Showing posts with label Mantra. Show all posts
Showing posts with label Mantra. Show all posts

06 September 2013

Heart Sutra Mantra Epithets

The material in this essay has been rewritten, peer-reviewed, and published as
Attwood, Jayarava. (2017). ‘Epithets of the Mantra’ in the Heart Sutra. Journal of the Oxford Centre for Buddhist Studies, 12, 26–57.

Karaṇḍamudrā dhāraṇī
My last essay mined the footnotes of Jan Nattier's excellent article 1992 on the provenance of the Heart Sutra. Her article is a remarkable piece of scholarship and repays close study. The footnotes are no less interesting and in this essay I want to expand on a single long footnote: 54a (211-213). The 'a' is added because this information was included just as the article was going to press and the note, amounting to two full pages, had to be squeezed in, sans any Chinese characters (which in any case were hand written on a separate page at the end of the article).

The subject of this note is the epithets of the mantra. The section we're interested in reads:
tasmāj jñātavyam prajñāpāramitā mahāmantro mahāvidyāmantro ‘nuttaramantro ‘samasama-mantraḥ
Therefore, it should be known that the perfection of wisdom is a great mantra, a mantra of great insight, an unexcelled mantra, an unequalled mantra
For Conze these are epithets of the Buddha applied to a mantra as a way of conveying the magical power of the mantra: "The prañāpāramitā... is here envisaged as a spell" (1973: 101-104). The epithets in question are those from the familiar itipi so gathā that Triratna Buddhist Community members will know as the Buddha Vandana. In Pāli:
iti pi so bhagavā arahaṃ sammāsambuddho vijjācarana sampanno sugato lokavidū anuttaro purisadammasārathi satthā devamanussānaṃ buddho bhagavā ti
As we can see by simple comparison Conze is stretching things somewhat with this comparison. Of the Heart Sutra terms only anuttara 'unexcelled' has an actual parallel and it is a rather common superlative applied to any and all Buddhist ideals.

Nattier cites two letters sent to her by Nobuyoshi Yamabe. Yamabe San completed a PhD at Yale in 1999 and is the author of several books on Buddhism. Yamabe identified a number passages in Chinese which closely parallel the Heart Sutra epithets. Nattier adds two extra passages to those identified by Yamabe. We'll begin with the passage found in the Aṣṭasāhasrikā Prajñāpāramitā Sūtra (Aṣṭa). This text is the basis for the Pañcaviṃśatisāhasrika Prajñāpāramitā Sūtra (Pañcaviṃśati) and is therefore of some interest. Also the existence of a clear Sanskrit text allows us some insight into another matter.

The Chinese Heart Sutra (T 8.251) reads:
故知般若波羅蜜多,是大神咒 ,是大明咒,是無上咒,是無等等咒, 
Gùzhī bōrěbōluómìduō, shì dàshén zhòu, shì dàmíng zhòu, shì wúshàng zhòu, shì wúděngděng zhòu, 
Therefore know the perfection of wisdom, the great magical mantra, it is the great knowledge mantra, unsurpassed mantra, an unequalled mantra,
般若波羅蜜多 bōrěbōluómìduō is a transliteration of prajñāpāramitā. A short digression here. The Middle Chinese pronunciation of 般若波羅蜜多, reconstructed from rhymes, but lacking information on tones, would have been ban ya ba ra mil da. As we will see shortly the Aṣṭa is written in Classical Sanskrit. However the transliteration banya suggests a spelling more like Pāli paññā than Sanskrit prajñā. Baum and Glass's interim Gāndhārī Dictionary record several spellings of prajñā from the Gāndhārī Dhammapada: praña, prañaï, prañaya. The transliteration of prajñā is quite standard across genres. I can find only one variant: 鉢若 bōruò, Middle Chinese balya. It seems the initial syllable was not heard or seen as a conjunct /pra/ by early Chinese translators even when we can be reasonably sure the text used it.

shén is a term from Daoism that is sometimes used to translate Sanskrit ṛddhi 'supernatural power' or even deva. Generally is means 'supernatural, divine' or 'magical'. It's missing from all of the Sanskrit versions of the text, which opens the possibility that it was added to the Chinese after the Sanskrit text was created.

Yamabe identified a counterpart from the Chinese Aṣṭa, early 5th century CE, translation by Kumārajīva (T 8.227 843b25-27) reads:

Bōrěbōluómì shì dà míngzhòu,
bōrěbōluómì shì wúshàng zhòu,
bōrěbōluómì shì wúděngděng zhòu.

Prajñāpāramitā is a great vidyā (明呪),
Prajñāpāramitā is an unsurpassed vidyā (呪),
Prajñāpāramitā is an unequalled vidyā (呪).
As in the last essay, one doesn't need to know Chinese to see that these are the identical characters, except that the anomalous 是大神咒 shì dà shén zhòu is absent. If one knows that Chinese languages, like English, are subject-verb-object languages, one can even guess that 是 means 'is'. Also note that in the Aṣṭa the last syllable of prajñāpāramitā is left off, which is typical. The reason for translating 明呪 míngzhòu and 呪 zhòu as vidyā becomes apparent when we look at the Sanskrit text below. Note also the substitution of 呪 zhòu for zhòu, on which I will say more below.

The Sanskrit version of this text has been edited by Vaidya (p.36, line 30-p.37 line 7 = Conze 's translation p.108-109). This is one of the best attested texts of Buddhist Sanskrit literature. I have seen and handled the beautiful Cambridge manuscript (Add 1643) dated to 1015 CE, which forms the basis of the critical edition. It's written in Classical Sanskrit with just a few Prakritisms. The edition by Vaidya has been digitised, from which I take the following (placing each sentence on a new line to facilitate reading):
mahāvidyeyaṁ kauśika yad uta prajñāpāramitā|
apramāṇeyaṁ kauśika vidyā yad uta prajñāpāramitā|
aparimāṇeyaṁ kauśika vidyā yad uta prajñāpāramitā|
anuttareyaṁ kauśika vidyā yad uta prajñāpāramitā|
asameyaṁ kauśika vidyā yad uta prajñāpāramitā|
asamasameyaṁ kauśika [vidyā] yad uta prajñāpāramitā|

O Kauśika, the perfection of wisdom is certainly a great spell.
O Kauśika, the perfection of wisdom is certainly an immeasurable spell.
O Kauśika, the perfection of wisdom is certainly a measureless spell.
O Kauśika, the perfection of wisdom is certainly an unsurpassed spell.
O Kauśika, the perfection of wisdom is certainly an unequalled spell.
O Kauśika, the perfection of wisdom is certainly a peerless spell.
Kauśika is one of the epithets of the Vedic God Indra, usually called Śakra (Pāli Sakka) in Buddhist texts, who plays an important role in early Buddhism and is one of the main interlocutors of the Aṣṭa. The context here is the Perfection of Wisdom per se. Both apramāṇa and aparimāṇa mean 'not-measured or measureless'. Similarly both asama and asamasama mean 'without equal'. I translate vidyā here as 'spell', as the context shows that the idea is something to be spoken or chanted that has magical powers. There is an irreducible element of magical thinking in these texts that is inherent in their pre-scientific world view. It's nothing to be embarrassed about.

Note that the word in Sanskrit is vidyā throughout, and not mantra or dhāraṇī. Here we see 明呪 míngzhòu translating vidyā. Note that in the Heart Sutra epithets we get the sequence 大明咒,無上咒,無等等咒. In the context of the Heart Sutra the tendency is to see 明 as an extra character: the great  knowledge  mantra 咒. We know from the Aṣṭa passages that 明呪 means vidyā, so we ought to read 大明咒 as 'great vidyā'. And this means that  is a shorthand reference to vidyā. The character 明 is being dropped from the other epithets, not added to only one of them. 

This passage from the Aṣṭa is a slightly more elaborate version of what we find in the Heart Sutra. Now compare the parallel passage in Kumārajīva's translation of the Pañcaviṃśatisāhasrikā Prajñāpāramitā Sūtra (T 8.223).
Shì bōrěbōluómì shì dàmíngzhòu, shì wúshàng míngzhòu.
The prajñāpāramitā is a great vidyā, an unsurpassed vidyā.
Though Nattier notes that the relevant chapter is missing from earlier editions of the Sanskrit, it is found twice in the more recent Sanskrit edition produced by Takayasu Kimura (vols 2&3). Kimura has edited the earlier Sanskrit text of Dutt and referenced both the Chinese and Tibetan translations to produce a new Sanskrit edition based on the same late Sanskrit manuscripts used by Dutt. So we cannot be entirely sure that Kimura has not, once again, back translated an existing Chinese passage into Sanskrit to fill a perceived void. In any case the two passages are:
mahāvidyaiṣā kauśika yad uta prajñāpāramitā, anuttaraiṣā kauśika vidyā yad uta prajñāpāramitā. (Vol. 2-3:55)
evam ukte bhagavān śakraṃ devānām indram etad avocat: evam etat kauśikaivam etat, mahāvidyeyaṃ kauśika yad uta prajñāpāramitā, anuttareyaṃ kauśika vidyā yad uta prajñāpāramitā, asamasameyaṃ kauśika vidyā yad uta prajñāpāramitā.
(Vol. 2-3:70)
The second of these more closely matches what we find in the Sanskrit Heart Sutra with three epithets: mahāvidyā, anuttara vidyā, and asamasama vidyā. It also alerts us to a further occurrence in Kumārajīva's Pañcaviṃśati (T. 223) at p. 286b28 (unnoticed by Yamabe or Nattier)
Bōrěbōluómì shì dàmíngzhòu, wúshàng míngzhòu, wúděngděng míngzhòu.
The prajñāpāramitā is a great vidyā, an unsurpassed vidyā, an unequalled vidyā.
Again we see from comparing Chinese with Sanskrit, that 明呪 translates vidyā and here it is not abbreviated to 呪 but spelt out each time. If the core part of the Heart Sutra comes from the earlier passage of the Pañcaviṃśati then this passage suggests that the epithets were also borrowed, probably from this passage. Except that it is clear from the context that these epithets are not describing the mantra, but the perfection of wisdom itself. We associate the epithets with the mantra because the word mantra appears in the Sanskrit Heart Sutra. The word is used just twice in the Aṣṭa and not at all in the Pañcaviṃśati (suggesting perhaps that the Aṣṭa occurrences are interpolations).

