Mantra in the Early Prajñāpāramitā Literature

One of the loose ends that needs tying up in thinking about the context of the Heart Sutra is the reference to mantra in the Sanskrit text. Of course, I have shown that the word doesn't occur in Chinese, but still, it does occur in the Sanskrit, so whoever translated the text into Sanskrit felt it was relevant. What we need to show is that it doesn't relate to the Prajñāpāramitā tradition, per se. 

The word mantra does occur in early Prajñāpāramitā texts, but not in the tantric sense and not in reference to Buddhist practices. Prajñāpāramitā makes it clear that mantra are not used by bodhisatvas because they are associated with trivial magic. 

A survey of all the uses of the word mantra in the extant Sanskrit texts is very manageable though identifying all the Chinese counterparts is more difficult due to lack of standardised translations. But the Chinese texts are important. Even the earliest Sanskrit texts come from the last century or two of Buddhism in India and although we now have a 1st Century CE Gāndhārī manuscript it only covers two chapters and has suffered a lot of damage. The Chinese translations from the Tang and before represent an earlier phase of development that is far more relevant to the creation of the Heart Sutra than, say, the late Nepalese manuscripts. If we want to know how mantra was seen in 7th Century China, we will need to take the Chinese texts of that period into account. 

As previously, this essay will survey the occurrence of the word mantra in the text now known as the Aṣṭasāhasrikā Prajñāpāramitā Sūtra (Aṣṭa) and the text now known as the Pañcaviṃśatisāhasrikā Prajñāpāramitā Sūtra (Pañc). The Chinese names for these vary and it's not clear that there was a distinction in the early translations. My principal points of reference in Chinese will be Kumārajīva's early 5th Century translations: Xiǎopǐnbōrě jīng 《小品般若經》(T. 227) and  Móhēbōrěbōluómì jīng《摩訶般若波羅蜜經》(T. 223). I will also include references to Dàoxíngbōrě jīng《道行般若經》(T. 224) by Lokakṣema (179 CE) — a translation of the Prajñāpāramitā Sūtra before there were smaller and larger recensions.* I will use Xiaojing a generic term for Chinese translations of the smaller sutra and Dajing for the larger. So as usual in the Nattier method we have four texts: Aṣṭa, Pañc, Xiaojing, and Dajing. We expect that occurrences in Aṣṭa will be copied into Pañc, and that Xiaojing and Dajing will reflect this (although, spoiler, this pattern is broken with respect to the word mantra). 

* Note the title of Lokakṣema's text translates as The Way of Practising Gnosis Sutra or something like prajñācāryamarga sūtra.

There is no Mantra in the Heart Sutra

By finishing a project begun by Yamabe Nobuyoshi and published by Jan Nattier (1992: n.54a) my article on the "epithets" passage (Attwood 2017) showed that the word mantra does not occur in the Xīnjīng. We know, from comparing his translations to the surviving Sanskrit versions of the same texts, that Kumārajīva translated Sanskrit vidyā as míng zhòu 明咒. But when Xuánzàng compiled the Xīnjīng in 656 CE, he read míng zhòu 明咒 as two words: bright dhāraṇī. One way we know this is that Xuánzàng included two epithets—dà shénzhòu 大神咒  and dà míngzhòu大明咒—that in Kumārajīva's Chinese both mean mahāvidyā. Keep in mind that Xuánzàng compiled the Xīnjīng from Kumārajīva's Prajñāpāramita five years before he began his own translations. Xuanzang also included a dhāraṇī (咒) incantation from the recently translated Dhāranīsamuccaya (trans. 654 by Atikūṭa), probably because he knew that Wu Zetian liked magic.

As suggested by Abé Ryūichi (1999), Tantra is a context - something that I think is much clearer in Shingon Buddhism than in Tibetan Tantra. The presence of isolated elements, such as a mantra, outside of that context cannot be considered tantric. Specifically, Tantra requires the communication of the cosmic body, speech, and mind of the Dharmakāya Buddha in the abhiṣekha ritual via mudrā, mantra, and maṇḍala. By replicating the cosmic body, speech, and mind the sādhaka transforms themself into a tathāgata. Nothing of this context is present in the Heart Sutra, or in the broader Prajñāparamitā tradition that it draws on. But Tantric Prajñāpāramitā texts were composed later on, potentially confusing matters.  

As a reflection of the translator's source text, mantra is obviously incorrect. Still, the choice of mantra in the Sanskrit translation is relevant to understanding the context. The monk who translated the Xīnjīng into Sanskrit either thought that zhòu 咒 meant mantra, or he wanted us to think that it did (i.e. he wanted to expressly link the Heart Sutra to the newly arrived Tantric Buddhism). He may have been unaware of the Sanskrit Prajñāpāramitā texts (and thus that the source has vidyā), but he must have been aware of the potential ambiguity of the character zhòu 咒. Like many Buddhist technical terms it has a straightforward use in Medieval Chinese, i.e. "incantation" as well as the specific uses mantra, dhāraṇī, vidyā.

With this in mind, and beginning with the Sanskrit, we can now look for mantra in the Prajñāpāramitā.

In the Prajñāpāramitā

I can identify three passages in the Aṣṭasāhasrikā that use the word mantra. However, not all of them have parallels in the Xiaopin. Curiously, only one of the three occurrences has a parallel in Pañcaviṃśātisāhasrikā although it also has another use but this is clearly much later. Where possible, I have tried to identify where the word occurs in Lokakṣema's translation of the Xiaopin (T 224).

Passage 1

teṣu ca susthitāḥ samāhitāśca bhaviṣyanti asyāṃ prajñāpāramitāyām | māreṇāpi te na śakyā bhedayitum, kutaḥ punar anyaiḥ sattvaiḥ, yad uta cchandato vā mantrato vā | tat kasya hetoḥ? yathāpi nāma tad dṛḍha-sthāmatvād anuttarāyāṃ samyaksaṃbodhau | te ca kulaputrāḥ kuladuhitaraś ca śrutvā enāṃ prajñāpāramitām udāraṃ prītiprāmodyaprasādaṃ pratilapsyante |(Vaidya 1960: 113)

They will be stable and concentrated in this perfect insight. They cannot be separated from it because of a verse or mantra, even by Māra, much less by other beings. Why is that? Precisely because of their resolute steadfastness with respect to ultimate complete awakening. And that disciple having heard this of this perfect insight will partake in excellent rapture, joy, and tranquility. 

Conze treats cchandato as "willpower" (1973. p. 160) i.e. reading chandataḥ "at will, according to desire". Paired with mantra the more obvious reading is chandas "sacred hymn; metre, metred verse". I think we have to see both words as being ablatives of cause, in -taḥ. Bhedayitum is an infinitive of the causative √bhid "break, injure, separate". Monier-Williams makes a distinction here by relating chandas to the Atharvaveda and mantra to the Ṛgveda, Samaveda, and Yajurveda. The Pāli texts separate the three Vedas and the Atharvaveda, treating the latter as something apart and characterise it negatively (See Who Were the Atharvans?).

The counterpart passage in Xiaopin begins at T 8.555.b06 but there is no mention of mantra. 

Passage 2

punar aparaṃ subhūte avinivartanīyasya bodhisattvasya mahāsattvasya vajrapāṇir mahāyakṣo nityānubaddho bhavati |... sa yānīmāni strīṇāṃ vaśīkaraṇāni mantra-jāpyauṣadhi-vidyā-bhaiṣajyādīni, tāni sarvāṇi sarveṇa sarvaṃ na prayojayati | (Vaidya 1960: 166)

Furthermore Subhūti, there is an eternal connection of the irreversible bodhisatva mahāsatva to the great yakṣa Vajrapāṇi...  He does not employ any of the mantras, recitations, herbs, spells, potions, etc for the subjugating of women.

阿惟越致菩薩，執金剛神常隨侍衛，不令非人近之。... 不以呪術藥草引接女人。(T 227, 8.565a24)

The irreversible bodhisatva, Vajrapāṇi (執金剛神), is always bound (常隨) to serve and protect, he does not command nonhumans to draw near him, ... doesn't use incantations [and] herbs to attract women.

Conze got this translation wrong, making strīṇāṃ vaśīkaraṇāni mean "the work of women" (399), but vaśī√kṛ means to "subdue", "subjugate". Women are not the agents of this, the agent of the sentence is, i.e. the bodhisatva.

The Sanskrit phrase mantra-jāpyauṣadhi-vidyā-bhaiṣajyādīni could be treated differently i.e. as mantrajāpya-oṣadhividyā-bhaiṣajya-ādīni "mantra-recitation, herb-lore, potions, and so on". Either way, these are practices associated with vulgar magic and sex (which for a community of monks is off limits). 

Kumārajīva's Chinese is quite different. Here zhòushù 呪術 means incantation and has been used in the past to translate vidyā, dhāraṇī, mantra, mantra-vidhi, and jāpya. "Non-humans" (fēi rén 非人) = Skt. amanuṣya and refers to devas, nāgas, yakṣas, it's not clear to me why not allowing (bù lìng 不令) them near him (jìn zhī 近之) is a good thing.

There is a Chinese counterpart in Lokakṣema's translation at T 224 (8.455.b26, but with Seishi Karashima's corrections 2010). I'm quite sure this is the right passage but to be honest I'm struggling to make sense of it - it is very different from Kumārajīva.

是菩薩，和夷羅洹化諸鬼神，隨後，亦不敢近附。... 終不誘他人婦女。若有治道符祝，行藥，身不自為，亦不教他人為，見他人為者心不喜也。終不說男子若女人為。 (T. 224; 8.455b.28-c.3). 

Here 和夷羅洹 Vajrapāṇī is actually a transliteration of a Middle Indic form of the name: *Vajiravāṇi. One of many indications that Lokakṣema was translating from Gāndhārī rather than Sanskrit. The phrase is zhōng bù yòu tā rén fùnǚ. 終不誘他人婦女。It means something like: "In the end he does not seduce these women". I don't think this should be followed by "。" since what follows are the means associated with the seduction in all the other texts, i.e. zhì dào 治道 "uses witchcraft", fú zhòu 符祝 "incantations", xíng yào 行藥 "practising medicine".

There is a counterpart of this passage in the Large Sutra that doesn't mention Vajrapāṇi.

punar aparaṃ subhūte bodhisattvo mahāsattvo bodhimanasikāraiḥ samanvāgato yāni tāni strīṇām āveśanāni vaśīkaraṇāni mantravidyauṣadhibhaiṣajyāni tāni sarvāṇi sarveṇa sarvaṃ sarvathā sarvan na prayukte na ca prayojayati na ca strīṇām āveśanam anyatarānyataraṃ karoti, na striyāḥ puruṣasya vā ādeśanāprātihāryaṃ karoti, putro vā te bhaviṣyati dhītā vā te bhaviṣyati, kulodgato vā bhaviṣyati, dīrghāyuṣko vā bhaviṣyati. (Kimura 4:157)

Furthermore, Subhūti, the bodhisatva mahāsatva endowed with attention to awakening does not employ, in any way, shape, or form the mantras, spells, herbs, potions (mantra-vidyā-oṣadhi-bhaiṣajyāni) etc. used to magically subdue women; and he does not engage in doing other magic on women: he will not declare mind-reading of a woman or a man, he will not predict the sex of children, or lineage, or lifespan.

