Summary

The Uṣṇīṣavijayā Dhāraṇī with Its Ritual Manual is a short work in which the Buddha Amitāyus teaches the uṣṇīṣavijayā dhāraṇī along with its benefits and a number of short rites for its recitation.

Acknowledgements

This text was translated by Catherine Dalton, who also wrote the introduction.

The translation was completed under the patronage and supervision of 84000: Translating the Words of the Buddha.

The generous sponsorship of May, George, Likai, and Lillian Gu, which helped make the work on this translation possible, is most gratefully acknowledged.

Introduction

The Uṣṇīṣavijayā Dhāraṇī with Its Ritual Manual opens in Sukhāvatī, where the Blessed One Amitāyus is residing. Amitāyus addresses the bodhisattva Avalokiteśvara, informing him that there are beings who suffer from illnesses and short lifespans, and introducing the uṣṇīṣavijayā dhāraṇī as a remedy for such painful circumstances. Avalokiteśvara immediately asks Amitāyus to pronounce the dhāraṇī, which the Tathāgata does from within a state of samādhi.

After he pronounces the dhāraṇī, Amitāyus addresses Śakra and explains that reciting it purifies all lower rebirths and obscurations and leads beings to higher rebirths. Amitāyus lists a number of lower births that will be avoided by someone who recites this dhāraṇī, as well as the various prominent families within the human realm into which the reciter will be born on their way to attaining the state of awakening. As he continues to elaborate upon the ways in which one can use the dhāraṇī to benefit beings, Amitāyus explains that it can be hung from the top of a flagstaff, a mountain, or a tall building so that those in the surrounding areas will be purified and prevented from taking lower rebirths. He also notes that it may be carved into wood, placed at a caitya, and worshiped. This inspires Yama, the Lord of Death, to pledge to Amitāyus to protect beings who recite the dhāraṇī.

The Four Great Kings then request instructions on the rites associated with the dhāraṇī, and Amitāyus proceeds to teach a number of rites, including ways the dhāraṇī can be used to benefit the dead, animals, and those who are ill, as well as to prevent lower rebirths and extend life. These rituals involve the preparation of a maṇḍala in which the dhāraṇī is installed, making various offerings, the construction of caityas, and hundreds, thousands, or even a hundred thousand recitations of the dhāraṇī. In some cases the rituals involve fasting or performing the ritual at specific times of the lunar month. Śakra is instructed to teach one of these rites to the god Supratiṣṭhita to extend his life, which he does, and the god is delighted with the results. Finally, Amitāyus teaches his own dhāraṇī along with its related sādhana and fire offering rite, along with a number of additional applications of the dhāraṇī to obtain results such as glory and kingship.

This work is one among a group of texts in the Kriyātantra section of the Tibetan Kangyurs that contain the uṣṇīṣavijayā dhāraṇī and its related rituals (kalpa). The present text is the longest of four short dhāraṇī texts‍—three of which have the same title‍—that present the uṣṇīṣavijayā dhāraṇī with its ritual manual (kalpa). These four works share a similar narrative opening (nidāna) up through the presentation of the dhāraṇī proper, and several among them also share additional passages. Beyond its close relationship with the other ritual manuals connected with the uṣṇīṣavijayā dhāraṇī, this text shares a number of parallel passages with what we will call‍—simply for the purpose of distinguishing it from the present group of dhāraṇī-kalpas‍—the “primary” uṣṇīṣavijayā dhāraṇī text (Toh 597, which is titled Sarva­durgati­pariśodhana-uṣṇīṣavijayā- dhāraṇī rather than Uṣṇīṣavijayā-dhāraṇī-kalpasahitā). Despite these significant textual parallels, that text has an entirely different opening narrative and, unsurprisingly, lacks nearly all of the ritual instructions contained in our text.

