Notes

n.1A natural reconstruction from the Tib. dga’ skyes. However, there seems to be no information on this individual. There is, for example, nothing relevant in G.P. Malalasekera’s Dictionary of Pali Proper Names or in Negi.

n.2The sūtra regularly uses the term gshin gyi ’jig rten, which is not a usual term, but is probably the same as the more usual gshin rje’i ’jig rten, i.e., the Sanskrit yamaloka , the world of Yama, Death, or the lord of death and ruler over the various departed ancestors (pitṛ). Cf. Negi s.v. gshin rje’i ’jig rten (= yamaloka ).

n.3Note that the Tibetan term mes po is often used in the sūtra to simply mean “grandfather,” but in other places the sūtra clearly uses it to mean “ancestor.” It is attested in the Mahāvyutpatti 3880 and in other glossaries as the equivalent of pitāmaha (paternal grandfathers), which can also mean simply the pitṛ or “ancestors.” See Monier-Williams, s.v. pitāmaha.

n.4On the four śrāddha-rites, see Sayers (2013), chapter 4. Brahmanical texts, like the Āpastamba Dharma­sūtra, advocate other offerings too, such as beans, barley, water, roots, fruits, cattle, buffalo, fish, and even rhinoceros. See Sayers (2013), pp. 107–8. Āpastamba’s list does not, however, seem to match the offerings of the family of Nandaja, viz., “horses, elephants, clothes, and a variety of ornaments, gold and silver, pearls, crystals, and other jewels, and a variety of delicious and sweet food and drink.”

n.5Cf. Renou and Filliozat (1985) vol. 1, §670: “All the images of future lives are physical; it is said in the Atharvaveda that cremation produces a new body that is revitalized, free from imperfections.” Cf. also ibid. §674: “The term fathers (pitṛ) designates, in the Vedas, the first ancestors, the founders of the human race, those who gave their names to the brahmanical families. But more generally it designates the dead as a whole, providing that they have been cremated or buried according to the rituals.” (Our translation from the original French of Louis Renou). Sayers (2013) provides a recent and thorough treatment of brahmanical ancestor worship; see also Hopkins (1992) as well as Jamison and Witzel (1992/2003) on the role of Vedic ritual and the offerings of food (piṇḍa) in death and the afterlife. See also Knipe (1977) on sapiṇḍīkaraṇa and Bodewitz (1999) on the Vedic afterlife and the role of the god of Death (yama) in particular. Shushan (2011) gives an overview of the basic Vedic eschatology. Doniger O’Flaherty (1983) looks at the early Vedic and Purānic antecedents of theories of karma and reincarnation, notably the idea of redeath (punarmṛtyu), a type of dissolution of the afterlife that is to be prevented by ritual.

n.6Sayers (2013), pp. 86–99.

n.7See Namai (1991). The argument of Āryaśūra and Dharmakīrti alike is that consciousness can only exist when preceded by a previous consciousness; thus the initial consciousness of a baby must have an anterior consciousness in a previous life.

n.8The Sanskrit of verse five of the Pratītya­samutpāda­hṛdaya­kārikā, as cited in Candrakīrti’s Prasannapadā 428.11–12 and 551.14 (ed. La Vallée Poussin), reads: svādhyāya­dīpa­mudrā­darpaṇa­ghoṣārkakānta­bījāmlaiḥ / skandha­pratisaṃdhir asaṃkramaś ca vivadvadbhir avadhāryau* // *Prasannapadā 428: upadhāryau. The translation is our own, informed by the sūtra’s own explanation of the eight. Cf. the translations of this verse in May (1959), p. 259, and Skilling (1997), p. 253. For further canonical sources for each of the eight examples, see May (1959), n. 908 and n. 933. On the “sunstone” (arkakānta = sūryakānta ), which is contrasted with the “moonstone” (candrakānta), see Lamotte (1949), p. 446, n. 1. Cf. Kālidāsa’s Abhi­jñāna­śākuntala II, 7: “Indeed, burning fiery energy lies hidden in ascetics focussed on calm, just as sūryakāntas which are cool enough to be touched spit out their [fiery energy] when another such energy prevails.” (śama­pradhāneṣu tapodhaneṣu gūḍaṃ hi dāhātmakam asti tejaḥ / sparśānukūlā iva sūrya­kāntās tad anya­tejo’bhibhavād vamanti). Less poetically put, a sunstone functions as a magnifying glass: it emits heat when aligned with the blazing sun.

n.9Skilling (1997), p. 255: “The eight similes are not only identical to those of the Pratītya­samutpāda­hṛdaya-kārikā, but also occur in a very similar order: this is sufficient to establish a relationship between the two texts. We may therefore conclude that one of the texts is referring to the other. Since the similes are not only listed in the sūtra, but also described at length, and since the similes are only a part of the long sūtra, of which they form a natural component, I suggest that it is Nāgārjuna who has based his verse on the sūtra, and not the composer or editor of the sūtra who has adapted Nāgārjuna’s verse into his text. It is indeed characteristic of Nāgārjuna’s style to give brief paraphrases of canonical passages in his important works, such as the Madhyamaka-kārikās, the Ratnāvalī, and the Suhṛllekha.” Skilling points out that Sanskrit and Tibetan texts of Nāgārjuna’s verse 5 have the eight analogies in slightly different orders, but that the order in the sūtra is exactly the same as that in the Tibetan of verse 5.

