Guhyagarbha Tantra (with Commentary)

by Gyurme Dorje | 1987 | 6,373 words

The English translation of the Guhyagarbha Tantra, including Longchenpa's commentary from the 14th century. The whole work is presented as a critical investigation into the Nyingma School of Tibetan Buddhism, of which the Guhyagarbhatantra is it's principle text. It contains twenty-two chapters teaching the essence and practice of Mahayoga, which s...

6. Origin of the Guhyagarbhatattvaviniścayamahātantra

Any discussion of the historical position of the Guhyagarbhatattvayiniścayamahātantra must take note of the controversy regarding its origin which prevailed in Tibet in the centuries immediately following the later propagation of the Buddhist teachings. We have already noted that dissemination of the ancient tantras was restricted in consequence of their secrecy and the denser of their misapplication. In the eleventh century, Lha bLa-ma Ye-shes-'od and others sought to outlaw the teachins and practice of tantra, accusing the adherents of this tradition of engaging in sbyor-sgrol practices. The Guhyagarbhatattvaviniścaya. one of the main texts expounding these methods, was subjected to criticism. Other such texts, including the Guhyasamājatantra. were paradoxically exempted from this attack. Nonetheless, as we shall see in our examination of the text itself, the expression of these techniques in the Guhyagarbha appears to have a particularly subtle intention when contrasted with the overt sexual and macabre descriptions found in certain other tantras.[1] Further study of the Collected Tantras of the rNying-ma-pa, particularly of its Anuyoga and Atiyoga texts, would, it has been suggested, reveal that the the ancient translations appear to have their own distinct terminology and a literary style better suited to the Tibetan language than the rigid formalism present in many of the later translations, giving some weight to Rong-zom-pa's early critique.[2]

An Incident from the life of Zur-chung-pa Shes-rab Grags alludes to this controversy with some humour. When four students of the bKa'-gdams-pa teacher Khyung-po Grags-se were defeated by Zur-chung-pa in debate and agreed to become his disciples, having understood the profundity of his view, Khyung-po Grags-se announced:[3]

“Anyone who kills one like Zur-chung-pa, who harbours perverse opinions and leads everyone astray, will certainly attain buddha-hood!”

Zur-chung-pa, on hearing this, remained silent without thought of anger and was later seen smiling. On being asked the reason for his mirth, he answered,

“As for doctrines, this, my secret mantra-tradition of the greater vehicle, is it! For it is the tradition of secret mantras that maintains that buddha-hood may be attained by ‘liberation’; the dialecticians do not think so. Now, even such a great dialectician as Khyung-po Grags-se has said that anyone who kills one like Zur-chung-pa will attain buddha-hood. So, in his innermost heart, he has turned to my doctrine. Therefore, I am delighted!”[4]

Another eleventh century figure, 'Gos Khug-pa Lhas-btsas, reputedly nursing a grudge because he had been refused instruction by Zur-po-che,[5] in his Broadside ('byams-yig) sought to refute the authentic origin of the tantra, imputing it to have “four faults” (skyon-bzhi), and claiming that it was not known in India.[6] The tantra was generally considered by 'Gos to lack the five excellencies (phun-sum tshogs-pa lnga), i.e. those of teacher, retinue, location, teaching and time. He imputed it to have a flawed introduction (klong-log). i.e., that unlike other tantras it had no audience of bodhisattvas; a flawed time (dus-log). i.e., that it speaks of four times instead of three; a flawed maṇḍala (dkyil-'khor lag), i.e., that Vajrasattva appears at the centre of the maṇḍala instead of Vairocana; and a flawed text (rgyud-log) because it refers to other tantras when indicating the auspicious times and days for its practice.[7]

Slight variations on these “four faults” have been reported in the later writings of Sog-bzlog-pa bLo-gros rGyal-mtshan, dPa'-bo gTsug-lag 'Phreng-ba, and others. Sog-bzlog-pa, in his dris-lan lung-dang rig-pa'i 'brug-agra. p. 33. holds 'Gos Lhas-btsas to have imputed the Guhyagarbha to be flawed in word (sgra-akyon). flawed in meaning (don-skyon), flawed by contradiction (gal-skyon) and flawed by disconnection (ma-'brel-ba'i skyon).

