The Essence of all Mantras

I declare that A
is the essence of all mantras,
and from it arise mantras without number;
and it produces in entirety the Awareness
which stills all conceptual proliferations.

The Mahāvairocana Abhisaṃbodhi Tantra
translated by Stephen Hodge (XVIII.3, p.326-7)

I was fortunate enough to attend a lecture by Professor Richard Salomon recently. He heads up the Early Buddhist Manuscripts Project which is based around Kharoṣṭhī script manuscripts from Gāndhāra and in the Gāndhāri language. These texts which are held in the British Library are very old, dating to the 1st or 2nd century common era. Gāndhāra is a very interesting area, having been the entry point to India for immigrants, traders, and invaders for many centuries. So it was a very rich and diverse culture. Kharoṣṭhī was the first script used to write India languages, and that it was derived from the version of the Aramaic script used by various Persian conquerors. In Kharoṣṭhī there is one sign for an initial vowel - the short a. To indicate other vowels one uses diacritic marks, in the same was that medial and final vowels are indicated by diacritic marks on consonant signs. Kharoṣṭhī was later displaced by Brahmī from which all modern Indian scripts (as well as most South-east Asian scripts, and the various forms of Tibetan writing)

Regular readers will be aware that I've been interested in the Arapacana alphabet for a while. One of the features of the Arapacana is that is has only one initial vowel sign. Professor Salomon has shown that this is almost certainly because it was the alphabet of the Gāndhāri language which was written in Kharoṣṭhī. It seems that this is a related to the absence of initial vowels in the Aramaic script - they are not used in Semitic languages. When designing a script to write Buddhist texts one needs to be able to write initial vowels, for instance: evam mayā śutam (Thus have I heard which begins all Buddhist sūtras). Brahmī scripts use a different sign for each vowel (although long vowels are indicated with diacritics marks in most cases).

Kharoṣṭhī vowels
a i u e o ṛ aṃ

Kharoṣṭhī created a single vowel sign on the model of the consonant signs - it is simply 'a' if unadorned, but can become any vowel with diacritic marks.


The quote at the beginning of this post may not be familiar, but the sentiment might be. The letter a has this special place in Buddhist thought and practice. One explanation is that the letter a, when added to the beginning of most Sanskrit nouns, it turns them into their opposite: vidya is knowledge, while avidya, is ignorance. This allows us to use the letter a to stand for the Truth which cannot be fully comprehended by language: it is possible to negate any definite statement about the transcendental (including this one!).

However I don't think this alone accounts for the notion that the letter a is the source of all mantras, if only because the a- prefix for verbs usually indicates the imperfect past tense rather than any sense of negation. Another idea relates to the way that Indic alphabets attach an inherent short letter a to each consonant. So the Sanskrit consonants are written as syllables or phonemes - called akṣara - (e.g. ka kha ga gha ṅa); not simply letters (e.g. k kh g gh ṅ). As in Kharoṣṭhī, medial and final vowels are indicated by diacritic marks. This is quite a good way of looking at it, but there is still a slight flaw which involves the vowels.

Sanskrit vowels in Siddhaṃ script
a ā i ī u ū e ai
o au aṃ aḥ ṛ ṝ ḷ ḹ
अ आ इ ई ए ऐ
ओ औ अं अः ऋ ॠ ऌ ॡ

The vowels, except for ā, aṃ and aḥ , can't really be considered to derive from the letter a. All vowels are similar in that they are voiced similarly - differences in sound are due to shifts in the tongue and lips changing the resonant frequency of the vocal track, but it doesn't seem to be enough to consider, say, the letter ī to derive from the letter a. Graphically the vowels are mostly not related to the shape of the letter a either. This is all true of the Brahmī derived scripts. It is not quite true for Kharoṣṭhī however because of the single initial vowel.

My suggestion is that the special function of the letter a in Buddhism is a relic of the Gāndhāra area. It is only in Kharoṣṭhī that all signs for letters derive from, or contain, the short a.