Vidyā has a number of connotations. Clearly both Aṣṭa and Pañcaviṃśati are applying the word to the prajñāpāramitā per se, not to the mantra (as we typically read the Heart Sutra). Vidyā derives from the verbal root √vid 'to know, to discover' (cognate with 'wise, wisdom' etc). Sometimes you'll see vidyā translated as 'science' but the whole context is pre-scientific so this is anachronistic. No body of knowledge before ca. 1700 fits today's definition of science, which is not to say that there was no valid knowledge, only that it could not be considered scientific until the scientific method ha been invented during the European Enlightenment. Vidyā means knowledge in a particular field: knowledge of the Vedas, knowledge of political governance etc. Knowledge cultivated through learning and experience, rather than divinely inspired knowledge or insight. It also have a magical connotation. Knowledge in the sense of vidyā bestows control over the subject studied, when one thoroughly knows a subject one is said to have "mastered" it. Ironically we are stuck using 'wisdom' for prajñā, which means (and is cognate with) knowledge; and 'knowledge' for vidyā, which is cognate with wisdom.

Although vidyā later becomes, at times, almost synonymous with mantra, at the time the Aṣṭa was composed, and probably even the Pañcaviṃśati, Indian Buddhists still probably thought of mantras as the spells mumbled by Brahmins (for money) at ceremonies. The Pāli texts contain a few passages making it clear that the chanting of mantras is un-Buddhist (DN 1 [i.9]; SN 7.8, SN 28.10, Sn 480). By contrast the chanting of parittās, or protective texts, was already established as a popular Buddhist practice in the Milindapañha, which predates the Aṣṭa.

The parittā practice may well be connected to the idea of the saccakiriyā (Skt satyakriyā) or 'truth act'. This practice, attested in for example the Pāli Aṅgulimālā Sutta, insists that plainly and clearly stating a truth can alter reality. Aṅgulimālā, for example, uses a saccakiriya to ease the pain of a women and baby experiencing a difficult childbirth. Many other examples are found in Pāli. Some scholars have attempted to link the practice to similar ideas in Vedic culture. There is even a suggestion that some aspects of the power of truth are Indo-European. Holding a red-hot axe-head is a test of truth in both Vedic and Celtic literature for example. It may be that by chanting a sacred text aloud, sacred texts being true by definition, that one might avoid calamity or avert disaster. As mentioned last week, this was how Xuánzàng used the Heart Sutra.

Nattier cites the example of the word for mantra as an example of a back translation. Her thesis is that the order of textual production was like this:
  1. Sanskrit Pañcaviṃśati
  2. Chinese translation Pañcaviṃśati
  3. Chinese Heart Sutra - short text
  4. Sanskrit Heart Sutra - short text
  5. Sanskrit Heart Sutra - long text
  6. Chinese Heart Sutra - long text
We can see that Nattier's theory explains the changes that occur in the word vidyā. In this case the Sanskrit Pañcaviṃśati (itself based on Aṣṭa) uses the word vidyā. Kumārajīva translated this as 明呪 míngzhòu, the usual translation of vidyā. The Heart Sutra first uses 明咒 míngzhòu then abbreviates to zhòu; where zhòu is a homonym for zhòu meaning dhāraṇī (or mantra). This is then back translated as Sanskrit mantra. The change from 呪 zhòu to 咒 zhòu might have occurred for any number of reasons, not excluding simple error based on similarities of sound and graphic form.

It is interesting to note here that T 250 (attributed to Kumārajīva) has 明呪 míngzhòu in each of the epithets, which conforms to the general pattern of Kumārajīva's translations noted above. Nattier's conclusion regarding T 250 is "[it] was based not directly on his version of the Large Sūtra, but on citations from the sūtra contained in the Ta chih-tu lun*" (187).
* i.e. T 25.1509 大智度論 Dàzhìdù lùn (Mahāprajñāpāramitāśastra) Attributed to Nāgārjuna and translated by Kumārajīva.
Dàzhìdù lùn itself shows signs of partly Chinese authorship: "Some of the most notable evidence provided by Chou is that the Dazhidu lun’s commentary on the Mahaprajñaparamita Sutra follows Chinese word order rather than Indian..." (McBride 332-333)

At the time the Heart was composed in China we might expect the key term to be dhāraṇī, since the mid seventh century date proposed by Nattier slightly predates the arrival of Tantra in China, while dhāraṇī texts, such as the Karaṇḍamudra Dhāraṇī depicted above, were and to some extent still are, a central aspect of Chinese Buddhism. The first Sanskrit version of the Heart Sutra was produced in India, probably in the late seventh or early eighth century at a time when Tantra was in full swing. These dates coincide for example with Stephen Hodges' proposed dates for the composition of the Sarvatathagata-tattvasaṃgraha. In such an environment mantra might have be the natural translation of 咒. Hence find a mantra where we expect not to and, according to my own definitions, where we might expect to find a dhāraṇī.

This is further evidence that the Heart Sutra is synthetic, which is to say it was constructed in China from a variety of sources, probably by a devotee of Avalokiteśvara in the 7th century. Now on the basis of a comparison with the Sanskrit sources, there is an argument for revising this portion of the Sanskrit text:
tasmāj jñātavyam prajñāpāramitā mahāmantro mahāvidyāmantro ‘nuttaramantro ‘samasama-mantraḥ,
tasmāj jñātavyam prajñāpāramitā mahāvidyā anuttaravidyā asamasamavidyā.
It should be understood that the perfection of wisdom is great knowledge, supreme knowledge, peerless knowledge.



  • Conze, Edward (1973). The Perfection of Wisdom in 8000 Lines and its Verse Summary. San Francisco: City Lights.
  • McBride, Richard D, II. (2004) 'Is there really "Esoteric" Buddhism?'  Journal of the International Association of Buddhist Studies. 27(2): 329-356.
    • Vaidya, P. L. (1960) Aṣṭasāhasrikā Prajñāpāramitā. The Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning. Online:

    30 August 2013

    Heart Sutra Mantra

    My calligraphy of Heart Sutra
    Siddhaṃ script
    The Heart Sutra is a synthetic text composed in China from three main elements:
    1. Extracts from Kumārajīva's translation of the Pañcaviṃśatisāhasrikā Prajñāpāramitā Sūtra (T 8.223; ca. 5th century).
    2. Elements drawn from the devotional cult of Avalokiteśvara (觀自在 Guānzìzài). 
    3. The cult of dhāraṇī chanting, and a mantra probably drawn from existing Chinese texts. 
    The first element is quite well covered in the literature, especially as Jan Nattier (1992) focusses on this part of the text in her reconstruction of its provenance. My next essay will address a lesser known aspect of this issue which is buried in Nattier's footnotes. The second element deserves a little more attention, but is covered briefly in Nattier (174-5). This essay will largely focus on the third element. 
      In her long essay on the origins of the Heart Sutra, Jan Nattier notes (footnote 52 & 53) that two other scholars have found mantras similar to the Heart Sutra mantra in other places in the Chinese Tripiṭaka. One of these references is particularly significant as it seems to pre-date the composition of the Heart Sutra itself. This essay will present the Chinese source texts for these mantras. Before dealing with these mantras, we need to pay some attention to the dhāraṇī cult itself, and try to establish some terminological boundaries. 

      The cult of dhāraṇī chanting is sometimes placed in the context of Tantric Buddhism, but I think this is a mistake. It is true that mantras are a feature of Tantric Buddhism. However, as Ryūichi Abé. has shown, Tantric Buddhism requires certain elements to be present in order to be Tantric. In The Weaving of Mantra he emphasises the abhiṣeka or initiation in particular because the abhiṣeka is the ür-ritual which underpins all of Tantric Buddhist practice. In Japan, prior to the arrival of Kūkai and Saichō with genuine Tantric Buddhism, some Tantric elements were present: images, dhāraṇī and mantra, and even texts such as the Mahāvairocana Abhisaṃbodhi Tantra. However, in the absence of the Tantric paradigm and organising principles, these elements did not add up to Tantric Buddhism.