There are many Sanskrit witnesses of the uṣṇīṣavijayā dhāraṇī proper, and the primary uṣṇīṣavijayā dhāraṇī text (Toh 597) survives in at least one, an incomplete early manuscript. While our text appears to no longer be extant in Sanskrit, there is at least one surviving Sanskrit uṣṇīṣavijayā dhāraṇī work that is closely related to it and belongs to the same group of related dhāraṇī texts described above. This work shares the same opening narrative and some of the ritual material with the texts from this group.

The primary uṣṇīṣavijayā text (Toh 597) was first translated into Chinese by Buddhapāli in the late seventh century, and then at least five times subsequently. Several ritual manuals for the dhāraṇī’s recitation were also translated into Chinese, but our text does not appear to be among them. One ritual manual (Taishō 978), translated into Chinese by Dharmadeva between 973 and 981, is among the group of uṣṇīṣavijayā dhāraṇī texts to which the present work belongs. The primary uṣṇīṣavijayā text was significant in East Asia, and one scholar has even identified it as the most important esoteric Buddhist scripture translated into Chinese in the seventh century. Practices connected with the uṣṇīṣavijayā dhāraṇī were important in China, in particular in conjunction with funerary rites, where the dhāraṇī was written on pillars near tombs, especially from the mid-Tang to Ming dynasties (ca. 800–1600 ᴄᴇ). In addition to its ritual uses, in China this dhāraṇī receives mention in poems and tales of miracles and is analyzed in philosophical commentaries.

The uṣṇīṣavijayā dhāraṇī also appears to have been popular in Dunhuang. A number of Tibetan manuscripts from Dunhuang include just the dhāraṇī on its own, both in Tibetan transliteration (dhāraṇīs, like mantras, are commonly left untranslated in Tibetan texts) and in Tibetan translation. The primary uṣṇīṣavijayā dhāraṇī text also appears in several Dunhuang manuscripts. Interestingly, several drawings from Dunhuang show maṇḍala (altar) arrangements corresponding to uṣṇīṣavijayā dhāraṇī texts, and one in particular, which has labels written in Chinese, depicts a maṇḍala that is nearly identical to a maṇḍala described in one of the rites in our text, even though no known ritual manual surviving in Chinese describes such a maṇḍala. There is also a woodblock print from tenth-century Dunhuang that has an image of Amitābha and a dhāraṇī written in Sanskrit, but with Chinese writing on the side. The Amitābha dhāraṇī from this print is very similar to (but not identical with) the second short dhāraṇī transmitted in the present text. The records of uṣṇīṣavijayā-related works at Dunhuang, then, suggest a close relationship between Tibetan- and Chinese-speaking Buddhist practitioners there.

In Nepal, uṣṇīṣavijayā dhāraṇī rituals continue to be performed as part of modern Newar Buddhist practice, where their practice is sometimes prescribed for Wednesdays in particular. Practices connected to the uṣṇīṣavijayā dhāraṇī likewise continue in modern Tibetan Buddhism. The so-called Tongchö (stong mchod)‍—the thousandfold offering practice of Uṣṇīṣavijayā, a version of which is mentioned briefly in our text‍—is currently performed in Tibetan monasteries, sometimes using a ritual manual composed by the nineteenth-century polymath Jamyang Khyentsé Wangpo. Other notable Tibetan works on the uṣṇīṣavijayā dhāraṇī and its associated practices include commentaries by the great Sakya lama Butön (bu ston rin chen grub, 1290–1364) and the fourth Panchen Lama, Losang Chökyi Gyaltsen (blo bzang chos kyi rgyal mtshan, 1570–1662).