dPa'-bo gTsug-lag 'Phreng-ba, mkhas-pa'i dga'-ston. p. 179, speaks of “four errors” (mi-rigs-pa bzhi). namely, the error of the Guhyagarbha’s introductory statement “At the time of this explanation” ('di-skad bshad-pa'i dus-na ces ma-rigs-pa), the error of its maṇḍala which is said to have an immeasurable ground (gzhi tshad-med mi-rigs-pa).[8] the error of its explaining the three times as four times (dus-gsum-la dus-bzhir bshad-pa mi-rigs-pa) and the error of Vajrasattva being the central deity of the maṇḍala instead of Vairocana (dkyil-'khor-gyi gtso-bo rdo-rje sems-dpas byas-pa mi-rigs-pa).

The rNying-ma response to these four flaws, faults or errors is disclosed in the course of the appended commentary by kLong-chen Rab-'byams-pa. Vigorous counter-refutations have also been made, in particular by bCom-ldan Rig-pa'i Ral-gri, and the aforementioned authors—dPa'-bo gTsug-lag 'Phreng-ba and Sog-bzlog-pa bLo-gros rGyal-mtshan.

bCom-ldan Rig-pa'i Ral-gri's commentary is entitled Proof of the Secret Nucleus (gsang-snying sgrub-pa rgyan-gyi me-tog). No longer extant as a distinct work, most of it survives or is cited in other texts, viz. rGyal-sras Thugs-mchog-rtsal, chos-'byung rin-po-che'i gter-mdzod. II, f. 357-61, Dalai Lama V, gan-ga'i chu-rgyun. Vol. 4, p. 397. the Collected Writings of Sog-bzlog-pa. Vol. 1. PP. 500-509 and 524-526 (= nges-don 'brug-sgra). and kLong-chen III bKra-shis rNam-rgyal, legs-bshad dri-med gan-ga'i chu-rgyun. p. 20. References to this treatise are also found in dPa'-bo gTsug-lag 'Phreng-ba, mkhas-pa'i dga'-ston. p. 178.[9]

In Sog-bzlog-pa’s version, Rig-pa'i Ral-gri is quoted as follows:[10]

This tantra is genuine for the following reasons: The master Viśvamitra in his Great Commentary on the Glorious Guhyasamāja (dpal gsang-ba 'dug-pa 'grel-chen. T. 1844), in the course of his comments on the passage (from the Guhyasamāja):

“How far does the Being of Pristine Cognition reach?...” cites the Guhyagarbha as follows: In the abode of Akaniṣṭha without extremes or centre, on the radiant wheel of pristine cognition that is the limitless ground, there is his celestial palace blazing forth with jewels of pristine cognition, completely uninterrupted throughout the ten directions... (Ch. 1, section 3)

And also:

In every Inconceivable (world-system), he appears universally as diverse buddha-body, speech, and mind. (Ch. 1. section 6)

Then, in commenting on the (Guhyasamāja) passage:

“The stūpa should be known to be
The palatial abode of all buddhas...”

He cites the Guhyagarbha as follows:

Its spire is the pristine cognition central to all, in which all maṇḍalas of the buddhas of the ten directions and four times without exception are not distinct from one another, and are of a single essence. (Ch. 1. section 3)

Then, commenting on the passage. “Substantial existence is based on insubstantiality...”, he gives:

Emaho! This wondrous marvelous reality
Is the secret of all the perfect buddhas.
All is created through the uncreated.
At creation itself there is no creation.
(Ch. 2, 6)

Then, while explaining the meaning of “secret” he says. “The Guhyagarbha speaks of five empowerments.” (Ch. 10)

Moreover, he quotes the passage beginning:

Their [body-colours] are dark blue, white, yellow, scarlet... up to:
... [Pervasive] without extremes or centre,
[It is an unthinkable] spontaneously present [maṇḍala];
(Ch. 6, 9-11)

And he says, “According to the Guhyagarbha. there are three realities.”[11]

In these and all other such instances Viśvamitra begins by mentioning the title Guhyagarbha.

