In the process of revising the history of the Heart Sutra, it has become clear that Conze's Prajñāpāramitā chronology is faulty in many respects. In this essay, I will discuss some of the main faults with the existing chronology and then propose a substantial revision, notably deprecating his use of the term "abbreviation".
In "The Development of Prajñāpāramitā Thought" (1967) Conze outlined nine stages of development:
- The initial formulation represented by the first two chapters of Ratnaguṇasamcayagāthā (Rgs).
- Chapters 3-28 of Rgs.
- Incorporation of matter from the Abhidharma.
- Concessions to the Buddhism of Faith.
- The last third of the Large Prajñāpāramitā
- The short Sutras.
- Yogacarin commentaries.
- Tantra
- Chan.
Later, Conze (1978) boils this down to a fourfold chronology:
- The basic text (ca. 100 BCE – 100 CE)
- Expansion (ca 100 – 300 CE)
- Abbreviation (300 – 500 CE)
- Tantric/Magical influence (600 – 1200 CE).
There is some revision of this chronology in Stefano Zacchetti's (2020) authoritative encyclopedia entry for Prajñāpāramitā Sūtras in the Brill Encyclopedia of Buddhism, but even Zacchetti largely accepts Conze's views on the Heart Sutra. Unfortunately, neither Huifeng (2014) nor any of my research on the Heart Sutra made it into his article. Zacchetti's contribution to Prajñāpāramitā Studies was huge and his early death was a great loss to the profession. However, he did not go far enough in critiquing Conze and revising the chronology.
Ratnaguṇasamcayagāthā
Conze's rationale for dating the Ratnaguṇasamcayagāthā very early was never sound. The first translation of Rgs into Chinese was in 991CE, by Fǎxián 法賢 (T 229). As Zacchetti (2020) says "... external evidence clearly points to a much later date of composition of this text, as all its witnesses, including the Chinese translation, are comparatively late". Zacchetti also notes that Rgs is not included in Xuanzang's compendium of Prajñāpāramitā texts.
Far from being an early exemplar, the Rgs played little or no part in the development of Prajñāpāramitā thought or literature and was a decorative after-thought.
Aṣṭasāhasrikā Prajñāpāramitā
This makes Aṣṭasāhasrikā clearly the oldest Prajñāpāramitā text. The Split Collection manuscript of Aṣṭasāhasrikā has been carbon-14 dated to 74 CE with a two-sigma range of 47-147 CE (Falk 2011: 20). This date is consistent with paleographic dating of the manuscript. Thus we can place the existence of a written text in the first century, but we don't know when it was first written down, or when it was first composed.
Aṣṭasāhasrikā was first translated into Chinese in 179 CE by Lokakṣema (T 224). The procedure of subtracting a century or two from the date of the earliest Chinese translation was never credible, despite being widely applied. Note, for the record, that the oldest extant Pāli text is from the fifth or sixth century. Which makes the oldest physical evidence for Prajñāpāramitā some 300-400 years older than the oldest Pāli text. As far as I know, the idea that the Pāli canon was written down around the beginning of the Common Era first occurs in a fifth-century Sri Lankan text: the Mahāvaṃsa.
My sense of Aṣṭa is that the written text reflects an older, probably oral tradition. Everyone seems to agree that the first Chapter of Aṣṭa is likely older than the rest and possibly represents the "original" Prajñāpāramitā text. The first chapter has no narrative structure but is an episodic (almost disparate) collection of independent dialogues. Aṣṭa often seems like a roughly edited selection of brief stories featuring fragments of dialogues between figures that feature in early Buddhist texts.
The fragments in Aṣṭa show a clear affinity to ideas found in early Buddhist texts, suggesting that Prajñāpāramitā is conceptually much older than its oldest texts. Anālayo (2021) argues that the practice of withdrawing attention from sensory experience (so that it ceases) seems to predate Buddhism since this is precisely what the Buddha learns from his pre-awakening teachers: Āḷāra Kālama and Udaka Ramaputta.