The question of what, or who, exactly, Uṣṇīṣavijayā is is a complex one that cannot be clearly answered here. In short, like a number of uṣṇīṣa deities, she is sometimes identified as a protective deity, in this case a goddess, emanated from the Buddha’s uṣṇīṣa. Indeed, Uṣṇīṣavijayā is clearly depicted as a goddess in a number of short sādhanas included in Indian anthologies such as the Sādhanamāla, compiled from the works of many authors probably during the period of the Pāla kings (eighth to twelfth century). Three closely similar sādhanas of a three-faced, eight armed form of the goddess are included in the Tengyur, one in each of the three related anthologies translated from the Indian collections into Tibetan in the eleventh to fourteenth centuries respectively. A variety of other forms are depicted or described in Chinese, Japanese, Tibetan, Mongolian, and Kashmiri sources. In the later Tibetan tradition Uṣṇīṣavijayā can even appear as one of a group of three long-life deities along with the Buddha Amitāyus and White Tārā. However, in our Kangyur text, and indeed in all but one (Toh 598) of the uṣṇīṣavijayā works in this section of the Kangyur, while the dhāraṇī itself uses the feminine vocative form throughout, the name uṣṇīṣavijayā is not rendered into Tibetan in the feminine, and the word uṣṇīṣavijayā is not used to refer to anything apart from the name of the dhāraṇī‍—the dhāraṇī of the crown victory. The single instance in this text in which we could interpret uṣṇīṣavijayā to refer to a goddess is a sentence in one of the rites for an Amitāyus sādhana that also involves reciting the uṣṇīṣavijayā dhāraṇī. After having properly followed the rite, the text notes that “in the early morning at dawn you will see the face of the Bhagavatī, and she will give you whatever accomplishments you desire.” The name of this Bhagavatī is not mentioned in our text, but we might well presume that she is Uṣṇīṣavijayā herself.

The range of possible answers to the question of what the name Uṣṇīṣavijayā refers to is enlarged even further by the existence of a group of related texts widely used in Southeast Asia, sharing the Pali title Uṇhissa-vijaya-sutta (or in some cases simply Uṇhissa-vijaya) but found in a number of different forms, some in Pali but others in Siamese, Lao, Yuon, and Khmer. Some refer at least briefly to the story of the god Supratiṣṭhita (Pali Supatiṭṭhita) which, though somewhat secondary in the present text, is the central frame story of Toh 597. But instead of the dhāraṇī of the Sanskrit and Tibetan texts they contain a set of verses (gāthā) to be recited whose content is unrelated to that of the Sanskrit dhāraṇī. The gāthā are also found alone in several ritual compilations. Even in the vernacular versions, the verses are written in Pali. In these texts, in their own opening lines, it seems to be the verses themselves that are referred to as the Uṇhissa-vijaya.

This text was translated into Tibetan by the Indian paṇḍita Dharmasena and Khampa Lotsāwa Bari Chödrak, and it is therefore an eleventh- or twelfth-century translation. However, the imperial Phangthangma catalog lists one Uṣṇīṣavijayā-dhāraṇī-vidhisahitā, which, even if not the same as the present text, is certainly a work of a similar type. Thus, along with the records of Uṣṇīṣavijayā texts at Dunhuang, its presence in the Phangthangma catalog at the very least indicates the early presence of parts of the Uṣṇīṣavijayā corpus, including not just the dhāraṇī but also some of its associated rites, in Tibet.

The present translation is based on the Tibetan translation of the text found in the Tantra Collection (rgyud ’bum) section of the Degé Kangyur, in consultation with the Stok Palace Kangyur and the notes in the Comparative Edition of the Kangyur (dpe bsdur ma). The text is stable across all the Kangyurs consulted, with the same title and colophon and only minor variants. We also consulted the Sanskrit from Schopen’s transcription of the “Los Angeles Manuscript” of the primary dhāraṇī text and Hidas’ edition and translation of the surviving Sanskrit Uṣṇīṣavijayā-dhāraṇī text for the passages that are parallel with the present text.

The main dhāraṇī is not identical in every detail across the five different versions in the Degé Kangyur (Toh 594–598), and the existence of further variations across different Kangyurs and versions in extra-canonical collections further complicates the picture. Reference to the dhāraṇī as presented in Hidas’ edition of the Sanskrit yields useful orthographic confirmation, but may be misleading as a model given that the ten different Nepalese Sanskrit manuscripts on which it is based are of much later date than any of the present Tibetan witnesses. Here and in the other works in the group we have therefore chosen to transcribe the dhāraṇī as it appears in the Degé version of each text, making only minor choices of orthography and adding annotations to point out the most significant discrepancies.

