The Vajrasattva Mantra: Sanskrit Edition and English Translation

The Vajrasattva Mantra: Sanskrit Edition and English Translation by David Reigle, April 2024 The Vajrasattva mantra is very widely used, so it comes as no surprise that many variants are found in its Sanskrit wording and in how it is understood. Despite its importance, I am not aware of any detailed investigation of its wording and its meaning.1 In view of this need, I have made an attempt to do so here. 2 The results of my investigation are given first, since these are perhaps all that some readers will require. oṃ vajrasattva, samayam anupālaya, vajrasattvatvenôpatiṣṭha, dṛḍho me bhava, sutoṣyo me bhava, supoṣyo me bhava, anurakto me bhava, sarva-siddhim me prayaccha, sarva-karmasu ca me cittaṃ śreyaḥ kuru hūṃ, ha ha ha ha hoḥ, bhagavan sarva-tathāgata-vajra, mā me muñca, vajrībhava mahā-samaya-sattva, āḥ | Oṃ. Vajrasattva, keep [your] pledge. As Vajrasattva, stand near [me]. Be steadfast toward me. Be well pleased with me. Be well fulfilled on account of me. Be attached to me. Grant me all accomplishments; and in all actions make my thought beneficent, hūṃ. Ha ha ha ha hoḥ. O Blessed One, diamond-nature of all the tathāgatas, do not abandon me. Be of diamond-nature, O great pledge being. Āḥ. This mantra, like all mantras, is in Sanskrit. It was retained in Sanskrit, and only transliterated into Tibetan characters but not translated into the Tibetan language, in the many Tibetan Buddhist texts where it is found. It is usually from these Tibetan texts that it has been taken when it appears in English language books. When transliterated into one script, and then from that into another, there is much room for error. In fact, this is what we regularly find. By contrast, when Benoytosh Bhattacharyya edited the Sādhanamālā from original Sanskrit manuscripts, he was able to establish an almost perfect text of this mantra, which occurs there eight times.3 Among the errors in this mantra as occurring in English language books where it is derived from Tibetan sources, the most common is taking the second occurrence of Vajrasattva as a vocative; that is, as addressing Vajrasattva twice. The majority of the published translations of this mantra that I know of, made from Tibetan sources, wrongly put a space after vajrasattva before -tvena, thus making it a vocative. However, vajrasattvatvena is a single word, not the vocative vajrasattva and a hypothetical word tvena.4 In -tvena we have the perfectly regular Sanskrit suffix -tva, making vajrasattvatva, and then the whole word is declined in the instrumental case, vajrasattvatvena. It will be useful to now look at the word by word meanings for this mantra, and to provide some grammatical analysis. The Vajrasattva Mantra: Sanskrit Edition and English Translation 2 Word by Word Meanings and Grammatical Analysis oṃ, (mantra syllable), untranslatable. A sacred syllable that begins most mantras. vajra-sattva, (noun, masculine vocative singular), O Vajrasattva; O Diamond-like Being (in the sense of having a diamond-like or indestructible nature). This is a karmadhāraya or adjective-noun compound, consisting of the noun vajra, thunderbolt, diamond, here used as an adjective meaning diamond-like, having a diamond-like nature, in the sense of the hardness of a diamond, therefore indestructible, impenetrable, indivisible, referring to the ultimate nature of something, and the noun sattva, being, living being, sentient being. samayam, (noun, masculine accusative singular), pledge, vow, commitment, agreement, oath. anupālaya, (verb, imperative, second person singular, anu + √pāl, “to protect”), [you] protect, preserve, guard, watch over, keep, keep watch over (the prefix anu, meaning “after,” with the root pāl, meaning “protect,” yields the sense of “keep protecting”). vajrasattvatvenôpatiṣṭha = vajrasattvatvena + upatiṣṭha, (when two vowels come together, here the final “a” of vajrasattvatvena and the initial “u” of upatiṣṭha, they are required to merge into a single vowel, here “o”, in accordance with the rules of Sanskrit sandhi.) vajrasattvatvena, (noun, neuter instrumental singular), as Vajrasattva; literally, by being Vajrasattva. This word is vajrasattva with the suffix -tva. The suffix -tva means the state or condition of being something, here Vajrasattva. When decined in the instrumental case, -ena, it means “as” something, here, as Vajrasattva. A good example of this usage is found in the Niṣpanna-yogāvalī near the beginning of the description of the Kālacakra body maṇḍala with the word nāyaka, “leader,” in the form of nāyakatvena : karṇikāyāṃ tu pūrṇimā 'māvāsī ca nāyakatvena sthitā. There, surrounded by twenty-eight lunar day deities on the petals of the lotus, the full moon and new moon stand on the central receptacle of the lotus, nāyakatvena, “as the head of the group” (translated by Lokesh Chandra and Nirmala Sharma, Niṣpanna-yogāvalī, p. 291). upatiṣṭha, (verb, imperative, second person singular, upa + √sthā, “to stand”), [you] stand near, stand by (the side of), remain near, stay with, dwell, reside (the prefix upa, meaning near or by, with the root sthā, meaning stand or remain, yields the sense of stand near, remain by).5 dṛḍho → dṛḍhaḥ, (adjective, masculine nominative singular, past passive participle, √dṛh/dṛṃh, “to be fixed or firm”), steadfast, firm, steady. me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. bhava, (verb, imperative, second person singular, √bhū, “to be”), [you] be. sutoṣyo → sutoṣyaḥ, (adjective, masculine nominative singular, future passive participle, su + √tuṣ, “to be pleased or satisfied”), well pleased, well satisfied, content, happy (the prefix su, well, with the root tuṣ, to be pleased or satisfied, yields the sense of well pleased, well satisfied). me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. bhava, (verb, imperative, second person singular, √bhū, “to be”), [you] be. The Vajrasattva Mantra: Sanskrit Edition and English Translation 3 supoṣyo → supoṣyaḥ, (adjective, masculine nominative singular, future passive participle, su + √puṣ, “to nourish or develop”), well nourished, well nurtured, increased, made to thrive, (by extension →) well fulfilled (the prefix su, well, with the root puṣ, to nourish, yields the sense of well nourished). me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. bhava, (verb, imperative, second person singular, √bhū, “to be”), [you] be. anurakto → anuraktaḥ, (adjective, masculine nominative singular, past passive participle, anu + √rañj, “to be colored, to be attracted by, to be attached to”), attached to, devoted to, fond of, affectionate toward. me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. bhava, (verb, imperative, second person singular, √bhū, “to be”), [you] be. sarva-siddhim, (noun, feminine accusative singular), all accomplishments. This is a karmadhāraya or adjective-noun compound, consisting of the adjective sarva, all, and the feminine noun siddhi, accomplishment, achievement, attainment, perfection, feat, supernormal power. Note that when a compound begins with a word such as “all,” even though the compound is declined in the grammatical singular, it is appropriate to translate it according to its plural sense. me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. prayaccha, (verb, imperative, second person singular, pra + √yam, “to give”), [you] grant, bestow.6 sarva-karmasu, (noun, neuter locative plural), in all actions. This is a karmadhāraya or adjective-noun compound, consisting of the adjective sarva, all, and the neuter noun karma, action.7 ca, (conjunction), and. In Sanskrit, unlike in English, the “and” normally follows rather than precedes what it adds. me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. cittam, (noun, neuter accusative singular), mind, thought, or metaphorically, heart. The usual meaning is “mind,” but sometimes it clearly means “thought” or “thoughts.” For example, Bhagavad-gītā, chap. 16, verse 16, aneka-citta-vibhrāntāḥ, “bewildered by many thoughts.” śreyaḥ, (adjective, neuter accusative singular), better, more of something good such as more fortunate, or simply fortunate, auspicious, propitious, with the context here indicating virtuous, beneficent. The great majority of old Sanskrit sources have śreyaḥ, while the great majority of modern Tibetan sources have the variant: śriyaṃ, (feminine accusative singular of the word śrī, which is normally a noun, and less often an adjective as it would have to be here), radiant, glorious, splendid, with the context here indicating virtuous. Note that there would be a gender disagreement with an adjective declined in the feminine, śriyam, describing a neuter noun, cittam. kuru, (verb, imperative, second person singular, √kṛ, “to make”), [you] make. hūṃ, (mantra syllable), untranslatable. This syllable is often found at the end of mantras. ha ha ha ha hoḥ, (mantra syllables), untranslatable. In a widespread Tibetan tradition they symbolize the wisdoms (jñāna) of the tathāgatas. This idea goes back to a text from the early translation period in Tibet, preserved in the bKa' ma collection of rNying ma teachings, so it likely goes back to an Indian The Vajrasattva Mantra: Sanskrit Edition and English Translation 4 source. The same text was included in works by dGe lugs pa writers, and the explanation of these syllables as the wisdoms was widely promulgated by dGe lugs pa writers.8 bhagavan, (noun, masculine vocative singular), O Blessed One. This word consists of bhaga and the possessive suffix vat. So it means possessing bhaga. In the traditional meaning, the word bhaga refers to six good qualities, including lordliness (aiśvārya), glory (yaśas), splendor (śrī ), wisdom (jñāna), and two others that differ in the many various sources, both Hindu and Buddhist. One of these is either strength (vīrya), virtue (dharma), or beauty (rūpa), and the other one is either dispassion (vairāgya) or effort (prayatna).9 One who possesses these qualities is a blessed one. This is how it is most widely understood and translated. However, although the compilers of the Madhya-vyutpatti cited this traditional meaning, they instead chose to follow an etymology based on bhagna, “conquer,” rather than bhaga.10 Hence they translated bhagavat into Tibetan as bcom ldan 'das, and this became the standard Tibetan translation of bhagavat. The first two syllables mean possessing victory, so “victor.” The last syllable adds the idea of transcendence. Thus, Jeffrey Hopkins translates bcom ldan 'das as Supramundane Victor. sarva-tathāgata-vajra, (noun, masculine vocative singular), O diamond-nature of all the tathāgatas. This is a double compound, i.e., a compound within a compound. The first compound is a karmadhāraya or adjective-noun compound, consisting of the adjective sarva, all, and the noun tathāgata, literally meaning thus (tathā) gone (gata, past passive participle from √gam, “to go”), sometimes instead understood as thus (tathā) come (āgata). Tathāgata is a commonly used descriptive term for a buddha. The compound sarva-tathāgata means all the tathāgatas. This compound taken as a single unit then forms the first member of a second compound, sarvatathāgata-vajra. This second compound is a tatpuruṣa or case-relation compound. The first member of these compounds is understood to be in a particular case, here the genitive case. So sarvatathāgata is understood to mean “of” all the tathāgatas. The final member of this second compound is the noun, vajra, diamond, having the sense of diamond-nature. So the whole double compound means diamond-nature of all the tathāgatas. Since it is declined in the vocative case, it means O diamond-nature of all the tathāgatas. mā, (particle), makes the following verb negative, whether or not there is an intervening word or words. me, (pronoun, first person, genitive or dative, singular), of me, for me, to me. muñca, (verb, imperative, second person singular, √muc, “to set free, leave”), [you] abandon, leave, release, forsake, desert. mā me muñca, (imperative verb phrase with negative mā and intervening pronoun me), [you] do not abandon me. vajrī-bhava, (verb, imperative, second person singular), [you] be or become a vajra or like a vajra, a diamond, diamond-like, of diamond-nature, i.e., indestructible, impenetrable, indivisible, the ultimate nature of something. In this grammatical construction the noun vajra takes the “cvi” secondary affix (taddhita pratyaya), which means that its final vowel is replaced by long “ī”, making vajrī. This word is then joined with the imperative verb bhava, [you] be or become. It is possible to take this as two 5 The Vajrasattva Mantra: Sanskrit Edition and English Translation separate words: the possessive noun vajrin, possessing a vajra, declined in the masculine nominative singular, vajrī, and the imperative verb bhava, [you] be or become. In the Sarva-tathāgata-tattva- saṃgraha, where the Vajrasattva mantra first occurs, it is immediately preceded by three occurrences of the cvi construction, and is shortly followed by another occurrence of the cvi construction. This makes it likely that the verb in the cvi construction vajrī-bhava is intended here, rather than the separate noun and verb, vajrī bhava.11 mahā-samaya-sattva, (noun, masculine vocative singular), O great pledge being. This is a double compound, i.e., a compound within a compound. The first compound is a karmadhāraya or adjectivenoun compound, consisting of the adjective mahā, great, and the second compound as its second member, consisting of the last two words taken as a unit, samayasattva. The second compound, samaya-sattva , is also a karmadhāraya or adjective-noun compound. It consists of the noun samaya, pledge, here used as an adjective, and the noun sattva, being or living being, together meaning a pledge being. Then this second compound forms the last member of the first compound, beginning with mahā, great. So the whole double compound means great pledge being. Since it is declined in the vocative case, it means O great pledge being. In Buddhist tantra, a pledge being is a deity that is visualized in the sādhana meditation practice. Then the jñāna-sattvas, the wisdom beings, understood as the actual deities, are invited to come and merge with the visualized deities, the pledge beings. āḥ, (mantra syllable), untranslatable. Tibetan tradition typically regards this as the seed syllable of the speech vajra of all the buddhas. This is in relation to oṃ as the seed syllable of the