Notes

n.1The four texts are Toh 594, 595 (the present text), 596, and 598. The first three share the same title: Crown Victory of All Tathāgatas: The Uṣṇīṣavijayā Dhāraṇī with Its Ritual Manual (de bzhin gshegs pa thams cad kyi gtsug tor rnam par rgyal ba zhes bya ba’i gzungs rtog pa dang bcas pa). The fourth has an ever-so-slightly different title: Crown Victory of All Tathāgatas: A Ritual Manual for the Uṣṇīṣavijayā Dhāraṇī (de bzhin gshegs pa thams cad kyi gtsug tor rnam par rgyal ma’i gzungs zhes bya ba’i rtog pa).

n.2The surviving Sanskrit work seems, more properly, to be titled the Sarvagati (rather than Sarvadurgati)-pariśodhana-uṣṇīṣavijayā-dhāraṇī, but either way the title provides evidence of the relationship between the Uṣṇīṣavijayā and Sarvadurgatipariśodhana corpuses. For more on this relationship see J. Dalton 2016 and forthcoming. The point here is simply that Toh 597 is the Uṣṇīṣavijayā-dhāraṇī, rather than the Uṣṇīṣavijayā-“ dhāraṇī with its ritual manual” (kalpasahitā).

n.3Hidas 2020, p. 141. See also Hidas 2021b, which catalogs a number of Indic dhāraṇīsaṃgraha collections, many of which include the uṣṇīṣavijayā dhāraṇī.

n.4The Sanskrit of this work is preserved in what Gregory Schopen calls the “Los Angeles Manuscript,” though it appears to be held currently in Japan. This is an early manuscript from Bamiyan-Gilgit that Schopen transcribed and translated into English in an unpublished work, which we are grateful to Jacob Dalton for sharing. In addition to being incomplete, probably due to the loss of a folio, the manuscript lacks several passages that are found in the Tibetan translation of Toh 597 and contains a few passages that are absent in that translation, including two passages that are found in Toh 594. Nonetheless, the Sanskrit manuscript is by and large the same work that is translated into Tibetan as Toh 597. More recently, the Sanskrit of the very same manuscript was studied by Gudrun Meltzer in a 2007 “limited distribution report” (Silk 2021, p. 108), to which we have not had access, as well as by Unebe Toshiya, who published the Sanskrit along with a Japanese translation in a 2015 article.

n.5This text has been edited on the basis of ten Nepalese Sanskrit manuscripts and translated into English in Hidas 2020. From among the works belonging to this group that are preserved in the Tibetan canon, the Sanskrit text is most closely parallel, though not identical, with the present text, Toh 595.

n.6The first translation is Taishō 967, followed by Taishō 968–971 and Taishō 974 (Chou 1945, p. 322).

n.7According to Chou, the ritual manuals surviving in Chinese are Taishō 972–973 (Chou 1945, p. 322). Hidas 2020 notes that the full set of Uṣṇīṣavijayā-related texts found in the Taishō canon includes Taishō 968–974, 978, and 979.

n.8Hidas mentions that Taishō 978 “stands closest to the Nepalese tradition” of the Sanskrit work that he has edited, which is also how he describes the relationship between the Sanskrit work and Toh 595 (Hidas 2020, p. 156n6–7). A comparison of Toh 595 and Taishō 978 shows that while neither exactly matches the Sanskrit text that Hidas edited, the Tibetan and Chinese are indeed translations of the same Sanskrit work and contain identical material apart from the Chinese translation’s inclusion of a single, very short passage about a toothbrush that is absent in the Tibetan translation (Toh 595) but present in some of the other uṣṇīṣavijayā texts in the Tibetan canon.

n.9Sørensen 2011a, p. 165.

n.10Sørensen 2011b, p. 386. See also Silk 2021 for further mention of the usages of the uṣṇīṣavijayā dhāraṇī, often alongside the Heart Sūtra, in China.

n.11Copp 2005, p. 4. For further details see Copp 2005, which addresses the topic of dhāraṇīs in medieval China using the uṣṇīṣavijayā dhāraṇī as a case study.

