Notes

n.1As discussed below, this is not the title given in the sūtra’s colophon.

n.2That yang dag par sdud pa reflects the translation choice of the original translation team, rather than later editors or compilers of the Tibetan canon, is suggested by the use of this title in the imperial-period translation catalogs.

n.3The use of yang dag par brjod pa to translate dharmasaṅgīti is confirmed by extant Sanskrit citations from the sūtra in Indic sources such as Śāntideva’s Śikṣāsamuccaya.

n.4Though attributed to Nāgārjuna, there is some doubt if he is the true author of this work, or if this is a pseudonym used by an otherwise unidentified author.

n.5A number of these citations appear to be taken not from the sūtra itself, but from Śāntideva’s Śikṣāsamuccaya.

n.6Ringpapontsang 2017, pp. 75 and 193.

n.7Sanskrit citations of the Dharma­saṅgīti­sūtra are found in Śāntideva’s Śikṣāsamucaya, Kamalaśīla’s Bhāvanākrama, and Prajñākaramati’s Bodhi­caryāvatāra­pañjikā. Prajñākaramati appears to cite the Śikṣāsamuccaya rather than the Dharmasaṅgīti itself, so while there is some minor variation in the cited passages his work contains no unique citations. Most of the citations from the Bhāvanākrama are also shared with the Śikṣāsamuccaya, though it contains a few unique citations as well. All citations from the Śikṣāsamucaya and the unique citations from Bhāvanākrama have been noted below. A citation purported to be from the Dharma­saṅgīti­sūtra is found in Haribhadra’s Abhi­samayālaṅkārālokā and Abhāyakaragupta’s Amnāyamañjarī and Abhayapaddhati, but an equivalent passage could not be identified in Tibetan translation of the sūtra.

n.8Fo shuo fa ji jing 佛說法集經 (Dharma­saṅgīti­sūtra), Taishō 761 (CBETA; SAT).

n.9See Hermann-Pfandt 2008, p. 54, and Kawagoe 2005, p. 9.

n.10A Sanskrit witness of the Śikṣāsamuccaya is available in a fourteenth- to fifteenth-century palm leaf manuscript written in a Bengali script. It is presently held by the Cambridge University Library (MS Add. 1478;). The manuscript was edited and published by Cecil Bendall in 1902. A transcript of this manuscript prepared by Jens Braarvig was the main source consulted for this translation, but the manuscript was checked regularly to confirm or correct uncertain readings.

n.11Tib. khyad par lung ston pa. This is understood as a translation of the Skt. term akṣūṇavyā­karaṇa.

n.12Tib. chos dang don. This translation understands the phrase to refer to phenomena and what designates them conceptually. The phrase could be interpreted in a number of ways.

n.13Following F, C, J, K, and Y in reading shes pa. D reads shes pas.

n.14Tentative for dbang gcig pa. S uniquely reads dpang gcig pa, “sole witness,” which is also a viable interpretation.

n.15The following passage describes the ten strengths of a thus-gone one, though in a slightly different order than other texts.

n.16Tib. pha rol. This may refer specifically to those who do not follow the Buddhist path.

n.17This is not typically counted as one of the ten strengths, and appears to replace “knowledge of death and birth” found in the more common set of ten.

n.18This translation follows F, K, Y, N, and S in reading sangs rgyas rnam pa du yod pa. Degé reads sangs rgyas rnams du yod pa, “how many buddhas there are.” In the corresponding section below, this question is phrased more simply as “understand the Buddha” (sangs rgyas khong du chud pa).

n.19This translation follows F, K, Y, and S in reading dpe rnams kyis. D reads de rnams kyis, “by them.”

n.20This translation follows F, H, K, Y, N, and S in reading sangs rgyas kyi mthu dang sangs rgyas kyi byin gyis brlabs kyis. D omits sangs rgyas kyi mthu dang.

n.21This translation follows H, K, Y, and N in reading blo gros rgya chen po, to align with the use of the phrase above. D omits rgya, and thus would read “great intelligence.”

n.22This translation follows F, H, K, Y, N, and S in reading rtsol ba’i rnam par rtog pa. D reads rtsom pa’i rnam par rtog pa, “concepts of commencement.”

n.23The Tib. term legs par ’gro ba is understood to translate the Skt. sugati.

n.24It is possible, but not certain, that Kamalaśīla cites an abridged version of this line in the Bhāvanakramā. See Tucci 1958, p. 199.

n.25Tib. yang dag ma yin pa, here understood as a translation of the Skt. abhūta.

n.26The translation of points 9 and 10 are tentative.

n.27F, H, K, and S, among others, make it clear that the break between points 4 and 5 are as translated here. In D the line break (shad) is omitted.

