Introduction

Overview

i.1In this profound Mahāyāna sūtra, The Dhāraṇī of the Jewel Torch, the Buddha Śākyamuni explains, with the help of the bodhisattvas Mañjuśrī, Samanta­bhadra, and Dharma­mati, how bodhisattvas progress toward awakening.

i.2Although seen as a sūtra in its own right, it is closely connected to the family of texts belonging to the Avataṃsakasūtra, two chapters of which it shares. As its title suggests, it can also be seen as a dhāraṇī, or as a sūtra about a dhāraṇī.

i.3Substantial passages were quoted by Śāntideva in the Śikṣāsamuccaya, and these extracts are now the only remnants of the Sanskrit text. The Tibetan translation, by the Indian master Surendra­bodhi and the chief-editor translator the monk Yeshé Dé, dates to the early, imperial translation period, and its verses on faith later had a wide impact in Tibetan works. The Chinese translation, by Fatian, dates to the late tenth century and is classified as an Avataṃsaka text.

Narrative and Doctrinal Content

i.4The setting of the text is the Vulture Peak in Rājagṛha. Its audience is a great gathering of highly accomplished monks and bodhisattvas, headed by Samanta­bhadra who, as the initial interlocutor, asks the Buddha how dharmadhātu should be understood. A brief but profound exchange follows.

i.5Mañjuśrī then appears and requests the Buddha to teach the dhāraṇī of the jewel torch (1.­11). The Buddha insists that Mañjuśrī should request Samanta­bhadra to teach it instead, and Samanta­bhadra’s dialog with Mañjuśrī starts with the meaning of buddha. A brief interlude follows (1.­40–1.­54) in which Śāriputra (here Śāradvatī­putra) compares his own understanding unfavorably with Mañjuśrī’s vast wisdom, and professes his inability and unwillingness to debate with him; similar brief conversations between Śāriputra and Mañjuśrī recur at several points in the text.

i.6The bodhisattva Dharma­mati then makes his appearance (1.­55) and enters the meditative absorption called the infinite application of the bodhisattva’s jewel torch . Blessed and encouraged by millions of buddhas to summon the eloquence to teach, Dharma­mati sets out the ten categories of bodhisattva (1.­59–1.­84) in the long passage that follows. A number of wonders then occur, after which Dharma­mati summarizes the ten categories in verse (1.­88–1.­178).

i.7The Buddha, in response to several ensuing requests to teach, briefly teaches the mantra of the dhāraṇī (1.­213) and comments on its meaning. At Samanta­bhadra and Mañjuśrī’s request, he then explains the benefits that hearing this sūtra will have for future disciples (1.­228–1.­256). Here several stark warnings are given to future hearers (mainly future monks) who might one day criticize or reject this sūtra.

i.8The final main section of the sūtra is a very long passage (twenty folios in the Degé edition) of versified encouragement for bodhisattvas, introduced by a prologue featuring Ānanda. The main versified part (2.­20–2.­397) is spoken by the Buddha as a narrative that introduces, relates, and concludes teachings given by the bodhisattva Bhadraśrī on the good qualities and modes of conduct of the bodhisattva. Bhadraśrī first eulogizes the thought of awakening (bodhicitta) and then a few verses later praises faith in a well-known passage, parts of which were quoted by Śāntideva (see below) and subsequently by many Tibetan authors. Bhadraśrī then describes many of the other qualities of bodhisattvas and their ability to manifest miraculously in different ways, including the astounding visual and other sensory content of their meditative absorptions, the many kinds of miraculous rays of light with which they bring benefit to the world and beings, and comparisons with the powerful magical displays of the ordinary gods such as Indra and the king of the nāgas.

i.9At the end of the teaching, its importance and future benefits are expressed by Subhūti and others, and Ānanda promises to retain it.

The Sūtra, the Avataṃsaka, and the Chinese Translation

i.10Although it is found in the Kangyur among other Mahāyāna sūtras in the General Sūtra section (as Tōh 145 in the Degé Kangyur) and is listed as belonging to that general category in the Denkarma inventory of translated texts (as well as to the Dhāraṇī section, see below), the sūtra also belongs to the family of texts related to the Avataṃsakasūtra (phal po che, “A Multitude of Buddhas,” Toh 44). Indeed, in the other imperial period inventory, the roughly contemporary Phangthangma, it is listed under the heading of “the works included in the group of sūtras of the noble, great, very extensive Buddhāvataṃsaka.”

i.11The prominent role of the bodhisattva Samanta­bhadra; the centrality of the dharmadhātu; the vast numbers of buddhas who gather and the mention of the Buddha Vairocana in the pivotal passage about the absorption and blessing of Dharma­mati; the tenfold division and subdivision of the categories of bodhisattva; the repeated vocative “O sons of the victors”; and many other features of this work, above all the central theme of how bodhisattvas first emerge in the presence of a tathāgata and progressively develop access to the buddha qualities, culminating in their regency and consecration, are all strongly reminiscent of the Avataṃsaka.