Later chapters of Aṣṭa still lack any sense of narrative, except for the story of Sadāprarudita (Chapters 31 and 32), but are more thematically consistent. Huifeng's PhD thesis identified several "chiastic" structures in the earliest Chinese translation of Aṣṭa. A chiastic structure involves a mirror image. Topics are introduced from the beginning to the middle and then recapitulated in reverse order from the middle to the end. Huifant identified chiastic structures in Chapters 1 and 2 and in Sadāpraruidita chapters. He also proposes that the entire text has a chiastic structure. This conflicts with the broadly accepted notion that the text was built up piecemeal from a core variously said to comprise Chapters 1 and 2, just Chapter 1, or some part of Chapter 1. Joseph Walser (2018: 130-134) gives a good overview of the "quest for the ur-sūtra"
While I don't know enough to comment on this, it does seem to me that no chapter of Aṣṭa has just one topic.
Vajracchedikā Prajñāpāramitā
The Vajracchedikā—which Conze placed in a later period—is now widely believed to be from an earlier period. However, the rationale for this early dating is unclear and difficult to check. Curiously, Zacchetti's (2020) account of the Vajracchedikā ignores the issue of its place in Conze's chronology, which risks people interpreting this as approval.
It seems likely that Vajracchedikā was initially considerably shorter since there is in fact a clear ending at verse 13a. More text seems to have been tacked onto the end, which fits the same pattern of evolution of other Prajñāpāramitā texts.
Harrison (2006) does not mention dates, but merely alludes to comments made by Gregory Schopen (1975: 153):
It is, however, worth noting that a number of Japanese scholars have suggested a date for the Vaj which is considerably earlier than the one suggested by Conze, and that the exact nature of the relationship between the Vaj and the [Aṣṭa] is far from clear.
The main text then points to a footnote which lists several sources of interest:
Two of these sources are repeatedly cited by other scholars also.
Nakamura, H. (1964). "A Critical Survey of Mahāyāna and Esoteric Buddhism chiefly based on Japanese Studies." Acta Asiatica 6: 64-65.
Ui, Hakuju. (1958). "Chronological Survey of the Vajracchedikā-Prajñāpāramitā." Nagoya-Daigaku-Bungakubu Kenkyu-Ronshū XXI: 49-51.
Ui (1958) is not available in any UK library as far as I can see. Cambridge, Oxford, and SOAS libraries all have this periodical, but their runs are incomplete and Vol XXI is missing from all three. Cambridge also lacks the early numbers of Acta Asiatica. None of these is online as far as I can see. So as of today, I have been unable to consult any of these (I haven't given up).
The citation, "Trans, by Hanayama Shōyū, 'SVRPL' (cf. n. 8) 55-61." refers to
Hanayama, Shōyū 花山勝友. (1966) "A summary of various research on the Prajñāpāramitā literature by Japanese scholars." Acta Asiatica 10: 55-6
Again, I have yet to get access to this. Schopen notes that:
"Ui says 'judging from its contents, this sūtra gives us the impression that it is a very old sūtra" (p. 56); and "... the latest date of the establishment of the Diamond Sutra will be 200 A.D. or probably 150 A.D., though we cannot decide the earliest possible date of this sūtra" (p. 60)."
However, it's not clear on what basis Ui is making this judgement. Although the article is widely cited, no one seems to say what was Ui's rationale for this date. And I can't find any more recent research.
The first translation into Chinese is Kumārajīva's T235 Jīngāng bōrě bōluómì jīng «金剛般若波羅蜜經» in 402 CE. However, extant commentaries have been attributed to both Asaṅga and Vasubandhu. This would help except that the dates of these two are not certain and such attributions cannot be taken at face value: numerous texts are attributed to both of them apocryphally. Asaṅga is assumed to have lived in the fourth century. Jonathan C. Gold asserts that Vasubandhu lived in the late fourth or early fifth centuries (Stanford Encyclopedia of Philosophy). But there does not seem to be any consensus on their dates.