(Chapter 50 of Conze's translation). 

I think the latter part of the passage in Lokakṣema is quite similar to the latter part of this.

Passage 3

punaraparaṃ subhūte avinivartanīyā bodhisattvā mahāsattvāḥ kāmāvacarebhyo devebhyaścyutā rūpāvacarebhya ārūpyāvacarebhyo vā devebhyaścyutāḥ santaḥ ihaiva madhyadeśe jambūdvīpe pratyājāyante / yatra sattvāḥ kalāsu kovidāḥ, kāvyeṣu kovidāḥ, mantreṣu kovidāḥ, vidyāsu kovidāḥ, śāstreṣu kovidāḥ, nimitteṣu kovidāḥ, dharmārthakovidāḥ / Vaidya 167)

Furthermore, Subhūti, irreversible bodhisatvas, being fallen from the sphere of desire, or from the gods of the form sphere, or the gods of the formless sphere, are reborn right here in the middle country (madhyadeśe) where beings are learned (kovida) in the arts, verse, incantations, spells, exegesis, etymology, and understanding duty. 

須菩提！阿惟越致菩薩，多於欲界、色界命終來生中國，善於伎藝，明解經書， 呪術占相，悉能了知。(T. 227; 8.565b.11-15)

Subhūti. The irreversible bodhisatva, exceeding the kāmadhātu (欲界) and the rūpadhātu (色界) after death he is born in the middle country (中國), [where people are] good at the arts (善於伎藝), experienced (明解) in exegesis (經書), divination (占相), all kinds of learning (悉能了知)

The region of madhyadeśa is roughly speaking the Ganges Valley border to the north by the Himalaya mountains, to the south by the Vindhya Hills. In other words this is the Buddhist heartland. People there are learned in kalā, kāvya, vidyā, śāstra, nimitta, dharmārtha.  

• kalā is ambiguous, literally "a sixteenth" (of unknown etymology) but "the arts" seems to fit, later kalā formalised as the 64 kinds of performing arts; 
• kāvya is the art of metered verse especially as found in the Vedas and Epics, 
• mantra is ambiguous since it can refer to magical spells generally or it is a way of referring to verse from the Vedas used within rituals (this was the original sense), as we have seen Tantra is definitely not intended; 
• vidyā in this context is the practical arts, but also the soteriological arts;
• śāstra is the art of explaining the content of religious and/or grammatical texts; 
• nimitta is ambiguous and could be related to divination ("signs") or grammar where it refers to etymology roughly a synonym of nirukta;
• dharmārtha is also ambiguous and could refer either to "the meaning of the Dharma", or to the contrast between the letter (dharma) and the spirit (artha) of, for example, a religious teaching, or to religion (Dharma) and to wealth (artha), i.e. to what Christians call the spiritual and temporal realms.  

This is the long way of saying that the people of Madhyadeśa were educated. Probably not everyone, but everyone you'd expect to be educated was - in this case male landowners and their sons. (Note this is just a description of the times). Either Kumārajīva's text was shorter or he felt that a few examples followed by "all kinds of learning" (xī néng le zhī 悉能了知) got the point across. 

The phrase "being fallen" (cyutāḥ santaḥ) is a euphemism for dying.

The passage is found in Lokakṣema's translation at T 224; 8.455c.17-18. (從欲處、色處、空處，從彼間來生中國，常於善人黠慧中生，在工談語曉經書家生。)

Passage 4

There is one further passage that occurs in Pañc, but it evidently late and a reference to tantric Buddhism since it mentions that the superior man (satpuruṣa) "protects the secret mantras" (guhyamantrarakṣaṇāc). Note that Conze's translation does not include it where we expect it from Kimura's Sanskrit text (i.e. at p.584) but he does include the parallel translation in Appendix II  (p.660), which deals with the reasons why a Buddha has the thirty marks of the superior man.

Given that this passage is an interpolation we need not dwell on it and can now move to concluding remarks.

Conclusions

We have some simple and obvious conclusions from this material:

1. Mantra was not considered part of the bodhisatva path.
2. Mantra was considered vulgar magic (used for attracting women, etc), 
3. There is no sign of a Tantric context in our source texts.
4. The fact that mantra occurs less often in Pañc than in Aṣṭa suggests that perhaps such references were added to Aṣṭa after the creation of Pañc (evidence of something similar happened to the epithets passage - see Attwood 2017)

It is simply not possible that if the gate gate incantation were a mantra, that an Indian Buddhist writing in the first half of the first millennium of the Common Era would have included a mantra in a Prajñāpāramiā text. Ergo, the Heart Sutra was not composed in India when Conze suggests it was. 

Furthermore, we know that the so-called mantra is, in fact, a dhāraṇī and dhāraṇī were added to texts in India. However, there is still no evidence of the Heart Sutra outside of China before the 8th Century. What we can say is that Indians who went to Tibet wrote commentaries on it (Lopez 1988, 1996). However, while Lopez assumes that the commentaries were composed in India, the evidence does not support this. We can really only say that this is evidence for the text in 8th-12th Century Tibet. It is not evidence for the presence of the Heart Sutra in India. Rather, the earliest evidence for the text anywhere near India is a 13th Century Nepalese ms., Cambridge ADD 1680 (see my transcription of this ms).

We know that there are many copies of the Heart Sutra at Dunhuang, including many in Tibetan. Both Chinese and Tibetan Buddhists were capable of composing the extended version, taking it to Tibet. We can only hope that someone studies these texts at some point (I have a conference paper on this issue by Ben Nourse that is not for publication, but Nourse has not returned to the topic of the Heart Sutra at Dunhuang). I think this would be a great PhD topic for someone well versed in Tibetan and Chinese. 

If I write this up for publication at some point, I'll need to look at Mokṣala's translation of the Large Sutra as well. 

This is all confirmation of the revisionist history of the Heart Sutra proposed by Nattier and which has been my main focus for eight years. The Heart Sutra was not composed in Sanskrit. It was composed, probably by Xuanzang, in Chinese, using excerpts from Kumārajīva's Large Sutra translation. Xuanzang added a dhāraṇī onto the end of the text. No one at the time confused this for a mantra until the monk who translated the Heart Sutra into Sanskrit made the mistake of translating zhòu 咒 as mantra. Once that happened and the fraud was successful, everyone started thinking of the gate gate dhāraṇī as the gate gate mantra. 

~~oOo~~

Bibliography

Attwood, Jayarava. (2017). ‘Epithets of the Mantra’ in the Heart Sutra.’ Journal of the Oxford Centre for Buddhist Studies, 12, 26–57.

Abé, Ryūichi (1999). The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse. New York: Columbia University Press.

Karashima Seishi. 2010. A Glossary of Lokakṣema's Translation of the Aṣṭasāhasrikā Prajñāpāramitā. Tokyo: IRIAB, Soka University.

Karashima Seishi. 2011. Critical edition of Lokakṣema's translation of the Aṣṭasāhasrikā Prajñāpāramitā. Tokyo: IRIAB, Soka University.

Kimura, Takayasu. 2009. Pañcaviṃśatisāhasrikā Prajñāpāramitā. Tokyo: Sankibo Busshorin.

Vaidya, P.L. 1960. Aṣṭasāhasrikā Prajñāpāramitā Sūtra. Darbhanga: The Mithila Institute. (Gretil Archive, 2014. Including Karashima, S. (2013) On the "Missing" Portion in the Aṣṭasāhasrikā Prajñāpāramitā. ARIRIAB, 16: 189-192).


Note: I had made some statements about mantra in the Prajñāpāramitā in the past when Alexander O'Neill wrote to me in February 2018. He challenged my conclusions, which admittedly were based on a rather cursory reading and my Pāli bias. I have had in the back of my mind to do a close reading of the relevant passages since then but have only just gotten around to it. My thanks to Alexander for prompting me to go the extra mile and look closely at the details (wherein the Devil lurks). 

Dhammaniyāmata and idappaccayatā

I was going over my notes on niyāma and comparing some Pāli and Chinese texts a couple of years ago and started writing this essay. I discussed it on the Sutta Central forums in 2018. I noticed it sitting in the draft folder and thought it would be worth finishing and putting on the blog.

My interest in the concept of niyāma is long-standing and ongoing because Sangharakshita employed the word in his teaching. However, he based his use of niyāma mainly on the ideas of Carolyn Rhys Davids who seems to have concocted a narrative that didn't relate to what we find in Pāli. Rhys Davids was aided by Ledi Sayadaw although they disagreed on some details. Ever since I discovered the way Buddhaghosa actually used niyāma I've been interested to flesh it out. One of the avenues for doing so is to refer to the several uses of the word niyāma in the compound dhamma-niyāmatā and one of the places we find this word used is the Paccaya Sutta (SN 12:20). When we compare the Pāli with its counterpart in the Samyuktāgama, i.e. SĀ 296, we notice something quite interesting with respect to another word idappaccayatā.

A Samyuktāgama manuscript was translated into Chinese by Guṇabhadra (求那跋陀羅) in the Liu Song 劉宋 period (435–443 CE). However, there is also a Sanskrit text from a cache found at Turfan. It was copied much later, probably around the 13th Century. We begin with the Pāli.

Paccaya Sutta (SN 12:20)

The Pāli passage with my translation:

uppādā vā tathāgatānaṃ anuppādā vā tathāgatānaṃ, ṭhitāva sā dhātu dhammaṭṭhitatā dhammaniyāmatā idappaccayatā. (SN II.25)

“Whether anyone is awakened or not, the principle remains: the fact of mental events being conditioned, the fixed course of mental phenomena, and specific conditionality [of mental phenomena].”

Tathāgata is how the Buddha referred to himself. It means someone who has realised nirvāṇa or attained awakening (etc). More literally the phrase is "arising of a tathāgata or non-arising of a tathāgata".

Here thitāva is ṭhitā eva. Ṭhita (Skt sthita) is the past participle of the verb tiṭṭhati (Skt tiṣṭhati from the root √sthā "to stand, remain"). A ṭhita is something lasting or enduring. The long final ā tells us that ṭhita is being used as an adjective of dhātu, a feminine noun. Dhātu can have a range of meanings including "element; natural condition, property; factor, item, principle." If we take ṭhitā dhātu as a unit, then we expect meanings such as "abiding principle", "established property", or "enduring natural condition". In other words, it is a state of affairs that remains in play. The particle eva emphasises the endurance of the principle.

