases, but the lack of any discussion of variants is suspicious. My experience of finding multiple errors in Conze's 60 year old, well scrutinised version of the Heart Sutra has put me on guard. Thus it was of considerable interest to see Huifeng making an argument precisely on the basis of variant readings recorded in the Taishō Tripiṭaka. We first need to set the background for the point Huifeng makes.

The argument about the attribution of T 250 begins at Nattier p.184 (Huifeng places the argument at p.164, which may be a typo). The importance of the argument is that it eliminates Kumārajīva as a potential translator of the Heart Sutra and thus makes an important chronological point. Had Kumārajīva been a plausible translator we would have to push back the creation of the Heart Sutra to the early 5th century. We've already pointed out that the attribution to Kumārajīva is not made until 730 CE - this makes the attribution "highly suspect" (184). Nattier makes four numbered points and adds a fifth as after thought (though it is not inconsequential).

1. Near the beginning T 250 contains an extra passage:
   舍利弗！色空故無惱壞相，受空故無受相，想空故無知相，行空故無作相，識空故無覺相。何以故？
2. In the middle T 250 adds: 是空法，非過去、非未來、非現在。which has no counterpart in T 251.
3. The wording of the phrase 'form is not different from emptiness' is worded differently.
4. T 250 and 251 vary in their rendering of certain key terms: prajñāpāramitā, skandha, bodhisattva, and the names Avalokiteśvara and Śārputra.

The fifth point is that no translation by Xuánzàng ever eclipsed a translation by Kumārajīva in Chinese Buddhism. In other words where Kumārajīva has translated a text, it becomes the standard, right down to the present, despite the existence of an (arguably) better translation by Xuánzàng. And yet the most popular version of the Heart Sutra the one attributed to Xuánzàng and all of the Chinese commentaries, which begin to be produced only in the latter half of the 7th century, are on that single version (188).

Fukui Fumimasa, cited by Nattier, has used the similarity between extra parts of T 250 (points 1 and 2 in Nattier's discussion) and Kumārajīva's Large Wisdom Text (T 233) to argue that Kumārajīva must have translated the text, but Nattier counters that "a nearly verbatim agreement between two Chinese texts should instead arouse our suspicions" (196). The odds of choosing the exact same translation in two different texts (and recall that the wording of the Sanskrit texts is different) are "enormous". (186). The obvious conclusion is not that T 250 is a translation by Kumārajīva, but a portion copied from his T 233 translation. "And this is especially true of a translator like Kumārajīva, who is renowned not for wooden faithfulness to the Sanskrit original but for his fluid and context sensitive renditions." (186).

The different wording at point 3 Nattier attributes to the text being taken not directly from T 233 the Large Wisdom Text, but from his translation of the Mahāprajñāpāramitāśastra: T 25.1509 大智度論 Dàzhìdù lùn (Attributed to Nāgārjuna).

T 251

色不異空，空不異色
sè bù yì kōng, kōng bù yì sè 
For is not different from emptiness, emptiness is not different from form

T 250

非色異空，非空異色
fēi sè yì kōng, fēi kōng yì sè
It is not that form is different from emptiness, it is not the emptiness is different from form.

With respect to the passage under discussion here T 1509 reads:

非色異空，非空異色

Nattier concludes: "In other words, the Heart Sutra may be viewed as the creation of a Chinese author who was more familiar with the Large Sutra as presented in this widely popular commentary than with the text of the sutra itself" (187). In my view is true of T 250, but I don't think it can apply to T 251 which is entirely consistent with the sutra (i.e. with T 233) except where a few of Xuánzàng's translations of individual words have been used (i.e. those mentioned in point 4), which suggests that it was altered post-composition.

However Nattier also sees this scenario as consistent with point 4. Where the terms mentioned above are used, T 251 has the characters typical of a translation by Xuánzàng, while T 250 uses characters typical of Kumārajīva. Of course if the Heart Sutra is drawn from another text translated by Kumārajīva then we are not surprised to find his terminology in use. T 251 can then be explained as having been edited by someone familiar with Xuánzàng's vocabulary. 