n.12See IOL Tib J 307/PT 54, PT 6, and PT 368 for Tibetan translations of the work, IOL TIB J 322 and IOL Tib J 349/3 for a Tibetan translation of the dhāraṇī alone (not the whole text), and IOL Tib J 466/2, IOL Tib J 547, IOL Tib J 1134, IOL Tib J 1498, PT72, and PT73 for Tibetan transliterations of the Sanskrit dhāraṇī alone (J. Dalton and van Schaik 2006; accessed through The International Dunhuang Project: The Silk Road Online). The translations of the primary uṣṇīṣavijayā dhāraṇī text (Toh 597) appearing at Dunhuang include at least one passage parallel with rites described in our text but missing from the primary text in its Tibetan canonical translation, though present in the surviving Sanskrit manuscript corresponding to Toh 597 studied by Schopen (see J. Dalton forthcoming; Schopen unpublished).

n.13Schmid 2011, pp. 372–73.

n.14J. Dalton 2016 and forthcoming.

n.15See Hidas 2014, pp. 110–11.

n.16Bühnemann 2014; Rospatt 2015, p. 821.

n.17See Bhattacharyya 1928, vol. 2.

n.18The three are Toh 3377, 3248, and 3580, translated respectively by Khampa Lotsāwa Bari Chödrak (khams pa lo tsA ba ba ri chos grags, eleventh century, also translator of Toh 594 and possibly of this text, see n.­52); Patshap Lotsāwa Tsültrim Gyaltsen (pa tshab lo tsA ba tshul khrims rgyal mtshan, twelfth century); and Yarlung Lotsāwa Trakpa Gyaltsen (yar klungs lo tsA ba grags pa rgyal mtshan, late thirteenth or early fourteenth century).

n.19See Chandra 1980.

n.20These Southeast Asian texts are not included in the official Pali Canon of the Theravāda tradition and are unknown in Sri Lanka. In mainland Southeast Asia, however, they are popular in rituals for extending life and in funeral rites. Whether they reflect the diffusion of texts and practices directly from India prior to the relatively recent evolution of Theravāda orthodoxy, or were transmitted via Chinese along with Chinese migrations and cultural influence in the region, remains an open question. For a detailed study of these texts and their possible origins, see Cicuzza (ed.) 2018.

n.21Phangthangma (2003), p. 23. While the phrase cho ga dang bcas pa (Skt. vidhisahitā) is functionally equivalent to the phrase rtog pa dang bcas pa (Skt. kalpasahitā), we unfortunately have no way of knowing whether this text was or resembled the primary uṣṇīṣavijayā dhāraṇī text (Toh 597) with a ritual manual attached to it, or if it resembled the present text or any of the other works in the modern canons titled Uṣṇīṣavijayā-dhāraṇī-kalpasahitā (i.e., Toh 594 or 595); Toh 596 is too short to correspond with the text identified in the Phangthangma as having 120 ślokas, and Toh 598 is an unlikely candidate because, while it shares the opening narrative with the other texts in this set, it seems to represent a separate, and later, ritual system. The Phangthangma also lists what may be a copy of the dhāraṇī alone, outside of the framework of a sūtra (Phangthangma, p. 31). The other imperial catalog, the Denkarma, lists only the primary uṣṇīṣavijayā dhāraṇī text, identified clearly as the text included in the later canons as Toh 597 by its full title in that catalog: the Sarvadurgatipariśodhana-uṣṇīṣavijayā-dhāraṇī (Lalou 1953, p. 327).

n.22Unlike many dhāraṇī texts (including Toh 597), which tend to appear both in the Tantra Collection and the Dhāraṇī Collection sections of the Kangyurs, the texts in the genre of dhāraṇī-kalpas seem to appear exclusively in the Tantra Collection section of the Kangyurs.

n.23The text is F 631, Phukdrak Kangyur, vol. 117 (rgyud, dza), F.224.a–231.a. It should be noted, however, that the version of the dhāraṇī preserved in F 631 differs from the dhāraṇī in the present text much more substantially than any of the versions in the mainstream Kangyurs.

n.24Probably Sumpa Lotsāwa Dharma Yontan (sum pa lo tsA ba dhar ma yon tan), a translator and teacher of Sakya Jetsun Drakpa Gyaltsen, but possibly his uncle, also called Sumpa Lotsāwa, Palchok Dangpö Dorje (dpal mchog dang po’i rdor rje). Both studied in Nepal. See Treasury of Lives.

