Notes

n.1For example, Étienne Lamotte once called the Vimalakīrti­nirdeśa “perhaps the crown jewel of the Buddhist literature of the Great Vehicle,” in L’Enseignment de Vimalakīrti (Lamotte 1987, p. v), while more recently Jonathan Silk has made a softer claim, describing it as “one of the most radiant stars in the firmament of Mahāyāna sūtra literature,” on the back cover of Vimalakīrtinirdeśa: The Teaching of Vimalakīrti (Gómez and Harrison 2022). On the appreciation (or lack thereof) of the Lalitavistara as a literary work, see Silk 2022, especially pp. 285 and following. We would also like to acknowledge here our indebtedness to Gómez and Harrison’s translation of the Vimalakīrti­nirdeśa. We have used many of their translation choices of terms and phrases in this translation, as well as their explanations for our glossary.

n.2For the Sanskrit text of these passages, see Study Group on Buddhist Sanskrit Literature 2005, which also includes the Tibetan and Chinese translations. For a translation from Tibetan of the above passages, see Robert A. F. Thurman, trans., The Teaching of Vimalakīrti , Toh 176, 6.­13 and 4.­1 (2017).

n.3The relevant passages are noted in the body of the translation. For the Sanskrit editions of these works, see the following: Lévi 1907; La Vallée Poussin 1903; and Tucci 1971; and for a translation of the final work, which is not extant in Sanskrit, see Gareth Sparham, trans., The Long Explanation of the Noble Perfection of Wisdom in One Hundred Thousand, Twenty-Five Thousand, and Eighteen Thousand Lines , Toh 3808 (2022).

n.4The relevant passages are cited in the body of the translation. For the former, a partial Sanskrit manuscript has recently been identified in Tibet, on which see Wang et al., 2020. A complete translation of it was made from Tibetan into English by Bhikkhu Pāsādika and published serially in the journal Linh-Son publication d’études bouddhiques, beginning with “The Sūtrasamuccaya – An English Translation from the Tibetan Version of the Sanskrit Original (I).” For the latter, the classical Sanskrit edition is Cecil Bendall’s Çikshāsamuccaya: A Compendium of Buddhistic Teaching; it has also been translated into English on two occasions: Cecil Bendall’s and W. H. D. Rouse’s Śikṣā Samuccaya (1922) and Charles Goodman’s The Training Anthology of Śāntideva (2016).

n.5The relevant passages are cited in the body of the translation. Sanskrit edition: Bunyiu Nanjio, The Laṅkāvatāra Sūtra (1923); English translation: Daisetz Teitaro Suzuki, The Lankavatara Sutra: A Mahayana Text (1932).

n.6The relevant passages are cited in the body of the translation. It is also possible that the Tathāgataguhya reworks the passages as they are found in the Bodhisattvapiṭaka, or that both texts share a common source or sources. Parts of both passages are also quoted in the Śikṣāsamuccaya and Sūtrasamuccaya, and there they are explicitly attributed to the Tathāgataguhya. A Sanskrit edition of the Bodhisattvapiṭaka will be published in Liland et al. (forthcoming).

n.7The relevant passages are cited in the body of the translation. For the list of citations of the sūtra in this text, see Lamotte 1970, p. 1638, n. 1. On the question of authorship and the nature of this important work, see Zacchetti 2021.

n.8For a recent approach to the study of Buddhist literature as a form of literature, see Shaw 2021. A classic in the field of literary studies of the Bible is Robert Alter’s The Art of Biblical Narrative (2011).

n.9See, for example, Anesaki 1911, and Radich 2015, especially p. 105 ff.

n.10See Lamotte 1966 and Zin 2009.

n.11For an English translation of the former, see Walshe 1995, p. 114; for the latter, see Bodhi and Ñāṇamoli 1995, p. 326.

n.12This translation is based on the Sanskrit, for which see Vaidya 1987, p. 50 (or p. 54 of the second edition); for an alternate English translation based on the Tibetan translation, see The Play in Full, Toh 95, 6.­47, (Dharmachakra Translation Committee 2013).

n.13See Hopkins 1915, pp. 10, 31, 61, and 142–48.

n.14See 18.­22

n.15For a study and translation of this passage, see Bodhi 1978.

n.16This translation is based on the Sanskrit, for an alternative translation of which, see Gómez and Harrison2022, p. 71; another alternative translation, based on the Tibetan translation, may be found in The Teaching of Vimalakīrti, Toh 176, 6.­3 (Thurman 2017).

n.17For a discussion of the different versions, see The Good Eon (Bhadrakalpika), Toh 94, i.­15–i.­18 (Dharmachakra Translation Committee, 2022).

n.18For the quotation in the latter, see The Long Explanation , Toh 3808, 1.­8 (Sparham 2022). It is also worth comparing this statement with what is found in The Teaching of Vimalakīrti, Toh 176, 1.­24-1.­27 (Thurman 2017).

n.19See n.­102.

n.20On this point, see the remarks by Cecil Bendall on Cambridge MS Adds. 901, 1365, and 1617 in his Catalogue of the Buddhist Sanskrit Manuscripts in the University Library, Cambridge (1883), pp. 15–17, 70–73, and 140–41.

n.21Winternitz 1933, pp. 394–95 and 635.

n.22Lewis 2000, pp. 15–16; Tuladhar-Douglas 2014, pp. 86 and 130 ff., especially pp. 132–33, and ch. 4, n. 46.

n.23The bibliographic information for these articles by Ikuma and for several more by other Japanese and Chinese scholars may be found in the bibliographic entry for this sūtra on the website of the Open Philology project.

n.24Étienne Lamotte, “Vajrapāṇi en Inde,” pp. 140–44, wherein Lamotte gives a three-page summary of the past-life story told in chapter 5 of the sūtra (according to the chapter divisions of the eleventh-century Chinese translation, which seems to follow those of the Tibetan translation), in which Brahmā and Vajrapāṇi each take a vow. The former vows to request each of the thousand buddhas of this fortunate era to teach the Dharma, and the latter vows to become the constant companion of each and every one of these buddhas. However, it seems as though Lamotte did not finish reading the sūtra, because after his synopsis of the story he claims that no mention is made in the sūtra of when or how Vajrapāṇi attains future awakening as a buddha, but this information is given in a later chapter of the sūtra (chapter 16 according to the Tibetan chapter divisions).

n.25Shingan 2021.

n.26Szántó 2021.

n.27Shāstri 1917, pp. 17–21.

n.28Denkarma, folio 295.b; see also Herrmann-Pfandt 2008, pp. 18–19. Phangthangma 2003, p. 6.

n.29For the description, see Marcel Lalou 1961, p. 200. The scan of this Dunhuang manuscript can be seen here. We have not yet identified the quotations of the sūtra in this manuscript or noted them in this translation.

n.30On this date and various scholarly opinions on the accuracy of its attribution to Dharmarakṣa, see the entry on Taishō 310 in the Chinese Buddhist Canonical Attributions database.

n.31For more on this version of the text, see Taishō 312 in the Chinese Buddhist Canonical Attributions database. See also the entry on K 1486 in Lewis R. Lancaster’s Descriptive Catalogue of the Korean Buddhist Canon.

n.32“Acceptance” (bzod pa, kṣānti) likely refers here to anutpattika­dharmakṣānti, “acceptance of the fact that things do not arise,” which is said to constitute a definitive understanding of the emptiness of all phenomena. Possibly the same expression is used to describe Vimalakīrti in the Vimalakīrti­nirdeśa, section 2.1, the Sanskrit for which reads prati­labdhakṣāntika. The Tibetan translation here and there is identical: bzod pa thob pa.

n.33The Tibetan translation reads mngon par shes pa’i ye shes gyis rnam par rtsen pa, which we understand to be translating a Sanskrit compound similar to abhijñā­jñāna­vikrīḍita.

n.34The Tibetan translation here is identical to what is also found in the Vimalakīrti­nirdeśa, section 1.3, for which the underlying Sanskrit is likely nihatamāra­pratyarthika. The Tibetan renders this compound as a dvandva, “x and y,” whereas one could interpret it differently to say that the adversaries actually are Māra in all his forms. The idea behind the translation of “Māra in all his forms” is that Buddhist literature generally recognizes several types of Māra. See, for instance, the entry on Māra in Buswell and Lopez, The Princeton Dictionary of Buddhism.

n.35This appears to be a minor translation variant upon a phrase also found in the Vimalakīrti­nirdeśa 2.1, where it describes Vimalakīrti: sarva­buddha­stutastomita­praśaṃsitaḥ.

n.36A similar statement describing the bodhisattva is found in The Play in Full (Toh 95), 2.­1.

n.37A similar expression is found among the opening descriptions of the bodhisattvas in the Vimalakīrti­nirdeśa, section 1.3: sarva­parṣadanabhi­bhūtavaiśaradya­vikrāmibhiḥ. Comparing the two passages, the Tibetan here suggests a reading of sarva­parṣadabhi­bhūta­vaiśaradyakulaśaḥ.

n.38zla ba dang nyi ma mog mog por byed pa; literally “they darkened the sun and the moon.” This expression is also used to describe an eclipse.

n.39“The three times” refers to the past, present, and future.

n.40Following Stok palace chos zab mo rnam par nges pa ye shes kyis ston pa la mkhas pa. Degé has a vertical line after rnam par nges pa, but this doesn’t require that we split the phrases into two.

n.41This sentence and the preceding one are quoted in the Sūtrasamuccaya. For an English translation, see Pāsādika 1978b, p. 28. The Tibetan expression found here, dkon mchog gsum gyi rigs rgyun mi ’chad par byed pa, is identical to one also found in the Vimalakīrti­nirdesa, section 1.3, for which the Sanskrit reads triratnavaṃ­śānupacchetṛbhiḥ. The same expression is attested later in the Sanskrit manuscript of the Tathāgataguhya, too.

n.42Following Stok Palace mtha’ yas rnam par gnon pa. Degé lacks this name but instead reads blo gros mtha’ yas, which duplicates and transposes the next name: mtha’ yas blo gros. Yet, blo gros mtha’ yas is the name of a bodhisattva found later in chapter 4, for which the extant Sanskrit equivalent is Anantabuddhi.

n.43This sentence is quoted by Kamalaśīla in the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, p. 12.

n.44This sentence is quoted by Kamalaśīla in the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, p. 12.

n.45The fragmentary Sanskrit manuscript kept in the library of the Asiatic Society in Bengal, MS G10765, preserves a block of text that begins with the second half of this sentence and runs to the beginning of folio 104.b of the Tibetan translation.

n.46The Sanskrit manuscript suggests the alternative that the supply here is of “serving” (sevana) companions in what is good.

n.47This sentence is cited by Kamalaśīla in the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, pp. 12–13.

n.48The Lankāvatāra Sūtra asks about the meaning of the fact that Vajrapāṇi always accompanies the Buddha, using different terminology from what is found here, but touching on a similar theme. See Nanjio, pp. 240 and 242, for the Sanskrit, and the English translation in Suzuki, pp. 207 and 209.

n.49The Sanskrit manuscript reads pratibhātu, perhaps an imperative form of pratibhā, “to shine light on,” but which is connected to the concept of pratibhāna, commonly translated as “eloquence” or “inspired speech.”

n.50The first block of text in the Sanskrit manuscript ends here, right in the middle of this sentence at the following point: “… Blessed One, through the majestic pow-.” The next line of the manuscript continues another fragmentary passage beginning slightly further in the text on folio 105.b of the Tibetan translation.

n.51Here the text seems to use the term “mystery” (bsam gyis mi khyab pa) or “inconceivable thing or quality,” acintya in Sanskrit, as almost a synonym for the term gsang ba (“secret”). This may help us to understand these terms better and even to see how they came to be combined so closely in the iteration of the title of this sūtra.

n.52In the Śikṣāsamuccaya, a quotation is given ostensibly from the Ratnamegha Sūtra that discusses the concept of kāyakuhanā or “artifice of the body,” as well as verbal and mental forms of artifice, such as flattery (lapanā) and dissimulation (Bendall 1902, pp. 267–68), which are described as actions of body, speech, and mind that are intended to persuade donors and patrons to give gifts.

n.53We see here a play on words between yongs su mi rtog pa (aparikalpa), rendered in the previous sentence as “involuntarily,” and rnam par mi rtog (avikalpa), rendered here as “[they] do not form concepts.” Both terms involve the lack of any supposition or application of false ideas or dualistic thinking onto the true nature of reality.