body vajra and hūṃ as the seed syllable of the mind vajra. The Variant Readings As said above, the most common error is taking the second occurrence of Vajrasattva in this mantra as a vocative. However, this error does not result from any variant reading, but only from how the words are understood to divide. All words are written together in the Sanskrit manuscripts, with no space between them. Likewise in the Tibetan, although each syllable is separated by a dot (tsheg), the words are not separated. So how to divide the words is up to the reader. This is normally done by recognizing the words. But of course Tibetans would not be expected to recognize that vajrasattvatvena is a Sanskrit word. We see that in some Tibetan transcriptions the -tvena part, -tveno- when joined with the following word, is found as -teno-. This is an old error, occasionally (but not usually) found in the Narthang and Peking Tengyurs, and in these cases the sDe-dge and Co-ne Tengyurs correctly have -tveno-. It is easily explained as a scribal error, since the “v” (Tibetan “w”) is only small mark on the “t” in Tibetan. This error continued up to the present in some Tibetan texts. For example, it is found seven times each in both editions of Lama Thubten Yeshe’s book on Vajrasattva (listed below). Therefore, when giving the readings from Tibetan sources below, I have always reported it. The Vajrasattva Mantra: Sanskrit Edition and English Translation 6 Much more common is the variant reading śriyaṃ for śreyaḥ, to the point of being the normal form in later Tibetan sources. In the old sources śriyaṃ is an occasional variant, but the normal form is śreyaḥ. These words are related, and are similar enough in their general meaning to not significantly affect the overall sense of the sentence. When giving the readings from Tibetan sources below, I have always reported this word. We sometimes find in the Tibetan sources the mantra syllable huṃ instead of hūṃ, and the mantra syllable ho instead of hoḥ. Such variants for mantra syllables are frequent in Tibetan texts, since these are Sanskrit syllables, and as such are not part of normal Tibetan vocabulary. Benoytosh Bhattacharyya in his Sanskrit edition of the Sādhanamālā has huṃ throughout, but after realizing his error he changed it to hūṃ in his later writings. When giving the readings from Tibetan sources below, I have always reported these two mantra syllables. The vocative bhagavan is most often found as bhagavān in the Tibetan sources, which erroneously makes it a nominative. I have usually reported it, but not always, since it is clearly an error. Among the sources I have checked, this error occurs only once in a Sanskrit manuscript, and there it is more likely to be a scribal error than an actual variant reading. It is basically a Tibetan error, since Tibetans would not be expected to know the difference between these two. Another common error found in the Tibetan sources is vajribhava for vajrībhava. Tibetan does not have a long vowel “ī”, so this is an easy error to make. It is shown in Tibetan script by an added sign below the vowel “i”. In Sanskrit, this word can only be vajrī, whether or not compounded with bhava. I have usually reported it, but not always, since it is clearly an error. In later Tibetan sources this mantra normally ends with two added final syllables hūṃ phaṭ, after āḥ. These expand the hundred-syllable mantra to one hundred and two syllables. In the old Tibetan canonical sources they are normally only found in the Heruka version of the Vajrasattva mantra, where they are needed to make one hundred syllables. This mantra usually ends correctly in the old Tibetan canonical sources. I have usually reported this by saying “ends āḥ.” When the mantra syllable āḥ is found as aḥ, or even as a, I have reported this. For the sake of clarity and ease of comparison, I have usually reported variant readings from the Tibetan texts as they would be in Sanskrit. Thus, I report vajri, not badzri, as it is written in Tibetan script. One must know that the Sanskrit letter “v” is transliterated into Tibetan script as “b”, and in some contexts as “w”, and the Sanskrit letter “j” is transliterated into Tibetan script as “dz”. Likewise, the Sanskrit letter “c” is transliterated into Tibetan script as “ts”. So, for example, the Sanskrit word citta is transliterated into Tibetan script as tsitta. In the variant readings given below, I have occasionally given the Tibetan as written in Tibetan script. This was done primarily to remind readers of what they would see in the Tibetan texts. Note that I am using standard transliteration systems when representing Sanskrit and Tibetan words and letters in the Roman or Latin script that English is written in. This is the IAST (International Alphabet of Sanskrit Transliteration) system for Sanskrit, and the Wylie system for Tibetan. 7 The Vajrasattva Mantra: Sanskrit Edition and English Translation Summary The Vajrasattva mantra is of fundamental importance in the Mahāyāna Buddhist tradition. Since it is so widely used from Tibetan sources into which errors have crept over the centuries, it was necessary to consult and cite a large number of authoritative sources to show the correct original reading. These many authoritative sources leave no doubt as to its correct original wording. The variant readings serve only to indicate errors, such as -teno- for -tveno-, or to give a form that is not original, such as śriyaṃ instead of śreyaḥ. From the correct wording, the correct meaning can be ascertained, as given above for example. There are other equally correct ways to translate it. This mantra, like the majority of mantras, consists of grammatically correct Sanskrit sentences. So its meaning is clear. A variant meaning worth noting is “make me” instead of “you be” for the four phrases ending with “me bhava.” In Sanskrit, the imperative verb bhava can only mean “you be,” and this is how it was translated in the oldest and most authoritative Tibetan translation known to me: gyur cig (see below, “The Rare Tibetan Translations”). However, there are also Tibetan translations of this word in this mantra as “make,” mdzad or mdzod. Tshul khrims rgyal mtshan in his rather anomalous translation of this mantra in the Hālāhala Sādhana (see below) has dgongs for the first two occurrences, and mdzad du gsol for the last two. The later Tibetan writer Bhadanta Indra (see below) has mdzad du gsol for the first occurrence, and mdzod for the other three occurrences. This, of course, changes the meaning of the mantra rather dramatically. Instead of requesting Vajrasattva to be steadfast toward me, well pleased with me, well fulfilled on account of me, and attached to me, it requests Vajrasattva to make me steadfast, make me well pleased, make me well fulfilled, and make me attached, presumably to Vajrasattva. Thus, Jeffrey Hopkins in the book, Tantric Practice in Nying-ma, p. 146, translates these four phrases in the Vajrasattva mantra as “Make me firm. Make me satisfied. Fulfil me. Make me compassionate,” and the same in the later book, Kalachakra Tantra: Rite of Initiation, p. 403.12 The earlier book gives phrase-by-phrase meanings, explaining: “bhava : be (imperative), having the sense of ‘make’.” Why bhava, “be,” is here thought to have the sense of “make,” is not stated. With all due respect, I must regard this as an error that was introduced long ago into the Tibetan tradition and perpetuated up to the present.13 The Tibetan Recensions Consulted Any occurrence of the Vajrasattva mantra in the Kangyur was checked in the Comparative Kangyur: bKa' 'gyur, dpe bsdur ma. Beijing: Krung go'i bod rig pa dpe skrun khang, 2006-2009. The Comparative Kangyur gives the Tibetan text in the sDe-dge recension, and variant readings from seven other recensions in backnotes. The seven other recensions are: Yongle (g.yung lo), Lithang, Peking, Narthang, Co-ne, Urga (khu-re), and Lhasa (zhol). These eight recensions will be regularly referred to below. Any occurrence of the Vajrasattva mantra in the Tengyur was checked in the Comparative Tengyur: bsTan 'gyur, dpe bsdur ma. Beijing: Krung go'i bod rig pa dpe skrun khang, 1994-2005. The Comparative The Vajrasattva Mantra: Sanskrit Edition and English Translation 8 Tengyur gives the Tibetan text in the sDe-dge recension, and variant readings from three other recensions in backnotes. The three other recensions are: Peking, Narthang, and Co-ne. These four recensions will be regularly referred to below. The Sanskrit Sources The Vajrasattva mantra is found in the Sarva-tathāgata-tattva-saṃgraha, “Compendium of the True Reality of All the Buddhas.” This seems to be its first known occurrence.14 Moreover, this seems to be its only occurrence in a text that is regarded as the word of the Buddha.15 In this text it is found at the end of the first chapter of the first part. This lengthy chapter was translated into Chinese by Amoghavajra in 753-754 CE as an independent text (Taisho no. 865), which became one of the two foundation texts of the Shingon school of Buddhism. Selected material from this text had been translated into Chinese earlier by Amoghavajra’s teacher, Vajrabodhi, in 723 CE, shortly after their arrival in China from India in 720 CE. Vajrabodhi is reported to have been taught the Sarva-tathāgata-tattva-saṃgraha in southern India in 702708 CE.16 In this translation of selected material (Taisho no. 866), the Vajrasattva mantra is included.17 The whole text was later translated into Chinese by Dānapāla, in 1012-1015 CE (Taisho no. 882). The whole text was also translated into Tibetan (Tohoku no. 479), by Śraddhākara-varman and Rin-chen bzang-po (958-1055 CE). It is there regarded as the primary text of the yoga-tantra division. We are fortunate to have the Sanskrit original of this text in a very old manuscript, estimated to date from the 9th10th century on the basis of its archaic script. As described by J. W. de Jong in his review of the 1974 second part of Horiuchi Kanjin’s edition of this text, “In 1932 Giuseppe Tucci announced the discovery in Nepal of a palm leaf manuscript of the Sarvatathāgatatattvasaṃgraha in the collection of general Kesar Sham Sher Jang” (Indo-Iranian Journal, vol. 19, 1977, p. 125). He references this to Tucci’s 1932 Italian book, Indo-Tibetica, vol. 1, p. 93. This book is now available in an English translation by Uma Marina Vesci, as Stupa: Art, Architectonics and Symbolism, where we read, also on p. 93: “. . . the Sarva-tathāgata-vajra-samaya-mahākalpa-rāja, a large fragment of which has been found in Nepal in a manuscript on palm-leaves, written in late Gupta script, possibly of the IX century” (New Delhi: Aditya Prakashan, 1988). Tucci adds in a footnote: “. . . the copy sent to me by General Kesar Sham Sher Jang Bahadur, who has traced and preserved the manuscript and with his usual courtesy has acceded to my desire of letting me have it and copy it.” The transcript in devanāgarī script, so obtained, was microfilmed. In this form it became available to Horiuchi, then forming the basis of his edition. Then in 1956 this same manuscript was again discovered, by John Brough and David Snellgrove. In his Introduction to the 1981 facsimile reproduction of it, Snellgrove wrote, p. 5: “. . . we discovered this manuscript early in 1956 in the private library of Field Marshall Kaiser Shamsheer in Kathmandu, who used generously to leave us free with his librarian to seek out whatever we might find of his considerable collections and then to make photographs. Thus one day we came upon this bundle of Indian palm 9 The Vajrasattva Mantra: Sanskrit Edition and English Translation leaves, written in a script which suggested at once its origin in 9th-10th century Bihar, completely disordered and listed under a title that bore little relationship to its apparent contents.” They made two sets of microfilms of it. Several years later they made them available to Isshi Yamada. This manuscript then formed the basis of his 1981 edition. As reported in de Jong’s review, Horiuchi obtained a photocopy of this manuscript “in April 1973 when the second part of his edition had already been handed over to the printer but he was able to add supplementary notes recording the readings” of it. He then used it to revise the first part of his edition, which had been previously published serially in Japanese journals, and was now published in book form in 1984. Following upon the occurrence in this text, the Vajrasattva mantra seems to have gradually come into use in Sanskrit sādhanas, until its use became the norm in sādhanas written in Tibetan. What may be considered its standard form was recorded by Abhayākaragupta (11th-12th century CE) in his Vajrāvalī, an influential tantric manual. It is also found in another tantric manual, the Kriyā-saṃgraha [-pañjikā] by Kuladatta, in both its standard form and with some small variants. In the collection of 312 short sādhanas published as the Sādhanamālā we find eight occurrences of the Vajrasattva mantra, usually in its standard form, with occasional small variants. The Vajrasattva mantra is known as the “hundred-syllable” (śatākṣara) mantra. All words are written together in Sanskrit manuscripts, without spaces between them. The Sanskrit sources, given below, leave no doubt about which hundred syllables this mantra consists of. However, there are some possible differences in how these syllables break into words. How this mantra is printed in the various Sanskrit editions shows how the various editors understood the word breaks. The Vajrasattva mantra as found in the various Sanskrit editions is given below, followed by my comments. Sarva-tathāgata-tattva-saṃgraha There are three printed editions of the Sarva-tathāgata-tattva-saṃgraha. They are all based on the one known Sanskrit manuscript, or a transcript of it. As said above, it is a very old manuscript, reported by Giuseppe Tucci to be written in late Gupta script, possibly of the ninth century (Indo-Tibetica, vol. 1, 1932, p. 93). In brief, summarized from above, Tucci obtained a devanāgarī transcript of it. From Tucci, Horiuchi Kanjin obtained a microfilm of this devanāgarī transcript. Horiuchi used this transcript as the basis of his edition, the first part published serially in Japanese journals, 1968-1973, and the second part as a book, in 1974. As the second part went to press, Horiuchi obtained a photocopy of the original manuscript, and was able to add supplementary notes based on it to his 1974 