Notes

n.1The ideas that actions committed in the past appear to the mind at the moment of death, and that a person’s last thought determines their next existence, are not frequently encountered in Mahāyāna literature. There are, however, precedents for the second idea‍—the karmic efficacy of final thoughts‍—in Hindu and early Buddhist texts. See McDermott (1980), pp. 177–78. For an example from the Pali suttas, see MN 3.214 in Bhikkhu Ñāṇamoli and Bhikkhu Bodhi (1995), pp. 261–62.

n.2Willis (1982), p. 119.

n.3Denkarma folio 299.b; Phangthangma (2003), p. 17. Both catalogs describe the sūtra as seventy ślokas in length.

n.4Mahāvyutpatti, no. 1379.

n.5Currently, the manuscript is kept in the Potala. Vinītā’s critical edition is based on a copy of the manuscript that is kept in the China Tibetology Research Center. For further details on the state of this manuscript see Vinītā (2010), pp. xv–xvii.

n.6Vinītā (2010), pp. xxvii–xxix.

n.7Kawa Paltsek, gsung rab rin po che’i gtam rgyud dang shAkya’i rabs rgyud. BhSS is quoted on pp. 805–7 in the Pedurma edition.

n.8There are numerous Sanskrit fragments of this text. See Vinītā (2010), pp. 410–47, for a detailed critical edition, which compares the various sources. Vinītā includes transcribed sections of Candrakīrti’s Madhyamakāvatāra­bhaṣya from a copy of a Sanskrit manuscript from the Potala kept at the China Tibetology Research Center (Box 136). The critical edition also includes Sanskrit fragments from Candrakīrti’s Prasannapadā, Prajñākaramati’s Bodhi­caryāvatāra­pañjikā, Asaṅga’s Bodhisattva­bhūmi, and Śāntideva’s Śikṣāsamuccaya, all of which quote the BhSS or parallel passages of the PPSS (as is the case with the Śikṣāsamuccaya).

n.9See n.­8.

n.10See Degé Tengyur, vol. 96 (dbu ma, tsa), folio 158.a. The authorship of these texts is uncertain. See Hakamaya (1977), p. 482; see also Tola and Dragonetti (1986), p. 13, fn. 1. The Tibetan and restored Sanskrit editions of these texts, along with a translation of the śāstras and summary of the ṭīka, are given in Sastri (1938).

n.11Lindtner states that Nāgārjuna had adopted these verses from the BhSS, while Tola and Dragonetti suggest that it may have been the other way around, and that while the prose section certainly predates Nāgārjuna’s work, it is possible that some of the verses, being later additions, were drawn from the śāstra. See Lindtner (1992), pp. 264–66; see also Tola and Dragonetti (1986), pp. 4–5.

n.12Tola and Dragonetti (1986), p. 3.

n.13Skilling (2021), pp. 393–408.

n.14Hakamaya (1977), pp. 480–82. This very informative paper is merely an abstract from a larger, more detailed article. Unfortunately for us it was only available in Japanese. See “The Bhava­saṅkrānti­sūtra‍—Introduction and Translation,” Journal of Buddhist Studies (Komazawa daigaku bukkyō gakubu ronshū) 8, Tokyo: Komazawa University, 1977: 276–302.

n.15See Buswell and Lopez (2014) under “pudgala” and “pudgalavāda.” See also McDermott (1980), pp. 168–69. There may also be some theoretical discrepancy between the causality described in the Bhava­saṅkrānti­sūtra, in which the action from the past ceases before the first consciousness of the next birth arises (although the two are causally connected), and the general Sarvāstivādin view that maintains that past, present, and future actions exist simultaneously.

n.16See n.­38.

n.17Sastri (1938), pp. xvi–xvii.

n.18The Viṃśatikā of Vasubandhu (Toh 4056) describes the causal efficacy of external objects using an example, similar to that of BhSS, of a person having intercourse in a dream, but in which the perception is compared to the resulting seminal emission. Nāgārjuna’s Mahā­prajñā­pāramitā­śāstra gives another similar account, in which three brothers wonder why dreaming of three courtesans who were described to them brought them pleasure and woke them up from sleep. This story is used to teach the emptiness of phenomena. Lamotte (1944), p. 425.

n.19The Theravādins and several other Buddhist schools likewise refute an intermediary state. For a discussion of the different theories of transmigration in early Buddhist schools, see McDermott (1980), pp. 165–92, and Wayman (1984), p. 251.