Ṭhitatā is an abstract noun from the same past participle (ṭhita). PED suggests it means "the fact of standing or being founded on". In other words, the connotation is somewhat different here. The word is mainly used in precisely this context as a quality of dhammas. As we know, conditioned (saṅkhata) dhammas arise (samuppāda) in dependence on a condition (paccaya). Dependence (literally "hanging down from") is an inversion of the cognitive metaphor involved in the Pāli word paṭicca (Skt pratītya), which is from the root prati√i and means "going back to, returning". This also gives us the title of the sutta, paccaya (Skt. pratyaya). A dhamma (literally "support") springs-up (samuppāda) when the condition (paccaya) that supports it (from below) is in place. I take dhamma-ṭṭhitatā to be a reference to the principle of conditionality. And that we can take it to mean that the principle of conditionality is an abiding principle.

Next, niyāma means "a fixed course; constrained; inevitably". In the context of dhamma-niyāmatā "the fact of the fixed course of dhammas", this means that dhammas don't get a choice. When the conditions are in place, dhammas must arise; when the conditions are absent dhammas must either not arise or having arisen they must cease. Buddhaghosa relates this to the inevitability and inescapability of the ripening karma (cf Attwood 2014). So again this is a reference to the principle of conditionality.

Finally, the abiding principle is also idappaccayatā (Skt idampratyayatā). The etymology is fairly obvious (more so in Sanskrit) but difficult to articulate in English. The whole thing is an abstract noun, so refers to an abstraction from the idea of idaṃ pratyāya "this condition" or "whose condition is this". However we get there, the word refers to the specificity of the conditions: specific conditions give rise to specific results. In other words, there is an order to how dependent arising functions: it has to function and it has to function in a particular way that relates consequence to action: A kusala cetanā gives rise to a kusala phala; if there is a kusala phala we can infer a kusala cetanā as condition. We can see, therefore, that idappaccayatā is the same quality as Buddha-ghosa's bījaniyāma (like for like), perhaps combined with utuniyāma (timeliness).

We might call this a law of nature. A law of nature is always applicable, always gives the same result given the same causes. It is a ṭhitā dhātu or a niyāma. Initially out of idle curiosity, I wanted to see what the Chinese equivalents of these words were so I used Sutta Central to identify the Chinese counterpart in the Samyuktāgama. And things started to get interesting because it rapidly became apparent that the Chinese text was corrupt in an interesting way.

The Chinese. SĀ 296

The Samyuktāgama parallel with Choong Mun-keat's translation is

若佛出世，若未出世，此法常住，法住法界 (T2.84.b17-18)

“Whether a Buddha arises in the world, or not, this is the unchangeable nature of dharma, the status of dharma, the element of dharma.” (Choong 2010: 45)

Note that in the Sutta Central metadata this translation is credited as "originally published in" Choong (1999) but it is not translated in that book as far as I can tell. Rather, it is translated in Choong (2010).

Choong (1999: 19) leads us to believe that certain Pāli and Middle Chinese terms are equivalents, i.e.

• dhammaṭṭhitatā = fǎ zhù 法住
• dhammaniyāmatā = fǎ dìng 法定
• idappaccayatā = fǎ jiè 法界

But this cannot be right and furthermore what we actually have in the Chinese Pratyāya Sūtra is: fǎ cháng zhù 法常住, fǎ zhù 法住, fǎ jiè 法界. Under 法常住 the Digital Dictionary of Buddhism suggests "the Dharma that is eternally abiding". But I think the pronoun cǐ 此 in the text implies dharmāḥ in the plural, meaning mental phenomena rather than the Dharma (singular) - this is also my reading of the Pāli compounds. So a translation like "dharmas abide eternally" is more likely, though obviously this is quite problematic from a Theravāda point of view (I'll return to this point at the end of this section).

Note that the text has taken a shortcut. There is no equivalent to ṭhitāva sā dhātu "the principle remains". Although we do actually have the characters we need for it, i.e. zhù jiè 住界 "abiding principle".

T2 p.84
17-18

The problem here is that fǎ cháng zhù 法常住 "dharmas eternally remain" is followed by more or less the same word, i.e.fǎ zhù 法住 "dharmas remain". We are expecting to find 法定 the equivalent of dhammaniyāmatā. In other words, this appears to be a scribal error with repetition/substitution of 住 for 定. Such errors are very common in copied manuscripts and called an "eye-skip"). From the imaged page accompanying the CBETA reader the error occurs in the printed version of the Taishō Tripiṭaka as well (see also image right).

Choong glosses over (and thus hides) the repetition by choosing different translations for the two phrases, viz. "the unchangeable nature of dharma" and "the status of dharma". Here "nature" and "status" both translate zhù 住. The character 住—meaning "stay behind, remain; pause, halt"—is a commonly used to translate words deriving from Sanskrit √sthā "abide, stand, remain" since the semantic fields substantially overlap. It is also used to translate the verb viharati "dwelling, abiding".

Moreover, the characters fǎ zhù 法界 do not translate idappaccayatā but rather usually translate dharmadhātu, i.e "the realm of dharmas qua experience " or the "experiential realm" rather than the later idea of a realm of the Dharma). A modern translation of idappaccayatā is 此縁性 but this is no help to us here precisely because it is modern and not found in the Āgama texts. And the reason for this took some time to appreciate.

It turns out that in the whole Saṃyutta Nikāya the word idappaccayatā only occurs here in the Paccaya Sutta, so there is no easy way to find out if Guṇabhadra used it in other contexts. In fact, this word is very uncommon. Across all the Nikāyas it only occurs one other time, in the Ariyapariyesanā Sutta (MN 26), and there only once. The Chinese version of MN 26, i.e. MĀ 204 omits the passage that includes the word idappaccayatā. The story of Brahmā’s request to teach (Section 20 of MN 26) is recounted in (1st) Ekottarikāgama (19.1) but does not use this word. Ekottarikāgama 24.5 recounts part of the story but also misses out the passage of interest. So there appears to be no Chinese Āgama text that uses this expression. At this point, I do not even have any Chinese characters that I could search for in the Chinese Canon.

We do find the expression in Dàzhìdù lùn《大智度論》 (T. 1509) a voluminous commentary on the Large Prajñāpāramitā Sutra, where our passage of interest is cited.

有佛無佛，是因緣法相續常在世間 (T1509.253c.1-2)

“Whether there is a Buddha or there is no Buddha, this causality (idaṃpratyayatā), this nature of things (dharmatā), is always present in the world.” (Ani Migme’s translation of Lamotte's French translation)

However, here 因緣 seems to correspond, not to idaṃpratyāya, but to another common compound, >hetu-pratyāya “causes and conditions”. The character yīn 因 means "cause" and routinely translates hetu in contrast to yuán 緣 which routinely translates pratyāya "condition". The combination f ǎ xiāng 法相 does, at least, correspond to dharmatā, but I think Lamotte has fudged the translation of yīn yuán 因緣 because he knew what it ought to say. With hindsight the expression idaṃpratyayatā is not used here.

Sarvāstivāda?

I noted above a little anomaly in the Chinese translation with respect to fǎ cháng zhù 法常住 "dharmas abide eternally". The Saṃyuktāgama manuscript translated into Chinese is thought to have belonged to a Sarvāstivāda sect. This phrase—fǎ cháng zhù 法常住—may betray a Sarvāstivāda point of view.

The late David Bastow (1995) outlined why the Sarvāstivādins came to the conclusion that dharmas must be eternal (always existent, i.e. sarva-asti). They began by taking the formula of dependent arising seriously (cited here in the less familiar Sanskrit form that Sarvastivādins used)

yaduta asmin satīdaṃ bhavaty asyotpadād idam utpadyate |
yaduta asmin asatīdaṃ na bhavaty asya nirodhād idaṃ nirudhyate ||

When this is present, that exists,
with the arising of this, that arises.
When this is absent, that does not exist,
When this ceases, then than ceases.

Let us suppose that a citta or dharma characterised by greed (lobha) arises and then ceases. If we are to make progress as Buddhists, we have to know that we just had a lobha-dharma. In order to know this, the citta itself must be a condition for the knowledge. But according to the formula "that" (result) arises only when "this" (condition) is present. Therefore, the lobha-dharma must still be present. If this relation holds true, then the logical outcome is that dharmas must always be present, they must always exist: sarva-asti. This is where the logic of the formula takes us and is by no means illogical or stupid. It was the dominant Buddhist view in the northwest of Greater India for some centuries.

Of course, this is eternality is problematic, but the Sarvastivādins got around it by positing that dharmas are always present but only active in the present. This was their metaphysical manoeuvre. All Buddhists ended up having metaphysical manoeuvres to try to link consequences to actions over time (this is one of the main topics of my book on karma and rebirth). The sarva-asti manoeuvre is actually quite metaphysically conservative compared to what made it through into the present. Buddhists tended to proliferate supernatural entities and processes to make karma work, e.g. bhavaṅgacitta or ālayavijñāna.

With this digression complete. we can now turn to the Sanskrit manuscript of the Pratyāya Sūtra, which is the latest of all our sources but may be of some use.


Pratyāya Sūtra

Part of a Sanskrit Nidānasaṃyukta was found at Turfan and edited by Chandrabhāl Tripāṭhī. It was probably copied in the 13th or 14th Century and we know very little about the provenance of it. This gives us a third version of the passage

ity utpādād vā tathāgatānām anutpādād vā sthitā eveyaṃ dharmatā dharmasthitaye dhātuḥ (Sutra 14, line 5)

Thus, whether or not a tathāgatā arises, this principle remains: naturalness and the stability of dharmas.

Cf P. uppādā vā tathāgatānaṃ anuppādā vā tathāgatānaṃ, ṭhitāva sā dhātu dhammaṭṭhitatā dhammaniyāmatā idappaccayatā. (SN II.25)

It appears to me that the Sanskrit text has become garbled. The word order has changed, partially obscuring the relation between sthitā and dhātuḥ. And we are lacking any equivalent of idappaccayatā.

The word dharmatā appears to be out of place. We are expecting to see dharmaniyāmatā. Meanwhile, dharmasthitaye has a case ending which is unexpected, since we expect to see a nominative singular. Given that the word is a feminine noun in -ā we don't expect to see -aye at all. According to Egerton's Buddhist Hybrid Sanskrit Grammar (1953: 63) it does occur as a case ending for -ā in the Mahāvastu, but it is used obliquely, i.e. for all the cases from instrumental to locative. Which in any case is not the nominative case that we expect (1953: 61). It is likely that -āye was intended, since this is more common, but again, this is the oblique case ending and it cannot be correct.

If I take the Pāḷi and render it into directly into Classical Sanskrit and Middle Chinese, it looks like this (followed by the actual texts for comparison):

P. ṭhitāva sā dhātu dhammaṭṭhitatā dhammaniyāmatā idappaccayatā
S. sthitaiva tāḥ dhātu dharmasthitatā dharmaniyāmatā idaṃpratyayatā
C. 住即界法住法定 XXX*

Turfan: sthitā eveyaṃ dharmatā dharmasthitaye dhātuḥ

SĀ 296: ... 此法常住，法住法界.