n.54The second text block of the Sanskrit manuscript picks up right at the end of this sentence with the words ojo dadāti, “he [the bodhisattva] gives his physical vitality,” which is strangely affixed to part of the line that construes with the previous text block. In any case, the Sanskrit fragment makes it clear that the bodhisattva is being spoken of here in the third-person singular masculine form, he/him, and the Tibetan translation reflects this, too, but for the sake of style and inclusivity we have rendered the sequence using the plural, they/them. This tension between the default masculine pronouns and the attempt at a more inclusive rendering persists throughout the sūtra.

n.55The mention made here of covering the cosmos with a jeweled parasol is quite similar to the opening marvel in the first chapter of the Vimalakīrti­nirdeśa (1.­13).

n.56The translation here reflects the extant Sanskrit manuscript insofar as the Tibetan text seems to have taken the offering of flowered parasols with the previous sentence, whereas the Sanskrit suggests that we read it together with the offering of the baskets of flowers, but it also partly reflects the Tibetan translation in the parsing of this and the next three sentences.

n.57This paragraph is one of several that is quoted in the Śikṣāsamuccaya. For the Sanskrit passage, see Bendall 1902, pp. 158–59.

n.58Beginning with this sentence and running until the end of this past life story and then a bit further, there is a parallel found in chapter 9 of the Bodhisattvapiṭaka (Toh 56). A comparison of the Sanskrit manuscript and the Tibetan translation of that text with the preserved fragments of the Sanskrit of this sūtra as well as its complete Tibetan translation reveals numerous differences in the precise language despite great similarities between the two texts. For instance, in the Bodhisattvapiṭaka, the Buddha tells this story to Śāriputra in first-person voice, while in this sūtra Vajrapāṇi tells the story to Śāntamati. For this reason, we have based the translation of this section primarily on the Tibetan translation of this sūtra and any available Sanskrit fragments.

n.59The terms “air” (vāta, lung), “heat” (pitta, mkhris pa), and “ phlegm ” (śleṣman, bad kan) refer to the three doṣas, the basic categories or properties of the human body and mind according to the traditional Indian medical system known as Ayurveda, which also influenced the development of traditional Tibetan medicine.

n.60The end of this sentence marks the end of the second text block of text in the Sanskrit manuscript.

n.61In the Bodhisattvapiṭaka, Śakra first speaks in prose for about a paragraph before continuing with a series of six verses, three of which are similar but not identical to what is found in this sūtra.

n.62Here Vajrapāṇi resumes his narration of the story, but the text suggests that he does so with a verse. Therefore, we have rendered it as such in the translation. In the Bodhisattvapiṭaka, too, the voice shifts in the final verse from Śakra to the narrator, which in that text is the Buddha himself.

n.63Here we have two verb forms, mi sems and mi rtog, which are likely translating the Sanskrit expressions na vicintayati and na vikalpayati, “he does not ponder” and “he does not ruminate,” respectively.

n.64From this sentence through the penultimate sentence of the following paragraph, the Śikṣāsamuccaya also preserves the Sanskrit, for which see Bendall 1902, p. 159.

n.65The Vimalakīrti­nirdeśa may owe something here again to the sūtra, or vice versa, since the language used to describe this medicine-made girl is the same language the goddess uses therein to describe the flowers to Śāriputra. In section 6.14 of the Vimalakīrti­nirdeśa, the goddess says that the flowers do not form ideas (na kalpayanti) and they do not form conceptions (na vikalpayanti). For an alternative translation of the passage made from Tibetan see The Teaching of Vimalakīrti, Toh 176, 6.­14 (Thurman 2017). Though the Sanskrit preserved in the Śikṣāsamuccaya appears corrupted here, the Tibetan translation of these two sūtra passages is nearly identical, and again, we can take note of the use of the terminology related to the term vikalpa (rnam par rtog pa), and also its connection here (and there) to the forming or fashioning (kalpana) of bodies. Winternitz 1933, p. 394, n. 5 notes that this image appears to be a counterpart to the narrative trope of the “poison girl” (viṣakanyā). This trope is mentioned in the Arthaśāstra and in later narrative literature such as the Kathāsaritsāgara, as well as in modern Indian literature and film.

n.66With the first half of this sentence, the series of close parallel passages found in chapter 9 of the Bodhisattvapiṭaka (Toh 56) appears to come to an end.

n.67This same list of metaphors for the body is found in the Vimalakīrti­nirdeśa, chapter 2, for which the Sanskrit reads tṛṇakāṣṭhakuḍyaloṣṭa­prati­bhāsasadṛśah. The Tibetan translation of the two passages is identical except for the fourth member of the list, for which the Tibetan translation of the Vimalakīrti­nirdeśa gives bong ba and here the translation of the Tathāgataguhya has lam, or “road,” which seems as though it may be an error or corruption. The Vimalakīrti­nirdeśa explains that all these things are examples of “inanimate” (jaḍa) objects. In general, it is worth comparing the descriptions of the Dharma body here in this sūtra with the descriptions of the physical body and of the Dharma body at the end of chapter 2 of the Vimalakīrti­nirdeśa, sections 2.9–12. For a translation from Tibetan, see The Teaching of Vimalakīrti, Toh 176, 2.­8-2.­12 (Thurman 2017).

n.68The Śikṣāsamuccaya quotes the sūtra from the beginning of this sentence up to and including this word, and then it selectively includes several more terms from this long list of attributes of the realized one’s speech. For the Sanskrit, see Bendall 1902, p. 126.

n.69On the suspicion (perhaps wrong) that the Tibetan here, lung bstan pa la thogs pa, may be rendering the phrase kṣunavyākaraṇā, found in the passage quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 126. See also Edgerton’s entries on kṣuṇa and kṣūna.

n.70The previous four phrases are quoted in a passage in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 126.

n.71Reading brlang following the Narthang and Zhol editions of Kangyur. Pedurma p. 320, n. 20.

n.72This phrase, kha ’og tu lus ’khrus par mi byed pa’i tshigs, appears to correspond to the phrase prati­jñottāraṇ­avacanā in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 126. See also Edgerton’s entry on uttāraṇa, which discusses this phrase.

n.73This phrase, mngon pa’i nga rgyal gyi brda sprod pa’i tshig, appears to correspond to the phrase ābhi­mānikavyākaraṇ­avacanā in the quotation in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 126.

n.74The last several words in this list are hard to discern as specific numbers and have therefore been given as a list increasing by powers of three.

n.75Here the sūtra is playing on two of the primary uses of the word dharma. It refers in a general way to “things” or “phenomena” and their component factors, but is also often used to mean “quality” or “attribute,” as it does here and elsewhere in the phrase buddhadharmāḥ, “the qualities of a buddha.”

n.76Though the Sanskrit is not extant here, Ikuma 2013 argues that the first chapter, according to the divisions of the Sanskrit manuscript, must have concluded here.

n.77The sign of auspiciousness meant here is the svastika .

n.78The voice shifts back to the Buddha of the present narrating the story of the past, but he seems to continue in verse so we have likewise rendered it in verse.

n.79This sentence and the preceding question are quoted in the Sūtrasamuccaya where they are connected to the passage noted in n.­249. See Pāsādika 1978a, part 1, p. 26.

n.80This whole paragraph is quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 316.

n.81Here again the reader can note the different sense of the word dharma, which is sometimes translated both here and above as “quality,” and left untranslated here in the expression “king of the Dharma.” One might consider translating the latter as “righteous king,” but that would likely not exhaust the range of meanings of the word dharma here.

n.82Another fragmentary passage from the Sanskrit manuscript begins in the middle of this sentence and continues nearly to the end of this chapter.

n.83The Tibetan translation here includes a sentence with a pair of names not found in the Sanskrit manuscript. The striking similarity of these names with the previous pair suggests that the Tibetan may have been an alternate translation of the previous name incorporated at some point into the text. Since this is not clear, however, we have included this pair in our translation, too, while giving the prince’s name, dri ma med pa’i ’od, as Vimalaprabhāsa, which is attested as an equivalent in the Lalitavistara, to signal the similarity with the previous pair. This name could also be Vimalaprabhā. The name of the tathāgata, spyan legs pa, is not attested, so far as we know, as a translation of the name Sunetra , but the meaning is the same.

n.84The Sanskrit manuscript reads gaganadhātusamā here, which suggests that we interpret nam mkha’ in the Tibetan translation as nam mkha’i khams. Also, the Sanskrit has jina (“victor”), whereas the Tibetan has sangs rgyas (“buddha”). The translation follows the Sanskrit.

n.85This sentence explains the name of the tathāgata, which is attested as Roca in the Sanskrit manuscript. Negi’s entry on the Tibetan term mos pa, used to translate the name here, makes it clear that the term translates both adhimukti (“strong inclination”) and ruci (“light”), but also “zest” or “longing.”

n.86The third portion of this sūtra preserved in Sanskrit ends in the middle of this sentence.

n.87Though the Sanskrit here is not extant, Ikuma 2013 argues that the second chapter, according to the divisions of the Sanskrit manuscript, would have ended here.

n.88The translation of this sentence reflects an emendation of the Tibetan text, which reads de bzhin (“thus”) whereas every other sentence has de bzhin gshegs pa (“the realized one(s)”) in the same position in the sentence. It has been translated as if it read the latter.

n.89The fourth segment of the text preserved in the Sanskrit manuscript begins in the middle of this sentence and continues for about ten pages of the Tibetan translation.

n.90The *Mahā­prajñā­pāramitopadeśa quotes this passage among several in a compilation of quotes and paraphrases from this sūtra in describing the secret of the Buddha’s body, speech, and mind. See Lamotte 1981, p. 560.

n.91The Sanskrit here reads dhṛtarāṣtrahaṃsa­rāja, perhaps referring to the story in the Mahābhārata explaining King Dhṛtarāṣṭra’s blindness. The Tibetan simply reads ngang pa’r rgyal po ngang skya (“light orangish white like the king of swans”).

n.92The translation here reflects some of the particular syntax of the Tibetan translation. The Sanskrit suggests that we could alternatively translate this passage as follows: “Moreover, in this respect, the realized one would not be produced by a body or a mind” (na ca tatra tathā[ga]taḥ kāyena vā cittena vā vyāvṛto bhaved). Both translations would seem to be resolvable around the same basic idea, however.

n.93The number here follows the Sanskrit manuscript. The Tibetan translation says here that he circled him three times, but slightly down further the Tibetan agrees with the Sanskrit that he circles the Buddha Padmaśrīrājagarbha seven times.

n.94Following Degé. The Sanskrit manuscript says he only walked around him once.

n.95The translation here reflects the Tibetan translation, which also suggests a possible emendation of the third line of the verse in the Sanskrit manuscript from buddhānantaṃ na paśyāmi to mūrddhānante na paśyāmi or something similar.

n.96The equivalent verse in the Sanskrit manuscript has some variances with the Tibetan here. They both have “moral conduct” (śīla, tshul khrims), “concentration” (samādhi, ting nge ’dzin), and “liberated knowledge and vision” (vimuktir jñānadarśana, rnam grol ye shes mthong), but the Sanskrit lacks “liberation” (vimokṣa, rnam grol) and includes “wisdom” (prajñā, shes rab) and “knowledge” (jñāna, ye shes). It is not entirely clear whether the Tibetan translators were working from a different iteration of the verse here or if perhaps the translation reflects the challenge of squeezing the Sanskrit into seven syllable lines of Tibetan.

n.97This verse and the next one are quoted in the Sūtrasamuccaya, where they are connected to the passage noted in n.­40. See Pāsādika 1978b, part 3, p. 28.

n.98This verse and the next one are quoted in the Sūtrasamuccaya in a different section from the one in which the preceding verses are quoted. See Pāsādika 1981, p. 27.

n.99Here, the state of the buddhas is the “inconceivable (or mysterious) state” (acintyaṃ sthānam [sic]) and “the place of the secrets” or “the secret place” or “hidden place” (guhyasthānaṃ [sic]).

n.100The beginning of this paragraph is cited explicitly in the *Mahā­prajñā­pāramitopadeśa. On this passage therein, see Lamotte 1970, p. 1681.

n.101The Tibetan translation includes the phrase “and the actions of the body” (sku’i phrin las te). It is not in the Sanskrit manuscript and could represent an interpolation. There seems to be a slight discrepancy or corruption here in the Sanskrit manuscript, which does not include this passage, but begins the next sentence with the phrase “the purity of actions that is the secret of the body…” (kāyaguhya­karma­pari­śuddhir).