n.10The other entry for tshe ’phos pa in the Mahāvyutpatti, i.e., 2980, gives the Sanskrit as jāti­vyativṛttaḥ, “one who has left birth”; the Tibetan given here is tshe ’phos pa’am tshe rjes pa, “leaving life or changing lives.”

n.11See n.­40.

n.12Note too that in this version of the title found at the end of the text, bstan pa’i mdo (“The Sūtra Teaching…”) replaces zhus pa’i mdo (The Sūtra of Questions Regarding…”). This title given at the end of the text is the one that figures in Butön’s list of canonical translations (F.144.b), Situ Paṇchen’s catalogue (dkar chag) of the Degé Kangyur (F.134.b), and quite frequently (or as the variant ’chi ’pho ba ji ltar ’gyur ba lung bstan pa’i mdo) in later commentarial literature citing the sūtra.

n.13The them spangs ma recension Kangyurs have Āyuḥyathabhutāgraha­paripṛcchā­sūtra (S, L: Āyuḥ­yatha­bhutāgraha­sūtra), which also remains dubious even when corrected to Āyuḥ­yathābhūta­graha­paripṛcchā­sūtra (S: Āyuḥ­yathābhūta­graha­sūtra). Cf. Skilling (1997), p. 257: “Both titles seem awkward and unlikely, and may be later concoctions.”

n.14The sūtra is absent from both early ninth century inventories, the Denkarma (ldan dkar ma) and Phangthangma (’phang thang ma), but is mentioned in Butön’s fourteenth century list of canonical texts. On Tibetan institutions of translation and their procedures, before and during the ninth century, see Scherrer-Schaub (2002), which also provides, inter alia, a bibliography of the main research on these subjects.

n.15See Bibliography and Abbreviations for details.

n.16We find sman pa in all editions. It clearly needs to be read in its attested sense of phan pa (“benefit”) and not in the sense of “a doctor.” See Zhang Yisun et al. (1985), s.v. sman pa.

n.17gsol ba = honorific for lto chas.

n.18ris ’thun pa. The term is later replaced by mes po, “ancestors.”

n.19It is clear in the reply later that the type of “switching” being discussed is one that would occur arbitrarily, without karmic or other causes.

n.20phyir zhing mang ba, literally, “once again many/more.”

n.21mang ba, “more.”

n.22rang gi sems kyang rang gis mi mthong na. Here the point seems to be that they do not literally see their minds. Seeing is reserved for physical objects.

n.23thog ma mes po gcig las gyes pa yin na.

n.24The argument seems to be as follows. Among our presently existing kin some are enemies (dgra), unfriendly to each other, while some are not. If we take the collection of kinfolk from the first ancestor on, the same would hold. We thus could not determine that our kinfolk are precisely the ones associated as friends (grogs), and other people’s kin are the ones that are not.

n.25phyogs gcig nas. The textual passage is long and difficult, and the translation is therefore unsure. It appears that the argument is essentially an elaboration on the argument of the previous paragraph. A collection of ancestors, some of whom are very different from and even antagonistic to the others, would have to be somehow apprehended and befriended as a unified harmonious party.

n.26rnyed pa = lābha.

n.27dri za ’chi ka ma’i sems la nye bar ’jug pa zhes bya ba’i rigs. We unfortunately have no information on this type of gandharva.

n.28Read with D, A, S, L: da ltar tshe ma ’phos pa’i rang gi pha ma. Cᴅs: da ltar tshe ’phos pa’i rang gi pha ma.

n.29mi des rmis pa bzhin du gang rmis pa’i pha ma’am / spun zla’am / bran khol lam gzhan su yang rung ba rmi lam na snang ba der* tshor ba yod na ni rmis pa de bden pa yin na / des rmis pa’i pha ma’am / spun zla’am / bran khol lam / gzhan su yang rung rmi lam na snang ba de bden par ji ltar bzung /. *S, L: der; D, A: des. Cᴅs appears to be truncated: mi des rmis pa bzhin du gang rmis pa’i pha ma’am / spun zla’am / bral [sic] khol lam / su yang rung ba rmi lam na snang ba des tshor ba yod na ni rmis pa de bden pa yin na / des rmis pa ltar med pas bden par ji ltar bzung. “If the parents, relatives, and servants they dreamed of, or any others appearing in their dream, actually were to have the feelings in question, just as that person dreamed they did, then that which they dreamed would have been real. They do not exist as they dreamed them, so how could they be thought to be real?”

n.30D, A, S, L: tshe ’phos pa’i rmis pa tshe ’phos pa de yin par ga la ’gyur; Cᴅs: tshe ma ’phos pas rmis pa de tshe ’phos pa de yin pa ga la ’gyur. “How could what is dreamed by a non-deceased person ever be that deceased person?”