edition. He then revised the first part, which was published as a book in 1983. He was able to use two Chinese translations (the partial one by Amoghavajra and the complete one by Dānapāla), the Tibetan translation, and two commentaries in Tibetan translation, by Ānandagarbha and Śākyamitra, to prepare his edition. The second published edition was prepared by Isshi Yamada, and published in 1981. It is based on the microfilms of the original palm-leaf manuscript made by David Snellgrove and John Brough (see p. 5 The Vajrasattva Mantra: Sanskrit Edition and English Translation 10 of his edition). He also used two Chinese translations (by Amoghavajra and by Dānapāla) and the Tibetan translation, in the Peking, Snar-thaṅ, and Sde-dge editions (p. 6). The third published edition, by Lokesh Chandra, is “a revised Devanagari edition” of Yamada’s romanized edition (see p. 9 of this edition). It was published in 1987. The original manuscript was published in facsimile by Lokesh Chandra and David Snellgrove in 1981. It had been “discovered” by David Snellgrove and John Brough in 1956, as Snellgrove wrote in his Introduction (quoted above), not knowing that it had been previously discovered by Tucci in 1932. Indeed, Snellgrove went on to say in his Introduction, p. 6: “Professor Giuseppe Tucci drew attention long ago to its importance as a fundamental work of the Yoga-Tantra class, although the original Sanskrit version of the work had not yet been rediscovered.” He adds in a footnote here: “See his Indo-Tibetica, vol. 3, part I, pp. 38 onwards.” But Tucci there wrote, as we now have it in English translation: “In fact the Sarvatathāgata-tattva-saṅgraha, is one of the most ancient tantra, and most diffused in Buddhist schools, . . . Of the Tattva-saṅgraha the translation remains, . . . ; the original Sanskrit seemed however lost. But in Nepal I have found a collection called Sarva-tathāgat-ābhisamaya-mahākalparāja, which coincides with the Chinese translation of the Tattva-saṅgraha and must therefore be considered as a particular redaction of it.” (The Temples of Western Tibet and their Artistic Symbolism , English version of Indo-Tibetica III.1: The Monasteries of Spiti and Kunavar, p. 38. New Delhi: Aditya Prakashan, 1988.) 1. Horiuchi, Kanjin. Bon-Zō-Kan taishō Shoe Kongōchō-gyō no kenkyū, bonpon kōtei hen, Vol. 1. Kōyasan: Mikkyō Bunka Kenkyūjo, 1983 (this is an edition of the Sarva-tathāgata-tattva-saṃgraha). The vajrasattva mantra occurs on pages 203-204: oṃ vajra-satva samayam anupālaya! vajra-satva-tvenôpatiṣṭha!. dṛḍho me bhava! su-toṣyo me bhavânurakto me bhava!! su-poṣyo me bhava! sarva-siddhiṃ ca me prayaccha! sarva-karmasu ca me citta-śreyaḥ kuru! hūṃ ha ha ha ha ho! bhagavan sarva-tathāgata-vajra mā me muñca! vajrī- bhava! mahā-samaya-satva! āḥ! // Comments: (1) The anurakto me bhava phrase occurs between rather than after the phrases sutoṣyo me bhava and supoṣyo me bhava. For this phrase, Yamada notes that “Ch = Skt.” (see below), meaning that the two Chinese sources agree with the Sanskrit in its placement. For its placement in the Tibetan translation, see the Yamada listing, below. (2) The addition of ca after sarva-siddhiṃ must be regarded as a scribal error, since it is not in the Chinese and Tibetan sources, as noted by Yamada: “C & T omit CA.” (3) The absence of final “ṃ” on citta before śreyaḥ is very likely to be a scribal error, since it is not supported in other sources. (4) The spelling of satva and similar words with a single “t” rather than a double “tt” is common in old manuscripts. It is inconsequential and need not be retained. 11 The Vajrasattva Mantra: Sanskrit Edition and English Translation 2. Yamada, Isshi. Sarva-Tathāgata-Tattva-Saṅgraha Nāma Mahāyāna-Sūtra. New Delhi: Sharada Rani, 1981, Śata-Piṭaka Series, Vol. 262. The vajrasattva mantra occurs on page 95: oṃ vajra-satva-samayam anupālaya vajra-satvatvenopatiṣṭha dṛḍho me bhava su-toṣyo me bhavānurakto me bhava su-poṣyo me bhava sarva-siddhiñ ca me prayaccha sarva-karmasu ca me citta-śreyaḥ kuru hūṃ ha ha ha ha hoḥ bhagavan sarva-tathāgata-vajra mā me muṃca vajrī bhava mahā-samaya-satva āḥ // Comments: (1) This printing has vajra-satva-samayam rather than vajra-satva samayam, meaning with anupālaya, “Keep the pledge of Vajrasattva” rather than “O Vajrasattva, keep the pledge.” (2) Yamada notes that for the anurakto me bhava phrase: “Ch = Skt. But in T su-poṣyo me bhava comes before anurakto me bhava.” However, according to the Comparative Kangyur, this is true only of the Yongle and Peking recensions, not of the other six Tibetan recensions that were compared (see below). These six agree with the Sanskrit in its placement of this phrase. We may note that these six break bhava | anurakto. This is necessary to make one hundred syllables. Once we eliminate the superfluous ca after sarva-siddhiñ, there would only be ninety-nine syllables. (3) For the addition of ca after sarva-siddhiñ, he notes that “Ch & T omit CA.” His “Ch” refers to both Chinese sources, and is true of all eight Tibetan sources recorded in the Comparative Kangyur, as well as all four sources of the mantra compilation by Bu-ston that are recorded in the Comparative Kangyur. (4) For bhagavan, he notes that “Only T gives bhagavān na instead of bhagavan.” However, according to the Comparative Kangyur, the added na is true only of the Narthang and Lhasa recensions. All eight recensions have spellings with bhagavān (ending with: bā na, bān, or wān) rather than bhagavan. (5) By putting a space between vajrī and bhava, this printing has the noun vajrī and the verb bhava rather than the verb vajrī-bhava. 3. Lokesh Chandra. Sarva-Tathāgata-Tattva-Saṅgraha. Delhi: Motilal Banarsidass, 1987. The vajrasattva mantra occurs on page 32 in devanāgarī script. I have here transliterated it from devanāgarī script: oṃ vajra sattvasamayamanupālaya, vajrasattvatvenopatiṣṭha, dṛḍho me bhava, sutoṣyo me bhavānurakto me bhava, supoṣyo me bhava, sarvasiddhiñca me prayaccha, sarvakarmasu ca me cittaśreyaḥ kuru hūṃ ha ha ha ha hoḥ bhagavan sarvatathāgatavajra mā me muṃca vajrībhava mahāsamayasatva āḥ | | The Vajrasattva Mantra: Sanskrit Edition and English Translation 12 Comments: (1) This printing has vajra sattvasamayamanupālaya rather than vajra-sattva samayam anupālaya, meaning, “O Vajra, keep the pledge of living beings” rather than “O Vajrasattva, keep the pledge.” (2) The spelling satva in Yamada’s edition was changed to sattva in the first two occurrences, but not the third. (3) This printing changes Yamada’s noun and verb vajrī bhava to the verb vajrī-bhava. The facsimile reproduction of the original manuscript: Lokesh Chandra and David L. Snellgrove. Sarva-tathāgata-tattva-saṅgraha: Facsimile Reproduction of a Tenth Century Sanskrit Manuscript from Nepal. Śata-Piṭaka Series, Vol. 269. New Delhi: Mrs. Sharada Rani, 1981. The manuscript reproduction has four folio sides on each page. The Vajrasattva mantra is found in the second folio side on p. 15, lines 5-7. The Tibetan translation of the Vajrasattva mantra in the Sarva-tathāgata-tattva-saṃgraha : The Vajrasattva mantra in the Sarva-tathāgata-tattva-saṃgraha occurs in the Comparative Kangyur, volume 84, page 80. In addition to the readings given in the next paragraph, it has the readings -twe no-, shre yaḥ (in the sDe-dge, Lithang, Narthang, and Urga recensions), shre ya (in Yongle), shre yaṃ (in Peking), and shri yaḥ (in Co-ne and Lhasa), hūṃ, hoḥ, badzrī- (in sDe-dge, Lithang, Co-ne, and Urga), badzri- (in Yongle, Peking, Narthang, and Lhasa), ends aḥ (in sDe-dge, Lithang, and Co-ne), āḥ (in Yongle, Peking, Narthang, Urga, and Lhasa). Comments: (1) It spells satwa rather than sattwa in all three places. (2) As noted above in the Yamada listing, six of the eight Tibetan recensions agree with the Sanskrit in its placement of the anurakto phrase between rather than after the sutoṣyo and supoṣyo phrases. These are sDe-dge, Lithang, Narthang, Co-ne, Urga, and Lhasa. In the Yongle and Peking recensions the anurakto phrase comes after the supoṣyo phrase. (3) As noted above, it does not have the added ca after sarva-siddhiñ. (4) It has tsittaṃ (cittaṃ) rather than tsitta (citta) before śreyaḥ, with no variants. Note: The colophons show that the Yongle and Peking recensions have the same translation as the other recensions. So the transposition of the anurakto phrase was apparently found in the manuscript used for the Yongle recension. The Yongle recension was printed from carved wood blocks in 1410 CE, making it by far the oldest available blockprint of the Kangyur. This reading was then copied in the Peking recension, since the Peking is well known to be a direct copy of the Yongle. There is another source for the Tibetan translation of the Vajrasattva mantra in the Sarva-tathāgata- tattva-saṃgraha. Four of the eight recensions compared in the Comparative Kangyur, namely, Yongle, Peking, Lithang, and Co-ne, include a text by Bu-ston in which he compiled the mantras from Kangyur texts. It has, from the [Sarva-tathāgata-] Tattva-saṃgraha, two occurrences of the Vajrasattva mantra, vol. 104, pp. 208-209, 238-239. These are also found in The Collected Works of Bu-ston, Part 16 (ma), folio sides numbered 163, 187. However, I have not found a second occurrence of the Vajrasattva mantra in the Sarva-tathāgata-tattva-saṃgraha, either in the Sanskrit or in the Tibetan translation. The Vajrasattva Mantra: Sanskrit Edition and English Translation 13 For the first occurrence, which must correspond to the Vajrasattva mantra in the Sanskrit text we have (because it is found in the first chapter of that text), it has the readings: satwa, -twe no-, anurakto phrase before supoṣyo phrase, no added ca after sarva-siddhiñ, tsittaṃ, shre yaḥ, hūṃ, ho (in all four recensions recorded here in the Comparative Kangyur), hoḥ (in his gSung 'bum, i.e., his Collected Works), bha ga wan (in Yongle, Peking, Lithang, and gSung 'bum), bha ga wān (in Co-ne), badzrī- (in all four Kangyur), badzri- (in gSung 'bum), ends ā (in Yongle, Peking), a (in Lithang, Co-ne), aḥ (in gSung 'bum). For the second occurrence, it has: satwa (in Yongle, Peking, Co-ne, gSung 'bum), sato (in Lithang, first occurrence only), -twe no-, anurakto phrase after supoṣyo phrase, no added ca after sarva-siddhiñ, tsittaṃ (in Yongle, Peking, Lithang), tsitta (in Co-ne, gSung 'bum), shre yaḥ, hūṃ, hoḥ (in Yongle, Peking, Lithang), ho (in Co-ne, gSung 'bum), bha ga wān (in Yongle, Peking), bha ga win (in Lithang), bha ga bāṃ (in Co-ne), bha ga bān (in gSung 'bum), badzṝ- (in Yongle, Peking), badzrī- (in Lithang), badzri- (in Co-ne, gSung 'bum), ends āḥ (in Yongle, Peking, Lithang), aḥ (in Co-ne, gSung 'bum). English translations of the Vajrasattva mantra in the Sarva-tathāgata-tattva-saṃgraha : Two translations of the Sarva-tathāgata-tattva-saṃgraha include translations of the Vajrasattva mantra. They are as follows: Giebel, Rolf W. Two Esoteric Sutras: The Adamantine Pinnacle Sutra, The Susiddhikara Sutra. Berkeley: Numata Center for Buddhist Translation and Research, 2001. The Adamantine Pinnacle Sutra is the Vajra-śekhara-sūtra, another name of the Sarva-tathāgata-tattva-saṃgraha. The Vajrasattva mantra is found on p. 99, in Sanskrit derived from Horiuchi’s edition, followed by an English translation: Oṃ vajrasattva samayam anupālaya, vajrasattvatvenopatiṣṭha, dṛḍho me bhava, sutoṣyo me bhava, anurakto me bhava, supoṣyo me bhava, sarvasiddhiṃ me prayaccha, sarvakarmasu ca me cittaśreyaḥ kuru, hūṃ ha ha ha ha ho, bhagavan sarvatathāgatavajra mā me muñca, vajrībhava, mahāsamayasattva āḥ. (Oṃ, O Vajrasattva, keep [your] pledge! be close at hand as Vajrasattva! be firm for me! be wellpleased in me! be attached to me! be well-nurturing for me! grant me all success and bring about happiness of mind for me in all actions! hūṃ ha ha ha ha ho! O Lord, Vajra of All the Tathāgatas, do not abandon me! be like a vajra! great pledge-being! āḥ! ) Todaro, Dale Allen. An Annotated Translation of the Tattvasaṃgraha (Part 1) with an Explanation of the Role of the Tattvasaṃgraha Lineage in the Teachings of Kukai. PhD. thesis, Columbia University, 1985. The Vajrasattva mantra is found on p. 444 in Sanskrit from Horiuchi’s edition (not quoted here by me), and in English translation on pp. 321-322 (the final aḥ here may be a typographical error, because on p. 444 it is āḥ ): The Vajrasattva Mantra: Sanskrit Edition and English Translation 14 Oṃ Vajrasattva! Keep the pledge! Serve Vajrasattva! Make me firm! Make me very happy! Make me pleased! Make me very prosperous and grant all accomplishments to me! Produce in every action the best of thought for me! Hūṃ ha ha ha ha hoḥ! Bhagavat Thunderbolt of all Tathāgatas! Do not abandon me! Be firm! Great pledge being! aḥ! Vajrāvalī by Abhayākara-gupta The one published edition of the Vajrāvalī was prepared by Masahide Mori on the basis of eight palm-leaf manuscripts and five paper manuscripts. This provides a solid basis for ascertaining the form of the Vajrasattva mantra as found in this text. This text was written at the peak of the flourishing of tantric Buddhism in India (11th-12th century CE). Abhayākaragupta was the abbot of Vikramaśīla monastery, the center of tantric study and practice in India. His Vajrāvalī is the most authoritative compendium or manual of tantric ritual. Mori, Masahide. Vajrāvalī of Abhayākaragupta: Edition of Sanskrit and Tibetan Versions , 2 volumes. Tring, U.K.: The Institute of Buddhist Studies, 2009. Buddhica Britannica Series Continua XI. The Vajrasattva mantra found in this careful edition of this influential text matches the mantra as edited at the beginning of this paper. It is found in volume 2, pp. 353-354: oṃ vajrasattva samayam anupālaya vajrasattvatvenopatiṣṭha dṛḍho me bhava sutoṣyo me bhava supoṣyo me bhava anurakto me bhava sarvasiddhiṃ me prayaccha sarvakarmasu ca me cittaṃ śreyaḥ kuru hūṃ ha ha ha ha hoḥ bhagavan sarvatathāgatavajra mā me muñca vajrībhava mahāsamayasattva āḥ | The variant readings: The old palm-leaf manuscripts are consistent in giving this mantra as edited by Mori and given above. In his book, Mori gave the variant readings only from the eight palm-leaf manuscripts in order to avoid lengthy footnotes, even though in the editing process he had also recorded the variant readings from the five paper manuscripts (see p. 40). For this mantra, he lists eight variant readings. All eight of these are found in manuscript A, Cambridge University Library Add. 1703, and only two of these