Notes

n.1In a non-Buddhist context grahas and grahīs primarily appear in texts related to Āyurveda, non-Buddhist Tantric systems, and astrology. However, illness-causing spirit possessors also appear in classical literature such as the Ṛgveda, Atharvaveda, and the Mahābhārata. The interested reader is referred to The Self Possessed, Frederick M. Smith’s study on possession in South Asia that deals extensively with this subject and includes a discussion of traditional understandings of the ontology of disease-causing entities like grahas and bhūtas. See especially his chapter “The Medicalization of Possession in Āyurveda and Tantra” (Smith 2006, pp. 471–578).

n.2The Tibetan translation of both graha and grahī is the ungendered term gdon, though in some texts the gender of gdon is specified by adding the masculine (pho) or feminine (mo) prefix, e.g. pho gdon, mo gdon. In our text, however, all are referred to simply as gdon, with their gender determined by associated adjectives.

n.3See Smith 2006 for a discussion of grahīs and their function in (primarily non-Buddhist) Indic literature. Smith writes that “[o]ne of the most prominent varieties of possession in Indian literature is demonic, disease-producing possession” (Smith 2006, p. 272), and “[t]he most common word for such invasive disease-causing spirits in Āyurveda is grahī…” (Smith 2006, p. 476).

n.4Ibid, p. 482.

n.5See Tidwell, Nianggajia, and Fjeld 2023.

n.6Ibid., note 16.

n.7Smith 2006, Chapter 12; and Tidwell, Nianggajia, and Fjeld 2023.

n.8Ibid.

n.9See Norbu, p. 166 for further details on the srid pa ’pho dgu.

n.10For more on The Threefold Invocation Ritual, see The Threefold Invocation Ritual (spyan ’dren rgyud gsum pa, Toh 846), Introduction.

n.11Dalton 2016, p. 204.

n.12Ibid and Dalton and van Schaik 2006, p. 309, as well as The Threefold Invocation Ritual and The Threefold Ritual . Peter Skilling (1992, p. 122) has also observed a similar tripartite pattern among collections of Pali paritta texts used for protective ritual functions. It is therefore notable that none of the eighteen Sanskrit Dhāraṇī­saṃgraha texts cataloged and studied by Gergely Hidas (2021) include an opening invocation and closing dedicatory section framing the dhāraṇīs included in those collections. A more detailed study of the structure of Chinese Dhāranī­saṃgrahas, of which there are several, would surely prove interesting, but the scholars who have studied these Chinese works have not remarked on such a structure (see Shinohara 2014). It thus appears that this tripartite structure in Dhāraṇī­saṃgraha literature may remain unique to Tibetan collections.

n.13See the 84000 Knowledge Base page on the Compendium of Dhāraṇīs for further details on these collections and their texts. In Tshalpa Kangyurs that lack a Compendium of Dhāraṇīs section, most of the same texts are found at the end of the Tantra section. This is precisely the case with the present texts. In such Kangyurs, The Tantra of the Play of the Goddess Uṣṇīṣā and both The Threefold Ritual texts are found in the Tantra section.

n.14See the introduction to The Threefold Ritual. We also find a number of the very same Indic elements from The Threefold Invocation Ritual referenced in The Tantra of the Play of the Goddess Uṣṇīṣa: the seven mothers, the four sisters, Ekajatī, Indra, Yama, Vaiśravaṇa, grahas, yakṣīnīṣ, rākṣasīs, piśācīs, nāgas, garuḍas, bhūtas, kumbhāṇḍas, etc. The Threefold Invocation Ritual refers to these deities and other beings as “held by the hook of the vidyāmantra,” (1.­17) while The Tantra of the Play of the Goddess Uṣṇīṣa prescribes vidyāmantras as methods to overcome illnesses caused by such spirits.

n.15Dalton and van Schaik 2006, p. 309.

n.16In two extracanonical collections, the mdo rgyud gsungs rab rgya mtsho’i snying po gces par btus pa ’dod ’byung nor bu’i phreng ba compiled by Tāranātha, and the dpe rnying rtsa chen bris ma’i skor phyogs bsdus (provenance unknown), the title is given on the title page as The Play of Uṣṇīṣādevī (gtsug tor lha mo rol ma), but the title given in both the incipit and explicit of the text itself is The Uṣṇīṣā of the Playful Goddess Mahākālī (lha mo nag mo chen mo rol par byed pa’i gtsug tor zhes bya ba’i gzungs). The latter is the title under which it is preserved in the Western Tibetan canonical collections from the monasteries at Barden, Stagrimo, Stongde, and Egoo. However, in the Barden collection, the only one among these four collections that we have had access to, the short title given by the editors at the front of the text is also The Play of Uṣṇīṣādevī (gtsug tor lha mo rol ma) (https://www.istb.univie.ac.at/kanjur/rktsneu/sub/index.php.

