What Have We Proven About the Heart Sutra?

My thinking on the Heart Sutra was probably due for an overhaul, and I have been fortunate recently to correspond with an academic who has taken an interest in the Heart Sutra. This has led me to rethink some aspects of the "Chinese origins" thesis.

In particular, I have been thinking about what has been proven or (better) established about the history of the Heart Sutra. By "established" here, I think more of the legal criteria of proving something beyond reasonable doubt. I don't mean perfect and eternal certainty, since nothing in ancient history is ever that certain. I mean "established" in the sense that if a fair-minded person with the appropriate background knowledge and skill, would examine the evidence we have presented, they would agree that our methods are both practicable and applicable, and would come to the same, or very similar, conclusions. With the caveat that all conclusions are subject to revision if new information comes along.

So what have Watanabe Shōgo, Jan Nattier, Matthew Orsborn (aka Huifeng), and I actually proven, in this sense? I think three propositions can be regarded as well-established historical facts.

1. Two passages in the Chinese Heart Sutra were copied from the Chinese Large Sutra translation by Kumārajīva (T 223).
2. The Sanskrit Heart Sutra was translated from a Chinese source.
3. The Heart Sutra enters the historical record in a Chinese document dated to 25 Dec 656 CE.

While I have made larger claims in the past, my recent correspondence has prompted me to clarify what seems certain and what is speculative.

1. The Chinese Heart Sutra copied two passages in Chinese
from the Large Text translation by Kumārajīva.

It was common knowledge amongst the early Tang commentators—by Kuījī 窺基 (T 1710), Woncheuk 圓測 (T 1711), Jìngmài 靖邁 (X 522), Fǎzàng 法藏 (T 1712), Huìjìng 慧浄 (X 521) and Zhishen 智詵 (Dunhuang ms.*)—that (at least some of) the text of the Heart Sutra was extracted from another Prajñāpāramitā text, with some of them specifically naming the Large Sutra. So this part of the thesis should be uncontroversial (and as far as I know, it is).

* The text of Zhishen's commentary has appeared in a Japanese publication by Yanagida and was used in a Japanese language study by Cheng (2007). On which see Notes on Zhishen's Heart Sutra Commentary (23 August 2024).

The two passages that we have identified are what Nattier called the "core section" and the "epithets section". Below I put the versions of the core section alongside each other, further broken down into three sections. As Nattier noted, you don't actually need to understand Chinese or Sanskrit at this point, you can follow this argument on pattern recognition alone.

Chinese

When we compare the two versions of the passage in the Chinese Heart Sutra and Kumārajīva's Mohe jīng, they are nearly identical on a character-by-character basis. The two texts have been punctuated differently in CBETA, so I have standardised this to make the comparison easier. I have highlighted the differences.

T 251	T 223
舍利子，色不異空，空不異色， 色即是空，空即是色。 受想行識亦復如是。 舍利子，是諸法空相，不生不滅，不垢不淨，不增不減。 是故空中無色，無受想行識，無眼耳鼻舌身意，無色聲香味觸法，無眼界 乃至 無意識界，無無明亦無無明盡 乃至 無老死亦無老死盡，無苦集滅道，無智亦無得，以無所得故。	舍利弗，非色異空，非空異色， 色即是空，空即是色， 受想行識亦如是。 舍利弗，是諸法空相，不生不滅、不垢不淨、不增不減。 [是空法非過去、非未來、非現在]*， 是故空中無色，無受想行識，無眼耳鼻舌身意，無色聲香味觸法，無眼界 乃至 無意識界，亦無無明亦無無明盡，乃至 [亦]無老死亦無老死盡，無苦集滅道，[亦]無智亦無得，[亦無須陀洹無須陀洹果，無斯陀含無斯陀含果，無阿那含無阿那含果，無阿羅漢無阿羅漢果，無辟支佛無辟支佛道，無佛亦無佛道。] * bracketed parts are omitted from T 251

There are two main types of difference here: omissions and amendations.

Omissions

There are two large omissions: (1) A whole line in the middle is omitted. (2) The final section is cut short, so that, after mentioning zhì 智 "knowledge" and dé 得 "attainment", T 223 carries on to give examples of each. This section is complicated by Kumārajīva appearing to mistranslate na prāptir nābhisamayaḥ “no attainment, no realisation”. The significance of the phrase is still visible in T 223, however, which carries on to say:

亦無須陀洹無須陀洹果，無斯陀含無斯陀含果，無阿那含無阿那含果，無阿羅漢無阿羅漢果，無辟支佛無辟支佛道，無佛亦無佛道。
"There is no stream-enterer, no fruit of stream-entry; no once-returner, no fruit of once-returning; no non-returner, no fruit of non-returning; no arhat, no fruit of arhatship; no pratyekabuddha, no path of pratyekabuddha; no buddha, and no path of buddhahood."

This fits well with the Sanskrit text of the Gilgit manuscript (see below). The "stream-enterer" (須陀洹), "once-returner" etc are examples of prāpti "attainment", while the fruit (guǒ 果) of Stream Entry, etc are examples of abhisamaya "realisation".

We are more used to seeing the terms "path" (marga) and "fruit" (phala) here or, in Chinese, the sìxiàng sìguǒ 四向四果 "four accesses and four realisations". The sìxiàng 四向 are: 須陀洹 "stream-enterer", 斯陀含 "'once-returner", 阿那含 "non-returner", and 阿羅漢 "arhat", and sìguǒ 四果 are the guǒ 果 (fruit) of these. To this list, T 223 adds the pìzhī fó 辟支佛 "pratyeka buddha" and the fó 佛 "buddha". The list in Pañc also has an entry for the bodhisatva and the fruit of bodhisatva-hood.

Amendations

Where the text of T 251 is amended, words and phrases are changed from Kumārajīva's translation idiom to Xuanzang's translation idiom, i.e. Shèlìfú 舍利弗 becomes Shèlìzi 舍利子; and fēi sè yì kōng, fēi kōng yì sè 非色異空，非空異色 becomes sè bù yì kōng, kōng bù yì sè 色不異空，空不異色. The first expression negates the idea that 色 and 空 are different, while the second states that 色 and 空 are not different. It's just two ways of saying the same thing.

Note that my view on the latter change has shifted from what I wrote in Attwood (2021: 25-26). It has become clear from new notes added by CBETA to T 1858 that 非色異空，非空異色 is the version of the phrase in T 223. Thus, this difference appears to be better explained as another case of the text being amended to fit Xuanzang's idiom. See also the discussion on CBETA.

In any case, even without knowing Chinese, we can see that the two Chinese texts are substantially the same and, where they are different, we have a simple explanation of why they are different. As we will see, the relationship between the two Sanskrit texts is very different indeed.

Sanskrit

Below is the core passage in Sanskrit, using the revised edition I published earlier this year and the text of Pañc transcribed by Zacchetti (2005:393) from the Gilgit manuscript (actually one of three, but only one has been published and then only in a facsimile edition). Again, even without knowing any Sanskrit, one can see that the two texts have some very different choices of words and idioms.