One piece of supporting evidence comes from the Sūtra of Perfect Wisdom in 25,000 Lines. This sutra was probably composed in the 2nd or 3rd century, and is preserved in a variety of Sanskrit originals, as well as in Tibetan and Chinese translations. In the sūtra the alphabet is used as a mnemonic for a series of reflections on the nature of phenomena. Each letter is indicated by a keyword starting with that letter; and each word is the basis for a line of verse. Being a Sanskrit text one might expect the Sanskrit alphabet to be used, but it is not. The alphabet is a partially Sanskritised version of the Arapacana alphabet. Even in the fully Sanskritised version of this practice - present for example in the Mahāvairocana Abhisaṃbodhi Tantra - the vowels are sometimes left off so we have the Sanskrit consonants, but the letter a as the only vowel. The tradition is preserved and the trail seems to lead back to Gāndhāra, at least on Indian soil.

I say "on Indian soil" because the use of alphabetical verses, that is to say verses in which the first letter of the first word of each line are in alphabetical order (a kind of acrostic) is unknown in pre-Buddhist India. Verses were organised by length, and by numerical schemes, but not alphabetically. Verses were arranged alphabetically in Semitic cultures, so there are Old Testament psalms and Manichean hymns with verses in alphabetical order. Which brings us around in a circle to the Semitic origins of Kharoṣṭhī.

The letter a, then , is the source of all the other letters in the alphabet; and the alphabet is the source of all the mantras - hence the composer(s) of the Mahāvairocana abhisaṃbodhi Tantra could say that "from [a] arise mantras without number".

If you'd like to learn to write the letter a in the Siddhaṃ script then visit my other website: visiblemantra.org

image: Siddhaṃ letter a from AKARA : The Quest for Perfect Form
(although it looks identical to one in John Steven's book Sacred Calligraphy in the East, p43.)

The Heart Sūtra - Indian or Chinese?

In this post I want to call attention to an important article, now over 15 years old, but with hardly any recognition outside academic circles. The article is:

Nattier, Jan. (1992). "The Heart Sutra : a Chinese apocryphal text?"Journal of the International Association of Buddhist Studies. Vol. 15 (2), p.153-223. https://journals.ub.uni-heidelberg.de/index.php/jiabs/article/view/8800

Jan Nattier (left) is Associate Professor of Religious Studies, Indiana University, and a scholar of great merit and interest. The article is a fine example of contemporary scholarship, meticulously reasoned, well structured, and typically for Nattier, well written. This last is a strong feature of Nattier's published work - she can write very well. However the article also offers a startling conclusion with wide implications for Buddhists.

The main argument of the article is that the Heart Sutra was composed in China, incorporating some verses from the Chinese version of the Large Prajñāpāramita text, and back translated into Sanskrit sometime in the 7th century. Nattier also offers an explanation for the two different versions, one longer and one shorter, of the Heart Sutra. Page references are to Nattier's article.

Nattier focuses initially on the shorter version of the Heart Sutra. This has several problematic features which distinguish it from sutras generally and the other Prajñāpāramita sutras in particular:

• it does not begin with "thus have I heard"
• there is is no audience reaction at the end of the sutra
• the Buddha makes no appearance 
• Avalokiteśvara, the bodhisattva of compassion makes an unexpected appearance, while the usual characters of Prajñāpāramita sutras (such as Subhuti) do not
• the sutra contains a mantra, which few other Prajñāpāramita sutras do, and then only the later tantric sutras. 

Any explanation of the origin of the Heart Sutra should provide some insights into these oddities, and Nattier's article does just this.

It has been known for centuries that the lines beginning with "form is not other than emptiness" and ending with "no knowledge and no attainment" are quoted from the Pañcaviṃśatisāhasrikā Prajñāpāramita , or Perfection of Paragnosis in 25,000 Lines (hereafter the Large Sutra). The first reference to this borrowing is in a Chinese commentary from the 7th century [note: this is a reference ot the commentary by Kuiji, T 1710]. 