One noticeable difference across both Tibetan and Sanskrit versions of the dhāraṇī is the presence or absence of the syllable oṁ at the beginning of certain phrases. In the present work and Toh 595, 596, and 597 there are only three such oṁ syllables, while in Toh 598 oṁ appears no less than nine times, as it does in Hidas’ edition from Sanskrit sources and in extra-canonical liturgies. The Tibetan translation of Toh 598 was made at a significantly later date than the other works of the group, and may possibly signal a change in usage that is also reflected in the Nepalese Sanskrit texts of even later date. This is corroborated by the absence of extra oṁ-s in the Dunhuang manuscripts. The colophon of Toh 597 found in the Phukdrak (phug brag) Kangyur includes a note claiming that the texts with only three oṁ-s are to be considered more correct. The note also states that although there may have been Sanskrit sources with as many as nine oṁ-s, the twelfth-century translator Sumpa Lotsāwa reported that all the Sanskrit texts he had seen contained only three, and that the Sanskrit manuscripts of the texts held at Sakya monastery had no more than that. Because Sumpa Lotsāwa is known to have lived and studied in Nepal, his comment on the “correct” number of oṁ-s in the Sanskrit manuscripts available to him offers a glimpse of the evolution of the text in the Nepalese tradition. As Hidas’ edition of the Nepalese manuscripts suggests, the number of oṁ-s in the dhāraṇī seem to proliferate, eventually reaching a total of nine.

Over the centuries, the textual transmission of the dhāraṇī has preserved the major portion of it with remarkable fidelity. Nevertheless, the few anomalies to be seen across all these closely related texts are a reminder that here, as with other dhāraṇī works, some variations over time and place are to be expected.

Summary

The Uṣṇīṣavijayā Dhāraṇī with Its Ritual Manual is a short work in which the Buddha Amitāyus teaches the uṣṇīṣavijayā dhāraṇī along with its benefits and a number of short rites for its recitation.

Acknowledgements

This text was translated by Catherine Dalton, who also wrote the introduction.

The translation was completed under the patronage and supervision of 84000: Translating the Words of the Buddha.

The generous sponsorship of May, George, Likai, and Lillian Gu, which helped make the work on this translation possible, is most gratefully acknowledged.

Introduction

The Uṣṇīṣavijayā Dhāraṇī with Its Ritual Manual opens in Sukhāvatī, where the Blessed One Amitāyus is residing. Amitāyus addresses the bodhisattva Avalokiteśvara, informing him that there are beings who suffer from illnesses and short lifespans, and introducing the uṣṇīṣavijayā dhāraṇī as a remedy for such painful circumstances. Avalokiteśvara immediately asks Amitāyus to pronounce the dhāraṇī, which the Tathāgata does from within a state of samādhi.

After he pronounces the dhāraṇī, Amitāyus addresses Śakra and explains that reciting it purifies all lower rebirths and obscurations and leads beings to higher rebirths. Amitāyus lists a number of lower births that will be avoided by someone who recites this dhāraṇī, as well as the various prominent families within the human realm into which the reciter will be born on their way to attaining the state of awakening. As he continues to elaborate upon the ways in which one can use the dhāraṇī to benefit beings, Amitāyus explains that it can be hung from the top of a flagstaff, a mountain, or a tall building so that those in the surrounding areas will be purified and prevented from taking lower rebirths. He also notes that it may be carved into wood, placed at a caitya, and worshiped. This inspires Yama, the Lord of Death, to pledge to Amitāyus to protect beings who recite the dhāraṇī.

