Summary

The Perfection of Wisdom in Twenty-Five Thousand Lines is among the most important scriptures underlying both the “vast” and the “profound” approaches to Buddhist thought and practice. Known as the “middle-length” version, being the second longest of the three long Perfection of Wisdom sūtras, it fills three volumes of the Kangyur. Like the two other long sūtras, it records the major teaching on the perfection of wisdom given by the Buddha Śākyamuni on Vulture Peak, detailing all aspects of the path to enlightenment while at the same time emphasizing how bodhisattvas must put them into practice without taking them‍—or any aspects of enlightenment itself‍—as having even the slightest true existence.

Acknowledgements

Translation by the Padmakara Translation Group. A complete draft by Gyurme Dorje was first edited by Charles Hastings, then revised and further edited by John Canti. The introduction was written by John Canti. We are grateful for the advice and help received from Gareth Sparham, Greg Seton, and Nathaniel Rich.

This translation is dedicated to the memory of our late colleague, long-time friend, and vajra brother Gyurme Dorje (1950–2020), who worked assiduously on this translation in his final years and into the very last months of his life. We would also like to express our gratitude to his wife, Xiaohong, for the extraordinary support she gave him on so many levels.

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

The generous sponsorship of Kris Yao and Xiang-Jen Yao, which helped make the work on this translation possible, is most gratefully acknowledged.

Introduction

The Perfection of Wisdom in Twenty-Five Thousand Lines is one of the three so-called “long” sūtras on the Perfection of Wisdom, or Prajñā­pāramitā. It fills three complete volumes of the Degé Kangyur, and of all the Prajñā­pāramitā sūtras it is second in length only to the massive Perfection of Wisdom in One Hundred Thousand Lines (Toh 8), which fills twelve volumes. The third and shortest of the three “long” sūtras, the Perfection of Wisdom in Eighteen Thousand Lines (Toh 10), fills two and a half volumes.

All three sūtras have a similar structure, closely parallel content, and convey the same teaching: detailed presentations of everything that causes, conditions, and propagates the state of suffering, and of everything that either brings about, or is constituted by, the awakening from that state of suffering. Pervading these presentations is the constant message characteristic of the Prajñā­pāramitā sūtras: that despite the importance of understanding all the phenomena of defilement and of putting into practice all the phenomena of the path leading to purification, none of the phenomena that are known, practiced, or attained must ever be taken as having any ultimately real existence if true awakening is to be attained.

The phenomena that are comprised by defilement and purification are grouped in the numerous sets or lists (mātṛkā or “matrices”) that have characterized Buddhist teachings from their earliest origins. These sets and subsets of phenomena (dharma) became, of course, a particular feature of the systematizing Abhidharma texts, but in parallel also came to constitute an important component of the Prajñā­pāramitā sūtras. Their inclusion, which in the “long” sūtras is even more extensive than in the earlier Perfection of Wisdom in Eight Thousand Lines (Toh 12), may have served to confirm the continuity of the Prajñā­pāramitā’s more profound and difficult perspective on Buddhist practice with earlier forms, as well as to highlight its distinctiveness. In each of the three long sutras, the discussions between the Buddha and the other interlocutors on these sets of dharmas follow one another in almost identical sequence, and the differences in length among the three are almost entirely due to the different degrees to which each set is unpacked into the individual items that it comprises.

While the exhaustive presentation of dharmas that these texts contain provides an important scriptural basis for the classic Buddhist understanding and categorization of phenomena in general, their intent goes much farther than the descriptive itemizing characteristic of the Abhidharma. Most obviously, in his dialectical treatment of each topic the Buddha explicitly undermines any tendencies on the part of his disciples, however subtle they may be, to take any dharma as real or truly existing, or even to adopt it as a point of reference. It is from this “profound” theme of the Prajñā­pāramitā sūtras that the fullest forms of Buddhist philosophical understanding of emptiness developed, as exemplified by the great Madhyamaka treatises of Nāgārjuna and later scholars; it is also this theme that led to the practice traditions of Chan, Thiền, Sŏn, and Zen.

Rather less obviously, for it remains largely implicit (especially in the progressive order in which he sets them out), the Buddha’s instructions on how to overcome and abandon those tendencies form an integrated series of practices‍—the actual path that the bodhisattva must take toward complete awakening‍—and by extension a description of how a practitioner’s mind may be progressively brought to a direct realization of the ultimate. These “vast” instructions were explained to Asaṅga by Maitreya and are an underlying element of the great Yogācāra treatises as well as the key by which many later Indian commentaries explain the Prajñā­pāramitā sūtras.

The Early Spread of the Prajñā­pāramitā Sūtras

Buddhist tradition holds that the Prajñā­pāramitā was taught by the Buddha on Gṛdhrakūṭa (“Vulture Peak”), the craggy hill near the city of Rājagṛha, capital of the Indian kingdom of Magadha. Some accounts say that this teaching took place over a period of some thirty years, while others say twelve years. Nevertheless, not only the three long sūtras but all five of the longer so-called “mother sūtras” are traditionally said to be accounts of the same teaching by the Buddha given on a single occasion. This assertion is made on the basis of two points: first that the Buddha’s interlocutors are the same, and second that a prophecy made by the Buddha included in all of them could only have been made once. Indeed, the multiplicity of versions of this single teaching is not even limited to the five long sūtras in the canon, for several even longer versions are said to have been recorded for the needs of nonhuman beings. While the longest of the versions preserved in the Kangyur, The Perfection of Wisdom in One Hundred Thousand Lines (Toh 8), is said to have been destined for the nāgas‍—from whose realm it was retrieved by Nāgārjuna‍—it is also said that there is a sūtra for the gods with ten million lines, and that the longest version of all, destined for the gandharvas, is a sūtra in one billion lines.

From a historical perspective, versions of Prajñā­pāramitā sūtras in writing seem to have first appeared in the first centuries ʙᴄᴇ and ᴄᴇ. The sūtra’s own exhortation to readers to write it out in the form of books may be associated with the early Mahāyāna’s embrace of written texts. Modern scholars have disagreed about which of the geographically dispersed Buddhist communities may have first given rise to the Prajñā­pāramitā literature, some favoring its origin among the Mahāsāṅghikas of Andhra in the south of India, while others point to evidence of its early flourishing in the northwest regions such as Gandhāra. Whichever may be the case, a birch-bark scroll from the northwest, found in Bajaur (a district of present-day Pakistan near the Afghan border), radiocarbon-dated to the first century ᴄᴇ, is the oldest known Prajñā­pāramitā manuscript. Like most of the earliest birch-bark scrolls from the region it is in the Gāndhārī language, written in Kharoṣṭhī script. The fragmentary portion of the Prajñā­pāramitā sūtra it contains cannot be matched to any recension among the known versions of the sūtras, and may therefore represent an intermediate stage during their evolution from even earlier Prakrit versions into the differentiated sūtras we know today, in their Chinese and Tibetan translations and in the later Sanskrit texts preserved mainly in Nepal.

Different hypotheses regarding the steps and order in which that evolution took place have been proposed, one being that The Perfection of Wisdom in Eight Thousand Lines (Aṣṭasāhasrikā) represents an early version that first led to further long “mother” versions by a process of expansion in the first three centuries ᴄᴇ, and subsequently to the many short versions‍—notably the “Heart” and “Diamond Cutter” sūtras (Hṛdaya and Vajracchedikā)‍—by an opposite contraction from the fourth to the sixth century, after which even more compressed and sometimes “tantric” versions appeared, as well as texts in which the Prajñā­pāramitā is worshiped as a female “mother” deity. The neatness of this schema, however, may overlook some of the evidence. For instance, there are reasons for suspecting that the Vajracchedikā may date to as early as the second century ᴄᴇ.

It is important to bear in mind that the naming of the different versions by the number of lines they contain is likely to have been a later development, applied as a means of classifying the profusion of circulating texts of different lengths. It was already in use by the time these texts were first translated into Tibetan in the late eighth and early ninth centuries, but it is not a feature of the earlier Chinese translations. The earliest evidence of this nomenclature appears to be in the Chinese literature, in the record of a lecture by the early sixth century translator Bodhiruci, and its widespread adoption in the centuries that followed may have served to limit further profusion and even reduce the variety of different texts by fixing their number. Nevertheless, its retrospective application to earlier texts may obscure rather than clarify their recensional affinities, and the comparison of parallel passages across recensions of all the sūtras remains a valid means of exploring qualitative textual differences regardless of the quantitative affiliation defined by the particular title.

This is particularly true of the longest versions of the sūtra, i.e., the sūtras in Ten Thousand Lines, Eighteen Thousand Lines, Twenty-Five Thousand Lines, and One Hundred Thousand Lines, which are better seen as a group, often termed “the Long Prajñā­pāramitā Sūtras.” It includes texts that exist variably in Sanskrit, Chinese, and Tibetan, different recensional relationships between which can be distinguished independently of the “length” denominations into which they are categorized.

The Perfection of Wisdom in Twenty-Five Thousand Lines in Central Asia and China

Coming to The Perfection of Wisdom in Twenty-Five Thousand Lines itself, we do not know‍—apart from placing it with reasonable probability within the process of the expansion of longer texts as mentioned above‍—when or where it first appeared as a distinct redaction of the Prajñā­pāramitā sūtra, or even in what language. As is the case with many canonical works, the earliest historical mentions predate any surviving physical texts, and come from accounts of its first translation into Chinese.

Perhaps unsurprisingly, physical evidence that might corroborate the earliest of these historical records is unavailable, since the very oldest surviving manuscripts that are identifiable witnesses of The Perfection of Wisdom in Twenty-Five Thousand Lines (with the cautions mentioned above) are considerably later fragments in Sanskrit or Buddhist Hybrid Sanskrit from Khotan, Gilgit, and from unknown Central Asian locations, dating from the fifth or sixth century‍—by which time earlier recensions of the sūtra had already been well established for three or four centuries in China and in multiple translations.

There are separate accounts relating how two copies of such recensions, both representing a long Prajñā­pāramitā sūtra of between eighteen and twenty thousand lines, were brought from Khotan to China to be translated into Chinese, both at around the same time in the second half of the third century ᴄᴇ. These copies of the sūtra would certainly have been in an Indic language, and at least by this period such versions of the sūtra were presumably known and used at many sites in the area of Buddhist influence along the cultural and trade routes extending from northwest India up into Central Asia and the Silk Roads skirting the Tarim Basin. Nevertheless, the sūtra’s presence in Khotan at this date may represent early evidence of Mahāyāna influence in what had not long before been a predominantly non-Mahāyāna Buddhist community.

The story begins in 260 ᴄᴇ when the Chinese monk Zhu Shixing, who had studied Lokaksema’s early translation of The Perfection of Wisdom in Eight Thousand Lines (Taishō 224) in Luoyang, set out westward to find the “more complete” text that had been heard about in China but had not been seen. He had been determined to travel all the way to India if necessary but, finding the text in Khotan, did not have to go any further. After various trials and tribulations, he sent a copy to China with his Khotanese disciple Puṇyadhana, but himself remained in Khotan for the rest of his life. Puṇyadhana set out from Khotan with the manuscript in 282 but it was only some five years later that he arrived in Cangyuan, where Mokṣala and others translated it in 291 (Taishō 221).

Mokṣala’s translation, despite the earlier start, ended up being the second Chinese translation to be completed, for in the meantime a Khotanese monk called Gītamitra had brought another Sanskrit manuscript from Khotan to China, this time to Chang’an, where he assisted the great translator Dharmarakṣa to produce the earliest Chinese translation (Taishō 222) in 286.

These accounts underline the importance of the Central Asian oasis states of the Tarim Basin in the spread of Buddhism to China in general, as well as in the early transmission of the long Prajñā­pāramitā sūtras. It is noteworthy that all these scholar-monks responsible in their different ways for the transmission of these texts to China‍—Puṇyadhana, Gītamitra, and Mokṣala himself‍—were all Khotanese, and that Dharmarakṣa was himself a Central Asian born in Dunhuang, probably of Yuezhi ethnicity.

What remains of Dharmarakṣa’s translation today is incomplete and includes only the first twenty-seven chapters, while Mokṣala’s is more complete and contains ninety chapters. The considerable differences that can be discerned between the two texts (in addition to the incomplete nature of the former) are partly due to these two translators’ quite different styles, but also reflect what must have been substantial differences between their Indic source texts despite the close proximity in which they seem to have coexisted in Khotan.