n.17It is also possible that the redactors of some canonical collections simply did not have access to these works.

n.18This text, Toh 987, and all those contained in this same volume ( gzungs ’dus, waM ), are listed as being located in volume 101 of the Degé Kangyur by the Buddhist Digital Resource Center (BDRC). However, several other Kangyur databases‍—including the eKangyur that supplies the digital input version displayed by the 84000 Reading Room‍—list this work as being located in volume 102. This discrepancy is partly due to the fact that the two volumes of the gzungs ’dus section are an added supplement not mentioned in the original catalog, and also hinges on the fact that the compilers of the Tōhoku catalog placed another text‍—which forms a whole, very large volume‍—the Vimala­prabhā­nāma­kālacakra­tantra­ṭīkā (dus ’khor ’grel bshad dri med ’od, Toh 845), before the volume 100 of the Degé Kangyur, numbering it as vol. 101, although it is almost certainly intended to come right at the end of the Degé Kangyur texts as volume 102; indeed its final fifth chapter is often carried over and wrapped in the same volume as the Kangyur dkar chags (catalog). Please note this discrepancy when using the eKangyur viewer in this translation.

n.19The text nonetheless lists nineteen.

n.20In this list we read the numbers as ordinal rather than cardinal numbers based on a list of these grahīs later in the text that uses ordinal numbers.

n.21This translation follows C, J, K, Q, and Y in reading dbang byed. D omits dbang.

n.22dbyigs dug, a poison made of precious substances.

n.23While the Sanskrit name of this goddess is usually rendered as Ekajaṭā, all of the Tibetan translations here render her name Ekajaṭī, which is indeed how she is commonly referred to in Tibet.

n.24grong rgyun gi gdon. Literally the “continuity” of a dead person.

n.25We emend here and below to shugs ’gro ma from shugs sgrol ma on the basis of TLR, which reads shugs ’gro byams ma.

n.26We emend here to smyo byed gdon yin pas na from smyo byed gdon du za ba yin pas na, following a structure that repeats itself in the subsequent passages, in which illnesses or problems are attributed to one of the grahīs described in the first part of this text (using the same grammatical pattern: X gyi gdon yin pas…).

n.27This translation follows TLR in reading sgrib byed nag mo, which better fits the description in the verse and is one of the grahīs mentioned at the beginning of the text. D, and all other versions of The Tantra of the Play of the Goddess Uṣṇīṣā, read sgrol byed nag mo, “Black Liberator.”

n.28Here we emend to phung byed nag mo from bud med nag mo following the name and sequential position of this grahī in the initial list of eighteen primary emanations. Most, though not all, of the eighteen emanations appear in order in this section of the text.

n.29phya phya. Translation tentative.

n.30sems la mi dga’ phya phya. Translation tentative.

n.31Translation tentative. The Tibetan reads rkang lag bya btsag mang po, the meaning of which is unclear. In TLR the corresponding line reads rkang lag brkyang bskum bya smyang byed, which roughly means, “their legs and arms extend, retract, and stretch out.” This suggests a condition such as spasms.

n.32This reading follows K, Q, and Y in reading dbang ldan nag mo; D reads dbang ldan nag po.

n.33We emend to gnod sbyin nag mo from gnod sbyin nag po, given the name and sequential position of this grahī (not graha) given in the initial list of eighteen primary emanations. Not all of the eighteen emanations appear in order in this section of the text, but many of them do.

n.34According to Nebesky-Wojkowitz (1975, pp. 277 and 315), the ma bdud are a Tibetan class of “lesser” (ma) demons, who are distinct from the ya bdud, “greater demons.” On the ma bdud and ya bdud, see also Blondeau 2009, pp. 204, n. 13 and pp. 231–34.

n.35This translation follows TLR in reading ’dre’u; D reads dre.

n.36’jam dpal ro tan. The meaning of ro tan is ambiguous, and so has simply been phoneticized here.

n.37khram la ’debs pa. This phrase can be used to refer to suppressing those who harm the Dharma.

n.38rda bran. The meaning of this term could not be determined.

n.39This and the remaining mantras mix Tibetan words with transliterations of Sanskrit terms. The Sanskrit in each case means “pacify svāhā!”

n.40Emending ri tshe to ri rtse.

n.41yul sa. On the equivalence of the terms yul sa and yul lha (local deity), see Karmay 1996, p. 393.