Hṛd	Pañc (Gilgit)
[iha] śāriputra rūpameva śūnyatā śūnyataiva rūpaṃ rūpān na pṛthak śūnyatā śūnyatāyā na pṛthag rūpaṃ | evam eva vedanā-saṃjñā-saṃskāra-vijñānaṃ ||	na hi śāradvatīputrānyad rūpam anyā śunyatā nānyā śunyatānyad rūpam rūpam eva śunyatā śunyataiva rūpam evaṃ nānyā vedanānyā śunyatā | nānyā saṃjñā nānyā śunyatā | nānye saṃskārā anyā śunyatā | nānyad vijñānam anyā śunyatā | nānyā śunyatānyad vijñānaṃ | vijñānam eva śunyatā śunyataiva vijñānaṃ |
[iha] śāriputra sarvadharmāḥ śūnyatālakṣaṇā anutpannā aniruddhā amalā avimalā anūnā aparipūrṇāḥ ||	yā śāradvatīputra śunyatā na sā utpadyate na nirudhyate | na saṃkliśyate na vyavadāyate | na hīyate na vardhate | [nātītā nānāgatā na pratyutpannā |]*
[tasmāc chāriputra śūnyatāyāṃ]* na rūpaṃ na vedanā na saṃjñā na saṃskārāḥ na vijñāna,	na tatra rūpaṃ na vedanā na saṃjñā na saṃskārā na vijñānaṃ
na caksuḥśrotraghrāṇajihvākāyamanāṃsi; na rūpaśabdagandharasaspraṣṭavyadharmāḥ	na cakṣur na śrotraṃ na ghrāṇaṃ na jihvā kāyo na manaḥ na rūpaṃ na śabdo na gandho na raso na sparśo na dharmāḥ
	[na tatra skandhā na dhātavo nāyatanāni]*
na cakṣurdhātur yāvan na manovijñānadhātuḥ	na tatra cakṣurdhātur na rūpadhātur na cakṣurvijñānadhātur na śrotradhātur na śabdadhātur na śrotravijñānadhātuḥ na ghrāṇadhātur na gandhadhātur na ghrāṇavijñānadhātur na jihvādhātur na rasadhātur na jihvāvijñānadhātuḥ na kāyadhātur na spraṣṭavyadhātur na kāyavijñānadhātur na manodhātur na dharmadhātur na manovijñānadhātur
nāvidyā nāvidyākṣayo yāvan na jarāmaraṇaṃ na jarāmaraṇakṣayo	na tatrāvidyā nāvidyānirodhaḥ na saṃskārā nna saṃskāranirodhaḥ na vijñānaṃ na vijñānanirodhaḥ na nāmarūpaṃ na nāmarūpanirodhaḥ na ṣaḍāyatanaṃ na ṣaḍāyatananirodhaḥ na sparśo na sparśanirodhaḥ na vedanā na vedanānirodhaḥ na tṛṣṇā na tṛṣṇānirodhaḥ nopādānaṃ nopādānanirodhaḥ na bhavo na bhavanirodhaḥ na jātir na jātinirodhaḥ na jarāmaraṇaṃ na jarāmaraṇanirodhaḥ
na duḥkhasamudayanirodhamārgā	na duḥkhaṃ na samudayo na nirodho na mārgaḥ
na jñānam na prāptiḥ || * not present in Pañc	na prāptir nābhisamayaḥ [na srotaāpanno na srotaāpattiphalaṃ na sakṛdāgāmī na sakṛdāgāmiphalaṃ nānāgāmī nānāgāmiphalaṃ nārhan nārhatvam na pratyekabodhir na pratyekabuddhaḥ na tatra mārgākārajñatā na bodhisatvaḥ na tatra bodhir na buddhaḥ]* *bracketed sections omitted from Hṛd

One of the main developments within the Prajñāpāramitā literature was the expansion of abbreviated passages. So for example, an abbreviated list of dhātu found in Hṛd: na cakṣurdhātur yāvan na manovijñānadhātuḥ (where yāvan means something like et cetera or "and so on, up to"). In Pañc we get the unabbreviated list:

na tatra cakṣurdhātur na rūpadhātur na cakṣurvijñānadhātur
na śrotradhātur na śabdadhātur na śrotravijñānadhātuḥ
na ghrāṇadhātur na gandhadhātur na ghrāṇavijñānadhātur
na jihvādhātur na rasadhātur na jihvāvijñānadhātuḥ na
kāyadhātur na spraṣṭavyadhātur na kāyavijñānadhātur
na manodhātur na dharmadhātur na manovijñānadhātur.

This kind of difference is not problematic, since we know our witnesses of Pañc are ~250 years later than the version that Kumārajīva had. And this kind of change is exactly what we expect. Moreover, the general gist of the Gilgit passage is identical in the later Nepalese manuscripts edited in the Kimura edition of Pañc, with only minor differences.

The point I wish to make is that, even allowing for expansion over time, the core passage found in Hṛd cannot have been copied from Pañc or any text like Pañc. I will look at the details of some of the differences below, but here we are simply looking at whether Hṛd could have conceivably been copied from a version of Pañc (likely one with more abbreviations). If you are comparing them, focus on the first two parts above. See how the words and phrasing are considerably different,

There's only one plausible conclusion here since only one chain of events can explain this pattern of differences. The copying cannot have happened in Sanskrit because the Hṛdaya is clearly not a copy of any known text of Pañc; nor could it be a copy of some unknown text affected by known processes such as the expansion of abbreviated passages. The copying happened in Chinese.

An argument can be made that "any known text" is too narrow. Some people appear to believe that a version of the Large Sutra text must have existed in which the copied passage is the same as the Heart Sutra. It is true that these texts changed over time. However, it's they did not change at random. Rather the Large Sutra changed in relatively predictable ways, by far the most common was the expansion of abbreviated passages. Importantly, however, no matter what changes we find between versions, the whole text of Pañc was always in idiomatic Buddhist Sanskrit.

The changes we find between Sanskrit Heart Sutra and Sanskrit Large Text don't fit such known patterns of change. The Sanskrit Heart Sutra frequently diverges from well-known Sanskrit idioms. If it had to draw from a Sanskrit original, then we have to posit a whole lineage with nonstandard locutions which spawned their own lineage of Prajñāpāramitā but died out before anyone noticed them, but not before the Heart Sutra was composed and left no trace of its existence in India. Even if we didn't already have a much simpler answer, this scenario is far too convoluted to be plausible.

Arguments from absence are weak, but especially weak when the absent thing has been invented in the absence of evidence in the first place. Occam's Razor counsels us against inventing new entities to explain our evidence.

Everyone accepts that the copying took place. We know that Chinese Buddhists had already been copying passages from Buddhist texts for centuries. They copied passages to make inscriptions, for example, and to make handy summaries of long texts, or to carry around with them for the putative magical benefits. We also know that extracts of texts—either in the form of digest texts (chāo jīng 抄經) or parts of texts that circulated independently (bié shēng 別生) such as the Guānyīn jīng «觀音經» (a chapter of the Miàofǎ Liánhuá jīng «妙法蓮華經» or Lotus Sutra)—were common. Copying passages is an everyday routine operation in Chinese Buddhism. There is absolutely nothing strange about Chinese Buddhists copying passages.

The simplest solution is the best solution: the copied passages were copied in Chinese. This is straightforward, historically plausible, and explains all the facts that we have.

Of course, it is tempting to expand on this, and in the past, I have done so without much hesitation. But such expansions are conjectural. It is logical to assume that if one part of the text was composed in Chinese then the rest of it must have been also. But this is not proven, this is an open conjecture.

The likelihood is that the whole text was composed in Chinese. The fact is that the copied passages were copied in Chinese. Since these copied passages amount to about half the text, we have also proved that this half of the text was composed in Chinese.

2. The Sanskrit Heart Sutra was translated from a Chinese source.

We now shift focus to certain details of the Sanskrit texts, comparing the idioms in Hṛd with its parallels in Pañc with the aim of showing that the Sanskrit text in Hṛd is more plausible as a translation of a Chinese source.