In the early years of the fifth century yet another Central Asian, the great Kuchean translator Kumārajīva, produced his own complete translation (Taishō 223) comprising ninety chapters. This third Chinese translation became the best known and most studied, partly because of Kumārajīva’s readable style and literary renown, but also because it was accompanied by his parallel translation of selections from an encyclopedic commentarial work, known as the Dazhidu lun (*Mahā­prajñā­pāramitopadeśa, Taishō 1509) and attributed to Nāgārjuna. This translation of the commentary became an authoritative and detailed source of Mahāyāna doctrine for Chinese scholars, but its presumed Sanskrit original has left no mention at all in any Indic source. If its traces are to be found at all in India, as indeed they may be, they are mostly indirect.

Finally, in the mid-seventh century Xuanzang translated a massive compilation of Prajñā­pāramitā texts he had brought from his travels in India, known as the Large Perfection of Wisdom Sūtra (Taishō 220), its content and structure probably based on a collection compiled in India rather than being of his own devising. The first section or “assembly” of the compilation represents a text similar to The Perfection of Wisdom in One Hundred Thousand Lines, although the original Sanskrit text that Xuanzang used was probably even longer. It is the second section, containing eighty-five chapters, that represents the fourth and last Chinese translation of The Perfection of Wisdom in Twenty-Five Thousand Lines. It is worth noting here that in Xuanzang’s Chinese translations, as in the Tibetan translations, the language and structure of the two longest sūtras, The Hundred Thousand Lines and The Twenty-Five Thousand Lines, are noticeably similar except for the degree of repetition and expansion of the lists of dharmas, whereas the third section, representing The Perfection of Wisdom in Eighteen Thousand Lines, has more differences in content and may reflect a more elaborated and possibly later text.

It is important to emphasize again here that the categorization of these early translations into sūtras denominated as having a certain length in terms of “lines” (śloka) had not yet become the more definitive naming system that evolved in later centuries. All these texts can most usefully be seen as versions of the “Large Perfection of Wisdom” (Mahā­prajñā­pāramitā) sūtra, as distinct from the mid-length Perfection of Wisdom in Eight Thousand Lines and the many short versions of the sūtra, each of which followed their own evolutionary path. The texts, even in different languages, of the “Large Perfection of Wisdom,” differentiated as they are by greater or lesser degrees of expansion of the lists of dharmas, show complex patterns of textual proximity that do not necessarily follow their numerical denominations.

The work that, despite this reserve, we can nevertheless call The Perfection of Wisdom in Twenty-Five Thousand Lines, particularly in the form of Kumārajīva’s translation and its accompanying commentary, has arguably been among the most important influences on Chinese Buddhism. Yet the ways in which the sūtra was interpreted and used in China led in strikingly different directions from those taken by Indian and Tibetan scholars. This is partly because the massively detailed Dazhidu lun commentary, so influential in China, had remained largely unknown in India (see above), and partly because of the comprehensive way in which Kumārajīva and his followers presented Mahāyāna thought and practice, integrating the Prajñā­pāramitā with Nāgārjuna and Āryadeva’s Madhyamaka treatises as well as with other Mahāyāna sūtras in the textual corpus of the San lun zong or “Three Treatises” school, the influence of which later spread to Korea and Japan. But another major determinant in this respect is simply that the sūtra’s influence and spread in China was early, and effectively predated the influential new turns that‍—while Kumārajīva was still at work on his translations in Chang’an‍—the study of the sūtra was just beginning to take in India.

Meanwhile in India…

At an unknown date in the mid-fourth century, a brilliant young Gandhāran scholar in Puruṣapura (modern Peshawar in Pakistan) is said to have attained mastery of all the available Buddhist scriptures and treatises except‍—to his frustration‍—the Prajñā­pāramitā sūtras, which he could not fully understand. He sought advice from his teachers, one of whom recommended that he should seek the divine help of the bodhisattva and future buddha Maitreya in retreat. He therefore set off to Central India and spent twelve years arduously but (in his own estimation) fruitlessly practicing in a cave until, at the point of giving up his quest, he had a vision of Maitreya in person and was taken by him to the Tuṣita heaven. There he received and wrote down a set of treatises known as the “Five Teachings of Maitreya” (byams chos sde lnga).

Such, in brief, is the legendary account transmitted in Tibetan Buddhist tradition about how Asaṅga (ca. 320–90) wrote down the five important treatises that Tibetan tradition attributes to Maitreya. All five treatises are said‍—in the legendary account, and by some commentators, too‍—to be Maitreya’s explanations of the Prajñā­pāramitā, but one of them in particular, The Ornament of Clear Realization (Abhisamayālaṃkāra), bears an explicit relationship to it and seems to have had a powerful influence on the exegesis of the Prajñā­pāramitā texts, both in the great Buddhist monastic universities of Nālandā, Vikramaśilā, and Odantapuri in India, and after their decline in Kashmir and Tibet where their scholastic traditions continued. Surprisingly, however, and despite Xuanzang’s familiarity with Asaṅga’s other works, it was not translated into Chinese and had no influence at all on Chinese scholarship.

The Ornament of Clear Realization is a short and somewhat cryptic text, not a commentary in the usual sense but rather a key to the implicit structure and essential points of the longer Prajñā­pāramitā sūtras. It can be applied to any of the “six mother sūtras,” which indeed are collectively so called because they all follow that very structure and all include the full set of all eight “clear realizations” (abhisamaya, mngon par rtogs pa) that constitute the main divisions of the Ornament. Each of these eight divisions is subdivided into a hierarchy of further divisions, making seventy points in all. The seventy points follow the order of the topics discussed in the long versions of the sūtra and, at a low resolution, can be mapped more or less closely to them all‍—although at higher resolutions the hypothetical version of the text that is inferentially discernible from the Ornament’s details does not correspond exactly to any one of the extant witnesses, even those of The Perfection of Wisdom in Twenty-Five Thousand Lines to which it is most closely related and on which it is often said to be based.

The influence that The Ornament of Clear Realization exerted on the interpretation of the Prajñā­pāramitā is evident from the range of Prajñā­pāramitā treatises by Indian authors that have survived, either in Sanskrit or in Tibetan translations in the Tengyur. There are no fewer than twenty-one such treatises based on the Ornament: twelve that apply its enumeration of topics as commentaries to the different “mother” versions of the sūtra, and nine that are (in various senses) commentaries on the Ornament itself. The long commentary attributed in the Chinese tradition to Nāgārjuna, the Dazhidu lun (see above)‍—which would, of course have predated the appearance of The Ornament of Clear Realization by a century or two if the attribution is correct‍—seems to have left virtually no evident traces in the Indian literature, and Nāgārjuna’s Madhyamaka works, which can be viewed as general treatises on the doctrines of the Prajñā­pāramitā, are not usually counted as part of the Prajñā­pāramitā literature as such. Only Dignāga’s Piṇḍārtha­saṅgraha (Toh 3809) and two other treatises, both incorporating the Sanskrit term bṛhaṭṭīkā in the titles by which they are most commonly known (Toh 3807 and 3808), comment on the long Prajñā­pāramitā sūtras without specific reference to The Ornament of Clear Realization, although all three use terms and concepts from the other Maitreya-Asaṅga treatises and the Yogācāra system in general.

Toward the end of the flowering of Buddhist learning that took place in India between the sixth and twelfth centuries ᴄᴇ in the monastic universities of the Gupta and Pāla periods, it is clear that The Ornament of Clear Realization had become the preeminent focus for the study of the Prajñā­pāramitā sūtras. Whether this happened early all at once, gradually over a long period, or in different times in different places remains less clear, although the snapshots provided by the Tibetan translations of this literature, reflecting as they presumably do the availability and perceived importance of different texts at different periods, may provide some clues.

The Prajñā­pāramitā Takes Root in Tibet

The great cultural transfer of Buddhist literature, practice, and scholarship that began in the Tibetan imperial period of the late eighth and early ninth centuries, particularly during the reigns of Tri Songdetsen and Ralpachen, naturally included translating the Prajñā­pāramitā texts available in India at the time. By the time the two surviving inventories of the early ninth century, the Denkarma and Phangthangma, had been compiled, all the Prajñā­pāramitā sūtras had been translated, but only some of the treatises and commentaries.

The accounts of Tibetan historians concerning the early translation period focus principally on the various translators, translations, and manuscripts of The Perfection of Wisdom in One Hundred Thousand Lines, but most of the other sūtras, long and short, are listed in the two inventories, including The Perfection of Wisdom in Twenty-Five Thousand Lines.

Of the treatises, the two bṛhaṭṭīkā commentaries on the long sūtras (mentioned above) are listed in the inventories, as are a number of commentaries on the short sūtras. It is not entirely certain whether or not there was an early translation of The Ornament of Clear Realization itself, as the entry that appears to mention it may be read in different ways, but there were early translations of one, and possibly two, of Haribhadra’s most important commentaries on it, and of a commentary by one of his students, Buddhaśrījñāna. It is worth remembering that Haribhadra himself may have studied with Śāntarakṣita and is likely to have been still alive and active when these early translations were made.

It was nevertheless only in the later period of translation, under the successive influences of Rinchen Sangpo, Atiśa Dipaṃkaraśrījñāna, and particularly Ngok Lotsāwa Loden Sherab, that the full panoply of Prajñā­pāramitā and Ornament of Clear Realization treatises appear to have been extensively studied. Ngok and his circle replaced or revised many of the older translations and supplemented them with translations of works by later Indian scholars, including Smṛtijñānakīrti, Dharmaśrī, Ratnākaraśānti, Abhayākaragupta, and Atiśa himself, which had not even been composed at the time of the early translation period.

In other words, while the prevailing state of Prajñā­pāramitā study in India at the time was reflected in the early period Tibetan translations, its evolution had by no means come to an end and was continuing, contemporaneously with the implantation of Buddhism in Tibet. Indeed, that evolution continued within Tibet in subsequent centuries, following the rather abrupt decline of Buddhism in northeastern India and, some time later, in the northwest and Kashmir, too. Several major exegetical lineages and systems of Prajñā­pāramitā study were founded in both eastern and central Tibet, particularly to begin with at Sangphu but also at Sakya, Tsurphu, Jonang, Mindröling, and the three great Gelukpa monasteries around Lhasa. Study of the Prajñā­pāramitā scriptures in the light of The Ornament of Clear Realization, although often in parallel with study of the Madhyamaka, became a distinct, specialist topic pursued by innumerable great scholars down to the present day. Over the centuries almost two hundred and fifty ever more detailed commentaries and synthetic treatises have been composed in Tibetan by almost all the best-known scholastic authors.

In contrast to the great importance attached to interpretations based on The Ornament of Clear Realization, meanwhile, the commentaries not based on it‍—Dignāga’s and the two bṛhaṭṭīkā commentaries‍—were much less studied in Tibet.

The Perfection of Wisdom in Twenty-Five Thousand Lines in Tibet

In common with the other long versions of the sūtra, The Perfection of Wisdom in Twenty-Five Thousand Lines was translated in the early period of translation, as mentioned above. Although its colophon does not mention the translators who worked on it, the various Kangyur catalogs agree that it was translated “at the time of Yeshé Dé and others,” and the fact that it is mentioned in the two early imperial inventories is evidence that its translation was completed by around 813 ᴄᴇ at the latest.

The Tibetan version of the text found in the Kangyur‍—the version translated here‍—contains certain archaic spellings (used in an unusual, inconsistent pattern), certain archaic terms, and renders some bodhisattvas’ names in nonstandard ways; these features may indicate that it was not as extensively revised as were some other canonical texts in the early ninth-century revisions. Its language is strikingly similar to that of The Perfection of Wisdom in One Hundred Thousand Lines, both in these respects and in the exact correspondence of the very numerous passages common to both sūtras. Indeed, it is evident that the two texts‍—at least as they are found in the Degé Kangyur today‍—must have been translated in close relationship, whether in parallel or in sequence, and very likely that the same translators were responsible for both. In contrast, the Tibetan translation of the third of the long sūtras, The Perfection of Wisdom in Eighteen Thousand Lines, diverges noticeably from the other two in its language and use of certain terms and names, even though its content and structure are essentially the same and it is quite plausibly the work of the same translators. It is difficult to be sure whether or not those differences in the Tibetan reflect differences in the source from which The Perfection of Wisdom in Eighteen Thousand Lines was translated, but a parallel discrepancy can be seen in the Chinese of Xuanzang’s translation (see above i.19), which‍—to the extent that there is an equivalence of the three long sūtras in Tibetan with Xuanzang’s three “assemblies”‍—suggests that The Perfection of Wisdom in Eighteen Thousand Lines may have diverged in its textual evolution relative to the other two versions some time before it was translated into Chinese and Tibetan.

Kangyur and Tengyur Versions of the Sūtra

The version of The Perfection of Wisdom in Twenty-Five Thousand Lines as found in the Kangyur and translated here (Toh 9 in the Degé Kangyur) no doubt reflects one of several Sanskrit versions that were circulating in India at the time it was translated‍—i.e., in the early period‍—and may even include elements taken from more than one source. It is not, however, the only canonical version of the sūtra in Tibetan, for another was adopted by scholars in the later period and is now found in the Tengyur (Toh 3790).