* Re XXX: by the end of the section on the Chinese text we still had not identified the Chinese equivalent of the word idappaccayatā.

There was nothing very difficult or complicated about either of these translations except the lack of a Chinese translation for idappaccayatā. The words in the Turfan ms. seem to have gotten jumbled up and fragmented so they no longer make sense. Something similar has happened to the Chinese text.

Second Phrase

If we read a little further on in the Chinese text we find a similar phrase which uses some of the same terminology (and the Pāli version once again uses the word idappaccayatā). Starting with the Chinese text and Choong's translation:

此等諸法，法住、法空、法如、法爾，法不離如，法不異如，審諦真實、不顛倒 (T 2.82.b22-4)

All these dharmas are the status of dharma, the standing of dharma, the suchness of dharma; the dharma neither departs from things-as-they-are, nor differs from things-as-they-are; it is truth, reality, without distortion. (Choong 2010: 45-6)

Choong's note (2010: 45) reads:

法定 The unchangeable nature of dharma. Original Taishō texts has 法空, but according to CSA*, it should be 法定 (vol.2, p.36).(2010: 45)

*CSA = Yin (1983).

In other words, the problem here is that 法空 (dharma-śūnyatā) should be 法定 (dharma-niyāmatā). This is consistent with the Sanskrit text. Note that the opening block of text 此等諸法 is literally "these 此 [plural] 等 many 諸 dharmas 法", hence Choong "all these dharmas".

The Turfan Sanskrit text counterpart reads

iti yātra dharmatā dharmasthititā dharmaniyāmatā dharmayathātathā avitathatā ananyathā bhūtaṃ satyatā tattvatā yāthātathā aviparītatā aviparyastatā (14.6)

Here, yātra is not to be confused with the locative adverbial pronoun yatra "where". It must either be a mistake for yatra, or the result of sandhi from yā atra "which here", where yā is the Prakrit relative pronoun in the feminine nominative singular. The Classical Sanskrit is yāḥ which would be yā followed by a vowel, but would not undergo further sandhi, i.e. yāḥ atra > yā atra. The Pāli text has yā tatra, which appears to confirm the yā atra reading. However, overall, the Pāli has a very different vocabulary at this point.

Pāli: Iti kho, bhikkhave, yā tatra tathatā avitathatā anaññathatā idappaccayatā—ayaṃ vuccati, bhikkhave, paṭic­ca­samup­pādo.

Thus, monks, that which is actual, not unactual, not otherwise—this is called dependent arising.

As Bodhi (2000: 742 n.54) elsewhere notes (i.e. in relation to SN 56:20 = 56:27), the Four Noble Truths are said to be tatha, avitatha, and anaññatha or in his translation "actual, unerring, and not otherwise". Note these are adjectives rather than nouns. Buddhaghosa's commentary on SN 12:20 in the Sāratthappakāsinī (or Saṃuyttanikāya Aṭṭhakathā) gives these terms a very specific meaning: tathatā refers to phenomena arising when conditions are present; avitathatā refers to this being a non-repeating process - one set of conditions gives rise to one phenomena; and anaññathatā means that each set of conditions gives rise to a phenomenon that is specific to those conditions, i.e. anaññatha is synonymous with idappaccayatā (which makes me suspect that the latter was added after the fact). Although Bodhi links this commentarial gloss with SN 56:20, the commentary he is translating is not the one that comes up when I look at the commentary on this text. Unfortunately, Bodhi does not say what he is translating.

Note that once again neither the Chinese nor the Sanskrit has an equivalent of idappaccayatā. It's not clear why they have a completely different pericope at this point. Clearly, this goes beyond a simple translation issue. The text appears to have been constructed with a different pericope at this point.

Conclusion

I started off exploring the meaning of dhamma-niyāmatā in a sutta with a view to better understanding Buddhaghosa's later use of the term niyāma "fixed course, constraint". This quality of dhamma-niyāmatā is said to be an abiding principle (ṭhitā dhātu). It seems that it refers to the conditionality of dhammas qua phenomena (a word that originally means "appearances"). And thus is it not related to the way that Buddhaghosa uses the term dhammaniyāma to describe the miracles that accompany the milestones in the career of a buddha. This is quite important for the concept of niyāma in Buddhism.

We have to be careful when thinking about phenomena in early Buddhism lest we inadvertently transpose a modern understanding of the concept or the terms we use. It is true that early Buddhists seem to make a distinction between mental and physical phenomena. It seems quite likely to me that they considered physical objects to be independent of their minds, but understood that phenomena associated with such objects (i.e. how such objects appear to the mind) as like mental phenomena. But the insight is connected in their case to meditative experience in which sensory and cognitive experience cease without the loss of consciousness, what is sometimes referred to as contentless awareness. Buddhists were not Idealists in the classical sense, but they might have had some sympathies with Kant's Transcendental Idealism had they come across it.

As I understand early Buddhists, they were making an epistemic argument about the conditions under which we experience phenomena. And they were not making a metaphysical argument about the nature of objects that we experience through phenomena. Indeed, the cessation of experience in meditation eclipsed any and all metaphysical speculation in importance. Which is not the same as saying that early Buddhists did not have metaphysics or that their epistemic conclusions did not have metaphysical implications. Rather, they simply never systematically developed metaphysics as a branch of philosophy. Phenomena and the nature of experience were the focus.

I argued that dhammaṭṭhitatā is a reference to fact of conditionality, what the early Buddhists considered to be an enduring law of nature, that phenomena arise when the condition is present. And that the term dhammaniyāmatā is a synonym with only a slightly different connotation. That dhamma (qua mental event) has a fixed course (niyāmatā) is a reference to the inevitability of a phenomenon arising when the condition for it is present, and the ceasing or non-arising when the condition is absent. This suggests that dhamma-niyāmatā is more like Buddhaghosa's kamma-niyāma.

The original phrase in the sutta was probably: ṭhitāva sā dhātu dhammaṭṭhitatā dhammaniyāmatā "this fact remains: the fact of mental events being conditioned, the constraints on mental events". And since the text refers to only one fact (sā dhātu), we have to read dhamma-ṭṭhitatā and dhamma-niyāmatā as being synonyms, not two different terms. Contrarily, we have to read dhamma here as dhammā (plural) and as being related to phenomena qua how things appear to us, not the existence of things or the nature of their existence.

It seems that idappaccayatā only occurs in Theravāda texts and may be a late insertion. I could find no early Buddhist texts in other languages that contain this word, even when there are translations that seem to be direct parallels. There seems to be no entry in the Gāndhārī dictionary that corresponds to this word. This seems to me to be quite a significant preliminary result that should be followed up on, but it probably won't be me that does it as I have my hands full with other things for the foreseeable future.

We also see how errors build up to render a text confusing or even meaningless and how Chinese texts read in isolation are often misleading.

~~oOo~~

Bibliography

Attwood, Jayarava. (2014). "Escaping the Inescapable: Changes in Buddhist Karma." Journal of Buddhist Ethics, 21,503-535. http://blogs.dickinson.edu/buddhistethics/2014/06/04/changes-in-buddhist-karma

Bastow, David. (1995). The First Argument for Sarvāstivāda. Asian Philosophy 5(2), 109-125. Online: http://buddhism.lib.ntu.edu.tw/FULLTEXT/JR-ADM/bastow.htm

Choong Mun-keat (1999). The Notion of Emptiness in Early Buddhism. [2nd. Ed.] Motilal Banarsidass.

Choong Mun-keat (2010). Annotated Translation of Sutras from the Chinese Samyuktāgama relevant to the Early Buddhist Teachings on Emptiness and the Middle Way. [2nd Rev. Ed.] Thailand: International Buddhist College.

Edgerton, F. (1953). Buddhist Hybrid Sanskrit Grammar and Dictionary. Vol I. Reprinted: Munshiram Manoharlal, 2004.

Tripāṭhī, C. (1962). Fünfundzwanzig Sūtras des Nidānasaṃyukta. (Sanskrittexte aus den Turfanfunden, VIII). Berlin : Akademie-Verlag. Online at GRETIL.

Yin Shun. (1983) Combined Edition of Sūtra and Śāstra of the Saṃyuktāgama. Shanghai: Zhonghua Book Company. [印順. (1983) 雜阿含經論會編. 中華書局.]

Revisiting Avalokiteśvara in the Heart Sutra

In my 2019 article on Xuanzang and the Heart Sutra, I argued that is was implausible for Xuanzang to have been involved in any clandestine attempt to pass the Heart Sutra off as a genuine sūtra. By contrast, Jeffrey Kotyk (2020) makes a good case for the Heart Sutra having been openly composed by Xuanzang as a condensation of the Prajñāpāramitā (i.e. a chāo jīng) and given to Gaozong and Wu Zhao as a gift on the birth of a son. On 6 January, 656, Xuánzàng sent a letter to the emperor celebrating the birth of a new prince the month before. He wrote in a letter: "I dare to offer a copy of the Prajñā Heart Sutra in gold letters, one scroll and a case." (輒敢進金字《般若心經》一卷并函 T 2053; 50.272b.12).

We still lack an explanation for the process of the text becoming an "authentic" sutra, although I have identified many of the components of the received myth and shown that they emerge over several decades. There must have been a point when the "fact" that the Heart Sutra was a translation by Xuanzang became established. If we accept Kotyk's thesis (and I am inclined to) then this transition occurred within five years because the Fangshan Stele, which credits Xuanzang as translator, is dated 13 March 661. 

Such considerations are tied up with questions of the historicity of sources. In the same article, Kotyk argued against the uncritical use of Xuanzang's Biography published in 688 CE* as an historical source because it is a hagiography with all that this implies: the religious and political agendas of the author are far from hidden. Unfortunately, when we strip out the magical and mystical elements we do not arrive at a narrative that tallies with the other historical sources (although, of course these also have their biases). In particular, the Biography appears to distort the relationship of Xuanzang and Taizong in ways that are favourable to the Buddhist community but not entirely plausible.

* i.e. Huìlì 慧立 and Yàncóng 彥悰. Da tang da ci'en si sanzang fasha chuan 大唐大慈恩寺三藏法師傳 (Biography of the Tripiṭaka Master [of the] Great Ci’en Monastery [of the] Great Tang), T 2053, 50.

Partly inspired by correspondence with Kotyk, I have also been critical of the use of the hagiography as history (Attwood 2019). By sheer bad luck my article was published before Kotyk (2020) whereas he was finished first and I had read a draft and corresponded with him about it while writing my article. Preceding us both, Max Deeg has been critical of naive readings of the Xuanzang's travelogue, Notes on the Western Regions (西域記 Xīyù jì), composed ca 645 or 646 CE.