The four perverse faults, etcetera, (criticised by 'Gos Lhas-btsas), are also to be rejected:

1. (When texts begin with the words) Thus I have expounded,[12] it traditionally means that they were compiled by the buddhas themselves, for it is impossible for even the tenth level bodhisattvas to compile all the teachings of the buddhas. As it says in the Verification of the Secret (Śrīguhyasiddhi, T. 2217), composed by master Saroruha as a commentary on the Guhyasamāja:[13]

Most masters claim
That the most radiant tantra.
The glorious Guhyasamāja.
Had as its compiler
The spiritual warrior called Lokeśvara.
But by the kindness of my venerable Guru
I know that the compiler of the glorious Guhyasamāja
Could not have been any other.
And so the being who propounded it
Was that tantra's author.
The indestructible reality of mind.

In accord with this explanation, there is a tradition whereby the exponent himself is the compiler.[14]

2. As for the Immeasurable ground: the Abhidharma, too, explains that Akaniṣṭha is immeasurable.

3. Concerning the four times: Viśvamitra’s Great Commentary (T. 1844) says: “Thus, the fourth time should be known to be sameness...” Moreover, the phrase, By all the lords of ten directions And four times is also found In the new translations. Buddhaguhya explains that it refers to the four aeons.

4. Regarding Vajrasattva’s appearance at the centre (of the maṇḍala): even the new translations explain that the foremost figure in the maṇḍala may change positions.

Concerning the passage: The final punctuation dots (tig) are discriminative awareness through which names are applied (Ch. 4, 15): the Indian manuscript of the Guhyagarbha reads sūtrī prajñātiṣyati. Sūtrī (“thread”) is the Sanskrit word for the (“measuring line”). Sūryaprabhāsiṃha’s commentary (Guhyagarbhatattvanirṇaya-vyākhyānaṭīkā. P. 4719) explains (tig as being equivalent) to this. Tig is an archaicism.[15]

As for the reference to other tantras (which is found in the Guhyagarbha): All the tantras expounded later on, such as the Hevajra (T. 417-8), also refer to the Summation of the Real (Tattvasaṃgraha. T. 479) which had been delivered first.[16]

Rig-pa'i Ral-gri's argument thus seeks to establish the authenticity of the Guhyagarbha. citing quotations from it which occur in celebrated Indian texts of the Guhyasamāja cycle. While certain tantras may have been written down directly in the Tibetan language, there is no evidence to include the Guhyagarbha among these.

Further criticisms levelled by 'Bri-gung dPal-'dzin at the Atiyoga system In general and at Padmasambhava’s man-ngag lta-'phreng, a celebrated commentary on chapter thirteen of our text, have been examined by Sog-bzlog-pa and 'Jigs-med gLing-pa, and more recently by N. Norbu and S.G. Karmay.[17] The latter has noted passages from this commentary which occur in gNubs-chen's early work, bsam-gtan mig-sgron, and translated the entire text as reproduced by Rong-zom-pa. In addition, he has brought to our attention certain passages from Sūryaprabhāsiṃha’s Indian commentary, the Śrīguhyagarbhatattvanirṇayavyākhyānaṭīkā, among the Tun Huang documents.[18] The early literary and historical sources thus bring us closer to the traditional view that the text was Introduced in the eighth century.

That many of the rNying-ma tantras were unknown in eleventh century India is not surprising when one considers that their translations are attributed to the eighth century and that the majority of them were considered to have been Imported Into eighth century Tibet, not from the Magadha heartland of North India, but from Oḍḍiyāna and adjacent regions in the north-west. Atiśa, on a visit to the library of Pehar dKor-mdzod gLing at bSam-yas, is known to have marvelled at the existence of tantras which no longer survived in Central India.[19]

The arguments raised by Ye-shes 'od and 'Gos Lhas-btsas against the Guhyagarbha lost their impact and controversy by the fourteenth century. Indeed, they became dead issues for Tibetan historians such as 'Gos Lotsâwa gZhon-nu dPal (1392-1481) who personally acquired the Sanskrit manuscript of the root-tantra which had been rediscovered at bSam-yas in the interim.[20]