The section of the text that Huifeng draws attention to, dealt with by Nattier point 3, is the part that translates as "form is not different from emptiness, emptiness is not different from form" (Huifeng 10-11, n.34). As above this is different in the two Heart Sutra texts.

T 251 色不異空，空不異色
T 250 非色異空，非空異色 

Kumārajīva's Large Wisdom Text (T 233) here is different to T 250 :

T 223 色不異空空不異

Punctuation aside, and all Chinese punctuation comes from modern editors, the characters in T 251 and T 223 are identical. While those in T 250 are different. On this basis Nattier argues that T 250 is drawn from T 1509 and probably not the work of Kumārajīva.

However, Huifeng points out that the Taishō edition footnotes record that in the earlier Sòng, Yuán, Míng and Gōng editions of the Tripiṭaka T 223 reads 非色異空，非空異色. This is exactly the same as T 250. Thus the third point regarding the attribution of T 250 to Kumārajīva is not valid and the Heart Sutra found in T 250 precisely matches the earlier editions of the Large Wisdom Sutra (T 223) which is known with certainty to be the work of Kumārajīva. This invalidates Nattier's comments regarding point 3, which is an important part of her argument about the authorship of T 250. There is now an imperative to examine other such arguments in more detail.

Problematically Huifeng infers from this what he considers to be a reasonable doubt about the entire Chinese Origin hypothesis.

"Detailed analysis would be needed to determine whether or not the rest of her argument and conclusions, and thus [the] conclusion that the Heart Sūtra is a Chinese apocryphal work, are still valid" (11, n.34)

I think Huifeng is overstating the doubts that emerge from his valid point. If he had provided evidence that definitely linked T 250 with Kumārajīva then that would be a game changer. But there is a weight of other evidence provided in Nattier's article that undermines the attribution. Particularly the lateness of the first attribution, which is in fact some 300 years late. Kumārajīva is not an obscure translator whose work went unnoticed. Where he translated a text it is usually still the standard version even in the present day (this is true of the Lotus Sutra for example). If he translated the Heart Sutra the lack of any mention of this before 730 CE requires an explanation. Also there are the extra passages. It is axiomatic that texts are frequently added to, but seldom reduced in scope. This is plainly evident in the extant Sanskrit manuscripts that I have examined for example. One passage of 37 characters and another of 12—which incidentally appear in no Sanskrit source either—do not just disappear from a Buddhist sutra.

Nattier goes further, saying "nor is there any evidence that Kumārajīva himself had any role in the production of the 'translation' associated with his name" (184). And this is the considered opinion of previous scholars working in the field where "significant progress [in understanding the origins of T 250] has recently been made by Japanese and Western scholars" (184).

There is no doubt whatever that the Heart Sutra is largely based on the Pañcaviṃśati. This fact is already acknowledged in some of the Nepalese mss., which title the text: ārya-pañcaviṃśatikā-prajñāpāramitā-hṛdaya nāma dhāraṇī (The Dhāraṇī entitled the Heart of the 25,000 Line Perfection of Wisdom).

If the creator of the Heart Sutra took their extracts from a Sanskrit text then they either a) had a previously unknown version of the Sanskrit Pañcaviṃśati with an entirely different wording, though largely the same meaning; or b) the composer deliberately changed the wording of the Pañcaviṃśati to create a paraphrase without substantially changing the meaning, a procedure unknown in any other case. In either a) or b) the result was a Sanskrit Heart Sutra in which the wording, but not the meaning, was different from the extant versions of Sanskrit text it was extracted from (i.e. Pañcaviṃśati) and also contained one or two phrases that make better sense in Chinese than in Sanskrit. Neither of these two options are very likely. By contrast the differences between the Large Wisdom Text and the Heart Sutra in Chinese are minimal. For the most part the wording (including choice of characters) is identical. The obvious conclusion, the conclusion demanded by Occam's Razor, is that the extraction happened in Chinese and in China; and that the Sanskrit translation of the short text Heart Sutra was done by a Chinese speaker. Unlike the short-text, the extra passages found in the long text Heart Sutra seem to have been composed by someone well versed in Sanskrit and the conventions of Sanskrit Buddhist literature.