n.20It should be noted that theories of the nominalist critique of “self” go back to Pali sources like the Milindapanha and were later used to support the śūnyavāda view in Madhyamaka works. We are grateful to David Higgins for providing this insight along with the following note: “A shared premise of such [Madhyamaka nominalist] critiques is that the ‘self’ is a mere nominal designation (prajñāpti) for what is in reality a heterogeneous assemblage of partite factors. Although nominalism is often associated with Buddhist epistemological views concerning the ontological status of universals, early Buddhist thinkers already regarded composite entities such as selves and persons as having a mere nominal existence. Along these lines, early Abhidharma traditions distinguished between substantially existent (dravyasat) and nominally existent (prajñaptisat) entities and regarded the ‘person’ as nothing more than a merely conventional denomination established on the basis of the five psychophysical aggregates and having a merely conventional reality. According to Mi bskyod rdo rje, […] Madhyamaka thinkers such as Candrakīrti denied even ‘nominal existence’ to selves and persons within the tenets of their own system.” Higgins (forthcoming).

n.21That is, “seeing” cannot occur independent of a number of causes and conditions.

n.22Hakamaya (1977), p. 479, fn. 4. The trisvabhāva theory (mtshan nyid gsum) is a creation of the Yogācāra school used to address what it perceived to be insufficiency (and nihilism) of the two truths of the Madhyamaka school. The three natures are (1) parikalpita­svabhāva (kun brtags pa’i mtshan nyid), the “imagined” nature, that is our mistaken conception of phenomena; (2) para­tantra­svabhāva (gzhan dbang gi mtshan nyid), or “relative” nature, which refers to the causal, interdependent nature that gives rise to that conceptuality; and (3) pari­niṣpanna­svabhāva (yongs grub pa’i mtshan nyid), or the “perfect” nature, which is the true, ultimate nature of phenomena.

n.23To quote Hakamaya, “The most part of the BhSS is quoted in the Madhyamakāvatara­bhāṣya […] of Candrakīrti in order to refuse the ālayavijñāna of Yogācāra. Though it is difficult to explain the reason why Candrakīrti would be better pleased to take the BhSS than the PPSS, he would perhaps expected [sic] the more efficacious refutation by means of employing the BhSS as the scriptural authority, which belonged to the just opposite of him, i.e. the Yogācāra. It was also to the Yogācāra tradition that the Chinese translators of the BhSS had belonged.” Hakamaya (1977), p. 480.

n.24For a more detailed comparison of the various editions, see Vinītā (2010), pp. 409–51. See also Sastri (1938), pp. 8–19, for a side-by-side translation of the three Chinese versions.

n.25In Taishō 575 and Taishō 576, the Buddha initiates the conversation by immediately describing the example of a man dreaming of a beautiful woman. This corresponds to the parallel passage found in PPSS.

n.26See n.­1.

n.27The following section from here up until the verse section at 1.­16 has a parallel in the PPSS.

n.28In the Chinese versions, the content of the dream varies slightly. Taishō 575: “Suppose a man had a dream in which he saw a crowd of people, and among them was a beautiful woman taking part in amusement.” Taishō 577: “Suppose a man, while asleep and dreaming, saw some people, and among them was a very beautiful woman, and he became completely overwhelmed.” The other versions more closely resemble the scenario here in D.

n.29In the Sanskrit, the Buddha’s question here varies slightly: “Great king, what do you think? Would the man be wise who would reminisce about the most beautiful woman in the land, whom he has enjoyed in his dream and consequently become passionate for?” (tat kiṃ manyase mahā­rāja, api tu sa puruṣaḥ paṇḍita­jātīyo bhavet, yaḥ svapnāntare bhuktāṃ jana­pada­kalyāṇīṃ striyam anusmaret, tataś cāsyāḥ pratiharṣanam). There are also slight variations in the way the question is worded in the Chinese and the editions of the parallel passage found in PPSS, although the meaning is essentially the same.

n.30Sanskrit reads “pleasant and unpleasant” (saumanasya­daurmanasya).

n.31PPSS reads “enacted by three types of conditioning of the body, four types of conditioning of speech, and three types of conditioning of mind” (lus kyis rnam pa gsum/ ngag gis ni rnam pa bzhi/ yid kyis ni rnam pa gsum mngon par ’du byed).