Nattier spends quite some space looking at the various versions of these lines. They occur in four places:

• Sanskrit Large Sutra (using the oldest extant manuscript from Gilgit)
• Chinese Large Sutra (trans. by Kumarajiva)
• Sanskrit Heart Sutra (Conze's critical edition)
• Chinese Heart Sutra (trans. Hsuan-tsung)

Nattier makes several comparisons. Firstly the Chinese Heart Sutra and the Chinese Large Sutra. These are laid out side by side and even without being able to understand the Chinese characters, it is obvious that they are virtually identical. Next the Sanskrit Heart Sutra and the Chinese Heart Sutra are compared and we find a "virtual word for word correspondence" (p.160). However comparing the Sanskrit Large Sutra and the Sanskrit Heart Sutra we find many differences of vocabulary and word order, although the meaning is synonymous. An example is:

Sanskrit Large Sutra : (na)anyad rūpam anyā śunyata / nānya śunyatānyad rūpa
Sanskrit Heart Sutra : rūpān na pṛthak śunyatā / śunyatāyā na pṛthag rūpam

In the list of negated dharmas [i.e. not arising, not ceasing], the Sanskrit Large Sutra uses singular verbal forms, is more repetitious and slightly longer; while the Sanskrit Heart Sutra uses plural adjectival forms, and is shorter and more succinct. Almost every word, barring some very well known technical terms such as śunyata, are different. Conze explains the differences in repetition as a process of summarising, however Nattier contends that this runs counter to the general Indian tendency to elaboration. In any case the changes in vocabulary are unprecedented and "there is no straight forward way to derive the Sanskrit Heart Sutra from the Sanskrit Large Sutra, or vice versa." (p.167)

The best way to understand the progression is that the verses moved from the Sanskrit Large Sutra to the Chinese Large Sutra, and thence into the the Chinese Heart Sutra, and finally into the Sanskrit Heart Sutra. Which is to say that it is far more plausible on philological grounds that the Sanskrit Heart Sutra is a translation of the Chinese rather than the other way around.

Nattier proceeds to marshal supporting evidence for this conclusion beginning by considering known examples of back-translation - these are plentiful in Mongolian scriptures apparently. An important sign of back-translation is the choice of "unmatched but synonymous terms" (p.170). Also there may be occurrences of incorrect word order, grammatical errors point to the under lying language. In this case the evidence points to the Chinese Heart Sutra as being a likely intermediary between the Sanskrit Large Sutra and the Sanskrit Heart Sutra: where the former has nirodha (extinction), the latter has kṣayo (destruction) while the Chinese Heart Sutra has chin [i.e. jìn 盡] which can be a translation of either. This turns out to be true for each synonym in the Sanskrit texts.

Historical evidence also supports the argument. Indian commentaries* cannot be dated to before the 8th century, while there is no independent evidence such as quotes in other texts which might place it earlier. By contrast Chinese commentaries are definitely dated in the 7th century, and "..the existence of the Heart Sutra is attested in China at least a century before its earliest known appearance in India" (p.174)

* i.e. the Indo-Tibetan commentaries; there are no Indian commentaries. 

However, there are still some problems. In particular, the Chinese were usually very particular when composing apocryphal texts, taking a lot of effort to make them look like Indian sutras, and yet the Heart Sutra clearly lacks many important features. Nattier cites a Japanese study (by FUKUI Fumimasa) which she says make a strong case for reconsidering the Chinese title of the Heart Sutra: hsin ching [i.e. Xīn jīng 心經]. Fukui says this should be understood not as saying that the text is the heart, or essence of the Prajñāpāramita tradition, but rather represents a dhāraṇī scripture, i.e. simply a text to be chanted. It is clear that this has indeed been the function of the text since its earliest mentions. The missing attributes (such as the "thus have I heard") are less of a problem if we accept that the text is not even attempting to be a sutra.