The Four Great Kings then request instructions on the rites associated with the dhāraṇī, and Amitāyus proceeds to teach a number of rites, including ways the dhāraṇī can be used to benefit the dead, animals, and those who are ill, as well as to prevent lower rebirths and extend life. These rituals involve the preparation of a maṇḍala in which the dhāraṇī is installed, making various offerings, the construction of caityas, and hundreds, thousands, or even a hundred thousand recitations of the dhāraṇī. In some cases the rituals involve fasting or performing the ritual at specific times of the lunar month. Śakra is instructed to teach one of these rites to the god Supratiṣṭhita to extend his life, which he does, and the god is delighted with the results. Finally, Amitāyus teaches his own dhāraṇī along with its related sādhana and fire offering rite, along with a number of additional applications of the dhāraṇī to obtain results such as glory and kingship.

This work is one among a group of texts in the Kriyātantra section of the Tibetan Kangyurs that contain the uṣṇīṣavijayā dhāraṇī and its related rituals (kalpa). The present text is the longest of four short dhāraṇī texts‍—three of which have the same title‍—that present the uṣṇīṣavijayā dhāraṇī with its ritual manual (kalpa). These four works share a similar narrative opening (nidāna) up through the presentation of the dhāraṇī proper, and several among them also share additional passages. Beyond its close relationship with the other ritual manuals connected with the uṣṇīṣavijayā dhāraṇī, this text shares a number of parallel passages with what we will call‍—simply for the purpose of distinguishing it from the present group of dhāraṇī-kalpas‍—the “primary” uṣṇīṣavijayā dhāraṇī text (Toh 597, which is titled Sarva­durgati­pariśodhana-uṣṇīṣavijayā- dhāraṇī rather than Uṣṇīṣavijayā-dhāraṇī-kalpasahitā). Despite these significant textual parallels, that text has an entirely different opening narrative and, unsurprisingly, lacks nearly all of the ritual instructions contained in our text.

There are many Sanskrit witnesses of the uṣṇīṣavijayā dhāraṇī proper, and the primary uṣṇīṣavijayā dhāraṇī text (Toh 597) survives in at least one, an incomplete early manuscript. While our text appears to no longer be extant in Sanskrit, there is at least one surviving Sanskrit uṣṇīṣavijayā dhāraṇī work that is closely related to it and belongs to the same group of related dhāraṇī texts described above. This work shares the same opening narrative and some of the ritual material with the texts from this group.

The primary uṣṇīṣavijayā text (Toh 597) was first translated into Chinese by Buddhapāli in the late seventh century, and then at least five times subsequently. Several ritual manuals for the dhāraṇī’s recitation were also translated into Chinese, but our text does not appear to be among them. One ritual manual (Taishō 978), translated into Chinese by Dharmadeva between 973 and 981, is among the group of uṣṇīṣavijayā dhāraṇī texts to which the present work belongs. The primary uṣṇīṣavijayā text was significant in East Asia, and one scholar has even identified it as the most important esoteric Buddhist scripture translated into Chinese in the seventh century. Practices connected with the uṣṇīṣavijayā dhāraṇī were important in China, in particular in conjunction with funerary rites, where the dhāraṇī was written on pillars near tombs, especially from the mid-Tang to Ming dynasties (ca. 800–1600 ᴄᴇ). In addition to its ritual uses, in China this dhāraṇī receives mention in poems and tales of miracles and is analyzed in philosophical commentaries.

The uṣṇīṣavijayā dhāraṇī also appears to have been popular in Dunhuang. A number of Tibetan manuscripts from Dunhuang include just the dhāraṇī on its own, both in Tibetan transliteration (dhāraṇīs, like mantras, are commonly left untranslated in Tibetan texts) and in Tibetan translation. The primary uṣṇīṣavijayā dhāraṇī text also appears in several Dunhuang manuscripts. Interestingly, several drawings from Dunhuang show maṇḍala (altar) arrangements corresponding to uṣṇīṣavijayā dhāraṇī texts, and one in particular, which has labels written in Chinese, depicts a maṇḍala that is nearly identical to a maṇḍala described in one of the rites in our text, even though no known ritual manual surviving in Chinese describes such a maṇḍala. There is also a woodblock print from tenth-century Dunhuang that has an image of Amitābha and a dhāraṇī written in Sanskrit, but with Chinese writing on the side. The Amitābha dhāraṇī from this print is very similar to (but not identical with) the second short dhāraṇī transmitted in the present text. The records of uṣṇīṣavijayā-related works at Dunhuang, then, suggest a close relationship between Tibetan- and Chinese-speaking Buddhist practitioners there.