That Tengyur version of the sūtra is known colloquially in Tibetan as the “eight-chapter version of the Twenty-Five Thousand” (nyi khri le’u brgyad ma), and the compilation of the Sanskrit text that was its source is attributed to Haribhadra. Although it is sometimes presented as a commentary written by Haribhadra (hence its inclusion in the Tengyur), it is better described as a slightly different version of the sūtra, with some relatively minor differences of order, modifications in certain places, and added structural divisions in the form of brief headings that follow the eight principal topics and their subheadings as set out in The Ornament of Clear Realization. Western scholars, following Conze, have often referred to it as the “recast” or “revised” version, but such a description is misleading in implying the primacy of some “unrevised” version entirely free from exegetical redaction. In the evolution of the many coexisting versions of the long sūtra, there has been no such clear distinction between scriptural transmission and exegetical development.

The Tengyur version is a Tibetan translation that, according to its colophon, was made from a Sanskrit manuscript in a Yambu (Kathmandu) monastery by the Newar paṇḍita Śāntibhadra and the Tibetan translator Naktso Lotsāwa Tsultrim Gyalwa (nag tsho lo tsA ba tshul khrims rgyal ba, 1011–64, closely associated with Atiśa). Nevertheless, the Tibetan translation appears to have been made using the early translation of the Kangyur version (i.e., Toh 9) as the basis upon which the relatively small number of changes necessary to align it with the Sanskrit manuscript in question were made. A large majority of passages are common to the two versions and nearly identical; to describe the Tengyur version as a different translation is therefore somewhat misleading. At the same time, the fact that the later translation was probably made using the earlier as its starting point does not necessarily imply the same chronological relationship between the two Sanskrit source texts. The existence of this alternative, Tengyur version again bears witness to the continuing evolution in the interpretation and study of the Prajñā­pāramitā sūtras in India, and increasingly in parallel in Tibet itself, during the four or five centuries that mark the introduction of Buddhist literature to Tibet.

In the monastic colleges still active today that specialize in Prajñā­pāramitā studies, it is the Tengyur (“eight-chapter”) version of The Perfection of Wisdom in Twenty-Five Thousand Lines to which reference is predominantly made, either directly or because it is the version effectively integrated in the commentaries. To the extent that the curricula of colleges today reflect those of their forebears back through the centuries, the Tengyur version may well have been the principal focus of detailed study since the time it was translated‍—which, although in the later translation period, still predated by several centuries the emergence of the Kangyur and Tengyur as the established canonical collections.

The present Kangyur version has nevertheless remained in place, and in particular represents the sūtra in terms of its being the sacred, scriptural “words of the Buddha” (buddhavacana), along with the other sūtras, for reading and recitation if not for detailed analysis and study. The existence of an alternative may even have contributed to the fact that this version seems to have preserved more features of its early translation than many other canonical texts.

Sanskrit Texts of The Perfection of Wisdom in Twenty-Five Thousand Lines

To have in mind the versions of the sūtra that exist in Tibetan makes it easier to understand the relationship of the surviving Sanskrit texts to those versions.

In brief, the Sanskrit manuscripts we have of this specific sūtra‍—as denominated by its length, i.e., The Perfection of Wisdom in Twenty-Five Thousand Lines itself‍—are of a later date than the Tibetan translations, being copies of manuscripts preserved by the Newar Buddhist tradition in Nepal and not generally more than a few centuries old. Manuscripts of that recension are preserved in Cambridge, Tokyo, Paris, and Kolkata. A critical edition of the section corresponding to the first of the eight abhisamaya topics was published by Nalinaksha Dutt in 1934, and more recently a complete edition in several volumes by Takayasu Kimura has appeared (1986–2007), with concordances to the various manuscripts and the Tibetan and Chinese translations.

These manuscripts of the so-called “Nepalese recension,” and the editions made from them, correspond most closely to the Tengyur “eight-chapter” version of the sūtra, and like it are explicitly aligned to the order of topics in The Ornament of Clear Realization. However, since most passages of the Kangyur and Tengyur versions are either identical or very similar (as noted above), these complete Sanskrit manuscripts are still useful and informative references for reading the present Kangyur version.

The Sanskrit manuscripts of The Perfection of Wisdom in One Hundred Thousand Lines are also of Nepalese origin and are of relatively recent date. Nevertheless, from a recensional perspective they are closer to the present, Kangyur version of The Perfection of Wisdom in Twenty-Five Thousand Lines than the Sanskrit manuscripts of The Perfection of Wisdom in Twenty-Five Thousand Lines just mentioned, which correspond more to the eight-chapter Tengyur version. These Hundred Thousand Sanskrit manuscripts are thus an important reference for the study of this text. In the first decade of the twentieth century, Pratāpacandra Ghoṣa produced an edition of the first twelve chapters from three manuscripts kept in Kolkata and one in Cambridge. More recently, Kimura has been publishing an edition from manuscripts kept in Tokyo of further portions of the text, so far in four volumes, the last in 2014.

Several other incomplete or fragmentary Indic manuscripts are also important references, in part because they are older than the “Nepalese” recensions. The oldest known Prajñā­pāramitā text of all, radiocarbon-dated to the first century ᴄᴇ, is the Gāndhāri birch-bark manuscript from Bajaur, as already mentioned above (i.7), which is not closely identifiable with any surviving version but perhaps belongs to the hypothetical period when precursor versions resembling The Perfection of Wisdom in Twenty-Five Thousand Lines had not yet given rise to any of the “Larger Perfection of Wisdom” series.

The earliest surviving manuscript of a “Larger” version is another birch-bark scroll, this one found along with a large number of other texts in Gilgit in 1931. It is in Sanskrit, is almost complete, and can be dated by details of its script to the sixth or seventh century ᴄᴇ. Although it was thought at first by Edward Conze to be a hybrid consisting of parts of The Perfection of Wisdom in Twenty-Five Thousand Lines and parts of The Perfection of Wisdom in Eighteen Thousand Lines, it is now taken as one among the many coexisting versions of the “Larger” sūtra series. As the most complete of the Gilgit Prajñā­pāramitā manuscripts, it is a very important source; the others are incomplete fragments. Another, similar Sanskrit manuscript of a generic “Larger” version was found in Dunhuang.

Somewhat later manuscripts include a set of ninety-one fragments inscribed on copper, identifiable as belonging to a version of The Perfection of Wisdom in Twenty-Five Thousand Lines, found in Sri Lanka in 1923 in the remains of a stūpa. These Indikuṭasāya Copper Plaques are written in a Sinhalese script of the eighth or ninth century. More substantial fragments, found more recently in Sri Lanka at Anuradhapura, are written on several large gold sheets in a ninth century script. These Sri Lankan fragments of The Perfection of Wisdom in Twenty-Five Thousand Lines contain passages in an order that conforms to The Ornament of Clear Realization but do not include the exegetical subdivision headings characteristic of the Tibetan translation in the Tengyur or the later “Nepalese” Sanskrit versions. In other words, they can be taken as representing an intermediate phase in a multiplicity of recensions ranging between these latter versions on the one hand, and on the other those earlier recensions‍—like the hypothetical Sanskrit text from which the present Kangyur version must have translated‍—that appear to have been the least influenced by the Ornament.

Finally, this discussion of the Sanskrit manuscripts related to the present sūtra would be incomplete without a mention of the last four chapters in this Kangyur version of the text, chapters 73 to 76, which, as mentioned below, are not present in any Indic version directly identifiable with this particular recension. Nevertheless, the Sanskrit of The Perfection of Wisdom in Eight Thousand Lines closely matches them in content, and Vaidya’s 1960 edition provides a convenient reference.

Structure and Content

The Structure and Its Correspondences with the Other Long Sūtras

The sūtra, which fills three volumes in the Degé Kangyur, is divided into seventy-six chapters of quite unequal length. The chapter divisions correspond, to a greater or lesser extent, to those in the other long versions of the Prajñā­pāramitā. Of the three longest sūtras, only The Perfection of Wisdom in Eighteen Thousand Lines provides chapter titles for all its eighty-seven chapters; in the present text, only seventeen of the seventy-six chapters have explicit titles. As mentioned below (i.60 and i.61), some of these chapter titles (which in the source texts are found in the chapter colophons at the end of each chapter rather than as initial titles) are probably intended to encompass a group of preceding chapters as well.

At the most basic level, the structure common to the three long sūtras can be divided into three parts:

• The first chapter which, as in many sūtras, provides the setting or context (nidāna, gleng gzhi), and is common to all the long versions of the sūtra, with differences in length due to different degrees of expansion and some other minor differences in content;

• The main subject matter of the Buddha’s dialogues with his disciples, covered in the bulk of the subsequent chapters, providing the parallel in content of all the sūtras; and

• Several final chapters variably included (and not included at all in The Perfection of Wisdom in One Hundred Thousand Lines), which can be viewed as supplementary elements‍—this is where the three long versions differ most, apart from the extent of their repetitive expansion or contraction.

Thus, the first seventy-one chapters of the present text match in content all seventy-two chapters of The Perfection of Wisdom in One Hundred Thousand Lines, but in more condensed form. The Hundred Thousand ends at the equivalent of this point.

Many of the chapter breaks in the present text also correspond to those in The Perfection of Wisdom in Eighteen Thousand Lines with its yet more condensed subject matter, but since the shorter text has more breaks, the first seventy-one chapters in the present text correspond to the first eighty-two of the eighty-seven chapters of the Eighteen Thousand.

Coming now to the variably included final chapters, the seventy-second chapter of the present text is known as “The Maitreya Chapter” and is found only here in this text (both the Kangyur and Tengyur versions) and in the Eighteen Thousand. More will be said about it below.

The next chapter in the Eighteen Thousand (chapter 84) is the “Verse Summary of the Perfection of Wisdom,” which is also found in all Kangyurs as a separate text (Toh 13), but is not included in the present text or in any other of the long sūtras.

The final four chapters of the present text, 73 to 76, cover the story of Sadāprarudita and Dharmodgata, and also (at the end of chapter 76), the Buddha’s entrustment of the text to Ānanda. These four chapters are not included in the Nepalese Sanskrit manuscripts or the Tengyur version of the text, nor in the Hundred Thousand, but are included (as three final chapters rather than four) in both the Eighteen Thousand and the Eight Thousand. The content of the equivalent chapters in those two texts is almost identical to the content of these chapters here, although the Tibetan translation in this text is clearly a different one made by a different translator team. It nevertheless matches closely, as mentioned above, the surviving Sanskrit of that part of the Eight Thousand. More will be said about the content of these narrative chapters below.

The Content and Its Topical Divisions

As already mentioned, the sūtra opens with the setting of the context for the Buddha’s teaching of the Prajñā­pāramitā. He is on Vulture Peak near Rājagṛha, surrounded by a gathering of arhat monks, nuns, laymen, laywomen, and countless bodhisattvas, of whom thirty-seven are named at this point. After a sequence of meditative stabilities and miraculous displays in which he emanates lights that allay the sufferings of beings in world systems throughout the trichiliocosm and make visible the buddhas teaching in the buddhafields in the ten directions, buddhas and bodhisattvas in each of those buddhafields become aware that a teaching is about to take place, and amid further miraculous displays the bodhisattvas arrive in this world with their offerings to attend the teachings.

The main subject matter, the teachings that the Buddha then gives, unfold as the rest of the text from the beginning of the second chapter. A very broad description of the sūtra’s principal theme, which is common to all the Prajñā­pāramitā texts, has been sketched above in the opening paragraphs of this introduction.

At a first reading and without exegetical guidance, the sūtra’s very extensive presentation of its subject matter may seem somewhat disorganized and unstructured. There are nevertheless several much-studied ways of understanding how the teaching can be classified into different sections and topics. What follows is an extremely abbreviated outline of three such ways among those mentioned in the commentaries: the “three approaches” (or “gateways”); the “eleven discourses”; and The Ornament of Clear Realization’s eight principal topics or “clear realizations” and their seventy subtopics.

