Articles by alphabetic order
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0


Uihuu

From Tibetan Buddhist Encyclopedia
Jump to navigation Jump to search
that [[[Buddhajñānapāda]]] practiced [with her] for eighteen months.35 “I have not realized it”

means that the waves of [Buddhajñānapāda’s] realization had not poured forth. When the great guru, as well, said “I have also not realized it,” he was somewhat discouraged. Thinking, “Until I realize this, anything else is useless,” he placed his seal on a volume of the Samājatantra36 and, tying this around his neck, he set off to the north.37 Behind Vajrāsana is the forest called Kuvaca Which is full of tigers and bears—a terrifying place. There I spent six months, and thus realized the suchness of phenomena. I met an emanated monk together with two gurus. |10| [[[Buddhajñānapāda]]] went to a forest called Kuvaca, which is behind Vajrāsana. [His] intention was as follows: “I remain among sentient beings who turn their gaze away from the Essence of Enlightenment. There are many tigers and bears, and so forth, [which are the manifestations] of desire and the other [[[afflictions]]]; it is a truly terrifying place. Since I want to be free from that, I will remain [here] for six months invoking [the deity?]38 and practicing, by means of which I will realize the suchness of phenomena.” And how did he realize that? [This is explained] in the lines beginning with I met an emanated monk… This monk was an emanation of the Great Vajra Holder. His lower robe was open,39 he had made a turban out of his dharma robe and was plowing a field. And the two gurus were an ugly woman with a small child and a white female dog with markings [on her coat].40 When he met them, since he did not [yet] have waves Vimalamudrī? Or, taking Tāranātha’s reporting into account, Vimalāmodā or Vimalāmodinī? Thanks to Harunaga Isaacson for suggesting these possibilities on what this name may have originally been. 35 Lotsāwa’s periphrasis of Vaidyapāda supports my translation of this phrase (gnas der rnal ‘byor ma rnams dang lhan cig pa’i spyod pa yang zla ba bco brgyad kyi bar du mdzad do//) (Deb ther sngon po, 448; Roerich 1976, 368-9). 36 This is one of the more enigmatic passages in Vaidyapāda’s text, and here I have not followed Lotsāwa’s reading. Vaidyapāda’s commentary reads rang gi phyag rgya ‘dus pa’i glegs bam du byas nas. Lotsāwa (Deb ther sngon po, 448; Roerich 1976, 369) has understood this to mean “he transformed his consort into the form of a volume of text” (rang gi phyag rgya ma glegs bam gyi gzugs su bsgyur te/), and Tāranātha (Bka’ babs bdun, 105) follows suit: “He had there a consort named Mālamodi whom he transformed into a volume of the Samāja[-tantra] and affixed to his neck...” (der mā la mo di zhes bya ba’i phyag rgya zhig yod pa ‘dus pa’i glegs bam du bsgyur te mgul du btags nas/). (I believe that Templeman (1983, 72) has mistranslated this passage in Tāranātha.) Both readings of Vaidyapāda are grammatically possible, but I am somehow hesitant to translate following Gö’s and Tāranātha’s interpretation of the phrase, in part because a consort does not figure in any later part of the account. 37 ‘dus pa’i rgyud chen zhes pa la sogs pa ni bla ma’i dgongs pa ste/ de la ‘dus pa’i rgyud ni rnal ‘byor rnams so// de’i ‘grel pa ni rnal ‘byor bslabs pa sme (sme] D, dme P) sha can gyi bu mo’i bi ma la mu tri’o (tri’o] P, dri’o D)// de dang bcas par bco brgyad bar du mnyan pa ni zla ba bco brgyad kyi bar du bsgrub pa’o// bdag gis ma rtogs pa zhes pa ni rtogs pa’i dba’ rlabs ma ‘phros pa’o// bla ma chen pos kyang bdag gis kyang ma rtogs zhes gsungs pa dang/ thugs cung zad chad nas ‘di ma rtogs par gzhang ni don med do bsams nas/ rang gi phyag rgya ‘dus pa’i glegs bam (bam] D, baṁ P) du byas nas mgul du btags nas de las byang phyogs su bgrod de/ (Sukusuma, D 90a.7- 90b.2; P 108a.7-108b.2). 38 Lotsāwa definitely takes this to mean invoking the deity, and he specifies that it is done by means of a wrathful ritual (lha drag tu skul ba’i cho ga la brtson pas) (Deb ther sngon po, Vol. I, 448-9; Roerich 1976, 369 has omitted this detail in his translation.) 39 Here I have emended byi ba’i sham thabs can to bye ba’i sham thabs can following Lotsāwa and Tāranātha who both have this reading (Deb ther sngon po, Vol, I, 449; Bka’ babs bdun, 104). 40 This account is further embellished in Chögyal Phagpa’s 13th-century version of the encounter, which I have translated below, but already in Vaidyapāda’s telling, nearly everything that could be wrong with this “emanated monk” is already there: he is accompanied by a woman (monks are celibate!) who has a son (monks are celibate!!), is plowing a field (monks are prohibited from tilling the soil and other such farmwork!), and wearing his dharma robe on his head with his lower robe open (monks are to dress in a respectful and seemly fashion!). 9 of realization the guru [[[Buddhajñānapāda]]] felt no shame in front of them. Then, the monk knowing that [[[Buddhajñānapāda]]] was engaged in the supreme mantra conduct, in order to bring forth his vision [of true reality?],41 emanated the maṇḍala of Mañjuśrī.42 On the eighth day of the seventh month, during [the constellation] Puṣya At the time when Mṛgaśīrṣa and Hasta are fading,43 in the early morning, right at dawn, Towards the emanated maṇḍala-cakra of Mañjuśrī44 I made a fervent supplication to understand the meaning. |11| What was the date [when this happened]? On the eighth day during [the constellation] Puṣya, at the time when Mṛgaśīrṣa and Hasta are fading. What was the month? The seventh month. What was the time? In the early morning, right at dawn. [[[Buddhajñānapāda]]] was asked if he had faith in the emanated maṇḍala or the guru, and when he replied that he had faith in the maṇḍala the monk together with the gurus immediately [left and] entered a small house.45 Then 41 de la spyan ras kyis bca’ ba’i phyir. This line is also puzzling, and I am unsure of the translation. Lotsāwa has paraphrased his understanding quite straightforwardly, “in order to benefit him...” (de la phan pa’i phyir) (Deb ther sngon po, Vol I, 449). 42 rdo rje gdan gyi rgyab na ku ba tsa zhes bya ba’i tshal yod de der phyin pa’o// de yi dgongs pa ni byang chub kyi snying po las kha phyir bltas pa’i sems can rnams kyi nang na bdag gnas te/ de na ‘dod chags la sogs pa’i stag dang dred la sogs pa mang zhing shin tu ‘jigs pa’i sa ste (ste] D, te P) / bdag de las thar par ‘dod pa’i phyir der zla ba drug bskul ba dang bcas pa’i bsgrub pas gnas pas (pas] P, pa’i D) chos rnams kyi de bzhin nyid rtogs so zhes so// ji ltar rtogs she na/ sprul pa’i dge slong zhes pa la sogs pa’o// de yang rdo rjedzin pa chen pos sprul pa’i dge slong bye (bye] sugg. em. based on Deb ther sngon po and Bka’ babs bdun, byi D, P) ba’i sham thabs can chos gos las thod byas pa gcig zhing rmo zhing gnas pa dang/ bla ma gnyis te bu chung dang ldan pa’i bud med ngan pa (pa] D, ma P) dang khyi mo dkar ba mtshan ma can no// de rnams dang phrad pa las rtogs pa’i rlabs (rtogs pa’i rlabs] D, rtog rlabs P) mi mnga’ bas bla mas de rnams la ma khrel to// de nas dge slong gis sngags kyi spyod pa’i mchog la gnas par shes nas/ de la spyan ras kyis btsa’ ba’i phyir/ ‘jam dbyangs kyi dkyil ‘khor sprul lo// (Sukusuma, D 90b.2-5; P 108b.2-7). 43 Puṣya is the eighth lunar mansion in Indian astrology; Mṛgaśīrṣa is the fifth; Hasta is the thirteenth. 44 ‘jam dpal dbyangs kyi (kyi] S P N, kyis D C) dkyil ‘khor ‘khor lo (lo] S P N, lor D C) sprul pa la. 45 Lotsāwa reports the account nearly verbatim from Vaidyapāda, but Roerich has understood it differently and translated it as follows: “(His teacher) asked him: “Do you have faith in the teacher or the maṇḍala?” and he replied: “I have faith in the maṇḍala.” (The maṇḍala then vanished), and he found himself and the teacher staying inside a small house.” (Deb ther sngon po, 449; Blue Annals, 369). This appears to be a misreading of the text on Roerich’s part, as neither Vaidyapāda nor makes any indication that the maṇḍala vanished, nor indeed does report that Buddhajñānapāda entered the small house. The account, in Lotsāwa’s rendering simply states that “He replied that he had faith in the maṇḍala and then the monk together with the two gurus entered into a small house.” (dge slong bla ma gnyis dang bcas pa khang pa chung ngu zhig gi nang du zhugs par gyur to//) (Deb ther sngon po, 449). Because did not earlier follow Vaidyapāda in clarifying that the “two gurus” referred to the woman and the dog, Roerich presumably had not seen the phrase “two gurus” before, and apparently took it to mean Buddhajñānapāda and Mañjuśrīmitra. However, in Vaidyapāda’s account, which has in this section reproduced almost exactly, it was clear from the earlier reference that the two gurus are the woman and the dog and do not include Buddhajñānapāda. goes on to explain that after Buddhajñānapāda made his supplication, the lord of the maṇḍala—and here the term used, dkyil ‘khor gyi gtso bo, more likely refers to a deity rather than a guru—gave him instructions. There is no indication in Gö’s account that the maṇḍala was somehow re-emanated, because he never indicates that it disappeared. In Gö’s account, just as in Vaidyapāda’s, the monk and the woman and dog simply responded to Buddhajñānapāda’s preference for the maṇḍala rather than the guru by leaving and going inside a house, and Buddhajñānapāda then received his instructions directly from Mañjuśrī, the main deity of the maṇḍala. The disappearance and reappearance of the maṇḍala is clearly articulated in the account by Chögyal Phagpa, who reports two versions of the story, the first in which Buddhajñānapāda says he wishes to receive initiation from the maṇḍala and the monk says, “Fine, receive it from the maṇḍala!” and leaves, and another version in which the maṇḍala vanishes after Buddhajñānapāda says he wishes to receive initiation from the deity, upon which Buddhajñānapāda supplicates the monk who then re-emanates the maṇḍala from his heart center at dawn (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 614). I have translated this full episode from Chögyal 10 the great guru [[[Buddhajñānapāda]]] made the following supplication to the maṇḍala of Mañjuśrī in order to [be able to] receive suchness.46 47 Then, the great bodhisattva Mañjuśrī Looked upon me with a smiling face and said, “Excellent” three times. With this vajra song, like an echo, he taught to me The playful dance and the suchness of all phenomena. |19| Then, as an introduction to Mañjuśrī’s speech the great guru said Then... Then means immediately after the supplication. He is called Mañjuśrī (“the gentle voiced one”) [because] he satisfies beings with his gentle and sweet voice, since he is the pure form of the great wisdom of all the buddhas. He is called a bodhisattva because he is integrated with awakening (bodhi), not because awakening is his goal. For that very reason he is called great, and is distinguished from the [[[bodhi]]]sattvas on the ten bhūmis. He looked upon me with a smiling face means he was quite delighted because of having realization of the ultimate state. [The fact that] he said “Excellent” three times indicates that he was pleased by [Buddhajñānapāda’s display of] various modes of conduct that accord with having obtained suchness, by his supplications made with speech that accords with that meaning, and by his having observed everything to be profound and genuinely luminous. The rest was already explained. Like an echo has the sense of being like an echo, which makes a sound but is not truly established. A song that is like a vajra is a vajra song, which is a pleasing song. With the words he taught [this] to me, the great guru makes others feel confident.48 [What follows, amounting to ninety percent of the text of the Dvitīyakrama, are Mañjuśrī’s instructions to Buddhajñānapāda, recorded in Mañjuśrī’s first-person speech, and concluding with a prediction and command given by Mañjuśrī, in which he addresses Buddhajñānapāda directly in the second person. With the conclusion of these teachings, and the dissolution of his vision, Buddhajñānapāda returns to his autobiographical account.] Phagpa’s account below. Amye Zhab gives both versions of the story from Chögyal Phagpa’s account (Gshin rje chos ‘byung, 48a.1-4) and Dudjom reports only the version of the account where the maṇḍala disappears (Dudjom 1991, 494-96). 46 tshes gang zhe na/ mgo dang lag gnyis yol dang tshed brgyad rgyal la bab ces (ces] D, zhes P) so// nam zla gang zhe na/ ston zla ra ba zhes so// dus gang zhe na/ tho rangs skya rengs shar dus su zhes’o// der sprul pa’i dkyil ‘khor dang bla ma la mos pa dris pa dang/ sprul pa’i dkyil ‘khor la mos par bka’ tsal pa dang/ dge slong bla ma dang bcas pa de nyid du khang pa chung du cig gi nang du zhugs so// de nas bla ma chen pos ‘jam pa’i dbyangs kyi dkyil ‘khor la de bzhin nyid blang bar bya pa’i phyir gsol ba ‘di skad du btab bo// (Sukusuma, D 90b.5-7; P 108b.7- 109a.2). 47 I have omitted here the seven verses of Buddhajñānapāda’s supplication to Mañjuśrī, as these verses (v 12-18) do not contain autobiographical content. See Part II for the full translation of the root text. 48 da ni ‘jam pa’i dbyangs kyi gsung la ‘jug pa’i tshig bla ma chen po’i zhal snga nas gsungs pa/ de nas zhes pa la sogs pa’o// de nas zhes pa ni gsol ba btab pa’i de ma thag pa’o// ‘jam dbyangs zhes pa ni ‘jam zhing mnyen pa’i dbyangs kyis ‘gro ba rnams tshim par byed pa ste/ sangs rgyas thams cad kyi shes rab chen po rnam par dag pa’i phyir ro// de nyid byang chub dang ‘dres pa’i phyir byang chub sems dpa’ ste/ byang chub la dmigs pa ni ma yin no// de nyid kyis ni chen po zhes te sa bcu’i sems dparnams dgar ba’o// ‘dzum pa’i bzhin bltas zhes pa ni shin tu rangs pa ste/ mthar thug pa’i gnas rtogs pa’i phyir ro// legs zhes lan gsum gsungs zhes pa ni/ de bzhin nyid thob pa dang rjes su mthun pa’i spyod pa ji snyed pa dang/ don gyi rjes su ‘brang ba’i gsung ji snyed pas gsol ba ‘debs pa dang/ thams cad zab mo dang yang dag par gsal bar dmigs pa la thugs rangs pa’o// gzhan ni bshad zin to/ /sgra brnyan lta bur zhes pa ni brag ca lta bu ste grags kyang ma grub ces pa’i don to// rdo rje lta bu dang ldan pa’i glu ni rdo rje glu (ni rdo rje glu] D, P om.) ste dga’ bar byed pa’i glu’o// de lta bus bdag la bstan zhes bla ma chen pos gzhan yid brtan par mdzad pa yin no// (Sukusuma, D 93a.1-5; sP 111b.1-7). 11 In this way with the vajra song like an echo, together with the playful dance And the [[[maṇḍala]]-]cakra, right then49 he sang and praised me. Then, right there, he disappeared like a cloud into the sky And the monk and two gurus also likewise disappeared. |374| Then, in order to conclude Mañjuśrī’s speech, the master spoke about the dissolution of the maṇḍala with the verse beginning, In this way... The playful dance and the rest have already been explained. And the [[[maṇḍala]]-]cakra refers to Akṣobhya and the others. Right then means at that very time. As for, He sang and praised me [the words of that song of praise] should be known from the Treasury of Verses.50 Right there means in that very place. Into the sky means into suchness.51 Disappeared like a cloud into the sky is said in order to indicate that, just like clouds and moisture arise from the sky and dissolve back into it, likewise the Bhagavan, as well, through the yoga of great compassion, appears out of suchness and dissolves back into it. This being the case, his ‘causal emanations’ should be known to [do] the same. Having understood that, in order to tell the story of how he carried out the benefit of fortunate [[[disciples]]] he said I... and the rest.52 Realized a little bit is said in order to abandon [the act of] holding back the teachings out of avarice from those who are suitable recipients, [since Buddhajñānapāda had, in fact] exhausted [the obscurations to realizing] the ultimate suchness of all phenomena together with their latent traces, and had, by means of the stages of mudrā as explained above, gained realization.53 54 In a place fifty krośas behind Vajrāsana I lived in the Parvata cave. In order to benefit beings I compiled this [text, the Dvitīyakrama], composed and taught all of the treatises, and so forth. Since excellent beings made extensive supplications, I was delighted [to do so]. |375| 49 de nyid. I am following Vaidyapāda in interpreting this as referring to the immediate moment (Sukusuma, D 134b.6). 50 This text is mentioned by Vaidyapāda earlier in the commentary as a composition of Buddhajñānapāda’s. To the best of my knowledge, it is unfortunately not extant. I address Vaidyapāda’s list of Buddhajñānapāda’s writings below. 51 nam mkha’i khams su zhes te de bzhin nyid [+ nyid sugg. em.; P and D om.] du’o//. I have emended the text very slightly here, adding nyid, where it seems to have been left out. This is because without the emendation the content of the sentence does not make much sense; it would simply read Into the sky means the same.” Also this emendation brings the meaning of the sentence in accord with what follows. 52 Unusually, this short section of Vaidyapāda’s commentary appears to be commenting on a line or lines of the root text that are not extant in our version of the Dvitīyakrama. 53 I remain unsure about the meaning of this last sentence and suspect that the text may be corrupt. 54 da ni ‘jam pa’i dbyangs kyi gsung bsdu ba’i phyir bla mas sprul pa’i dkyil ‘khor bsdu ba gsung pa/ de ltar zhes pa la sogs pa’o// rol pa’i gar zhes pa la sogs pa ni bshad zin to// ‘khor lor bcas pas zhes pa ni rtag pa la sogs pa’o// de nyid ces pa ni dus der ro// glu dbyangs kyis bdag la bstod pa ni tshigs su bcad pa’i mdzod las shes par bya’o// der zhes pa ni gnas de nyid du’o// nam mkha’i khams su zhes te de bzhin nyid [+ nyid sugg. em.; P and D om; see note 51] du’o// sprin rnams med pa lta bur thim (thim] sugg. em. based on Dvitīyakrama; shes D, P) par ‘gyur/ zhes pa ni ji ltar sprin rlan (D, P add las; I suggest omitting) nam mkha’ las byung zhing der zhi ba bzhin du/ bcom ldan ‘das ‘di yang thugs rjes chen po’i sbyor bas de bzhin nyid las (las] sugg. em.; la D, P) snang zhing yang der zhi bar bstan pa’i phyir ro// de bas na de’i rgyu’i sprul pa yang de bzhin du shes par bya’o// de ltar rang gi de shes nas skal ldan gyi don ji ltar byas pa’i lo rgyus gsungs ba bdag gi zhes pa la sogs pa’o// cung zad rtogs pa zhes pa ni/ dngos po thams cad kyi mtha’i pha rol du son pa’i de bzhin nyid du ni bag la nyal du bcas pa zad pa ste/ de phyag rgya’i rim pas gong nas gsungs pa ltar rtogs nas dpe mkhyud snod rung ‘ga’ la yang spang pa’i phyir ro// (Sukusuma, D 134b.5-135a.2; P 162a.5-162b.3 ) 12 Behind Vajrāsana means to its northeast. A krośa is fifty fathoms. Fifty of those is six yojanas plus two krośas. The Parvata cave is [also] called Ma ta hra ni tra, the Dharma Sprout, and is a place where great lords of practice of former times stayed. I lived there means it was [his] residence. For what reason? In order to benefit beings, which means those who stayed nearby. Since there were many who were suitable recipients, the master mentions that it was for their benefit that he compiled this [text], meaning the [Dvitīyakramatattvabhāvanā]-Mukhāgama. All of the treatises refers to those that were mentioned above.55 Composed means produced. Taught means explained. The words and so forth include bestowing samayas and other activities. The cause for doing this was that excellent being made extensive supplications, just like those above. I was delighted [to do so] means that [it was done] with confidence. And [thus] in this way he engaged in the composition of those [texts].56 Living there together, my retinue and I [received] necessities, Clothing, food, a treasury of jewels, and various vast offering substances for gaṇacakra. [From] the tenth-ground bodhisattva, the treasure guardian,57 great Jambhala Each day we regularly received seven hundred kārṣāpaṇa. |376| Living there means there in that cave. Together [with] my retinue refers to the disciples who followed him. Among them there were eighteen who acted as his regents, and among those there were four who attained nirvāṇa in this very life: Dīpaṃkarabhadra, *Praśāntamitra,

stages [of practice] of the great master just as [he taught them].58 All of their necessities— clothing, food, a treasury of jewels like gold and so forth, a vast array of substances for making offerings to the Heart of Awakening,59 and the necessities for himself and his students to engage in gaṇacakra practice—were provided by the bodhisattva of the ten bhūmis, who is himself the lord of treasures and is therefore [called] the Treasure Guardian. He appears in the form of a yakṣa and is therefore called the yakṣa Jambhala. Each day he provided each of them with seven hundred karṣāpaṇa of cowries.60 55 This refers to a list of texts in the Dvitīyakrama that Mañjuśrī commanded Buddhajñānapāda to compose and the further elaboration of that list found in Vaidyapāda’s commentary. Many, but not all, of these texts can be identified and are extant. I discuss this list below. 56 rdo rje gdan gyi rgyab (rgyab] D, ‘gab’ P) ni byang shar gyi mtshams na’o// rgyang grags ni ‘dom lnga brgya’o// de lnga bcu ni dpag tshad drug dang rgyang grags gnyis so// parba (parba] D, spar ba P) ta’i phug ces pa ni ma ta hra ni tra (ma ta hra ni tra] D, ma ta hrin dra P) zhes te chos kyi myu gu zhes pa sngon gyi grub pa’i dbyang phyug chen po’i gnas so (so] D, P om.)/ de la brten te zhes pa ni gnas bcas pa’o// ci’i phyir sems can don bya’i phyir/ zhes pa ni de’i nye ‘khor rnams ni khyad par du snod du rung ba mang bas/ bla mas kyang de’i don du zhes so// ‘di bsdus zhes pa ni zhal gyi lung ngo// rab tu byed pa thams cad ces pa ni gong du smos pa rnams so// rtsom (rtsom] D, rtsam P) pa ni byed pa’o// ston pa ni bshad pa’o// sogs kyi sgras bsdus pa ni dam tshig sbyin pa la sogs pa’o// de’i rgyu yang dam pas gsol ba rgya chen po btab pas zhes gong ma ltar ro// bdag ni shin tu brod ces pa ni rang yid ches nas so// de rnams rtsom pa’i sbyor ba la zhugs pa’o// (Sukusuma, D 135a.2-5, P 162b.3-7). 57 srung] D C V (D), gsung S P N V (P). 58 de rnams bdag cag gi dang po’i gzhung ltar de bla ma chen po’i rim pa ji bzhin pa’i phyir ro//. I am unsure of the meaning of this line, which seems to be corrupt in some way. 59 Presumably here this term refers to Vajrāsana. 60 der gnas ni phug der ro// ‘khor bcas rnams zhes pa ni rang gi rjes su spyod pa’i slob ma rnams kyi nang na rgyal tshab kyis pa’i gang zag bco brgyad yod de de rnams kyi nang nas mthong ba’i chos la mya ngan las ‘da’ ba bzhi yod de/ mar me mdzas bzang po dang/ rab tu zhi ba’i bshes gnyen dang/ sgra gcan ‘dzin bzang po dang/ rdo rje bde ba chen po’o// de rnams bdag cag gi dang po’i gzhung ltar de bla ma chen po’i rim pa ji bzhin pa’i phyir ro// de thams cad kyi yo byad dang gos dang/ zas dang gser la sogs pa’i nor gyi mdzod dang/ byang chub kyi snying po la mchod pa’i yo byad rgya chen po dang/ rang dang slob ma’i tshogs kyi sna tshogs ‘khor lo’i bya ba rnams sbyor bar byed pa ni/ sa bcu’i byang chub sems dpa’ ste de nyid gter rnams kyi bdag po yin pas (yin pas] D, bas na P) 13 Then I traveled to meet the great guru Pālitapāda61 In order to please that guru, I compiled62 some short sādhanas And the guru and all the others there were pleased. I returned to the place I had come from and63 joyfully performed the benefit of some64 fortunate [[[Wikipedia:individuals|individuals]]]. |377| Then he tells the account of having been invited by his guru, who had come to know of his blessings with the verse beginning, Then... The statement I compiled some short sādhanas refers to those mentioned above. There means there in that place in the south [of India]. The guru was Pālitapāda. By the others, we should understand those who were gathered there, that is, those [[[dharma]]] relatives who were present. Were pleased means [[[pleased]]] by his dharma teachings and so forth. The place I had come from means the Parvata [[[cave]]]. I performed the benefit of some fortunate [[[Wikipedia:individuals|individuals]]] means those who hadn’t been included in his previous activity.65 Thus, in this way everyone, having come to know the detailed accounts [of my life], Should use all methods to please the sublime and sincere learned one, And listen to and contemplate his teachings, compositions, and so forth. |378| Thus, having generated faith in that way (i.e. by means of telling the story of his own encounter with suchness), he teaches about the training in nondual wisdom and its result with the verse beginning, Thus... Having come to know the detailed accounts means the detailed accounts about the great master: the taming of Nālandā, making offerings at Vajrāsana, the [account of] the consecration and the others.66 Through these accounts the faith of those who have fortune is mdzod srung ngo (srung ngo] D, gsungs so P) / gnod sbyin gyi cha lugschang bas na gnod sbyin gyi gnas so (gnod byin gyi gnas so] D, gnod gnas so P) // des nyin re ‘gron bu kā rṣā pa ṇa (kā rṣā pa ṇa] D, ka rṣa pa na P) bdun brgya re re la sbyor zhes so// (Sukusuma, D 135a.5-135b.1; P 162b.7-163a.4). 61 bā li pā da’i] D C , bha li pa trī S P N. Vaidyapāda’s commentary reads bsrung ba’i zhabs. I follow Szántó in giving his name as Pālitapāda, based on the presence of this guru’s name in an 11th-century Sanskrit manuscript of Samantabhadra’s Sāramañjarī (Szántó 2015, 542). Tāranātha has rendered it more or less correctly, as well, as Pā li ta pa da (Bka’ babs bdun, 104). In my edition of the Dvitīyakrama, however, I have left the rendering from the Derge and Cone Tengyurs—Bā li pā da—because to “correctly” phoneticize the teacher’s name would make the line unmetrical. 62 It is worth noting that Buddhajñānapāda uses the word “compile” (bsdus) rather than “compose” (rtsom). In an earlier verse, he also uses the term “compile” to describe the compilation of the Dvitīyakrama, but that is presumably because it is in fact Mañjuśrī’s teaching, which he is only compiling within the framework of his own narrative. In this case “compiling” rather than “composing” these sādhanas may hint at a process more revelatory than compositional, but more likely it is simply an acknowledgement that the sādhana was compiled, at least in part, from other sources, most prominently the Guhyasamāja-tantra itself. 63 nas] S P N, gnas D C. 64 ‘ga’] D C V (D and P), dga’ S P N 65 de nas byin rlabs shes pas bla mas spyan drangs pa’i lo rgyus gsungs pa/ de nas zhes pa la sogs pa’o// cung zad bsdus pa zhes pa ni gong du gsungs pa rnams so// de ru zhes pa ni lho phyogs kyi gnas der ro// bla ma ni bsrung ba’i zhabs so// sogs kyi sgras bsdus pa der rtogs so// (D + zhes) ‘tshal spun zlar gyur ba rnams so// mnyes par byas te zhes pa ni chos kyi gtam la sogs pa’o// sngon gnas zhes pa ni parba (parba] D, par pa P) ta’o// skal ldan don ‘ga’ byas zhes pa ni sngon las ma gtogs (ma gtogs] D, rtogs P) pa rnams kyi’o// (Sukusuma, D 135b.1-3; P 163a.4-6). 66 Vaidyapāda here refers to several accounts of Buddhajñānapāda’s life as if they are already well-known stories that will be understood by anyone reading his text. These same accounts are described in the later Tibetan histories in much more detail, though unfortunately only one such supportive detail is, to my knowledge, found in an extant Indian source, Atīśa’s *Bodhipathapradīpapañjikā, which I discuss below. Some of the Tibetan historians who 14 further increased. Then, the learned one who is learned in those scriptures that we uphold is the guru [[[Buddhajñānapāda]]] himself. Since he himself has overcome mental doubts he is sincere. [[[Beings]] should] please him, using all methods which were taught above. In order to familiarize themselves with these, they [should listen to and reflect upon] his teachings, which means his compositions, and so forth,—which [are called] compositions because they are very excellently composed—like the Samantabhadrī[-sādhana] and so forth. The and so forth includes the commentary on the tantra and other [texts]. Listen[ing] to these means also bring about attainment, since the stages of the grounds and paths come about through attainment. Contemplat[ing] them means repeatedly bringing about mental certainty through valid engagement [with them].67 Through relying upon that, remaining in isolated places and the rest, Training one’s mind in suchness, and genuinely realizing the way things are, [One can] attain awakening in this very life, or [even] in [just] six months, and so forth—who could refute this?! |379| Through relying upon that means relying upon those contemplations. In order to bring about suchness in a unique way one is meant to stay in isolated places, and so forth, as described above. Through training one’s mind in suchness means by means of the two stages, like the first [stage] and so forth. Through genuinely realizing the way things are means that through encountering signs of realization, realizing a little bit, genuinely realizing suchness, [and] by means of vratas and the like, one exhausts the remainders [of defilements] in this life, meaning during this very life. As for [even] in [just] six months, the text [also] states and so forth, which indicates an inferior [[[attainment]], i.e. longer time periods]. [Within the various time frames mentioned, one can] attain awakening, which is the realization of the ultimate state. Who could refute this achievement, enacted through such unique methods? Indeed, this being the way things are,68 [it] is difficult to refute, like a cascade of raindrops [falling] through the empty sky.69 provide the more detailed accounts of Buddhajñānapāda’s life, like Tāranātha, do list Indian sources that are no longer known to us. 67 da ni de lta bus dad pas byas te/ gnyis su med pa’i ye shes bsgom pa ‘bras bu dang bcas pa gsungs pa/ de bas zhes pa la sogs pa’o// gtam rgyud rgyas par shes byas nas/ zhes pa ni bla ma chen po’i gtam rgyud rgyas pa na landa ([landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus kyis skal ba dang ldan pa cher dad par byas nas/ des kyang rang gi ‘dod pa’i gzhung la mkhas pa ni bla ma ste (ste] P, sta D)/ de nyid kyis blo’i som nyi bzlog pas na gzu po’o// (P +de) dgong du gsungs pa’i thabs kun gyis mnyes par byas te (te] P, ta D) zhes so// de la (la] sugg. em., las D, P)/ goms pa’i phyir na lung ste/ rab tu byed pa la sogs pa’o// rab tu byed pa ni shin tu legs par byad pa’i phyir na ste/ kun du bzang mo la sogs pa’o// sogs kyi sgras bsdus pa ni rgyud kyi rnam par bshad pa la sogs pa’o// de rnams nyan pa ni thob byed dang bcas pa sa lam gyi rim pa thob pa las byung bas so// bsam par byas zhes pa ni ‘thad sgrub kyis yang dang yang du blo nges par bya’o// (Sukusuma, D 135b.3-6; P 163a.6-163b.3). 68 de’i chos nyid 69 de la rab brtan zhes pa ni bsam pa de la rab tu brten te de nyid khyad par can du bya ba’i phyir dgon sogs rab tu brten (brten] D, bsten P) byas zhes te gong ma ltar ro// rang gi sems de nyid bsgoms pas/ zhes pa ni/ rim pa gnyis kyis zhes pa dang po la sogs pa ltar ro// ji bzhin rab tu rtogs par (par P] pa D) byas pa yis/ zhes pa ni rtogs pa’i rtags rnyed pa dang cung zad rtogs pa dang/ de nyid yang dag bar rtogs pas brtul zhugs la sogs pas lhag ma zad pas tshe ‘di nyid la zhes te/ mthong ba’i chos nyid la’o// zla ba drug gis zhes pa ni sogs pa zhes pa tha ma’i tshig tu’o// byang chub thob pa ni mthar thug pa’i gnas rtogs pa ste/ thabs khyad par can gyis byed pa ‘di ni su yis bzlog ces te de’i chos nyid dgag dka’ ba ste/ bar snang la char gyi rgyun ltar ro// (Sukusuma, D 135b.6-136a.2; P 163b.4- 7). 