In consequence, Sog-bzlog-pa could credibly present the following sixteenth century account of its introduction and translation in his slob-dpon sangs-rgyas gnyis-pa padma 'byung-gnas-kyi rnam-par thar-pa yid-kyi mun-sel. p. 128: Therein, perhaps following Sangs-rgyas gLing-pa's bka'-thang gser-'phreng. p, 400, he states that the Sanskrit manuscripts of the sgyu-'phrul sde-brgyad were taken from Nālanda Vihāra by Padmasambhava and then translated through miraculous ability at rGya-dkar sGra-bsgyur gLing, south of bSam-yas.[21] The texts were then kept at the Ke-tshang in bSam-yas when no longer extant in India.

In his dris-lan lung-dang rigs-pa'i 'brug-sgra. p. 12, Sog-bzlog-pa then repeats 'Gos Lotsâwa’s account of the discovery of the Sanskrit manuscript in bSam-yas by the great paṇḍita Śākyaśrī (1127-1225). The latter entrusted it to rTa-ston gZi-brjid, from whom it passed into the hands of Sha-ge Lotsâwa and thence to bCom-ldan Rig-pa'i Ral-gri who composed the aforementioned commentary in defence of the tantra. Subsequently, Thar-pa Lotsâwa retranslated the Sanskrit version of the root-text known as the rgyud phyi-ma. with two additional chapters (Chs. 23 & 24) for the first time, and these were revised by 'Gos Lotsâwa gZhon-nu dPal in person.[22]

Later rNying-ma writers like 'Jigs-med gLing-pa refuse to debate the specific points of 'Gos Lhas-btsas, considering that the past refutations of bCom-ldan Rig-pa'i Ral-gri and Sog-bzlog-pa were unanswerable. That this view was also held by followers of the new translation schools is evidenced by the following dismissive response of the Sa-skya-pa scholar, Zi-lung-pa Śāk-ya mChog-ldan (1428-1507):[23]

It is not necessary to prove laboriously that
The rNying-ma-pa doctrines were translated from Indian originals.
It is enough that they are proven to be
The teaching of the emanational master (Padmasambhava).
Although they do not conform with the mantras and symbols
Of those translated from India later on.
The proof of their validity is infallible accomplishment
Through their supreme and common attainment.[24]
They may be compared with the doctrines taken
By supreme, accomplished masters from various, great lands.
And which were not translated in India
From their respective volumes;
For it is said that with Vajrasattva’s consent
The compilers of those transmitted precepts
Were themselves permitted to teach them
In the language of each different country.
The rNying-ma-pa doctrinal traditions that definitely were
Translated from India require no proof.
Having formulated arguments one might prove
The indefinite ones to be treatises.
But the great ones who came before in Tibet,
Discovering this to be an artificial, conceptual path.
Have avoided wandering upon it.
As they themselves have explained.

Footnotes and references:


Reference has already been made to these polemics, p. 28, note 54. 'Jigs-med gLing-pa, rnying-ma'i rgyud-'bum-gyi rtogs-brjod. pp. 147-148, however, cites a relevant passage from chapter sixteen of the Guhyasamājatantra, indicating that the sbyor-sgrol practices are also prominent in tantras respected by the later schools. Yet there is clearly a distinction in purpose between the apparently shocking coded or twilight language (sandhyābhāṣa) found in tantras like the Caṇḍamahāroṣaṇa. and the Guhyagarbha's elaborate and lucid presentation of the sbyor-sgrol practices within the context of advanced meditation. One can only speculate that it was the clear and direct expression of secret teaching in this text which caught the attention of Lha bLa-ma Ye-shes-'od, thereby causing him to link it with the abuse of sbyor-sgrol in eleventh century Tibet. Incidentally, coded language also occurs in the early translations, and a detailed examination of its usage and metaphor would be a valuable study in itself.