n.28This is a slightly alternate list of the twelve traditional branches of Buddhist scripture. In most lists, the twelfth genre is “instruction” (upadeśa), but this text has “expositions” (viniścaya). The traditional list in Sanskrit consists of sūtra , geya, vyākaraṇa, gāthā, udāna, nidāna, itivṛttaka, avadāna, jātaka, vaipulya, adbhūtadharma, and upadeśa.

n.29Tib. phur ma, which is here understood as a translation of the Skt. puṭa or puṭika. This is then inferred to refer to the container filled with the mixed colors, rather than the individual colors that were poured into it.

n.30Tib. khri lam, which is here understood to refer to a road along which palanquins, chariots, and other conveyances of the nobility can pass.

n.31As noted above, the topic of this passage is described differently here than it is in the passage listing Mativikrama’s questions. There, this question concerns “the different aspects of a buddha” (sangs rgyas rnam du yod pa).

n.32The “essential buddha, “enjoyment buddha,” and “emanation buddha” appear to refer to the three bodies of a buddha, the dharmakāya , sambhogakāya, and nirmāṇakāya.

n.33Ācārya (slob dpon) and upādhyāya (mkhan po), respectively.

n.34Tib. chos tsam du zad pa, which is here understood as a translation of the Skt. dharmamātra.

n.35This translation follows C, H, F, J, K, Y, N, and S in reading yul med pa. D has yul med pas.

n.36Tib. chags pa; Skt. rajyate.

n.37The preceding passage in which the eight worldly concerns are enumerated, is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 264, and Goodman 2016, p. 251. The terminology and syntax of the English translation are informed by the Sanskrit.

n.38This translation follows the D reading de bzhin gshegs pa’i sku rtag pa mi zad pa. C, F, J, K, and Y omit rtag pa. S adds chos, giving the reading de bzhin gshegs pa’i sku rtag pa chos mi zad pa.

n.39These are the four correct discernments (catuḥ­pratisaṃvid; so sor yang dag par rig pa bzhi).

n.40“Propriety” (apatrāpya; khrel yod pa) differs from “conscientiousness” (hrī; ngo tsha shes pa) insofar as the former involves concern with what others will think, and the latter indicates a personal, internal conviction about moral behavior.

n.41These are the four means of attracting disciples.

n.42This is the set of four māras (caturmāra; bdud bzhi) that are obstacles to those on the Buddhist path.

n.43This is connected to the previous point insofar as followers of the Great Vehicle hold that śrāvakas and pratyekabuddhas only give up the obscuration of the afflictive emotions, not the cognitive obscurations.

n.44This line is cited in Kamalaśīla’s Bhāvanākrama. See Tucci 1958, p. 197

n.45There are, inexplicably, eleven points in this section of the “ten approaches.” All versions of the Tibetan translation consulted include these eleven points.

n.46That is to say those who neglect to address the activities of Māra, not those who neglect to engage in them.

n.47Tib. phung po gsum pa’i yongs su sngo ba. The precise referent of “the three sections” is unclear. The term often refers to a set of confessional practices associated with the bodhisattva path, and it can refer to the acts of confession, rejoicing, and requesting teachings.

n.48The section that follows presents a version of the “seven jewels” (saptaratna; rin chen bdun) of a universal monarch.

n.49From the context it appears that the term garuḍa (nam mkha’ lding) refers to Viṣṇu’s mount rather than garuḍas in general. This interpretation is based on the preceding reference to Indra’s mount Airāvaṇa.

n.50This tentative translation follows F and S in reading mi rin po che. D has mig rin po che, “precious eye.”

n.51Tib. blon po rin po che ni rin po che rigs ma lta bu. Translation tentative.

n.52This phrase seems to imply that he keeps an eye on matters surreptitiously.

n.53Referring to the afflictive obscurations and obscurations to meditative attainment.

n.54Referring to the images that seem to appear in the spot on a peacock’s tail feather.

n.55This translation follows C, F, J, K, and Y in reading dge ba zag pa med pa’i tshogs. D reads dge ba zad pa med pa’i tshogs, “the collection of inexhaustible virtue.”

n.56There are traditionally a set of ten objects of meditative immersion (Skt. kṛtsna; Pali kasiṇa; Tib. zad pa), but only nine are given here. Additionally, this list is somewhat at variance with other presentations by listing the elements based on their properties. Thus we have “solidity” instead of “earth,” “wetness” instead of “water,” “heat” instead of “fire,” and “movement” instead of “air.” The full set of ten objects of meditative immersion, using their more standard terminology, appears in this text at F.29.b. See 1.­147.

n.57Tib. phyad par lung ston pa. Translation tentative.

n.58The following passages restate the structure used above for desirous minds and minds free of desire, applying it only to the last binary in the list, “minds that are liberated and minds that are not liberated.” This same formula is meant to be applied to the intervening binaries as well.