n.102Following the Tibetan translation. The Sanskrit manuscript reads kāyaguhya­karma­pari­śuddhir [sic] here, which could be translated “the purity of actions that is the secret of the body,” but it then reads tathāgata­kāya­karma­pari­śuddheḥ [sic] in the next sentence, which suggests the possibility of a corruption here.

n.103Following the Sanskrit phrase here, tathāgata­kāyaguhyācintya­nirdeśe nirdiśyamāne [sic]. The Tibetan has de bzhin gshegs pa’i gsang ba bsam gyis mi khyab pa bstan pa ’di bshad pa na, “when this teaching on the mystery and secret of the realized ones was being taught.” The phrasing here is identical to the order of the words in the Tibetan translation of the full title of the sūtra, and the word order is supported by the Sanskrit manuscript.

n.104The Sanskrit manuscript differs from the Tibetan translation here in that the former says that eight thousand bodhisattvas attained acceptance.

n.105The Sanskrit manuscript also preserves a chapter ending here, demonstrating that it numbers the chapters differently from the Tibetan translation. The manuscript says here, “The chapter of the secret of the realized one’s body, being the third in the Great Extended Discourse (mahāvaipulye).”

n.106This sentence is quoted, but without explicit reference to the Tathāgataguhya, and placed in the voice of the Buddha in a passage in the Lankāvatāra Sūtra, for which see Nanjio, pp. 142–43 and 240, for the Sanskrit, and the English translation by Suzuki, pp. 123–24 and 207. This passage is also quoted twice in Candrakīrti’s Prasannapadā, where it is explicitly said to come from the Tathāgataguhya Sūtra. The first citation is in the commentary on chapter 18, verse 7 of the Mūla­madhyamakakārikā; the second citation is in the commentary on chapter 25, verse 24, the last verse in the chapter on the analysis of nirvāṇa. For the Sanskrit text, see La Vallée Poussin 1903, pp. 366–67 and 539–40. There are slight differences between the sentence as it is quoted in the Prasannapadā and the passage as found in the Sanskrit manuscript and the Tibetan translation of this sūtra, but after quoting this sentence, Candrakīrti appears to continue to cite or paraphrase this sūtra. In neither case, however, does there seem to be an exact equivalency to what is found in the extant Sanskrit manuscript or the Tibetan translation. In this respect, it would be interesting to compare the Prasannapadā with the earlier Chinese translation of the sūtra. After the first quotation of the sentence above, Candrakīrti says, “How, then, is the teaching of the Dharma taught by the Blessed One to all the various kinds of folk who are ready to the trained, beginning with gods (sura), asuras, humans (nara), kinnaras, siddhas, vidyādharas, and uragas (i.e., nāgas)? With the mere utterance of speech for a single moment, [a speech that is] a great light, reddish like the sun in autumn, a light that removes the darkness from the minds of those beings (tatajjanamanastamoharaṇī), opens the manifold thickets of the lotuses of their intellects, dries up the oceans and rivers of old age and death, and surpasses the magnitude of the light rays from the seven suns of the fiery conflagration [at the end] of the eon.” Then follows the quotation of the three metaphors in verse from this sūtra for which the references are given when they occur. What follows the second quotation by Candrakīrti of the sentence above is the following: “Moreover, all beings, whose aspirations and constituent element s vary, understand the speech that comes from the Realized One variously in accordance with their level of dedication. And each and every one of them has the thought, ‘The Blessed One is teaching this Dharma to us; we are hearing the Realized One’s teaching of the Dharma.’ In this regard, the Realized One does not form an idea; he does not form a concept. For, indeed, Śāntamati, the Realized One is devoid of all vain imaginings or lingering traces of the web of thoughts and concepts.” This second passage seems quite similar to parts of the passage that follows this one in the main text and part of a passage found a few paragraphs further below in the sūtra. After what was just translated, the second quotation in the Prasannapadā continues with a verse and several more lines that pursue a similar theme, at the end of which Candrakīrti says, “All of this has been explained at length in ‘The Chapter on the Secret of the Speech of the Realized One’ (tathāgata­vāgguhya­parivarta).” The main statement above is also quoted at the beginning of The Long Explanation of the Noble Perfection of Wisdom in One Hundred Thousand, Twenty-Five Thousand, and Eighteen Thousand Lines‍—see section 1.­81.8‍—where again the quotation appears to differ slightly from the passage as found in the extant Sanskrit manuscript of this sūtra.

n.107The Mahāyāna­sūtrālaṃkāra­bhāṣya makes reference to a list of sixty such qualities, which it attributes to The Teaching of the Lord of Guhyakas (Guhyakādhi­pati­nirdeśa). The Mahāyāna­sūtrālaṃkāra­bhāṣya then defines or explains many of the terms. See Lévi, pp. 79–80, for the Sanskrit, and Jamspal et al., pp. 156–58, for an English translation of the passage. The different iterations of this list contain a few variations.

n.108The Tibetan reads snyan cing ’jebs pa, for which the Mahāyāna­sūtrālaṃkāra­bhāṣya has valgu. The Sanskrit manuscript reading of varṇa, perhaps meaning “praiseworthy,” would seem to be a corruption of valgu.

n.109The Sanskrit reads luḍitā, which should likely be corrected to laḑitā. The Mahāyāna­sūtrālaṃkāra­bhāṣya reads lalitā. The Tibetan here reads ’brel pa, which seems to repeat the earlier term sahitā (“relevant”).

n.110This sentence contains what amounts to an expanded list of classical genres or categories of Buddhist literature, including the traditional lists of nine and twelve “divisions” (aṅga) of the Dharma.

n.111Following the Sanskrit manuscript. The Tibetan translation reads “one hundred thousand eons.”

n.112Following the Sanskrit manuscript here, which seems to read parivijñaptitā and should be emended to parivijñaptito. Vijñapti means “information” or an “announcement,” “request,” or “report.” Here, the English word “report,” used in the sense of the report a firearm makes when it is fired, gives the appropriate sense. The prefix pari- may suggest the idea that the echo is perceived as producing a “successive” sound, or perhaps that the sound is perceived as coming from “all around,” or perhaps it simply intensifies it‍—a “loud” report. In any case, the Tibetan translation here, gzhan rnam par rig par byed par yang, suggests an alternate reading of paravijñaptitaḥ, which then suggests another possible interpretation that the sound is perceived as coming from “another” (para, gzhan) direction or from “something (or somewhere) else” (para, gzhan), as it sometimes does when one hears a loud noise. In both cases, the analogy appears intended to show that the agency or source of the sound is ambiguous.

n.113The clever play continues here on the different senses of the word vijñapti or rnam par rig par byed pa, a “report” or “something informative.” The ambiguity of the compound in Sanskrit and Tibetan is such that the Buddha’s speech here can be said to inform the motivations of all beings, or it is informative of the motivations of all beings, or both at the same time.

n.114The following story about the extent of the Buddha’s voice can be fruitfully compared with the story of Abhibhū in the Pali Saṃyutta Nikāya, for a translation of which see Bodhi 2000, p. 250, and with the Buddha’s remarks on this story in the Pali Aṅguttara Nikāya, for a translation of which see Bodhi 2012, pp. 313–14.

n.115This sentence and part of the previous few sentences are cited or paraphrased in the *Mahā­prajñā­pāramitopadeśa. On this passage therein, see Lamotte 1976, p. 1985.

n.116In the middle of this sentence the fourth segment of text preserved in the Sanskrit manuscript comes to an end. The fifth segment of Sanskrit text, another long segment, resumes after about ten Tibetan pages.

n.117This part of the story is cited in a longer passage from the *Mahā­prajñā­pāramitopadeśa that mentions this story in the context of discussing the secrets of the Buddha’s body, speech, and mind in connection with this sūtra. See Lamotte 1981, pp. 560–61.

n.118This verse and the next two verses are quoted by Candrakīrti in the Prasannapadā. Here is the Sanskrit, as quoted in that text: yathāyantrakṛtaṃ tūryaṃ vādyate pavaneritam | na cātra vādakaḥ kaścinniścaranty atha ca svarāḥ. For the Sanskrit text, see La Vallée Poussin 1903, p. 366.

n.119Here is the Sanskrit of this verse, as quoted in Candrakīrti’s Prasannapadā: evaṃ pūrvasuśuddhatvāt sarvasattvāśayeritā | vāgniścarati buddhasya na cāsyāstīha kalpanā. For the Sanskrit text, see La Vallée Poussin 1903, p. 367.

n.120Here is the Sanskrit of this verse, as quoted in Candrakīrti’s Prasannapadā: pratiśrutkādayaḥ śabdā nādhyātmaṃ na bahiḥ sthitāḥ | vāgapyevaṃ narendrasya nādhyātmaṃ na bahịh sthitāḥ. For the Sanskrit text, see La Vallée Poussin 1903, p. 367.

n.121In the following list, the attempt has generally been made to follow the conventions of the Tibetan translation and repeat the common expression of the four noble truths in each sentence, and also to translate the equivalent expressions of the truths in the languages of the divine realms into English when the Tibetan also attempts to do so. Nevertheless, the meaning of many of the Tibetan translations is unclear, and thus the English translations should be viewed as somewhat tentative.

n.122Here, as in a few other instances in this list, the Tibetan translation simply transliterates what must have been found in the Sanskrit manuscript. This transliteration has been rendered here without an attempt to identify the possible sense, if any. One suspects that a word like ohita may be a non-Sanskritic form of the Sanskrit word avahita, which can mean “plunged into,” but this is speculative.

n.123Here, the transliteration of the Sanskrit manuscript has been reproduced as it is found in the Tibetan translation. The meaning, if any, of the underlying manuscript is unclear, but could perhaps be rendered as follows: “ ‘mine, mine’; in regard to ‘mine, mine’; the discipline of conceit (mānānunaya); and pursuing conceit (mānānugama).”

n.124The Sanskrit manuscript resumes in the middle of this sentence. The final term in the second list of terms reads muktavatānu­sandhi in the Sanskrit, which the Tibetan translation translates as grol ba’i bye brag phyed par mtshams sbyor ba, perhaps meaning “conducive to thoughts about freedom.” The presence of bye brag phyed par in the Tibetan translation suggests the Tibetans may have read muktamata, “thoughts about freedom,” instead of muktavatā, “a place characterized by freedom.” Also, it is interesting to note that the Tibetan here translates the term anusandhi, “conforming or conducive to,” with the term mtshams sbyor, whereas earlier in the same list the Tibetan transliterates the term in its rendering of haranusandhi, translated above as “what sounds like ‘hara!’ ” In both cases, we have interpreted the term anusandhi to mean that Vajrapāṇi is saying that there is an approximate correspondence between the actual expression in the language of the gods and the way he has rendered it, but this is a somewhat speculative interpretation. Perhaps in both cases the idea behind the term anusandhi is closer to the Tibetan translation of the second instance, “what conforms or is conducive to.”

n.125From this point through the remainder of Vajrapāṇi’s speech in this chapter, the Sanskrit manuscript reads kulaputra (“noble son”), whereas in the Tibetan translation Vajrapāṇi addresses him as zhi ba’i blo gros = Śāntamati. This seems to be a pervasive difference between the Sanskrit manuscript and the Tibetan translation.

n.126Ikuma 2020 compares the list of ethnic names with a similar list of names found in the Chinese translation of the *Abhidharma-mahāvibhāṣā.

n.127Following the Sanskrit manuscript. The Tibetan transliterates the term pilin or possibly pilina. The name and its referent are unclear. One is tempted to correct to Pulinda, but that term is already also present later in the list. Pilindagāma is given in the Pali Vinaya as the nickname of a village where the monk Pilindavaccha begged for alms. See Malalasekera, Dictionary of Pāli Proper Names, under ārāmikagāmā.

n.128Following the Sanskrit manuscript. The Tibetan transliterates the term soma. The name and its referent are unclear.

n.129Following the Sanskrit manuscript. The Tibetan transliterates the term dacipa. The name and its referent are unclear.

n.130Following the Sanskrit manuscript. The Tibetan transliterates the term pun dra, which looks like a legitimate variant reading.

n.131This is a conjectural emendation. Ikuma 2020 suggests emending to puruṣa. The Sanskrit manuscript reads puruca and the Tibetan gives a transliteration, phururava.

n.132Following the Tibetan transliteration here on the assumption that it is rendering a term in the manuscript on which the Tibetan translation is based. A corresponding term does not seem to be attested in the extant Sanskrit manuscript. The name and its referent are unclear.