So Vajracchedikā seems to have existed by the end of the fourth century. If there is evidence compelling us to push this date back, I cannot find it (yet). I occasionally correspond with Paul Harrison and I know that he is writing a book on Vajracchedikā and has his own ideas on dates, partly based on the Central Asian manuscripts. Nothing he has published so far makes this any clearer.
Note 5 Feb 2024. Harrison has responded to an email saying that he believes Vaj was originally composed in a Prakrit based on Central Asian fragments. And that composition in the second century, per Ui (1958), seems likely.
Expansion
Dates for the composition of the extended Prajñāpāramitā texts, particularly Pañcaviṃśatisāhasrikā, are uncertain. There are two Chinese translations from the third century: Guāng zàn jīng «光讚經» (T 222) by Dharmarakṣa in 286 CE; and Fàng guāng bānrě jīng «放光般若經» (T221), by Mokṣala in 291 CE. The earliest Sanskrit witness is the Gilgit Manuscript dated to the sixth or seventh century.
Pañcaviṃśatisāhasrikā retains the basic organisational plan of Aṣṭa, and includes more or less all of the text, but intersperses new material that is twice the length of the shorter text. Jan Nattier (2003: 62, n. 19) has likened the process to slicing bread and filling the spaces.
Abbreviations
While Aṣṭasāhasrikā certainly underwent massive expansion, notably producing Pañcaviṃśatisāhasrikā, (and several variants) there never was a period of "abbreviation" of Prajñāpāramitā. No Indian text was ever abbreviated. The closest we get to this in Indian Buddhist literature is passages quoted in anthologies.
Moreover, the Heart Sutra definitely does not fit Conze's paradigm since it was composed in China in the mid-650s. It is incorrect to think of Vajracchedikā as an "abbreviation". Rather it is a short text targeting a particular type of wrong view about the relation between experience and abstract concepts.
It is not plausible to think of the short Tantric texts as "abbreviated" either. As with Vajracchedikā, these are not "abbreviations" because this would imply the existence of some longer text that they were abbreviations of. No such longer text exists. They are simply short texts composed by Tantric Buddhists, with a Prajñāpāramitā flavour to them.
In my view, the development of Madhyamaka metaphysics was a wholly separate and unrelated process, though the two do share some antecedents. The mythology linking Nāgārjuna to Prajñāpāramitā is a late tradition and not entirely coherent, given the strikingly different emphases and conclusions of the two traditions.
Provisional Revised Chronology
In retrospect, we can speculatively identify a zeroeth phase of Prajñāpāramitā, as found in texts such as the Cūḷasuññatā Sutta (MN 121) and the Kaccānagotta Sutta (SN 12.15). Prajñāpāramitā has numerous antecedents in early Buddhist literature and, if legends can be taken as indicative, this form of practice predates Buddhism. For example, the Ariyapariyesanā Sutta (MN 26) points to the Buddha learning precisely this style of meditation, aimed at the cessation of sensory experience, from his pre-enlightenment teachers.
There is a major caveat here. While we can identify antecedents to Prajñāpāramitā it is a mistake to think of these as proto-Prajñāpāramitā. The prefix proto- implies a teleology: the idea that these earlier ideas and practices necessarily became Prajñāpāramitā (and only Prajñāpāramitā). In fact, there is no such necessity, and these particular antecedents are also antecedent to other forms of Buddhism. Buddhism is notable for constantly diversifying into new forms and converging from time to time in new syntheses (on the limitations of the branching tree model of evolution compared to a braided stream model, see: Evolution: Trees and Braids 27 December 2013).
