The Pañcaśatikā Prajñāpāramitā sūtra, the sūtra on Perfection of Wisdom in Five Hundred Lines, when describing the aggregates, etc., uses three terms that apparently refer to the three natures (svabhāva) taught in Yogācāra texts. As a Prajñā-pāramitā sūtra, it would be part of the second promulgation of the Dharma, while the sūtras behind the Yogācāra texts are part of the third promulgation of the Dharma. Because of this, the Tibetan teacher Dolpopa regarded the Pañcaśatikā Prajñāpāramitā as a text of definitive meaning (nītārtha), and characterized it as one of the Buddha’s own auto-commentaries (rang ‘grel ) on the extensive Prajñā-pāramitā sūtras. Dolpopa taught that the Prajñā-pāramitā sūtras should be understood by way of the three natures found in these “auto-commentaries.” However, one of the three terms used in the Pañcaśatikā Prajñāpāramitā in its Tibetan translation does not seem to fit well as referring to the three natures. The original Sanskrit text was long lost, and with no Indian commentary to consult even in Tibetan translation, there was no way to determine what was actually meant by this term. Fortunately, the Sanskrit original was recovered in Tibet and published in 2016 as number 20 of the important series, Sanskrit Texts from the Tibetan Autonomous Region.¹

The three terms in the Tibetan translation of the Pañcaśatikā Prajñāpāramitā, near the beginning, are dngos po med pa, dngos po ngan pa, and dngos po yod pa, translated by Edward Conze in 1973 as “non-existence,” “a poorish kind of existence,” and “existence,” and translated by Cyrus Stearns in 2010 as “nonexistent,” “an inferior existence,” and “existent.”² These are supposed to correspond to the three natures: the imagined (parikalpita, kun brtags), the dependent (paratantra, gzhan dbang), and the perfect (pariniṣpanna, yongs grub). As may be seen, the second term in the Pañcaśatikā Prajñāpāramitā, dngos po ngan pa, “a poorish kind of existence,” or “an inferior existence,” does not seem to fit well in this scheme. Yet these English terms are fully accurate translations of the Tibetan term. With the Sanskrit now available, we can see what happened. The three Sanskrit terms are: abhāva, “non-existent,” nâbhāva (na abhāva), “not non-existent,” and sad-bhāva, “truly existent.”³ These correspond well to the three natures taught in Yogācāra texts: the imagined, the dependent, and the perfect.

The Tibetan translator, perhaps to avoid the double negative that is in the Sanskrit, na abhāva, “not non-existent,” chose dngos po ngan pa to translate this second term, ostensibly “a poorish kind of existence,” or “an inferior existence.” The common meaning of ngan pa is indeed “poorish” or “inferior,” as Conze and Stearns translated it. However, here the Tibetan translator apparently intended one of the uncommon meanings of ngan pa, namely, asat, “not true,” thus yielding “not truly existent” in contrast with the third term, “truly existent.” This meaning of ngan pa as asat can be found in the Bodhisattvabhūmi (Nalinaksha Dutt edition, 1966, p. 98): asat-saṃkathā, ngan pa’i gtam, “untrue conversation.” Another example of this meaning can be found in the Jātakamālā (P. L. Vaidya edition, 1959, p. 159): asad-dṛṣṭiḥ, lta ba ngan pa, “false view.”⁴

With the help of the original Sanskrit, we can now see that these three terms in the Pañcaśatikā Prajñāpāramitā do in fact correspond well to the three natures taught in Yogācāra texts. Three other terms that apparently refer to the three natures taught in Yogācāra texts are used in another Prajñā-pāramitā text that Dolpopo regarded as being of definitive meaning (nītārtha), and that he characterized as one of the Buddha’s own auto-commentaries (rang ‘grel ) on the extensive Prajñā-pāramitā sūtras. The Maitreya Paripṛcchā or “Questions of Maitreya” chapter of the Prajñā-pāramitā sūtras in 25,000 and 18,000 lines, when describing the aggregates, etc., uses parikalpita, “imagined,” vikalpita, “conceptually differentiated,” and dharmatā, “true nature” (Tibetan kun brtags pa, rnam par brtags pa, and chos nyid ). These, too, correspond well to the three natures: the imagined, the dependent, and the perfect.

An extensive commentary on all three of the large Prajñā-pāramitā sūtras, those in 100,000 lines, 25,000 lines, and 18,000 lines, directly equates the three natures taught in Yogācāra texts with the three terms found in the “Questions of Maitreya” chapter, and uses these terms throughout in its explanations.⁵ Dolpopa drew heavily upon this commentary, called in short the Bṛhat-ṭīkā, “Large Commentary,” and known in Tibet as the Yum gsum gnod ‘joms, “Destruction of Objections to the Three Mother Sūtras.”⁶ Most of Tibetan tradition, including Bu-ston who edited the Tengyur, regarded it as being written by the early Indian teacher Vasubandhu, famous for his Yogācāra treatises. Tsongkhapa, however, held that it was written by the much later writer Daṃṣṭrāsena, because it included some late references. It is of course possible that Daṃṣṭrāsena merely added some things to the earlier text by Vasubandhu. In any case, the method of understanding the Prajñā-pāramitā sūtras by way of the three natures taught in Yogācāra texts goes back at least to Dignāga, who is traditionally regarded as a direct disciple of Vasubandhu. Dignāga wrote in his Prajñāpāramitā-piṇḍārtha, verses 27-29:⁷

prajñā-pāramitāyāṃ hi trīn samāśritya deśanā |
kalpitaṃ paratantraṃ ca pariniṣpannam eva ca || 27 ||

The teaching in the Perfection of Wisdom is based on three:
the imagined, the dependent, and the perfect.

nâstîty-ādi-padaiḥ sarvaṃ kalpitaṃ vinivāryate |
māyôpamâdi-dṛṣṭāntaiḥ paratantrasya deśanā || 28 ||

By the words, “does not exist,” etc., all the imagined is refuted.
By the examples, like an illusion, etc., the teaching of the dependent [is given].

caturdhā vyavadānena pariniṣpanna-kīrtanam |
prajñāpāramitāyāṃ hi nânyā buddhasya deśanā || 29 ||

By the fourfold purification, the perfect is taught.
For in the Perfection of Wisdom there is no other teaching of the Buddha.