n.133Following the Sanskrit manuscript. The Tibetan here gives sbal ka, which looks like a corruption of sā la ka.

n.134This name and its referent are unclear. The Sanskrit manuscript reads kṛviṣamukha. The Tibetan gives the transliteration/translation kri ba sha’i gdong da. Some kind of emendation seems necessary.

n.135This name and its referent are unclear. The Sanskrit manuscript reads paścānudhāyin, but the Tibetan translation of rjes su rgyug suggests a possible emendation to paścānudhāvin, which is what has been used as the basis for this translation.

n.136Many of these last several names and their referents are unclear, and it is also unclear whether some of the earlier names, such as “Downward-Facing People,” refer to actual or imaginary groups, or if the list descends intentionally into ambiguous characterizations. The conclusion of the list is also unclear, and the Tibetan translation suggests a number of possible emendations to the Sanskrit text. For instance, where the manuscript reads lokottarikāṇāṃ nāsikānām evaṃpramukhānāṃ…, the Tibetan translation of ’jig rten gyi bar na gnas pa’i yul la sogs pa… suggests lokāntarikāṇām. More work is needed to determine whether the Sanskrit word nāsikānām refers to a group of people or requires some kind of emendation. It does not seem to be represented in the Tibetan translation.

n.137From this sentence to the end of this paragraph is cited at the very beginning of The Long Explanation of the Noble Perfection of Wisdom in One Hundred Thousand, Twenty-Five Thousand, and Eighteen Thousand Lines (Toh 3808), at section 1.­5.

n.138Following the Tibetan translation, which reads ’di ni gsung rmad du byung ba’o, for which the underlying Sanskrit ostensibly would be vāgadbhutam iyam. The extant Sanskrit manuscript actually reads vāgbhūtatā iyaṃ, which could be translated, “This is the true nature of a realized one’s speech.”

n.139Following the Sanskrit manuscript, which reads tathāgatasyācintyatā. The Tibetan translation has “the inconceivable nature of a realized one’s speech.”

n.140Following the Sanskrit manuscript. The Tibetan translation could be interpreted to say “the true nature of a realized one and the qualities of a realized one,” which seems like an attempt to interpret what is a difficult expression in the Sanskrit.

n.141Following the Tibetan translation. This sentence provides another good example of the kind of minor differences one sometimes finds between the extant Sanskrit manuscript and the Tibetan translation. Here, the Sanskrit manuscript reads only “incalculable worlds,” rather than the fuller expression found in the Tibetan translation, but notice how the slightly more restrained expression here in Sanskrit conforms a bit better with the next several sentences in both the Tibetan and the Sanskrit.

n.142Following the Tibetan translation. In keeping with its numbering and divisions, the Sanskrit manuscript says here “the fourth, namely the chapter on the secret of speech” (vāgguhyaparivarto nāmaś caturthaḥ).

n.143Following the Sanskrit manuscript, which reads cittavismṛti. This phrase seems to be absent from the Tibetan translation, as the next Tibetan phrase reads thugs la ’gyur ba mi mnga’, which looks like a translation of the Sanskrit phrase cittapariṇāma or “alteration of mind,” which is absent in the Sanskrit manuscript.

n.144This phrase is added from the Tibetan translation. It seems to be absent from the Sanskrit manuscript, as stated in the previous note.

n.145Following the Sanskrit manuscript, which reads citta­saṃharṣanaṃ. This phrase seems to be absent from the Tibetan translation, as the next phrase reads thugs la ’grug pa or thugs la ’grul pa, “mental conflict” or “mental confusion.”

n.146This phrase is added from the Tibetan translation. It seems to be absent from the Sanskrit manuscript, as stated in the previous note.

n.147This phrase, thugs la sel ba mi mnga’, is added from the Tibetan translation. It seems to be absent from the Sanskrit manuscript.

n.148Following the Tibetan rnam par ’khrug pa, which suggests emending the Sanskrit manuscript to vikopana from its present reading, vilokana.

n.149This phrase follows the Tibetan. It is lacking in the Sanskrit manuscript.

n.150Following the Tibetan, which reads gzigs pas kun gzigs pa dang rnam par gzigs pa mi gzigs so. The Sanskrit manuscript, which reads only paśyanayā ca na paśyanāṃ paśyati, suggests that we understand kun gzigs pa (saṃpaśyanā?) and rnam par gzigs pa (vipaśyanā?) as the direct objects of the verb “to see.” On the face of it, the Sanskrit seems to say “he does not see seeing by means of seeing,” but perhaps the Sanskrit would tolerate some kind of emendation.

n.151This would seem to be one of the passages quoted explicitly from this sūtra in the Mahā­prajñā­pāramitopadeśa. For this passage, see Lamotte 1970, p. 1638.

n.152Following the Tibetan. The Sanskrit manuscript here has only “effortlessly” (anābhogena), whereas the Tibetan again lists all three adverbs as it has repeatedly above.

n.153Following the Tibetan. The Sanskrit manuscript, once again, has a slightly less elaborate sentence that nonetheless scans. It can be translated as follows: “In this way, noble son, the realized one is one whose knowledge has no foundation; his unattached knowledge and vision of the thoughts, actions, and motivations of all beings arises without impediment, and does so effortlessly and spontaneously.”

n.154The Sanskrit manuscript reads sarva­satva­saṃjñaptiṃ ca karoti. The Tibetan renders this passage as sems can thams cad la sems rnam par rig par yang mdzad pa, apparently reading vijñapti rather than saṃjñapti. The term saṃjñapti can mean “information” (like vijñapti sometimes does) and also “appeasement” or “mollification,” but it may sometimes have the sense of a request or demand, as vijñapti can, too. The expression could thus be rendered a bit more literally, perhaps, as “he does what is requested by all beings.”

n.155Following the chapter breaks and enumeration of the Tibetan translation. The Sanskrit manuscript does not indicate any chapter break here at all.

n.156Following the Tibetan bdag de bzhin gshegs pa la yid ches par mchi’o. Śāriputra’s whole response here is absent in the Sanskrit manuscript.

n.157Following the Tibetan. Śāriputra’s claim that he had not previously seen Vajrapāṇi is absent in the Sanskrit manuscript.

n.158Following the Sanskrit manuscript, which reads acintya­jñānābhi­jñādhiṣṭhānabalena ­samanvāgato. The Tibetan translation does not include a term for acintya (“inconceivable”), and its interpretation of the compound may be translated as follows: “endowed with the power of the empowering authority of supernormal faculties and knowledge.”

n.159Following the Tibetan translation. The Sanskrit manuscript also makes a chapter break here, but it reads as follows: “the fifth chapter, the secret of the mind” (citta­guhyaparivartta pañcamaḥ).

n.160This chapter develops a theme also explored in other Buddhist literature on the life of the Buddha. Compare, for instance, with the Mahāvastu, translated by Jones, vol. 2, pp. 114–29 and 182–99. Another, somewhat closer parallel to the portrayal here is found in the Lalitavistara, chapter 17, on the austerities of the Bodhisattva. In fact, in this section of the sūtra on the Buddha’s life, there are several similarities and even shared passages between this sūtra and the Lalitavistara, which have been studied in Ikuma 2018.

n.161Following the order of the Tibetan translation. The Sanskrit manuscript has an equivalent sentence, sūryam anu­pari­vartamāna (Tib. nyi ma yongs su sgyur ba), but includes it after the next group of types of clothing.

n.162Following the Sanskrit, which reads phalacīvaraṃ. The Tibetan translation gives shing leb gyon pa (“clothes made of wooden boards”), which suggests that the Tibetan translators understood phala to be synonymous with phalaka, which is found just above in the list of types of beds. Phala can be synonymous with phalaka, but it is also a general term for fruit or nut. This would give the translation “clothes made from fruits (or nuts),” but according to Meulenbeld’s dictionary of Sanskrit, plants named phala can be used more specifically for certain types of fruiting trees, such as the Indian nut palm, Areca catechu.

n.163Following the Sanskrit manuscript cīvaravasanaṃ. The Tibetan says rtsal tsi ra gyon pa (“wearing clothes made of cira grass”), but this seems to be an error, possibly resulting from a scribal error in the manuscript used for the Tibetan translation. For one thing, cira does not seem to be the name for a type of grass.

n.164Following the Tibetan translation shing shun gyi gos kyon pa. The Sanskrit manuscript appears to be lacking a corresponding expression.

n.165Following the Sanskrit manuscript and reading śyāmākabhakṣaṃ. The Tibetan translation is sre da za ba.

n.166The Sanskrit here reads kodravāhāraṃ, while the Tibetan reads nas shin te za ba, which may suggest a translation of “poor people’s grain.” Kodrava or Kodo millet is sometimes called “ditch millet,” and seems to have been associated with people who live in poverty and cannot afford more expensive grains. See the entry on kodrava in Monier Williams. Now, the next term found in the Tibetan translation is tsi tse, which is given by Lokesh Candra as a translation for kodrava, too, but the Sanskrit manuscript appears only to have a single term here. So, one wonders whether nas shin te refers to something else or if the Tibetan translation preserves alternate translations of a single Sanskrit term. We have chosen the latter option and not translated tsi tse separately.

n.167The Sanskrit term is lājāhāra for which the corresponding Tibetan translation seems to be bra yos za ba.

n.168Following the Sanskrit. The Tibetan translation suggests an alternative of “the most distinguished vows of the noble ones.”

n.169The name for this meditation is āsphānaka in Sanskrit, mkha’ khyab in Tibetan, and appānaka or appāṇaka in Pali. In the Mahāsaccaka Sūtta of the Majjhima Nikāya (Bodhi and Ñāṇamoli, p. 337), it is described as a type of meditation in which one does not breathe. It is also mentioned in the Mahāvastu (Jones, vol. 2, p. 124), as well as in The Play in Full (Toh 95), chapter 17, sections 17.­22–17.­26, where a different explanation of this form of meditation is given. It is on the basis of this alternative explanation that it has been translated here as all-pervading meditation .

n.170Following the Sanskrit, which reads gośatakṣīra. The Tibetan translation says that the milk came from a thousand cows.

n.171Following the Sanskrit, first, in that the Tibetan translation includes not only gods, nāgas, yakṣas, and gandharvas, but also asuras, garuḍas, kinnaras, and mahoragas in this list of types of beings. Yet just below, this passage in the Tibetan translation accords with the extant Sanskrit manuscript in listing only the first four types of beings. However, in another respect we also follow the Tibetan translation here. It duplicates the beings’ request, seemingly for emphasis, whereas the Sanskrit manuscript only has a single request.

n.172Following the Tibetan. The Sanskrit manuscript does not make a chapter break here.

n.173Following the Tibetan translation, which includes this verse that is absent from the Sanskrit manuscript. It fits well with the order of progression up the various heavenly realms. Yāma’s Heaven is said to be above the Heaven of the Thirty-Three and below the Heaven of the Contented.

n.174This paragraph and the following speech have a close, nearly identical parallel in chapter 19 of The Play in Full (19.­8–19.­11). Other versions of this episode with the nāga king are found in the Mahāvastu, where he is called Kāla. See Jones, vol. 2, pp. 249 ff., 284 ff., and 354 ff.

n.175Following the Sanskrit, which reads vimuktipaṭṭabaddho, “he is bound with the paṭṭa of liberation.” Among its several meanings, paṭṭa can refer to a turban or diadem worn by a king or other royal person as a symbol of authority. The Tibetan translation has something rather different here: rnam par grol ba’i thabs brnyes pa, “he has acquired skill in the method of liberation.”

n.176Following the interpretation of the Tibetan translation. The Sanskrit could be interpreted to say a bit more straightforwardly that he uses the Dharma to defeat the proponents of other doctrines.

n.177Following the Tibetan. This paragraph is absent from the Sanskrit manuscript at this point, but another paragraph is found in both the Sanskrit manuscript and in the Tibetan translation after the next set of verses below, which contains some of the same language.

n.178Following the Tibetan, which includes the word “then” and supplies the speaker, “Great Brahmā, sovereign of the Sahā world.” Otherwise, the Tibetan mirrors the Sanskrit manuscript, which transitions into Brahmā’s verses and leaves the speaker implied.

n.179“Paternal grandfather” (pitāmaha) is sometimes used as another name for Brahmā. The singular construction here, alongside the plural verbs and pronouns, suggests the dialogue of answer and response. The Tibetan translation of this verse renders it with a somewhat simpler grammatical construction than the Sanskrit does, but seems to get at the same basic meaning.