More specifically Deeg (2016: 126-8) has pointed to historical inaccuracies in how Xuanzang portrays the Indian king, Harṣavardhana (606 to 647 CE), of the Puṣpabhūti Dynasty (henceforth King Harṣa). Deeg plausibly argues that these inaccuracies appear to be deliberate narrative devices on the part of Xuanzang. He seems to have tried create a sympathetic protagonist for the Tang Emperor Taizong to identify with, so that he might take a moral lesson worked into the story. In this essay, I will extend Deeg's argument: if we accept that Xuanzang took a didactic approach in writing Notes on the Western Regions and gave Avalokiteśvara an educational role, then it is worth reconsidering the unexpected appearance of Avalokiteśvara in the Heart Sutra in this light.

Avalokiteśvara

Name

The bodhisattva first appears in Chinese translations from the 2nd Century CE under a range of names (Nattier 2007). The various Chinese forms reflect two forms of the name in Sanskrit:, i.e. Avalokitasvara and Avalokiteśvara. The two names and the relative chronology were first noticed by Nikolaĭ Dmitrievitch Mironov (1927). The principal Chinese forms are:

• 廅樓亘  (Èlóuxuān). “Sound-Observer”
• 闚音      (Kuīyīn) “Sound-Observer”
• 見音聲  (Jiànyīnshēng) “He Sees Sounds” 
• 光世音  (Guāngshìyīn) “Sounds of the World of Light”
• 觀世音  (Guānshìyīn) “He Observes Sounds of the World”

Èlóuxuān 廅樓亘 might have been an attempt at a transliteration, perhaps of an even more primitive form of the name, i.e. Avaloka-svara. A possibility Nattier did not consider was a Prakrit form of the name: avalokita-svara in Pāḷi would be spelled olokita-sara.  The Gāndhārī form of the name is Ologispara.* 

* The Gāndhāri Dictionary) lists Olo'iśpare as representing Avalokeśvara (i.c. avaloka-īśvara). However, Salomon and Schopen (2002) have cast doubt on this reading of the inscription without being able to clarify what the correct reading should be. It is probably the locale the donor lived in. 

Up to about the 6th Century, Chinese translators were evidently encountering avalokita-svara since the translations all refer to having "observed" (avalokita) a "sound" (svara). This has a flavour of synaesthesia about it and I'm not aware of any convincing explanation of the name that deals with the fact that one does not usually observe sounds, one hears sounds and observes visual phenomena. The "spelling" Guāngshìyīn 光世音 is probably the result of having misheard the name as ābhā-loka-svara "light world sound".

It's sometimes suggested that the name Guānshìyīn 觀世音 was shortened during the Tang to Guānyīn 觀音 after the death of Emperor Taizong (r.  626 to 649), to avoid the wordshì 世 from his personal name 李世民 Lǐ Shìmín. Such taboos were common in China after Emperors died. However, the practice of shortening the name began long before the birth of Lǐ Shìmín. For example, Kumārajīva frequently uses the two character name in his Lotus Sutra translation (T. 262) dated 403 CE. In any case, the taboo usually required a substitution rather than a simple excision. For example, in some expressions 世 shì was substituted with 代 dài (Kroll 2015: 73).

From the 6th Century translations of Bodhiruci onwards, a new form of the name started appearing in which the word svara was replaced with īśvara "Lord, Master". These include Guānshìzìzài 觀世自在 (“Sovereign of the Observed World”), and Xuanzang's translation Guānzìzài 觀自在 (“Sovereign of the Observed”). Since Avalokitasvara absorbed some of the iconography of Śiva around this time it is assumed that he also absorbed one of Śiva's principle epithets, Maheśvara "Great Lord" to become Avalokita-īśvara (a-ī > e). Alexander Studholme includes a detailed discussion of the relationship between Avalokiteśvara in the Kāraṇḍavyūha Sūtra and Śiva Mahēśvara as we meet him in the Purāṇa texts (2002: 37 ff) 

In translating Xuanzang's Travelogue, Samuel Beal explained Xuanzang's choice of characters for the name. Beal correctly back translated 自在 as īśvara, but understood it to mean “self-existent” and interpreted its meaning as “god.” This apparently influenced many later interpretations of the name. In fact Sanskrit īsvara more straightforwardly means "Lord" or "Master" from √īṣ, which is related to the PIE root *aik- "be master of, possess." Cognate words are English "own" (as in possess) and "owe" and German "eigen". In the Chinese Āgama texts, 自在 simply means "master". Reading Chinese Buddhist texts without reference to the Indic sources can lead us astray, even when they are composed in Chinese.

Sanskrit texts and fragments noted by Mironov (1927) confirm that the name starts off as Avalokitasvara and transforms into Avalokiteśvara at some point. This change is more recently documented by Jan Nattier (2007) and Seishi Karashima (2016). It has also been noted by all and sundry that the latter name never caught on in China where the, now female, figure is still principally known as Guānyīn 觀音. The gender-change came much later than the period that interests me. 

What is not much discussed is the kind of compound that the words avalokita-svara and avalokita-īśvara might be. It is important to note that avalokita is a part participle, i.e. "seen, viewed, observed", not "seeing, viewing, or observing"; it comes from a root √lok meaning "look", i.e. it is rooted in the visual sense. The Chinese translation, guān 觀, also means "observe, consider"; the character combines the semantic radical xiàn 見 meaning "see" with a phonetic radical guàn 雚. As far as I can tell, few of the common  English translations correspond to possible grammatical analyses of the compound. The form avalokita-īśvara seems obviously to be a tatpuruṣa, "Lord of the seen [world]" or "Lord with [a compassionate] gaze."

However, avalokita-svara could be any of: "viewed sound" (karmadhāraya), "sound of the seen" (tatpuruṣa), or "whose sound is observed" (bahuvrīhi). None of these particularly makes sense to me, but then none of the traditional explanations follow the rules for interpreting Sanskrit compounds. There are certainly folk etymologies that sound plausible, but if you approach the compound from a purely grammatical point of view, then this name is strange.

One possibility is that the name was not composed in Sanskrit, but in Prakrit. So svara could be a wrong Sanskritisation of a Prakrit word. We know several examples of this (e.g. bodhisatva, sūtra, mahāyāna). We might note for example that Skt svara is Pāli sara "sound, voice". But Pāli sara is also Sanskrit:

• śara "reed, arrow" 
• śara "going" (√sṛ) 
• saras "lake"
• sara "remembering" (√smṛ)

Another root, √śṛ "crush", might also have given rise to sara (but this is not listed in the PTS Dictionary). So Pāli sara could stand for Sanskrit words śara, saras, sara, or svara. And only context can disambiguate them. With a name, the context could easily remain ambiguous. For example avalokitaśara "the one whose going is observed" is not entirely stupid as a name. The same root also gives us P. saraṇa "refuge" as in saraṇagamana "going for refuge", which could give the name a Buddhist flavour. I'm not saying this is the answer, but I am saying that answers we do have don't make much sense and this might be a way to seek a better explanation.

Roles

Nattier (2007) further summarises the roles that Guanyin tends to play in these early Mahāyāna texts:

1. A passive audience member. The name Guanyin crops up in lists of those present when doctrines are preached. 
2. As Èlóuxuān, the bodhisatva becomes an object of devotion. This is unusual because usually texts admonish us to become bodhisatvas, not the worship them. Paul Harrison has suggested that this role may be a Chinese invention. 
3. Receives a prophecy to Buddhahood.
4. Successor to Amitābha.

His first significant appearance is the Larger Sukhāvativyūha Sūtra. He was popularised in the translations of Dharmarakṣa, especially the Saddharmapuṇḍarīka Sūtra. In none of these texts does Guanyin play the active role of teacher. In other words, what should surprise us about the appearance of Guanyin in the Heart Sutra is not that its a so-called wisdom text, since that compassion/wisdom distinction is anachronistic in relation to bodhisatvas at that time.  Of course, Guanyin is associated with compassion in the sūtras but not exclusively. For example, Kuījī describes him as "possesses wisdom and compassion, universally practices kindness, perpetuates pure lands, and rescues the defiled worlds" (Heng-Ching and Lusthaus 2001: 15. Translating 有具悲智遍行慈愍。紹隆淨剎府救穢方。T. 1710; 33.524c.10).

Before moving on I should say that Kuījī expresses no surprise at finding Guanyin in this text.  He tackles the name in his commentary as though it is just another set of characters. Woncheuk does spend some time establishing that Guanyin is fully enlightened, so we might infer that he uncomfortable about the absence of the Buddha. He notes "There is no introduction or conclusion in this [sūtra]. Since [this text] selects the essential outlines from all the Prajñāpāramitā-sūtras, it has only the main chapter, without introduction and conclusion, just as the Kuan-yin ching (Avalokiteśvara-sūtra) is not composed of three sections." (Choo 2006: 138)

The idea that Guanyin's presence is unexpected may be partly due to expectations that grew up later, perhaps as a result the tantric practice of dividing deities up into demarcated "families". What ought to stand out is the fact that Avalokiteśvara is giving instruction, using words that—in the Large Sutra—were put into the mouth of the Buddha. However, there is a text in which Avalokiteśvara does have such a role and that is Xuanzang's Notes on the Western Regions (西域記 Xīyù jì). Before we turn to this text, we need to consider some generalities about the politics of early medieval China.

The Politics of Buddhism in China

Politics is an important aspect of the historiography of Buddhism in China, especially in Tang China. In order to flourish, in the ancient world, any religion has to negotiate a relationship with state power. There is no right of free of religion, though the Chinese were often tolerant of heterodoxy at this time. This relationship with the state has political, economic, and social dimensions. We may say that, in the ancient world, Buddhism flourishes because of these relations with governments and rulers, if only because monks are economically unproductive and supporting large numbers of them requires surplus wealth. A small community may produce surplus food to feed an extra person or two. But the building of, for example, large monasteries for hundreds of monks to live in one place requires the kind of wealth and resources that usually only states have access to. Rulers expect return on investment, even if that return is an intangible like the promise of a good afterlife. But religion can be a double edged sword, because it comes with obligations, both personal and political. A ruler has to be seen to be pious and to support the institutions of religion. In Tang China, even Taizong gave imperial support to Buddhism though it is clear that he did not like it. 

The dynamic with respect to Buddhism is particularly interesting because of the social structure of Buddhism: the distinction between full-time monks and the devout laity is not absolute. People could move between these two worlds and the monastic sangha was (at least in theory open to anyone). Increasingly, women were excluded from the monastic side Buddhism so that by the Tang women play a marginal role in Buddhist history (with one very notable exception).

The relationship between Buddhist monks and the Chinese state is fascinating because monks are economically unproductive, eschew social norms (especially the Confucian ethos of filial piety), refuse to acknowledge the superiority of the emperor (monks refuse to bow to him), and yet rely on patronage for their existence. Confucians saw Buddhists as deeply immoral for these reasons. While Buddhism did evangelise and attract largesse from the merchant class, it was their appeal to rulers that ensured that Buddhism flourished. This is all the more apparent in the light of periodically anti-Buddhist sentiment and purges such as occurred in China during the Tang. In addition, religious institutions were exempt from paying taxes and so tended to accumulate wealth. Although there are technical restrictions on individual monks from handling money or owning property, in practice Buddhist monasteries in the Tang Capital of Chang'an had incalculable wealth, were involved in usury and commerce, and as a result caused economic imbalances in the Chinese economy. We could see the purges in 845 CE in which the wealth of Buddhist monasteries was appropriated by the state and the scale of Buddhist institutions was drastically reduced (although only briefly) as a rebalancing of the economy. The expansion of Buddhist monasticism is often an economic disaster for the countries in which it happens (more especially where they also capture the reins of governance). 