15 2. Tantric Buddhism in Late 8th-century India Playfully dancing the great dance, with your various arms twisting and holding tight, you open the eight soft lotus petals and insert the vajra, the cause of nondual bliss. The secret suchness, undefiled, becomes clear. The moon which is born from the vajra and petals is perfectly gathered: this is the supreme suchness of all phenomena, born from means and wisdom. Revered master, in order to benefit me, explain what is hidden! -Buddhajñānapāda, Dvitīyakrama This extraordinary autobiographical narrative of the yogin and tantric exegete Buddhajñānapāda’s travels throughout the Indian subcontinent studying with different gurus, receiving teachings directly from Mañjuśrī in a visionary encounter, and setting up a hermitage with his disciples in the latter part of his life provide us a rich picture of Indian tantric Buddhist practice in the late 8th and early 9th centuries. His account is further enriched by the additional details provided in Vaidyapāda’s 9th-century70 commentary. This was a period of immense creativity and development within tantric Buddhist traditions, and many doctrinal and especially ritual developments from precisely this period continue to frame the structure of tantric Buddhist practice up to the present day. While the cadence and timbre of Buddhajñānapāda’s own voice clearly emerges from his surviving writings—and not only from the autobiographical narratives therein—we can better appreciate his individuality when it is approached from within the context of the world in which he lived and wrote, so it is to this that we will first turn. The Political, Social, and Religious Climate of Early Medieval India It is at our peril that scholars of Indian Buddhist traditions have often focused too narowly on textual sources and developments only within the Buddhist world, and neglected to consider the wider social, political, and religious climate of the Indian subcontinent in which Buddhist doctrinal and ritual developments emerged. When we do thus widen our perspective, as is fortunately increasingly the case in recent scholarship, we have access to a much more holistic, and therefore deeper as well as broader, view of the traditions we seek to understand. Having let Buddhajñānapāda himself, and his disciple Vaidyapāda, speak their stories first, I would like to begin my own account of Buddhjñānapāda’s life and writings by widening the lens to take in the broader world of the Indian subcontinent into which his voice emerged and was first heard. The early medieval period in India (roughly the mid-6th to the early13th century) was a time of upheaval, change, and immense creativity. Following the fall of the “golden age” Gupta empire in the 6th century of the common era there was a period of significant political restructuring. This was described by earlier scholars using a rhetoric of decline and decentralization, in which the process was termed “feudalization,”71 while more recent scholarship describes the same period using more positive language as a process of incorporation, in which newly founded state polities were both incorporating new territories and expanding into territories that had previously not been touched by a state polity.72 The general political climate of the period was neither a centralized state nor fragmented regional kingdoms, but rather, “a series of diverse and uneven political orders which, while regionally based, sought 70 Vaidyapāda was likely a direct disciple or at furthest a grand-disciple of Buddhajñānapāda, thus placing him squarely in the 9th century. I address the dates of both in more detail below. 71 e.g. Kosambi 1956. 72 e.g. Chattopadhyaya 1994. 16 to relate themselves, in diverse ways, to ever more integrated political hierarchies which had as their ideal the notion of an imperial polity ruled over by a single supreme overlord, a king over kings.”73 While the political rhetoric championed the idea of digvijaya, “conquest over the directions” as frequently mentioned in praśasti, the eulogistic poems dedicated to leaders that became an important literary mark of the political culture of the day, in actual point of fact such “conquest” often did not involve direct rule of the conquored lands by the overlord.74 More commonly, the conquered areas continued to be ruled by their own, now “lesser” lords, who submitted to the “greater” lord, thus creating a complex system of social and political relationships. An important aspect of this system was the gifting of land by new or established rulers, as a way of showing favor to their constituents, and sometimes also to encourage the expansion of agriculture into uncultivated areas. These land grants were often given to religious institutions— initially to monasteries, or individual brahmin families or communities, and increasingly, with the development of what is often called “temple Hinduism” to temples. In fact, religious grants seem, in most areas, to have been far more frequent than grants to non-religious beneficiaries.75 This political situation of the early medieval period resulted in a number of important developments in the religious sphere. The first of these is a direct result of the expansion of state polities into new areas. This expansion involved the movement of peoples, which resulted in the meeting of more established and pan-regional religious traditions, such as the Vaiṣṇava, Śaiva, Buddhist, and Jain traditions, with more local forms of religion.76 This process sometimes occasioned the adoption of local deities into translocal traditions.77 In addition, this political 73 Ali 2004, 33. 74 Pollock emphasizes the fact that the political styles and the tradition of composing praśasti was common to rulers from diverse religious traditions—Buddhist, Vaiṣṇava, Śaiva, and Jain. He notes that, in fact, they “all wrote more or less similar poetry and engaged in identical political practices” (Pollock 2009, 572). 75 Thapar 2002, 451. The most common explanation of this land-granting practice, given in much of the historiography, is that this was a practice done for purposes of “legitimation.” The new rulers, because many of them were not from traditional ruling (kṣatriya) families, needed to justify their rule, and thus gave grants to brahmins who then wrote important “fictitious” genealogies (vaṃśa) legitimizing the rule of these families. While these developments which, described as “purāṇic,” may seem to apply only to religions that developed out of the brahmanical tradition, in fact the Buddhists and Jains both integrated themselves into these structures, as well, claiming descent from the so-called “Solar Dynasty” that is one of the two important lineages in the purāṇic genealogies (Samuel 2008, 68). What is more, the 8th-century Buddhist tantra, the Mañjuśrīmūlakalpa, also contains a predictive royal genealogy, very much in the purāṇic model (Sanderson 2009, 94). Sheldon Pollock’s critique of the rhetoric of legitimation found in earlier historiography—essentially arguing that such legitimation does not make sense in the pre-modern period when rulers could (and did) simply force their rule upon people and did not need to rely on documents like genealogies to convince the populous that their rule was legitimate—is well taken (Polluck 2009, 521). However, Pollock himself admits that if legitimation does anything at all it builds ruling class consensus, rather than that of the larger populous (Pollock 2009, 523). This observation, in fact, makes perfect sense of the popular practice of granting land and receiving genealogical confirmation of one’s right to rule. These genealogies, which begin to appear from the 6th century, just as the Gupta empire was falling, insist on birth into certain types of lineages as a requirement for being part of the ruling class. Thus there was indeed a need for legitimizing oneself as belonging to a certain type of family in order to engage in the elite political culture of the time. Because of the structure of the political order, with its enmeshed polities and the important and intricate relationships that involved the exchange of gifts and women (which Daud Ali has carefully described in his 2004 work), it was impossible for a ruler to exist as a completely independent polity—one could not rule in a social and political vacuum. Political relationships were crucial to the maintenance of power, and to engage in these relationships, it was necessary to hold claim to a certain type of birth. Thus while the rhetoric of legitimation does not, as Pollock suggested, make sense as legitimation on behalf of the larger populous, it does make perfect sense when understood as a requirement for participation in the elite political culture of the time. 76 Thapar 2002, 389. 77 See e.g. Granoff 2004. 17 reorganization also involved significant warfare, which displaced populations and may have had some influence on de-urbanization in certain areas. When people and communities are on the move, they meet with other groups, leading to the intermingling of beliefs and practices, and creating a perfect environment for religious and cultural creativity. We can see this in the mutual influence between Buddhist and Śaiva tantric traditions78 where we find extensive Buddhist textual borrowing from Śaiva tantras,79 as well as examples of traditional Buddhist iconography passing into the Śaiva tradition.80 Moreover, there is also documented evidence of Śaiva borrowing from earlier Buddhist tantras,81 and the Buddhist tantric use of transgression specifically as a method to cultivate nondual gnosis was later adopted by Śaiva authors.82 Certain techniques, such as the practice of utkrānti, in which the consciousness of a yogin is ejected from his body, and sometimes transferred into the body of another individual (or more frequently a corpse), are known in Śaiva, Buddhist, and Jain texts, indicating a culture in which yogic techniques were shared.83 Many pilgrimage sites were also shared commonly among multiple traditions. Another important way in which the political environment affected the religious trends in early medieval India involves what has been described as the parallel developments of the “apotheosis of the king” and the “feudalization of the gods.”84 The early medieval period is thus characterized by the emergence of the idea of divine kingship. This process also involved the shift in ritual practices connected with kingship in which earlier Vedic rituals of royal consecration were replaced by purāṇic (Vaiṣṇava, Śaiva), and Buddhist versions.85 The purāṇic legends of the time describe the gods in ways that reflect the political culture of the time—they marry, live in fortresses, and so forth, like kings and queens.86 Unsurprisingly, rulers were enthusiastic patrons of these religious developments. The specific medieval Buddhist response to the political developments of the time has been studied by Ronald Davidson in his important work Indian Esoteric Buddhism, in which he asserts that the metaphor of kingship is the defining metaphor of Buddhist tantric systems. Davidson’s analysis of the role of the political environment in informing the ritual world of the tantric maṇḍala remains an astute and important observation. In fact, Daud Ali’s work on early medieval court culture provides a number of examples, easily visible to the scholar of Buddhist tantra, which further corroborate Davidson’s thesis.87 Indeed it does appear that one factor in the development of these particular forms of Buddhist practice was the current political climate; they appear to constitute a method for securing a place, as well as patronage, for Buddhism in the 78 The relationships between these two traditions has been studied by Sanderson and Davidson, who take different, if not exactly opposing, perspectives (See Sanderson 2001 and 2009 and Davidson 2002). 79 See Sanderson 2001 and 2009 and Hatley 2016. 80 Davidson 2002, 86. 81 Hatley 2016. 82 Wedemeyer 2013, 166-67. 83 See Smith (2006, 289) for an excellent description of this process in a Jain text, and for a description of the practice as allegedly performed by the 8th-century yogin Śaṅkara (Smith 2006, 294). The earliest description of the process of the ejection of consciousness that I am familiar with in a Buddhist text is in Buddhajñānapāda’s Dvitīyakrama, which has strong echos in the Catuṣpīṭha-tantra (On which see Szántó 2012a, esp. pp. 455-62). I address the topic of utkrānti in Buddhajñanapāda’s writings in Chapter Six. 84 Davidson 2002, 71. 85 Davidson 2002, 127-31. 86 ibid., 71. 87 For example, the king possesses a “seven-walled palace” (Ali 2004, 42); he has messengers (Skt. dūtaka, =Tib. pho nya) who carry out his business and doorkeepers (dvārapāla) guarding each door of the palace (ibid., 45); he bestows favor on a supplicant if he is pleased or satisfied (ibid., 106); sits on a lion throne and is fanned by whisk bearers (ibid., 112). 18 changing political and social environment. However, it is also important to take seriously the soteriological aims that these writings themselves explicitly claim to pursue. Another arena in which the social and political climate influenced religious developments, and one that is just beginning to be studied, is the realm of courtly culture. As Ali has shown with his groundbreaking work, the early medieval court “formed a key context for the production of knowledges that have more commonly been attributed to a generalised ‘society’ in ancient India.”88 Many of these developments have been an important influence in the religious sphere, as well, and Ali has documented significant contact between the courtly and religious worlds. It seems that a substantial number of men of the court came from monasteries, hermitages or brahmin households that were supported by the king and many prominent courtly gentlemen became hermits or monks when their masters died, or when they themselves entered old age.89 This certainly indicates a climate in which ideas could move freely between those two domains. Indeed Vatsyāyana, in his well-known Kāmasūtra, which typifies the pursuit of erotic and aesthetic pleasure central to the courtly life of the day, suggests that a young woman should learn erotic skills discreetly from an older sister, or from a nun (presumably one who had an earlier adult life as a laywoman!).90 Ali also shows that aspects of court protocol “intersected with codes of conduct from a wider domain, particularly a religious one.”91 These religious developments in turn influenced the political culture, because, in large part, religious masters and institutions were successful in their aims to secure patronage. That is to say, the kings of the time supported these traditions, and incorporated their rituals into rituals of state, effectively replacing the earlier Vedic model.92 Kings spent tremendous amounts of wealth supporting religious institutions and religious specialists. Royal preceptors often became wealthy in their own right, allowing them to support the development of their own traditions, to build temples, or support monasteries. Indeed the enormous Buddhist monasteries of the early medieval period such as Nālandā and Vikramaśīla, and the breathtaking Śaiva, Vaiṣṇava, and Śākta temples constructed during this period—the incredible outpouring of religious art, architecture, and literature—was primarily possible due to royal patronage. We will see resonances of many of these broader developments as we look more closely at Buddhjñānapāda’s life, writings, and thought: clear evidence of his movement throughout the subcontinent; engagement with the large state-supported Buddhist institutions of his time and 88 Ali 2004, 25. 89 ibid., 49. 90 ibid., 218. 91 ibid., 103. The influence of courtly culture (rather than politics, which Davidson has examined) on specifically Buddhist literature and ideas is a tantalizing but little explored area. We can easily see the influence of the sumptuary culture of the court on the Mahāyāna sūtras, with their imagery of worlds of jewels and gems, and the posture of royal ease adopted by the bodhisattva imagery of the period (Ali 2002, 159). This culture continues to appear in later śāstras, as in the elaborate bathing and dressing pavilions described by the 7th-century Buddhist author Śāntideva in his famed Bodhicāryāvatāra. The literary theory of rasa, developed in the nātyaśāstra literature and very important to the courtly aesthetic, was brought into the Buddhist tantras in the 8th-century Sarvabuddhasamāyoga-tantra (Smith 2006, 333). Moreover, with regards to erotology—another important cultural development that was made and refined in the courtly context— as we will see, Buddhajñānapāda, in his Dvitīyakrama, uses the classic four-fold female erotic typology from kāmaśāstra in his description of tantric Buddhist consorts (Dvitīyakrama, verses 50-67), in what is a very early instance of this classification system in Indian literature, preceding its appearance in extant kāmaśāstra literature by several hundred years (I discuss this further in Chapter Six). The same four-fold typology is also found in the later Saṃvarodaya-tantra (See Tsuda 1994, 155). Ali (2011, esp. pp 54-55) has explored some of the ways in which Buddhist tantric literature and practice seems to have influenced kāmaśāstra. 92 This is not to say that the earlier Vedic rituals were completely left behind. Indeed, many aspects of these much older rituals were incorporated into newer purāṇic, Buddhist, and Jain rituals. See, for example, Marko Geslani’s work on the incorporation of śānti rites into post-Vedic ritual contexts (Geslani 2011 and 2012). 19 with the political elite, as well as with other types of patrons and systems of patronage; evidence of his participation in a doctrinally and ritually eclectic millieu; and the incorporation of aspects of courtly culture into religious doctrine and ritual. Let us now narrow our focus one notch to survey the specifically Buddhist doctrinal and ritual context in which Buddhajñānapāda lived and wrote. 8th-Century Indian Buddhism While his writings indicate, through their many references to and much terminology from non-Buddhist traditions, that Buddhajñānapāda was living in a religiously eclectic environment, they leave no doubt about his self-identification as a Buddhist practitioner. Within the Buddhist tradition, he also clearly identifies as a practitioner of Buddhist tantra, and indeed appears to have been participating in, and likely even contributing to, the cutting edge of Buddhist tantric ritual technologies of his time. His life and work, therefore, must also be understood within the framework of the Buddhist tradition as it existed in India in the 8th century. The rise of Mahāyāna Buddhism around the turn of the common era brought with it a revelatory and visionary turn in the Buddhist tradition,93 expanding the scope of Buddhist worlds, and bringing newly expressed philosophical orientations to the fore. The Prajñāpāramitā-sūtras and Nāgārjuna’s Madhyamaka philosophy articulated a vision of reality whose emptiness of inherent nature allowed for infinite possibility, while the slightly later sūtras on buddha nature and those incorporatingYogācāra thought, along with their accompanying commentarial literature, attended more closely to the identity and nature of the mind that had access to such a reality. It was within the context of these doctrinal systems that Buddhist tantra began gradually to develop in the 7th century of the common era. However, while certainly it was in terms of such Buddhist doctrinal systems that the tantras were interpreted, the texts themselves appear to have emerged out of a more practical, that is to say a ritual, context.94 By the 8th century when Buddhajñānapāda lived and wrote, Buddhist tantric traditions had developed to the point that contemporary authors were beginning to classify the tantras into different categories.95 Buddhaguhya, a contemporary of Buddhajñānapāda’s, divides the tantras into Kriyā tantras, which involve more outward practice, and Yoga tantras, which involve more inner yogic practices.96 He also mentions what has sometimes been interpreted as a separate category, Ubhaya, or “both” (also sometimes called “Upa” or “Caryā”), which involve a mixture of external and internal practice.97 The Yoga tantras, the most prominent early example of which is the 7th-century Sarvatathāgatatattvasaṃgraha-tantra, showed a significant turn towards concern with 93 See, e.g. Harrison 1978 and 2003. 94 See J. Dalton 2016; Dharmachakra Translation Committee http://read.84000.co/translation/toh498.html, i.21; and Shinohara 2014. 95 See J. Dalton 2005, 121-31 for a summary of 8th-century Indian authors Buddhaguhya’s and Vilāsavajra’s tantric doxographies. These systems were almost certainly known to Buddhajñānapāda as Vilāsavajra is named in the Dvitīyakrama as Buddhajñānapāda’s own guru, and Buddhaguhya is sometimes mentioned, at least in the later Tibetan histories, as his disciple, though modern scholars have questioned this claim. Buddhajñānapāda’s own Dvitīyakrama contains a doxography, though not a tantric one (Dvitīyakrama, 126-43). He simply places tantra as a whole above all non-Buddhist and Buddhist philosophical systems, which is rather unique given that, as Dalton points out in the article just referenced, the Indian systems of classifying different systems of tantra rely primarily on ritual, rather than doctrinal, distinctions (J. Dalton 2005, 119-20). See also note 100 below for the wide variety of doxographical categories found in Vaidyapāda’s 9th-century Sukusuma. 96 J. Dalton 2005, 123-4. 97 ibid., 123-5. 20 soteriological goals,98 set forth the “mature” five-family maṇḍala system, and demonstrated full self-awareness of being a unique system of Buddhist practice.99 However, in the 8th century Buddhist tantric systems began to undergo a futher shift, with the development of what came to be called the Mahāyoga tantras. These texts, the most prominent of which is the Guhyasamājatantra, bring antinomian elements of sex and violence that were marginal in the Yoga tantras into the fore.100 The Mahāyoga tantras are furthermore characterized by the quite literal shift in the five-family maṇḍala arrangement from the centrality of Vairocana, the main buddha of the socalled buddha family, to that of Akṣobhya, of the vajra family. The 8th century also saw the emergence of the important Sarvabuddhasamāyoga-tantra, sometimes classified as a Yoga or Mahāyoga tantra, and other times as pertaining to the other newly emerging category of the Yoganiruttara or Yoginī tantras, which came into their full flourit in the 9th and 10th centuries with the Cakrasaṃvara and Hevajra-tantras, among others. The 8th and early 9th centuries were also a particularly important period in terms of the unfolding of the ritual structures and frameworks within which tantric Buddhist practice took place. That is, the system of tantric initiations was developing precisely in this period from the five-fold series of initiations that characterized the earlier Yoga tantras,101 to the addition of the later sexual initiations: the guhyābhiṣeka, then the prajñājñānābhiṣeka, and finally the so-called “fourth” initiation.102 The addition of these higher initiations corresponded with, and was likely necessitated by, the appearance of new modes of practice. The 8th century saw the use and development within the Buddhist tradition of sexual yogas, and the further development of practices that involved the manipulation of subtle energies within the body, which were often performed within the context of sexual practice. The addition of these new techniques into the tantric practitioner’s repertoire resulted in the division of tantric practice into the now ubiquitous categories of the generation (utpattikrama, bskyed rim) and perfection stages (niṣpannakrama, utpannakrama, rdzogs rim).103 Though of course these newly developed categories were in flux, the generation stage can be loosely characterized by the visualization of oneself as a deity and the worship thereof, and the perfection stage by the manipulation of internal energies while in 98 Tribe 2000, 209. 99 See Weinberger 2003 (esp. pp. 185-89) which draws attention to the rewriting of the Buddha Śākyamuni’s awakening story as a narrative of tantric practice in the Sarvatathāgatatattva-saṃgraha. Weinberger argues that this represents tantra’s “coming out party” or its “declaration of independence” as something distinct from earlier Buddhist traditions (Weinberger 2003, 189). 100 While the term Mahāyoga tantra was certainly used to describe the Guhyasamāja-tantra and other tantras pertaining to this class, they also continued to be referred to by some authors as Yoga tantras. The tantric doxographical categories of the time were indeed so variable that even within a single commentary by a single author, such as Vaidyapāda’s 9th-century commentary on the Dvitīyakrama, tantras are distinguished in multiple ways including those tantras “that emphasize wisdom” and those “that emphasize method” (Sukusuma, D 90a.2-4); Yoga and Mahāyoga tantras (Sukusuma, D 107a.6-7); Yoga and Yoganiruttara a.k.a. Dākiṇī tantras (Sukusuma, 108a.6-108b.1); and Krīya, Caryā, and Yoga tantras (Sukusuma, D 112a.4-5). 101 These are the water, crown, vajra, bell, and name initiations. See Isaacson 2010b, 264. When the later initiations were added, these five were re-classified as the “first” or vase initiation (kalaśābhiṣeka). I discuss initiation in Buddhajñānapāda’s writings in Chapter Seven 102 Isaacson (2010b) gives a summary of the gradual development of these initiatory systems. I address this topic in more detail in the context of Buddhajñanapāda’s writings on initiation, along with those of several of his direct disciples, in Chapter Seven 103 The well-known scriptural locus classicus of the two stages of tantric practice is the Samājottara, though as I have pointed out in an earlier conference paper and will discuss further in Chapter Eight, this important verse in fact seems to have originated in Buddhajñānapāda’s writings (C. Dalton 2013). The terms seem to have been in somewhat general use, though, by the middle of the 8th century, as they are found in other texts from the time, such as Padmasambhava’s Man ngag lta ba’i phreng ba (one of the few texts that scholars accept as having been composed by the historical Padmasambhava). 21 that form, often performed while in sexual union with a consort. As we will see, Buddhajñānapāda’s writings are important for refining our understanding of both of the development of initiatory rituals and the two stages of tantric practice. While these emerging systems certainly became popular modes of Buddhist practice that survive to this day as a central component of particularly the Tibetan and Newar Buddhist traditions, it is also clear that Buddhist tantra, like the Mahāyana before it, was not universally accepted among Buddhist communities. The tantras themselves and their exegetical commentaries, including Buddhajñānapāda’s own writings, include a number of features and strategies that appear to be aimed at legitimizing these newly emerging and unsurprisingly controversial practice systems. There are, moreover, records of discord between those who accepted the new traditions and those who did not, including reports of Buddhist monks publicly burning tantric Buddhist scriptures and destroying tantric Buddhist images at Vajrāsana in the late 8th/ early 9th century.104 From his writings we can see that Buddhajñānapāda, like his contemporary tantric exegetes, was well versed in a great deal of the Buddhist literature, doctrine, and practice systems that preceded him. This includes both exoteric Mahāyana sūtras and philosophical systems as well as earlier tantric traditions like those of the Mahāvairocana-tantra, the Sarvatathāgatatattvasaṃgraha-tantra and the Sarvabuddhasamāyoga-tantra, citations from the latter two of which appear in his writings. But it is the Guhyasamāja-tantra that is most central to his oeuvre, and especially to the ritual systems he set forth. Indeed, Buddhajñānapāda became known as the founder of the eponymous Jñānapāda School of Guhyasamāja practice, one of several lineages of practice associated with this tantra that first flourished in India and were later brought to Tibet. Thus, in order to further contextualize Buddhajñānapāda’s life and writings, we must again narrow our lens even more, to look at the emergence of and practice systems associated with the Guhyasamāja-tantra. The Guhyasamāja-tantra One indication of the Guhyasamāja-tantra’s importance is its consistent inclusion in every known version of an otherwise variable list of eighteen quasi-canonical tantric compositions, that circulated from India into both China and Tibet during the 8th century.105 The so-called Vajroṣṇīṣa (erroneously rendered as Vajraśekhara in earlier scholarship)106 scriptures were transmitted to China by Amoghavajra and Vajrabodhi, and are described in a short summary text by Amoghavajra as eighteen “assemblies,” of which the fifteenth has been identified as the Guhyasamāja.107 The idea of such a group of eighteen tantras, or tantric cycles, was also passed on to Tibet, where they were known there as the Māyājāla tantras and ascribed to the class of Mahāyoga tantra. However, Orna Almogi has argued, on the basis of the great variety in terms of both content and organization of thee lists preserved in Tibetan literature, both historical and doctrinal, that an actual list of the specific tantras that the group of eighteen 104 Chimpa and Chattopadhyaya 2010, 279; Maclean 1989, 12; and Flood 2009, 34. Incidentally, Flood interprets this as an act of hylotheism, which it clearly was not. The monks would not have been concerned that the tantric Buddhists were “confusing a transcendental god with matter” as Flood suggests (they certainly would have made and revered images of the Buddha, too), but that they were worshipping deities and engaging in modes of practice that the monks deemed non-Buddhist. Indeed, Tāranātha reports the monks to have said that the texts were “composed by Māra” (Chimpa and Chattopadhyaya 2010, 279). 105 It is, in fact, one of only three texts that appear in all known lists of this group; the other two are the Śrīparamādya-tantra and the Sarvabuddhasamāyoga-tantra. 106 On this issue see Geibel 1995, 109 and Davidson 2011, 24. 107 See Eastman 1981 and Geibel 1995. 22 contained was likely not.108 Nonetheless, all of the different Tibetan lists studied by Almogi contain the Guhyasamāja-tantra.109 An Indian version of this list is found in a commentary on the Āryaprajñāpāramitānayaśatapañcāśataka (Tōh. 2647) by the late 8th-century Indian scholar Jñānamitra who mentions “eighteen sections” headed by the Sarvabuddhasamayoga-tantra, but also including the Guhyasamāja.110 The Guhyasamāja-tantra survives in a number of Sanskrit manuscripts, at least three Tibetan translations, and one Chinese translation, again attesting to its importance.111 Yukei Matsunaga’s studies of the historical development of the tantra (1964, 1977a, 1977b, 1978) still serve as the primary basis for research in the field. Like most Buddhist tantras, the Guhyasamāja is an accretive text: the first twelve chapters of the tantra comprise an earlier level of textual composition, chapters thirteen through seventeen constitute an additional level,112 and the socalled eighteenth chapter, the Samājottara, first circulated in India as an independent text before being appended to the root tantra.113 Based on the presence of the summary of the “Guhyasamāja-yoga” among the eighteen “assemblies” noted in the account translated into Chinese (or perhaps composed)114 by Amoghavajra, who travelled in India between 744 and 746, it is clear that some form of the Guhyasamāja-tantra was in circulation in the first half of the 8th century.115 However, the Guhyasamāja as described in Amoghavajra’s “eighteen assemblies” only covers the basic maṇḍala structure and neglects the more antinomian elements of the 108 Almogi 2014, 51. 109Almogi 2014. On versions of the groups of eighteen tantras see also Eastman 1981 and J. Dalton 2005, 126 n32. 110 Geibel 1994, 114. 111 The extant Tibetan translations are the canonical translation in the Kangyur (including the so-called eighteenth chapter of the tantra, the Samājottara); the translation preserved in the Collected Nyingma Tantras (Rnying ma rgyud ‘bum v 12, 89a-157a), which also includes the Samājottara; and a manuscript of the translation of the root tantra alone, without the Samājottara, from Dunhuang. Kenneth Eastman (1979) has studied these different recensions of the Tibetan translations, concluding that the Dunhuang translation is the basis for the other two. The Sanskrit of the tantra was first edited by Bhattacharya (1931) on the basis of four Sanskrit manuscripts, and subsequently by Bagchi (1965). The first Western-language translation was made by Fremantle, whose doctoral dissertation (1971) included a Sanskrit edition and a romanized transcription of the Tibetan translation of the root tantra from the Peking Kangyur, as well as an English translation of the root tantra. Her work does not address the Samājottara. More recently Matsunaga (1978) has made a more comprehensive Sanskrit edition, including the Samājottara. His edition takes into account not only the Sanskrit and Tibetan, but also the Chinese translation of the tantra. 112 Indeed, even a quick glance at the composition of the Guhyasamāja-tantra bears witness to the fact that the chapters, starting from chapter thirteen onwards suddenly become much longer than the first twelve. 113 Matsunaga 1977b. In fact, the Samājottara is still preserved as an independent text in the Derge edition of the Tibetan Kangyur, where it is entitled the Rgyud phyi ma (Tōh. 443). According to Matsunaga the Samājottara also underwent stages of development. He notes that, “...the Uttaratantra [=Samājottara] text which is quoted in Viśvāmitra’s commentary and which remains as an old Tibetan translation differs with the present text. Accordingly, it is likely that a small process of development occurred before the present form of the Uttaratantra was completed” (Matsunaga 1977, 117). Matsunaga’s observations with respect to the Samājottara as preserved in the Nyingma Canon may be accurate. However, with respect to Viśvamitra’s commentary, Dpal gsang ba 'dus pa'i rgyud kyi man ngag gi rgya mtsho thigs pa (Tōh. 1844), my reading of this text has led me to the conclusion (which I hold with considerable certainty) that it is, in fact, not a translation from the Sanskrit at all, but rather an indigenous Tibetan composition. In addition to lacking both a Sanskrit title at the beginning and a translator’s colophon at the end (which would not in and of itself preclude its being an Indic text), the commentary, which deals only with the Samājottara and not with the root tantra, is nearly twice the length of most Indic commentaries on the tantra and shows a number of linguistic features that I believe could only have arisen in an indigenous Tibetan composition commenting on a Tibetan translation of the Samājottara, rather than on the Sanskrit text. 114 Even traditional Japanese Shingon sources consider this text to be a composition by Amoghavajra rather than a translation (Geibel 1995, 108). Nonetheless, he is understood to be summarizing Indian sources with which he was familiar. 115 Matsunaga 1977b, 112. See also Geibel 1995. 23 present day form of the tantra, sharing more features with earlier Yoga tantras such as the Sarvatathāgatatattva-saṃgraha.116 The text transmitted to Tibet in the period of the early translations (prior to the collapse of the Tibetan empire in 843), however, was the full root tantra.117 It is thus likely that while the Guhyasamāja-tantra originated in the early part of the 8th century, the text as we know it today was developed in the later part of that century. Buddhajñānapāda’s life falls directly towards the end of the period in which we surmise that the root tantra took its final form and given that the Samājottara first circulated somewhat later than the root tantra, his relationship to both of these texts is an important question that I will address below. The format of the tantras, with their compilatory nature, diverse topics, and swift jumps from narrative, to doctrinal content, to ritual, and back makes it difficult to summarize their contents.118 I will, however, just briefly outline some of the contents of the Guhyasamāja-tantra to give a sense of the scripture that Buddhajñānapāda references in many of his works. The tantra begins rather dramatically with the statement that the Bhagavan was abiding in the bhaga (vagina) of the Vajra Consort, which was so novel that just the tantra’s opening section was the subject of an entire commentary in its own right.119 The first chapter continues with the emergence of the Guhyasamāja maṇḍala which is generated by the speaker of the tantra, alternately called the Bhagavan, Mahāvairocana, and Bodhicittavajra. The deities produced are the five buddhas, beginning with Akṣobhya, who takes the central place,120 the four buddha consorts, and the four wrathful gatekeepers, each of whom is generated by means of his or her own mantra.121 The tantra alternately discusses doctrinal questions, like the nature of awakening (usually in the narrative frame of a discussion between the main promulgator of the tantra and his 116 Matsunaga 1977b, 112. 117 Also of note for the late 8th-century dating of the root tantra is the fact that the translation of the Guhyasamājatantra (but not the Samājottara) preserved in the Rnying ma rgyud ‘bum was, according to its colophon, translated by the early-period translators Vimalamitra and Kawa Paltsek (Ka ba dpal brtsegs), though Eastman notes that it was extensively altered after the thirteenth century (Eastman 1979, 3). The Rnying ma rgyud ‘bum translation of the Samājottara states, confusingly, that it was translated by the translators Buddhaguhya, who lived in the 8th century, and Drogmi Palgyi Yeshe (‘Brog mi dpal gyi ye shes) who lived in the 11th—while the translation of the Samājottara preserved in the Derge Kangyur states that it was translated by Śraddhākaravarman and Rinchen Zangpo (Rin chen bzang po) (958-1055), of the later translation period. It thus seems likely that the Samājottara was not translated until the later period. The Blue Annals also confirms the early-period translation of the Guhyasamāja-tantra, but attributes the translation to the translator Che Tashi (Lce bkra shis), and makes no mention of the translation of the Samājottara (Roerich 1976, 359). Additionally, Campbell notes that a translation of Vajrahāsa’s commentary on the root tantra is preserved in the Ldan dkar ma catalogue, thus dating the root tantra definitively before the early 9th century (Campbell 2009, 46). 118 Indeed most scholarship even on specific tantras makes no attempt to do so. The format of the Introductions to the 84000 Project translations of the Tibetan canon, which require the translator to provide at least some summary of the text she has translated, are therefore a welcome addition to the scholarship on Buddhist canonical literature (See http://read.84000.co/). 119 This is Vilāsavajra’s Śrīguhyasamājatantranidāṇagurūpadeśabhāsya (Tōh. 1910). The introductory narrative of the Guhyasamāja-tantra of course follows the traditional sūtric narrative stating “Thus I have heard at one time, the Bhagavan was residing at....” Usually the location of his residence in the sūtras is a location in India, such as Rajgir in the instance of many of the Prajñāpāramitā-sūtras. The Guhyasamāja-tantra thus shakes up this traditional narrative structure and brings tantric sexual imagery immediately to the fore by locating the Bhagavan’s residence not in an identifiable geographical location in India, but rather in his consort’s bhaga. 120 Other sections of the tantra, however, clearly prioritize Vairocana (see eg. Chapter 3). In some places we can see signs of the earlier three-family system, with just Vairocana, Akṣobhya, and Amitayus (see e.g. Chapter 9), as well as of a version of the five-family system where Vairocana is primary (see e.g. Chapter 12). 121 This basic thirteen-deity maṇḍala is not represented in either the sādhana literature of the Jñānapāda School or the Ārya School of the Guhyasamāja-tantra, which both have more elaborate maṇḍala structures. I have not, in fact, seen any Guhyasamāja sādhana with this simple thirteen-deity maṇḍala that directly accords with the tantra. 24 retinue of tathāgatas and bodhisattvas), and sets forth short sādhana-like practices including rites for the visualization of the buddhas, akin to those found in generation stage practice manuals, and yogic techniques like the sūkṣma-yoga that we find in perfection stage manuals. However, no full-fledged sādhana for either stage of practice appears in the tantra.122 Additionally, there are maṇḍala rituals, descriptions of sexual yogas, and practices including the injunction to consume the sexual fluids produced from union; injunctions to engage in antinomian behaviors like the consumption of traditionally impure substances; and injunctions to engage in behaviors associated with the three poisons of passion, aggression, and delusion, normally strictly proscribed in Buddhist practice.123 The tantra also describes the practice of mantra recitation, as well as a number of wrathful rituals.124 Like many tantras, the Guhyasamāja-tantra also contains “recipes” and instructions for the ritual production of certain substances that are to be consumed in order to gain (primarily) worldly accomplishments. There is also mention, but not a clear description, of initiation, as well as references to four stages of practice—sevā, upasādhana, sādhana, and mahāsādhana—that have been variously interpreted and used by different authors, including Buddhajñānapāda, to structure tantric sādhana. Given the almost chaotic nature of the Buddhist tantras, many of which are eclectic just like the Guhyasamāja-tantra, commentaries providing interpretive frameworks and liturgical manuals providing practical details were necessary to give practitioners avenues for engagement in these complex systems. These were always supplemented by oral instructions, as well, passed down through a lineage.125 According to the later Tibetan tradition, there were two main schools of Guhyasamāja exegesis and practice in India: the Jñānapāda School, eponymously named for Buddhajñānapāda, and the Ārya School, presumably named after its main exponent (Ārya) Nāgārjuna. These two were, however, not the only Guhyasamāja traditions practiced in India. A Guhyasamāja sādhana by the Indian tantric exegete Candrakīrti, extant in its original Sanskrit, refers to four schools of Guhyasamāja practice distinguished by the number of deities in their maṇḍalas; however, only the two mentioned above seem to have been passed on into the Tibetan tradition and we have little evidence of these other systems.126 Among these two major 122 The tantra does, however, contain many short ritual sequences describing practices that seem to pertain to what later became classified as both of the two stages of practice. The nature of the relationship between sādhanas and other such authored ritual manuals and the tantras to which they are connected remains a topic that merits further study. 123 These more antinomian elements of the Buddhist tantras have been interpreted variously by both traditional and modern scholars alike. Wedemeyer (2013) provides a helpful analysis of the various ways in which such practices have been, and ought to be read. 124 Wrathful rituals gain prominence from chapters thirteen onwards. 125 Indeed, it is said that the tantras are actually intentionally structured in a chaotic and confusing way specifically so that their content cannot be approached or practiced without the assistance of not only commentaries, but also oral instructions on the practices received directly from a guru. 126 Tomabechi 2008, 171n1. His reference is to Candrakīrti’s Vajrasattvaniṣpādanasūtra. See Hong and Tomabechi 2009, 35 for the Tibetan edition. The Guhyasamāja-based sādhanas found at Dunhuang, however, are exceptional in displaying no evidence of a distinction between the Jñānapāda and Ārya schools, and thus perhaps represent a stage of Guhyasamāja practice before this distinction developed. See J. Dalton and Van Schaik 2006 for a catalogue of the Tibetan tantric manuscripts from Dunhuang. I have, however, noticed what I find to be a curious distinction in ritual structure between the Indian Guhyasamāja sādhanas (and in fact almost all other tantric sādhanas) preserved in the Tibetan Tengyur and those preserved at Dunhuang, which I believe suggests that the Dunhuang Guhyasamājarelated manuscripts somehow relate to a different “strand” of liturgical works than those preserved in the Indic sources. That is, throughout Dalton and van Schaik’s catalogue there are references to the three samādhis “of Mahāyoga,” by which Dalton and van Schaik mean the samādhi of suchness (de bzhin nyid kyi ting nge ‘dzin), the all-illuminating samādhi (kun tu snang ba’i ting nge ‘dzin) and the causal samādhi (rgyu’i ting nge ‘dzin). Dalton (2004, 9) takes these same three samādhis to be characteristic of the generation stage practice of the Mahāyoga tantras on the whole. Indeed, in the Tibetan Mahāyoga practices of the Nyingma tradition up until the present day 25 Guhyasamāja practice traditions, Buddhajñānapāda’s is the earlier one. The Ārya School’s later date is determined both by its reliance not on the Guhyasaṃāja-tantra itself for its maṇḍala and ritual system, but on the later Guhyasamāja explanatory tantras—the Vajramālā, the Caturdevīparipṛcchā, the Saṃdhivyākaraṇa, and the Vajrajñānasamuccaya—and also by its more developed exegetical tradition.127 these three are exactly the three samādhis that pertain to the generation of the deity in generation stage practice, and indeed these are the three samādhis as they are represented in sādhanas and other Mahāyoga works at Dunhuang, including those that pertain to the Guhyasamāja-tantra. But when we come to Indic liturgical texts, we find a different story. A search of the Tibetan Kangyur and Tengyur shows only four texts—two tantras and two commentaries—that set forth the particular system of three samādhis common at Dunhuang and in the later Tibetan Nyingma tradition. One of these four, though, is the Dgongs pa ‘dus pa’i mdo, which Dalton (2002) has shown to be a Tibetan composition rather than an Indian text. The other three are the Śrīherukakaruṇākrīḍita-tantra, translated by Śrīkirti, which is one of the Rnying rgyud; Mañjuśrīmitra’s Bodhicittabhāvanādvādaśārthanirdeśa (Tōh. 2578), which has no translator’s colophon; and Vajravarman’s Vajravidāraṇādhāraṇīvṛtti, translated by Śraddhākaravarman and Rinchen Zangpo. (The Śrīherukakaruṇākrīḍita-tantra and Mañjuśrīmitra’s text both list only the second and third samādhis—the kun tu snang ba’i ting nge ‘dzin and the rgyu’i ting nge ‘dzin by name— but both also describe a samādhi prior to these two. The description of that first samādhi in both of these texts can, however, be correlated to the de bzhin nyid kyi ting nge ‘dzin.) Thus we find this particular set of three “Mahāyogasamādhis in only three Indic texts surviving in the Tibetan canon, one of which is a from the Rnying rgyud section of the Kangyur, and another of which lacks a translator’s colophon. What we do find frequently in the Kangyur and Tengyur, as well as in the surviving Sanskrit manuscripts, like those of the Sāramañjarī, is a different set of three samādhis, those already well known from the Yoga tantras. These three samādhis continue to be used throughout the liturgical literature preserved in Sanskrit and in Tibetan translation of the so-called Mahāyoga tantras (including every single reference in the Kangyur and Tengyur to the three samādhis within the Guhyasamāja corpus), all the way up through works on the later Yoginī tantras like the Cakrasaṃvara, Hevajra, Sampuṭa-tantras. These other three samādhis, which are used in the Jñānapāda School and the Ārya School alike to structure their generation stage practices are the preliminary yoga samādhi (ādiyoga-samādhi), the supremely victorious maṇḍala samādhi (maṇḍalarājāgrī-samādhi), and the supremely victorious action samādhi (karmarājāgrī-samādhi). Tanaka (1996, 259), who works primarily with Indic texts, has observed that, “The stage of generation in late Tantric Buddhism is divided into three further stages, called ādiyoga nāma samādhi, maṇḍalarājāgrī nāma samādhi, and karmarājāgrī nāma samādhi...” We can compare this to the observation of van Schaik, who works primarily with Dunhuang texts, that “Meditation in Mahāyoga sādhanas tends to proceed along the structure of the three concentrations (ting nge ‘dzin, Skt. samādhi)...These three are: (i) The concentration on suchness (de bzhin nyid), (ii) the concentration on total illumination (kun tu snang ba), and (iii) the concentration on the cause (rgyu)” (van Schaik 2012, 13). The fact that there is such a clear difference in the set of three samādhis that are consistently used in Tibetan Dunhuang tantric texts and those that are consistently used in the Indic tantric texts suggests to me that these two groups of texts pertain to, for lack of a better term, different ritualstrands.” Certainly, more research is necessary to determine more about the relationship between the Guhyasamāja-related texts from Dunhuang and those preserved in the Tibetan canon. Christian Wedemeyer is currently working on some of the Dunhuang Guhyasamāja sādhanas; his work will be a very welcome addition to the very little research that has thus far been done on this topic. 127 See Matsunaga 1977, 115-16. Despite the fact that it is historically later, the Ārya School has received by far the most scholarly attention, both within the Tibetan scholastic tradition and the modern academy. There are several reasons for this. First, it was the Ārya School’s practice tradition of Guhyasamāja that became popular in Tibet, especially in the Tibetan Gelugpa school, and consequently a number of scholarly commentaries and sādhanas have been composed following that tradition up to the present day. The presence of a living Tibetan tradition and indigenous Tibetan commentarial literature on the Ārya School has made its Indian source texts (whether in the original Sanskrit or in Tibetan translation) much more approachable for modern scholars. Additionally, the Ārya School has fascinated modern scholars because of the curious names of its major exponents—Nāgārjuna, Āryadeva, and Candrakīrti—which are the very names of earlier Indian philosophers associated with the Madhyamaka philosophical tradition founded by the “original” Nāgārjuna, i.e. the non-tantric author of the Mūlamadhyamikakārikās. See Wedemeyer 2007, 7-43 for a detailed treatment of this issue. Moreover, a larger number of Indian texts pertaining to this tradition survive in their original Sanskrit. Following Alex Wayman’s (1977) pioneering inquiry into the Ārya School of Guhyasamāja exegesis and practice—an informative if somewhat rambling monograph—more recent scholarship (Tanaka 1999-2004c; Tomabechi 2006; Wedemeyer 2007; Campbell 2009; Hong and Tomabechi 2009; Wright 2010; Bentor 2010; Kittay 2011; and Columbia University team, forthcoming) has devoted some further attention to the study and translation Ārya School texts. With respect 26 The two systems can be generally distinguished by their distinct generation stage maṇḍalas, their perfection stage practices, and the number of initiations in their respective systems. The Jñānapāda School maṇḍala has nineteen deities and centers on Mañjuvajra, as laid out in Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, whereas the Ārya School has a thirty-two deity maṇḍala centering on Akṣobhyavajra as laid out in Nāgārjuna’s Piṇḍīkṛtasādhana. Regarding the perfection stage practices, Buddhajñānapāda’s system (which we must again note has been far, far less studied in the modern academy) is less formally structured than that of the Ārya School but is characterized by a system of three bindu yogas that include the practice of vajrajāpa and several (but not all) of the yogas from the classical ṣaḍanga or sixbranch yogas. These practices are detailed both in his Dvitīyakrama and his Muktitilaka. The Ārya School’s perfection stage system follows the structure of the five stages (pañcakrama) according to Nāgārjuna’s Pañcakrama and Candrakīrti’s Caryāmelāpakapradīpa.128 The early texts of the Jñānapāda School preserve a system of just three initiations, concluding with the prajñājñāna initiation (as it seems the fourth initiation had not yet fully developed in Buddhajñānapāda’s time), whereas the Ārya School initiatory rituals include the fully developed system of four initiations.129 Given the paucity of modern scholarship on the Jñānapāda School our understanding of the relationship between the two traditions continues to develop, and the present study hopes to make some steps toward that understanding. 3. Buddhajñānapāda’s Life Everyone having come, in this way, to know the detailed accounts of my life, should use all methods to please the sublime and sincere learned one, and listen to and contemplate his teachings and compositions. Through relying upon this, remaining in isolated places, training one’s mind in suchness, and genuinely realizing the way things are, it is possible to attain awakening in this very life, or even in just six months— who could refute this!? -Buddhajñānapāda, Dvitīyakrama Name to the study of Ārya School texts Wright (2010) and Tomabechi (2006) have made editions, translations, and studies of, respectively, Nāgārjuna’s Piṇḍīkṛta, the central Ārya School generation stage sādhana, and his Pañcakrama, a perfection stage manual. Their studies of these systems are supplemented by Wedemeyer’s (2007) Tibetan and Sanskrit edition, translation, and study of Āryadeva’s Caryāmelāpakapradīpa, which further elaborates Nāgārjuna’s perfection stage system, and Tomabechi’s (2006) Tibetan edition of the same text. In a series of articles Tanaka has presented an edition of Nāgabodhi/Nāgabuddhi’s Śrī-guhyasamājamaṇḍalopāyikā-viṃśati-vidhi, a maṇḍala ritual according to the Ārya School’s Guhyasamāja maṇḍala (See Tanaka 1999-2004c). Hong and Tomabechi (2009) present an edition of and brief introduction to Candrakīrti’s Ārya School Vajrasattvasādhana. Campbell (2009), Bentor (2010), Kittay (2011), and the team from Columbia (forthcoming) all deal primarily with the hermeneutical approach of the Ārya School, rather than its ritual systems. Campbell’s work considers the hermeneutics of Candrakīrti’s Pradīpoddyotana, an Ārya School commentary on the Guhyasamāja root tantra, an edition and translation of which are currently being prepared by a team at Columbia University. Bentor examines the Ārya School interpretations of a single verse from the root tantra, while Kittay looks at the question of hermeneutics with respect to the Vajramālā, an explanatory tantra of the Guhyasamāja-tantra that sets out the maṇḍala and ritual system of the Ārya School. Kittay’s study includes a complete English translation of the Vajramālā from the extant Tibetan translation of the text (the Sanskrit is no longer extant) but does not include a Tibetan edition. 128 Although the Tibetan tradition asserts the Caryāmelāpakapradīpa to be a “meaning commentary” (don ‘grel) on Nāgārjuna’s Pañcakrama, Tomabechi’s analysis of the relationship between these texts has led him to conclude that the Caryāmelāpakapradīpa is actually earlier than the Pañcakrama (Tomabechi 2006, 36-38). 129 Vaidyapāda’s Yogasapta, which I examine in Chapter Seven, shows that “the fourth,” though not yet identified as a separate initiation was indeed part of early Jñānapāda School practice. 27 Even before we address the details of his life and work, a few words ought to be said about Buddhajñānapāda’s name, which is reported with quite some variety in the works available to us which include Buddhajñānapāda’s own writings, the colophons of the Tibetan translations of his works, the writings of his direct disciples and other Indian authors (in Tibetan translation and in Sanskrit), and indigenous Tibetan texts.130 One feature of some, but not all, of Buddhajñānapāda’s writings, is the inclusion of his own name, cleverly inscribed within the dedicatory verses. In all five of the works in which he does this, he uses the words buddhajñāna (sangs rgyas ye shes), making it likely that this was the name he used for himself.131 Pāda is added as an honorific suffix to the names of quite a number of Indian masters, including many in Buddhajñānapāda’s lineage. Although Buddhajñānapāda (sangs rgyas ye shes zhabs), 132 and Buddhajñāna (Sangs rgyas ye shes),133 are the most common names found in the colophons to his works, we find other versions of his name in the Tibetan colophons, as well: Buddhaśrījñāna (Sangs rgyas dpal ye shes),134 and Buddhaśrījñānapāda (Sangs rgyas dpal ye shes zhabs/ Sangs rgyas dpal kyi ye shes zhabs).135 In Indic works by his direct disciples and later writers we find a similar variety: Buddhajñanapāda, Buddhajñāna, Jñānapāda (ye shes zhabs), Śrījñānapāda (Dpal ldan ye shes zhabs), and Buddhaśrījñānapāda. Tibetan authors similarly run the gamut, naming him Buddhajñānapāda, Buddhajñāna, Buddhaśrījñāna, and Jñānapāda. Adding śrī to various places in his name was presumably done out of respect, and shortening names is also common practice. We can certainly not assume that all of the works in the canon ascribed to authors under these names were written by one and the same Buddhajñānapāda (and I will address the attribution of each of the extant works under the whole variety of these names below), but it is clear that a wide variety of names was used in various contexts to refer to the particular individual whom I here refer to as Buddhajñānapāda. Not only does this seem to be the most commonly used of his various names (though the shortened version, Jñānapāda may also be in the running for common usage), but since it seems he called himself Buddhajñāna, I have settled on Buddhajñānapāda. Dates and Early Life as a Student We have already heard the story of Buddhajñānapāda’s life as he himself tells it, along with the details that Vaidyapāda adds to his guru’s autobiographical narrative. These early Indian accounts were supplemented by later Tibetan histories whose reports of Buddhajñānapāda’s life, albeit written hundreds of years after his death,136 contain additional details not found in the 130 In the secondary literature he is usually referred to as Jñānapāda, Buddhaśrījñāna, Buddhaśrījñānapāda, or Buddhajñānapāda. 131 The use of his name in the dedicatory verses is found in his Caturaṅga/Samantabhadra-sādhana, Muktitilaka, Ātmasādhanāvatāra, *Gativyūha, and Sañcayagāthā-pañjikā. 132 In the colophons of his Guhyajambhalasādhana (D and P), *Gativyūha (D), and Mahāyānalakṣaṇasamuccaya (D and P). It is also found in the colophon of Śākyamitra’s Mukhāgama, which is often attributed to Buddhajñānapāda. 133 In the colophons of his Caturaṅga/Samantabhadra-sādhana (D and P), Ātmasādhanāvatāra (D and P), Śrīherukasādhana (D and P), Bhaṭṭārakāryajambhalajalendra-sādhana (D and P), and *Gativyūha (P). 134 In transliteration as Bud dha śrī jñā na in his Sañcayagāthā-pañjikā (D), and probably erroneously in transliteration as Bud dha śrī ka jñā na in the Sañcayagāthā-pañjikā (P). 135 In the colophons of the Dvitīyakrama (D and P), the Muktitilaka (D and P), and the Caturaṅgasādhana (D and P). 136 The Tibetan histories I have relied upon in this study are Chögyal Phagpa’s 13th-century Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa; Lotsāwa Zhonnu Pel’s 15th-century Deb ther sngon po (see also Roerich 1976); Tāranātha’s 17th-century Rgya gar chos ‘byung (See also Chimpa and Chattopadhyaya 1970) and Bka’ babs bdun (See also Templeman 1983); Amye Zhab’s 17th-century Gsang ‘dus chos ‘byung and his Gshin rje chos ‘byung; and Dudjom Jigdral Yeshe Dorje 20th-century History of the Nyingma School. This is certainly not an exhaustive list of the Tibetan accounts of Buddhajñānapāda’s life. I have not been able to include here all of the 28 Indian sources and appear to thus be based both on the Indian sources (i.e. Buddhajñānapāda, Vaidyapāda, and other Sanskrit sources no longer available to us), as well as possibly on oral history.137 Vaidyapāda seems to assume knowledge of such oral histories already in his commentary on the Dvitīyakrama when he suggests that the “detailed accounts” that Buddhajñānapāda mentions include “the taming of Nālandā, making offerings at Vajrāsana, [the account] of the consecration and others.”138 Since Vaidyapāda himself offers no further details, it seems he assumes his reader will know of the events he references simply by giving these names. As we shall see, although some of the Tibetan histories elaborate on and interpret aspects of Buddhajñānapāda’s life story in ways that seem to serve an apologetic or polemical function, we have no outright reason to doubt some of the details that are added in others of them, and they certainly add richer texture to what we know of Buddhajñānapāda’s life. We are unusually fortunate in the realm of medieval Indian history to be able to give fairly precise dates for Buddhajñānapāda’s life. Of the six human teachers he mentions in the Dvitīyakrama (setting aside for the moment his most important guru, Mañjuśrī himself), three are historically identifiable figures—Haribhadra, Vilāsavajra, and Pālitapāda—all of whom can be dated to the 8th century. In the colophon of his Abhisamayālaṃkārālokā Haribhadra states that it was completed at the Trikaṭuka Monastery under the reign of the Pāla king Dharmapāla (r. ca. 775-812),139 and Tāranātha notes that Dharmapāla was a patron of both Haribhadra and Buddhajñānapāda, describing both masters as Dharmapāla’s gurus.140 Vilāsavajra likewise lived in the latter part of the 8th century.141 Moreover one of Buddhajñānapāda’s own works, completed in the early part of his life, appears in the lDan kar ma catalogue which was completed in 824.142 On this basis we can reliably place Buddhajñānapāda’s life in the latter part of the 8th and early part of the 9th century. In a recent article Péter Szántó has drawn attention to an account in Atīśa’s

an offering of his kingdom, his queen, and himself, to Buddhajñānapāda, which he later ransomed back with their weight in gold.143 This would mean that Buddhajñānapāda’s activity details included in these histories that differ from or supplement Buddhajñānapāda’s and Vaidyapāda’s accounts— they are divergent enough that such an endeavor would require a separate study, and my main concern here is with Buddhajñānapāda’s life and thought as we find it presented in the Indic tradition, rather than the later Tibetan interpretations of his life and tradition. However, I have included some accounts from the Tibetan sources, as the Tibtean historians do significantly supplement our knowledge of Buddhajñānapāda’s life, particulary several accounts from his later life that are only briefly mentioned in Vaidyapāda. 137 With regard to other no longer extant or accessible Indic sources, Tāranātha, whose accounts of Buddhajñānapāda’s life include quite a number of details not found in the Indian texts known to us, explains that his work was based on several Sanskrit sources that now appear to be lost, including one of the Magadhan scholar Sa dbang bzang po, whose work covers the history up to the reign of Rāmapāla (r. ca. 1072-1126) (Sanderson 2009, 89). The later Tibetan accounts, moreover, unsurprisingly draw from the earlier Tibetan accounts, as well. 138 na landa ([landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus (Sukusuma, D 135b.4; P 163a.8). Elsewhere in his commentary Vaidyapāda references and cites from a number of textual sources, so it seems that these references to stories of Buddhajñānapāda’s life, which are mentioned by topic rather than by referening a particular text, do probably refer to oral accounts. 139 Chimpa and Chattopadhyaya 1970, 266n4; Ruegg 1981, 101n320; and Sanderson 2009, 93-4. 140 Rgya gar chos ‘byung, 262; Chimpa and Chattopadhyaya 1970, 274. The Pālas were, to quote Sanderson, a “robustly” Buddhist East Indian dynasty who used the Buddhist dharmacakra as a royal emblem, whose inscriptions began with praise to the Buddha, and a number of whose rulers are described in the manuscript and inscriptional record with the epithet paramasaugataḥ (Sandserson 2009, 87). 141 Tribe 1994, 9-23. 142 Tomabechi (2008, 175) cites Hadano on this point. This may refer to his Sañcayagāthā-pañjikā, the translation of which is indeed attributed to the Imperial era translators Kawa Paltsek and Vidyākarasiṃha. 143 Szántó 2015, 539. 29 stretched perhaps further into the 9th century than previously thought, that is at least until some time after 812 CE when Devapāla ascended the throne. However, an identical account of the king offering himself and his queen to Buddhajñānapāda and later ransoming themselves back by paying their weight in gold is also recorded in Tāranātha’s Seven Transmissions (Bka’ babs bdun) and a different version of the story in which the king builds a temple in gratitude to Buddhajñānapāda is found in Chögyal Phagpa’s Biography and Lineage History of Jñānapāda’s Guhyasamāja. In both of those sources, though, the king in question is identified as Dharmapāla, rather than Devapāla.144 While these accounts at least place in question Atīśa’s version of the story involving King Devapāla, Tāranātha seems to have either received mistaken information or misunderstood his sources on the Pāla succesion, as he incorrectly places Dharmapāla after Devapāla rather than before him and reverses some (but not all!) information about their respective reigns and activities.145 Moreover, Atīśa was both geographically and historically much closer to the events than either Chögyal Phagpa or Tāranātha, which perhaps makes his report more likely to be accurate. Another factor to consider is that Buddhajñānapāda is asserted by Tāranātha to have been the first tantric ācārya at Vikramaśīla, and it seems likely that it was Devapāla, not Dharmapāla (as Tāranātha states) who founded Vikramaśīla Monastery. Devapāla’s having constructed Vikramaśīla is mentioned in the colophon of Anupamavajra’s Ādikarmapradīpa,146 and this is further substantiated by the colophon of Atīśa’s Ratnakaraṇḍodghāta, where he states that he composed the work “At the great temple called Vikramaśīla, the commitment of Devapāla,” presumably indicating that the construction of Vikramaśīla was the fulfilment of one of King Devapāla’s tantric commitments.147 If that is the case, and if Buddhajñānapāda was indeed Vikramaśīla’s first tantric ācārya, then he did very likely live into Devapāla’s reign. The late 8thcentury dating of his gurus Haribhadra and Vilāsavajra would also suggest the likelihood of Buddhajñānapāda’s having lived at the end of the 8th century, and well into the 9th.148 But if Tāranātha’s account of his life, in which Buddhajñānapāda is said to have lived for 80 years, is correct, then his life could easily have spanned a good part of both of those centuries. Buddhajñānapāda’s own account of his life tells us nothing of his birth or childhood but begins only with his studies. It appears not to be until an account of his life written by Chögyal Phagpa in the 13th century, that we learn more about his early life, though the information in our sources here varies significantly. Chögyal Phagpa tells us that Buddhajñānapāda was born in a place called Sindhura in southeast India to a king called Gyaparuprabhava (!?),149 whereas Tāranātha, in the 17th century, reports that according to some sources (to which we unfortunately no longer have access) Buddhajñānapāda was a brahmin who was ordained at Nālandā into the Mahāsāṃghika school, while according to others he was a kṣatriya “reader” (scribe?).150 As 144 (Bka’ babs bdun, 106; Templeman 1983, 74; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 617 145 See Sanderson 2009, 90-1. Tāranātha does, though, correctly assert Dharmapāla to roughly be a contemporary of King Trisong Deutsen (r. ca. 755-797) (Rgya gar chos ‘byung, 264; Chimpa and Chattopadhyaya 1970, 276). 146 Sanderson 2009, 91. 147 de wa pā la’i thugs dam bi kra ma/ shī la zhes bya’i gtsug lag khang chen du (Ratnakaraṇḍodghāta, D 116b.4). 148 See Sparham 1989, 3 on Haribhadra’s dates and Tribe 1994, 9-23 on Vilāsavajra’s. 149 Rgyal po gya pa ru pra bha wa. (Gsang ‘dus ye shes zhabs kyi rnam thar dang brgyud pa’i rim pa, 610). Amye Zhab’s 17th century Gshin rje chos ‘byung repeats this information (Gshin rje chos ‘byung, 304a.4). His account begins as a slightly shortened, but otherwise word-for-word, copy of Chögyal Phagpa’s account though there are places where Amye Zhab follows Vaidyapāda more or less word-for-word rather than Chögyal Phagpa. When he comes to the account of Buddhajñānapāda’s meeting with the monk in the Kuvaca forest, however, Amye Zhab’s account diverges from both. 150 Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 610; Bka’ babs bdun, 103; Templeman 1983, 71. On Tāranātha’s sources, now lost to us, see note 137. Tāranātha’s statement about Buddhajñānapāda’s ordination 30 Buddhajñānapāda himself tells us, he spent a number of years in the earlier part of his life traveling quite extensively throughout the subcontinent studying and practicing under the guidance of different gurus with whom he remained for varying amounts of time—he mentions staying with one guru for just eight months and another for nine years. The teachers Buddhajñānapāda mentions in his account are Haribhadra, with whom he studied in the town of Takṣaśilā, in the area of Khapir, in Magadha;151 Vilāsavajra, Guṇeru and Jātig Jālā, with whom he studied in Uḍḍiyāna;152 Bālipāda, who lived in Ko no dze in the area of Jālandhara;153 and Pālitapāda, who stayed at “the place with sky trees” in the Koṅkan, most probably at modern-day in the Mahāsāṃghika school is the only one I have seen that identifies him as having taken monastic ordination, though his early career studying with Haribhadra and teaching at Nālandā are suggestive of his having monastic status, at least in the early part of his life. As for Chögyal Phagpa’s account of Buddhajñānapāda’s life, he closely follows Vaidyapāda’s narrative from the account of Buddhajñānapāda’s studies with Haribhadra onward, but also includes episodes not reported in Vaidyapāda. His source for the earlier information on Buddhajñānapāda’s birth and parentage is unclear, but its style of delivery (in addition to his being the son of a king, he also notes that from a young age Buddhajñānapāda was handsome and intelligent, etc.) is classically hagiographical. Phagpa also implausably states that from Pālitapāda (who he styles Balipata) Buddhajñānapāda received initiation into the thirtytwo deity Guhyasamāja maṇḍala (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 612). (Tāranātha also states that in the Koṅkan Pālitapāda heard teachings directly from Candrakīrti and received the text of the Pradīpodyotana, so there do appear to be some accounts of Pālitapāda’s having been an Ārya School practitioner (Rgya gar chos ‘byung, 260; I disagree with the translation of this passage in Chimpa and Chattopadhyaya 1970, 273).) Phagpa’s account also contains the first version that I have seen of the expanded (from Vaidyapāda’s narrative) story of Buddhajñānapāda’s meeting with Mañjuśrī, in which he travels toward Wutaishan in hopes of meeting Mañjuvajra. I have translated this episode in full below. Tāranātha, Amye Zhab, and Dudjom all report abbreviated forms of this same account, which seem to be based on Phagpa’s, or else on the same sources he was using, though Phagpa’s hagiography contains several accounts not found in either of the others. See Templeman 1983, 72-3 and Dudjom 1991, 495-6 for English translations of Tāranātha’s and Dudjom’s versions of this story, respectively. (Templeman’s translation, in particular, suffers from a number of errors, some of which are understandable given that he was apparently not familiar with Vaidyapāda’s text, from which Tāranātha derives much of his material). While Phagpa’s account of Buddhajñānapāda’s life in some places seems implausible, and in this particular episode is so detailed as to seem suspicious (it even includes dialogue between Buddhajñānapāda, the monk, and the woman!), he reports multiple versions of certain portions of the account (some of which are included in Tāranātha’s, Amye Zhab’s, and Dudjom’s later tellings, and others of which are not), suggesting he was relying on multiple sources. At one point Phagpa details an account of Buddhajñānapāda’s visit to his guru Pālitapāda’s residence which is reported in Samantabhadra’s Sāramañjarī. It is possible, then, that other of these more detailed accounts of certain episodes of Buddhajñānapāda’s life were present in Indian sources we no longer have access to, or are at present buried in extant texts, like the account of the visit to Pālitapāda in the Sāramañjarī, and are simply waiting to be unearthed. 151 The toponyms mentioned in the Dvitīyakrama and the Sukusuma are difficult to ascertain with certainty. The place names mentioned here in the context of Buddhajñānapāda’s studies with Haribhadra are Dbu kyi yul chen, Kha pir (pir] P, bir D), and Rdo ‘jog. Rdo ‘jog is a common translation of Takṣaśilā (see C. Dalton and Szántó, forthcoming). However, the region in which the town is said to be located, Kha pir, calls into question this being the commonly-known city of Takṣaśilā located in modern-day Pakistan. On one hand, Kha pir may be a corrupted rendering of Kaspir, i.e. Kaśmir (see C. Dalton and Szántó, forthcoming). However, since this particular Khapir is specified as being in Magadha, such an identification is only possible if Magadha is understood to mean the Indian subcontinent more broadly, rather than the region of Magadha, which is not near Kaśmir. But Vaidyapāda describes Magadha as “in the area of Nālandā,” which again renders the identification of Khapir as Kaśmir difficult (Sukusuma, D 89a.7). 152 We may assume Uḍḍiyāna to be the region in the northwest of the subcontinent, often identifed as the Swat Valley in modern-day Pakistan. 153 Again here we have some difficulty identifying these locations. At first glance Ko no dze, or in the Derge edition of Vaidyapāda’s commentary Ka no dze (P reads ko no dze), does seem to be a transliteration of Kannauj, and Davidson (2002, 312) has rendered it as such. However, Szántó (2015, 542-3) places some doubt on this identification, since modern-day Kannauj is not near the modern-day city of Jalandhar, and C. Dalton and Szántó (forthcoming) note that at the time Kannauj was referred to as Kanyākubja, making the identification even less likely. 31 Kadri.154 His most important guru was Mañjuśrī himself, who appeared to Buddhajñānapāda as an “emanated monk” who then emanated Mañjuśrī and his maṇḍala for Buddhajñānapāda in the Kuvaca forest behind Vajrāsana.155 While the exact locations of these encounters are difficult to ascertain with certainty based on the toponyms given in his account, what is clear is that Buddhajñānapāda’s travels took him for thousands of kilometers across a wide swath of the subcontinent from its north-central area to the far northeast, to the southeast, and finally to the northeast. Buddhajñānapāda’s Gurus and Their Possible Influence on his Thought Haribhadra As for his teachers and his studies, among those figures we are able to identify and about whom we know more from other sources, Haribhadra, the only non-tantric guru that Buddhajñānapāda mentions studying under,156 is a well-known late 8th-century scholar of Prajñāpāramitā literature whose Abhisamayalaṃkārālokā remains important in many Tibetan monastic curricula even up to the modern day. Despite his own preference for writing on tantric topics, Buddhajñānapāda, who is considered Haribhadra’s principle disciple, was clearly influenced by this master. One of Buddhajñānapāda’s two non-tantric works, the Sañcayagāthāpañjikā, composed while he was staying at Nālandā early in his life and which is by far the longest of all his extant writings, is an extensive Prajñāpāramitā commentary. In this work he follows in Haribhadra’s footsteps in synthesizing Madhyamaka philosophy with the Abhisamayālaṃkāra.157 Vilāsavajra Vilāsavajra, the first tantric guru that Buddhajñānapāda mentions, is likewise a wellknown late 8th-century author158 who wrote treatises on a number of tantras including the Mañjuśrīnāmasaṃgīti, the Guhyagarbha-tantra, and the introductory section (nidāna) of the Guhyasamāja-tantra.159 If indeed the Vilāsavajra who authored the Nāmamantrārthāvalokinī 154 nam mkha’i shing ldan. Vaidyapāda’s etymology for the toponym reads: de la nam mkha’i shing ldan zhes bya ba ste ci phyir zhe na/ rtsa ba med par shing rnam ‘khril (P] ‘khril; D la ‘khris) shing steng du bras (P] bras; D bris) pa lta bur gnas pa’o// (D 90a.4). According to Khenchen Chodrak Tenphel the term “sky tree” (nam mkha’i shing) means mangrove (or perhaps banyan?) (personal communication, March, 2016). The toponym was earlier identified by Davidson as Kāṇherī, based upon his reading of Vaidyapāda’s gloss of the name (2002, 313). I am more convinced, however, by Szántó’s more recent work, which suggests that the place mentioned may rather be Kadri, which is in the Koṅkan near Mangalore, rather than Kānherī (i.e. Kṛṣṇāgiri) which is not usually understood to be part of the Koṅkan. See Szántó (2015, 550-52) for the full details of his assessment, which I also discuss in more detail below in conjunction with Buddhajñānapāda’s guru Pālitapāda. 155 Vajrāsana, we can safely say, is at Bodhgaya in modern-day Bihar. 156 Haribhadra wrote four Prajñāpāramitā texts, but no tantric work attributed to him survives either in Sanskrit or in Tibetan translation in the Tengyur. He was, however, certainly familiar with tantric systems, as he advocates in one of his works that practitioners should visualize their meditational deity in the form of Vajradhara (Sparham 1989, 3). Haribhadra was himself the disciple of the unknown Vairocanabhadra and of the well-known Śāntarakṣita, under whose tutelage he was a co-disciple of Kamalaśīla (ibid., 2-3). 157 Ruegg 1981, 101-2. However, it appears that Buddhajñānapāda did not follow his guru Haribhadra’s interpretation of there being four kāyas instead of the three taught in the Abhisamayālaṃkāra (Makransky 1997, 6). The Sañcayagāthā-pañjikā is the only one of Buddhajñānapāda’s extant works that I did not read in full as part of this study. This work, and particularly its relationship to Haribhadra’s Prajñāpāramitā writings, is a topic that certainly deserves further attention. 158 On Vilāsavajra’s dates see Tribe 1994, 9-23. 159 Vilāsavajra’s Mañjuśrīnāmasaṃgīti commentary is the Nāmamantrārthāvalokinī, his Guhyagarbha commentary is the Spar khab, and his Guhyasamāja commentary the Śrīguhyasamājatantranidāṇagurūpadeśabhāṣya. In his 32 commentary on the Mañjuśrīnāmasaṃgīti was Buddhajñānapāda’s guru, we find in that text the unusual instance of a guru citing his disciple’s work.160 However, this indeed appears to be the case here; the work cited by Vilāsavajra, the Mahāyānalakṣaṇasamuccaya, was composed by Buddhajñānapāda early in his career, most probably prior to his discipleship under Vilāsavajra.161 Moreover, we can say with certainty that Buddhajñānapāda knew Vilāsavajra’s writing on the Mañjuśrīnāmasaṃgīti, since in his Ātmasādhanāvatāra Buddhajñānapāda reproduces (without telling us he is doing so) a lengthy section of Chapter Five of Vilāsavajra’s Nāmamantrārthāvalokinī concerning the correspondences between the deities from the Mañjuśrīnāmasāṃgīti maṇḍala and a number of Mahāyāna categories.162 We can discern a number of possible ways in which his studies with Vilāsavajra may have influenced Buddhajñānapāda’s thought. Vilāsavajra’s Nāmamantrārthāvalokinī is written with an emphasis on Vijñānavādin perspectives combined with an acknowledgement of Madhyamaka systems163 that is likewise prominent in Buddhajñānapāda’s writings, and the central figure of Mañjuśrījñānasattva in the sādhana found in the fourth chapter of that work, which we now know that Buddhajñānapāda was familiar with, may perhaps have influenced Buddhajñānapāda’s own use of Mañjuvajra as the central figure of the generation stage maṇḍala in his Samantabhadra-sādhana.164 In any case, it seems likely that the identification of Mañjuśrī as an Ādibuddha figure in the Mañjuśrīnāmasaṃgiti and the fact that this practice system was taught by his guru may have been a factor in Buddhajñānapāda’s important personal connection with Mañjuśrī. And yet while Vilāsavajra also composed a Guhyasamāja-tantra commentary, and the Guhyasamāja was clearly the most important tantra for Buddhajñānapāda, there is no mention of Buddhajñānapāda’s having studied the Guhyasamāja with him.165 The most intriguing possibility of Vilāsavajra’s thought influencing Buddhajñānapāda’s concerns the former’s relationship with the Guhyagarbha-tantra. While Buddhajñānapāda makes no mention at all of the Guhyagarbha in his oeuvre, he is mentioned in some later Tibetan histories and polemical treatises as an Indian author who uses the term “great perfection” (rdzogs chen) in his works. The term indeed appears in two instances in Buddhajñānapāda’s writings in Tibetan translation (unfortunately the Sanskrit of neither passage is extant), but the context of its usage in these instances is less suggestive of a connection with great perfection traditions than are a number of other strains of his thought, including references to the immediate and direct study of the Nāmamantrārthāvalokinī Tribe notes that while the ascription of the Guhyasamāja commentary and the Guhyagarbha commentary cannot be definitively made to the author of the Nāmamantrārthāvalokinī, there are also no substantial reasons for doubting the ascription (Tribe 1994, 14). J. Dalton likewise asserts the same with regard to the attribution of authorship of the Spar khab to Vilāsavajra (J. Dalton 2005, 125n28). 160 This was first noted by Tribe (1994, 16). 161 Szántó 2015, 541 and C. Dalton and Szántó forthcoming. 162 This passage in the Ātmasādhanāvatāra reproduces not a continuous segment of Chapter Five of the Nāmamantrārthāvalokinī, but rather a number of shorter passages from that chapter that Buddhajñānapāda has strung together to create several pages of the Ātmasādhanāvatāra. The presence of this passage is especially significant since the Nāmamantrārthāvalokinī is extant in Sanskrit, so we now have access to yet another passage of the Ātmasādhanāvatāra in its original Sanskrit. I address this further below in my discussion of the Ātmasādhanāvatāra. 163 Tribe 2016, 11. 164 Tribe (1994, 8n20) suspects as much. Obviously Buddhajñānapāda’s vision of Mañjuśrī would have been a more important factor in this decision. 165 Vaidyapāda mentions only that Buddhajñānapāda studied “many Kriyā and Yoga tantras” with Vilāsavajra (Sukusuma, D 89b.5). As mentioned above, Vaidyapāda classifies the tantras in a number of different ways in the Sukusuma (see note 100), and it is not clear whether he would have included the Guhyasamāja-tantra under the category of Yoga tantras or held it to be exclusively a Mahāyoga tantra. In any case the Guhyasamāja is not explicitly mentioned as a topic of study with Vilāsavajra as it is in the case of Buddhajñānapāda’s later guru Pālitapāda. 33 pointing out of reality by a guru to a disciple, an emphasis on the immediacy of awakening, and overtly anti-ritual rhetoric in the midst of a deeply ritual practice system.166 Pālitapāda The guru with whom Buddhajñānapāda studied the Guhyasamāja-tantra and whom he reports attending for nine years and then returning to visit again later in his life, Pālitapāda, has only recently emerged as a figure about whom more is known. Buddhajñānapāda describes studying with this guru in the Koṅkan in a location called nam mkhashing ldan, literally “the place with sky trees,” and reports that he was surrounded by disciples who had miraculous abilities and that the whole entourage regularly recieved extensive support for their livelihood and practice. The unusual toponym mentioned for the location in the Koṅkan where Pālitapāda resided has been identified by Davidson as Kāṇherī and more recently by Szántó as Kadri, near present-day Mangalore.167 Szántó’s assessment has the advantage of being the location of a Śaiva temple whose deity is called Mañjunātha, and an ancient inscription on the site identifies the place as a vihāra.168 These details suggest that the temple was originally a Buddhist site devoted to Mañjuśrī, which of course fits well with Buddhajñānapāda’s system of Guhyasamāja whose central deity is Mañjuvajra, and suggests the possibility that yet another of Buddhajñānapāda’s gurus, in addition to Vilāsavajra, may have taught tantric practices centered on Mañjuśrī. Such a possibility is furthered by Szántó’s report of a Sanskrit initiation manual called the Parikramapadopāyikā composed by a certain Śrīkīrtipāda who identifies his own guru as Pālitapāda, and mentions that guru as having taught maṇḍala rituals, presumably an initiation manual, in whose spirit Śrīkīrtipāda composed his own work.169 Szántó further reports that the anonymous Mañjuvajrodaya (Tōh. 2590) shares significant parallels with Śrīkīrti’s manual, leaving the alluring possibility that the Mañjuvajrodaya might possibly be Pālitapāda’s work, on which his disciple Śrīkīrti based his own initiation manual.170 Though Szántó does not report this, the Mañjuvajrodaya appears to be a maṇḍala rite for the Mañjuśrīnāmasaṃgīti.171 The fact that its title, and a number of instances in the text, refer to Mañjuvajra places the Mañjuvajrodaya at a very interesting crossroads between the Mañjuśrīnāmasaṃgīti tradition and Buddhajñānapāda’s tradition of the Guhyasamāja. None of our sources report Pālitapāda to have taught the Mañjuśrīnāmasaṃgīti, but that certainly does not preclude his having done so. That the Pālitapāda mentioned by Śrīkīrti and Buddhajñānapāda’s guru by that name are one and the same teacher is especially likely, given that the Indian author Samantabhadra states that his Sāramañjarī, a commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, was composed at the command of one Kīrtipāda.172 Presumably the Kīrtipāda (a.k.a. Śrīkīrti; adding pāda as a marker of respect to the names of members of Buddhajñānapāda’s lineage was 166 I discuss these issues in more detail in Chapter Three. 167 Davidson 2002, 313; Szántó 215, 550-52. 168 Szántó 2015, 551-2. 169 ibid., 552-3. 170 ibid.,553. 171 I have not had the opportunity to study the text in detail, but its location in the Tengyur among quite a few other works devoted to the Mañjuśrīnāmasamgīti (and none devoted to the Guhyasamāja) suggests as much. In addition, the Tohoku Catalogue notes to Buton’s Mtshan brjod kyi dkyil ‘khor gyi bkod pa specify that Buton’s text, a commentary on the Mañjuśrīnāmasaṃgīti from the Yoga tantra perspective and including a presentation of its maṇḍala structure, makes reference to a number of works, including the Mañjuvajrodaya (databases.aibs.columbia.edu). 172 Sāramañjarī, D 1a.3; Szántó 2015, 554. And, yes, it is confusing that a person named Samantabhadra composed a commentary on a sādhana called the Samantabhadra, but he did. 34 common173) who commanded Samantabhadra to compose a commentary on Buddhajñānapāda’s sādhana was a figure in his community who was senior to Samantabhadra, and a co-disciple of Buddhajñānapāda’s under Pālitapāda would fit that role well. Moreover, Szántó reports that Śrīkīrtipāda’s Parikramapadopāyikā is devoted in particular to the choreographical details of the maṇḍala ritual, and Buddhajñānapāda composed an entire text, the *Gativyūha, devoted to such choreographical details, suggesting that this topic may have been a speciality of their common guru. Buddhajñānapāda reports leaving Pālitapāda’s company after nine years when his guru admitted that he himself did not have full realization of the Guhyasamāja-tantra. He then traveled back to North India where he eventually had a vision of Mañjuśrī in the Kuvaca forest behind Bodhgaya, in which the instructions that formed the basis for Buddhajñānapāda’s unique system of Guhyasamāja practice were revealed to him. Nonetheless, Pālitapāda must have had a great influence on his disciple, as Buddhajñanapāda describes returning to visit Pālitapāda in the Koṅkan later in his life, on which occasion he composed some sādhanas to please his guru. This visit took place subsequent to his vision of Mañjuśrī, but presumably prior to his composition of the Dvitīyakrama, given that the visit is mentioned in that text. Vaidyapāda reports that it was the Samantabhadra-sādhana that Buddhajñānapāda composed at Pālitapāda’s request, though Mañjuśrī himself had already commanded its composition in Buddhajñānapāda’s vision.174 Other Gurus The other three human gurus Buddhajñānapāda mentions studying with—Guṇeru, and Jātig Jvālā (the latter also seems to have been his consort), and Bālipāda—are at this point only known to us from their names and the details given in Buddhajñānapāda’s own account. It is worth noting, however, that two of these (unsurprisingly) historically unidentifiable gurus—that is, two among the six human teachers that Buddhajñānapāda names—Guṇeru and Jātig Jvālā, were women. Vaidyapāda also mentions two female co-disciples of Buddhajñānapāda’s under his guru Pālitapāda, the prostitutes Ālokī and Sādhuśīlā; a tantric consort trained in the Guhyasamāja yogas with whom he practiced also under Pālitapāda’s guidance, the butcher girl Vimalamutrī; and one female student of Buddhajñānapāda’s, the nun Guṇamitrā who requested him to compose a text at Nālandā early in his career. Thus we find women occupying the full variety of tantric roles in Buddhajñānapāda’s life—as gurus, consorts, co-disciples, and disciples. Mañjuśrī[-mitra!?] The most important of Buddhajñānapāda’s gurus, whose direct first-person speech makes up ninety percent of the Dvitīyakrama and whose instructions form the basis for Buddhajñānapāda’s system of Guhyasamāja practice, is none other than Mañjuśrī himself who appeared to Buddhajñānapāda in a vision—or so it seems in the Dvitīyakrama and in Vaidyapāda’s Sukusuma. In both of their accounts, the “emanated monk” (in Vaidyapāda’s rendering he is further identified as an emanation of the “Great Vajra Holder”175) who revealed Mañjuśrī and his maṇḍala to Buddhajñānapāda’s is eclipsed by the figure of Mañjuśrī himself as the direct source of Buddhajñānapāda’s revelation. In some (but, tellingly, not all) of the later Tibetan histories, though, the identity of this monk becomes much more central, and it is he who 173 There are several instances of this in addition, of course, to Buddhajñānapāda’s own name. There is his disciple Vaidyapāda, as well as repeated references to Dīpaṃkarabhadra as “Bhadrapāda” in Samantabhadra’s Sāramañjarī. 174 Sukusuma, D135b.1; D 133b.6; Samantabhadrī-ṭīkā, D 131b.6-7. See also Szántó 2015, 547. 175 It is not impossible that this could be translating Vajradhara, as Davidson (2002, 313) has rendered the appellation in English, but the term that the Tibetan translators used, rdo rjedzin pa, is not the common one for Vajradhara, which is rdo rjechang. 35 is celebrated as Buddhajñānapāda’s most important teacher, rather than Mañjuśrī. The story of their meeting is, in several of the Tibetan accounts—starting with Chögyal Phagpa’s and including Tāranātha’s, Amye Zhab’s, and Dudjom’s—significantly expanded upon from Vaidyapāda’s account, on which it is clearly based. Let us first witness this encounter in Chögyal Phagpa’s delightful telling, the earliest expanded version I am aware of, before we address the question of the monk’s identity.176 The episode is preceded by his teacher Pālitapāda being unable to cut through Buddhajñānapāda’s doubts about the meaning of the Guhyasamāja-tantra. [[[Buddhajñānapāda]]] went to Vajrāsana and made supplications to attain great awakening. Then he heard a voice in the sky that made a prophecy: “Son of good family, you must search for Mañjuvajra and through his blessings you will be freed from all of your doubts.” He asked the Ācārya Pālitapāda177 for permission to go to Wutaishan in China, since that is where Mañjuvajra resides, and permission was granted. Setting off from Vajrāsana he headed north, and in the forest called Kuvaca178 he saw that there was a woman and a female dog in front of a hut. Nearby a monk who had made his dharma robe into a turban was plowing a field. [[[Buddhajñānapāda]]] thought to himself, “Alas! There is a monk with a woman plowing a field! The Buddha’s teaching is certainly in decline!” and his heart was heavy. However, it was almost noon, so he was thinking to go there to request alms when the monk said, “Go get the ācārya some food for his alms.” The ācārya [[[Buddhajñānapāda]]] sat down and the monk commanded the woman, “Serve this monk his lunch.” The woman had taken a fishing net to a stream, caught a fish, and cooked it. Then she had placed [it upon] a leaf in front of the female dog. [The monk] having commanded “Serve [his] alms,” the dog vomited [onto the leaf] and [the woman] brought that [vomit] along with the fish to the ācārya. This [meal], which was meant specifically for him, the ācārya [[[Buddhajñānapāda]]] regarded as flesh and filth, and he did not eat it. Another version of the story says that the woman killed many birds and cooked their flesh and brought it [to Buddhajñānapāda] who did not eat it. The woman then snapped her fingers and the cooked [birds] flew away, upon which the ācārya began to have some doubts [about his previous judgments regarding his lunch companions]. Then the monk said, “Alas [he is a] so-called worldly being; give him some ordinary food.” [The woman] then brought some cooked rice and yogurt, which the ācārya did eat. He then thought to depart, but the monk said, “If you leave from here now you won’t find a place to stay tonight; go tomorrow.” So [[[Buddhajñānapāda]]] stayed, but the monk went off somewhere else. The ācārya was reciting the Guhyasamāja aloud, and whenever he came to a place [in the text] that he did not understand, the woman showed a displeased and sad 176 Other versions of this account are found in Tāranātha’s Bka’ babs bdun (104-6), Amye Zhab’s Gshin rje chos ‘byung (46a.3-48a.4) and Dudjom’s History of the Nyingma School. See Templeman (1983, 72-3) for an English translation of Tāranātha’s version of the same encounter, which unfortunately has a number of translation errors. See Dudjom (1990, 494-6) for an English translation of Dudjom Rinpoche’s version. Amye Zhab’s version of the account, in his Gshin rje chos ‘byung, is is reported to be translated in an unpublished article by Hubert Decleer (Decleer, upublished), and is also discussed in Decleer 1998. Decleer’s 1998 article notes the strong parallels between the account in Tāranātha and Amye Zhab’s works and the narrative about Dharmaśrīmitra in the ca. 15th Century Svayambhūpurāṇa (on the Svayambhūpurāṇa’s dates see von Rospatt 2015, 827). 177 Chögyal renders the name Pa li pa ta, but I have corrected it here to what we know to be the name of Buddhajñānapāda’s Guhyasamāja teacher from the Koṅkan. 178 Chögyal renders the name of the forest as ku pa, but I have corrected to the form given in the Dvitīyakrama. 36 expression. The ācārya thought, “This woman certainly is clairvoyant and knows the minds of others! She will be able to cut through my doubts!” He prostrated to her and supplicated [for instruction] but the woman said, “I don’t know. But that monk from before, because he is my husband, is extremely learned. You should ask him.” “Where did he go?” asked Buddhajñānapāda, and [the woman] replied, “[He went] to buy beer.” “When is he coming back?” he asked, and [she] replied “He’ll be back in the late afternoon.” So [[[Buddhajñānapāda]]] waited. In the late afternoon [the monk] came back staggering179 drunk. Seeing that, [[[Buddhajñānapāda]] felt] a lack of faith, but he swallowed his pride, prostrated and made the request, “Please teach me the Guhyasamāja.” “Ask for the initiation,” [the monk] commanded. “I’ve already gotten the initiation,” [[[Buddhajñānapāda]]] replied, upon which [the monk] said, “You must receive the initiation directly from me.” The ācārya searched for the ritual articles and requested initiation. According to a different account he gave [all] the money180 he had to the woman and she emanated the [necessary] ritual articles, so he obtained them [that way]. Then, at midnight on the eighth day of the seventh month [the monk] emanated a celestial palace and at its center was the nineteen-deity maṇḍala of Mañjuvajra, which he emanated and showed directly [to Buddhajñānapāda]. The monk, however, remained in front of the maṇḍala in the very same form in which he had appeared before and asked the ācārya, “Will you receive the initiation from me or from the maṇḍala?” The ācārya, even though he knew that the maṇḍala had been emanated by the monk himself, was inspired181 by the form of the deity and replied, “I’ll receive it from the maṇḍala.” “Fine, take it, then,” said [the monk], and [[[Buddhajñānapāda]]] received the full initiation from the maṇḍala. According to a different account when [[[Buddhajñānapāda]]] said “I’ll receive it from the maṇḍala,” the maṇḍala disappeared and [the monk said] “I emanated the maṇḍala,” upon which Buddhajñānapāda understood the maṇḍala to be the monk’s emanation and prostrated [to him] saying, “You are the father and the mother of all beings...”182 In this way he praised him, asked his forgiveness,183 and supplicated him. At dawn, from [the monk’s] heart-center the maṇḍala was [re-]emanated, and with a smile he said “Excellent!” and bestowed the initiation on him. Then [the monk] began [to explain]184 the condensed meaning of the Guhyasamāja, the [Dvitīyakramatattvabhāvanā]-Mukhāgama and other [instructions], and [thus] brought [[[Buddhajñānapāda]]] to realize all of the key points of the tantra. The ācārya was satisfied and delighted, so he said, “I want to make an offering; what will you accept?” “I don’t want anything,” the monk replied. “Take anything at all!” [[[Buddhajñānapāda]]] beseeched, so [the monk] said, “Alright, make the offering that you will prostrate every time you see [me].” The ācārya made that promise right on the spot as his offering. Then the monk said, “Because of [your] conduct with respect to food/ And holding a slight delusion with respect to me/ You will not, in this very life, gain accomplishment/ With those embodied aggregates/ But your mind will take on the vajra 179 ‘khyor] sugg. em. based on Gshin rje chos ‘byung which follows Chögyal Phagpa’s account word for word here; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa reads ‘khyol. 180 Lit. kārṣāpāṇa, a unit of currency mentioned in Buddhajñānapāda’s autobiography in the context of the offerings received by Buddhajñānapāda and his retinue from the wealth deity Jambhala when they were residing in the Parvata cave later in his life. 181 mos 182 This is the first line of Dvitīyakrama, verse 12. 183 bzod par gsol 184 brtsams. This could also mean that he “composed” the texts mentioned. 37 body/ And you will be liberated in the bardo.”185 Then he said, “Although you practice you will not attain buddhahood in this lifetime, but benefit others and you will be liberated in the bardo.” And then he disappeared.186 All accounts of Buddhajñānapāda’s life up to the 15th century with which I am familiar— that is, Buddhajñānapāda’s own, Vaidyapāda’s, and Chögyal Phagpa’s—refer to the figure that he met in the Kuvaca forest outside of Bodhgaya simply as a “monk,” or an “emanated monk.” It is Lotsāwa in the 15th century who gives him a name: the master Mañjuśrīmitra.187 I am unsure of Lotsāwa’s source for this information, but I can (and will, below) speculate on his possible motivations for including it. This identification of the monk who was Buddhajñānapāda’s teacher as Mañjuśrīmitra persists in some of the later Tibetan biographies, but not all. Tāranātha who, as mentioned above reports having access to several Indic records that no longer survive, does not identify the monk as Mañjuśrīmitra, referring to him only with the unusual term “householder monk” (khyim btsun), presumably because in Vaidyapāda’s Sukusuma, on which Tāranātha explicitly states he relied,188 he is both identified as a monk and as accompanied by a woman.189 The Sakyapa scholar Amye Zhab, who like Tāranātha was writing in the 17th century, provides a somewhat conflicted account. In his Gsang ‘dus chos ‘byung he reports on the one hand that the Jñānapāda tradition “was bestowed upon Jñānapāda by the Ācārya Mañjuśrīmitra who was indivisible from Mañjuśrī,” 190 but when reporting the encounter with the monk in the forest he makes the strange statement that Buddhajñānapāda “directly saw the face of the emanated monk together with two gurus—Vilāsavajra and Mañjuvajra.”191 Amye Zhab then goes on to report that it was directly from the mouth of Mañjuvajra that Buddhajñānapāda received the teachings recorded in the Dvitīyakrama.192 In his Gshin rje chos ‘byung, which contains an even more extensive account of Buddhajñānapāda’s life, Amye Zhab provides an equally confusing report that in the Kuvaca forest Buddhajñānapāda “had a vision of the maṇḍala of Mañjuśrī together with *Vilāsalīla (?! ‘jol sgeg rol pa)193 185 This is a rephrasing of Mañjuśrī’s comments to Buddhajñānapāda in verse 366 of the Dvitīyakrama, with an extra half verse about the bardo that is not present in the Dvitīyakrama added on here. 186 Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 612-15. The edition of Chögyal Phagpa’s text that I have translated from here includes readings from a second manuscript that occasionally adds short phrases with additional details. I have translated these, as well, wherever they are present. 187 Deb ther sngon po, 449; Roerich 1976, 369. 188 He notes in an earlier portion of the biography that Vaidyapāda’s commentary on the Dvitīyakrama identifies Guṇamitrā as a bhikṣunī (Bka’ babs bdun, 103). 189 Tāranātha, in his Bka’ babs bdun (104-6), follows Chögyal Phagpa, or perhaps a common Indic source, in providing a version of the much more detailed account of Buddhajñānapāda’s meeting with this monk, translated above. 190 ye shes zhabs lugs ‘phags yul du dar tshul ni/ mgon po ‘jam pa’i dbyangs dang dbyer med pa’i/ slob dpon ‘jam dpal bshes gnyen gyis ye shes zhabs la gnang ba yin te/ (Gsang ‘dus chos ‘byung, 56b.3-4). 191 sprul pa’i dge slong bla ma gnyis ‘jol sgeg rdo rje ‘jam rdo dang bcas pa’i zhal mgon sum du gzigs te/. (Gsang ‘dus chos ‘byung, 56b.4.) Earlier in his text Amye Zhab uses the same uncommon moniker for Vilāsavajra. 192 ibid., 56b.4. Amye Zhab’s rather odd account of Buddhajñānapāda’s life is further complicated by his assertion that Buddhajñānapāda received teachings on the Guhyasamāja-tantra from some twenty-five masters, including every guru Amye Zhab mentions in the context of his life story apart from Haribhadra—that is Vilāsavajra, Guṇeru (who he styles Gu ni ni), Bālipada (whose name he inexplicably renders Ba mo la tsam pa ta), and Pālitapāda (who he calls Baliṃta Ācārya) (Gsang ‘dus chos ‘byung, 56b.6-57a-3). 193 In his Gsang ‘dus chos ‘byung Amye Zhab uses the unusual ‘jol sgeg rdo rje for Vilāsavajra. I am assuming ‘jol sgeg rol pa here is an error for ‘jol sgeg rdo rje. 38 together with two emanated nuns (!) in front of an ācārya’s small hut.”194 In several short references to Buddhajñānapāda Jamgön Kongtrül, closely associated with the nonsectarian (ris med) movement in 18th-century Tibet identifies Buddhajñānapāda’s teacher in two different ways, in one instance as Mañjuśrīmitra, and two others as Mañjuśrī himself.195 Dudjom Jigdral Yeshe Dorje, the great 20th-century Nyingma scholar, follows Lotsāwa in very clearly identifying the monk who is Buddhajñānapāda’s guru as Mañjuśrīmitra.196 Indeed, Dudjom’s biography of Buddhajñānapāda occurs in his encyclopedic History of the Nyingma School in the context of a list of lineage biographies of Great Perfection masters, where Buddhajñānapāda’s primary role appears specifically to be a disciple of Mañjuśrīmitra in the Great Perfection lineage, given that no attention is given to Buddhajñānapāda’s own disciples in the subsequent lineage history.197 This brief, and certainly incomplete, survey of Tibetan historians’ accounts of Buddhajñānapāda’s life suggests that (with the exception of the Sakya scholar Amye Zhab, whose accounts of this particular episode of Buddhajñānapāda’s life are anyway somewhat perplexing) it is primarily authors who are connected with traditions of the Great Perfection who identify the monk that Buddhajñānapāda met in the Kuvaca forest as Mañjuśrīmitra. Historians who are not so connected do not make this identification, and Kongtrül—who had a strong relationship to both the Nyingma and Sarma traditions—appears to have asserted both positions. This propensity of Tibetan scholars connected to the Great Perfection traditions to identify Buddhajñānapāda’s teacher as Mañjuśrīmitra is almost certainly connected to the importance for these authors of further connecting Buddhajñānapāda—two of whose compositions contain in their Tibetan translations the wordgreat perfection” (rdzogs chen)—to the Great Perfection tradition. Given that the Indian origins of this tradition have been questioned by scholars from other Tibetan traditions and that the wordgreat perfection” is not found in many other Indic sources, placing Buddhajñānapāda in the Great Perfection lineage would indeed have been compelling. While we do not have the Sanskrit for either of the two passages where the term “great perfection” appears in the Tibetan translations of Buddhajñānapāda’s writings,198 on my 194 ‘jam pa’i dbyangs ‘jol sgeg rol pa dang bcas pa gtso bor gyur pa’i dkyil ‘khor dang/ sprul pa’i dge slong ma gnyis dang bcas pa slob dpon gyi khan gpa chung du zhing tu snang bar gyur to// (Gshin rje chos ‘byung, 47a.2-3) In this account of Buddhajñānapāda’s life Amye Zhab relies heavily on both Vaidyapāda and Chögyal Phagpa, both of whose accounts he follows word-for-word at different parts of his narrative. However, with his unusual version of the account of the encounter in the forest behind Vajrāsana, he departs from both. 195 In his great Treasury of Knowledge, in the volume on Buddhism’s Journey to Tibet (2010, 229), Jamgön Kongtrül states that Buddhajñānapāda received Guhyasamāja instructions from Mañjuśrīmitra, while in his Torch of Certainty, Kongtrül states that “Buddhajñānapāda’s faithless perception caused him to see Mañjuśrī as a married monk with children,” thus appearing to identify the master as none other than Mañjuśrī himself (Kongtrül 1994, 130), and in the Treasury of Knowledge in the volume on The Elements of Tantric Practice (2009, 145) he states directly that “The Guhyasamaja completion phase in the tradition of Jnanapada (Buddhashrijnana) [stems from] what are called the Oral Teachings of Manjushri. These Teachings were directly transmitted by Arya [Mañjuśhri] himself to the master Buddhashrijnana.” 196 Dudjom 1990, 494-6. 197 The next biography given is that of Śrī Siṅgha, who is also said to be Mañjuśrīmitra’s disciple, and who Dudjom reports may have even been one and the same individual as Buddhajñānapāda (!) (Dudjom 1990, 496). The only mention of Buddhajñānapāda’s disciples is their own role as Great Perfection practitioners: “...it is implicit that the host of his followers and disciples belonged to the lineage of the Great Perfection” (Dudjom 1990, 496). 198 What is more, in the Caturaṅga/Samantabhadra-sādhana which was translated twice into Tibetan under two different titles, the word rdzog chen is found in only one of the two translations of the text, the one by Smṛtijñānakīrti. (But it also appears in one of the commentaries on the Samantabhadra, by Vaidyapāda, which was translated by a different set of translators, Kamalaguhya and Ngadak Yeshe Gyaltsen (Snga bdag ye shes rgyal mtshan)). The other translation of that same passage in the sādhana, by Śraddhākaravarman and Rinchen Zangpo (Rin chen bzang po), instead uses the word bsam yas. I address this point in further detail in Chapter Four. 39 reading, both of these instances seem to use the term in a sort of general way to refer to the result of practice, which is certainly not antithetical to, but also not precisely the way it was being used in the early (or proto-) Great Perfection literature of the late 8th and early 9th centuries (some of which is Indic, but all of which survives only in Tibetan). Vaidyapāda’s commentaries on those passages, however, do interpret the term in a way that brings it closer to—and in one case precisely in line with—contemporary 8th/early 9th-century usage of the term “great perfection.” I examine this point in more detail in Chapter Four, where I conclude that it is nonetheless unlikely that Buddhajñānapāda used a Sanskrit word with the semantic content of “great perfection” at any point in his writings. But whatever we make of the infrequent use of that particular term in the Tibetan translations of his oeuvre, we would certainly be too quick to write off as the mere apologetics of later Tibetan authors what does appear to be a genuine doctrinal affinity between Buddhajñānapāda’s writings and several early or proto- Great Perfection works from around the same period. I return to this connection below in Chapter Four. Later Life as a Teacher and Author Following his visionary encounter with Mañjuśrī, Buddhajñānapāda reports setting up residence with his students in the Parvata cave not far from Bodhgaya. The details that Vaidyapāda gives regarding the location of this residence, which he further specifies as the practice place of “great practitioners of former times,” enable us to identify it as being in the region of the Rajgir hills.199 There, Buddhajñānapāda himself tells us, he and his disciples received daily donations from the wealth deity Jambhala, and it was there that Buddhajñānapāda compiled the instructions he had received from Mañjuśrī in the Dvitīyakrama and composed other works. He likewise reports traveling back to the Koṅkan to visit his guru Pālitapāda who requested Buddhajñānapāda to compose a sādhana, which Vaidyapāda identifies as the Samantabhadra.200 As noted before, Buddhajñānapāda further mentions some “detailed accounts” of his life, which Vaidyapāda specifies to be “the taming at Nālandā, making offerings at Varjāsana, the consecration, and the others,” but gives no further information. It is only in the Tibetan histories, starting with Chögyal Phagpa, that we begin to find descriptions of these and other events from his later life. Some of the most detailed accounts of Buddhajñānapāda’s later life are found in the writings of Tāranātha. His Seven Transmissions contains a biography which references Vaidyapāda’s Sukusuma directly and elaborates on each of the three episodes mentioned by Vaidayapāda by name. Tāranātha’s account of the offerings at Vajrāsana201 is as follows: Once at a time when the ācārya [[[Buddhajñānapāda]]] had built a straw hut near Vajrāsana and was staying there, King Dharmapāla came to Vajrāsana to make offerings, and all the other Buddhist ācāryas also came to offer. Seeing that this ācārya [[[Buddhajñānapāda]]] had not made offerings, [the king] thought he ought to punish him. He entered into the ācārya’s hut, but the ācārya was not there, and [instead] he saw an image of Mañjuśrī. He came outside and asked the [[[master’s]]] retinue [where he was], and they replied that he was right in [that hut]. [The king] again entered [the hut] and [the image of Mañjuśrī] 199 See C. Dalton and Szántó, forthcoming. 200 Samantabhadra and later Chögyal Phagpa report that on that journey Buddhajñānapāda initially followed proper decorum and refused to teach in the presence of his guru and agreed only once Pālitapāda had himself given his assent (Sāramañjarī, D 2b.6-3a.2; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 615). This episode is also discussed in Szántó (2015, 548-9). 201 An earlier and slightly variant version of this account is found in Chögyal Phagpa (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 617). 40 appeared as the ācārya. When [the king] asked “Why did you not make offerings?” [[[Buddhajñānapāda]]] replied, “I made them from here.” [The king] asked “How did you offer [from here]?” upon which the ācārya entered into equipoise and all of the images from Vajrāsana appeared directly in front of the ācārya like invited guests. [The king] then saw the ācārya make vast offerings to them. At that point the king felt confidence [in him] and requested initiation. Since he didn’t have anything else to offer as the initiation fee, he offered himself and his wife as [the master’s] servants. The next day, from the palace, gold equal in weight to both of their bodies [was brought] and offered as ransom.202 This account, elsewhere recounted as involving king Devapāla rather than Dharmapāla, as discussed above, is important in linking Buddhajñānapāda with the Pāla royalty, suggesting he may indeed have been a royal guru to one (or more) of the Pāla kings. In his History of Buddhism in India (Rgya gar chos ‘byung) Tāranātha further describes a close relationship between Buddhajñānapāda and King Dharmapāla in which the master gave predictions and advice to the king, advising him to have a great homa performed regularly to ensure the longevity of his dynasty as well as of the Buddhadharma. King Dharmapāla is said to have taken this advice and maintained the regular homa ceremonies at great expense.203As described above, royal patronage of monasteries and religious masters was common in the medieval period, and the wealth and renown such patronage must have brought would have helped Buddhajñānapāda spread his teachings more widely. Such a position in connection to the king might also partly explain the great respect Buddhajñānapāda is said to have been shown by his own gurus, Vilāsavajra and Pālitapāda, who respectively cite and are said to have requested his writings. Atīśa’s brief reference to this episode (which he links to king Devapāla, rather than Dharmapāla) is important, as it is the only Indic account I am familar with that links Buddhajñānapāda to the Pāla kings.204 Tāranātha goes on to recount “the consecration,” mentioned by Vaidyapāda, and specifies that it was the consecration of the great Vikramaśīla monastery: The four temples at Vikramalaśīla,205 Odyantapūri, Śrī Nālandā, and Somapuri were cut off from one another by days of travel. Vikramalaśīla had been newly constructed, Somapuri had undergone reconstruction, and the other two [[[monasteries]]] had temples that had been newly built, so the king requested consecration for these many [[[temples]]]. The ācārya [[[Buddhajñānapāda]]] emanated four bodies simultaneously and performed the consecration of all four at once.”206 Tāranātha, in his History of Buddhism in India identifies Buddhajñānapāda as not only having consecrated Vikramaśīla but also having served as its first tantric ācārya, though to my knowledge this is not reported in any of the extant earlier histories.207 Buddhajñānapāda’s Guhyasamāja tradition was passed down by a number of masters who were likewise connected 202 Bka’ babs bdun, 106-7. My translation here differs only in minor points from Templeman’s (1983, 74). 203 Rgya gar chos ‘byung, 266-67; Chimpa and Chattopadhyaya 1970, 278-79; See also C. Dalton and Szántó forthcoming. 204 Bodhipathapradīpapañjikā, D 288b.7-289a.1; See also Szántó 2015, 539. 205 As in much of the literature in Tibetan, Vikramaśīla is referred to by Tāranātha as Vikramalaśīla. 206 Bka’ babs bdun, 107. Again my translation differs only in minor points from Templeman’s (1983, 75). Tāranātha continues with a further account of a non-Buddhist yoginī who attempts to derail the consecration at Vikramaśīla, but whose attempts are magically foiled by Buddhajñānapāda. For this account see Templeman 1983, 75. 207 Rgya gar chos ‘byung, 7, 265-66; Chimpa and Chattopadhyaya 1970, 18, 278. 41 with Vikramaśīla, including Dīpaṃkarabhadra, Vaidyapāda, Ratnākaraśānti and Abhayākaragupta. I must admit that I find it somewhat discouraging that neither Buddhajñānapāda nor Vaidyapāda makes any direct statement about Buddhajñānapāda’s having had a connection with the Pāla royalty or having acted as the vajrācārya of Vikramaśīla, both of which seem accomplishments worthy of mention in an auto-/biography. Perhaps if these events did occur they happened after the composition of the Dvitīyakrama and were thus not described there by Buddhajñānapāda, and Vaidyapāda felt compelled to constrain his elaborations in the Sukusuma to events Buddhajñānapāda had mentioned directly. The fact that at least one Indic source, Atīśa’s *Bodhipathapradīpapañjikā discussed above, recounts an encounter between Buddhajñānapāda and the Pāla king Devapāla is, however, supportive of the later accounts in the Tibetan histories. The third episode mentioned by Vaidyapāda, the “taming at Nālandā” is elaborated upon in Tāranātha’s Seven Transmissions, and the same events are likewise mentioned in his History. The episode centers around the criticism of Vajrayāna practices by śrāvaka monks. The great ācārya was acting as the head of both Nālandā and Vikramaśīla. At that time some śrāvaka Sendhapas208 who resided at Udyantapuri and some monks who were distracted by conceptuality spoke negatively about him. One time when the ācārya was residing at Nālandā those monks repeatedly criticized him, saying “Buddhajñāna lacks discipline and it is therefore not suitable for him to be a preceptor who presides over the monastic saṇgha.” They also criticized the Vajrayāna, and it is said that [also at that time] many Singhalese Sendhapas at Vajrāsana destroyed a silver image of Heruka. The king [subsequently] killed many Singhalese from Vajrāsana and as he was beginning to impose punishment on some other Sendhapas the ācārya, out of great compassion, successfully protected them from being harmed by the king. In order to overcome their lack of faith he miraculously traveled under the earth, and on several occasions many nonhuman beings made offerings to the ācārya: he displayed quite a variety of miracles. Moreover, he composed many treatises on the supreme conduct of [secret] mantra, establishing it as not being in contradiction to the śrāvaka-piṭaka.209 The reference in this episode to Buddhajñānapāda’s lack of discipline presumably refers to his having a consort, which Buddhajñānapāda and Vaidyapāda both mention being the case at several points earlier in his life. As regards the general antipathy toward the Vajrayāna and Buddhajñānapāda’s efforts to overcome this, we certainly see in his writings an effort to present the tantric teachings within the context of the general Mahāyāna and to validate Vajrayāna doctrine and practices, though in his extant work there is nothing that particularly addresses the śrāvaka teachings.210 In his History Tāranātha further mentions that Buddhajñānapāda principally taught five among the “inner tantras:” the Guhyasamāja-tantra, the Māyājāla-tantra, the Sarvabuddhasamāyoga-tantra, the Guhyendratilaka-tantra, and the Yamāri-tantra,211 with 208 Sendha pa. The referent of this term is unclear, but from the way that Tāranātha uses it, it appears to refer to a particular group or sect of śrāvaka monks, primarily from Ceylon. See Templeman 1983, 143n144 for a further discussion of their possible identity. 209 Bka’ babs bdun, 108. Again, my translation differs only in minor points from Templeman’s (1983, 75-6). 210 The possible exception to this are his and Vaidyapāda’s use of the unusual terms thal byung blo can, thal byung zab mo and thal byung gyi stong pa nyid, all of which seem to refer to a śrāvaka-like fixation on the quiescent aspect of meditative equipoise, a position that Buddhajñānapāda rejects. See note 148 in Chapter Three. 211 The latter is rendered ‘jam dpal ‘kros pa, but presumably refers to the Krṣṇayamāri-tantra. Rgya gar chos ‘byung, 267; Chimpa and Chattopadhyaya 1970, 279. 42 the most emphasis on the Guhyasamāja. Though his surviving tantric writings are primarily Guhyasamāja-based, other Tibetan scholars including Amye Zhab include him as an important figure in the Kṛṣṇayamāri lineage.212 I will discuss Buddhajñānapāda’s surviving oeuvre in more detail below, but let us first turn our attention to his principal disciples. Disciples The earliest reference we have to Buddhajñānapāda’s disciples is in Vaidyapāda’s Sukusuma, which states that Buddhajñānapāda had eighteen disciples who functioned as his regents, among whom there were four who attained nirvāṇa in this lifetime: Dīpaṃkarabhadra (Mar me mdzad bzang po), Praśāntamitra (Rab tu zhi ba’i bshes gnyen), *Rahulabhadra (sgra gcan ‘dzin bzang po), and *Vajramahāsukha (Rdo rje bde ba chen po).213 These same four principal disciples are mentioned in Śrīphalavajra’s commentary on the Caturaṅga/Samantabhadra-sādhana.214 Dīpaṃkarabhara Of the four, we know the most about Dīpaṃkarabhadra, who composed the well-known and influential Guhyasamājamaṇḍalavidhi, an initiatory ritual according to the Jñānapāda School that continued to influence later ritual manuals such as Abhayākaragupta’s Vajrāvalī and 212 Gshin rje chos ‘byung, 16a.4. As regards the Yamāri tantra connection, in one lineage description Amye Zhab explains that Buddhajñānapāda’s Kṛṣṇayamāri lineage originated with Mañjuśrī and was passed from Buddhajñānapāda to the siddha Śrīdhara (Dpal ‘dzin), who passed it on to Nāropa (Gshin rje chos ‘byung, 16a.4). Later in the same work, at the conclusion of a biography of Buddhajñānapāda Amye Zhab writes that Buddhajñānapāda received from (emending la to las to accord with the verb gsan) Mañjuśrī teachings on the cycles of the Guhyasamāja as well as Vajrabhairava. He continues, explaining that Buddhajñānapāda passed the Vajrabhairava cycle on to Dīpaṃkarabhadra, who gave it to the siddha Śrīdhara, who composed many Yamārirelated works, which he then passed on to Nāropa (Gshin rje chos ‘byung, 48a.4-5). (Dīpaṃkarabhadra is credited with at least one Yamāri-tantra related text in the Tibetan canon, a work on the protection circle according to that tradition (Tōh. 1928), and many Yamāri-related compositions attributed to Śrīdhara survive in the canon.) Jamgön Kongtrül also briefly mentions Buddhajñānapāda’s view on the perfection stage yogas of the Yamāri-tantra as being consistent with that of Saraha (Kongtrül 200, 149). Buddhajñānapāda’s connection with the Yamāri-tantra tradition has been almost entirely ignored in the secondary literature (at least in English; there may exist references in Japanese of which I am unaware), with the exception of an unpublished article by Hubert Decleer that notes Buddhajñānapāda’s importance for the Sakyapas of the Ngor tradition as being the source of both their Guhyasamāja lineage and their Kṛṣṇayamāri lineage. In that article Decleer identifies an image from a Sakyapa thanka as depicting scenes from Buddhajñānapāda’s life (Decleer, unpublished). I am grateful to Hubert Decleer for sharing this unpublished work with me. Three Yamāri-tantra related compositions (Tōh. 2084, 2085, and 2086) are attributed to Buddhajñānapāda in the Tibetan canon, but I believe it is not likely that these are by the same Buddhajñānapāda who composed the many Guhyasamāja works that we know. I assess the attribution of these three texts to Buddhajñānapāda below in the section on his writings. To my knowledge there is not any other mention of Buddhajñānapāda’s having taught the other three tantras mentioned by Tāranātha—the Māyājāla, Sarvabuddhasamāyoga, or Guhyendratilaka—but he certainly knew the Sarvabuddhasamāyoga as several lines of his Dvitīyakrama (v. 50; see also notes 123 and 276 of my Dvitīyakrama translation) are a paraphrasis of verses in that tantra, and his disciple Praśāntamitra probably wrote a commentary on it, as well as one on the Māyājāla-tantra (which I discuss below). And Buddhajñānapāda’s guru Vilāsavajra taught parts of the Mañjuśrīnāmasaṃgīti in accordance with the Māyājāla (See Tribe 1994, 24). 213 Sukusuma, 135a.5-6. 214 Samantabhadrasādhana-vṛtti, D141a.6. The order of the list of the four main disciples is different here from Vaidyapāda’s Sukusuma, perhaps indicating that one list is not merely derivative of the other. Śrīphalavajra here also mentions three, but then names only two, of Buddhajñānapāda’s co-disciples: *Dharmākara (Chos kyi ‘byung gnas) of the Koṅkan and *Uṣṇīṣavajra (Gtsug tor rdo rje) of Mount Hasara (ri bo ha sa ra) (ibid., D 141a.5-6). The list of four principal disciples is repeated in several of the Tibetan histories, as well, and Tāranātha cites Śrīphalavajra’s statement about Buddhajñānapāda’s co-disciples (Bka’ babs bdun, 109; Templeman 1983. 76). 43 Jagaddarpaṇa’s Kriyāsamuccaya.215 A very substantial portion of the Guhyasamājamaṇḍalavidhi is a direct paraphrasis of Buddhajñanapāda’s Caturaṅga/Samantabhadra-sādhana and a good ninety percent of the Caturaṅga/Samantabhadra’s verses appear rephrased therein. Tāranātha reports that Dīpaṃkarabhadra succeded his guru Buddhajñānapāda as the second tantric ācārya at Vikramaśīla,216 and that he was said to have achieved even higher realization than his master.217 Tāranātha also identifies Vaidyapāda as Dīpaṃkarabhadra’s disciple.218 Praśāntamitra Buddhajñānapāda’s disciple Praśāntamitra is likely the author of a commentary on the Sarvabuddhasamāyoga-tantra, one on the Māyājāla-tantra, and a third on the Vajramaṇḍālaṃkāra.219 Tāranātha reports that Praśāntamitra received initiation from Buddhajñānapāda and had a meditative vision of Yamāri,220 and places him as the first disciple after Buddhajñānapāda in Jñānapāda’s lineage of the “word” instruction.221 He also identifies Vaidyapāda as a disciple of Praśāntamitra’s.222 About the other two among the four principal disciples mentioned in Vaidyapāda’s and Śrīphalavajra’s accounts, *Rahulabhadra and

  • Vajramahāsukha, we unfortunately can say nothing more.

Buddhaguhya and Buddhaśānti Although I am not aware of any such reference in an extant Indic text, Lotsāwa and Tāranātha mention Buddhaguhya and Buddhaśānti as disciples of Buddhajñānapāda’s, with Tāranātha specifying that this discipleship took place early in Buddhajñānapāda’s life.223 Buddhaguhya is a well known early 8th-century commentator on a number of Kriyā and Yoga tantras, including an important commentary on the Mahāvairocanābhisambodhi. A number of works on the Guhyagarbha-tantra are also attributed to him, though it is unclear whether it is the same author who wrote the Kriyā and Yoga tantra works and those on the Guhyagarbha.224 215 Szántó 2015, 554. This work survives in Sanskrit and has been edited. It was first circulated in an e-text (Klein- Schwind and Isaacson), edited in Dhīḥ (2006), and later by Bahulkar (2010). Szántó 2015, 556n34 gives a diplomatic transcript of verses that were missing the Dhīḥ edition. Daisy Cheung at Hamburg University is working on this text for her dissertation. Quite a number of other works in the Tengyur are ascribed to Dīpaṃkarabhadra (Mar me mdzad bzang po), including a short sādhana for the protection circle in the Yamāri-tantra tradition (Tōh. 1928), but I have not had the opportunity to look at these and assess their content or authorship. 216 Rgya gar chos ‘byung, 7. 217 Bka’ babs bdun, 110. Tāranātha gives a short biographical sketch of Dīpaṃkarabhadra (Bka’ babs bdun, 109- 112; Templeman 1983, 76-8). 218 Bka’ babs bdun, 112; Templeman 1983, 78. 219 Szántó 2015, 547n22. The commentaries are Tōh. 1663, 2514, and 2515, respectively. I agree with Szántó here on the likelihood that these commentaries are indeed composed by the Praśāntamitra who was Buddhajñānapāda’s disciple, especially given his note that this author includes a line in the dedicatory verses at the end of each text identifying himself as the author, a stylistic feature of many of Buddhajñānapāda’s own writings and those of his other disciples. 220 Rgya gar chos ‘byung, 267-68; Chimpa and Chattopadhyaya 1970, 279-80. 221 tshig gi brgyud pa’i brgyud pa’i bka’ babs. Bka’ babs bdun, 117; Templeman 1983, 66. It is unclear from Tāranātha’s description precisely what this “word lineage” is, but it is perhaps worth noting that the fourth initiation is often referred to as the “precious word initiation” and Vaidyapāda’s Yogasapta shows evidence of the early development of this initiation in Buddhajñānapāda’s tradition. 222 ibid. 223 Deb ther sngon po, Vol I, 451; Roerich 1976, 372; Rgya gar chos ‘byung, 269; Chimpa and Chattopadhyaya 1970, 280. Dudjom (1990, 465-6) repeats Tāranātha’s assertion. 224 Hodge 2003, 23 and 1995, 69. Scholarly opinion on whether the Yoga tantra and Mahāyoga tantra author were one and the same or not appears to be divided between those who assert “maybe no” and those who assert “maybe yes,” but there seems to be no certainty on the topic. Hodge (1995, 69) notes that Buddhaguhya does not mention the Guhyagarbha in any of his other tantric works and that the two corpi seem to him different on stylistic grounds. 44 Buddhaguhya is reported in a number of Nyingma sources to have been a student of Vilāsavajra’s, from whom he received teachings on the Māyājāla-tantra and especially the Guhyagarbha-tantra.225 He was also invited to Tibet by King Trisong Deutsen, but declined the invitation on the advice of his tutelary deity Mañjuśrī. 226 Some modern scholars have placed doubts on Buddhaguhya’s having been a disciple of Buddhajñānapāda’s, given that he appears to have been, if anything, Buddhajñānapāda’s senior contemporary rather than his junior, but the question remains unresolved.227 Of Buddhaśānti we know only of his association with Buddhaguhya as described in the Tibetan histories.228 Vaidyapāda There is considerable question about whether or not Vaidyapāda, Buddhajñānapāda’s most prolific Indian commentator, was his direct disciple. However, like with Buddhajñānapāda, the first issue we must address is that of his name. In almost all of the secondary literature mentioning this master, he is referred to as Vitapāda.229 This is no doubt due to the fact that in the preponderance of colophons of the Tibetan translations of his work this author’s name is given as Vitapāda.230 However, we also find his name in other colophons as Vitapāta,231 Vaidyapāda,232 Vidyapāda,233 Viryapata,234 and even Hahitapāda (!).235 When this name is Guarisco (Kongtrül 2005, 75n12) likewise notes that while traditional Nyingma scholarship identifies the author of these different works as a single figure, it is more likely that the author of the Guhyagarbha treatises is a different individual. Takahashi (2009, 198) however, opines that the claim of Nyingma authors that the author of the Yoga tantra and Mahāyoga tantra works are one and the same Buddhaguhya is worth taking seriously, especially given that both corpi seem to have been composed around the same period. She suggests that perhaps Buddhaguhya’s involvement with the Guhyagarbha teachings was less publicized by the author himself given their controversial nature, which would explain the absence of reference to the Guhyagarbha-related teachings in his other works. Van Schaik (2004, 187) simply notes that they “may be” one and the same. Germano (2002, 229-232) reports from Nyingmapa sources in which it is assumed to be one and the same Buddhaguhya who wrote both corpi, but does not express his own opinion one way or another. 225 Hodge 1995, 68-9. 226 ibid. 227 Hodge (2003, 22 and 2012, 68-9) and Weinberger (2003, 83) place doubts on the discipleship. 228 Rgya gar chos ‘byung, 269-71; Chimpa and Chattopadhyaya 1970, 280-83. Tāranātha relates several accounts in which Buddhaśānti appears as a companion of Buddhaguhya. Contrary to the letter in which Buddhaguhya declines to travel to Tibet, the two yogins are reported as companions traveling near Mount Kailash during the time of King Trison Deutsen in the Sba bzhed, where Buddhaguhya’s name is reported as Buddhagupta (Kapstein 2000, 26). The two names Buddhaguhya and Buddhagupta seem to refer to the same person (See Hodge 2003, 70; Weinberger 2003, 84). Germano cites Nyingmapa accounts stating that the exchange between Buddhaguhya and Trisong Deutsen took place precisely during Buddhaguhya’s journey to Kailash, and that he declined to visit central Tibet (Germano 2002, 229). 229 However, at least two modern scholars have addressed the issue of his name, one of whom has concluded that it is probably better rendered as Vaidyapāda. Szántó (2015, 540n6) reports a presentation given by Leonard van der Kuijp at Oxford in 2008 in which he suggested as much. I am unfamiliar with the details of van der Kuijp’s assessment, which has not been published. Kikuya (2012a, 1276n3) likewise reports that he himself has written about “the problems of Vitapāda’s transmission and his name,” but as the article referenced is in Japanese I have not been able to check it. 230 Bi ta pā da in the colophons of his Sukusuma (Tōh. 1866) (in D and P), Samantabhadrī-ṭīkā (Tōh. 1872) (P), Guhyasamājamaṇḍalopāyikā-ṭīkā (Tōh. 1873) (in D and P), Siddhisaṃbhavanidhi (Tōh. 1874) (in D and P), Yogasapta (Tōh. 1875) (in D and P), Mahābalividhi (Tōh. 1876) (in D and P), Ratnamati (Tōh. 1877) (in D and P), and Ātmārthasiddhikara (Tōh. 1878) (in D and P). 231 Bi ta pā ta in the colophon of his Samantabhadrī-ṭīkā (Tōh. 1872) (D). 232 Bai dya pā da, in the colophon to his Samyagvidyākara (Tōh. 1850) in the Derge Tengyur; the Peking here reads Bi dya pā da. 233 Bi dya pā da in the colophon to his Samyagvidyākara (Tōh. 1850) in the Peking Tengyur. 45 translated into Tibetan—in the Tibetan translations of Indic works and in indigenous Tibetan writings alike—it is consistently rendered as Sman pa’i zhabs, which fact already lends support to the Sanskrit Vaidyapāda, and perhaps even suggests that he may have been a physician (Skt. vaidya, Tib. sman pa).236 Lotsāwa gives his name in translation as Sman pa’i zhabs, but in his Seven Transmissions Tāranātha—who mentions a number of Indic sources on which his writings relied—gives the name both in transliteration as Vaidyapāda (Bai dya pā da) and in translation as Sman pa zhabs.237 In the dedicatory verses of his Yogasapta (in the colophon of which his name, incidentally, is given as Vitapāda), this master writes [I] Vaidyapāda (Sman pa’i zhabs) have received This supreme nectar of the seven yogas Accomplished through practice in the presence of the gurus Of the ocean of the Glorious Samāja. Having drunk this nectar May the fatal illness of Mistaken conceptuality Be completely dispelled! Freed from that [[[Wikipedia:illness|illness]]] may all beings Perfectly unfold the genuine aggregates And attain the suchness that is the result: The supreme nature of the seven yogas!238 It seems to me that in these verses Vaidyapāda is loosely playing on his name (“Mr. Doctor”) in reference to conceptuality as a fatal illness that is healed by the nectar of the seven yogas. Of course, the use of a medical metaphor is not unique—medical metaphors have been used in Buddhist texts from the very earliest literature. My sense, though, is that Vaidyapāda is using it here as a way of integrating his name more smoothly into the dedicatory verse. Vaidyapāda frequently emulates Buddhajñānapāda’s writing,239 and as we will see below, Buddhajñānapāda himself often wove his own name cleverly into the dedicatory verses of his writings. For a master whose name was “Mr. Buddha Wisdom,” that was, however, an easier task than for his disciple “Mr. Doctor.” It seems to me that here in the Yogasapta, Vaidyapāda found his chance.240 Of course, the “Vitapāda” in so many of our Tibetan colophons could certainly be a 234 Martin (2011, 2078) notes that the Black Hat Tanjur catalogue his Yogasapta reports the author of the Yogasapta as Birya pa ta. 235 Ha hi ta pā da in the colophon of the Derge edition of his Multitilaka-vyākhyāna (Tōh. 1870). The Peking reads Bi ta pā da. 236 Another possible understanding of the term, one not taken up by the Tibetan translators, is vaidya as in someone well versed in the Vedas. Though none of our extant Indic sources report anything about Vaidyapāda’s life, Tāranātha states that he was born a brahmin and was not only learned in the non-Buddhist doctrines but had become powerful due to them, presumably through practice, before becoming a Buddhist (Bka’ babs bdun, 112; Templeman 78-9) 237 Neither Chögyal Phagpa nor Amye Zhab mention this master. 238 dpal ldan ‘dus pa’i rgya mtsho las// bla ma’i zhal snga (snga] P, sngas D) bsgrub pas na// sbyor ba bdun gyi bdud rtsi mchog// sman pa’i zhabs kyis thob pa’o// de ‘dra’i bdud rtsi de la ni// ‘thung bye de ni log pa yi (yi] P, yin D)// rtog pa yis ni rab ‘chi (‘chi] P, ‘cing D) ba’i// nad ni kun nas med gyur cig// de med pas na sems can kun// yang dag phung po rgyas ‘gyur te// sbyor ba bdun gyi rang bzhin mchog// ‘bras bu de nyid rtogs par shog// (Yogasapta, D 75b.1-3; P 89b.6-8) 239 See, for example, the opening verse dedicated to the buddha, dharma, saṇgha, and gurus in his Siddhisaṃbhavanidhi, which loosely emulates Buddhajñānapāda’s own opening verse, likewise dedicated to the three jewels and gurus, in the Dvitīyakrama. In his compositions Vaidyapāda likewise replicates much of the terminology that is especially characteristic of Buddhajñānapāda’s writings. 240 Vaidyapāda inscribes his own name into only one other of his ten extant compositions, his Siddhisaṃbhavanidhi, but in one of the introductory verses, rather than the conclusion. This verse actually has some parallels with the 46 vernacular rendering of the name Vaidyapāda.241 And though there is no single piece of definitive evidence, and thus the question still remains open, I feel there is sufficient reason to depart from the more common usage in the secondary literature and refer to this master as Vaidyapāda. To return to the question of whether or not Vaidyapāda was a direct disciple of Buddhajñānapāda’s, again there is no definitive piece of evidence one way or the other, but I believe it is likely that he was.242 Three of his ten surviving works are commentaries on Buddhajñānapāda’s compositions, and he directly mentions Buddhajñānapāda or his tradition in five of the remaining seven. In these references he makes statements such as having composed treatises, “in order to remember the stages of the pith instructions of my gurus who uphold the lineage, Buddhajñānapāda and so forth...,”243 “in order to remind myself and others of the stages of the pith intructions of my gurus who uphold the lineage, Buddhajñānapāda and so forth...,”244 and others which contain “the complete teachings that have come down in stages from Buddhajñāna,”245 or which are explained “in terms of the suchness of the instructions that have come down in stages from our great guru Buddhajñāna.”246 In one dedicatory verse he notes, “Since the appearance of the kindness of the unsurpassed great compassion of Śrībuddhajñanapāda has been veiled, although I am a fool, I have uncovered it slightly.”247 Admittedly none of these statements settles the case, but one does get the sense that Buddhajñānapāda was one among Vaidyapāda’s gurus with whom he had some direct personal connection, though certainly not his only master. This is upheld by Lotsāwa248 and Tāranātha, who both clearly write that Vaidyapāda was Buddhajñānapāda’s direct disciple. But the way that Tāranātha states this is more telling: he lists Vaidyapāda in the lineage as a student of Dīpaṃkarabhadra, who he has already reported as Buddhajñānapāda’s disciple, but adds that Vaidyapāda, “trained in the tantras after having heard [teachings], beginning with the Prajñāpāramitā and continuing up to the outer and inner [[[tantras]] of the] secret mantra, at Nālandā in Madhya from both Dīpaṃkarabhadra and concluding verse from the Yogasapta, but here Vaidyapāda was working with a maritime, rather than a medical, metaphor: “From the great ocean of the Glorious Samāja/ By means of the ship of the guruslineage/ [I] Vaidyapāda (Sman pa’i zhabs) have obtained/ The three wish-fulfilling gems! dpal ldan gsang ‘dus mtsho chen las/ bla ma’i brgyud rim gru gzings kyis/ yid bzhin nor bu rnam pa gsum/ sman pa’i zhabs kyis rnyed pa’o// (Siddhasaṃbhavanidhi, D 2a.1; P 2b.1-2).) 241 Could it have been written Vaidyapāda, but colloquially pronounced Vitapāda? 242 Modern scholars have asserted both positions. Tomabechi (2008, 172-3) and Klein-Schwind (2012. 17) hold Vaidyapāda to be Buddhajñānapāda’s direct disciple, while J. Dalton (2004, 17) and Kikuya (2012, 1264) assert that he was a later commentator, and Davidson (2002, 311) also seems to suggest that he was not a direct disciple. Szántó (2015, 547) hedges his bets, reporting only that Vaidyapāda was “supposedly Jñānapāda’s direct student.” In a recent article written together with Péter Szántó, (C. Dalton and Szántó, forthcoming) we suggest that Vaidyapāda was probably not a direct student, a position I have since come to revise. 243 rgyud don ‘dzin bdag bla ma la’ang/ sangs rgyas ye shes zhabs la sogs pa’i rim pa’i man ngag rang nyid kyis dran phyir/ (Ātmārthasiddhikara, D 84b.3-4; P 100b.2) 244 rgyud don ‘dzin bdag bla ma la’ang/ sangs rgyas ye shes zhabs sogs rim pa’i man ngag rang dang gzhan gyis dran pa’i phyir/ (Samyagvidyākara, D 180a.5”) 245 sangs rgyas ye shes zhal snga’i rim ‘ongs pa/ ma lus bstan pa (Mahābalividhi, D 75b.5-6; P 90a.3-4). 246 bdag cag gi bla ma chen po sangs rgyas ye shes kyi zhal snga nas kyi rim pa nas ‘ongs pa’i rlung gi de nyid gyi sgo nas... (Ratnamati, D 81a.5; P 96a.8). 247 This verse admittedly has some problems, and its transmission may be corrupt. dpal ldan ye shes zhabs ni bla na med pa thugs rje chen po yis// drin gyi (gyi] P, gyis D) snang ba (ba] sugg. em., bas D P) khebs phyir bdag ni blun yang cung zad tsam du phye// (Ātmārthasiddhikara, D 94a.7; P112a.5) 248 Deb ther sngon po, vol 1, 451; Roerich 1976, 271. 47 Dīpaṃkarabhadra’s guru Śrījñānapāda.”249 Given that Vaidyapāda wrote a commentary on Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, we know that he was certainly junior to this master.250 Moreover, in Vaidyapāda’s commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana he gives a citation from Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi and identifies the author there simply as “the (my?) guru.”251 Tāranātha, elsewhere in the Seven Transmissions, lists Vaidyapāda as a disciple of Praśāntamitra, who was also a direct disciple of Buddhajñānapāda’s.252 Given all of these considerations, I believe it to be likely that Vaidyapāda was indeed a direct student of Buddhajñānapāda’s, but that he met him when the latter was already quite aged, while Vaidyapāda was likely still quite young. Thus several among Buddhajñānapāda’s disciples, including Dīpaṃkarabhadra and possibly Praśāntamitra, also count among Vaidyapāda’s gurus. This would place Vaidyapāda’s life squarely in the early to mid 9th century, a date which is further supported by the works he does and does not cite in his oeuvre, and the content thereof.253 Unfortunately we know nothing of Vaidyapāda’s life from Indic sources. Turning to the later Tibetan histories, Tāranātha gives a short biographical sketch in his Seven Transmissions in which he notes that Vaidyapāda was born a brahmin in a border region and was learned and accomplished in non-Buddhist practices before gaining faith in the Buddhist teachings. He is also said there to have been a practitioner of the wrathful deity Hūṃkāra, who receives mention in several of Buddhajñānapāda’s short tantric writings.254 Tāranātha goes on to note that it is 249 yul dbus nālandār slob dpon mar me mdzad bzang po dang/ mar me mdzad bzang po’i yang bla ma dpal sangs rgyas ye shes gnyis la pha rol tu phyin pa nas brtsam te/ gsang sngag phyi nang gi bar gyi thos pas rgyud sbyang/ (Bka’ babs bdun, 112; Templeman 1983, 79). 250 The text is his Guhyasamājamaṇḍalopāyikā-ṭīkā. 251 bla ma’i zhal snga nas (Samantabhadra-ṭīkā, D134b.7). Samantabhadra, another commentator on the Caturaṅga who seems to have been reading Vaidyapāda’s commentary (I write more about the relationship between their commentaries in Chapter Five), cites the very same passage from Dīpaṃkarabhadra at this point in his Caturaṅga commentary, but Samantabhadra identifies the author of the citation as Bhadrapāda, a commonly used name for Dīpaṃkarabhadra (Sāramañjarī, D5b.4). This may be an indication that this master was not Samantabhadra’s personal guru, whereas Vaidyapāda may indeed have had such a personal relationship with Dīpaṃkarabhadra. 252 Bka’ babs bdun, 117; Templeman 1983, 66. 253 Vaidyapāda knows several texts that Buddhajñānapāda does not, most crucially the Samājottara, the so-called eighteenth chapter of the Guhyasamāja-tantra, and Śākyamitra’s Anuttarasandhi, which was incorporated into Nāgārjuna’s Pañcakrama as its second stage, the sarvaśuddhiviśuddhikrama. I discuss the possibility of Śākyamitra's discipleship with Buddhajñānapāda below, and Buddhajñānapāda’s and Vaidyapāda’s relationship with the Samājottara in Chapter Eight. 254 Tāranātha’s short (and charming!) biographical sketch reads: “The master Vaidyapāda was born into the brahmin caste in a border region. He became learned in the tīrthika doctrines and gained power through their practice, but later he gave rise to faith in the Buddha’s teachings. He trained in the tantras after having heard [teachings] beginning with the Prajñāpāramitā and continuing up to the outer and inner [[[tantras]] of the] secret mantra at Nālandā in Madhya from both Dīpaṃkarabhadra and Dīpaṃkarabhadra’s guru Śrījñānapāda, as well. [They] bestowed initiation on him and gave him the complete instructions. In particular, during initiation into the maṇḍalas of the Samāja and Heruka, his flower fell on wrathful Hūṃkāra. After meditating for a long time he gave rise to a unique samādhi of the two stages. He knew that after practicing for six months he would attain siddhi, but that he needed to rely as a practice support on a vajra-family consort, a doṃbi girl who was blue in color like an utpala flower, so he searched for her in all directions and [finally] found her. When requesting her from her parents they said, “Are you crazy, brahmin ācārya!? Since we are of the doṃbi caste, won’t [this] bring punishment upon both of us?” He answered, “Since I need a practice support, the ordinary castigations of [my association with] lower castes and so forth will not apply.” They said, “Well, then, we need gold and silver equal in weight to the girl’s body,” upon which the ācārya immediately brought forth a treasure from below the earth and gave it to them. Then the ācārya together with his consort practiced in a cave for six months upon which on the eighth day of the waxing moon at dawn a great sound hūṃ resounded from the sky and [he] directly saw all the maṇḍalas of Śrīheruka and so forth. [He] also attained the state of the siddhi of supreme mahāmudrā. He benefited many beings, primarily by means of the path of the Samāja, composed many treatises, and finally, in that very body, set off for Buddha Akṣobhya’s pure 48 sometimes reported that Vaidyapāda is the same individual as the master Hūṃkāra renowned in the Nyingma School.255 There are ten compositions attributed to Vaidyapāda (under a variety of names, see above) that survive in Tibetan in the Tengyur.256 While I have read only five of these in full, I have surveyed all of them and feel confident that on grounds of both style (some points related to which have also been noted above) and content, all of these works can be attributed to a single author whose main interest was clearly the elucidation and propagation of Buddhajñānapāda’s Guhyasamāja tradition.257 I will address the contents of several of his works in the succeeding chapters. Śākyamitra The final disciple of Buddhajñānapāda’s we will consider here is Śākyamitra. He receives no mention in the Tibetan histories in connection with Buddhajñānapāda, but rather is identified in several sources as a disciple of (the tantric) Nāgārjuna.258 Yet one Śākyamitra is the author of the Mukhāgama (Tōh. 1854, not to be confused with Buddhajñānapāda’s Dvitīyakramatattvabhāvanā-mukhāgama (Tōh. 1853), which is often referred to as the Mukhāgama in the secondary literature!), which he claims in both its opening and closing verses to be a record of the oral instructions of Buddhajñānapāda. This text is identified by both traditional and modern scholars alike as Buddhajñānapāda’s composition, presumably because the colophon reports it to be the “Oral Instructions (mukhāgama) on the sādhana of Buddhajñānapāda, [the master] who came from Glorious Uḍḍiyāna.”259 But Śākyamitra clearly states within the text itself that it was he who composed the treatise,260 and in a passage that is admittedly difficult to interpret, he seems to claim to have met Buddhajñānapāda and received land, taking off in flight like the king of Garuḍas” (Bka’ babs bdun, 112-12). My translation here differs only slightly from that of Templeman (1983, 78-9). 255 Bka’ babs bdun, 112; Templeman 1983, 79. If this is indeed the case, Tāranātha continues, it means he was born in Nepal and visited Tibet in the time of king Senalek. 256 The ten are the Samyagvidyākara (Tōh. 1850), Sukusuma (Tōh. 1866), Muktitilaka-vyākhyāna (Tōh. 1870), Samantabhadrī-ṭīkā (Tōh. 1872), Guhyasamājamaṇḍalopāyikā-ṭīkā (Tōh. 1873), Siddhisaṃbhavanidhi (Tōh. 1874), Yogasapta (Tōh. 1875), Mahābalividhi (Tōh. 1876), Ratnāmati (Tōh. 1877), and Ātmārthasiddhikara (Tōh. 1878). 257 A glance at the full titles of Vaidyapāda’s works listed in the previous note may appear alarming to some, specifically given the presence of the Yogasapta-nāma-caturabhiṣekaprakaraṇa, The Seven Yogas: An Explanation of the Four Initiations. Certainly this is unexpected, given that the early Jñānapāda tradition as found in Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, and even up until the 11th-century commentary on that text by Ratnākaraśānti, is well known in modern scholarship to preserve a tradition of just three initiations, rather than four (See Isaacson 2010a, 269; Wedemeyer 2014, unpublished, and Wedemeyer forthcoming). Vaidyapāda’s position in this treatise, and indeed its entire content, are worthy of serious further study, but for now it will suffice to say that I do feel confident that this text was authored by him. I address the Yogasapta in some detail in Chapter Seven. 258 Lotsāwa holds Śākyamitra to be one of the four main disciples of Nāgārjuna (Roerich 1976, 359-60) and Tāranātha mentions that “Ācārya Śākyamitra the great was certainly a disciple of Ācārya Nāgārjuna, though [I] have not seen or heard his story.” slob dpon śākya bshes gnyen chen po yang slob dpon klu grub kyi slob ma yin par nges mod kyi lo rgyus ma mthong zhing ma thos so// (Rgya gar chos ‘byung, 114; Chimpa and Chattopadhyaya, 128). 259 Mukhāgama, D 28b.6. 260 Unlike Tomabechi (2008, 174), who writes that Śākyamitra is the “compiler of the Mukhāgama,” which he attributes to Buddhajñānapāda, I believe it is clear in the pledge to compose at the beginning of the text Śākyamitra holds himself to be the author, rather than simply having compiled Buddhajñānapāda’s oral instructions. In the admittedly difficult verses at the beginning of the Mukhāgama, Śākyamitra writes that he will, in this text, explain with clear words and without rhetoric or philosophy the profound meaning of the sādhana of Buddhajñānapāda (Mukhāgama, D 27b.3-6). This claim by Śākyamitra, which seems to amount to a pledge to explain Buddhajñānapāda’s complicated teachings in more common language, is very much in support of Tomabechi’s argument in his 2008 article that Āryadeva in the Caryāmelāpakapradīpa is referring to the Jñānapāda School authors with his comments on the unsuitability of the rhetorical complexity of their compositions. 49 his oral instructions, and even to be his primary disciple.261 In any case, there is no doubt that the contents of the Mukhāgama derive from Buddhajñānapāda’s teachings, as this text, like Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, contains a summary and rephrasing of a good portion of the Caturaṅga/Samantabhadra-sādhana. Śākyamitra, however, departs from Buddhajñānapāda’s work significantly more than did Dīpaṃkarabhadra, in particular adding an extensive instruction on the protection circle that is absent in the Caturaṅga/Samantabhadra.262 Other compositions attributed to Śākyamitra include the Koṣalālaṃkāra, a commentary on the Sarvatathāgatatattvasaṃgraha and the Anuttarasandhi, which was integrated into Nāgārjuna’s Pañcakrama as its second stage.263 Both of these works seem to show the influence of Buddhajñānapāda’s writings. The Kośalāṃkāra contains an autobiographical section extremely reminiscient of Buddhajñānapāda’s in the Dvitīyakrama describing Śākyamitra’s travels and studies under gurus, including to the Koṅkan and Uḍḍiyāna, places where Buddhajñānapāda also reports traveling.264 This record makes no mention of Buddhajñānapāda,265 but several of the lines in the autobiographical section sound so strikingly similar to those in the Dvitīyakrama, that it seems likely that one account has inspired the other. If the author of the Mukhāgama and the Kośalālamkāra are indeed the same individual, it would appear that in his autobiographical account Śākyamitra was emulating his guru, 261 As Tomabechi (2008, 174) notes, the passage in question “presents some difficulty in interpretation,” to say the least. (In fact, unfortunately much of the Mukhāgama presents difficulty in interpretation, and my sense is that either the original manuscript that the translator was working with had problems or a number of difficulties arose in its translation. The colophon states that the Mukhāgama was translated by Rinchen Zangpo alone, without mention of the assistance of an Indian paṇḍita, which may have been part of the problem.) But I nonetheless agree with Tomabechi that these verses in question do suggest that the compiler of the Mukhāgama claims to have met Buddhajñānapāda and received instruction directly from him. Unlike Tomabechi, I will (perhaps unwisely!) hazard a tentative translation of the passage in question (I include here also the two verses that precede those cited by Tomabechi (2008)), “These instructions that I have composed/ I received with great faith from the lotus of my guru’s mouth/ Let scholars treat it as a spectacle [if they wish],/ [From my side] I wrote it in order to benefit all beings./ Just like [even] someone who has achieved something through lies/ Is [still] praised a bit by his [[[own]]] father/ Just like that, I have found a bit of merit—/ Through it may all the world become [like] Mañjuvajra!/ [He who was] born in glorious Uḍḍiyāna,/ Knower of the meaning of the countless tantras without exception,/ The glorious one, completely pacified and peaceful, spoke these words./ The supreme main disciple upon whom he bestowed initiation,/ Who was born [in a place located] in the direction of the Sindh from there [i.e. from Uḍḍiyāna],/ Who victoriously resided in the place called the abode of Vaiśravana,/ I, Śākyamitra, myself, awakened and/ [Recorded] his (i.e. Buddhajñānapāda’s) perfectly liberating oral instructions.” bdag gis man ngag brtsams ‘di rab dad pas// bla ma’i zhal gyi padma las rnyed de// mkhas pa rnams kyis ‘di la ltad mo (ltad mo] D, brtag mod P) gyis// ‘gro ba kun la phan phyir bdag gyis byas// ji ltar log pa’i rdzun (rdzun] P, brdzun D) gyis byas pa yis// pha yi drung nas cung zad rnyed pa ltar// de bzhin bdag gis bsod nams cung zad rnyed// de nas ‘jig rtan ‘jam pa’i rdo rjer shog// dpal ldan u rgyan yul du brten (brten] D, bstan P) skyes shing// dpag med rgyud don ma lus rab mkhyen pa// dpal ldan rnam par dul zhi’i zhal snga nas// de yis dbang bskur slob ma’i gtso bo mchog// de las sin dhū’i ngos su de skyes shing// rnam thos bsti gnas zhes bshad rgyal zhugs nas// shā kya’i bshes gnyen bdag nyid sangs rgyas shing// de yis yongs bkrol zhal gyi man ngag go// (Mukhāgama, D 28b.3-5; P 33a.7-33b.2). A second passage earlier in the text also clearly references Buddhajñānapāda as the source of the instructions contained therein (dpal ldan ye shes zhal bshad sgrub thabs kyi// man ngag...) (Mukhāgama, D 17b.5-6). 262 The Mukhāgama does appear to include what indeed may have been oral instructions on the practice of the Caturaṅga/Samantabhadra-sādhana, as it also parallels in some places comments that are made by Vaidyapāda in his commentary on the Samantabhadra. 263 The Anuttarasandhi is cited by Vaidyapāda in his Muktitilaka-vyākhyāna, adding further evidence to the connection between Śākyamitra and Buddhajñānapāda’s tradition (Tomabechi 2008, 173). 264 Kośalālaṃkāra, D 1b.5-2a.5. This short passage is translated in Davidson 2002, 159-60. 265 It does, however, mention a master named Dharmākara as one of Śākyamitra’s teachers in the Koṅkan. One “Dharmākara of the Koṅkana” is also identified by Śrīphalavajra as Buddhajñānapāda’s co-disciple (Samantabhadrasādhana-vṛtti, D 141a.5-6). 50 Buddhajñānapāda.266 The Anuttarasandhi likewise appears to have been influenced by another of Buddhajñānapāda’s works, the Muktitilaka. Here Śākyamitra recounts the story of Buddha Śākyamuni’s awakening by the river Nairañjana, but his account seems to be based not on that in the Lalitavistara (which he mentions), nor even the so-called tantric retelling of this account in the Sarvatathāgatatattvasaṃgraha, but on Buddhajñānapāda’s own version of this account, given in his Muktitilaka.267 The account from the Muktitilaka is central to understanding Buddhajñānapāda’s thought, and I will address it in detail in Chapter Three, but for our present purposes, it is enough to note that it is Buddhajñānapāda’s version of the account of Śākyamuni’s awakening that seems to serve as the basis for Śākyamitra’s in the Anuttarasandhi. The Mukhāgama, Kośalālaṃkāra, and Anuttarasandhi also all include dedicatory verses in which the author has inscribed his own name into the verse, a technique favored by Buddhajñānapāda. Four other works in the Tibetan canon are attributed to Śākyamitra, though Wedemeyer has shown one of these, a commentary on the Caryāmelāpakapradīpa (Tōh. 1834) to be a Tibetan composition.268 Further study of the works attributed to Śākyamitra is certainly necessary to determine more about this influential269 but somewhat elusive author (or authors; indeed, the question of whether the remainder of these works can even be attributed to a single author needs to be addressed). However, given what is currently known of the writings attributed to him, I think we can suggest as likely that one Śākyamitra, author of at least the Mukhāgama, the Kośalālaṃkāra, and the Anuttarasandhi, was a disciple of Buddhajñānapāda’s.270 Death We know little of Buddhajñānapāda’s later life, including his death, but in the Dvitīyakrama Mañjuśrī makes a prediction that has been interpreted by commentators to mean that Buddhajñānapāda would not attain an awakening that involved the full transformation of his bodily aggregates in his lifetime, but only at the time of his passing. Mañjuśrī said: 266 Moreover, I will argue in Chapter Two that Buddhajñānapāda’s autobiographical narrative, which culminates in a vision of Mañjuśrī from whom he directly recieved the instructions that constitute the primary contents of the Dvitīyakrama, serves an important legitimizing function for Buddhajñānapāda’s system of practice outlined therein. Śākyamitra’s account of his studies with human gurus in India does not serve such a function, and thus is more likely to have simply been added in emulation of Buddhajñānapāda’s autobiographical account. Such autobiographies are extremely uncommon in Indian texts of this period—these two are the only ones of which I am aware. 267 Tomabechi (2006, 139n157) notes this parallel and suggests that Śākyamitra’s passage is based upon Buddhajñānapāda’s. There is another half-verse in the Anuttarasandhi that parallels one in Buddhajñāpāda’s Dvitīyakrama, but in that instance both verses appear to be based on a passage in the Sarvabuddhasamāyoga-tantra (See note 123 in my Dvitīyakrama translation.) Nonetheless, the fact that Śākyamitra paraphrases the same passage that Buddhajñānapāda has paraphrased remains telling. 268 These are the Bhadracaryāpraṇidhānarājaṭīkā (‘Phags pa bzang po spyod pa’i smon lam gyi rgyal po’i rgya cher ‘grel pa) (Tōh. 4013), the Krodhamahābalasādhana (‘phags pa khro bo stobs po che’i sgrub thabs) (Tōḥ. 3636), the Mahāmudrāyogāvatārapiṇḍārtha (phyag rgya chen po’i rnal ‘byor la ‘jug pa’i man ngag tu bshad pa), and the Caryāsamuccayapradīpa-nāma-ṭīkā (Spyod pa bsdus pa’i sgron ma zhes bya ba’i rgya cher bshad pa) (Tōḥ. 1834). The latter is the one demonstrated by Wedemeyer (2009) to be a Tibetan composition. 269 Both the Kośalālaṃkāra and the Anuttarasandhi are seminal texts in their own fields. The Kośalālaṃkāra is one of three major Indic commentaries on the Sarvatathāgatatattvasaṃgraha and is important as its earliest word-byword commentary (Hopkins 2005, 19); it influenced another important commentary on the tantra by Ānandagarbha (Kwon 2002, 26). The Anuttarasandhi is important in first introducing the theory of prakṛti and āloka, which serves as the “ontological and epistemilogical foundation of the yogic practice of the Ārya School in its entirety” (my translation from the French) (Tomabechi 2006, 49-50). 270 Tomabechi (2008) has already suggested as much as regards to the author of the Anuttarasandhi and what he refers to as the “compiler” of the Mukhāgama (he does not address the Kośalālaṃkāra) and places this Śākyamitra at a critical juncture between the Jñānapāda and Ārya traditions. 51 However, because of [your] conduct regarding food, And holding a slight delusion with respect to me You will not, in this very life, Bring about a complete transformation of the state of Your body—the aggregates including form. |365| However, you will accomplish consciousness, Which is indestructible, as the mahāmudrā.271 |366| As we have seen in Chögyal Phagpa’s account translated above,Vaidyapāda was followed by the later Tibetan historians in interpreting this prediction to mean that due to Buddhajñānapāda’s earlier lack of faith in the monk who emanated the maṇḍala of Mañjuśrī and his refusal of some foods that the monk’s female companion served to him he would attain the final result of the path only at death.272 This position is stated most clearly in Tāranātha’s Seven Transmissions: If he had previously not given rise to any lack of faith at all with regard to the emanation of Mañjuśrī, who was a practitioner of the avadhūti, he would have transformed in that very [[[body]]] into the vajra rainbow body. However, since he had some minor disrespectful thoughts [towards him], at the age of eighty he left behind the body [produced by karmic] ripening and attained the body of unity.273 Here we have the only mention of which I am aware of Buddhajñānapāda having lived such a long life, passing away only at the age of eighty, no small feat in 8th-century India. Buddhajñānapāda’s Writings You should compose with a genuine intention a sādhana, homa, bali, gaṇacakra, summary, commentary, maṇḍala-vidhi, and so forth for the first stage of the tantra that is the gathering of all the buddhas, which is greatly secret, secret, and supremely secret—this great scripture, surpassed by none—to be like a scalpel for sentient beings who are obscured by the darkness of ignorance. - Mañjuśrī, addressing Buddhajñānapāda directly, Dvitīyakrama Fortunately for our study of his thought, Buddhajñānapāda was a fairly prolific author. Unfortunately, few of his works survive in their original Sanskrit, and our only access to most of his oeuvre is through their Tibetan translations, made primarily in the 11th century.274 The earliest list we have of Buddhajñānapāda’s writings is in the Sukusuma, in which Vaidyapāda comments on Mañjuśrī’s command to Buddhajñanapāda to compose a number of texts (see the 271 khyod kyang zad kyi spyod pa dang// nga la cung zad ‘khrul rtogs pas// khyod kyis (kyis] D C S, kyang P, kyi N) tshe ‘di nyid la ni// gzugs bcas phung po rang lus ni// gnas ni yang dag mi ‘gyur te// |365| rnam par shes pa mi shigs pa// phyag rgya chen por rab tu ‘grub// |366| (Dvitīyakrama, v 365-66). 272 Vaidayapāda’s own comments on the point of not transforming the aggregates of the body in this life are somewhat confusing, but their interpretation by later authors is clearer. See note 504 in my translation of the Dvitīyakrama for more detail on Vaidyapāda’s comments on this passage. 273 Bka’ babs bdun, 108. Here my translation differs significantly from that of Templeman (1983, 76), who I believe has misunderstood the passage. sngon ‘jam dbyangs sprul pa a ba dhū ti’i spyod pa can de la ma dad pa gtan nas ma skyes na/ de nyid du ‘ja’ lus rdo rje’i skur gnas ‘gyur ba yin pa las/ der ma gus pa’i rnam par rtog pa cung zad skyes pas dgungs lo brgyad cu lhan cig bzhes pa na rnam par smin pa’i sku lus bor te/ zung ‘jug gi sku brnyes pa yin no// 274 The translation of Buddhajñānapāda’s Sañcayagāthā-pañjikā, a Prajñāpāramitā commentary that is one of his early works, however, is attributed to the 8th-century Imperial Era translators Kawa Paltsek and Vidyākarasiṃha. 52 quotation above), by listing the master’s compositions that fulfill this command. The sādhana, Vaidyapāda says, refers to the “three Samantabhadrīs” (kun tu bzang mo gsum); the homa is (for?) the generation stage, and he notes that there are two such homa rituals; the bali ritual is that of Unfaltering Tārā (mi nub pa’i sgrol ma); the gaṇacakra text is the Mahāgaṇacakra (though it is unclear if this is meant to be the name of a text or simply stating that it is a ritual for the practice of the mahāgaṇacakra);275 the summary is the Blazing Gem (rin po chebar ba); and the commentary “he did not compose.” As for the maṇḍala-vidhi, Vaidyapāda notes that this vidhi in two hundred and fifty verses was taken to Kaśmir and that he had not seen it. He explains that the “and so forth” includes the Great Root Wisdom (rtsa ba’i ye shes chen po) and the Treasure of Verses (tshigs su bcad pa’i mdzod),276 the Muktitilaka (grol ba’i thig le), and the Ātmasādhanāvatāra (bdag nyid grub par ‘byung ba),277 the *Bodhicittabindu (byang chub sems kyi thig le), the Great Commentary on Glorious Auspiciousness (dpal bkra shis kyi rnam par bshad pa chen po), The Method for Engaging in the Fourth (bzhi pa la ‘jug pa’i thabs),278 and three Jambhala sādhanas.279 Many of these texts cannot be identified among Buddhajñānapāda’s extant works,280 but some of them fortunately can: his Samantabhadra-sādhana,281 Muktitilaka, Ātmasādhanāvatāra, and the three Jambhala sādhanas all survive. 275 Although unfortunately this text is not extant, it does seem to have been translated into Tibetan, as a gaṇacakravidhi (tshogs kyi ‘khor lo’i cho ga) attributed to “Jñānapāda” is listed in the Black Hat Tanjur catalogue (p. 434): tshogs kyi ‘khor lo’i cho ga ye shes zhabs kyi o rgyan gyi yul nas spyan drangs nas mdzad pa smri ti’i ‘gyur (Martin 2011, 650). Smṛtijñānakīrti also translated Buddhajñānapāda’s Caturaṅgasādhana, so it seems likely that this may indeed have been Buddhajñānapāda’s gaṇacakravidhi that he translated, as well. 276 There is at least the outside possibility that this could refer to Buddhajñānapāda’s bsdud pa tshigs su bcad pa’i dka’ ‘grel, the Sañcayagāthā-pañjikā. 277 The title of this text is usually translated into Tibetan as bdag nyid grub pa la ‘jug pa, but presumably it refers to the same text here. 278 Among Buddhajñānapāda’s compositions mentioned by Vaidyapāda that are no longer extant, this title in particular is tantalizing. As noted above, the early Jñānapāda School is generally known to have preserved a system of three, rather than four, initiations, but Vaidyapāda composed the Yogasapta, a treatise on the “seven yogas” of the “the fourth.” We may guess that this Method of Engaging in the Fourth (if it did ever exist—but we have no good reason to doubt Vaidyapāda’s claim that it did) may have had something to do with Buddhajñānapāda’s own position regarding a fourth initiation. I discuss Vaidyapāda’s Yogasapta in Chapter Seven. 279 Sukusuma, D 133b.7-134a.3. Vaidyapāda then notes that these “fourteen teachings” were composed in accordance with Mañjuśrī’s prediction. The only way I have been able to make this list total fourteen is by counting each of the texts listed in the root text as one (7; ignoring the fact that Vaidyapāda says that the sādhana actually refers to three texts, and the homa to two), subtracting the commentary that Vaidyapāda says was not composed (-1), and adding the texts Vaidyapāda lists in as part of the “etc.” (+8; again ignoring the fact that the “three Jambhala sādhanas” counts only as one of the eight). Lotsāwa also gives the list of Buddhajñānapāda’s compositions, clearly drawn from Vaidyapāda, and engages in a similar (but not identical) mathematical endeavor regarding this list of “fourteen”! (Deb ther sngon po, Vol I, 550). In the Blue Annals Roerich has mistakenly identified several of the members of the list that gives with texts that are not Buddhajñānapāda’s. Regarding Roerich’s misidentification of the third “Samantabhadra” text, see note 281. Roerich also mistakenly identifies the “commentary on the tantra” (which Vaidyapāda and both report that Buddhajñānapāda did not compose(!)) as the Candraprabhā (Tōh. 1852), which was actually composed by Pramuditākaravarman. Roerich later reports that this was not Buddhajñānapāda’s work, but it seems that his translation is confused here because he was not aware of Vaidyapāda’s passage from the Sukusuma, which was clearly paraphrasing in this section. Roerich also reports Buddhajñānapāda’s 250-verse maṇḍala-vidhi to be the Guhyasamājamaṇḍalavidhi (Tōh. 1865), which was actually composed by Dīpaṃkarabhadra and has closer to 450 verses. The list of “fourteen” is likewise reported (obviously relying on Vaidyapāda) in Chögyal Phagpa’s biography (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 615-16). 280 There are twelve texts listed by Vaidyapāda here that we do not know: 1. the third among the “three Samantabhadrīs,” (on which see note 279) 2. and 3. the two homa rituals, 4. the bali ritual of Unfaltering Tārā, 5. the gaṇacakra rite, 6. the Blazing Gem, 7. the maṇḍalavidhi in 250 verses, 8. the Great Root Wisdom, 9. the Treasury of Verses, 10. the *Bodhicittabindu, 11. the Great Commentary on Glorious Auspiciousness, and 12. the Method for Engaging in the Fourth. 53 It seems unlikely that Vaidyapāda intends for the list of fourteen works given in the Sukusuma to encompass the entirety of Buddhajñānapāda’s oeuvre, as he is merely commenting on the specific set of Guhyasamāja-related texts that Buddhajñānapāda composed to fulfill Mañjuśrī’s command. Indeed, in addition to the six works from Vaidyapāda’s list that do survive, a number of other compositions in the Tibetan canon (a few of which also survive in their entirety or in part in Sanskrit) are attributed to Buddhajñānapāda under a variety of names. Many of these compositions do indeed appear to be the works of the 8th-century master, though at least one is certainly not his, and the attribution of others remains in question. The following sixteen works in the Tibetan canon are attributed to authors named Buddhajñānapāda, Buddhaśrījñāna, or one of the other above-mentioned variants of the name; I will present a very brief summary and assess the attribution of each of them below: the Dvitīyakramatattvabhāvanāmukhāgama (Tōh. 1853); Mukhāgama (Tōh. 1854); Samantabhadra-sādhana (Tōh. 1855); Caturaṅgasādhanopāyikā-samantabhadrī (Tōh. 1856); Śrīherukasādhana (Tōh. 1857);282 Muktitilaka (Tōh. 1859); Ātmasādhanāvatāra (Tōh. 1860); Bhaṭṭārakāryajambhalajalendrasādhana (Tōh. 1861); Guhyajambhalasādhana (Tōh. 1862); Vistarajambhalasādhana (Tōh. 1863);283 *Gativyūha (Tōh. 1864); *Trikāyavākcittādhiṣṭḥanoddeśa. (Tōh. 2085); Traisattvasamādhisamāpatti (Tōh. 2086); Mahāpratisarārakṣā (Tōh. 3124); Sañcaya-gāthāpañjikā (Tōh. 3798); and the Mahāyānalakṣaṇasamuccaya (Tōh. 3905).284 In assessing the 281 It is interesting that Vaidyapāda refers to this text(s) as the: “three Samantabhadrīs,” as two translations of the sādhana into Tibetan survive in the canon under two different names, the Samantabhadra-sādhana and the Caturaṅga-sādhanopāyikā-samantabhadrī, which upon comparison are not actually distinct texts. There is also a third “Samantabhadra” text, the Kun tu bzang po bsdus don listed in the Peking Tengyur catalogue, but the text itself, however, is strangely absent from the place where it should be in that Tengyur, and receives no mention whatsoever in the catalogues of the other Tengyurs. Lotsāwa (Deb ther sngon po, Vol 1, 450) mentions the Kun tu bzang po bdus don as the third among the “three Samantabhadrīs,” and were the text to be found it would indeed be a good candidate, given its placement in the Peking Tengyur catalogue together with other Jñānapāda Schoool works. Roerich (1976, 370), however, inexplicably identifies the Kun tu bzang po bsdus don as Buddhajñānapāda’s Śrīherukasādhana (Tōh. 1857). I address the issue of the two different extant translations/versions of the Caturaṅga/Samantabhadra-sādhana in Chapter Five. 282 Also in Sādhanamālā, No. 243. 283 Also in Sādhanamālā, No, 285. This short sādhana contains no statement of authorship, but I have included it in the list because I believe we can attribute it to Buddhajñānapāda. I discuss this further below. 284 I have not included in this list three compositions from the canon (and a fourth that is a short prayer extracted from one of the other three) that are clearly attributed to the “Kaśmiri paṇḍita Buddhaśrījñāna,” who lived around 1200 and worked with the Tibetan Lotsāwa Nub Jampai Pal (Gnubs byams pa’i dpal, 1173-1236) on a number of translations preserved in the Tibetan canon, including those of his own writings. Makransky (1997, 268) and C. Dalton and Szántó (forthcoming) mention this Buddhaśrījñāna as the author of an Abhisamayālaṃkāra commentary, the Prajñāpradīpāvalī, and as a namesake of Buddhajñānapāda’s, and warn against confusing the two. In addition to the Prajñāpradīpāvalī (Tōh. 3800), two other works in the Tengyur—the Cittaratnaviśodhanamārgaphala (Tōh. 2465), and the Jinamārgāvatāra (Tōh. 3964) are clearly attributed to this Kaśmiri paṇḍita; indeed in the colophons of all three (and even in several of the colophons of the works he collaborated on as a translator) he is specifically referred to as “the Kaśmiri” Buddhaśrījñāna, presumably to distinguish him from Buddhajñānapāda who as we have seen was occasionally also called by that name. A fourth work, the Jinamārgāvatārodbhavapraṇidhāna (Tōh. 4391), attributed in its colophon to one Buddhaśrījñāna (Sangs rgyas dpal ye shes) (but with no specification that he is the Kaśmiri paṇḍita of that name, nor any translator's colophon) is in fact a prayer that has simply been extracted from his Jinamārgāvatāra (See Jinamārgāvatāra, D 234a.3-235a.1), with a single introductory verse added at the beginning. Buddhaśrījñāna is said in the colophon of his Citta-ratnaviśodhanamārgaphala to be of Kaśmiri blood but to have been born in Nepal, where several of his translations with Nub Jampai Pal were carried out. Besides the Kaśmiri Buddhaśrījñāna’s own three compositions, the two collaborated on the translation of a somewhat eclectic collection of other works in the canon: Ratnākaraśānti’s Vajratārāsādhana (Tōh. 1324); Kālapāda’s Śrīkālacakrasahajasādhana (Tōh. 1361); Maitrīpa’s Śrīcakrasaṃvarasādhana-ratnadīpa (Tōh. 1484); Śrīdhara’s Krodhavārāhīvajrayoginīsādhana (Tōh. 1586); Līlāvajra’s Śrīsahajaguhyasamājasādhana (Tōh. 1913) (That this short sādhana was composed by the later Līlāvajra and not Buddhajñānapāda’s guru Vilāsavajra, whose names in 54 authorship of these works, I am taking into account a number of factors: their presence in Vaidyapāda’s list of Buddhajñānapāda’s writings, their colophonic attribution, their style— including the presence of a dedicatory verse in which the author has inscribed his name, Buddhajñāna, in several of the works—and their content.285 In the course of my research I was able to read all of the sixteen works above in full, with the single exception of the Sañcayagāthāpañjikā, Buddhajñānapāda’s long Prajñāpāramitā commentary. Brief Summary and Assessment of Authorship of works Attributed to Buddhajñānapāda Non-Tantric Works Buddhajñānapāda’s Sañcayagāthā-pañjikā (Tōh. 3798), a commentary on the Sañcayagāthā Prajñāpāramita sutra, seems to have been composed quite early in his career. In the Dvitīyakrama Buddhajñānapāda mentions having composed some treatises at Nālandā in response to the request “the one of noble birth called *Guṇamitra/ā (yon tan bshes gnyen)”286 described by Vaidyapāda as a brahmin nun,287 prior to continuing his travels around the subcontinent studying with tantric gurus. The concluding verses of the Sañcayagāthā-pañjikā, in addition to including a dedicatory verse into which Buddhajñānapāda inscribes his own name, mention as its petitioner the very same *Guṇamitra/ā (here her name is transliterated as gu ṇa mi tra).288 In the Sañcayagāthā-pañjikā Buddhajñānapāda relates each section of the Abhisamayālaṃkāra to passages of the Ratnaguṇasañcayagāthā.289 Here he seems to be emulating his guru Haribhadra’s method of relating the Abhisamayālaṃkāra to one of the shorter Prajñāpāramitā texts; Haribhadra, in his well-known Abhisamayālaṃkārālokā, relates the Abhisamayālaṃkara to the Aṣṭasāhasrikā.290 The Sañcayagāthā-pañjikā was translated into Tibetan during the early translation period, very likely before 824, making it possible that this work may even have been translated during Buddhajñānapāda’s own lifetime.291 The Mahāyānalakṣaṇa-samuccaya (Tōh. 3905) was likewise composed early in Buddhajñānapāda’s career. As noted above, this text has the unusual distinction of having been cited by Buddhajñānapāda’s guru Vilāsavajra in his Nāmamantrārthāvalokinī.292 However, the Mahāyānalakṣaṇa-samuccaya is attributed to Buddhajñānapāda in Samantabhadra’s mid-late 9thcentury (?) Sāramañjarī,293 which cites portions of the text; thus we do have, besides the Tibetan Tibetan are often rather interchangable, is clear from a reference in the work to the four joys (dga’ bzhi). Buddhajñānapāda and even Vaidyapāda only spoke of three.); Niṣkalaṅka’s Śrībandhavimukta-śāstra (Tōh. 2463); Candrakumāra’s Śrīvajrasarasvatīdevyupāyikā (Tōh. 3699); and Vasudhara’s Āryajambhalajalendraviśeṣastotra (Tōh. 3747). 285 For another brief summary of Buddhajñānapāda’s surviving works see C. Dalton and Szántó forthcoming. 286 Dvitīyakrama, verse 4. 287 Sukusuma, D 89b.2. 288 This verse clearly gives the impression that the Sañcayagāthā-pañjikā is one of Buddhajñānapāda’s juvenilia. It reads, “When someone else, her gaze on the result (i.e. awakening), / entrusts one [with a task], why not try?/ [I] wrote this commentary on the difficult points/ for Guṇamitrā alone.” ‘bras bu la lta gzhan gyis kyang// bcol na ci phyir ‘bad mi bya// gu ṇa mi tra kho na’i ngor// dka’ ‘grel ‘di ni ‘dir byas so// (Sañcayagathā-pañjikā, D 189a.4-5). 289 Makransky 1997, 259-60. 290 ibid., 270. See Makransky 1997, 259-63 for a summary of Buddhajñānapāda’s position on the number of kāyas in the Sañcayagāthā-pañjikā, which suggests that here he did not follow the innovative four-kāya theory in his guru Haribhadra’s Abhisamayālaṃkārālokā. 291 See Tomabechi 2008, 175. 292 Tribe 1994, 16; Szántó 2015, 541; C. Dalton and Szántó, forthcoming. 293 Samantabhadra’s mid-late 9th-century dates are suggested by the fact that he mentions that he received the command to compose the Sāramañjarī by one Kīrtipāda, who is likely the same Śrīkīrti who was a disciple of 55 colophonic attribution, an early attestation of its attribution to Buddhajñānapāda by an Indian author who upheld his tradition not long after Buddhajñānapāda’s life.294 The text itself is a relatively short compilation of definitions of basic Buddhist terminology and important aspects of the Mahāyāna path like the aggregates, elements, links of dependent origination, pāramitās, bodhisattva bhūmis, sixteen emptinesses, and so forth.295 The format of the Mahāyānalakṣaṇasamuccaya, in which Buddhajñānapāda begins with a short invocation, then simply lists the topics he will discuss and proceeds to address each in turn, is also found in his *Gativyūha, a text that does bear a signature dedicatory verse. A fragment of the original Sanskrit of the Mahāyānalakṣaṇa-samuccaya survives and has been published, and more of the text can be reconstructed from its citations in the longest available recension of the aforementioned Sāramañjarī.296 Major Tantric Works The Dvitīyakramatattvabhāvanā-mukhāgama297 (Tōh. 1853), edited in Appendix I and translated in Part II of this dissertation, is one of Buddhajñānapāda’s most well-known compositions. It contains his autobiographical account, which serves as the text’s narrative frame, details his vision of Mañjuśrī, and reports the entire contents of the instructions Mañjuśrī gave him in the vision. The instructions given by Mañjuśrī to Buddhajñānapāda reported in the Dvitīyakrama are themselves quite eclectic and the text contains a diverse collection of doctrinal and ritual sections nestled within Buddhajñānpāda’s unique narrative framework. Almost all of the ritual material in the text pertains to the practices of the perfection stage, the “second” stage of tantric practice, as well as the initiatory rituals that permit the practitioner to engage in those practices. This work shares strong similarities and an overlap in vocabulary and content with the Muktitilaka, including a number of parallel passages, and to a lesser degree with the Caturaṅga/Samantabhadra-sādhana, and the Ātmasādhanāvatāra. The Dvitīyakrama is Buddhajñānapāda’s only major extant work that lacks a dedicatory verse with his name inscribed into it. I presume that in the case of the Dvitīyakrama, Buddhajñānapāda found this unnecessary given that the work itself contains his autobiographical details, and that in fact ninety percent of the content is, technically speaking, not actually Buddhajñānapāda’s own composition, but simply his report of Mañjuśrī’s direct instructions, recorded in Mañjuśrī’s own first-person speech. Chapter Two examines its structure and provides a summary of this unique composition, while different aspects of the Dvitīyakrama’s content—its doctrine, generation-stage ritual, perfection-stage ritual, and initiatory sequences—are explored in more detail in each of the remaining chapters. The Muktitilaka (Tōh. 1859) is another important work of Buddhajñānapāda’s, which is known, along with the Dvitīyakrama with which it shares much vocabulary and content and multiple parallel passages, as presenting the perfection stage practices according to Buddhajñanapāda’s Guhyasamāja tradition. While the Muktitilaka certainly does contain some perfection stage instructions—specifically, on the three bindu yoga and vajra recitation—those same practices are presented much more extensively in the Dvitīyakrama. Moreover, perfection Buddhajñānapāda’s guru Pālitipāda, and possibly a co-disciple of Buddhajñānapāda’s. He thus seems to have been one-and-a-half or two generations later than Buddhajñānapāda (See Szántó 2015a, 554). 294 Szántó 2015, 541; C. Dalton and Szántó, forthcoming. 295 Yonezawa 1998 summarizes its contents. 296 Yonezawa 1998; Szántó 2015a, 545. Szántó has generously shared with me his draft edition of the long recension of the Sāramañjarī. 297 I explain my departure from the title usually given for this work, the Dvikramatattvabhāvanā-mukhāgama, in Chapter Two, and also in note 3 of my translation of the Dvitīyakrama in Part II of the dissertation. 56 stage instructions make up only a small portion, maybe fifteen percent, of the Muktitilaka’s content. In fact, the text contains much more doctrinal than ritual material, some of which is presented in innovative narrative ways. The Muktitilaka places a special emphasis on nondual nonconceptual suchness, which can be known instantaneously through relying upon a realized lineage guru, and which subsumes all other outer Vajrayāna practices. Buddhajñānapāda has inscribed his name here in the concluding verses. Chapter Three examines the doctrinal content of the Muktitilaka in more detail, while Chapter Six focuses on the perfection stage practices found here and in the Dvitīyakrama. The Samantabhadra-sādhana (Tōh. 1855) and the Caturaṅgasādhana (Tōh. 1856) are in fact two translations of the same sādhana, which is undoubtedly Buddhajñānapāda’s composition.298 This important work details the rituals of the generation stage practice according to Buddhajñānapāda’s tradition, a nineteen-deity maṇḍala centered on Mañjuvajra, which became distinctive of his Jñānapāda School of Guhyasamāja practice. The Caturaṅga/Samantabhadra-sādhana shares significant parallels, including several parallel passages, with Buddhajñānapāda’s Ātmasādhanāvatāra, particularly in the philosopical section towards the end of the sādhana. It also bears one of his signature dedicatory verses. The sādhana was obviously popular, as it is the subject of five extant commentaries, one of which survives in three different recensions.299 Moreover, almost the entire sādhana was rephrased in his disciple Dīpaṃkarabhadra’s influential Guhyasamajamāṇḍalavidhi. The Samantabhadra survives in its entirety in Sanskrit, but is unfortunately not available for study.300 However, a short portion of the Sanskrit has been photographed and edited, and some has been reconstructed and published on the basis of the Sāramañjarī, an extant Sanskrit commentary on the sādhana.301 This sādhana, including the rituals of Buddhajñānapāda’s generation stage practice, the details of its two different surviving translations into Tibetan, and an overview of the extant commentaries and their relationships, is discussed in detail in Chapter Five. The Ātmasādhanāvatāra (Tōh. 1860) is Buddhajñānapāda’s only surviving tantric work written in prose, rather than verse. It is a complex treatise in which he sets a Yogācāra- Madhyamaka philosophical foundation for arguments that support and defend the tantric path, notably a defense of the practice of deity yoga. The treatise goes so far as to make the claim that full awakening is only possible through the tantric path of deity yoga. The work also deals, however, with a number of Mahāyāna topics approached without a tantric lens. It additionally seeks, as do several of Buddhajñanapāda’s other works, to homologize several important Mahāyāna concepts with tantric concepts or practices, and to identify all of these with, or subsume them within, suchness, which is also described here as the nature of the mind. This text is particularly important as it contains a number of citations of other Buddhist works, giving us 298 Some scholars have sometimes considered these two to be two different works, but more recent scholarly consensus (including Kikuya 2012a, Szántó 2015, and C. Dalton and Szántó forthcoming), which my own reading of the sādhana strongly supports, is that they are indeed two translations of the same work. 299 C. Dalton and Szántó (forthcoming). The commentaries are Vaidyapāda’s Samantabhadrī-ṭīkā, Śrīphalavajra’s Samantabhadrasādhana-vṛtti, Thagana’s Śrīsamantabhadrasādhana-vṛtti, Samantabhadra’s Sāramañjarī, and an unidentified commentary in Sanskrit mentioned in Kawasaki, 2004. The Sāramañjarī survives in three recensions, two in Sanskrit and one in Tibetan. 300 Kawasaki 2004 describes the manuscript that contains the Samantabhadra, but which remains unavailable to scholars to study. 301 Kanō (2014) has published an edition of the verses from a short section of this manuscript which was photographed. The manuscript is on display in the Tibet Museum at the Norbulingka in Lhasa. Other verses survive in several Nepalese ritual manuals (one at the Institute for Advanced Studies of World Religions, IASWR MBB-I- 11; one in the Nepal National archives, NAK 1/1697 = NGMPP A 936/1; and one at the Cambridge University Library Add. 1708.III, f. 2r4-5), and are edited in Tanaka 1996. Szántó (2015, 543n14) has published a revision of Tanaka’s edition of these verses. 57 an idea of the scriptural resources Buddhajñānapāda drew on in his thought. These include the Sarvatathāgatatattvasaṃgraha and the Sarvabuddhasamāyoga-tantra, but the Ātmasādhanāvatāra is unusual in lacking any reference to the Guhyasamāja-tantra, which serves as the central reference for most of his other tantric writings. This absence raises the question of whether the Ātmasādhanāvatāra may have been composed prior to Buddhajñānapāda’s Guhyasamāja-focused tantric works. Moreover, while the Ātmasādhanāvatāra’s only specific reference to Mañjuśrī is in the work’s homage “to the bodhisattva Mañjughoṣa,” as I noted above Buddhajñānapāda has reproduced a lengthy part of his guru Vilāsavajra’s Nāmamantrārthāvalokinī commentary on the Mañjuśrīnāmasaṃgīti towards the end of the Ātmasādhanāvatāra.302 The section in question equates the deities of the Mañjuśrīnāmasaṃgīti maṇḍala described in Vilāsasvajra’s commentary with a number of Mahāyāna principles. So while the Ātmasādhanāvatāra on the whole cannot be said to be a Mañjuśrīnāmasaṃgīti-centered work, the presence—albeit without any introduction or explanation—of the maṇḍala from that tantra in this work confirms Buddhajñānapāda’s familiarity with the Mañjuśrīnāmasaṃgīti, as well as suggesting that the work was indeed likely composed before the Guhyasamāja-tantra became the focus of his writings. In any case, the Ātmasādhanāvatāra is undoubtedly Buddhajñānapāda’s work, given both the presence of his “signature” in the dedicatory verse, as well as the parallels with his Caturaṅga/sSamantabhadrasādhana. Like the Caturaṅga/Samantabhadra, the Sanskrit of the Ātmasādhanāvatāra also survives in full but is likewise unavailable for study,303 but quite a large portion of the Sanskrit text can be reconstructed from the long recension of the Sāramañjarī where it is cited at length.304 The presence of the passage from the Nāmamantrārthāvalokinī, which is extant in Sanskrit, now provides us further access to another portion of the Ātmasādhanāvatāra in its original Sanskrit. The availability of a large portion of the Ātmasādhanāvatāra in Sanskrit is most fortunate, especially given that the Tibetan translation of the text is incomprehensible in a number of places.305 Chapter Three addresses some of the doctrinal points found in the Ātmasādhanāvatāra. Minor Tantric Works In addition to these important tantric compositions, Buddhajñānapāda wrote a few shorter works dealing with tantric practice, mostly sādhanas. The *Gativyūha (Stang stabs kyi bkod pa, Tōh. 1864) is the most unique among these works, as it is not a sādhana, but rather a short text detailing the postures and mudrās of deities, as well as postures and mudrās to be assumed in certain tantric ritual contexts. The first and last sections of the text are in verse, with a prose 302 A number of disparate short sections are taken directly from Chapter Five of the Nāmamantrārthāvalokinī, and incorporated into an almost continuous segment of the Ātmasādhanāvatāra, with very little interjection on Buddhajñānapāda’s own part. The Sanskrit edition of this section of the Nāmamantrārthāvalokinī is found in Tribe 2016 pp. 268-281 and corresponds with the Tibetan of the Ātmasādhanāvatāra found in the Derge recension of that text, D 57a.3-58b.6. The segments of the Nāmamantrārthāvalokinī incorporated into the Ātmasādhanāvatāra are: Nāmamantrārthāvalokinī, Chapter 5, lines 14-20, 36-43 71-76, 81-87, 104-113, 131-139, 156-159, 161-152, 178- 179, 194-195, and 200-203 (following Tribe’s Sanskrit edition, pp. 268-281). 303 See Kawasaki 2004, 51. 304 Szántó 2015, 545-46. Szántó has generously shared with me his draft edition of the Ātmasādhanāvatāra based on its citations in the Sāramañjarī. 305 The colophon states that the translation was done by Śāntibhadra (shānti bha dra) and Khukpa Lhetse (‘go lhas btsas). The Sanskrit text is also difficult, so it is possible that the translators had some trouble with the text, or that they were working with a corrupt manuscript. In any case, the Ātmasādhanāvatāra as it stands in Tibetan translation alone is quite unapproachable in more than one place. 58 section in the middle. While the text lacks the traditional translator’s opening with a Sanskrit title and translator’s homage, as well as a translator’s colophon, it does have a dedicatory verse in which Buddhajñānapāda has inscribed his own name. The *Gativyūha also begins with an homage to Vajra Hūṃkāra, who is mentioned twice in Buddhajñānapāda’s Jambhala sādhanas as the self-visualization for the more wrathful methods of accomplishment.306 As noted above, a composition by Śrīkīrti, who seems to have been a fellow disciple of Buddhajñānapāda’s under their master Pālitapāda likewise focuses on postures among the other details of the initiatory ritual, so it is possible that this was something emphasized by their common guru.307 The Śrīherukasādhana (Tōh. 1857) is a short work which is, as its name suggests, focused on the figure of the wrathful Heruka. The general structure of the sādhana, albeit in an extremely condensed form, is the same as that of Buddhajñanapāda’s Caturaṅga/Samantabhadra-sādhana, and the causal deity (described here in the Śrīherukasādhana as the vajrasattva and in the Samantabhadra as Vajrabhadra308), is identical in terms of form, color, and implements, to the causal deity as described in the Caturaṅga/Samantabhadra. These shared features suggest that the attribution to Buddhajñānapāda (his name is given as the Ācārya Buddhajñāna (slob dpon sangs rgyas ye shes) in the colophon) may indeed be correct.309 Moreover, the sādhana is grouped with others of his works in the Tibetan canon, suggesting that the redactors of the canon may also have accepted the attribution.310 Although the gate guardians depicted in the Caturaṅga/Samantabhadrasādhana are certainly wrathful in aspect, the specifically cremation ground aesthetic of the Heruka in the Śrīheruka-sādhana—holding a skull garland, his body smeared with ash, garlanded by bones—is not reflected in Buddhajñānapāda’s other writings, and is more often associated with the later Yoginī tantras.311 The short commentary (Tōh. 1858) that follows this sādhana in the Tengyur clearly associates this work with the Guhyasamāja-tantra, and, as seen in the 9th-century Guhyasiddhi (a text that is probably slightly later than Buddhajñānapāda) that aesthetic was already associated with Guhyasamāja practitioners engaging in vrata practices, even if it was not representated in the aesthetic of the primary deities of the Guhyasamājamaṇḍala itself.312 What is more, the idea of and the term heruka were certainly in use in Buddhajñānapāda’s time, even in conjunction with the Guhyasamāja cycle, as both Vaidyapāda and Samantabhadra’s commentaries specify the herukas among the “others” in whose accumulations of merit Buddhajñānapāda rejoices in a verse from the Caturaṅga/Samantabhadra-sādhana.313 If the Śrīheruka-sādhana is indeed a composition of Buddhajñānapāda’s, it would thus seem to be an early example of the genre of a Heruka sādhana with a cremation ground aesthetic. Most of the sādhana, with the exception of four and a half 306 This practice is mentioned in both the Bhaṭṭārakāryajambhalajalendra-sādhana and the Vistarajambhalasādhana. 307 See Szántó 2015a, 552-3. 308 rdo rje bzang. A combination of the terms Vajradhara and Samantabhadra? The Caturaṅga reads rdo rje dam pa. 309 The Śrīherukasādhana is followed in the Tengyur by a short commentary (Tōh. 1858) on the work that is anonymous. The commentary clearly identifies the sādhana as pertaining to the Guhyasamāja tradition, and comments on the full sādhana as contained in the Tibetan canon, including the final four and a half verses on the Heruka vrata missing in the Sanskrit. There is nothing in the content of the commentary that would absolutely preclude its having been composed by Buddhajñānapāda, nor is there anything that strongly suggests that it was. 310 However, this was not always the case when works were included in the Tibetan canon as Wedemeyer 2009 has shown. 311 The wrathful deities of the protection cakra surrounding the maṇḍala in Śākyamitra’s Mukhāgama, which he says is based on oral instructions from Buddhajñānapāda, come closer to this aesthetic, garlanded by bones and snakes, but still there is no mention of cremation ground ash or skulls in that imagery. 312 See Krug 2018, 266. 313 Caturaṅga-sādhana-ṭīkā, D 135b.4; Sāramañjarī, D 6b.6. 59 verses at the end of the Tibetan translation, survives in Sanskrit, as No. 243 in the Sādhanamālā.314 The final verses that are absent in the Sanskrit are not part of the samādhi of the deity but rather briefly describe the Heruka vratas that the practitioner is enjoined to undertake. The Bhaṭṭārakāryajambhalajalendra-sādhana (Tōh. 1861), Guhyajambhala-sādhana (Tōh. 1862), and Vistarajambhala-sādhana (Tōh. 1863), are three short Jambhala sādhanas, the first two of which are attributed to Buddhajñānapāda, and all three of which share a common final colophon with the injunction that the three sādhanas may not be given to disciples who have not received initiation.315 While the Vistarajambhala-sādhana has no authorial attribution, nor a translator’s colophon, there are a number of factors that suggest it to be Buddhajñanapāda’s composition: Vaidyapāda mentions in the Sukusuma that Buddhajñānapāda composed three Jambhala sādhanas, the two Jambhala sādhanas immediately preceding the Visatarajambhalasādhana in the Tengyur are attributed to Buddhajñānapāda, and the three sādhanas, despite the first two having been translated by different teams of translators, all share a common colophon, as mentioned above. Moreover, the Vistarajambhala-sādhana shares some features with each of the two preceding sādhanas, and the translator’s colophon of the Bhaṭṭārakāryajambhalajalendra-sādhana suggests following the Guhyajambhala-sādhana and the Vistarajambhala-sādhana on some details regarding the visualization of the forms of the retinue deities. As noted above, Bhaṭṭārakāryajambhalajalendra-sādhana and the Vistarajambhala-sādhana both advocate the practice of Vajra Hūṃkāra, to whom homage is paid in the initial verses of Buddhajñānapāda’s *Gativyūha. Jambhala and Vasudharā, god and goddess of wealth, respectively, played important roles in Buddhajñānapāda’s life story: as noted above Buddhajñānapāda reports that Jambhala himself provided a daily stipend of sorts to Buddhajñānapāda and his disciples when he was living at the Parvata cave behind Vajrāsana, and Vaidyapāda explains that the daily provisions for Pālitapāda’s disciples were provided by the goddess Vasudharā. Vaidyapāda also notes that it was during his discipleship under the yoginī Guṇeru in Oḍḍīyāna that Buddhajñānapāda achieved the accomplishment of Jambhala. The Vistarajambhala-sādhana also survives in Sanskrit, as Sādhanamālā No. 285,316 though there are a number of places where the Tibetan translation includes passages not present in the Sanskrit, a few instances of the opposite, and one place where the two include divergent versions of the same passage.317 Works of Questionable Attribution, Unlikely to be Buddhajñānapāda’s Among the sixteen extant works attributed to Buddhajñānapāda under the variety of his names in the Tibetan canon, there are four that remain of questionable attribution, and one that we can definitively rule out as being his composition. Among the works whose attribution to Buddhajñānapāda it is difficult to be certain about is the Mahāpratisarārakṣā (Tōh. 3124). This is a short protective ritual centered around the goddess Mahāpratisarā, one of the figures in the well-known pan-Asian Pañcarakṣā tradition. While there is nothing in the ritual that would preclude its composition by Buddhajñānapāda, there is no suggestion that he composed such a work in any of the sources describing his life and his writings. What is more, the composition, which is written primarily in prose combined with short verses that the ritual officiant 314 Bajracharya 1928, No. 243. 315 Vistarajambhala-sādhana, D 66b.2. 316 Bajracharya 1928, No. 285. 317 The Tibetan translation of the text is also generally problematic, including instances where, for example, the Sanskrit phonetics for a word are included right in the middle of the two syllables of the word that is its Tibetan translation! For example, the Tibetan at one point reads sa bon bī ja pū ra gang ba, when in fact “sa bon gang ba” is itself the Tibetan translation of the Sanskrit word bījapūra! 60 presumably is to recite, along with a number of mantras, does end with dedicatory verses— precisely the kind of verses into which Buddhajñānapāda was wont to inscribe his name—but the author of the Mahāpratisarārakṣā has not done so. Of course Buddhajñānapāda did not always write such dedicatory verses, but in his extant works that do contain such dedicatory verses, he did always include his name.318 The Śrīraktayamārisādhana (Tōh. 2084), *Trikāyavākcittādhiṣṭhanoddeśa (Tōh. 2085), and Trisattvasamādhisamāpatti (Tōh. 2086) are three short Yamāri-related works ascribed in their colophons to Buddhajñānapāda.319 The colophon of the first of these is followed by a list of of its lineage masters which begins: Yamāri, Vajrayoginī, Buddhajñānapāda, Śrīdhara,320 Līlāvajra, and continues through a number of further masters up to “myself,” presumably one of the translators.321 In the authorial colophon of two of these short works Buddhajñānapāda is associated with Vajrayoginī or Vajravārāhī, and said to have had her vision or to have directly received her blessings, which is never reported in any of the life accounts of the 8th/9th century master. Two of the works also have short dedicatory verses at the end, but neither has Buddhajñanapāda’s characteristic signature within them. The vocabulary and style of the works (which is common among the three) is not reminiscent of that in Buddhajñānapāda’s other writings, and particularly the very casual use of the terms “generation stage” and “perfection stage”322 in the Trisattvasamādhisamāpatti gives the sense of terms that were in common usage, whereas we know that Buddhajñānapāda was one of the early masters to employ these terms and we find them introduced and used very deliberately, rather than casually, in his other writings. As mentioned above, though, Buddhajñānapāda is very clearly associated with Yamāri practices in the later Tibetan tradition, especially within the Ngor tradition of the Sakyapas, who also uphold his Guhyasamāja lineage, so it is not unexpected to find Yamāri-related texts attributed to Buddhajñānapāda. However, the Buddhajñānapāda who is the author of these particular Yamāri texts is, in my estimation and for the reasons noted above, unlikely to be the same as the Buddhajñānapāda who is the author of the Dvitīyakrama, the Caturaṅga/Samantabhadrasādhana and the other works mentioned above. 318 The single exception to this rule is the Dvitīyakrama, discussed above. In possible support of the attribution of the Mahāpratisarārakṣā to Buddhajñānapāda, the paṇḍita Sumatikīrti, who was involved in the translation of the Mahāpratisarārakṣā, also translated Samantabhadra’s Jñānapāda School Mañjuśrī-sādhana (Tōh. 1880), and we know that canonical translators often worked on multiple works by a particular author or that were connected to a particular tradition. The works that Sumatikīrti translated, however, appear to be a rather eclectic collection, so his involvement in these two works may not be related. 319 His name in the colophon reads bu ddha jñā na pā da in Tōh. 2084 and sangs rgyas ye shes zhabs in Tōh. 2085 and Tōh. 2086. 320 As noted above, in Amye Zhab’s Gshin rje chos ‘byung, Śrīdhara is the master listed after Buddhajñānapāda in one of the Yamāri lineages given there, but Amye Zhab’s lineage sequence begins with Mañjuvajra, rather than Yamāri and Vajrayoginī, and passes from Śrīdhara to Nāropa, not to Līlāvajra (Gshin rje chos ‘byung, 16a.3-4). But the lineage Amye Zhab writes about is a Kṛṣṇayamāri lineage, rather than a Raktayāmari lineage. In another lineage description in the same work Amye Zhab explains that Mañjuśrī passed the teachings on Vajrabhairava to Buddhajñānapāda, who gave them to Dīpaṃkarabhadra, who passed them to Śrīdhara, who then composed many Yamāri-related treatises (Gshin rje chos ‘byung, 48a.4-5). Śrīdhara does mention Buddhajñānapāda in at least three of his Yamāri-related works, but in all of these cases he cites Buddhajñānapāda specifically in his association with the Guhyasamāja tradition, rather than the Yamāri tradition. In one case the reference is to an iconograpical issue, an in two others to a doctrinal point (See Sahajāloka, D 86b.6-7, Kṛṣṇayamārisādhana, D 6b.5, and Śrīraktayamārisādhana, D 93b.6, respectively). 321 According to the translator’s colophon of all three of these short works—the initial translators were the paṇḍita from Madhyadeśa Ānandabhadra and the Tibetan lotsāwa Sonam Gyaltsen (bsod rnams rgyal mtshan), but the texts were later revised or retranslated at Nālandā by the Siddha Kaṛnaśrī and Neten Nyima Gyaltsen (gnas brten nyi ma rgyal mtshan). The three colophons are not entirely identical, but all convey the same information. 322 For example, the text is described in the colophon as skye rdzog zung du ‘jug pa’i man ngag. 61 Works not by Buddhajñānapāda While the Mukhāgama (Tōh. 1854), as mentioned above, is attributed to Buddhajñānapāda in both traditional histories and modern scholarship due probably to the presence of his name and the absence of Śākyamitra’s in the work’s colophon, the dedicatory verses of text itself clearly state that it was written by Śākyamitra, and I therefore do not consider it to be Buddhajñānapāda’s work. Moreover, in the initial verses of the text Śākyamitra goes so far as to distinguish his own style from Buddhajñānapāda’s. The Mukhāgama is, however, obviously based on Buddhajñānapāda’s Samantabhadra-sādhana and may very well be based on his oral instructions, as its title—and perhaps even its contents—indicate.323 Clearly Śākyamitra was referencing the Dvitīyakrama-tattvabhāvanā-mukhāgama when he selected the title of the Mukhāgama, and obviously wanted his own composition to be associated with Buddhajñānapāda and his tradition, despite stating in the concluding verses that he (Śākyamitra) himself composed it. Concluding Summary: Extant Works Reasonably Attributable to Buddhajñānapāda Out of the sixteen surviving works in the Tibetan canon attributed to Buddhajñānapāda under the variety of his names, I believe we can conclude with a reasonable degree of certainty that eleven are the writings of a single author. While five of these works also survive fully or in part in Sanskrit, as described in their summaries above, I know of no work attributed to Buddhajñānapāda that is extant in Sanskrit but not in Tibetan translation.324 The extant works we can confidently attribute to our 8th/9th-century author Buddhajñānapāda are: 1. Dvitīyakramatattvabhāvanā-mukhāgama (Tōh. 1853) 2a. Samantabhadra-sādhana (Tōh. 1855) 2b. Caturaṅgasādhanopāyikā-samantabhadrī (Tōh. 1856) 3. Śrīheruka-sādhana (Sādhanamālā No. 243, Tōh. 1857)325 4. Muktitilaka (Tōh. 1859) 5. Ātmasādhanāvatāra (Tōh. 1860) 6. Bhaṭṭārakāryajambhalajalendra-sādhana (Tōh. 1861) 7. Guhyajambhala-sādhana (Tōh. 186) 8. Vistarajambhala-sādhana (Sādhanamālā No. 285, Tōh. 1863) 9. *Gativyūha (Tōh. 1864) 10. Sañcayagāthā-pañjikā (Tōh. 3798) 11. Mahāyānalakṣaṇa-samuccaya (Tōh. 3905) A Life Remarkably Lived Buddhajñānapāda was a remarkable figure about whom we are able to glean a surprising amount of information for an individual who lived in early medieval India. In particular, his autobiographical narration in the Dvitīyakrama gives us a sense of the life of a well-educated yogin whose determination to attain awakening through tantric methods drove him to travel vast distances and serve many teachers in what appear to have been thriving tantric communities 323 See note 260. 324 If there were such an extant work, however, a very likely place for it to be found would be in the manuscript described in Kawasaki 2004. 325 Among the works in this list, this is the one about which I feel the most hesitation in making the attribution to Buddhajñānapāda, for the reasons outlined above. 62 across the breadth of the subcontinent. In many ways Buddhajñānapāda’s life was extraordinary—his transformative vision of Mañjuśrī, composition of popular and lasting tantric works, and possible connections with Pāla royalty and position of note in one of the large monasteries of his time. Yet his own account of his life also somehow gives a flavor of the ordinary lived experience of a Buddhist tantric practitioner in his time—studying with many different gurus, acknowledging his own lack of understanding early in his career, and persevering in his study and practice until he encountered what he felt to be genuine truth, which he then felt compelled to share with others. In the next chapter I will examine in more detail the work that contains this remarkable account, Buddhajñānapāda’s Dvitīyakrama, addressing the unusual structure of the text and how Buddhajñānapāda used that as a way to convey its equally unique contents. The great variety of topics addressed in the Dvitīyakrama will give us a helpful overview of Buddhajñānapāda’s thought before moving on, in the subsequent chapters, to an indepth assessment of some of its aspects. 63 Chapter Two Narrating Revelation: The Dvitīyakrama’s Unique Framing of Doctrine and Ritual On the eighth day of the seventh month, during the constellation Puṣya at the time when Mṛgaśīrṣa and Hasta are fading, in the early morning, right at dawn, towards the emanated maṇḍala-cakra of Mañjuśrī, I made a fervent supplication to understand the meaning. -Buddhajñānapāda, Dvitīyakrama The Dvitīyakrama is certainly the most unique of Buddhajñānapāda’s compositions, but it also stands out among Indian Buddhist writings as a whole. While descriptions of visionary encounters are not at all uncommon in Indian Buddhist literature, especially from the rise of the Mahāyāna onward, autobiographical descriptions of visionary encounters—and indeed any type of autobiographical writings at all—are. Buddhajñānapāda’s claims to visionary inspiration positioned him among important earlier Buddhist figures like Nāgārjuna and Asaṅga whose visionary experiences are regarded by the tradition as the source of new—or, more precisely, newly revealed—Buddhist doctrinal and practice systems. The accounts of the visionary experiences of these earlier authors, however, were recorded and passed down by subsequent members of their traditions, whereas Buddhajñānapāda documented his own, along with other details of his life. Such autobiographical writings are extremely rare; apart from the Dvitīyakrama and Buddhajñānapāda’s disciple Śākyamitra’s short autobiographical account in his Kośalālaṃkāra, which the Dvitīyakrama seems to have inspired, I know of no other autobiographical narratives in early Indian Buddhist literature.1 In the Dvitīyakrama Buddhajñānapāda crafts a narrative frame for the work’s sometimes innovative contents that puts him in a role that is both central and peripheral. That is, on the one hand, the Dvitīyakrama is Buddhajñānapāda’s account of his own life story, but on the other hand the primary content of the work is not his at all: a full ninety percent of the Dvitīyakrama is simply a record of Mañjuśrī’s direct (sometimes even first-person) speech, so it is he who is in some sense the author of—and perhaps more importantly the authority behind—that content. Buddhajñānapāda’s opening reference to the contents of the Dvitīyakrama as “the words of the guru Mañjuśrī,” Vaidyapāda explains, is meant to refute the idea that Buddhajñānapāda himself had composed the instructions.2 Throughout the history of Buddhist literature Buddhist authors—following a general trend in Indian traditions at large—have upheld this trope of “not having made up anything new,” explaining innovation as nothing more than the correct interpretation of what was already presented in the original scriptural sources, or in the compositions of lauded philosophers like Nāgārjuna (who themselves claimed simply to be correctly interpreting scripture). The production of new scriptures has thus been an important way in which Buddhist traditions have grown and developed over time. Paul Harrison’s work on the development of the early Mahāyāna scriptural corpus describes the pratyutpannasamādhi, advocated in the Pratyutpanna-sūtra, in which the practitioner engages in prolonged visualization of a buddha and his buddhafield in order to gain a visionary encounter in which he receives teachings from that buddha, that the practitioner 1 Janet Gyatso has written that “First-person discourse about one’s life is virtually nonexistent in Indian Buddhist literature; we can only mention the Therī- and Tharagāthā, which contain a few poems that may, be autobiographical, and occasional statements attributed to the Buddha.” (Gyatso 1998, 115). 2 rang bzo dgag pa (Sukusuma, D 89a.3). 64 subsequently remembers and shares with others.3 The Mañjuśrīnāmasaṃgīti, which Buddhajñānapāda’s guru Vilāsavajra wrote a commentary on, and which seems to have had some influence on Buddhajñānapāda and his work, likewise prescribes a similar, but tantricised buddhānusmṛti type of practice focused on Mañjuśrī: “He who...recites [from memory] this crest jewel called the Nāmasaṃgīti three times each day, or who recites it from a book [and] who, taking the form of the Fortunate One, Mañjuśrījñānasattva, reflects and meditates on that form...will before very long see him [i.e. Mañjuśrī] in his Form Body (rūpakāya).”4 The practice of revelation through visionary encounter was thus already well established in Buddhajñānapāda’s time, and, in fact, at the close of his teaching in the Dvitīyakrama, Mañjūśrī explicitly commands Buddhajñānapāda to compile his instructions and pass them on. Buddhajñānapāda was unique, however, in so directly narrating this personal encounter in his own writings. The title of the Dvitīyakramatattvabhāvanā-mukhāgama, Oral Instructions on the Meditation on the Reality of the Second Stage, already implies that its instructions were received from a source other than the author/compiler of the text—here, of course, Mañjuśrī. The term mukhāgama functions in the title as a sort of genre marker, in the place where one often sees such genre markers as sādhana, vidhi, ṭīkā, or vṛtti. It is not a popular genre in Buddhist literature; one finds just a few instances in the titles of Buddhist texts, where it seems to indicate instructions that have been received from an authority other than the author.5 As we saw above, Śākyamitra uses the term as the title of his own Mukhāgama, which purports to record Buddhajñānapāda’s oral instructions on the Caturaṅga/Samantabhadra-sādhana.6 In any case, the mukhāgama genre is an uncommon one, and Buddhajñānapāda seems to have been particularly unusual in using it to label instructions received directly from a buddha/bodhisattva teacher such as Mañjuśrī. The precise identity or nature of Mañjuśrī as the source of revelation in the Dvitīyakrama is not as straightforward as it might seem. In Chapter One I discussed the claims of some later Tibetan authors that the monk who emanated the maṇḍala of Mañjuśrī for Buddhajñānapāda in the Dvitīyakrama was the master Mañjuśrīmitra, a claim not found in any of the Indic texts (nor in many of the Tibetan works) relating to Buddhajñānapāda, nor in the Dvitīyakrama. Buddhajñānapāda and Vaidyapāda seem to understand this “emanated monk” simply as an emanation of the bodhisattva/deity who was the true source of Buddhajñānapāda’s revelation. But this still leaves the question of how to understand the identity of the visionary form of Mañjuśrī who taught Buddhajñānapāda directly. Buddhajñānapāda himself refers to him in the Dvitīyakrama as the “guru Mañjuśrī”7 and the “great bodhisattva Mañjuśrī.”8 Vaidyapāda, however, takes pains to clarify that Mañjuśrī is called a bodhisattva “because he is integrated 3 Harrison 1978, 54-5; 2003, 120. 4 Tribe 1997, 124. 5 These include the Nandyāvartatrayamukhāgama (Tōh. 2415) of Kaṅkālā and Mekalā, which states at the outset that it is based on “the guru’s oral instructions;” the Grub pa’i dbang phyug paṇḍita chen po shrī ba na ratna’i zhal lung rin po che’i snying po’i phreng ba (P 5096), and Zhal lung rin po che’i phreng ba (P 5099), which are two nearly identical short compilations of citations from Indic texts that the 15th-century Indian yogin Vaṇaratna often cited in his teachings (thanks to Ryan Damron for sharing with me his understanding of the content and function of the latter two works); and the Mgon po dmar po’i tshe bsgrub kyi zhal gdams (P 4927) of Śavaripa is a short longlife practice associated with red Mahākāla. The authorial attribution to Śavaripa is followed by a list of the instruction’s lineage, which begins with Vajradhara, and has its “author” Śavaripa next in the lineage. 6 In that text Śākyamitra does claim some agency, suggesting that he will deliver these instructions with more clarity and simplicity than Buddhajñānapāda did in the Caturaṅga/Samantabhadra. See Chapter One, note 258. 7 Dvitīyakrama, verse 2. 8 Dvitīyakrama, verse 19. 65 with awakening (bodhi), not because awakening is his goal.”9 This suggests that Vaidyapāda understands, and wants his readers to understand, Mañjuśrī as elevated above the level of an “ordinary” bodhisattva abiding on the bhūmis, and instead as representing a form of full awakening.10 Similarly, in his Muktitilaka-vyakhyāna Vaidyapāda identifies Mañjuśrī as the “foundation” of the unique qualities of the Bhagavan, both in terms of their cause and their effect.11 Such a portrayal is not at all outside of the range of Buddhist conceptions of the figure of Mañjuśrī in the 8th and 9th centuries. Early Mahāyāna sūtras associate him with wisdom, and often portray him as a tenth-bhūmi bodhisattva, but sometimes even as a fully awakened buddha or a teacher of buddhas.12 In the 8th-century Mañjuśrīnāmasaṃgīti, however, Mañjuśrī is named as the Ādibuddha.13 The Nāmamantrārthāvalokinī, a commentary on the Nāmasaṃgīti composed by Buddhajñānapāda’s guru, Vilāsavajra, likewise equates Mañjuśrī, here as Mañjuśrījñānasattva, with nondual wisdom itself, and thus the source of the buddhas’ awakening.14 Indeed, Vaidyapāda’s statement mentioned above distinguishing Mañjuśrī from a mere bodhisattva on the bhūmis echoes a similar statement made by Vilāsavajra about Mañjuśrījñānasattva in his commentary to the Nāmasaṃgīti: “The gnosis-being Mañjuśrī is not the bodhisattva who is the master of the ten stages (bhūmi). Rather, he is non-dual gnosis (advayajñāna), the perfection of wisdom (prajñāpāramitā) itself.”15 Such a perspective seems to underlie Buddhajñānapāda’s understanding of Mañjuśrī in the Dvitīyakrama. Though referring to him as a “bodhisattva,” when Buddhajñānapāda supplicates the visionary Mañjuśrī for instructions, he addresses him in quite elevated terms as “the father and the mother of all beings,” the “emptier of the three realms, greatest of the great,” as “beginningless, unvoiced, lacking the upper part of the bindu, the revered, the letterless, producer of nectar, the empty bliss of great joy,” and the “great protector.”16 These exalted epithets suggest that Buddhajñānapāda likewise identifies Mañjuśrī here as a fully awakened buddha, the ultimate source of the ultimate truth. In the homage at the beginning of his Muktitilaka, Buddhajñānapāda describes Mañjuśrī as the “emanation of boundless [[[buddha]]] families.”17 It is no wonder, then, that Mañjuśrī, in the form of Mañjuvajra, is the central figure in Buddhajñānapāda’s generation stage Guhyasamāja sādhana, the Caturaṅga/Samantabhadra, despite Mañjuśrī/Mañjuvajra’s peripheral role in the Guhyasamāja-tantra itself.18 In the Dvitīyakrama Buddhajñānapāda makes supplications “towards the emanated maṇḍala-cakra of Mañjuśrī,” thus implying he saw Mañjuśrī as the central figure in a maṇḍala of deities. Presumably this was the nineteen-deity maṇḍala of Mañjuvajra that Buddhajñānapāda describes 9 de nyid byang chub dang ‘dres pa’i phyir byang chub sems dpa’ ste/ byang chub la dmigs pa ni ma yin no/ (Sukusuma, D 93a.2). 10 In his Muktitilaka-vyākhyāna, when citing a passage from the Dvitīyakrama Vaidyapāda refers to the speaker of the instructions in the Dvitīyakrama simply as “the bodhisattva Mañjuśrī” (Muktitilaka-vyākhyāna, 46b.5). It is interesting, though, that when citing the passage he specifies Mañjuśrī rather than Buddhajñānapāda as the source of the quotation. 11 Muktitilaka-vyākhyāna, D 46b.6-7. 12 Tribe 2016, 15n27. Here Tribe, referencing the work of Lamotte and others, gives a useful short overview of Mañjūśrī’s role in the Mahāyāna sūtras. 13 ibid.; Tribe 1997, 109. 14 ibid. 15 Tribe 2016, 8. 16 In Dvitīyakrama, verses 12 and 13. 17 Muktitilaka, D 47a.1-2. This identification of Mañjuśrī with all five of the buddha families is found in a converse (but presumably complimentary) form in Vilāsavajra’s Nāmamantrāvalokinī. 18 Mañjuśrī receives two brief mentions in the Guhyasamāja-tantra, in Chapters 13 and 15, and Mañjuvajra likewise has two brief mentions in Chapters 12 and 16.