n.180Though they are not identical, the following section with Kālika bears some strong similarities to a parallel episode found in chapter 19 of The Play in Full, from 19.­24. For the Sanskrit, see Vaidya, p. 202 ff.

n.181Following the Sanskrit manuscript here, which reads mahāvegavat. The term simultaneously conveys both swiftness, like a swiftly flowing river, and a sense of great urgency, alacrity, feeling impelled to action. The epithet would seem to allude to the story about the Buddha Śākyamuni that he took his vow as a bodhisattva later than Maitreya did, but progressed along the bodhisattva path at a more rapid pace, as one finds in The Questions of Maitreya, Toh 85 from 1.­51, and elsewhere. The Tibetan translation here, gzi brjid che ldan pa, could be rendered “one who possesses great vigor (or energy).” This might suggest that the manuscript used for the Tibetan translation had a different term here, perhaps mahātejovat, but this would disrupt the meter of the verse and vega and tejas can both convey a sense of vigor.

n.182With certain slight variations of terminology and syntax, this paragraph and the five verses that follow are found in The Play in Full, 19.­61–19.­66. For the Sanskrit, which is almost verbatim, see Vaidya, p. 206 ff.

n.183The following episode with the grass seller Svastika is also found in The Play in Full, from 19.­68. The episode has a longer history, too. See, for example, Mahāvastu, vol. 2, pp. 126, 249, and 355. There are also references to Sotthiya the grass seller in Pali Buddhist literature, such as the Buddhavaṃsa.

n.184Following the Tibetan, which reads sems can rnams kyi mos pa ji lta ba. Mos pa implies an underlying Sanskrit adhimukti (“motivation”), while the Sanskrit manuscript we have reads yathā vimuktānāṃ satvānāṃ, perhaps “according to the beings’ status of liberation (vimukta).” It could be that the phrases here refer to the spiritual faculties and qualities of liberation of the bodhisattvas themselves and not to those of other beings.

n.185Following the Tibetan. The Sanskrit manuscript does not make a chapter break here.

n.186Compare the following with the quite similar (though not identical) passage in The Play in Full, 21.­192–21.­200.

n.187Following the Sanskrit, which reads adharmarājā i[va] rājyāt cyutaḥ. The Tibetan reads chos kyi rgyal po rgyal srid nyams pa ltar, perhaps “like a righteous king whose kingdom is ruined.”

n.188Following the Sanskrit. The Tibetan translation seems to render the Sanskrit onomatopoeic expressions phucchakkāra and picikkāra as ’ur sgra, “a roaring sound,” and tug chem gyi sgra, “a clippity-clop sound,” respectively. The Tibetan also includes the imperative expression sod cig pa, “kill him,” which is lacking in the Sanskrit.

n.189Following the Sanskrit. The Tibetan has the slightly expanded expression skyes bu chen po yongs su bskyab tu gsol, yongs su bskab tu gsol, “Great person, protect us! Protect us!”

n.190Following the Tibetan. The Sanskrit manuscript does not make a chapter break here.

n.191The Sanskrit manuscript here reads kecid sarveṇa sarvaṃ māraṃ nopasaṃkrāmantam iti saṃjāṃte [sic], and the Tibetan translation seems to reflect this reading: kha cig gis ni bdud yongs thams cad du ma ’ongs par shes so. The Tibetan interprets the verb form upasaṃkrāmanta in the common sense of “approaching” or “arriving,” and it is used several times in this way previously in this sūtra. However, as Edgerton notes in his entry on the term upasaṃkramati, the same verb can be used in the sense of “violently attack,” and thus it is possible to see a play on words here or simply to translate it as follows: “some thought that Māra was not assaulting him in any way at all.”

n.192Following the Sanskrit, which reads kecit pārijātaṃ kecit kovidāraṃ, seemingly suggesting that there is a distinction to be made between the two. The Tibetan translation reads kha cig gis ni byang chub kyi shing yongs ’du sa brtol du mthong ngo, which suggests that the translators understood pārijāta and kovidāra to be referring to a single tree, perhaps a huge banyan tree with multiple trunks forming a single canopy. See Edgerton’s entries on pāriyātra and kovidāra for explanation and disambiguation.

n.193The Sanskrit manuscript here reads ekacittakṣaṇa­samāyuktayā prajñayā, and above it there is the similar expression ekakṣaṇa­samāyoga. The Tibetan translation renders both these expressions in the same way, skad cig gcig dang ldan pa’i shes rab kyis and skad cig gcig dang ldan pa. Indeed, two similar expressions are used more or less interchangeably in chapter 7 of the Pañca­vimśati­sāhasrikā Prajñāpāramitā Sūtra, for a discussion and edition of which see Watanabe 1990.

n.194Following the Tibetan, which reads kun tu ye shes kyis mngon par rdzogs par byang chub bam. This sentence is missing from the Sanskrit manuscript, but the underlying Sanskrit of kun tu ye shes could be sarvatrajñāna.

n.195After this sentence, we have not included two sentences that are found in the Tibetan translation but not in the Sanskrit manuscript. These sentences in the Tibetan translation can be translated, “Who are the four? The Four Great Kings are Dhṛtarāṣṭra, Virūḍhaka, Virūpākṣa, and the great king Vaiśravaṇa.”

n.196Following the Tibetan translation, which reads tshong pa ga gon dang bzang po dang (“the merchants Trapuṣa and Bhallika”). The Sanskrit manuscript, however, does not name them, but says only “the merchants who are worthy of receiving praise in the three [worlds]” (triṣu saṃvarṇṇīyakā vaṇijaḥ). Given the relative similarity in the required number of syllables and their nature, one wonders if there may have been a corruption in the history of this Sanskrit manuscript, but it also seems possible that the manuscript version operates on the assumption that the reader will know who is meant here and the Tibetan and/or a different manuscript tradition supplies the names.

n.197The terms translated here as “withdrawal” and “withdrawn,” viveka and vivikta respectively, and both translated into Tibetan as dben pa, have the sense of solitariness or isolation, as well as, indeed, the sense of freedom from all conditioned things. See also n.­255.

n.198Following the Sanskrit manuscript, which has animitta, here translated as “groundless” in the sense of having no cause or ground of perception. Nimitta can have the sense of an object on which a perception is based. The Tibetan here reads differently: mi rtag pas ’dul ba rnams kyis ni mig mi rtag pa’o zhes thos so, “Those who could be trained by impermanent heard ‘the eye is impermanent.’ ” This would be redundant, however, because the sūtra has already mentioned impermanence in the first statement above.

n.199Following the Tibetan, which has chu zla lta bu, rendered here as “moon reflected in water,” for which the closest Sanskrit term is udakacandra. The Sanskrit manuscript does not seem to have an equivalent term here, but the situation is complicated by the fact that the next term found in both the Tibetan and Sanskrit is another of several near synonyms: pratibhāsa, “reflection” or possibly “mirror image,” which seems to correspond in the Tibetan translation to mig yor. A third term not found here but often closely aligned with these two expressions is pratibimba, which is sometimes rendered into Tibetan as gzugs brnyan.

n.200The translation takes the cue here from the Tibetan translation, which seems to spell out what may be an abbreviated expression in the Sanskrit. The Sanskrit manuscript reads simply yāvat manaḥ pratītyasamutpannam, “up to the mind is dependently arisen,” while the Tibetan translation has mig la ji lta ba de bzhin du yid kyi bar du yang rten cing ’brel par ’byung ba. Of course, it is possible that the manuscript on which the Tibetan was based may have had a more elaborate expression, too.

n.201Following the Sanskrit. The Tibetan translation says here, “Some heard that the aggregates are impermanent, some that the element s are impermanent, and some heard that the sense spheres are without self.”

n.202Following the Tibetan translation here. The last two sentences are missing from the extant Sanskrit manuscript, but the closeness of the expressions to what is found in the extant manuscript suggests that these sentences were present in the manuscript that was used as the basis for the Tibetan translation.

n.203Following the Tibetan, which begins bcom ldan ’das kyis. There seems to be a confusion in the Sanskrit manuscript regarding the speaker of this sentence. In the manuscript, the sentence begins asmiṃ khalu punaḥ śāntamate (“Moreover, Śāntamati, while this…”), as if Vajrapāṇi were still speaking here about the past, but it seems that in this sentence the unnamed narrator is describing the effect of Vajrapāṇi’s teaching upon the audience in the narrative present. In this regard, perhaps śāntamate may have crept in as an error for bhagavatā, which has the same number of syllables? The Sanskrit and the Tibetan align fairly well for the remainder of the sentence. The Sanskrit manuscript could be construed to say “while this (asmiṃ) chapter of the Dharma (dharma­parivartte), that is, the teaching of the wheel of Dharma (dharma­cakra­nirdeśe), [and] the child’s play of the lion’s roar … was being taught (nirdiśyamāne) …” However, one wonders if there has been a corruption here, too, in the transmission of this Sanskrit manuscript.

n.204Following the Tibetan. The Sanskrit manuscript does not indicate a chapter break here.

n.205Following the Tibetan for the last few sentences, where it seems to indicate a slightly more elaborate phrasing than the Sanskrit manuscript does, but without significantly altering the meaning. For instance, the Sanskrit manuscript does not contain the first interrogative expression above, “Why so?” It also lacks words corresponding to the verbs “understand” and “hear about” in this sentence, as well as the phrase “or to speak of the secrets of the realized ones,” in the earlier sentence.

n.206Following the Tibetan for this paragraph. The Sanskrit manuscript is somewhat less elaborate and repetitious here, and it requires that one carry down the correlative construction from several sentences above. The whole paragraph in the Sanskrit may be translated, “Moreover, Lord of the Guhyakas, those beings will not only have served one buddha; they will have served them for many hundreds of thousands of millions and billions of eons, done the preparations, and put down the roots of virtue here in the Great Vehicle.” At the end of this passage, the Sanskrit manuscript also indicates a chapter break that draws to a close this long chapter of the Sanskrit version, which includes chapters 11 through 14 of the Tibetan translation, as well as chapter 15 of the Tibetan version up to this point. The Sanskrit manuscript’s chapter colophon reads, “The sixth chapter, the vision of the wondrous transformations of the Realized One” (tathāgata­vikurvaṇasandarśana­pari­varttaḥ ṣaṣṭhaḥ).

n.207As will be seen, the use of the term translated here and below as “calming,” upaśama in Sanskrit and nye bar zhi ba in Tibetan, stretches the meaning of the English word “calming.” The original term has the sense of calming down strong emotions, quieting the mind, putting one’s thoughts or emotions to rest, making them stop in such a way that they do not arise again. One might also note that the term is connected with the word śānta in the bodhisattva Śāntamati’s name.

n.208A passage beginning here with Śāntamati’s question is quoted at some length in the Prasannapadā in a commentary on chapter 18, verse 6 of the Mūlamadhyamaka­kārikā on the critique of the self (ātman). For the Sanskrit text, see La Vallée Poussin 1903, pp. 361–63.

n.209The terms saṃkalpa, vikalpa, and parikalpa occur here together.

n.210Here the Tibetan translation and the Sanskrit manuscript are in alignment, whereas the Prasannapadā is different. The latter says, “From the calming of all views comes the calming of all wishes. From the calming of all wishes comes the calming of all afflictions.”

n.211Following the Sanskrit manuscript and the Tibetan translation, as well as the Śikṣāsamuccaya, which also quotes this metaphor. For the Sanskrit, see Bendall 1902, p. 242. The Prasannapadā give a slight variant, saying “all the fruits (phala), leaves, and branches.”

n.212The longest continuous portion of text preserved in the Sanskrit manuscript comes to an end in the middle of this sentence. This portion begins approximately forty pages earlier at folio 141.b in the Tibetan translation. The citation preserved in the Prasannapadā continues for several more sentences.

n.213Following the Prasannapadā, which appears to have a slightly more complete reading here. The Tibetan translation lacks the term for “emptiness” in the list.

n.214Here the sūtra begins to draw out a rendering of the term satkāya by drawing upon its literal meaning in a way one might even call poetic in its literalism. The Prasannapadā reads satkāya iti śāntamate akāya eṣaḥ, and the Tibetan translation, zhi ba’i blo gros ’jig tshogs zhes bya ba de ni tshogs ma yin pa, clearly reflects this underlying Sanskrit. One might alternatively translate the passage, “This so-called real person, Śāntamati, is without personhood.”