n.25The title of this text and the first part, through the presentation of the dhāraṇī, are exactly parallel with the opening of the longer Toh 594 and closely parallel with the opening of the very short Toh 596, though that text has made a number of emendations, improving upon some readings and making others even less smooth and more complicated. The text is likewise parallel with the opening of Toh 598, though that text has more extensively edited the opening and even some aspects of the dhāraṇī itself.

n.26chos yang dag par sdud pa’i phug khang bzangs mchog. The Sanskrit in the closely parallel text edited by Hidas reads dharmasaṃgītimahāguhyaprāsāde (Hidas 2020, p. 152). The Tibetan phrase is awkward, and it seems that there may have been some textual corruption. What has been rendered in Tibetan as phug seems to be guhya in the Sanskrit parallel; perhaps the Tibetan translators were reading guhā‍—which does translate to phug‍—rather than guhya. Although we cannot be sure that the surviving Sanskrit witnesses represent the older reading, they provide a more coherent reading than the one in our Tibetan witnesses, so we have translated this word following the Sanskrit, rather than the Tibetan witness.

n.27The text here is corrupt and appears to have transmitted a line slip, where a line from slightly lower in the text made its way incongruously to a place where it does not belong, rendering this sentence nonsensical. The Sanskrit text and the parallel passage in Toh 598 both lack this error, confirming that it is a textual corruption. We have relied upon Hidas’ Sanskrit edition to repair the Tibetan text here. The Tibetan reads de rnams kyi phyir ’khor gyi dkyil ’khor la nye bar gzigs te/ kun du gzigs pa’i dpal gyi mtshan de bzhin gshegs pa thams cad kyi gtsug tor rnam par rgyal ba zhes bya ba’i gzungs ’di bcang ba dang / bklags pa dang / gzhan la rgya cher yang dag par ston pa ni tshe ring ba sgrub par byed pa’i ched du’o. The passage in bold has been incongruously lifted from its proper place several lines down in the text and added here. The Sanskrit passage lacks this error but also includes several additional words absent in the Tibetan. However, as the Sanskrit and Tibetan texts are not identical in other places in this parallel passage either, and since the Tibetan text makes perfect sense without these additional elements, we have not taken the liberty of adding them in the English translation. The Sanskrit passage, in Hidas’ edition (with the elements absent in the Tibetan text indicated in bold), reads teṣāṃ arthāya hitāya sukhāya imāṃ sarvatathāgatoṣṇīṣavijayā-nāma-dhāraṇīṃ dhārayed vācayed deśayet paryavāpnuyāt parebhyaś ca vistareṇa samprakāśayet | dīrghāyuṣkāṇām upādāyeti (Hidas 2020, p. 152). The passage in Toh 598 reads de rnams kyi don du tshe ring bar nye bar bsgrub par bya ba’i phyir/ de bzhin gshegs pa thams cad kyi gtsug tor rnam par rgyal ma zhes bya ba’i gzungs ’di gzung bar bya/ gzhan la rgya cher yang dag par bstan par bya’o.

n.28thams cad dang ldan pa.

n.29mama is not present in Hidas’ edition of the Sanskrit manuscripts.

n.30Hidas’s edition of the Sanskrit reads sarvatathāgatamātre, a plausible variant unattested in Tibetan sources.

n.31There is some variation in this phrase across the Tibetan and Sanskrit sources. Toh 594, 597, and 596 read sadā me; this text, Toh 598, and Toh 984 read me sadā; and, Hidas’ Sanskrit edition has mama sadā. The meaning is the same in all cases.

n.32The Yongle and Kangxi versions of this text include the line sarvatathāgatasamayādhiṣthānādhiṣṭhite here. The Degé version of Toh 597 includes the phrase samantān mocaya mocaya ādhiṣṭhāna, though it is absent in other canonical recensions of the same translation. Hidas’s Sanskrit edition includes sarvatathāgatahṛdayādhiṣṭhānādhiṣṭhite at this point.

n.33māṃ is absent in Hidas’ Sanskrit edition.