n.59This translation follows F, H, N, and S in reading sems thams cad shes kyang. D reads sems can thams cad shes kyang, “Even though they know all beings…”

n.60Tib. sems la sems das snyems pa yang med rlom sems kyang med. Translation tentative.

n.61This translation follows C, F, H, J, K, Y, N, and S in reading brtan pa. D reads bstan pa, “teach/show.”

n.62Tib. rdzu ’phrul bya shes pa, which is understood here as a translation of the Skt. ṛddhividhi­jñāna.

n.63The Tib. term shes pa po is understood from the context to be short for ’du shes pa po.

n.64These are also described in detail in The Transcendent Perfection of Wisdom in Ten Thousand Lines (Toh 11, 1.­55).

n.65Linum usitatissimum. Tib. zer ma’i me tog; Skt. umakapuṣpa. The identification of this flower and other plants in this translation is informed by Singh and Chunekar 1999 and the Pandanus Database of Indic Plants .

n.66Cassia fistula. Tib. dong ka’i me tog; Skt. karṇikārapuṣpa.

n.67Pentapetes phoenicea. Tib. ban du dzi ba ka’i me tog; Skt. bandhu­jivakapuṣpa. This flower is more typically named bandhuka in Skt.

n.68Unidentified. This translation is based on the attested Skt. uśanastāraka (Tib. skar ma pa ba sangs). This is possibly the white flower of the pipal tree.

n.69As used in this text, the term “jewel-in-hand” (lag na rin po che) appears to convey a meaning similar to “sky treasury” (nam mkha’i mdzod), referring to a bodhisattva’s ability to manifest any desired object at will.

n.70C, F, and S read gzhon pa, “youth.”

n.71A similar set of ten truths, with slight variation, is presented in The Ten Bhūmis (Toh 44-31, 1.­384–1.­385.

n.72That is, the first of the four truths of the noble ones. Each of the characteristics that follow are similarly identified with one of the four truths.

n.73Tib. mtshan nyid las gyur pa. This is understood here as equivalent to the Skt. lakṣaṇika.

n.74S uniquely reads “the truth of the characteristic of classification” (rnam par dbye ba’i mthsan nyid kyi bden pa).

n.75C, F, J, K, Y, and N omit “emptiness” (stong pa nyid).

n.76This translation of the description of the first dhyāna follows the Tibetan syntax, but is informed by numerous similar, but not identical, Sanskrit citations, such as this from the Pañca­viṃśatisāhasrikā Prajñā­pāramitā: viviktaṃ kāmair viviktaṃ pāpakair akuśalair dharmaiḥ savitarkaṃ savicāraṃ vivekajaṃ prītisukhaṃ prathamaṃ dhyānam upasaṃpadya viharati (Kimura 1992, vol. 5, p. 93).

n.77All versions of the Tibetan translation consulted read kun tu longs spyod pa, though this term seems out of place in this context.

n.78This translation of the description of the second dhyāna follows the Tibetan syntax, but is informed by numerous similar, but not identical, Sanskrit citations, such as this from the Pañca­viṃśatisāhasrikā Prajñā­pāramitā: vitarkavicārāṇāṃ vyupaśamād adhyātmaṃ saṃprasādāc cetasa ekotībhāvād avitarkam avicāraṃ samādhijaṃ prītisukhaṃ dvitīyaṃ dhyānam upasaṃpadya viharati (Kimura 1992, vol. 5, p. 93).

n.79This translation of the description of the third dhyāna follows the Tibetan syntax, but is informed by numerous similar, but not identical, Sanskrit citations, such as this from the Pañca­viṃśatisāhasrikā Prajñā­pāramitā: prīteś ca virāgād upekṣako viharati, smṛtimān saṃprajānaṃ sukhaṃ ca kāye na pratisaṃvedayati yat tad āryā ācakṣate, upekṣakaḥ smṛtimān sukhavihārī tṛtīyaṃ dhyānam upasaṃpadya viharati (Kimura 1992, vol. 5, p. 93).

n.80The precise point where the content of the bodhisattva’s thought begins is not clear from the Tibetan syntax.

n.81This translation of the description of the fourth dhyāna follows the Tibetan syntax, but is informed by numerous similar, but not identical, Sanskrit citations, such as this from the Pañca­viṃśatisāhasrikā Prajñā­pāramitā: sa sukhasya ca prahāṇāt duḥkhasya ca prahāṇāt pūrvam eva saumanasyadaurmanasyayor astaṃgamād aduṣkhāsukham upekṣāsmṛtipariśuddhañ caturthaṃ dhyānam upasaṃpadya viharati (Kimura 1992, vol. 5, p. 93–94).