n.215Following the Prasannapadā here, which reads na kasati na vikasati, na cinoti na vicinoti. The Tibetan translation reads ’byung ba ma yin pa, rnam par ’byung ba ma yin pa ste. The terms kasati and vikasati would seem to be middle-Indic forms of the Sanskrit verb kṛṣ, and perhaps are used to suggest a semantic connection to the term kāya. In the Pali canon, kasati has the sense of making a furrow or plowing, and vikasati is used in the sense of the blooming of a flower, but in Sanskrit the corresponding verbs also can have a sense of drawing something to oneself or extending mastery over something. Such latter meanings may be behind the choices made in the Tibetan translation, since ’byung ba and rnam par ’byung ba render forms of the verb bhū and vibhū, which can also convey the sense of mastery. Cinoti and vicinoti are forms of the verb ci, and also seem meant to suggest a poetic etymological explanation of the concept of kāya. Attempted translations of these latter two verb forms are not found at all in the Tibetan and thus may have been absent in the manuscript used as a basis for the translation.

n.216Following the Prasannapadā’s reading of na adhyavasyate, which has the sense of dwelling in a place, but also clinging to it or desiring it. The Tibetan translation here, shes par bya ba ma yin pa’o, is more along the lines of something to be understood, which makes sense but doesn’t follow the primary meanings of the metaphors being given in the list.

n.217Following the Tibetan translation, which has only zhi ba, not nye bar zhi ba. Accordingly, the translation has been modified slightly. The Prasannapadā has upaśama (“calming”), as above.

n.218This sentence and several in the section that follows are cited by Kamalaśīla toward the end of the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, p. 28. The term rendered here as “stops burning” is the same term, śāmyati, rendered here as “one calms down,” and is connected to the main terms under discussion here, “calming” and “a calmed one.” This metaphor illustrates the concept of “calming” with the idea that a fire will “calm down” or “become extinguished” when its source of fuel is eliminated. Similarly, the mind will calm down when the mental objects that form the basis for cognition are eliminated.

n.219This sentence is the last one quoted in the long citation in the Prasannapadā.

n.220The previous two sentences in this paragraph are part of the quotation by Kamalaśīla in the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, p. 28.

n.221This sentence and the previous one are part of the quotation by Kamalaśīla in the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, p. 28.

n.222This translation assumes a reading of the underlying Sanskrit as buddhasama­citta. The Tibetan translation, sangs rgyas dang mnyam pa’i sems, may suggest a different interpretation of the compound phrase that could be rendered as “a mind that is the equal to that of buddhas.” However, samacitta typically has the meaning of being even-minded.

n.223From the beginning of this paragraph up to the end of this sentence is including in the long quotation by Kamalaśīla toward the end of the third Bhāvanākrama. For the passage in Sanskrit, see Tucci 1971, pp. 28–29.

n.224The contrast being made in the last two sentences is between those who are said to possess “the power of conditions” (pratyayabala, rkyen gyi stobs) and those who possess “the power of causes” (hetubala, rgyu’i stobs).

n.225The Tibetan term rendered here as “freedom” is ’byung ba, which we understand to be a translation here of the Sanskrit term niḥśaraṇa, that is, escape from saṃsāra. While the Sanskrit is not extant here, this is an attested translation equivalent even in this sūtra, where it is found near the beginning of folio 178.b of the Tibetan translation. However, another strong possibility is that ’byung ba here is translating the Sanskrit term udaya, which in this case would refer to the reality of change or the fact that things arise and pass away.

n.226Though this section is not extant in Sanskrit, the numbering in the Sanskrit manuscript makes it clear that the Sanskrit would not have a chapter break here, but would rather combine this chapter with the next one to form a single chapter.

n.227The Sanskrit manuscript resumes in the middle of the first line of this verse. For “craving” (tṛṣṇā) in the second line, the Tibetan seems to read “becoming” (srid pa), which may suggest a variant, but is more likely an error for sred pa. In the third line, too, where the Sanskrit reads “refine our eyes” (nayanaśodhanā), the Tibetan may contain a variant, as it suggests “engage our eyes” (mig ni spyod mdzad pa, perhaps nayanacaraṇā).

n.228The translation reflects the Sanskrit, but emending the apparent manuscript reading of parvata to parṣadā on the basis of the Tibetan. The Tibetan translation of the last two lines can be translated as follows: “The assembly is looking at your face, Great Seer. / Why have you, Supreme Human, displayed your smile?”

n.229Following the Tibetan, which reads lo legs pa, for which the Sanskrit is subhikṣa, whereas the Sanskrit manuscript does not have this term and instead reads kṣama (“peaceful”), which is sometimes also found in similar lists of these terms.

n.230This list of terms in both the Sanskrit and the Tibetan contains a few words that are difficult for us to understand with confidence, both in themselves and as they potentially correspond (or not) to one other. In the Sanskrit, after the term kubja (“hunchbacked”), which has a clear correspondence in the Tibetan translation sgyur po, the Sanskrit manuscript reads ladaro (seemingly, perhaps, wrong for lolita, “afflicted by tremors”?) and viśastro (“weaponless,” but maybe wrong for viśākha, “without hands”?) before ending with mūko and paṃgur, which are clear enough, while the Tibetan has gtum po (“savage”?), tig pa, mi srun pa (“deranged”?), gla glo can (“possessing barbarous speech”?), lkugs pa, and grum po. Now, grum po would seem to correspond to paṃgu (“lame”), which is the last member of both lists, and either tig pa or lkugs pa could correspond to mūka (“tongue-tied” or “mute”), but the other terms remain unclear and therefore we have hesitated to include them in the main body of the translation.

n.231The Tibetan translation clarifies that the purity under discussion here is the purity of “moral conduct” (tshul khrims), or it used a manuscript that had the term śīla here in place of śuddha in the extant Sanskrit manuscript.

n.232Following the Tibetan, but the Sanskrit manuscript makes a chapter break here, too. It reads, “The seventh, entitled ‘The Prediction Chapter.’ ” So, according to the sequence of chapters in the Sanskrit manuscript, this seventh chapter would include all of chapter 16 and part of chapter 15, according to the Tibetan chapter divisions, from the point at which Śāntamati asks the Buddha about the meaning of the word “calming.” The other difference to note here is that the Sanskrit manuscript ends the chapter immediately after Vajrapāṇi’s last verse, and it begins the next chapter with the identical narration we find here closing this chapter in the Tibetan version.

n.233Following the Tibetan, which adds the term skyes bu (“human being”), for which the Sanskrit is puruṣa, to this list.

n.234Following the Sanskrit here, which reads kiṃ punar jñānaṃ dvayaprabhāvitam, which seems to make better sense of the back-and-forth flow of the dialogue. The Tibetan translation suggests instead that Śāntamati asks, “What knowledge has been predicted?” (ye shes gang lung bstan). The answer is then given, “A kind that is not produced by duality” (gang gnyis kyis rab tu ma phyi ba’o). The Tibetan then picks up right along with the Sanskrit. If one were to follow the Tibetan, then who is saying what becomes confused. Consequently, we have added the names of the speakers at certain points to help the reader stay clear on who is speaking.

n.235Following the Sanskrit, which reads advayakoṭimatās te. The Tibetan translation here has de dag ni gnyis su med pa’i mtha’ la gnas pa’o, “they ‘stand’ on the position of nonduality,” which may reflect a different reading of koṭisthita, rather than koṭimata. We understand koṭi (mtha’) here in the sense of an alternative or option or position that one could take in a conceptual analysis. As the conversation continues, the phrase koṭisthita, “stood (or was situated or fixed) on a position,” will be used, too.

n.236Following the Sanskrit manuscript here and correcting it to read vijñaptisthānaṃ. The Tibetan translation here reads rnam par rig pa ma yin pa’i mtha’ ni gnas so, “It is standing on the position of what is not made known,” which suggests an underlying Sanskrit phrase of avijñaptikoṭisthāna. However, the question that immediately follows in both Sanskrit and Tibetan asks about making something known, and not about what is not made known.

n.237Following the Tibetan here and in the following two sentences, since the Sanskrit manuscript appears to be corrupted here.

n.238Here is another instance of wordplay in this dialogue, on the polyvalent term dharma, which is left untranslated when it refers to the Buddha’s teachings, and then translated here simply as “thing” when it refers to the dharmas or the building blocks of experience or the phenomenal world. For the mainstream, non-Mahāyāna Abhidharma, dharmas are the real things out of which conditioned phenomena are made, and that sense is in the background here and being undermined by Vajrapāṇi’s argument from emptiness.

n.239The second sentence of Vajrapāṇi’s response here is translated from the Tibetan. It is not in the Sanskrit manuscript.

n.240The Sanskrit manuscript reading of tṛṣṇā supports the reading of sred pa found in the Narthang Kangyur, rather than srid pa, which is found in Degé, Stok, and most of the other Kangyurs.

n.241Following the Tibetan. The Sanskrit manuscript does not make a chapter break here, but it is not entirely clear why it should be linked to the chapter(s) that follow it. Ikuma 2013 seems to indicate that all of this chapter in the Tibetan should be included with the previous two Tibetan chapters to constitute a single chapter in the Sanskrit about Vajrapāṇi’s prediction.

n.242The Sanskrit manuscript cuts off here in the middle of this sentence and resumes again after about three pages of the Tibetan translation.

n.243The Sanskrit manuscript resumes at this sentence, which is actually a phrase in the middle of a very long sentence in the original that has been broken up into several sentences in this translation.

n.244Following the Sanskrit, which seems to fit a bit better here. The Tibetan translation say that Vajrapāṇi spoke to “Śakra, Brahmā, the Lokapālas, and the great assembly of gods.”

n.245Following the Sanskrit manuscript here, and emending it to read saṃtarpya saṃprayā­panī­tapātraṃ dhautapāṇiṃ bhagavantaṃ viditvā…, which brings it in line with a closely parallel expression found, for instance, in the Divyāvadāna, p. 53. The Tibetan translation of the Tathāgataguhya, following Degé, reads bstabs te tshim par byas nas lhung bzed gyu te | bcom ldan ’das phyag bcabs par rig nas… For lhung bzed gyu, translating apanītapātra (“put away [his] bowl”), Stok Palace and Narthang read the expression lhung bzed bkrus, which suggests a variant of dhautapātraṃ, “washed [his] bowl.” It is not entirely certain what the Tibetan phrase phyag bcabs is intended to translate or say: “[his] hands hidden”?

n.246Following the Tibetan, as the Sanskrit manuscript does not make a chapter break here.

n.247Following the Sanskrit here, which reads viśeṣagāmitāyāṃ yogam abhyudgamiṣyanti. This varies from the earlier expression viśeṣatāyām abhyudgaccheyuḥ, whereas the Tibetan translation here repeats the same translation of the earlier phrase: khyad par mngon du ’phags par ’gyur.

n.248Beginning with this sentence, there appears to be another rather close parallel with a long passage of the Bodhisattvapiṭaka (Toh 56), this one in chapter three of the latter text. This parallel, though not exact, continues for approximately eight folios of the Tibetan translation.

n.249This paragraph and the next one are quoted in the Sūtrasamuccaya. See Pāsādika 1978a, part 1, pp. 25–26. Therein, they are connected to the citation from this same sūtra noted in n.­79.

n.250Following the Sanskrit here, even though the Tibetan does not include the first half of this sentence about the ten forms of good conduct. It is possible that it was added at some point to an earlier iteration of the sūtra, since the context right here is primarily about what the noble son or daughter does not do.

n.251Following the Sanskrit manuscript here, which reads anapriyarupinaḥ [sic]. The Tibetan reads gcam bur mi smra ba, “who do not utter flattery,” which suggests the underlying apriyavādinaḥ. One can also imagine a hypothetical reconstruction from the Sanskrit and Tibetan, taken together, of anapriyavādinaḥ (“who do not utter disagreeable words”). We have chosen the prioritize the extant Sanskrit reading because of the apparent triad here of speech, body, and mind.

n.252Following the Sanskrit manuscript. This sentence is absent from the Tibetan translation, and its absence is perhaps justified by the fact that this series of statements is structured around the perfections (pāramitā), yet a few sentences later both the Sanskrit and the Tibetan include a statement about the opposite of wisdom or discernment (prajñā): “the state of one whose discrimination is faulty” (duṣprajñatā).

n.253Following the Sanskrit manuscript here. The second half of this sentence is absent from the Tibetan.