To Nattier's extensive evidence I have added the example of the word vidyā in Pañcavimśati becoming the word mantra in the Heart Sutra, which is difficult to explain in purely Sanskrit terms, but makes sense given a Chinese intermediary and the Chinese practice of abbreviating technical terms. I've tried to show how the Heart Sutra might look had it been extracted from a Sanskrit Pañcavimśati from that time period rather than from the Chinese (An Alternate Sanskrit Heart Sutra). I also showed that Nattier's arguments about satyam amithyatvād in relation to the Chinese text are better explained by Conze's mistaken choice for his critical edition: the text should have read samyaktva-amithyātvāt.

Thus I think we can say that, though Nattier's argument is incorrect in one or two small details, the case for the Heart Sutra being assembled in China is very strong and that extraordinary evidence would be required to refute this theory. Huifeng's main point about versions is certainly true, and it does change the picture, but only a little. That said a still more nuanced story might still emerge from paying attention to the footnotes and variant readings. Huifeng is correct to wonder what other variations might reveal if studied. And variations are the norm rather than the exception when dealing with Buddhist texts.

~~oOo~~

Bibliography 

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

• Orsborn, M. B.  [aka 釋慧峰 Shi Huifeng] (2008) A Survey Of Prajñāpāramitā Sūtra Translations In Chinese. Online: http://prajnacara.blogspot.co.uk/2011/06/survey-of-prajnaparamita-sutra.html.

Why is there a Dhāraṇī in the Heart Sūtra?

"[The Dhāraṇī Chapter is] another later addition probably when Dhāraṇī was extensively taken into the body of Buddhist literature just before its disappearance from the land of its birth. Dhāraṇī is a study by itself. In India where all kinds of what may be termed abnormalities in religious symbology are profusely thriving, Dhāraṇī has also attained a high degree of development..." - D T Suzuki. The lankavatara Sutra. p.223, n.1.

D. T. Suzuki

D T Suzuki is a key figure in the translation of Buddhism to the West. He is cited by David McMahan as representative of the Romantic influence in Modernist Buddhism. However, this quotation and those below show that he was also under the influence of Scientific Rationalism -- though perhaps his response to dhāraṇī as an "abnormality in religious symbology" is rather more emotional than rational.

Suzuki is not alone in struggling to find a place for dhāraṇī in Buddhism. In his History of Buddhist Thought, E. J. Thomas suggests that dhāraṇī has "infected popular Buddhism" and "spells form an important part of popular Buddhism, but they have nothing in themselves peculiarly characteristic of Buddhism. They are a form of sympathetic magic” (186. Emphasis added). So far from the Western conception of Mahāyāna Buddhism is the dhāraṇī that it gets no mention in Paul Williams' oft cited textbook, Mahāyāna Buddhism, despite being a central element in Mahāyana Buddhism as it is practiced around the world.

Suzuki is equally vehement when it comes to the dhāraṇī in the Heart Sutra which he refers to as "apparently a degradation or a degeneration" (228). [1]

“A Mantram or Dhāraṇī is generally supposed, when uttered to effect wonders… Can we say, then, that the end of the Buddhist disciplines can be attained by means of a mere mystic phrase?” (Suzuki Essays. 229)

"Another thing which makes this presence of a Mantram in the Hṛidaya more mystifying is that the concluding Mantram is always recited untranslated as if the very sound of the Sanskrit-Chinese were a miracle working agency."

His polemic against the Heart Sutra dhāraṇī is quite lengthy and concludes:

"taken in itself has no meaning, and its vital relation to the Prajñāpāramitā is unintelligible”. (Suzuki Essays. 236)

The conceit that if I cannot understand a subject then it is "unintelligible" is common in scholars of Suzuki's generation and before. However, contemporary scholar, Donald Lopez, is likewise puzzled, and, so he tells us, are the Indian commentators whose works are preserved in Tibetan. He says:

"The question still remains of the exact function of the mantra within the sūtra, because the sūtra provides no such explanation and the sādhanas make only perfunctory references to the mantra". (Lopez. The Heart Sutra Explained. p.120)