n.82This translation follows C, F, K, Y, and S in reading mngon par smon. D reads mngon par sman.

n.83This translation follows D and other versions in reading sems can thams cad kyir, but it is noteworthy that C, H, J, K, and N read sems can thams cad kyis, while F has sems can thams cad kyi.

n.84Conjectural for mang ba, the sense of which is unclear. Of the sources consulted, only F omits mang ba.

n.85Cf. Pañca­viṃśatisāhasrikā Prajñā­pāramitā: sa sarvaśo rūpasaṃ­jñānāṃ samati­kramāt pratighasaṃ­jñānām astaṃgamān nānātvasaṃ­jñānām amanasikārād anantam ākāśam ity ākāśānantyāyatanam upasaṃpadya viharat[i] (Kimura 1992, vol. 5, p. 100).

n.86This translation follows C, F, J, K, and Y in omitting don de’i, which is attested in D. S similarly reads de’i don.

n.87Cf. Pañca­viṃśatisāhasrikā Prajñāpāramitā: sa sarvaśa ākiṃcanyāyatana­samatikramād naivasaṃ­jñānāsaṃ­jñāyatanam upasaṃpadya viharati (Kimura 1992, vol. 5, p. 101).

n.88Tib. mu stegs can. This term refers to any non-Buddhist tradition that holds views in opposition to those held by Buddhists.

n.89Tib. dmigs pa can, Skt. aupalambhika. This term refers to traditions that invest perception and objects of perception with reality. Such a view could be held by Buddhists and non-Buddhists.

n.90Perhaps more literally “they would not form any latent impressions about them.”

n.91This translation follows C, F, H, J, K, Y, N and S in reading thos par byed. D reads thogs par byed, “create obstacles.”

n.92Magnolia champaca. Tib. tsam pa ka; Skt. campaka.

n.93Jasimum grandiflorum. Tib. sna ma; Skt. jātī.

n.94Śāntideva cites this statement listing the three types of patience in the Śikṣāsamuccaya, though in his sequence the second and third types of patience are reversed. See Bendall 1902, p. 179, and Goodman 2016, p. 179. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.95Tib. sbyor ba’i gnas. Yoga (sbyor ba) is understood here in its more generic sense of “mastery” of any activity, secular or spiritual, to which one dedicates oneself.

n.96This translation follows F, H, and S in reading shes la. D reads shes pa.

n.97Tib. stong pa. The attested Skt. śūnyatā has been followed in the translation.

n.98Tib. kyang. Skt. eva.

n.99The preceding portion of Nirārambha’s discussion with Mativikrama on suchness is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 263–64, and Goodman 2016, p. 250. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.100This translation follows S in reading sangs rgyas rnam rtogs pa. D has sangs rgyas rnams rtogs pa.

n.101Translation speculative.

n.102The preceding two verses are cited by Kamalaśīla in the extant Sanskrit witnesses of the Bhāvanākrama. See Tucci 1958, p. 219, and 1971, p. 29. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.103This statement is reminiscent of the Buddha’s discourses that a person is not considered brahmin based on their birth, but because of their virtuous deeds and qualities. See, for example, the Soṇadaṇḍa Sutta of the Dīghanikāya (Walshe 1995, pp. 125–32) and the Vāseṭṭha Sutta of the Suttanipāta (Bodhi 2017, pp. 260–69).

n.104“Bodhisattvas” (Tib. byang chub sems dpa’ rnams kyi) is not attested in the Sanskrit witness of this citation from Śāntideva’s Śikṣāsamuccaya. It reads, “A discourse on the truth, child of good family, is a discourse on the Dharma (satya­saṅgītiḥ kula­putra dharma­saṅgītiḥ).

n.105The preceding passage is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 12, and Goodman 2016, p. 14. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.106It seems clear from the citation marker zhes gsungs so that another quote is being cited in the following passage, but it is not clear from the syntax where precisely it begins.

n.107Tib. so sor nges pa. This is understood here as equivalent to the Skt. pratiniyama.

n.108The Tibetan text does not indicate that Nirārambha’s direct speech ends here, but the content and context of the following lines suggest that it does.

n.109The Śīkṣāsamuccaya preserves a slightly different syntax here: nādyantamadhye pratiṣṭhitamūlaḥ, “its root is not fixed in a beginning, end, or middle.”

n.110The Śikṣāsamuccaya includes the term śarīra here, which appears not to be attested in the Tibetan. While śarīra could be a translation of the Tib. term ro, it is more likely a translation of kuṇapa, which is attested in the Śikṣāsamuccaya.

n.111The content of the bodhisattva’s reflection described in this passage is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 229, and Goodman 2016, pp. 222–23. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.112The translation of anatomical terms in the following passage is based on the use of the Skt. terms in the classical Āyurvedic treatises as informed by Meulenbeld 1999.