The three approaches (sgo gsum) are the brief, intermediate, and detailed teachings, destined respectively for those whose faculties allow them to understand terse, middling, or extensive explanations:

• (1) The brief teaching comes at the start of chapter 2, and consists only of the Buddha’s statement:

  “Śāradvatīputra, bodhisattva great beings who wish to attain consummate buddhahood with respect to all phenomena in all their aspects should persevere in the perfection of wisdom.” (2.2)

(2) The intermediate teaching follows immediately and continues through the discussions between Śāriputra, Subhūti, and the Buddha to the end of chapter 13. The Buddha responds to Śāriputra’s question about what the brief teaching means in terms of the four topics into which it can be subdivided: what a bodhisattva great being is, what it is to attain consummate buddhahood with respect to all phenomena in all their aspects, what persevering means, and what the perfection of wisdom is. Four practices are taught (armor-like, engagement, accumulation, and deliverance), and then, in some detail, eight aspects related to the “persevering.” The last of these eight is a discussion, starting with chapter 8, that arrives at an authoritative conclusion, including twenty-eight or twenty-nine questions, further dialogue between Subhūti and Śāriputra, and in chapters 11, 12, and 13 a long discussion of the Great Vehicle, its attributes, and its results. This entire intermediate teaching is sometimes referred to as “the chapter of Subhūti,” which is also the chapter title this text gives to the last chapter in this section, the thirteenth; that chapter title may be intended to cover the entire group of chapters 3 through 12, too. The intermediate teaching corresponds to the first chapter of The Perfection of Wisdom in Eight Thousand Lines, is centered on all-aspect omniscience, and by teaching nonconceptual perfection of wisdom focuses on ultimate truth.

(3) The detailed teaching is covered by most of the rest of the text, from chapter 14 to the end of chapter 71 (it does not include the Maitreya chapter or the final Sadāprarudita and Dharmodgata chapters). It contains a long series of points to be explained and is divided into two parts: (a) The first part begins with the questions put to Subhūti by Śakra, lord of the gods. The first ten of its chapters, 14 through 23, in which Śakra figures prominently, are probably intended to be covered by the title “Śakra” given in the chapter colophon of chapter 23, and among many important points they explain are how a bodhisattva’s knowledge encompasses that of śrāvakas and pratyekabuddhas, and the benefits, protections, and vast merit conferred by appreciating, honoring, and knowing the perfection of wisdom. Subsequent chapters cover a range of topics, from the dedication of merit, in chapter 24, up to how a bodhisattva’s progress to awakening becomes irreversible, in chapter 40. (b) The second part comprises the discussions prompted by Subhūti’s two hundred and seventy-seven questions, which start in chapter 41 and end in chapter 71. It is centered on knowledge of the paths, and in teaching both conceptual and nonconceptual perfection of wisdom it explains both relative and ultimate truths.

The eleven formulations (rnam grangs bcu gcig) are mentioned in several commentaries detailing the interlocutor concerned, but are not explicitly correlated with particular locations in the texts. Following the mentions in the two bṛhaṭṭīkā commentaries (Toh 3807 and 3808) and Sparham, we may speculatively identify some of them in the present text as follows:

• (1) to Śāriputra, from the beginning of chapter 2 as far as 2.245;

• (2) by Subhūti, the rest of chapter 2 to the end of chapter 13;

• (3) to Śakra, chapter 14, but possibly including several subsequent chapters as well;

• (4) to Subhūti, difficult to identify;

• (5) to Maitreya, the first part of chapter 24;

• (6) to Subhūti, perhaps the rest of chapter 24 from 24.46;

• (7) to Śakra, perhaps chapter 25 to 25.17;

• (8) to Subhūti, perhaps starting in chapter 25 at 25.18, but very likely chapters 41 to 71;

• (9) to Maitreya, chapter 72;

• (10) to Subhūti, again difficult to identify with any certainty; and

• (11) the narrative of Sadāprarudita and Dharmodgata and entrustment of the sūtra to Ānanda, chapters 73 to 76.

It is possible that the list is not intended to follow a sequential order in all instances. In any case, further study on this approach to the sūtras would be desirable.

The eight topics and seventy points of The Ornament of Clear Realization cannot be mapped with complete precision to the content of this version of the sūtra, as explained above (i.24)‍—and indeed that is the raison d’être of the Tengyur version. Nevertheless, included in the following list of the eight principal topics and their subdivision into seventy points is an approximate matching of the eight topics themselves to the content of the chapters of this sūtra, based on how the equivalent passages of the Sanskrit text can be identified in the Tibetan of this version. The seventy points of the Ornament are more difficult to match with the present text, and indeed many of them are not even mentioned in the sutra; some, however, can be discerned‍—not always in the same order as in the Ornament.

(1) All-aspect omniscience (sarvākārajñatā, rnam pa thams cad mkhyen pa nyid) is covered in chapters 2 to 13. Its ten topics are (i) setting of the mind on enlightenment; (ii) the instructions concerning its application within the Great Vehicle; (iii) the four aspects of becoming established on the path of preparation, comprising warmth, peak, acceptance, and supremacy; (iv) the naturally present affinity with the spiritual family, which is the basis for attaining the Great Vehicle; (v) the referents through which the Great Vehicle is attained; (vi) the goals attained through the Great Vehicle; (vii) the armor-like practice; (viii) practice through engagement; (ix) practice through the provisions of merit and wisdom; and (x) practice of definitive deliverance.

(2) Knowledge of the path (mārgajñatā, lam gyi rnam pa shes pa nyid) is covered in chapters 14 to 27. Its eleven topics are (i) essential aspects for understanding the path; (ii) knowledge of the path which is that of the śrāvakas; (iii) knowledge of the path which is that of the pratyekabuddhas; (iv) the beneficial path of insight, which accords with the Great Vehicle; (v) the functions of the path of cultivation; (vi) the aspirational path of cultivation; (vii) the path of cultivation resulting in eulogy, exhortation, and praise; (viii) the path of cultivation resulting in dedication; (ix) the path of cultivation resulting in sympathetic rejoicing; (x) the path of cultivation resulting in attainment; and (xi) the path of meditation resulting in purity.

(3) Knowledge of all the dharmas (sarvajñatā, thams cad shes pa nyid) is covered in chapter 28. Its nine topics are (i) the basic understanding that wisdom leads to not remaining in cyclic existence; (ii) the basic understanding that compassion leads to not remaining in quiescence; (iii) the basic understanding that lack of skillful means leads to distance from the perfection of wisdom; (iv) the basic understanding that skillful means lead to proximity to it; (v) the basic understanding of the discordant factors associated with the fixation of śrāvakas and pratyekabuddhas; (vi) the remedial factors countering those fixations; (vii) training in the aforementioned basic understandings; (viii) training in the sameness of those basic understandings; and (ix) the path of insight which integrates these basic understandings.

(4) Clear realization of all aspects (sarvākārābhisambodha, rnam kun mngon rdzogs rtogs pa) is covered in chapters 29 to 44. Its eleven topics are (i) the aspects of the three kinds of knowledge (1–3 above); (ii) training in those aspects; (iii) the qualities acquired through those trainings; (iv) the defects to be eliminated during training; (v) the defining characteristics of training; (vii) the path of provisions in accord with liberation; (vii) the path of preparation in accord with the four degrees of penetration (1.iii above); (viii) the signs of irreversibility in bodhisattva trainees; (ix) training in the sameness of cyclic existence and quiescence; (x) the training associated with the pure realms; and (xi) training in skillful means for the sake of others.

(5) Culminating clear realization (mūrdhābhisamaya, rtse mor phyin pa’i mngon rtogs) is covered in chapters 45 through to the first few paragraphs of chapter 59. Its eight topics are (i) the culminating training of warmth on the path of preparation; (ii) the culminating training in the peak on the path of preparation; (iii) the culminating training in acceptance on the path of preparation; (iv) the culminating training in supremacy on the path of preparation; (v) the culminating training on the path of insight; (vi) the culminating training on the path of cultivation; (vii) the culminating uninterrupted training on the path of cultivation, comprising the adamantine meditative stability; and (viii) the mistaken notions that are to be eliminated.

(6) Serial clear realization (ānupūrvābhisamaya, mthar gyis pa’i mngon rtogs) is covered by most of chapter 59, from after the first few paragraphs, through to the first few paragraphs of chapter 60. Its thirteen topics are (i–vi) the serial trainings in the six perfections of generosity, ethical discipline, tolerance, perseverance, meditative concentration, and wisdom; (vii–xii) the serial training in the six recollections of the spiritual teacher, the Buddha, the Dharma, the Saṅgha, ethical discipline, and generosity; and (xiii) the serial training in the realization that phenomena are without essential nature.

(7) Instantaneous clear realization (eka­kṣaṇābhisamaya, skad cig ma gcig gis mngon par rtogs pa) is covered in most of chapter 60, from after the first few paragraphs, through to the end of chapter 61. Its four topics are (i) instantaneous training in terms of maturation, (ii) instantaneous training in terms of nonmaturation, (iii) instantaneous training in terms of the lack of defining characteristics, and (iv) instantaneous training in terms of nonduality.

(8) The fruitional buddha body of reality (dharma­kāyābhi­samaya, ’bras bu chos sku) is covered in chapters 62 to 71. Its four topics are (i) the buddha body of essentiality, (ii) the buddha body of wisdom and reality, (iii) the buddha body of perfect resource, and (iv) the buddha body of emanation.

The Protagonists: Śāriputra, Subhūti, Śakra, and the Others

Most of the interest in the Perfection of Wisdom sūtras shown by commentators, whether traditional or modern, has understandably been focused on the doctrinal content rather than the narrative structure, which is sometimes even dismissed as contrived and largely irrelevant. But while the narrative content may not be these sūtras’ most essential feature, it has both purpose and meaning. An appreciation of how the teachings are said to unfold, and of who is saying what, why, and in answer to what question, adds a rich and rewarding layer of understanding for anyone reading these difficult texts closely. Who, then, are the protagonists in this work, and why are they the protagonists and not others?

The most obvious feature of all the longer Perfection of Wisdom sūtras in this regard is that much of the discussion occurs between the Buddha and his śrāvaka disciples, although the subject matter lies at the heart of the Mahāyāna, the path of bodhisattvas.

A number of Mahāyāna sūtras feature conversations between the Buddha, bodhisattvas, and śrāvaka disciples in which the śrāvaka disciples are depicted as rigid and limited in their views compared to the bodhisattvas. But in the Prajñāpāramitā sūtras this is not the case. Substantial statements about the perfection of wisdom are made not only by the Buddha in answer to the śrāvaka interlocutors’ questions, but also directly by the śrāvaka interlocutors themselves. Indeed very little is said by any bodhisattvas, despite their explicit presence at the teaching recounted in these sūtras. Even Mañjuśrī, who could be expected to have a leading role, is simply said to be present among the other bodhisattvas, but no more. Avalokiteśvara too, so central in the Heart Sūtra, is merely present in the audience. The sole exception is Maitreya, who takes part in the dialogue a few times and‍—most notably‍—is the Buddha’s main interlocutor for the important “Maitreya chapter” (chapter 72 here in the Twenty-Five Thousand), on which see below (i.111–i.113). In the final chapters, Dharmodgata and his disciple Sadāprarudita could perhaps be said to be bodhisattva protagonists (see below, i.114); but their status is different, as figures from the past in a narrative related by the Buddha.

Instead, much of the teaching on the Perfection of Wisdom is set out in the form of exchanges between the Buddha and his śrāvaka disciples. Seven or eight names of well-known disciples are variously mentioned in the sūtra collectively as small groups concerned in the discussions on particular topics, sometimes along with the bodhisattvas, but the only ones among them who speak as individuals are Śāriputra, Subhūti, Pūrṇa Maitrāyaṇīputra, and Ānanda.

Of these, it is to Śāriputra‍—without Śāriputra having asked any question‍—that the Buddha addresses the brief statement at the beginning of chapter 2 that starts off the entire teaching. In the canonical literature, Śāriputra (whose name takes the longer form Śāradvatīputra in the sūtra) is “foremost of those with great wisdom” of the disciples, the specialist in the distinction and analysis of dharmas characteristic of the Abhidharma, and it is therefore fitting, as the Dazhidu lun points out, that the teaching should initially be addressed to him. Śāriputra then asks for more detail, and the Buddha’s response forms the entirety of the rest of the long chapter, which is accordingly known as the “Śāriputra chapter” and forms a survey of what is meant by the bodhisattva’s practice of the perfection of wisdom that is relatively easy to understand‍—suggesting that it is the material particularly appropriate to Śāriputra’s perspective. Nevertheless, in subsequent chapters, Śāriputra continues to participate. He does so mainly by asking questions, many of them put to Subhūti rather than to the Buddha himself, and makes few statements of his own. Nevertheless, he is wise and experienced enough even to be able to correct a misunderstanding of Śakra’s at one point (25.6). Śāriputra is appreciative and respectful of the teaching, and wants to understand it in terms of his own framework. The other interlocutors treat Śāriputra’s perspective on it respectfully. Implicitly, however, as the sūtra proceeds it becomes clear that Śāriputra’s category-based understanding lacks the profundity of Subhūti’s.