Nattier (1992: 170) noted some diagnostic criteria for back-translations. These include “unmatched but synonymous equivalents” for some Sanskrit terms and “incorrect word order, grammatical errors that can be traced to the structure of the intermediary language, and incorrect readings (due to visual confusion of certain letters or characters in the intermediary language)”.

2.1. The first passage to look at is part of the famous phrase usually translated as "form is not different from emptiness".

Hṛd	Pañc (Gilgit)
rūpān na pṛthak śūnyatā śūnyatāyā na pṛthag rūpaṃ	nānyā śunyatānyad rūpam rūpam eva śunyatā śunyataiva rūpam

Again, even without a detailed knowledge of Sanskrit, one can see that, apart from certain jargon terms, both the words and the phrasing of Hṛd are quite different from Pañc. And this is definitely not an example of the kind of change that we expect to see over time. The manuscripts show that this type of idiom is stable over time and Prajñāpāramitā has a very strong preference for one over the other.

Hṛd has the syntax Y na pṛthak X, where Y is a noun in the ablative case (indicating "from Y") and na pṛthak means "not different". Hence, Y na pṛthak X means "X is not different from Y" (though English translations often get the word order wrong). This expression is not used in Pañc or Aṣṭa, though contra my previous assertions, pṛthak does occur as a standalone term in Pañc. Rather such comparisons are made using the formula: na anya X anya Y. Here anya "other" is a pronoun. A literal translation is not that helpful, but the idiom means "X is not other than Y". Sometimes I translate this as "It's not that X is one thing and Y another".

The two texts use different idioms to say the same thing. More than this, however, the idiom found in Pañc is widely used in Prajñāpāramitā literature, and other idiom is found only in Hṛd (To the best of my knowledge). As noted, this means that the copying happened in Chinese. The synonymous expression with unidiomatic wording is Nattier's "leading indicator" of a back-translation.

The Chinese expression fēi sè yì kōng 非色異空 is a conceivable translation of either rūpān na pṛthak śūnyatā or nānyā śunyatānyad rūpam.

2.2. The second example shows a much greater set of differences, not all of which are simply a choice of idiom.

iha śāriputra sarvadharmāḥ śūnyatālakṣaṇā anutpannā aniruddhā amalā avimalā anūnā aparipūrṇāḥ ||

yā śāradvatīputra śunyatā na sā utpadyate na nirudhyate | na saṃkliśyate na vyavadāyate | na hīyate na vardhate

We need not dwell on the two different forms of the name—Śāriputra and Śāradvatīputra—since these are known to vary in Mahāyāna texts, meaning that this is not a diagnostic criterion.

The subjects of the two phrases are different. The sentence in Hṛd is about "all phenomena" (sarvadharmāḥ), while the sentence in Pañc is about "absence" (śūnyatā). Hṛd is generally read as saying that "all phenomena" are qualified by a series of three pairs of adjectives, necessarily in the plural. In Pañc, by contrast, "absence" is qualified by a series of three pairs of verbs in the singular present indicative. However, it is also possible to read Hṛd as saying that all dharmas are qualified by (-lakṣana) an absence (śūnyatā) and that absence is unarisen, unceased, etc.

And in the light of Pañc, I think this latter reading is preferable, but it doesn't change the point here.

The set of six adjectival qualifiers in Hṛd is found nowhere else in Prajñāpāramitā literature. One pair, anutpannā aniruddhā does occur elsewhere, but the other two adjectives are not found paired anywhere in the literature. By contrast, the set of six verbs does occur elsewhere in Prajñāpāramitā, and all of the pairs often occur independently as well.

Again, we have two distinct idioms. The idiom of Hṛd is confined to Hṛd alone, and the idiom in Pañc is simply a standard Buddhist Sanskrit idiom. As Nattier (1992: 170) pointed out, this pattern of differences in Hṛd bears all the hallmarks of a “back-translation”.

2.3 Huifeng (2014) noted, amongst other pertinent observations that yǐwúsuǒdégù 以無所得故 was Kumārajīva's translation of anupalambhayogena, whereas Hṛd has aprāptivāt.

While aprāptitvāt looks plausible based on the idea that dé 得 "attain" was used to translate Sanskrit words from pra√āp "attain", in fact, closer analysis shows that the verb here is a binomial, suǒdé 所得, which both Kumārajīva and Xuanzang use to translate words from √labh "obtain", particularly upa√labh "apprehend". In Sanskrit to apprehend an idea is to grasp it, obtain it, and understand it mentally. The cognitive metaphor here is IDEAS ARE OBJECTS (we have this same cognitive metaphor in English also). Once we make this equivalence we can metaphorically apply to ideas any action that applies to objects.

This word anupalambhayogena also provides an important key to understanding Prajñāpāramitā and forms the basis of my exegesis of the text (Attwood 2022)

2.4 I've made two contributions to this part of the project. Attwood (2017) shows, following a note in Nattier's article, that the original vidyā becomes mantra, with the Chinese míngzhòu 明呪 or zhòu 呪 providing the intermediary. And Attwood (2018) showed that the Sanskrit phrase tryadhvavyavasthitāḥ sarvabuddhāḥ "all buddhas of the three times" is a calque of the Chinese Sānshì zhū fó 三世諸佛, where we expect: atītānāgatapratyutpanna "past, future, and present" (Attwood 2018).

2.5 Compare also Hṛd nāvidyā nāvidyākṣayo and Pañc: na tatrāvidyā nāvidyānirodhaḥ. For me, this is the most iconic difference. We know that avidyā "ignorance" is the root condition for the whole nidāna sequence in the forward (anuloma) direction that leads to birth and death. In in the reverse direction, avidyā-nirodha the cessation of ignorance is the final step in liberating oneself from birth and death. And Indian Buddhist texts universally (across Pāli, Sanskrit, and Gāndhārī), with only one exception, speak of avidyā-nirodha. The exception is the Sanskrit Heart Sutra, which has avidyā-kṣaya.

Note that although Harada Waso (2002) claims to have found a discussion of avidyā-kṣayatva in the Aṣṭasāhasrikā, in fact, the whole passage he cites concerns the opposite of kṣaya, namely akṣaya "indestructible". (Cf Conze's translation 1973: 271 where he translates akṣaya as "inexhaustible").

Conclusion

There is a pervasive pattern of differences between Hṛd and Pañc. Not only is Hṛd not a copy from Pañc, Hṛd is largely an unidiomatic paraphrase of Pañc, where only well-known (and stable) Buddhist technical terms are unchanged. This is consistent with Hṛd being a back-translation from a Chinese Heart Sutra. It is not consistent with any scenario in which the Hṛd was composed in India or in Sanskrit. Moreover, in each case, the Chinese text is a plausible intermediary between the two different Sanskrit phrases and explains why the Sanskrit in the Heart Sutra is so peculiar.

However, there are a number of passages that make the source text uncertain. For example, in the first sentence, there is no Sanskrit counterpart of the Chinese expression dù yīqiè kǔ è 度一切苦厄. The omission is problematic, but in such a way as to complicate the picture, not as a refutation.

It is, therefore, a well-established historical fact that the core section and the epithets sections of Hṛd were translated into Sanskrit from Chinese. We have reason to be confident that the rest of the Sanskrit text is translated from Chinese as well since neither anupalambhayogena nor tryadhvavyavasthitāḥ sarvabuddhāḥ occurs in the core or the epithets section.