n.113Tib. gru do. The Skt. attested in the Śikṣāsamuccaya is trika.

n.114Tib. lag pa’i thal mo. The Śikṣāsamuccaya reads hasta, “hand.”

n.115Tib. pho ba. The term is omitted in the Śikṣāsamuccaya.

n.116The preceding passage, beginning with the paragraph on the description of the parts of the body, is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, pp. 228–29, and Goodman 2016, p. 222. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.117This translation follows F, K, Y, and S in reading rig pa. D reads reg pa; C, J, and N read rigs pa.

n.118This translation follows C, F, J, K, Y, and N in reading tshor ba yin na. D reads tshor ba ma yin na, “if it did not involve sensation.”

n.119This translation follows C, H, J, K, Y, N, and S in reading sems ma gtogs par. D reads sems ma rtogs par, “without realizing the mind.”

n.120“Refuge” (Tib. skyabs su gyur pa) is absent in the Sanskrit attested in the Śikṣāsamuccaya.

n.121Śāntideva elides the lines from “…have surpassed the obscuration of the afflictive emotions” until “they have the exalted intentions that come from great compassion.”

n.122It is unclear what “the three paths” (Tib. lam gsum) refers to here.

n.123An abridged version of the preceding passage, beginning with “furthermore,” is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 322, and Goodman 2016, pp. 299–300. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.124Tib. ste’u dang tsan dan du mtshungs pa. This is understood here as equivalent to the Skt. phrase vāsīcandana­kalpa, an idiomatic expression that is difficult to capture in English. It is understood to refer to an attitude of indifference exemplified through the juxtaposition of extreme opposites of experience.

n.125The Skt. attested in the Śikṣāsamuccaya reads sarva­lokānabhi­bhūtāḥ, “they are not surpassed by anything in the world.”

n.126It is unclear what the “great yoga” (mahāyoga; rnal ’byor chen po) is in this context.

n.127“Domains of great meditation who possess the wonder of great liberation” is not attested in the Śikṣāsamuccaya.

n.128Śāntideva elides the lines from “and are like the Great Brahmā” to “provide relief through the great Dharma.”

n.129An abridged version of the preceding passage, beginning with “furthermore,” is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 322, and Goodman 2016, p. 300. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.130This final line is translated based on the attested Sanskrit syntax: evaṃ ca tān anusmṛtya tadguṇa­pari­niṣpattyarthaṃ smṛtim upasthāpayati.

n.131Tib. chos kyi bgo skal la spyod pa. This is understood to be equivalent to the Skt. term dharmadāyāda.

n.132This translation follows F and S in reading chos ni, which also follows the structural pattern used in the rest of this passage. D reads gnas ni.

n.133The term “field” (Tib. zhing) can perhaps be understood here as “buddhafield” (sangs rgyas kyi zhing), but could also refer more generically to anyone or anything that is a fitting focal point of an action, such as generosity and so forth.

n.134The preceding passage, beginning with “resting in meditative composure,” is cited with some variation by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 119, and Goodman 2016, pp. 117–18. The terminology and syntax of the English translation are informed by the Sanskrit.

n.135This translation follows C, H, J, K, Y, and N in reading kun spyod. D and S read kun dpyod.

n.136This follows the attested Sanskrit term vicakṣaṇaiḥ, which has been translated into Tibetan with mkhas rnams kyis.

n.137The preceding two verses are cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, pp. 233–34, and Goodman 2016, p. 226. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.138This translation follows C, H, J, K, Y, and N in reading lus la sems. D and S read lus la sems, “mind apart from the body.”

n.139Translation tentative for zhing dang skye la rab gnas pa. S uniquely reads zhi dang skye la rab gnas pa.

n.140Tib. bdag bcag gnyis kyis chos yang dag par rjod pa dang ’thun pa bgyis pa lags bsam. The “Dharma discourse” (dharmasaṅgīti) referred to here is understood to be the teaching by the Buddha that opens this sūtra and which prompted the dialogue between Nirārambha and Mativikrama. This statement could be read, by extension, to refer to the Dharma­saṅgīti­sūtra itself: “Is what we have said here in accord with The Dharma Council?”

n.141In other words, “why are you called nirārambha ?”

n.142This translation follows K and S in reading med pa’i phyir. D reads med par gyur na.

n.143A type of false image caused by ocular distortions, typically as the result of cataracts or other eye disorders.

n.144This translation follows D and S. C, H, J, K, Y, and N read sems can thams cad gyi dam pa, “are sublime among all beings.” This reading would result in sgrib pa med pa’i chos can being a separate adjectival phrase.

n.145Tib. zug rngu ’byin pa. This is understood as equivalent to the Skt. term śalyahartṛ, a word used for surgeons in classical Indian medical literature.