Indeed, it is Subhūti who is by far the most prominent protagonist in these sūtras. At the beginning of chapter 3, the Buddha asks Subhūti to address the bodhisattvas and tell them how a bodhisattva will become emancipated in the perfection of wisdom. This important passage sets the scene for the rest of the intermediate teaching (on which see i.60 above), all of which (up to the end of chapter 13) is referred to as the “Subhūti chapter” since much of the teaching is either delivered by Subhūti himself, or given by the Buddha in response to Subhūti’s questions. This does not mean, however, that Subhūti’s participation as a protagonist is limited to that section of the sūtra. His conspicuous presence and his major statements continue throughout, and there are only a few (mostly short) chapters in the rest of the text in which Subhūti does not figure at all.

Subhūti is an intriguing figure. As an arhat, he is declared by the Buddha in the canonical literature to hold two distinctions of the “foremost among…” type: “foremost among those worthy of donations” and “foremost among those dwelling free of afflicted mental states.” In the Chinese canon he has another distinction, that of being “foremost among those dwelling in meditation on emptiness.” Despite these mentions of his distinctions Subhūti figures surprisingly rarely in the Pali canon. In the Mūlasarvāstivāda vinaya corpus he seems to be mentioned even less, but the one story about him, found in the Vinayakṣudrakavastu (Toh 6) as well as in the Pali paracanonical literature, is perhaps significant in explaining his interest in the most profound views; it has also traditionally been used to place the timing of the Prajñāpāramitā teachings in the main events of the Buddha’s life. In the Lokottaravāda Mahāvastu he does not appear at all. In the Mahāyāna sūtras, however, he figures much more frequently, and his appearances as a protagonist mostly show him in a respectful light that conforms to the personal characteristics alluded to in the Pali and Mūlasarvastivāda literature: those of someone interested above all in meditative practice and the profound view of impermanence, emptiness, nonself, and the ultimate that transcends conceptual notions. In the Lotus Sūtra he is one of the small group of śrāvakas whose future full awakening to buddhahood is prophesied by the Buddha. On the other hand, there are also a few Mahāyāna works, as there are for other śrāvaka disciples, in which he is demonstratively outclassed by bodhisattvas.

That Subhūti’s role in the Perfection of Wisdom sūtras is not presented as paradoxical‍—a śrāvaka arhat who teaches the perfection of wisdom to bodhisattvas‍—is perhaps surprising at first sight, but is an important key to understanding the long sūtras.

First of all, to practice the perfection of wisdom, as is made clear in the sūtra, is not itself the exclusive domain of the bodhisattva path. For example, Śāriputra himself confirms to Subhūti at the end of chapter 4 (4.24) that śrāvakas and pratyekabuddhas, as well as bodhisattvas, “should earnestly study, take up, uphold, recite, master, and focus their attention correctly on this very perfection of wisdom.” Moreover, as Subhūti starts his explanations to Śakra in chapter 14 (14.3) by exhorting the gods to become bodhisattvas, he also implies that although śrāvaka arhats are (from their own perspective) in their last life, there is nothing to prevent them from spending their remaining time practicing the perfection of wisdom.

Is it a different matter for a śrāvaka to teach the perfection of wisdom‍—especially to bodhisattvas? The important role of śrāvakas in the past teaching the six perfections to future bodhisattvas, and thus ensuring the emergence of future buddhas, is mentioned by Subhūti in chapter 14 (14.27). If there is nevertheless any basis for a śrāvaka teaching the perfection of wisdom to be seen as paradoxical, it is exposed and dispelled early on, at least from a formal perspective. No sooner has the Buddha asked Subhūti to teach than some of the arhats and bodhisattvas present wonder whether he will teach using his own understanding and inspired speech, or through the power of the Buddha (3.2); Subhūti replies immediately (3.3) that a śrāvaka like him can only teach anything through the power of the Buddha. A good deal of room for interpretation is nonetheless left in this passage‍—first in what is meant by “inspired speech” (pratibhāna, spobs pa), then also in the details of Subhūti’s disclaimer, but most of all in the way Subhūti begins to respond to what the Buddha has asked him to do (3.4). For instead of launching straight into a teaching of his own, he asks how he could possibly teach bodhisattvas a perfection of wisdom when neither bodhisattvas nor a perfection of wisdom can be identified other than their mere designations. This key prompt elicits a long teaching by the Buddha on just that topic that lasts for several chapters, punctuated by further questions from Subhūti and long statements by him addressed either to the Buddha, or to Sāriputra when the latter again asks for clarification. Indeed that central question continues to echo throughout the text.

On a few particular occasions we are reminded that Subhūti, or in other places Śāriputra, speaks by the power of the Buddha. But each of the two arhats remains firmly in character, and it is clear that‍—while both are technically on the same level‍—Subhūti is someone of a very different category and status compared to Śāriputra. He also has a quite different role. While both at times seek clarifications from the Buddha, Śāriputra does little more than elicit such clarifications, whether from the Buddha or from Subhūti‍—while Subhūti, on the other hand, makes long and extensive statements of his own. Indeed, at one point in chapter 13 (13.63) Śāriputra exclaims that Subhūti should have been declared “foremost among those who teach the Dharma,” which of course was not one of his official distinctions.

Summary

The Perfection of Wisdom in Twenty-Five Thousand Lines is among the most important scriptures underlying both the “vast” and the “profound” approaches to Buddhist thought and practice. Known as the “middle-length” version, being the second longest of the three long Perfection of Wisdom sūtras, it fills three volumes of the Kangyur. Like the two other long sūtras, it records the major teaching on the perfection of wisdom given by the Buddha Śākyamuni on Vulture Peak, detailing all aspects of the path to enlightenment while at the same time emphasizing how bodhisattvas must put them into practice without taking them‍—or any aspects of enlightenment itself‍—as having even the slightest true existence.

Acknowledgements

Translation by the Padmakara Translation Group. A complete draft by Gyurme Dorje was first edited by Charles Hastings, then revised and further edited by John Canti. The introduction was written by John Canti. We are grateful for the advice and help received from Gareth Sparham, Greg Seton, and Nathaniel Rich.

This translation is dedicated to the memory of our late colleague, long-time friend, and vajra brother Gyurme Dorje (1950–2020), who worked assiduously on this translation in his final years and into the very last months of his life. We would also like to express our gratitude to his wife, Xiaohong, for the extraordinary support she gave him on so many levels.

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

The generous sponsorship of Kris Yao and Xiang-Jen Yao, which helped make the work on this translation possible, is most gratefully acknowledged.

Introduction

The Perfection of Wisdom in Twenty-Five Thousand Lines is one of the three so-called “long” sūtras on the Perfection of Wisdom, or Prajñā­pāramitā. It fills three complete volumes of the Degé Kangyur, and of all the Prajñā­pāramitā sūtras it is second in length only to the massive Perfection of Wisdom in One Hundred Thousand Lines (Toh 8), which fills twelve volumes. The third and shortest of the three “long” sūtras, the Perfection of Wisdom in Eighteen Thousand Lines (Toh 10), fills two and a half volumes.

All three sūtras have a similar structure, closely parallel content, and convey the same teaching: detailed presentations of everything that causes, conditions, and propagates the state of suffering, and of everything that either brings about, or is constituted by, the awakening from that state of suffering. Pervading these presentations is the constant message characteristic of the Prajñā­pāramitā sūtras: that despite the importance of understanding all the phenomena of defilement and of putting into practice all the phenomena of the path leading to purification, none of the phenomena that are known, practiced, or attained must ever be taken as having any ultimately real existence if true awakening is to be attained.

The phenomena that are comprised by defilement and purification are grouped in the numerous sets or lists (mātṛkā or “matrices”) that have characterized Buddhist teachings from their earliest origins. These sets and subsets of phenomena (dharma) became, of course, a particular feature of the systematizing Abhidharma texts, but in parallel also came to constitute an important component of the Prajñā­pāramitā sūtras. Their inclusion, which in the “long” sūtras is even more extensive than in the earlier Perfection of Wisdom in Eight Thousand Lines (Toh 12), may have served to confirm the continuity of the Prajñā­pāramitā’s more profound and difficult perspective on Buddhist practice with earlier forms, as well as to highlight its distinctiveness. In each of the three long sutras, the discussions between the Buddha and the other interlocutors on these sets of dharmas follow one another in almost identical sequence, and the differences in length among the three are almost entirely due to the different degrees to which each set is unpacked into the individual items that it comprises.

While the exhaustive presentation of dharmas that these texts contain provides an important scriptural basis for the classic Buddhist understanding and categorization of phenomena in general, their intent goes much farther than the descriptive itemizing characteristic of the Abhidharma. Most obviously, in his dialectical treatment of each topic the Buddha explicitly undermines any tendencies on the part of his disciples, however subtle they may be, to take any dharma as real or truly existing, or even to adopt it as a point of reference. It is from this “profound” theme of the Prajñā­pāramitā sūtras that the fullest forms of Buddhist philosophical understanding of emptiness developed, as exemplified by the great Madhyamaka treatises of Nāgārjuna and later scholars; it is also this theme that led to the practice traditions of Chan, Thiền, Sŏn, and Zen.

Rather less obviously, for it remains largely implicit (especially in the progressive order in which he sets them out), the Buddha’s instructions on how to overcome and abandon those tendencies form an integrated series of practices‍—the actual path that the bodhisattva must take toward complete awakening‍—and by extension a description of how a practitioner’s mind may be progressively brought to a direct realization of the ultimate. These “vast” instructions were explained to Asaṅga by Maitreya and are an underlying element of the great Yogācāra treatises as well as the key by which many later Indian commentaries explain the Prajñā­pāramitā sūtras.

The Early Spread of the Prajñā­pāramitā Sūtras

Buddhist tradition holds that the Prajñā­pāramitā was taught by the Buddha on Gṛdhrakūṭa (“Vulture Peak”), the craggy hill near the city of Rājagṛha, capital of the Indian kingdom of Magadha. Some accounts say that this teaching took place over a period of some thirty years, while others say twelve years. Nevertheless, not only the three long sūtras but all five of the longer so-called “mother sūtras” are traditionally said to be accounts of the same teaching by the Buddha given on a single occasion. This assertion is made on the basis of two points: first that the Buddha’s interlocutors are the same, and second that a prophecy made by the Buddha included in all of them could only have been made once. Indeed, the multiplicity of versions of this single teaching is not even limited to the five long sūtras in the canon, for several even longer versions are said to have been recorded for the needs of nonhuman beings. While the longest of the versions preserved in the Kangyur, The Perfection of Wisdom in One Hundred Thousand Lines (Toh 8), is said to have been destined for the nāgas‍—from whose realm it was retrieved by Nāgārjuna‍—it is also said that there is a sūtra for the gods with ten million lines, and that the longest version of all, destined for the gandharvas, is a sūtra in one billion lines.

From a historical perspective, versions of Prajñā­pāramitā sūtras in writing seem to have first appeared in the first centuries ʙᴄᴇ and ᴄᴇ. The sūtra’s own exhortation to readers to write it out in the form of books may be associated with the early Mahāyāna’s embrace of written texts. Modern scholars have disagreed about which of the geographically dispersed Buddhist communities may have first given rise to the Prajñā­pāramitā literature, some favoring its origin among the Mahāsāṅghikas of Andhra in the south of India, while others point to evidence of its early flourishing in the northwest regions such as Gandhāra. Whichever may be the case, a birch-bark scroll from the northwest, found in Bajaur (a district of present-day Pakistan near the Afghan border), radiocarbon-dated to the first century ᴄᴇ, is the oldest known Prajñā­pāramitā manuscript. Like most of the earliest birch-bark scrolls from the region it is in the Gāndhārī language, written in Kharoṣṭhī script. The fragmentary portion of the Prajñā­pāramitā sūtra it contains cannot be matched to any recension among the known versions of the sūtras, and may therefore represent an intermediate stage during their evolution from even earlier Prakrit versions into the differentiated sūtras we know today, in their Chinese and Tibetan translations and in the later Sanskrit texts preserved mainly in Nepal.

Different hypotheses regarding the steps and order in which that evolution took place have been proposed, one being that The Perfection of Wisdom in Eight Thousand Lines (Aṣṭasāhasrikā) represents an early version that first led to further long “mother” versions by a process of expansion in the first three centuries ᴄᴇ, and subsequently to the many short versions‍—notably the “Heart” and “Diamond Cutter” sūtras (Hṛdaya and Vajracchedikā)‍—by an opposite contraction from the fourth to the sixth century, after which even more compressed and sometimes “tantric” versions appeared, as well as texts in which the Prajñā­pāramitā is worshiped as a female “mother” deity. The neatness of this schema, however, may overlook some of the evidence. For instance, there are reasons for suspecting that the Vajracchedikā may date to as early as the second century ᴄᴇ.