3. The Heart Sutra enters the historical record in a Chinese document dated to 25 Dec 656 CE.

Establishing this proposition as a fact is more difficult because we have conflicting historical records and scholarly opinions. The majority of scholars put the composition of the Heart Sutra in the third or fourth century CE, but we regularly see other dates from the first to the seventh century CE.

However, unbeknownst to the English-speaking world, the year before Nattier published her paper, the Japanese scholar Watanabe Shōgo (1991) had already made a huge contribution. Using entries in Chinese catalogues of Buddhist texts in translation, Watanabe was able to show that we can discount all the stories of older or lost translations. The oldest text we know of is T 251.

This is confirmed by two surveys of the oldest Heart Sutra inscriptions in China: Wang (2010) and He & Xu (2019). Of the two dozen or so Tang Dynasty Heart Sutra inscriptions, all have the text of T 251, though many have minor character substitutions that don't change the meaning or the pronunciation. This is followed by a rapid expansion in the number of inscriptions, commentaries, catalogue entries, and other kinds of evidence in China. These surveys confirm that the first physical evidence of the Heart Sutra is the stele from Fangshan dated 13 March 661 CE.

There is a mention of the title in Xuanzang's biography by Huili and Yancong (T 2053), in a letter from Xuánzàng 玄奘 (602–664) to Emperor Gāozōng 高宗 (15 July 649 – 27 December 683), dated 25 December 656 CE. Since Jeffrey Kotyk (2019) has shown that this letter is independently preserved elsewhere, we can credit this as the earliest mention of the title of the Heart Sutra, though note that this tells us nothing about the content of the text mentioned.

There is no physical evidence of the Heart Sutra from India, not a single scrap of manuscript has ever been identified (and given how idiosyncratic Hṛd is, a scrap with, say, avidyākṣaya on it would be sufficient). No other kind of evidence has turned up either: the title Prajñāpāramitāhṛdaya is never mentioned in any canonical or post-canonical Indian text, including several anthologies made using extracts of popular sutras. Absence of evidence is not evidence of absence, but if the Heart Sutra were an Indian text we would strongly expect to at least find some mention of it somewhere.

The Heart Sutra next turns up in the late eighth century in both Dunhuang and Tibet. This was at a time when Tibet ruled Dunhuang. While seven commentaries in Tibetan are credited to Indian Buddhists, some of these are clearly commenting on a Tibetan text (Horiuchi 2021). And the attributions have never been tested. Is the commentary attributed to, say, Kamalaśīla, consistent with other works known to have been composed by him? We don't know. And I suspect that when someone finally looks at this, they will find reasons to doubt all of these attributions. Part of the problem with these commentaries is that none of them includes a whole embedded text.

Thus in terms of the historical record, the first reliable evidence of the Heart Sutra's existence is the mention of the title by Xuanzang, on 25 December 656 CE. The first evidence of the content of the Heart Sutra is the Fangshan stele from 13 March 661 CE.

I have argued that the Heart Sutra cannot have existed before 654 CE when the Tuóluóní jí jīng «陀羅尼集經» (T 901). was translated by Atikūṭa. This is based on the idea that the Heart Sutra copied the dhāraṇī from this text, though my academic interlocutor has argued for an alternative conjecture that is not implausible. I therefore do not include this as an established fact, until we can get some clarity on this issue.

Conclusion

Based on these methods applied to this evidence, then, we can say that Watanabe, Nattier, Orsborn, and I have established three historical facts beyond reasonable doubt:

1. Two passages in the Chinese Heart Sutra were copied from the Chinese Large Sutra translation by Kumārajīva (T 223).
2. The Sanskrit Heart Sutra was translated from a Chinese source.
3. The Heart Sutra enters the historical record in a Chinese document dated to 25 Dec 656 CE.

If anyone has a genuine way to dispute these facts I have yet to see it. I understand that many Buddhists find these facts unpalatable, but this is insufficient to falsify the facts. Polemics by Fukui, Harada, and Ishii make no dent in these facts.

To be clear, I think these facts are falsifiable. One simply has to find, for example, a manuscript or scrap of manuscript of the Sanskrit Heart Sutra from before the seventh century, or a much older Chinese witness. Or a previously unknown version of Pañc that is unidiomatic in the way that Hṛd is. Such evidence would be near miraculous, but if it turned up, we'd have to rethink things.

However, a historian composes a history based on the evidence at hand, not based on what might turn up in the future. "Black swan" events happen all the time, but the whole point of them is that we cannot predict them.

I believe that any open and fair-minded person would come to the same conclusion if they examined the evidence that we have published to date, using the (simple) methods that Nattier developed. Moreover, we now have quite a bit more evidence than Nattier had, and we have better access to the Sanskrit Pañc witnesses (Kimura's edition of the Nepalese and Karashīma et al's new facsimile edition of the Gilgit ms.)

These facts are consistent with comments made by the early Tang commentators, who were all associates of Xuanzang, but all took distinctly different approaches to the text. However, these facts are not consistent with the well-documented myth of the Heart Sutra. And so we are left with some mysteries: who created the Heart Sutra in the first place and for what purpose? Who translated the text into Sanskrit? Who added the extensions (and when and where)? We can fill some of these gaps with plausible conjectures, but in this essay, I wanted to focus on historical facts that can be considered established.

We could wish for more secondary literature on back-translations that was relevant. The problem is that back-translation has become a major tool for assessing the validity of translations, especially in sensitive areas such as translated medical advice pamphlets and for assessing machine translations. The literature on the application of back-translation is extensive, but I cannot find any other literature on back-translations in a situation such as we have with the Heart Sutra (I haven't given up, but wading through hundreds of irrelevant database hits is time-consuming and boring).

~~oOo~~

Bibliography

Dà Táng dà Cí’ēnsì sānzàng fǎshī chuán xù «大唐大慈恩寺三藏法師傳序» A biography of the Tripiṭaka Master of the Great Ci’en Monastery of the Great Tang Dynasty (T 2053). Translated into English by Li Rongxi (1995).

Attwood, Jayarava. (2017). "‘Epithets of the Mantra’ in the Heart Sutra." Journal of the Oxford Centre for Buddhist Studies 12: 26–57. http://jocbs.org/index.php/jocbs/article/view/155

Attwood, Jayarava. (2018). "The Buddhas of the Three Times and the Chinese Origins of the Heart Sutra." Journal of the Oxford Centre for Buddhist Studies 15: 9-27.

Attwood, Jayarava. (2021) "The Chinese Origins of the Heart Sutra Revisited: A Comparative Analysis of the Chinese and Sanskrit Texts." Journal of the International Association of Buddhist Studies. 44: 13-52.

Attwood, Jayarava. (2022) "The Cessation of Sensory Experience and Prajñāpāramitā Philosophy" International Journal of Buddhist Thought and Culture 32(1):111-148.

Cheng, Zheng. 程, 正 [テイ, セイ]. (2007). “智詵撰『般若波羅蜜多心經疏』の譯注研究 (2)”. 駒澤大学仏教学部研究紀要 65: 139-156. http://repo.komazawa-u.ac.jp/opac/repository/all/31124/

Horiuchi, Toshio. (2021). “Revisiting the ‘Indian’ Commentaries on the Prajñāpāramitāhṛdaya: Vimalamitra’s Interpretation of the ‘Eight Aspects’.” Acta Asiatica 121: 53-81.

Huifeng, Shi. [aka Orsborn, M.] (2014). "Apocryphal Treatment for Conze’s Heart Problems: “Non-attainment”, “Apprehension”, and “Mental Hanging” in the Prajñāpāramitā Hṛdaya." Journal of the Oxford Centre for Buddhist Studies 6: 72-105.