i.12Two long passages in the text represent two complete chapters of the Avataṃsaka­sūtra. In terms of content they are close to being exact matches, although the translations in Tibetan are different. The long passage recounting Dharmamati’s absorption and his ensuing revelations in both prose and verse (from 1.­55 to 1.­178) is almost identical to the whole of chapter 20 of the Tibetan Avataṃsaka­sūtra, “The Ten Categories of Bodhisattvas” (chapter 15 of the Chinese), while almost the entirety of the final verse section recounting the teachings by Bhadraśrī (from 2.­27 to 2.­397 near the conclusion of the text) matches the whole of chapter 17 of the Tibetan Avataṃsaka­sūtra, “Bhadraśrī” (chapter 12 of the Chinese).

i.13The Chinese translation of this text, Taishō 299, made by Fatian almost a hundred years later than the Tibetan, in the year 983, is also classified as a sūtra of the Avataṃsaka family. It is placed in the Taishō in the Huayan volume, volume 10, along with the Avataṃsaka­sūtra itself and the other standalone texts related to it.

Why Is the Sūtra Also a Dhāraṇī?

i.14The text is classified not only as a sūtra, but also as a dhāraṇī, and in those Kangyurs that have an additional Dhāraṇī section it is duplicated there (as Tōhoku no. 847 in the Degé Kangyur). Indeed, the title itself includes the word dhāraṇī, and the teaching requested of the Buddha is referred to as “the dhāraṇī of the jewel torch.”

i.15The term dhāraṇī is derived from the Sanskrit root √dhṛ (“to hold” or “to maintain”), and among its wide range of meanings most are closely related to the retaining‍—in the mind, in memory, in words, or in writing‍—of a particular teaching, realization, or approach to awakening. Perhaps the two most widespread senses in which the term is used are when it refers to a mantra-like formula that “encodes” its meaning without necessarily expressing it in comprehensible speech, or when it describes the highly developed capacity of advanced practitioners to memorize and accurately retain a set of detailed and profound instructions. But as well as signifying the means by which such meanings or sets of instructions are retained (i.e., what holds them), it can also designate a specific meaning or instruction itself (i.e., what is held).

i.16Furthermore, by extension from these senses of the term, a text that either contains a (mantra-like) dhāraṇī, or is about a dhāraṇī in any of these senses, may itself be referred to as a dhāraṇī. This is the basis for the term dhāraṇī having also come to designate a whole scriptural genre of Mahāyāna texts‍—well represented in the Kangyur, which contains some two hundred fifty texts in that category. However, as a genre it is both quite diverse in its composition and shares most of the texts it contains with other genres. It is often not entirely clear whether any one text is labeled a dhāraṇī because the text itself is a dhāraṇī, contains a dhāraṇī, or is about a dhāraṇī.

i.17For all these reasons, each text placed in this genre deserves its own analysis of what makes it “a dhāraṇī.” In the case of the present text, mentions are made throughout to a “dhāraṇī of the jewel torch,” but it is difficult to determine whether they all have the same reference, or whether they variously refer to a particular realization of bodhisattvas, to a teaching on that realization, or to the text itself.

i.18In the first chapter, there are four separate occasions on which the dhāraṇī seems to be taught. Although the corresponding mentions could conceivably all be understood as referring to one and the same instance of the dhāraṇī, three of the four occasions end with a statement that the dhāraṇī has now been taught, in the past tense. In the first of the four instances, the exchange between Mañjuśrī, Samanta­bhadra, and the Buddha (starting at 1.­11) is termed an explanation of the dhāraṇī in the initial request. In the second instance, Dharma­mati’s long teaching on the ten levels of bodhisattvas is also described as a dhāraṇī immediately afterward by Samanta­bhadra (1.­179). The third instance is a dialog between Mañjuśrī and Śāri­putra (starting at 1.­196) in response to the latter’s request for an explanation of the dhāraṇī, which is lauded as a teaching on that dhāraṇī afterward (1.­205). The fourth instance is when yet another request is made to the Buddha, this time by Samanta­bhadra, to teach the dhāraṇī (1.­211); the Buddha teaches what is described as a mantra, and in the discussions that follow it is made clear that the meaning it carries is that of the ineffable ultimate nature of reality.

i.19Along with dhāraṇīs, a number of sūtras mention gateways (Skt. mukha, Tib. sgo), meditative absorptions (Skt. samādhi, Tib. ting nge ’dzin), and liberations (Skt. vimokṣa, Tib. rnam par thar pa) as different kinds of qualities attained by bodhisattvas. That some of the mentions of the dhāraṇī in this sūtra fall into the category of such attained qualities is suggested by the first of the four instances instance here being also termed an “access” or gateway (1.­34), and by the second instance being described as arising from the gnosis that Dharma­mati has attained while immersed in a meditative absorption called “the infinite application of the jewel torch.” Nevertheless, this second instance, the long teaching on the ten levels of bodhisattvas, is clearly also seen as a teaching, in the sense of presenting specific doctrinal content. The third instance is heralded by Śāriputra’s announcement that a sūtra is to be taught, yet what happens turns out to be a short and somewhat cryptic dialog equating explanation with emptiness, and demonstrating how neither can be taught. Only the fourth instance, the mantra, can be reasonably clearly placed in the category of dhāraṇīs that are encoding formulae, and the meaning that the mantra can be assumed to express is linked to the first and third instances in the teaching by Samanta­bhadra that follows it, on how the dhāraṇī should be “retained” and cultivated as a teaching on thatness, the ultimate (Skt. tathatā, Tib. de bzhin nyid).