n.31We read with S and L: ’grams. This is a variant of gram, the intransitive verb meaning “to be spread, dispersed, scattered.” See Golstein (2004), s.v. ’grams. The numerous implausible variants in editions (D, A: ’drams; K, Y: ’drangs; C: ’drems; S, L: ’grams; Cᴅs: missing) suggest that the word in question was not understood by scribes. A search of the BDRC site shows that ’drams (homophonous with ’grams) is not attested elsewhere.

n.32We follow D and A: grong dang khang pa de la sems yod na ni khang pa de’i sems snang du rung na. More literally, “If the city and house had a mind” (sems yod na ni). All witnesses in A, as well as S and L, have sems (“mind”), except for C as recorded in A, which has sems can (“being”). However, C does not have sems can in any of the subsequent text of this argument.

n.33A reads des rdo, but one should no doubt read de sa rdo. The intersyllabic dot (tsheg) appears quite clearly in S and L.

n.34Read snang with D; A has mistakenly recorded snad.

n.35We follow D, A: bstangs; S: btang; L: gtang.

n.36D, A: bstangs; S, L: btang.

n.37gshin gyi ’jig rten is not a usual term, but is probably the same as the more usual gshin rje’i ’jig rten, i.e., the Sanskrit yamaloka , the world of Yama, the lord of Death and ruler over the various departed ancestors (pitṛ). Cf. Negi s.v. gshin rje’i ’jig rten (= yamaloka ). On the Vedic afterlife, yamaloka , see Bodewitz (1999), Shushan (2011).

n.38Read with S, L, Cᴅs: tshe ’phos pa la; D, A: tshe ’phos pa las.

n.39Read with S, L: … rgyan thogs so zhes smra na / de ltar snang ba de yang…; D, A, Cᴅs: … rgyan thogs so // zhes smra zhing de ltar snang ba de yang….

n.40We have translated on the basis of S and L. Here is the whole passage in those editions: kye rgyal po chen po ’jig rten pa dag gang gis mthong ba dang / pha ma la sogs pas rmi lam du rmis te / tshe ’phos pa la bsngos pas gshin de bza’ btung gis mgu’o // bzhon pa zhon no // *gos gon no* // rgyan thogs so zhes smra na / de ltar snang ba de yang bza’ btung gis mi mgu’o // zhon pa med do // **gos mi gon no** // rgyan mi thogs so zhes smra zhing de ltar snang ba dag dri za ’chi ka’i sems la nye bar ’jug pa dang / mi ma yin pa dag gis de ltar snang bar byed pa yod do //. *…* omitted in L. **…** omitted in L. The Tibetan text may be corrupt in all editions; it has a seeming over-use of the phrase de ltar snang ba several times in the same sentence. We have tried to take them into account as best as possible, but the translation remains tentative. The basic point, however, seems to be as earlier (see 1.­25), viz., people think that offerings will enable the dead to eat, ride, be clothed, etc., but what actually happens is that certain sorts of gandharvas create an apparition of a dissatisfied deceased person so that they themselves can profit from the clothes, food, etc., offered by relatives to the deceased.

n.41There are quite divergent readings of more or less equal plausibility. We read with D and A: tshe ma ’phos pa’i mtshan ma dang dan rtags ’jig rten pa gang dag smra ba ni. Cᴅs: tshe ’phos pa’i mtshan ma dang dan rtags ’jig rten pa gang dag smra ba ni; S, L: tshe ma ’phos pa’i mtshan ma dang dan rtags dang bcas pa’i ’jig rten pa gang dag smra ba ni. “What one calls worldlings who have characteristics and distinctive signs of the living….”

n.42The Sanskrit names for the sorts of gandharva and bhūta spirits are given with variants in the canonical editions as well as in Cᴅs. We have followed D and A here, but are unable to ascertain the correct Sanskrit names for these spirits and have simply given the Tibetan transcription as is. Cf. S, L: ba tsi na and ba ra hin ti.

n.43bsgrub.

n.44shing tha dad pa brgya stong bye ba dag; literally, “the one hundred thousand times ten billion different sorts of wood.”

n.45’dod sred. This term has Sanskrit equivalents, but it is not sufficiently clear what the equivalent would be if it is being used as a proper name.

n.46Here the term ’pho ba clearly has the sense of “transmigrate,” “pass from one life to the next.” When the Tibetan reads rtag pa mi ’pho here, i.e., “it is not something permanent that transmigrates/passes,” the Tibetan mi ’pho is to be understood like the Sanskrit term asaṃkrama (= Tib. mi ’pho ba) in verse 5 of the Pratītya­samutpāda­hṛdaya­kārikā. See i.­10 in our introduction.

n.47The point seems to be that the sūryakānta itself is cool to the touch and yet it emits heat. See n.­8 for a verse from Kālidāsa to this effect.

n.48The term mu chags su is obscure. We have taken it as having a sense like that of mu mthud kyis, “continually.”

n.49The title given here at the end of the sūtra, in all Kangyurs, is ’chi ’pho ba ji ltar ’gyur ba bstan pa’i mdo, and differs from the main title at the beginning. See i.­11 and n.­12.