n.146This line could also be interpreted to say “mastery over the Dharma.”

n.147Reading khur rtsa ba (D) as khur tsa ba (S).

n.148This line is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 143, and Goodman 2016, p. 138.

n.149This translation follows D and S in reading drang ba. C, J, K, Y, and N read drang srong (ṛṣi), “sage.”

n.150Tentative for rgyun sman du gyur pa, which follows D and the majority of texts cited in the Comparative Edition. S uniquely reads rgyun rmad du gyur pa, which is more syntactically plausible and means “in a constant state of astonishment.” It is worth noting that the modern concept of “addiction” would be anachronistic to this text.

n.151Tentative for g.ye’i phyogs. S reads dbye’i phyogs.

n.152The phrase “not nonexistent” is not attested in the Sanskrit citation from the Śikṣāsamuccaya.

n.153Tib. shin tu bzung ba. The Sanskrit reads svārādhitam.

n.154This phrase is not attested in the Sanskrit citation from the Śikṣāsamuccaya.

n.155The Tibetan shin tu mnyam par brtsam pa appears to be a misinterpretation of the attested Sanskrit term susamārabdha. The Sanskrit term is most naturally parsed as su+sam+ā+√rabh, whereas the Tibetan translators seem to have read it as su+sama+ā+√rabh.

n.156The preceding paragraph is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 122, and Goodman 2016, p. 120. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.157This translation follows C, J, K, and Y, as well as the Sanskrit cited by Śāntideva in reading sems can (sattva). D and S read sems, “mind.”

n.158The passage beginning with “A field of beings…” is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 153, and Goodman 2016, pp. 148–49. The terminology and syntax of the English translation are informed by the Sanskrit.

n.159Tib. bdud rtsi; Skt. amṛta. Here the term amṛta is interpreted literally, rather than as “nectar,” “ambrosia,” etc.

n.160This translation follows C, F, J, K, Y, and N in reading yid gnas pa ma mchis pa rnams. D and S read yid dang gnas pa ma mchis pa rnams, “those who are without mind and location.’

n.161Tib. bsod nams dang bsod nams ma lags pa dang mi g.yo ba’i ’du byed. This is understood to be equivalent to the Sanskrit puṇyā­puṇyāniñjya­saṃskāra.

n.162This translation follows D in reading bdag la sman pa dang gzhan la sman pa. F and S read bdag la phan pa dang gzhan la phan pa, “[unable to] benefit themselves or benefit others.” H and N read bdag la phan pa dang gzhan la sman pa, “[unable to] benefit themselves or heal others.”

n.163As above, F and S read phan pa, “benefit,” where D reads sman pa, “heal.”

n.164The Sanskrit cited by Śāntideva reads advitīyasya bhagavan dharmasya raṇam nāsti, “Blessed One, a nondual phenomenon does not have a sound.”

n.165The Sanskrit cited by Śāntideva reads naiṣāṃ kaścit, “nothing belongs to them.”

n.166The Sanskrit cited by Śāntideva does not include an equivalent of “nor attachment” (chags par yang mi bgyid de).

n.167This translation follows F and S and the Sanskrit cited by Śāntideva in reading dge sbyong (śramaṇa). D reads dge slong (“monk”) here.

n.168This translation follows C, F, H, K, Y, N, and S in reading yang dag par mthong ba. D reads yang dag pa mthong ba.

n.169An abridged and slightly divergent version of the preceding passage, beginning with “the eyes do not emit sound to forms…,” is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 263, and Goodman 2016, p. 251. The terminology and syntax of the English translation of this passage are informed by the Sanskrit.

n.170A pun is used here that is lost in translation. The Sanskrit term for “syllable” is akṣara, which also means “imperishable.”

n.171This translation follows D in reading mngon par dad pa. F, H, N, and S read mngon par dga’, “take delight in.”

n.172This line is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 122, and Goodman 2016, p. 122.

n.173This passage follows the set of the twelve links of dependent arising.

n.174This translation follows C, F, H, J, K, Y, N, and S in reading rnam par rig pa. D reads rnam par reg pa.

n.175The phrase “based in great compassion” (snying rje chen po sngon du ’gro) is not attested in the Sanskrit citation from the Śikṣāsamuccaya.

n.176This translation follows C, F, H, K, Y, S, U and the Sanskrit cited by Śāntideva in omitting the repetition of chen po attested in D.

n.177An abridged citation of the passage that begins, “Blessed One, it is any appropriate action of a bodhisattva’s body…,” and ends here is cited by Śāntideva with some variation in the Śikṣāsamuccaya. See Bendall 1902, p. 117, and Goodman 2016, p. 115. An unabridged version of the same passage is cited by Kamalaśīla in the Bhāvanākrama, for which see Tucci 1958, p. 222. The terminology and syntax of the English translation are informed by the Sanskrit.

