Monthly Archives: May 2019

Tilopa’s famous Gangama Mahamudra‒according to tradition taught by Tilopa to Naropa at the banks of the River Ganges and later transmitted to Marpa Lotsawa‒has been translated before, e.g., by Trungpa, Tiso and Torricelli, Brunnhölzl, and Khenpo Thampel.

Why now another translation? The answer is: because we have different texts.

More then a decade ago I started collecting manuscript editions of the Gangama, that is, text editions that were written by hand and transmitted outside of the “official” canon (i.e., the Tibetan Tangyur). I soon noticed that these paracanonical manuscripts contained a text that differed in many ways from the canonical version of the Peking, Derge, Narthang, and Cone Tangyur editions of the same text.

The greatest surprise for me was to discover the vast extent of structural intervention undertaken by the redactors of the canon. The paracanonical (manuscript) versions are structured as follows (numbers refer to the lines of the Tibetan text):

1‒4: advice to listen
5‒29: introduction to the nature of the mind through the examples of space, clouds, and the sun
30: view
31: conduct
32‒37: meditative practice
38‒47: pledges
48‒52: benefits of practicing this path
53‒55: defects of not practicing like that
56‒59: practice of relying on the guru and renunciation
60‒66: ascertaining the result of view, meditation, and conduct
67‒79: abandoning distractions in the solitude
79‒86: benefits of such practice
87‒100: practice of the individuals of highest capacity
101‒104: types of individuals
105‒113: practice of the individuals of lower capacities
114‒118: results and qualities

One of the key features of this structure is that the text directly introduces with 25 lines the nature of the mind. The key feature of the structure of the canonical version as found in the Tangyur, on the other hand, is that the text teaches first a gradual teaching of 28 lines before it offers an introduction to the nature of the mind. The redactors of the canon thus have changed the very nature of the text, turning this teaching of the Indian Siddha tradition into a mainstream “gradual path” type of teaching (lam rim).

Secondly, the paracanonical manuscript tradition presents the Gangama as a text with irregular numbers of syllables per line, generally nine or eleven, but occasionally also seven and thirteen. This allows for a natural expression as we find it in many instructional texts on mahamudra, such as the poetical and spontaneous songs of the mahasiddhas and also in many of Jigten Sumgön’s songs and instructions. The editors of the Tangyur, on the other hand, have changed this into a uniform pattern of nine syllable verses with mostly four lines, thereby streamlining it to fit the style of other versified teachings.

The text that I present here as the main text is taken from the Oral Transmission of Cakrasamvara and the Oral Transmission of the Dakini, both edited and arranged by the great Drukpa Kagyü master Padma Karpo (1527‒1592). According to the tradition, these teachings were received by Tilopa directly from the Dakinis when he was staying in Uddiyana. Tilopa transmitted them to Naropa and the latter to his Tibetan disciple Marpa Lotsawa, who passed them on to Milarepa. At this point, there is some confusion that was recently cleared up in an article by Marta Sernesi (2011). According to her, people have confused this oral transmission (or, as she calls it, aural since that is closer to the meaning of the Tibetan word snyan, “ear”) with the Nine Instructions of the Formless Dakini. The story of Milarepa receiving only five of these instructions from Marpa and sending his disciple Rechungpa to India to bring back the remaining four has actually nothing to do with the transmission of the Oral Transmission of Cakrasamvara (or the Oral Transmission of the Dakini, which are alternative terms) itself since the Nine Instructions are only supplemental teachings to the actual Oral Transmissions. Larsson (2012: 86) and Quintman (2014: 41) still repeat the mistake, and this shows once again how dangerous it is to take such legends at face value. In truth, in the writings of the Oral Transmission, it is clear that the transmission from Marpa to Mila has been complete.

Sernesi also points out that the Oral Transmission of Cakrasamvara or the Dakini are (alternative) names for the teaching, and the Oral Transmission of Rechung and Oral Transmission of Ngendzong, who was the other cotton clad yogi to receive the transmission from Mila, are names for particular lineages in which the instructions were transmitted. There also appears to be a further lineage through Gampopa and Phagmodrupa (Larsson 2012: 88), the Oral Transmission of Dakpo, possibly with abridged or essential instructions (Sernesi 2011: 180, n. 2).

The Oral Transmission of Cakrasamvara contains teachings by Vajradhara, Vajrayogini and other Dakinis, Tilopa, Naropa and other Indian masters, as well as by Marpa, Milarepa, and later disciples. According to Padma Karpo’s introduction and catalog of the Oral Transmission of the Dakini (mKha’ ’gro snynan brgyud kyi dpe tho, Torricelli 2000: 361), the Gangama Mahamudra is the first text of the collection (as arranged by him), and it is its essential instruction.

Of the already existing translations of the Gangama, Trungpa Rinpoche’s and Tiso and Torricelli’s are made from the canonical version of the Tangyur. Brunnhölzl’s translation is based on a paracanonical transmission as it appears in the 5th Shamarpas’s commentary. Within the paracanonical transmission, I observed two groups with the Shamarpa’s and the rGya gzhung manuscripts on the one side and the Oral Transmissions on the other (this is described in the introduction to my edition). The Shamarpa’s commentary and Brunnhölzl’s translation show some particular features that I pointed out in the introduction to my translation of the Oral Tradition manuscripts. The commentary of H.H. Drikung Kyabgön Chetsang Rinpoche is based on my edition of ten paracanonical and four canonical editions. Khenpo Thampel’s translation of the root text, however, seems to be based on a canonical version.

My translation and the accompanying edition are not supposed to present a definitive edition or translation of the Gangama. I aim above all to document the hitherto neglected Oral Transmission and to make the many interesting variant readings of the different manuscript families visible.

You can download my translation and edition in the downloads section in the lower part of the right column of your screen: “Tilopa: Gangama Mahamudra (Translation)” and “Tilopa: Gangama Mahamudra (Edition of Tib. text).” Enjoy!


For bibliographical references to the other translations, see my translation of the Oral Tradition.

Larsson, Stefan (2012) Crazy for Wisdom: The Making of a Mad Yogin in Fifteenth-Century Tibet, Leiden: Brill.

Sernesi, Marta (2011) “The Aural Transmission of Samvara: An Introduction to Neglected Sources for the Study of Early Bka’ brgyud,” Mahamudra and the Bka’ brgyud Tradition, Andiast: International Institute for Tibetan and Buddhist Studies.

Torricelli, Fabrizio (2000) “Padma dkar-po’s Arrangement of the ‘bDe-mchog snyan-brgyud,’ East and West, 50(1/4), 359-386.

Quintman, Andrew (2014) The Yogin and the Madman: Reading the Biographical Corpus of Tibet’s Great Saint Milarepa, New York: Columbia University Press.