      Abé is trying to revise the history of Japanese Buddhism, but he has enunciated an important hermeneutic for discussing the presence or absence of a mode of Buddhist thought. For example: if a person bows before a Buddha statue, burns incense, and chants oṃ maṇipadme hūṃ, but has no knowledge of why Buddhists do such things, does this make them a Buddhist? These are simply decontextualised actions with no intentional underpinning. They are Buddhist in externals only. A similar argument is simmering away with respect to the Jon Kabat Zinn inspired mindfulness treatments. Does the teaching of mindfulness amount to teaching Buddhism, or does it lack key elements, such as "going for refuge", that render the teaching non-Buddhist? Some Buddhists who teach mindfulness argue that they are teaching Buddhism when they teaching mindfulness. Others argue that the lack of context for the practice, particularly the absence of Buddhist metaphysics, means this is a beneficial secular practice that does not conduce to liberation.

      In any case, the point is that although dhāraṇīs were incorporated into Tantric Buddhism, there is nothing in the dhāraṇī sūtras or the chapters inserted into larger texts such as the Saddharma-puṇḍarīka or Survabhāṣotama, to indicate a Tantric context. The first hints of Tantra associated with a mantra seem to be found in the Kāraṇḍavyūha Sūtra according to Alexander Studholm's study of that text, which includes an account of something like an initiation, though it still lacks the central features of the abhiṣeka ritual. The point of chanting dhāraṇīs seems largely to have been protection from malign forces or entities. And thus they have much more in common with the Theravāda practice of parittā chanting than with Tantric practice (at least with respect to the Tantric Buddhism practised by Kūkai). It's not until they are incorporated into rituals centred on the abhiṣeka, that they become Tantric. This criteria is common to other elements that were incorporated, not least the elements from Vedic ritual. No one, to my knowledge, argues that Vedic fire rituals were "proto-Tantric". 

      Nattier points to the opinion of Fukui Fumimasa (1981. Source text is in Japanese) that the name of the Heart Sutra in Chinese is 心經 Xīnjīng, literally "heart sūtra", but that 心 xīn (heart) here connotes dhāraṇī rather than 'pith' and that the text might well be a chanting text, i.e. a dhāraṇī text. We know from Xuánzàng's record of his journey to India that he used the text as a protective measure against unseen malevolent spirits. The title of the short text in Sanskrit, Prajñāpāramitāhṛdaya, does not include the word sūtra, and it seems likely that the original, short text was not considered as a sūtra. That transition probably happened in India when the traditional elements of a sūtra, such as the beginning evaṃ maya śrutaṃ... and the appreciation at the end were added. Another reading of 心 is "gist" with the idea that rather than Heart Sutra, the meaning is Gist Text, with the text representing the gist of Prajñāpāramitā.

      The dhāraṇīs of the pre-tantric Mahāyāna texts are often radically different in form from the mantras of later Tantric Buddhism. Of course there is a huge amount of variation and cast-iron definitions are difficult to construct.

      Defining Mantra and Dhāraṇī

      Tantric mantras have a number of structural features in common: a beginning (usually oṃ); a name or function; and a final seed-syllable. 

      Typically Tantric mantras begin with oṃ (not auṃ) which served to mark what follows as a mantra. However in the earliest fully-fledged Tantra, the Mahāvairocana Abhisaṃbodhi Tantra, the mantras all begin namas samanta-buddhānāṃ or namas samanta-vajrānāṃ. The ending -ānām indicated the genitive plural case (of the Buddhas). However, in Prakrits (including Pāli) the dative case (to or for the Buddhas) endings began to be replaced by the genitive case endings. Here ending is the usual genitive, but the sense is dative and the words mean "homage to all Buddhas/vajras".

      What follows oṃ can be the name of a deity (oṃ amideva hrīḥ, oṃ vajrapāṇi hūṃ, oṃ vagiśvara muṃ) or relate to a function in the ritual, especially purification with the śūnyatā mantra or the Vajrasattva mantra. Names of deities are sometimes in the dative case, or in a kind of faux dative created by the addition of -ye to the end of the word: oṃ muni muni mahāmuni śākyamuniye svāhā. The correct dative of śākyamuni is śākyamunaye (final i is replaced by aye)

      Tantric mantras typically end with a seed-syllable (bījākṣara) related to the deity or with svāhā. Sometimes the seed-syllable is specific to the deity, or to the "family" they belong to. Mantras of the vajra family typically end in hūṃ, while the padma family often end in hrīḥ. At other times it seems unconnected to other considerations. For example  oṃ maṇīpadme hūṃ is a padma family mantra. Some mantras incorporate dhāraṇī style features into them which would include the Heart Sūtra mantra and the Tārā mantra (oṃ tāre tuttāre ture svāhā). The variety of mantras is partly due to their being a number of systems existing in parallel. 

      Dhāraṇī by contrast seldom begin with oṃ and almost never end in a seed-syllable. They almost always end with svāhā. The word svāhā is the Vedic equivalent to the Hebrew amen. It is used in the Yajurveda to solemnise offerings: one makes an offering of rice mixed with ghee to the fire while chanting, for example "agnaye svāhā" or 'For Agni, amen' (Taittirīra Saṃhitā The content of the dhāraṇī is a string of words or sounds which seldom reference names of deities, and frequently include nonsense words such as hilli, huru often with repetition and ringing the changes of the first syllable: hilli hilli milli milli. There is a tendency to use words ending in -e. Various theories have been proposed to explain this phenomenon, but my opinion is that the -e ending is a Prakrit masculine nominative singular. This probably also applies to the well known oṃ maṇipadme hūṃ mantra. A feature of dhāraṇī, then, is the use of Prakrit or Buddhist Hybrid Sanskrit. 

      Typical dhāraṇīs from the Saddharmapuṇḍarikā Sūtra:
      anye manye mane mamane citte carite same samitā viśānte mukte muktatame same aviṣame samasame jaye kṣaye akṣaye akṣiṇe śānte samite dhāraṇi ālokabhāṣe pratyavekṣaṇi nidhiru abhyantaraniviṣṭe abhyantarapāriśuddhimutkule araḍe paraḍe sukāṅkṣi asamasame buddhavilokite dharmaparīkṣite saṁghanirghoṣaṇi nirghoṇi bhayābhayaviśodhani mantre mantrākṣayate rute rutakauśalye akṣaye akṣayavanatāye vakkule valoḍra amanyanatāye svāhā.
      iti me iti me iti me iti me iti me; nime nime nime nime nime; ruhe ruhe ruhe ruhe ruhe| stuhe stuhe stuhe stuhe stuhe svāhā.
      There is a world of difference between these two dhāraṇī and most Tantric mantras.

      The Heart Sutra Mantra

      The Heart Sutra mantra is clearly referred to as a mantra by the text. But it has more features in common with dhāraṇī in form and content. It's lacks the opening oṃ for example, though some traditions have simply added one. The repetition and play of sounds in gate gate pāragate pārasaṃgate is typical of dhāraṇī. Why then does the text refer to this as a mantra? I will look more closely at this issue in the next essay.

      Meanwhile let us compare the Heart Sutra mantra with the three mantra/dhāraṇī listed in Nattier's footnotes as being similar. Of these T 12.387  大方等無想經  Dàfāngděng wúxiǎng jīng (Mahāmegha Sūtra), identified by Fukui (1981), is important because it was translated in the early fifth century, two centuries before the proposed date for the composition of the Heart Sutra.

      Mantras and dhāraṇīs are typically not translated by the Chinese, but the sounds are represented using characters for their pronunciation.  Unfortunately it can be very difficult to reconstruct the Sanskrit from a Chinese transliteration. For example the character 卑 bēi has been used to transliterate the Sanskrit syllables pra, pre, pe, pi, vi, and vai. Note that I'm using Pinyin Romanisation in these posts, which often does not reflect pronunciation at the time the texts were composed. The language of the day is referred to as Middle-Chinese (MC). Where relevant and possible I will indicate the MC pronunciation 

      The dhāraṇī in question is:
      竭帝 波利竭帝 僧竭帝 波羅僧竭帝波羅卑羅延坻 
      三波羅卑羅延坻 婆羅 婆羅 波沙羅 波娑羅 摩文闍 摩文闍 
      遮羅帝 遮羅坻 波遮羅坻 波遮羅坻 三波羅遮羅坻
      比提 嘻利 嘻梨 薩隷醯 薩隷醯 富嚧 富嚧 莎呵
      jiédì bōlìjiédì sēngjiédì bōluósēngjiédì bōluóbēiluóyánchí
      sānbōluóbēiluóyánchí póluó póluó bōshāluó bōsuōluó mówéndū mówéndū
      zhēluódì zhēluóchí bōzhēluóchí bōzhēluóchí sānbōluózhēluóchí
      bǐtí xīlì xīlí sàlìxī sàlìxī fùlú fùlú shā hē
      Fortunately for us some markers are clear at the beginning.  The Mantra in the Heart Sutra in Chinese is:
      揭帝 揭帝 般羅揭帝 般羅僧揭帝 菩提 僧 莎訶
      jiēdì jiēdì bānluójiēdì bānluósēngjiēdì pútí sēng shāhē
      One does not need to understand the characters to see that many of them graphically match up between the two mantras above, especially at the beginning. The opening characters of both are very similar. Both 竭帝 jiédì and  帝  jiēdì are transliterations of Sanskrit gate (the difference in pronunciation is a matter of tone). MC pronunciation in both cases was gal (with a hard g sound).