n.254Following the Sanskrit here, which reads abhūtasaṃjñā, whereas the Tibetan translation says that one examines “the bewilderment caused by what is unreal” (yang dag pa ma yin pa kun tu gti mug pa de, abhūtasaṃmoha).

n.255A single term, vivikta, dben pa in Tibetan, has been double-translated here as “isolated and free,” to bring out some of its polyvalence. For more on this term, see n.­197.

n.256The Sanskrit manuscript here attests to the somewhat unusual term parigodhi (“one is who attached to or greedy for something”). The Tibetan translation has yongs su spyod pa, perhaps suggesting an underlying paribhoktā (“an enjoyer”). In either case, the bodhisattva has no interest in forming “a household” (parigraha, yongs su ’dzin pa).

n.257Following the Sanskrit manuscript, which reads pratiśaraṇa (“placing confidence in”), which is related to śaraṇa (“refuge”), three times. The Tibetan translation rton pa agrees with the Sanskrit in the first two uses of the term, but in place of the third use, where the Sanskrit manuscript has prati­śaraṇabhūta (“one in whom confidence is placed”), most Kangyurs read ston pa (“the teacher”). However, the more difficult reading of rton pa is preserved in the Peking and Yongle Kangyurs. The evidence from the Sanskrit manuscript and the close similarity between the Tibetan terms suggests that the more difficult reading was altered to the easier one at some point.

n.258The microfilm of the Sanskrit manuscript is quite difficult to read here, but Szántó reads mahātmānaṃ (“the great self”), and this looks like the correct reading. The Tibetan translation here reads theg pa chen po, implying an underlying mahāyānaṃ (“the Great Vehicle”), and while this seems plausible, it also seems that the sūtra is making a contrast here between the endeavor for oneself alone and the examination, for the sake of all beings, of the “great self,” which construes here with the knowledge of the buddhas. This reading is further supported by the discussion four paragraphs later of “greatness of character (or self)” (mahātmatā, che ba nyid, also translated herein as “greatness”), where it is connected again with the knowledge of the buddhas.

n.259This whole paragraph and part of the next paragraph are quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 357.

n.260This paragraph up to and including this sentence is quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 357.

n.261After this sentence, the Sanskrit manuscript reads punar aparam asti saṃvṛtyā (“moreover, [if] they exist conventionally”), which is absent from the Tibetan translation, but the same phrase is found just below in both the Sanskrit and the Tibetan, which suggests that perhaps the Sanskrit manuscript contains an accidental duplication of the phrase here. Alternatively, one could read it as a qualification of the intended meaning of “existence,” both here and below.

n.262The Śikṣāsamuccaya quotes this final sentence at the end of the quotation that includes the portions noted above. For the Sanskrit, see Bendall 1902, p. 357. The Sanskrit manuscript of the sūtra also breaks off again here and does not resume until the middle of folio 187.b of the Tibetan translation.

n.263The translation of the term ’byung ba here is “freedom,” assuming an underlying niḥśaraṇa. Alternatively, one could translate, “For those who are proud, there is no escape.” See also n.­225. The parallel passage in the Bodhisattvapiṭaka does not seem to contain this sentence or the preceding two sentences, and thus it could not be consulted for comparison.

n.264It is at this point that the long parallel with the passage in chapter three of the Bodhisattvapiṭaka seems to come to an end.

n.265This is a transcription of the reading of this mantra in the Degé edition with only light editing in the form of spaces between some of the possible words. The Pedurma indicates quite a number of variant readings among the several different editions it reports, mostly for particular syllables and most commonly for a lengthening of the vowel. As an example of the general pattern of variations one can find, here is a full transcription of the Stok Palace reading: āhula | hulavatī | hularaṇā | aguhā | ca ca ca ca ca | niṣuranā | kśuta | kśatā | kśaya | kśaya | akśayasamā | śamamanikarā | hahula huvati | khilavati | kinābiṣatissaraṇī | jajājajamati | avatā | nivatā | varta vartānusāriṇī | bhutānasmṛti ho | manuśānusmṛti ho | devānusmṛti ho | nāgānām | yakṣāṇām | gandharvāṇām | kinnāraṇām | mahoragaṇām | svastir bhavatu sarvasatvānām | vikirnatu mahāparrvatā | samvyathatan haraṇi | garjatu samudravega iti.

n.266The Tibetan gives the somewhat unusual expression here and just above, ’dar pad pad, which has been translated with the verb “to tremble,” as if it were simply a form of ’dar ba. A bit further below one also finds the term rab du ’dar ba, which also means “to tremble.”

n.267There is a play on words here between adhimokṣa, mos pa in Tibetan, rendered as “ardent devotion,” and vimokṣa, or rnam par grol ba in Tibetan, rendered as “liberation.”

n.268The main text transcribes the reading of this mantra in the Degé edition, with only light editing in the form of introducing spaces between some of the possible words. The Pedurma indicates about twenty-five variant readings among the several different editions it reports, and here is a full transcription of mantra in the Stok Palace edition: tadyathā | hulu | hulā | hule | dhīre | hārīte | akuśalakṣaye | cale | calavati | sāre | kśure | kśāntivale | kśame | kśāntikare | śame | hulu | alikhi | khirikhare | āveśanisaraṇe | jaye | jayavati | sthire | avarte | vivarte | avartani | maitrabhutānan sanggrahe | dame | damasamvanne | tīrṇetarayase | svayan | svasti devebhyo | nāgebhyo | yakśebhyo | gandharvarākśasebhyo | manuṣe amanuṣebhya svastir bhavatu sadāsukham.

n.269Though the Sanskrit is not extant here, Ikuma 2013 suggests that there would be a chapter break here according to the division of the Sanskrit manuscript.

n.270The term translated here as “apathy,” btang snyoms in Tibetan, upekṣā in Sanskrit, is the same term often translated in Buddhist literature as “equanimity,” and while it retains this sense here, too, the context also foregrounds other shades of the term’s ordinary meaning: “indifference,” “disregard,” “impassivity.”

n.271The Sanskrit manuscript resumes toward the end of this sentence.

n.272The Sanskrit manuscript makes a chapter break here, too. It reads, “The ninth chapter, entitled ‘On Śūrabala.’ ”

n.273The Sanskrit manuscript breaks off again toward the end of this sentence and resumes on folio 192.a of the Tibetan translation.

n.274This whole paragraph is quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 274.

n.275The Sanskrit manuscript resumes in the middle of this sentence.

n.276Once again, there is a play on words here in Sanskrit between “ardent devotion” (adhimukti, mos pa) and “liberation” (vimukti, rnam par grol ba). See n.­185.

n.277This sentence and several that follow are quoted early in the first chapter of the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, pp. 7–8.

n.278The quotation in the Śikṣāsamuccaya ends with this sentence. For the Sanskrit, see Bendall 1902, pp. 7–8.

n.279Following the Tibetan interpretation here, zlog par bgyid, which seems to suggest an emendation of the Sanskrit to udghātayet, which possibly goes a bit better with the other two verbs used here. However, one should still strongly consider the extant Sanskrit manuscript reading of udghāṭayed [sic] (“[who] would reveal/explain [it]”), which is admittedly the more difficult term, on the meaning of which see Edgerton’s entries on udghaṭaka and udghāṭaka.

n.280The “royal plural,” which is evident in the Sanskrit manuscript, has been retained in translation here.

n.281This may be an oblique reference to the sūtra Eliminating Ajātaśatru’s Remorse (Ajātaśatrukaukṛtyavinodana, Toh 216).

n.282Following the Tibetan chapter title and number. The Sanskrit manuscript also makes a chapter break here, and reads, “Thus is the tenth chapter, ‘On Ajātaśatru.’ ”

n.283The verbal phrases “retain in their minds” and “hold on to” both translate the same verb, dhārayanti (rendered as ’dzin in the Tibetan translation). This term can mean “to hold,” “to memorize,” and “to remember,” and is a form of the verb from which is also derived the important and difficult term dhāraṇī (gzungs in Tibetan), “powerful memory and the powerful mnemonic formulas that support it,” which becomes the focus of the conversation that follows below. One may also note here again, as so often throughout this sūtra, that one can see a play on the meaning of the word dharma in the phrase, saddharma, “the true Dharma”‍—that is, the Buddha’s teaching‍—and dharma in the phrase sarvadharma, “everything”‍—that is, all things and all the building-blocks out of which things are made.

n.284Following the Sanskrit manuscript, which is difficult to read here and probably requires emendation, but seems to say kāya ca vijñaptau [sic], whereas the Tibetan translation reads gzhan gyi rnam par rig pa la (“regarding the imposition of another”), which also makes sense.

n.285Following the Sanskrit manuscript reading of yaḥ pratiṣṭhito [sic], whereas the Tibetan reads su gnas pa med pa (“one who does not stand still”).

n.286Following the Tibetan translation. The Sanskrit manuscript does not make a chapter break here.

n.287There are several layers of punning that occur in this paragraph. The first point is simply to note that the term rendered here and below as “accessing,” anupraveśa in Sanskrit and translated into Tibetan as rjes su ’jug pa, has the primary sense of entering, and the strong secondary meaning here of understanding. Both meanings are implied here simultaneously. Secondly, in this sentence there seems to be a play on the word akṣara, first in the meaning of what is “imperishable” or “unchangeable” (that is, nirvāṇa), and secondly in the sense of “a syllable” or a written character of a syllabary, what an English speaker would call a letter of the alphabet. This pun seems to have been missed by the Tibetan translation, which translates both uses of the term with yi ge (“syllable”), and thus the Tibetan could be translated as “the knowledge that accesses the syllables in the syllables.”

n.288There is a pun here on the word ākāra, which can mean both “aspect,” as it does in the name of this dhāraṇī , and “the syllable ā.” The Tibetan translation also picks up on this pun and translates here accordingly, a shes bya ba. There is also the play again on the two meanings of the word dharma, “teaching” and “thing.” Also, we alternated the translation of anupraveśa here with “point of entry.”

n.289“Right” and “wrong” are translations here of dharma and adharma, respectively, with the latter having an especially strong meaning of incorrectness of thought and action, though the other meanings of dharma remain active here, too. The Tibetan translation says, “The syllables do not form the idea that ‘this is a dharma’ or ‘this is not a dharma’ ” (yi ge de dag ni chos zhe’am | chos ma yin zhes bya bar rnam par mi rtog go), with the same ambiguity between “thing” and “teaching” in the meaning of the word dharma.

n.290The Tibetan translation understands the word ākāra here in the compound ākārānupraveśa in its other meaning of “the syllable ā.”

n.291This interpretation of the Sanskrit, which reads tat kasmād acchedā tena dharmatānugatā akṣayā tena dharmatānugatā [sic], takes acchedā and akṣayā here as feminine nominative singular words going with the prior feminine noun gaṇanā (“count”), which was paired earlier with the verb upagam (“to reach”). The Tibetan translation, de ci’i phyir zhe na | rgyun chad pa med pas des chos nyid rjes su rtogs so | des cho nyid mi zad par rjes su rtogs so, would seem to understand the passage a bit differently: “What is the reason for that? The one who is not interrupted has thereby reached an understanding of the nature of things; he has reached an understanding of the nature of things as inexhaustible.” The Sanskrit term anugata means “followed,” “arrived,” “entered,” “reached,” and also “understood” or “realized.”

n.292Following the Tibetan, which reads nyong mongs pa rnams kyis mi rdzi ba yin no | pha rol rgyol ba rnams gyis mi thub pa yin no for the last two sentences. The Sanskrit manuscript seems to transpose the order of the verbs. Thus, the Sanskrit reads durdhakṣo [durdharṣo?] bhavati kleśaiḥ | anavamṛdyo bhavati pratyarthikaiḥ.

n.293Following the Sanskrit asaṃśayo bhavati jñānena, and taking the instrumental ending here as in many of these sentences in a locative sense. The Tibetan reads ye shes gyis chags pa med pa yin no (“their knowledge is unattached”). Perhaps it reflects a variant reading of asaṃsaktaḥ here?

n.294The term paripācana in Sanskrit, yongs su smin par byed pa in Tibetan (“bringing [beings] to maturity” or “ripening”), also has the equally primary meanings of cooking and applying heat, which are seen here.

n.295There is a play on words here, because the Sanskrit term sudānta, “well trained” (shin tu dul ba in Tibetan), also has the meaning in Sanskrit of “having fine teeth (or tusks).”