n.34Hidas has translated the dhāraṇī based on his edition, and rather than retranslate it, we give his translation here. Substantive variants between the Sanskrit basis for his translation and the Degé have been noted above. “Oṁ veneration to the glorious Buddha distinguished in all the Three Worlds. Namely, oṁ bhrūṃ bhrūṃ bhrūṃ, purge, purge, purify, purify, O Unequalled Enveloping Splendor Sparkle Destiny Sky, O the One of Purified Nature, O the One Purified by the Topknot Victory, let all Tathāgatas consecrate me with consecrations of the nectar of the excellent Sugata’s words along with great seals and mantrapadas, oṁ bring, bring, O the One who Nourishes Life, purge, purge, purify, purify, O the One Purified by Sky Nature, O the One Purified by the Topknot Victory, O the One Impelled by Thousand Rays, O the One Beholding all Tathāgatas, O the One Fulfilling the Six Perfections, O Mother of all Tathāgatas, O the One Established in the Ten Stages, O the One Empowered by the Empowerment of the Heart of all Tathāgatas, oṁ O Seal, O Seal, O Great Seal, O the One Purified by the Firmness of the Vajra Body, O the One Purged of all Obscurations Resulting from Actions, turn back for me O Life-purged One, O the One Empowered by the Empowerment of the Vow of all Tathāgatas, oṁ muni muni, mahāmuni, vimuni vimuni, mahāvimuni, mati mati, mahāmati, mamati, sumati, O the One Purified by Truth and the True Goal, O the One Purged by a Burst Open Mind, oṁ he he, triumph triumph, succeed succeed, recollect recollect, manifest manifest, expand expand, O the One Empowered by the Empowerment of all Buddhas, oṁ O Pure One, O Pure One, O Awakened One, O Awakened One, O Vajra, O Vajra, O Great Vajra, O Vajra-essence, O Victory-essence, O Triumph-essence, O Vajra-flame-essence, O Vajra-born, O Vajra-produced, O Vajra, O the One with a Vajra, let my body become a vajra and that of all beings, let there be body-purification for me and purification of all destinies, O the One Empowered by the Empowerment of the Heart of all Tathāgatas, let all Tathāgatas provide encouragement, oṁ awake awake, succeed succeed, awaken awaken, wake up, wake up, liberate liberate, release release, purge purge, purify purify, liberate completely, O the One Purified by an Enveloping Ray, O the One Empowered by the Empowerment of the Heart of all Tathāgatas, oṁ O Seal O Seal, O Great Seal, O Great Seal and Mantrapada svāhā” (Hidas 2020, p. 154). The exact parallel with Toh 594 that began with the start of the text ends here (Toh 594, folio 230.b.6), but further parallels with a different section of Toh 594 (folio 233.b.7) begin again immediately with the line starting “Child of noble family…”

n.35The Tibetan text reads ’dzin pa (“holder”), while the parallel Sanskrit text edited by Hidas reads -nivāraṇī (“destroyer”) (Hidas 2020, p. 153). Moreover, three of Hidas’ Sanskrit witnesses here read -harā, also meaning “destroyer” (ibid., p. 161n142). Given the lack of sense in the Tibetan passage and the consistency of meaning in the Sanskrit witnesses, this translation follows the Sanskrit reading.

n.36kham phor kha sbyar gyi nang du. Here we follow the Stok, Narthang, Choné, Urga, and Lhasa Kangyurs, which read kham phor (“clay cup”), rather than the Degé, which reads the nonsensical kham por.

n.37All recensions consulted repeat “twenty-one” twice, reading nyi shu rtsa gcig rtsa gcig, but when the same thing happens in a passage below the extraneous rtsa gcig is not present in the Yongle and Kangxi Kangyurs, so we have emended the text and omitted it here, too.

n.38Here the text first brings up the name of the second dhāraṇī that is taught in this text: the dhāraṇī of limitless life (tshe dpag tu med pa), which is also the name of the Buddha Amitāyus. Presumably the benefits mentioned from here on refer to that dhāraṇī, which is itself taught by the Buddha Amitāyus below.