      The first words in the Mahāmegha Sutra mantra are: jiédì bōlìjiédì sēngjiédì bōluósēngjiédì which most likely represent Sanskrit: gate parigate saṃgate paragate. The Mahāmegha mantra ends 莎呵 shāhē; the Heart Sutra has 莎訶 shāhē; both represent svāhā. Note the graphic similarity of 呵 and 訶 which have the same pronunciation, he, in MC.

      A little note here that the mantra in Xuánzàng's version of the Heart Sutra (T 8.251) has an extra out-of-place character, 僧 sēng, between bodhi (菩提 pútíand svāhā (莎訶 shāhē). Even though this is probably the oldest version of the text, it is not without problems! 

      A similar dhāraṇi is also found in T. 21.1353 東方最勝燈王陀羅尼經 Dōngfāng zuìshèng dēngwáng tuóluóní jīng (First-radiance Knowledge King Sūtra = Sanskrit Agrapradīpadhārāṇīvidyarāja-sūtra). As in T 12.387 the dhāraṇī shares opening elements with the Heart Sutra mantra using the same transliterating characters.
      阿  竭帝 波羅竭帝 波羅僧竭帝     
      a    jiédì    bōluójiédì  bōluósēngjiédì   
      a gate paragate parasaṃgate
      Here the character 阿 is often used for the Sanskrit short 'a' vowel and thus may reference the idea of the perfection of wisdom  in one letter, or more precisely the fact that all dharmas are empty of self existence (sarvadharmāḥ svabhāvaśūnyatāḥ) because they are unarisen (anutpanna). See also The Essence of All Mantras; and Sound, Word, Reality.

      The gate gate mantra itself, with the same transliteration, is found in T 18.901 陀羅尼集經 Tuóluóní jí jīng (Dhāraṇī Collection Sūtra). This was translated ca. 653 CE which is around the same time that Nattier proposes for the composition of the Heart Sutra. Note also that it is a collection of dhāraṇī (陀羅尼 Tuóluóní) rather than mantra. The presence of a dhāraṇī in a collection is not conclusive evidence that it existed detached from the Heart Sutra before its composition, but it at least shows that dhāraṇīs can be detachable. It's quite possible that similar examples may turn up with further examination. 

      It seems that, not only is the core of the Heart Sutra an extract, but the "mantra" might also be an extract from a dhāraṇī. It might be thought that the fact that the Heart Sutra is a mash-up of bits from other texts invalidates the text. However the composition method closely resembles many Pāli texts which are clearly constructed from pre-existing elements that can be found scattered around the Canon. Far from being unusual, the Heart Sutra is following standard Buddhist procedure. Even the subsequent addition of a proper sūtra introduction is in keeping with general Buddhist practice. 

      In my discussion of cladistic methods applied to studying manuscripts, I argued that it would help to iron out biases. Another bias that Buddhist Studies faces is the prejudice in favour of texts with Indian "originals". In my essay Which Mahāyāna Texts? I outlined an observation made in another publication by Jan Nattier about which Mahāyāna texts are prominent in the West. The existence of a Sanskrit manuscript is one of the influential factors likely to bring a Mahāyāna text to prominence. The fact is that the Heart Sutra is broadly accepted as a genuine masterpiece of Buddhist thought. Commentaries from across the spectrum of Buddhist schools adopt the Heart Sutra as an epitome of their thought. Is a text any less authentic because it was not composed in India? It is true that Buddhists believed that the text was of Indian origin and that was an element in popularising it. Now that we know differently will Buddhists have to abandon this text? I think there is no question of abandoning the text, but the necessary adjustments might be quite difficult. One sign of this is the rejection of the Chinese origin thesis by Red Pine in the introduction to his translation and commentary on the Heart Sutra. Though his reasoning is spurious, it is none-the-less interesting to see how difficult Buddhists find it to absorb information like this.

      One of the reasons for writing about Nattier's work is that it has yet to penetrate to the heart of popular imagination and the discussion about textual origins is in its infancy. Such writing raises questions for Buddhists. If we take scholarship seriously, then we are forced to examine our own beliefs and sometimes to admit that our beliefs are based on false assumptions such as authenticity being related to India. 


      A further note 26 sept 2013.
      The Tibetan canonical versions of the Heart Sutra both include tadyathā in the mantra itself. I've looked at this generally in Tadyathā in the Heart Sūtra - the inclusion of tadyathā 'like this' in the mantra is like actors speaking stage directions out loud. One of the versions also interpolates oṃ into the mantra as do some of the Nepalese manuscripts. 


      I've already written about the mantra of this text a couple of times:

      All Chinese texts from CBETA.

      • Abé, Ryūichi (1999). The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse. New York: Columbia University Press.
      • Fukui Fumimasa (1981) Hannya shingyô no rekishiteki kenkyû. [= Historical studies of the Buddhist scripture Prajñaparamita-hrdaya or Heart Sutra.] Tōkyō: Shunjūsha. 
      • Studholme, Alexander (2002). The origins of oṃ manipadme hūṃ : a study of the Kāraṇḍavyūha Sūtra. Albany: State university of New York Press.

      29 June 2012

      Canonical Sources for the Vajrasattva Mantra

      I've mentioned that Maitiu O'Ceileachair and I have identified the earliest textual occurrence of the Hundred Syllable Vajrasattva Mantra in the Chinese Tripiṭika. Circumstances have meant that Maitiu and I have not been able to write up our notes formally. I know there is considerable interest in this mantra, and the Vajrasattva Mantra continues to be the most popular page on my mantra website. So I thought I would write up some of the basic stuff that we've found, along with transcriptions of the mantra from various Canonical sources. This blog post represents our collaborative effort, but credit for all the observations on the Chinese goes to Maitiu.

      The earliest occurrence in the Chinese Canon, which is really the only candidate for the earliest literary use of the mantra, since only the Chinese dated their texts, is in T.866, a collection of mantras related to the Sarvatathāgata-tattvasaṃgraha (STTS). T.866 was translated into Chinese by Vajrabodhi (ca 671-741) in 723 CE. Stephen Hodge (2003) says that Vajrabodhi acquired his STTS manuscript circa 700 CE, so it had to have been composed before that date.

      Two Sanskrit manuscripts of STTS are extant, though both are relatively recent copies. One has been published in facsimile edition (Candra & Snellgrove), and another forms the basis of a critical edition by Yamada (which means that he compares his Sanskrit manuscript with other versions).  *see comments. We also looked at two versions in printed editions of the Tibetan Canon (the Peking and Derge editions) and several other Chinese versions from the Taisho Edition of the Tripiṭaka (e.g. T.873, 875, 884, 1224, 1320, 1956), including Amoghavajra's translation into Chinese (T. 873).

      The mantra occurs in the context of a brief introductory paragraph and is followed by another brief paragraph.

      Sanskrit text

      atha sarvamudrāṇāṁ sāmānyaḥ svakāyavākcittavajreṣu vajrīkaraṇavidhivistaro bhavati| yadā mudrādhiṣṭhānaṁ śithilībhavati, svayaṁ vā muktukāmo bhavati, tato'nena hṛdayena dṛḍhīkartavyā|
      oṃ vajra-satva-samayam anupālaya
      dṛḍho me bhava su-toṣyo me bhavānurakto me bhava
      su-poṣyo me bhava sarva-siddhiñ ca me prayaccha
      sarva-karmasu ca me citta-śreyaḥ kuru hūṃ
      ha ha ha ha hoḥ
      Bhagavan sarva-tathāgata-vajra mā me muṃca
      vajrī bhava mahā-samaya-sattva āḥ ||
      anenānantaryakāriṇo'pi sarvatathāgatamokṣā api saddharmapratikṣepakā api sarvaduṣkṛtakāriṇo'pi sarvatathāgatamudrāsādhakā varjasattvadṛḍhībhāvādihaiva janmanyāsu yathābhirucitāṁ sarvasiddhimuttamasiddhiṁ vajrasiddhiṁ vajrasattvasiddhiṁ vā yāvat tathāgatasiddhiṁ vā prāpsyantī-tyāha bhagavāṁ sarvatathāgatavajrasattvaḥ||

      Todaro's translation of the Sanskrit.
      (except for the mantra which is my translation)

      "Now an explanation of the rite of the strengthened of all mudrās alike in one's own body, speech and mind thunderbolt is given. When the mudrā empowerment becomes weak or when there is a desire for liberation by oneself, then one should be made firm with this mantra:
      O Vajrasattva honour the agreement!
      Reveal yourself as the vajra-being!
      Be steadfast for me!
      Be fully nourishing for me!
      Be very pleased for me!

      Be passionate for me!
      Grant me all success and attainment!
      And in all actions make my mind more lucid!
      ha ha ha ha hoḥ
      O Blessed One, vajra of all those in that state, don't abandon me!
      O great agreement-being become real!
      "The Bhagavat Vajrasattva of all the Tathāgatas said: "Notwithstanding continuous killing, the slander of all the Tathāgatas, the repudiation of the true teaching and even all evil and injury, (by this) the perfection of all the Tathāgata's mudrās from the strengthening of Vajrasattva, in the present life as you desire, and all accomplishments, the supreme accomplishment, the thunderbolt accomplishment or the accomplishment of Vajrasattva, up to the accomplishment of the Tathāgata, will be attained quickly."