The question about the role of the dhāraṇī in the Heart Sutra raises a deeper question. We know that from the second century CE a large number of dhāraṇī sūtras were composed and that they were clearly very important to Buddhists of the time and some of them continue to be important in Asian Mahāyāna Buddhism (A Zen practising friend recently sent me a dhāraṇī for removing obstacles to help me find a new place to live!). The collection of Nepalese manuscripts at Hamburg University lists hundreds of dhāraṇī sūtras! We know that from around that time or a little later, dhāraṇīs began to be added to Mahāyāna sūtras such as the Laṅkāvatāra, the Saddharmapuṇḍarīka and the Suvarṇaprabhāsottama. Given the apparent prominence of dhāraṇī in Mahāyāna Buddhism from relatively early in the development of it, why are modern commentators on Mahāyāna surprised to find a magical chant in a Buddhist sūtra? Let's try to deal with this question first, and then look at the specifics of the Heart Sutra.

Buddhism is inseparable from magic. Many of the most interesting stories in the life of the Buddha involve magic and miracles. The story of the conversion of the Kassapas at Gaya involves a whole series of magical feats, for example. The very awakening experience is often retold in terms of the development of specific super-normal powers such as knowledge of past-lives or clairvoyance (and plenty of living Buddhists claim to have experienced these magical powers). As time goes on, more and more miracles and magic are added to the stories, but there is no point at which magic and the supernatural is not evident. There is no absolutely no reason to assume that 'original' Buddhism was a purely rational affair. This is a Western conceit that has distorted Buddhism from at least the 19th century onwards. The idea of "rational Buddhism" is ironic, because the initial European view of Buddhism was that it was a form of irrational, idolatrous, superstitious, heathen religion (this is brought out in Philip C. Almond's The British Discovery of Buddhism).

As Westerners we have developed an ambivalent attitude to magic since the European Enlightenment. There seems to be a spectrum of belief from entirely credulous to entirely sceptical but, on the whole, we consider magic to be a form of entertainment, not to be taken too seriously. Or we treat it as a vulgar relative of the religious miracle.

Be that as it may, there is no getting away from the facts: Buddhist texts are full of magic. And, in particular, they are filled with that magic which involves spoken sounds. I've recently made reference to the paritta and to the sacca-kiriya or 'truth act', both of which are encountered in early Buddhism. I'll be saying more about the sacca-kiriya in a couple of weeks.

The most obvious answer to the question of why anyone would be surprised to find a magic spell in a Buddhist text is that they are fooling themselves. Magic and superstition is exactly what we ought to expect. But, more specifically, the presence of a dhāraṇī in a late Mahāyāna text is exactly what we expect to find. Mahāyāna texts are full of dhāraṇī. 

From as early as the second century, sūtras were being composed specifically as a vehicles for dhāraṇī. By the fourth century those texts which did not have one, had them added (which is what Suzuki is referring to in the first quote). A number of dhāraṇī continue to be popular in China and Japan today: e.g., the Nīlakaṇṭha Dhāraṇī or the the Karaṇḍamudrā dhāraṇī. Several times a year people write to me to ask if I have the Uṣṇīṣavijaya Dhāraṇī in Siddhaṃ script (I don't). If you look around there are dozens of YouTube videos with the dhāraṇī being chanted in Chinese transliteration.

The presence of dhāraṇī in Buddhist texts is used by some to argue that Tantric Buddhism dates from as early as the second century, but it really doesn't. Tantra is a synthesis of a huge variety of religious ideas and forms which happened in India in the 6th-7th centuries CE. Ronald M. Davidson plausibly argues that the synthesis was given impetus by the breakdown of socio-political structures across India with the fall of the Gupta Empire and the resulting chaos. This places the Tantric synthesis in the late 5th century at the earliest, but the 6th century is more likely. The first  fully developed Tantric text, Mahāvairocana Abhisambodhi Tantra, dates to the mid 7th century (probably the 640's) according to Stephen Hodge. The cult of dhāraṇī predates Tantra by several centuries.