n.178This translation follows F, K, Y, and S in reading mang ba. D reads med pa.

n.179This translation follows F, H, N, and S in reading bdag gi sems rnam par grol ba. D reads bdag gi sems can rnam par grol ba.

n.180This translation follows the reading mi mchi ba in H and N, and mi ’gro ba in F and S. D reads mi ’chi ba, “not die.”

n.181This translation follows F, K, Y, and S in reading mi skye mi rga. D reads mi skye mi dga’o, “not born and not joyful.”

n.182D: legs pa’i blo gro. F, N, and S report the alternate translation dge ba’i blo gros. This very likely translates the same Sanskrit term, sādhumatī.

n.183Priyadarśana’s statement to this point is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 124, and Goodman 2016, p. 122. The terminology and syntax of the English translation are informed by the Sanskrit.

n.184This translation follows F and S in reading don spyod pa. D reads don sbyor ba.

n.185This translation follows F, H, K, Y, N, and S in reading rnam par mi rtog pa. D reads nam par rtog pa, “conceptual.”

n.186This translation follows F, H, K, Y, N, and S in reading mnyen pa. D reads mnyan pa, “listening.”

n.187This line is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 145, and Goodman 2016, p. 141. It appears that the version of the text that Śāntideva read included the line “Blessed One, this is a discourse on the Dharma,” following the line noted here. Or, perhaps more likely, Śāntideva appended the final line of this discourse to his citation of the first line.

n.188This translation follows F, H, K, Y, N and S. D reads gang ’khor ba nyid mya ngan ’das par yang dag par rjes su mthong la yang dag par sgrib pa’i brtson ’grus kyang mi dor ba ’di ni byams pa chen po’o, “Their great love is to recognize saṃsāra itself to be nirvāṇa while not giving up obscured diligence.”

n.189D: lus kyi ring bsrel. This is understood to be equivalent to the Sanskrit compound śarīradhātu, which in this passage line seems to be used to refer both to the main physical constituents of the body as left over after cremation and to the identification of those remains as spiritual relics. The term has been translated slightly differently in each clause to capture this dual valence.

n.190The term lus is understood here as a translation of vāpa, as it is in the passage on bodily constituents at, which has an attested equivalent in the Śikṣāsamuccaya.

n.191There is a pun on the term dharma/chos that comes through clearly in the Sanskrit and the Tibetan that translates it, but is lost in English translation. This sentence uses two valences of the term dharma, “phenomena” and “the Dharma.” To capture this pun, the sentence could be rendered, “Bodhisattvas also attain accomplished dhāraṇī, through which they directly reveal all dharmas, revealing the entire expanse of dharmas just as they would a single word of the Dharma.”

n.192This translation relies on the Sanskrit of the Śikṣāsamuccaya to identify the Tibetan term gdugs as equivalent to āditya, “sun,” rather than its other common equivalent chattra, “parasol.”

n.193The version of this passage cited by Śāntideva in the Śikṣāsamuccaya includes a line not attested in the Tibetan translations consulted. It reads, “When all of the faculties are governed by the mind, there is a large increase in the perception of their respective objects. Likewise, Blessed One, when governed by great compassion, there is a large increase in the respective functioning of the other factors that bring about awakening” (Bendall 1902, p. 287, and MS Add. 1478, f.126.a: tad yathā bhagavan sarveṣām indriyāṇāṃ manasādhi­ṣṭhitānāṃ svasvaviṣaye grahaṇ­aprācuryaṃ bhavati | evam eva bhagavan mahākaruṇādhiṣṭhitānām anyeṣāṃ bodhikarāṇāṃ dharmāṇāṃ svasmin svasmin karaṇīye prācuryaṃ bhavati).

n.194This entire passage is cited by Śāntideva, with some variation, in the Śikṣāsamuccaya. See Bendall 1902, pp. 286–87, and Goodman 2016, p. 271. The terminology and syntax of the English translation are informed by the Sanskrit.

n.195Śāntideva’s citation of this line reads buddha­dharmāṇāṃ mūlam, “the root of the Buddhadharma.” D reads chos rnams kyi rtsa ba, and S reads chos thams cad kyi rtsa ba.

n.196Śāntideva’s citation of this line reads dharmaśabda, “the sound of the Dharma.”

n.197The phrase “but they do not where there is no intention” is not present in the Sanskrit citation from the Śikṣāsamuccaya.

n.198Śāntideva’s citation in the Śikṣāsamuccaya includes svadhiṣṭhita, “maintain it well.”

n.199The preceding passage is cited by Śāntideva, with some variation, in the Śikṣāsamuccaya. See Bendall 1902, pp. 284–85, and Goodman 2016, pp. 269–70. The terminology and syntax of the English translation are informed by the Sanskrit.