      The reconstructed version of the mantra created on the basis of Sthiramati's work in Jayarava (2010) reflects the extant Sanskrit and Chinese texts of STTS quite well, with only minor differences. It may be that the Tibetans were working from a different source text.

      The mantra explicitly allows that someone who has done evil, more or less any kind of evil, will not be prevented from making progress. The Chinese version includes the five atekicca or unforgivable actions. (Giebel p.99). This represents that last phase of turning a tenet of Early Buddhism on its head, i.e. that the consequences of actions are inescapable. This role of the mantra--usually referred to as 'purifying karma'--remains central in the narratives surrounding its use in Tibetan Buddhism. The mantra seems much less prominent in Sino-Japanese Tantric Buddhism, and Vajrasattva (Japanese: 金剛薩埵 Kongosatta) plays quite a different role than in Tibet.

      The text refers to the mantra as hṛdaya, i.e, 'heart mantra' or 'heart essence'.

      Both extant Sanskrit versions spell sattva with one t, i.e. satva; which may indicate some Middle-Indic influence, although the language of this passage appears to conform to Classical Sanskrit norms.

      The main difference between this mantra text and the one reconstructed from the Tibetan in Jayarava (2010) is that Yamada has su-toṣyo me bhavānurakto me bhavasu-poṣyo me bhava; where as the Tibetan (and the Chinese texts) transpose the last two phrases:  sutoṣyo me bhava, supoṣyo me bhava, anurakto me bhava. Note that bhavānurakto is a coalescence of bhava anurakto forced by Sanskrit sandhi rules (-a a- > -ā-).

      Tibetan Versions of the Mantra

      The Tibetan texts below are transcribed as they appears in the printed text, including punctuation marks, see also note at the end of this section. The lines of woodblock prints are long, and the mantra goes over a couple of long lines in both cases--difficult to reproduce in this medium so I haven't tried.

      Derge Ed.
      ཨོཾ་བཛྲ་སཏྭ་ས་མ་ཡ། མ་ནུ་པཱ་ལ་ཡ། བཛྲ་ས་ཏྭ་ཏྭེ་ནོ་པ། ཏི་ཥྛ་དྲྀ་ཌྷོ་མེ་བྷ་བ། སུ་ཏོ་ཥྱོ་མེ་བྷ་བ། ཨ་ནུ་ར་ ཀྟོ་མེ་བྷ་བ། སུ་བོ་ཥྱོ་མེ་བྷ་བ། སརྦྦ་སི་ དྡྷི་མྨེ་པྲ་ཡཱ་ཙྪ། སརྦྦ་ཀརྨྨ་སུ་ཙ་མེ་ཙི་ཏྟཾ་ཤྲེ་ཡཿ་ཀུ་རུ་ཧཱུྂ། ཧ་ཧ་ཧ་ཧ་ཧོཿ། བྷ་ག་བཱན། སརྦྦ་ཏ་ཐཱ་ག་ཏོ། བཛྲ་མཱ་མེ་མུཉྩ་བཛྲི་བྷ་བ་མ་ཧཱ་ས་མ་ཡ་སཏྭ་ཨཿ།

      oṃ badzra satva sa ma ya| ma nu pā la ya| badzra satva tve no pa| ti ṣṭha dṛ ḍho me bha ba| su to ṣya bha ba| a nu ra kto me bha ba| su po ṣyo me bha ba| sa rbba siddhi mme pra ya tsatsha| sa rbba ka rmma su tsa me tsi ttaṃ śre yaḥ kuru hūṃ| ha ha ha ha hoḥ| bha ga vān| sa rbba ta thā ga ta| badzra mā me nu ñca ba drī bha ba ma hā sa ma ya satva aḥ

      Peking Ed.
      །ཨོཾ་བཛྲ་སཏྭ་ས་མ་ཡ། །མ་ནུ་པཱ་ལ་ཡ།བཛྲ་ས་ཏྭ་ཏྭེ་ནོ་བ།ཏི་ཥྛ་ཌི་ཌྷོ་མེ་བྷ་བ་སུ་ཏོ་ཥྱོ་མེ་བྷ་བ།སུ་བོ་ཥྱོ་མེ་བྷ་བ།ཨ་ནུ་རག་ཏོ་མེ་བྷ་བ།སརྦྦ་སིད་དྷི་མྨེ་པྲ་ཡཱ་ཙྪ་་་་་་་་་་སརྦ་ཀརྨ་སུ་ཙ་མེ།ཙི་ཏྟཾ་ཤྲེ་ཡཾ་ཀུ་རུ་ཧཱུྂ།ཧ་ཧ་ཧ་ཧ་ཧོཿ་བྷ་ག་བཱན།སརྦྦ་ཏ་ཐཱ་ག་ཏོ། །བཛྲ་མཱ་མེ་མུཾཙ་་་་་་བཛྲི་བྷ་བ་མ་ཧཱ་ས་མ་ཡ་སཏྭ་ཨཱཿ

      | oṃ badzra satva sa ma ya | | ma nu pā la ya | badzra sa tva tve no ba | ti ṣṭha ḍi ḍho me bha ba su to ṣyo me bha ba | su po ṣyo me bha ba | a nu rag to me bha ba | sa rbba sid dhi mme pra ya tsatsha ……….. sarva karma su tsa me | tsi ttaṃ śre yaṃ ku ru hūṃ | ha ha ha ha hōḥ bha ga vān | sa rbba ta thā ga to | | badzra mā me muṃtsa……badzri bha ba ma hā sa ma ya satva āḥ
      Peking ed. shows signs of being slavishly copied from a woodblock of a different size. The repeated shad | | (not to be confused with a nyis shad || ), for example in the first line 'ya | | ma' indicates that the original line ends with ya | and the new line starts with | ma. The groups of multiple tsheg indicate space filling. We've included the exact number of tsheg as in the printed text (C.f. Beginning and End Markers in Buddhist Texts).

      Tibetan regularly makes several substitutions: va > ba; ja > dza; ca > tsa. In addition rva > rbba; rma > rmma (Derge). Medial nasals are sometimes replaced by anusvāra, e.g. muñca > muṃtsa. Both have satva for sattva, but so do extant Sanskrit texts.

      General anomalies in the Tibetan versions of the mantra are discussed in Jayarava (2010). Particularly the break between samayam anupālaya becoming samaya manupālaya from an Indic original that would have written individual syllables with no word breaks: e.g. स म य म नु पा ल य sa ma ya ma nu pā la ya (See also the Chinese Siddhaṃ script preserved in T. 875 below.) This is quite simply an error, and was probably a mistake of reading rather than listening.

      Both texts incorrectly add a shad in the middle of vajrasattvatvenopatiṣṭha. The words are vajrasattvatvena upatiṣṭha with a sandhi  -a u- > -o- (See Jayarava 2010 for more on this).

      Chinese Versions of the Mantra

      Reconstructing Sanskrit from Chinese is an imprecise art and often relies on knowing what the Sanskrit 'should' say. Chinese transcriptions are not very good at representing visarga and anusvāra can go missing as well (though this might be the Buddhist Hybrid Sanskrit source material rather than the translators). Some translators indicate vowel length and some don't. Generally Amoghavajra is pretty good and many translators followed his conventions.

      The earliest occurrance is T. 866.

      T. 866
      A Summary of Recitations Taken from the Sarvatathāgatatattvasaṃgraha(sūtra)
      Translation by Vajrabodhi: 11th year of Kaiyuan (開元), Tang dynasty (CE 723) in Zisheng Monastery (資聖寺). (fasc. 2)

      [each section of the mantra is transliterated and then followed, in parentheses, by a gloss in Chinese]
      [0239a12] 唵 跋折囉 薩埵三摩耶 麼奴波邏耶。(金剛薩埵三摩耶願守護我)跋折囉薩埵 哆吠奴烏(二合)播底瑟吒(以為金剛薩埵)涅哩茶烏(二合)銘婆嚩(為堅牢我)素覩沙揄(二合)銘婆嚩(於我所歡喜)阿努囉(上 )訖覩(二合)婆銘縛 素補使榆(二合)銘婆嚩 薩婆悉地 含銘般囉野綽(授與我一 切悉地)薩婆羯磨素遮銘(及諸事業)質多失唎耶(令我安隱)句嚧吽呵 呵呵呵護(引)薄伽梵(世尊)薩婆怛他揭多(一切如來)跋折囉麼迷悶遮(願金剛莫捨離我)跋折哩婆嚩(令我為金剛三摩耶薩埵)摩訶三摩耶薩埵阿(去 引)
      oṃ vajra sattvasamaya manupālaya (vajrasattvasamaya please protect me) vajrasattvatvenopatiṣṭha (become vajrasattva) dṛḍho me bhava (be strong [for] me) sutoṣyo me bhava (be pleased with me) anurakto me bhava supoṣyo me bhava sarvasiddhi [there is an extra syllable here gam/kam] me prayaccha (bestow on me all siddhis) sarvakarmasu ca me (and all karmas) citta śreyaḥ (make me at peace) kuru hūṃ ha ha ha ha hoḥ bhagavan sarvatathāgata vajra mā me muñca (please Vajra do not abandon me) vajrībhava (make me the vajra samayasattva) mahāsamayasattva āḥ
      Vajrabodhi gives glosses for some parts of the mantra that make it clear that he understands sarvasiddhi to mean 'all the siddhis'. I suspect that the punctuation of this line is incorrect and 含 has been moved from directly behind 悉地 and that these characters should be read together as siddhiṃ or siddhaṃ. It is possible that siddhiṃ is a Middle-Indic form of siddhīn. According to Edgerton (BHSD) when the nasal of -īn is retained the vowel is shortened.