The fact is that our information about Buddhism is curtailed (to be generous, or censored to be cynical). Thus, dhāraṇī centred Buddhism does not feature much in Western narratives because it not only does not put aside superstition, but places magic at centre stage. I've argued at some length that most Buddhism is sharply dualistic, with belief in a distinction between matter and spirit, and a profound bias for spirit. Magic is largely, though not completely, excised from modernist Buddhism. It is particularly rigorously rooted out when it has any connection to the mundane or physical world - spirit magic is more acceptable. Using Buddhist (magical) techniques for mundane purposes is frowned on, even though this has been the norm for millennia. Many Western Buddhists are dismissive of Soka Gakkai for precisely this reason. Mindfulness therapies are also attacked for commodifying the Dharma and because their goals are oriented towards material well-being rather than transcendence of suffering (though some practitioners, with varying degrees of plausibility, deny these charges).

Something to keep in mind is that mature Mahāyāna held that awakening required three incalculable aeons to accomplish. There were times and places where this made awakening seem impossibly far off. When awakening is absolutely transcendental, and thus of infinite value, it is also beyond the reach of ordinary mortals. Under these conditions intellectualism and scholasticism seems to flourish. Buddhism becomes a rather academic exercise for monks, and a way to ward off evil for lay people; and, after all, lay people are not expected to be actively pursuing awakening in most Buddhist societies, anyway. This was certainly true of, for example, the Mahāyāna Buddhism in late 8th century Japan. It stands out because it was against this backdrop of impossibly far off awakening, and absolutely separate Dharmakāya, that Kūkai and Saichō introduced Tantra to Japan. Kūkai's claim that tantra was a communication directly from the Dharmakāya and his maxim 'awakening in this very life' at first caused confusion and resistance in the Japanese Buddhist establishment, before setting off a religious revival that continues to inform Buddhism in that country.

Our pre-scientific forbears lived in dangerous world at the mercy of disease and elements. They knew, as we know, if we are honest, that we are not in control. We don't even control our own bodies, which age and get sick against our will. The attraction of any technique which claimed to exert control over nature and the cosmos, which held chaos and danger at bay, ought to be obvious. In our own time a man was recently jailed because he sold a device consisting of an aerial connected to a plastic handle to the Iraqi military as a bomb detector. The units sold for more than US$30,000 each, and the Iraqis are said to have bought about 6000 of them. Faced with uncontrollable forces we are desperate for some protection.

When my appendix grumbled and caused me intense pain, I did not resort to chanting the Heart Sutra, I high-tailed it to a hospital and had them remove the offending organ under a general anaesthetic before it killed me. But in the ancient world the Heart Sutra might have been all that I had. It would not have saved me, but it might have made me feel better for the short time before I died of septicaemia. It might have soothed my relatives as they watched me die in agony. I know that many moderns argue for the efficacy of ancient medicine. It's true that chewing willow bark would ease pain because it contains salicylic acid. But without any knowledge of the mechanisms of the body or of the chemistry of such remedies there was no way to be truly systematic about it. We see this in the medieval medical approaches such as humours, acupuncture, and āyurveda. All of these approaches to wellbeing are unrelated to the way the body actually works and rely on magical explanations to explain their efficacy. There is an irreducible element of magic because when they were invented magic was the most potent counteractive to the uncontrollable forces in those peoples' lives. At least they believed it was. And in the face of violent or painful death who can blame them? It's harder to explain their persistence or the emergence of a non-system like homeopathy. 

How Should We Understand Magic?

What is magic? By which I mean what is it really? I don't take magic on its own terms. Indeed, I don't think we should take any form of tradition on its own terms. However, understanding the persistence of magic requires careful consideration. On the face of it, magic simply does not do what it says. Hence, its value almost certainly lies elsewhere. Over the years a number of Western attempts to understand magic have suggested that believers were simply childish, idiots, delusional, or wishful thinkers; or that magic was meaningless symbolic action. But none of these theories are enough to explain the persistence of magic, particularly in the realm of healing. Most of them reflect Western intellectual laziness more than informed comment. My own preference is for the work of Ariel Glucklich. His in-aptly named book The End of Magic, offers a critique of Western attempts to explain magic and, on the basis observation of modern Tantric healers in India, tries to do better. I touched on this back in 2008 in an essay called Mantra, Magic, and Interconnectedness. Glucklich summarises:

"Magic is based on a unique type of consciousness: the awareness of the interrelatedness of all things in the world by means of simple but refined sense perception... magical actions... constitute a direct, ritual way of restoring the experience of relatedness in cases where that experience has been broken by disease, drought, war, or any number of other events." (12)

The theme of interrelatedness is one that is deeply resonant for Buddhists. Even before the direct theme of interconnectedness emerges in Buddhism, we find descriptions of the experience of radiating love, treating all beings as a mother treats her only child, and making no distinctions between self and other, etc. These are redolent of one of the most common of all so-called mystical experiences, oceanic boundary loss, in which one feels at one with everything, completely selfless, and blissful. I think this experience is central to understanding Indian religion and seems to underlie a lot of the religious impulses, certainly in Buddhism. We reify and cheapen the experience with maxims like 'all is one', but the experience of being at one with everything ought not to be dismissed lightly. I suggest that if we are seeking a rational explanation for the purpose and persistence of magic in Buddhism this is where we might usefully start.

Another observation I've made about dhāraṇī is that the early dhāraṇī show every sign of not being in Sanskrit, but of being from a Prakrit (or vernacular) tradition. When we see a dhāraṇī like the one that Suzuki was complaining about in the Laṅkāvatāra...

Tuṭṭe tuṭṭe vuṭṭe vuṭṭe paṭṭe paṭṭe kaṭṭe kaṭṭe amale amale vimale vimale nime nime hime hime vame vame kale kale kale kale aṭṭe maṭṭe vaṭṭe tuṭṭe jñeṭṭe spuṭṭe kaṭṭe kaṭṭe laṭṭe paṭṭe dime dime cale cale pace pace badhe bandhe añce mañce dutāre dutāre patāre patāre arkke arkke sarkke sarkke cakre cakre dime dime hime hime ṭu ṭu ṭu ṭu ḍu ḍu ḍu ḍu ru ru ru ru phu phu phu phu svāhā.

... our first thought is not Sanskrit literature or any other kind of high culture. The doubled retroflex consonants and -e endings point away from Sanskrit as the language of composition and towards a Prakrit like Māgadhī. This form is all about rhythm, repetition and alliteration, and lacks any formal grammar or semantics. The point of this is not what it means, because this is not language as such. The point is that in chanting it one has an experience that will be partly determined by culture and conditioning and partly by circumstances. This dhāraṇī is meant to be chanted aloud, repeatedly, probably collectively in a ritual context. And the experience of chanting it is the point of chanting it. We have very little information about the way these sounds were understood or used. 

These sounds were probably made to protect the chanter from misfortune, not to gain any rarefied spiritual attainment. This is about survival, about invoking chthonic forces to come to your aid in times of trouble. It is something to chant when feelings of being isolated or alienated from your people, family, or tribe threaten to overwhelm you. And we know from the literature of China that the Heart Sutra itself was used this way. If Glucklich is right, then the chant protects the chanter by (re)establishing their sense of connection to everything.

We don't, or no longer, live in a culture where it makes any sense to approach interrelatedness in quite this way. However, we are not a million miles from it, either. For example, we still sing together and through that experience, especially in large gatherings, can experience the relatedness that music or chanting facilitates. Rhythmic chanting stimulates endorphin production, especially when done in groups, but the effect must have gone beyond a feel good factor. I've done exercises with groups which encourage a careful listening to the sounds of chanting, and stimulated receptivity to those sounds, and achieved a heightened sense of both calm and connectedness. It's possible to use vocal sound in this way because of the way that it affects us. Non-word vocal sounds both evoke and communicate emotions without involving the intellect (Compare the study of conversational grunts). I believe that there are good evolutionary explanations for this, but the experience itself is far more persuasive. It is enough for most people to chant together and experience a sense of connectedness to their fellows.

This experience is overlaid with various ideas in whatever culture it is experienced. Ancient Indians no doubt believed in the magical efficacy of chanting and with the development of Tantra layered meaning onto the spoken or chanted sound. In a reading culture we forget that the Dharma was primarily an aural experience for most of the last 2500 years. To be well versed in the Dhamma was to be a sutavant, 'one who has heard' or a savaka, 'a hearer'. The divine revelations of the Veda were called 'śruti' 'hearing'. One did not read silently in our culture until relatively recently, and even now we hear the words in our heads as we silently read them. We probably all know that a well spoken poem, for example, is a very different aesthetic experience to a silently read one.