are also found in other manuscripts. It is obvious from these that ms. A is faulty for this mantra. Its variants include erroneous readings -opratiṣṭha for -opatiṣṭha, suposo for supoṣyo, bhavi for bhava, an omission, and a partial repetition. The two that are also found in other manuscripts are citta for cittaṃ in ms. A and ms. B, Asiatic Society of Bengal No. G4835, and for śreyaḥ, ms. A has śriyaḥ, and ms. E, Asiatic Society of Bengal No. G3855, has śreya. Mori’s book also includes an edition of the Tibetan translation, based on three recensions: Peking, sDe-dge and Narthang (see p. 45). The edition of the Tibetan transliteration of the Vajrasattva mantra is The Vajrasattva Mantra: Sanskrit Edition and English Translation 15 found on p. 355. From it, relevant variant readings are: For śreyaḥ, Narthang and sDe-dge have śreyaṃ, and Peking has śrīyaṃ. All three recensions incorrectly have bhagavān for bhagavan, and all three end with aḥ rather than āḥ. There is a small discrepancy regarding the source of the variant ho for hoḥ, listing sDe-dge as the source for it in the Sanskrit edition, and listing Peking as the source for it in the Tibetan edition. According to the Comparative Tengyur, vol. 39, p. 160, ho is in Peking and Narthang, while hoḥ is in sDe-dge and Co-ne. Also, the Comparative Tengyur shows śreyaṃ in sDe-dge and Co-ne, and śrīyaṃ in Peking and Narthang. There is an apparent typographical error in Mori’s text having supuṣyo rather than supoṣyo, with no variant listed. The Comparative Tengyur has supoṣyo, with no variant listed. The Vajrāvalī also includes the Vajrasattva mantra as adapted to use the term Heruka. In this version the syllables hūṃ and phaṭ are added at the end in order to make one hundred syllables. It is found in the bali-vidhi section of the Vajrāvalī, while the standard Vajrasattva mantra is found in the pratimādi-pratiṣṭhā-vidhi section. For another example of the Heruka version from a Sanskrit source, see Elizabeth English, Vajrayoginī, 2002, p. 296, given below under “Related Materials in English.” The Heruka version is found in Mori’s edition, volume 2, p. 504, and the edition of the Tibetan transliteration is on the same page. oṃ vajraheruka samayam anupālaya herukatvenopatiṣṭha dṛḍho me bhava sutoṣyo me bhava anurakto me bhava supoṣyo me bhava sarvasidhiṃ me prayaccha sarvakarmasu ca me cittaṃ śreyaḥkuru hūṃ ha ha ha ha hoḥ bhagavan vajraheruka mā me muñca heruko bhava mahāsamayasattva āḥ hūṃ phaṭ | The relevant variant readings: Manuscript D, National Archives, Kathmandu, no. 3-402, has the phrase supoṣyo me bhava before anurakto me bhava. This is significant because, although this manuscript is not dated, Mori says “this manuscript may well be one of the earliest existent manuscripts of the VA” (p. 33). The placement of the supoṣyo phrase before the anurakto phrase is usual in the standard Vajrasattva mantra, and is also found in the Heruka version edited by Elizabeth English from an old palm-leaf manuscript. The four other variants listed by Mori from Sanskrit manuscripts are not relevant. Variants from the Tibetan transliteration, all of them found in all three recensions, are sidhi for sidhiṃ, bhagavān for bhagavan, heruka bhava for heruko bhava, and aḥ for āh. I do not know why Mori has śreyaḥkuru written together here, in both the Sanskrit and Tibetan editions. Kriyā-saṃgraha aka Kriyā-saṃgraha-pañjikā by Kuladatta The Kriyā-saṃgraha, also called the Kriyā-saṃgraha-pañjikā, is a later tantric manual, not as famous as the Vajrāvalī, and used specifically as a manual for the construction of a monastery and for the consecration of everything to be put in it. It has eight sections (prakaraṇa). The Vajrasattva mantra occurs twice in the sixth section: first in its second sub-section (devatā-yoga), and again in its eighth sub-section The Vajrasattva Mantra: Sanskrit Edition and English Translation 16 (pratiṣṭhā), there in its eighth and last sub-sub-section (pratimādi-pratiṣṭhā). Both of these parts have been published as or within partial Sanskrit editions, the first by Hitoshi Inui, 1991-1994, and the second by Ryugen Tanemura in 2004. Hitoshi Inui’s edition of the devatā-yoga sub-section of the sixth section was published in three installments in Kōyasan Daigaku Mikkyō Bunka Kenkyūjo Kiyō, vol. 4, 1991, pp. 184-152; vol. 5, 1992, pp. 160-133; and vol. 7, 1994, pp. 112-91 (reverse pagination). His edition is based on three very early palm-leaf manuscripts, designated A, B, C, two later palm-leaf manuscripts designated T5, T6, and four paper manuscripts designated T2, T3, T4, T8, besides the Tibetan translation in two recensions, Peking and sDe-dge, designated P and D. The Vajrasattva mantra is found in vol. 5, 1992, p. 140: oṃ vajrasatva samayam anupālaya vajrasatvatvenopatiṣṭha dṛḍho me bhava supoṣyo me bhava sutoṣyo me bhava anurakto me bhava sarvvasiddhim me prayaccha sarvvakarmmasu ca me citaṃ śreyaḥ kuru hūṃ ha ha ha ha hoḥ bhagavan sarvvatathāgatavajra mā me muñca vajrībhava mahāsamayasatva āḥ || Comments: (1) The supoṣyo me bhava phrase occurs before the sutoṣyo me bhava phrase. (2) The doubling of consonants after “r” is very common in old manuscripts. It is inconsequential. (3) The spelling citaṃ for cittaṃ is anomalous, since citaṃ has a different meaning. However, the variant readings show that two of the three very early manuscripts, B and C, have cittaṃ. This occurrence in the Tibetan translation of the Kriyā-saṃgraha is found in the Comparative Tengyur, vol. 31, p. 1688. Its readings are: satwa (all three occurrences), -twe no-, supoṣyo phrase before sutoṣyo phrase, sarba (not sarbba, all three occurrences), karmma, tsittaṃ (not tsitaṃ, i.e., citaṃ), shre yaḥ, hūṃ, hoḥ, bha ga bān (in all four recensions), ma me muṃ tsa in the sDe-dge and Co-ne recensions, mā me muṃ tsa in the Peking and Narthang recensions, badzrī (i.e., vajrī) in sDe-dge and Co-ne, badzri (i.e., vajri) in Peking and Narthang, ends aḥ in sDe-dge and Co-ne, ends āḥ in Peking and Narthang. This mantra in this location of the abridged English version of this book by Tadeusz Skorupski, Kriyāsaṃgraha: Compendium of Buddhist Rituals, An abridged version (Tring, UK: Institute of Buddhist Studies, 2002), pp. 84-85, was translated by him as: Oṃ Vajrasattva protect my pledge. Vajrasattva be in attendance, be firm for me, be favourable to me, remain attached to me, promote me, and grant me all attainments. Ascertainer of all karma, make superior my mind, Hūṃ Ha Ha Ha Ha Hoḥ. O Lord Vajra of all the Tathāgatas do not abandon me. Vajra-Essence, Great Pledge-being Āḥ. Comments: Skorupski incorrectly translates the second Vajrasattva as a vocative, like in the Sanskrit of the mantra when he incorrectly gives it later in the book with a space after it (p. 154 fn. 245): vajrasattva 17 The Vajrasattva Mantra: Sanskrit Edition and English Translation tvenopatiṣṭha. He translates the mantra as having the order “ sutoṣyo me bhava, anurakto me bhava, supoṣyo me bhava,” as in the Sanskrit mantra when he gives it later in the book, unlike it is found here. He has obviously translated sarva-karmasu ca as if it was sarva-karma-sūca, “Ascertainer of all karma,” even though when he gives the Sanskrit mantra later in the book he correctly has sarvakarmasu ca. He translates vajrībhava as a vocative, “Vajra-Essence,” rather than as a verb, perhaps reading vajrabhava, even though when he gives the Sanskrit mantra later in the book he has vajrībhava. We may note that Skorupski’s intention with this book, an abridged version, was to show what is in the Kriyāsaṃgraha rather than to provide a philologically precise translation, as was Tanemura’s intention. The second occurrence of the Vajrasattva mantra in the Kriyā-saṃgraha[-pañjikā] is found in the partial Sanskrit edition by Ryugen Tanemura: Kuladatta’s Kriyāsaṃgrahapañjikā: A critical edition and annotated translation of selected sections (Groningen: Egbert Forsten, 2004). His edition was established on the basis of nine palm-leaf manuscripts (see pp. 100-103), of which eight were available for this section, and four Tibetan recensions (Sde dge, Peking, Snar thang, and Golden Manuscript, see pp. 112113), although he says that the quality of the Tibetan translation is extremely poor (p. 114). The Vajrasattva mantra occurs in the pratimādipratiṣṭhā sub-sub-section, on p. 208: oṃ vajrasattvasamayam anupālaya vajrasattvatvenopatiṣṭha dṛḍho me bhava sutoṣyo me bhava supoṣyo me bhava anurakto me bhava sarvasiddhiṃ me prayaccha sarvakarmasu ca me cittaṃ śreyaḥ kuru hūṃ ha ha ha ha hoḥ bhagavan sarvatathāgatavajra mā me muñca vajrībhava mahāsamayasattva āḥ Comments: (1) Reading vajrasattvasamayam altogether before anupālaya yields the meaning, “Obey the pledge of Vajrasattva,” as he translated it, rather than “O Vajrasattva, keep the pledge.” (2) The supoṣyo phrase is before the anurakto phrase in all eight palm-leaf manuscripts that he used, unlike in the Tibetan translation (in all four recensions). This occurrence in the Tibetan translation of the Kriyā-saṃgraha is found in the Comparative Tengyur, vol. 31, p. 1847. Its readings are: satwa (all three occurrences), -twe no-, anurakto phrase before supoṣyo phrase, shre yaḥ, hūṃ, hoḥ, bha ga bān, ends aḥ in sDe-dge and Co-ne, ends āḥ in Peking and Narthang. The variant readings for the Tibetan recensions reported by Tanemura show it ending in aḥ in Peking and Narthang, besides sDe dge and Golden Manuscript. I have checked the Peking recension, and it ends in āḥ as reported in the Comparative Tengyur. This mantra was accurately translated into English by Tanemura, p. 305: Oṃ. Obey the pledge of Vajrasattva. Be present as Vajrasattva. Be firm for me. Be well-satisfied for me. Be well-nourished for me. Be pleased for me. Give me all success. Make my mind excellent in The Vajrasattva Mantra: Sanskrit Edition and English Translation 18 all actions. Hūṃ. ha ha ha ha hoḥ. O, Blessed one! O, All the tathāgatas that are vajra ! Never leave me. Become [unbreakable like] vajra. O, great pledge being. Āḥ. This mantra in this location of the abridged English version of this book by Tadeusz Skorupski, Kriyāsaṃgraha: Compendium of Buddhist Rituals, An abridged version, is given in Sanskrit on p. 154, footnote 245: oṃ vajrasattva samayam anupālaya vajrasattva tvenopatiṣṭha dṛḍho me bhava sutoṣyo me bhava anurakto me bhava supoṣyo me bhava sarvasiddhiṃ me prayaccha sarvakarmasu ca me cittaṃ śreyaḥ kuru hūṃ ha ha ha ha hoḥ. bhagavan sarvatathāgatavajram me muñca vajrībhava mahāsamayas tvam āḥ. Comments: The second vajrasattva is incorrectly separated from tvenopatiṣṭha by a space. Then, the placement of the anurakto phrase before the supoṣyo phrase would indicate that this is derived from the Tibetan translation here rather than from the Sanskrit. All eight palm-leaf manuscripts that Tanemura used, and the 1977 manuscript reproduction that Skorupski used (folio side number 237, line 4) have the anurakto phrase after the supoṣyo phrase. Lastly, there appear to be two typographical errors near the end of the mantra, making it two syllables short: sarvatathāgatavajram me instead of sarvatathāgatavajra mā me ; and mahāsamayas tvam instead of mahāsamayasatva. Sādhana-mālā The one published edition of the Sādhana-mālā was prepared by Benoytosh Bhattacharyya on the basis of eight manuscripts. Three of these are palm-leaf, although it appears that Bhattacharyya used copies rather than the originals of two of these. He wrote, regarding Cambridge ms. Add. 1686, dating from 1165 CE (vol. 1, p. xii): “An authenticated copy of this MS is preserved in the CENTRAL LIBRARY, BARODA.” Then, regarding a manuscript from the Library of the Asiatic Society of Bengal (vol. 1, p. xiii): “An authenticated copy of this MS is also preserved in the CENTRAL LIBRARY, BARODA, and the present text has been collated therewith.” In regard to the palm-leaf manuscript that he used directly, also from the Library of the Asiatic Society of Bengal, he wrote (vol. 1, p. xii): “It is by far the best MS of Sādhanamālā and is mostly correct.” Of the five paper manuscripts he used, two were so corrupt he says, that they were not fully collated. As is well known, his edition is a composite edition, taking sādhanas from more than one Sanskrit collection of them. Bhattacharyya, Benoytosh. Sādhanamālā, 2 volumes. Gaekwad’s Oriental Series, volumes 26 and 41. Baroda: Central Library, 1925; Baroda: Oriental Institute, 1928. 19 The Vajrasattva Mantra: Sanskrit Edition and English Translation I have noted eight occurrences of the Vajrasattva mantra in this published collection of 312 short sādhanas, in the sādhanas numbered 6, 29, 56, 71, 142, 195, 218, and 247. Other than some omissions of commas, here follows how it is printed in all eight places, with two exceptions. In sādhana number 142, p. 295, it has supoṣyo me bhava after anurakto me bhava, and in sādhana number 247, p. 480, it has supoṣyo me bhava before sutoṣyo me bhava. Note that he always had huṃ in this edition, but later changed it to hūṃ (see note 3, below). I have here transliterated it from the devanāgarī script. oṃ vajrasattva samayamanupālaya, vajrasattvatvenopatiṣṭha, dṛḍho me bhava, sutoṣyo me bhava, supoṣyo me bhava, anurakto me bhava, sarvvasiddhiṃ me prayaccha, sarvvakarmmasu ca me cittaṃ śreyaḥ kuru, huṃ hahahaha hoḥ bhagavan sarvatathāgatavajra mā me muñca, vajrībhava mahāsamayasattva āḥ | Two palm-leaf manuscripts of sādhana collections were published in facsimile in 1994, compiled by Gudrun Bühnemann. One, the *Sādhanaśatapañcāśikā, is the palm-leaf manuscript from Cambridge University Library that was used by Bhattacharyya in his edition, Cambridge ms. Add. 1686. The other, the *Sādhanaśataka, is a palm-leaf manuscript from the China Library of Nationalities that had been preserved in Tibet, made available for the first time in this publication. The Vajrasattva mantra occurs seven times in these two manuscripts: in each manuscript in the sādhanas numbered 6, 29, and 71 in Bhattacharyya’s edition, and in only one of the two manuscripts in the sādhana numbered 218. So I have been able to check the readings in these seven occurrences. Bühnemann, Gudrun, compiled by. *Sādhanaśataka and *Sādhanaśatapañcāśikā: Two Buddhist Sādhana Collections in Sanskrit Manuscript. Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 32. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien, 1994. There are three different sādhana collections preserved in Tibetan translation in the Tengyur. Seven of the eight sādhanas that we have in Sanskrit in Bhattacharyya’s edition are found in Tibetan in these collections. The sādhanas numbered 6, 29, and 56 in Bhattacharyya’s edition are found in all three of these Tengyur collections. These provide eight different occurrences of the Vajrasattva mantra, since one of these gives a translation rather than a transliteration of it. The sādhana numbered 71 is also found in all three collections, but it only includes the Vajrasattva mantra in one of the three. However, in another one of these three collections the Vajrasattva mantra is found in the closely related immediately following sādhana. The sādhanas numbered 142 and 195 are found in one collection in the Tengyur, providing two more occurrences of the Vajrasattva mantra. The sādhana numbered 218 is found in two different collections in the Tengyur, providing two more occurrences of the Vajrasattva mantra. Altogether fourteen occurrences of the Vajrasattva mantra in the Tengyur have been checked. The Comparative Tengyur The Vajrasattva Mantra: Sanskrit Edition and English Translation 20 gives variant readings from four different recensions of the Tengyur. So the variant readings from these four recensions for the fourteen occurrences of the Vajrasattva mantra have been noted below. The individual sādhanas in the three collections found in the Tengyur are identified by their Tohoku numbers, abbreviated as Toh. The Tohoku numbers refer to the standard reference work, A Complete Catalogue of the Tibetan Buddhist Canons (Bkaḥ-ḥgyur and Bstan-ḥgyur), published by the Tohoku Imperial University, Sendai, Japan, in 1934. From it, the location of each text in the sDe-dge recension of the Tengyur (Bstan-ḥgyur) is given below. I have scanned and posted the Tohoku catalogue here: http://prajnaquest.fr/downloads/BookofDzyan/Tohoku%20Catalogue%20of%20Tibetan%20Buddhist%20 Canons.pdf Also given are the numbers of the individual sādhanas from the Peking catalogue, abbreviated as Pek. This refers to The Tibetan Tripitaka, Peking Edition, Catalogue & Index, published by Suzuki Research Foundation, Tokyo, 1962. For a long time, the Peking edition in its reprint from Japan was the only Kangyur and Tengyur available. Then follows full reference to the location of each text in the Comparative Tengyur, described above in the section, “The Tibetan Recensions Consulted.” From it, variant readings in the sDe-dge, Peking, Narthang, and Co-ne recensions of the Tengyur are given below. In these listings, the abbreviation Pek. = Peking, and Nar. = Narthang. Note that -tveno- as found in the texts is usually given below as -tvena, as dissociated from the following upatiṣṭha. 1. Sādhanamālā #6. Ārya-ṣaḍakṣarī-mahāvidyā-sādhana, pp. 27-28: variant for vajrasattvatveno-, ms. A: ºsattvo tenoº; variant for hoḥ, mss. AN: ho. Bhattacharyya comments about manuscript A (Preface, p. xi): “It is a complete MS though full of spelling mistakes, omissions and repetitions. Spelling mistakes are ten per line on an average.” =*Sādhanaśataka #23, folio 15a2-3, omits -sattva in the first vajrasattva, anurakto me bhava is added in the margin before supoṣyo me bhava, has -tvena, śreyaḥ, hūṃ, hoḥ, bhagavan, vajrībhava, ends āḥ, includes daṇḍas as follows: . . . anupālaya | vajrasattvatvenopatiṣṭha | dṛḍho me bhava | sutoṣyo me bhava | [added here in the margin: anurakto me bhava] supoṣyo me bhava | sarvvasiddhiṃ me prayaccha | sarvvakarmmasu ca . . . [no more daṇḍas]. =*Sādhanaśatapañcāśikā #18, folio 51Aa2-5, has -tvena, śreyaḥ (ḥ added after the line was written), hūṃ, hoḥ, bhagavan, vajrībhava, ends āḥ, omits ca after karmasu. Toh. 3150: Phu 173b2-174b1; Pek. 3971; Comparative Tengyur, vol. 39, p. 488, has -tvena (Pek. Nar. tena), śriyaṃ (Pek. Nar. śrīyaṃ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ (Pek. Nar. aḥ). Other variants are inconsequential. Toh. 3332: Mu 24b7-25b4; Pek. 4153; Comparative Tengyur, vol. 40, p. 108, has -tvena, śreyaḥ (Pek. Nar. śriyaṃ), hūṃ, hoḥ, bhagavān, ends āḥ. Other variants are inconsequential. Toh. 3405: Mu70b1-71a6; Pek. 4226; Comparative Tengyur, vol. 40, p. 347, has -tvena, śreyaḥ (Pek. Nar. śrīyaḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ. Other variants are inconsequential. The Vajrasattva Mantra: Sanskrit Edition and English Translation 21 2. Sādhanamālā #29. Hālāhala-sādhana, p. 74. =*Sādhanaśataka #21, folio 14a1-2, has -tvena, dṛḍho mo bhava, śreyaḥ, hūṃ, hoḥ, bhagavān, vajrībhava, ends āḥ, includes daṇḍas as follows: . . . sutoṣyo me bhava | supoṣyo me bhava | anurakto me bhava | sarvvasiddhiṃ me prayaccha | sarvvakarmmasu ca . . . [no other daṇḍas]. =*Sādhanaśatapañcāśikā #9, folio 44Aa4-6, has -tvena, śreya, hūṃ, hoḥ, bhagavan, vajrībhava, ends āḥ, has ca after karmasu. Toh. 3159: Phu 179a7-180b3; Pek. 3980; Comparative Tengyur, vol. 39, p. 518, has translation instead of transliteration, translated by Tshul khrims rgyal mtshan: oṃ rdo rje sems dpa' dam tshig rjes su bskyad (Pek. Nar. brkyang) du gsol | | bdag rdo rje sems dpa' nyid du nye bar gnas pa brtan par dgongs | | bdag la shin tu dgyes par dgongs | | bdag la rgyas par mdzad du gsol | | bdag la rjes su chags par mdzad du gsol | | dngos grub thams cad bdag la legs par stsol | | bdag gi las thams cad grub par mdzod | | sems nges pa (Pek. Nar. par) legs par mdzad du gsol | | hūṃ ha ha ha ha hoḥ bcom ldan 'das de bzhin gshegs pa rnams thams cad kyi rdo rje'i rang bzhin las bdag la grol bar mdzod (Pek. bral bar mdzod, Nar. bral bar mdzad) | | rdo rje'i ngo bo'i dam tshig sems dpa' chen po āḥ | | Toh. 3330: Mu 23b7-24b1; Pek. 4151; Comparative Tengyur, vol. 40, p. 100, has -tvena (Pek. Nar. tena), anurakto phrase before supoṣyo phrase in sDe-dge and Co-ne, but not in Pek. and Nar., śreyaḥ (Pek. Nar. śriyaḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ. Other variants are inconsequential. Toh. 3421: Mu 84b1-85b2; Pek. 4242; Comparative Tengyur, vol. 40, p. 409, has -tvena, śreyaḥ (Pek. Nar. śriyaḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ. Other variants are inconsequential. 3. Sādhanamālā #56. Arapacana-sādhana, p. 119: variant for cittaṃ śreyaḥ, ms. Ab: cittaśriyaḥ. Not in *Sādhanaśataka. Not in the extant folios of *Sādhanaśatapañcāśikā #29. Toh. 3173: Phu 190a2-192a4; Pek. 3994; Comparative Tengyur, vol. 39, p. 569, has -tvena, anurakto phrase before supoṣyo phrase in sDe-dge and Co-ne, but not in Pek. and Nar., śreyaḥ, hūṃ, hoḥ (Pek. Nar. ho), bhagavān, ma for mā in sDe-dge and Co-ne, but mā in Nar. and Pek., vajrī- (Pek. Nar. vajri-), ends aḥ (Pek. Nar. āḥ). Other variants are inconsequential. Toh. 3310: Mu 8a7-10a6; Pek. 4131; Comparative Tengyur, vol. 40, pp. 30-31, has -tvena, śreyaḥ (Pek. śriya, Nar. śriyaṃ), hūṃ, hoḥ (Pek. Nar. ho), bhagavān, ma for mā only in Pek. and Nar., vajrī(Pek. Nar. vajri-bhāva), ends aḥ (not āḥ). Other variants are inconsequential. Toh. 3450: Mu 107a4-109a2; Pek. 4271; Comparative Tengyur, vol. 40, p. 517, has -tvena, śreyaḥ (Pek. Nar. śrīyaḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ. Other variants are inconsequential. The Vajrasattva Mantra: Sanskrit Edition and English Translation 22 4. Sādhanamālā #71. Ārya-siddhaikavīra-sādhana, p. 145: footnote regarding the Vajrasattva mantra, which occurs at the end of this sādhana: “This Śatākṣara Mantra is dropped in C and Ba.” =*Sādhanaśataka #12, folio 8b5-9a1, has -tvena, anurakto phrase before supoṣyo phrase, śriyaḥ, hūṃ, hoḥ, bhagavan, vajrībhava, ends āḥ, includes daṇḍas as follows: . . . sutoṣyo me bhava | anurakto me bhava | supoṣyo me bhava | . . . [no other daṇḍas], has ca after karmasu. =*Sādhanaśatapañcāśikā #39, folio 52b4-6, has -tvena, śreyaḥ, hūṃ, hoḥ, bhagavan, vajrībhava, ends āḥ, omits ca after karmasu. Toh. 3184: Phu 197b3-199a2; Pek. 4005; Comparative Tengyur, vol. 39, p. 603, has -tvena, śreyaḥ (Pek. śrīya, Nar. śrīyaṃ), hūṃ, hoḥ, bhagavān, ends āḥ (Pek. Nar. aḥ). Other variants are inconsequential. Toh. 3322: Mu 17b2-18b4; Pek. 4143; Comparative Tengyur, vol. 40, pp. 69-72, this version omits the Vajrasattva mantra. Toh. 3465: Mu 118a4-119b6; Pek. 4286; Comparative Tengyur, vol. 40, pp. 565-568, this version omits the Vajrasattva mantra; but it is found in the next sādhana, Toh. 3466 (which includes sādhana #72), Comparative Tengyur, vol. 40, p. 570, has -tvena, anurakto phrase before supoṣyo phrase only in Nar. and Pek., śreyaḥ (Pek. Nar. śrīyaḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-bhāva), ends āḥ. Other variants are inconsequential. 5. Sādhanamālā #142. Kalpokta-mārīcī-sādhana, p. 295: has supoṣyo me bhava after anurakto me bhava with no variants; has variant: mss. ANC omit ca after sarvvakarmmasu. Not in *Sādhanaśataka. Not in *Sādhanaśatapañcāśikā. rTog-pa-las gsuṅs-pa’i ’od-zer-can-gyi sgrub thabs: Toh. 3532: Mu 173a1-175a7; Pek. 4354; Comparative Tengyur, vol. 40, pp. 820-821, has -tvena (Nar. tena), śreyaḥ, hūṃ, hoḥ (Pek. Nar. ho), bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ (Pek. Nar. aḥ). Other variants are inconsequential. Sādhanamālā, vol. 2: 6. Sādhanamālā #195. Mahā-pratisarā-sādhana, p. 398: variant for me cittaṃ, ms. B: cittaṃ me. Not in *Sādhanaśataka. Not in *Sādhanaśatapañcāśikā. So-sor-’braṅ-ba-chen-mo’i sgrub thabs: Toh. 3584: Mu 216b1-217a4; Pek. 4406; Comparative Tengyur, vol. 40, p. 1018, has -tvena, śreyaḥ (Pek. śriyaḥ, Nar. śriyaṃḥ), hūṃ, hoḥ, bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ. Other variants are inconsequential. The Vajrasattva Mantra: Sanskrit Edition and English Translation 23 7. Sādhanamālā #218. Prajñāloka-sādhana, p. 430: Bhattacharyya has in parentheses (oṃ āḥ) at the beginning before oṃ vajrasattva, and (huṃ) at the end after āḥ. Not in *Sādhanaśataka. =*Sādhanaśatapañcāśikā #154, folio 147b1-2, has -tvena, anurakto phrase after dṛḍho phrase, then supoṣyo phrase, then sutoṣyo phrase, śreyaḥ (ḥ added after the line was written), hūṃ, ho, bhagavan, vajrībhava, ends āḥ hūṃ, omits ca after karmasu. rDo-rje phag-mo’i sgrub thabs: Toh. 3295: Bu 35b7-39a3; Pek. 4118; Comparative Tengyur, vol. 39, p. 944, has -tvena, śreyaṃ (Pek. Nar. śrīyaṃ), hūṃ, hoḥ, bhagavān, ma for mā only in Pek. and Nar., ends āḥ hūṃ (Pek. Nar. aḥ). Other variants are inconsequential. Śes-rab-kyi snaṅ-ba sgrub-pa’i cho-ga: Toh. 3608: Mu 230b5-233b4; Pek. 4430; Comparative Tengyur, vol. 40, p. 1096, has -tvena, śreyaḥ, hūṃ, hoḥ (Pek. Nar. ho), bhagavān, vajrī- (Pek. Nar. vajri-), ends āḥ hūṃ. Other variants are inconsequential. 8. Sādhanamālā #247. Pūjā-vidhi-saṃgraha, p. 480: has supoṣyo me bhava before sutoṣyo me bhava with no variants. Not in *Sādhanaśataka. Not in *Sādhanaśatapañcāśikā. Not in Tengyur. Other Sanskrit Texts in Tibetan Translation Besides these occurrences of the Vajrasattva mantra, it is found in five of eleven Tibetan texts in the Tengyur that have the title Vajrasattva-sādhana, or similar such titles. These eleven are: Vajrasattva-sādhana, Toh. 1628, Comparative Tengyur, vol. 13, pp. 631-632, not found in it. Vajrasattva-guhyārtha-dhara-vyūha, Toh. 1664, Comparative Tengyur, vol. 14, pp. 297-337, found on p. 306, has -tvena, anurakto phrase before supoṣyo phrase, śriyaḥ (Pek. śriya), hūṃ, hoḥ, bhagavān (Pek. Nar. bhagavan), vajrī- (Pek. Nar. vajri-), ends āḥ, followed by oṃ āḥ hūṃ. Vajrasattva-sādhana, Toh. 1680, Comparative Tengyur, vol. 14, pp. 801-804, not found in it. Bhagavac-chrī-mahāsukha-vajrasattva-sādhana, Toh. 1681, Comparative Tengyur, vol. 14, pp. 805-819, not found in it. The Vajrasattva Mantra: Sanskrit Edition and English Translation 24 Vajrasattvānusmṛti-nāma-vidhi, Toh. 1682, Comparative Tengyur, vol. 14, pp. 822-838, found on pp. 827828, has -tvena, ḍiḍho, śreyaḥ, hūṃ, hoḥ (Pek. Nar. ho), bhagavān, ends āḥ. Vajrasattva-sādhana, Toh. 1814, Comparative Tengyur, vol. 18, pp. 537-558, not found in it. Vajrasattva-sādhana-nibandha, Toh. 1815, Comparative Tengyur, vol. 18, pp. 561-571, not found in it. Vajrasattva-pūjā-vidhi, Toh. 1820, Comparative Tengyur, vol. 18, pp. 709-717, found on p. 716, has -tvena, anurakto phrase before supoṣyo phrase, śreyaḥ (Pek. Nar. citaṃ śriyaṃḥ), hūṃ, hoḥ, bhagavān, muñja for muñca only in Pek. and Nar., ends āḥ. Vajrasattva-sādhana-vyākhyā, Toh. 1835, Comparative Tengyur, vol. 18, pp. 1515-1527, not found in it. Vajrasattvodaya-nāma-sādhana, Toh. 2517, Comparative Tengyur, vol. 31, pp. 1084-1112, found on p. 1102, has -twe no-, anurakto phrase before supoṣyo phrase, shre yaḥ in sDe-dge and Co-ne, shri yaḥ in Pek. and Nar., hūṃ, hoḥ (Pek. Nar. ho), bha ga bān, badzrī- (Pek. Nar. badzri-), ends aḥ (not āḥ). Vajrasattva-sādhana, Toh. 2518, Comparative Tengyur, vol. 31, pp. 1117-1128, found on p. 1122, has -twe no-, anurakto phrase before supoṣyo phrase, shre yaḥ, hūṃ, hoḥ, bha ga bān, badzrī (Pek. Nar. badzri), ends aḥ (not āḥ). The Rare Tibetan Translations When translating the Buddhist scriptures into Tibetan, it was the practice of the Tibetan translators to only transliterate Sanskrit mantras into Tibetan characters, rather than translate them into the Tibetan language. However, there exist a few Tibetan translations of the Vajrasattva mantra. One of these was given above, as found in the Tibetan translation of the Hālāhala-sādhana by Tshul khrims rgyal mtshan. This is very unusual. Although he translated the mantra here, he transliterated it in his other translations: the Ārya-ṣaḍakṣarī-mahāvidyā-sādhana, the Arapacana-sādhana, and the Ārya-siddhaikavīra-sādhana. His translation is also unusual in some of its translation choices that depart from the norm. In addition to the Tibetan translation of the Vajrasattva mantra found in the sādhana text from the Tengyur, I am aware of two sources that give the Vajrasattva mantra phrase by phrase along with a Tibetan translation of each phrase. One of these was sent to me by the late Edward Henning, as from the bCom ldan 'das thams cad rig pa'i dkyil 'khor gyi cho ga'i rnam bshad legs par bshad pa thams cad kyi snying po, by 'Dul 'dzin Grags pa rgyal mtshan, 1374-1434.18 When later searching for this text at the Buddhist Digital Resource Center, I saw that this same explanation is found in a text in the sNga 'gyur bKa' ma, the collected teachings of the Nyingma from the early period.19 This would indicate that it is 25 The Vajrasattva Mantra: Sanskrit Edition and English Translation centuries older than the time of the Gelugpa lama Grags pa rgyal mtshan, a student of Tsong kha pa. I also saw that it is found another time, in the collected writings, gSung 'bum, of the Gelugpa lama Co ne Grags pa bshad sgrub, 1675-1748.20 Being incorporated in the writings of these later Gelugpa lamas would indicate that it was regarded as authoritative. The other source is the Sūtra-tantrodbhavāḥ Katipayadhāraṇīmantrāḥ (with Tibetan Translation), by Bhadanta Indra, Transcribed and Edited by Thinlay Ram Shashni. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies, 1997. Despite the Indian name, Bhadanta Indra was a Tibetan. I could not find his dates. The Buddhist Digital Resource Center record for this book reports “unknown author.” This book is based on a collection of mantras and dhāraṇīs made by Zhwa lu lo tsā ba Chos skyong bzang po, 1441-1527. Bhadanta Indra later supplied the Tibetan translations of the mantras. Here follows the Tibetan translation of the Vajrasattva mantra as found in the sNga 'gyur bKa' ma, with variants from 'Dul 'dzin Grags pa rgyal mtshan’s text in brackets (sources given in notes 18 and 19). oṃ [ni] sngar bzhin no [minus no] | badzra satwa sa ma ya ni rdo rje sems dpa'i dam tshig | ma nu pa [/pā] la ya ni rjes su skyongs shig | badzra sa twa ti [/twe] no pa tiṣṭha ni rdo rje sems dpa' nyid kyis [minus kyis] nye bar zhugs shig | dṛ ḍho me bha wa ni bdag la brtan par gyur cig pa'o [minus pa'o] | su po ṣyo mi [/me] bha wa ni bdag la [shin tu] rgyas par gyur cig | su to ṣyo me bha wa ni bdag la shin tu dgyes par gyur cig | [su to ṣyo phrase before su po ṣyo phrase] a nu rakto me bha wa ni bdag la rjes su chags par gyur cig | sarba siddhi mme pra yatstsha ni dngos grub thams cad bdag la rab tu stsol [zhig] | sarba karma su tsa me ni bdag la las thams cad rtsol par [/gtsang bar] mdzod cig | tsittaṃ śre yaḥ [/tsitta śri yaṃ] ku ru ni sems kyi dpal du mdzod cig | hūṃ ni bzhugs [/gzhug] pa'i sa bon yin [la] | ha ha ha ha ni ye shes [bzhi] | ho [/hoḥ] ni mnyes shing [/zhing] dbang du gyur cig [/byed pa'o] | bha ga wān ni bcom ldan 'das | sarba ta thā ga ta ni de bzhin gshegs pa thams cad | badzra mā [/ma] me muñtsa ni rdo rjes bdag rab tu grol bar mdzod cig | badzrī [/badzra] bha wa ni rdo rje can du gyur cig | ma hā sa ma ya [/yā] satwa āḥ [/aḥ] [ni] dam tshig chen po'i sems dpa' zhes pa'o | | Here follows the Tibetan translation of the Vajrasattva mantra by Bhadanta Indra, 1997, pp. 52-53: oṃ badzra sattwa samayamanupālaya | rdo rje sems dpa dam tshig rjes su skyongs | badzra sattwatwenopatiṣṭha | rdo rje sems dpa' nyid kyi nyer gnas mdzod | The Vajrasattva Mantra: Sanskrit Edition and English Translation 26 dṛḍho me bhaba | bdag la brtan par mdzod du gsol | sutoṣyo me bhaba | bdag la shin tu dgyes par mdzod | supoṣyo me bhaba | bdag la shin tu rgyas par mdzod | anurakto me bhaba | bdag la rjes su chags par mdzod | sarba siddhimme prayatstsha | bdag la dngos grub thams cad stsol | sarba karmasu tsa me tsittaṃ śreyaḥ kuru hūṃ | las kun la yang bdag sems la | dge legs su mdzod hūṃ | ha ha ha ha hoḥ bhagaban sarba tathāgata badzra ma me muñtsa | ha ha ha ha hoḥ bcom ldan de bzhin gshegs kun gyi | rdo rjer bdag ma 'dor | badzrī bhaba | rdo rje can mdzod | mahā samaya sattwa āḥ | dam tshig sems dpa' chen po āḥ | | Notes 1. However, a serious attempt to accurately ascertain its wording and its meaning was made by Dharmacārī Jayarava [Attwood] and published as “The Hundred Syllable Vajrasattva Mantra” in Western Buddhist Review, vol. 5, Oct. 2010, pp. 57-74. It is available at westernbuddhistreview.com/archive. It discusses a previous attempt to do so by Dharmachari Sthiramati (Dr. Andrew Skilton) in his article, “The Vajrasattva Mantra: Notes on a Corrected Sanskrit Text,” The Order Journal, vol. 3, 1990, pp. 60-73. Dharmachari Sthiramati used eighteen sources that with two possible exceptions drew only on Tibetan sources, and a woodblock print giving the Sanskrit in rañjanā script. I was unable to see either of these articles when I wrote my article in 2006 (see next note). I got access to Dharmachari Sthiramati’s article only in March, 2024. I see that, even with rather inadequate sources from Tibetan, his knowledge of Sanskrit enabled him to produce a very accurate Sanskrit edition and English translation. Dharmacārī Jayarava added grammatical analyses beyond what Dharmachari Sthiramati gave. Their editions are almost identical. There are only two real differences. Sthiramati has “vajrasattva samayam anupālaya,” while Jayarava has “vajrasattva samayamanupālaya.” This allows the possibility of breaking it as “samaya manupālaya,” as is commonly found in Tibetan transliterations, even though in his explanation he chose to break it as “samayam anupālaya.” For as indicated by Sthiramati, “samaya manupālaya” is not really a possible option, grammatically speaking. Sthiramati has “vajrī bhava,” while Jayarava has “vajrībhava.” Both of these are equally possible grammatically, but “vajrībhava” is more likely (see below in the “Word by Word Meanings and Grammatical Analysis”). Among the otherwise helpful grammatical analyses added by Jayarava, I noticed a few that require correction. See notes 5, 6, 7, and 8 below. 2. This article was nearly finished, or so I thought, in 2006. At that time I had checked all the occurrences of the Vajrasattva mantra in the published edition of the Sādhanamālā, the two published palm-leaf manuscripts of it, and the Comparative Tengyur. I had also gathered nine published English translations of it, and other English materials on it that had been published up to that time. However, I The Vajrasattva Mantra: Sanskrit Edition and English Translation 27 was not then able to finalize it. From this unfinished article I included the edited mantra and its English translation in my article, “Sanskrit Mantras in the Kālachakra Sādhana,” published in As Long as Space Endures: Essays on the Kālacakra Tantra in Honor of H.H. the Dalai Lama , 2009, p. 309. The sources that I have gathered since then only confirm the edition and translation made earlier. 3. The only error in Bhattacharyya’s text of the Vajrasattva mantra is “huṃ” for “hūṃ.” This error occurs throughout both volumes of his edition of the Sādhanamālā, in whatever mantra this syllable is found. He seems to have thought at first that the correct form was “huṃ,” so he applied this there consistently, despite the readings of the manuscripts he used. That he later recognized this error is shown by the fact that he changed it to “hūṃ” in his revised second edition of The Indian Buddhist Iconography, even there quoting his edition of the Sādhanamālā as if it had “hūṃ.” In the first edition of The Indian Buddhist Iconography (London: Oxford University Press, 1924), Benoytosh Bhattacharyya describes Vajrahuṅkāra on pp. 143-144, there quoting the then unpublished Sādhanamālā and always giving “huṃ” (or “huṅ” in “huṅkāra”). In the second edition, revised and enlarged (Calcutta: Firma K. L. Mukhopadhyay, 1958), he describes Vajrahūṅkāra on pp. 181-182, now always giving “hūṃ” (or “hūṅ” in “hūṅkāra”). He does this even in the quotation from the Sādhanamālā (p. 506), despite the fact that his edition of this volume (which had been published in 1928) actually has “huṃ” (or “huṅ” in “huṅkāra”). The two old Sanskrit palm-leaf manuscripts of this text that have now become available in facsimile, dating from the time of the Tibetan translations, and the many Tibetan transliterations themselves, all invariably have “hūṃ,” showing that this is in fact the correct reading. 4. There is no Sanskrit word tvena. It is sometimes taken to be a form of the pronoun “by you” (so Geshe Ngawang Dhargyey et al. as translated by Brian Beresford, in Mahāyāna Purification, and Khetsun Sangpo Rinbochay as translated by Jeffrey Hopkins, in Tantric Practice in Nying-ma, see listing below). However, the Sanskrit second person pronoun declined in the instrumental case meaning “by you” is tvayā, not tvena. It cannot be some variant form of tvayā. Besides the obvious reason that tvena is the perfectly regular suffix -tva declined in the instrumental case, there are also other reasons. First, it would not make sense to use a second person imperative verb, upatiṣṭha, “(you) stand near,” along with “by you.” Second, if tvena was a variant form of the second person pronoun in the instrumental case, we would expect to find it in some texts written in what has been called “Buddhist Hybrid Sanskrit.” But we do not. Franklin Edgerton does not record any such form in his standard Buddhist Hybrid Sanskrit Grammar. Nor is such a form found in Vedic Sanskrit, which has some similarities with Buddhist Hybrid Sanskrit. 5. Jayarava (see note 1 above) wrote, p. 60, that upatiṣṭha is a passive past participle, but it is an imperative verb. 6. Jayarava (see note 1 above) wrote, p. 61, that prayaccha is from the root iṣ, “to desire, to wish,” but it is from the root yam, “to give.” 7. Jayarava (see note 1 above), p. 66, after accepting the correct sarvakarmasu ca, discusses why Tibetans might have taken it as sarvakarma suca. He suggests that they might have understood suca as śoca, from the root śuc, in order to get the meaning, “Purify all my karma.” He notes that this cannot be The Vajrasattva Mantra: Sanskrit Edition and English Translation 28 the case, because the palatal sibilant ś is not interchangeable with the dental s. In fact, the interchange of these two sibilants is a regular feature of old Sanskrit Buddhist manuscripts. So this is not the reason why suca is incorrect. Stephen Beyer, 1973, and Martin Willson, 1986, give the equally incorrect sūca, from the root sūc, taking the phrase as sarva-karma-sūca, translated by Beyer as “Indicator of all karma.” 8. The text preserved in the bKa' ma collection of rNying ma teachings includes an explanation of the Vajrasattva mantra. It there says: ha ha ha ha ni ye shes, “ha ha ha ha are the wisdoms.” This same text as included in the Kun rig rnam bshad by the dGe lugs pa writer 'Dul 'dzin Grags pa rgyal mtshan, 1374-1434, has: ha ha ha ha ni ye shes bzhi, “ha ha ha ha are the four wisdoms.” Its explanation of the whole mantra in Tibetan is quoted below. The same text was also included in the gSung 'bum, collected writings, of the dGe lugs pa writer Co ne Grags pa bshad sgrub, 1675-1748. The explanation of these syllables as the wisdoms was taught by Panchen Lama Losang Chokyi Gyaltsen, 1570-1662, in his Essence of the Ocean of Attainments, who there says: “The four has and the hoḥ signify the five wisdoms” (translated by Yael Bentor and Penpa Dorjee, 2019, p. 67). This idea is found in many Tibetan sources. For example, Gyumé Khensur Lobsang Jampa says: “the four ha syllables, followed by hoḥ, stand for the five forms of transcendent knowledge” ( Guhyasamāja Practice in the Ārya Nāgārjuna System, Volume One, p. 117). Lama Thubten Yeshe says: “These five syllables symbolize the five wisdoms of the five dhyani buddhas” (The Tantric Path of Purification, 1995, p. 280; Becoming Vajrasattva, 2004, p. 262). Geshe Ngawang Dhargyey et al. as translated by Brian Beresford says: “The syllables ha ha ha ha ho symbolise the five types of pristine awareness, or wisdom.” (Mahayana Purification, p. 98; 2nd ed. p. 69). Despite all these Tibetan sources, Elizabeth English in her 2002 book, Vajrayoginī: Her Visualizations, Rituals, & Forms , p. 297, understood the syllables ha ha ha ha as laughter. Dharmachari Sthiramati in his 1990 article (see note 1 above), pp. 72 and 73, understood the syllables ha ha ha ha hoḥ as laughter. Jayarava Attwood in his 2010 article (see note 1 above), p. 61, wrote that the string of syllables ha ha ha ha hoḥ is generally understood to be laughter. But as we have seen, these syllables are generally understood to signify the wisdoms of the tathāgatas. There is an alternative explanation of ha ha ha ha as “the pure nature of the four joys” (Khetsun Sangpo, Tantric Practice in Nying-ma, p. 148). This may be seen in Tibetan in the text by the Nyingma lama Glag bla bSod nams chos 'grub, 1862-1944 (see note 13 below), p. 517: ha ha ha ha bzhi dga' ba bzhi rnam par dag pa'i rang bzhin. But I have not seen a traditional explanation of these syllables as laughter. 9. The traditional meaning of bhaga is given in a verse that is widely found or quoted, with some variants. For example, as found in the Viṣṇu-purāṇa, 6.5.74, in both the 1997-1999 critical edition and in the 1910 Veṅkateśvara Press edition, this verse is: aiśvaryasya samagrasya vīryasya yaśasaḥ śriyaḥ | jñāna-vairāgyayoś caiva ṣaṇṇāṃ bhaga itīraṇā || H. H. Wilson in his 1840 translation of this text translated these six qualities as “dominion, might, glory, splendour, wisdom and dispassion.” McComas Taylor in his 2021 translation of this text translated them as “supremacy over all, power, glory, majesty, knowledge and asceticism.” There is the variant reading 29 The Vajrasattva Mantra: Sanskrit Edition and English Translation dharmasya for vīryasya in four of twenty-seven manuscripts used for the critical edition. Also, this verse supplied by editor Fitzedward Hall in a footnote to the second edition of Wilson’s translation, vol. 5, 1870, p. 212, has dharmasya, even though Wilson’s “might” clearly translates vīrya. This verse as quoted by Śaṅkarācārya in his commentary on Śvetāśvatara Upaniṣad, 3.11, has dharmasya in the Ānandāśrama edition, following four of its five sources. The other source has vīryasya. The earlier Bibliotheca Indica edition has vīryyasya, i.e., vīryasya, here. In Buddhist texts the usual form of this verse, where rūpasya replaces vīryasya or dharmasya, and prayatnasya replaces vairāgya, is as follows: aiśvaryasya samagrasya rūpasya yaśasaḥ śriyaḥ | jñānasyâtha prayatnasya ṣaṇṇāṃ bhaga iti smṛtiḥ || This is as found, for example, in Vilāsavajra’s commentary on the Mañjuśrī-nāmasaṃgīti, verse 6 of chapter 1, in the 2016 edition and translation of chapters 1-5 of this text by Anthony Tribe, Tantric Buddhist Practice in India, p. 232. He translated these six qualities, p. 108, as “complete sovereignty, beauty, fame, glory, knowledge and activity.” As found in N. H. Samtani’s 1971 Sanskrit edition of the Arthaviniścaya-sūtra, p. 77, this verse ends with śrutiḥ instead of smṛtiḥ, but the six qualities are the same. In his translation, published in 2002 as, Gathering the Meanings: The Compendium of Categories , p. 49, these are: “[full] lordship, beauty, fame, wealth, knowledge, and effort.” Among the many other Buddhist books that give this verse, such as Haribhadra’s Ālokā on Abhisamayālaṃkāra 1.1-2, worth noting is its occurrence in a quotation from the lost Kālacakra mūla-tantra, found in the Sekoddeśaṭīkā, commenting on verse 1, pp. 64-65 in the 2006 critical edition by Francesco Sferra (the pioneering 1941 edition by Mario Carelli has the faulty reading artha instead of atha in pāda c, and this is listed as a variant reading in Sferra’s edition). 