      Note that the Chinese appears to read vajra sattvasamaya manupālaya rather than vajrasattva samayam anupālaya in line with the Sanskrit mss. If this is correct then the error could have occurred on Indian soil and been transmitted to Tibet and China as it was.

      (translated by Amoghavajra 753 CE. 1st chapter only)
      唵日羅 薩 怛 三 摩 耶 麼 努 波 (引) 耶
      日羅 薩 怛 怛 尾 怒 波 底 瑟 奼
      捏 哩 濁 寐 婆 蘇 都 使 庾 寐 婆
      阿 努 囉 羯 都 寐 婆
      蘇 布 使 庾 寐 婆
      薩 悉 朕 寐 缽 囉 也 車
      薩 羯 摩 素 者 寐 質 多 室 哩 藥 矩嚕 吽
      呵呵呵呵 斛 (引)
      婆 伽 梵 薩 怛 他 櫱 多 日囉 摩 弭 悶 遮
      日哩 婆 摩 訶 三 摩 耶薩怛 噁(引)

      ǎn rì luó sà dá sān mā yē me nǔ bō (yǐn) yē
      rì luó sà dá dá wěi nù bō de sè chà
      niē li zhuó mèi pó sū dōu shǐ yǔ mèi pó
      ā nǔ luo jié dōu mèi pó
      sū bù shǐ yǔ mèi pó
      sà xī zhèn mèi bō luo yě chē
      sà jié mā sù zhě mèi zhì duō shì li yào ju lū hōng
      a a a a hú (yǐn)
      pó gā fàn sà dá tā niè duō rì luó mā mǐ mēn zhē
      rì li pó mā hē sān mā yē sà dá ě (yǐn)

      Amongst the Chinese versions are two which preserve a (corrupt) Siddhaṃ version of the mantra. We include one of these for comparison. (The Siddhaṃ is written using the CBETA Font which is not aesthetically pleasing but gives us an idea of what Chinese Siddhaṃ looks like.)

      蓮華部心念誦儀軌 [平安時代寫東寺三密藏藏本]
      A Ritual Manual of the Mental Recitation of the Lotus Section.
      Written during the Heian Period (794-1185 CE). From the Sanmitsu Collection of the Tō-ji.
      [0326a26] 金剛三昧。

      oṃ va jra sa tva sa ma ya ma nu pā la ya va jra sa tva nve no pa ti ṣṭa dṛ ho me bha va mi su tu ṣuo me bha va a nu ra kto me bh ba sup u ṣo me bha va sa rva si ddhiṃ me pra ya ccha sa rva ka rma su ca me cit ta śre ya ku ru hūṃ ha ha ha ha hoḥ bha ga vaṃ sa rva ta thā ga ta va jra mā ma muṃ ca va jrī bha va ma hā sa ma ya sa tvā āḥ
      Be aware that this mantra is corrupted and contains many introduced errors. It is provided for comparison purposes only.


      These then are principle canonical sources of the Hundred Syllable Vajrasattva Mantra in the Sanskrit, Tibetan and Chinese versions of the Sarvatathāgata-tattvasaṃgraha. Since the Chinese accurately recorded the date of their translations we can be confident that T. 866 is the earliest translated text in the Chinese Tripiṭika to contain this mantra. The differences between the various versions are relatively minor, though they suggest that even at the earliest times this text existed in several versions containing these minor differences, i.e. not all the differences are due to translations or scribal error.

      All of these canonical versions tend confirm the notion that the mantra was originally written in good Sanskrit rather than the somewhat garbled version in the received Tibetan tradition. The garbling of the mantra forms part of the discussion in Jayarava (2010), as does the tension created by received tradition vs. other forms of authority. However T. 866 suggests that at least some of the errors were present in the Indian tradition already. The fact of the difference between the canonical and received versions of the mantra highlights the conflict of sources of authority in the Buddhist tradition. Though Tantric Buddhism places great emphasis on guru to disciple transmission, which tends to outweigh textual authority; the fact that we now have much greater access to the Tripiṭika and the knowledge that the mantra has been partially garbled are difficult to ignore for Western converts unconsciously inculcated with the valorisation of textual authority.

      The Vajrasattva mantra was set free from this context in the Tibetan Tantric tradition where it performs an important role in purifying karma that might otherwise impede progress on the Buddhist path.  In its self this is a fascinating aspect of the history of ideas in Buddhism.



      大正新脩大藏經 [Taishō Revised Tripiṭaka]

      Chandra, Lokesh and Snellgrove, David L. Sarva-tathāgata-tattva-saṅgraha : facsimile reproduction of a tenth century Sanskrit manuscript from Nepal. New Delhi : Sharada Rani, 1981. Online transcription Digital Sanskrit Buddhist Canon.

      Chinese Buddhist Electronic Text Association (CBETA)

      'De-bshin-gśegs-pa thams-cad-kyi de-kho-na-ñid bsdusp-pa shes-bya-ba theg-pa chen-poḥi mdo (Sarvatathāgata-tattvasaṃgraha-nāma-mahāyāna Sūtra).' The Tibetan Tripitaka Peking Edition. (Ed. D. T. Suzuki) Tokyo: Tibetan Tripitaka Research Institute, 1956. Vol.4, p.233. (Ña 37a-b)

      ‘De-bshin-gśegs-pa thams-cad-kyi de-kho-na-ñid bsdusp-pa shes-bya-ba theg-pa chen-poḥi mdo (Sarvatathāgata-tattvasaṃgraha-nāma-mahāyāna Sūtra).’ Derge: The Sde-dge Mtshal-par Bka’-’gyur: A Facsimile Edition of the 18th Century Redaction of Si-tu Chos-kyi-’byuṅ-gnas Prepared under the Direction of H.H the 16th Rgyal-dbaṅ Karma-pa. Delhi: Delhi Karmapae Chodhey Gyalwae Sungrab Partun Khang, 1976-1979.

      Giebel, R. W. (2001) Two Esoteric Sutras. Numata.

      Hodge, Stephen. The Māhvairocana-Ambhisaṃbodhi Tantra: With Buddhaguhya's Commentary. London: RoutledgeCurzon, 2003.

      Jayarava. 'The Hundred Syllable Vajrasattva Mantra.' Western Buddhist Review, 5, Oct 2010. Online:

      Tadaro, Dale Allen. An Annotated Translation of the Tattvasamgraha (Part 1) with an Explanation of the Role of the Tattvasamgraha Lineage in the Teachings of Kukai. Doctoral dissertation Columbia University, 1985.

      Weinberger, Steven Neal. The significance of yoga tantra and the "Compendium of Principles" ("Tattvasamgraha Tantra") within tantric Buddhism in India and Tibet. Ph.D. Dissertation, University of Virginia, 2003.

      Yamada, Isshi. Sarva-tathāgata-tattva-saṅgraha nāma mahāyāna-sūtra : a critical edition based on a Sanskrit manuscript and Chinese and Tibetan translations. New Delhi : Sharada Rani, 1981 p 95.

      30 July 2010

      Some Additional Notes

      Here are two follow up notes to previous essays, one on the -e ending in mantras, another on the name Gotama; and lastly a brief note on dating the Canon.

      1. The -e Ending in Mantras.

      In March 2009 I wrote Words in Mantras That End in -e. In that essay I revisited some of the ideas about what the -e ending might signify, especially with respect to the Heart Sūtra mantra. Kern, Conze and other Sanskritists have seen it as a feminine vocative singular, though of course there are other grammatical possibilities. [1] I speculated that the -e ending was simply a masculine nominative singular, and that the mantras were composed in a region of India which employed that ending as opposed to Classical Sanskrit -as/-aḥ or Pāli -o. Recently I stumbled on an article by Signe Cohen which adds something to the picture. I know Cohen from her excellent linguistic analysis of the Upaniṣads: Text and Authority in the Older Upaniṣads. This book is particularly important for the understanding it brings of the internal struggles apparent especially in the Bṛhadāranyka Upaniṣad which has Yajurveda sages in direct competition and victorious over Ṛgveda sages. However in 2002 Cohen published a short article on the -e ending:
      On the Buddhist Hybrid Sanskrit/Middle Indic Ending "-e" as a 'Magadhism', Acta Orientalia Vol. 63 (2002), p.67-9.
      This article points out that although the -e form for the masculine nominative singular does indeed occur in the North East of India, it is in fact far more widespread. This has partly been obscured as editors of Sanskrit texts have 'corrected' the text for critical editions. Patrick Olivelle complains of the same problem with the Upaniṣads in his article:
      'Unfaithful transmitters: philological criticism and critical editions of the Upaniṣads,' in Language Texts and Society, Firenze University Press, 2005. (p. 285f) [originally published in Journal of Indian Philosophy 26, 1998: 173f.]
      Western Editors, believing Indian pandits to be incompetent, silently emended unusual spellings. However as Olivelle points out, those pandits were far from incompetent, likely to be well versed in Pāṇini, and to know a 'wrong' form when they saw one. Indian scholars tended to preserve dialectical and archaic variants, being inherently more conservative in relation to texts they saw as sacred. To the European scholar of a certain era nothing but their own objectivity was sacred. While we may not accept the pandits explanations of such variant forms (which are frequently ascribed to the peculiarities of Vedic or given mystical significance) they were at least not so over-confident as to 'correct' them. As such, modern critical and printed editions of the Upaniṣads often obscure the history of the text by removing evidence, and reproducing previously corrected texts without question.