It is important to bear in mind that the naming of the different versions by the number of lines they contain is likely to have been a later development, applied as a means of classifying the profusion of circulating texts of different lengths. It was already in use by the time these texts were first translated into Tibetan in the late eighth and early ninth centuries, but it is not a feature of the earlier Chinese translations. The earliest evidence of this nomenclature appears to be in the Chinese literature, in the record of a lecture by the early sixth century translator Bodhiruci, and its widespread adoption in the centuries that followed may have served to limit further profusion and even reduce the variety of different texts by fixing their number. Nevertheless, its retrospective application to earlier texts may obscure rather than clarify their recensional affinities, and the comparison of parallel passages across recensions of all the sūtras remains a valid means of exploring qualitative textual differences regardless of the quantitative affiliation defined by the particular title.

This is particularly true of the longest versions of the sūtra, i.e., the sūtras in Ten Thousand Lines, Eighteen Thousand Lines, Twenty-Five Thousand Lines, and One Hundred Thousand Lines, which are better seen as a group, often termed “the Long Prajñā­pāramitā Sūtras.” It includes texts that exist variably in Sanskrit, Chinese, and Tibetan, different recensional relationships between which can be distinguished independently of the “length” denominations into which they are categorized.

The Perfection of Wisdom in Twenty-Five Thousand Lines in Central Asia and China

Coming to The Perfection of Wisdom in Twenty-Five Thousand Lines itself, we do not know‍—apart from placing it with reasonable probability within the process of the expansion of longer texts as mentioned above‍—when or where it first appeared as a distinct redaction of the Prajñā­pāramitā sūtra, or even in what language. As is the case with many canonical works, the earliest historical mentions predate any surviving physical texts, and come from accounts of its first translation into Chinese.

Perhaps unsurprisingly, physical evidence that might corroborate the earliest of these historical records is unavailable, since the very oldest surviving manuscripts that are identifiable witnesses of The Perfection of Wisdom in Twenty-Five Thousand Lines (with the cautions mentioned above) are considerably later fragments in Sanskrit or Buddhist Hybrid Sanskrit from Khotan, Gilgit, and from unknown Central Asian locations, dating from the fifth or sixth century‍—by which time earlier recensions of the sūtra had already been well established for three or four centuries in China and in multiple translations.

There are separate accounts relating how two copies of such recensions, both representing a long Prajñā­pāramitā sūtra of between eighteen and twenty thousand lines, were brought from Khotan to China to be translated into Chinese, both at around the same time in the second half of the third century ᴄᴇ. These copies of the sūtra would certainly have been in an Indic language, and at least by this period such versions of the sūtra were presumably known and used at many sites in the area of Buddhist influence along the cultural and trade routes extending from northwest India up into Central Asia and the Silk Roads skirting the Tarim Basin. Nevertheless, the sūtra’s presence in Khotan at this date may represent early evidence of Mahāyāna influence in what had not long before been a predominantly non-Mahāyāna Buddhist community.

The story begins in 260 ᴄᴇ when the Chinese monk Zhu Shixing, who had studied Lokaksema’s early translation of The Perfection of Wisdom in Eight Thousand Lines (Taishō 224) in Luoyang, set out westward to find the “more complete” text that had been heard about in China but had not been seen. He had been determined to travel all the way to India if necessary but, finding the text in Khotan, did not have to go any further. After various trials and tribulations, he sent a copy to China with his Khotanese disciple Puṇyadhana, but himself remained in Khotan for the rest of his life. Puṇyadhana set out from Khotan with the manuscript in 282 but it was only some five years later that he arrived in Cangyuan, where Mokṣala and others translated it in 291 (Taishō 221).

Mokṣala’s translation, despite the earlier start, ended up being the second Chinese translation to be completed, for in the meantime a Khotanese monk called Gītamitra had brought another Sanskrit manuscript from Khotan to China, this time to Chang’an, where he assisted the great translator Dharmarakṣa to produce the earliest Chinese translation (Taishō 222) in 286.

These accounts underline the importance of the Central Asian oasis states of the Tarim Basin in the spread of Buddhism to China in general, as well as in the early transmission of the long Prajñā­pāramitā sūtras. It is noteworthy that all these scholar-monks responsible in their different ways for the transmission of these texts to China‍—Puṇyadhana, Gītamitra, and Mokṣala himself‍—were all Khotanese, and that Dharmarakṣa was himself a Central Asian born in Dunhuang, probably of Yuezhi ethnicity.

What remains of Dharmarakṣa’s translation today is incomplete and includes only the first twenty-seven chapters, while Mokṣala’s is more complete and contains ninety chapters. The considerable differences that can be discerned between the two texts (in addition to the incomplete nature of the former) are partly due to these two translators’ quite different styles, but also reflect what must have been substantial differences between their Indic source texts despite the close proximity in which they seem to have coexisted in Khotan.

In the early years of the fifth century yet another Central Asian, the great Kuchean translator Kumārajīva, produced his own complete translation (Taishō 223) comprising ninety chapters. This third Chinese translation became the best known and most studied, partly because of Kumārajīva’s readable style and literary renown, but also because it was accompanied by his parallel translation of selections from an encyclopedic commentarial work, known as the Dazhidu lun (*Mahā­prajñā­pāramitopadeśa, Taishō 1509) and attributed to Nāgārjuna. This translation of the commentary became an authoritative and detailed source of Mahāyāna doctrine for Chinese scholars, but its presumed Sanskrit original has left no mention at all in any Indic source. If its traces are to be found at all in India, as indeed they may be, they are mostly indirect.

Finally, in the mid-seventh century Xuanzang translated a massive compilation of Prajñā­pāramitā texts he had brought from his travels in India, known as the Large Perfection of Wisdom Sūtra (Taishō 220), its content and structure probably based on a collection compiled in India rather than being of his own devising. The first section or “assembly” of the compilation represents a text similar to The Perfection of Wisdom in One Hundred Thousand Lines, although the original Sanskrit text that Xuanzang used was probably even longer. It is the second section, containing eighty-five chapters, that represents the fourth and last Chinese translation of The Perfection of Wisdom in Twenty-Five Thousand Lines. It is worth noting here that in Xuanzang’s Chinese translations, as in the Tibetan translations, the language and structure of the two longest sūtras, The Hundred Thousand Lines and The Twenty-Five Thousand Lines, are noticeably similar except for the degree of repetition and expansion of the lists of dharmas, whereas the third section, representing The Perfection of Wisdom in Eighteen Thousand Lines, has more differences in content and may reflect a more elaborated and possibly later text.

It is important to emphasize again here that the categorization of these early translations into sūtras denominated as having a certain length in terms of “lines” (śloka) had not yet become the more definitive naming system that evolved in later centuries. All these texts can most usefully be seen as versions of the “Large Perfection of Wisdom” (Mahā­prajñā­pāramitā) sūtra, as distinct from the mid-length Perfection of Wisdom in Eight Thousand Lines and the many short versions of the sūtra, each of which followed their own evolutionary path. The texts, even in different languages, of the “Large Perfection of Wisdom,” differentiated as they are by greater or lesser degrees of expansion of the lists of dharmas, show complex patterns of textual proximity that do not necessarily follow their numerical denominations.

The work that, despite this reserve, we can nevertheless call The Perfection of Wisdom in Twenty-Five Thousand Lines, particularly in the form of Kumārajīva’s translation and its accompanying commentary, has arguably been among the most important influences on Chinese Buddhism. Yet the ways in which the sūtra was interpreted and used in China led in strikingly different directions from those taken by Indian and Tibetan scholars. This is partly because the massively detailed Dazhidu lun commentary, so influential in China, had remained largely unknown in India (see above), and partly because of the comprehensive way in which Kumārajīva and his followers presented Mahāyāna thought and practice, integrating the Prajñā­pāramitā with Nāgārjuna and Āryadeva’s Madhyamaka treatises as well as with other Mahāyāna sūtras in the textual corpus of the San lun zong or “Three Treatises” school, the influence of which later spread to Korea and Japan. But another major determinant in this respect is simply that the sūtra’s influence and spread in China was early, and effectively predated the influential new turns that‍—while Kumārajīva was still at work on his translations in Chang’an‍—the study of the sūtra was just beginning to take in India.

Meanwhile in India…

At an unknown date in the mid-fourth century, a brilliant young Gandhāran scholar in Puruṣapura (modern Peshawar in Pakistan) is said to have attained mastery of all the available Buddhist scriptures and treatises except‍—to his frustration‍—the Prajñā­pāramitā sūtras, which he could not fully understand. He sought advice from his teachers, one of whom recommended that he should seek the divine help of the bodhisattva and future buddha Maitreya in retreat. He therefore set off to Central India and spent twelve years arduously but (in his own estimation) fruitlessly practicing in a cave until, at the point of giving up his quest, he had a vision of Maitreya in person and was taken by him to the Tuṣita heaven. There he received and wrote down a set of treatises known as the “Five Teachings of Maitreya” (byams chos sde lnga).

Such, in brief, is the legendary account transmitted in Tibetan Buddhist tradition about how Asaṅga (ca. 320–90) wrote down the five important treatises that Tibetan tradition attributes to Maitreya. All five treatises are said‍—in the legendary account, and by some commentators, too‍—to be Maitreya’s explanations of the Prajñā­pāramitā, but one of them in particular, The Ornament of Clear Realization (Abhisamayālaṃkāra), bears an explicit relationship to it and seems to have had a powerful influence on the exegesis of the Prajñā­pāramitā texts, both in the great Buddhist monastic universities of Nālandā, Vikramaśilā, and Odantapuri in India, and after their decline in Kashmir and Tibet where their scholastic traditions continued. Surprisingly, however, and despite Xuanzang’s familiarity with Asaṅga’s other works, it was not translated into Chinese and had no influence at all on Chinese scholarship.

The Ornament of Clear Realization is a short and somewhat cryptic text, not a commentary in the usual sense but rather a key to the implicit structure and essential points of the longer Prajñā­pāramitā sūtras. It can be applied to any of the “six mother sūtras,” which indeed are collectively so called because they all follow that very structure and all include the full set of all eight “clear realizations” (abhisamaya, mngon par rtogs pa) that constitute the main divisions of the Ornament. Each of these eight divisions is subdivided into a hierarchy of further divisions, making seventy points in all. The seventy points follow the order of the topics discussed in the long versions of the sūtra and, at a low resolution, can be mapped more or less closely to them all‍—although at higher resolutions the hypothetical version of the text that is inferentially discernible from the Ornament’s details does not correspond exactly to any one of the extant witnesses, even those of The Perfection of Wisdom in Twenty-Five Thousand Lines to which it is most closely related and on which it is often said to be based.

The influence that The Ornament of Clear Realization exerted on the interpretation of the Prajñā­pāramitā is evident from the range of Prajñā­pāramitā treatises by Indian authors that have survived, either in Sanskrit or in Tibetan translations in the Tengyur. There are no fewer than twenty-one such treatises based on the Ornament: twelve that apply its enumeration of topics as commentaries to the different “mother” versions of the sūtra, and nine that are (in various senses) commentaries on the Ornament itself. The long commentary attributed in the Chinese tradition to Nāgārjuna, the Dazhidu lun (see above)‍—which would, of course have predated the appearance of The Ornament of Clear Realization by a century or two if the attribution is correct‍—seems to have left virtually no evident traces in the Indian literature, and Nāgārjuna’s Madhyamaka works, which can be viewed as general treatises on the doctrines of the Prajñā­pāramitā, are not usually counted as part of the Prajñā­pāramitā literature as such. Only Dignāga’s Piṇḍārtha­saṅgraha (Toh 3809) and two other treatises, both incorporating the Sanskrit term bṛhaṭṭīkā in the titles by which they are most commonly known (Toh 3807 and 3808), comment on the long Prajñā­pāramitā sūtras without specific reference to The Ornament of Clear Realization, although all three use terms and concepts from the other Maitreya-Asaṅga treatises and the Yogācāra system in general.

Toward the end of the flowering of Buddhist learning that took place in India between the sixth and twelfth centuries ᴄᴇ in the monastic universities of the Gupta and Pāla periods, it is clear that The Ornament of Clear Realization had become the preeminent focus for the study of the Prajñā­pāramitā sūtras. Whether this happened early all at once, gradually over a long period, or in different times in different places remains less clear, although the snapshots provided by the Tibetan translations of this literature, reflecting as they presumably do the availability and perceived importance of different texts at different periods, may provide some clues.