In Nepal, uṣṇīṣavijayā dhāraṇī rituals continue to be performed as part of modern Newar Buddhist practice, where their practice is sometimes prescribed for Wednesdays in particular. Practices connected to the uṣṇīṣavijayā dhāraṇī likewise continue in modern Tibetan Buddhism. The so-called Tongchö (stong mchod)‍—the thousandfold offering practice of Uṣṇīṣavijayā, a version of which is mentioned briefly in our text‍—is currently performed in Tibetan monasteries, sometimes using a ritual manual composed by the nineteenth-century polymath Jamyang Khyentsé Wangpo. Other notable Tibetan works on the uṣṇīṣavijayā dhāraṇī and its associated practices include commentaries by the great Sakya lama Butön (bu ston rin chen grub, 1290–1364) and the fourth Panchen Lama, Losang Chökyi Gyaltsen (blo bzang chos kyi rgyal mtshan, 1570–1662).

The question of what, or who, exactly, Uṣṇīṣavijayā is is a complex one that cannot be clearly answered here. In short, like a number of uṣṇīṣa deities, she is sometimes identified as a protective deity, in this case a goddess, emanated from the Buddha’s uṣṇīṣa. Indeed, Uṣṇīṣavijayā is clearly depicted as a goddess in a number of short sādhanas included in Indian anthologies such as the Sādhanamāla, compiled from the works of many authors probably during the period of the Pāla kings (eighth to twelfth century). Three closely similar sādhanas of a three-faced, eight armed form of the goddess are included in the Tengyur, one in each of the three related anthologies translated from the Indian collections into Tibetan in the eleventh to fourteenth centuries respectively. A variety of other forms are depicted or described in Chinese, Japanese, Tibetan, Mongolian, and Kashmiri sources. In the later Tibetan tradition Uṣṇīṣavijayā can even appear as one of a group of three long-life deities along with the Buddha Amitāyus and White Tārā. However, in our Kangyur text, and indeed in all but one (Toh 598) of the uṣṇīṣavijayā works in this section of the Kangyur, while the dhāraṇī itself uses the feminine vocative form throughout, the name uṣṇīṣavijayā is not rendered into Tibetan in the feminine, and the word uṣṇīṣavijayā is not used to refer to anything apart from the name of the dhāraṇī‍—the dhāraṇī of the crown victory. The single instance in this text in which we could interpret uṣṇīṣavijayā to refer to a goddess is a sentence in one of the rites for an Amitāyus sādhana that also involves reciting the uṣṇīṣavijayā dhāraṇī. After having properly followed the rite, the text notes that “in the early morning at dawn you will see the face of the Bhagavatī, and she will give you whatever accomplishments you desire.” The name of this Bhagavatī is not mentioned in our text, but we might well presume that she is Uṣṇīṣavijayā herself.

The range of possible answers to the question of what the name Uṣṇīṣavijayā refers to is enlarged even further by the existence of a group of related texts widely used in Southeast Asia, sharing the Pali title Uṇhissa-vijaya-sutta (or in some cases simply Uṇhissa-vijaya) but found in a number of different forms, some in Pali but others in Siamese, Lao, Yuon, and Khmer. Some refer at least briefly to the story of the god Supratiṣṭhita (Pali Supatiṭṭhita) which, though somewhat secondary in the present text, is the central frame story of Toh 597. But instead of the dhāraṇī of the Sanskrit and Tibetan texts they contain a set of verses (gāthā) to be recited whose content is unrelated to that of the Sanskrit dhāraṇī. The gāthā are also found alone in several ritual compilations. Even in the vernacular versions, the verses are written in Pali. In these texts, in their own opening lines, it seems to be the verses themselves that are referred to as the Uṇhissa-vijaya.