10. The etymology of bhagavat is found in the Madhya-vyutpatti, the sGra sbyor bam po gnyis pa, in the Comparative Tengyur, vol. 115, p. 316, line 17, to p. 317, line 8. 11. In the Sarva-tathāgata-tattva-saṃgraha, the Vajrasattva mantra is immediately preceded by three occurrences in prose of the cvi construction: atha sarva-mudrāṇāṃ sāmānyaḥ sva-kāya-vāk-cittavajreṣu vajrī-karaṇa-vidhi-vistaro bhavati | yadā mudrādhiṣṭhānaṃ śithilī-bhavati, svayaṃ vā moktukāmo bhavati, tato 'nena hṛdayena dṛḍhī-kartavyā |. It is shortly followed by another occurrence in prose of the cvi construction: vajrasattva-dṛḍhī-bhāvād (Horiuchi edition, pp. 203-204; Yamada edition, p. 95). 12. A related mantra used to invoke Vajrasattva as part of the Kālacakra initiation, found in the initiation (abhiṣeka) chapter of the Vimalaprabhā Kālacakra commentary, was translated by the early Tibetan writer Bu-ston in his annotated edition. It has the word bhava three times, followed by the word kuru, “make,” one time. All four of these were translated by Bu-ston as mdzod. This related mantra is found in the Vimalaprabhā Kālacakra commentary on chapter 3, verse 35, Sanskrit edition, vol. 2, p. 37: oṃ a ā aṃ aḥ vajrasattva mahā-sukha-vajra kālacakra śiṣyasyābhimukho bhava santuṣṭo bhava varado bhava kāya-vāk-cittādhiṣṭhānaṃ kuru kuru svāhā. It was translated into Tibetan by Bu-ston in his annotated edition of the Vimalaprabhā found in his Collected Works of Bu-ston, vol. 2, p. 298: rdo rje The Vajrasattva Mantra: Sanskrit Edition and English Translation 30 sems dpa' chen po bde ba rdo rje dus kyi 'khor lo slob ma'i mngon du phyogs par mdzod yang dag par mnyes par mdzod mchog stsol bar mdzod sku gsung thugs byin gyis brlab par mdzod. It was incorporated into Kay-drup’s Mandala Rite of the Glorious Kālachakra, translated by Jeffrey Hopkins in the book, Kālachakra Tantra: Rite of Initiation, p. 195 (also pp. 440-441): “Vajrasattva, Vajra Kālachakra of great bliss, approach the student, thoroughly please (the student), bestow the supreme, bless into magnificence exalted body, speech, and mind.” As may be seen, Bu-ston here translated bhava as mdzod, the imperative of mdzad, the honorific of byed, which translates Sanskrit words from the root kṛ, such as the imperative kuru, “make.” Since he then also translated kuru as mdzod, he apparently took bhava and kuru as synonyms in this mantra, even though bhava means “be” rather than “make.” 13. Thus, for example, the explanation of the Vajrasattva mantra by the Nyingma lama Glag bla bSod nams chos 'grub, 1862-1944, titled Yi ge brgya pa'i tshig 'grel nyung ngu gnad kyi thur ma , glosses bhava as mdzod cig. This text, 12 folios, is found in his gSung 'bum, vol. 3, pp. 501-524, Buddhist Digital Resource Center W22389, images 515-538. 14. Ryugen Tanemura had noted in his 2004 partial translation, Kuladatta’s Kriyāsaṃgrahapañjikā, p. 261, fn. 116: “Probably its earliest source is the Sarvatathāgatatattvasaṃgraha,” after which he gave the location to it in Horiuchi's edition. Then Jayarava confirmed this in his 2010 article (see note 1 above), p. 71, fn. 23: “Maitiu O’Ceileachair and I have identified the earliest textual reference to the Vajrasattva Mantra in the Chinese Tripiṭaka, in ‘A Summary of Recitations’ (Taishō 866). The summary is a selection of mantras from the Sarvatathāgata-tattvasaṃgraha translated into Chinese in 723 CE by Vajrabodhi, based on a text obtained in India around 700 CE.” 15. An electronic search of the Comparative Kangyur at the Buddhist Digital Resource Center yielded only one actual occurrence of the Vajrasattva mantra within a Kangyur text, the occurrence in the Sarva-tathāgata-tattva-saṃgraha, cited below. Three other occurrences came up in the search. Two of these are the Vajrasattva mantra from the Sarva-tathāgata-tattva-saṃgraha found in the text by Bu-ston in which he compiled the mantras from Kangyur texts, also cited below. As there noted, Bu-ston’s text is only found in four of the eight recensions compared in the Comparative Kangyur, namely, Yongle, Peking, Lithang, and Co-ne. The other occurrence is where the Vajrasattva mantra is added after the conclusion of the Amogha-pāśa-kalpa-rāja in four of the eight recensions compared: Yongle, Peking, Lithang, and Co-ne (vol. 92, pp. 927-928, note 756). It has the anurakto phrase before the supoṣyo phrase, shri yaṃ in Yongle, Peking, and Lithang, shrī yaṃ in Co-ne, ho, ends a (not āḥ). Other variants are inconsequential. 16. See: “The Life of the Tang Court Monk Vajrabodhi as Chronicled by Lü Xiang: South Indian and Śrī Laṅkān Antecedents to the Arrival of the Buddhist Vajrayāna in Eighth-Century Java and China,” by Jeffrey Sundberg, in collaboration with Rolf Giebel, pp. 134-135. Pacific World: Journal of the Institute of Buddhist Studies, Third Series, Number 13, Fall 2011. 17. See Jayarava in note 14 above, and Jayarava’s blog post, June 29, 2012: “Canonical Sources for the Vajrasattva Mantra”: https://jayarava.blogspot.com/2012/06/canonical-sources-for-vajrasattva.html. He here quotes the Vajrasattva mantra in Chinese characters from Vajrabodhi’s translation of selected 31 The Vajrasattva Mantra: Sanskrit Edition and English Translation material (Taisho no. 866) from the Sarva-tathāgata-tattva-saṃgraha, made in 723 CE. Vajrabodhi gives a gloss in Chinese for each phrase, and these glosses are included and also given in English translation in Jayarava’s blog post. All the observations on the Chinese are there credited to Maitiu O’Ceileachair. 18. Buddhist Digital Resource Center W12837. The explanation of the Vajrasattva mantra is found on folio sides numbered 343-344. The main title or short title is: Kun rig rnam bshad. It is described as: Survey of the practices associated with the Sarvadurgatipariśodhana Tantra. Work by one of the most important students of Tsongkhapa. 19. This text is the sNgon 'gro me tog phreng mdzes kyi 'grel pa . It is found in volume 16 of the 120volume set, bKa' ma shin tu rgyas pa, 1999, Buddhist Digital Resource Center W25983, pp. 5-158, images 7-160. The explanation of the Vajrasattva mantra is found on pp. 56-57. This text is also found in volume 15 of the 133-volume set, sNga 'gyur bka' ma shin tu rgyas pa, Chengdu, 2009, Buddhist Digital Resource Center W1PD100944, pp. 441-568, images 451-578. The explanation of the Vajrasattva mantra is found in images 493-494. Access is restricted because it is in copyright. 20. Buddhist Digital Resource Center IE1PD90129 (the first letter is capital “i”). The text is titled: bCom ldan 'das rnam par snang mdzad kyi cho ga rgyud don gsal ba 'grel pa dang bcas pa nas 'byung ba'i sngags kyi don rnam par bshad pa'i zur bkol bzha . It is the third text in volume 3 of this 18-volume set. The explanation of the Vajrasattva mantra is also found in his text titled: Lho sgo'i cho ga'i rgyas 'grel gzhan phan nyi 'od, in volume 16 of his collected writings. Access is restricted because of copyright. Related Materials in English (listed by date of publication) Blofeld, John. The Tantric Mysticism of Tibet: A Practical Guide . New York: E. P. Dutton & Co., 1970. “The Vajrasattva Purification,” pp. 159-163. Gives a translation of a brief Vajrasattva practice text from the Nyingma order under the heading, “The Vajrasattva Meditation and Recitation,” but does not include the Vajrasattva mantra. A further part of this practice is given on pp. 186-187. Willis, Janice Dean. The Diamond Light of the Eastern Dawn: A Collection of Tibetan Buddhist Meditations. New York: Simon and Schuster, 1972. “The Yoga Method of Dorje Sempa,” pp. 83-86. Dorje Sempa is Tibetan for Vajrasattva. Gives the Vajrasattva mantra, p. 85, according to Tibetan phonetics, e.g., Benzar sato for Vajra-sattva. Later in the book as part of a longer practice, p. 113, it is given as in the Tibetan transliteration of each Sanskrit syllable, e.g., va jra sa tva sa ma ya ma nu pa la ya. Beyer, Stephan. The Cult of Tārā: Magic and Ritual in Tibet. Berkeley / Los Angeles / London: University of California Press, 1973. Gives on p. 144 the Sanskrit text of the Vajrasattva mantra, and English translation; also gives its use in a Tārā practice, pp. 434-436. The Sanskrit edition incorrectly The Vajrasattva Mantra: Sanskrit Edition and English Translation 32 breaks vajrasattva tvenôpatiṣṭha, has anurakto me bhava between sutoṣyo me bhava and supoṣyo me bhava, incorrectly has sarva-karma-sūca, has śreyaḥ, has vajrī-bhava, ends āḥ. Accordingly, the English translation incorrectly has “Vajrasattva, let them be firm!” and “Indicator of all karma”: “OṂ Vajrasattva, guard my vows! Vajrasattva, let them be firm! Be steadfast for me, be satisfied, be favorable, be nourished for me! Grant me all the magical attainments! Indicator of all karma: make glorious my mind HŪṂ! HA HA HA HA HOḤ! Blessed One, diamond of all the Tathāgatas: do not forsake me, make me diamond! Great being of the vow ĀḤ!” Rabten, Geshe. The Preliminary Practices of Tibetan Buddhism (first published under the shorter title, The Preliminary Practices), translated by Gonsar Tulku. Dharamsala: Library of Tibetan Works and Archives, 1974; reprint, 1975 (now under the fuller title); reprints (with differing pagination): 1982, 1986, 2003. American edition, Burton, Washington: Tusum Ling Publications, 1974; reprint 1975 (both under the fuller title). “Vajrasattva Sadhana,” pp. 52-56 in 1974/1975 eds., pp. 79-85 in 1982 ed. (see also next listing). Gives the Vajrasattva mantra with words divided as in the Tibetan transliteration, e.g., vajrasattva samaya / manu palaya. Wallace, B. Alan. The Life and Teaching of Geshé Rabten. London: George Allen & Unwin, 1980. “Vajrasattva Meditation,” pp. 137-142. Includes the same sādhana found in The Preliminary Practices of Tibetan Buddhism listed above, but is translated differently. Also gives the Vajrasattva mantra with words divided as in the Tibetan transliteration, e.g., vajrasattva samaya / manu pālaya, but with added diacritics. In addition it gives the meaning of the Vajrasattva mantra phrase-by-phrase, and then “as a whole may be paraphrased as” (p. 140): “Oṃ The vow of Vajrasattva must be guarded. Make me of such a nature that I might dwell steadily in your presence, Courageous Bearer of the Vajra! Take pleasure in me; expand your presence in me, grant me your love. Bestow on me all wholesome actions and transform my mind into one of bliss. Hūṃ ha ha ha ha hoḥ Exalted One, Vajra of the Transcendent Beings, do not forsake me until I am of one nature with the Vajra. Great Being, abide by your pledge to me! Āḥ hūṃ phaṭ!” Kongtrul, Jamgon. The Torch of Certainty, translated from the Tibetan by Judith Hanson. Boulder: Shambhala Publications, 1977; reprint, Boulder: Prajñā Press, 1983. “Second of the Four Special Foundations: The Hundred-Syllable Mantra of Vajrasattva Which Purifies Harmful Deeds and Removes Obscurations,” pp. 79-90. Gives detailed practice instructions, but does not give the Vajrasattva mantra itself, presumably taking for granted that it is already known. Beresford, Brian C., translated and edited by. Mahāyāna Purification: The Confession Sūtra, with commentary by Ārya Nāgārjuna, & The Practice of Vajrasattva, with Sādhana; Supplemented by verbally transmitted commentaries from Geshe Ngawang Dhargyey, Geshe Rabten, Gegen Khyentse, Thubten Zopa Rinpoche. Dharamsala: Library of Tibetan Works and Archives, 1980; reprints 1993, 2003, with different pagination and changed title: The Confession of Downfalls: The Confession Sūtra, with 33 The Vajrasattva Mantra: Sanskrit Edition and English Translation commentary . . . . Gives the Sanskrit text of the Vajrasattva mantra, phrase-by-phrase meanings, and English translation, pp. 96-100 of the 1st ed., pp. 67-71 of the reprint ed. The Sanskrit edition incorrectly breaks vajrasattva tveno patistha and sarva karma succha me. Accordingly, the phrase-by-phrase meanings incorrectly have “tven[a] means ‘by you’,” and “succha me means ‘make me good’.” The English translation is given as: Om Vajrasattva Protect my commitment Vajrasattva, may I be upheld by you Remain firmly with me May you be pleased with me May you be happy with me Have affection for me Bestow on me all powerful attainments Make all my actions good Make my mind most glorious Hum! Ha ha ha ha ho O Endowed Transcendent Destroyer The Vajra of all Ones Thus-Gone Do not abandon me You whose nature is adamantine One with the great commitment Ah hum phat Sangpo, Khetsun, Rinbochay, translated and edited by Jeffrey Hopkins. Tantric Practice in Nying-ma. Ithaca, New York: Snow Lion Publications, 1982; Revised format 1996 (same pagination). “Vajrasattva Meditation,” pp. 141-153. Gives the Sanskrit text of the Vajrasattva mantra, English translation, and phrase-by-phrase meanings with grammatical notes, pp. 146-149. Leaving aside the intrusive second “n”, the Sanskrit edition incorrectly breaks Vajrasattva, tvenopatiṣnṭha. Accordingly, the English translation incorrectly has “Vajrasattva, reside [in me].” Likewise, the phrase-by-phrase meanings incorrectly have “tvena : by you, Vajrasattva.” They also have “bhava : be (imperative), having the sense of ‘make.’” So the English translation of the mantra is given as: “Oṃ Vajrasattva, keep [your] pledge. Vajrasattva, reside [in me]. Make me firm. Make me satisfied. Fulfil me. Make me compassionate. Grant me all siddhis. Also, make my mind virtuous in all actions. Hūṃ ha ha ha ha hoḥ all the blessed Tathagatas, do not abandon me, make me indivisible. Great Pledge Being, āḥ hūṃ.” Mullin, Glenn H., compiled and edited by. Meditation on the Lower Tantras, from the collected works of the previous Dalai Lamas. Dharamsala: Library of Tibetan Works & Archives, 1983. “A Vajrasattva Guru-Yoga Method,” by the Seventh Dalai Lama, translated by Kevin Garrett with Chomdze Tashi The Vajrasattva Mantra: Sanskrit Edition and English Translation 34 Wangyal, pp. 138-143. Gives the Sanskrit text of the Vajrasattva mantra on p. 142 with some incorrect word breaks; for example: samaya, manu palaya. Has the suposhyo phrase before the sutoshyo phrase. Combines Vajrasattva meditation with guru-yoga meditation by seeing the guru as Vajrasattva. McDonald, Kathleen. How to Meditate: A Practical Guide. London: Wisdom Publications, 1984. “Vajrasattva