n.296Following the Sanskrit, which reads apatyaprāptaḥ svacittavaśavarttitayā. In the classical literature of India, Brahmā is sometimes considered the progenitor of all beings, some of whom are born from his mind. However, the corresponding Tibetan translation here, dbang thob ste rang gi sems la dbang sgyur bas, could be rendered as “[they have] attained mastery through becoming a master of their own mind,” perhaps reading pātya in place of apatya.

n.297Following the Sanskrit, which reads niraṃgaṇo [sic], a term that can mean coloring or painting, but also refers to dancing, public theater performances, and the like, the avoidance of which is one of the eight or ten precepts (śīla). The Tibetan translation here reads nyon mongs pa med pa (“[they are] without affliction”).

n.298The Sanskrit manuscript here reads nirupakramo, which may contain a play on words. Upakrama in Sanskrit has several meanings. Perhaps its most basic sense is the simple act of going near or approaching someone or something, but it also means attending to a sick person and the cure that heals a sick person. So, nirupakrama here could mean “they need no remedy,” but at the same time, in Buddhist literature the term also can mean a violent attack, similar to the sense of upakkama in Pali. The Tibetan translation gnod pa med pa (“they are not harmed” or perhaps “they do no harm”), picks up this level of meaning.

n.299“To arrive at the end” (paryantaṃ adhigantum) also has the sense of “to understand or realize something,” which is brought out by the Tibetan translation mtha’ rtogs pa.

n.300The logic of this exchange relies to a large extent on the flexibility of the word varṇa, a flexibility demonstrated by the fact that the Tibetan uses two words to translate it here: bsngags pa (“praise”), and kha dog (“color”). The term in Sanskrit derives from a verb that means “to paint,” “to color,” “to describe,” “to depict,” and “to praise,” among other meanings. The translation of varṇa into English has been modulated using “description,” “description in praise of,” “praise,” “color,” and “praise the form of,” in order to give a sense of the logic.

n.301Following the Sanskrit manuscript. The Tibetan translation specifies a list of the seven jewels here: “the wheel, the elephant, the horse, the jewel, the queen, the steward, and the minister.”

n.302Following the Sanskrit manuscript. The Tibetan translation includes here the standard list of the four means of drawing others to oneself.

n.303There are several puns and poetic elements involved in this analogy. First of all, the entire comparison relies on the similarity between the words dharaṇī (“the earth”), literally that which is the support, and dhāraṇī , rendered here by the phrase “a powerful memory and the formulas that support it.” In addition, the parallel verbs and verbal phrases used in the analogy say that the earth emits or injects (utsṛjati) life-giving juice (rasam) into the flora of the earth and in the same way the bodhisattva emits (samutsṛjati) detached thought (cittam asaṅgaṃ) to all beings by an emission (utsarga) of motivation for the Dharma.

n.304Following the Tibetan translation here. The Sanskrit lacks this phrase.

n.305“Creation” is a translation of the Sanskrit term prabhāvana, whereas the Tibetan translation of this term, rab tu brjod pa, shades closer to “expression,” which is also a viable translation of prabhāvana here.

n.306“Daily concerns” is a translation of the Sanskrit phrase uṣavicārāḥ, whereas the Tibetan reads here spyod pa, which is attested as a translation of vicāra, but has more of a sense of “conduct,” which also works for vicāra here.

n.307This is a challenging passage, and it seems to us that there may be some kind of discrepancy between the Sanskrit manuscript and the Tibetan translation. The translation above is based on our best guess about the sense of the Sanskrit, which reads anugataḥ so kṣayatām anantatāṃ sarva­tragamanatāṃ sarva­praveśānāṃ tasya nādhiṣṭhānaṃ manaso… [sic]. The Tibetan translation reads de mi zad pa dang | mtha’ yas pa dang | thams cad kyi rjes su ’gro ba dang | thams cad la ’jug pa’i rjes su song ba yin no | de la rgyun chad pa med do | chags pa med do, and could be translated as follows: “[The bodhisattva] has become inexhaustible, unlimited, all-pervasive, and engaged with everything. For him, there is no ruin. He is not attached.”

n.308Following the Tibetan here for the second and third types of empowerment. The Sanskrit manuscript has “the empowerment in the use of the Dharma and the empowerment in the Dharma” as the second and third types.

n.309The translation here reflects the Sanskrit manuscript. The Tibetan translation says “…the teaching on the inexhaustible nature of the analogies in praise of the virtues of powerful memory and the formulas that support it…,” which is the same phrase that is used just below as the title of this chapter in the Tibetan translation.

n.310Following the Tibetan translation. The Sanskrit manuscript does not make a chapter break at this point.

n.311Following the Sanskrit manuscript here, and translating the verb pracaret as “may be circulated,” as it was translated in a similar context earlier. The Tibetan translation of spyod pa is closer to “may be practiced.” The Tibetan translation here also has the slightly more elaborate expression in the second half of the sentence, similar to what was seen earlier in both Sanskrit and Tibetan, and may be translated as “…so that it may be practiced (spyod pa) in Jambudvīpa in the latter time, that is, in the final five-hundred-year period of the Dharma.”

n.312On the translation of gupti here as “form” in the phrase “words that protect the form of the Dharma” (dharma­guptyāra­kṣaṇapada), see Edgerton’s entry on the term gupti. The Tibetan translation of this term is sba ba (“hiding” or “place of concealment”), which reflects the ordinary meaning of the term in Sanskrit.

n.313The main text of this translation gives the transliteration of the Degé edition of the Kangyur, and a translation of the Tibetan, also based on the Degé edition, with only light editing in the form of spaces between some of the possible words. As one can see, the mantra has been rendered in the Tibetan translation partly by transliterating some Sanskrit syllables and partly by translating some words and phrases. Other editions of the Kangyur show several variant readings of the transliteration and translation. When compared to the extant Sanskrit manuscript, both the transliteration and the Tibetan translation demonstrate an obvious relationship to the Sanskrit text, but at the same time it is possible to construe the Sanskrit differently in places. No attempt has been made here at a critical reconstruction of the mantra. Here is a transcription of the Sanskrit manuscript’s reading of the mantra for comparison (the commas represent comma-like dashes in the manuscript and the bracketed asterisk represents an illegible character): jaya jayamati | jayaśatru, āle | amale | aluḍite | name | namayati | nāmasandhi | utte | uttamati | uttaraṇi | āre, āramaṇī | āriśamani | āmule | mūlābaḍhḍhite | mūlānugate | aṭṭe, maṭṭe | maṭṭiṭhe | ābhimāḍite, khurukhuru, khurusandhi | dharmānugate | dharmatakṣe | dharmapraveśe | sare, sara, sare | abhede | bhedasandhi | ehi, ehani | enanugate nigraho mārāṇāṃ [*]nirghātanaṃ, tīrthānāṃ mohaṃnaṃ | dharmadveṣināṃ vidhamanaṃ kleśānāṃ ujvālanaṃ dharmanetrīnāṃ | ārakṣā kathitānāṃ | āvaiṣaṇaṃ nirvāṇasya praho bodhisatvaparicārakānāṃ parisaṃsthāpanā parṣadaḥ kāyānupradānaṃ dharmaśravaṇikānāṃ samanvāharatvaṃ saṃyaggatānām avalokanaṃ samyakpratipannānāṃ āmukhībhāvatvaṃ mantrapadāni mā praṇaśyantu mantrasandhi ajānanatvaṃ udāharaṇajñānaṃ | akṣuṇatā | anavamṛdyatā | deśanāsvabhāvatvaṃ. It should be noted that this mantra circulated separately along with a second one below in the Dhāraṇī­saṃgraha, for an edition of which see Hidas, Powers of Protection, pp. 83–84. The illegible syllable in this manuscript is not rendered therein.

n.314There is a play on words here. The phrase rendered as “keep … safe,” is guptiṃ kariṣyāmo [sic] in Sanskrit, sba bar bgyi in the Tibetan translation, also means “to conceal,” “to hide away,” “to put somewhere for safekeeping.” The expression “anyone else who looks for a way to penetrate it” (‘nyaṃścāvatārā prekṣiṇo [sic] in the Sanskrit, glags lta ba gzhan dag in the Tibetan), has the double meaning of “looking for a way to understand it” as well as “looking for a weak spot in the defenses that one can penetrate.”

n.315As with the previous mantra, the treatment of this mantra in the Tibetan translation is a hybrid of transliteration and translation, and other editions of the Kangyur have variants. No attempt has been made to construct a critical edition. The main text of this translation gives the transliteration of the Degé edition of the Kangyur, and a translation of the Tibetan, also based on the Degé edition, with only light editing in the form of spaces between some of the possible words. Here is a transcription of the Sanskrit manuscript’s reading of the mantra for comparison: jaye, durjjaye, jayamati | same, śatrunirghātani, amūle mūlaparicchinne mārasainyavitrāsani | muktemati | śuddhe abhede | bhaya, mohani bhārohaṇi vante | vidya vidyavarottame, nigrahaṃ paravādināṃ | dharmavādināṃ saṃgrahaṃ ārakṣā dharmagaṃjasya vidye buddhaprakāśite | ame ame, mamacchede | arthe | arthanistaraṇe caturṇṇāṃ lokapālānāṃ āveśanapadāni bhāṣitāni | vīre | vīramati | gupte | śubhe | śubhavatī same | śakrasya devarājasyehāveśanaṃ kṛtaṃ | maitre śomavati | kṣānti kṛtsnakaruṇodāhṛte prītiupekṣasaṃpanne | brahmāpyāveśito iha | araḍe | varaḍe | kha kha | amūle | mūlaśodhani | mārasya nigrahārthāya ime mantrāḥ prakalpitāḥ | adhiṣṭhitaṃ narendreṇa idaṃ sūtraṃ subhāṣitaṃ | pracariṣyati tatkāle yatra jñātā bhaviṣyati | ime ca bhāṣitā mantrā medinī ca prakaṃpitā | samāgatāḥ sarvamārā idaṃ vacanam abravit | vayam ārakṣiṣyāmas tāṃvijñāṃ dharmabhānakāṃ | yeṣāṃ hasta idaṃ sūtraṃ kāle yāsyati paścime. There are metrical patterns interspersed within the mantra, along with some phrases the grammatical form of which makes it easier to discern possible meaning. The passage ends with three apparent verses that have a clearer grammatical form in Sanskrit, and thus were translated into Tibetan, and now into English, but there is no indication of any change in voice throughout the passage. These verses are also found in the version of this mantra that is preserved in the Dhāraṇī­saṃgraha, for the edition of which see Hidas, Powers of Protection, pp. 83–84.

n.316Following the Sanskrit manuscript, which has less elaborate phrasing here. The Tibetan translation includes the standard description of this realized one “as a worthy one, a perfectly awakened one, one perfected in knowledge and conduct, a sublime one, a knower of the world, an unsurpassable trainer of those ready to be trained, a teacher of gods and humans, a blessed buddha.”

n.317Following the Sanskrit. The Tibetan translation has “support and protect (yong su bsrung ba) this awakening …” Also, the Sanskrit manuscript breaks off at the end of this sentence and picks up again only at the very end of the sūtra, where it has a final chapter title and colophon.

n.318This verse is quoted in the Śikṣāsamuccaya. For the Sanskrit, see Bendall 1902, p. 357.

n.319Following the Tibetan translation. The Sanskrit manuscript includes two lines of writing in what appears to be a different hand, including a chapter title for what appears to be the last chapter of the sūtra, as well as a colophon for the manuscript. It reads: api tu khalu punar bahu­kalpa­koṭīnayutaśata­sahasraparyupāsitās te bhaviṣyanti kṛta­pari­karmāṇo varopita­kuśamūlā iha mahāyāne || iti tathāgata­guhyakāvi­kurvaṇa­sandarśana­pari­vartta ekādaśapaṭala samāptā || || śubham astu || saṃvat 224 caitrasukladaśami ādityavāra siddha yakājula ||. One might translate as follows: “…what is more, those who have attended over many hundreds of thousands of millions and billions of eons will have completed their preparations and will have put down the roots of virtue in this Great Vehicle. Thus is completed the eleventh division, the ‘Chapter of the Vision of the Wondrous Transformations of the Guhyaka and the Realized One.’ Finished on Sunday, the tenth day of the half-month of the waxing moon in the month of Caitra in the year 224…” The year 224 is equal to approximately 1104 ᴄᴇ, and as Szántó points out, this early date probably indicates that the colophon was copied from an earlier manuscript. We remain uncertain about the meaning of the final syllables yakājula.