n.39Here we follow the Choné Kangyur, which reads lan nyi shu rtsa gcig, rather than the Degé Kangyur, which reads lan nyi shu rtsa gcig rtsa gcig.

n.40The term ’od dpag tu med pa, which is translated here as “light rays” because it is part of a larger phrase, is also the name of the Buddha Amitābha, of whom Amitāyus is generally understood to be simply a different form.

n.41There appears to have been considerable confusion‍—especially, but not exclusively, in Tibetan‍—in the transmission of this dhāraṇī, which also appears in the parallel passage in Toh 594. Here, in Toh 595, the Degé recension correctly transmits the dhāraṇī with the words amṛte, amṛtobdhave, and so forth, whereas the Degé recension of Toh 594 consistently reads amite, amitodbhave, etc. The dhāraṇī as preserved in these two works in other Kangyurs varies between the two readings, sometimes even giving both readings at various places in the dhāraṇī, e.g., amṛte amitodbhave. A Sanskrit recension of a very similar (but not identical) Amitābha dhāraṇī preserved in a tenth-century manuscript from Dunhuang consistently reads amṛte amṛtodbhave, etc. (see Hidas 2014, pp. 110–11). In his edition of that dhāraṇī, Hidas also cites other Sanskrit sources (the Mahāpratisarā, Sādhanamālā, and Sarvadurgatipariśodhana, which include portions of the dhāraṇī with the forms amṛte, etc. Hidas 2014, p. 111n42 and n43 and p. 112n47). He also, however, cites the Uṣṇīṣavijayāsādhana in the Sādhanamālā, which preserves the readings amite amitodbhave, etc. (See Hidas 2014, p. 112n47). The version preserved in the modern Tibetan tradition, in most cases, of a short part of this dhāraṇī (also found below in the present text) that is associated with Uṣṇīṣavijayā reads oṁ amṛta āyurdade svāhā, likewise using the form amṛta. On the preponderance of the evidence, we have thus adopted the reading amṛte here.

n.42The parallel passage in Toh 594 here adds amṛtasaṃbhave, but this is absent in all recensions of our text.

n.43We have emended from the Degé reading here, āyurdate, to the theoretically more “correct” form āyurdatte on the basis that the latter is attested later in this same text, in the extract from this dhāraṇī repeated below at 1.­26. However, it should be noted that the alternative spelling āyurdade, attested only in the Stok Palace and Phukdrak Kangyurs for this text at this point, is common elsewhere in this group of Uṣṇīṣavijayā texts, and seems to be the preponderant form in use in numerous sādhanas and other texts in the Tibetan Buddhist tradition. See also n.­49.

n.44We read kariye following the Yongle, Lithang, Peking, and Choné Kangyurs and the parallel passage from Toh 594, rather than karīye in the Degé.

n.45de’i ring bsrel gyi snying po can du byas te. This translation is tentative.

n.46tshe dang blo ster ba’i yon byin te. This translation is tentative.

n.47stong phrag gcig nas brtsams te ji srid lnga’am gcig gi bar du’o. This line is perplexing, and we remain unclear about the number of caityas that the text is instructing the practitioner to make.

n.48rang dang ’dra ba.

n.49The spelling amṛtāyurdade, which is the preponderant form for this mantra in later Tibetan works, is attested for this text only in the Stok Palace, Shey, and Phukdrak Kangyurs. See also n.­43.

n.50The fact that “life-extending and intelligence-sharpening” here refers to the additional dhāraṇī oṁ amṛtāyurdatte svāhā is suggested by the passage above, which refers to that dhāraṇī with precisely that phrase in the context of adding it to the main uṣṇīṣavijayā dhāraṇī.

n.51The Sanskrit word for “glory” is śrī, and the Sanskrit word for “wood apple” is śrīphala (“the śrī fruit”).

n.52This text has no translator’s colophon. However, Toh 594, which is nearly identical to this text apart from some variations (it has been ever-so-slightly edited for clarity by employing infrequent substitutions of different types of connective particles, using the less archaic form of a word, and so forth), does have a translator’s colophon, which reads, “It was translated, edited, and finalized by the scholar Dharmasena and by Khampa Lotsāwa, the monk Bari.” This suggests that Toh 595 was translated by the same team.