The Prajñā­pāramitā Takes Root in Tibet

The great cultural transfer of Buddhist literature, practice, and scholarship that began in the Tibetan imperial period of the late eighth and early ninth centuries, particularly during the reigns of Tri Songdetsen and Ralpachen, naturally included translating the Prajñā­pāramitā texts available in India at the time. By the time the two surviving inventories of the early ninth century, the Denkarma and Phangthangma, had been compiled, all the Prajñā­pāramitā sūtras had been translated, but only some of the treatises and commentaries.

The accounts of Tibetan historians concerning the early translation period focus principally on the various translators, translations, and manuscripts of The Perfection of Wisdom in One Hundred Thousand Lines, but most of the other sūtras, long and short, are listed in the two inventories, including The Perfection of Wisdom in Twenty-Five Thousand Lines.

Of the treatises, the two bṛhaṭṭīkā commentaries on the long sūtras (mentioned above) are listed in the inventories, as are a number of commentaries on the short sūtras. It is not entirely certain whether or not there was an early translation of The Ornament of Clear Realization itself, as the entry that appears to mention it may be read in different ways, but there were early translations of one, and possibly two, of Haribhadra’s most important commentaries on it, and of a commentary by one of his students, Buddhaśrījñāna. It is worth remembering that Haribhadra himself may have studied with Śāntarakṣita and is likely to have been still alive and active when these early translations were made.

It was nevertheless only in the later period of translation, under the successive influences of Rinchen Sangpo, Atiśa Dipaṃkaraśrījñāna, and particularly Ngok Lotsāwa Loden Sherab, that the full panoply of Prajñā­pāramitā and Ornament of Clear Realization treatises appear to have been extensively studied. Ngok and his circle replaced or revised many of the older translations and supplemented them with translations of works by later Indian scholars, including Smṛtijñānakīrti, Dharmaśrī, Ratnākaraśānti, Abhayākaragupta, and Atiśa himself, which had not even been composed at the time of the early translation period.

In other words, while the prevailing state of Prajñā­pāramitā study in India at the time was reflected in the early period Tibetan translations, its evolution had by no means come to an end and was continuing, contemporaneously with the implantation of Buddhism in Tibet. Indeed, that evolution continued within Tibet in subsequent centuries, following the rather abrupt decline of Buddhism in northeastern India and, some time later, in the northwest and Kashmir, too. Several major exegetical lineages and systems of Prajñā­pāramitā study were founded in both eastern and central Tibet, particularly to begin with at Sangphu but also at Sakya, Tsurphu, Jonang, Mindröling, and the three great Gelukpa monasteries around Lhasa. Study of the Prajñā­pāramitā scriptures in the light of The Ornament of Clear Realization, although often in parallel with study of the Madhyamaka, became a distinct, specialist topic pursued by innumerable great scholars down to the present day. Over the centuries almost two hundred and fifty ever more detailed commentaries and synthetic treatises have been composed in Tibetan by almost all the best-known scholastic authors.

In contrast to the great importance attached to interpretations based on The Ornament of Clear Realization, meanwhile, the commentaries not based on it‍—Dignāga’s and the two bṛhaṭṭīkā commentaries‍—were much less studied in Tibet.

The Perfection of Wisdom in Twenty-Five Thousand Lines in Tibet

In common with the other long versions of the sūtra, The Perfection of Wisdom in Twenty-Five Thousand Lines was translated in the early period of translation, as mentioned above. Although its colophon does not mention the translators who worked on it, the various Kangyur catalogs agree that it was translated “at the time of Yeshé Dé and others,” and the fact that it is mentioned in the two early imperial inventories is evidence that its translation was completed by around 813 ᴄᴇ at the latest.

The Tibetan version of the text found in the Kangyur‍—the version translated here‍—contains certain archaic spellings (used in an unusual, inconsistent pattern), certain archaic terms, and renders some bodhisattvas’ names in nonstandard ways; these features may indicate that it was not as extensively revised as were some other canonical texts in the early ninth-century revisions. Its language is strikingly similar to that of The Perfection of Wisdom in One Hundred Thousand Lines, both in these respects and in the exact correspondence of the very numerous passages common to both sūtras. Indeed, it is evident that the two texts‍—at least as they are found in the Degé Kangyur today‍—must have been translated in close relationship, whether in parallel or in sequence, and very likely that the same translators were responsible for both. In contrast, the Tibetan translation of the third of the long sūtras, The Perfection of Wisdom in Eighteen Thousand Lines, diverges noticeably from the other two in its language and use of certain terms and names, even though its content and structure are essentially the same and it is quite plausibly the work of the same translators. It is difficult to be sure whether or not those differences in the Tibetan reflect differences in the source from which The Perfection of Wisdom in Eighteen Thousand Lines was translated, but a parallel discrepancy can be seen in the Chinese of Xuanzang’s translation (see above i.19), which‍—to the extent that there is an equivalence of the three long sūtras in Tibetan with Xuanzang’s three “assemblies”‍—suggests that The Perfection of Wisdom in Eighteen Thousand Lines may have diverged in its textual evolution relative to the other two versions some time before it was translated into Chinese and Tibetan.

Kangyur and Tengyur Versions of the Sūtra

The version of The Perfection of Wisdom in Twenty-Five Thousand Lines as found in the Kangyur and translated here (Toh 9 in the Degé Kangyur) no doubt reflects one of several Sanskrit versions that were circulating in India at the time it was translated‍—i.e., in the early period‍—and may even include elements taken from more than one source. It is not, however, the only canonical version of the sūtra in Tibetan, for another was adopted by scholars in the later period and is now found in the Tengyur (Toh 3790).

That Tengyur version of the sūtra is known colloquially in Tibetan as the “eight-chapter version of the Twenty-Five Thousand” (nyi khri le’u brgyad ma), and the compilation of the Sanskrit text that was its source is attributed to Haribhadra. Although it is sometimes presented as a commentary written by Haribhadra (hence its inclusion in the Tengyur), it is better described as a slightly different version of the sūtra, with some relatively minor differences of order, modifications in certain places, and added structural divisions in the form of brief headings that follow the eight principal topics and their subheadings as set out in The Ornament of Clear Realization. Western scholars, following Conze, have often referred to it as the “recast” or “revised” version, but such a description is misleading in implying the primacy of some “unrevised” version entirely free from exegetical redaction. In the evolution of the many coexisting versions of the long sūtra, there has been no such clear distinction between scriptural transmission and exegetical development.

The Tengyur version is a Tibetan translation that, according to its colophon, was made from a Sanskrit manuscript in a Yambu (Kathmandu) monastery by the Newar paṇḍita Śāntibhadra and the Tibetan translator Naktso Lotsāwa Tsultrim Gyalwa (nag tsho lo tsA ba tshul khrims rgyal ba, 1011–64, closely associated with Atiśa). Nevertheless, the Tibetan translation appears to have been made using the early translation of the Kangyur version (i.e., Toh 9) as the basis upon which the relatively small number of changes necessary to align it with the Sanskrit manuscript in question were made. A large majority of passages are common to the two versions and nearly identical; to describe the Tengyur version as a different translation is therefore somewhat misleading. At the same time, the fact that the later translation was probably made using the earlier as its starting point does not necessarily imply the same chronological relationship between the two Sanskrit source texts. The existence of this alternative, Tengyur version again bears witness to the continuing evolution in the interpretation and study of the Prajñā­pāramitā sūtras in India, and increasingly in parallel in Tibet itself, during the four or five centuries that mark the introduction of Buddhist literature to Tibet.

In the monastic colleges still active today that specialize in Prajñā­pāramitā studies, it is the Tengyur (“eight-chapter”) version of The Perfection of Wisdom in Twenty-Five Thousand Lines to which reference is predominantly made, either directly or because it is the version effectively integrated in the commentaries. To the extent that the curricula of colleges today reflect those of their forebears back through the centuries, the Tengyur version may well have been the principal focus of detailed study since the time it was translated‍—which, although in the later translation period, still predated by several centuries the emergence of the Kangyur and Tengyur as the established canonical collections.

The present Kangyur version has nevertheless remained in place, and in particular represents the sūtra in terms of its being the sacred, scriptural “words of the Buddha” (buddhavacana), along with the other sūtras, for reading and recitation if not for detailed analysis and study. The existence of an alternative may even have contributed to the fact that this version seems to have preserved more features of its early translation than many other canonical texts.

Sanskrit Texts of The Perfection of Wisdom in Twenty-Five Thousand Lines

To have in mind the versions of the sūtra that exist in Tibetan makes it easier to understand the relationship of the surviving Sanskrit texts to those versions.

In brief, the Sanskrit manuscripts we have of this specific sūtra‍—as denominated by its length, i.e., The Perfection of Wisdom in Twenty-Five Thousand Lines itself‍—are of a later date than the Tibetan translations, being copies of manuscripts preserved by the Newar Buddhist tradition in Nepal and not generally more than a few centuries old. Manuscripts of that recension are preserved in Cambridge, Tokyo, Paris, and Kolkata. A critical edition of the section corresponding to the first of the eight abhisamaya topics was published by Nalinaksha Dutt in 1934, and more recently a complete edition in several volumes by Takayasu Kimura has appeared (1986–2007), with concordances to the various manuscripts and the Tibetan and Chinese translations.

These manuscripts of the so-called “Nepalese recension,” and the editions made from them, correspond most closely to the Tengyur “eight-chapter” version of the sūtra, and like it are explicitly aligned to the order of topics in The Ornament of Clear Realization. However, since most passages of the Kangyur and Tengyur versions are either identical or very similar (as noted above), these complete Sanskrit manuscripts are still useful and informative references for reading the present Kangyur version.

The Sanskrit manuscripts of The Perfection of Wisdom in One Hundred Thousand Lines are also of Nepalese origin and are of relatively recent date. Nevertheless, from a recensional perspective they are closer to the present, Kangyur version of The Perfection of Wisdom in Twenty-Five Thousand Lines than the Sanskrit manuscripts of The Perfection of Wisdom in Twenty-Five Thousand Lines just mentioned, which correspond more to the eight-chapter Tengyur version. These Hundred Thousand Sanskrit manuscripts are thus an important reference for the study of this text. In the first decade of the twentieth century, Pratāpacandra Ghoṣa produced an edition of the first twelve chapters from three manuscripts kept in Kolkata and one in Cambridge. More recently, Kimura has been publishing an edition from manuscripts kept in Tokyo of further portions of the text, so far in four volumes, the last in 2014.

Several other incomplete or fragmentary Indic manuscripts are also important references, in part because they are older than the “Nepalese” recensions. The oldest known Prajñā­pāramitā text of all, radiocarbon-dated to the first century ᴄᴇ, is the Gāndhāri birch-bark manuscript from Bajaur, as already mentioned above (i.7), which is not closely identifiable with any surviving version but perhaps belongs to the hypothetical period when precursor versions resembling The Perfection of Wisdom in Twenty-Five Thousand Lines had not yet given rise to any of the “Larger Perfection of Wisdom” series.

The earliest surviving manuscript of a “Larger” version is another birch-bark scroll, this one found along with a large number of other texts in Gilgit in 1931. It is in Sanskrit, is almost complete, and can be dated by details of its script to the sixth or seventh century ᴄᴇ. Although it was thought at first by Edward Conze to be a hybrid consisting of parts of The Perfection of Wisdom in Twenty-Five Thousand Lines and parts of The Perfection of Wisdom in Eighteen Thousand Lines, it is now taken as one among the many coexisting versions of the “Larger” sūtra series. As the most complete of the Gilgit Prajñā­pāramitā manuscripts, it is a very important source; the others are incomplete fragments. Another, similar Sanskrit manuscript of a generic “Larger” version was found in Dunhuang.

Somewhat later manuscripts include a set of ninety-one fragments inscribed on copper, identifiable as belonging to a version of The Perfection of Wisdom in Twenty-Five Thousand Lines, found in Sri Lanka in 1923 in the remains of a stūpa. These Indikuṭasāya Copper Plaques are written in a Sinhalese script of the eighth or ninth century. More substantial fragments, found more recently in Sri Lanka at Anuradhapura, are written on several large gold sheets in a ninth century script. These Sri Lankan fragments of The Perfection of Wisdom in Twenty-Five Thousand Lines contain passages in an order that conforms to The Ornament of Clear Realization but do not include the exegetical subdivision headings characteristic of the Tibetan translation in the Tengyur or the later “Nepalese” Sanskrit versions. In other words, they can be taken as representing an intermediate phase in a multiplicity of recensions ranging between these latter versions on the one hand, and on the other those earlier recensions‍—like the hypothetical Sanskrit text from which the present Kangyur version must have translated‍—that appear to have been the least influenced by the Ornament.