Most of the remaining problems occur in the portion of the text which surrounds the quoted verses - what Nattier calls "the frame". She seeks to show that it is plausible for the frame to have been composed in China. For instance the presence of Avalokiteśvara: this is quite consistent with devotional Buddhism in South West, 7th century China, and his presence is less surprising if the text is a devotional text for chanting rather than the essence of the Prajñāpāramita tradition. The presence of the mantra also marks out the Heart Sutra as different. Nattier points out that the mantra is present in at least three other Chinese texts, and the epithets of the mantra also exist independently. (p.177). The point being that the presence of a mantra need not rule out a Chinese origin.

I think this is the only place where Nattier misses a trick. Donald Lopez, for instance, has commented on the lack of coherence between the mantra and the text.

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

The mantra is not of a piece with the sutra, but appears to have been tacked on. Further Alex Wayman has noted that commentaries on the text lack coherence:

"The [commentators] seemed to be experiencing some difficulty in exposition, as though they were not writing through having inherited a tradition about the scripture going back to its original composition" - Secret of the Heart Sutra p.136

These observations only strengthen the impression of a text appearing suddenly without a history of exegesis to be referred to. But, back to Nattier's article...

Another feature which supports the idea that the frame was written in China relates to phrases such as satyam amithyavāt which Conze translates as: "[It is] true. For what could go wrong". This is clearly an awkward phrase both in Sanskrit and in English translation. The Chinese—chen shih pu hsü  [i.e. Zhēnshí bùxū 真實不虛] or "genuine, not vain"—however is "entirely natural in Chinese". As Nattier says:

"The Heart Sutra thus diverges from anticipated Sanskrit usage, offering instead a precise replication of the word order of the Chinese" (p.178)

The final mystery is the existence of the two versions of the sutra. The evidence is good that the short version was the one which was most prominent version in China. All of the extant Chinese commentaries are based on the Hsüan-tsang's (or Xuanzang) 'translation' of the short version. If we accept the idea that the sutra was back-translated into Sanskrit after being composed in China, then the long version makes sense in the face of Indian criteria for authenticity - which include the appropriate opening, the presence of the Buddha, and the audience reaction to the discourse. The long version supplies all these features that are missing from the short version. From the Indian point of view the short version is not a sutra at all - which fits with the idea that it was not intended to be one.

On purely philological grounds it seems that the Heart Sutra was composed in China around the lines quoted from the Chinese version of the Pañcaviṃśatisāhasrikā Prajñāpāramita. Internal textual evidence supports this conclusion, as do historical considerations. In short everything points in the direction of the Heart Sutra being a Chinese liturgical text which only became a sutra on being back translated into Sanskrit, probably in India in the late 7th century. What is more, the most problematic features of the sutra become comprehensible if we accept this view.

Nattier spends several pages exploring the role of Hsüan-tsang in the popularisation of the text: it was certainly a favourite of the pilgrim/translator, and he did know it before he left on his trip to India. It seems likely, though it is not proven, that it was Hsüan-tsung himself who introduced the text to India and translated it into Sanskrit when he discovered that the Indians lacked it. We know that exactly this happened in the case of another Chinese apocryphal text, the Awakening of Faith in the Mahāyāna, which he translated into Sanskrit during his stay at Nālandā.

To those amused, or perhaps alarmed, by this apparent forgery, Nattier points out that "it is now becoming clear that the Chinese were avid producers as well as consumers of Buddhist sutras... and indeed evidence is accumulating for an important backwash of Chinese Buddhist influence into Central Asia" (p.181). Though the Heart Sutra may be an apocryphal text:

...this in no way undermines the value that the text has held for Buddhist practitioners. "Whatever is conducive to liberation" - so the Buddha is said to have told his followers - "that is my teaching" (p.199).

Nattier's article is a fantastic example of the kind of careful and exacting scholarship which marks her out. The conclusions are monumental, and yet eminently accessible. I highly recommend reading the article. Her work deserves a wider audience, and her conclusions should be informing our understanding of Buddhist history, both social and textual. One thing is clear from this, and her other publications, we Buddhists cannot afford to be fundamentalists when it comes to texts!

~~oOo~~

[lightly edited for style, 8 Feb 2025]