This text was translated into Tibetan by the Indian paṇḍita Dharmasena and Khampa Lotsāwa Bari Chödrak, and it is therefore an eleventh- or twelfth-century translation. However, the imperial Phangthangma catalog lists one Uṣṇīṣavijayā-dhāraṇī-vidhisahitā, which, even if not the same as the present text, is certainly a work of a similar type. Thus, along with the records of Uṣṇīṣavijayā texts at Dunhuang, its presence in the Phangthangma catalog at the very least indicates the early presence of parts of the Uṣṇīṣavijayā corpus, including not just the dhāraṇī but also some of its associated rites, in Tibet.

The present translation is based on the Tibetan translation of the text found in the Tantra Collection (rgyud ’bum) section of the Degé Kangyur, in consultation with the Stok Palace Kangyur and the notes in the Comparative Edition of the Kangyur (dpe bsdur ma). The text is stable across all the Kangyurs consulted, with the same title and colophon and only minor variants. We also consulted the Sanskrit from Schopen’s transcription of the “Los Angeles Manuscript” of the primary dhāraṇī text and Hidas’ edition and translation of the surviving Sanskrit Uṣṇīṣavijayā-dhāraṇī text for the passages that are parallel with the present text.

The main dhāraṇī is not identical in every detail across the five different versions in the Degé Kangyur (Toh 594–598), and the existence of further variations across different Kangyurs and versions in extra-canonical collections further complicates the picture. Reference to the dhāraṇī as presented in Hidas’ edition of the Sanskrit yields useful orthographic confirmation, but may be misleading as a model given that the ten different Nepalese Sanskrit manuscripts on which it is based are of much later date than any of the present Tibetan witnesses. Here and in the other works in the group we have therefore chosen to transcribe the dhāraṇī as it appears in the Degé version of each text, making only minor choices of orthography and adding annotations to point out the most significant discrepancies.

One noticeable difference across both Tibetan and Sanskrit versions of the dhāraṇī is the presence or absence of the syllable oṁ at the beginning of certain phrases. In the present work and Toh 595, 596, and 597 there are only three such oṁ syllables, while in Toh 598 oṁ appears no less than nine times, as it does in Hidas’ edition from Sanskrit sources and in extra-canonical liturgies. The Tibetan translation of Toh 598 was made at a significantly later date than the other works of the group, and may possibly signal a change in usage that is also reflected in the Nepalese Sanskrit texts of even later date. This is corroborated by the absence of extra oṁ-s in the Dunhuang manuscripts. The colophon of Toh 597 found in the Phukdrak (phug brag) Kangyur includes a note claiming that the texts with only three oṁ-s are to be considered more correct. The note also states that although there may have been Sanskrit sources with as many as nine oṁ-s, the twelfth-century translator Sumpa Lotsāwa reported that all the Sanskrit texts he had seen contained only three, and that the Sanskrit manuscripts of the texts held at Sakya monastery had no more than that. Because Sumpa Lotsāwa is known to have lived and studied in Nepal, his comment on the “correct” number of oṁ-s in the Sanskrit manuscripts available to him offers a glimpse of the evolution of the text in the Nepalese tradition. As Hidas’ edition of the Nepalese manuscripts suggests, the number of oṁ-s in the dhāraṇī seem to proliferate, eventually reaching a total of nine.

Over the centuries, the textual transmission of the dhāraṇī has preserved the major portion of it with remarkable fidelity. Nevertheless, the few anomalies to be seen across all these closely related texts are a reminder that here, as with other dhāraṇī works, some variations over time and place are to be expected.