Those who invest want a good return. In the case of Buddhism, the beneficiaries promise that generosity goes towards good fortune in the present life and a good rebirth for the donor. In a pre-modern world where life and death appear to be entirely a matter of fate, the promise of good fortune and a good rebirth attracted considerable largesse. Buddhists also provided pageantry in the form of large-scale ritual performances. The key to survival in early Medieval China was to have the ruling family on side, and while the Sui Dynasty Emperors had been great supporters of Buddhism, the early Tang Emperors, Gaozu, Taizong, and Gaozong were all indifferent or, in the case of Taizong, hostile, to Buddhism. 

For Xuanzang to be a favourite of the Buddhism-hating emperor, Taizong, then, is a not inconsequential historical fact. In the Biography, Xuanzang is first portrayed as defying the emperor to seek the Dharma in the West, although Kotyk (2019) shows that this defiance may have been invented since the imperial ban on travel was lifted before Xuanzang set out. On his return from the West, Xuanzang is welcomed and feted by the same Emperor (i.e. his defiance has no negative consequences). This is the Buddhist struggle with temporal power in a nutshell. 

We also have to look at the audience for the Biography. Xuanzang's influence as a translator is facet that is often exaggerated. In fact very few works attributed to him were influential except for some of his translations of Yogācāra works for which there were no previous translations. When it came to sūtras, none of Xuanzang's translations displaced those of Kumārajīva from 250 years earlier. The Fǎxiàng 法相 School of Yogācāra Buddhism that he founded lasted only about a century and was never very influential (although Yogācāra per se was very influential). By 688, some 24 years after his death, Xuanzang's lack of influence must have started to be obvious. Yancong's Biography seems to be tuned to giving Xuanzang's remaining followers a boost and perhaps generating some positive PR amongst other Buddhists. It is, however, unlikely that the Biography was widely read outside of Buddhist circles. This circle may or may not have included the Empress Dowager Wu Zhao (her husband Gaozong died in 683 CE) although in 688, Wu Zhao had her hands full suppressing a rebellion by members of the ruling Li family, paving the way to becoming Emperor herself. Wu Zhao was not beyond inducing Buddhist monks to engage in conspiracies to promote the idea of a female emperor. 

Teachable Moments

With this overview, we can now consider the political dimension of Xuanzang's Notes on the Western Regions (西域記 Xīyù jì) and in particular the role played by Guanyin in the story of King Harṣa. Max Deeg (2009, 2012, 2016) has made a considerable contribution to our understanding of the Notes. He highlights and critiques the naive, positivist use of the text, in which everything is taken at face value. He also points out that little or no attempt has been made to position the Notes as one of a genre of Buddhist travelogues (2009: 35-41). In addition, while Xuanzang likely wrote notes for the book, the fincal composition was actually compiled and redacted by Biànjī 辯機. 

Deeg has shown that Xuanzang manipulated his narrative to make it more palatable to Taizong. 

"In the light of Taizong's sensitivity to his own standing, reputation and the impression he would make on future generations, it becomes clear that Xuanzang had to manoeuvre and act quite adroitly to convey the politically and morally critical message directed to his emperor" (Deeg 2016: 98).

As noted above, Deeg (2016) concentrates on King Harṣa. This is partly because Harṣa is quite well documented. We have inscriptions, three plays that are attributed to him, and a biography, Harṣacarita, composed in Sanskrit by Bāṇabhaṭṭa (Bāṇa). So we can directly compare Xuanzang's narrative with the Indian evidence. Deeg argues "that the Indian king is portrayed not as a historical person, but as an idealized Buddhist ruler and—as I have argued elsewhere [Deeg 2009: 51]—as a speculum, or a 'mirror,' held before Taizong." (2016: 100). Xuanzang has two political purposes in the Notes. Firstly to flatter Taizong and secondly to quietly admonish him by presenting kingdoms in Indian in ideal Buddhist terms. And to this end Xuanzang presents Harṣa as relatable, but also as a Buddhist (an ideal Buddhist) king. 

A clearly Buddhist embellishment in Xuanzang’s story is the episode of Avalokiteśvara’s advice to Harṣa to take up the royal or imperial duties without assuming the “lion throne” (shizi zhi zuo 師子之座, Skt. siṃhāsana) and the title “great king” (dawang大王, Skt. mahārāja, or mahārājādhirāja) which does not have any direct correspondence in any of the other sources on Harṣa. (Deeg 2016: 126-7).

In the story of Harṣa, as Xuanzang tells it, the reigning king is killed by a neighbouring kingdom. His son is dead, but his younger brother (Harṣa) is proposed as king instead. This idea is greeted with popular acclaim and the job is offered to Harṣa. However, Harṣa hesitates, protesting that he is hardly qualified and lacks virtue. Something Deeg does not comment on, I think, is that this level of modesty is a Chinese virtue that is not so prominent in India (Compare my discussion of Ajātasattu's meeting with the Buddha. Attwood 2010).

"The public opinion considers (me) suitable (for the throne, but how could forget (my own) shortcomings? Now, at the banks of the river Gaṅgā there is a statue of the bodhisattva Avalokiteśvara, Since it has (already shown) a lot of wonderful signs I wish to go there and ask for advice." (Deeg 2009: 52)

Translating: 物議為宜，敢忘虛薄？今者殑伽河岸，有觀自在菩薩像，既多靈鑒，願往請辭。(T 2087; 51.894.b8-10 ff.)

The Bodhisatva counsels Harṣa to take the job, predicting that because of his previous merit he will be a great king. This is where he advises Harṣa not to assume the “lion throne” (Shīzǐ zhī zuò 師子之座, Skt. siṃhāsana) and the title “great king” (dàwáng 大王, Skt. mahārāja, or mahārājādhirāja). Harṣa takes the throne and eschews the titles, but his first act is to vow vengeance on the neighbours who killed his elder brother. He goes on to conquer them and the rest of India in a sweeping military conquest. 

In his presentation of this material, Xuanzang is at pains to make Harṣa recognizable to Taizong, to make Harṣa a "mirror" for Taizong to see himself.

"The intention in the context of the [Notes] is clear: both rulers are lauded because of their pacification of the realm, the construction of stūpas and monasteries (vihāra), and the convocation of donation parties. This was certainly meant as a propagandistic and 'pedagogical' hint directed to the address of the emperor Taizong..." (Deeg 2009: 57)

Deeg (2016) returns to the Notes and draws out further reasons to think that this is so. For example, he draws parallels between Xuanzang's Harṣa narrative and the facts of how Taizong gained the throne, i.e. by murdering his brothers and the heir apparent, and forcing his father to Abdicate (2016: 125). Deeg notes that there was an ongoing power struggle between Taizong and his chief ministers over who would succeed him. In the end it was Li Zhi , his 9th son, who became Emperor Gaozong in 649 CE. Court factionalism raged on until 655 CE, when Wu Zhao became Empress Consort and decisively brought the still powerful Yang family in on the side of Gaozong (see for example the account in Eisenberg 2012).

In Xuanzang's narrative, Harṣa's older brother is killed and his taking the throne is encouraged by Avalokiteśvara. Taizong had murdered his own brother to take the throne. Deeg seems to argue that Xuanzang is offering Taizong a justification for his fratricide in the form of adopting Buddhist ideals of rulership. But this is achieved indirectly and Taizong is left to draw his own conclusions. Deeg speculates that Taizong might have felt reluctance to assume the throne given his means of ascension. I find this aspect of his account less plausible. A man who murders his brother and forces is father to retire does not seem the type to then have doubts. Taizong is, above all, decisive. However, as Deeg points out (2016: 128) the Harṣacarita does portray Harṣa as reluctant to assume the throne, so perhaps the comparison was intended to flatter Taizong (the man with no doubts). The other parallel between the two rulers is that Harṣa goes on to conquer all of India unifying it under his rule. This was ever the ideal for a Chinese emperor and something that Taizong was quite successful at.

"I think that the narrative of Harṣa's royal lineage and ascension to the throne is directed towards the ruling emperor Taizong—and maybe also towards the ambitious crown prince, and later emperor, Gaozong—as a reminder of the pious and correct behavior of an ideal ruler" (Deeg: 2016: 125).

Although Gaozong is mentioned in passing, and is not prominent in Deeg's articles, it is worth considering that Gaozong was part of the intended audience of the Notes.


Īśvara

Indian records show that Harṣa was not a Buddhist, he was a devotee of the benevolent forms of Śiva, particularly Maheśvara or Paramameśvara. Deeg suggests that if there were an historical event behind the story, that Maheśvara could mutatis mutandis become Avalokiteśvara for Xuanzang's purposes. Especially in Xuanzang's Chinese where the names are Dàzìzài 大自在 and Guānzìzài 觀自在 respectively (128). It is not that Buddhism was entirely foreign to Harṣa, Buddhists were a major presence in India at the time. The Harṣacarita, authored by "stern Śaiva" Bāṇabhaṭṭa, used Buddhist elements in his description of the king. 

For my purposes, what is significant is not simply that Xuanzang has altered the story to serve a political purpose, so much as that he has Avalokiteśvara step outside his role of saviour and protector to become a political advisor. One whose advice led to the annihilation of Harṣa's enemies (who had killed his brother) but which also led to a massive subcontinent spanning war of conquest. The model here, of course is Asoka. The key difference is that Asoka became a Buddhist only after being repulsed by his bloody wars of conquest. Asoka renounced violence to become the ideal Buddhist king, whereas Harṣa embraces violence on the advice of Avalokiteśvara. Taizong was also involved in pacifying remaining pockets of rebellion in the newly reforged Chinese Empire, but was also actively extending the boundaries. 

Summary and Conclusion

Max Deeg has argued that we need to be aware of the political and didactic elements in Xuanzang's Notes on the Western Regions (Xīyù jì 西域記). Focussing on the events that Xuanzang links to the historical figure of King Harṣa, we can see from Indian sources that this story has been changed (by Xuanzang) in ways that can be interpreted as manipulation for political ends. The story has been recast so as to reflect Chinese values. It makes flattering comparisons between Harṣa and Taizong, but reflects Xuanzang's views on ideal governance and regal deportment. Xuanzang is a Buddhist while Taizong is rather unsympathetic to Buddhism. Xuanzang therefore uses the medium of an historical morality take, based on a real story, to get his message across. In this cause, Śaivite Harṣa becomes a Buddhist who consults and receives political advice from Avalokiteśvara, the bodhisatva of compassion. This advice is apparently calculated to resonate with Taizong and to justify his wars of conquest. Xuanzang cannot come right out and chastise Taizong for usurping the throne, but he can show Taizong a way to atone for his usurpation by being a "good" emperor on Xuanzang's terms. This is a rather bold project on Xuanzang's part, but since no one in China at that time had the ability to fact-check his account, it was taken at face value.  It is only now that we can compare the Indian accounts and see the discrepancies.