n.200D: tshul bzang. A Sanskrit equivalent could not be verified. H, N, and S read dpal bzang.

n.201Tib. ’khor go ba ma lags. This translation is informed by the Sanskrit cited by Śāntideva in the Śikṣāsamuccaya, which reads na paurī. The term paura/paurī can be used to refer to the ordinary citizens of a realm and the language they speak, rather than the sophisticated language used by members of the court, intellectuals, literati, and so forth.

n.202This preceding passage is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 127, and Goodman 2016, p. 124. The terminology and syntax of the English translation are informed by the Sanskrit.

n.203Tib. nam mkha’ mdzod; Skt. gaganagañja . The bodhisattva Gaganagañja here incorporates his own name into his discourse.

n.204This translation follows F, Y, and S in reading this as a distinct line with its own final punctuation.

n.205Tib. tshangs pa’i yul. This term is not found in any lexicons consulted, and so is taken here as equivalent to tshangs pa’i yul sa, an attested translation of brahmāṇḍa, the “Egg of Brahmā,” a brahmanical term for the universe.

n.206Tib. nam mkha’ rnyeb pas rnyeb pa. This translation is tentative. The term rnyeb is here understood as a variant form of bsnyabs/snyob/bsnyab, “reach for” or “to stretch out toward.” C and F read rnyed pas rnyed, but this variant seems less plausible. The full phrase appears to be idiomatic, and thus is difficult to translate literally.

n.207This translation follows F, Y, and S in reading bla’i, which is understood to form the construction bla’i…de lta ma yin pa “It would not be better….” D reads sla’i.

n.208This translation follows C, F, H, J, K, Y, and N in reading rab tu myur bar rnam par thar bar ’gyur. D reads rab tu myur ba de rnam par thar bar ’gyur.

n.209This translation follows F and S in reading tha na sangs rgyas kyi byang chub ’dod par ’gyur yang med pa, as well as the essentially synonymous cher na sangs rgyas kyi byang chub ’dod par ’gyur yang med pa attested in C, J, K, and Y. D reads chad na sangs rgyas kyi byang chub ’dod par ’gyur yang med pa, the meaning of which is ambiguous.

n.210Tib. mu tig gi do shal rgya stongs ri ba. It is unclear what unit “worth a hundred thousand” refers to.

n.211D: phan ran. This translation is tentative, and understands ran pa to be equivalent to madhya in the Buddhist Hybrid Sanskrit sense of “neutral,” “impartial,” etc. There is significant variation across the Tibetan versions consulted, none of which provide a clear reading.

n.212This translation follows C, N, F, and S in reading mi khyer. D reads me khyer.

n.213This translation follows C, F, H, J, K, Y, N, and S in reading khyod ni gcig bu. D reads khyod mi gcig bu.

n.214Tib. byang chub sems dpa’ de dag. It is unclear from the context who “those” bodhisattvas are, but presumably they are the bodhisattvas that follow the Buddha Great Light.

n.215This preceding passage, beginning with “How do you know that one is maintaining correct behavior?” is cited by Śāntideva in the Śikṣāsamuccaya. See Bendall 1902, p. 146, and Goodman 2016, pp. 141–42. The terminology and syntax of the English translation are informed by the Sanskrit.

n.216There is a play on the term pāramitā, “perfection,” in this line. Creative etymologies of pāramitā often explain it to mean “gone (ita) to the far (pāram) shore,” hence “perfection.” This line engages this interpretation by rejecting the binary of “far” (para) and “near” (avāra).

n.217This translation follows C, F, H, K, Y, N, and S in reading dad pa mtshung pa. D reads dang ba mtshungs pa.

n.218Reading ’ged pa following H and N for bged pa in D. F reads dge ba and S reads ’gyed pa.

n.219Tib. dran pa rab tu byin pa. The precise meaning of this statement is unclear.

n.220Tib. mdoms kyi sba ba sbubs su nub pa; Skt. kośopaga­tavastiguhya. This is one of the major marks of a buddha.

n.221Tentative for brtson ’grus brtsams pa mngon par shes par ’gyur ba.

n.222This translation follows C, F, H, K, Y, N, and S in reading sbyor ba. D reads spyod pa.

n.223F, H, N, and S read shes pa where D reads ’tshal ba. The meaning is essentially the same.

n.224In the following section, each bodhisattva makes a statement about understanding the Dharma that reflects the meaning of their name or a well-known quality of their character. Here, for example, Nirārambha (rtsom pa med) declares that those who are engaged (ā√rabh; rtsom pa) are not capable of understanding the Dharma, but he is because he is not engaged (nir+ā√rabh; rtsom pa ma mchis pa).