Dolpopa, then, was not innovating when he advocated understanding the Prajñā-pāramitā sūtras by way of the three natures taught in Yogācāra texts. He was merely following a much older Indian tradition. This led him to find correspondences to these three natures in the Prajñā-pāramitā sūtras themselves, such as the Pañcaśatikā Prajñāpāramitā. He quoted the whole opening section of this sūtra at the beginning of his concise text, Ngo sprod khyad ‘phags, “Exceptional Introduction.”⁸ He then equated its three terms with the three natures taught in Yogācāra texts. He said the same thing, again equating its three terms with the three natures, in his Autocommentary to the “Fourth Council”.⁹ Thus, the Pañcaśatikā Prajñāpāramitā with its three terms corresponding to the three natures was regarded by Dolpopa as a text of considerable importance for understanding the Prajñā-pāramitā sūtras.

Notes

1. Pañcaśatikā Prajñāpāramitā: Sanskrit and Tibetan Texts, critically edited by Li Xuezhu and Fujita Yoshimichi. Beijing: China Tibetology Publishing House, and Vienna: Austrian Academy of Sciences Press, 2016.
2. “The Perfection of Wisdom in 500 Lines,” in The Short Prajñāpāramitā Texts, translated by Edward Conze (London: Luzac & Company, 1973), p. 108. Relevant sentence quoted by Cyrus Stearns in The Buddha from Dölpo: A Study of the Life and Thought of the Tibetan Master Dölpopa Sherab Gyaltsen (Ithaca, N.Y.: Snow Lion Publications, 2010), p. 101, with reference to Dolpopa’s comment on it in his Autocommentary to the “Fourth Council”, p. 233. In the 1999 first edition this quotation is on pp. 96-97, and the three terms are translated as: “a nonexistent entity, a base entity, and an existent entity.”
3. These three terms first describe the neuter word rūpam, “form” (p. 1), so according to their masculine gender they would be nouns rather than adjectives; e.g., “non-existence” rather than “non-existent.” However, to call form “non-existence” does not make sense to me. So bhāva is probably used here as the noun, “an existent” (an existing thing). The sentence, then, would say: “form is a non-existent, not a non-existent, and a truly existent.” This is rather awkward English. I think the same idea is conveyed by translating these terms as if they were adjectives: “form is non-existent, not non-existent, and truly existent.” This is what I have done, even though it is not a literally accurate translation.
4. These examples are found in J. S. Negi, Tibetan-Sanskrit Dictionary, Sarnath, Varanasi: Central Institute of Higher Tibetan Studies, vol. 3, 1995. I have only added the English translations.
5. Ārya-śata-sāhasrikā-pañcaviṃśati-sāhasrikâṣṭādaśa-sāhasrikā-prajñāpāramitā-bṛhaṭ-ṭīkā.
6. For the English translation of this title, I follow Stearns, 2010 (see note 2 above), p. 97.
7. The original Sanskrit was first edited by Giuseppe Tucci and published in the Journal of the Royal Asiatic Society in 1947, which I have scanned and posted here: http://www.downloads.prajnaquest.fr/BookofDzyan/Sanskrit%20Buddhist%20Texts/prajnaparamita_pindartha_1947.pdf. It was published again in 1959 by Erich Frauwallner in the Wiener Zeitschrift für die Kunde Süd- und Ostasiens in 1959, which I have scanned and posted here: http://www.downloads.prajnaquest.fr/BookofDzyan/Sanskrit%20Buddhist%20Texts/prajnaparamita_pindartha_1959.pdf. Although Tucci also included an English translation, I have here re-translated these verses more literally.
8. The Ngo sprod khyad ‘phags is found in volume 12 of the 13-volume modern typeset edition of the collected writings of Dolpopa, pp. 40-52 (jo nang kun mkhyen dol po pa shes rab rgyal mtshan gyi gsung ‘bum, [Beijing:] krung go’i bod rig pa dpe skrun khang, 2011). For the English translation of this title, “Exceptional Introduction,” I follow Stearns, 2010 (see note 2 above), p. 422. Matthew Kapstein describes it as: “An ‘introduction’ (ngo-sprod ) to the ultimate and definitive significance (nges-don mthar-thug) of the doctrine.” (The ‘Dzam-thang Edition of the Collected Works of Kun-mkhyen Dol-po-pa Shes-rab Rgyal-mtshan: Introduction and Catalogue, p. 66. Delhi: Shedrup Books, 1992). The opening section of this sūtra that Dolpopa quoted (pp. 40-43) corresponds to the Sanskrit edition (see note 1 above), sections 1 and 2, pp. 1-4.
9. Translated by Stearns, 2010 (see note 2 above), p. 233, and quoted by him on p. 101. In the 1999 first edition this is quoted on p. 96.