Lo-chen Dharmaśrī, gsang-bdag zhal-lung. pp. 102-103, discusses the distinctive terminology of the Anuyoga and Atiyoga texts. His argument that the language employed in higher vehicles is not the same as that current in lower vehicles is taken up by bDud-'joms 'Jigs-bral Ye-shes rDo-rje, NSTB, Book 2, Pt. 7. p. 735, who emphasises that the doctrinal terminology of the prajñāpāramita is absent in the śrāvakapiṭaka and in the lower mantra texts, while the language of the lower mantra-texts is not found in the Anuttaratantras such as Guhyasamāja. and the terms of the latter do not much occur in Anuyoga and Atiyoga. Reference has already been made, p. 17, n. 12-13; P- 28, n. 53, to the early translators who sought to render meaning rather than word. Their original translations of tantra-texts are contrasted with the revised translations of sūtra-texts, e.g., in 'Jigs-med gling-pa, rgyud-'bum dris-lan. pp. 285-288.


NSTB, Book 2, Pt. 5. PP- 352-353-


Zur-chung-pa, in jest, associates Khyung-po Grags-se's mundane self-interested desire to have him killed with the forceful rite of "liberation” (sgrol). which 18 explained to transfer the consciousness of another from the body into a buddha-field, acting out of compassion. See phyogs-bcu mun-sel. pp. 396-400; also NSTB. Book 2. Pt. 5, PP- 281 ff.


NSTB. Book 2. Pt. 5. PP. 340-341, Pt. 7. p. 764.


'Gos Khug-pa Lhas-btsas's work is contained in sngags-log sun-'byin skor. pp. 18-25: Thimpu, 1979. India, as stated above, p. 132. note 11. refers in this context to the Magadha region alone.


The tantra clearly has no audience of bodhisattvas because it is held to be a self-manifesting expression of buddha-nature, i.e. it manifests in and of itself (rang-snang) to the buddhas alone. This problem is discussed by kLong-chen Rab-'byams-pa in phyogs-bcu mun-sel. Ch. 1. pp. 16-28. where he also explains that the introductory phrase 'di-skad bshad-pa'i dus-na refers to the fourth time. i.e. sameness throughout past, present and future. On the third of 'Gos Lhas-btsas's points. kLong-chen-pa states the central deity of a maṇḍala may rotate. Interestingly, he is the only commentator on the Guhyagarbha who insists on Vairocana being the central deity, rather than Vajrasattva. His reasons for so doing are outlined in phyogs-bcu mun-sel. Ch. 1, pp. 60-72. As to the fourth point, he claims, phyogs-bcu mun-sel. Ch. 11, pp. 417-418, that there many precedents for tantras referring to others which had been delivered earlier in time.


kLong-chen-pa, op. cit.. Ch. 1, asserts that in the view of the new translation schools the ground of Akaniṣṭha is also considered to be immeasurable.


Refer to S.G. Karmay, op. cit.. p. 277. note 23. The first of these texts is commonly attributed to kLong-chen-pa.


The following reproduces almost the entire text of this short work as preserved in Collected Writings Sog-bzlog-pa, Vol. 1, pp. 524-526.


The five empowerments referred to are also known as the five empowerments of ability (nus-pa'i dbang lnga), which are included among the fifteen ordinary sacraments of empowerment. See Phyogs-bcu mun-sel. Ch. 10. pp. 372-376. They are namely, the empowerment of the listener (nyan-pa'i dbang) which is that of Ratnasambhava, the empowerment of the meditator (bsgom-pa'i dbang) which is that of Akṣobhya. the empowerment of the expositor (chad-pa'i dbang) which is that of Amitābha, the empowerment of enlightened activity (phrin-las-kyi dbang) which is that of Amoghasiddhi. and the empowerment of the king of indestructible reality (rdo-rje rgyal-po'i dbang) which is that of the five enlightened families. The three realities (de-kho-na-nyid gsum) in question are explained in Sūryaprabhāsimha's commentary (P. 4719). pp. 2-3. to be the uncreated reality which is the causal basis of the maṇḍala, the resultant reality which is the spontaneous Samantabhadra, and the reality which appears as a chain of seed syllables and is the causal basis of the secret enlightened mind.


One should note that these are not the actual words of introduction employed in this tantra-text. The particular reason behind the Guhyagarbha's usage of the words: Thus, at the time of this explanation is discussed by kLong-chen Rab-'byams-pa in phyogs-bcu mun-sel. Ch. 1, pp. 16-28.