The Biography by Huìlì 慧立 and Yàncóng 彥悰 portrays Taizong undergoing a deathbed conversion to Buddhism under the skilful guidance of Xuanzang. This is a kind of apotheosis for Taizong, since in embracing Buddhism he becomes in his last few days exactly the ideal ruler that Xuanzang had wanted. It is also the ultimate vindication of Buddhism to bring around the notoriously hostile emperor. However, again, the non-Buddhist Chinese historical sources make it very unlikely that Taizong did convert to Buddhism and few historians accept this account as factual. There is no supporting evidence from non-Buddhist (non-hagiographical) sources and it seems rather too convenient.

Avalokiteśvara is the bodhisatva par excellence in 7th Century China. Maitreya and Manjuśrī are also important but Avalokiteśvara's role in the Lotus Sutra and the Guanyin Sutra make him the most prominent "bodhisatva" in that context. And this alone could explain Xuanzang's use of Avalokiteśvara in the Heart Sutra. However, he had to have known that the words he copied from the Large Sutra were mostly from the mouth of the Buddha and that the principal protagonists of the Prajñāpāramitā were the Buddha, Subhūti, Śāriputra, and Śakra, Lord of the Gods; not bodhisatvas. The instruction in the Large Sutra begins with the Buddha speaking to Śāriputra, although in the Small Sutra the Buddha asks Subhūti to instruct the bodhisatvas in Prajñāpāramitā. 

If Deeg's conjectures about Xuanzang's relations with Taizong and his attempts to create teachable moments in the Notes are right, then the unexpected appearance of Avalokiteśvara might be explained by his role in Xuanzang's narrative of Harṣa. While the Heart Sutra is more or less what it appears to be—i.e. a short summary of Prajñāpāramitā doctrine—and lacks the obvious political overtones of the Notes, the mere reference to Avalokiteśvara could be enough to invoke that earlier narrative for Gaozong and Wu Zhao. Gaozong took the throne more conventionally than his father, although in a parallel to Harṣa, only after his two older brothers had been eliminated (although in this case they but Wu Zhao (if any story about her can be believed) may well have emulated Taizong in murdering rivals for her position and purging the opposition once she gained power.

It's worth emphasising this last point since it is seldom even mentioned: Wu Zhao was not some kind of psychopathic anomaly. She has to be seen in the context of Taizong's murder of his brothers and competitors, his ruthless suppression of opposition, and relentless wars of conquest. Both unexpectedly rose to high office. Both murdered those who stood in their way. Both were astute leaders and politicians.

Kotyk has proposed that Xuanzang composed the Heart Sutra ca February 656 CE as a gift for the birth of Wu Zhao's son, Lǐ Xián 李賢 (29 January 655 – 13 March 684) who would later go on to be Crown Prince. Around the same time, Lǐ Hóng 李弘 (652 – 25 May 675) was made Crown Prince, an event that also gave rise to the founding of Ximing and Jing'ai monasteries (in Chang'an and Luoyang respectively) and to projects to copy the entire Tripiṭaka and to catalogue all Buddhist texts in each city (which bore fruit in 664 and 666 CE respectively).

Xuanzang had to be very careful in expressing his criticism of Taizong. Wu Zhao was already a Buddhist and perhaps more likely target of the gift of a new condensed sutra which emphasised the ephemeral nature of experience than Gaozong. Perhaps Xuanzang felt less comfortable expressing criticism, but still managed to create a pointer back to the Notes by unexpectedly placing Guanyin where he was least expected. The gift of the sutra happened at a time when Wu Zhao had eliminated the most vehement opposition and cemented her grip on power. This did not end the factionalism that had begun during the reign of Taizong, but it was a decisive moment in bringing it to an end. Perhaps in retrospect the naming of Li Hong as Crown Prince is more significant than the birth of Li Xian.

I'm aware that the conclusion here is tenuous. As I revise the received tradition of the Heart Sutra I have to gently remove the layers of accreted myth and legend. What remains is fragmented and partial. It is not yet possible to clearly the shape of it. What is needed is for a qualified, preferably young, Sinologist to take up the enquiry and see what else may be discerned in the Chinese sources by someone with an open mind. As an enthusiastic amateur, who started this adventure far too late in life, I can only go so far with this. There are many questions about the Heart Sutra still to be answered, but we tend not to answer a question before it is asked. If I can contribute anything it is to show that there are open questions.

~~oOo~~

Bibliography

Attwood, J. (2010). "Did King Ajātasattu Confess to the Buddha, and did the Buddha Forgive Him?" Journal of Buddhist Ethics, 15, 279-307.

Deeg, Max . (2009). ‘Writing for the Emperor. Xuanzang between Piety, Religious Propaganda, Intelligence and Modern Imagination’, In Straube, Martin, et al. (eds), Pāsādikadānam. Festschrift für Bhikkhu Pāsādika. Marburg: Indica et Tibetica Verlag, 32–60 (Indica et Tibetica 52).

Deeg, Max. (2012) “‘Show Me the Land Where the Buddha Dwelled …’—Xuanzang’s ‘Record of the Western Regions’ (Xiyu ji): A Misunderstood Text?,” China Report 48 (2012): 89–113.

Deeg, Max (2016). 'The political position of Xuanzang: the didactic creation of an Indian dynasty in the Xiyu ji.' In: Juelch, Thomas ed. The Middle Kingdom and the Dharma Wheel: Aspects of the Relationship between the Buddhist Saṃgha and the State in Chinese History, Vol. 1. (Sinica Leidensia, vol. 133). Leiden, Boston: Brill, pp. 94-139.

Eisenberg, Andrew. (2012) 'Emperor Gaozong, the Rise of Wu Zetian, and factional politics in the Early Tang.' Tang Studies 30, 45-69.

Karashima, Seishi. (2016) “On Avalokitasvara and Avalokiteśvara”, in Annual Report of The International Research Institute for Advanced Buddhology at Soka University (ARIRIAB), vol. 20 (2017): 139-165.

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

Mironov, N. D. (1927). 'Buddhist Miscellanea'. The Journal of the Royal Asiatic Society of Great Britain and Ireland. 2 (Apr., 1927): 241-279.

Nattier, Jan. (2007) ‘Avalokiteśvara in Early Chinese Buddhist Translations: A Preliminary Survey.’ In Magee, W and Huang, Y.H. (Eds). Bodhisattva Avalokiteśvara (Guanyin) and Modern Society. Proceedings of the Fifth Chung-Hwa International Conference on Buddhism, 2006: 191-212. Taiwan: Dharma Drum Publishing.

Salomon, R. and Schopen, G. (2002) 'On an Alleged Reference to Amitābha in a Kharoṣṭhī Inscription on a Gandhārian Relief.' Journal of the International Association of Buddhist Studies. 25(1-2): 3-31.

On the Pronunciation of Jña

One of the pleasures of learning Sanskrit for an English speaker is that it is written almost exactly as it is pronounced. I say almost, because there are one or two irregularities. The most striking irregularity is probably the consonant cluster in jña. This comes out variously as gya, gnya, dya, dnya, and nya and, at face value, there is no rime or reason for this. However, it turns out to be a regional feature influenced by how languages deal with Sanskrit loan words.

In a Buddhist context the word jñā is fairly common as it occurs in technical terms like vijñāna, saṃjñā, and prajñā as well as more general terms like jñāna and jñātavyam. Prajñāpāramitā texts use words deriving from the root √jñā very often. The meaning of jñā is relatively straight forward; it means: "knowledge" (noun) and "to know" (verb). But when one interacts with Sanskritists, one realises that jña is pronounced in a number of more or less counter-intuitive ways, none of which reflect the written word. 

In this essay I will try to explain where these differences come from and try to identify how jña should be pronounced. I will argue that it should, like the rest of the language, be pronounced as written. My evidence will come from two main sources the reconstructed historical phonology of Sanskrit and the historical writing of Sanskrit.

We know that Sanskrit is part of the Indo-European family, an idea first proposed by William Jones in his famous address to the Royal Asiatic Society in 1786:

"The Sanskrit language whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either, yet bearing to both of them a stronger affinity, both in the roots of verbs and in the forms of grammar, than could possibly have been produced by accident; so strong, indeed, that no philosopher could examine them all three, without believing them to have sprung from some common source, which, perhaps, no longer exists."

Note that the relationship is primarily based on similarities of grammar (the conjugations of the verbs and declensions of nouns), but one may also see it in cognate words, such as English know, Greek γνῶσις (gnōsis), Avestan: zan- (v. to know), and so on. Comparative linguistics allows us to compare existing languages and reconstruct the mother tongue of both. In the case of Sanskrit we can trace it back to a common root with Iranian languages, called Proto-Indo-Iranian (PII) , and then one step further to the common root Indo-Iranian and European languages called Proto-Indo-European (PIE). The reconstructed phonology has a logic to it and is internally consistent but no recordings or documents of any kind exist. All that exists are daughter languages.

Phonology

Articulation

We think we have a fairly accurate idea of the phonology of attested ancient languages in the region of Iran and North Western India: Vedic, Old Persian, and Avestan. And we can match this to information about articulation - the physical motions and points of contact of the tongue, mouth, and throat that produce vocal sounds. We can also determine qualities such as whether the motion stopped the air flow completely (if briefly) or only partially. A consonantal stop or affricate requires the tongue to completely block the air flow momentarily, i.e. Sanskrit k g ṭ ḍ t d. A fricative does not completely block the air, i.e. Sanskrit j c v ś ṣ s h and creates a turbulent air flow that changes the timbre of the sound. There is also an approximant which constricts the airflow but not enough to create turbulence.

In addition, a vocal sound may be accompanied by vibrations of the vocal cords or not, which we call voiced (Skt: g j ḍ d b) and unvoiced (Skt: k c ṭ t p). Or the sound may resonate mainly in the nasal cavity which we call nasal (Skt: ṅ ñ ṇ n m).

The /j/ and /ñ/ sounds of Sanskrit are made with the blade of the tongue pressed against the hard palate and are thus called palatal (Skt: c ch j jh ñ). Further forward is the alveolar region which is where the palate merges into the gums. English speakers pronounce /t/ and /d/ by contact of the tip of the tongue on the alveolar. Sanskrit distinguishes a true dental, i.e. tongue touches the teeth (t th d dh n) and a retroflex, i.e. tongue curls back so the tip touches the edge of the hard palate (Skt: ṭ ṭh ḍ ḍha ṇ). Indians tend to hear our English alveolar consonants as retroflex. So "doctor" becomes डक्टर् ḍakṭar. When the back of the tongue contacts the soft palate we call these sound velar region (Skt: k kh g gh ṅ). 