Nattier, Jan (1992). "The Heart Sūtra: a Chinese apocryphal text?" Journal of the International Association of Buddhist Studies. 15 (2) 153-223.

Watanabe, Shōgo 渡辺章悟 (1991). 経録からみた『摩訶般若波羅蜜神呪経』と『摩訶般若波羅蜜大明呪経』」印度学仏教学研究 39-1: 54–58. [“Móhē bānrě bōluómì shénzhòu jīng and Móhē bānrě bōluómì dàmíngzhòu jīng, As Seen in the Sutra Catalogues.” Indogaku Bukkyōgaku Kenkyū 39-1: 54–58.]

Notes on Finding Buddhists in Global History

Jonathan Walters was kind enough to correspond with me and send me a pdf copy of his extended essay in book form: Finding Buddhists in Global History (1998). The book has three chapters

1. Problematizing the Buddhological Construct
2. Historicizing the Buddhological Construct
3. Beyond the Buddhological Construct

Walters is a historian, and thus his work can be bracketed with other historians of Buddhism such as David Drewes, Bernard Faure, Charles Hallisey, Frank Reynolds, Gregory Schopen, and Joseph Walser. All of these historians have been critical of how the history of Buddhism has been handled within Buddhist Studies. Note that the book and its argument are aimed at historians who wish the see Buddhists in their global context. Other approaches to history are available and valid.

What follows are my notes on reading this essay (thus my understanding of Walter's argument). Numbers in parentheses, e.g. (4), are page numbers in Walters 1998. My comments are in [square brackets].

Note that the preface has footnotes but the main text uses endnotes.

Problematizing the Buddhological Construct

The Buddhological Construct is a master narrative of Buddhist history in five key moments:

1. The historical Buddha founded Buddhism
2. Early Buddhism
3. Asoka and general propagation through Asia
4. Medieval decline
5. Early modern revival (19th and 20th Centuries)

Within this framework, various Buddhological arguments have raged over:

1. The dates of the Buddha and the authenticity of the texts
2. The character of early Buddhism
3. The chronology of Buddhist missions
4. The causes of medieval decline
5. The details of the modern revival

This master narrative was a product of nineteenth-century European colonialism (5). It produced massive overviews of Buddhism [such as Etienne Lamotte's History of Indian Buddhism]. Since then Buddhological scholarship has become increasingly specialized, but it has continued to presuppose the master narrative. Walters is critical of this narrative:

"The Buddhological Construct is a crude and misleading interpretation of the evidence, and we now need to move beyond it... the Buddhological Construct obscures [the global significance of Sri Lankan texts and monuments] with pat generalizations that have little or no relationship to the evidence itself." (4)

Rather than being updated due to more specialized studies that undermine it, the Buddhological Construct has remained in force and become a dogma.

Historicizing the Buddhological Construct

Many of the ideas and concepts that seem basic to Buddhist Studies—such as "Buddhism" and "the historical Buddha"—were invented in the period 1820–1840. By the 1840s Europeans had discovered that Buddhism was a religion, though they saw it as "moribund" [i.e. dead] (8). In the 1850s, Buddhists themselves were challenging the idea that Buddhism was moribund.

19th-century scholars were far from systematic in their approach.

Scholars did not begin by deliberating about what should be asked of the primary evidence, which questions should be answered on the basis of which evidence, nor what sort of evidence it was in the first place. Instead, being thinkers of their age, these scholars dived right into the evidence as though its nature and the questions to which it represented answers should be transparent and obvious. (8).

The Buddhological Construct reflects 19th-century preoccupations. Walters suggests that it "was little more than the extrapolation to Buddhist history of 19th-century Protestant historical self-understanding." (9). So the five key moments of Buddhist history are modeled on a Protestant understanding of the history of Christianity, i.e. founder, original church, missions, decline, and revival (9). The history of Buddhism was shoehorned into this model by 19th-century Buddhologists.

Walters suggests that this outline of history was so ingrained in 19th-century European scholars that it was presupposed before they ever opened a Buddhist text. And the "discovery" of this same pattern in Buddhist history was more or less inevitable. This led to subsequent questions about Buddhist history being framed in the same Christian terms. Walters suggests that Buddhologists and Theologians were concerned with more or less the same questions, with only the names being different. (10)

At the same time, Buddhologists insisted that their methods were "strictly historical", though this relied on "that nineteenth-century oxymoron 'scientific history' (also known as empiricism, historicism, or positivism)" (10). This approach "entailed numerous presuppositions about the nature of historical evidence and historical knowledge."

[I think Walters could have expanded on what these presuppositions were because they are not obvious to me. OTOH, it's obvious, for example, that Wynne (2019), Hinuber (2019), and Levman (2019) are all still operating largely in this 19th-century positivist mode when they write about the "historical Buddha"].

The problem with this is that Buddhist texts were not written with history or historicism in mind. The questions that concerned nineteenth-century scholars were, on the whole, completely unrelated to the concerns expressed by the authors of the Buddhist canonical literature.

This question has led an increasing number of Buddhologists today to argue that we are gravely mistaken to read Buddhist texts as though they had been written by nineteenth-century Europeans. (10)

In a note (41 n.4), the only name that Walters cites is Gregory Schopen, particularly his 1991 article "Archaeology and Protestant Presuppositions in the Study of Indian Buddhism".

[Given the "increasing number" I would have expected more citations here, but see also note 3.]

If Buddhist literature is conceived of as a series of answers to questions, then the questions asked in that literature are utterly unrelated to the questions being asked by 19th-century European historians. Moreover, European historical methods have changed significantly in the meantime. As Walters (11) says: "Our questions, presuppositions, historiographies, and sociopolitical realities are drastically different than those early nineteenth-century European men."

The classic example of this is the mining of Buddhist texts for historical details, while at the same time excluding any and all non-historical details as examples of "self-edification", "bias", "superstition", and appeal to "the psychological needs of the masses" (11). The insistence on this approach, Walters argues, "was rooted in one of the most basic Enlightenment European presuppositions: What I call the imperial dogma of universal human nature." (11).

This involves Europeans taking themselves to be exemplars of universal human nature; and assuming that "all people ask the same basic questions because all people experience the same basic reality in the same basic way" (11). And this goes to explain why European historians saw their own history as paradigmatic of world history.

[This view of history is still prevalent in "scientific" works of history such as Yuval Harari's Sapiens. It is directly challenged by David Graeber and David Wengrow in their book The Dawn of Everything (2021), which makes the argument that European culture is very far from being representative.]

Moreover, the picture of Buddhist history that premodern Buddhist texts do present varies wildly (12). Different Buddhist groups had different ideas about questions such as the dates of the Buddha and never arrived at a consensus (unlike modern scholars). [This section of Walters' essay—Philosophical and Political Problems—is more difficult to follow and understand]

Walters again argues that a majority of Buddhologists have left behind these outdated methods, embracing modern approaches to Buddhist literature and art, while leaving the "historicists" to argue about the details of their anachronistic approach.

[And again, I wish he had thought to provide some evidence to support this sweeping generalization because it's not self-evident].

Walters asserts that the "stale and unresolved debates of 19th-century historicists" effectively stymie efforts to place Buddhist history in a global context. (13)

He then returns to consider the epistemological presuppositions of the Buddhological Construct and the "scientific" paradigm in history.

[The latter is important because both Drewes and his recent detractors refer to taking a "scientific" approach to history.]