Śrīguhyasiddhi. T. 2217. P. 3061, Vol. 68, 228.4.7-8. Note that the last verse of the Peking version reads thugs-rje che for thugs rdo-rje. The author Sororuha or Padmavajra is regarded as a form of Padmasambhava.


The exponent is identified with the compiler in the sense that such advanced tantras are said to be manifested in and of themselves (rang-snang). See the above explanation of this term, p. 42, note 90.


The punctuation dots, which delimit or measure the Sanskrit word are, of course, the visarga. See the explanation in phyogs-bcu mun-sel. p. 189.


See the above note 109. The Yogatantras, exemplified by the Tattvasamgraha (T. U79). are considered to be earlier than the Anuttarayogatantras. See D.L. Snellgrove, The Hevajra Tantra. II, v. 57 for this specific reference.


'Bri-gung dPal-'dzin's text, chos-dang chos ma-yin-pa rnam-par dbye-ba'i rab-tu byed-pa. is reproduced in Sog-bzlog-pa, nges-don brug-sgra. p. 265, where his contention that the man-ngag lta-'phreng is a commentary on the Guhyasamājatantra and his rejection of the rdzogs-chen terminology are refuted. 'Jigs-med gLing-pa, rgyud-'bum-gyi rtogs-brjod, p. 133. repeats the refutation, pointing out that Buddhajñānapāda's Mukhāgama. P. 2716, Vol. 65, p. 10.1.2, refers to Atiyoga in a celebrated line: rdzogs-pa chen-po ye-shes spyi-vi gzugs. In the rgyud-'bum dris-lan, p. 127. he also quotes mNga'-ris Pan-chen Padma dBang-rgyal's rejection of 'Bri-gung dPal-'dzin's position. On this verse and mNga'-ris Pan-chen's repetition of it at the beginning of his sdom-gsum rnam-nges, a commentary on the integration of prātimokṣa. bodhisattva and mantra vows, see NSTB, Book 1, Pt. 4, pp. 194b-195b. Book 2. Pt. 6, pp. 706 ff. For recent criticisms, see N. Norbu, The Small Collection Hidden Precepts. pp. 8-9. and S.G. Karmay, op. cit., pp. 246-251.


Karmay notes, op. cit.. p. 234, that two passages from man-ngag lta-phreng are cited in gNubs-chen Sangs-rgyas Ye-shes's bsam-gtan mig-sgron. Ch. 6. The entire text is reproduced in Selected Writings of Rong-Zom chos-kyi bzang-po. For the Tun Huang references, see above p. 37, note 81.


The comments of Atiśa on this subject are recorded in his biography. See H. Elmer, Rnam thar rgyas pa, vol. 2, p. 53. passage 076.


For perhaps the earliest surviving account of its rediscovery, see 'Gos Lotsawa gZhon-nu dPal, deb-ther sngon-po, stod-cha. p. 1365 and G. Roerich, Blue Annals, pp. 103-104. 'Gos himself was Involved in the retranslation of the roottext.


However the colophons of the sgyu-'phrul sde-brgyad and biographical sources concur that Vimalamitra was the translator of the whole collection. See NSTB, Book 2, Pt. a. p. 179.


According to G. Roerich, Blue Annals. p. 104, the twenty-third and twenty-fourth chapters of this later version are contained in the sDe-dge xylograph edition of the rnying-ma'i rgyud-'bum. Vol. XII (Na). The text does not appear to be contained in NGB.


On Zi-lung-pa Sak-ya mChog-ldan and his contributions to philosophical controversy in Tibet, see L.W.J. van der Kuijp, Contributions to the Development at Tibetan Buddhist Epistemology. pp. 10-22. This particular passage is cited by Sog-bzlog-pa in his Collected Writings. vol. 1, pp. 519-520.


I.e., supreme and common accomplishments (mchog-dang thun-mong-gi dngos-grub), on which see above, note 75.

Help me keep this site Ad-Free

For over a decade, this site has never bothered you with ads. I want to keep it that way. But I humbly request your help to keep doing what I do best: provide the world with unbiased truth, wisdom and knowledge.

Let's make the world a better place together!

Like what you read? Consider supporting this website: