Literary Sources of the Gunakarandavyuha


- W. B. Douglas

  1. Introduction

  2. Identifying the sources

  3. Literary sources

  4. Borrowing authority

  5. References


1. Introduction

The Gunakarandavyuha, a Buddhist Sutra composed in Sanskrit by Newars probably in the 15th century, shows a broad familiarity with the literature of later Indian Buddhism. Its title, which points toward its most important source, the Karandavyuha, has led some scholars to assume it is only a verse expansion of that text, but this would be a mistake. Other important sources include the Bodhicaryavatara, the Manjushrinamasamgiti, and late Indian manuals describing the Poshadha ritual. There are also as yet unidentified poetic sections, and narrative episodes which are expansions of Avadana material.


A study of the sources of the Gunakarandavyuha (hereafter GKV) and the way in which they are adapted begins to show something of the sophistication of its 15th century composer (or composers). The process of writing new Buddhist texts in Sanskrit had ended everywhere else by this time. Right through the 12th century, we can say that the Buddhists living in the Valley of Nepal were part of a broader community of Indian Buddhists stretching from Kashmir to Indonesia. Newar Buddhist pandits taught in the great Indian monastic universities; other pandits came to Nepal to visit sacred sites or to study.1 In the tumultuous decades when Buddhism lost most of its material foundation in India, the valley of Nepal became a safe heaven for the continued practice of Sanskrit Buddhism. There were a few other regions where Sanskrit Buddhism still hung on - Kashmir, Bali, eastern Bengal - but only in the valley of Nepal was the Buddhist culture strong enough to take root and flourish, as indeed it still does today. In the centuries after the collapse of mainstream Indian Buddhism, the Newar Buddhists had to adapt and localise the great tradition, now bereft of its pilgrimage sites, its great universities, its oceanic trade routes, and its political patronage. The GKV, along with a handful of other distinctively Newar Buddhist texts,2 was written as a part of this process of localization. To the extent that it draws on Indian sources, and indeed claims to be a version of an old and respected Indian Buddhist text, it is asserting community with and demonstrating knowledge of that vanishing tradition.


In this paper, I will outline some of the identifiable sources of the GKV. These can be divided by the degree to which they literally form a part of the text; some are clear influences, while others actually contribute material to the body of the text.


I will begin by identifying some of the Indian sources which are definite influences on the GKV. then, I will break up the GKV schematically into three categories: entirely original material, material which is directly traceable to previous texts, and the majority of the text, in which the author(s) built on a dominant source, transforming it into poetry and weaving new material into the older sources.


In order to look at how the GKV handles its sources, we will then look at two examples of these different processes: chapter eighteen which consists largely of material lifted directly from the Bodhicaryavatara; and the episode of Avalokiteshvara rescuing the worms in Varanasi, where a story from the Karandavyuha (hereafter KV) is rendered into poetry with the addition of new information on the advanced stages of the path. Finally I will return to the claim, implicit in the title that the GKV is just another version of the KV and consider what force this would have given the project of the GKV.3


2.  Identifying the sources

The limitations of literary sources. We can divide the sources for the GKV into three: literary sources, iconographic sources, and ritual sources. These are not mutually exclusive categories; the GKV itself is an example of a mahatmya, which is not primarily a ritual manual, but it nonetheless contains some ritual information. It is only the literary sources which can be said to contribute textual material directly to the GKV. The two primary literary sources for the GKV are the KV and the Bodhicaryavatara.


Yet in neither of these texts will we find any mention of the Poshadha Vrata, a ritual which is the centre of the GKV. The most important epithet of the Buddha, Shri Ghano Buddha, is similarly absent from these classical sources; and Yama’s marvelous outpouring of names and forms in his greeting to Avalokiteshvara in chapter two draws on a wide range of iconographic forms are largely late Indian in derivation. The importance of these non-literary elements indicates the influence of specific texts and traditions which could not otherwise be inferred. A full account of all the iconographic elements of the GKV would go far beyond the remit of this paper, however.


2.1 Avadana materials

The KV and the GKV inherit some of their episodes from the Divyavadana and other avadana collections. Best known is the episode of Simhalasarthavaha, the merchant who travels to Shri Lanka with 500 men and encounters there the rakshinis who charm his crew. This storu took strong root in Nepal and, with the substitution of Tibet for Shri Lanka became the story of Shri Sangha Bo. Siegfried Lienhard has traced the evolution of this story as well as that of the magical horse Balaha.4 Although the author(s) of the GKV were working in a cultural context which prized the avadana stories and indeed created vast compilations of them as part of the same rituals that the GKV endorses, there is no new avadana material in the GKV which is not supported by the pre-existing narrative of the KV.


2.2 The Manjushrinamasamgiti

The MNS is apparently the source for the epithet of the Adibuddha used in the GKV, Shri Ghano Buddha. The term ghanah first appears in the Manjushrinamasamgiti. Although its appearance in the GKV does not necessarily reflect a direct infuence from that text, it is an unusual epithet in the later Indian Sanskrit texts5 and does not occur in any other Avalokiteshvara texts that I know of. The term is, however shared with the Svayambhu Purana in the longer recension and also appears in later Newari songs.6 The first verse of the GKV7 reads:


            1. yah shrighano buddhah sarvalokadhipo jinah |

            te natham sharanam gatva vakshye lokeshasatkatham* ||


            * T1 N1 N2 C1 H1; J1 has a totally reworked Nath mangalam.


In the Svayambhupurana of 12 chapters,8 the closing verse of the first chapter reads


            ity adishtam munindrena shrighanena nishamyate |

          maitreyadisabhalokah sarve ‘pi sampramoditah* ||


            * ed. sampramodite


In the MNS the term ghanah is found at 6.20:9


            6.20 ghanaikasaro vajratma sadyojato jagatpatih |

          gaganodbhavah svayambhuh prajnajnanalao mahan ||



The term ghano is glossed by Ravishrijnana as nividatvat, “impenetrable” or “dense”. It is unclear whether the term is first used in the Swayambhu Purana or the GKV, but it would appear that the epithet is a distinctive  feature of Newar Buddhism, derived from the MNS, which is in palce by the 15th century and preserved as an epithet of the Adibuddha thereafter.


Further evidence for influence of at least the vocabulary of the MNS is the occurence of the unusual term rad, meaning “crown” or “king”. In the MNS it occurs in the compound dharmarad; in the GKV we find it in chapter two in the name Yamarad.


2.3 Poshadha ritual material

The central concern of the GKV is the promotion of the Poshadha Vrata. I will not here go into details of this claim, or even begin to list the number of places in the text where it is recommended, or where its benefits are described. However, the form of the ritual understood and described in the text is clearly not simply the lay folk coming to hear stories; it is an elaborate ritual of Amoghapasha as described in the modern case by John Locke,10 and as I have documented in the fourteenth century.11 This is the well known ritual fast also called the Ashtami Vrata. The oldest manuals describing this form of the Poshadha Vrata exist in Tibetan translations of the thirteenth century, and are perhaps eleventh century in origin. There are a handful of texts describing the Poshadha ritual in the bsTan ‘gyur,12 and there are Poshadha riutal texts in Sanskrit in Nepalese collections which have not been dated or properly edited.13 The cult of Amoghapasha with its Poshadha Vrata seems to have been a late development in Indian Buddhism, perhaps with Kamirian roots, which blossomed in Nepal and became the basis for the cult of Karunamaya. Indeed, I believe the GKV was written as part of the expansion and promulgation of this cult, which was originally centred at Bungadyah.


Given that the original purpose the Poshadha, or upavastha, ritual was the edification and instruction of lay Buddhists through oral teachings, it is not at all surprising that the two major works associated with the Poshadha vow are collections of stories. The GKV, like the KV, is a collection of episodes celebrating the activities of Avalokiteshvara in saving all sentient beings without exception. Some of these episodes, as I mentioned above, are drawn from the vast pool of avadana literature. The other major corpus of texts associated with the Poshadha ritual, the Ashtamivrata Katha texts, are collections of avadana and Jataka material which were intended for narration during the performance of the fast.



It is thus interesting, as I noted above, that the GKV is surprisingly conservative in not taking over some of the same story material found in the Ashtamivrata Katha. The close association of those stories with performances of the ritual is demonstrated by the illustrations in the paubha or painted scrolls which were sometimes produced to commemorate the performance of the Poshadha Vrata.14 Nonetheless, the GKV promotes the efficacy of the Poshadha Vrata without drawing in any stories like the Virakusha Avadana that exemplify its power; instead, preserves the body of Avalokiteshvara miracle stories from the KV. The two collections of stories seem to complement each other, and indeed the GKV’s essentially conservative nature is not matched either by the Ashtamivrata Katha manuscripts, which can contain a selection of different avadanas, or the Svayambhu Purana which has developed several different recensions. Brinkhaus (1993)


3. Literary sources

Literary style     

Before commencing a discussion of the textual heritage of the GKV, it will help us to establish the style of the text, which determines the degree to which it can borrow directly, and explains why it is often easy to notice borrowings from older Sanskrit texts. The style of the GKV swings between a rather mechanical use of metric units and bursts of accomplished and beautiful poetry. In form it is almost entirely shlokas, the most important exceptions being the trishtubh section of the first chapter and the more complex verses, largely borrowed from the Bodhicaryavatara, in the 8th and 18th chapters.


The repetitive use of formulaic material dominates much of the work, and in this it is reminiscent of other late, non-Buddhist, puranas. Stylistically it is most similar to its sister text, the Swayambhupurana. As the entire text is a series of nested dialogues in praise of Avalokiteshvara, these standard padas often fall into one of four categories: dialogue framing narrative, epithets of a speaker, descriptions of corrective devotion to Avalokiteshvara, or formulaic descriptions of the path along with Avalokiteshvara propels the countless beings whom he is determined to rescue from samsara.



So for example the pada tam alokyaivam abravit, occurs frequently introducing dialogue; the variation samalokyaivam adishat occurs three times in the first chapter. In the flow of shloka metre, this fits the even padas. Similarly, the epithet jinatmajah occurs at the end of even padas describing Buddhas, Avalokiteshvara, or other great Bodhisattvas, where odd padas will often have bodhisattvo mahasattvah. At this level, the similarity of the GKV to the Svayambhu Purana is most apparent; they share many of these repeated padas and the limited vocabulary used to build them. We can see this clearly by comparing the closing verse from the Svayambhu Purana cited above at 2.2, with the closing verses from chapter XII of the GKV.

            23cd. ity adishtam munindrena vishvabhuva nishamya te |

            24. sarve sabhashrita lokah prabhyanandan prabodhitah||


On a larger scale, lists of the various creatures such as yakshas, rakshas, kumbhandas, and so on recur, as do descriptions of the progress of beings toward enlightenment. Half verses such as kramena bodhisambharam purayitva samahitah in chapter 19 or parishuddhatrimandalah15 are repeated with variations in similarity constructed lists of varying lengths which describe, in greater lr lesser detail, the progress of beings from devotion to Avalokiteshvara expressed in meditation, remembrance, and uttering his name, through rebirth in Sukhavati and their eventual attainment of enlightenment.


For the historian, one of the most interesting repeated units is the standard description of the social order. This is repeated in chapters I, XVII, and XVIII at least. First the professional religious are listed: bhikshunyash cailakash caivam upasaka upasikah vratino ‘pi mahasattvah; then the social order is traversed: brahmanah kshatriyash capi rajano mantrino janah amatyah shresh hinah paurah sarthavaha mahajanah; the urban-rural order is outlined: janapada gramyah parvatikash ca naigamah ; and finally the question of origins: tathanyadeshika likah.16


This sort of repetition is often said to be characteristic of oral poetry; the metric units aid in composition and recitation, and the lists with their fixed elements and variable length allow for a degree of embellishment appropriate to the situation. Yet in the GKV, these devices, which in a written work are not usually regarded as a sign of high art, are mixed with more carefully crafted verses. Rather than an artifact of oral tradition, this repetition is more the mark of hack versification and a need for an efficient means to churn out a long narrative poem. The quality of the verse varies directly with the interest of the passage, and when describing miraculous wishing trees appearing in the Jetavana garden, the author turns out much more artful verses.


3.1 Analysis of the GKV

In overview, the completely original matter in the GKV can be summarised as follows:


The narrative frame       

It has an original narrative framework which is considerably more complex than that of the KV. With other Newar Sanskrit Buddhist texts, it has a double framing narrative which is a dialogue between Jinashri and jayashri enclosing and mirroring a dialogue between Ashoka and Upagupta. This in turn encloses a set of narrative frames, the most important of which is Shakyamuni talking to Sarvanivarananishkambhi, and this in turn sometimes encloses or shifts to similar dialogues between previous Buddhas (Vishvabhu, Vipashvi, and so on) and their respective interlocuters. The structure is quite a puzzle box, and allows, for instance, for one assembly to experience a direct visitation from Padmapani Lokeshvara (in ch. VI) while preserving the unbearable longing of Sarvanivarananishkambhi, in the enclosing narrative, to himself encounter Avalokiteshvara.


Beginning and end           

Its opening and closing chapters, chapters I and XIX, are entirely new material, the first a long description the merits of worshipping the triple jewel and the last a rather political exhortation to the Jayshri/Ashoka characters to run their kingdoms in confirmity with the Buddha Dharma. The internal chapters are to varying degrees adaptations of episodes from the KV, occurring in the same order. Chapters VIII and XVIII, however, are largely selected extracts from the Bodhicaryavatara, along with a set of as yet unidentified verses in longer metres which repeat between the two chapters. The most important block of original material is probably contained in the first chapter, which contains an internal poem of some 56 verses (I.102-158) in trishtubh, addressed to Ashoka by his preceptor Upagupta, recommending performance of the Poshadha ritual.


GKV XVIII     BCA citation    

53-111      V.1-59 

112-144    V.70-102     

145           V.108  

146           V.107  

148-152   VI.1-6         

153-6       VI.8-12

157-8       VI.112-3  

159-60     VI.126-7  

161-5       VI.130-4  

166-8       VII.1-3

176-7       VII.74-5  

178-93     VII.1-16

200-204   IX.1-5         


Table 1: Citations of the Bodhicaryavatara in chapter XVIII of the Gunakarandavyuha


3.2 Direct borrowings

The clearest examples of direct borrowing in the GKV are the eighth chapter, the Balisambodhanabodhimargavatarana or Introduction to the Path and Enlightenment of Bali, and the eighteenth chapter, the Shikshasamvarasamuddesha or Instruction concerning the student’s vows. Tucci in his rather dismissive study of the GKV has given a preliminary list of these citations; he notes that nearly half of the Bodhicaryavatara is inserted into these two chapters of the GKV. For the purposes of this paper, I will look only at the eighteenth chapter, as it contains further material which may have been borrowed, but which cannot be traced to the Bodhicaryavatara.


There is an identically titled chapter in the KV, the last chapter of the second section, but in the GKV this chapter consists largely of extracts from the Bodhicaryavatara. The introductory verses describe the awful fate awaiting monks whose behaviour is less than exemplary, following a similar section of the KV17 but after XVIII.52 the citations begin. I have detailed which verses of the GKV cite which of the Bodhicaryavatara in table 1.


Differences between the GKV and its source. The text in these citations shows a few variations against the edited modern Sanskrit text18. In some cases, this is the result of scribal error somewhere in the transmission. For example, BCA VIII.11ab is omitted, and XVIII.66a has vakya where BCA 5.13a reads bahya. In other cases, an alternate reading is preserved. XVIII.58 reads

            58. yasmad bhayani sarvani duhkhany apramitany api |

            cittad eva samudyanti sarvesham bhavacarinam* ||


 * kathitam tattvavadina: C1 adds


While the printed edition, which unfortunately has no critical apparatus, has

            V.6 yasmad bhayani sarvani duhkhany apramitany api |

            cittad eva bhavantiti kathitam tattvavadina ||


The scribe of C1 apparently had the Bodhicaryavatara in front of him. Two verses previously, where the GKV tradition reads vyagrah siaha gaja rikshah sarvam ca dushtashatravah,19 he alone writes vyagrah simha gaja rikshah sarpa sarve ca shatravah with the Bodhicaryavatara. It is not surprising, then, that confronted with the variant reading in verse 58 he decides to keep both readings.


As an example, this variant reading is actually somewhat perplexing; the meaning is distinct between the two, yet there is nothing crucially different; there does not appear to be a doctrinal point being defended for instance. The version of the Bodhicaryavatara might be translated, "Because all fears and limitless suffering arise from the mind alone, says the speaker of truth." The GKV version reads, "Because all fears and limitless suffering spring up only from the mind of all those passing through existence." I have not consulted older manuscripts of the Bodhicharyatara from the National Archives to see if this variation is limited to the GKV or if, in fact, it is traceable to a particular recension of the Bodhicharyatara. If so, it would give us valuable information about the date of composition of the GKV.


3.2.1 Citations

Occasionally one encounters sections of the text which must be citations, although I have not yet been able to identify most of them. As an example, there is a short stotra at the end of the fifth chapter on the Asuras. Avalikiteshvara has been teaching them in the form of an acharya, and when he is satisfied that they are established on the Buddha’s path (bodhimarge pratish hapya V.48, 62, 98, etc.) he leaves; just before he disappears he reveals himself.


tatah so ‘ntarhitah khe sthah prabhasayan samantatah |

dhritva lokeshvaro murtim sarvans tan samadarshayan ||

tam akashe prabhasantam lokeshvaramjinatmajam |

drish va te danavah sarve babhu vismayanvitah ||


They respond by worshipping him as he disappears with the following verses:20


namaste bhagavan natha sada te sharane sthitah |

bodhicaryavratam dhritva carama tat prasidatu ||

yad asmad aparadham tat kshantavyam bhavata sada |

evam asman samalokya sampalayitum arhati ||


I cannot be sure that this is not the work of the composer of the GKV, but I suspect that it is either a prayer common at the time of the composition of this text, or a citation from a popular source.


The verse material of chapters 8 & 18 Tucci21 asserts that the other ornate verse material contained in chapters 8 and 18  must also be a borrowing from some as yet unidentified source. In this opinion, the compilers of the GKV, an 'interminable litany' which "monotonously repeats the basic core of the work,"22 were incapable of producing such verses. his alternative hypothesis is that they must come from some as yet unknown work of Shantideva on the subject of the six perfections.


I am not as convinced as Tucci was that the compiler or the GKV were completely incapable of producing such verse. The quality of the verse in the GKV ranges from the merely mechanical upwards. The closing verse of the first chapter, much of which is in anushtubh, is shragdhara. Furthermore, the continuous aspiration of Newar Sanskrit composition is expressed in a steady production of Sanskrit plays and poems in the medieval period. In the 15th century we know that the Indian Buddhist pandit Vanaratna, who had trained with a Sanskrit poet in the south if India, came to settle in the kathmandu valley. thus both of the cultural milieu and the presence of expertise argues against Tucci’s flat denial.


However, demonstrating the possibility that these verses might have been composed in Nepal is not at all the same as proving that they were composed by the author(s) of the GKV. Two facts make it rather difficult to argue the case. First, these verses are aesthetically the best material in the work (outside the extracts from the Bodhicharyatara, which they certainly rival); and second, they occur only within the two blocks of borrowed material and in fact are repeated between the two chapters 8 and 18. Compared with the shragdhara verse in the first chapter, they show a richness of vocabulary and technique; and I am forced to agree with Tucci that they probably are borrowed material.


Tucci’s proposal, that they belong to an unknown composition fo Shantideva, seems to me wishful thinking. Collections of ornate Buddhist devotional verse are a staple of the Newar Buddhist literary field. The Lokeshvarashatakam, a poem consisting of one hundred shragdhara verses in praise of Avalokiteshvara, has remained popular in the valley although it is dense and filled with puns which appear to have defied the Tibetan translator; its manuscripts usually include a commentary or Newari gloss. Although I agree with Tucci that these verses are taken from a single source, I do not think there is good evidence to argue that they must be necessarily from Shantideva’s pen. If they are, in fact, the only remaining verse from an otherwise lost work then the skillful use by the author of the GKV of his Indian Buddhist textual heritage will have an additional value for us as scholars.23


3.3 Adaptations

The GKV tends to follow the narrative line of the KV, converting the high Sanskrit of the older text into verse form, often simplifying the language and adding embellishments. When the GKV has new material to add, such as a recommendation to perform the Poshadha vow or a description of the benefits of Avalokiteshvara’s name, it usually inserts this this material into natural breaks in the flow of the KV narrative. This process can be seen at its simplest in a comparison of a rather short episode in which Avalokiteshvara manifests as a bee to rescue the worms living in Varanasi. In the KV this is the fifteenth chapter of the first section; it is the twelfth chapter of the GKV. An uncritical editiona and English translation of chapter 12 can be found as an appendix to this article.


The story is simple. Avalokiteshvara decides to rescue the worms living in the sewage of Varanasi. He takes the form of a bee, and hums “Namo Buddhaya, namo Dharmaya, namo Sanghaya.” The worms, hearing him, begin to repeat and remember this phrase, and by the merit acruing from that are reborn in Sukhavati.


Here is the moment of the worms hearing him and taking up the sound from the KV:

     tad esham shabdam nishcarayati namo buddhaya namo dharmaya namo sanghaya iti | tac chrutvate ca sarve pranakah namo buddhaya namo dharmaya namo sanghayeti namam anusmarayanti |


And here it is in the GKV:

8.  namo buddhaya dharmaya sanghayeti pranoditam |

madhurashabdam uccarya bhramate sa viyaccaran ||

9. tam khebhramantam alokya sarve te pranakash capi* |

tat kalaravam akarna cintayen hy evam utsukah ||

10. aho yam sujhavan pakshi bhramate khet ‘pi** yatheccaya |

kim anena kritam punyam yenaivam carate sukham ||

11. kim asmabhih kritam papam yenamedhyashrita vayam |

iti vicintya te sarve krimayas tatsukhucchitah ||

12. tad viracam anishrutva samtish hante tad unmukhah |

tatha te krimayah sarve tannamasmritibhavitah ||


* pranaka api: T

** ti: T


The author of the GKV has embellished and enlivened the story in the process of setting it to verse. This is counterbalanced by a tendency to simplify the language of the KV, sometimes quite drastically. The KV goes on to tell us that the reason for their rebirth as worms, at the same time as it expresses the power of their liberation:

     te ca sarve buddhanamasmaranamatrena vimshatishikharasamudgatam stkayadrish ishailam jnanavajrena bhittva, sarve te sukhavatyam lokadhatav upapannah


The GKV leaves this out entirely, however; and in general it does omit much that is complex or even witty in the KV. It does, however, insert new doctrinal material where it can, and in this case the career of the ex-worms is expanded not only by the addition of an intermediate incarnation as flying ensects, but once in Sukhavati, their progress in attaining various stages in explored in more detail: in particular their attainment of the “triple Awakening” (trividham bodhim) is noted. In other contexts the progress through accumulating the requisites of enlightenment (bodhisambharam) and the six perfections is detailed. This reflects a concern with path-doctrine typical of later Indian Buddhism that has little place in the KV.


4. Borrowing authority

I mentioned above the essentially conservative nature of the GKV. This is, I think, tied to the claim implicit in its title to be a version of the KV. Indeed, scholars working in this material have frequently been badly confused by relation between the two texts, and it is only Tucci and Regamy among Western scholars of the last generations who have correctly understood it to be a later, Newar production. This confusion is an unintentional side effect of a deliberate attempt to identify the Newar text with the older.


In the classical Indian scholastic milieu, plagiarism within a textual tradition was not an act of intellectual theft but the assertion of membership within a valid tradition. the paradigm of learning was repetition and commentary. The old texts accumulated new explanations and paraphrases which depended for their position and authority on the original text, still repeated. This claim to authority, which simultaneously denies the originality of the present work while deriving effective authority for its truth claims from the status of the parent work (often at the expense of that work’s original position) is well known in scholastic literature as the production of commentaries.


The problem is more difficult for the would-be authors of sutras. Discussions of buddhavacanam occupied the scholars of classical Indian Buddhism, but the cannon was never truly closed. Rob Mayer writes “ It is certainly true that fresh revelation of new scripture continued unbroken in Buddhist India right up until the end.”24 Indeed, as we see from this material, the process of generating Buddhist Sanskrit scriptures continued on well after the end of Indian Buddhism as part of the process whereby Newar Buddhists localised and grounded their own form of Buddhism after its historical support in Indian Buddhism had withered away.


Some sutras did accumulate commentaries,25 and those which were understood as tantras often developed great trees of commentary comparable to the shastras. This process kept the schools founded on those root tantras alive and relevant.


The earliest manuscripts of the KV are from Gilgit, written no later than 630 CE. Adelhard Mette  has argued on this basis of textual divergence in the two oldest manuscripts that the original text is from perhaps the fourth century.26 By the time of the composition of the GKV some thousand years later, the mantra and sadhana of Avalokiteshvara contained in the second nirvyuha of the KV would have marked it as a Vajrayana text, comparable to those classified as Kriya Tantras. The KV is thus a late Mahayana text and an early Tantra, probably the earliest tantra of Avalokiteshvara.27 It appears, however, not to have attracted any Indian commentaries in the millennium between its emergence and the composition of the GKV. This makes it less likely that the author of the GKV could have positioned his own work as a commentary on the earlier text.


Their strategy, instead was to claim that the GKV was in some sense identical to the KV, or a simple enhancement of it. Both are usually known in their colophons as Arryavalokiteshvara-guna-karandavyuha-mahayana-sutraraja. The difference in style and size of the two texts means that scribes and priests working with the two texts would see them as unambiguously distinct. The claim to the name of the older text was, however, accompanied by a close adherence to its subject matter and narrative sequence.


The claim by the GKV to be an extension of the earlier text has the double effect of relying on the older textual tradition and revitalising it. Thematically, the GKV imports a great deal of new material into the narrative structure in inherited from the KV. Its description of the path, understanding of the social order, iconography, and emphasis on the Poshadha Vrata are all new. Utilising the existing structure of the KV was a way of authorising validating this new material, and especially of lending weight to its prescription of the Poshadha Vrata and its particular cult of Amoghapasha. At the same time, rooting this new text in the Indian Buddhist tradition has the effect of asserting the continuity and vitality of the lost Indian tradition in this new medieval Newar scripture. In fact, through the breadth of its sources, the GKV demonstrates the continuity and range of the Indo-Newar Buddhist heritage from the fourth century to the fifteenth.




1. atha gaganaganjo ‘sau bodhisattvah kritanjalih

vishvabhuvam munindram ca pranatvaivam avocat

2. bhagavan sa mahasattvo lokeshvaro jinatmajah

nadyapiha samayati kadagacchet tad adisha

3. iti tenoditam shrutva vishvabhuh sa munishvarah

gaganaganjam alokya tam punar evam abravit

4. tatah samprathitash casau lokeshvaro vilokayan

varanasyam samuddhartum sattvan samabhigacchati

5. drishtva sa pranino ‘nekan asamkhyeyan suduhkhitan

savidmutramridalagnan tish haty evam vicintayan

6. ha papam katham etani savidmutrashritan aham

krimyasamkhyasahasrani proddhareyam prabodhayan

7. tatra sa cintayan matva kripaya samvilokayan

bhramararupam Adhaya bhramate tad upacaran

8. namo buddhaya dharmaya samghayeti pranoditam

madhurashabdam uccarya bhramate sa viyyacaran

9. tam khebhramamtam alokya sarve te pranakash capi

tatkalaravam akarna cintayen hy evam utsukah

10. aho yam sukhavan pakshi bhramate khe ‘pi yatheccaya

kim anena kritam punyam yenaivam carate sukham

11. kim asmabhih kritam papam yenamedhyashrita vayam

iti vicintya te sarve krimayas tatsukhecchitah

12. tad viracam anushrutva samtish hante tad unmukhyah

tatha te krimayah sarve tannamasmribhavitah

13. tat triratnam namaskaram dhritva tish hanti cetasa

tatha caivah samuccarya triratnanamacetasa

14. smritva kritva namaskaram tish hanti trimaner muda

etatpunya viliptas te sarve samjatapakshakah

15. tata uddiya gangayam nipatantas tyajanty asun

tatas te vimalatmanah samprayatah sukhavatim

16. sarve sugandhamukha nama bodhisattva ‘bhavanty api

te tatramitabhasasya pitva dharmamritam sada

17. trividham bodhim asadya nivritipadam apnuyuh

evam asau mahasattvah lokeshvarah krimin api

18. prayatnena samuddhritya preshayati sukhavatim

tatha tasya jagacchastuh punyaskandham mahattaram

19. aparameyam asamkhyeyam ity akhyatam munishvaraih

ye casya lokanathasya shraddhya sharane sthitah

20. dhyatva smritva sammuccarya nama bhajanti sarvada

sada te sadgatau jata durgatau na kada cana

21. saddharmashrisukhapannah samprayayuh sukhavatim

tatramitabhanathasya pitva dharmamritam sada

22. trividham bodhim asadya samprapsyanti jinalayam

iti matva sada bhadra saukhyam icchanti ye bhave

23. te ‘sya trailokanathasya bhajantu sharane sthita

ity adishtam munindrena vishvabhuva nishamya te

24. sarve sabhashrita lokah prabhyandan prabodhitah

iti varanashikrimiki oddharanaprakaranam samaptam


2b : adisha] adishah T1

4b : samabhigacchati] samabhigacchati T1

5b : tish haty] sama hy J1

9a : pranakash capi] pranaka api T1

10a : khe ‘pi] kheti T1

13a : namaskaram] askaram T1


B  Transcription of KV chapter 15

This is only a transcription for the purposes of comparison from Vaidya’s text. It must be

regarded as provisional until A. Mette’s edition comes out.28

     atha aryavalokiteshvaro bodhisattvo mahasattvas tasmat simhaladvipad avatirya varanasyam mahanagaryam uccaraprasravasthane gato yatranekany anekani krimikulashatasakasrani prativasanti | tato ‘valokiteshvaro bodhisattvo mahasattva upasamkramya tatrasa tani pranishatasahasrani drishtvaatmanam bhramararuupam abhinirmaya ghunaghunayamanam | tad esham shabdam nishcarayati – namo buddhaya namo dharmaya namah samghaya iti | tac chrutva te ca sarve buddhanamasmaranamatrena vimshatishikharasamudgatam satkayadrish shailam jnanavajrena bhittva sarve te sukhavatyam lokadhatav upapannah sugandhamukha nama bodhisattva babhuvuh | sarve te bhaga vato ‘mitabhasya tathagatasyantikad idam karndavyuham nama mahayanam shtutva anumodya ca nanadigbhyo vyakaranani pratilabdhani || atha aryavalokiteshvaro bodhisattvo mahasattvo sattvaparipakam kritva tasya varanasya mahanagaryah prakrantah ||


C  English translation

1.  Now this bodhisattva Gaganaganjo, respectfully folding his hands, bowed and addressed the great sage Vishvabhu as follows:

2.  ‘Lord, great Lokeshvara, born from the Victor himself, is not now coming here. Please tell us this, when might he come ?

3.  Hearing what he had said, Vishvabhu tha great sage looked at Gaganaganja and spoke him again in this way:

4.  ‘Next (after the event of the last chapter) Lokeshvara set out looking about to rescue beings in Varanasi, and went there.

5.  He saw countless miserable beings covered in feces, urine, and mud; and he paused, thinking.

6.  “Oh, evil! How should I rescue these countless thousands of worms, dwelling in the sewage, and enlighten them?”

7.  Thinking there, he considered, and gazed on with pity; then taking the form of a bee he wandered along towards them.

8.  He advances singing “Namo Buddhaya Dharmaya Sanghaya” in a sweet voice and wandered irregularly along.

9.  When all those being saw him wandering in the air, and heard his rhythmic sound, they longingly thought:

10. Oh! Such a happy bee wanders in the air as it pleases. What merit did he do that he can travel so happily?

11. What evil have we done we dwell in sewage?” Thinking this all the worms wanted his joy.

12. Listening to his song, they stayed staring up at him. Then all the worms began to recollect his name.

13. Doing homage to the triple gem, they carried it in their minds, and then also sang it, with their minds on the triple gem.

14. Remembering it, revering it, they dwelt in the triple gem with joy; and anointed by this merit they were all reborn as flies.

15. Then flying over Ganga they fell, abandoning their lives. With stainless personalities, they progressed to Sukhavati.

16. They all became bodhisattvas with the name of “Good-scented face.” There, they drank the Dharma-nectar of Amitabha constantly.

17. They achieved the threefold Awakening, and attained the stage of Nirvana. Thus did Lokeshvara rescue even the worms by (his) exertions,

18. And conveyed them to Sukhavati. That is the greatness of this World-teacher’s mass of merit.

19. Immeasurable and uncountable, so say the best sages. And those who take faithful refuge with the Lokanatha

20. Contemplating, saying, recollecting his name and worshipping him always, they always have good rebirths and never bad.

21. Attaining the joy of the lustre of the Dharma, they come to Sukhavati. There they drink the Dharma-nectar of Amitabha-nath.

22. They attain the tripple Awakening and arrive at the realm of the Victors. Considering this constantly, those who want goodness and pleasure in life

23. Take refuge with the Lord of the Triple World and worship him. Saying this, the best of sages fell silent

24. And all the people in this assembly were delighted and enlightened.




Horst Brinkhaus. The Textual History of the Different Versions of the Svayambhu Purana. In Gerard Toffin, editor, Nepal Past and Present, page ? CNRS, Paris, 1993.


Ronald M. Davidson. The Litany of Names of Manjushri - Text and Translation of Manjushrinamasamgiti. MCB, 20:1-69, 1981.


W.B. Douglas. Washing your neighbour’s god. EBHR, in press.


David N. Gellner. Monk, Householder, and Tantric Priest: Newar Buddhism and its Hierarchy of Ritual. Number 84 in Cambridge Studies in Social and Cultural Anthropology. Cambridge University Press, Cambridge, 1992.


Karanhdavyuha. In P.L. Vaidya, editor, Mahayana-sutra-sangrahah, pages 205-308. Mithilavidyapithapradhana, Darbhanga, 1961.


Banarasi Lal, editor. AryaManjushrinamasamgiti with Amritakanika-tippani and Amritakanikodyota-nibandha. Number XXX in Bibliotheca Indo-Tibetica. CIHTS, Sarnath, 1994.


Todd T. Lewis. Mahayana Vratas in Newar Buddhism. JIABS, 12, 1:109-38.,1989.


Todd T. Lewis. A Chronology of Newar-Tibetan Relation in the Kathmandu Valley. In Siegfried Lienhard, editor, Change and Continuity: Studies in the Nepalese Culture of the Kathmandu Valley, number VII in Orientalia, pages 149-166. CESMEO, Turin, 1996.


Siegfried Lienhard. Nevarigitimanjari: Religious and Secular Poetry of the Nevars of the Kathmandu Valley. Alquist & Wilksell, Stockholm, 1974.