The dhāraṇī cult left behind little or nothing about the mechanics of dhāraṇī - so far as I am aware. While the Arapacana alphabet and the idea of mnemonics are often invoked, they are seldom relevant except in that specific context. Nor is the idea that the dhāraṇī represents a condensation of the text found in actual texts to my knowledge (except perhaps in the later dhāraṇī texts). Neither explanation obviously applies in the case of the Heart Sutra. By the time dhāraṇī start appearing in Buddhist texts it's almost as if their origins are already forgotten. One can see the forgetfulness in the development of the Arapacana as it is expressed in the Sanskrit language - the seeming ignorance that it is the Gāndhārī alphabet is already apparent in the Sanskrit Pañcaviṃśati. Indeed, it's quite difficult to tell what the word dhāraṇī means from the way it is used in texts. The etymology is clear enough, it's from √dhṛ, 'to bear', and thus cognate with dhara, 'bearer', and dharma, 'foundation'. However, for example, when we try to understand it in context, in a text like the Pañcaviṃśati, the impression we get is quite vague. We only know that it is associated with these strings of syllables, but not how it is associated or why.

Conclusion

To sum up, the Heart Sutra has a dhāraṇī in it because it was de rigueur for a Buddhist text of that era to have one. To have composed a text and not included a dhāraṇī would have been odd by the standards of the time and place. Every Buddhist, particularly in China, would have had a favourite dhāraṇī or two that they chanted for protection and prosperity. Dhāraṇīs were a prominent feature in monastic liturgy.

The early dhāraṇī probably emerged from a Prakrit speaking milieu and to some extent they retained Prakrit features (such as the -e ending) but were gradually Sanskritised as Sanskrit became the standard language for Buddhist texts in India. Dhāraṇīs invoke a kind of magic which we no longer understand - partly because no records were made of how it worked, and partly because it was later overlaid by Tantra. However, Ariel Glucklich's theory of magic gives us a way to understand dhāraṇī chanting with a degree of rationality that does not diminish the experience of the practice.

The Prajñāpāramitā texts the Heart Sutra was made from lack dhāraṇīs (pre-dating the cult by a century or two), so a suitable dhāraṇī was found in other popular texts and adapted for the purpose. Since it had no contenders, it became the de facto Prajñāpāramitā dhāraṇī. The relationship to Prajñāpāramitā does, indeed, seem to be arbitrary, though I suspect the initial choice had meaning for the monk who chose it.

~~oOo~~

Notes

1. I have tried to show that the so-called mantra in the Heart Sutra is in fact a dhāraṇī, and that the word 'mantra' itself a substitute for 'vidyā' and is used in the text only because of a confusion caused by being translated back and forth between Sanskrit and Chinese at different times.

For more reading on the role of dhāraṇī in Chinese Buddhism I recommend this article which turned up after I finished writing this essay:

Mcbride, Richard D. II (2005) 'Dhāraṇī And Spells In Medieval Sinitic Buddhism.' Journal of the International Association of Buddhist Studies. 28(1): 85-114. Online: http://archiv.ub.uni-heidelberg.de/ojs/index.php/jiabs/article/viewFile/8958/2851

Had I read this earlier I would have incorporated it into the previous essays.


Bibliography

• Davidson, Ronald M. (2003) Indian Esoteric Buddhism: A Social History of the Tantric Movement. Columbia University Press.

• Glucklich, Ariel. (1997) The End of Magic. New York, Oxford University Press. 

• Hodge, S. (trans.) (2003) The Mahā-vairocana-abhisambodhi tantra : with Buddhaguhya’s commentary. London : Routledge Curzon. 

• Suzuki, D. T. (1971) Essays in Zen Buddhism: third series. Red Wheel/Weiser. [First published 1934]

• Suzuki, D. T. (1991) The lankavatara Sutra: a Mahayana text. Taipei : SMC Publishing. [First published 1932]