Walters makes a strong claim: that "Buddhism" did not exist before the nineteenth century. At one level he is talking about the term "Buddhism", coined by English-speaking scholars. On another, he points out that 19th-century accounts of global history did not see Buddhism as significant. He argues that, around the 18th century, there was a sea change in the way European writers refer to Buddhism and how they see it in a global context. Before this time, Buddhism is seen as serving the agency of Asian kings and their subjects. Afterward, Buddhism becomes a "transcendental essence" (13) with its own agency that followers merely enact. Again this reflects the idea of European Christianity as the model for all religions. As a result, Walters says: "I avoid the term Buddhism whenever possible" (13).

For Walters, the term Buddhism represents a hypostatization of a complex social phenomenon. We see this in phrases such as "Buddhism teaches that life is impermanent and without essence" (14). In this view, "Buddhism" is conceived of as a monolithic agent. Walters would rather that Buddhists, in all their diversity be seen as agents, rather than

[The obvious parallel for me is the routine hypostatization of science, in phrases like "Science says that the supernatural doesn't exist" (which is something I might have said). To my mind, this hypostatization is not peculiar to Buddhism but reflects a general pattern in European thinking. Walters (14) notes the same pattern in religious concepts like God, the Church, and the Holy Spirit as the agents of Christianity].

[I think here Walters is presaging Joseph Walser's (2022) idea of "Buddhism without Buddhists". Walser notes that while scholars assert that "Buddhism teaches anātman", if you go to Buddhist countries and ask Buddhists about this, they (A) have not heard of this doctrine, and (B) assert some kind of doctrine that would be considered ātmavāda.]

Next Walters draws attention to the political consequences of "hypostatization of 'Buddhism' as the agent of Buddhist history" (14). This view sees Buddhists as "passive victims of their own beliefs and practices".

The overwhelming evidence that Buddhist thought and practice were closely tied to political fortunes in separate kingdoms and across the premodern interregnal Buddhist world---Buddhist wars and warriors, Buddhist economies and monopolies, Buddhist courts and diplomacy, royal sponsors of Buddhist activities and Buddhist participants in royal activities---was dismissed as "un-Buddhistic", a mere accretion of what "Buddhism" was essentially supposed to be. (14)

The depoliticization of "Buddhism", or its deliberate dislocation from its socio-political context served the ends of European Christians and politicians. Since where Buddhists were involved in politics, this could be portrayed as a degeneration of the pure religion and contributed to the "decline" narrative.

What Buddhologists did for "empire" (wittingly or unwittingly)—providing its intellectual justification because Buddhist political statements and actions are not really political, or else they are not really Buddhist—they simultaneously did for Christian Mission. Making Buddhist agency into a "religion", showing that Buddhists do not really live up to their own "gospels" and insisting on the decadence and dormancy of contemporary "Buddhism" at that time, all dovetailed quite nicely with missionary goals. (15)

The Buddhological Construct obstructs the historian. It fails to answer historical questions [e.g. What was the relationship of Buddhists to their polity/king?] but it prevents such questions from even being asked by placing them out of bounds.

The invention of Buddhism was thus part and parcel of colonialism, the intellectual counterpart to a military regime that denied the sovereignty of those who were imperialized. (14).

As a historian, Walters is concerned to know how specific Buddhists saw themselves fitting into the world. He's interested in knowing how they saw themselves in relation to competing groups of Buddhists. Was there, in fact, a sense of shared identity amongst Buddhists? And so on. The standard kinds of questions that modern historians ask, but which never seem to have occurred to the scholars who constructed and promoted the Buddhological Construct.

Beyond the Buddhological Construct

For Walters, the way to go beyond the constraints and obstruction of the Buddhological Construct is provided by the British philosopher and archaeologist, R. G. Collingwood. Walters recommends reading An Autobiography and An Essay on Philosophical Methods rather than the better-known The Idea of History (this was published posthumously against Collingwood's wishes and supplemented with one of his student's notes that are "quite out of line with Collingwood's larger systemic writings" (40-41 n. 4)

Walter's essay undergoes a major change in style in this last chapter. The precise criticisms of the formation and application of the Buddhological Construct are replaced by more impressionistic statements about "global history" and "global perspective".

We need to bring a global perspective to bear on each bit of primary evidence we study, rather than to expect to cobble all the bits together into a global perspective. (18)

[This chapter introduces a suite of novel ideas and practices, along with several neologisms for concepts (whether they need them or not). For anyone unfamiliar with Collingwood, which I suspect is the majority of us, the endnotes provide important contextualisation. A lot of these notes could have been included in the body of the text]

Interregnal Buddhism

A term that is important for Walters is interregnal Buddhism (20-23). It is used several times before it is defined (on p. 20), which is confusing for the unfamiliar reader. Interregnal is being used where we might have expected to find international and reflects relations not between modern nation-states but between medieval kingdoms. However, "interregnal" and "global" are somehow not the same thing and the two terms continue to be used.

Walters argues that the societies in which Buddhism thrived, and for which we have primary sources, were kingdoms marked by constantly shifting relations (20). He argues that (unnamed) "orientalists" viewed the changing nature of the political landscape as evidence of disorder. Curiously, here, Walters introduces an emic concept "Buddhists... described change itself as a form of order" (20) and suggests that this is the key to seeing order in ever-changing Asian kingdoms.

At times we are deep in an unfamiliar jargon:

At any particular point in time, this "medieval" Asian world can be understood as a scale of forms being constantly revised across space, within which the entire interregnal Buddhist world... was but one large part. (20)

But what is a "scale of forms"? This is presumably evident to those who know Collingwood. The rest of us have to go back to the Preface and consult footnote 1.

The scale of forms, embodying hierarchical relationships of kind and degree among overlapping classes, is a technical term in R. G. Collingwood's philosophy of history. (xv n.1.)

But there is still more information in the endnotes (44 n 4). Here we find that the term "scale of forms" refers to the way an individual is embedded in overlapping social units:

An individual always acts as a member of a family, a resident of a village or town, a participant in a region or district, a subject-citizen of the king, of the king's circle of allies, and ultimately of the entire world. (44 n.4)

Note that in use, Walters defines "the scale of forms" and in his text "a scale of forms". The former suggests something monistic, the latter something pluralistic. Which is it? It's not clear to me what we gain from using this unfamiliar, rather counter-intuitive, terminology for facts that seem to be self-evident because, for example, I'm familiar with the works of Robin Dunbar, who has attempted to put upper limits on the size of these categories for human societies. Given this, the "scale of forms", which tells us nothing at face value, is a poor choice.

What is the point of this? (Unnamed) "Scholars" who take the view that there is something called "Buddhism" (qua hypostatized transcendental essence with agency) see "Buddhism" as all the claims to be Buddhist. Walters wants to contrast this with his concept of the "interregnal Buddhist world" which is...

a shifting, dynamic context within which Buddhists were the agents of their own history and their essence was contested, to be fully known only on the basis of what particular Buddhists in particular times and places thought and did. (21)

Walters asserts that every individual action reveals its global context (22). For example:

Each particular claim made by whatever particular Buddhist about one of the countless areas of contestation among Buddhists themselves and between Buddhists and non-Buddhists, when taken seriously as an informed action of a self-conscious human mind, reveals its global context wherever and whenever we look for it. (22)

[I think this is similar to my feeling that Buddhist history is written in isolation from world history. For example the periodisation of Buddhism doesn't relate to any periodization of Indian history more generally. Buddhist histories often play down the extent to which Buddhists were still embedded in wider social dynamics. It's almost as though Buddhists existed in a hermetically sealed environment in which external forces play no role. When examining changes in doctrine over time, we see descriptions and generalizations but not explanations: positivists work inductively to produce generalizations; historians work abductively to produce causal explanations.].