Siegfried Lienhard. Avalokiteshvara in the Wick of the Night-Lamp. IIJ, 36:93-104.,1993.


Erberto F. Lo Bue. The role of Newar scholars in transmitting the Indian Buddhist heritage to Tibet (c. 750 – c. 1200). In Samten Karmey and Philippe Sagant, editors, Les habitants du Toit du monde, pages 629-658. SociŽtŽ dÕethnologie, Nanterre, 1997.


John Locke. The uposhadha Vrata of Amoghapasha Lokeshvara in Nepal. L’Ethnographie, 83:159-89., 1987.


John K. Locke S.J. Karunamya: The Cult of Avalokitesvara-Matsyendranath in the Valley of Nepal. Sahayogi Prakashan, Kathmandu, 1980.


John K. Locke S.J. Buddhist Monasteries of Nepal. Sahayogi Prakashan, Kathmandu, 1985.


Robert Mayer. A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis. Kiscidale, Oxford, 1996.


R. O. Meisezahl. Amoghapasha. Some Nepalese Representations and their Vajrayanic Aspects. Monumenta Serica, 26:455-97, 1967.


Adelhard Mette. Remarks on the Tradition of the Karandavyuha. In K.N. Mishra, editor, Aspects of Buddhist Sanskrit: Proceedings of the International Symposium on the Language of Sanskrit Buddhist Texts, Oct 1-5, 1991, pages 510-519. CIHTS, Sarnath, 1993.


Hari Prasad Shasri. Svayambhu Purana. Asb, 1896-1900.


Svami Dvarikadas Shastri, editor. Bodhicaryavatara. Bauddha Bharati 21. Bauddha Bharati, 1988. Bodhicaryavatara; contains the Panjika of Prajnakaramati.


Guiseppe Tucci. La redazione poetica del Karandavyuha. Atti della Reale Accad. Delle Scienze di Turino, Lviii:605-30, 1923.


Alex Wayman. Chanting the names of Manjusri: the Manju ‘srinamasamgiti. Shambala, Berkley, 1985.




1  This material has helpfully been summarized in LoBue (1997).

2  These include the Svayambhu Purana in its various versions, the Bhadrakalpavadana (recently discussed in a 1998 Oxford dissertation by Joel Tatelman), and the Shringabheri Avadana.

3  For the purposes of this paper, I am constrained by the available editions of the Karandavyuha. Adelhard Mette is preparing an edition using the old Gilgit sources which will appear soon, I hope. However, KV (1961) in the Buddhist Sanskrit Text Series, for all its faults, is based in the Nepalese textual tradition and is thus useful here as it reflects the text as the author(s) of the GKV would have found it. See Mette (1993)

4  Lienhard (1993)

5  In an email from Harunaga Isaacson, he writes “ The term/epithet Shrighanah hardly occurs in the late Indian tantric texts.”

6  See song 7 in Lienhard (1974)

7  GKV manuscript sigils for the remainder of this paper: T1, Tokyo 33 (1705). N1 NGMPP G14/23 (1632). N2 NGMPP 49/4 (1771). C1 Bendall 1270 (1770?)/ L1 Cowell & Eggeling 19. J1 Jodhpur 1190 (1805?).

8  Shastri (1896-1900, p 44). See Brinkhaus (1993) for a discussion of the redaction of this text.

9  Davidson (1981, 54), Wayman (1985, p. 80), and with Ravishrijnana’s commentary Lal (1994, p. 45)

10  Locke (1987)

11  Douglas (in press)

12  The poshadhakaraniya or gso sbyong blang ba and the Aryamoghapashaposhadhavidhyamnaya or ‘phags pa don yod zhangs pa’i gso sbyong gi chog man dag, both by Shakyashri Bhadra.

13  The Ashtamivrata Vidhi as well as the somewhat indeterminate texts all called Ashtamivrata Katha. Whether the ritual section of these texts can successfully be correlated to the Tibetan translations of Sanskrit originals remains to be seen.

14  See, for example, the scroll discussed in Meisezahl (1967), which has scenes from the Virakusha Avadana around the central mandala.

15  1.175. 179, etc.

16  This particular version list can be found at I.9-11, but see XVII for a longer version.

17  KV (1961, 307-8)

18  Shastri (1988)

19  XVIII.56ab: T33 J1 N1

20  T33 51v-52r

21  Tucci (1923)

22  Tucci (1923, p. 615)

23  As part of editing the entire text I am preparing an edition of just these verses, which I hope will aid in their identification.

24  Mayer (1996, p. 9)

25  The Prajnaparamita texts gave rise to several different commentarial traditions, including the Abhisamayalankara corpus. The Lankavatara also attracted commentaries, although not to the same degree.

26  Mette (1993, pp. 514-5)

27 In Tibetan doxography, the root tantra of Avalokiteshvara, or rather of the Lotus family, is the Padmajalatantra or pad ma dra ba. The Karandvyuha is not mentioned in mKhas drub rje’s discussion of the categories of tantra, and Bu sTon filed it under mDo, the sutras; yet on purely formal grounds it is similar to other texts which are known to the tradition as tantras.

28  KV (1961, pp. 281-2).