Finally, this discussion of the Sanskrit manuscripts related to the present sūtra would be incomplete without a mention of the last four chapters in this Kangyur version of the text, chapters 73 to 76, which, as mentioned below, are not present in any Indic version directly identifiable with this particular recension. Nevertheless, the Sanskrit of The Perfection of Wisdom in Eight Thousand Lines closely matches them in content, and Vaidya’s 1960 edition provides a convenient reference.

Structure and Content

The Structure and Its Correspondences with the Other Long Sūtras

The sūtra, which fills three volumes in the Degé Kangyur, is divided into seventy-six chapters of quite unequal length. The chapter divisions correspond, to a greater or lesser extent, to those in the other long versions of the Prajñā­pāramitā. Of the three longest sūtras, only The Perfection of Wisdom in Eighteen Thousand Lines provides chapter titles for all its eighty-seven chapters; in the present text, only seventeen of the seventy-six chapters have explicit titles. As mentioned below (i.60 and i.61), some of these chapter titles (which in the source texts are found in the chapter colophons at the end of each chapter rather than as initial titles) are probably intended to encompass a group of preceding chapters as well.

At the most basic level, the structure common to the three long sūtras can be divided into three parts:

• The first chapter which, as in many sūtras, provides the setting or context (nidāna, gleng gzhi), and is common to all the long versions of the sūtra, with differences in length due to different degrees of expansion and some other minor differences in content;

• The main subject matter of the Buddha’s dialogues with his disciples, covered in the bulk of the subsequent chapters, providing the parallel in content of all the sūtras; and

• Several final chapters variably included (and not included at all in The Perfection of Wisdom in One Hundred Thousand Lines), which can be viewed as supplementary elements‍—this is where the three long versions differ most, apart from the extent of their repetitive expansion or contraction.

Thus, the first seventy-one chapters of the present text match in content all seventy-two chapters of The Perfection of Wisdom in One Hundred Thousand Lines, but in more condensed form. The Hundred Thousand ends at the equivalent of this point.

Many of the chapter breaks in the present text also correspond to those in The Perfection of Wisdom in Eighteen Thousand Lines with its yet more condensed subject matter, but since the shorter text has more breaks, the first seventy-one chapters in the present text correspond to the first eighty-two of the eighty-seven chapters of the Eighteen Thousand.

Coming now to the variably included final chapters, the seventy-second chapter of the present text is known as “The Maitreya Chapter” and is found only here in this text (both the Kangyur and Tengyur versions) and in the Eighteen Thousand. More will be said about it below.

The next chapter in the Eighteen Thousand (chapter 84) is the “Verse Summary of the Perfection of Wisdom,” which is also found in all Kangyurs as a separate text (Toh 13), but is not included in the present text or in any other of the long sūtras.

The final four chapters of the present text, 73 to 76, cover the story of Sadāprarudita and Dharmodgata, and also (at the end of chapter 76), the Buddha’s entrustment of the text to Ānanda. These four chapters are not included in the Nepalese Sanskrit manuscripts or the Tengyur version of the text, nor in the Hundred Thousand, but are included (as three final chapters rather than four) in both the Eighteen Thousand and the Eight Thousand. The content of the equivalent chapters in those two texts is almost identical to the content of these chapters here, although the Tibetan translation in this text is clearly a different one made by a different translator team. It nevertheless matches closely, as mentioned above, the surviving Sanskrit of that part of the Eight Thousand. More will be said about the content of these narrative chapters below.

The Content and Its Topical Divisions

As already mentioned, the sūtra opens with the setting of the context for the Buddha’s teaching of the Prajñā­pāramitā. He is on Vulture Peak near Rājagṛha, surrounded by a gathering of arhat monks, nuns, laymen, laywomen, and countless bodhisattvas, of whom thirty-seven are named at this point. After a sequence of meditative stabilities and miraculous displays in which he emanates lights that allay the sufferings of beings in world systems throughout the trichiliocosm and make visible the buddhas teaching in the buddhafields in the ten directions, buddhas and bodhisattvas in each of those buddhafields become aware that a teaching is about to take place, and amid further miraculous displays the bodhisattvas arrive in this world with their offerings to attend the teachings.

The main subject matter, the teachings that the Buddha then gives, unfold as the rest of the text from the beginning of the second chapter. A very broad description of the sūtra’s principal theme, which is common to all the Prajñā­pāramitā texts, has been sketched above in the opening paragraphs of this introduction.

At a first reading and without exegetical guidance, the sūtra’s very extensive presentation of its subject matter may seem somewhat disorganized and unstructured. There are nevertheless several much-studied ways of understanding how the teaching can be classified into different sections and topics. What follows is an extremely abbreviated outline of three such ways among those mentioned in the commentaries: the “three approaches” (or “gateways”); the “eleven discourses”; and The Ornament of Clear Realization’s eight principal topics or “clear realizations” and their seventy subtopics.

The three approaches (sgo gsum) are the brief, intermediate, and detailed teachings, destined respectively for those whose faculties allow them to understand terse, middling, or extensive explanations:

• (1) The brief teaching comes at the start of chapter 2, and consists only of the Buddha’s statement:

  “Śāradvatīputra, bodhisattva great beings who wish to attain consummate buddhahood with respect to all phenomena in all their aspects should persevere in the perfection of wisdom.” (2.2)

(2) The intermediate teaching follows immediately and continues through the discussions between Śāriputra, Subhūti, and the Buddha to the end of chapter 13. The Buddha responds to Śāriputra’s question about what the brief teaching means in terms of the four topics into which it can be subdivided: what a bodhisattva great being is, what it is to attain consummate buddhahood with respect to all phenomena in all their aspects, what persevering means, and what the perfection of wisdom is. Four practices are taught (armor-like, engagement, accumulation, and deliverance), and then, in some detail, eight aspects related to the “persevering.” The last of these eight is a discussion, starting with chapter 8, that arrives at an authoritative conclusion, including twenty-eight or twenty-nine questions, further dialogue between Subhūti and Śāriputra, and in chapters 11, 12, and 13 a long discussion of the Great Vehicle, its attributes, and its results. This entire intermediate teaching is sometimes referred to as “the chapter of Subhūti,” which is also the chapter title this text gives to the last chapter in this section, the thirteenth; that chapter title may be intended to cover the entire group of chapters 3 through 12, too. The intermediate teaching corresponds to the first chapter of The Perfection of Wisdom in Eight Thousand Lines, is centered on all-aspect omniscience, and by teaching nonconceptual perfection of wisdom focuses on ultimate truth.

(3) The detailed teaching is covered by most of the rest of the text, from chapter 14 to the end of chapter 71 (it does not include the Maitreya chapter or the final Sadāprarudita and Dharmodgata chapters). It contains a long series of points to be explained and is divided into two parts: (a) The first part begins with the questions put to Subhūti by Śakra, lord of the gods. The first ten of its chapters, 14 through 23, in which Śakra figures prominently, are probably intended to be covered by the title “Śakra” given in the chapter colophon of chapter 23, and among many important points they explain are how a bodhisattva’s knowledge encompasses that of śrāvakas and pratyekabuddhas, and the benefits, protections, and vast merit conferred by appreciating, honoring, and knowing the perfection of wisdom. Subsequent chapters cover a range of topics, from the dedication of merit, in chapter 24, up to how a bodhisattva’s progress to awakening becomes irreversible, in chapter 40. (b) The second part comprises the discussions prompted by Subhūti’s two hundred and seventy-seven questions, which start in chapter 41 and end in chapter 71. It is centered on knowledge of the paths, and in teaching both conceptual and nonconceptual perfection of wisdom it explains both relative and ultimate truths.

The eleven formulations (rnam grangs bcu gcig) are mentioned in several commentaries detailing the interlocutor concerned, but are not explicitly correlated with particular locations in the texts. Following the mentions in the two bṛhaṭṭīkā commentaries (Toh 3807 and 3808) and Sparham, we may speculatively identify some of them in the present text as follows:

• (1) to Śāriputra, from the beginning of chapter 2 as far as 2.245;

• (2) by Subhūti, the rest of chapter 2 to the end of chapter 13;

• (3) to Śakra, chapter 14, but possibly including several subsequent chapters as well;

• (4) to Subhūti, difficult to identify;

• (5) to Maitreya, the first part of chapter 24;

• (6) to Subhūti, perhaps the rest of chapter 24 from 24.46;

• (7) to Śakra, perhaps chapter 25 to 25.17;

• (8) to Subhūti, perhaps starting in chapter 25 at 25.18, but very likely chapters 41 to 71;

• (9) to Maitreya, chapter 72;

• (10) to Subhūti, again difficult to identify with any certainty; and

• (11) the narrative of Sadāprarudita and Dharmodgata and entrustment of the sūtra to Ānanda, chapters 73 to 76.

It is possible that the list is not intended to follow a sequential order in all instances. In any case, further study on this approach to the sūtras would be desirable.

The eight topics and seventy points of The Ornament of Clear Realization cannot be mapped with complete precision to the content of this version of the sūtra, as explained above (i.24)‍—and indeed that is the raison d’être of the Tengyur version. Nevertheless, included in the following list of the eight principal topics and their subdivision into seventy points is an approximate matching of the eight topics themselves to the content of the chapters of this sūtra, based on how the equivalent passages of the Sanskrit text can be identified in the Tibetan of this version. The seventy points of the Ornament are more difficult to match with the present text, and indeed many of them are not even mentioned in the sutra; some, however, can be discerned‍—not always in the same order as in the Ornament.

(1) All-aspect omniscience (sarvākārajñatā, rnam pa thams cad mkhyen pa nyid) is covered in chapters 2 to 13. Its ten topics are (i) setting of the mind on enlightenment; (ii) the instructions concerning its application within the Great Vehicle; (iii) the four aspects of becoming established on the path of preparation, comprising warmth, peak, acceptance, and supremacy; (iv) the naturally present affinity with the spiritual family, which is the basis for attaining the Great Vehicle; (v) the referents through which the Great Vehicle is attained; (vi) the goals attained through the Great Vehicle; (vii) the armor-like practice; (viii) practice through engagement; (ix) practice through the provisions of merit and wisdom; and (x) practice of definitive deliverance.

(2) Knowledge of the path (mārgajñatā, lam gyi rnam pa shes pa nyid) is covered in chapters 14 to 27. Its eleven topics are (i) essential aspects for understanding the path; (ii) knowledge of the path which is that of the śrāvakas; (iii) knowledge of the path which is that of the pratyekabuddhas; (iv) the beneficial path of insight, which accords with the Great Vehicle; (v) the functions of the path of cultivation; (vi) the aspirational path of cultivation; (vii) the path of cultivation resulting in eulogy, exhortation, and praise; (viii) the path of cultivation resulting in dedication; (ix) the path of cultivation resulting in sympathetic rejoicing; (x) the path of cultivation resulting in attainment; and (xi) the path of meditation resulting in purity.

(3) Knowledge of all the dharmas (sarvajñatā, thams cad shes pa nyid) is covered in chapter 28. Its nine topics are (i) the basic understanding that wisdom leads to not remaining in cyclic existence; (ii) the basic understanding that compassion leads to not remaining in quiescence; (iii) the basic understanding that lack of skillful means leads to distance from the perfection of wisdom; (iv) the basic understanding that skillful means lead to proximity to it; (v) the basic understanding of the discordant factors associated with the fixation of śrāvakas and pratyekabuddhas; (vi) the remedial factors countering those fixations; (vii) training in the aforementioned basic understandings; (viii) training in the sameness of those basic understandings; and (ix) the path of insight which integrates these basic understandings.

(4) Clear realization of all aspects (sarvākārābhisambodha, rnam kun mngon rdzogs rtogs pa) is covered in chapters 29 to 44. Its eleven topics are (i) the aspects of the three kinds of knowledge (1–3 above); (ii) training in those aspects; (iii) the qualities acquired through those trainings; (iv) the defects to be eliminated during training; (v) the defining characteristics of training; (vii) the path of provisions in accord with liberation; (vii) the path of preparation in accord with the four degrees of penetration (1.iii above); (viii) the signs of irreversibility in bodhisattva trainees; (ix) training in the sameness of cyclic existence and quiescence; (x) the training associated with the pure realms; and (xi) training in skillful means for the sake of others.

(5) Culminating clear realization (mūrdhābhisamaya, rtse mor phyin pa’i mngon rtogs) is covered in chapters 45 through to the first few paragraphs of chapter 59. Its eight topics are (i) the culminating training of warmth on the path of preparation; (ii) the culminating training in the peak on the path of preparation; (iii) the culminating training in acceptance on the path of preparation; (iv) the culminating training in supremacy on the path of preparation; (v) the culminating training on the path of insight; (vi) the culminating training on the path of cultivation; (vii) the culminating uninterrupted training on the path of cultivation, comprising the adamantine meditative stability; and (viii) the mistaken notions that are to be eliminated.