Purification,” pp. 178-186. Gives on p. 182 the Vajrasattva mantra partly in transliteration and partly as pronounced by modern Tibetans, e.g., vajrasattva samaya manu palaya, but suto kayo may bhawa. Then, pp. 185-186, very generalized phrase-by-phrase meanings, e.g., “Bhawa Please grant me the ability to realize the nature of phenomena.” Concludes with “In summary, the mantra means:” “O great courageous one whose holy mind is in the vajra nature of all buddhas, having destroyed every obscuration, attained all realizations and passed beyond all suffering, gone just as it is—do not forsake me but liberate me, please, according to your pledge.” Gyatso, Tenzin, edited, translated and introduced by Jeffrey Hopkins. The Kālachakra Tantra: Rite of Initiation for the Stage of Generation. London: Wisdom Publications, 1985. Gives on p. 403 the Vajrasattva mantra as found in “The Guru Yoga of Kālacakra in Connection with the Six Sessions in Completely Facilitating Form”: Oṃ vajrasatva, samayam anupālaya, vajrasatva, tvenopatiṣhṭha, dṛḍho me bhava, sutoṣhyo me·bhava, supoṣhyo me bhava, anurakto me bhava, sarva-siddhiṃ me prayachchha, sarva-karmasu cha me chittaṃ shrīyam kuru, hūṃ ha ha ha ha hoḥ, bhagavan-sarva-tathāgata-vajra, mā me muñcha, vajrī bhava, mahāsamaya-satva, āḥ hūṃ phaṭ. [Oṃ Vajrasattva, keep (your) pledge. Vajrasattva, reside (in me). Make me firm. Make me satisfied. Fulfil me. Make me compassionate. Grant me all feats. Also, make my mind virtuous in all actions. Hūṃ ha ha ha ha hoḥ all the supramundane victorious Ones Gone Thus, do not abandon me, make me indivisible. Great Pledge being, āḥ hūṃ phaṭ.] Kalu Rinpoche. The Gem Ornament of Manifold Oral Instructions, Which Benefits Each and Everyone Appropriately. San Francisco: KDK Publications, 1986. “Ngöndro: Dorje Sempa Meditation,” pp. 49-55. Dorje Sempa is Tibetan for Vajrasattva. Gives the Vajrasattva mantra, p. 53, according to Tibetan phonetics, e.g., Bedzra sa to for Vajra-sattva. Willson, Martin. In Praise of Tārā: Songs to the Saviouress. London: Wisdom Publications, 1986; reprint, 1996 (this is a somewhat corrected second edition, see p. 31). Gives on p. 350 the Sanskrit text of the Vajrasattva mantra found at the end of the Tibetan translation of the Tārā-bhaṭṭārikā-sādhana by Atīśa (i.e., Atiśa), as corrected from its Tibetan transliteration. It is the same in both printings. It incorrectly has vajrasattva tenôpatiṣṭha, incorrectly has sarva-karma-sūca (this was corrected to sarva-karmasu ca in his 2000 book, Deities of Tibetan Buddhism, see below), has śreyaḥ, has vajrī-bhava, ends āḥ. Dharmachari Sthiramati (Dr. Andrew Skilton), “The Vajrasattva Mantra: Notes on a Corrected Sanskrit Text,” The Order Journal, Issue 3, November1990, pp. 60-73. Gives the corrected Sanskrit text on pp. 71 and 72, after explaining the reasons for the corrections to the version that was previously used 35 The Vajrasattva Mantra: Sanskrit Edition and English Translation by the Western Buddhist Order. The Vajrasattva mantra found in this article matches the mantra as edited at the beginning of this paper, minus hyphens, with only one difference: vajrī bhava, instead of vajrībhava. Gives an English translation on p. 73: 1 OṂ Vajrasattva! Preserve the bond! 2 As Vajrasattva, stand before me. 3 Be firm for me. 4 Be greatly pleased for me. 5 Deeply nourish me. 6 Love me passionately. 7 Grant me siddhi in all things, 8 And in all actions make my mind most excellent. HŪṂ 9 Ha Ha Ha Ha Hoḥ 10 Blessed One! Vajra of all the Tathāgatas! Do not abandon me. 11 Be the Vajra-bearer, Being of the great bond! ĀḤ (HŪṂ PHAṬ) Patrul Rinpoche, translated by the Padmakara Translation Group. The Words of My Perfect Teacher. San Francisco: HarperCollins Publishers, 1994. “Chapter Three: Meditating on the teacher as Vajrasattva to cleanse all obscurations. III. The Actual Meditation on Vajrasattva,” pp. 267-280. Gives detailed practice instructions, but does not give the Vajrasattva mantra itself, presumably taking for granted that it is already known. Thurman, Robert A. F. Inside Tibetan Buddhism: Rituals and Symbols Revealed. San Francisco: Collins Publishers San Francisco, 1995. “Vajrasattva Practice,” pp. 50-53. Gives a brief description of the practice, with pictures. Then on p. 53, “The mantra can roughly be translated as follows:” “OM (Come to me from heaven!) I invoke the bonding with Vajrasattva, may he come close to me! Make me brave! Make me happy! Protect me! Energize me! Grant me all powers! Purify my evolutionary process! Make my mind glorious! I rejoice in the five wisdoms—ha ha ha ha hoh! O Lord, Diamond Thunderbolt of all Realized Ones! Do not abandon me! Make me a diamond thunderbolt holder, a great devotee hero! AH (May it happen!) HUNG (May it be integrated!) PHAT (May interferences be gone!).” Yeshe, Lama Thubten. The Tantric Path of Purification: The Yoga Method of Heruka Vajrasattva . Boston: Wisdom Publications, 1995. As can be seen by comparison, the 2004 book with changed title, Becoming Vajrasattva: The Tantric Path of Purification , is a slightly revised reprint of the 1995 book, even though it does not say so or make any reference to the 1995 book. The book is partly based on a teaching given in April 1974. It includes “Translation and Explanation of the One Hundred Syllable Mantra,” 1995 edition, pp. 279-280, 2004 edition, pp. 261-262, giving the mantra in which Vajrasattva is replaced by Vajra Heruka. This has wrong word breaks samaya manupalaya (1995 edition only) and heruka tenopatishtha (teno- is also wrong), and wrong sarva karma sucha. The 1995 edition has shriyam, changed to shreyam in the 2004 edition. The mantra ends ah hum phat, as it must in order to make the The Vajrasattva Mantra: Sanskrit Edition and English Translation 36 hundred syllables. Then follows very generalized phrase-by-phrase meanings. For example, manupalaya (1995) or anupalaya (2004) is explained as “The power of divine loving kindness, which can bring you to eternal happiness.” Khyentse, Dilgo, Rinpoche, translated and edited by The Padmakara Translation Group. The Excellent Path to Enlightenment: Oral teachings on the root text of Jamyang Khyentse Wangpo. Ithaca, New York: Snow Lion Publications, 1996. “Vajrasattva,” pp. 51-61. Gives the practice in this section. Then gives the Vajrasattva mantra in the Appendix, pp. 103-104, in Tibetan script and also according to Tibetan phonetics, e.g., Benzar satto for Vajra-sattva. Padmasambhava, Commentary by Gyatrul Rinpoche, translated by B. Alan Wallace. Natural Liberation: Padmasambhava’s Teachings on the Six Bardos . Boston: Wisdom Publications, 1998. “Reciting the One Hundred Syllables to Purify Sins and Obscurations,” pp. 67-69. Gives the Sanskrit text of the Vajrasattva mantra, p. 68. It incorrectly breaks vajrasattva tvenopatiṣṭha, has śrīyaṃ, incorrectly has vajra bhava, ends āḥ. Lamrimpa, Gen, translated by B. Alan Wallace. Transcending Time: The Kālacakra Six-Session Guru Yoga. Boston: Wisdom Publications, 1999. “Vajrasattva Purification,” pp. 20-21. Gives the Sanskrit text of the Vajrasattva mantra, p. 20. It incorrectly breaks vajrasattva tvenopatiṣṭha, has śrīyaṃ, incorrectly has vajri bhava, ends āḥ hūṃ phaṭ. Willson, Martin, and Martin Brauen, editors, translated by Martin Willson. Deities of Tibetan Buddhism: The Zürich Paintings of the Icons Worthwhile to See (Bris sku mthoṅ ba don ldan). Boston: Wisdom Publications, 2000. “13. Yellow Vajrasattva (rDo-rje-sems-dpa’ ser-po),” p. 243. Gives the Sanskrit text of the Vajrasattva mantra found in this sādhana, as corrected from its Tibetan transliteration. The Tibetan text can be found in: Sadhana-mala of the Panchen Lama, Part 1, reproduced by Lokesh Chandra, New Delhi: International Academy of Indian Culture, 1974, folio side 135. It spells satva (all three occurrences), has tve no (not te no), śri yaṃ, hūṃ, hoḥ, bha ga wan (not bha ga wān), ma me muñtza (not mā me muñtza), badzra bha ba (not badzrī bha ba), āḥ hūṃ phaṭ. His Sanskrit edition spells sattva, incorrectly breaks vajrasattva twenôpatiṣṭha, has sarva-karmasu ca (corrected as compared to his 1985 book, In Praise of Tārā, see above), has śriyaṃ, hūṃ, hoḥ, bhagavan, mā me muñca, vajrī-bhava, ends āḥ hūṃ phaṭ. Then follows what he calls “just a possible literal meaning”: “Oṃ Vajrasattva, preserve the pledge, Vajrasattva be here close by, be fixed for me, be my excellent satisfaction, be my excellent nourishment, be my attachment, grant me all realizations, and in all actions make my mind glorious, hūṃ, ha ha ha ha hoḥ, Vajra of all the Tathāgatas, do not abandon me, Vajra-wielder, Great Pledge Being, āḥ hūṃ phaṭ !” English, Elizabeth. Vajrayoginī: Her Visualizations, Rituals, & Forms . Boston: Wisdom Publications, 2002. This book includes a Sanskrit edition of the Vajravārāhī Sādhana by Umāpatideva, based primarily on an old palm-leaf manuscript that had been used in Tibet. This sādhana includes the Heruka version of the Vajrasattva mantra, given on p. 296. For another example of the Heruka version from a Sanskrit source, see Vajrāvalī, above. An English translation of this version is given in a footnote on p. 297, as The Vajrasattva Mantra: Sanskrit Edition and English Translation 37 follows. Notice that the syllables ha ha ha ha were taken as laughter, whereas the Tibetan tradition regards them as seed-syllables of the tathāgatas. oṃ vajraheruka samayam anupālaya, herukatvenopatiṣṭha, dṛḍho me bhava, sutoṣyo me bhava, supoṣyo me bhava, anurakto me bhava, sarvasiddhiṃ me prayaccha, sarvakarmasu ca me cittaṃ śreyaḥ kuru hūṃ, ha ha ha ha hoḥ bhagavan vajraheruka mā me muñca, heruko bhava mahāsamayasattva āḥ hūṃ phaṭ “O Vajraheruka! Guard the pledge! Be present to me as Heruka! Be firm for me! Be very glad for me! Be very abundant for me! Love me deeply! Grant me all siddhi! And in all actions, make my intention better! . . . (laughter) . . . O blessed one, Vajraheruka! Do not desert me! Be a Heruka, great samaya being! . . .” Jayarava, Dharmacārī [Attwood]. “The Hundred Syllable Vajrasattva Mantra.” Western Buddhist Review, vol. 5, Oct. 2010, pp. 57-74. Gives the Sanskrit text of the Vajrasattva mantra as corrected, referring to a previous article by Sthiramati (Dr. Andrew Skilton), word meanings with grammatical notes, and English translation. The Vajrasattva mantra found in this article matches the mantra as edited at the beginning of this paper, minus hyphens, with only one difference: samayamanupālaya is written together. This is because he allows the possibility of breaking it as samaya manupālaya, as is commonly found in Tibetan transliterations, even though he accepts it as samayam anupālaya. In fact, samaya manupālaya is not really a possible option. In his otherwise helpful grammatical analyses, there are a few that require correction. He wrote that the verb upatiṣṭha is a passive past participle. But it is an imperative verb. He wrote that prayaccha is from the root iṣ. But it is from the root yam. He wrote that sarvasiddhiṃ is singular, so it cannot mean ‘all attainments’ (plural). But it can. He wrote that ha ha ha ha hoḥ is generally understood to be laughter. But it is not. There are other statements pertaining to the Sanskrit that are misleading, such as about the Siddhaṃ script. However, his English translation is not inaccurate: oṃ O Vajrasattva, honour the agreement! Manifest as Vajrasattva! Be steadfast for me! Be very pleased for me! Be fully nourishing for me! Be passionate for me! Grant me all success and attainment And in all actions make my mind more lucid! hūṃ ha ha ha ha hoḥ O Blessed One, diamond of all those in that state, do not abandon me! Become real, O great agreement-being. āḥ The Vajrasattva Mantra: Sanskrit Edition and English Translation 38 Phuntsok, Yeshe, Khenpo. Vajrasattva Meditation: An Illustrated Guide. Somerville, Mass.: Wisdom Publications, 2015. Gives on p. 70 the Vajrasattva mantra “according to the Tibetan pronunciation of the Sanskrit syllables,” e.g., benza sato for Vajra-sattva, and then “according to Sanskrit phonetics.” The source for this latter is not stated. It has the same syllables as the Vajrasattva mantra edited at the beginning of this paper, which is the same as the one published in my paper, “Sanskrit Mantras in the Kālachakra Sādhana,” in As Long as Space Endures, 2009, p. 309. But as given in this 2015 book it has added capital letters, added exclamation marks, deleted hyphens, deleted ḥ (visarga) on shreya and ho, deleted diacritics, phoneticized letters (e.g., sh for ṣ and ś, ch for c, ri for ṛ, and ta for ṭha), wrong word separation by putting a space between vajrasattva and tvenopatishta, and by putting a comma after tathagata before vajra, and wrong phrase separation by putting punctuation after “cha me!” before “Chittam.” Bentor, Yael, and Penpa Dorjee. The Essence of the Ocean of Attainments: The Creation Stage of the Guhyasamāja Tantra according to Panchen Losang Chokyi Gyaltsen . Somerville, Mass.: Wisdom Publications, 2019. “Recitation and Meditation on Vajrasattva,” pp. 66-68. The Panchen Lama’s text gives the meaning of the words and phrases of the Vajrasattva mantra one by one, as he divided them. The Sanskrit mantra according to Thinlay Ram Shashni’s edition of Bhadanta Indra’s Sūtra-tantrodbhavāḥ Katipayadhāraṇīmantrāḥ is given in footnote 301 on p. 67. Note the probable typographical error here, vajrī bhāva instead of his vajrī bhava. It is given as vajrī bhava in the translation of the Panchen Lama’s explanations on that page. The example phrase sutoṣyo me bhava is translated here as “render me pleased,” and in a footnote as “make me pleased.” Jampa, Lobsang, Translated and Annotated by Artemus B. Engle. Guhyasamāja Practice in the Ārya Nāgārjuna System, Volume One: The Generation Stage. Boulder: Snow Lion, 2019. The edited Vajrasattva mantra is given on pp. 75, 111, 440, 541, 547-549, 649, and 677. The only differences from the mantra as edited at the beginning of this paper are the space between vajrī bhava, and the added final syllables hūṃ phaṭ. “The Meaning of the Mantra” is found on pp. 116-118, giving the meaning of the words and phrases of the Vajrasattva mantra one by one. The example phrase sutoṣyo me bhava is translated here as “Please be well satisfied with me.” ओं वज्रसत्त्व समयमनुपालय वज्रसत्त्वत्वेनोपतिष्ठ दृढो मे भव सुिोष्यो मे भव सुपोष्यो मे भव अनुरक्तो मे भव सववससतधं मे प्रयच्छ सववकमवसु च मे सचत्तं श्रेयः कु रु हं ह ह ह ह होः भगवन्सवविथागिवज्र मा मे मुञ्च वज्रीभव महासमयसत्त्व आः ||