On the other hand, Walters notes that most particular Buddhists were not involved in the production of texts and thus we know little or nothing about them (22). This is important because the people who produced texts were an educated elite who could be expected to have a broader interregnal (perhaps even global) perspective of where Buddhists fit into the world.

Shared Paradigms and Opposing Interpretations.

As important as Buddhist interregnal interactions, were intra-Buddhist conflicts in which both sides were "claiming to represent the Buddhist whole". Walters notes that "Buddhist sources greatly disagree on virtually every topic of interest to historians, including the topics that constitute the five key moments of the Buddhological Construct" (23). Note that through this section, Walters works through each of the moments in the Buddhological Construct one by one.

1. The historical Buddha founded Buddhism
2. Early Buddhism
3. Asoka and general propagation through Asia
4. Medieval decline
5. Early modern revival (19th and 20th Centuries)

1 & 2. The Buddha and Early Buddhism.

The 19th-century approach to this pluralism was to attempt to adjudicate between them. In practice, this usually meant championing one of the medieval Buddhist polemics. But no evidence from the historical Buddha or early Buddhism as imagined by Buddhologists survives, so there is no historical basis for adjudicating. "Given that none of the evidence survives from the 'historical Buddha' or 'early Buddhism' as imagined by Buddhologists."(23).

Moreover, rather than trying to eliminate the contested nature of Buddhist history, we should be focused precisely on conflict:

When we ask what this contestation was intended to achieve, what rules made it possible, and how it proceeded in any particular instance, we are well on the way to finding Buddhists in global history (23).

Rather than take Buddhists at face value, and a pluralistic community riven by internal and external conflicts, 19th century historians sifted Buddhist documents for the one true Buddhism. Modern historians should focus on those periods for which we have reliable, dateable primary evidence, since the kinds of questions that historians ask can only be answered by documents of how Buddhists thought. And such documents as we have date from the "medieval" period (24).

3. Asoka.

Similarly, with respect to the variety of Buddhist stories about Asoka, Walters asserts

Rather than ask which version [of many] depicts "the historical Ashoka", we should wonder what it meant to be talking about Ashoka at all many centuries after he and his entire empire had crumbled to dust. The global historian should ask: Why did all Buddhists maintain their own version of the Ashoka Maurya story? (24-25)

[It annoys me that people use the Sanskrit form of the name Asoka Moriya, given that we know the man himself communicated in Prakrit and Buddhists did not use Sanskrit until several centuries after his death. Sanskrit in this context is an anachronism at best.]

Walters introduces a spatial metaphor for the five key moments identified as constituting the Buddhological Construct. We should think of these as "sites" (25). He argues "Viewing the five key moments... as key sites in the interregnal Buddhist world is to take seriously the contestation over them." (25). [On 34 Walters confusingly gives a completely different list of "five key sites": see "Conclusion" below]

Rather than focusing on the content of conflicting Buddhist accounts of Buddhism, we should stay with the fact that there were conflicting versions of these stories. Noting that Buddhists have argued amongst themselves about such issues, and trying to explain why, is what should interest the historian. The fact that Buddhists (one-sidedly) recorded intra-Buddhist debates and composed polemics should not be glossed over. 

[No historian would think of taking the approach of adjudicating which of the different Greek myths was the most accurate version of history].

Walters notes that all the various sects tended to claim to be the one true Buddhist path and criticized or even denigrated the others. They vied with each other to be the official representatives of Buddhism. They all claimed that their identity as Buddhists was the only true Buddhist identity. Walters points to a shared paradigm with contested interpretations.

The claim to be Buddhist was always a particular claim about the Buddha; which Buddha claimed was a vital factor in the sociopolitical import of those claims. All others may have lesser or even heretical understandings of the details, but even they share in taking the Buddha as the paradigm of all authentic human existence. (26).

The many contradictory biographies of the Buddha are not "windows into the time of the historical Buddha", which is itself restricted to a brief period in Northern India. Rather, such competing stories reflect medieval Buddhists attempting to define their identity and their place in the world.

As such the Buddhological Construct is not helpful because it projects the global Buddha back in time to "an ultimately irretrievable and decidedly nonglobal situation" (26). The Buddha imagined as a man living during the second urbanization is diminished compared to the medieval Buddha. [Note that this approach is exemplified in Bernard Fuare's (2022) recent book on the life of the Buddha].

In this manner, the Buddhological Construct distracts from one of the most central bodies of evidence we have for reconstructing the roles Buddhists played in global history. (26)

Much the same can be said for the subject of the number and chronology of the Great Buddhist Councils, the contents of the authentic Tripiṭaka. Common standards of ethics and etiquette, common "church" languages, and shared epistemic frameworks (like karma and rebirth) made a broader Buddhist identity possible, and this, in turn, created pan-Asian institutions (like Nālanda), artistic movements, economic exchanges, and the justification for empires (via the idea of the cakravartin). But the study of these well-documented aspects of Buddhist history is minimized by the Buddhological Construct and the focus on "early Buddhism", a non-global history which Walters again refers to as "irretrievable" (27).

Stories about Asoka vary but there is a common core to the story. Asoka presents us with the first global paradigm for a Buddhist ruler. And yet Buddhists contested the details. Walters makes the comparison with different interpretations of the intent of the "founding fathers" of the USA.

...the agreement that there is a (more or less potential) Buddhist imperial claim of which Ashoka is the ordinary paradigm, made championing one's own version of the Ashoka story consequential, generating a potentially imperial prize worth fighting for. (28)

[It's rare to see ancient Buddhists being described in such human terms. It reminds us that they were human beings.]

4. Decline

A similar case can be made for the idea that the Buddhist teachings will one day disappear. Note that this belief seems to have been ubiquitous amongst Buddhists. Again, the details are contested, but the decline narrative is also different from the previous three aspects of the Buddhological Construct. (28)

Ideas like the decline of Buddhism became more prominent when Buddhists felt their world to be in decline, be it a natural disaster, military invasion, or a loss of status for the saṅgha under a non-Buddhist king (and most kings in Indian history were not Buddhist), or simply a perception of declining standards of morality within the saṅgha.

Walters argues that all these kinds of calamities tended to make Buddhists unite (rather the disagree). He notes that a sense of decline in China drove Chinese Buddhists to visit India, where they witnessed the decline of Indian Buddhism. However, he also asserts that Buddhist kingdoms often existed in relative isolation from each other:

The main reason for disagreements in Buddhist accounts of the Dharma's decline may be a lack of much interaction of any kind. Buddhist kingdoms became entities unto themselves, exercising the potential political agency of the claim to be Buddhist, if at all, only in the limited region. (29)

[This is by far the weakest point in the article. Walters begins by arguing that the Buddhological Construct reflects 19th-century preoccupations. But now he tells us, and I agree, that the decline of the Dharma is a pan-Buddhist concern over an extended period of time, and as well as leading to the usual divergent accounts, it also led to a kind of trans-sectarian unity in the face of a common enemy].

5. Revival

The last plank in the Buddhological Construct is the revival of Buddhism beginning in the 19th century. This revival is conceived as coming out of the blue: "it has no premodern precedent" (29). The putative "revival" supposedly takes the same form as revivalist Protestant Christian movements from 19th-century Europe and the USA. In fact, we "are hard-pressed to find any premodern counterpart" (30). In essence

The revival of Buddhism is thus the impossible revival of something ("Buddhism" itself) that did not previously exist. (29)

This reflects Walters' earlier contention that "Buddhism" per se was an invention of the 19th century.