Historical Phonology

The reconstructed articulation of PIE is largely settled, though there are some controversial aspects to it. Fortunately, we don't need to consider the controversial parts. In order to explain the palatal consonants in Vedic, Old Persian, and Avestan languages we need to postulate three palatals in PIE that in fact were pronounced somewhere between the palatal and velar: (1) an unvoiced palatovelar affricate, /k̂/ (2) a voiced palatovelar affricate, /ĝ/ and (3) an aspirated voiced palatovelar affricate /ĝʰ/ (Burrow 1973). This notation is indicative of articulation rather than sound. As I understand it we cannot give precise pronunciations of these using the International Phonetic alphabet (I've found no sources that use IPA to notate PIE phonology).

To get to Sanskrit these sounds underwent two parallel processes of "palatalisation". In most cases these consonants drifted towards the front of the mouth slightly and became fricative. In the first palatalisation (using the notation from Burrow 1973) PIE: k̂ ĝ ĝʰ first became Proto-Indo-Iranian: ć ȷ́ ȷ́ʰ and then ś ź źʰ. At this point the PII language family started to break up.

In the Avestan form of Old Iranian these sounds became s z z. In the Persian form of Old Iranian ś and źʰ became θ and z, while from the earlier form ȷ́ we get (directly) d. Similarly in Vedic ś and źʰ become ś and h, while from ȷ́ we get j. We also see j deriving another way, which is evident in some sandhi differences, but exploring this would take me too far from my aim to explore jñā. The j in jñā derived this way.

Sanskrit jñā is cognate with Avestan zan- and Old Persian dān-. Note the insertion of a vowel between the initial consonant and the nasal. This is something we also see in practice in Sanskrit. The verbal form of the root jñā is jānāti, though we also see jñāta, jñātum, jñātva, and jñāyate.

We can now start to put some more precise values on these sounds using the International Phonetic Alphabet (IPA is conventionally put in square brackets). The "palatals" in Sanskrit are c ch j jh ñ, but in IPA notation are actually alveolo-palatal and in IPA are written: [tɕ] [tɕʰ] [dʑ] [dʑʱ] [ɲ]

If you want to know how these sound, cut and paste the symbol into a search on YouTube. Each one is demonstrated in a short video

Here we are focused on the voiced palatals.

[ʒ] voiced post-alveolar fricative
[dʒ] voiced post-alveolar affricate
[dʑ] voiced alveolo-palatal affricate
[dʑʱ] breathy voiced alveolo-palatal affricate
[j]   voiced palatal approximant
[ɟ]   voiced palatal affricate
[ɲ]  voiced palatal nasal
[nʲ] or [ɲ̟] voiced palatised nasal

We would expect jña to be pronounced [dʑɲɐ] or [ɟɲɐ] (most sources seem to prefer the latter). As noted above, jña is almost never pronounced this way. Ashok Aklujkar points out, in his Sanskrit textbook, that this is partly due to regional variations in India and partly due to Anglisation. Other authors point out that words like jñāna were taken into modern languages as loan words and that this is the source of regional variants.

In the yellow and orange coloured regions, and in Panjab, the pronunciation is /gya/. In the green and red regions and Sindh, it is /dnya/, although in Gujarat it is /gnya/. South India has a mixture. English speakers often opt for /nya/. We can organise these variants: 

IAST
jña	→ g ña (gnya)	→ gya
	→ d ña (dnya)	→ dya
	→ ña (nya)
IPA
[ɟɲɐ]	→ [g nʲɐ]	→ [gʲɐ]
	→ [d nʲɐ]	→ [dʲɐ]
	→ [nʲɐ]

The change from [ɟ] → [g] is a move back in the mouth towards the velar articulation while the change from [ɟ] → [d] is a move forward towards the alveolar. These are clearly different processes and one could not lead to the other. However, with the shift in the initial consonant, the nasal changes from the plain palatal nasal [ɲ] to the voiced palatised nasal [nʲ]. This is why we see it written as "nya". In fact, what happens is the tongue makes contact at the alveolar ridge and rolls back along the palate. In Sanskrit we'd say it was ī blending into a (or a backwards diphthong ai). Phonetically we'd write the vowel change as [ɨ] → [ɐ]. In a sense, then, [nʲɐ]  sounds like [ɲɨɐ].

We can infer that gya [gʲɐ] could be a simplified version of gnya [g nʲɐ]. Thus the Gujaratis who pronounce jñā as gnya [g nʲɐ] are probably conservative with respect to the northern and eastern Indians who have reduced this to gya. Since English speakers have difficulty with [ɟɲə] they simply reduced it to [nʲə]. Pāli generally reduced conjunct consonants to geminate consonants thus jña becomes ñña. But the ñ is pronounced by English speakers as a palatalised nasal not as patal, i.e. as [nʲə] rather than [ɲə], hence names like Nyanaponika (Ñāṇapoṇika) and Nyanamoli (Ñāṇamoḷi).

Since old Persian contains the regular change [dʒ] →  [d] we might suspect from Persian influence on the modern regional pronunciation in parts of India most influenced by the later Persians. However, this does not quite add up since Persian speaking Mughals also ruled in Delhi where /gya/ is standard.

One thing to keep in mind is that in Latin -gn- was pronounced [ŋn], i.e. they did not pronounced gnōsis as /g-nosis/ but as /no-sis/ (like Skt ṅosis). However, the sandhi rules of Sanskrit tend to rule out this possibility.

Sandhi

All spoken languages exhibit changes in pronunciation as words and sounds run together. Some of the languages that use a phonetic writing system notate them and some don't. Sanskrit does and there are a set of rules called sandhi (joining). In classical Sanskrit the rules are compulsory and applied uniformly, but in vernacular languages they became more optional. Madhav Deshpande pointed out to me (in an email) that certain sandhi involving j make it clear that it was heard as [ɟ] even in the conjunct jñ.

The Sanskrit Heart Sutra gives us an excellent example. Towards the end it says "therefore it should be known", in Sanskrit without sandhi: tasmāt jñātavyam; with sandhi tasmāj jñātavyam. The rule is t followed by t becomes j. Similarly in the Mahābharata (179.1) we see  tasmāj jīvo nirarthakaḥ "Therefore this life is useless", and in the Bhagavatā Purāṇa (11.19.5) we see tasmāj jñānena "therefore by this knowledge",

The rules are very specific and became part of the language long before they were codified by Pāṇini ca 4th Century BCE. This tells us that the j in jña was being heard as [ɟ] in Sanskrit when the rule was made or else an exception would be noted.

Chinese

Another potential source of information is Chinese transliterations. Prajñā, for example, was a word that was generally transliterated rather than translated. The transliteration was established early on, by Lokakṣema ca 179 CE, as 般若; however, this was based on Gāndhārī texts. In modern Buddhist circles this is pronounced bōrě although the standard Pinyin transliteration is bānruò. The Middle Chinese reconstruction is complex. There are multiple choices and multiple encoding systems.

• 般 has been reconstructed as /puɑn/ or /pʷɑn/ or /pan/ or /pɐn/.
• 若 has been reconstructed as /ȵɨɐk̚/. The symbol ȵ is mainly used in Sinological circles and represents the alveolo-palatal nasal, i.e. [nʲ].

The Gāndhāri spelling of prajñā was praṃña or praña (c.f. Pāḷi paññā) And thus we can see that the standard Chinese transliteration presents a Prakrit pronunciation without the conjunct. Unfortunately, it was never updated to reflect the Sanskrit pronunciation when Sanskrit texts started to appear. And thus we cannot use the Chinese to reconstruct the received pronunciation at any later period.

Writing

Since Sanskrit is usually pronounced as written we can also look at historical writing to see what it tells us. In modern Devanāgarī the akṣaras for ja and ña combine to form a special sign for the conjunct jña, i.e. ज् + ञ > ज्ञ. Hiralal Rasikdas Kapadia (1936: 289), an expert on Jaina literature, points out that there are two types of Devanāgarī and the Jaina variant as another way of writing jña. In the Jaina variant we get more of a sense of the conjunct. This becomes clearer in old Indian scripts.

Older scripts make a clear distinction between the j and the ña when writing jñā. A good point of reference for us is the Hōryūji manuscript which probably dates from the 9th Century. This manuscript contains a syllabary. The image on the right shows the palatals: ca cha ja jha ña.

We can also isolate the akṣara jñā from this manuscript and see that it has two quite distinct ligatures and is thus written as a conjunction of j and ñā. Note that when written as a ligature the shape of ña is quite different. Also, the scribe has adapted the the basic ja shape in order to indicate the long vowel ā. I will provide some handwritten examples that make the process clearer.

In the collection of akṣaras below I try to show the evolution of ja from Brahmī which gives us two distinct forms of ja, one of which informs the Devanāgarī script and one of which is preserved in Japanese Siddham.

Left to right: Brahmī, Gupta, Siddham (Hōryūji ms.);
upper: Siddham and alternate; jñā (Hōryūji);
lower: Nāgarī, Devanāgarī.

In the formal Siddham script of medieval Japan, the ñā ligature is often written vertically, but this makes for a very tall character. The Hōryūji scribe turns the ligature through 90°. Here are some original and copied examples of jña and jñā. In modern Siddham we often see jña written for jñā.

Some of this is superfluous detail, but it makes the point that until the invention of the (non-Jaina) Devanāgarī script jñā was written as a standard conjunct made of ja and ña. Had the Sanskrit pronunciation been meant to be something else, then the scribes would have used another akṣara.

Note: 18 March 2020. The Creation and Spread of Scripts in Ancient India. Harry Falk

Conclusion

On the grounds of historical phonology, sandhi, and graphology we can say that jña was intended to be pronounced [ɟɲɐ]. The sandhi rule for tasmāt jñātavyam makes it clear that at some point this was the pronunciation and there was no parallel of the Latin pronunciation of [ŋn] for gn. The regional pronunciations are logical developments from an original [ɟɲɐ] and influenced by the use of Sanskrit loan words. However, we don't really know why the pronunciation changed, only that it did. 

For Indian pandits and academic Sanskritists alike, the pronunciation they learned from their teachers will likely be the way they keep pronouncing it. Furthermore, the sound combination [ɟɲɐ] is difficult for English-speakers and they will inevitably tend to drop the [ɟ] and end up with [ɲɐ] or [nʲɐ]. So probably few will follow me in adopting [ɟɲɐ]. And thus all this is for nought. But I find it fascinating and it has kept me amused for a week. It was nice to get my calligraphy pens out again after a long break. In these difficult times, that's not a bad thing, eh? 

~~oOo~~

Bibliography

Burrow, T. (1973). The Sanskrit Language (3rd edition). London: Faber and Faber.

Kapadia, H. R. (1936). 'A Note on Kṣa and Jña.' Annals of the Bhandarkar Oriental Research Institute. 17(3), 1935-36: 289-296.