n.32Sanskrit: “ceases in the mind” (manasi nirudhyate); PPSS: “are exhausted and cease from the beginning” (thog ma nyid nas zad cing ’gags).

n.33Taishō 575: “Just as with the man waking from a dream who does not see the beautiful woman and the crowd.” (Note that in Taishō 575, the man dreams of a crowd; see n.­28.) In all the other versions, the woman appears in the mind of the person waking from the dream; here perhaps it is implied that while she is not seen in actuality, the mental attachment to her continues in the waking state. It is also possible that this is a scribal error in the Chinese.

n.34PPSS adds: “Great king, on the basis of two conditions, the last consciousness as the empowering [condition] and the action as the object [condition], the first consciousness pertaining to birth appears” (rgyal bo chen po de la rnam par shes pa tha ma’i dbang dang / las dmigs pa de dag dang / rkyen gnyis kyis skye ba’i char gtogs pa’i rnam par shes pa dang po ’byung). This phrase is not found in the quotation of PPSS found in the Śikṣāsamuccaya.

n.35Y, K, Taishō 575, and Taishō 577 omit being born among the asuras. All the other versions, including PPSS, mention six realms of birth. Hakamaya suggests that this indicates that this prose section originated from a non-Mahāyāna discourse. See Hakamaya (1977), p. 481.

n.36On the fact that this sūtra does not implicate an intermediate state between lives, see n.­19.

n.37Sanskrit and Taishō 577 have “arises” (pradurbhūta). PPSS also has “ceases” (’gag).

n.38Here Taishō 575 is unique among all the sources, reading “there is one phenomenon that transmigrates from this life into the future life.” Based on this statement, Sastri theorizes that this edition may indicate that the sūtra originates from schools, such as the Vātsiputrīya and Saṃmitīya, that assert the transmigration of a “person” (pudgala) or some similar phenomenon from one life to the next. See introduction, i.­10; and Sastri (1938), pp. xvi–xvii. However, with little context or commentary for this statement in Taishō 575, it is difficult to say exactly what is meant by this curious statement.

n.39The Chinese version of the PPSS (Taishō 310) reads (quoted from Vinītā’s translation), “It is perceived that both the performed karma and the experienced result do not perish. There exists neither a doer, nor an experiencer.” Vinītā (2010), p. 433, note b.

n.40Sanskrit has sarvadharmāṇaṃ (“all phenomena”), as does Candrakīrti (Madhyamakāvatāra­bhaṣya).

n.41Taishō 576 omits this part of the list (“transmigration after death is empty of transmigration after death”).

n.42Sanskrit omits this part of the list (“karma is empty of karma”).

n.43Sanskrit reads, “Then, great king, when the first consciousness pertaining to birth has consecutively arisen, there arises, without interruption, the mindstream concordant with one’s fortune, in which the experience of the result of the previous actions is perceived” (tasya ca mahārāja aupapattyaṃ­śikasya prathama­vijñānasya samanantaropapannasya tatsabhāgāvicalā cittasaṃtatiḥ pravartate yatra karmaṇāṃ vipākasya prati­saṃvedanā prajñāyate).

n.44In the following verse section Taishō 576 is significantly different; it is only six verses long and, in several instances, quite difficult to decipher. Taishō 575 and 577 are more coherent and generally accord with the Tibetan and Sanskrit, albeit with some differences in phrasing. See Sastri (1938), pp. 16–19 for a comparative translation of the three Chinese versions of this verse section.

n.45In addition to being quoted extensively among Tengyur treatises, the following verse section contains many parallels to other sūtras and treatises. These citations are provided in the appendix.

n.46The Tib. nga can (N, H: ngang can) is likely a shortened form of nga rgyal can, which conforms with the Sanskrit abhimānin. This is supported by the translation of the last two lines of this stanza in Prajñā­pradīpaṭīkā (Toh 3859), vol. 100, folio 8.a5: log pa’i nga rgyal ’jig rten la / / de ni kun rdzob bden par bstan / /.

n.47Meaning a combination of causes and conditions such as the object, the eye, the consciousness, etc.

n.48Translated from H: mi dpog, in reference to the Sanskrit na gāhate (“does not penetrate”). C, D, K, Y, J, N, and S read dpogs, which is the same verb but usually only used for the imperative tense.