      Cohen notes that in fact the -e form is found all over North India, and especially in Sanskrit loan words in Tocharian. She concludes:
      "The common assumption that the -e ending is an Eastern Dialect form must be seriously questioned. Rather than being a specifically Eastern Dialectical feature found sporadically in other parts of India due to eastern influence, it appears that the -e ending was widespread, especially in Buddhist Sanskrit, that it must be considered a standard form, next to the -o ending." [p. 68; my italics]
      My conjecture is that Buddhist mantras were composed in Prakrit or Buddhist Hybrid Sanskrit rather than Classical Sanskrit, and that words ending in -e in mantras are simply nominative singular forms, the gender of the words in the mantra having no relationship to the gender of the deity - and in the case of the Heart Sūtra there is no deity anyway.


      2. The Name Gautama

      In my essay What Was the Buddha's Name? I drew attention to the quirk of history which left the Buddha, a kṣatriya by tradition but possibly a non-āryan, with an ostentatiously Brahmin gotra-, or clan-name: Gautama (meaning 'descended from Gotama, the one with the most cows go'). However more than half a century ago D.D. Kosambi offered a different take on this subject in a review published in 1953:
      D.D. Kosambi. 'Brahmin Clans'. Journal of the American Oriental Society, Vol. 73, No. 4 (Oct. - Dec., 1953), pp. 202-208.
      He points to two brief Pāli passages which suggest that Gautama (Pāli: Gotama) is not the Buddha's gotra name. The first is from the Therīgāthā verses of the Buddha's maternal aunt and foster mother. She says (Th 2 162)
      Bahūnaṃ vata atthāya, māyā janayi gotamaṃ;

      Truly for the many, Māyā gave birth to Gotama
      Kosambi's point here is that the names Māyā and Gotama are on the same level - i.e. they are both first names. This is to read the text quite literally, and I'm a bit doubtful about doing that. Compare for instance the case of the Brahmin boy Uppatissa, son of Rūpasārī, better known as Sāriputta 'son of (Rūpa)sārī'.[2] However Kosambi points out that neither does the Buddha's wife become known as Gotamī in any tradition. The fact that Mahāpajāpati, his mother's sister, is called Gotamī also suggests that it is not the Buddha's clan-name since the names pass pass down patrilineally (though I think Kosambi here is thinking in terms of Brahminical social rules which required Brahmins to marry outside their gotra). Kosambi also notes that bhikkhus are sakiyaputta not gotamaputta. He does not attempt to explain why the future Buddha might be named after Vedic sages however, which still strikes me as odd.

      Kosambi's other text is the Pabbajjā Sutta [Sn 3.1] in which King Bimbisāra asks the Buddha where he is from. The Buddha replies that he comes from the country of Kosala, and:
      Ādiccā nāma gottena, sākiyā nāma jātiyā;
      Tamhā kulā pabbajitomhi, na kāme abhipatthayaṃ.

      Called Ādiccā by clan, called Sākiya by caste [jāti]
      I went forth from that family, not longing for pleasures.
      The phrase only occurs once in the canon, but elsewhere the Buddha says that the Sākiya consider rājā okkāka their ancestor [Ambaṭṭha Sutta, DN 3, PTS D i.92-3] and Pāli okkāka is Sanskrit ikṣvāku a king of the ādityā [P. ādiccā] gotra. The suggestion then is that the Buddha's name was in Sanskrit Gautama Ādityā; and Pāli Gotama Ādiccā. The Buddha is also sometimes called Āṅgirasa which according to the Dictionary of Pāli Names was a tribe which included the Gautama gotra. My reading of some of the DOPN references suggests that āṅgirasa was being used as an adjective (e.g. 'shiny like the sun') rather than a name. Against the passage above Kosambi also cites the Mahāpadāna Sutta [Dn 14, PTS ii.3]
      Ahaṃ, bhikkhave, etarahi arahaṃ sammāsambuddho gotamo gottena ahosiṃ.

      I bhikkhus, now worthy, fully awakened, was of the Gotama gotra. [3]
      This phrase occurs 3 times in the suttas, all in the Mahapadāna. Kosambi refers to this as "the first interpretation of Gotama as the Buddha's gotra name... obviously a late formation under Brahmin influence". Indeed it is so obvious that Kosambi provides no evidence for his conjecture, nor does he consider the possibility that both statements about gotra are "late formations". Contrarily we find the name Gotama being used in the last two chapters of the Sutta-nipāta which are generally considered to be the oldest layers of the Pāli Canon.

      It is still a puzzle as to why the Buddha even has a gotra name, let alone a Brahmin one (which both Gautama and Ādityā are). He was not a Brahmin. I don't think Kosambi solved the mystery, but he provided an interesting additional view point. One last observation of my own is that though the Buddha meets Brahmins from many other gotra lineages, he never seems to meet a Gautama Brahmin. This is despite the fact that the two ancestors Gotama and Bharadvāja are mentioned together in Bṛhadāranyka Upaniṣad 2.2.4, and Gautama the Buddha meets more than a dozen Brahmins from the Bhāradvāja lineage, who mostly seem to live in Kosala (see e.g. DN 3, 13, 27, 32, but throughout the nikāyas).

      18 Aug 2011
      I've been looking at Brahmins in the Canon and thinking about the Buddha's Brahmin surname. No other males with the gotra name "Gautama" are found in the Pāli Canon, though there are several women. I think the facts we have might be explained if the Buddha's mother and her sister were of the Gautama clan, and married Sudhodana who was a Śākya. Gautama in other words is actually Gautamaputra, Son of Gautamī; on the same model as Śāriputra is the Son of (his mother) Śārī.

      6 Sept. Extra note

      Snodgrass, vol.2 p. 471
      In the Garbhadhātu Maṇḍala of the Shingon school, associated with the Mahāvairocana Abhisaṃbodhi Tantra (MAT), there is a figure called Gautama (or Gotama; Japanese: Kudonsen 瞿曇仙). This is the Vedic Gautama and he attends on the god Agni. Adrian Snodgrass suggests that he is the subject of many hymns in the Ṛgveda, though this is not correct as far as I know. He is, however, credited as the author of some of them. Snodgrass translates from Dainichikyōsho (Śubhakarasiṃha's commentary on the MAT):
      "The hermit-ascetic Gotama [sic], flying in the sky at well, let fall two drops of sweat upon the earth, and the earth gave birth to sugar cane. Warmed by the sun, the sugar cane gave birth to two children, who became Śākya kings"*
      These two are the progenitors of the clan which 'Siddhartha' was born into. Gautama has a consort called Gautamī. I have not yet found the connection between the 'sugar cane' clan (Kansho) and the Śākya clan, though it may rest on a Chinese (mis)translation. In any case it that MAT includes the Vedic Gautama alongside the many other Vedic gods and important figures. Note that this story glosses over the fact that Gautama is a brahmaṇa, while the Buddha is usually referred to as a kṣatriya.

      * see Snodgrass, Adrian. The Matrix and Diamond World Mandalas in Shingon Buddhism, vol. 2, p. 470.


      3. Dating the Canon.

      The Assalāyana (MN 93) is a lengthy discussion between the eponymous Brahmin and the Buddha about the claim by Brahmins to be the best class (brāhmaṇo'va seṭṭho vaṇṇo). [4] Amongst the various arguments the Buddha puts forward is the relativist argument that some countries only have two classes, viz. ayyo and dāsa, i.e. noble and slave. [MN 93.5] These two countries (janapada) referred to are Yona and Kamboja. Various maps put Kamboja in different places, but it was supposedly north and west of Gandhāra. Shrimali centres it on the Kabul River (which flows through the Hindu Kush mountains from what is now Afghanistan to join the Indus) [5] Yona is thought to refer to Bactrian Greeks even further west. As the DOPN says:
      The name is probably the Pāli equivalent for Ionians, the Baktrian Greeks. The Yonas are mentioned with the Kambojas in Rock Edicts v. and xii of Asoka, as a subject people, forming a frontier district of his empire.
      These Greeks are thought to have been descendants of garrisons left by Alexander of Macedon. And this gives us our date. [6] At the time of the Buddha the Persian Achaemanids ruled as far east as the Indus River - i.e. including Gandhāra. We can confidently date Alexander's Indian campaign as part of his assault on and destruction of the Achaemanid Empire, to 327-326 BCE. If yona means 'Greek', then MN 93.6 cannot have been written before this date. Dates of the Buddha are less certain but the most recent research points to his death being circa 400 BCE, some 70 odd years before Alexander. Greek cultural influence remained for some time with post-Mauryan Dynasty Gandhāra being ruled by what is termed an 'Indo-Greek' dynasty from ca. 180 BCE - 10 CE. Greek aesthetic ideals heavily influenced Gandhāra art for some centuries, so that the first anthropomorphic images of the Buddha, produced in that region during the Kushan period (ca. 75-241 CE) showing obvious Hellenistic features.