i.20Most of the subsequent mentions of the dhāraṇī as such, in what remains of the first chapter and at the beginning of the second (it is not mentioned at all in the long verse section), are made in the context of its future holders and of its past history, intermingled with descriptions of it as a Dharma discourse. In other words, as a teaching‍—but also, in the kind of internal self-reference that is a common feature of many Mahāyāna sūtras, designating this very text itself.

i.21The frequent mentions in this text of the “dhāraṇī of the jewel torch” are therefore quite varied in terms of the sense in which the term is being used. We have made no attempt to use capitalization or punctuation to distinguish those that may refer to the text itself, to a teaching, to the mantra, or to a realization.

i.22Neither of the two long sections that appear as chapters in the Avataṃsaka­sūtra make any mention of a dhāraṇī. None of the excerpts in Sanskrit quoted by Śāntideva (see below) include passages where the dhāraṇī is mentioned in the Tibetan text, but the title Śāntideva uses to introduce his citations does include the designation dhāraṇī.

The Title and Its Variants

i.23The Sanskrit title transliterated in the Tibetan text, Ratnolkādhāraṇī in its short form, is the same as the title that appears in the Sanskrit manuscripts of the Śikṣāsamuccaya (see below). The Sanskrit ulkā can mean a fiery phenomena in the sky, i.e., a meteor, and also a firebrand or torch.

i.24Of the title in Tibetan, however, there are several different renderings. In all Kangyurs, the title is dkon mchog ta la la’i gzungs, incorporating the unusual, archaic word ta la la, meaning “lamp” or “torch.” In some of the twenty or so Tengyur treatises that quote the text (including the Tibetan translation of the Śikṣāsamuccaya), the ta la la title is used, even if in some cases the word gzungs (dhāraṇī) is dropped or replaced by the word mdo (sūtra). In others, however, the title is rendered in various forms that use, instead of ta la la, the more usual Tibetan term for “lamp” or “torch,” sgron ma or sgron me. Probably as a consequence, later Tibetan authors of indigenous works (see below) use sometimes one version of the title, sometimes the other, and only some authors who use the sgron ma variants seem to be aware that the canonical work they are quoting is in the Kangyur under a different title.

The Sūtra in Śāntideva’s Śikṣāsamuccaya and Other Treatises

i.25The sūtra is quoted a little over twenty times in treatises in the Tengyur, notably by Atiśa, Vimalamitra, and Śāntideva, but also by lesser known authors. As noted above, both the dkon mchog ta la la and dkon mchog sgron ma forms of the title can be found, and there are considerable minor variations. Most, but not all, of the quotations are from the long verse section of the second chapter.

i.26The most extensive extracts appear in Śāntideva’s Śikṣāsamuccaya (Training Anthology), and indeed the sūtra appears to have been among Śāntideva’s favorite texts, as he quotes from it more than from any other work. His text contains two short extracts, one longer passage, and one very extensive section of the verses from the second chapter that makes up more than half of one of his chapters. The Śikṣāsamuccaya has survived in Sanskrit, as well as in its Tibetan translation in the Tengyur, and its Sanskrit text thus contains the only known remnants of the sūtra in Sanskrit.

The Sūtra’s Impact on Tibetan Works

i.27The sūtra is listed in the Mahāvyutpatti as one of the hundred or so Dharma texts that were presumably best known at the time, and is frequently quoted by Tibetan authors of all the main traditions. The passages on the importance of faith are the most commonly quoted, and for some authors it is the scriptural source for there being‍—variously‍—three, four, or six kinds of faith. Other parts of both chapters are also cited.

i.28Identifying quotes from the sūtra is made more difficult by the variety of titles used. In the case of several authors, including Chomden Rikpa Raltri (bcom ldan rig pa ral dri, thirteenth century), Longchen Rabjampa (klong chen rab ’byams pa, fourteenth century), and many of the early Sakya scholars, quotes using both the dkon mchog sgron ma and the dkon mchog ta la la forms of the title can be found in the same work, suggesting that in some cases they may have been consulting treatises or other sources that used these different titles as well as the canonical text itself without always recognizing that both titles designate the same sūtra. Shākya Chokden (shA kya mchog ldan, fifteenth century) specifically mentions the identity of both titles.

The Translation

i.29This translation is based principally on the Degé block print and the Comparative Edition (dpe bsdur ma) of the Kangyur. Yeshé Dé’s early-ninth-century translation contains a few archaic words that have survived subsequent editing, including the ta la la in the title, mentioned above. A few other noteworthy archaic spellings, recorded in the notes, are byin as a verb of the Buddha’s speech (see 1.­31); dbung, “center” (see 1.­84); and the spelling nod pa for mnod pa (prahaṇam, “to receive”). In a few passages we have suggested a change in the text reading in an endnote, often in consultation with the Stok Palace version.