On the other hand, Walters argues that the changes we see in Buddhist doctrine and practice in this period reflect attitudes that are evident in the earliest evidence: Buddhists asserting that they are the Buddhists.

"Revivals" of the premodern sort did not just happen; rather, the powerful complex agents consisting of kings, courtiers, monks, and nuns set out to make them happen, articulating and effecting the particular shared vision of the interregnal Buddhist world, and their own place within it, which proved true at the particular place and time in the interregnal Buddhist world in which they found themselves. (30)

The point seems to be that Buddhists were constantly evaluating their attitudes and practices and changing them in response to changing conditions. Buddhists have pursued "revivals" throughout global history, and these have to be understood in their own context. Walters again emphasizes that change is not random or imposed on Buddhists by "Buddhism" qua "transcendental essence". Agency resides in people, not institutions.

Changes were made not as an inevitable result of creeping "Jesuitism" but as a result of reasoned argument, multireligious debate, political realism, enforced submission, and wished for and realized power. Human agency, Buddhist agency, made such "revivals" happen. (31).

 

Conclusion

Walters' concern throughout is with global history, the broadest view of human history. Buddhists clearly played an important role in this history. But the study of Buddhists in global history is hampered by the insistence on the 19th-century imperialist Buddhological Construct because so much effort goes into shoring it up and the results tell us nothing about global history.

He contrasts the Buddhological Construct with the records left by Chinese pilgrims who visited India.

The Chinese pilgrims cataloged the interregnal Buddhist world as a system of overlapping pictures of the whole—the whole Buddha's corpse and its worship, the whole sangha and its divisions, the whole potential empire and its constituent kingdoms—whose landmarks were the actual stupas, monasteries, and royal courts that the visited and described. (34)

From the pilgrim's accounts, we can say that Buddhist identity is twofold: to the outside world, Buddhists typically present as members of a single unified religion. At the same time, amongst themselves, Buddhists incessantly debate and negotiate what it means to be "Buddhist".

[There is a confusing use of "five key sites" on p 34 that are not the "five key sites" of the Buddhological Construct. But when "the five keys sites are mentioned again on p.35, it seems that the original set is indicated]

To understand the role of Buddhists in global history we have to take this into account. However, the global context is seldom so explicit. And recovering it is not easy. For his part, Walters endorses the view of Collingwood.

We cannot access the past; we can only imagine it, which is an activity of the present that will be as rich or poor as the imaginations of the historians who engage in it. (35)

Walters tries to address a potential criticism by stating that the Buddhological Construct is also a product of the imagination.

[The Buddhological Construct] is an interpretation of the very same evidence that all Buddhologists now know, and from the perspective of global history, it is not a particularly believable or well-thought-out interpretation at that. (35)

For Collingwood, according to Walters, the only hard facts that historians have are artifacts. The rest is a matter of interpretation. Collingwood's approach is not the only one, but it is a lot better than the "dead-end" Buddhological Construct. Moreover, not everyone need be concerned with global history. But for those who are concerned with seeing the place of Buddhists in global history, this approach allows primary sources to suggest answers to our questions. This is in contrast to the Buddhological Construct which aims to limit which questions can be asked and to provide standardised answers to those questions.

Still the global cannot simply be reduced to the particular and vice versa. One has to see the stupa both as a reflection of Buddhist soteriology and as a product of the (usually royal) patronage that paid for it. Building a stupa to worship has global ramifications.

Buddhist philosophy was both a quest for truth and a quest for power, each quest making the other more pressing; religious experiences were not only transcendent human realities but also self-conscious enactment of human agency, even on a global scale. (37)

 

—Notes—

Walters' essay makes a compelling case for rejecting the (19th century) establishment view of Buddhist history which hypostatizes Buddhism, giving it agency over Buddhists. He makes the case that agency resides in human beings; in this case, in Buddhists themselves. And that Buddhists were in constant dialogue with both non-Buddhists and other Buddhists over what it meant to be Buddhists.

Buddhists have played an important role in global history, but this role is effectively obscured and the history of Buddhists has been cut off from the global context by the imposition of the Buddhological Construct by 19th century imperialists.

That said, I think Walters (writing in 1999) has missed the extent to which reconstructing "early Buddhism" has become a Buddhist project, with prominent Theravāda bhikkhus and Tibetan Lamas taking leading roles within academia. The infiltration of Buddhist Studies by religious Buddhists since Walters composed this essay has only further entrenched certain biases and fallacies. While not entirely in sympathy with their imperialist predecessors, they nevertheless use the idea in the classic way that Walters describes: that is to say they use the construct of "early Buddhism" to mount a defence of an original "pure" teaching and claim to be representatives and defenders of that original teaching.

Rather than the idealized picture of unified Buddhism that Buddhists themselves present to non-Buddhists, historians should pay attention to what Buddhists were saying about each other in writing. And the most reliable sources we have are "medieval" (or at least post-Asoka).

I noted that Walters can be bracketed with other historians, amongst whom there has been a paradigm shift away from the Buddhological Construct and arguments based on it. Amongst the things these historians insist on is that a "primary source" is a first-hand account, written down by someone who lived through the events in question. There are no such sources for Buddhism before Asoka. Importantly, the idea of "early Buddhism" also known as "primitive Buddhism" is an invention of the non-Buddhist European scholars who first presented Buddhism to a modern European audience. (My next blog post will be a deeper exploration of this issue).

Whether or not Walter's use of Collingwood's ideas and the associated jargon moves us forward is moot. Looking at citations of Walter's essay I don't see any great enthusiasm for Collingwood.

For me, the great value of this long essay is that it speaks in a new voice about a new paradigm for doing Buddhist history that actually takes historical theories and methods into account. In the last few months, it has become apparent to me that this is rare in Buddhist Studies. Many of the combatants in the conflict over the historicity of the Buddha, for example, seem to have no idea that their methods are at odds with the whole field of history (see my critique of Levman 2019, which is representative of the positivist-historicist approach).

Ironically, this whole conflict, while being mainly carried on by people with doctorates in Buddhist Studies or Philology, pivots on approaches to history—primary sources, corroboration, etc—that are taught explicitly in high school, and have become background assumptions even in undergraduate history courses. Countering the Buddhological Construct in terms of (the field of) History is difficult because historians are not interested in publishing high-school level critiques of their professional colleagues in Buddhist Studies, and those professionals in Buddhist Studies are not interested in History because it contradicts their articles of faith (which combine religious and scholarly conceits).

I am already following up on some of the recommended reading and processing the change in paradigm that is demanded by paying attention to historians. What I found during this exercise is that it deflated my belief that we know anything about "early Buddhism". I've come around to seeing arguments over early Buddhism as an imperialist European conceit.

—Bibliography—

Faure, Bernard. (2022) The Thousand and One Lives of the Buddha. University of Hawai'i Press.

Graeber, D. and Wengrow, D. (2021). The Dawn of Everything: A New History of Everything. Allen Lane.

Harari, Yuval Noah. (2011). Sapiens: A Brief History of Humankind. Random House.

Schopen, Gregory. (1991). "Archaeology and Protestant Presuppositions in the Study of Indian Buddhism". History of Religions Journal 31(1): 1-23. https://www.jstor.org/stable/1062872

Walser, Joseph. (2022). "Buddhism without Buddhists? Academia & Learning to See Buddhism Like a State". Pacific World. Series 4, Vol. 4: 103-170. https://www.shin-ibs.edu/documents/pwj4/3/4-3-4-Walser.pdf

Walters, Jonathan S. (1998). Finding Buddhists in Global History. American Historical Association.