      Note 7 May 2017 - to the best of my knowledge the Greeks never used Ionian as a general label. It was always a specific reference to Greeks who lived in Ionia - modern day Turkey. Moreover, the Greeks in question were Macedonians from Macedonia and that is probably how they referred to themselves. However, the Persians may have used Ionian as a general term for Greeks. If the Pāḷi Canon is using a Persian term for Greeks then this suggests that it was incorporated before Alexander. In which case the date goes back to being very vague indeed. 


      1. for instance -e can signify a masculine or neuter locative singular of a noun or past-participle in -a, such as gata (past-participle of gacchati).
      2. I don't want to multiply examples needlessly but Moggallana's given name was Kotila (after his village, just as Upatissa was called after his village). Kassapa (tortoise) is a very common name in Pāli perhaps because it was a gotra name as well. It seems that calling people by clan or family names, or epithets was a common practice.
      3. Note that Walsh translates this as a present (I am) when the verb is clearly past-tense; the Buddha left his clan, class, and caste behind when he went forth.
      4. D ii.148. Note that he continues "the other class is defective" (sometimes in this pericope the plural is used 'the other classes'). The Pāli being: hīno añño vaṇṇo. Here the term hīna is clearly being used pejoratively in a caste context. See also my Hīnayāna Reprise.
      5. Shrimali, Krishna Mohan. The Age of Iron and the Religious Revolution : c.700-c.350 BC. (A People's History of India: 3A). New Dehli, Tulika Books. 2007. Map p.85.
      6. I haven't found any reference to this fact, but I presume someone else has already noticed this.

      08 January 2010

      Mystical Grammar - oṃ & auṃ

      oṃ in the Siddhaṃ script
      oṃ in the
      Siddhaṃ script
      For the ancient Indians grammar was one of the major paradigms for understanding how the universe functioned. One product of this is in the understanding of the seed syllable (bījākṣara) oṃ in the Vedic and then the Hindu traditions.[1] The earliest references to oṃ are in the Yajur Veda. This Veda was composed sometime after 1000 BC but before the Buddha. In some rituals the hotṛ Brahmin shouts oṃ at the end of the invocation to the god being sacrificed to (anuvākya) as an invitation to partake of the sacrifice.

      The analysis of oṃ as being made up of three parts (a + u + ṃ) originates in the Sanskrit grammarian tradition but is given ritual or religious significance in the post-Buddhist early Upaniṣads, especially the Māṇḍūkya and Praśna Upaniṣads. Let's look at how this works.

      Vowels may be monophthong or diphthong - made up on a single sound (short or long), or made up of two sounds (short or long). Linguists might describe a diphthong as beginning on one vowel sound and ending on another. The vowels of Sanskrit can divided up like this:

      monophthongsa i u ṛ ḷ
      अ इ उ ऋ ऌ
      ā ī ū ṝ
      आ ई ऊ ॠ
      diphthongse o
      ए ओ
      ai au
      ऐ औ

      Note that the anusvāra (ṃ) and visarga (ḥ) are often counted as vowels, but practically they are modifications of existing vowels: nasalisation and aspiration respectively. They can be applied to any of the vowels. The long vowel ḹ (ॡ) is a theoretical possibility but in practice is never used.

      The vowel o is a diphthong which is made up of two sounds: a + u. However note that the vowel au is a long diphthong which is analysed as ā + u. The two vowels o and au sound quite different: o sounds like o in hope; au sounds like ou in sound. Similarly e can be thought of as a + i: and ai as ā + i. Technically (and metrically) e and o are long vowels. In Sanskrit the Proto-Indo-European short vowels e and o converged with a (which helps to explain why a is far more common than other vowels). Since there is no short e or short o in Sanskrit there is no need to write the long vowels as ē and ō, though this would be more consistent.

      It is necessary to understand these distinctions in order to understand some sandhi phenomena, because in some cases o actually behaves as a+u. The conjugations of the verbal root √bhū 'to be' offer a good example. This is a class 1 verb and forms a stem in -a with guṇā (strengthening) of the root vowel: so bhū (with guṇā) > bho; and when we add the stem vowel -a we get the stem form bhava; and the 3rd person singular is bhavati. What happens here is that the o in bho is treated as a + u, and the addition of -a invokes sandhi rules governing when two vowels meet - in this case u + a > va: ie bha+u+a > bhava. (This kind of thing is what makes learning Sanskrit difficult).

      auṃ written in the Siddhaṃ script

      in the
      Siddhaṃ script
      From a purely technical point of view we can see that the analysis of o as a + u does not justify writing oṃ as auṃ using the long diphthong. Note that the syllable auṃ written in Siddhaṃ (left) looks like the modern Hindu ॐ which is frequently transliterated as auṃ, and suggests that some confusion about this crept into Hindu discourse. Buddhist texts did not adopt the practice of writing oṃ as auṃ as far as I have been able to discover.

      Another purely technical point is that the notation oṃ indicates a nasalised o vowel. This should rhyme with the French 'bon', not with the English 'bomb'. In fact this distinction seems to have been lost for some time, and oṃ (ओं) is regularly pronounced as om (ओम् ) even in India (i.e. with the bilabial rather than the pure nasal). Note also that au is the vowel sound in the English word 'sound'. So auṃ should not sound like oṃ and vice verse.

      These jejune distinctions were important to the Indian grammarians because it was thought that the Vedas were divinely inspired, eternally unchanging and true, texts. They were transmitted orally, and after some centuries the vernacular Sanskrit language was significantly different from Vedic [2] which lead to scholars making a thorough investigation of the language - both canonical and vernacular at around the time of the Buddha. It was important to get the pronunciation right if the meaning was to be preserved. Changing the pronunciation was unthinkable.

      By the time the early Upaniṣads were being composed (beginning ca 800 BCE) there was quite a lot of interest in the relationship between words and reality. The existence of eternal, true words gave this a particular flavour. Also note that the word for 'true' and 'real' was the same: sat. It was the authors of the Upaniṣads, especially the Chāndogya (CU), who began to make the connections between syllables and aspects of the cosmos, though this seems to have been a natural development of the idea of correspondences (bandhu) between the macrocosm and the microcosm which was also a preoccupation in the Vedas. Oṃ in CU text is seen as a single syllable and equated with the udgītha, that is with the chanting of the sāman or hymns of the Sāma Veda. In other texts oṃ is associated with brahman. Then later, in the last century BCE, the technical breakdown of o into a+u was given esoteric significance. A key passage from the Māṇḍūkya reads:
      so 'yamātmādhyakṣaramoṅkāraḥ | adhimātra pādā mātrā mātrāś ca pādā akāra ukāro makāra iti || ManU 1.8

      On the subject of syllables, this syllable 'oṃ' is the ātman; on the subject of metre, the feet are the metre, and the feet are the syllables a, u, and ma. (my translation)
      Here oṃ, the ātman, is likened to 'śloka' the poetic metre (mātra [3]) consisting of four lines or 'feet' (pādā) of eight syllables, with each of the lines likened to a constituent phoneme. The Māṇḍūkya then spells out the esoteric correspondences of the constituent phonemes. The fourth foot (pādā) is said to be without a phoneme (amātra) and ineffable (avyavahāraya).

      I'm not aware of any canonical Buddhist text which restates the Vedic breakdown of oṃ into a+u+ṃ, though Kūkai does break hūṃ into ha+a+ū+m suggesting that the technique was not unknown to him. [4] For Buddhists the esoteric significance is typically based on the Arapacana acrostic which was originally a mnemonic for remembering aspects of an extended reflection on śūnyatā, for example: akāra (the syllable a) is the first syllable; which reminds us of the key word anutpanna (non-arising); and the full reflection subject is akāro mukhaḥ sarvadharmāṇāṃ ādyanutpannavāt (the syllable 'a' is an opening because of the primal quality of non-arising of all mental phenomena). Various versions of the Arapacana exist, the earliest date from around the 1st or 2nd century CE.

      This method of analysing mantras is far more significant in understanding the function of a mantra than the words in the mantra. For instance Kūkai always seems to have broken down words (even sūtra titles) into syllables in order to understand their esoteric significant. In Tibetan Buddhism the fact that the Avalokiteśvara mantra has six syllables which enables it to match up with the six realms of conditioned existence is probably more important than the understanding of the word maṇipadme (which has been central to Western exegesis of the mantra).



      1. This distinction is a bit vague. I call the religion Vedic which is primarily based directly on the three Vedas (Ṛg, Sāma, Yajur) excluding the Atharva (which is a distinct tradition I think) and which is focussed on the sacrifice: it's main gods were Indra, Agni, and Soma - though several dozen deities were propitiated. Hinduism is a complex of various religious ideas and practices where the Vedas have faded into the background and practice is focussed on devotion (bhakti) or with Tantric rites (śakti): prominent Gods are Śiva, Viṣṇu, Brahmā and the mother goddess in many forms especially Lakṣmi and Kāli. This is of course a massive over-simplification. What seems important is to mark that there have been tectonic shifts in India religions over the millennia.
      2. Vedic is the most common name for the language of the Vedas. It has a number of differences from Classical Sanskrit which was codified by Pānini in the 5th or 4th century BCE.
      3. Note the phonetic similarity with the English word - both come from the same Indo-European root meaning 'to measure'.
      4. see Ungi gi in Hakeda Major Works p.246ff

      Further Reading