The first phase of Prajñāpāramitā literature proper is what we now call the Aṣṭasāhasrikā. The first Prajñāpāramitā text was likely considerably shorter (based on extant witnesses that grow over time) and was probably originally just called Prajñāpāramitā. Arguments about an ur-text underlying the extant manuscript witnesses are speculative. This period is likely still intimately connected with an ancient tradition of meditative practice focussed on attaining cessation and dwelling for long periods (up to seven days) in the absence of sensory experience. As noted, this approach has clear antecedents in early Buddhist texts and likely predates Buddhism (based on Buddhist accounts).
The second phase involves both the expansion of Aṣṭa and the composition of the Vajracchedikā. In the first case, the basic Prajñāpāramitā text (in roughly 8000 lines) was expanded into versions of 10,000, 18,000 and 25,000 lines, incorporating a great deal of new material as well as unpacking some of the abbreviated expressions. Contra Conze, I don't think this involved Abhidharma or "concessions to the Buddhism of faith". A subsequent expansion of the 25,000-line text to 100,000 lines did not add new material, but mainly consisted of a full and complete expansion of all the abbreviated expressions. If there was still a distinct Prajñāpāramitā practice community, it had begun to adopt an approach that was both more eclectic and more scholastic.
A third phase was the composition of commentaries such as the Abhisamayalaṅkāra and the Mahāprajñāpāramitopadeśa, which were highly influential in Tibet and China respectively. These commentaries are quite clearly scholastic rather than practical. Thus I assume that Prajñāpāramitā as a distinct form of practice is no longer visible, though I think some form of this practice probably still existed and was passed on within other traditions (I think here particularly of Mahāmudra).
The fourth and final phase reflected the absorption of elements of Prajñāpāramitā into Tantric Buddhism and the disappearance or submergence of any remnants of Prajñāpāramitā as a distinct approach to Buddhist practice. This coincides with the production of a number of short texts expressing a Tantric worldview and the emergence of Prajñāpāramitā as a Tantric deity. Whether these are really Prajñāpāramitā texts is moot and I am inclined to lump them with Tantra.
~~oOo~~
Bibliography
Anālayo. (2021) “Being Mindful of What is Absent.” Mindfulness 13: 1671-1678.
Conze, Edward. (1967). "The Development of Prajñāpāramitā Thought." In Thirty Years of Buddhist Studies, 123-147. New Delhi: Munshiram Manoharlal. [First published In Buddhism and Culture, Dedicated to Dr. Daisetz Teitaro Suzuki in Commemoration of His Ninetieth Birthday, ed. S. Yamaguchi, 24–45. Kyoto: Suzuki Gakujutsu Zaidai].
———. (1978) The Prajñāpāramitā Literature. (2nd Ed.) Tokyo: The Reiyukai.
Falk, Harry. (2011). The ‘Split’ Collection of Kharoṣṭhī Texts. ARIRIAB XIV (2011), 13-23. https://www.academia.edu/3561702/split_collection
Nattier, Jan. (2003). A Few Good Men: The Bodhisattva Path according to The Inquiry of Ugra (Ugraparipṛcchā). Honolulu: University of Hawai’i Press.
Schopen, G. (1975). "The phrase ‘sa pthivīpradeśaś caityabhūto bhavet' in the Vajracchedikā: Notes on the cult of the book in Mahāyāna." Indo-Iranian Journal 17(3/4): 147-181.
Walser, Joseph. (2018). Genealogies of Mahāyāna Buddhism: Emptiness, Power, and the Question of Origin. London and New York: Routledge.
Zacchetti, Stefano. (2020) “Prajñāpāramitā Sūtras.” In Brill's Encyclopedia of Buddhism Online, edited by Jonathan A. Silk, Oskar von Hinüber, and Vincent Eltschinger (Unpaginated). Leiden: Brill. https://referenceworks.brillonline.com/entries/encyclopedia-of-buddhism/prajnaparamita-sutras-COM_0017 [accessed 6 Sept 2023].
